Precision Medicine 2020

Methods in
Molecular Biology 2204
Tao Huang Editor
Precision
Medicine
ALGrawany
METHODS IN MOLECULAR BIOLOGY
Series Editor
John M. Walker
School of Life and Medical Sciences
University of Hertfordshire
Hatfield, Hertfordshire, UK
For further volumes:

http://www.springer.com/series/7651
For over 35 years, biological scientists have come to rely on the research protocols and
methodologies in the critically acclaimed Methods in Molecular Biology series. The series was
the first to introduce the step-by-step protocols approach that has become the standard in all
biomedical protocol publishing. Each protocol is provided in readily-reproducible step-by-
step fashion, opening with an introductory overview, a list of the materials and reagents
needed to complete the experiment, and followed by a detailed procedure that is supported
with a helpful notes section offering tips and tricks of the trade as well as troubleshooting
advice. These hallmark features were introduced by series editor Dr. John Walker and
constitute the key ingredient in each and every volume of the Methods in Molecular Biology
series. Tested and trusted, comprehensive and reliable, all protocols from the series are
indexed in PubMed.
ALGrawany
Precision Medicine
Edited by
Tao Huang
Shanghai Institute of Nutrition and Health, Chinese Academy of Sciences, Shanghai, China
Editor
Tao Huang
Shanghai Institute of Nutrition and
Health, Chinese Academy of Sciences
Shanghai, China
ISSN 1064-3745 ISSN 1940-6029 (electronic)

Methods in Molecular Biology
ISBN 978-1-0716-0903-3 ISBN 978-1-0716-0904-0 (eBook)
https://doi.org/10.1007/978-1-0716-0904-0
© Springer Science+Business Media, LLC, part of Springer Nature 2020

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is
concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction
on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation,
computer software, or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply,
even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations
and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to
be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty,
expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been
made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This Humana imprint is published by the registered company Springer Science+Business Media, LLC, part of Springer
Nature.
The registered company address is: 1 New York Plaza, New York, NY 10004, U.S.A.
ALGrawany
Preface
With the development of omics technologies, especially next-generation sequencing, disease

progression can be investigated in an unprecedented way. The multi-omics approaches
reveal the essence of disease pathology and make precision diagnosis and therapy possible.
Precision medicine will transform the medical practice fundamentally. In these 19 chapters,
we will introduce the technologies and applications of precision medicine.
The first four chapters present emerging experimental and bioinformatics technologies
that are widely used in precision medicine, such as T cell receptor repertoire sequencing,
nanopore sequencing, microsatellite instability, and network analysis. The T cell receptor
repertoire sequencing can help monitoring immune responses and predicting the prognosis
of disease. Nanopore sequencing as third-generation sequencing has great potential for
point-of-care testing (POCT) due to its fast turn-around time, portable and real-time data
analysis. Microsatellite instability is a key indicator for predicting response to anti-PD-1
inhibitors. Network analysis can integrate multi-omics big data and reveal the molecular
mechanisms underlying complex diseases.
The next six chapters are focused on cancer studies. There have many successful
applications of precision medicine in cancers. Breast cancer, prostate cancer, lung cancer,
nasopharyngeal carcinoma, and cancer of unknown primary origin are discussed. Biomedical
imaging, methylation, immunohistochemistry, and gene expression can all be used for
cancer diagnosis and treatment optimization. Even cancer of unknown primary origin can
be traced based on their molecular characteristics.
Cardiovascular diseases (CVDs) are the most common diseases and cause many deaths
each year widely. There are five chapters for cardiovascular diseases, such as atrial fibrillation
(AF), atherosclerosis, dilated cardiomyopathy, and total anomalous pulmonary venous
connection (TAPVC). DNA methylation has potential value in being biomarkers and
underlying the diagnosis and prognosis of atrial fibrillation. Proprotein convertase subtili-
sin/kexin type 9 (PCSK9) plays an important role in atherosclerosis and shows therapeutic
potentials.
The next two chapters are focused on other complex diseases beside cancers and
cardiovascular diseases, such as chronic obstructive pulmonary disease (COPD) and sys-
temic lupus erythematosus (SLE). Chronic obstructive pulmonary disease (COPD) is a
common disease with high morbidity and mortality in the world. Airway inflammation
biomarkers can facilitate precise manage of neutrophil-predominant COPD. Systemic
lupus erythematosus (SLE) is a complex autoimmune disease which faces difficulties in
treatment. Stratification of SLE patients based on genetic profiling will enable us to make
more effective and precise choices for treatment plans.
The last two chapters introduce the latest engineering and surgical developments in
precision medicine. Scientists seek various engineering approaches, such as 3D printing, to
harness stem cells, scaffolds, growth factors, and the extracellular matrix to promise
enhanced and more reliable bone formation. Ex Vivo Lung Perfusion (EVLP) is a technique
for extending lung preservation time and repairing lung injury in the field of lung transplan-
tation. EVLP can increase the number of lungs that meet the transplant criteria and, to some
extent, alleviate the current shortage of donor lungs.
v
vi Preface
As mentioned above, this book covers most perspectives of precision medicine, from
basic research to clinical surgeries, from cancer to cardiovascular disease, from sequencing
technology to big data analysis. It is hoped that this book can broaden the horizons for
researchers, engineers, and clinicians and accelerate the interdisciplinary precision medicine
research and applications.
Shanghai, China Tao Huang

ALGrawany
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
PART I EMERGING EXPERIMENTAL AND BIOINFORMATICS TECHNOLOGIES
1 T Cell Receptor Repertoire Sequencing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Huixin Lin, Yonggang Peng, Xiangbin Chen, Yuebin Liang,
Geng Tian, and Jialiang Yang
2 Nanopore Sequencing and Its Clinical Applications . . . . . . . . . . . . . . . . . . . . . . . . . 13
Xue Sun, Lei Song, Wenjuan Yang, Lili Zhang, Meng Liu,
Xiaoshuang Li, Geng Tian, and Weiwei Wang
3 The Clinical Significance of Microsatellite Instability in Precision
Treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
Zhenyu Huang, Xiaojian Chen, Chenying Liu, and Long Cui
4 Applications of Network Analysis in Biomedicine . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Steven Wang and Tao Huang
PART II PRECISION MEDICINE OF CANCERS
5 Diagnosis and Treatment of Breast Cancer in the Precision

Medicine Era . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Jing Yan, Zhuan Liu, Shengfang Du, Jing Li, Li Ma,
and Linjing Li
6 Application and Analysis of Biomedical Imaging Technology
in Early Diagnosis of Breast Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
Lin Chen, Nan Jiang, and Yuxiang Wu
7 Recent Advances in DNA Repair Pathway and Its Application
in Personalized Care of Metastatic Castration-Resistant Prostate
Cancer (mCRPC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Chenyang Xu, Shanhua Mao, and Haowen Jiang
8 Methylation in Lung Cancer: A Brief Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Chang Gu and Chang Chen
9 Epstein–Barr Virus DNA in Nasopharyngeal Carcinoma:
A Brief Review. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Fen Xue and Xiayun He
vii
viii Contents
10 A Review on Cancer of Unknown Primary Origin: The Role

of Molecular Biomarkers in the Identification of Unknown Primary
Origin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Na Yan, Yanxiang Zhang, Xuejie Guo, Dawei Yuan,
Geng Tian, and Jialiang Yang
PART III PRECISION MEDICINE OF CARDIOVASCULAR DISEASES

11 DNA Methylation in Atrial Fibrillation and Its Potential Role
in Precision Medicine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Mengwei Lv, Wen Ge, Zhi Li, Chao Wang,
and Yangyang Zhang
12 PCSK9 Inhibition and Atherosclerosis: Current Therapeutic
Option and Prospection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Pratik Pandey, Cuimei Zhao, and Ban Liu
13 Precise Drug Sequential Therapy Can Improve the Cardioversion
Rate of Atrial Fibrillation with Valvular Disease after Radiofrequency
Ablation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145
Tao Li and Yongjun Qian
14 Precision Medicine and Dilated Cardiomyopathy . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
Xiang Li and Wenyan Zhu
15 Research Progress in Pathogenesis of Total Anomalous Pulmonary
Venous Connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Xin Shi, Yanan Lu, and Kun Sun
PART IV PRECISION MEDICINE OF OTHER COMPLEX DISEASES
16 Airway Inflammation Biomarker for Precise Management

of Neutrophil-Predominant COPD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
Xue Liang, Ting Liu, Zhiming Zhang, and Ziyu Yu
17 Genome Variation and Precision Medicine in Systemic Lupus
Erythematosus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Ru Yang, Yaqi Hu, and Lin Bo
PART V ENGINEERING AND SURGICAL DEVELOPMENTS

IN PRECISION MEDICINE
18 Precision Medicine in Tissue Engineering on Bone . . . . . . . . . . . . . . . . . . . . . . . . . 207

Bingkun Zhao, Qian Peng, Rong Zhou, Haixia Liu,
Shengcai Qi, and Raorao Wang
19 Progress of Clinical Application for Ex Vivo Lung Perfusion
(EVLP) in Lung Transplantation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
Chang Gu, Xufeng Pan, and Jianxin Shi
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
ALGrawany
Contributors
LIN BO • Department of Rheumatology, The Second Affiliated Hospital of Soochow

University, Suzhou, Jiangsu Province, China
CHANG CHEN • Department of Thoracic Surgery, Shanghai Pulmonary Hospital, Tongji
University School of Medicine, Shanghai, China
LIN CHEN • Department of Kinesiology, Jianghan University, Wuhan, China
XIANGBIN CHEN • Geneis (Beijing) Co., Ltd., Beijing, People’s Republic of China
XIAOJIAN CHEN • Department of Colorectal and Anal Surgery, Xinhua Hospital, Shanghai
Jiao Tong University School of Medicine, Shanghai, China
LONG CUI • Department of Colorectal and Anal Surgery, Xinhua Hospital, Shanghai Jiao
Tong University School of Medicine, Shanghai, China
SHENGFANG DU • Department of Anesthesiology, The Second Hospital of Lanzhou University,
Lanzhou, People’s Republic of China
WEN GE • Department of Thoracic and Cardiovascular Surgery, Shuguang Hospital,
Shanghai University of TCM, Shanghai, China
CHANG GU • Department of Thoracic Surgery, Shanghai Pulmonary Hospital, Tongji
University School of Medicine, Shanghai, China; Department of Thoracic Surgery,
Shanghai Chest Hospital, Shanghai Jiao Tong University, Shanghai, China
XUEJIE GUO • Geneis (Beijing) Co., Ltd., Chaoyang District Beijing, People’s Republic of
China
XIAYUN HE • Department of Radiation Oncology, Fudan University Shanghai Cancer
Center, Shanghai Medical College, Shanghai, China
TAO HUANG • Shanghai Institute of Nutrition and Health, Chinese Academy of Sciences,
Shanghai, China
ZHENYU HUANG • Department of Colorectal and Anal Surgery, Xinhua Hospital, Shanghai
YAQI HU • Department of Rheumatology, The Second Affiliated Hospital of Soochow
HAOWEN JIANG • Department of Urology, Huashan Hospital, Fudan University, Shanghai,
China
NAN JIANG • Department of General Surgery, First Hospital of Tsinghua University, Beijing,
China
XUE LIANG • The Fifth Affiliated Hospital, Guangzhou Medical University, Guangzhou,
Guangdong, People’s Republic of China; Key Laboratory of Molecular Target & Clinical
Pharmacology, School of Pharmaceutical Sciences, Guangzhou Medical University,
Guangzhou, People’s Republic of China; State Key Laboratory of Respiratory Disease, School
of Pharmaceutical Sciences, Guangzhou Medical University, Guangzhou, People’s Republic
of China
YUEBIN LIANG • Geneis (Beijing) Co., Ltd., Beijing, People’s Republic of China
JING LI • Department of Clinical Laboratory Center, The Second Hospital of Lanzhou
University, Lanzhou, China
ix
x Contributors
LINJING LI • Department of Clinical Laboratory Center, The Second Hospital of Lanzhou

HUIXIN LIN • Geneis (Beijing) Co., Ltd., Beijing, People’s Republic of China
TAO LI • Department of Cardiovascular Surgery, National Clinical Research Center for
Geriatrics, West China Hospital, Sichuan University, Chengdu, Sichuan Province, China
BAN LIU • Department of Cardiology, Shanghai Tenth People’s Hospital, Tongji University
School of Medicine, Shanghai, China
CHENYING LIU • Department of Colorectal and Anal Surgery, Xinhua Hospital, Shanghai
HAIXIA LIU • Department of Stomatology, Shanghai 10th People’s Hospital, Tongji University
MENG LIU • Geneis (Beijing) Co., Ltd., Beijing, People’s Republic of China
TING LIU • The Fifth Affiliated Hospital, Guangzhou Medical University, Guangzhou,
Guangdong, People’s Republic of China
ZHUAN LIU • Department of Clinical Laboratory Center, The Second Hospital of Lanzhou
XIANG LI • Department of Cardiology, The First Affiliated Hospital of Chongqing Medical
University, Chongqing Medical University, Chongqing, China
XIAOSHUANG LI • Geneis (Beijing) Co., Ltd., Beijing, People’s Republic of China
ZHI LI • Department of Cardiovascular Surgery, Jiangsu Province Hospital, The First
Affiliated Hospital of Nanjing Medical University, Nanjing, China
YANAN LU • Department of Pediatric Cardiology, Xin Hua Hospital, School of Medicine,
Shanghai Jiao Tong University, Shanghai, China
MENGWEI LV • Shanghai East Hospital of Clinical Medicine College, Nanjing Medical
University, Shanghai, China; Department of Cardiovascular Surgery, Shanghai East
Hospital, Tongji University School of Medicine, Shanghai, China
LI MA • Department of Clinical Laboratory Center, The Second Hospital of Lanzhou
SHANHUA MAO • Department of Urology, Huashan Hospital, Fudan University, Shanghai,
China
PRATIK PANDEY • Department of Cardiology, Shanghai Tenth People’s Hospital, Tongji
XUFENG PAN • Department of Thoracic Surgery, Shanghai Chest Hospital, Shanghai Jiao
Tong University, Shanghai, China
QIAN PENG • Department of Stomatology, Shanghai 10th People’s Hospital, Tongji University
YONGGANG PENG • Geneis (Beijing) Co., Ltd., Beijing, People’s Republic of China
YONGJUN QIAN • Department of Cardiovascular Surgery, National Clinical Research
Center for Geriatrics, West China Hospital, Sichuan University, Chengdu, Sichuan
Province, China
SHENGCAI QI • Department of Stomatology, Shanghai 10th People’s Hospital, Tongji
JIANXIN SHI • Department of Thoracic Surgery, Shanghai Chest Hospital, Shanghai Jiao Tong
University, Shanghai, China
XIN SHI • Department of Pediatric Cardiology, Xin Hua Hospital, School of Medicine,
LEI SONG • Geneis (Beijing) Co., Ltd., Beijing, People’s Republic of China
ALGrawany
Contributors xi
KUN SUN • Department of Pediatric Cardiology, Xin Hua Hospital, School of Medicine,
XUE SUN • Geneis (Beijing) Co., Ltd., Beijing, People’s Republic of China
GENG TIAN • Geneis (Beijing) Co., Ltd., Chaoyang District Beijing, People’s Republic of
China
CHAO WANG • Tongji University School of Medicine, Shanghai, China
RAORAO WANG • Department of Stomatology, Shanghai 10th People’s Hospital, Tongji
STEVEN WANG • Department of Biological Sciences, Columbia University, New York, NY,
USA
WEIWEI WANG • Geneis (Beijing) Co., Ltd., Beijing, People’s Republic of China
YUXIANG WU • Department of Kinesiology, Jianghan University, Wuhan, China
CHENYANG XU • Department of Urology, Huashan Hospital, Fudan University, Shanghai,
China
FEN XUE • Department of Radiation Oncology, Fudan University Shanghai Cancer Center,
Shanghai Medical College, Shanghai, China
JIALIANG YANG • Geneis (Beijing) Co., Ltd., Chaoyang District Beijing, People’s Republic of
China
RU YANG • Department of Rheumatology, The Second Affiliated Hospital of Soochow
WENJUAN YANG • Geneis (Beijing) Co., Ltd., Beijing, People’s Republic of China
JING YAN • Department of Clinical Laboratory Center, The Second Hospital of Lanzhou
NA YAN • Geneis (Beijing) Co., Ltd., Chaoyang District Beijing, People’s Republic of China
DAWEI YUAN • Geneis (Beijing) Co., Ltd., Chaoyang District Beijing, People’s Republic of
China
ZIYU YU • The Fifth Affiliated Hospital, Guangzhou Medical University, Guangzhou,
Guangdong, People’s Republic of China
LILI ZHANG • Geneis (Beijing) Co., Ltd., Beijing, People’s Republic of China
YANGYANG ZHANG • Department of Cardiovascular Surgery, Shanghai East Hospital, Tongji
YANXIANG ZHANG • Geneis (Beijing) Co., Ltd., Chaoyang District Beijing, People’s Republic
of China
ZHIMING ZHANG • The Fifth Affiliated Hospital, Guangzhou Medical University,
Guangzhou, Guangdong, People’s Republic of China
BINGKUN ZHAO • Department of Stomatology, Shanghai 10th People’s Hospital, Tongji
CUIMEI ZHAO • Department of Cardiology, Tongji Hospital, Tongji University School of
Medicine, Shanghai, China
RONG ZHOU • Department of Stomatology, Shanghai 10th People’s Hospital, Tongji
WENYAN ZHU • Department of Oncology, Chongqing (CHN.USA) Hygeia Hospital,
Chongqing, China; Yidu Cloud (Beijing) Technology Co., Ltd., Beijing, China
Part I
Emerging Experimental and Bioinformatics Technologies

ALGrawany
Chapter 1
T Cell Receptor Repertoire Sequencing

Huixin Lin, Yonggang Peng, Xiangbin Chen, Yuebin Liang, Geng Tian,
and Jialiang Yang
Abstract
The status of T cell receptors (TCRs) repertoire is associated with the occurrence and progress of various
diseases and can be used in monitoring the immune responses, predicting the prognosis of disease and other
medical fields. High-throughput sequencing promotes the studying in TCR repertoire. The chapter focuses
on the whole process of TCR profiling, including DNA extraction, library construction, high-throughput
sequencing, and how to analyze data.
Key words T cell receptor repertoire, Library construction, High-throughput sequencing, Data
analysis
1 Introduction
T lymphocytes play essential roles in the adaptive immune system. A

unique T cell receptor (TCR) is expressed on the surface of each
individual T lymphocyte. The interaction of TCR with the antigen-
major histocompatibility complex (MHC) molecules is the basis of
T lymphocytes recognizing antigen [1]. TCRs are highly variable
heterodimer molecules consisting of either a combination of alpha
and beta chains (αβ TCR), the major TCR, or a combination of
gamma and delta chains (γδ TCR) [2]. The extreme diversity results
from randomized combinations of DNA during T cell develop-
ment, which contain distinct variable diversity, joining (V(D)J)
gene segments, and deletion and/or insertion of nucleotides at
the junctions of these segments [3]. The hypermutation leads to
such strong combinatorial and junctional diversity that the resulted
TCRs are able to recognize a marvelous variety of antigens.
The sum of all TCRs is termed the TCR repertoire or TCR
profile. Researchers are becoming more and more interested in
determining the TCR repertoire status for its changing greatly
with the occurrence or progression of various diseases [4–7]. The
main difficulty in studying the TCR profile is its extreme diversity.
Tao Huang (ed.), Precision Medicine, Methods in Molecular Biology, vol. 2204, https://doi.org/10.1007/978-1-0716-0904-0_1,
3
4 Huixin Lin et al.
After DNA random combining, the diversity is up to a mathemati-

cal prediction of 1015 different TCR αβ molecules. Even with
thymic selection, an estimate of about 2 107 TCR αβ molecules
will be remained [2]. However, the latest high-throughput
sequencing methods can help us face this big challenge.
The whole process of TCR profiling includes DNA/RNA
extraction, library construction, high-throughput sequencing, and
data analysis. Both genomic DNA (gDNA) and RNA can be used as
the starting materials for TCR sequencing. Library construction
approaches contain multiplex PCR [8], target enrichment [9], and
50 RACE cDNA synthesis plus nested PCR [10]. The choices of
starting materials and library construction methods depend on the
purposes of studying. In the real world, qualified FFPE DNA
samples are relatively easier to be obtained. And given the diversity
of the target and the convenience of PCR amplification, multiplex
PCR is the most popular approach to construct library.
Here, we show the methods of the TCR sequencing, including
gDNA extraction from FFPE samples, multiplex PCR for library
construction, and the bioinformatics analysis of data.
2 Materials
Unless indicated, the reagents with analytical grade and ultrapure

water are necessary to prepare the following solutions.
2.1 Genomic DNA QIAamp DNA FFPE Tissue Kit (QIAGEN, Hilden, Germany) or
Isolation from FFPE other similar products on the market are recommended to isolate
Samples genomic DNA from FFPE tissue samples. Besides, the reagents and
equipment which need to be supplied by user are listed as below:
1. Xylene.
2. Ethanol (96–100%).
3. 1.5 mL or 2 mL microcentrifuge tubes.
4. Pipet tips (pipet tips with aerosol barriers recommended).
5. Thermomixer, heated orbital incubator, heating block, or
water bath capable of incubation at 90 C.
6. Microcentrifuge with rotor for 2 mL tubes.
7. Vortex mixer.
8. Fluorimetry (Qubit) and spectrophotometry (NanoDrop).
2.2 TCRβ CDR3 After amplification and purification of the TCRβ-CDR3 targeted
Library Construction region, ABclonal Rapid DNA Lib Prep Kit (ABclonal, Boston,
USA) or other similar products are recommended to construct
TCRβ CDR3 library. Besides, the reagents and equipment which
need to be supplied by user are listed as below (see Note 1):
ALGrawany
T Cell Receptor Repertoire Sequencing 5
1. Thermocycler.
2. Multiplex adapters compatible with Illumina® platforms.
3. Ethanol.
4. Nuclease-free water.
5. PCR strip tubes or plates.
6. Magnetic stand.
7. Agencourt™ AMPure XP bead (stored at 4 C).
8. Pipettes and multichannel pipettes.
9. Aerosol-resistant pipette tips.
10. Microcentrifuge.
11. Vortex mixer.
12. Agilent Bioanalyzer.
2.3 Data Analysis The following software are used to analyze the sequencing data.
1. Raw data processing: Cutadapt.
2. Analysis of clean data: MixCR.
3. Characteristics of TCR results: tcR.
3 Methods
3.1 Genomic DNA Recommend using QIAamp DNA FFPE Tissue Kit (QIAGEN,
Isolation from FFPE Hilden, Germany) or other similar products to isolate genomic
Samples DNA from FFPE tissue samples. The following are the main steps
to isolate genomic DNA according to the standard procedure of
QIAamp DNA FFPE Tissue Kit (QIAGEN, Hilden, Germany) (see
Note 2).
1. Trim excess paraffin off the sample block using a scalpel (see
Note 3).
2. Immediately place up to 8 sections (5–10 μm thick) into a
1.5 mL microcentrifuge tube, and 1 mL xylene was added to
the tube. Then close the lid and vortex thoroughly for 10 s.
3. Centrifuge at 20,000 g for 2 min.
4. Discard the supernatant by pipetting, but do not remove any of
the pellet.
5. Add 1 mL ethanol (96–100%) to the tube and vortex
thoroughly.
6. Centrifuge at 20,000 g for 2 min.
7. Discard the supernatant by pipetting, but do not remove any of
the pellet.
8. Open the tube and place it at room temperature until all
residual ethanol has evaporated.
6 Huixin Lin et al.
9. Resuspend the pellet with 180 μL Buffer ATL. Add 20 μL

proteinase K to the pellet and fully mix by vortexing.
10. Incubate at 56 C until the sample has been lysed completely.
(This process usually needs about 1 h.)
11. Then transfer the tube to 90 C and continue to incubate for
1 h (see Note 4).
12. Briefly centrifuge to collect all drops from the lid.
13. Add 200 μL Buffer AL to the tube and mix vigorously by
vortexing. Then add 200 μL ethanol (96–100%) and mix vig-
orously again.
14. Briefly centrifuge to collect all drops from the lid.
15. Transfer all the lysate to the QIAamp MinElute column (in a
2 mL collection tube) without wetting the rim, then close the
lid and centrifuge at 6000 g for 1 min. Place the column in a
new 2 mL collection tube and the old collection tube contain-
ing the flow-through should be discarded.
16. Carefully open the lid of the column and add 500 μL Buffer
AW1 without wetting the rim, then close the lid and centrifuge
at 6000 g for 1 min. Place the column in a new 2 mL
collection tube, and the old collection tube containing the
flow-through should be discarded.
17. Carefully open the lid of the column and add 500 μL Buffer
AW2 without wetting the rim, then close the lid and centrifuge
at 6000 g for 1 min. Place the column in a new 2 mL
collection tube, and the old collection tube containing the
flow-through should be discarded.
18. Centrifuge at 20,000 g for 3 min to dry the membrane
completely.
19. Place the column in a new 1.5 mL microcentrifuge tube.
Carefully open the lid of the column, then add 20–100 μL
Buffer ATE to the center of the membrane of the column.
20. Close the lid and incubate at room temperature for 1 min.
Centrifuge at 20,000 g for 1 min.
21. Qubit dsDNA assay and NanoDrop are used to assess the
concentration and purity of the isolated genomic DNA,
respectively.
3.2 TCR CDR3 Use primers for the J alleles of the TCR β chains together with a
Library Construction mix of primers for all known V alleles to amplify the TCR across the
CDR3 region by multiplex PCR.
3.2.1 Target Region
Amplification by 1. Add components including 25 μL Platinum™ Multiplex PCR
Multiplex PCR Master Mix, 16 μL Forward Primers (10 μM), 4.8 μL Reverse
Primers (10 μM), 100 ng genomic DNA, and appropriate
volume of water ensuring 50 μL total reaction volume to
EP tube.
ALGrawany
2. Place the EP tube on thermocycler (95 C for 2 min; 30 cycles

of 98 C for 30 s, 64.3 C for 30 s, 72 C for 20 s; 72 C for
5 min; 4 C hold).
3. Add 50 μL (ratio 1.0) of Agencourt™ AMPure XP beads to
reaction tube, mix thoroughly by pipetting. Incubate at RT for
5 min.
4. Pellet the beads on a magnetic stand at RT for 2 min.
5. Carefully remove and discard the supernatant.
6. Wash the beads with 200 μL fresh 80% ethanol. Pellet the beads
on a magnetic stand and carefully remove the ethanol.
7. Repeat step 6 in Subheading 3.2.1 for a total of two washes.
8. Resuspend the magnetic beads in 37 μL of low-EDTA TE
buffer. Mix thoroughly by pipetting, and then incubate at RT
for 1 min to release the DNA from the beads.
9. Pellet the beads on a magnetic stand at RT for 2 min. Transfer
clear supernatant to a new PCR tube.
After amplification and purification of the TCRβ-CDR3 tar-
geted region, recommend using ABclonal Rapid DNA Lib Prep Kit
(ABclonal, Boston, USA) or other similar products to construct
TCRβ CDR3 library. The following are the main steps to construct
library according to the standard procedure of ABclonal Rapid
DNA Lib Prep Kit (ABclonal, Boston, USA).
3.2.2 End Preparation 1. Prepare end-preparation reaction mix including 10 μL End

Prep Buffer, 3 μL End Prep Enzymes, and 37 μL Fragmented
DNA in PCR tubes on ice.
2. Mix thoroughly by pipetting.
3. Place the EP tube on thermocycler (20 C for 30 min; 65 C
for 30 min; 4 C hold) with a heated lid at 75 C.
3.2.3 Adaptor Ligation 1. Prepare the ligation reaction mix, including 50 μL End Prep
Reaction Mix (Subheading 3.2.2), 16.5 μL Ligation MM,
2.5 μL Working Adaptor, and 3 μL Ligase Mix in PCR tubes
on ice.
2. Incubate the reaction at 22 C for 15 min in a thermocycler
without a heated lid, and then hold at 4 C.
3.2.4 Clean Up 1. Add 56 μL (ratio 0.8) of Agencourt™ Ampure XP beads and

Ligated DNA mix well by pipetting.
2. Incubate the mixture at room temperature for 5 min.
4. Carefully remove and discard the supernatant.
8 Huixin Lin et al.
5. Wash the beads with 200 μL fresh 80% ethanol. Pellet the beads
on a magnetic stand and carefully remove the ethanol.
6. Repeat step 5 in Subheading 3.2.4 for a total of two washes.
7. Resuspend the magnetic beads in 21 μL of low-EDTA TE
buffer. Mix thoroughly by pipetting, and then incubate at RT
for 1 min to release the DNA from the beads.
9. Transfer 20 μL of the supernatant to a new PCR tube. Store the
library at 20 C until it is ready for QC, library quantification,
or sequencing.
10. Take appropriate amount of library to satisfy the sequencing
platform.
3.3 Data Analysis In general, the sequencing data with Fastq format from the Illu-
mina platform are raw data. Raw reads in sequencing data may
3.3.1 Clean Data
contain primer, adaptor, and low-quality reads. It is necessary to
Obtained from
perform quality control to obtain clean reads for the further data
Sequencing Data
analysis (see Note 5).
Cutadapt is a wide-used sequencing data filtering tool for high-
throughput sequencing platform at a certain fault tolerance. It can
recognize, cut, and remove the sequences of adapters, primers, and
poly-A tails, and other types of unwanted sequence. The following
steps are needed for obtaining the clean data with Cutadapt
software.
1. Install the Cutadapt using the pip command in the Linux
system:
Input the command “pip3 install –user –upgrade cutadapt”
for installing the software. (https://cutadapt.readthedocs.io/
en/stable/).
2. Trim low-quality reads: type the command “cutadapt -q 15, 10
-o output.fastq input.fastq”, then the 50 end will be trimmed
with a cutoff of 15, and the 30 end will be trimmed with a cutoff
of 10. The parameter “q” means that both the 50 end and the 30
end of each read were trimmed. The parameter “output.fastq”
means the output file, and the parameter “input.fastq” refers to
the input file.
3. Trim the adapter sequences for paired-end reads: type the
command “cutadapt -b ADAPTER_1 -b ADAPTER_2 -o
out.1.fastq -p out.2.fastq reads.1.fastq reads.2.fastq”. “b” repre-
sents that the sequence for the 5’or 30 (both possible) adapter
will be trimmed. “o” refers to output file after trimming the
adapter sequence for reads.1.fq; p means that another output
file after trimming the adapter sequence for reads.2.fq.
ALGrawany
3.3.2 Analysis of Data Primary analysis for immune repertoire usually includes three main
Obtained from processing steps: (a) align sequencing reads to reference V, D, J,
Amplification Sequencing and C genes of T-cell receptors; (b) assemble clonotypes using
for T Cell Receptor alignments results; (c) export alignment to text format file.
Repertoire MiXCR software was used for primary analysis in this part [11].
1. Install the MiXCR software
Browse the website (https://github.com/milaboratory/
mixcr/) and download latest MiXCR. MiXCR needs JAVA
environment, so we download the Java SE Development Kit
(https://www.oracle.com/technetwork/java/javase/
downloads/jdk12-downloads-5295953.html) and install it
according to the instruction. After the installation is complete,
we then start analysis of the TCR sequence by MiXCR. The
detailed procedure for the TCR analysis is as follows:
2. Primary analysis for T cell receptor repertoire
We select single analyze amplicon command in MiXCR for
T-cell receptor repertoire analysis. The command for this anal-
ysis is as follows:
java –jar mixcr.jar analyze amplicon -s <species> --

starting-material <startingMaterial> --5-end <5End> --3-end
<3End> --adapters <adapters> input_file1 [input_file2] ana-
lysis name
The parameter “species” refers to the organism, and the

value “hs” refers to HomoSapiens. Possible values for parame-
ter “starting-material” is “rna” or “dna.” The values for “5-
end” is “no-v-primers” that means no V gene primers or
v-primers refers to V gene single primer/multiple. The values
for “3-end” can be j-primers, j-c-intron-primers, c-primers,
indicating J gene single primer/multiplex, J-C intron single
primer/multiplex, C gene single primer/multiplex, respec-
tively. Two possible values are no-adapters and adapters-
present. Both input files are input_file1 and input_file2. The
parameter “analysis name” refers to the output file.
After command running is complete, a tab-delimited text
file is produced with information about all clonotypes assem-
bled by CDR3 sequence.
3.3.3 Characteristics of After analysis by MiXCR, we get the TCR information that include
TCR Results clone count,clone fraction, target sequences, and so on. In order to
obtain the characteristics of CDR3, further data analysis of the
immune repertoire is required by the R package “tcR” [12] (see
Note 6). It involves the calculation of diversity indices, calculation
of V and J gene usage. Totally, it includes the following steps:
10 Huixin Lin et al.
1. Install R language and load the R package “tcR”:

We choose one mirror to download R (https://cran.r-proj
ect.org/mirrors.html), and install it in windows system accord-
ing to the instructions. Download the R package “tcR” from
CRAN mirror (https://cran.r-project.org/web/packages/
tcR/index.html) and install the latest release version. Before
starting the program, we load the “tcR” packages, the com-
mand is as follows:
library(tcR)
2. Input data and data manipulation:

The input data for tcR are tab-delimited files which is a
default output of the MiXCR software. The following com-
mand can parse the input files (see Note 7).
immdata1 <- parse.file("~/mixcr_data/immdata1.txt",

’mixcr’)
The parameter “immdata1.txt” is the output documents of

MiXCR analysis, which needs to specify the path. The parame-
ter “immdata1” contains the input data for tcR in an available
format.
3. Descriptive statistics:
Load the data and obtain information for clonotypes and
read counts, the command is:
data(immdata1)
cloneset.stats(immdata1)
After analysis, a summary of counts of nucleotide and

amino acid clonotypes and read counts will be given.
4. Diversity evaluation:
A diversity index can be used to evaluate the repertoire
diversity. It provides more information about community com-
position than simply species richness. It computes the ecologi-
cal diversity index; the command is as follows:
repDiversity(immdata1, ’div’, ’read.count’)
It computes the Inverse Simpson Index, the command is as

follows:
repDiversity(immdata1, ’inv.simp’, ’read.prop’)

ALGrawany
It computes the Gini-Simpson index, the command is as

follows:
repDiversity(immdata1, ’gini.simp’, ’read.prop’)
It computes diversity of repertoire using Chao index, the

command is as follows:
repDiversity(immdata1, ’chao1’, ’read.count’)
4 Notes
1. Till now, the β chain is the main target studied because of its
higher combinatorial potential compared to the α chain, which
is due to the presence of the D gene component. And the
CDR3 region is the preferential target for most of TCR reper-
toire studies, due to its direct interaction with TCR peptide.
Thus, the protocol only involves the TCRβ-CDR3 region.
2. All the centrifugation steps are recommended to perform at
room temperature (15–25 C) unless indicated.
3. If the surface of FFPE sample has been exposed to air, discard-
ing the first 2–3 sections is necessary.
4. Longer incubation times or higher incubation temperatures in
Buffer ATL may make DNA more fragmented.
5. The data analysis is based on paired-end sequencing data and
suitable for DNA sequence using amplicon sequencing
method.
6. Although MiXCR and tcR are excellent software for analysis of
TCR repertoires, there are other tools that can be used in this
kind of data analysis.
7. When the output results from MiXCR are used as input data for
tcR analysis, the cloneCount number should be the
integer data.
References
1. Rosati E et al (2017) Overview of methodolo- 3. Burtrum DB et al (1996) TCR gene recombi-
gies for T-cell receptor repertoire analysis. nation and alpha beta-gamma delta lineage
BMC Biotechnol 17(1):61 divergence: productive TCR-beta rearrange-
2. Turner SJ et al (2006) Structural determinants ment is neither exclusive nor preclusive of
of T-cell receptor bias in immunity. Nat Rev gamma delta cell development. J Immunol
Immunol 6(12):883–894 157(10):4293–4296
12 Huixin Lin et al.
4. Cui JH et al (2018) TCR repertoire as a Novel 8. Carlson CS et al (2013) Using synthetic tem-
indicator for immune monitoring and progno- plates to design an unbiased multiplex PCR
sis assessment of patients with cervical cancer. assay. Nat Commun 4:2680
Front Immunol 9:2729 9. Linnemann C et al (2013) High-throughput
5. Jia Q et al (2015) Diversity index of mucosal identification of antigen-specific TCRs by TCR
resident T lymphocyte repertoire predicts clin- gene capture. Nat Med 19(11):1534–1541
ical prognosis in gastric cancer. Onco Targets 10. Mamedov IZ et al (2013) Preparing unbiased
Ther 4(4):e1001230 T-cell receptor and antibody cDNA libraries for
6. Lin KR et al (2018) T cell receptor repertoire the deep next generation sequencing profiling.
profiling predicts the prognosis of Front Immunol 4:456
HBV-associated hepatocellular carcinoma. 11. Bolotin DA et al (2015) MiXCR: software for
Cancer Med 7(8):3755–3762 comprehensive adaptive immunity profiling.
7. Liu YY et al (2019) Characteristics and prog- Nat Methods 12(5):380–381
nostic significance of profiling the peripheral 12. Nazarov VI et al (2015) tcR: an R package for
blood T-cell receptor repertoire in patients T cell receptor repertoire advanced data analy-
with advanced lung cancer. Int J Cancer 145 sis. BMC Bioinform 16:175
(5):1423–1431. https://doi.org/10.1002/
ijc.32145
Chapter 2
Nanopore Sequencing and Its Clinical Applications

Xue Sun, Lei Song, Wenjuan Yang, Lili Zhang, Meng Liu, Xiaoshuang Li,
Geng Tian, and Weiwei Wang
Abstract
Nanopore sequencing is a method for determining the order and modifications of DNA/RNA nucleotides
by detecting the electric current variations when DNA/RNA oligonucleotides pass through the
nanometer-sized hole (nanopore). Nanopore-based DNA analysis techniques have been commercialized
by Oxford Nanopore Technologies, NabSys, and Sequenom, and widely used in scientific researches
recently including human genomics, cancer, metagenomics, plant sciences, etc., moreover, it also has
potential applications in the field of healthcare due to its fast turn-around time, portable and real-time
data analysis. Those features make it a promising technology for the point-of-care testing (POCT) and its
potential clinical applications are briefly discussed in this chapter.
Key words Nanopore sequencing, point-of-care testing, Structure variation, Tumor suppressor gene,
Integration site
1 Introduction
Nanopore sequencing technology integrates nanopores with

semiconductor-based electronic system to distinguish among bio-
molecules including DNA, RNA, peptide, etc. by recognizing elec-
tric current variation patterns. The DNA/RNA may be forced
through the hole in different ways, for instance, electrophoresis or
enzymes that attached to the nanopore to guide the DNA/RNA
through the hole, as a long ssDNA/RNA string, one base at a time.
The amount of current varies when nanopore is blocked by A, T, C,
or G, as well as nucleotides with different modifications. The nano-
pore sequencing technology allows directly reading of long nucle-
otide sequences with modifications including methylation sites in
both DNA and RNA. Recently, nanopore-based techniques have
been widely used in scientific researches, including human geno-
mics, cancer, metagenomics, etc., its advantages, for instance, fast
13
ALGrawany
14 Xue Sun et al.
turn-around time, portable and real-time data analysis, etc. also

make it a promising technology for the point-of-care testing
(POCT) and its potential clinical applications are briefly discussed
in this chapter.
2 Identification of Cancer-Related Structure Variations Using Nanopore Sequencing
2.1 The Advantage Structure variations (SVs) include translocations, large deletions,
of Nanopore amplifications, and inversions and are closely related to human
Sequencing diseases. SVs are often driver alterations, with translocations and
in Identifying Structure amplifications to activate oncogenes, as well as deletions and inver-
Variations sions to inactivate tumor suppressor genes (TSGs). Therefore,
accurate detection of SVs is important for cancer diagnosis and
treatment. Although the current next-generation sequencing
(NGS)-based assays have been widely used in research and diagno-
sis, the short-read of NGS brings challenges to efficiently identify
SVs. The third-generation nanopore sequencing relies on Coulter
counter to read the DNA sequence by measuring the changes in
electrical current, resulting from a DNA strand being forced
through a nanometer-sized pore. The nanopore sequencing can
generate long reads and provide a promising methodology to effi-
ciently detect SVs, especially SVs contain or locate around large
Indels, tandem, and simple repeats. Although nanopore sequenc-
ing recently has a relatively high sequencing error that needs to be
fixed by error polishing/correction using bioinformatic tools, its
accuracy has been proved to be sufficient for SV detection [1–5].
2.2 Materials The work flow for nanopore sequencing with gDNA sample is
and Methods based on protocols provided by Oxford Nanopore Technologies
(https://community.nanoporetech.com/protocols/gDNA-sqk-
lsk109/v/GDE_9063_v109_revT_14Aug2019).
2.2.1 DNA Fragmentation 1. Transfer 1 μg (or 100–200 fmol) genomic DNA into a DNA
LoBind tube.
2. Adjust the volume to 49 μL with nuclease-free water.
3. Mix thoroughly by inversion, spin down, and transfer to the
Covaris g-TUBE.
4. Spin the g-TUBE for 1 min at room temperature at the speed
for the fragment size required.
5. Transfer the 49 μL fragmented DNA to a clean 1.5 mL Eppen-
dorf DNA LoBind tube.
2.2.2 DNA Repair 1. Thaw DNA CS (DCS) at room temperature, spin down, mixed
and End-Prep by pipetting, and place on ice.
Nanopore Sequencing and Its Clinical Applications 15
2. In a 0.2 mL thin-walled PCR tube, mix the following in

Table 1.
3. Mix gently by flicking the tube and spin down.
4. Using a thermal cycler, incubate at 20 C for 5 min and 65 C
for 5 min.
5. AMpure XP bead clean-up.
2.2.3 Adapter Ligation 1. In a 1.5 mL Eppendorf DNA LoBind tube, mix in the follow-
and Clean-Up ing order (Table 2).
2. Mix gently by flicking the tube and spin down.
3. Incubate the reaction for 10 min at room temperature.
4. Add 40 μL of resuspended AMPure XP beads to the reaction
and mix by flicking the tube.
5. Incubate on a Hula mixer (rotator mixer) for 5 min at room
temperature.
6. Spin down the sample and pellet on a magnet. Keep the tube
on the magnet and pipette off the supernatant.
Table 1
DNA repair reaction
Reagent Volume (μL)

DNA CS 1
DNA 47
NEBNext FFPE DNA repair buffer 3.5
NEBNext FFPE DNA repair mix 2
Ultra II End-prep reaction buffer 3.5
Ultra II End-prep enzyme mix 3
Total 60
Table 2
Library construction with nanopore adaptors (AMX)

DNA sample from the previous step 60
Ligation buffer (LNB) 25
NEBNext quick T4 DNA ligase 10
Adapter mix (AMX) 5
Total 100
16 Xue Sun et al.
7. Wash the beads by adding either 250 μL Long Fragment Buffer

(LFB) or 250 μL Short Fragment Buffer (SFB). Flick the beads
to resuspend, then return the tube to the magnetic rack and
allow the beads to pellet. Remove the supernatant using a
pipette and discard.
8. Repeat the previous step once.
9. Spin down and place the tube back on the magnet. Pipette off
any residual supernatant. Allow to dry for ~30 s, but do not dry
the pellet to the point of cracking.
10. Remove the tube from the magnetic rack and resuspend pellet
in 15 μL Elution Buffer (EB). Incubate for 10 min at room
temperature.
11. Pellet the beads on a magnet until the eluate is clear and
colorless.
12. Remove and retain 15 μL of eluate into a clean 1.5 mL Eppen-
dorf DNA LoBind tube.
2.2.4 Priming 1. Thaw the Flush Tether (FLT), Flush Buffer (FLB), Sequencing
and Loading the Flow Cell Buffer (SQB), and Loading Beads (LB) at room temperature.
When thawing is complete, place the tubes on ice.
2. Open the MinION and slide the flow cell.
3. Open the priming port and draw back a small volume of buffer
to remove bubbles using P1000.
4. Prepare the priming mix: add 30 μL of mixed Flush Tether
(FLT) to the tube of mixed Flush Buffer (FLB) and vortex
to mix.
5. Load the priming mix 800 μL into the priming port and wait
for 5 min.
6. Prepare the library for loading as below, mix the Loading Beads
(LB) before use (Table 3).
7. Open the SpotON sample port cover and load 200 μL of the
priming mix into the priming port.
Table 3
Library for loading preparing

Sequencing buffer (SQB) 37.5
Loading beads (LB) 25.5
DNA library 12
Total 75
Fig. 1 The bioinformatic pipeline to analyze nanopore sequencing data for

identification of structure variations. Reads are aligned to the reference genome
using LAST. Repeat copy number changes are identified by tandem-genotypes.
By comparing to population data, the method is prioritized to improve the
sensitivity of tandem repeat identification
8. Add 75 μL mixed sample to the SpotON sample port in a

dropwise fashion.
9. Gently replace the SpotON sample port cover, priming port
cover and the MinION lid.
10. Start sequencing using MinKNOW on a computer.
2.2.5 Data Analysis LAST, Sniffles, NanoSV, and some computational pipelines have
been used to call structural variations from nanopore data
[2, 6]. For example, tandem-genotypes were designed to identify
tandem repeat copy number [6]. The flow chart of tandem-
genotypes [6] is shown as an example (Fig. 1, [6]). This pipeline
combines LAST packages for error corrections and split alignment
to improve sensitivity of detecting the sequences with simple/
tandem repeats. By integrating multiple datasets for repeats anno-
tation and copy number correction, we are able to accurately iden-
tify the tandem repeat changes associated with diseases.
2.3 Discussion Accurate detection of tumor-related variations in a timely fashion,

including SVs, is important for patient management from early
detection to monitoring recurrence. To identify or predict cancer-
related SVs are more complicated when they locate in repeat
18 Xue Sun et al.
regions [7]. The ability of long-read sequencing methods, for

instance, nanopore sequencing, could be an ideal tool for identify-
ing cancer-associated SVs. Comparing with NGS methods, the
main advantages of nanopore sequencing are to sequence through
repeating regions, rapid and low cost.
Firstly, the long-read of nanopore sequencing (up to 2 Mb)
allows for the reading through repeated regions, while the short-
read generated with second-generation sequencing methods do
not accurately and efficiently map repeat regions. Previous studies
have shown that long-read sequencing can increase the detection
rate of SVs and identify novel SVs, in one of the studies, more than
30,000 SVs can be detected in the somatic gDNA of a single
individual [1, 8].
Secondly, real-time nanopore sequencing and data analysis can
provide results in minutes, allowing rapid diagnosis and treatment.
In order to get 99% confidence in the variant at 1:1000 of the
sample, we need about 450 coverage in the area of interest. Each
of the 512 channels can generate a read individually, and the
sequences can be generated and analyzed at the same time. By
contrast, NGS-based methods can produce millions of reads, but
usually take hours or days. The analysis cannot be carried out before
the completion of sequencing [1].
Lastly, nanopore sequencers can be low cost (about $1000 per
device) and portable, allowing point-of-care testing, while second-
generation methods require substantial up-front investment
(>$100,000) and sufficient laboratory facility and more computing
resources are required [1].
3 Identification of Cancer-Related Fusion Genes Using Nanopore Sequencing
3.1 Accurate There are about 25,000 different genes in the human genome. In
Identification of Fusion the presence of tumors, genome-level breaks and reassembles often
Genes Using occur, fusion genes are created. In most cases, fusion genes can lead
Long-Read Nanopore to abnormal transcripts and proteins, or gene expression disorders.
Sequencing Detection of gene fusion events is important for clinical diagnosis
and prognosis. Some fusion genes are reported to be drug targets,
and patients with such mutation may achieve complete remission.
Tests for fusion genes can guide clinicians to develop a personalized
therapy and avoid excessive or inadequate treatment. Fusion gene
detection plays an important role in the selection of tumor-targeted
drugs [9].
Recently, the nanopore sequencing technology is widely
accepted on the market and MinION nanopore sequencing instru-
ment from Oxford Nanopore Technologies (ONT) is one of the
commercialized products. Its characteristics are single molecule
sequencing, long sequencing reading, fast sequencing speed, real-
time monitoring of sequencing data, and convenient carrying [10].
ALGrawany
3.2 Materials This work flow describes a method of targeted nanopore sequenc-
and Methods ing by combining protocols of the Ligation Sequencing Kit 1D and
the IDT xGen Lockdown. The overall work flow includes steps
3.2.1 Work Flow
from DNA fragmentation, target sequence enrichment, to nano-
of Probe-Based Enrichment
pore library construction [11] (Fig. 2).
of Target Sequences
DNA was extracted using QIAamp DNA Blood Mini Kit (Qia-
gen, Germany) and fragmented to 1–2 kb using fragmentase or
ultrasound instrument. DNA library was constructed with the
Ligation Sequencing Kit 1D (SQK-LSK108). The enrichment of
DNA is using xGen Hybridization and Wash Kit. A customized
blocking oligo is designed based on nanopore library adaptors and
used in the hybridization [11].
3.2.2 Bioinformatic Tools Traditional NGS sequence alignment software cannot meet the
for Fusion Junction requirements of MinION sequence alignment because the error
Identification rate of MinION sequencing data is relatively high, and even adjust-
ing parameters cannot achieve any better alignment rate. Margin-
Align is an alignment software, which is optimized based on LAST
or BWA mem. MarginAlign can evaluate and correct the source of
Minion sequencing errors, thus increasing the alignment efficiency
with reference genome [12, 13]. High-confidence alignments
ensure accurate identification of gene fusion and rearrangement.
3.3 Discussion The short reads of NGS affect the accuracy of fusion gene detec-
tion. The problem is particularly severe in cancer samples. Based on
previous studies, structural variations from long reads with
hundreds of copies measured by MinION were more reliable than
those from the millions of short reads sequenced by the NGS
platform. Currently, the accuracy of MinION sequencing is about
92%. For the discovery of pathogenic bacteria and alternative splic-
ing, such sequencing accuracy can meet the demands, but for
clinical tests require higher accuracy [12], with the continuous
optimization of sequencing-related chemistry and base calling soft-
ware, we can expert better sequence quality with longer read length
and nanopore sequencing technology will have broader clinical
applications.
4 Applications of Nanopore Sequencing in Human Genetic Diseases
4.1 Improved Next-generation sequencing technologies provide rapid and rela-

Diagnostics in Genetic tively cost-effective genomic sequencing. The technologies have
Diseases revolutionized the field of human genetics. However, there are
with Long-Read still limitations due to the short-read, for instance, making the
Nanopore Sequencing detection of repetitive regions and large structural variations very
challenging. Unlike short-read sequencing platforms, the nanopore
sequencing technology can process complete fragments and offer
ultra-long-read lengths over 2 Mb, thus provides a more complete
20 Xue Sun et al.
gDNA
Fragmentation
50 min End-prep
A
A
T
PCR adapter
ligation
PCR with rapid F

attachment primers R
User-
defined
Capture of probe-
template duplexes
onto beads
Elute and PCR with rapid F

attachment primers R
Attachment of rapid 1D
sequencing adapters
10 min
Loading
Fig. 2 The work flow of probe-based enrichment of target sequences and nanopore library construction.
Specific probes were designed to capture long fragments containing gene fusion junction regions. The
enriched target fragments were re-amplified and built into nanopore library
view of genetic variation. Researches can benefit from the nanopore

technology to carry out a variety of experiments including whole
genome and targeted sequencing.
For researchers who are interested in specific areas at high
depth, the targeted sequencing approach is commonly employed.
There are a range of targeted sequencing methodologies which are
compatible with nanopore sequencing. Besides typical PCR and
probe capture-based enrichment strategies, CRISPR/Cas9-based
enrichment method also been developed for long, targeted DNA
molecules [14].
4.2 Long-range A team of researchers from the UK and USA utilized the Oxford
PCR-Based Targeted Nanopore MinION in combination with long-range PCR to
Sequencing amplify and sequence the entire ~8 kb gene GBA [15]. Homozy-
gous or biallelic mutations in the GBA gene can cause Gaucher
disease (GD), the most common lysosomal storage disorder. PCR
and traditional short-read DNA sequencing of GBA gene is com-
plicated by the nearby pseudogenes (with up to 96% homology).
The team design and validate a method for sequencing GBA using
nanopore long-read sequencing. They extracted DNA, amplified an
8.9 kb sequence, then carried the barcoding step, library prepara-
tion, and sequencing. The data analysis work flow is illustrated in
Fig. 3.
According to the results, the team concluded that the nanopore
sequencing technology can detect missense mutations and an
exonic deletion in the difficult gene GBA, with the added advantage
of phasing (Fig. 4).
4.3 Targeted, Conventional amplification-based enrichment methods can be lim-

Amplification-Free ited by base composition (e.g., GC-rich content), bias (e.g., allele
DNA Sequencing Using bias), and PCR product length. The CRISPR/Cas9 techniques to
CRISPR/Cas9 enrich for specific regions of interest can solve these challenges with
preserving the epigenetic modification information at the
same time.
Timothy Gilpatrick from Johns Hopkins University utilized the
CRISPR/Cas9 technique to detect 10/11 genomic loci with a
median length of 18 kb by using the Oxford Nanopore MinION
BASECALLING Reads base called and Reads mapped to reference

& DEMULTIPLEXING MAP TO REFERENCE
demultiplexed (Albacore now genome (hg19) using NGMLR.
Raw data.
recommended platform). FASTQ Coverage calculated using
file output. Only ‘pass’ reads bedtools.
further analysed.
VARIANT CALLING
SNVs detected using PHASING True variants phased

Nanopolish and structural using WhatsHap.
variants detected using Sniffles.
Fig. 3 Bioinformatic pipeline designed for identifying SNVs and long Indels in GBA gene
22 Xue Sun et al.
Fig. 4 Detection and phasing of a 55‐base pair exonic deletion in one sample. Part of the reads are shown
(Reference sequence NM_000157.3). The arrows point to eight selected SNVs (red: coding SNVs; blue:
noncoding SNVs). Red box in this figure is the deletion. Red-colored reads and blue-colored reads are the
different haplotypes
and Flongle [16]. The sequencing depth generated by the two flow
cells were 165X and 30X, respectively. Their study showed that the
CRISPR/Cas9 technique can simultaneously assessing single
nucleotide variants (SNVs), structural variants (SVs), and CpG
methylation. Figure 5 shows the schematic of Cas9 enrichment
operation. Figure 6 demonstrates the structural variation results.
5 Detection of Integration Sites of Cancer-Related Viruses
5.1 Higher Detection Approximately 15–20% of all cancers worldwide are associated with
Rate of Viral viral infections. To date, at least five DNA viruses, Epstein–Barr
Integrations Using virus (EBV), Human papilloma virus (HPV), Merkel cell polyoma-
Long-Read Nanopore virus (MCV), Kaposi’s sarcoma-associated herpesvirus (KSHV or
Sequencing HHV-8), and Hepatitis B virus (HBV), and three RNA viruses,
Hepatitis C virus (HCV), Human T lymphotropic virus type-1
(HTLV-1), and human immunodeficiency virus (HIV) have been
shown to contribute to the development of human cancers though
this number is likely to increase over time [17–20].
Previous studies have identified viruses in the tissues of cancer
patient [21–26]. HPV sequences were detected in nearly all cervical
carcinomas as well as in a subset of squamous cell carcinomas of the
head and neck; HBV and HCV were reported to be associated with
a subset of liver cancers; EBV gene expressed in a subset of gastric
cancers. Oncogenic viruses contribute to tumorigenesis by induc-
ing transformation of the infected cells. Viruses may induce sus-
tained disorders of host cell growth and survival. Viral integration
may also trigger DNA damage response (DDR) that many viruses
need for their replication, and increases host genome
instability [17].
ROI
P
P
P
P
P
P
Dephosphorylate
DNA ends Cas9
Introduce cuts
with Cas9
P
P
P P
End-Prep motor
+ protein
Adaptor Ligation adaptor
P dA
dA P
Load to 5’
Sequencer
5’
Fig. 5 Procedure of the Cas9 enrichment. ROI region of interest
The integration of viral DNA into the host genome is a critical

step in their life cycle. Viral DNA integration into the host genome
is considered one of the most important risk factors for the devel-
opment of carcinoma [27–29]. It has been of great interest to
understand the implications of integration and to determine
whether it is involved in tumor formation [30–35].
The application of next-generation sequencing (NGS) has dra-
matically enhance our ability to explore the landscape of viral
integration of both DNA and RNA viruses [32, 35]. However,
there is a particularly challenge in identifying integrations within
repeats or with structural rearrangements by using short-read
sequencing. The development of long-read sequencing technology
has the potential to increase the detection rate of viral integrations
in repeat regions. We introduce two target-enrichment methods for
identification of viral integration sites with long-read sequencing.
24 Xue Sun et al.
Breast Cancer: 6kb deletion, chr5

a GM12878: GM12878: b coverage [0-80]
MCF-10A
72kb deletion, chr5 69kb deletion, chr6 reads
400 72 kb 400 coverage [0-200]

200 200 69 kb
Log2 Coverage
Log2 Coverage
100 100 reads
MDA-MB-231
50 50
25 25
12 12
6 6 coverage [0-80]
3 3 MCF-7 reads
0 0 18 kb
105100000 105160000 78260000 78320000 6 kb
Genomic Coord (Chr5) Genomic Coord (Chr5)
Breast Cancer: 8kb deletion, chr7
paternal coverage
maternal coverage coverage [0-400]
GM12878:
155kb deletion, chr8 MCF-10A reads
400
200 155 kb
Log2 Coverage
100 coverage [0-400]

50
reads
25 MDA-MB-231
12
6
3
coverage [0-150]
0 MCF-7 reads
39400000 39500000
Genomic Coord(Chr8) 20 kb
8 kb
Fig. 6 Structural variation results: (a) There are three deletions in the GM12878 lymphoblast cell line and the
reads are segregated into paternal and maternal allele. Yellow triangles are the Cas9 cut site. (b) Two
deletions detected in cell lines MDA-MB-231 and MCF-7 except for MCF-10A cell line
5.2 Materials The Xdrop technology is applied for identification of HPV18 inte-
and Methods gration sites (Fig. 7), and main steps were described in details as
follows.
5.2.1 Xdrop Technology
for Detecting HPV18 1. PCR and droplet chemicals
Integration Sites Carcinoma cell line DNA (New England Biolabs) was
diluted with DNase-free water (Gibco) to 0.5 ng per μL prior
to use. PCR-mix for 20 μL was set up as Table 4.
All reagent provided by Thermo Scientific. The sequences
of primers were shown in Table 5.
For the primary droplet production 3% fluorosurfactant
(RAN Biotechnologies) in Novec HFE-7500 was used as car-
rier phase. For the secondary droplets, a DE-buffer containing
1.5 Optima buffer (40 mM Tris–HCl, 60 mM Trizma-base,
25 mM (NH4)2SO4, 0.015% Tween 80, 45 mM NaCl) and 3%
glycerol was used as carrier phase.
2. Droplet production
Double emulsion droplets were produced using a two-step
emulsification procedure, initially creating water in oil (W-O)
droplets followed by second emulsification to create water-in-
oil-in-water (W-O-W) droplets.
Primers
Sample DNA
1. PCR reagents and DNA 2. Encapsulate 3. PCR detection 4. Sort 5. Collect 6. Amplify DNA
Fig. 7 Overview of Xdrop enrichment work flow. PCR reagents including primers are mixed with sample DNA
(1) before being encapsulated in DE droplets (2). Droplet PCR allows fluorescence-based detection of the DNA
molecules of interest (3) that are then sorted out on a cell sorter (4). The DNA from the sorted droplets is
initially collected (5) and amplified using droplet MDA (6). The orange DNA helixes depicts the target DNA of
interest and grey DNA helixes depicts non-target DNA
Table 4
Reaction mix for droplet amplification

10 PCR-buffer without detergent 2
25 mM MgCl2 2
2 mM dNTP 2
50% Glycerol 1.2
GoTaq polymerase (5 U/μL) 0.4
Bovine serum albumin (2 mg/mL) 0.25
Forward/reverse primer (10 μM) 0.8
TP1 Probe (μM) 0.6
0.5 ng/μL DNA 1
Water 8.95
Total 20
Table 5
Specific primers for target viral sequences
Primer Sequences (50 –30 )

Forward primer TGTGCTGGAGTGGAAATTGG
Reverse primer GGCATGGGAACTTTCAGTGTC
TP1 probe FAM-CAACACCTAAAGGCTGACCACGG-BHQ1
26 Xue Sun et al.
3. Primary droplets (W-O)

The initial chip used to prepare the primary emulsion was a
14 μm etch depth hydrophobic “Small Droplet Chip, 14 μm”
(Dolomite Microfluidics). Liquids were pushed into the micro-
fluidic chip using MFCS-EZ pressure controller (Fluigent,
Germany) applying pressures of 640 mbar on the primary
sample (PCR) and 650 mbar to the secondary liquid (Oil).
Droplet production was done for about 40 min, processing a
total of 40 μL PCR mixture.
4. Secondary emulsions (W-O-W)
Immediately following primary production, droplets were
collected in PTFE-tubes using a 1 mL syringe (Scientific Glass
Engineering, Australia) ensuring an air-free liquid system to
pull the droplets into the tube. The tube was then connected to
the inlet-position of the 4-way Linear Connector (Dolomite
microfluidics, UK connector (Part number 3000024). Dro-
plets were pushed into the chip at 0.25 μL/min using a Legato
110 syringe pump (KD Scientific). During secondary droplet
production, spacer oil was applied into the chip using a syringe
system identical to that carrying the droplets, delivering oil to
space the introduced droplets prior to the second emulsifica-
tion. Spacer oil was connected to position 2 in the connector
using a syringe pump set to deliver a flow of 0.40 μL/min.
Double emulsion buffer (DE) was introduced to the chip using
a Legato 100 (KD Scientific) single syringe pump applying
pressure to a 10 mL syringe (Scientific Glass Engineering,
Australia). Pump speed of the DE-buffer was set to 28 μL/
min. Second emulsification was performed for 160 min until all
primary droplets had passed the junction of the DE-chip.
5. Droplet sorting and gating
Sorting was carried out on a single laser 488 nm, S3e cell
sorter (BioRad inc.) using ProSort software (v. 1.3b). Instru-
ment PMTs were adjusted to: FSC ¼ 239, SSC ¼ 261,
FL1 ¼ 590, and FL2 ¼ 367. FSC was used as primary thresh-
old and the value was set to 1.00. Gating of positive droplets
was performed in three consecutive gating events. First gate
was set to discriminate between double emulsion droplets and
“other” elements in the carrier buffer. The second gate was
used to split the double emulsion droplets from the first gate
into fluorescent and non-fluorescent double emulsion droplets.
The third gate was applied to ensure that only positive droplets
with the expected properties were sorted. Sorting purity was
set to “Enrich” and event rate was kept as close to 4000
events/second as possible throughout the experiment. Prior
to sorting droplets, 5 μL Tris (10 mM) was placed at the
bottom of the 1.5 mL collection tube to avoid disrupting the
sorted droplets. Sorting was done for a period of 27 min and a
total of 143 positive double emulsion droplets were sorted.

Upon completed sorting, the collection tube was centrifuged
at 1000 g for 10 s to collect any liquid from the side of the
tube, arising from the splash impact of sorted droplets hitting
the liquid surface of inside tube.
6. DNA amplification
The collected droplets were coalesced by adding 20 μL of
PicoBreak (Sphere Fluidics), mixing, and centrifuging the sam-
ple. 3 μL of the resulting aqueous-phase was used as template
for a multiple displacement amplification (MDA) reaction. The
MDA reaction mix was kindly provided by Samplix. The MDA
reaction was emulsified on a x-junction droplet generator chip
(ChipShop) using 1% PicoSurf in 7500-Novec oil as carrier
phase. The droplet production was driven by air pressure con-
trolled by pressure regulator (Fluigent). The droplets contain-
ing the MDA reaction were incubated for 16 h at 30 C
followed by 10 min at 65 C to terminate the reaction. 6ul
from the MDA reaction was used as template for a second
droplet MDA reaction. After each MDA round the emulsions
where coalesced using 20 μL PicoBreak.
7. Nanopore sequencing
An ONT library was produced from 400 ng of enriched
DNA using the Rapid Sequencing (SQK-RAD003) protocol.
The library was sequenced on 1 MIN106 flowcell (R9.4) for
17.5 h with subsequent Albacore basecalling (v2.3.4). All was
performed using standard settings according to manufacturer’s
recommendations.
8. Detection of HPV18 integration sites
HPV18 integration sites were identified by mapping all
sequence reads to the HPV18 reference genome. The reads
mapping to HPV18 were subsequently re-mapped to the
human reference genome GRCh38 (hg38). Reads mapping
to both genomes were considered HPV18/Chr8 fusion-reads
(Fig. 8).
9. Sanger sequencing
PCR primers located on each side of the fusion points were
designed and used to generate PCR amplicons across the fusion
point. These PCR products were Sanger sequenced using one
of the PCR primers.
5.2.2 Pooled CRISPR 1. Work flow of RCIP-seq

Inverse PCR Sequencing DNA isolation was extracted using the Qiagen AllPrep
(PCIP-seq) for Sequencing DNA/RNA/miRNA kit. High molecular weight DNA was
the Integration Sites and Its sheared to ~8 kb using Covaris g-tubes™ (Woburn, MA) or a
Associated Provirus Megaruptor (Diagenode), and then was used to make the
libraries by end-repair using the NEB Next End Repair Module
28 Xue Sun et al.
Fig. 8 Detection of HPV18 integration sites by Xdrop enrichment and long-read sequencing. (a) Overview of
the complete HPV18 genome with genes depicted in grey. Below the three types of integrated HPV18 found in
the HeLa genome. The breakpoints are shown with numbered circles. (b) Overview of the HPV18 fusion points
and the suggested structure of the integrations identified in the HeLa genome. The positions of the
chromosome 8 integration sites refer to the GRCh38 genome assembly. The position of the primers used
for enrichment is shown with red P’s. The numbered circles correspond to the numbering in (a)
(New England Biolabs). Intramolecular circularization was

incubated by overnight at 16 Cwith T4 DNA Ligase. Remain-
ing linear DNA was removed with Plasmid-Safe-ATP-Depen-
dent DNAse (Epicentre, Madison WI). Guide RNAs and the
final oligo sequence were designed using chopchop (http://
chopchop.cbu.uib.no/index.php) and the EnGen™ sgRNA
Template Oligo Designer (http://nebiocalculator.neb.com/
#!/sgrna), respectively. Oligos were synthesized by Integrated
DNA Technologies (IDT). Oligos were pooled and guide
RNAs synthesized with the EnGen sgRNA Synthesis kit,
S. pyogenes (New England Biolabs). Selective linearization
reactions were carried out with the Cas-9 nuclease,
S. pyogenes (New England Biolabs). PCR primers which
were tailed to facilitate the addition of Oxford Nanopore
indexes in a subsequent PCR reaction were designed using
primer3 (http://bioinfo.ut.ee/primer3/). The linearized frag-
ments were amplified with LongAmp Taq DNA Polymerase
(New England Biolabs) and a second PCR added the appropri-
ate Oxford Nanopore index. PCR products were verified via gel
electrophoresis and quantified on a nanodrop spectrophotom-
eter. Indexed PCR products were multiplexed and Oxford
Nanopore libraries performed either using the Ligation
Sequencing Kit 1D (SQK-LSK109). The resulting libraries
Fig. 9 Overview of the PCIP-seq method. (a) Simplified outline of method (b) A pool of CRISPR guide-RNAs
targets each region, the region is flanked by PCR primers. Guides and primers adjacent to 50 & 30 LTRs are
multiplexed. (c) As the region between the PCR primers is not sequenced, we created two sets of guides and
primers. Following circularization, the sample is split, with CRISPR-mediated cleavage and PCR occurring
separately for each set. After PCR, the products of the two sets of guides and primers are combined for
sequencing
were sequenced on Oxford Nanopore MinION R9.4 flow cells,

respectively, and basecalled using albacore 2.3.4 according to
the manufacturer’s instructions. Only the 1D reads from both
flow cell versions were used (Fig. 9a).
2. Bioinformatic pipeline for identification of viral
integration sites
Reads were mapped with Minimap (https://atom.io/
packages/minimap) to the host genome with the viral genome
as a separate chromosome (Fig. 9b). In-house R-scripts were
used to identify integration sites (IS). Briefly, chimeric reads
that partially mapped to at least one extremity of the viral
genome were used to extract virus-host junctions and shear
sites (Fig. 9c). Junctions within a 200 bp window were clus-
tered together to form an “IS cluster,” compensating for
sequencing/mapping errors. The IS retained corresponded to
the position supported by the highest number of virus-host
junctions in each IS cluster. Clone abundance was estimated
based on the number of reads supporting each IS cluster, reads
with the same shear site were considered PCR duplicates.
30 Xue Sun et al.
5.3 Discussions There is a wide application of target enrichment methods for short-
read sequencing technologies, but only few are compatible with the
long-read sequencing platforms [36]. In this report, we described
two enrichment methods that compatible with nanopore sequenc-
ing. The Xdrop technology isolates long DNA fragments by stan-
dard laboratory cell sorter (FACS) of double emulsion
(DE) droplets [37]. A unique advantage of the Xdrop work flow
is that the DNA amplification can be initiated from femtogram
amounts of target DNA, compared to other enrichment protocols
were nanograms, or even micrograms, of DNA is required. The
method described here can conveniently be applied to other tar-
gets, where structural information is sought, by design of a simple
primer-set. With an efficient droplet production and optimization
of the final amplification, the procedure can be completed in less
than 24 h.
PCIP-seq can be utilized to identify integration sites while also
sequencing the associated provirus [38]. For integration site iden-
tification, the method was capable of identifying more than ten
thousand BLV integration sites in a single sample, using ~4 μg of
template DNA. Even in samples with a PVL of 0.66%, it was
possible to identify hundreds of integration sites with only 1 μg of
DNA as template. The improved performance of PCIP-seq in
repetitive regions further highlights its utility, strictly from the
standpoint of integration site identification. In addition to its appli-
cation in research, high-throughput sequencing of virus integration
sites has shown promising clinical tool to monitor viral progression,
especially in a clinical context to track clonal evolutions in areas with
poor biomedical infrastructure. Other potential applications
include determining the integration sites and integrity of retroviral
vectors, detecting transgenes in genetically modified organisms or
identifying on target CRISPR–Cas9-mediated structural rearrange-
ment that could be missed by conventional long-range PCR.
Hybridization directly using short DNA- or RNA-probes are
also used to pull down the target sequences. Hybridization-based
methods can be used to enrich fragments with size up to 10 kb
using short probes (<200 bp), but require relatively large amounts
of input DNA (>500 ng) [39]. There are large variety of techni-
ques used to detect integration sites based on previous studies
[33]. Target enrichment can be a highly cost- and time-effective
and a promising technology for point-of-care testing to identify the
disease-associated variants [40].
References
1. Norris AL, Workman RE, Fan YF et al (2016) 15. Leija‐Salazar M et al (2019) Evaluation of the
Nanopore sequencing detects structural var- detection of GBA missense mutations and
iants in cancer[J]. Cancer Biol Ther 17 other variants using the Oxford Nanopore
(3):246–253 MinION. Mol Genet Genomic Med 7(3):e564
2. Gong L, Wong C, Cheng W et al (2017) Nano- 16. Gilpatrick T et al (2019) Targeted nanopore
pore sequencing reveals high-resolution struc- sequencing with Cas9 for studies of methyla-
tural variation in the cancer genome tion, structural variants, and mutations.
[J]. bioRxiv BioRxiv:604173
3. Sanchisjuan A, Stephens J, French C et al 17. Mesri E, Feitelson MA, Munger K (2014)
(2018) Complex structural variants in Mende- Human viral oncogenesis: a cancer Hallmarks
lian disorders: identification and breakpoint analysis. Cell Host Microbe 15(3):266–282
resolution using short- and long-read genome 18. Plummer M, Martel CD, Vignat J, Ferlay J,
sequencing[J]. Genome Med 10(1):1–10 Bray F, Franceschi S (2016) Global burden of
4. https://nanoporetech.com/applications cancers attributable to infections in 2012: a
5. Miao HF, Zhou JP, Yang Q et al (2018) Long- synthetic analysis. Lancet Glob Health 4(9):
read sequencing identified a causal structural e609–e616
variant in an exome-negative case and enabled 19. Tashiro H, Brenner MK (2017) Immunother-
preimplantation genetic diagnosis. Hereditas apy against cancer-related viruses. Cell Res 27
155(32) (1):59–73
6. Mitsuhashi S, Frith MC, Mizuguchi T et al 20. Tuna M, Amos CI (2017) Next generation
(2019) Tandem-genotypes: robust detection sequencing and its applications in
of tandem repeat expansions from long DNA HPV-associated cancers. Oncotarget 8
reads[J]. Genome Biol 20(1):58 (5):8877
7. Stancu MC, Van Roosmalen MJ, Renkens I 21. Strong MJ, Tina OG, Zhen L, Guorong X,
et al (2017) Mapping and phasing of structural Melody B, Chris P, Kun Z, Taylor CM, Fle-
variation in patient genomes using nanopore mington EK (2013) Epstein-Barr virus and
sequencing[J]. Nat Commun 8(1):1326–1338 human herpesvirus 6 detection in a non-Hodg-
8. Au CH, Dona N, Beca B (2019) K. Ip et al. kin’s diffuse large B-cell lymphoma Cohort by
Rapid detection of chromosomal translocation using RNA sequencing. J Virol 87
and precise breakpoint characterization in (23):13059–13062
acute myeloid leukemia by nanopore long- 22. Tang KW, Alaeimahabadi B, Samuelsson T,
read sequencing. Cancer Genet 239:22–25 Lindh M, Larsson E (2013) The landscape of
9. Soda M, Choi YL, Enomoto M et al (2007) viral expression and host gene fusion and adap-
Identification of the transforming EML4–ALK tation in human cancer. Nat Commun 4
fusion gene in non-small-cell lung cancerJ. (2513):2513
Nature 448(7153):561–566 23. Network CGA (2014) Comprehensive molec-
10. Jain M, Olsen HE, Paten B et al (2016) The ular characterization of gastric adenocarci-
Oxford Nanopore MinION: delivery of nano- noma. Nature 513(7517):202–209
pore sequencing to the genomics communityJ. 24. Network CGA (2014) Comprehensive molec-
Genome Biol 17(1):239 ular characterization of urothelial bladder car-
11. Nanopore Protocol 1D Sequence capture cinoma. Nature 507(7492):315–322
(SQK-LSK108) 25. Network CGA (2015) Comprehensive geno-
12. Jain M, Fiddes IT, Miga KH et al (2015) mic characterization of head and neck squa-
Improved data analysis for the MinION nano- mous cell carcinomas. Nature 517(7536):576
pore sequencer[J]. Nat Methods 12(4):351 26. Cantalupo PG, Katz JP, Pipas JM (2017) Viral
13. https://github.com/benedictpaten/ sequences in human cancer. Virology
marginAlign 513:208–216
14. Oxford Nanopore Technologies. Incorporat- 27. Hanahan D, Weinberg R (2011) Hallmarks of
ing sequence capture into library preparation cancer: the next generation. Cell 144
for MinION GridION and PromethION. (5):646–674
Online. Available at: https://nanoporetech. 28. Agalioti T, Lomvardas S, Parekh AB, Yie J,
com/resource-centre/incorporating- Maniatis T, Thanos D, Hanahan D, Weinberg
sequence-capture-library-preparation-minion- RA (2000) The hallmarks of cancer. Cell 100
gridion-and-promethion. Accessed 29 August, (1):57–70
2019
32 Xue Sun et al.
29. Gutschner T, Diederichs S (2012) The hall- maintenance of the reservoir. Curr Opin HIV
marks of cancer. RNA Biol 9(6):703–719 and AIDS 13(2):152
30. Xu M, Zhang WL, Zhu Q, Yao YY, Feng QS, 36. Ameur A, Kloosterman WP, Hestand MS
Zhang Z, Peng RJ, Jia WH, He GP, Feng L (2019) Single-molecule sequencing: towards
(2019) Genome-wide profiling of Epstein-Barr clinical applications. Trends Biotechnol 37
virus integration by targeted sequencing in (1):72–85
Epstein-Barr virus associated malignancies. 37. Madsen EB, Kvist T, Höijer I, Ameur A, Mik-
Theranostics 9(4):1115 kelsen MJ (2018) Xdrop: targeted sequencing
31. Murakami Y, Saigo K, Takashima H, of long DNA molecules from low input sam-
Minami M, Okanoue T, Bréchot C, Paterlini- ples using droplet sorting. bioRxiv:409086
bréchot P (2005) Large scaled analysis of hep- 38. Artesi M, Hahaut V, Ashrafi F, Marçais A,
atitis B virus (HBV) DNA integration in HBV Hermine O, Griebel P, Arsic N, van der
related hepatocellular carcinomas. Gut 54 Meer F, Burny A, Bron D et al (2019) Pooled
(8):1162–1168 CRISPR Inverse PCR sequencing (PCIP-seq):
32. Pett M, Coleman N (2010) Integration of simultaneous sequencing of retroviral insertion
high-risk human papillomavirus: a key event points and the associated provirus in thousands
in cervical carcinogenesis? J Pathol 212 of cells with long reads. bioRxiv:558130
(4):356–367 39. Wang M, Beck CR, English AC, Meng Q,
33. Pinatti LM, Walline HM, Carey TE (2018) Buhay C, Han Y, Doddapaneni HV, Yu F,
Human papillomavirus genome integration Boerwinkle E, Lupski JR (2015) PacBio-
and head and neck cancer. J Dent Res 97 LITS: a large-insert targeted sequencing
(6):691–700 method for characterization of human disease-
34. Zhang J, Huang T, Zhou Y, Cheng A, Yu J, To associated chromosomal structural variations.
KF, Kang W (2018) The oncogenic role of BMC Genomics 16(1):214
Epstein-Barr virus-encoded microRNAs in 40. Mamanova L, Coffey AJ, Scott CE, et al.
Epstein-Barr virus-associated gastric carci- (2010) Target-enrichment strategies for next-
noma. J Cell Mol Med 22(1):38–45 generation sequencing (vol 7, pg 111) [J]. Nat
35. Symons J, Cameron PU, Lewin SR (2018) Methods 7(6):479–479
HIV integration sites and implications for
Chapter 3
The Clinical Significance of Microsatellite Instability

in Precision Treatment
Zhenyu Huang, Xiaojian Chen, Chenying Liu, and Long Cui
Abstract
The recent years have seen the high heterogeneity of colorectal cancer (CRC) receiving increasing attention
and being revealed step by step. Microsatellite instability (MSI), characterized by the dysfunction of
mismatch repair gene, plays an important role in the heterogeneity of colorectal cancer. MSI status can
be identified by immunohistochemistry for MMR protein such as MLH1, MSH2, PMS2, and MSH6 or
PCR-based array for MMR gene. Recent studies have revealed MSI status is the only biomarker that can be
used to select patients with high-risk stage II colon cancer for adjuvant chemotherapy. Furthermore, it
always indicated better stage-adjusted survival when compared with microsatellite stable (MSS) tumors. For
immunotherapy, patients with MSI tumors exhibited significant response to anti-PD-1 inhibitors after the
failure to conventional therapy. In this chapter, we discuss the detection methods of MSI, the prognostic
value of MSI, and its clinical guiding value in the management of precision therapy.
Key words Microsatellite instability, colorectal cancer, microsatellite stable, DNA mismatch repair,
immunotherapy
1 Introduction
Recent study has demonstrated colorectal cancer (CRC) as the

third most common cancer in males and the second in females
[1]. Among these abundant cases, CRC is a heterogeneous and
molecularly complex disease [2]. The complexity of CRC leads to
different prognosis in patients with CRC, as well as different
response to conventional treatment and novel-target therapy,
which requires more accurate and individualized treatment towards
all kinds of subgroups.
In early-staged CRC, including stage I-II, high-risk stage II,
stage III which exist with regional lymph node metastases, we all
have an agreed account of that deficient on DNA mismatch repair
(dMMR) genes is the only biomarker for them to decide whether
they can get survival benefits through adjuvant chemotherapy
[3, 4]. MSI (microsatellite instability) which is mainly caused by
33
34 Zhenyu Huang et al.
dMMR is beyond doubt that the most valuable genetic characteris-

tic and an integral part of tumor heterogeneity. MSI related early-
staged CRC display high degrees of microsatellite instability as a
result of defects in genes involved in the DNA mismatch repair
(MMR) pathway such as MLH1, MSH2, MSH6, and PMS2 [5] or
a hypermethylation of the MHL1 promoter [6, 7]. Consequently
accumulation of high-level mutations leads to cancer susceptibility.
In this review, we discuss the detection methods of MSI, the
prognostic value of MSI, and its clinical guiding value in the man-
agement of precision therapy.
2 Cause of MSI
Microsatellite sequence, a short tandem repeat, is a repeat of a

single base or base fragment (1 to 6 bases) usually comprise of
10 to 60 repetitions, they are scattered throughout the genome,
mostly in non-coding regions [8]. These extensions constitute
frequent hotspots of DNA polymerase slip during DNA replication,
resulting in an increase or decrease in repeat nucleotides [5]. Typi-
cally, these errors result in unpaired nucleotides that are recognized
and excised by the DNA Mismatch Repair System (MMR) which is
a group of ribozymes that play a role in DNA replication and then
recombined into the affected portion [9]. If the DNA mismatch
base repair enzyme is lacking, the accumulation of replication error
DNA cannot be repaired in the proliferating cells and result in
cancer susceptibility. Missing mismatch repairs can cause abnorm-
alities in the genome to repeat tens to hundreds of nucleotide base
short sequences, called microsatellite instability (MSI), which cor-
responds to microsatellite stability (MSS). The core of the mis-
match repair protein consists of two heterogeneous proteins
(MSH2/MSH6 or MLH1/PMS2), so inactivation of one or
more MMR genes such as MLH1, MLH2, MSH6, and PMS2
can lead to MSI [10].
3 Detection of MSI
As one of the useful prognostic markers in patients with CRC, the

detection of microsatellite status is strongly recommended. The
common detection methods of microsatellite status are immuno-
histochemistry for MMR proteins and PCR-based assay for micro-
satellite markers. These methods identify MSI from different
aspects and each has its own advantages and limitations.
Immunohistochemistry identifies MSI by assessing the staining
intensity of specific MMR protein including MLH1, MSH2,
MSH6, and PMS2. If all MMR proteins are present, the microsat-
ellite status is defined as microsatellite stable (MSS). When there is a
The Clinical Significance of Microsatellite Instability in Precision Treatment 35
loss of MMR protein, a further detection such as PCR or IHC for

BRAF mutation is required, which assists in distinguishing MSI
from Lynch syndrome [11].
Microsatellite instability detection by PCR was firstly recom-
mended by a National institute workshop in 1997 [12, 13]. This
method evaluates the microsatellite status of tumors by amplifying
microsatellite repeats including three dinucleotide markers
(D5S346, D2S123, and D17S250) [12, 14] and two mononucle-
otide markers (BAT25 and BAT26) [12]. MSS was defined as no
instability at markers as above, where MSI-H represented there are
more than one marker exhibiting instability. And MSI-L was used
to describe the intermediate phenotype.
4 MSI: Prognostic Value
Microsatellite instability (MSI) phenotype—a defined subgroup of

CRC account for approximately 10–15% of patients shows a high
mutation rate in the genomic DNA sequence [15]. The identifica-
tion of MSI has important clinical significance for several reasons, of
which the prognostic value is most obvious. Although how MSI
status affects prognosis of CRC patients is not completely under-
stood, abundant studies have demonstrated that MSI always repre-
sents a better prognosis especially in cases of locally advanced stage
II and stage III CRC [16]. Furthermore, patients with the muta-
tions in MLH1 and MSH2 are at a higher risk of CRC and a more
frequent CRC monitoring including FOBT and colonoscopy is
required [17]. Prophylactic colectomy is considered if necessary.
Meanwhile, the detection of MSI contributes to the diagnosis of
Lynch syndrome, an inheritable disease. Individuals with Lynch
syndrome have significantly higher risks of developing extracolonic
malignancies in addition to early onset of CRC [18]. Intensive
cancer surveillance has been shown to substantially reduce cancer-
related death in this group of patients.
5 MSI: Response to Chemotherapy
The current 5-FU-based chemotherapy like FOLFOX and FOL-

FIRI is the recommended option for high-risk stage 2 and stage
3 CRC. However, a series of studies have shown that patients with
MSS colorectal cancer benefit more from 5-Fu chemotherapy
[19]. If MSI colorectal cancer is compared with untreated MSS
colorectal cancer patients, the former has a better prognosis. How-
ever, as long as the treatment is received, the situation is different
because the majority of patients observed to be affected by chemo-
therapy are stage III patients, but most patients with MSI colorectal
cancer are diagnosed in stage II. There is no complete study to fully
predict the different response rates of chemotherapy to MSI and

MSS colorectal cancer. Sargent et al. shows that in stage II colorec-
tal cancer, MSI patients receiving 5-Fu monotherapy had a worse
prognosis (recurrence rate 30%) than MSS patients, while untreated
MSI patients had a better prognosis (recurrence rate 15%), which
indicated MSI status as a strong predictive biomarker for nonre-
sponse to 5FU-based chemotherapy [4]. A subgroup analysis of
MOSAIC study has confirmed adjuvant therapy containing oxali-
platin is of benefit for patients with stage III MSI tumors
[20]. Therefore, it is urgent to investigate the inner mechanism of
ineffective 5FU-based chemotherapy. Other therapeutic biomar-
kers are currently being developed, ranging from immunohisto-
chemical analysis to high-end genomic approaches. Tajima et al.
revealed the reason may contribute to that the MSH2-MSH6
mismatch repair complex is required for binding 5FU after its
incorporation into DNA and for triggering cell death [21]. In the
MAVERICC trial, the expression of the excision repair cross-
complementation group 1 (ERCC1) [22] gene is being investi-
gated as a potential predictive marker of resistance to platinum
compounds [23]. And in the study of Tabernero, J. et al. detection
of mutations in circulating DNA can predict efficacy to regorafenib
[24]. Until now, the predictive and guiding value of MSI for
chemotherapy efficacy has not been shaken.
6 MSI: Response to Immunotherapy
The existing research has found that almost all tumors with mis-
matched repair gene defects have high activity of Th1/CTL in the
immune microenvironment, and tumor tissues selectively express a
variety of immune checkpoint factors including programmed
death-1 (PD-1), programmed death-1 ligand 1 (PD-L1), cytotoxic
T lymphocyte-associated antigen-4 (CTLA-4), lymphocyte-activa-
tion gene (LAG-3), and indoleamine 2,3-dioxygenase (IDO) in
order to balance the microenvironment [25, 26]. This is a good
explanation for why MSI tumor patients cannot rely on highly
active Th1/CTL in the tumor microenvironment for tumor killing.
At present, there are clinically relevant inhibitors for the above-
mentioned immunological checkpoint molecules, and selective
application of immunotarget inhibitors for MSI patients may
achieve better therapeutic benefit. Compared with traditional che-
motherapy, immunotherapy has the characteristics of low toxicity
and long-lasting effect.
With the success of targeted therapy using antibodies against
immune checkpoint, such as PD-1 in various tumors including lung
cancer and melanoma, the application and efficacy of immunother-
apy in CRC have received more attention [27].
The Clinical Significance of Microsatellite Instability in Precision Treatment 37
Recently, U.S. Food and Drug Administration (FDA) approval

of two PD-1 inhibitors—nivolumab [28](with or without
low-dose ipilimumab which is a CTLA-4 blockade [29]) and pem-
brolizumab for MSIH/dMMR mCRC after progression on che-
motherapy [27]. Several single-arm trials have shown impressive
response rates with nivolumab in patients with MSI-H and d-MMR
colorectal cancers [27, 28, 30–32]. Target molecules for other
checkpoint inhibitors such as lymphocyte activation gene-3
(LAG3) [33] and indoleamine 2,3-dioxygenase (IDO) are under
investigation. Currently, studies have been conducted on broad
spectrum combination of checkpoint inhibitors, even along with
traditional chemotherapeutics, to investigate potential survival
benefit for MMR proficient patients which were not sensitive to
immunotherapy.
7 Conclusion
The incidence of colorectal cancer worldwide is constantly

increased, and it has become much clear that CRC is a complex
and heterogeneous disease. DMMR/MSI CRC as an important
subtype responsible for approximately 15% of them, the molecular
target genes that are differentially regulated between MSI and MSS
cancers require further clarification. The more vertically refined the
molecular subtype classification of CRC, the better individualized
and precise treatment for different molecular types of patients, as
well as the substantial clinical benefit.
References
1. Siegel RL et al (2017) Colorectal cancer statis- unstable gastric cancers. Hum Pathol 45
tics. CA Cancer J Clin 67(3):177–193 (2):285–293
2. Athauda A et al (2019) Integrative molecular 7. Yamamoto H et al (2014) An updated review
analysis of colorectal cancer and gastric cancer: of gastric cancer in the next-generation
what have we learnt?, Cancer Treat Rev. sequencing era: insights from bench to bedside
73:31–40 and vice versa. World J Gastroenterol 20
3. Hutchins G et al (2011) Value of mismatch (14):3927–3937
repair, KRAS, and BRAF mutations in predict- 8. Vilar E, Gruber SB (2010) Microsatellite insta-
ing recurrence and benefits from chemotherapy bility in colorectal cancer-the stable evidence.
in colorectal cancer. J Clin Oncol 29 Nat Rev Clin Oncol 7(3):153–162
(10):1261–1270 9. Kim TM, Laird PW, Park PJ (2013) The land-
4. Sargent DJ et al (2010) Defective mismatch scape of microsatellite instability in colorectal
repair as a predictive marker for lack of efficacy and endometrial cancer genomes. Cell 155
of fluorouracil-based adjuvant therapy in colon (4):858–868
cancer. J Clin Oncol 28(20):3219 10. Li GM (2013) Decoding the histone code: role
5. Yamamoto H, Imai K (2015) Microsatellite of H3K36me3 in mismatch repair and implica-
instability: an update. Arch Toxicol 89 tions for cancer susceptibility and therapy. Can-
(6):899–921 cer Res 73(21):6379–6383
6. Kim KJ et al (2014) Prognostic implications of 11. Jass JR (2006) Hereditary non-polyposis colo-
tumor-infiltrating FoxP3+ regulatory T cells rectal cancer: the rise and fall of a confusing
and CD8+ cytotoxic T cells in microsatellite-
term. World J Gastroenterol 12 bevacizumab as first-line chemotherapy in met-

(31):4943–4950 astatic colorectal cancer. Clin Cancer Res 25
12. Boland CR et al (1998) A National Cancer (10):2988–2995
Institute Workshop on microsatellite instability 23. Li P et al (2013) ERCC1, defective mismatch
for cancer detection and familial predisposi- repair status as predictive biomarkers of survival
tion: development of international criteria for for stage III colon cancer patients receiving
the determination of microsatellite instability oxaliplatin-based adjuvant chemotherapy. Br J
in colorectal cancer. Cancer Res 58 Cancer 108(6):1238–1244
(22):5248–5257 24. Tabernero J et al (2015) Analysis of circulating
13. Rampino N et al (1997) Somatic frameshift DNA and protein biomarkers to predict the
mutations in the BAX gene in colon cancers clinical activity of regorafenib and assess prog-
of the microsatellite mutator phenotype. Sci- nosis in patients with metastatic colorectal can-
ence 275(5302):967 cer: a retrospective, exploratory analysis of the
14. Perucho M, Boland CR et al (1998) A National CORRECT trial. Lancet Oncol 16
Cancer Institute workshop on microsatellite (8):937–948
instability for cancer detection and familial pre- 25. Galon J et al (2006) Type, density, and location
disposition: development of international cri- of immune cells within human colorectal
teria for the determination of microsatellite tumors predict clinical outcome. Science 313
instability in colorectal cancer. Cancer Res (5795):1960–1964
58:5248–5257 26. Le DA-O et al (2017) Mismatch repair defi-
15. Ionov Y et al (1993) Ubiquitous somatic muta- ciency predicts response of solid tumors to
tions in simple repeated sequences reveal a new PD-1 blockade. Science 357(6349):409–413
mechanism for colonic carcinogenesis. Nature 27. Le DT et al (2015) PD-1 blockade in tumors
363(6429):558–561 with mismatch-repair deficiency. N Engl J Med
16. Moertel CG et al (1995) Fluorouracil plus 372(26):2509–2520
levamisole as effective adjuvant therapy after 28. Overman MJ et al (2017) Nivolumab in
resection of stage III colon carcinoma: a final patients with metastatic DNA mismatch
report. Ann Intern Med 122(5):321–326 repair-deficient or microsatellite instability-
17. Koessler T et al (2008) Common variants in high colorectal cancer (CheckMate 142): an
mismatch repair genes and risk of colorectal open-label, multicentre, phase 2 study. Lancet
cancer. Gut 57(8):1097–1101 Oncol 18(9):1182–1191
18. Lynch HT et al (2006) Phenotypic and geno- 29. Omuro A et al (2018) xNivolumab with or
typic heterogeneity in the Lynch syndrome: without ipilimumab in patients with recurrent
diagnostic, surveillance and management glioblastoma: results from exploratory phase I
implications. Eur J Hum Genet 14 cohorts of CheckMate 143. Neuro-Oncology
(4):390–402 20(5):674–686
19. Smyth EC et al (2017) Mismatch repair defi- 30. Le DT et al (2016) Programmed death-1
ciency, microsatellite instability, and survival: blockade in mismatch repair deficient colorec-
an exploratory analysis of the medical research tal cancer. N Engl J Med 34
council adjuvant gastric infusional chemother- (15_suppl):103–103
apy (MAGIC) trial. JAMA Oncol 3 31. Overman MJ et al (2018) Durable clinical
(9):1197–1203 benefit with nivolumab plus ipilimumab in
20. Andre T et al (2015) Adjuvant fluorouracil, DNA mismatch repair-deficient/microsatellite
leucovorin, and oxaliplatin in stage II to III instability-high metastatic colorectal cancer.
colon cancer: updated 10-year survival and Clin Oncol 36(8):773–779
outcomes according to BRAF mutation and 32. Andre T et al (2017) Combination of nivolu-
mismatch repair status of the MOSAIC study. mab (nivo) + ipilimumab (ipi) in the treatment
J Clin Oncol 33(35):4176–4187 of patients (pts) with deficient DNA mismatch
21. Tajima Y et al (2018) Prevalence and molecular repair (dMMR)/high microsatellite instability
characteristics of defective mismatch repair epi- (MSI-H) metastatic colorectal cancer (mCRC):
thelial ovarian cancer in a Japanese hospital- CheckMate 142 study. J Clin Oncol
based population. Jpn J Clin Oncol 48 35:3531–3531
(8):728–735 33. Anderson AC, Joller N, Kuchroo VK (2016)
22. Parikh AR et al (2019) MAVERICC, a rando- Lag-3, Tim-3, and TIGIT: Co-inhibitory
mized, biomarker-stratified, phase II study of receptors with specialized functions in immune
mFOLFOX6-bevacizumab versus FOLFIRI- regulation. Immunity 44(5):989–1004
Chapter 4
Applications of Network Analysis in Biomedicine

Steven Wang and Tao Huang
Abstract
The abundance of high-throughput data and technical refinements in graph theories have allowed network
analysis to become an effective approach for various medical fields. This chapter introduces co-expression,
Bayesian, and regression-based network construction methods, which are the basis of network analysis.
Various methods in network topology analysis are explained, along with their unique features and applica-
tions in biomedicine. Furthermore, we explain the role of network embedding in reducing the dimension-
ality of networks and outline several popular algorithms used by researchers today. Current literature has
implemented different combinations of topology analysis and network embedding techniques, and we
outline several studies in the fields of genetic-based disease prediction, drug–target identification, and
multi-level omics integration.
Key words Network analysis, Random walk, Heat diffusion, Network embedding, Multi-omics
1 Introduction
With the growing availability of large amounts of biological infor-

mation acquired using detection methods such as high-throughput
sequencing, methods such as machine learning and network analy-
sis is becoming the state-of-the-art technique in manipulating
large-scale datasets [1]. Implementations of network approaches
have shown its capabilities of uncovering interacting components
within a constructed network, such as protein–protein interactions
or disease–gene relationships [2–7]. Depending on the specific
case, there exists various methods of constructing these networks
based upon different mathematical models, such as the co-expres-
sion network, the Bayesian network, and regression-based net-
works. These networks are not only a mean of data storage, but
also the starting point of implementing various inference methods
that perform tasks like classification or clustering using machine
learning or even deep learning. Based on these networks, topology
analysis can be used to uncover important features. We introduce
the mathematical concept and several examples of common
39
40 Steven Wang and Tao Huang
techniques, including module identification, network characteris-

tics analysis, shortest path analysis, guilt by association, random
walks, and heat diffusion. Furthermore, we give an overview of
popular network embedding methods, which reduces the dimen-
sion of networks in order to increase computation efficiency and
ease of visualization. Using different combinations of the techni-
ques, recent studies have implemented network analysis in wide-
ranging fields from cancer genetics to drug development. Building
upon the current success of network analysis in biomedical fields,
network approaches have much room to grow, especially by inte-
grating multi-level omics data into integrative networks.
2 Methods of Constructing Networks
2.1 Co-expression Co-expression networks are most commonly used in genetics,

Network Construction where relationships between different genes can be represented by
a co-expression network (GCN). By modelling the rise and fall of
genetic expression across samples, interacting genes can be identi-
fied, which are often representative of specific pathways or regulator
mechanisms. Specifically, the nodes of a GCN represent individual
genes and the edges between nodes represent their co-expression
relationship [8]. The edges can be binary (weight 0 or 1) to denote
the presence of connection or can be a continuous value between
0 and 1 using soft thresholding methods such as the sigmoid
function [9]. Namely, co-expression networks calculate the weight
l þa
with ωij ¼ min k ij, k ijþ1a , where lij ¼ Σaiuauj and ki ¼ Σaiu. The
ð i jÞ ij

adjacency function calculates aij ¼ sigmoid s ij , α, τ0 1
αðs ij τ0Þ
1þe
using the similarity matrix sij, which is calculated using the absolute
value of the Pearson correlation coefficient sij ¼ j cor(i, j) j.
A GCN can be constructed using expression profiles of a list of
genes, where co-expressed genes show similar transcript levels.
These networks are of interest because co-expressed genes are
often functionally related or members of the same pathway, open-
ing new means of associating unknown genes with known
biological functions and processes. For example, a study by Ma
et al. implemented a GCN to investigate Bamboo growth and
identified 1896 functional modules associated with photosynthesis,
hormone biosynthesis, cell wall biosynthesis, and more [10]. A
commonly used tool to construct GCN is WGCNA [11].
2.2 Bayesian In general, Bayesian networks (BNs) are probabilistic graphical

Network Construction models that compute conditional dependence of random variables
using Bayesian inference. By translating these dependencies into
interconnected nodes, BNs provide a comprehensible and modular
framework for representing complex systems [12]. BNs consist of
Applications of Network Analysis in Biomedicine 41
two components: a network structure in the form of a directed

acyclic graph and a conditional probability for each node as defined
below [12]. Xi is the individual node and Pa(Xi) is the set of all
parents of Xi.
A BN is then trained to optimize the parameters, namely the set
of conditional probability distributions P(Xi | Pa(Xi)), to produce a
network that best matches the training set. Search procedures are
often used for efficiency in training, with a notable method being
the Sparse Candidate algorithm introduced by Friedman, Nach-
man, and Pe’er [13]. Commonly used tools include BNT (open-
sourced) [14] and BayesiaLab (closed source https://www.
bayesialab.com/).
2.3 Regression- Regression networks model the relationship between nodes using
Based Network linear models. Namely, a set of parameters a ¼ [a1, a2, ... an]
Construction models a linear weight between a node and its parents u ¼ [u1,
u2, ... un]. Applications of regression-based networks include Gen-
eReg, developed by Huang et al., which constructed a gene regu-
latory network modelling expression changes between regulators
and target genes [15]. The network optimizes a time delay linear
regression model by iteratively adding possible regulators of a
target gene under an AIC (Akaike information criterion) model
selection criteria [16].
3 Network Topology Analysis
Network topology is general describes the structural arrangement

of links and nodes within a network. In biomedicine, networks
often model interacting components such as gene–gene, protein–
protein, or drug–target interactions. Hence, the topology of a
network contains a large amount of information of interest, and
its analysis is key to the utilization of network approaches in various
biomedical fields. Here, we introduce several important concepts
that are associated algorithms adopted by researchers to perform
predictive tasks.
3.1 Module A prominent feature of real-world networks is their modular struc-

Identification tures, where the networks are often organized into distinct modules
of highly interconnected nodes [17]. Modules are also known as
groups, clusters, or communities in certain cases. In biology, mod-
ules typically correspond to sets of interacting components, such as
protein complexes, metabolic pathways, or a set of interacting
genes [18]. Thus, identifying modules can be of significant use in
analyzing functional associations in omics studies.
A common computational method to identify modules is by
locating “hotspots,” which are subnetworks with high aggregate
scores. Scores can be calculated and normalized using tools such as
jActiveModules [19]. Based on this information, there are many

computational algorithms that search for modular structures. In
genomics, for example, algorithms like PANOGA [20], dmGWAS
[21], and PinnacleZ [22] have been implemented in studies to map
gene or PPI networks [23–25]. In the fields of proteomics, on the
other hand, tools like CFinder [26], mfinder [27], and FANMOD
[28] are available for identifying novel protein modules and net-
work motifs.
3.2 Network Networks have several attributes and characteristics that are central
Characteristics to its formation and our understanding of it. There are three
Analysis defining characteristics to network topologies, being average path
length (APL), clustering coefficient, and degree distribution. Aver-
age path length (APL), defined as the least number of steps
between all possible pairs of nodes averaged over a number of
nodes, is a measure of how well connected a network
is. Clustering coefficient includes two variations, the global cluster-
ing coefficient and the local clustering coefficient. The global coef-
ficient provides a measure of the level of clustering in the whole
network. Its definition involves the concept of a triplet, which is
simply three nodes connected with two (open triplet) or three
edges (closed triplet). With that in mind, the global clustering
coefficient is defined as the number of closed triplets over the
total number of open and closed triplets. The local clustering
coefficient, on the other hand, measures the degree of connection
within a cluster. It is defined mathematically as the existing number
of edges over the possible number of edges. The concept of degree
distribution involves the term degree, which is simply the number
of connections that a node has. Degree distribution, therefore,
captures the degrees of all the nodes in a network and can often
be represented in histogram graphs.
With that in mind, an important network topology introduced
by Watts and Strogatz is the small world network, which have high
clustering coefficients, low APL, and small degrees of separation
[29]. In other words, most nodes in small world networks can be
reached within a small number of steps despite that most nodes are
not neighbors. Mathematically, the distance L between two random
nodes in a small world network is proportional to the log of the
number of nodes N [29].
L log N
An effective metric used is the small-world index (SWI). Lobs
and Cobs are observed APL and clustering coefficients, Llatt and Clatt
are obtained from a comparison lattice network, and Lrand and Crand
are obtained from a comparison random network [30].
L obs L latt C obs C rand
SWI ¼
L rand L latt C latt C rand
Another type of network topology that has sparked recent

interest is the scale-free network, specifically referring to networks
whose degree follows a power-law distribution [31].
P ðkÞ kλ
3.3 Shortest Path Shortest path analysis aims to find a path between two nodes that
Analysis minimizes sum of the weights in weighted networks or steps in
unweighted networks. A commonly used algorithm in shortest
path analysis is the Dijkstra’s algorithm, which has several variants
corresponding to different types of networks and graphs
[32, 33]. There are also other algorithms available that are opti-
mized for different types of networks [34, 35].
3.4 Guilt By Guilt by association, in principle, states functional similarities

Association between genes often suggests that they share expression profiles
or are protein interaction partners [36]. This is an immensely useful
concept in computational biology, where functions are assigned to
genes based upon prior expression profiles of known genes and thus
making tasks like predicting novel disease genes to be possible [37–
39]. Nevertheless, this principle has also been challenged in current
literature, describing the many scenarios where “guilt by associa-
tion” is ineffective or inapplicable [4, 40].
3.5 Random Walk The theory of random walk can be traced by the formulation of
Brownian motion and since then has been widely used to model
diffusion and random motion. The simplest and standard form of
random walk is an uncorrelated and unbiased model, where move-
ment in completely independent of previous motion. This model,
known as the simple isotropic random walk model, is the basis of
other random walk models [41]. On the other hand, correlated
random walks (CRWs) have directional biases, where each step
likely moves in a similar direction as the previous. In a directed
and weighted graph, a random walk is also known as a Markov
chain. Namely, Markov chains model the stochastic transition
between n states using a transition probability matrix.
An important use of random walk in networks is to locate
“central” nodes that are highly connected. These nodes often reveal
important protein interaction, pathways, or important regulator
and disease genes in the context of different biological networks.
Therefore, random walk can be found widely implemented in cur-
rent literature [42–46].
3.6 Heat Diffusion One of the important goals of network analysis is to identify com-
munities—sets of internally cohesive nodes that are separated from
the remainder of the network. Mathematically, the concept of a
community can be represented by the conductance measure, which
is the ratio of the number of edges leaving a set of nodes to the
number of internal edges [47, 48]. A small conductance set indi-

cates high number of internal edges and few edges leaving the set,
hence a community. Two notable methods of finding small con-
ductance sets are the personalized PageRank diffusion and the heat
kernel. These algorithms take an estimate of a diffusion and return
the set of smallest conductance. Implementations of heat diffusion
can be found in genomics, proteomics, and drug design [49–52].
4 Network Embedding
Since modern information networks can contain billions of nodes

and edges, it can become computationally inefficient to perform
inferences such as classification or clustering with the original scale.
Network embedding is a method that aims to reduce the
dimensionality of these networks and thus allowing the data to be
effectively visualized and used as features for further inference
[1, 53]. Important characteristics of good network embedding
methods include adaptability of implementing new data, scalability
to large-scale networks, low dimensionality, community awareness
of similarity between nodes, and continuous representation
[53]. We will introduce some popular network embedding meth-
ods and their respective features.
4.1 word2vec In general, word2vec refers to a group of word embedding models

that are used to construct linguistic information. Originally devel-
oped by researchers at Google, word2vec is a two-layer neural
network that take large corpora of text and outputs a vector space
where related words are placed in proximity with each other
[54]. This capability of analyzing textual data has helped computa-
tional biologist address the time-consuming process of curating
current and past literature. A group of researchers have developed
a machine learning approach to extract gene–disease relations from
available literature via word2vec [3]. Furthermore, the word2vec
approach of neural embedding have inspired computational biolo-
gists to develop tools like Gene2vec in order to model gene–gene
interaction networks [55].
4.2 node2vec Machine learning algorithms, in general, uses multiple independent

and discriminating features to perform tasks. Such features are
typically manually designed using expert knowledge. However,
this process is time-consuming and does not generalize across
different tasks. Node2vec is an algorithmic framework developed
by Grover and Leskovec that can learn continuous feature repre-
sentations for nodes in networks [56]. Concretely, node2vec pro-
duces features with maximal likelihood of preserving
neighborhoods of nodes in a network. These features can then be
used in machine learning inference methods to perform link
prediction, which has a range of applications in discovering gene–

gene, protein–protein, or drug–target interactions. Node2vec has
been shown in multiple studies to be the best performing embed-
ding method to model gene–gene interaction network [57] and
disease–gene networks [5].
4.3 DeepWalk DeepWalk, developed by Perozzi et al., is a network embedding

method that takes a network as input and returns a latent represen-
tation as output, which can then be used as features to machine
learning algorithms [58]. The central idea behind DeepWalk is
analogous to that of word2vec, where sentences are used as training
data to analyze words. In DeepWalk, short random walks in a
network are generated as training data, and features representations
of individual nodes are returned. In comparison to other network
embedding methods, DeepWalk extracts features well even with
small labelled training datasets and is very scalable.
Studies have applied DeepWalk and network-based methods to
uncover interaction between biological components. A study by
Zhang et al. implemented a DeepWalk-based network to predict
long noncoding RNA-associated disease [2]. Other works have
developed DeepWalk-based networks to investigate drug–target
associations [59, 60] and microRNA functions [61].
4.4 LINE The LINE network embedding method, introduced by Tang et al.,
is a scalable algorithm that can be applied to a range of networks
while maintaining local and global network structures [62]. Specifi-
cally, LINE addresses three major problems faced by network
embedding of large-scale networks: preserving first and second-
order proximity, scalability, and compatibility of different types of
edges (directed, undirected, and/or weighted). First-order prox-
imity refers to the local pairwise similarity between nodes, while the
second-order proximity indicates the structural similarity between
neighborhoods [63]. Hence, by combining the two, LINE can
preserve both local and global network structure. A study by Zitnik
and Lesovec incorporated LINE to construct networks used to
predict protein function [64].
4.5 SDNE The Structural Deep Network Embedding (SDNE) method, pro-
posed by Wang et al., is a network embedding technique that
addresses the nonlinear and sparse nature of real-world large net-
works [63]. Like LINE, SDNE proposes to preserve network
structure using two orders of proximity. Furthermore, SDNE dif-
fers from LINE in that it jointly optimizes the first- and second-
order proximities. SDNE also adopts a deep structure to capture
the nonlinearity of network structures. The authors show that
SDNE outperforms LINE using three testing sets [63].
4.6 Struc2Vec Struc2vec is a recent representation learning framework developed

by Riberio et al. that takes a slightly different approach [65]. A
common problem faced by available network embedding methods
is that while nodes with similar functions should have similar latent
representations, algorithms often fail to do so when the neighbor-
hoods of these nodes are far apart. This is because the neighbor-
hood concept is inherently defined by proximity within the
network. Struc2vec circumvents this limitation by defining struc-
tural similarity between nodes independent of their position in the
network. Other features of struc2vec include a hierarchical mea-
surement of structural similarity and the generation of random
contexts for nodes.
5 Application of Network Analysis
5.1 Genetic-Based Network analysis has allowed the prediction of disease causative
Prediction of Disease genes to be possible, which can significantly improve the under-
standing of the genetics basis of various diseases and guide their
treatments.
A study by Ata et al. implemented an integrative framework,
N2VKO, to predict disease genes [6]. The node embeddings were
learned from protein–protein interaction (PPI) network using the
node2vec learning method, and classification models were built
using node embeddings and various biological annotations. This
approach has proved to be effective by cross-validating their pre-
diction with literature on the predicted disease genes.
5.2 Drug–Target As the productivity of modern pharmaceutical research and drug

Identification development is slowing down, network approaches is showing
promise in overcoming these obstacles [66]. A common approach
of integrating network approaches into drug development is using
link prediction to simulate drug–target interaction. For example, a
study by Lee and Nam constructed three subnetworks for predict-
ing drug–target interaction. Specifically, random walks with restart
to calculate affinity scores between nodes and the model was shown
to outperform previous guilt by association models [67]. Other
literatures also highlight the promise of using network approaches
in drug–target identification [50, 59, 60, 68–70]. Alternatively,
repurposing and repositioning drugs offer a time- and cost-efficient
way to develop treatment. To this end, Zeng et al. developed
deepDR, which uses a multi-modal deep autoencoder to learn
low-dimensional representations [71]. The authors validate the
repurposing capabilities of deepDR using clinical data of Alzhei-
mer’s disease and Parkinson’s disease.
5.3 Multi-omics Data Applications of network analysis in genomics involves identifying

Integration gene–gene interactions, pathways, and disease-causing genes
[6, 10, 12, 23, 24, 39, 46, 57, 66, 72]. In proteomics, the most
common uses of network analysis are protein–protein interaction
(PPI) analysis and inference of protein functions [73–76]. The
increasing abundance of omics data poses the prospect of integrat-
ing multi-level omics to achieve higher predictive capabilities of
network models. Examples of these algorithms include SAMNet-
Web [77] and pwOmics [78], which integrate transcriptomic and
proteomic data. Other tools, such as MetScape [79], Cytoscape
[80], and Grinn [81], integrate metabolomic and genomic
information.
The integration of multi-level omics data has produced effec-
tive algorithms that have wide-ranging fields of implementation.
For example, in the study of ovarian cancer, Gevaert et al. devel-
oped an algorithm AMARETTO to identify driver genes [82],
while Zhang et al. integrated coding genomic and epigenomic
data to uncover potential genetic pathways [83].
6 Conclusions
The inherent highly interconnected nature of biological systems

and the growing available of various forms of sequencing data has
allowed network analysis to become the state-of-the-art technique
in modelling the interaction between biological components. The
work of many researchers has produced improved algorithms for
various tasks including topology analysis, network embedding, and
link prediction, all of which are essential to the growing capabilities
of network approaches in biomedicine. Using these algorithms,
researchers were able to further human knowledge in cancer,
omics studies, drug development, and more. In this brief review,
we covered the basic underlying concepts and important analysis
techniques of network analysis, as well as examples of studies
implementing these tools.
References
1. Nelson W et al (2019) To embed or not: net- 4. Guo Y et al (2013) Dissecting disease inheri-
work embedding as a paradigm in computa- tance modes in a three-dimensional protein
tional biology. Front Genet 10 network challenges the “Guilt-by-Association”
2. Zhang H et al (2019) Predicting lncRNA- principle. Am J Hum Genet 93(1):78–89
disease associations using network topological 5. Peng J, Guan J, Shang X (2019) Predicting
similarity based on deep mining heterogeneous Parkinson’s disease genes based on Node2vec
networks. Math Biosci 315:108229 and autoencoder. Front Genet 10
3. Bhasuran B, Natarajan J (2018) Automatic 6. Ata SK et al (2018) Integrating node embed-
extraction of gene-disease associations from lit- dings and biological annotations for genes to
erature using joint ensemble learning. PLoS predict disease-gene associations. BMC Syst
One 13(7):–e0200699 Biol 12(S9)
7. Robertson KD (2005) DNA methylation and 23. Hormozdiari F et al (2015) The discovery of
human disease. Nat Rev Genet 6(8):597–610 integrated gene networks for autism and
8. Stuart JM (2003) A gene-coexpression net- related disorders. Genome Res 25(1):142–154
work for global discovery of conserved genetic 24. Wang Q et al (2015) EW_dmGWAS: edge-
modules. Science 302(5643):249–255 weighted dense module search for genome-
9. Nagrecha S, Lingras PJ, Chawla NV (2013) wide association studies and gene expression
Comparison of gene co-expression networks profiles. Bioinformatics 31(15):2591–2594
and bayesian networks. In: Intelligent informa- 25. Smith CL et al (2014) Identification of a
tion and database systems. Springer, Berlin, human neonatal immune-metabolic network
Heidelberg associated with bacterial infection. Nat Com-
10. Ma X et al (2018) Co-expression gene network mun 5:4649
analysis and functional module identification in 26. Adamcsek B et al (2006) CFinder: locating
bamboo growth and development. Front cliques and overlapping modules in biological
Genet 9:574 networks. Bioinformatics 22(8):1021–1023
11. Langfelder P, Horvath S (2008) WGCNA: an 27. Kashtan N et al (2004) Efficient sampling algo-
R package for weighted correlation network rithm for estimating subgraph concentrations
analysis. BMC Bioinformatics 9(1):559 and detecting network motifs. Bioinformatics
12. Su C et al (2013) Using Bayesian networks to 20(11):1746–1758
discover relations between genes, environ- 28. Wernicke S, Rasche FJB (2006) FANMOD: a
ment, and disease. BioData Mining 6(1):6 tool for fast network motif detection. Bioinfor-
13. Friedman N, Nachman I, Peér D (1999) matics 22(9):1152–1153
Learning bayesian network structure from mas- 29. Watts DJ, Strogatz SH (1998) Collective
sive datasets: the “sparse candidate” algorithm. dynamics of ‘small-world’ networks. Nature
In: Proceedings of the Fifteenth conference on 393(6684):440–442
Uncertainty in artificial intelligence. Morgan 30. Neal Z (2015) Making big communities small:
Kaufmann Publishers Inc. using network science to understand the eco-
14. Murphy K (2001) The bayes net toolbox for logical and behavioral requirements for com-
matlab. Comput Sci Stat 33(2):1024–1034 munity social capital. Am J Community
15. Huang T et al (2010) Using GeneReg to con- Psychol 55(3–4):369–380
struct time delay gene regulatory networks. 31. Barabási A-L, Albert R (1999) Emergence of
BMC Res Notes 3(1):142 scaling in random networks. Science 286
16. Sakamoto Y, Ishiguro M, Kitagawa GJD (5439):509–512
(1986) Akaike information criterion 32. Dijkstra EW (1959) A note on two problems in
statistics. D. Reidel, The Netherlands, p 81 connexion with graphs. Numer Math 1
17. Guimerà R, Sales-Pardo M, Amaral LAN (1):269–271
(2007) Module idesdntification in bipartite 33. Fredman ML, Tarjan RE (1987) Fibonacci
and directed networks. Phys Rev E 76 heaps and their uses in improved network opti-
(3):036102 mization algorithms. J ACM (JACM) 34
18. Nikolayeva I, Guitart Pla O, Schwikowski B (3):596–615
(2018) Network module identification—A 34. Thorup M (2004) Integer priority queues with
widespread theoretical bias and best practices. decrease key in constant time and the single
Methods 132:19–25 source shortest paths problem. J Comput Syst
19. Ideker T et al (2002) Discovering regulatory Sci 69(3):330–353
and signalling circuits in molecular interaction 35. Gabow HN (1983) Scaling algorithms for net-
networks. Bioinformatics 18(suppl_1): work problems. In: 24th Annual Symposium
S233–S240 on Foundations of Computer Science (sfcs
20. Bakir-Gungor B, Sezerman OU (2011) A new 1983). IEEE
methodology to associate SNPs with human 36. Oliver SJN (2000) Proteomics: guilt-by-asso-
diseases according to their pathway related ciation goes global. Nature 403(6770):601
context. PloS One 6(10):e26277 37. Arabidopsis Interactome Mapping Consortium
21. Jia P et al (2010) dmGWAS: dense module (2011) Evidence for network evolution in an
searching for genome-wide association studies Arabidopsis interactome map. Science 333
in protein–protein interaction networks. Bioin- (6042):601–607
formatics 27(1):95–102 38. Mukhtar MS et al (2011) Independently
22. Chuang HY et al (2007) Network-based classi- evolved virulence effectors converge onto
fication of breast cancer metastasis. Mol Syst hubs in a plant immune system network. Sci-
Biol:3(1) ence 333(6042):596–601
ALGrawany
39. Wu X et al (2008) Network-based global infer- 55. Du J et al (2019) Gene2vec: distributed repre-
ence of human disease genes. Mol Syst Biol 4 sentation of genes based on co-expression.
(1) BMC Genomics 20(1):82
40. Gillis J, Pavlidis P (2012) “Guilt by Associa- 56. Grover A, Leskovec J (2016) node2vec: scal-
tion” is the exception rather than the rule in able feature learning for networks. arXiv
gene networks. PLOS Computat Biol 8(3): e-prints
e1002444 57. Kim M, Baek SH, Song M (2018) Relation
41. Blanchard P, Volchenkov D (2011) Fair and extraction for biological pathway construction
biased random walks on undirected graphs using node2vec. BMC Bioinformatics 19
and related entropies. Birkh€ause, Boston, Basel (8):206
42. Seah CS et al (2017) An enhanced topologi- 58. Perozzi B, Al-Rfou R, Skiena S (2014) Deep-
cally significant directed random walk in cancer walk: online learning of social representations.
classification using gene expression datasets. In: Proceedings of the 20th ACM SIGKDD
Saudi J Biol Sci 24(8):1828–1841 International Conference on Knowledge dis-
43. Roessler S et al (2015) Integrative genomic and covery and data mining. ACM
transcriptomic characterization of matched pri- 59. Zong N, Wong RSN, Ngo V (2019) Tripartite
mary and metastatic liver and colorectal carci- network-based repurposing method using
noma. Int J Biol Sci 11(1):88 deep learning to compute similarities for
44. Seah CS, Kasim S, Mohamad MS (2017) Spe- drug-target prediction. In: Vanhaelen Q
cific tuning parameter for directed random (ed) Computational methods for drug repur-
walk algorithm cancer classification. Interac- posing. Springer New York, New York, NY, pp
tion 114:1500 317–328
45. Suki B, Frey U (2017) A time-varying biased 60. Zong N et al (2017) Deep mining heteroge-
random walk approach to human growth. Sci neous networks of biomedical linked data to
Rep 7(1):7805 predict novel drug–target associations. Bioin-
46. Yang K et al (2019) HerGePred: heteroge- formatics 33(15):2337–2344
neous network embedding representation for 61. Deng L, Wang J, Zhang J (2019) Predicting
disease gene prediction. IEEE J Biomed Health gene ontology function of human MicroRNAs
Inform 23(4):1805–1815 by integrating multiple networks. Front Genet
47. Kloster K, Gleich DF (2014) Heat kernel based 10:3
community detection. arXiv e-prints 62. Tang J et al (2015) Line: Large-scale informa-
48. Chung F (2007) The heat kernel as the pager- tion network embedding. In: Proceedings of
ank of a graph. Proc Natl Acad Sci 104 the 24th International Conference on World
(50):19735–19740 Wide Web. International World Wide Web
Conferences Steering Committee
49. Paull EO et al (2013) Discovering causal path-
ways linking genomic events to transcriptional 63. Wang D, Cui P, Zhu W (2016) Structural deep
states using Tied Diffusion Through Interact- network embedding. In: Proceedings of the
ing Events (TieDIE). Bioinformatics 29 22nd ACM SIGKDD International Confer-
(21):2757–2764 ence on Knowledge Discovery and Data
Mining. ACM: San Francisco, CA,
50. Laenen G et al (2013) Finding the targets of a USA. p. 1225–1234
drug by integration of gene expression data
with a protein interaction network. Mol Bio- 64. Zitnik M, Leskovec J (2017) Predicting multi-
Syst 9(7):1676–1685 cellular function through multi-layer tissue
networks. Bioinformatics 33(14):i190–i198
51. Nitsch D et al (2010) Candidate gene prioriti-
zation by network analysis of differential 65. Ribeiro LF, Saverese PH, Figueiredo DR
expression using machine learning approaches. (2017) struc2vec: learning node representa-
BMC Bioinformatics 11(1):460 tions from structural identity. In: Proceedings
of the 23rd ACM SIGKDD International Con-
52. Vandin F (2011) et al Discovery of mutated ference on Knowledge Discovery and Data
subnetworks associated with clinical data in Mining. ACM
cancer Pac Symp Biocomputing 2012. 2011,
55–66 66. Moffat JG et al (2017) Opportunities and chal-
lenges in phenotypic drug discovery: an indus-
53. Chen H et al (2018) A tutorial on network try perspective. Nat Rev Drug Discov 16
embeddings (8):531
54. Mikolov T, et al (2013) Efficient estimation of 67. Lee I, Nam H (2018) Identification of drug-
word representations in vector space target interaction by a random walk with restart
method on an interactome network. BMC Bio- reveals hierarchical organization in large

informatics 19(8):208 integrated systems. PLoS One 6(4):e18209
68. Yildirim M et al (2007) Drug-target network. 76. Blondel VD et al (2008) Fast unfolding of
Nat Biotechnol 25(10):1119–1126 communities in large networks. J Stat Mech
69. Yamanishi Y et al (2010) Drug-target interac- 10:P10008
tion prediction from chemical, genomic and 77. Gosline SJ, Oh C, Fraenkel EJB (2014) SAM-
pharmacological data in an integrated frame- NetWeb: identifying condition-specific net-
work. Bioinformatics 26(12):i246–i254 works linking signaling and transcription.
70. Xu Z et al (2017) Seq2seq fingerprint: an unsu- Bioinformatics 31(7):1124–1126
pervised deep molecular embedding for drug 78. Wachter A, Beißbarth TJB (2015) pwOmics:
discovery. In: Proceedings of the 8th ACM an R package for pathway-based integration of
International Conference on Bioinformatics, time-series omics data using public database
Computational Biology, and Health Informat- knowledge. Bioinformatics 31(18):3072–3074
ics. ACM 79. Karnovsky A et al (2011) Metscape 2 bioinfor-
71. Zeng X et al (2019) deepDR: a network-based matics tool for the analysis and visualization of
deep learning approach to in silico drug repo- metabolomics and gene expression data. Bioin-
sitioning. Bioinformatics 35(24):5191–5198 formatics 28(3):373–380
72. Tillquist RC, Lladser ME (2019) 80. Shannon P et al (2003) Cytoscape: a software
Low-dimensional representation of genomic environment for integrated models of biomo-
sequences. J Math Biol 79(1):1–29 lecular interaction networks. Genome Res 13
73. Pons P, Latapy M (2005) Computing commu- (11):2498–2504
nities in large networks using random walks. 81. Wanichthanarak K, Fahrmann JF, Grapov DJBI
In: International symposium on computer and (2015 Sep 7) Genomic, proteomic, and meta-
information sciences. Springer, Berlin, bolomic data integration strategies. Biomark
Heidelberg Insights 10(Suppl 4):1–6
74. Raghavan UN (2007) Albert R, Kumara S, 82. Gevaert O et al (2013) Identification of ovarian
Near linear time algorithm to detect commu- cancer driver genes by using module network
nity structures in large-scale networks. Phys integration of multi-omics data. Interface
Rev E 76(3):036106 Focus 3(4):20130013
75. Rosvall M, Bergstrom CT (2011) Multilevel 83. Zhang Q, Burdette JE, Wang J-P (2014) Inte-
compression of random walks on networks grative network analysis of TCGA data for
ovarian cancer. BMC Syst Biol 8(1):1338
Part II
Precision Medicine of Cancers

Chapter 5
Diagnosis and Treatment of Breast Cancer in the Precision

Medicine Era
Jing Yan, Zhuan Liu, Shengfang Du, Jing Li, Li Ma, and Linjing Li
Abstract
Breast cancer is one of the most leading causes of death for women worldwide. According to statistics
published by the International Agency for Research on Cancer (IARC), the incidence of breast cancer is on
the rise year by year in most parts of the world. The existence of heterogeneity limits the early diagnosis and
targeted therapy of breast cancer. Nowadays, precision medicine brings a new perspective to personalized
diagnosis and targeted therapy, overcomes the heterogeneity of different patients, and provides an oppor-
tunity for screening of high-risk populations. As a clinician, we are committed to using genomic to provide a
favorable perspective in the field of breast cancer. The current review describes the recent advances in the
understanding of precision medicine for breast cancer in the aspect of the genomics which could be applied
to improve our ability to diagnose and treat breast cancer individually and effectively.
Key words Breast cancer, Precision medicine, Genomics, Prevention, Treatment
1 Introduction
Breast cancer is not only the most common malignancy in women,

but also one of the major causes of death among women all over the
world. Despite global mortality rates are falling, breast cancer is the
leading cause of death in some underdeveloped areas. Although
early screening programs and effective treatments have reduced
mortality rate, breast cancer still caused 521,900 deaths in 2012
accounting for 15% of all cancer deaths among women worldwide
[1]. Numerous studies suggested that hereditary factor causes less
than 10% of breast cancers, and the most common factors are
environmental, reproductive, and lifestyle factors. However, up to
25% of inherited breast cancers are linked to genetic mutations in
specific genes [2]. Recent years, considerable attention has been
paid to breast cancer care, while new advancements in applying
genomics and precision medicine to clinical treatment are still
facing challenges, and it is worthwhile devoting much effort
[3]. Genetic heterogeneity is the most common feature of the
53
54 Jing Yan et al.
breast cancer, which leads to significant differences in phenotype

and response to treatment of breast cancer patient and affects
prognosis of each individual, thus pose a critical challenges in
prediction, diagnosis, and treating breast cancer individually [4, 5].
As we know, the success of human genome sequencing has
brought opportunities for knowing the correlated genes with dis-
eases. On the basis of that, precision medicine for breast cancer,
which consider the genes patients are born with and the genes or
other biomarkers present in the cancer cells, provides strategy for
the prediction, diagnosis, treatment, and prevention. With this
approach, blood and tumor tissues of patients are collected for
analysis genetically. More importantly, precision medicine is based
on examining genes of each patient that will guide the clinicians in
every aspect [6]. Physicians can treat patients in a personalized way
by identifying the genes associated with the development and prog-
nosis of breast cancer [7]. Due to the diversity of drugs, give the
right drugs to the right patients at the right time to achieve the best
clinical results remains an ongoing challenge. Therefore, with
increasing understanding of the driver genes, it is possible to
improve the precision diagnosis and treatment of breast cancer.
2 New Perspective from Genetic Studies
The Human Genome Project (HGP) has provided a complete and

accurate sequence of the DNA base pairs of the human genome.
The first draft of the HGP was completed in 2001 and officially
released in 2003. HGP sequenced 99% of the whole human
genome and all data were published to support medical research
[8]. As we know, the symptoms of breast cancer vary from person
to person. One of the most important reasons is single nucleotide
polymorphisms (SNPs) among these patients are different. The
completion of HGP and the continuous advancement in genetic
research will facilitate the individual medicine which contributes to
the precision medicine in breast cancer.
As development of HGP, the BRCA1 susceptibility gene for
breast cancer was found on chromosome 17 in the 1990s, and the
sequence of BRCA1 was acquired few years later. Simultaneously,
another team of scientists discovered the BRCA2 gene using linkage
analysis. They found that a male breast cancer patient of the mam-
mary and ovarian cancer family did not have a mutation at 17q21
(BRCA1), mapped another breast cancer-related gene and named it
as BRCA2, which located in 13q12 [9, 10]. In the year 2000, the
HGP published the sequence of the BRCA2 gene. Currently, it was
well known that mutation of BRCA1/2 gene are associated with
inherited breast cancer, which are normally tumor suppressor genes.
Over 1500 mutations have been reported in the BRCA1 gene, and
the majority of them resulting in missense or non-functional
Diagnosis and Treatment of Breast Cancer in the Precision Medicine Era 55
proteins. Similarly, it was confirmed that more than 1800 muta-

tions in BRCA2 have been identified including frameshift dele-
tions, nonsense mutations, and insertions which lead to
premature truncation of proteins. These events could result in the
loss of function of tumor suppressor genes [11]. It was estimated
that most of the BRCA1 mutation carriers, and almost half of the
BRCA2 mutation carriers will develop breast cancer [12]. More
recently, mutations of 20 genes including CHEK2, PALB2, BRIP1,
RAD51C, NLRP2, BARD1, FGFR2, TOX3, etc. were published
as susceptibility genes of breast cancer which are frequently mutated
in general population and involve in developing of breast cancer
[13]. For example, CHEK2 plays a vital role in adjustment of p53
and BRCA1 function, and a single DNA building block at nucleo-
tide sequence 1100delC will lead to abnormality, consequently
enhance the risk of breast cancer and influence the response to
therapy [14, 15]. PALB2 gene was found as important in breast
cancer risk as BRCA1 and BRCA2 which can increase the risk 5 to
9 times higher than average [16, 17] Hence, these genes are sup-
posed to as promising mechanism and new markers which could be
widely applied in precision medicine.
As we mentioned before, breast cancer is a highly heteroge-
neous tumor with significant individual differences in molecular
immune-phenotype, biological behavior, histopathologic morphol-
ogy, and response to therapy. Such differences were believed to be
caused by molecular differences mainly in the abnormal expression
of key genes [18]. It was demonstrated that there are at least four
genetic subtypes named Luminal A (ER+, Her2-, G1/2), Luminal
B/Her2 negative (ER+, Her2-, G3), HER2-enriched (ER-, Her2
+), and basal-like/triple negative (ER-, PR-, Her2-) in breast can-
cer. Thus, identification of genetic subtypes and evaluated different
patterns of metastases in breast cancer is a major demand for
diagnosis, evaluation of prognosis, and generation of individualized
therapeutic schedule [19, 20]. It was reported that triple-negative
patients have a higher rate of distant metastasis and a poorer prog-
nosis than other breast cancer subtypes [21]. Another study using
proteomic profiles discovered that different expression of proteins
between subtypes were related to energy metabolism, mRNA trans-
lation, cell growth, and cell to cell communication [22].The Cancer
Genome Atlas(TCGA) Network also associated the profiling of
gene expression in each subtype with mutation profiles and DNA
copy number variations. For instance, both luminal A and B which
are classified as ER+ tumors demonstrated a high frequency of
PIK3CA mutations, whereas basal tumors showed a high frequency
of TP53 mutations [23].
56 Jing Yan et al.
3 Early Diagnosis and Prognostic Assessing Using Sequencing Approach
Diagnosis for breast cancer includes tumor biopsies, biomarkers as

well as imaging techniques which have yielded valuable information
for decades. However, these traditional diagnostic methods have
several drawbacks due to their inability to capture the variation in
genetic heterogeneity and invasiveness [24]. Nowadays, with the
rapid progress in massive parallel sequencing methods known as
next-generation sequencing (NGS), gene screening has been used
to discover relationships between genomic variations and diseases
of interest, which further facilitate the development of precision
medicine [25]. Principle of NGS is based on the reading of numer-
ous fragments randomly digested from genomic material repeatedly
(ranging from 100 to 10,000 times). This new approach makes it
possible for the sequencing of the entire human genome in several
hours with low price, which is hard to analyze using the Sanger
sequencing method [26]. To have precise information from the
sequencing for diagnosis purposes, the data first should be ana-
lyzed. However, the most challenge of NGS is the lack of sufficient
databases for analysis of incidental mutations as well as short of
sufficient bioinformatics software to deal with mass of data gener-
ated by sequencing [27].
As the increasing demand for personalized screening, it is more
and more significant to include molecular diagnosis into the screen-
ing framework for patients [28]. Genetic testing is necessary for
patients with a family history of hereditary cancer. Evidence sug-
gests that diagnosing patients with a family history of breast cancer
mutations has better psychosocial benefits [29]. Counseling and
genetic information can support the patient in making decision and
reduce risk [30]. With the development of NGS, it has become
possible to measure large numbers of genes with whole genome
sequencing (WGS) which facilitate precision diagnose and being
used as emerging standard for genetic testing. Even WGS is widely
used in tumor profiling frequently, other genetic screenings are also
in common use. On the condition that cancer types have been well-
studied, tumor-related genes or loci could be screened instead of
sequencing the whole genome. A study showed that other genes
besides BRCA1/2 which are associated with breast cancer could be
detected using 25-related-genes panel by NGS [31]. Physicians
typically provide genetic tests of BRCA1/2 genes for patients with
a family history of BRCA gene mutations [32]. Nowadays,
BRCA1/2 gene mutation detection has become a part of clinical
management in women with a family history in developed
countries [33].
Another detection being called “liquid biopsy” involves screen-
ing of enriched circulating tumor cells (CTCs) or circulating tumor
DNA (ctDNA) in peripheral blood using NGS. Due to the ability
of invasion and metastasis, breast cancer has a high clinical mortal-

ity. An early and accurate detection of metastasis status is of great
value. The liquid biopsy is believed to having the ability of higher
sensitivity, noninvasive, early tumor diagnosis, and better
recurrence monitoring. It can offer the potential significance of
“real-time” diagnosis [34]. Except for BRCA genes, high ctDNA
mutations of TP53, PIK3CA, and ESR1 were significantly asso-
ciated with lymph node metastasis, cancer recurrence, and poor
overall survival outcomes [35]. Some researchers declared that
others substance in peripheral blood of breast cancer patients
could also be biomarker of liquid biopsy. Fibronectin on the extra-
cellular vesicles secreted from breast cancer cells have high diagnos-
tic accuracy, sensitivity, and specificity [36] Circulation of
microRNAs (miRNA-133, miRNA-195) could also be tested as
potential biomarker to identify the metastatic breast cancer [37].
4 Personalized and Precise Therapy
The traditional treatment for breast cancer including radiation,

surgery, chemotherapy, and hormone therapy, which often fail to
eradicate tumor cells but still damage normal tissue, especially the
body’s immune system. Although rapid progress has been made in
conventional therapeutics, the treatment is still complicated with
development of resistance in cancer cells and advancement in stage
due to genetic heterogeneity. When the same treatment is imple-
mented, each patient response to treatment inconsistently. Early
study found the 127 gene mutations in a sample of tissue from a
43-year-old woman with breast cancer which is an example of the
genetic heterogeneity [38]. Currently, people pay more and more
attention to gene therapy of tumor which is also called personalized
and precise therapy.
The personalized and precise therapy of breast cancer is a
unique area owing to its special subtypes and unique driving
genes. It becomes more fascinating because the mutations that
cause progression of breast cancer also act as specific targets for
treatment. Recently, apart from the traditional chemotherapeutic
drugs, targeting the precise molecules according to molecular char-
acterization called targeted therapy are also being developed. In
breast cancer, genetic profile (ER+, HER2+, and triple negative) is
significant to choose special chemotherapeutic agent and individual
treatments [39]. It was reached agreement that the combination
with Trastuzumab (Herceptin) known as a HER2 inhibitors and
pertuzumab (also being called 2C4, trade name Perjeta) which is a
monoclonal antibody targeted HER2, as well as docetaxel could be
utilized for the treatment of metastatic HER-positive breast cancer.
This effective drug combos were also used as a neoadjuvant in early
HER2-positive breast cancer [40].
58 Jing Yan et al.
With the rapid progress of precision medicine, there are numer-

ous new targeted drugs which achieve satisfactory therapeutic
effects. For instance, endocrine therapies employing tamoxifen
and/or aromatase inhibitors are significant therapeutic schedule
for the targeted treatment in hormone-responsive breast cancer
[41]. It was well known that breast cancer cells rely on ER signaling
to drive tumor growth despite exposure to CDK4/6 inhibitors.
Importantly, Elacestrant known as a novel, nonsteroidal combined
estrogen receptor modulator (SERM) and estrogen receptor
degrader (SERD) can inhibit ER-dependent growth despite of
CDK4/6 inhibitor resistance observed such as CDK6 overexpres-
sion, upregulated cyclinE1 and E2F1 and Rb loss [42]. More
recently, the Food and Drug Administration (FDA) approved Tala-
zoparib (an adenosine diphosphate-ribose polymerase inhibitor)
for patients with deleterious germline BRCA-mutated, HER2-neg-
ative metastatic breast cancer. Researcher discovered that Talazo-
parib is shown antitumor effects in patients with advanced breast
cancer and germline mutations in BRCA1/BRCA2 which provided
a better effects than standard chemotherapy in progression-free
survival [43]. YBX1 which is briefly called nuclear expression of
Y-box binding protein has close relationship with clinical poor
outcomes and drug resistance in breast cancer, and phosphorylated
YBX1 (pYBX1) promotes expression of genes facilitate the drug
resistance and cell growth. It was confirmed that Everolimus
which is an mTORC1 inhibitor and a novel multikinase inhibitor
of AKT can suppress the phosphorylation with YBX1 and overcome
antiestrogen resistance in vivo and in vitro suggesting that it has
potential therapeutic value in treatment of progressive and
antiestrogen-resistant breast cancer [44]. Epertinib is an effective
inhibitor of HER2, EGFR, and HER4. It was reported that daily
oral Epertinib intake combined with trastuzumab, or with trastu-
zumab plus capecitabine have safety and encouraging antitumor
outcome [45]. In addition, there has been some progress in com-
bination therapies. Knudsen et al. combined cyclin-dependent
kinase (CDK) 4 and 6 inhibitors with treatment, and found that
the combination therapy inhibits tumor recurrence and improves
survival in breast cancer patients [46].
Gene therapy is another new technology through that normal
or therapeutic genes can be introduced into target cells in a specific
way to correct gene mutations. Compared with the traditional
treatment model, gene therapy has better targeting and pertinence,
and is more suitable for individualized treatment in breast cancer
patients. There are several methods of gene therapy, such as gene
replacement and gene knockout. Among that, suicide gene therapy
has bright prospect in individual therapy although it remains in the
research stage [47]. Mohseni et al. used the iC9 suicide gene to
induce apoptosis in McF-7 breast cancer cells, and the effect was
also shown to block cell cycle in combination with chemotherapy
drugs [48].
5 Conclusion
Precision medicine is a comprehensive approach that take into

account variability in genetic makeup, lifestyle of individual, envir-
onments et al. for personalized disease prevention, diagnosis, and
treatment. In recent years, the availability of large-scale omics data,
gene–lifestyle and gene–environment interaction information, big
data analytics as well as predictive algorithms have enabled us to
develop precision medicine strategies on case-by-case basis. How-
ever, it is just the beginning and it still has a long way to go.
Acknowledgements
This work was supported by the Internationally Technological

Cooperation Project of Gansu Province (18YF1WA117), Scientific
Research Project of Gansu Medical and Health Industry
(GSWSKY2016-14), and the Fundamental Research Funds for
the Central Universities (lzujbky-2017-81).
References
1. Torre LA, Bray F, Siegel RL et al (2015) Global BRCA1 locus on chromosome 17q. Nat Genet
cancer statistics, 2012. CA Cancer J Clinic 65 7(1):103–107
(2):87–108 11. Petrucelli N, Daly MB, Feldman GL (2010)
2. Rojas K, Stuckey A (2016) Breast cancer epide- Hereditary breast and ovarian cancer due to
miology and risk factors. Clin Obstet Gynecol mutations in BRCA1 and BRCA2. Genet
59(4):651–672 Med 12(5):245
3. Odle TG (2017) Precision medicine in breast 12. Chen S, Parmigiani G (2007) Meta-analysis of
cancer. Radiol Technol 88(4):401m–421m BRCA1 and BRCA2 penetrance. J Clin Oncol
4. Polyak K (2011) Heterogeneity in breast can- 25(11):1329–1333
cer. J Clin Invest 121(10):3786–3788 13. Rossing M, Sorensen CS, Ejlertsen B et al
5. Haukaas TH, Euceda LR, Giskeodegard GF (2019) Whole genome sequencing of breast
et al (2016) Metabolic clusters of breast cancer cancer. APMIS 127(5):303–315
in relation to gene- and protein expression sub- 14. Mellemkjaer L, Dahl C, Olsen JH et al (2008)
types. Cancer Metabol 4:12 Risk for contralateral breast cancer among car-
6. Kumar-Sinha C, Chinnaiyan AM (2018) Preci- riers of the CHEK2*1100delC mutation in the
sion oncology in the age of integrative geno- WECARE study. Br J Cancer 98(4):728–733
mics. Nat Biotechnol 36(1):46–60 15. Knappskog S, Chrisanthar R, Lokkevik E et al
7. Hayes DF (2016) Considerations for imple- (2012) Low expression levels of ATM may
mentation of cancer molecular diagnostics substitute for CHEK2 /TP53 mutations pre-
into clinical care. Am Soc Clin Oncol Educ dicting resistance towards anthracycline and
Book 35:292–296 mitomycin chemotherapy in breast cancer.
8. Green ED, Watson JD, Collins FS (2015) Breast Cancer Res 14(2):R47
Human genome project: twenty-five years of 16. Weitzel JN, Neuhausen SL, Adamson A et al
big biology. Nature 526(7571):29–31 (2019) Pathogenic and likely pathogenic var-
9. Miki Y, Swensen J, Shattuck-Eidens D et al iants in PALB2, CHEK2, and other known
(1994) A strong candidate for the breast and breast cancer susceptibility genes among 1054
ovarian cancer susceptibility gene BRCA1. Sci- BRCA-negative Hispanics with breast cancer.
ence (New York, NY) 266(5182):66–71 Cancer 125(16):2829–2836
10. Stratton MR, Ford D, Neuhasen S et al (1994) 17. Antoniou AC, Casadei S, Heikkinen T et al
Familial male breast cancer is not linked to the (2014) Breast-cancer risk in families with
ALGrawany
60 Jing Yan et al.
mutations in PALB2. N Engl J Med 371 randomized, controlled trial. Genet Med 19
(6):497–506 (3):330–336
18. Prat A, Perou CM (2009) Mammary develop- 31. Tung N, Battelli C, Allen B et al (2015) Fre-
ment meets cancer genomics. Nat Med 15 quency of mutations in individuals with breast
(8):842–844 cancer referred for BRCA1 and BRCA2 testing
19. Masood S (2016) Breast cancer subtypes: mor- using next-generation sequencing with a
phologic and biologic characterization. 25-gene panel. Cancer 121(1):25–33
Women’s Health (London, England) 12 32. Sefton P (2017) Testing for BRCA1/2 muta-
(1):103–119 tions. JAMA 318(20):2054
20. Silverstein A, Sood R, Costas-Chavarri A 33. Alemar B, Gregorio C, Herzog J et al (2017)
(2016) Breast cancer in Africa: limitations and BRCA1 and BRCA2 mutational profile and
opportunities for application of genomic medi- prevalence in hereditary breast and ovarian can-
cine. Int J Breast Cancer 2016:4792865 cer (HBOC) probands from Southern Brazil:
21. Howlader N, Cronin KA, Kurian AW et al are international testing criteria appropriate for
(2018) Differences in Breast Cancer Survival this specific population? PLoS One 12(11):
by Molecular Subtypes in the United States. e0187630
Cancer Epidemiol Biomark Prev 27 34. Neumann MH, Bender S, Krahn T et al (2018)
(6):619–626 ctDNA and CTCs in liquid biopsy–current sta-
22. Tyanova S, Albrechtsen R, Kronqvist P et al tus and where we need to progress. Comput
(2016) Proteomic maps of breast cancer sub- Struct Biotechnol J 16:190–195
types. Nat Commun 7:10259 35. Lee JH, Jeong H, Choi JW et al (2018) Liquid
23. Curtis C, Shah SP, Chin SF et al (2012) The biopsy prediction of axillary lymph node metas-
genomic and transcriptomic architecture of tasis, cancer recurrence, and patient survival in
2,000 breast tumours reveals novel subgroups. breast cancer: a meta-analysis. Medicine 97
Nature 486(7403):346–352 (42):e12862
24. Marrugo-Ramirez J, Mir M, Samitier J (2018) 36. Moon PG, Lee JE, Cho YE et al (2016) Fibro-
Blood-based cancer biomarkers in liquid nectin on circulating extracellular vesicles as a
biopsy: a promising non-invasive alternative to liquid biopsy to detect breast cancer. Oncotar-
tissue biopsy. Int J Mol Sci 19(10) get 7(26):40189–40199
25. Ng CK, Schultheis AM, Bidard FC et al (2015) 37. McAnena P, Tanriverdi K, Curran C et al
Breast cancer genomics from microarrays to (2019) Circulating microRNAs miR-331 and
massively parallel sequencing: paradigms and miR-195 differentiate local luminal a from
new insights. J Natl Cancer Inst 107(5) metastatic breast cancer. BMC Cancer 19
26. Gu W, Miller S, Chiu CY (2019) Clinical meta- (1):436
genomic next-generation sequencing for path- 38. Mukherjee S (2010) A distorted version of our
ogen detection. Annu Rev Pathol 14:319–338 normal selves. The Emperor of All
27. Bao R, Huang L, Andrade J et al (2014) Maladies:340–355
Review of current methods, applications, and 39. Wang J, Xu B (2019) Targeted therapeutic
data management for the bioinformatics analy- options and future perspectives for HER2-
sis of whole exome sequencing. Cancer Infor- positive breast cancer. Signal Transduct Target
mat 13(Suppl 2):67–82 Ther 4:34
28. Onega T, Beaber EF, Sprague BL et al (2014) 40. Wuerstlein R, Harbeck N (2017) Neoadjuvant
Breast cancer screening in an era of persona- therapy for HER2-positive breast cancer. Rev
lized regimens: a conceptual model and Recent Clin Trials 12(2):81–92
National Cancer Institute initiative for risk- 41. Shagufta AI (2018) Tamoxifen a pioneering
based and preference-based approaches at a drug: an update on the therapeutic potential
population level. Cancer 120(19):2955–2964 of tamoxifen derivatives. Eur J Med Chem
29. Meiser B, Quinn VF, Gleeson M et al (2016) 143:515–531
When knowledge of a heritable gene mutation 42. Patel HK, Tao N, Lee KM et al (2019) Elaces-
comes out of the blue: treatment-focused trant (RAD1901) exhibits anti-tumor activity
genetic testing in women newly diagnosed in multiple ER+ breast cancer models resistant
with breast cancer. Eur J Hum Genet 24 to CDK4/6 inhibitors. Breast Cancer Res 21
(11):1517–1523 (1):146
30. Metcalfe KA, Dennis CL, Poll A et al (2017) 43. Litton JK, Rugo HS, Ettl J et al (2018) Tala-
Effect of decision aid for breast cancer preven- zoparib in patients with advanced breast cancer
tion on decisional conflict in women with a and a germline BRCA mutation. N Engl J Med
BRCA1 or BRCA2 mutation: a multisite, 379(8):753–763
44. Shibata T, Watari K, Kawahara A et al (2019) CDK4/6 inhibition in relation to ER+/

Targeting phosphorylation of Y-box binding HER2-breast cancer. Oncotarget 7
protein YBX1 by TAS0612 and everolimus in (43):69111–69123
overcoming antiestrogen resistance. Mol Can- 47. Duzgunes N (2019) Origins of suicide gene
cer Ther 19(3):882–894 therapy. Methods Mol Biol (Clifton, NJ)
45. Macpherson IR, Spiliopoulou P, Rafii S et al 1895:1–9
(2019) A phase I/II study of epertinib plus 48. Mohseni-Dargah M, Akbari-Birgani S, Madadi
trastuzumab with or without chemotherapy in Z et al (2019) Carbon nanotube-delivered iC9
patients with HER2-positive metastatic breast suicide gene therapy for killing breast cancer
cancer. Breast Cancer Res 22(1):1 cells in vitro. Nanomedicine (London, Eng-
46. Knudsen ES, Witkiewicz AK (2016) Defining land) 14(8):1033–1047
the transcriptional and biological response to
Chapter 6
Application and Analysis of Biomedical Imaging Technology

in Early Diagnosis of Breast Cancer
Lin Chen, Nan Jiang, and Yuxiang Wu
Abstract
Breast cancer is the primary malignant tumor that endangers women’s health. The incidence of breast
cancer is increasing rapidly in recent years. Accurate disease evaluation before treatment is the key to the
selection of treatment options. Biomedical imaging technology plays an irreplaceable role in the diagnosis
and staging of tumors. Various imaging methods can provide excellent temporal and spatial resolution from
multiple levels and perspectives and have become one of the most commonly used means of breast cancer
early detection. With the development of radiomics, it has been found that early imaging diagnosis of breast
cancer plays an important guiding role in clinical decision-making. The purpose of this study is to explore
the characteristics of various breast cancer imaging technologies, promote the development of individua-
lized accurate diagnosis and treatment of imaging, and improve the clinical application value of radiomics in
the early diagnosis of breast cancer.
Key words Breast cancer, Biomedical imaging, radiomics, computer-aided diagnosis, Digital
mammography
1 Introduction
Breast cancer is the most common malignancy in women. Over the

past half century, many studies in various countries around the
world have confirmed that early screening of breast cancer imaging
is one of the most effective methods to improve the early diagnosis
rate, survival rate, and quality of life [1–4]. The World Health
Organization (WHO) has also clearly listed early breast cancer as
a curable disease. Early diagnosis and early treatment are the best
way to improve the cure rate of breast cancer [3, 5, 6]. The devel-
opment of mammography, ultrasound, MRI, nuclear medicine,
biomedical optics, and computer-aided diagnosis (CAD) technol-
ogies has improved the level of early diagnosis of breast cancer. The
application of emerging technologies such as electrical impedance,
electronic palpation, heat maps, and optical imaging, combined
with traditional diagnostic methods, can improve diagnostic
63
64 Lin Chen et al.
accuracy and enable flexibility and simplicity in rapid screening

applications [1, 6–9]. Based on a multifunctional nanoprobe that
integrates imaging and treatment, it provides effective solutions to
clinical problems such as early diagnosis of breast cancer, curative
effect evaluation, and relative lag in treatment. This promotes
individualized precise diagnosis and treatment of breast cancer
significantly.
2 Mammography
After the French doctor Gross developed the molybdenum target

anode X-ray machine in 1969, mammography technology devel-
oped rapidly [10]. Panoramic digital mammography system was
approved by the US FDA in 2000. Computer-aided diagnosis
(CAD) was used for breast imaging diagnosis in 2002. Three-
dimensional mammography technology was used in 2004. Digital
mammography (DBT) technology was used for breast examination
in 2006 [11, 12]. At present, the mammography technology has
become more mature, can clearly display tiny masses and fine
calcifications, and can be quasi-deterministic and localized. It is
currently recognized as the world’s preferred and most effective
method of mammography. Studies have shown that mammography
can detect 59% of noninvasive breast cancers with a diameter of less
than 1 cm and 53% of invasive breast cancers. However, mammog-
raphy has certain limitations, and its sensitivity and specificity are
affected by breast density. Because the breasts of young women are
in the radiation-sensitive period and the breast tissue is dense at this
time, it is not easy to detect the lesions. It is generally believed that
mammography is not suitable for screening of women under
40 years of age. Clinically, the application of molybdenum target
X-ray combined with spiral CT has significant effects in improving
the accuracy of patient diagnosis [13–15].
3 Ultrasound Imaging of Breast Cancer
With the development of digital signal processing technology and

the use of new ultrasonic imager, the quality of ultrasonic imaging
of breast and small organs has been significantly improved, making
the diagnosis and identification accuracy of breast cancer signifi-
cantly improved [16–18]. The optimization of operating condi-
tions of high-frequency linear array probes improves the transverse
resolution and the longitudinal resolution and reduces the noise.
Tissue harmonic imaging and composite imaging can better show
the edge and internal echo structure of breast tumors and improve
the detection rate of microcalcification. Three-dimensional ultra-
sound technology can visually display the shape, internal structure,
Application and Analysis of Biomedical Imaging Technology in Early. . . 65
and the relationship between the breast mass and the surrounding
tissue, can clearly show the invasion degree of the tumor to the
surrounding tissue and the shape and distribution of the blood
vessels inside the lesion, and can quantitatively evaluate the richness
of the blood supply of the tumor [19–21].
The advantage of ultrasound is that the patient has no pain, no
radiation damage, especially suitable for the examination of lacta-
tion, pregnancy, and young women, experts have suggested that
ultrasound as the preferred means of breast cancer screening in
China. It is believed that X-ray is more sensitive to detect calcified
breast cancer than high-frequency ultrasound, and high-frequency
ultrasound is better than X-ray in detecting blood flow signals.
Ultrasound imaging and X-ray photography are still the gold com-
bination of mammography, and the use of ultrasound imaging can
improve the sensitivity and specificity of mammography. Clinical
studies have also confirmed that combined ultrasound and molyb-
denum photography can significantly improve the early detection
rate of breast cancer patients [22].
4 Magnetic Resonance and Nuclear Medicine in Breast Imaging
In 1982, Ross et al. first applied magnetic resonance imaging to the

detection of breast lesions [4]. Following the development of the
mammary gland surface coil, high contrast, the introduction of
paramagnetic contrast agents and fast dynamic enhancement,
diffusion-weighted imaging, perfusion-weighted imaging and
magnetic resonance (NMR) spectroscopy analysis, fat suppression,
and other new technology application in breast imaging, magnetic
resonance (NMR) has made great improve signal-to-noise ratio,
can be obtained without overlapping images of 3 mm thick, can
effectively find small lesions, MRI breast check increasingly
brought to the attention of the two aspects of clinical and imaging
[4, 23, 24]. Compared with mammography, the detection of
lesions by MRI is not affected by gland density and can reflect the
characteristics of blood flow inside the lesions. Compared with
ultrasound, MRI provides higher spatial resolution images and is
not affected by operator dependence. From the perspective of
diagnosis, the sensitivity of MRI to early breast cancer, the accuracy
of breast cancer staging, and the consistency with the scope of
histological lesions are better than X-ray and ultrasound. MRI has
unique advantages for breast examination due to its high soft tissue
resolution and no radiation. MRI has a sensitivity of 88.4% to 100%
for the diagnosis of breast cancer and can detect breast cancer that
cannot be detected by clinical examination, mammography, and
ultrasound, but has a low specificity of 37% to 97%. In view of the
advantages and disadvantages of MRI, the high cost of examina-
tion, the lack of standardized operating techniques and standards,
66 Lin Chen et al.
and the limited detection of microcalcification, MRI is generally

not recommended as a screening tool for breast cancer screening
above the scale [9, 25, 26].
The principle of nuclear medicine imaging is to inject radioac-
tive tracer drugs into human body. Due to its own physiological
characteristics, the drug automatically concentrates on the organ to
be tested, and the nuclear medicine imaging device measures the
concentration distribution of radioactive drugs in the organ, so as
to achieve functional imaging of human tissues. The development
of nuclear medicine imaging has gone through several stages: cam-
era, single-photon emission-computed tomography, positron
emission-computed tomography (PET), and PET/CT. PET can
perform functional imaging at the molecular level with high sensi-
tivity and specificity, but its anatomical structure is unclear.
PET/CT imaging can complete both PET and CT examination at
the same time. The anatomical image information obtained by CT
can help distinguish the physiological uptake of tracer and the
uptake of diseased tissue and can detect and locate early breast
cancer lesions. PET/CT achieves the same image fusion of PET
molecular functional image and CT anatomical image, which can
reflect the morphological structure, pathological and physiological
changes of the lesion at the same time, and significantly improve the
accuracy of diagnosis [27, 28].
5 Biomedical Optical Imaging Technology
Near-infrared (NIR) mammary gland scanning technology is to

make use of its absorption characteristics of hemoglobin to form
an image, carry on the mammary gland full-field scanning, obser-
vation and diagnosis, has the advantages of simplicity, intuition,
convenience, and so on, is one of the important means of diagnos-
ing breast cancer, it can obtain the infrared mammary gland image
quickly, painless and lossless [29, 30]. As the mainstream diagnosis
technology of breast cancer, molybdenum target soft X-ray diagno-
sis has ray damage to human body, while near-infrared imaging
diagnosis is a non-invasive, repeatable diagnostic technology suit-
able for large-scale screening, but with a high false positive rate, it
has experienced a process from emergence to withdrawal in breast
cancer screening. In recent years, as people pay more and more
attention to molybdenum target X-ray diagnosis of damage and the
emergence of new technology, near-infrared optical imaging tech-
nology has found its value again [31–35].
5.1 Digital Infrared Early breast cancer and precancerous lesions mostly occur in the
Thermal Imaging stage of functional change, but have not developed into organic
lesions. Digital infrared thermal imaging (DITI) is a noninvasive
functional examination method, which is of great significance for
the diagnosis and differentiation of breast masses. DITI uses high

sensitivity, high speed low temperature infrared camera to detect
the body of the infrared thermal radiation and displays temperature
field of the human body surface, helping doctors determine the
parts of the lesion, the nature of the disease, the extent of the lesion.
This technique allows you to measure not only the depth of the
heat source in your body but also the shape and size of the heat.
Clinical trials have shown higher accuracy and specificity, especially
in the diagnosis of breast diseases. DITI reconstruction of the
image by using image analysis algorithm and mammary gland
images in different colors (red, orange, yellow), according to any
suspected regional hotspots (abnormal) can be marked on the
breast imaging. The application of thermal tomography in clinical
diagnosis has been accepted by western developed countries and
has been certified by FDA. Thermal tomography is a functional
information supplement for morphologic diagnostic methods such
as b-ultrasound, CT and MRI [36, 37].
5.2 Dynamic Optical The blood vessels in breast malignant tumors are tennis-shaped,
Breast Imaging with the characteristics of large total blood vessel cross section, slow
blood flow rate, strong tumor cell metabolism, and large oxygen
consumption, and present a special phenomenon of high blood and
low oxygen content. Deoxyhemoglobin has a high absorption rate
of light with a wavelength of 640 nm, and has strong light absorp-
tion sensitivity. Dynamic Optical Breast Imaging (DOBI) is an
advanced digital Dynamic functional breast cancer diagnosis device,
which is an early breast cancer diagnosis and screening instrument
launched by DOBI Medical International [38–40]. It records the
changes of blood volume and metabolic rate of new blood vessels in
breast tumors in real time by means of slight uniform pressure
pulse. Through continuous monitoring (45 s) and quantitative
analysis of these two indicators, early breast cancer above 2 mm
can be diagnosed. The DOBI shows the region with increased
blood volume through morphological images, indicating the pres-
ence of lesions. The characteristics of the lesion were determined by
the metabolic rate curve. It supports dynamic functional examina-
tion, continuous imaging to reflect physiological changes, and pre-
and post-correlation processing of the collected images. DOBI has
a specificity of 74%, a sensitivity of 92%, and a diagnostic accuracy of
79%. DOBI has a good correlation between breast cancer diagnosis
and MRI, and the diagnosis is 5–8 years earlier than mammogra-
phy, showing a significant advantage in the early diagnosis of breast
cancer. DOBI is not affected by breast density, and is suitable for
the screening and diagnosis of all female breast cancers, as well as
the efficacy evaluation and condition monitoring after breast cancer
treatment [40–43].
68 Lin Chen et al.
5.3 Blood Oxygen Using a combination of 805 nm (or 850 nm) infrared light and
Functional Imaging 735 nm (or 760 nm) red light, the functional imaging of mammary
gland blood oxygen can detect and compare the blood oxygen
content of the patient mass with that of the healthy tissue to
determine the nature of the mass, taking advantage of the charac-
teristics of the cancer tissue and the different absorption character-
istics in the near-infrared region. Blood oxygen functional imaging
detects the changes of oxygenated hemoglobin and deoxygenated
hemoglobin in the breast tissue, provides the metabolic status of
the breast, displays the information of the breast structure and
lesions, and realizes the integration of the information of the breast
anatomy and the information of the function, which greatly
improves the diagnosis level of the breast disease [44–46].
5.4 Near-infrared NIR fluorescence imaging technology is a new in vivo imaging

Fluorescence Imaging method. The wavelength of the fluorescent dye is 650–900 nm,
Technology and the absorption rate of nonspecific tissue is very low. Therefore,
compared with the traditional fluorescent dye with a wavelength
range of 300–500 nm, NIR fluorescence imaging can obtain fewer
background signals and clearer images. Targeted fluorescent probes
can accumulate in tumor cells in large quantities and enhance the
fluorescence signal, so they have the advantage of accurate identifi-
cation of breast cancer cells. NIR fluorescence imaging cannot only
provide a basis for the early diagnosis of tumors in a noninvasive
and efficient manner, but also can be used for intraoperative imag-
ing of breast-conserving breast cancer surgery, reducing normal
tissue damage, and achieving the purpose of accurate breast cancer
resection. At present, the main difficulty in applying NIR fluores-
cence imaging to breast cancer surgery is the lack of highly selective
fluorescent dyes designed for breast cancer cells [47–50].
5.5 Photoacoustic Accurate diagnosis and treatment of breast cancer at the molecular
Technology Based on and cellular level has always been the focus of early detection and
Multifunctional individualized treatment of breast cancer. Multifunctional nanop-
Nanoprobe robes can deliver targeted ligands, imaging agents, and drugs to the
tumor site at the same time by virtue of their versatility, realizing
“multimodal diagnosis and treatment“at the gene and molecular
level in vivo. Some nanoprobes (NPs) have unique physicochemical
properties and can be used as a common tool for “integrated
diagnosis and treatment.” The advantages of nano-drug delivery
system over traditional drugs also open up a new way for the
treatment of triple-negative and drug-resistant breast cancer.
Nanoprobes use different materials (such as metals, organics, and
semiconductor particles) as platforms to produce structures similar
in size to biological macromolecules. The surface of NP is often
coated with the hydrophilic polymer polyethylene glycol or func-
tionalized by facultative ionic groups to resist the adsorption of
serum proteins. Modified NP can be coupled with a variety of
molecules, including targeted molecules, fluorescent groups, radio-

isotopes, and therapeutic drugs. Most of the research on NP in
breast cancer is in the preclinical stage. NP has enhanced perme-
ability and residence effects through new blood vessels in breast
cancer, exuding from the blood vessel pores, while the lymphatic
system of the tumor is not perfect and cannot be cleared in time.
Therefore, NP can accumulate in the tumor tissues for a long time,
laying the foundation for further imaging detection. Np-based
breast cancer imaging includes magnetic resonance imaging, CT
imaging, radionuclide imaging, and photoacoustic imaging. NP
can also play a targeted therapeutic role by coupling chemotherapy
drugs or its own therapeutic effects. Although multimodal nanop-
robes have some advantages, they still have some limitations. The
particle size of multimodal polymer nanoprobe is relatively large,
which affects the penetrability of target cells. Multilayer nanop-
robes are difficult to be used in clinic because of their complex
structure and high cost. There were also significant differences in
probe circulation, tumor targeting, imaging, and therapeutic effi-
cacy. In addition, the mechanism of interaction between NP and
organism is still lacking, and biotoxicity is difficult to predict
[46, 51–55].
6 Conclusion
Radiomics technology can obtain real-time, dynamic, three-

dimensional, and functional imaging of normal breast and cancer
tissue. Through the multi-dimensional, multi-parameter, and
multi-mode functional imaging and molecular imaging, we can
combine the spatial relationship with the anatomical structure of
breast tissue and pathological tissue for comprehensive diagnosis
[5, 56, 57]. The application of radiomics technology in breast
cancer patients is an emerging translational research topic and is
often expected to improve diagnosis and characterization in experi-
mental design. Radiomics characteristics (such as intensity, shape,
texture, or wavelet) provide information about the cancer pheno-
type and tumor microenvironment that is different and comple-
mentary to other clinical and treatment-related data or genomic
data [58–61]. When radiomics technology-derived data are com-
bined with other relevant data to infer the outcome, an accurate
evidence-based and stable clinical decision support system is gen-
erated. There are broad methodological differences in single
research, so there is considerable room for improvement, such as
the use of standardized methods to improve the quality of research.
Just as artificial intelligence (AI) requires advanced computing and
statistical science, radiomics also requires highly collaborative inter-
disciplinary, inter-institutional, and international collaboration
[62, 63]. Currently, there is still little overlap between radiomics
70 Lin Chen et al.
and machine learning, but in the near future it will be promising to

integrate them into the study of tumor biology, to classify tumors,
and to predict response to treatment and prognosis. More careful
evaluation of prospective quantitative imaging characteristics
related to clinical outcomes is needed to improve the quality of
research and further explore the potential value of breast cancer-
related radiomics [62–66].
References
1. Jerome-D’Emilia B, Suplee P, Kushary D 7380 apparently healthy women. Dtsch Med
(2019) A 10-year evaluation of New Jersey’s Wochenschr 96:1207–1210
national breast and cervical cancer early detec- 11. Allweis TM, Nissan A, Spira RM, Sklair-
tion program: comparison of stage at diagnosis Levy M, Freund HR, Peretz T (2003) Screen-
in enrollees and nonenrollees. J Women’s ing mammography for early diagnosis of breast
Health 29(2):230–236 cancer: facts, controversies, and the implemen-
2. Klijn JG (2010) Early diagnosis of hereditary tation in Israel. Harefuah 142:281–286. 317
breast cancer by magnetic resonance imaging: 12. Kujiraoka Y, Murakami A (2006) Diagnosis of
what is realistic? J Clin Oncol 28:1441–1445 early stage of breast cancer by mammography.
3. Milosevic M, Jankovic D, Milenkovic A, Stoja- Nihon Rinsho Jap J Clin Med 64:486–490
nov D (2018) Early diagnosis and detection of 13. Gutnik LA, Castro MC (2016) Does spatial
breast cancer. Technol Health Care access to mammography have an effect on
26:729–759 early stage of breast cancer diagnosis? A
4. Tseshkovskii MS, Labetskii II (1986) Radio- county-level analysis for New York State. Breast
graphic methods of breast imaging in the J 22:127–130
early diagnosis of cancer. Med Radiol 14. Lacombe J, Mange A, Bougnoux AC, Prassas I,
31:66–70 Solassol J (2014) A multiparametric serum
5. Cardoso F, Kyriakides S, Ohno S, Penault- marker panel as a complementary test to mam-
Llorca F, Poortmans P, Rubio IT et al (2019) mography for the diagnosis of node-negative
Early breast cancer: ESMO Clinical Practice early-stage breast cancer and DCIS in young
Guidelines for diagnosis, treatment and women. Cancer Epidemiol Biomark Prev
follow-up. Ann Oncol 30:1674 23:1834–1842
6. Fouladi N, Pourfarzi F, Daneshian A, Alimo- 15. Nergiz-Eroglu U, Kilic D (2011) Knowledge,
hammadi S (2018) Mediating factors in early attitude and beliefs women attending mam-
diagnosis of breast cancer: from initial changes mography units have regarding breast cancer
in health to breast cancer detection. Asian Pac J and early diagnosis. Asian Pac J Cancer Prev
Cancer Prev 19:2751–2755 12:1855–1860
7. Bulut A, Bulut A (2017) Knowledge, attitudes 16. Salih A, Webb WM, Bates T (1999) Does
and behaviors of primary health care nurses and open-access mammography and ultrasound
midwives in breast cancer early diagnosis appli- delay the diagnosis of breast cancer? Breast
cations. Breast Cancer 9:163–169 8:129–132
8. Capelan M, Battisti NML, McLoughlin A, 17. Simmons R (2004) Ultrasound in the changing
Maidens V, Snuggs N, Slyk P et al (2017) The approaches to breast cancer diagnosis and
prevalence of unmet needs in 625 women liv- treatment. Breast J 10(Suppl 10):S13–S14
ing beyond a diagnosis of early breast cancer. 18. Tohno E, Ueno E (2005) Ultrasound
Br J Cancer 117:1113–1120 (US) diagnosis of nonpalpable breast cancer.
9. Jacobs MA, Wolff AC, Macura KJ, Stearns V, Breast Cancer 12:267–271
Ouwerkerk R, El Khouli R et al (2015) Multi- 19. Carlsen J, Ewertsen C, Sletting S, Vejborg I,
parametric and multimodality functional radio- Schafer FK, Cosgrove D et al (2015) Ultra-
logical imaging for breast cancer diagnosis and sound elastography in breast cancer diagnosis.
early treatment response assessment. J Natl Ultraschall Med 36:550–562. quiz 63-5
Cancer Inst Monogr 2015:40–46 20. Moon WK, Huang YS, Lo CM, Huang CS, Bae
10. Scherer E, Seifert J (1971) Mammography as a MS, Kim WH et al (2015) Computer-aided
method for early diagnosis of female breast diagnosis for distinguishing between triple-
cancer. Experiences with a screening test in negative breast cancer and fibroadenomas
ALGrawany
based on ultrasound texture features. Med diagnosis of breast cancer. Med Phys
Phys 42:3024–3035 35:2151–2161
21. Zhang YN, Wang CJ, Xu Y, Zhu QL, Zhou 32. Khanmohammadi M, Rajabi FH, Garmarudi
YD, Zhang J et al (2015) Sensitivity, specificity AB, Mohammadzadeh R, Mohammadzadeh
and accuracy of ultrasound in diagnosis of R (2010) Chemometrics assisted investigation
breast cancer metastasis to the axillary lymph of variations in infrared spectra of blood sam-
nodes in chinese patients. Ultrasound Med Biol ples obtained from women with breast cancer:
41:1835–1841 a new approach for cancer diagnosis. Eur J
22. Polat AV, Ozturk M, Polat AK, Karabacak U, Cancer Care 19:352–359
Bekci T, Murat N (2019) Efficacy of ultra- 33. Vreugdenburg TD, Willis CD, Mundy L,
sound and shear wave elastography for the Hiller JE (2013) A systematic review of elasto-
diagnosis of breast cancer-related lymphedema. graphy, electrical impedance scanning, and dig-
J Ultrasound Med 39(4):795–803 ital infrared thermography for breast cancer
23. Tozaki M (2008) Diagnosis of breast cancer: screening and diagnosis. Breast Cancer Res
MDCT versus MRI. Breast Cancer Treat 137:665–676
15:205–211 34. Gerasimova E, Audit B, Roux SG, Khalil A,
24. Morrow M, Waters J, Morris E (2011) MRI for Gileva O, Argoul F et al (2014) Wavelet-
breast cancer screening, diagnosis, and treat- based multifractal analysis of dynamic infrared
ment. Lancet 378:1804–1811 thermograms to assist in early breast cancer
25. Kawashima H, Inokuchi M, Furukawa H, Kita- diagnosis. Front Physiol 5:176
mura S (2012) Accuracy for a diagnosis of 35. Han F, Shi G, Liang C, Wang L, Li K (2015) A
breast cancer spread using 3.0T MRI. Nihon simple and efficient method for breast cancer
Rinsho Jap J Clin Med 70(Suppl 7):306–308 diagnosis based on infrared thermal imaging.
26. Banaie M, Soltanian-Zadeh H, Saligheh-Rad Cell Biochem Biophys 71:491–498
HR, Gity M (2018) Spatiotemporal features 36. Arora N, Martins D, Ruggerio D, Tousimis E,
of DCE-MRI for breast cancer diagnosis. Swistel AJ, Osborne MP et al (2008) Effective-
Comput Methods Prog Biomed 155:153–164 ness of a noninvasive digital infrared thermal
27. Brix G, Henze M, Knopp MV, Lucht R, Doll J, imaging system in the detection of breast can-
Junkermann H et al (2001) Comparison of cer. Am J Surg 196:523–526
pharmacokinetic MRI and [18F] fluorodeoxy- 37. Kosus N, Kosus A, Duran M, Simavli S, Turhan
glucose PET in the diagnosis of breast cancer: N (2010) Comparison of standard mammog-
initial experience. Eur Radiol 11:2058–2070 raphy with digital mammography and digital
28. Leithner D, Horvat JV, Bernard-Davila B, Hel- infrared thermal imaging for breast cancer
bich TH, Ochoa-Albiztegui RE, Martinez DF screening. J Turkish German Gynecol Assoc
et al (2019) A multiparametric [(18)F]FDG 11:152–157
PET/MRI diagnostic model including imag- 38. Athanasiou A, Vanel D, Balleyguier C,
ing biomarkers of the tumor and contralateral Fournier L, Mathieu MC, Delaloge S et al
healthy breast tissue aids breast cancer diagno- (2005) Dynamic optical breast imaging: a new
sis. Eur J Nucl Med Mol Imaging technique to visualise breast vessels: compari-
46:1878–1888 son with breast MRI and preliminary results.
29. Markel AL, Vainer BG (2005) Infrared ther- Eur J Radiol 54:72–79
mography in diagnosis of breast cancer (review 39. Fournier LS, Vanel D, Athanasiou A,
of foreign literature). Terapevticheskii Arkhiv Gatzemeier W, Masuykov IV, Padhani AR
77:57–61 et al (2009) Dynamic optical breast imaging:
30. Agostini V, Delsanto S, Molinari F, Knaflitz M a novel technique to detect and characterize
(2006) Evaluation of feature-based registration tumor vessels. Eur J Radiol 69:43–49
in dynamic infrared imaging for breast cancer 40. Cerussi AE, Tanamai VW, Mehta RS,
diagnosis. Conference proceedings: Annual Hsiang D, Butler J, Tromberg BJ (2010) Fre-
International Conference of the IEEE Engi- quent optical imaging during breast cancer
neering in Medicine and Biology Society neoadjuvant chemotherapy reveals dynamic
IEEE Engineering in Medicine and Biology tumor physiology in an individual patient.
Society Annual Conference 1:953–956 Acad Radiol 17:1031–1039
31. Ooi GJ, Fox J, Siu K, Lewis R, Bambery KR, 41. Flexman ML, Khalil MA, Al Abdi R, Kim HK,
McNaughton D et al (2008) Fourier transform Fong CJ, Desperito E et al (2011) Digital opti-
infrared imaging and small angle x-ray scatter- cal tomography system for dynamic breast
ing as a combined biomolecular approach to imaging. J Biomed Opt 16:076014
72 Lin Chen et al.
42. Sajjadi AY, Isakoff SJ, Deng B, Singh B, Wanyo 53. Yamaga I, Kawaguchi-Sakita N, Asao Y,
CM, Fang Q et al (2017) Normalization of Matsumoto Y, Yoshikawa A, Fukui T et al
compression-induced hemodynamics in (2018) Vascular branching point counts using
patients responding to neoadjuvant chemo- photoacoustic imaging in the superficial layer
therapy monitored by dynamic tomographic of the breast: a potential biomarker for breast
optical breast imaging (DTOBI). Biomed Opt cancer. Photo-Dermatology 11:6–13
Express 8:555–569 54. Nyayapathi N, Xia J (2019) Photoacoustic
43. Zimmermann BB, Deng B, Singh B, imaging of breast cancer: a mini review of sys-
Martino M, Selb J, Fang Q et al (2017) Multi- tem design and image features. J Biomed Opt
modal breast cancer imaging using coregistered 24:1–13
dynamic diffuse optical tomography and digital 55. Yao D, Wang Y, Zou R, Bian K, Liu P, Shen S
breast tomosynthesis. J Biomed Opt 22:46008 et al (2020) Molecular engineered squaraine
44. Vaupel P, Fortmeyer HP, Runkel S, Kallinowski nanoprobe for NIR-II/photoacoustic imaging
F (1987) Blood flow, oxygen consumption, and photothermal therapy of metastatic breast
and tissue oxygenation of human breast cancer cancer. ACS Appl Mater Interfaces 12
xenografts in nude rats. Cancer Res (4):4276–4284
47:3496–3503 56. Yu FH, Wang JX, Ye XH, Deng J, Hang J, Yang
45. Vaupel P, Mayer A, Briest S, Hockel M (2005) B (2019) Ultrasound-based radiomics nomo-
Hypoxia in breast cancer: role of blood flow, gram: a potential biomarker to predict axillary
oxygen diffusion distances, and anemia in the lymph node metastasis in early-stage invasive
development of oxygen depletion. Adv Exp breast cancer. Eur J Radiol 119:108658
Med Biol 566:333–342 57. Granzier RWY, van Nijnatten TJA, Woodruff
46. Zalev J, Richards LM, Clingman BA, Harris J, HC, Smidt ML, Lobbes MBI (2019) Explor-
Cantu E, Menezes GLG et al (2019) Opto- ing breast cancer response prediction to neoad-
acoustic imaging of relative blood oxygen satu- juvant systemic therapy using MRI-based
ration and total hemoglobin for breast cancer radiomics: a systematic review. Eur J Radiol
diagnosis. J Biomed Opt 24:1–16 121:108736
47. Abe K, Zhao L, Periasamy A, Intes X, Barroso 58. Braman N, Prasanna P, Whitney J, Singh S,
M (2013) Non-invasive in vivo imaging of near Beig N, Etesami M et al (2019) Association of
infrared-labeled transferrin in breast cancer peritumoral radiomics with tumor biology and
cells and tumors using fluorescence lifetime pathologic response to preoperative targeted
FRET. PLoS One 8:e80269 therapy for HER2 (ERBB2)-positive breast
48. Verbeek FP, Troyan SL, Mieog JS, Liefers GJ, cancer. JAMA Netw Open 2:e192561
Moffitt LA, Rosenberg M et al (2014) Near- 59. Antunovic L, De Sanctis R, Cozzi L,
infrared fluorescence sentinel lymph node Kirienko M, Sagona A, Torrisi R et al (2019)
mapping in breast cancer: a multicenter experi- PET/CT radiomics in breast cancer: promising
ence. Breast Cancer Res Treat 143:333–342 tool for prediction of pathological response to
49. Chi C, Zhang Q, Mao Y, Kou D, Qiu J, Ye J neoadjuvant chemotherapy. Eur J Nucl Med
et al (2015) Increased precision of orthotopic Mol Imaging 46:1468–1477
and metastatic breast cancer surgery guided by 60. Acar E, Turgut B, Yigit S, Kaya G (2019)
matrix metalloproteinase-activatable near- Comparison of the volumetric and radiomics
infrared fluorescence probes. Sci Rep 5:14197 findings of 18F-FDG PET/CT images with
50. Abbaci M, Conversano A, De Leeuw F, immunohistochemical prognostic factors in
Laplace-Builhe C, Mazouni C (2019) Near- local/locally advanced breast cancer. Nucl
infrared fluorescence imaging for the preven- Med Commun 40:764–772
tion and management of breast cancer-related 61. Lee SE, Han K, Kwak JY, Lee E, Kim EK
lymphedema: a systematic review. Eur J Surg (2018) Radiomics of US texture features in
45:1778–1786 differential diagnosis between triple-negative
51. Bhattacharyya K, Goldschmidt BS, Viator JA breast cancer and fibroadenoma. Sci Rep
(2016) Detection and capture of breast cancer 8:13546
cells with photoacoustic flow cytometry. J 62. Crivelli P, Ledda RE, Parascandolo N, Fara A,
Biomed Opt 21:87007 Soro D, Conti M (2018) A new challenge for
52. Wong TTW, Zhang R, Hai P, Zhang C, Pleitez radiologists: radiomics in breast cancer.
MA, Aft RL et al (2017) Fast label-free multi- Biomed Res Int 2018:6120703
layered histology-like imaging of human breast 63. Park H, Lim Y, Ko ES, Cho HH, Lee JE, Han
cancer by photoacoustic microscopy. Sci Adv 3: BK et al (2018) Radiomics signature on mag-
e1602168 netic resonance imaging: association with
disease-free survival in patients with invasive prediction of response to neoadjuvant chemo-

breast cancer. Clin Cancer Res 24:4705–4714 therapy in node-positive breast cancer patients.
64. Saha A, Yu X, Sahoo D, Mazurowski MA J Med Imaging 6:034502
(2017) Effects of MRI scanner parameters on 66. Marino MA, Pinker K, Leithner D, Sung J,
breast cancer radiomics. Expert Syst Appl Avendano D, Morris EA et al (2019)
87:384–391 Contrast-enhanced mammography and radio-
65. Drukker K, Edwards A, Doyle C, mics analysis for noninvasive breast cancer char-
Papaioannou J, Kulkarni K, Giger ML (2019) acterization: initial results. Mol Imaging
Breast MRI radiomics for the pretreatment Biol:1–8
Chapter 7
Recent Advances in DNA Repair Pathway and Its Application

in Personalized Care of Metastatic Castration-Resistant
Prostate Cancer (mCRPC)
Chenyang Xu, Shanhua Mao, and Haowen Jiang
Abstract
Prostate cancer (PCa) is one of the common malignancies in male adults. In the era of precision medicine,
many other novel agents targeting advanced prostate cancer, especially metastatic castration-resistant
prostate cancer (mCRPC), are currently being evaluated. Among all these candidate therapies, poly-ADP
ribose polymerase (PARP) inhibitors targeting DNA damage response (DDR) pathway has proven improv-
ing survival outcomes in clinical trials. In this review, we focus on recent advances in biology and clinical
implication of DDR pathway and aim to discuss the latest results in advanced prostate cancer, especially
mCRPC.
Key words Metastatic castration-resistant prostate cancer, DNA damage response, Prostate-specific
antigen, Androgen-deprived treatment, Homologous recombination deficiency
1 Introduction
Prostate cancer (PCa) is one of the common malignancies in male

adults. In men, prostate cancer ranks second in incidence and fifth
in mortality. It is estimated that there were almost 1.3 million new
cases of prostate cancer and 359,000 associated death worldwide
[1]. Thanks to the widespread application of prostate-specific anti-
gen (PSA) screening, more patients with low-grade and organ-
confined tumors have been detected early. Although the incidence
of advanced prostate cancer (T3 + NxMx) declined in the past
decades, the mortality of advanced prostate cancer patients
remained high [2]. Androgen-deprived treatment (ADT) has
been widely prescribed to patients with advanced prostate cancer
empirically. In the era of precision medicine, many other novel
agents targeting advanced prostate cancer, especially metastatic
castration-resistant prostate cancer (mCRPC), are currently being
evaluated [3–5]. These new options include the CYP17 inhibitor
75
76 Chenyang Xu et al.
abiraterone [6–8], the androgen receptor antagonist enzalutamide

[9, 10], the taxane cabazitaxel [11, 12], and immunotherapy such
as PD-1/PD-L1 antibody [13, 14]. Among all these candidate
therapies, poly-ADP ribose polymerase (PARP) inhibitors targeting
DNA damage response (DDR) pathway has proven improving
survival outcomes in clinical trials. It is estimated that at least ten
percent of advanced PCa patients have germline mutations in DNA
repair genes (DRG), and approximately one fifth of the mCRPC
patients carry aberrations of DRG [15]. Among these DRG,
homologous recombination deficiency (HRD) such as BRCA1/
2 mutant cells display sensitivity to PARP inhibition due to the
synthetic lethality of tumors [16].
In this review, we focus on recent advances in biology and
clinical implication of DDR pathway. Germline and somatic muta-
tion of DDR genes and PARP inhibitors have been widely studied
in preclinical research and clinical trials. We aim to discuss the latest
results in advanced prostate cancer, especially mCRPC.
2 Evidence Acquisition
A literature research for clinical trials and preclinical basic research

studies from January 2014 to December 2019 was conducted in
PubMed, Web of Science, EMBASE (Ovid), American Society of
Clinical Oncology (ASCO) Meeting Summary, Cochrane Database
and ClinicalTrials.gov (NLM). Keywords include “DNA repair,”
“homologous recombination,” “BRCA,” “ATM,” “prostate can-
cer,” “CRPC,” “PARP,” “predictive biomarkers,” “precision
medicine.”
3 Evidence Synthesis
3.1 DNA Damage Defective DNA repair causes replicative immortality, which is a
Response Pathway common hallmark of cancer [17]. There are two major pathways
to repair hazardous DNA double-strand breaks (DSB) in eukary-
3.1.1 Overview
otic cells, homologous recombination (HR), and nonhomologous
end-joining (NHEJ) [18]. Other mechanisms of DDR pathway
include base excision repair (BER), nucleotide excision repair
(NER), and mismatch repair (MMR), which mainly correct
single-strand breaks [19]. HR functions in the S and G2 phases of
the cell cycle, while NHEJ functions by ligating broken DNA ends
throughout the cell cycle [18]. HR played a central role in response
to DSBs repair, and deficiency of HR pathway may confer and
increase the risk of several tumors. Important mediators of HR
pathway include BRCA1, BRCA2, PALB2, ATM, RAD51,
MRE11, CHEK2, and XRCC2/3 [20].
Recent Advances in DNA Repair Pathway and Its Application in Personalized. . . 77
In classic HR pathway, a single-stranded DNA template was

resected from 50 terminal of a DNA break end. The resection of
50 -terminated DNA strand is mediated by RAD51 recombinase and
its ancillary factors. BRCA mediator complex (DSS1-BRCA2-
PALB2-BRCA1-BARD1) function in presynaptic filament forma-
tion, replication fork repair, and interstrand cross-link (ICL)
removal [20–22]. Mutations of these genes resulted in attenuation
of DSB repair capacity of cells.
3.1.2 BRCA1/2 Mutation The BRCA gene mutation was detected in familial ovarian and
breast cancer. The breast cancer type 1 susceptibility protein
(BRCA1) comprises of 1863 amino acid residues. It interacts with
BARD1 constituting a RING heterodimer to repair chromosome
damage with E3 ubiquitin ligase activity. BRCA1-BARD1 is
involved in DNA end resection, RAD51 recruitment, synaptic
complex assembly, and several other crucial steps of genome repair
[23]. In vitro research showed that BRCA1-deficient cells display
susceptibility to DNA damaging agents and chromosomal instabil-
ity [24–26]. The breast cancer type 2 susceptibility protein
(BRCA2) comprises of 3418 amino acid residues. Through inter-
action with RAD51, it demonstrates important function in replica-
tion fork maintenance [27].
In breast cancer, germline BRCA mutations resulted in the
expression of oncogenes [28]. The lifetime risk of breast cancer
increases from 60 to 70% with the presence of gBRCA mutation
[29, 30]. In triple-negative breast cancer (TNBC), germline muta-
tions in BRCA1 or BRCA2 occur in approximately 10% of TNBC
patients [31].
In epithelial ovarian cancer development, loss of function in
either BRCA1 or BRCA2 is an important risk factor. Recent sys-
temic analysis based on over 5000 tumor immunohistochemistry
samples reported that a 47.0 percent of ovarian cancers were asso-
ciated with loss of BRCA1 and 34.5 percent with the loss of
BRCA2 [32].
3.1.3 Ataxia-telangiectasia mutated kinase (ATM) is another extensively

Ataxia-Telangiectasia studied DDR protein. ATM is an important member of phospha-
Mutated Kinase (ATM) tidylinositol 3-kinase-like protein kinase (PIKKs) family [33]. ATM
activates checkpoint and DSB repair pathway by phosphorylation.
Either germline or induced ATM mutant leads to tumorigenesis
and metastasis [34]. According to immunohistochemistry studies
of primary and metastatic tumors including gastric, colon, and
breast cancer, ATM deficiency is common (approximately 20 per-
cent) and sensitizes PARP inhibitor therapy [35, 36]. Selective ATR
inhibitors like AZD6738 are currently in clinical development. The
preclinical data supporting the use of ATR inhibitors in synthetic
lethality with novel targeted agents such as PARP inhibitors
[37, 38].
3.2 DDR Biomarkers Studies of hereditary PCa reveal its relationship with inherited
for PCa Diagnosis DRG mutations. Germline BRCA2 mutations increase the risk of
and Prognosis PCa by 8.6 folds. The prevalence of germline BRCA2 mutations is
estimated to be 5.3% in advanced PCa [39]. The IMPACT study
3.2.1 Germline BRCA also found a higher incidence of PCa in gBRCA1/2 carriers than
Mutation noncarriers [40]. In a large-scale retrospective study including
and Hereditary PCa 61 BRCA2 and 18 BRCA1-mutated PCa patients, 23 percent of
gBRCA1/2 mutation carriers developed metastasis after 5 years of
radical treatment, significantly higher than noncarriers
( p ¼ 0.001).Of note, gBRCA1/2 were also associated with more
aggressive prostate cancer, higher risk of lymph node, and distant
metastasis at diagnosis [41]. Another retrospective case study of
313 patients found that BRCA1/2 and ATM mutation carrier rate
was significantly higher in lethal PCa patients (6.07%) than loca-
lized PCa patients (1.44%, p ¼ 0.0007). The mutation status of
gBRCA was associated with early age at death [42]. A prospective
cohort study of germline DDR gene mutation in mCRPC enrolled
419 patients and identified 107 genes in 68 carriers. Cancer-specific
survival (CSS) was only a half in gBRCA2 carriers (17.4 months
versus 33.2 months, p ¼ 0.027). Multivariable analysis identified
gBRCA2 mutations as an independent prognostic factor for CSS
(adjusted hazard ratio ¼ 2.11, p ¼ 0.033) [43]. Germline muta-
tions were reported to correlate with higher grade of PCa. Recent
study based on two cohorts included 1211 PCa patients character-
ized relation between PCa grade reclassification and germline
BRCA1/2 and ATM mutations. Eleven of 26 patients with muta-
tions experienced grade reclassification (adjusted hazard
ratio ¼ 2.74, p ¼ 0.01). Carrier rates for gBRCA2 were significantly
higher in the reclassified cohort (2.1%), especially in those reclassi-
fied from Gleason score 3 + 3 at diagnosis to Gleason score 4 + 3
(4.1%) [44].
3.2.2 Somatic Mutations Somatic mutations in DRG are common in sporadic prostate can-
of DRG and PCa cer. Approximately 23% of mCRPC patients carry somatic DDR
aberrations. BRCA1/2 and ATM gene mutations are the main
somatic aberrations with higher frequency in mCRPC compared
with primary PCa [45].
For PARP inhibitor therapy, ideal clinical biomarkers that not
only predict HRD in DRG wild-type tumors but also refine DRG
mutant population to account for phenotypic variants and rever-
sions remained controversial. Germline BRCA mutation has been
the most suitable biomarker for PARP inhibitor response although
the problems of variants of uncertain significance still exists. Other
PCa-associated mutations include BARD1, PALB2, RAD51B/C/
D, ATM, MLH1, MSH2, FANCA, and so on. These are potential
biomarkers that correlate with clinical response to PARP inhibitor
therapy. Prospective genomic profiling in PCa patients revealed that
somatic alterations of DDR genes arose early in mCRPC, including

ATM, RAD51C, and FANCA [46]. In metastatic PCa, a panel of
20 DRG next-generation sequencing found 11.8% of the patients
carried these mutations [39]. Several other studies reported the
prevalence of DRG defects in metastatic PCa within the range from
8.8% to 22.7% [45, 47, 48]. In patients with extremely aggressive
localized PCa, two somatic mutations in PRKDC which encodes
the catalytic subunit of the DNA-dependent protein kinase
involved in DNA double-strand break repair and recombination
were reported [49].
3.3 PARP Inhibitors Poly-ADP ribose polymerases (PARPs) are a family of enzymes
in mCRPC function in detecting and harboring other DNA repair proteins to
single-strand breaks (SSB). PARP1 and PARP2 repair DNA dam-
age by catalyzing the cleavage of nicotinomide adenine
dinucleotide-plus into nicotinamide and ADP-ribosyl action and
attaching ADP-ribose to form a complex [50]. In HR pathway,
PARP acts as a sensor and recruiting protein like MRE11 to start
DNA resection and replication. Theoretically, PARP inhibitor can
halt ligation of SSB before replication. Thus, accumulation of DSBs
due to defective BRCA1/2, ATM, and other DDR genes would
lead to loss of cell viability. Furthermore, when DDR genes are
inactivated, alternative repair pathways act in compensation, BER
DNA repair for instance. These alternative repair pathways could
also be inhibited through PARP [51–55]. Several PARP inhibitors
has been tested in various solid tumors, including olaparib, ruca-
parib, niraparib, veliparib, and talazoparib.
3.3.1 Olaparib Olaparib has gained U.S. Food and Drug Administration (FDA) in
treatment of advanced epithelial ovarian cancer and Her-2-negative
metastatic breast cancer with gBRCA mutations. It was also the first
PARP inhibitor that received FDA breakthrough therapy in 2016
to treat mCRPC patients carrying BRCA1/2 or ATM
mutations [56].
In a phase II clinical trial of olaparib for multiple tumors
(NCT01078662), eight mCRPC patients (one BRCA1 mutant
carrier and seven BRCA2 mutant carriers) were enrolled. The
median progression-free survival (PFS) for all eight patients was
7.2 month. Two patients responded for over 1 year [57]. Another
clinical trial of olaparib monotherapy enrolled 50 mCRPC patients.
Next-generation sequencing identified 16 patients carried DRG
deficiency. The response rate of these patients is higher than the
overall response rate (88% vs 33%) [58]. A multicenter, open-label,
randomized phase II trial (TOPARP, NCT01682772) recruited
711 mCRPC patients in the UK. Taken together, 98 patients had
DDR gene aberrations and were treated with olaparib. In 400 mg
cohort, the preliminary result showed a composite response rate of
54.3% (25 of 46; 95%CI 39.0–69.1). In 300 mg cohort, composite
response was confirmed in 39.1% (18 of 46; 95%CI 25.1–54.6)

[59]. As a monotherapy, olaparib is well tolerated. Major adverse
effects of olaparib were anemia, nausea, fatigue, and thrombocyto-
penia [58, 60].
Investigation of olaparib monotherapy helped understanding
of its efficacy, safety, and degree of PARP inhibition. More clinical
trials have been established to look at the response of olaparib plus
other cytotoxic drug. Combination of olaparib with abiraterone in
mCRPC patients were evaluated in a randomized, double-blind,
placebo-controlled phase II trial (NCT01972217) [61]. Altogether
142 patients were equally divided into two groups receiving ola-
parib plus abiraterone and abiraterone alone. The PFS was signifi-
cantly better in the experimental group (13.8 months versus
8.2 months, p ¼ 0.034). Meanwhile, olaparib combining with
immunotherapy also demonstrated a potential synergic effect
[62, 63]. Olaparib and pembrolizumab (anti-PD-L1 antibody)
combination therapy (Keynote-365, NCT02861573) was reported
to achieve a median overall survival of 14 months and a median
radiographic progression-free survival (rPFS) of 5 months. The
disease control rate was 32% (12 out of 41). Six patients (15%)
got composite response, none of whom had DDR deficiency
[64]. Similar efficacy was identified in durvalumab plus olaparib
combination therapy (KEYLINK-010, NCT03834519). Median
rPFS for all patients is 16.1 months (95% CI: 4.5–16.1 months)
with a 12-month rPFS of 51.5% (95% CI: 25.7–72.3%) [65].
Two ongoing phase II and III clinical trials include mCRPC
patients and treat them with olaparib monotherapy. Another two
studies focus on combination therapy of olaparib and abiraterone.
A phase II study (BRCAAway, NCT03012321) include
DDR-mutated mCRPC patients and randomly assigned to three
arms of olaparib alone, abiraterone alone, and combination therapy.
Multicenter, randomized clinical trial (PROpel, NCT03732820) is
a phase III trial aiming to recruit over 700 mCRPC patients and
compare olaparib versus placebo in combination with abiraterone.
Several other current trials examine the efficacy and safety of ola-
parib with radiation, immunotherapy, and testosterone (Table 1).
3.3.2 Rucaparib Rucaparib was first approved by FDA for BRCA-mutated advanced
ovarian cancer [66]. Its indication was then expended to mainte-
nance treatment of recurrent epithelial ovarian, fallopian tube, or
primary peritoneal cancer.
Rucaparib was tested in a phase II study (TRITON2,
NCT02952534). Fifty-two mCRPC patients with DDR mutation
that was previously treated with ADT or taxane-based chemother-
apy were recruited. The primary outcome confirmed objective
response rate and PSA response (over 50% decrease) rate. The latest
update demonstrated that 23 patients (51.1%) had PSA response
Table 1
Current clinical trials of Olaparib for castration-resistant prostate cancer
Combined Primary
NCT number intervention Cancer type Study design Phase outcome
NCT03434158 None Metastatic Multicenter, single 2 rPFS
castration- arm
resistant prostate
cancer after over
6 cycles of
docetaxel
NCT02987543 None Metastatic Multicenter, 3 rPFS
castration- randomized,
resistant prostate parallel control
cancer and treated with ADT
evidence of HRD
NCT03012321 Abiraterone Metastatic Multicenter, 2 PFS
castration- randomized
resistant prostate
cancer with
BRCA1/2 or
ATM mutation
NCT03732820 Abiraterone metastatic Multicenter, 3 rPFS
resistant prostate double-blinded,
cancer (mCRPC) placebo control
with no prior
cytotoxic
chemotherapy or
new hormonal
agents (NHAs)
NCT03317392 Radium Ra Castration-resistant Multicenter, 1 and MTD and
223 Dichloride prostate cancer randomized, 2 PFS
metastatic to the parallel control
bone treated with
olaparib or
radiation
NCT02893917 Cediranib Metastatic Multicenter, 2 rPFS
cancer treated with
olaparib
NCT03874884 177Lutetium- Metastatic Multicenter, single 1 DLTs and
Prostate- castration- arm, dose- MTD
Specific resistant prostate escalation
Membrane cancer
Antigen
(177 Lu-PSMA)
(continued)
ALGrawany
Table 1
(continued)
Combined Primary
NCT02861573 Pembrolizumab Metastatic Multicenter, 3 ORR and
(MK-3475) castration- non-randomized, RR-PSA
cancer treated with
pembrolizumab
plus enzalutamide
or docetaxel
NCT03834519 Pembrolizumab Metastatic Multicenter, 3 OS and PFS
(MK-3475) castration- randomized,
cancer progressed treated with
on an androgen abiraterone or
receptor signaling enzalutamide
inhibitor
NCT02484404 Anti-PD-L1 Metastatic Multicenter, 1 and ORR
antibody castration- non-randomized, 2
(MED14736) resistant prostate parallel treated
cancer, advanced with MED14736
ovarian/breast/ plus cediranib
colorectal cancer
NCT03787680 AZD6738 Metastatic Multicenter single 2 CR or PR in
castration- arm DNA
resistant prostate repair
cancer proficient
patients
NCT03516812 Testosterone Castration-resistant Single-center, single 2 RR-PSA
prostate cancer arm
rPFS radiographic progress-free survival, MTD maximum tolerated dose, DLT dose limiting toxicities, ORR objective
response rate, RR-PSA prostate-specific antigen response rate, CR complete response, PR partial response
and 11 patients (44.0%) had radiographic response [67]. Grades

3–4 anemia and thrombocytopenia were reported, suggesting that
toxicity of rucaparib was higher than olaparib. Currently, two phase
II studies aim at evaluating efficacy of rucaparib monotherapy in
mCRPC patients with HRD, and there are three clinical trials
exploring the potential survival benefit of rucaparib in combination
with nivolumab (anti-PD1 antibody) or chemotherapy (Table 2).
3.3.3 Niraparib Niraparib was the third PARP inhibitor that gained FDA approval
in 2017. It was the first PARP inhibitor to receive full approval by
the U.S. Food and Drug Administration (FDA) for the mainte-
nance treatment of recurrent ovarian cancer, regardless of a
patient’s germline or somatic mutational status [68].
Table 2
Current clinical trials of Rucaparib for castration-resistant prostate cancer
Combined Primary
NCT02952534 None Metastatic castration- Multicenter, single arm 2 ORR and
resistant prostate RR-PSA
cancer and evidence
of HRD
NCT02975934 None Metastatic castration- Multicenter, randomized, 3 rPFS
resistant prostate parallel control treated
cancer and evidence with abiraterone or
of HRD docetaxel
NCT03572478 Nivolumab Metastatic castration- Single-center, 1 and DLT
resistant prostate randomized, parallel 2
cancer and control treated with
metastatic/ monotherapy
recurrent
endometrial cancer
NCT03338790 Nivolumab Metastatic castration- Single-center, 2 ORR and
resistant prostate non-randomized, RR-PSA
cancer parallel control treated
with Nivolumab plus
arbiraterone or
docetaxel
NCT03442556 Docetaxel or Metastatic castration- Single-center, single arm 2 rPFS
carboplatin resistant prostate
cancer with
BRCA1/2 or ATM
mutation
ORR objective response rate, RR-PSA prostate-specific antigen response rate, rPFS radiographic progress-free survival,
DLT dose limiting toxicity
Two phase I dose-escalation studies of niraparib monotherapy

for advanced solid tumors have reported promising results. Among
the 21 mCRPC patients enrolled from the UK and the USA,
9 patients had stable disease for a median duration of 254 days
and a decrease in circulating tumor cells were found in 3 patients
(NCT00749502) [69]. A phase II study of niraparib (GALAHAD,
NCT02854436) evaluated in 123 mCRPC patients with patho-
genic mutations of BRCA1/2, ATM, FANCA, PALB2, CHEK2,
BRIP1, or HDAC2). Composite and objective response rates were
65% and 38%. In comparison, in 16 patients without BRCA1/
2 mutations, objective response rate was 11% [70]. Toxicities of
niraparib were limited to low-grade adverse effect. Another phase II
study is now open to assess the safety of niraparib in patients with
mCRPC. Apart from that, three undergoing clinical trials are
Table 3
Current clinical trials of Niraparib for castration-resistant prostate cancer
Combined Primary
NCT02854436 None Metastatic castration-resistant Multicenter, single 2 ORR
prostate cancer arm
NCT03748641 Abiraterone Metastatic prostate cancer Multicenter, 3 rPFS
acetate randomized,
placebo control
NCT03076203 Radium Ra Hormone-resistant prostate Multicenter, single 1 MTD
223 cancer metastatic to the arm
dichloride bone
NCT03431350 Abiraterone Metastatic castration-resistant Multicenter, 1 and ORR
acetate prostate cancer non-randomized 2
ORR objective response rate, rPFS radiographic progress-free survival, MTD maximum tolerated dose
recruiting mCRPC patients to examine the effect of combination

therapies, including niraparib plus abiraterone acetate and niraparib
plus Radium-223 (Table 3).
3.3.4 Veliparib Veliparib is another potent inhibitor of PARP1 and PARP2 that
demonstrated promising anti-PCa activity. Veliparib does not yet
have an FDA-approved label; nevertheless, there are currently
promising results available in preclinical and early clinical settings
[71–74].
Pilot study of 26 mCRPC patients treated with veliparib plus
temozolomide (TMZ) reported a median OS of 9.1 months and a
median PFS of 2.9 months [75]. Seventy mCRPC patients with
BRCA2-mutation were included in a dose-escalation phase I trial,
and three patients were recommended with phase II study of veli-
parib. The objective response rate was reported to be 37% and 66%
in phase I and phase II trial [76]. Another phase II randomized trial
of veliparib plus abiraterone for mCRPC (NCT01576172) exam-
ined the antitumor efficacy of this agent in combination therapy.
The PFS was slightly better in combination therapy compared with
abiraterone alone, but the result is of no statistical significance
(11 months versus 10.1 months, p ¼ 0.95) [77]. Ongoing phase
II and phase III trials are testing veliparib with cytotoxic chemo-
therapy in advanced solid tumors.
3.3.5 Talazoparib Talazoparib was more potent than the PARP inhibitors mentioned
above [78]. It gained FDA approval for gBRCA-mutated HER2-
negative locally advanced or metastatic breast cancer [79].
Started in 2018, TALAPRO-1 (NCT03148795) was the first
phase II randomized clinical trial involving talazoparib in post-
taxane mCRPC patients with DDR defects [80]. An initial safety
Table 4
Current clinical trials of Talazoparib for castration-resistant prostate cancer
Combined Primary
NCT03148795 None DDR-positive Multicenter, 2 ORR
metastatic single arm
castration-
resistant
prostate cancer
NCT03395197 Enzalutamide Metastatic Multicenter, 3 rPFS
resistant double-blind
prostate cancer placebo control
treated with
enzalutamide
NCT04052204 Avelumab, Metastatic Multicenter, 2 DLT,
bempegaldesleukin castration- non-randomized, ORR,
(NKTR-214) resistant parallel control RR-PSA
prostate cancer treated with
and locally avelumab plus
advanced bempegaldesleukin
squamous cell
carcinoma of the
head and neck
NCT03330405 Avelumab Locally advanced Multicenter, single- 1 and DLT,
or metastatic arm 2 ORR
solid tumorsa
NCT04019327 Temozolomide Castration- Single-center, single 1 and ORR
resistant arm 2
Prostate Cancer
ORR objective response rate, rPFS radiographic progress-free survival, DLT dose limiting toxicity, RR-PSA prostate-
specific antigen response rate
a
Including non-small cell lung cancer (NSCLC), triple-negative breast cancer (TNBC), hormone receptor-positive (HR
+) breast cancer, recurrent platinum-sensitive ovarian cancer, urothelial cancer (UC), and castration-resistant prostate
cancer (CRPC)
and efficacy analysis will be performed on 20 patients after over

8 weeks of treatment. Clinical efficacy and safety of talazoparib
combined with other therapies is yet to be determined in ongoing
clinical studies (Table 4).
3.3.6 Other PARP Other preclinical PARP inhibitors for patients with advanced solid
Inhibitors in Development tumors include pamiparib [81] (NCT03712930; NCT03150810),
IMP4297 (NCT03507543), HWH340 (NCT03415659),
SC10914 (NCT02940132), NMS-03305293 (NCT04182516),
LT-626 [82], etc. The efficacy and toxicity of these drugs remain
to be discovered.
4 Conclusion
In summary, more and more evidences reveal that DNA damage

response pathway is critical in carcinogenesis of prostate cancer.
DRG mutations such as BRCA1/2 and ATM are related with the
susceptibility of prostate cancer and tumor aggressiveness. Fortu-
nately, PARP inhibitors targeting these gene aberrations demon-
strate its efficacy in advanced prostate cancer to prolong
progression-free survival, especially mCRPC. Although yet not
available, olaparib offers an exciting new opportunity that opens
the gate of precision medicine for patients with mutation profile. In
future, large-scale clinical trials will test olaparib either as mono-
therapy or in combination with other treatments. Evaluation of
rucaparib (TRITON3, NCT02975934), niraparib
(NCT02854436), and several other PARP inhibitors for mCRPC
patients with DRG mutations is also on its way. In the era of
precision medicine, targeted mCRPC patients will benefit from
these novel therapeutic approaches.
References
1. Bray F, Ferlay J, Soerjomataram I et al (2018) abiraterone acetate plus prednisone in men

Global cancer statistics 2018: GLOBOCAN with chemotherapy-naı̈ve metastatic
estimates of incidence and mortality worldwide castration-resistant prostate cancer: stratified
for 36 cancers in 185 countries. CA Cancer J analysis based on pain, prostate-specific anti-
Clin 68(6):394–424 gen, and gleason score. Eur Urol 74(1):17–23
2. Fitzmaurice C, Abate D, Abbasi N et al (2019) 8. Fizazi K, Tran N, Fein L et al (2017) Abirater-
Global, Regional, and National Cancer Inci- one plus prednisone in metastatic, castration-
dence, Mortality, Years of Life Lost, Years sensitive prostate cancer. N Engl J Med 377
Lived With Disability, and Disability-Adjusted (4):352–360
Life-Years for 29 Cancer Groups, 1990 to 9. Saito K, Fujii Y (2018) Antitumor activity and
2017: A Systematic Analysis for the Global safety of enzalutamide after abiraterone ace-
Burden of Disease Study. JAMA Oncol 5 tate: seeking the optimal treatment sequence
(12):1749–1768 for castration-resistant prostate cancer patients.
3. Parker C, Lewington V, Shore N et al (2018) Eur Urol 74(1):46–47
Targeted alpha therapy, an emerging class of 10. Anker MS, Lehmann LH, Anker SD (2018)
cancer agents: a review. JAMA Oncol 4 Enzalutamide in castration-resistant prostate
(12):1765–1772 cancer. N Engl J Med 379(14):1380
4. Ritch CR, Cookson MS (2016) Advances in 11. Antonarakis ES, Tagawa ST, Galletti G et al
the management of castration resistant prostate (2017) Randomized, noncomparative, phase
cancer. BMJ (Clinical Research ed) 355:i4405 II trial of early switch from docetaxel to caba-
5. Nuhn P, De Bono JS, Fizazi K et al (2019) zitaxel or vice versa, with integrated biomarker
Update on systemic prostate cancer therapies: analysis, in men with chemotherapy-naı̈ve,
management of metastatic castration-resistant metastatic, castration-resistant prostate cancer.
prostate cancer in the era of precision oncology. J Clin Oncol 35(28):3181–3188
Eur Urol 75(1):88–99 12. Rouyer M, Oudard S, Joly F et al (2019) Over-
6. Almassi N, Reichard C, Li J et al (2018) all and progression-free survival with cabazi-
HSD3B1 and response to a nonsteroidal taxel in metastatic castration-resistant prostate
CYP17A1 inhibitor in castration-resistant cancer in routine clinical practice: the FUJI
prostate cancer. JAMA Oncol 4(4):554–557 cohort. Br J Cancer 121(12):1001–1008
7. Miller K, Carles J, Gschwend JE et al (2018) 13. Hansen AR, Massard C, Ott PA et al (2018)
The Phase 3 COU-AA-302 Study of Pembrolizumab for advanced prostate
adenocarcinoma: findings of the KEYNOTE- RAD51-mediated recombination. Nature 467

028 study. Ann Oncol 29(8):1807–1813 (7316):678–683
14. Gulley JL, Madan RA, Pachynski R et al (2017) 28. Kuchenbaecker KB, Mcguffog L, Barrowdale
Role of antigen spread and distinctive charac- D et al (2017) Evaluation of polygenic risk
teristics of immunotherapy in cancer treat- scores for breast and ovarian cancer risk predic-
ment. J Natl Cancer Inst 109(4):djw261 tion in BRCA1 and BRCA2 mutation carriers.
15. Lang SH, Swift SL, White H et al (2019) A J Natl Cancer Inst 109(7):djw302
systematic review of the prevalence of DNA 29. Antoniou A, Pharoah PD, Narod S et al (2003)
damage response gene mutations in prostate Average risks of breast and ovarian cancer asso-
cancer. Int J Oncol 55(3):597–616 ciated with BRCA1 or BRCA2 mutations
16. Ashworth A, Lord CJ (2018) Synthetic lethal detected in case Series unselected for family
therapies for cancer: what’s next after PARP history: a combined analysis of 22 studies. Am
inhibitors? Nat Rev Clin Oncol 15(9):564–576 J Hum Genet 72(5):1117–1130
17. Hanahan D, Weinberg RA (2011) Hallmarks 30. Mavaddat N, Peock S, Frost D et al (2013)
of cancer: the next generation. Cell 144 Cancer risks for BRCA1 and BRCA2 mutation
(5):646–674 carriers: results from prospective analysis of
18. Chang HHY, Pannunzio NR, Adachi N et al EMBRACE. J Natl Cancer Inst 105
(2017) Non-homologous DNA end joining (11):812–822
and alternative pathways to double-strand 31. Cancer Genome Atlas Network (2012) Com-
break repair. Nat Rev Mol Cell Biol 18 prehensive molecular portraits of human breast
(8):495–506 tumours. Nature 490(7418):61–70
19. Caldecott KW (2008) Single-strand break 32. Teixeira LA, Candido Dos Reis FJ (2020)
repair and genetic disease. Nat Rev Genet 9 Immunohistochemistry for the detection of
(8):619–631 BRCA1 and BRCA2 proteins in patients with
20. Her J, Bunting SF (2018) How cells ensure ovarian cancer: a systematic review. J Clin
correct repair of DNA double-strand breaks. J Pathol 73(4):191–196
Biol Chem 293(27):10502–10511 33. Li M, Yu X (2013) Function of BRCA1 in the
21. Schlacher K, Christ N, Siaud N et al (2011) DNA damage response is mediated by
Double-strand break repair-independent role ADP-ribosylation. Cancer Cell 23(5):693–704
for BRCA2 in blocking stalled replication fork 34. Rezaeian AH, Li CF, Wu CY et al (2017) A
degradation by MRE11. Cell 145(4):529–542 hypoxia-responsive TRAF6-ATM-H2AX sig-
22. Clauson C, Sch€arer OD, Niedernhofer L nalling axis promotes HIF1α activation,
(2013) Advances in understanding the com- tumorigenesis and metastasis. Nat Cell Biol
plex mechanisms of DNA interstrand cross- 19(1):38–51
link repair. Cold Spring Harb Perspect Biol 5 35. Miller RM, Nworu C, Mckee L et al (2020)
(10):a012732 Development of an immunohistochemical
23. Zhao W, Steinfeld JB, Liang F et al (2017) assay to detect the ataxia-telangiectasia
BRCA1-BARD1 promotes RAD51-mediated mutated (ATM) protein in gastric carcinoma.
homologous DNA pairing. Nature 550 Appl Immunohistochem Mol Morphol 28
(7676):360–365 (4):303–310
24. Rottenberg S, Jaspers JE, Kersbergen A et al 36. Jette NR, Radhamani S, Arthur G et al (2019)
(2008) High sensitivity of BRCA1-deficient Combined poly-ADP ribose polymerase and
mammary tumors to the PARP inhibitor ataxia-telangiectasia mutated/Rad3-related
AZD2281 alone and in combination with plat- inhibition targets ataxia-telangiectasia
inum drugs. Proc Natl Acad Sci U S A 105 mutated-deficient lung cancer cells. Br J Can-
(44):17079–17084 cer 121(7):600–610
25. Moynahan ME, Cui TY, Jasin M (2001) 37. Foote KM, Nissink JWM, Mcguire T et al
Homology-directed dna repair, mitomycin-c (2018) Discovery and characterization of
resistance, and chromosome stability is AZD6738, a potent inhibitor of ataxia telangi-
restored with correction of a Brca1 mutation. ectasia mutated and Rad3 related (ATR) Kinase
Cancer Res 61(12):4842–4850 with application as an anticancer agent. J Med
Chem 61(22):9889–9907
26. Alli E, Sharma VB, Hartman AR et al (2011)
Enhanced sensitivity to cisplatin and gemcita- 38. Bradbury A, Hall S, Curtin N et al (2019)
bine in Brca1-deficient murine mammary epi- Targeting ATR as cancer therapy: a new era
thelial cells. BMC Pharmacol 11:7 for synthetic lethality and synergistic combina-
tions? Pharmacol Ther 207:107450
27. Jensen RB, Carreira A, Kowalczykowski SC
(2010) Purified human BRCA2 stimulates
39. Pritchard CC, Mateo J, Walsh MF et al (2016) assembly or stability of XRCC1 nuclear foci at
Inherited DNA-repair gene mutations in men sites of oxidative DNA damage. Nucleic Acids
with metastatic prostate cancer. N Engl J Med Res 31(19):5526–5533
375(5):443–453 52. Luo X, Kraus WL (2012) On PAR with PARP:
40. Bancroft EK, Page EC, Castro E et al (2014) cellular stress signaling through poly
Targeted prostate cancer screening in BRCA1 (ADP-ribose) and PARP-1. Genes Dev 26
and BRCA2 mutation carriers: results from the (5):417–432
initial screening round of the IMPACT study. 53. Hegan DC, Lu Y, Stachelek GC et al (2010)
Eur Urol 66(3):489–499 Inhibition of poly(ADP-ribose) polymerase
41. Castro E, Goh C, Olmos D et al (2013) Germ- down-regulates BRCA1 and RAD51 in a path-
line BRCA mutations are associated with way mediated by E2F4 and p130. Proc Natl
higher risk of nodal involvement, distant Acad Sci U S A 107(5):2201–2206
metastasis, and poor survival outcomes in pros- 54. Schiewer MJ, Knudsen KE (2014) Transcrip-
tate cancer. J Clin Oncol 31(14):1748–1757 tional roles of PARP1 in cancer. Mol Cancer
42. Na R, Zheng SL, Han M et al (2017) Germline Res 12(8):1069–1080
mutations in ATM and BRCA1/2 distinguish 55. Langelier MF, Eisemann T, Riccio AA et al
risk for lethal and indolent prostate cancer and (2018) PARP family enzymes: regulation and
are associated with early age at death. Eur Urol catalysis of the poly(ADP-ribose) posttransla-
71(5):740–747 tional modification. Curr Opin Struct Biol
43. Castro E, Romero-Laorden N, Del Pozo A 53:187–198
et al (2019) PROREPAIR-B: a prospective 56. Helleday T (2016) PARP inhibitor receives
cohort study of the impact of germline DNA FDA breakthrough therapy designation in cas-
repair mutations on the outcomes of patients tration resistant prostate cancer: beyond germ-
with metastatic castration-resistant prostate line BRCA mutations. Ann Oncol 27
cancer. J Clin Oncol 37(6):490–503 (5):755–757
44. Carter H, Helfand B, Mamawala M et al 57. Kaufman B, Ronnie S-F, Schmutzler Rita K
(2019) Germline mutations in ATM and et al (2015) Olaparib monotherapy in patients
BRCA1/2 are associated with grade reclassifi- with advanced cancer and a germline BRCA1/
cation in men on active surveillance for prostate 2 mutation. J Clin Oncol 33(3):244–250
cancer. Eur Urol 75(5):743–749 58. Mateo J, Carreira S, Sandhu S et al (2015)
45. Robinson D, Van Allen EM, Wu YM et al DNA-repair defects and olaparib in metastatic
(2015) Integrative clinical genomics of prostate cancer. N Engl J Med 373
advanced prostate cancer. Cell 161 (18):1697–1708
(5):1215–1228 59. Mateo J, Porta N, Bianchini D et al (2019)
46. Abida W, Armenia J, Gopalan A et al (2017) Olaparib in patients with metastatic
Prospective genomic profiling of prostate can- castration-resistant prostate cancer with DNA
cer across disease states reveals germline and repair gene aberrations (TOPARP-B): a multi-
somatic alterations that may affect clinical deci- centre, open-label, randomised, phase 2 trial.
sion making. JCO Precis Oncol 2017:10 Lancet Oncol 21(1):162–174
47. Hart SN, Ellingson MS, Schahl K et al (2016) 60. Lafargue CJ, Molin GZD, Sood AK et al
Determining the frequency of pathogenic (2019) Exploring and comparing adverse
germline variants from exome sequencing in events between PARP inhibitors. Lancet
patients with castrate-resistant prostate cancer. Oncol 20(1):e15–e28
BMJ Open 6(4):e010332 61. Clarke N, Wiechno P, Alekseev B et al (2018)
48. Beltran H, Yelensky R, Frampton GM et al Olaparib combined with abiraterone in patients
(2013) Targeted next-generation sequencing with metastatic castration-resistant prostate
of advanced prostate cancer identifies potential cancer: a randomised, double-blind, placebo-
therapeutic targets and disease heterogeneity. controlled, phase 2 trial. Lancet Oncol 19
Eur Urol 63(5):920–926 (7):975–986
49. Grasso CS, Wu YM, Robinson DR et al (2012) 62. Heyman B, Jamieson C (2019) To PARP or
The mutational landscape of lethal castration- not to PARP?-Toward sensitizing acute mye-
resistant prostate cancer. Nature 487 loid leukemia stem cells to immunotherapy.
(7406):239–243 EMBO J 38(21):e103479
50. Wl K (2015) PARPs and ADP-ribosylation: 63. Césaire M, Thariat J, Candéias SM et al (2018)
50 years . . . and counting. Mol Cell 58 Combining PARP inhibition, radiation, and
(6):902–910 immunotherapy: a possible strategy to improve
51. El-Khamisy SF, Masutani M, Suzuki H et al the treatment of cancer? Int J Mol Sci 19
(2003) A requirement for PARP-1 for the (12):3793
64. Evan Y, Yu MC, Margitta R et al (2019) placebo for extensive-stage small-cell lung can-
Keynote-365 cohort a: pembrolizumab (pem- cer: ECOG-ACRIN 2511 study. J Clin Oncol
bro) plus olaparib in docetaxel-pretreated 37(3):222–229
patients (pts) with metastatic castrate-resistant 74. Loibl S, O’shaughnessy J, Untch M et al
prostate cancer (mCRPC). J Clin Oncol 37 (2018) Addition of the PARP inhibitor veli-
(7_suppl):145 parib plus carboplatin or carboplatin alone to
65. Karzai F, Vanderweele D, Madan RA et al standard neoadjuvant chemotherapy in triple-
(2018) Activity of durvalumab plus olaparib negative breast cancer (BrighTNess): a rando-
in metastatic castration-resistant prostate can- mised, phase 3 trial. Lancet Oncol 19
cer in men with and without DNA damage (4):497–509
repair mutations. J Immunother Cancer 6 75. Hussain M, Anthony CM, Slovin Susan F et al
(1):141 (2012) Pilot study of veliparib (ABT-888) with
66. Sabatucci I, Maltese G, Lepori S et al (2018) temozolomide (TMZ) in patients (pts) with
Rucaparib: a new treatment option for ovarian metastatic castration-resistant prostate cancer
cancer. Expert Opin Pharmacother 19 (mCRPC). J Clin Oncol 30(5_suppl):224–224
(7):765–771 76. Pahuja S, Joseph AL, Prakash BC et al (2015)
67. Abida W, Bryce AH, Vogelzang NJ et al (2019) Preliminary activity of veliparib (V) in BRCA2-
Genomic characteristics of deleterious BRCA1 mutated metastatic castration-resistant pros-
and BRCA2 alterations and associations with tate cancer (mCRPC). J Clin Oncol 33
baseline clinical factors in patients with meta- (7_suppl):170
static castration-resistant prostate cancer 77. Hussain M, Daignault-Newton S, Twardowski
(mCRPC) enrolled in TRITON2. J Clin PW et al (2018) Targeting androgen receptor
Oncol 37(15_suppl):5031 and DNA repair in metastatic castration-
68. Ison G, Howie LJ, Amiri-Kordestani L et al resistant prostate cancer: results from NCI
(2018) FDA Approval summary: niraparib for 9012. J Clin Oncol 36(10):991–999
the maintenance treatment of patients with 78. Murai J, Huang SY, Das BB et al (2012)
recurrent ovarian cancer in response to Trapping of PARP1 and PARP2 by clinical
platinum-based chemotherapy. Clin Cancer PARP inhibitors. Cancer Res 72
Res 24(17):4066–4071 (21):5588–5599
69. Sandhu SK, Schelman WR, Wilding G et al 79. Hoy SM (2018) Talazoparib: first global
(2013) The poly(ADP-ribose) polymerase approval. Drugs 78(18):1939–1946
inhibitor niraparib (MK4827) in BRCA muta- 80. De Bono JS, Higano Celestia S, Fred S et al
tion carriers and patients with sporadic cancer: (2019) TALAPRO-1: an open-label, response
a phase 1 dose-escalation trial. Lancet Oncol 14 rate phase II study of talazoparib (TALA) in
(9):882–892 men with DNA damage repair (DDR) defects
70. Smith MR, Kaur SS, Kevin KW et al (2019) and metastatic castration-resistant prostate
Phase II study of niraparib in patients with cancer (mCRPC) who previously received
metastatic castration-resistant prostate cancer taxane-based chemotherapy (CT) and pro-
(mCRPC) and biallelic DNA-repair gene gressed on greater than or equal to one novel
defects (DRD): preliminary results of GALA- hormonal therapy (NHT). J Clin Oncol 37
HAD. J Clin Oncol 37(7_suppl):202 (7_suppl):TPS342–TPS342
71. Boussios S, Karihtala P, Moschetta M et al 81. Friedlander M, Meniawy T, Markman B et al
(2020) Veliparib in ovarian cancer: a new syn- (2019) Pamiparib in combination with tisleli-
thetically lethal therapeutic approach. Invest zumab in patients with advanced solid
New Drugs 38(1):181–193 tumours: results from the dose-escalation
72. Tuli R, Shiao SL, Nissen N et al (2019) A phase stage of a multicentre, open-label, phase 1a/b
1 study of veliparib, a PARP-1/2 inhibitor, trial. Lancet Oncol 20(9):1306–1315
with gemcitabine and radiotherapy in locally 82. Mcpherson LA, Shen Y, Ford JM (2014) Poly
advanced pancreatic cancer. EBioMedicine (ADP-ribose) polymerase inhibitor LT-626:
40:375–381 Sensitivity correlates with MRE11 mutations
73. Owonikoko TK, Dahlberg SE, Sica GL et al and synergizes with platinums and irinotecan
(2019) Randomized phase II trial of cisplatin in colorectal cancer cells. Cancer Lett 343
and etoposide in combination with veliparib or (2):217–223
Chapter 8
Methylation in Lung Cancer: A Brief Review

Chang Gu and Chang Chen
Abstract
Despite the introduction of low-dose computed tomography (LDCT) and implementation of lung cancer
screening programs, lung cancer still maintains the leading cause of cancer-specific death all around the
world in terms of morbidity and mortality. Many studies demonstrated that the methylation status of
selected genes may act as prognostic biomarkers for lung cancer patients. Recently, the development of
high-throughput sequencing for methylation would help researchers better understand the role of methyl-
ation in the tumorigenicity or metastasis of lung cancer. This chapter reviews the progress of DNA
methylation in lung cancer.
Key words Lung cancer, Methylation, Methylation-specific polymerase chain reaction, Circulating
tumor DNA, Epigenetics
1 Introduction
Despite the introduction of low-dose computed tomography

(LDCT) and implementation of lung cancer screening programs,
lung cancer still maintains the leading cause of cancer-specific death
all around the world in terms of morbidity and mortality [1, 2]. The
detection of epigenetic changes in lung cancer started much later
than that of other genetic abnormalities. In lung cancer patients,
the methylation status of many genes, as well as the methylation in
blood or even in exfoliative airway epitheliums were investigated.
Many studies demonstrated that the methylation status of selected
genes may act as prognostic biomarkers for lung cancer patients.
Recently, the development of high-throughput sequencing for
methylation would help researchers better understand the role of
methylation in the tumorigenicity or metastasis of lung cancer. This
chapter reviews the progress of DNA methylation in lung cancer.
91
92 Chang Gu and Chang Chen
2 DNA Methylation and Gene Regulation
The regulation of transcriptional activity of genes is mainly owing

to epigenetic changes. Many chemical modifications of histone
proteins, including methylation, acetylation, phosphorylation, ubi-
quitination and ADP-ribosylation, as well as DNA methylation, act
as pivotal regulators by changing the chromatinic structure or
affecting the binding of transcription factors to DNA (resulting in
gene activation or silencing) [3, 4]. Furthermore, there is increas-
ing evidence that gene silencing is the result of DNA methylation
and chemical modifications of histone proteins [5]. Ikegami et al.
[6] demonstrated that DNA methylation has direct influence on
the methylation and the acetylation state of histone proteins.
3 DNA Methylation of Genes in Lung Cancer Patients
Bisulfite genomic sequencing and methylation-specific polymerase

chain reaction (MSP) are recognized as the most frequent methods
to identify the status of gene methylation [7, 8]. Besides, with the
development of polymerase chain reaction (PCR), quantitative real-
time PCR (RT-PCR) has been widely applied [9]. The methods
above are on the basis of unchanged structure of methylated cyto-
sines while unmethylated cytosine bases could be converted by
sodium bisulfite. Afterwards, specific primers can be applied to
detect methylated or unmethylated target sites.
The methylation status of many genes has been explored in
lung cancer patients, most of whom were non-small cell lung cancer
(NSCLC) patients. Genes associated with methylation have been
thoroughly probed, including RASSF1A [10], p16 [11], EGLN2
[12], SETDB1 [13], LRRC3B [14], and so on. RASSF1A and p16
play an important role in cell cycle regulation while EGLN2 and
LRRC3B DNA methylation and expression interact with hypoxia
inducible factor 1A to affect survival of early-stage NSCLC.
SETDB1 is crucial for cell membrane recruitment, phosphoryla-
tion, and activation of Akt following growth factor stimulation. In
primary NSCLC patients, the incidence of gene methylation was
proved up to 96% [15]. In small cell lung cancer (SCLC) patients,
ZAR1 acts as a novel epigenetically inactivated tumor
suppressor [16].
4 Different Methylation Patterns According to Histology
When compared with malignant lung tissue samples, the methyla-

tion of most genes was rarely detected in paired non-malignant
lung tissue samples, which implies that methylation of
ALGrawany
Methylation in Lung Cancer: A Brief Review 93
cancer-related genes is an important process in tumorigenesis of

lung cancer [17]. Toyooka et al. [18] conclude that small cell lung
cancer (SCLC), carcinoids, squamous cell carcinomas, and adeno-
carcinomas of the lung have unique profiles of aberrant methyla-
tion. They found the frequency of methylation of RASSF1A was
significantly higher in neuroendocrine tumor patients than in
NSCLC patients while the frequencies of methylation of CDH13,
APC, and p16 were significantly higher in NSCLC patients than in
those with neuroendocrine tumor. Besides, the frequencies of
methylation of RASSF1A, CDH1, and RARβ were higher in
SCLC patients than in bronchial carcinoids. As for squamous cell
carcinoma and adenocarcinoma, p16 is more frequent in lung
squamous cell carcinoma while adenocarcinoma has high frequency
of APC and CDH13.
5 DNA Methylation Acts as Biomarker in Lung Cancer
DNA methylation can be detected not only in tumor tissues, but in

blood, sputum, and bronchioloalveolar lavage (BAL) samples,
making DNA methylation detection of specific genes act as a trust-
worthy approach for diagnosis and prognosis in lung cancer
patients. Besides, DNA methylation sites can also serve as thera-
peutic molecule targets. In early days, some questions arose that is
there sufficient circulating tumor cell in early-stage cancer? What is
the specificity of blood DNA methylation detection for a certain
cancer?
In the early stages of lung cancer, DNA methylation occurs at a
high frequency, so it can be used as a new marker to help early
diagnosis and early screening of lung cancer. Methylated DNA
fragments can be detected in peripheral blood and bronchial epi-
thelial exfoliated cells, which are all less invasive diagnostic meth-
ods, so the detection of DNA methylation is very promising.
The most ideal test sample is peripheral blood. Since lung
cancer cells release circulating tumor DNA (ctDNA) into peripheral
blood by lysis, autocrine, apoptosis, and necrosis, ctDNA with
higher expression levels may be found in peripheral blood of lung
cancer patients, and they are usually associated with primary lung
cancer lesions, having the same type of gene expression. Some hot
DNA methylation sites in lung cancer tissues have been confirmed
in plasma or serum [19]. Studies have found that the methylation
status of p16, DAPK, SFRP1, and KLK10 in circulating blood of
NSCLC is significantly different from that in benign lung lesions
and normal lung tissues [20, 21]. Therefore, they can be used as
new markers that help with the diagnosis of early-stage lung cancer.
Some scholars have found that in the peripheral plasma of patients
with non-small cell lung cancer, the methylation of p16 gene
combined with circulating DNA content can increase the detection
sensitivity to 80%, which has significantly better advantage than

using single DNA methylation detection. However, everything
has its two sides, and peripheral blood testing is no exception.
The main disadvantages are: (1) the circulating DNA content in
some specimens is low and difficult to detect; and (2) the methy-
lated DNA fragments in peripheral blood are less organ specific.
The sputum sample test has certain organ specificity because
most of the cells from the respiratory and lung tissues, and DNA
methylation sites have been found in sputum samples [22]. The
population with the highest incidence of lung cancer are smokers.
The patient set has more sputum than the normal subjects due to
long-term and heavy smoking, and their sputum samples often can
be easily obtained without special induction. Therefore, the sputum
sample test is more suitable for the initial screening of lung cancer
in this part of the population. One study found that three or more
of the MGMT, p16, RASSF1A, DAPK, GATA5, and PAX5β sites
were detected having DNA methylation in the sputum samples
collected within 1.5 years prior to the diagnosis of lung cancer,
the probability of lung cancer in this group of patients will be 6.5
times higher than other patients [23]. The sputum test also has
limitations because sputum mainly comes from the central part of
the bronchial system and has limited ability to detect peripheral
type of lung neoplasms.
Bronchial lavage fluid is another alternative study sample. Sim-
ilar to sputum samples, bronchial lavage fluid may also contain
more specific lung cancer cells and release lung cancer DNA frag-
ments. Some scholars have tested DNA methylation in bronchial
lavage fluid in patients with non-small cell lung cancer and found
that one or more methylation site(s) of RARβ, p16, RASSF1A, and
H-cadherin can be detected in most (68%) samples. Another study
detected bronchial lavage samples from patients with suspected
lung cancer and found that the specificity was 99% and the sensitiv-
ity of 53%, when combined with methylation detection at the three
sites of p16, APC, and RASSF1.
All of the above three detection methods are mildly invasive,
and the detection of exhaled breath condensate samples is a
completely noninvasive detection method. The detection technol-
ogy detects biochemical components of the lung and the respira-
tory tracts, not affecting the pathophysiological processes in the
respiratory system [24]. Compared with sputum sample testing and
bronchial lavage sample testing, exhaled breath condensate sample
testing is simpler, more feasible, and repeatable. At present, this
technology can be widely applied in the analysis of intrapulmonary
inflammatory response, exploring lung respiratory function, devel-
oping lung cancer tumor markers, studying intrapulmonary oxida-
tive and stress responses, and monitoring the incidence of lung
cancer. At this stage, the research data on the diagnosis of lung
cancer by exhaled condensate is relatively poor and the data on the
methylation of target genes in exhaled condensate samples are

scarce [25]. Recently, studies have also found that DNA methyla-
tion analysis of the oral epithelium may also be a new method for
screening lung cancer [26].
6 DNA Methylation and the Treatment of Lung Cancer
DNA methylation is a reversible process, unlike irreversible genetic

information changes such as gene deletions or mutations. There-
fore, theoretically demethylation therapy on patients with lung
cancer or precancerous lesions may restore the function of some
tumor suppressor genes, thereby achieving the purpose of treating
lung cancer or preventing lung cancer.
Studies by Wu F et al. [27] showed that promoter methylation
of hMLH1 is involved in the sensitivity of cisplatin in NSCLC and
A549/DDP cells. Cisplatin-based adjuvant chemotherapy is more
beneficial to NSCLC patients with hMLH1-free methylation.
Methylation of hMLH1 may be a biomarker for individualized
treatment of non-small cell lung cancer. Unlike classical genetics,
epigenetic methylation changes are reversible, and DNMTs are key
enzymes in the process. Hypermethylation of the promoter region
results in changes in the expression of tumor suppressor genes,
which is the mechanism of most tumorigenesis. However, after
DNA combines with methylation inhibitors (decitabine,
5-azacytidine, etc.), stable complex is formed with DNMTs. The
complex, which in turn causes a decrease in the activity of DNMTs,
thereby reduces the methylation rate of the tumor suppressor
genome, inducing tumor cells to differentiate into normal cells or
causing apoptosis of tumor cells. Therefore, DNMTs and their
inhibitors may become research hotspots for the treatment of
tumors.
7 DNA Methylation and Prognosis of Lung Cancer
Some scholars have found that the degree of DNA methylation is

closely related to TNM staging and distant metastasis of non-small
cell lung cancer [28]. The study suggests that epigenetic abnorm-
alities such as DNA methylation may also be associated with pro-
gression of lung cancer and may be an indicator for predicting lung
cancer. Studies have shown that the methylation status of DLEC1
and hMLH1 genes in NSCLC patients is associated with TNM
staging of lung cancer and metastasis of mediastinal lymph nodes,
and is an independent prognostic factor for lung cancer
[29]. Another scholar pointed out that abnormal methylation of
RXRG is a risk factor in smoking non-small cell lung cancer
patients, and abnormal methylation of RXRG in non-smoker
non-small cell lung cancer patients is a protective factor [30]. A

comparative analysis of postoperative recurrence and recurrence-
free stage I NSCLC patients showed that the methylation status of
CDH13, p16, APC, and RASSF1A in lung cancer tissues and their
corresponding lymph node tissues was closely related to the recur-
rence of the corresponding patients. If methylation of the p16 and
CDH13 genes is detected in both tumor and mediastinal lymph
nodes, the odds ratio for lung cancer recurrence is 15.5. It is
suggested that if DNA methylation is detected in the dissected
lymph nodes, even if no metastasis is found in the histology of the
lymph nodes, it may indicate that local micrometastasis may have
occurred in the lung cancer patients. It can be seen that the hyper-
methylation of certain genes in NSCLC may confer the ability of
tumor cells to metastasize and spread, which may be an important
indicator for postoperative recurrence and poor prognosis.
8 Research Prospects
The epigenetics of lung cancer has been studied for nearly two
decades, during which time many important effects and mechan-
isms of lung cancer development have been accumulated. Liquid
biopsy of lung cancer is one of the current research hotspots. It is a
noninvasive test and its amount of information, especially DNA
methylation, has been extensively studied. DNA methylation mar-
kers are stable and easy to detect in tissues or body fluids (blood,
sputum, etc.) and may play a significant role in the early diagnosis,
prognosis, predictive treatment, and drug resistance of lung cancer.
However, the current lack of standardization of methylation detec-
tion has hindered the development of methylation, so the estab-
lishment of standardized schemes is particularly important. Perhaps
in the future, liquid biopsy can provide strong support for the
clinical work of treating lung cancer and also provide new insights
for the treatment of lung cancer.
References
1. Siegel R, Ma J, Zou Z et al (2014) Cancer 4. Shogren-Knaak M, Ishii H, Sun JM et al

statistics, 2014. CA Cancer J Clin 64(1):9–29 (2006) Histone H4-K16 acetylation controls
2. Gu C, Wang R, Pan X et al (2017) Sublobar chromatin structure and protein interactions.
resection versus lobectomy in patients aged Science 311(5762):844–847
35 years with stage IA non-small cell lung 5. Esteller M (2008) Epigenetics in cancer. N
cancer: a SEER database analysis. J Cancer Engl J Med 358(11):1148–1159
Res Clin Oncol 143(11):2375–2382 6. Ikegami K, Ohgane J, Tanaka S et al (2009)
3. Vettese-Dadey M, Grant PA, Hebbes TR et al Interplay between DNA methylation, histone
(1996) Acetylation of histone H4 plays a pri- modification and chromatin remodeling in
mary role in enhancing transcription factor stem cells and during development. Int J Dev
binding to nucleosomal DNA in vitro. EMBO Biol 53(2–3):203–214
J 15(10):2508–2518
7. Herman JG, Baylin SB (1998) Methylation- risk assessment in plasma. Cancer 110
specific PCR. Curr Protoc Hum Genet 16 (9):2019–2026
(1):10.6. 1–10.6. 10 20. Zhang YW, Miao YF, Yi J et al (2010) Tran-
8. Herman JG, Graff JR, Myöh€anen S et al (1996) scriptional inactivation of secreted frizzled-
Methylation-specific PCR: a novel PCR assay related protein 1 by promoter hypermethyla-
for methylation status of CpG islands. Proc tion as a potential biomarker for non-small cell
Natl Acad Sci 93(18):9821–9826 lung cancer. Neoplasma 57(3):228
9. Campan M, Weisenberger DJ, Trinh B et al 21. Zhang Y, Song H, Miao Y et al (2010) Fre-
(2009) MethyLight[M]//DNA methylation. quent transcriptional inactivation of Kallikrein
Humana Press, New York, NY, pp 325–337 10 gene by CpG island hypermethylation in
10. Hu H, Zhou Y, Zhang M et al (2019) Prog- non-small cell lung cancer. Cancer Sci 101
nostic value of RASSF1A methylation status in (4):934–940
non-small cell lung cancer (NSCLC) patients: 22. Shivapurkar N, Stastny V, Suzuki M et al
A meta-analysis of prospective studies. Biomar- (2007) Application of a methylation gene
kers 24(3):207–216 panel by quantitative PCR for lung cancers.
11. Liu Z, Lin H, Gan Y et al (2019) P16 methyla- Cancer Lett 247(1):56–71
tion leads to paclitaxel resistance of advanced 23. Belinsky SA, Liechty KC, Gentry FD et al
non-small cell lung cancer. J Cancer 10 (2006) Promoter hypermethylation of multiple
(7):1726 genes in sputum precedes lung cancer inci-
12. Zhang R, Lai L, He J et al (2019) EGLN2 dence in a high-risk cohort. Cancer Res 66
DNA methylation and expression interact (6):3338–3344
with HIF1A to affect survival of early-stage 24. Sriram KB, Larsen JE, Yang IA et al (2011)
NSCLC. Epigenetics 14(2):118–129 Genomic medicine in non-small cell lung can-
13. Wang G, Long J, Gao Y et al (2019) SETDB1- cer: paving the path to personalized care.
mediated methylation of Akt promotes its Respirology 16(2):257–263
K63-linked ubiquitination and activation lead- 25. Suda K, Tomizawa K, Yatabe Y et al (2011)
ing to tumorigenesis. Nat Cell Biol 21(2):214 Lung cancers unrelated to smoking: character-
14. Guo Y, Zhang R, Shen S et al (2018) DNA ized by single oncogene addiction? Int J Clin
methylation of LRRC3B: a biomarker for sur- Oncol 16(4):294–305
vival of early-stage non-small cell lung cancer 26. Hiroaki H, Narumi Y, Tomohiro K et al (2018)
patients. Cancer Epidemiol Prev Biomarkers Clinical implication of DNA methylation anal-
27(12):1527–1535 ysis in the oral epithelium for lung cancer risk
15. Heller G, Zielinski CC, Zöchbauer-Müller S stratification. Respirology 23(S2):69–78
(2010) Lung cancer: from single-gene methyl- 27. Wu F, Lu M, Qu L et al (2015) DNA methyla-
ation to methylome profiling. Cancer Metasta- tion of hMLH1 correlates with the clinical
sis Rev 29(1):95–107 response to cisplatin after a surgical resection
16. Richter AM, Kiehl S, Köger N et al (2017) in non-small cell lung cancer. Int J Clin Exp
ZAR1 is a novel epigenetically inactivated Pathol 8(5):5457
tumour suppressor in lung cancer. Clin Epige- 28. Qiu X, Zhou Q (2018) P093 DNA methyla-
netics 9(1):60 tion in CTC and ctDNA is a promising marker
17. Heller G, Fong KM, Girard L et al (2006) for early detection of lung cancer metastasis. J
Expression and methylation pattern of TSLC1 Thorac Oncol 13(12):S1082
cascade genes in lung carcinomas. Oncogene 29. Seng TJ, Currey N, Cooper WA et al (2008)
25(6):959 DLEC1 and MLH1 promoter methylation are
18. Toyooka S, Toyooka KO, Maruyama R et al associated with poor prognosis in non-small
(2001) DNA methylation profiles of lung cell lung carcinoma. Br J Cancer 99(2):375
tumors1. Mol Cancer Ther 1(1):61–67 30. Lee SM, Lee JY, Choi JE et al (2010) Epige-
19. Hsu HS, Chen TP, Hung CH et al (2007) netic inactivation of retinoid X receptor genes
Characterization of a multiple epigenetic in non-small cell lung cancer and the relation-
marker panel for lung cancer detection and ship with clinicopathologic features. Cancer
Genet Cytogenet 197(1):39–45
Chapter 9
Epstein–Barr Virus DNA in Nasopharyngeal Carcinoma:

A Brief Review
Fen Xue and Xiayun He
Abstract
Epstein–Barr virus (EBV) is a B-lymphocytes herpes virus and can transform B lymphocytes to malignant
tumor cells if infected. Nasopharyngeal carcinoma (NPC) is strongly associated with EBV. Circulating EBV
DNA in plasma has been recognized as an important biomarker of NPC. Much work has been done to
validate the ability of circulating EBV DNA for screening, diagnosis, risk stratification, monitoring, and
predicting prognosis. This chapter reviews the clinical progress of circulating cell-free EBV DNA in NPC.
Key words Epstein–Barr virus, Nasopharyngeal carcinoma, Circulating cell-free EBV DNA,
Reduced-dose radiation, Prognosis
1 Introduction
Although survival benefit was gained with the development of

radiological diagnostic techniques and chemoradiotherapy modal-
ities, distant metastasis and local recurrence remain the main failure
patterns for nasopharyngeal carcinoma (NPC). As the survival rates
for NPC patients decreased from stage I to stage IV, early diagnosis
may improve outcomes [1, 2]. Meanwhile, some patients with the
same stage also showed significant differences in treatment efficacy
and prognosis, which suggests that current anatomy-based stage
system is not enough to accurately predict prognosis [3]. Thus,
seeking other potential biomarkers for early detection and effective
monitoring would be of great importance for NPC patients.
Epstein–Barr virus (EBV) is a B-lymphocytes herpes virus and can
transform B lymphocytes to malignant tumor cells if infected
[4]. NPC is strongly associated with EBV. Since Lo et al. [5]
confirmed the feasibility of detecting circulating EBV DNA in
plasma, it has been recognized as an important biomarker of
NPC. Much work has been done to validate the ability of circulat-
ing EBV DNA for screening, diagnosis, risk stratification,
99
100 Fen Xue and Xiayun He
monitoring, and predicting prognosis [6, 7]. This chapter reviews

the clinical progress of circulating cell-free EBV DNA in NPC.
2 EBV DNA for Screening, Diagnosis, and Staging of NPC
With deep and concealed location of nasopharynx, clinical exam-

inations like nasopharyngoscopy or magnetic resonance imaging
(MRI) may not always detect the mucosal changes of nasopharynx
or small lesions in pharyngeal recess. Besides, early NPC is relatively
asymptomatic or with nontypical symptoms, so most of the patients
with NPC present with locally advanced disease or distant metasta-
sis at diagnosis. The extent of tumor at diagnosis is one of the most
important factors affecting survival. Therefore, identification of
patients with early-stage NPC through screening could potentially
improve treatment outcomes. Over the last few decades, circulating
cell-free EBV DNA has been proved as a potential biomarker for
screening and diagnosis [1, 8].
According to Lo et al. [5], circulating cell-free EBV DNA could
be detectable in the plasma of 96% of NPC patients and 7% of
normal individuals by using real-time quantitative PCR. The NPC
patients had much higher EBV DNA concentrations than the
normal individuals. And in the former group, patients with
advanced disease (stage III and IV) had significantly higher circu-
lating EBV DNA levels (about eight times in median levels) than
those with early disease (stage I and II). These findings suggest that
the liberation of EBV DNA into the blood may reflect the tumor
load in NPC patients. Then, were the amounts of circulating EBV
DNA released by small tumors still enough for detection? Chan
et al. [8] conducted a prospective study to verify its ability in
screening asymptomatic early NPC. In their study, nasopharyngo-
scopy and MRI were implemented for participants with persistently
positive EBV DNA in plasma. The results showed a 97.1% of
sensitivity and 98.6% of specificity. Compared to a historical cohort,
participants with NPC identified by screening had significantly
higher proportion of stage I or II disease (71 vs. 20%, P < 0.001)
and had superior 3-year progression-free survival (PFS, 97 vs. 70%;
hazard ratio, 0.10). Therefore, circulating cell-free EBV DNA may
be an ideal biomarker for screening. With early diagnosis, effective
treatments could be administrated timely to improve the treatment
outcomes and acute/late adverse effects caused by additional che-
motherapy could also be avoided.
In the latest eighth edition of the American Joint Committee
on Cancer (AJCC)/Union for International Cancer Control
(UICC) TNM staging system for NPC, patients present with cer-
vical lymph node metastasis and detectable EBV DNA but without
tumor in nasopharynx were defined as stage T0. Many studies have
been trying to incorporate circulating EBV DNA into current
Epstein–Barr Virus DNA in Nasopharyngeal Carcinoma: A Brief Review 101
TNM staging system to further segregate prognosis in NPC. Using

979 patients as primary cohort and 550 patients as the validation
cohort, Guo et al. [3] generated new stage groupings as follows:
stage RI (T1N0), RIIA (T2-T3N0 or T1-T3N1, EBV DNA
2000 copies/mL), stage RIIB (T2-T3N0 or T1-T3N1, EBV
DNA >2000 copies/mL; T1-T3N2, EBV DNA 2000 copies/
mL), stage RIII (T1-T3N2, EBV DNA >2000 copies/mL; T4N0-
N2), and stage RIVA (any T and N3), with 5-year PFS rates of
100%, 87.9%, 76.7%, 68.7%, and 50.4% for the validation cohort,
respectively. Their new stage groupings showed better prognostic
performance than current staging system. Another new staging
system was proposed by Lee et al. [9] as follows: RPA-I (T1–T4
N0–N2 & EBV DNA <500 copies per mL), RPA-II (T1–T4 N0–
N2 & EBV DNA 500 copies per mL), RPA-III (T1–T2 N3), and
RPA-IVA (T3–T4 N3). Their results also showed better survival
prediction than current staging system. Considering the varied
cut-off levels of EBV DNA in staging system, multicenter analysis
needed to be performed and quantitative assay needed to be stan-
dardized internationally. In this era of precision medicine, more
nonanatomic factors (like EBV DNA) are warranted to be
incorporated into staging system to improve survival prediction
and help stratify for individualized treatment strategies (intensified
treatment for high-risk patients and de-intensified treatment for
low-risk patients).
3 EBV DNA for Disease Monitoring in NPC
Circulating cell-free EBV DNA would drop dramatically with

tumor shrinkage and remission in NPC patients. Shotelersuk et al.
[10] reported that the detective rates of EBV DNA dropped from
69% before treatment to 46%, 23%, 7% at 1 week, 2 weeks or
3–6 weeks after radiotherapy, and 0% for patients with complete
remission. Other studies have shown that circulating EBV DNA
was rapidly cleared from the plasma with a median half-life of
139 mins after surgical resection of recurrent tumors, and with a
median half-life of 3.8 days after radiotherapy [11, 12]. Lo et al.
[12] also found that patients who remained in remission after
radiotherapy had continuous low or undetectable levels of circulat-
ing EBV DNA, while patients who developed local or distant failure
during follow-up had increased levels of circulating EBV DNA.
Similar phenomenon of correlation of EBV DNA load with disease
relapse and remission was found by Yip et al. [13]. In their study,
the elevation of EBV DNA load during follow-up was usually
corelated with further progression and often preceded clinical pro-
gression. Besides, a significant residual EBV DNA load after radio-
therapy often signaled residual disease in local or distant sites, or
predicted relapses or progression on further follow-up. Therefore,
the persistence or re-elevation of EBV DNA after radical treatment

often indicates tumor progression. It was further suggested that
patients with EBV DNA >500 copies/mL at 6 weeks after treat-
ment showed high risk for disease relapse or death compared with
those with undetectable level [14]. However, residual post-
treatment EBV DNA could also go spontaneous remission during
follow-up [15]. Almost three-quarters (72.4%) of the patients with
detectable post-treatment EBV DNA experienced a spontaneous
remission of EBV DNA. These patients still had higher risk of
disease relapse or death than those with undetectable post-
treatment EBV DNA level, and patients with persistently detectable
EBV DNA had the highest risk. Therefore, EBV DNA assays need
to be performed continuously at subsequent follow-up visits, and
early test of EBV DNA after treatment may help pick potential
high-risk patients for further intensified treatment. For future
post-treatment EBV DNA-related clinical trials, the optimal timing
of the EBV DNA assays needed to be uniformed in different
centers.
Currently, the diagnosis of locoregional recurrence or distant
metastasis of NPC patients is mainly based on imaging examination
and endoscopic biopsy. However, some of the patients may develop
mucosal ulcers, local edema, fibrosis, or osteoradionecrosis after
radiotherapy, which may lead to the interference of accurate detec-
tion of disease failure. Besides, it is hard to detect a recurrent or
metastatic lesion with diameter 5 mm with MRI or computed
tomography (CT). Therefore, regular EBV DNA measurement is
introduced to be a useful adjunct tool to endoscopy and imaging in
the surveillance for disease relapse and metastasis, despite the lack
of prospective data. Lo et al. [5] were the first to report that the
level of EBV DNA (median copy number: 32350 copies/mL) in
patients with recurrence was significantly higher than those without
(median copy number: zero copies/mL). It was further supported
by an imaging examination conducted by Makitie et al. [16], the
detection effects of EBV DNA load were consistent with the results
of PET/CT in the monitoring of locoregional recurrence or distant
metastasis. The meta-analysis including 25 studies also confirmed
the value of EBV DNA levels as a diagnostic tool for disease
progression, with a very high-level overall accuracy [17]. However,
its sensitivity in the detection of locoregional recurrences is rela-
tively lower than the detection of distant failures. With sensitive
tools, detection rate for distant metastasis ranged from 86% to 96%,
while that for locoregional recurrence varied from 51% to 67%
[13]. It was supposed that locoregional fibrosis and vascular occlu-
sion after radiotherapy may affect the EBV DNA released into
circulation in patients with locoregional recurrence. Researchers
have found that detection of EBV DNA by trans-oral nasopharyn-
geal brush biopsy was at high sensitivity and specificity for detecting
local NPC recurrence [18]. The same brush detection system has
been demonstrated by Lam et al. [19] to be clinical potential for

monitoring local recurrence in post-irradiated NPC patients. To
further verify its role in detecting salvageable early recurrent NPC, a
prospective study is ongoing by measuring the sensitivity and spec-
ificity of the combination of EBV and methylation marker genes in
body fluids of both plasma and nasopharyngeal brush
(NCT03379610).
4 EBV DNA and the Prognosis of NPC
As mentioned above, the circulating EBV DNA concentrations

seems to reflect the tumor burden in NPC patients. Researchers
[13] summarized studies and found that levels of circulating EBV
DNA increased from stage I to IV in NPC patients, with detection
rates of 50–86%, 94–95%, 91–100%, and 94–98%, respectively.
Therefore, it is reasonable to deduce that circulating EBV DNA
would be a potential prognostic marker. Considering the rapid
in vivo clearance rate of circulating EBV DNA, tumor load could
be reflected in almost real-time. As a result, researchers have inves-
tigated the prognostic value of circulating EBV DNA levels
measured at different time points (pre-, mid-, and post-treatment).
Leung et al. [20] found that pretreatment EBV DNA concen-
tration is more sensitive in predicting prognosis for early-stage
NPC than TNM staging. The prognosis is similar between stage
II NPC with low pretreatment EBV DNA concentration and stage
I NPC, while stage I NPC with high pretreatment EBV DNA
concentration showed even worse prognosis than stage III NPC.
Among locoregionally advanced NPC patients, Lin et al. [4]
reported an inferior overall survival (OS) and relapse-free survival
for NPC patients with a higher pretreatment EBV DNA concentra-
tion (with a cut-off of 1500 copies/mL). Another study also
showed that NPC patients with higher pretreatment EBV DNA
levels (with a cut-off of 4000 copies/mL) had a lower rate of 3-year
PFS, distant metastasis-free survival (DMFS) and OS rate, the
prognostic effects were proved in multivariate analysis [21]. In
metastatic NPC, patients with high circulating tumor cells (CTCs,
with a cut-off of 12 cells/7.5 mL) and EBV DNA levels (with a
cut-off of 10,000 copies/mL) at baseline also showed significantly
shorter PFS and OS [22]. These findings suggest that the pretreat-
ment EBV DNA level provides additional prognostic information
and is of great significance for the formulation of individualized
treatment plan when combined with clinical stage to develop risk
stratification. However, the detection method and standard values
need to be harmonized in future studies.
The dynamic changes of EBV DNA levels during treatment
may reflect the half-life of EBV DNA in plasma. Those with short
half-life and rapid negative EBV DNA conversion during treatment
may indicate that tumors respond well to current treatment. You

et al. [22] found that the conversion of pretreatment unfavorable
CTCs and EBV DNA (12 cells/7.5 mL, 10,000 copies/mL) to
favorable (1 cells/7.5 mL, 4000 copies/mL) after first-line chemo-
therapy was associated with significantly longer PFS and OS.
Other researchers found that detectable EBV DNA levels after
two cycles of neoadjuvant chemotherapy were proved to be poor
prognostic factors for PFS and DMFS [23]. To further explore the
relation between EBV DNA clearance rate and tumor radiosensi-
tivity, a prospective study was implemented in 107 NPC patients
receiving radiotherapy/chemoradiotherapy. Results showed that
detectable mid-treatment EBV DNA levels (measured at week
4 of radiotherapy) was associated with worse clinical outcome and
was the only independent prognostic factor for DMFS and OS
[24]. These findings may provide potentials for escalation or
de-escalation clinical studies based on molecular response of EBV
DNA during treatment.
Post-treatment EBV DNA levels were found to have better
prognostic effect than pre- or mid-treatment EBV DNA levels in
previous meta-analyses [13, 25]. It was further validated by Qu
et al. [6] that the risk of metastasis or mortality for patients with
high post-treatment EBV DNA levels was five- to six-fold higher
than those with low post-treatment EBV DNA levels. And the
prognostic effect of post-treatment EBV DNA levels was two- to
three-fold higher than pre- or mid-treatment EBV DNA levels.
Most studies supposed that any detectable levels of post-treatment
EBV DNA reflected residual tumor or subclinical metastasis and
had a worse DMFS and OS [4, 7]. Wang et al. [26] found that
almost all patients with persistent EBV DNA levels after systemic
treatment developed recurrence or metastasis. Therefore, consoli-
dation therapy should be considered for NPC patients with positive
post-treatment EBV DNA levels.
5 EBV DNA and the Treatment of NPC
Considering the importance of EBV DNA in screening, staging,

prognosis, and surveillance, its value in guiding treatment of NPC
has becomes a hot research. Twu et al. [27] retrospectively com-
pared the clinical outcomes of 85 NPC patients with or without
adjuvant chemotherapy (all with persistently detectable EBV DNA
after 1 week of radiotherapy). Those received adjuvant chemother-
apy showed reduced distant failure and improved OS. However, the
results of the first clinical trial using EBV DNA stratified therapy
were disappointing. In the Hong Kong NPC Study Group 0502
trial (NCT00370890) [28], adjuvant chemotherapy did not show
survival benefits in patients with detectable EBV DNA at 6–8 weeks
after radiotherapy. It was supposed that the timing of EBV DNA
screening and adjuvant chemotherapy may be too late, or some

patients may resistant to cisplatin-based adjuvant chemotherapy
after cisplatin-based concurrent chemotherapy. Whether additional
chemotherapy was necessary for patients with detectable post-
treatment EBV DNA remains controversial, we are anticipating
the results of NRG-HN001 trial (NCT02135042). This trial
divided NPC patients into low-risk and high-risk groups according
to post-radiotherapy EBV DNA levels and was initiated to test the
possibility of eliminating unnecessary adjuvant chemotherapy for
low-risk patients while introducing intensified adjuvant chemother-
apy for high-risk patients. Another three ongoing trials
(NCT02363400 by Taiwan National Health Research Institute,
NCT02874651 by Sun Yat-sen University, NCT00370890 by
Chinese University of Hong Kong) also tried to explore the neces-
sity of adjuvant chemotherapy for NPC patients with detectable
post-radiotherapy EBV DNA levels. Taiwan National Health
Research Institute (NCT03544099) also initiated a phase II trial
to examine the efficacy of pembrolizumab for prolonging the
one-year disease-free survival in nasopharyngeal carcinoma patients
with solely detectable EBV DNA after curative chemoradiation.
In addition, pre- or mid-treatment EBV DNA levels may also
be valuable in guiding treatment. An ongoing Phase II clinical trial
(NCT02871518) was designed to compare the efficacy of
two-cycle and three-cycle cisplatin-based concurrent chemotherapy
for low-risk patients (stage III-IVB, AJCC 7 edition, identified with
pretreatment plasma EBV DNA <4000 copies/mL). The
EP-STAR trial (NCT04072107) was designed to incorporate
on-treatment EBV DNA surveillance for guiding individualized
treatment adaptation. This trial aimed to investigate whether addi-
tional adjuvant metronomic capecitabine would benefit intermedi-
ate risk group (identified as patients with detectable EBV DNA
after one cycle of induction chemotherapy, which then drops to
undetectable levels during the following IC cycles), and whether
additional concurrent and adjuvant anti-PD-1 therapy would
benefit high-risk group (identified as patients with detectable
EBV DNA after three cycles of IC or with EBV DNA bounce
during the induction phase). Another Phase II clinical trial
(NCT03668730) aims to investigate the feasibility of reduced-
dose radiation (60 Gy) for low-risk NPC patients (identified as
stage III with pretreatment EBV DNA <4000 copies/mL, and
then drop to undetectable levels and develop radiographical com-
plete/partial response after two cycles of induction chemotherapy).
The results of these trials are anticipated for improving personalized
therapy.
6 Research Prospects
Circulating cell-free EBV DNA has been developed as a tumor

marker for NPC over the last two decades. As a simple, effective,
and economical method, the role in screening, early diagnosis,
staging, prognosis, and surveillance has been extensively investi-
gated. Incorporating EBV DNA levels at different time points and
monitoring its dynamic changes for risk-stratified treatment adap-
tation may have the potential to revolutionize NPC management.
However, it is important to note that current reported EBV
DNA-related researches mainly come from the epidemic areas of
NPC, whether it can be applicable for non-epidemic NPC patients
needs to be further confirmed. Besides, the current absence of
standardization of EBV DNA detection has hindered the compari-
son and interchange of the results from different institutions.
Therefore, the establishment of standardized assays is particularly
important for launching EBV DNA stratified clinical trials. We
believe the results of these trials have a promising future for perso-
nalized therapy and will help us move forward toward precision
medicine.
References
1. Chen YP, Chan ATC, Le QT et al (2019) 8. Chan KCA, Woo JKS, King A et al (2017)
Nasopharyngeal carcinoma. Lancet 394 Analysis of plasma Epstein-Barr virus DNA to
(10192):64–80 screen for nasopharyngeal cancer. N Engl J
2. Chua MLK, Wee JTS, Hui EP et al (2016) Med 377(6):513–522
Nasopharyngeal carcinoma. Lancet 387 9. Lee VH, Kwong DL, Leung TW et al (2019)
(10022):1012–1024 The addition of pretreatment plasma Epstein-
3. Guo R, Tang LL, Mao YP et al (2019) Pro- Barr virus DNA into the eighth edition of
posed modifications and incorporation of nasopharyngeal cancer TNM stage classifica-
plasma Epstein-Barr virus DNA improve the tion. Int J Cancer 144(7):1713–1722
TNM staging system for Epstein-Barr virus- 10. Shotelersuk K, Khorprasert C, Sakdikul S et al
related nasopharyngeal carcinoma. Cancer (2000) Epstein-Barr virus DNA in serum/
125(1):79–89 plasma as a tumor marker for nasopharyngeal
4. Lin JC, Wang WY, Chen KY et al (2004) Quan- cancer. Clin Cancer Res 6(3):1046–1051
tification of plasma Epstein-Barr virus DNA in 11. To EW, Chan KC, Leung SF et al (2003) Rapid
patients with advanced nasopharyngeal carci- clearance of plasma Epstein-Barr virus DNA
noma. N Engl J Med 350(24):2461–2470 after surgical treatment of nasopharyngeal car-
5. Lo YM, Chan LY, Lo KW et al (1999) Quanti- cinoma. Clin Cancer Res 9(9):3254–3259
tative analysis of cell-free Epstein-Barr virus 12. Lo YM, Leung SF, Chan LY et al (2000) Kinet-
DNA in plasma of patients with nasopharyn- ics of plasma Epstein-Barr virus DNA during
geal carcinoma. Cancer Res 59(6):1188–1191 radiation therapy for nasopharyngeal carci-
6. Qu H, Huang Y, Zhao S et al (2020) Prognos- noma. Cancer Res 60(9):2351–2355
tic value of Epstein-Barr virus DNA level for 13. Yip TT, Ngan RK, Fong AH et al (2014) Appli-
nasopharyngeal carcinoma: a meta-analysis of cation of circulating plasma/serum EBV DNA
8128 cases. Eur Arch Otorhinolaryngol 277 in the clinical management of nasopharyngeal
(1):9–18 carcinoma. Oral Oncol 50(6):527–538
7. Lam WKJ, Chan KCA, Lo YMD (2019) 14. Chan AT, Lo YM, Zee B et al (2002) Plasma
Plasma Epstein-Barr virus DNA as an arche- Epstein-Barr virus DNA and residual disease
typal circulating tumour DNA marker. J Pathol after radiotherapy for undifferentiated
247(5):641–649
nasopharyngeal carcinoma. J Natl Cancer Inst 22. You R, Liu YP, Lin M et al (2019) Relationship
94(21):1614–1619 of circulating tumor cells and Epstein-Barr
15. Zhang Y, Tang LL, Li YQ et al (2019) Sponta- virus DNA to progression-free survival and
neous remission of residual post-therapy overall survival in metastatic nasopharyngeal
plasma Epstein-Barr virus DNA and its prog- carcinoma patients. Int J Cancer 145
nostic implication in nasopharyngeal carci- (10):2873–2883
noma: a large-scale, big-data intelligence 23. Liu LT, Tang LQ, Chen QY et al (2015) The
platform-based analysis. Int J Cancer 144 prognostic value of plasma Epstein-Barr viral
(9):2313–2319 DNA and tumor response to neoadjuvant che-
16. Makitie AA, Reis PP, Irish J et al (2004) Corre- motherapy in advanced-stage nasopharyngeal
lation of Epstein-Barr virus DNA in cell-free carcinoma. Int J Radiat Oncol Biol Phys 93
plasma, functional imaging and clinical course (4):862–869
in locally advanced nasopharyngeal cancer: a 24. Leung SF, Chan KC, Ma BB et al (2014)
pilot study. Head Neck 26(9):815–822 Plasma Epstein-Barr viral DNA load at mid-
17. Peng H, Li Z, Long Y et al (2019) Clinical point of radiotherapy course predicts outcome
value of a plasma Epstein-Barr virus DNA in advanced-stage nasopharyngeal carcinoma.
assay in the diagnosis of recurrent or metastatic Ann Oncol 25(6):1204–1208
nasopharyngeal carcinoma: a meta-analysis. 25. Zhang J, Shu C, Song Y et al (2016) Epstein-
Biosci Rep:39(9) Barr virus DNA level as a novel prognostic
18. Ng RH, Ngan R, Wei WI et al (2014) Trans- factor in nasopharyngeal carcinoma: a meta-
oral brush biopsies and quantitative PCR for analysis. Medicine (Baltimore) 95(40):e5130
EBV DNA detection and screening of nasopha- 26. Wang WY, Twu CW, Lin WY et al (2011)
ryngeal carcinoma. Otolaryngol Head Neck Plasma Epstein-Barr virus DNA screening fol-
Surg 150(4):602–609 lowed by (1)(8)F-fluoro-2-deoxy-D-glucose
19. Lam JW, Chan JY, Ho WK et al (2016) Use of positron emission tomography in detecting
transoral nasopharyngeal brush biopsy for posttreatment failures of nasopharyngeal carci-
Epstein-Barr virus DNA detection of local noma. Cancer 117(19):4452–4459
recurrence of nasopharyngeal carcinoma after 27. Twu CW, Wang WY, Chen CC et al (2014)
radiotherapy. Head Neck 38(Suppl 1): Metronomic adjuvant chemotherapy improves
E1301–E1304 treatment outcome in nasopharyngeal carci-
20. Leung SF, Zee B, Ma BB et al (2006) Plasma noma patients with postradiation persistently
Epstein-Barr viral deoxyribonucleic acid quan- detectable plasma Epstein-Barr virus deoxyri-
titation complements tumor-node-metastasis bonucleic acid. Int J Radiat Oncol Biol Phys 89
staging prognostication in nasopharyngeal car- (1):21–29
cinoma. J Clin Oncol 24(34):5414–5418 28. Chan ATC, Hui EP, Ngan RKC et al (2018)
21. Chen WH, Tang LQ, Guo SS et al (2016) Analysis of plasma Epstein-Barr virus DNA in
Prognostic value of plasma Epstein-Barr virus nasopharyngeal cancer after chemoradiation to
DNA for local and regionally advanced naso- identify high-risk patients for adjuvant chemo-
pharyngeal carcinoma treated with cisplatin- therapy: a randomized controlled trial. J Clin
based concurrent chemoradiotherapy in Oncol 36(31):3091–3100
intensity-modulated radiotherapy era. Medi-
cine (Baltimore) 95(5):e2642
Chapter 10
A Review on Cancer of Unknown Primary Origin: The Role

of Molecular Biomarkers in the Identification of Unknown
Primary Origin
Na Yan, Yanxiang Zhang, Xuejie Guo, Dawei Yuan, Geng Tian,
and Jialiang Yang
Abstract
The primary site cannot be found after clinical and pathological evaluation, which are called cancers of
unknown primary origin (CUP). CUPs may resemble a specific primary tumor site which shares common
clinicopathological characteristics and prognosis. However, it may be present as a distinct disease entity with
undifferentiated pathological features. More than 4% of patients are diagnosed as CUP. These patients were
diagnosed as malignant tumors by cytology or pathology. And they were usually treated with empirical
chemotherapy and associated with a poor prognosis. How to accurately diagnose and treat a cancer of
unknown primary origin is a major clinical concern. To address this question, a complex assessment is
carried out which includes a complete medical history of the patient, physical examination, complete blood
count, urinalysis, serum chemistries, histologic evaluation, chest radiograph, computed tomography,
magnetic resonance imaging, and immunohistochemistry (IHC) studies. Molecular diagnostic information
reflects that CUP’s molecular characteristics are similar to primary tumors with the development of
genomics and the expansion of gene sequencing technology. Gene expression profiling is the most
commonly used molecular diagnostic method for CUP. In this chapter, we summarize the current
diagnostic methods and challenges of CUP, and the clinical value of the molecular-level tumor diagnostic
technique.
Key words Cancers of unknown primary origin, Metastasis, Immunohistochemistry, Gene expression
profiling, Molecular diagnostics
1 Introduction
The primary site cannot be found after clinical and pathological

evaluation, which are called cancers of unknown primary origin
(CUP) [1, 2]. These tumors may have features that are completely
undifferentiated, or they may be relatively well-differentiated, but
without an obvious primary origin characteristic [3]. More than 4%
of patients cannot be found the primary origin. 27% of the CUP can’t
be identified the primary site after autopsy series [4]. Metastatic
109
110 Na Yan et al.
tumors with an uncertain primary site is an important and difficult

clinical problem. How to tackle this problem? All of the patients
should be conducted the necessary investigations of the suspect
CPU cases. The necessary investigations for all patients include full
blood count, biochemistry, urinalysis, testing for occult blood in
stools, chest radiography, CT scan of thorax, abdomen, pelvis, histo-
logically confirmed metastatic cancer, detailed medical history, com-
plete physical (including pelvic and rectal) examination, and
histopathology review with specific immunohistochemical study.
The physical examination was recommended which includes palpa-
tion of breasts, rectal examination, and genitourinary. The patients
with different cancers’ typical characteristics should be conducted the
necessary investigations, that include mammography for all women,
breast magnetic resonance imaging (MRI), testicular ultrasonogra-
phy, PET or CT scan, concentrations of serum a-fetoprotein and b
human chorionic gonadotropin, concentrations of serum prostate-
specific antigen for all men, endoscopy, concentrations of serum
cancer antigen 125, and carcinoma antigen 15–3. With the develop-
ment of imaging technology, IHC studies and molecular identifica-
tion methods, the ability to identify the occult primary origin has
greatly improved in the last two decades [2, 5–7].
A CUP is associated with a poor prognosis as patients are
usually treated with a non-selective empirical therapy [8]. In the
current review, we summarize diagnostic challenges in CUP, with
emphasis on molecular profiling assays and their impact on man-
agement of this unique disease entity.
2 The Biology of CUP
There is a long-lasting debate about the biology of the CUP’s

metastatic process. As matters stand, the biological transfer of
cups is explained by two models. The first supposes is called the
linear progression. It goes through the process from a premalignant
lesion to malignant tumors, after which tumor cell dissemination
finds a metastasis. The linear migration hold that the migration
comes from completely deteriorated tumor cells and mainly
emerging in advanced cancer. “Genetic instability” leads to the
emergence of cancer cells. Cancer cells pass through multiple suc-
cessive rounds of mutation and selection for competitive fitness in
the context of the primary tumor. These tumor cell clones expand
and individual cancer cells leave the primary site to secondary
growths. The final tumor size is affected by the number of the
fully malignant clones. A second hypothesis postulates that metas-
tasis occurs early in the development of a malignant process. This
process is called parallel progression. Tumor cells undergo somatic
progression and metastatic growth at a distant site which depart the
primary origin before the acquisition of fully malignant
A Review on Cancer of Unknown Primary Origin: The Role of Molecular. . . 111
phenotypes. The parallel progression concludes that metastasis

must arise before the primary tumor was diagnosed, and the pri-
mary tumor was completely malignant. Parallel progression does
not question the opinion of linear migration. But it does not think
that metastasis was initiated at a late stage of primary tumor devel-
opment. Moreover, dissemination of malignant tumors was still
evolving. It may lead to allopatric selection that expansion of variant
cells adapted to specific microenvironments. But these two hypoth-
eses were derived indirectly and they lacked of incontrovertible
evidence [9–13].
3 Strategy of Diagnosis
The term cancer of unknown primary site presents as metastatic

disease. The origin of CUP is found only in a limited number of
cases. It is characterized by the presence of lymphatic and/or
hematogenous metastases and the inability to identify primary
tumors. The most common sites of origin included pancreaticobili-
ary tract, lung, and kidney. The commonest sites of involvement
include lungs, liver, lymph nodes, and bones. The conventional
diagnostic procedures include a complete medical history, physical
examination, complete blood count, urinalysis, serum chemistries,
histologic evaluation, chest radiograph, computed tomography,
magnetic resonance imaging, and immunohistochemistry (IHC)
studies [14–16].
3.1 Radiology The most frequently used modern imaging technology contains
CT, MRI, and PET-CT, which added substantially to the detection
of primary site in the past 30 years [17]. CT scan can be performed
for the chest, abdomen, and pelvis. CT scans performed a diagnos-
tic accuracy of 36–74% mainly in lung cancer, colorectal and pan-
creatic [6]. MRI was always recommended when occult carcinoma
presents with breast. In 70% of cases, MRI was very sensitive in
detecting primary breast cancers for women with axillary lymph-
adenopathy and negative mammography [18]. PET-CT was con-
sidered to be recommended when present with the investigation of
cervical lymphadenopathy and the head and neck cancers. PET-CT
performed better than MRI (22–44% vs. 20–27%) in which it
identifies correctly the origin of a CUP. However, PET-CT was
not the first choice which is considered from the medical expenses
[19–22].
3.2 Pathology and Histopathology plays a very important role in the CUP diagnosis. If
Immunohisto- clinical evaluation (specific evaluation of signs/symptoms and CT
chemistry scanning) cannot authenticate the primary site, the identification of
origin depends on the pathologic evaluation [23]. The first step is
to determine the type of cancer. 60% of cancers are identified as
112 Na Yan et al.
Table 1
Pathology staining provides a rough diagnosis for cancers of an unknown primary site
CK7 CK7+ Ovarian mucinous adenocarcinoma; pancreatic adenocarcinoma; cholangiocarcinoma;

CK20 CK20+ urothelial tumors;
CK7+ Lung adenocarcinoma; breast carcinoma; thyroid carcinoma; endometrial carcinoma;
CK20 cervical carcinoma;
CK7 Colorectal carcinoma; Merkel cell carcinoma;
CK20+
CK7 Hepatocellular carcinoma; Renal cell carcinoma;
CK20 Prostate carcinoma; squamous cell and small cell lung carcinoma; Head and neck
carcinoma;
adenocarcinomas. 30–35% of cancers are poorly differentiated or

undifferentiated carcinomas. Other types of cancers include 5% of
squamous cell carcinomas and 2% of neuroendocrine cancers [3]. It
is very important to identify the tumor lineage after determining
the type of cancers. Immunohistochemical (IHC) is the standard
pathological evaluation in CUP. IHC can lessen the diagnostic
spectrum to identify the tissue of origin. 60% of CUPs are adeno-
carcinomas. IHC stain for PSA is quite sensitive and specific for
prostate cancer [24]. CK5/6 has a nice performance in lung carci-
nomas that is negative for adenocarcinoma and positive for the
squamous cell carcinomas. However, individual IHC stains cannot
fully differentiate various adenocarcinomas. How to distinguish
specific adenocarcinomas is the most frequent diagnostic issue.
CK7 and CK20 are the most frequently used IHC staining.
Urothelial carcinomas and gastrointestinal (GI) will be tested posi-
tive with CK20, while CK7 positive for lung, endometrial, breast,
ovarian, or thyroid carcinomas [25, 26]. That provide a rough
diagnosis (Table 1). After that a series of highly specific staining
are performed (Table 2). We cannot use a great quantity of bio-
markers to test because the CUP specimens are small. In addition,
IHC staining has some limitations which include subjective inter-
pretation of results from pathologists, diversity of techniques in
performing the stains [5, 27].
3.3 Molecular Because of the development of molecular profiling, whole genome

Diagnosis or sets of gene expression in human tumors have enabled the
quantification. Gene expression profiling played an important role
in the classification and detection of primary tumor sites. Different
sites of origin have specific gene expression profiles that recognized
in cancers from different sites of origin. It reflects that the normal
tissues of origin can present the different gene expression profiles.
It can diagnose cancer types when we test the differential expression
of different gene sets. Early test performance was limited by the
number of gene expression markers that lead to relatively few
tumor types can be diagnosed [28, 29]. In order to solve this
Table 2
A combination of immunohistochemical markers provide further diagnosis for cancers of an
unknown primary site
Tumor type Marker

Carcinoma Cytokeratins
Lymphoma CLA, ALK1, CD30, CD43
Melanoma S-100, HMB45, Melan-A
Sarcoma Vimetin, Actin, Desmin, MyoD1, Myogenin, S100, CD34,
Squamous cell carcinoma CD99
Squamous cell carcinoma CK5/6, p63
(oropharynx) p16
Adenocarcinoma PSA, PAP
Prostate TTF-1, CK7þ/CK20
Lung Thyroid TTF-1, Thyroglobulin
Breast GCDFP-15, mammaglobin, ER
Colon CDX2, CK20þ/CK7
Pancreas/biliary CDX2, CK7þ/CK20þ or CK7þ/CK20
Germ cell tumors PLAP, OCT4, AFP, HCG
Neuroendocrine carcinoma Chromogranin, Synaptophysin, PGP9.5
problem, 40 different tumor types or subtypes can be detected by

using either reverse transcriptase polymerase chain reaction
(RT-PCR) or gene microarray techniques, combined with
improved bioinformatics [30, 31]. Hundreds of different tumors
had been tested gene expression profiles. The precision of molecu-
lar tests can be up to 75–85% [26, 27, 32]. Although many molec-
ular experiments have been previously evaluated in CUP, only two
assays had been developed commercially available. They are gene
RT-PCR assay, cancer TYPE ID, and the microarray for micro-
RNAs, Pathwork Tissue of Origin [27, 32]. Formalin-fixed paraffin
embedded (FFPE) tissue or cytology specimens can be used for
these assays [33, 34]. Cancer TYPE ID and Tissue of Origin require
a small amount of tissue which is different from IHC [5].
The Cancer TYPE ID assay can identify 30 main tumor types
and 54 subtypes, which is a 92-gene RT-PCR assay. There are three
methods to study the 92-gene RT-PCR assay: compared with
confirmatory IHC, direct comparison to specific primary sites,
and compared with single IHC diagnoses. When comparing the
results of the molecular tests with the IHC results the Cancer TYPE
ID showed a sensitivity of 72–95% in identifying the primary site of
the CUPs. Cancer TYPE ID reports a main cancer type with its
probability and one histological subtype with its probability, such as
ovary (93%) and subclass ovary serous (87%). It also reports other
main cancer classes with >5% probability. But that will be excluded
in those with <5% probability [26, 34, 35].
Tissue of Origin assay can analyze 2140 tumors of 58 types and
subtypes, which is a 2000 mRNAs microarray assay. The main
cancer types can be grouped into 15 classes that include kidney,
114 Na Yan et al.
non-small cell lung, breast, bladder, colorectal, gastric, testicular

germ cell, hepatocellular, non-Hodgkin’s lymphoma, melanoma,
thyroid, ovarian, pancreatic, prostate, sarcoma, and prostate. Tissue
of Origin reports similarity scores (SS) as cancer type probability.
Each of the 15 tumor classes have a similarity score. So, there are
15 SS for one sample, its total 100. It agrees with 90% of reference
diagnoses when a prediction with SS > 60. And it excludes with
90% of reference diagnoses when a prediction with SS < 5. In the
15 tumor classes, Tissue of Origin can rule out at least 12 tumor
classes. In most cases, it provides one highly probable diagnosis
[36, 37]. When comparing the results of the molecular tests with
the IHC results, Tissue of Origin showed a sensitivity of 87–94% in
identifying the primary site of the CUPs [27, 32, 38].
Although molecular tests require a small amount of tissue, on
average, 10% of specimens do not process successfully. It cannot
yield an RNA profile because of insufficient tissue, poor preserva-
tion, or extensive necrosis [5, 39]. Cancer TYPE ID is not very
sensitive for pancreatic, colorectal, and gastroesophageal cancers.
Tissue of Origin is not very sensitive for sarcoma [39]. On the other
hand, molecular tests cost several times as much as IHC [40].
The development of molecular tests require validation in prac-
tical cases and tumor sets, such as primary tumors or metastatic
cancer with known primary site, poorly differentiated tumors. Can-
cer TYPE ID individual showed 87% and 82% sensitivity in known
primary tumors site and metastatic cancer (790 cases) [35]. In
283 poorly differentiated or undifferentiated primary tumors, Tis-
sue of Origin showed 87% sensitivity. And it showed 91% sensitivity
in 179 known metastases. In addition, Tissue of Origin showed
94% and 95% sensitivity in malignant effusions by cytology
(17 cases) and laborious primary and metastatic cases [34, 41].
Most commercial tests are broad-based assays which class many
types tumors. They were devised pre-specified gene sets and have
been developed similarly. Another important test shows that
human papilloma virus (HPV) can identify the primary origin of a
squamous cell carcinoma of unknown primary in the metastatic
neck lesion. The HPV infection status was a strong predictor of a
primary cancer of the oropharynx, especially type 16. The overall
prevalence of HPV is 69.5% for tumors in head and neck sites, while
oropharyngeal SCC accounts for 47.7% [42].
3.4 IHC and Molecular tests and IHC have their own advantages and disadvan-
Molecular Methods tages. It can be seen that pathology with IHC and molecular
Complement profiling are not completely independent. There is the potential
Each Other for IHC and molecular methods to complement each other in the
diagnosis of primary unknown tumors. They were used many of the
same tissue-specific genes. When tumor types are easy to distin-
guish morphology and IHC, they are distinct on molecular testing
too. If patients are with a single special IHC marker diagnosis of the
ALGrawany
tissue of origin, molecular testing is unnecessary. When patients are

without a single IHC marker, molecular methods can add substan-
tially to the diagnostic evaluation of patients with cancer unknown
primary site. The molecular methods prediction of cancer unknown
primary origin was supported mostly by subsequent pathologic and
clinical evaluation [5, 43].
4 Treatment
Patients were usually treated with empirical chemotherapy, when

the tissue of origin was not known. Oncologists didn’t make special
treatment decisions based on IHC predictions, so that patients
were frequently treated with empiric chemotherapy too
[44, 45]. The empiric broad spectrum antineoplastic regimen is
always a platinum combined with either paclitaxel or gemcitabine.
80% patients treated with empiric chemotherapy, while it has been
reported a response rate of only 20% and a median survival of just
6 months. Only 20% of cancer unknown primary patients were
treated with relatively site-specific regimens by molecular methods.
These patients were treated according to guidelines for one of the
major known tumor types [8, 46].
CUPs are mainly divided between the favorable prognosis
group and the unfavorable prognosis group. Different subsets
patients should be supplied customized treatment which can be
achieved long-term disease control. Prognosis of favorable CUP
patients include women with papillary adenocarcinoma of the peri-
toneal cavity (Mean overall survival:36 months), women with ade-
nocarcinoma that involves only axillary lymph nodes(Mean 5-year
overall survival: 72%), squamous cell carcinoma that involves the
head and neck lymph nodes(5-year survival: 60–65%) and men with
blastic bone metastases, neuroendocrine carcinomas of unknown
primary site(Median survival: 15.5 months with 2-year survival:
33–50%. Long-term survivors:10–15%) [47]. Patients with a high
PSA can be managed as advanced prostate cancer, which will have a
favorable treatment prognosis. Resectable metastasis also has a
better response to therapy which can be treated by surgical excision.
It is unfavorable prognosis when adenocarcinoma metastatic to the
bone, brain, viscera (Tables 3 and 4) [18, 48–50].
In most of the cases, an isolated lymphadenopathy resides in
the clinical onset of CUPs. The pathology specimen of cervical,
supraclavicular, axillary, or inguinal metastasis should be obtained
in order to facilitate future direct targeted therapy. Excisional
biopsy or incisional biopsy/core-needle biopsy is the preferred
approach.
Patients achieved a response rate of 53% and an overall survival
of 13 months by a treatment with carboplatin and paclitaxel with or
without maintenance therapy with erlotinib and bevacizumab
[51]. A high PSA can be managed as advanced prostate cancer,
116 Na Yan et al.
Table 3
Subset stratification according to prognosis
Therapeutic
Group Tumor type Frequency management
Favorable prognosis Papillary adenocarcinoma of the 20% Treated according to
peritoneal cavity in women; guidelines for one of the
adenocarcinoma that involves only major known tumor
axillary lymph nodes; types
Adenocarcinoma that involves only
axillary lymph nodes;
Squamous cell carcinoma that involves
the head and neck lymph nodes;
Blastic bone metastases,
neuroendocrine carcinomas of
unknown primary site in men;
prostate cancer with a high PSA;
resectable metastasis
Unfavorableprognosis Adenocarcinoma metastatic to the 80% Treated with empiric
bone, brain, viscera; unknown cancer broad spectrum
primary origin after testing antineoplastic regimens
Table 4
Therapeutic management according to ESMO guidelines
Tumor type Recommended treatment

Poorly differentiated neuroendocrine Platinum + etoposide combination chemotherapy;
carcinoma;
Serous papillary peritoneal Optimal surgical debulking followed by platinum–taxane-
adenocarcinoma; based chemotherapy;
Isolated axillary nodal metastases; Axillary nodal dissection, mastectomy or breast irradiation,
and adjuvant chemohormonotherapy;
Squamous carcinoma involving cervical Neck dissection and/or irradiation of bilateral neck and head-
lymph nodes; neck axis. For advanced stages induction chemotherapy with
platinum-based combinationor chemoradiation;
Adenocarcinoma with a colon-profile; Chemotherapy regimens for colorectal cancer; Androgen
deprivation therapy RT
Men with blastic bone metastases and Single metastatic deposit from unknown primary Resection
IHC/serum PSA expression and/or RT systemic therapy
Unfavorable subsets Platinum-based empirical chemotherapy
which will have a favorable treatment prognosis [6]. The therapy

response rates of colon adenocarcinoma is similar to patients whose
CK20+, CK7, CDX2+ adenocarcinoma with a colon cancer profile.
Resectable metastasis also has a better response to therapy which
can be treated by surgical excision [52, 53].
5 Conclusion
It is still a challenge in diagnosing and treating an unknown primary

cancer. An unknown primary cancer could have a favorable treat-
ment prognosis or a dismal prognosis. Adenocarcinoma is the
commonest favorable treatment prognosis groups which is treated
with local or systemic treatment. Molecular assay can provide a
single diagnosis, when pathology and immunohistochemistry
methods cannot divide a single tissue of origin diagnosis. That is
very important for the selection of appropriate therapy. A better
performance of treatment will be approved in site-specific therapy
based on a tissue of origin diagnosis, which is compared with
empiric chemotherapy. Molecular assay diagnosis is sensitive, but
it still needs more further prospective research to testify the clinical
value.
References
1. Pavlidis N et al (2003) Diagnostic and thera- 10. Klein CA (2009) Parallel progression of pri-
peutic management of cancer of an unknown mary tumours and metastases. Nat Rev Cancer
primary. Eur J Cancer 39(14):1990–2005 9(4):302–312
2. Pentheroudakis G, Golfinopoulos V, Pavlidis N 11. Pavlidis N, Khaled H, Gaafar R (2015) A mini
(2007) Switching benchmarks in cancer of review on cancer of unknown primary site: A
unknown primary: from autopsy to microarray. clinical puzzle for the oncologists. J Adv Res 6
Eur J Cancer 43(14):2026–2036 (3):375–382
3. Yang M, Nonaka D (2010) A study of immu- 12. Klein G (1998) Foulds’ dangerous idea revis-
nohistochemical differential expression in pul- ited: the multistep development of tumors
monary and mammary carcinomas. Mod 40 years later. Adv Cancer Res 72:1–23
Pathol 23(5):654–661 13. Bross ID, Viadana E, Pickren J (1975) Do
4. Pentheroudakis G, Greco FA, Pavlidis N generalized metastases occur directly from the
(2009) Molecular assignment of tissue of ori- primary? J Chronic Dis 28(3):149–159
gin in cancer of unknown primary may not 14. Abbruzzese JL et al (1995) Analysis of a diag-
predict response to therapy or outcome: a sys- nostic strategy for patients with suspected
tematic literature review. Cancer Treat Rev 35 tumors of unknown origin. J Clin Oncol 13
(3):221–227 (8):2094–2103
5. Tomuleasa C et al (2017) How to diagnose and 15. Roh JL, Kim JM, Park CI (2008) Lateral cervi-
treat a cancer of unknown primary site. J Gas- cal lymph node metastases from papillary thy-
trointestin Liver Dis 26(1):69–79 roid carcinoma: pattern of nodal metastases
6. Pavlidis N, Pentheroudakis G (2012) Cancer of and optimal strategy for neck dissection. Ann
unknown primary site. Lancet 379 Surg Oncol 15(4):1177–1182
(9824):1428–1435 16. Begum S et al (2003) Detection of human
7. Kamposioras K, Pentheroudakis G, Pavlidis N papillomavirus in cervical lymph nodes: a
(2013) Exploring the biology of cancer of highly effective strategy for localizing site of
unknown primary: breakthroughs and draw- tumor origin. Clin Cancer Res 9
backs. Eur J Clin Investig 43(5):491–500 (17):6469–6475
8. Hainsworth JD et al (2009) Paclitaxel/carbo- 17. Fu TS et al (2016) The role of transoral robotic
platin plus bevacizumab/erlotinib in the first- surgery, transoral laser microsurgery, and lin-
line treatment of patients with carcinoma of gual tonsillectomy in the identification of head
unknown primary site. Oncologist 14 and neck squamous cell carcinoma of unknown
(12):1189–1197 primary origin: a systematic review. J Otolar-
9. Frost P (1991) Unknown primary tumors: an yngol Head Neck Surg 45(1):28
example of accelerated (type 2) tumor progres- 18. Pavlidis N, Pentheroudakis G, Plataniotis G
sion. Basic Life Sci 57:233–237. discussion (2009) Cervical lymph node metastases of
237-40 squamous cell carcinoma from an unknown
118 Na Yan et al.
primary site: a favourable prognosis subset of 33. Hainsworth JD, Greco FA (2014) Gene
patients with CUP. Clin Transl Oncol 11 expression profiling in patients with carcinoma
(6):340–348 of unknown primary site: from translational
19. Lee JR et al (2015) Detection of occult primary research to standard of care. Virchows Arch
tumors in patients with cervical metastases of 464(4):393–402
unknown primary tumors: comparison of (18) 34. Stancel GA et al (2012) Identification of tissue
F FDG PET/CT with contrast-enhanced CT of origin in body fluid specimens using a gene
or CT/MR imaging-prospective study. Radiol- expression microarray assay. Cancer Cyto-
ogy 274(3):764–771 pathol 120(1):62–70
20. Ettinger DS et al (2014) Occult primary, ver- 35. Kerr SE et al (2012) Multisite validation study
sion 3.2014. J Natl Compr Cancer Netw 12 to determine performance characteristics of a
(7):969–974 92-gene molecular cancer classifier. Clin Can-
21. Olson JA Jr et al (2000) Magnetic resonance cer Res 18(14):3952–3960
imaging facilitates breast conservation for 36. Boscolo-Rizzo P et al (2015) The prevalence of
occult breast cancer. Ann Surg Oncol 7 human papillomavirus in squamous cell carci-
(6):411–415 noma of unknown primary site metastatic to
22. Ambrosini V et al (2006) 18F-FDG PET/CT neck lymph nodes: a systematic review. Clin
in the assessment of carcinoma of unknown Exp Metastasis 32(8):835–845
primary origin. Radiol Med 111 37. Hainsworth JD et al (2013) Molecular gene
(8):1146–1155 expression profiling to predict the tissue of
23. Pentheroudakis G, Lazaridis G, Pavlidis N origin and direct site-specific therapy in
(2010) Axillary nodal metastases from carci- patients with carcinoma of unknown primary
noma of unknown primary (CUPAx): a system- site: a prospective trial of the Sarah Cannon
atic review of published evidence. Breast research institute. J Clin Oncol 31(2):217–223
Cancer Res Treat 119(1):1–11 38. Rosenfeld N et al (2008) MicroRNAs accu-
24. Lin F, Liu H (2014) Immunohistochemistry in rately identify cancer tissue origin. Nat Bio-
undifferentiated neoplasm/tumor of uncertain technol 26(4):462–469
origin. Arch Pathol Lab Med 138 39. Bochtler T, Loffler H, Kramer A (2018) Diag-
(12):1583–1610 nosis and management of metastatic neoplasms
25. Wittwer CT et al (1997) The LightCycler: a with unknown primary. Semin Diagn Pathol 35
microvolume multisample fluorimeter with (3):199–206
rapid temperature control. BioTechniques 22 40. Ryska A et al (2009) Paraganglioma-like med-
(1):176–181 ullary thyroid carcinoma: fine needle aspiration
26. Erlander MG et al (2011) Performance and cytology features with histological correlation.
clinical evaluation of the 92-gene real-time Cytopathology 20(3):188–194
PCR assay for tumor classification. J Mol 41. Pillai R et al (2011) Validation and reproduc-
Diagn 13(5):493–503 ibility of a microarray-based gene expression
27. Meiri E et al (2012) A second-generation test for tumor identification in formalin-fixed,
microRNA-based assay for diagnosing tumor paraffin-embedded specimens. J Mol Diagn 13
tissue origin. Oncologist 17(6):801–812 (1):48–56
28. Sotiriou C, Piccart MJ (2007) Taking gene- 42. Golfinopoulos V et al (2009) Comparative sur-
expression profiling to the clinic: when will vival with diverse chemotherapy regimens for
molecular signatures become relevant to cancer of unknown primary site: multiple-
patient care? Nat Rev Cancer 7(7):545–553 treatments meta-analysis. Cancer Treat Rev 35
29. Bender RA, Erlander MG (2009) Molecular (7):570–573
classification of unknown primary cancer. 43. Oien KA, Dennis JL (2012) Diagnostic work-
Semin Oncol 36(1):38–43 up of carcinoma of unknown primary: from
30. Yu J et al (2007) Feature selection and molec- immunohistochemistry to molecular profiling.
ular classification of cancer using genetic pro- Ann Oncol 23(Suppl 10):x271–x277
gramming. Neoplasia 9(4):292–303 44. Greco FA, Pavlidis N (2009) Treatment for
31. Vanneschi L et al (2011) A comparison of patients with unknown primary carcinoma
machine learning techniques for survival pre- and unfavorable prognostic factors. Semin
diction in breast cancer. BioData Min 4:12 Oncol 36(1):65–74
32. Rosenwald S et al (2010) Validation of a 45. Greco FA (2013) Molecular diagnosis of the
microRNA-based qRT-PCR test for accurate tissue of origin in cancer of unknown primary
identification of tumor tissue origin. Mod site: useful in patient management. Curr Treat
Pathol 23(6):814–823 Options in Oncol 14(4):634–642
46. Pentheroudakis G, Stoyianni A, Pavlidis N 50. Casali PG et al (2018) Gastrointestinal stromal

(2011) Cancer of unknown primary patients tumours: ESMO-EURACAN Clinical Practice
with midline nodal distribution: midway Guidelines for diagnosis, treatment and follow-
between poor and favourable prognosis? Can- up. Ann Oncol 29(Suppl 4):iv267
cer Treat Rev 37(2):120–126 51. Pavlidis N, Pentheroudakis G (2010) Cancer of
47. Stoyianni A, Pentheroudakis G, Pavlidis N unknown primary site: 20 questions to be
(2011) Neuroendocrine carcinoma of answered. Ann Oncol 21(Suppl 7):
unknown primary: a systematic review of the vii303–vii307
literature and a comparative study with other 52. Hainsworth JD et al (2012) A retrospective
neuroendocrine tumors. Cancer Treat Rev 37 study of treatment outcomes in patients with
(5):358–365 carcinoma of unknown primary site and a colo-
48. Piccioli A et al (2015) Bone metastases of rectal cancer molecular profile. Clin Colorectal
unknown origin: epidemiology and principles Cancer 11(2):112–118
of management. J Orthop Traumatol 16 53. Sorin T et al (2015) Occult cancer in specimens
(2):81–86 of reduction mammaplasty aimed at symmetri-
49. Conner JR, Hornick JL (2015) Metastatic car- zation. A multicentric study of 2718 patients.
cinoma of unknown primary: diagnostic Breast 24(3):272–277
approach using immunohistochemistry. Adv
Anat Pathol 22(3):149–167
Part III
Precision Medicine of Cardiovascular Diseases

Chapter 11
DNA Methylation in Atrial Fibrillation and Its Potential Role

in Precision Medicine
Mengwei Lv, Wen Ge, Zhi Li, Chao Wang, and Yangyang Zhang
Abstract
Atrial fibrillation (AF), a common arrhythmia, can cause many serious consequences, including stroke and
even death. The pathological mechanism of AF is very complicated. Epigenetic mechanisms, especially
DNA methylation, contribute to the pathogenesis and maintenance of AF. DNA methylation is an
important part of epigenetic and plays a significant role in human physiology and pathology. AF patients
possess specific methylation sites (e.g., Pitx2c, RASSF1A, SURs, SERCA2a, and LINC00472), which have
potential values of being biomarkers and underlie the diagnosis and prognosis of AF. These methylation
sites can also benefit accurate treatment of AF. With deeper understanding into the epigenetic mechanisms
of AF, the precision medicine for AF has also developed rapidly. In the future, DNA methylation omics and
other research methods will be integrated to explore the epigenetic mechanisms in AF.
Key words DNA methylation, Atrial fibrillation, Precision medicine, Epigenetic modification
1 Introduction
Atrial fibrillation (AF) is a common arrhythmia. Although AF has

long been considered as a cause of stroke, it was until recently
regarded as a relatively “benign” arrhythmia [1]. AF is an important
factor in the morbidity and mortality of cardiovascular diseases. To
further understand AF and its complications, researchers and doc-
tors have focused on demonstrating the mechanisms of electrical
and structural remodeling of the atria [2]. A number of epigenetic
mechanisms have been found to contribute to AF, including DNA
methylation and non-coding RNAs [3]. With deeper understand-
ing into the epigenetic mechanisms of AF, the precision medicine of
AF has also developed rapidly. At present, the difficulty in AF
treatment is that no effective drugs can be found to control the
chronic process of persistent AF. Networks of genetic variation that
explain biological differences among individuals can also interact
Mengwei Lv and Wen Ge contributed equally to this work.
123
124 Mengwei Lv et al.
with genetic, environmental, or both determinants of diseases to

produce highly personalized disease phenotypes. To deal with dis-
eases, people must understand all aspects of this complexity
[4]. Precision medicine is based on this demand. This review was
aimed to summarize and analyze current studies about DNA meth-
ylation in AF and estimate the potential value of DNA methylation
in precision medicine.
2 Overview of Precision Medicine
The definition of precision medicine varies among researchers.

Generally, precision medicine is a diagnostic and therapeutic
model that provides the most effective and accurate treatment for
patients depending on their individual conditions [5]. Much infor-
mation need to be collected and analyzed to implement precision
medicine decisions correctly. This collection and analysis process is
an ecosystem that integrates system biology, clinical research, labo-
ratory tests (including molecular and “genomic” data), image dis-
covery, and electronic health records [4]. The completion of the
human genome project, the launch of the international human
genome simple gram project, the availability of next-generation
sequencing, and reports of the importance of proteomics, metabo-
lomics, transcriptomics, and epigenomics have together stimulated
the interest in studying “omics” to understand human diseases.
With the development and application of molecular biology tech-
nology, a variety of biomarkers have been developed and applied to
the diagnosis, treatment, and prognosis of diseases [4]. DNA meth-
ylation is an important epigenetic regulatory mechanism underly-
ing the occurrence and development of cardiovascular diseases and
may become a potential target of precision medicine. However,
DNA methylation is rarely used for precision medicine in
AF until now.
3 Definition of DNA Methylation
DNA methylation is one of the most thoroughly-studied epigenetic

mechanisms and is essential for physiological development. During
DNA methylation, methyl (CH3) is covalently connected to the C5
position of cytosine residues to form 5-methylcytosine (5mC)
[6]. In normal mammalian cells, most of CpG dinucleotides are
methylated. Nonetheless, “CpG islands” are not evenly distributed
[7] and contain a mass of CpG sites in a quite short region in the
mammalian genome. Normally, these CpG sites are unmethylated.
But when in the pathological state, the methylation states of CpG
islands tend to be changed [8]. Although DNA methylation is an
important part of normal development and transcriptional regula-
tion, abnormal DNA methylation patterns are associated with many
heart diseases.
DNA Methylation in Atrial Fibrillation and Its Potential Role in Precision. . . 125
4 DNA Methylation and AF
AF is one of the most common arrhythmias in clinical practices, and

its genesis and development are attributed to many pathophysio-
logical factors. Since AF is a highly genetically heterogeneous dis-
ease, the presence of only a few chromosomal abnormalities or gene
mutations cannot account for its origin [9]. Epigenetic mechanisms
are momentous in regulating gene expression. DNA methylation
can act as an important epigenetic mechanism in many biological
processes, including cell proliferation, cell differentiation, and
other cytological processes [10]. The risk factors for AF include
genetic factors and environmental factors. AF has a genetic predis-
position [11], but DNA sequence variation may be absent in differ-
ent subtypes. This suggests that epigenetic mechanisms may be
involved in the occurrence and development of AF. A few research-
ers have worked with DNA methylation in AF, but have not found
the specific regulatory mechanism.
To summarize and analyze the research on DNA methylation
and AF, we searched potential studies in several databases, includ-
ing Pubmed, ScienceDirect, Elsevier, and Wiley Online Library.
The keywords of the search were “DNA methylation” and “Atrial
Fibrillation.” The literature survey results are listed in Table 1.
Table 1
Hypermethylated promoter of genes related to AF
Hypermethylated Published
Gene Source year Authors
Pitx2c Heart failure and angiotensin II modulate atrial Pitx2c 2013 YuHsun
promotor methylation Kao
et al.
RASSF1A DNMT3A silencing RASSF1A promotes cardiac fibrosis 2014 Hui Tao
through upregulation of ERK1/2 et al.
SURx Promoter DNA methylation regulates murine SUR1 2012 Naheed
(Abcc8) and SUR2 (Abcc9) expression in HL-1 Fatima
cardiomyocytes et al.
SERCA2a Tumor necrosis factor-αdecreases sarcoplasmic reticulum 2010 Yu-Hsun
Ca2+-ATPase expressions via the promoter methylation Kao
in cardiomyocytes et al.
LINC00472 LncRNA-LINC00472 contributes to the pathogenesis of 2019 Liao-yuan
atrial fibrillation (AF) by reducing expression of JP2 and Wang
RyR2 via miR-24 et al.
Pitx2c paired-like homeodomain 2, RASSF1A ras association domain family 1 isoform A, SURx sulfonylurea receptor,
SERCA2a sarcoplasmic reticulum Ca2+ -ATPases, LINC00472 long non-coding RNA 00472
4.1 Hypermethylated Some AF patients have a clear risk factor and may present with
Promoters primary heart diseases. As a common sense, heart failure (HF) can
of Paired-like increase the incidence of AF. HF can enhance hypermethylation of
Homeodomain the Pitx2c promoter region of the left atrium, accompanied by an
2 (Pitx2c) in AF increase in DNA methyltransferase (DNMT)1, leading to
decreased expression of Pitx2c [12]. Normally, Pitx2c is highly
expressed in the left atrium, maintaining sinus rhythm
[11]. Decreased Pitx2c expression leads to Kir2.1 down-expression
in the left atrium. Kir2.1, also known as potassium inwardly rectify-
ing channel subfamily J member 2, is more likely to allow potassium
to flow into, rather than out of, a cell. Kir2.1 can participate in
establishing action potential waveform and excitability of neuronal
and muscle tissues. Mutations in this gene can contribute to cardiac
arrhythmias [13].
HF can activate the renin angiotensin system, leading to an
increase of angiotensin II (AngII). AngII may cause structural
remodeling and abnormal electrical activity in the atrial and pulmo-
nary vein regions [14]. AngII can also lead to an increase in
DNMT1 and thereby DNA methylation resulting in hypermethyla-
tion of the Pitx2c promoter region [12]. This change will decrease
Pitx2c and Kir2.1 expressions so that electrophysiological remodel-
ing of the atrial region will give rise to AF. In contrast, angiotensin
receptor antagonist may benefit AF treatment. However, the
molecular mechanism of AngII was just proved to react in
HF-induced AF. Moreover, the treatment should be more precise.
This finding can contribute to precision medicine in AF.
4.2 Increased Collagen contents in normal atrial tissues are mixed at a certain
DNMT3A in Cardiac proportion to maintain the normal morphology and function of the
Fibrosis myocardium [15]. When atrial fibrosis occurs, the proportion of
various collagen in the extracellular matrix of cardiomyocytes is
disordered, and the collagen content significantly increases due to
excessive deposition of fibers. Trial fibrosis is a common pathologi-
cal manifestation of various heart diseases at a certain stage of
development and is the trigger of myocardial remodeling, which
can cause myocardial hypertrophy, thromboembolism, and even
sudden cardiac death [16]. Atrial fibrosis is a common pathological
process of AF, especially long-term persistent AF [17]. Patients
with simple AF experience significant collagen deposition com-
pared to those with normal sinus rhythm [15].
DNMT3A, one DNA methyltransferase, contributes much to
the epigenetic silencing of regulatory genes [18]. In the case of
myocardial fibrosis caused by hypertension and inflammation,
researchers detected the content of type I collagen (Col1A1) in
the heart of isoprenaline-treated rats and found that the inflamma-
tory response increased the expression of platelet-derived growth
factor-BB (PDGF-BB) and Col1A1 [19]. These results suggest that
inflammatory stress increases the accumulation of cardiovascular
vessels and collagen deposits in the heart, leading to myocardial

fibrosis. As an inhibitor of methyltransferase, 5-aza20-deoxycyti-
dine (5-AzadC) causes demethylation and gene resurrection.
5-AzadC normalizes the function of myocardial fibroblasts
in vitro and blocks the activation and proliferation of cardiac fibro-
blasts [20]. Overall, DNA methylation is suspected to affect gene
expression in fibroblasts. 5-AzadC can decrease cell proliferation
and fibroblast activity. In myocardial fibrosis, DNMT3A inhibits
the expression of Ras association domain family 1 isoform A
(RASSF1A) by hypermethylating the promoter region, which may
affect the activity of fibroblasts by affecting the ERK1/2 signaling
pathway [19]. These findings suggest that a decrease in DNMT3A
reduces fibrosis and the activation of cardiac fibroblasts. 5-AzadC
may be potentially used in AF by acting as an inhibitor of methyl-
transferase. The promoter of RASSF1A may also be a target of
precision medicine treatment.
4.3 Hypermethylated SUR1 and SUR2 genes together with Kir6.2 constitute the
Sulfonylurea Receptor ATP-sensitive potassium (KATP) channel [21]. Different subunit
(SUR) 1 in HL-1 isoforms may lead to different physiological characteristics of
Cardiomyocytes KATP. The expressions of SUR1 and SUR2 have not been fully
understood. Based on the identification of CpG island in the pro-
moter region of SUR1 and SUR2 genes, researchers speculated
that DNA methylation may be a regulator of SUR1 and SUR2
expression in cardiac myocytes [21]. The expression of SUR1 was
significantly higher than that of SUR2 in HL-1 cardiomyocytes.
These results of HL-1 cells are similar as atrial myocytes. In addi-
tion, 5-AzadC treatment can significantly increase the unmethy-
lated portion of SUR2 CpG island and the expression of SUR2
mRNA. In disease states caused by environmental or other factors,
abnormal DNA methylation may change the composition of KATP
and thereby alter the electrophysiological function and cardiac
rhythm. Although researchers have found significant role of DNA
methylation in constituting KATP, the result is not atria-specific
and whether it will work in AF has not been determined. More
studies should be carried out to clarify how it works in AF.
4.4 Tumor Necrosis TNF-α, a pro-inflammatory cytokine mainly produced by mono-

Factor-α (TNF-α) nuclear macrophages, is involved in the regulation of inflammatory
Increasing Methylation and immune responses. It is closely related with many cardiovascu-
in Sarcoplasmic lar diseases [22]. SERCA2a plays an important role in Ca2+ homeo-
Reticulum Ca2+ stasis and cardiac function and TNF-α can reduce the expression of
-ATPases (SERCA2a) SERCA2a [23]. Since the promoter region of SERCA2a contains
Promoter CpG islands, DNA methylation should be critical in regulating
SERCA2a. Failing human heart muscles are rich producers of
TNF-α. TNF-α can increase promoter methylation by raising
DNA methyltransferases and enhance methylation of the SERCA2a
promoter region, thereby downregulating SERCA2a. This
downregulation can result in changes in Ca2+ homeostasis, leading

to defects in cardiac systolic and diastolic functions. These findings
suggest that inhibiting hypermethylation may be a novel therapeu-
tic strategy for cardiac dysfunction. Although this result is not
AF-specific, it can also underlie precision treatment of AF.
4.5 Hypermethylated Long non-coding RNAs (lncRNAs) are transcripts in length of

LINC00472 more than 200 nucleotides and have a message-RNA-like structure,
Contributing to AF but lack a specific complete open reading framework. LncRNAs can
regulate the pathogenesis of many heart diseases, but the mechan-
isms regulating lncRNAs expression are still unknown. Reportedly,
the promoter of LINC00472 contains a CpG island, and high
DNA methylation leads to a decrease of LINC00472 [24]. As for
the role of LINC00472 in AF, the elevated level of its direct target,
called microRNA(miR)-24, leads to a downregulation of
Junctophilin-2 (JP2) protein expression [25]. JP2 plays a key role
in maintaining the spatial structure between the cardiac plasma
membrane and the sarcoplasmic reticulum [26]. With the depolar-
ization in cardiomyocytes, a small amount of extracellular Ca2+ can
induce the ryanodine receptor (RyRs) following the calcium-
induced calcium release mechanism, so it can release Ca2+ in sarco-
plasmic reticulum, causing the intracellular calcium concentration
to increase instantly [27]. When the JP2/RyR2 signaling pathway
fails, the normal heart rhythm will be affected and the incidence of
AF will increase. Hypermethylated LINC00472 can regulate the
origin of AF by modulating miR-24 level, affecting the JP2/RyR2
signaling pathway. This finding will contribute to the targeted
therapy of AF.
4.6 DNA Methylation DNA methylation is catalyzed and maintained by DNMTs, which
in Valvular AF are overexpressed in many diseases. The use of DNMT antagonists
can inhibit gene promoter methylation, suggesting that DNMTs
can be a therapeutic target [28]. DNMT1 ensures that the existing
methylation patterns during DNA replication are faithfully repli-
cated to the newly synthesized DNA strand, while DNMT3A and
DNMT3B are primarily responsible for introducing new cytosine
methylation at unmethylated CpG sites. Shen et al. explored the
global DNA methylation level in right atrial myocardial tissues
gathered after valve replacement surgery [29]. They collected
10 tissues from a sinus rhythm group and 10 tissues from an AF
group and found higher DNA methylation level in the AF group
[29]. Furthermore, DNMT3B was significantly increased in AF
patients, which was positively correlated with hypermethylation of
the natriuretic peptide receptor-A gene promoter, indicating
DNMT3B plays a dominant and more important role in the patho-
genesis of AF.
4.7 Genome-wide Abnormal DNA methylation is a common phenomenon in left

DNA Methylation atrial tissues of AF patients [30]. Although DNA methylation
Analysis differs only in a relatively small number of genes, DNA
in Persistent AF methylation-mediated regulation of gene expression in the patho-
genesis of AF may play an important role. In addition, methylation-
sensitive sites, which may be associated with the occurrence and
development of AF, may be therapeutic targets of persistent
AF. The results of genome-wide DNA methylation and gene
expression analysis were integrated to evaluate the effect of methyl-
ation on gene expression regulation [31]. In both datasets, only
20 of the 454 differentially methylated genes were identified. This
undesirable association between DNA methylation and gene
expression may be due to strict statistical setting or because addi-
tional elements are involved in regulating gene expression. Among
them, 20 genes of transcription factor RUNX1 interact most sig-
nificantly with other different methylated genes. RUNX1 activates
and promotes cardiac fibroblast proliferation through
TGF-receptor2, a signal transducer, and transcriptionally activates
epidermal growth factor receptors [32]. The latter two proteins are
downstream signaling molecules of AngII. In addition, RUNX1
inhibits the cardioprotective effects of inappropriate hypertrophy
through AKT serine-threonine kinase 3 and transcription factor
C/EBP [33].
4.8 Limitations Despite the significant findings about DNA methylation and AF,
of Current Research there are still some defects and disadvantages. The research about
on DNA Methylation Pitx2c focuses on HF-induced AF. Whether its result can be used to
and AF other subtypes of AF is unclear. RASSF1A can regulate cardiac
fibrosis, but it is not directly related to AF. Animal AF models
need to be used to further learn about RASSF1A and AF. The
result of SURx is not AF-specific, neither is SERCA2a. The study
about LINC00472 and AF has been demonstrated clearly, but
clinical research is needed to determine the diagnostic value of
LINC00472 or miR-24. The study about DNA methylation and
valvular AF is just omics study and lacks validation in vivo and
in vitro. Many informatics methods have been used in genome-
wide DNA methylation analysis in persistent AF, but these methods
are not based on experimental validation. In the future, DNA
methylation omics, molecular biology experiments, reliable animal
models, and large-scale clinical studies should be integrated to
explore DNA methylation mechanism in AF.
5 Conclusions
Epigenetic mechanisms, especially DNA methylation, play an

important role in AF pathomechanism and refer to non-genomic
changes with phenotypic change. Several genes have been found to
be closely related with AF due to their DNA methylation changes in
promoters, including Pitx2c, RASSF1A, SURs, SERCA2a, and

LINC00472. As DNA methylation can occur throughout lifetime
and in various tissues, it is important to determine the mechanisms
of these changes. More works are needed to determine the poten-
tial therapeutic targets of AF, including basic research and clinical
research. It is also an important part of precision medicine. Impor-
tant questions remain about the mechanisms by which environ-
mental exposure drives or alters DNA methylation. DNA
methylation mediates the expression of AF-related genes and reg-
ulates their biological role in AF. Among these mechanisms, the
processes that induce DNA methylation changes or these changes
can serve as biomarkers for the prognosis and diagnosis of
AF. Precision medicine will provide more precise individualized
treatment at the basis of DNA methylation-regulating mechanisms
in AF.
Acknowledgements
The present study was funded by National Key Research and

Development Program (Grant No. 2018-YFC-1312505 to Yan-
gyang Zhang). Mengwei Lv and Wen Ge contributed equally to
this work.
References
1. Ozturk C, Aparci M, Cakmak T, Metin S, 7. Jung M, Kadam S, Xiong W, Rauch TA, Jin SG,
Balta S, Sen A (2014) Atrial fibrillation pre- Pfeifer GP (2015) MIRA-seq for DNA methyl-
sented with syncope in a jet pilot during daily ation analysis of CpG islands. Epigenomics 7
briefing on squadron. Aviat Space Environ (5):695–706
Med 85(9):965–969 8. Guida F, Sandanger TM, Castagne R et al
2. Canpolat U, Oto A, Yorgun H et al (2015) (2015) Dynamics of smoking-induced
Association of plasma fibronectin level with genome-wide methylation changes with time
left atrial electrical and structural remodelling since smoking cessation. Hum Mol Genet 24
in lone paroxysmal atrial fibrillation: a cross- (8):2349–2359
sectional study. Turk Kardiyoloji Dernegi 9. Fuster V, Ryden LE, Cannom DS et al (2011)
arsivi: Turk Kardiyoloji Derneginin yayin orga- 2011 ACCF/AHA/HRS focused updates
nidir 43(3):259–268 incorporated into the ACC/AHA/ESC 2006
3. Gay MS, Li Y, Xiong F, Lin T, Zhang L (2015) guidelines for the management of patients with
Dexamethasone treatment of newborn rats atrial fibrillation: a report of the American Col-
decreases cardiomyocyte endowment in the lege of Cardiology Foundation/American
developing heart through epigenetic modifica- Heart Association task force on practice guide-
tions. PLoS One 10(4):e0125033 lines developed in partnership with the
4. Antman EM, Loscalzo J (2016) Precision med- European Society of Cardiology and in collab-
icine in cardiology. Nat Rev Cardiol 13 oration with the European heart rhythm asso-
(10):591 ciation and the Heart Rhythm Society. J Am
5. König IR, Fuchs O, Hansen G et al (2017) Coll Cardiol 57(11):e101–e198
What is precision medicine? Eur Respir J 10. Messerschmidt DM, Knowles BB, Solter D
50:1700391 (2014) DNA methylation dynamics during
6. Wang Y, Guan H (2017) The role of DNA epigenetic reprogramming in the germline
methylation in Lens development and cataract and preimplantation embryos. Genes Dev 28
formation. Cell Mol Neurobiol 37 (8):812–828
(6):979–984
11. Palatini P (2004) Parental atrial fibrillation as a chronic heart failure? Circulation 92
risk factor for atrial fibrillation in offspring. (6):1379–1382
JAMA 292(10):1174–1175 23. Kao YH, Chen YC, Cheng CC, Lee TI, Chen
12. Kao YH, Chen YC, Chung CC et al (2013) YJ, Chen SA (2010) Tumor necrosis factor-
Heart failure and angiotensin II modulate atrial alpha decreases sarcoplasmic reticulum Ca2+-
Pitx2c promotor methylation. Clin Exp Phar- ATPase expressions via the promoter methyla-
macol Physiol 40(6):379–384 tion in cardiomyocytes. Crit Care Med 38
13. Mulla W, Hajaj B, Elyagon S et al (2019) Rapid (1):217–222
atrial pacing promotes atrial fibrillation sub- 24. Shen Y, Wang Z, Loo LW et al (2015)
strate in Unanesthetized instrumented rats. LINC00472 expression is regulated by pro-
Front Physiol 10:1218 moter methylation and associated with
14. Chen YJ, Chen YC, Tai CT, Yeh HI, Lin CI, disease-free survival in patients with grade
Chen SA (2006) Angiotensin II and angioten- 2 breast cancer. Breast Cancer Res Treat 154
sin II receptor blocker modulate the arrhyth- (3):473–482
mogenic activity of pulmonary veins. Br J 25. Wang LY, Shen H, Yang Q et al (2019)
Pharmacol 147(1):12–22 LncRNA-LINC00472 contributes to the path-
15. Zhang YJ, Ma N, Su F, Liu H, Mei J (2015) ogenesis of atrial fibrillation (Af) by reducing
Increased TRPM6 expression in atrial fibrilla- expression of JP2 and RyR2 via miR-24.
tion patients contribute to atrial fibrosis. Exp Biomed Pharmacother 120:109364
Mol Pathol 98(3):486–490 26. Padmanabhan A, Haldar SM (2018) Unusual
16. Her AY, Choi EY, Shim CY et al (2012) Pre- transcription factor protects against heart fail-
diction of left atrial fibrosis with speckle track- ure. Science (New York, NY) 362
ing echocardiography in mitral valve disease: a (6421):1359–1360
comparative study with histopathology. Korean 27. Wehrens XH, Lehnart SE, Marks AR (2005)
circulation journal 42(5):311–318 Intracellular calcium release and cardiac dis-
17. Cochet H, Mouries A, Nivet H et al (2015) ease. Annu Rev Physiol 67:69–98
Age, atrial fibrillation, and structural heart dis- 28. Gros C, Fahy J, Halby L et al (2012) DNA
ease are the main determinants of left atrial methylation inhibitors in cancer: recent and
fibrosis detected by delayed-enhanced mag- future approaches. Biochimie 94
netic resonance imaging in a general cardiology (11):2280–2296
population. J Cardiovasc Electrophysiol 26 29. Shen K, Tu T, Yuan Z et al (2017) DNA meth-
(5):484–492 ylation dysregulations in valvular atrial fibrilla-
18. Challen GA, Sun D, Jeong M et al (2011) tion. Clin Cardiol 40(9):686–691
Dnmt3a is essential for hematopoietic stem 30. Tao H, Shi KH, Yang JJ, Li J (2016) Epigenetic
cell differentiation. Nat Genet 44(1):23–31 mechanisms in atrial fibrillation: new insights
19. Tao H, Yang JJ, Chen ZW et al (2014) and future directions. Trends Cardiovasc Med
DNMT3A silencing RASSF1A promotes car- 26(4):306–318
diac fibrosis through upregulation of ERK1/ 31. Zhao G, Zhou J, Gao J et al (2017) Genome-
2. Toxicology 323:42–50 wide DNA methylation analysis in permanent
20. Watson CJ, Collier P, Tea I et al (2014) atrial fibrillation. Mol Med Rep 16
Hypoxia-induced epigenetic modifications are (4):5505–5514
associated with cardiac tissue fibrosis and the 32. Tsai CT, Lai LP, Kuo KT et al (2008) Angio-
development of a myofibroblast-like pheno- tensin II activates signal transducer and activa-
type. Hum Mol Genet 23(8):2176–2188 tors of transcription 3 via Rac1 in atrial
21. Fatima N, Schooley JF Jr, Claycomb WC, myocytes and fibroblasts: implication for the
Flagg TP (2012) Promoter DNA methylation therapeutic effect of statin in atrial structural
regulates murine SUR1 (Abcc8) and SUR2 remodeling. Circulation 117(3):344–355
(Abcc9) expression in HL-1 cardiomyocytes. 33. Taniyama Y, Ito M, Sato K et al (2005) Akt3
PLoS One 7(7):e41533 overexpression in the heart results in progres-
22. Packer M (1995) Is tumor necrosis factor an sion from adaptive to maladaptive hypertrophy.
important neurohormonal mechanism in J Mol Cell Cardiol 38(2):375–385
Chapter 12
PCSK9 Inhibition and Atherosclerosis: Current Therapeutic

Option and Prospection
Pratik Pandey, Cuimei Zhao, and Ban Liu
Abstract
Low-density lipoprotein cholesterol (LDL-C) is a pivotal factor in atherosclerotic cardiovascular disease
(ASCVD), the leading cause of worldwide mortality. The limitations of statin therapy require alternative
treatment strategies to achieve target LDL-C level. Proprotein convertase subtilisin/kexin type 9 (PCSK9)
plays an important role in LDLR recycling, consequently regulating plasma cholesterol levels. Monoclonal
antibodies targeting PCSK9 increased expression of LDLRs at the cell surface and therefore decreased
circulating LDL-C. PCSK9 inhibitors have shown great efficacy in reducing plasma LDL-C levels, which
needs to inject once or twice monthly. Though SPIRE sponsors concern the immunogenicity and terminate
trials early, FOURIER and ODYSSER OUTCOME trials improved the efficacy of PCSK9 inhibitors in
LDL-C reduction. Inclisiran actually is a small interfering RNA (siRNA) developed to inhibit PCSK9
messenger RNA, leading to reduced concentrations of the PCSK9 protein and thereby lower concentra-
tions of LDL-C. Inclisiran is a latest alternative treatment to cholesterol-lowering therapeutics. Twice
injections of inclisiran durably reduced LDL-C levels over 1 year. siRNA therapeutics provided a simple,
novel, and less frequent approach to LDL-C reduction in phase I and II trials, which may be used either as in
combination with statin therapeutics or a stand-alone therapy in the future.
Key words PCSK9 inhibition, Atherosclerosis, Low-density lipoprotein cholesterol, Cardiovascular

disease, Small interfering RNA
1 Introduction
Cardiovascular disease (CVD) is the most common cause of death

all over the world, and atherosclerosis is considered as the primary
cause of the CVD. Atherosclerotic plaques grow when additional
apolipoprotein B (ApoB)-rich particles accumulates. The volume of
the atherosclerotic plaque is mainly affected by the levels of circu-
lating low-density lipoprotein cholesterol (LDL-C) and other
ApoB-rich particles, and by the exposure period of these lipids.
Pratik Pandey and Cuimei Zhao contribute equally to this chapter as co-first authors.
133
134 Pratik Pandey et al.
LDL-C measurement is an estimate of the cholesterol concen-

tration contained by LDL, which is abundantly containing ApoB.
The close relationship between the risk of arteriosclerotic cardio-
vascular disease (ASCVD) and the definite changes in serum
LDL-C has been proved by abundant Mendelian randomization
studies, epidemiological studies, and randomized clinical trials
(RCTs) [1, 2]. The development of coronary artery disease
(CAD) was substantially contributed by hypercholesterolemia,
especial an increase in LDL-C levels. Statins have been used as the
first-line pharmacotherapy for hypercholesterolemia for decades
and have been demonstrated to effectively reduce both LDL-C
levels and ASCVD events [3]. The synthesis of cholesterol is com-
plex and highly regulated. 3-hydroxy-3-methylglutaryl coenzyme
A reductase (HMG CoA-R) is the key limiting enzyme of the
cholesterol synthesis. Statins mainly reduce hepatic cholesterol syn-
thesis by antagonizing the activity of HMG CoA-R, which conse-
quently affects the sterol regulatory element binding protein
(SREBP) pathway and thereafter affects the genetic transcription
of LDL receptor (LDLR) [4]. A number of patients treated by
statin do not achieve the guideline recommended target LDL-C
levels, despite intensive-density statin treatment, and some still
develop to atherosclerosis progression and cardiovascular
events [5].
LDLR expression was reduced by function or elevated concen-
tration of plasma proprotein convertase subtilisin/kexin type
9 (PCSK9), which increases LDLR catabolism in the lysosome
and subsequently increases plasma concentrations of LDL-C
[6]. Recently, proprotein convertase subtilisin/kexin type
9 (PCSK9) regulates the membranal concentration of LDLR.
PCSK9 inhibitors, a new class of lipid-lowering agent, regulate
plasma cholesterol levels [7].
PCSK9 inhibitors were developed by using monoclonal anti-
bodies, which lower concentrations of circulating PCSK9 resulting
in increased LDLR expression at the cellular membrane and there-
fore decreasing circulating LDL-C [6]. Small interfering RNA
(siRNA), a new approach to diminish LDL-C, is developed to
regulate PCSK9 messenger RNA (mRNA). In the post-statin era,
PCSK9 inhibition will dominant the lipid-lowering domain for
their potent cholesterol-lowering efficacy [8], and we will discuss
the mechanism of PCSK9 inhibition and some large clinical rando-
mized trials.
2 PCSK9 and LDLR
In 1999, Varret et al. reported gene mutation identified in a kin-

dred [9], which was different from apoB or LDLR locus, and
linkage analysis suggested a closely related locus at 1p34.1-p32.
PCSK9 Inhibition and Atherosclerosis: Current Therapeutic Option and Prospection 135
In 2003, mutations of PCSK9 gene encoding were firstly identified

[10] as an familial hypercholesterolemia (FH) with autosomal dom-
inant inheritance. After this, other PCSK9 mutations in FH
patients were subsequently studied, treated, and followed up
worldwide [11–13]. After mutations studied in LDLR and apoB,
the PCSK9 mutations is the third locus in prevalence [14]. In the
newest report, variant D374Y in PCSK9 coding incrementally
exceeded the risk of coronary artery disease (CAD) correlating to
the LDLR gene mutation. Recently, many PCSK9 mutations were
reported and all the information can be easily consulted.
The human PCSK9 gene, about 22 kb in length, located on
chromosome 1p32.3, which is composed of signal peptide, predo-
main, catalytic domain, and carboxyl terminal junction connected
successively [15]. The PCSK9 gene is the third locus discovered in
FH, suggesting its importance in cardiovascular diseases (CVD)
besides apoB and LDLR genes. PCSK9 protein synthesis and secre-
tion were similar to other proprotein invertases. PCSK9 proenzyme
(apo-pcsk9,74 kD) was synthesized in the endoplasmic reticulum
and autocatalyze in the endoplasmic reticulum or Golgi body. The
PCSK9 zymogen is autocatalyzed between the locus of Gln152 and
Ser153. After autocatalysis, thse mature PCSK9 (60 kD) is secreted
into the plasma [16].
Statins, the first-line therapy for atherosclerosis presently
decrease the intracellular cholesterol synthesis, which consequently
promotes LDLR expression at the hepatocytic membrane, increas-
ing LDL uptake from the plasma, and thereby decreasing the
concentration of blood LDL-C and other ApoB-rich particles.
LDLR is the pivotal factors in LDL-C metabolism, by its capacity
to bind and subsequently eliminate LDL-C in the circulation. After
LDLR combined to the LDL, the complex was contained by
clathrin-coated pits, and consequently degraded by lysosome. In
this progress, LDLR dissociate from LDL-C according to the acidic
pH value of endosomes and then recycled to the cellular membrane
where it can clear more circulating LDL-C, and the dissociated
cholesterol was used by the cellular. PCSK9, as a pivotal regulator
of LDLR, prevents LDLR from recycling to the membrane [17].
PCSK9 regulated the LDLR protein levels as a serine protease
[18]. PCSK9 is present in human plasma [19, 20], which is majorly
expressed in the hepar, kidney, small intestine, and cerebrum
[10]. Studies demonstrated that PCSK9 contacted with cell surface
LDLR and response to its intracellular degradation [21], which is
consistent with the demonstration that the PCSK9 transcription
mutations decreased cell surface LDLR expression, consequently
increasing serum LDL-C levels.
The progress of PCSK9-inducing LDLR degradation requires
internalization of the PCSK-9 and LDLR, but the exact mechanism
is not clear [19]. Study has suggested that PCSK9 can directly
couple with the cellular surface LDLR [19]; however, the exact
binding domain of PCSK9 to the extracellular locus of the LDLR

has not been identified. ApoB-100 of LDL and apoE of VLDL are
two identified connecting ligands of the LDLR, which directly
interact with the LDLR [22]. The hepatic LDLR levels is reduced
by PCSK9 injecting into the vessel of mice, which suggested
PCSK9 interacted with the cell surface LDLR [19]. PCSK9 was
endocytosed and coimmunofluoresced with the LDLR all along the
endosomal pathway [19].
3 Randomized Clinical Trials of PCSK9 Inhibition
There is relatively small dyslipidemia population diagnosed with

FH. The prevalence rates of homozygous FH and heterozygous
FH are estimated as 1 in 1 million and 1 in 500 people, respectively,
owing to mutations of various gene [23]. However, up to 15% of
patients taking statins are reported to be intolerant of statin
therapy. Even with high-dose satins, only approximately one in
three patients achieved LDL-C target of less than 70 mg/dL
[24]. Till now, no special treatment may prevent statin-induced
muscular problems [25].
Statin treatment upregulated PCSK9 in patients and weakened
the effect of statin, which resulted in the research of PCSK9 inhibi-
tion [26]. PCSK9 inhibitors, including evolocumab, alirocumab,
and bococizumab, are monoclonal antibodies, combining to the
PCSK9 protein and subsequently reducing plasma LDL-C concen-
tration. Evolocumab (marketed by Amgen Inc.) and Alirocumab
(marketed by Sanofi-Aventis) have been both approved firstly by
FDA in 2015. Bococizumab trial (researched by Pfizer) showed
inadequate results of cardiovascular safety, which have been
stopped [27]. Inclisiran, a small interfering RNA (siRNA) mole-
cule, which is developed to reduce the hepatic production of
PCSK9. The mechanism of inclisiran is different from other
PCSK9 inhibitors mentioned above (Table 1).
3.1 FOURIER Trials The Further Cardiovascular Outcomes Research with PCSK9 Inhi-
bition in Subjects with Elevated Risk (FOURIER) trial was a large,
randomized, double-blind, and placebo-controlled trial involving
27,564 ASCVD patients with LDL-C concentrations of 1.8 mmol
per liter (70 mg per deciliter) or higher after receiving statin ther-
apy. This is a multinational clinical trial including 1242 sites in
49 countries [28].
The patients with a history of myocardial infarction (MI), symp-
tomatic peripheral artery disease (PAD), nonhemorrhagic stroke, or
high risk of ASCVD were defined as eligible. Eligible patients were
randomized to accept subcutaneous injection of evolocumab or
matching placebo in a 1:1 ratio. The administration dosage of evolo-
cumab was 140 mg twice monthly or 420 mg once monthly.
Table 1
Trials on PCSK9 inhibition
Comparison LDL-C
and reduction
Administration compared to Primary end points and
Trial Participants Frequency placebo outcome
FOURIER [28] n ¼ 27,564 mAb versus 59% at Major cardiovascular
MI, nonhemorrhagic placebo 48 months events, including
stroke, or symptomatic Once or cardiovascular death,
PAD with additional twice MI, stroke,
higher monthly hospitalization for
cardiovascular risk. unstable angina, or
coronary
revascularization
P < 0.001
ODYSSEY n ¼ 18,924 mAb versus 54.7% at Death from CHD,
OUTCOME ACS placebo 48 months nonfatal MI, fatal or
[30] Once or nonfatal ischemic
twice stroke, or
monthly hospitalization for
unstable angina
P < 0.001
SPIRE1 and 2 n ¼ 27,438 mAb versus 46.8% at Nonfatal MI, nonfatal
Cardiovascular event, or placebo 52 weeks stroke, unstable angina
high-risk ASCVD, Twice requiring urgent
including diabetes, monthly revascularization, or
CKD, or PAD with cardiovascular death.
additional cardiovascular P ¼ 0.08
risk conditions or FH
ORION-1 n ¼ 501 siRNA versus 46.4% at Efficacy, safety, and
ASCVD with LDL-C placebo 360 days tolerability of different
higher than 70 mg/dL Twice yearly 300 mg doses
or ASCVD risk twice __
equivalents, including yearly
type 2 diabetes or FH
with LDL-C higher than
100 mg/dL
MI myocardial infarction, ACS acute coronary syndrome, mAb monoclonal antibody, LDL-C low-density lipoprotein
cholesterol, siRNA small interference RNA, CKD chronic kidney disease, PAD peripheral vascular disease, FH familial
hypercholesterolemia, CHD coronary heart disease
Evolocumab is a human monoclonal antibody (IgG2). Its molecular

weight is approximately 141.8 kDa. The acting onset time is usually in
4 h, and the half-life period is about 11–17 days. Evolocumab is
metabolized through nonsaturable proteolysis [29]. LDL-C levels
were reduced from a median baseline value of 2.4 mmol per liter
(92 mg per deciliter) to 0.78 mmol per liter (30 mg per deciliter) after
48-week evolocumab treatment. The mean percentage reduction was
59% in evolocumab group. Compared to placebo, evolocumab ther-

apy greatly improved the primary end point (95% confidence interval
[CI], 0.79–0.92; hazard ratio, 0.85; P < 0.001) (Table 1) and the
secondary end point (95% CI, 0.73–0.88; hazard ratio, 0.80;
P < 0.001), which were concurrent in the lowest quartile group
with median baseline LDL-C level of 1.9 mmol per liter (74 mg per
deciliter). There is no significant difference reported between the
study groups with respect to adverse events, covering neurocognitive
incidents and newly onset diabetes, while the injection-site reactions
were more commonly reported in evolocumab therapy
(2.1% vs. 1.6%). In FOURIER trial, PCSK9 inhibitor, evolocumab,
was carried out after a certain duration of statin therapy and reduced
LDL-C concentration to a median of 0.78 mmol per liter (30 mg per
deciliter), which decreased the risk of cardiovascular events. This
study demonstrated that ASCVD patients had further benefit by
reducing LDL-C concentration below current recommended
targets [28].
3.2 ODYSSEY The Evaluation of Cardiovascular Outcomes After an Acute Coro-

OUTCOME Trial nary Syndrome During Treatment with Alirocumab (ODYSSEY
OUTCOME) trial is a randomized, multicenter, double-blind,
and placebo-controlled trial. This study involved 18,924 patients
with an acute coronary syndrome (ACS) 1–12 months earlier
[30]. Eligible patients had an LDL-C concentration of 1.8 mmol
per liter (70 mg per deciliter) or higher after treated with a high-
intensity dose of statin or maximum tolerated dose of statin.
Patients were randomized to subcutaneously alirocumab injection
at a dose of 75 mg twice monthly or matching placebo. The
molecular weight of alirocumab is approximate 146 kDa. Alirocu-
mab is also a human monoclonal antibody (IgG1), consisting of a
light chain which disulfide-linked to two heavy chains. The onset of
action is usually 4–6 h, and the half-life period is about 17–20 days.
It forms polypeptides and amino acids by undergoing proteolysis in
many tissues, and the LDL reducing effect is also unrelated to the
injection site [31]. This trial was followed up in a median 2.8 years.
Patients undergone coronary revascularization were 71.8% and
72.7% in alirocumab and control group, respectively. Most patients
received guideline-recommended treatments. Alirocumab declined
LDL-C level from 2.38 0.80 mmol per liter to 1.7 mmol per liter,
after 48-month treatment. In the placebo group, the average
LDL-C concentration changed from baseline 2.38 0.80 mmol
per liter to 2.7 mmol per liter after 48 months. A composite
primary end point event was reduced in the alirocumab group
(9.5%), compared to placebo group (11.1%) (95% confidence inter-
val [CI], 0.78–0.93; hazard ratio, 0.85; P < 0.001). The mortality
rates were 3.5% and 4.1% in the alirocumab and placebo group
(hazard ratio, 0.85; 95% CI, 0.73–0.98), respectively. Patients with
a baseline LDL-C concentration of 2.5 mmol or more per liter had
greater absolute benefit of the compound primary end point, com-

paring those with a lower basic LDL-C. The occurrence of adverse
events presented no difference between alirocumab group and
placebo group, excepting local reactions on injection site were
3.8% and 2.1% in the alirocumab group and the placebo group
(P < 0.001), respectively. This study concluded that ACS patients
receiving combined high-intensity statin and alirocumab therapy
had better primary end point than those who received high-
intensity statin therapy only [30].
3.3 SPIRE Trials Bococizumab is a third humanized monoclonal antibody (IgG2),

remaining 3% of the murine sequence, which is different from
evolocumab and alirocumab mentioned in the FOURIER trial
and ODYSSEY OUTCOME trial. SPIRE1 and SPIRE2 trials
were designated to evaluate cardiovascular outcomes of bococizu-
mab. SPIRE1 and SPIRE2 were parallel, multinational, and rando-
mized trials. The entry criteria of LDL-C concentrations were
higher in SPIRE2 trial than SPIRE1 trial. Comparing patients in
SPIRE1 trial, patients in SPIRE2 trial were included with higher
risk of ASCVD. 27,438 patients were randomized in the combined
trials to receive subcutaneous bococizumab injection at a dose of
150 mg twice monthly or placebo [32]. Molecular weight of boco-
cizumab is approximately 145.1 kDa. Patients with a previous
cardiovascular incidence were enrolled as secondary prevention
cohort, and patients with chronic kidney disease (CKD), diabetes,
or peripheral arterial disease (PAD) or an FH history were enrolled
as high-risk primary prevention cohort. Patients had an LDL-C
level at least 1.8 mmol per liter (70 mg per deciliter) in SPIRE1
trial and at least 2.6 mmol per liter (100 mg per deciliter) in
SPIRE2 trial. At 14 weeks, bococizumab treatment decreased the
mean LDL-C level 64.2% (P < 0.001) from baseline, compared to
+2.9% in the placebo group. The decreasement of LDL-C concen-
tration attenuated over time, and the mean percent change was
41.8% at 52 weeks and 38.3% at 104 weeks of the bococizumab
treatment, which were demonstrated in SPIRE lipid-lowering trials
[33]. There is no significant difference between the treatment
group and control group in SPIRE1 trial on the primary end
point, including nonfatal MI, cardiovascular death, hospitalization
for unstable angina requiring urgent revascularization, or nonfatal
stroke (P ¼ 0.94). In SPIRE2, patients with higher ASCVD risk
had better primary end point in the bococizumab group, compared
to the placebo group (P ¼ 0.02) The incidence of serious adverse
events were similar in both groups, but the injection-site reactions
were more frequent in bococizumab group (10.4%) than in control
group (1.3%) (P < 0.001) [32]. However, a vast proportion of the
patients treated by bococizumab developed high-titer antidrug
antibodies, which substantively attenuated the reduction of LDL-
C levels. In patients without antidrug antibodies, the decreasement

in LDL-C was widely variant after 12-week and 52-week bococi-
zumab therapy [33]. On the basis of the SPIRE trials’ immunore-
activity and outcome, the further development of bococizumab was
ceased [32].
3.4 ORION-1 Trial Inclisiran actually is a small interfering RNA (siRNA) developed to
inhibit PCSK9 mRNA, resulting in reduction of the PCSK9 protein
and thereby reduction of plasma LDL-C. The action mechanism of
inclisiran is different from the use of monoclonal antibodies
approaches to reducing circulating levels of PCSK9. The latter
approach prevents PCSK9 from binding to the LDLR by seques-
tering virtually all PCSK9 in the reticuloendothelial system
[34]. siRNA interferes cellular RNA pathways by binding and
cleaving target mRNA and, in turn, decreasing the target protein
synthesis. siRNA are short RNA molecules with double strand,
naturally occurring in the cell, binding to the RNA-induced silenc-
ing complex (RISC). After the siRNA binding, RISC then induces
mRNA cleavage and degradation by targeting the specified com-
plementary mRNA molecules [35]. In phase I clinical trial, health
volunteers with an LDL-C over 2.6 mmol per liter received a
subcutaneous injection of inclisiran or placebo. No patients
dropped out, due to adverse events. In this study, inclisiran, dosed
with 300 mg or more, reduced the 74.5% LDL-C level from
baseline to day 84 [8]. ORION-1 study is a phase II trial [36]. Eli-
gible patients, including 6% with FH and 69% with established
ASCVD, were randomized to a single dose (200 mg, 300 mg, or
500 mg) of inclisiran or placebo on day 1 (once yearly) or 2 doses
(100 mg, 200 mg, or 300 mg) of inclisiran or placebo on day 1 and
day 90 (twice yearly). All the patients received maximally tolerated
statin therapy and still had elevated LDL-C levels before randomi-
zation. The ORION-1 trial was supposed to discover the reduction
of LDL-C levels from baseline to day 180 and day 360. After
one-year follow-up, the average decreasement in LDL-C in
single-dose group were significant between groups, ranging from
29.5% to 38.7% (P < 0.001), and in two-dose group, ranging from
29.9% to 46.4% (P < 0.001) [36]. Inclisiran is the first and only
cholesterol-lowering siRNA. Inclisiran achieved durable and potent
LDL-C reduction with twice yearly injection in ASCVD patients on
appropriate lipid-lowering therapies over 12 months of follow-up
with a safety profile similar to placebo in a high-risk cardiovascular
population. Therefore, inclisiran potentially offers a novel new
treatment for LDL-C. ORION-1 is a phase II study, and trials of
cardiovascular outcomes will carry on. Phase III trials will provide
safety and efficacy in individuals at high risk of ASCVD, including
established familial hypercholesterolemia and ASCVD, assessing
the importance of inclisiran on cardiovascular outcomes.
4 Conclusion and Prospection
Statins have been recommended as first-line treatment approach for

patients with raised LDL-C and increased ASCVD risk. Trials
involving PCSK9 inhibitors and ezetimibe decreased LDL-C to
the lower level, even far below the current target 1.8 mmol per
liter, and demonstrated extra cardiovascular outcomes benefit. The
nonstatin agents, ezetimibe and PCSK9 inhibitors, adding to back-
ground statin therapy can further benefit cardiovascular outcomes.
With the development of new agents, nonstatin regimens with or
without statin therapy are promising options and should be consid-
ered if statin therapy with maximally tolerated dose does not
achieve target LDL-C concentration. Ezetimibe inhibits intestinal
cholesterol absorption and monotherapy can only decrease LDL-C
levels no more than 20%. The PCSK9 proteins involved in the
dynamics of lipid. PCSK9 inhibitors changed the lipid control in
hypercholesterolemic patients and reduce incidence of ASCVD by
reaching an extremely low LDL-C concentration in blood. PCSK9
inhibitors lower LDL-C very aggressively, and the decreasing of
plasma LDL-C level plays a pivotal role in atherosclerosis develop-
ment through plaque stabilization and regression. Large clinical
trials of these humanized monoclonal antibody have demonstrated
its efficacy, effectiveness, and safety in patients with ASCVD risk
owing to dyslipidemia. PCSK9 inhibitor treatments reduced
all-cause mortality in FOURIER trial and ODYSSEY OUTCOME
trial. Though SPIRE sponsors concern the immunogenicity and
terminate trials early, the immunogenicity of evolocumab and alir-
ocumab is extremely low. Inclisiran, an siRNA interfering PCSK9
mRNA, has also shown its efficacy, effectiveness, and long-term
action as phase II trial. Further studies will be carried out to
evaluate effects on substantial morbidity and mortality benefits.
Acknowledgements
We wish to thank the help in the paper given by Mengwei LV and

Yangyang Zhang.
References
1. Ference BA, Yoo W, Alesh I, Mahajan N, Mir- 2. Silverman MG, Ference BA, Im K, Wiviott SD,
owska KK, Mewada A, Kahn J, Afonso L, Wil- Giugliano RP, Grundy SM, Braunwald E,
liams KA Sr, Flack JM (2012) Effect of long- Sabatine MS (2016) Association between low-
term exposure to lower low-density lipoprotein ering LDL-C and cardiovascular risk reduction
cholesterol beginning early in life on the risk of among different therapeutic interventions: a
coronary heart disease: a Mendelian randomi- systematic review and meta-analysis. JAMA
zation analysis. J Am Coll Cardiol 316:1289–1297
60:2631–2639
3. Baigent C, Keech A, Kearney PM, Blackwell L, 12. Leren TP (2004) Mutations in the PCSK9
Buck G, Pollicino C, Kirby A, Sourjina T, gene in Norwegian subjects with autosomal
Peto R, Collins R, Simes R, Cholesterol Treat- dominant hypercholesterolemia. Clin Genet
ment Trialists’ (CTT) Collaborators (2005) 65:419–422
Efficacy and safety of cholesterol-lowering 13. Naoumova RP, Tosi I, Patel D, Neuwirth C,
treatment: prospective meta-analysis of data Horswell SD, Marais AD, van Heyningen C,
from 90,056 participants in 14 randomized Soutar AK (2005) Severe hypercholesterolemia
trials of statins. Lancet 366:1267–1278 in four British families with the D374Y muta-
4. Seidah NG, Prat A (2012) The biology and tion in the PCSK9 gene: long-term follow-up
therapeutic targeting of the proprotein conver- and treatment response. Arterioscler Thromb
tases. Nat Rev Drug Discov 11:367–383 Vasc Biol 25:2654–2660
5. Nozue T, Yamamoto S, Tohyama S, Fukui K, 14. Humphries SE, Whittall RA, Hubbart CS,
Umezawa S, Onishi Y, Kunishima T, Sato A, Maplebeck S, Cooper JA, Soutar AK,
Nozato T, Miyake S, Takeyama Y, Morino Y, Naoumova R, Thompson GR, Seed M, Dur-
Yamauchi T, Muramatsu T, Hibi K, rington PN (2006) Genetic causes of familial
Terashima M, Michishita I, TRUTH Investiga- hypercholesterolaemia in patients in the UK:
tors (2012) Comparison of arterial remodeling relation to plasma lipid levels and coronary
and changes in plaque composition between heart disease risk. J Med Genet 43:943–949
patients with progression versus regression of 15. Henrich S, Cameron A, Bourenkov GP,
coronary atherosclerosis during statin therapy Kiefersauer R, Huber R, Lindberg I, Bode W,
(from the TRUTH study). Am J Cardiol Than ME (2003) The crystal structure of the
109:1247–1253 proprotein processing proteinase furin explains
6. Norata GD, Tibolla G, Catapano AL (2014) its stringent specificity. Nat Struct Biol
Targeting PCSK9 for hypercholesterolemia. 10:520–526
Annu Rev Pharmacol Toxicol 54:273–293 16. Cunningham D, Danley DE, Geoghegan KF
7. Abifadel M, Varret M, Rabes JP, Allard D, (2007) Structural and biophysical studies of
Ouguerram K, Devillers M, Cruaud C, PCSK9 and its mutants linked to familial hyper-
Benjannet S, Wickham L, Erlich D, Derre A, cholesterolemia. Nat Struct Mol Biol 14
Villeger L, Farnier M, Beucler I, Bruckert E, (5):413–419
Chambaz J, Chanu B, Lecerf JM, Luc G, 17. Benjannet S, Rhainds D, Essalmani R, Mayne J,
Moulin P, Weissenbach J, Prat A, Krempf M, Wickham L, Jin W, Asselin MC, Hamelin J,
Junien C, Seidah NG, Boileau C (2003) Muta- Varret M, Allard D (2004) NARC-1/PCSK9
tions in PCSK9 cause autosomal dominant and its natural mutants: zymogen cleavage and
hypercholesterolemia. Nat Genet 34:154–156 effects on the low density lipoprotein (LDL)
8. Fitzgerald K, White S, Borodovsky A, Betten- receptor and LDL cholesterol. J Biol Chem
court BR, Strahs A, Clausen V, Wijngaard P, 279:48865–48875
Horton JD, Taubel J, Brooks A, Fernando C, 18. Horton JD, Cohen JC, Hobbs HH (2007)
Kauffman RS, Kallend D, Vaishnaw A, Simon A Molecular biology of PCSK9: its role in LDL
(2017) A highly durable RNAi therapeutic metabolism. Trends Biochem Sci 32:71–77
inhibitor of PCSK9. NEJM 376(1):41–51 19. Lagace TA, Curtis DE, Garuti R, McNutt MC,
9. Varret M, Rabès JP, Saint-Jore B, Cenarro A, Park SW, Prather HB, Anderson NN, Ho YK,
Marinoni JC, Civeira F, Devillers M, Hammer RE, Horton JD (2006) Secreted
Krempf M, Coulon M, Thiart R (1999) A PCSK9 decreases LDL receptors in hepatocytes
third major locus for autosomal dominant and livers of parabiotic mice. J Clin Investig
hypercholesterolemia maps to 1p34.1-p32. 116:2995–3005
Am J Hum Genet 64(5):1378–1387 20. Benjannet S, Rhainds D, Hamelin J,
10. Seidah NG, Benjannet S, Wickham L, Nassoury N, Seidah NG (2006) The propro-
Marcinkiewicz J, Jasmin SB, Stifani S, tein convertase (PC) PCSK9 is inactivated by
Basak A, Prat A, Chretien M (2003) The secre- Furin and/or PC5/6A functional conse-
tory proprotein convertase neural apoptosis- quences of natural mutations and post-
regulated convertase 1 (NARC-1): liver regen- translational modifications. J Biol Chem
eration and neuronal differentiation. Proc Natl 281:30561–30572
Acad Sci U S A 100:928–933 21. Maxwell KN, Breslow JL (2004) Adenoviral-
11. Timms KM, Wagner S, Samuels ME, Forbey K, mediated expression of Pcsk9 in mice results in
Goldfine H, Jammulapati S, Skolnick MH, a low-density lipoprotein receptor knockout
Hopkins PN, Hunt SC, Shattuck DM (2004) phenotype. Proc Natl Acad Sci U S A
A mutation in PCSK9 causing autosomaldomi- 101:7100–7105
nant hypercholesterolemia in a Utah pedigree. 22. Russell DW, Brown MS, Goldstein JL (1989)
Hum Genet 114:349–353 Different combinations of cysteine-rich repeats
mediate binding of low density lipoprotein ODYSSEY OUTCOMES Committees and

receptor to two different proteins. J Biol Investigators (2018) Alirocumab and cardio-
Chem 264:21682–21688 vascular outcomes after acute coronary syn-
23. Hovingh GK, Davidson MH, Kastelein JJP, drome. NEJM 379(22):2097–2107
O’Connor AM (2013) Diagnosis and treat- 31. Lunven C, Paehler T, Poitiers F et al (2014) A
ment of familial hypercholesterolaemia. Eur randomized study of the relative pharmacoki-
Heart J 2013(34):962–971 netics, pharmacodynamics, and safety of aliro-
24. Karalis DG, Subramanya RD, Hessen SE, cumab, a fully human monoclonal antibody to
Liu L, Victor MF (2011) Achieving optimal PCSK9, after single subcutaneous administra-
lipid goals in patients with coronary artery dis- tion at three different injection sites in healthy
ease. Am J Cardiol 107:886–890 subjects. Cardiovasc Ther 32(6):297–301
25. Young JM, Florkowski CM, Molyneux SL et al 32. Ridker PM, Revkin J, Amarenco P, Brunell R,
(2007) Effect of coenzyme Q10 supplementa- Curto M, Civeira F, Flather M, Glynn RJ,
tion on simvastatin-induced myalgia. Am J Gregoire J, Jukema JW, Karpov Y, Kastelein
Cardiol 100:1400–1403 JJP, Koenig W, Lorenzatti A, Manga P,
26. Norata GD, Tibolla G, Catapano AL (2014) Masiukiewicz U, Miller M, Mosterd A,
PCSK9 inhibition for the treatment of hyper- Murin J, Nicolau JC, Nissen S, Ponikowski P,
cholesterolemia: promises and emerging chal- Santos RD, Schwartz PF, Soran H, White H,
lenges. Vascul Pharmacol 62:103–111 Wright RS, Vrablik M, Yunis C, Shear CL,
Tardif JC, SPIRE Cardiovascular Outcome
27. Sible AM, Nawarskas JJ, Anderson JR (2016) Investigators (2017) Cardiovascular efficacy
PCSK9 inhibitors: an innovative approach to and safety of bococizumab in high-risk
treating hyperlipidemia. Cardiol Rev patients. NEJM 376(16):1527–1539
24:141–152
33. Ridker PM, Tardif J-C, Amarenco P et al
28. Sabatine MS, Giugliano R, Keech AC, (2017) Lipid-reduction variability and
Honarpour N, Wiviott SD, Murphy SA, antidrug-antibody formation with bococizu-
Kuder JF, Wang H, Liu T, Wasserman SM, mab. NEJM 376(16):1517–1526
Sever PS, Pedersen TR, FOURIER Steering
Committee and Investigators (2017) Evolocu- 34. Watts GF, Chan D, Dent R et al (2017) Facto-
mab and clinical outcomes in patients with car- rial effects of evolocumab and atorvastatin on
diovascular disease. NEJM 376 lipoprotein metabolism. Circulation
(18):1713–1722 135:338–351
29. Cicero AFG, Colletti A, Borghi C (2015) Pro- 35. Carthew RW, Sontheimer E (2009) Origins
file of evolocumab and its potential in the treat- and mechanisms of miRNAs and siRNAs. Cell
ment of hyperlipidemia. Drug Des Devel Ther 136(4):642–655
9:3073–3082 36. Ray KK, Stoekenbroek R, Kallend D,
30. Schwartz GG, Steg PG, Szarek M, Bhatt DL, Nishikido T, Leiter LA, Landmesser U, Wright
Bittner VA, Diaz R, Edelberg JM, Goodman RS, Wijngaard PLJ, Kastelein JJP (2019) Effect
SG, Hanotin C, Harrington RA, Jukema JW, of 1 or 2 doses of inclisiran on low-density
Lecorps G, Mahaffey KW, Moryusef A, lipoprotein cholesterol levels: one-year follow-
Pordy R, Quintero K, Roe MT, Sasiela WJ, up of the ORION-1 randomized clinical trial.
Tamby JF, Tricoci P, White HD, Zeiher AM, JAMA Cardiol 4(11):1067–1075
Chapter 13
Precise Drug Sequential Therapy Can Improve

the Cardioversion Rate of Atrial Fibrillation with Valvular
Disease after Radiofrequency Ablation
Tao Li and Yongjun Qian
Abstract
Objective: Based on pathogenesis of atrial fibrillation (AF), investigate the effects of precision drugs
continuous therapy on AF cardioversion rate after radiofrequency catheter ablation. Methods: We included
1334 patients who underwent mitral valve replacement with bipolar radiofrequency ablation due to mitral
valve disease with AF during June 2011 to July 2017. The data of clinical and related laboratory examina-
tions at discharge and follow-up were recorded. All patients were treated with or without angiotensin-
converting enzyme inhibitor (ACEI) and angiotensin II-receptor blocker (ARB) drugs according to their
conditions and doctor’s willingness. The heart rhythm was evaluated after treatment and follow-up of
6 months. Results: All 1162 cases were followed up, including 825 cases in mitral stenosis (MS) group,
337 cases in mitral regurgitation (MR) group. In MS group, left atrial diameter(LAD) and left ventricular
diameter(LVD) of the patients taking ACEI and ARB were significantly lower (P < 0.05), and they can
increase AF cardioversion rate from 79.1% of the control group to 83.7% and 82.8%, respectively (P ¼ 0.03
and 0.04). In MR group, the patients with ACEI compared with control group, there were no significant
differences in LAD, LVD, right atrial diameter (RAD), right ventricular diameter (RVD), left ventricular
ejection fraction(LVEF), and left ventricular fractional shortening(LVFS) (P > 0.05); but ARB group,
LAD, LVD decreased significantly (P < 0.05). And ACEI can increase AF cardioversion rate from 76.1% in
the control group to 77.2% (P ¼ 0.62), ARB to 81.6% (P ¼ 0.02). Conclusion: It does improve AF
cardioversion rate after radiofrequency catheter ablation that the precise anti-structural remodeling drugs
continuous therapy was adopted based on the pathogenesis of AF.
Key words Atrial fibrillation, Pathogenesis, Radiofrequency catheter ablation, Drug therapy, Struc-
tural remodeling
1 Introduction
Atrial fibrillation (AF) is one of the most common arrhythmias in

the clinic, and the risk of developing AF is one in four for patients
aged 40 years and above [1]. Valvular heart disease is often asso-
ciated with AF, especially in patients with mitral valve disease.
About 64% of mitral valve patients suffer from AF which increases
the risk of stroke, heart failure, and death [2–5]. Current
145
146 Tao Li and Yongjun Qian
treatments for AF include two methods: (1) controlling ventricular

rate (i.e., rate control) without trying to stop or prevent AF;
(2) trying to obtain and maintain sinus rhythm (i.e., rhythm con-
trol) [6]. Although large-scale randomized trials have not shown
that rhythm is better than rate control [7], it is generally believed
that the failure to benefit from more circadian rhythm control is
due to the limited and poor efficacy of antiarrhythmic drugs main-
tained by sinus rhythm reaction. Radiofrequency ablation is now
widely used to treat AF of valve disease in clinical practice, the rate
of maintenance of sinus rhythm is about 75–85% after 6 months,
only 62.3% reported recently [3–5, 8]. The maintenance rate of
sinus rhythm still falls far short of what doctors and patients
expects. Can the rate of AF conversion be further increased?
Through a series of studies of the AF pathogenesis, our
research group found that the structural remodeling of AF related
to valvular heart disease, and that the degree of structural remodel-
ing in different types of mitral valve lesions was different, and that
the mechanism of AF structural remodeling induced by Renin-
angiotensin-aldosterone system (RAAS) was not the same [9–
12]. Therefore, we assume that: (1) After the radiofrequency abla-
tion of AF relating to valvular heart disease, the treatment with
continuous anti-structural remodeling drugs may further improve
the cardioversion rate of AF; (2) With regard to different types of
mitral valve disease associated with AF, using different kinds of anti-
structural remodeling drugs for precise treatment after radiofre-
quency ablation may have different effects of AF cardioversion.
2 Data and Methods
2.1 Subjects There was a total of 1334 patients who underwent mitral valve
and Groups replacement with bipolar radiofrequency ablation in virtue of mitral
valve disease with AF from West China Hospital Sub-Database of
Chinese Adult Cardiac Surgical Database from June 2011 to
July 2017.
Inclusion criteria: (1) The MS or MR patients who needed
mitral valve replacement plus concomitant AF radiofrequency abla-
tion, and whose aortic valve is mild below or without stenosis or
regurgitation and does not require surgical intervention, and whose
tricuspid valves are moderate or less regurgitation or without val-
vuloplasty; (2) The patients who completed follow-up of the out-
patient department or telephone for 6 months and persisted in
using anti-structural remodeling-related drugs; (3) The patients
who were no complications of anticoagulation, membrane dysfunc-
tion, leakage of the valve, tricuspid moderate or more regurgita-
tion, and surgery-related cardiac.
Exclusion criteria: (1) The patients who existed two degree or
more conduction block or implanted permanent pacemaker after
operation; (2) The patients with aortic valve lesions of moderate or
Precise Drug Sequential Therapy Can Improve the Cardioversion Rate. . . 147
The patients of meeting the inclusion criteria from West

China Hospital Sub-Database of Chinese Adult Cardiac
Surgical Database from June 2011 to July 2017
(1334 cases)
During follow-up, patients who failed

to complete drug treatment, died and
did not complete 6 months follow-up
(-172 cases)
87% of patients who completed treatment or did not undergo anti-structural

remodeling treatment completed follow-up as required (1662 cases)
MS Group MR Group
(825 cases) (337 cases)
ACEI Group ARB Group Control Group ACEI Group ARB Group Control Group
(256 cases) (91 cases) (479 cases) (108 cases) (67 cases) (162 cases)
Fig. 1 Group of included patients diagram
more or requiring surgical treatment; (3) The patients with tricus-

pid moderate or more regurgitation or no tricuspid valvuloplasty
and requiring tricuspid valve replacement; (4) Other cases, such as
preoperative infective endocarditis, hyperthyroidism, hypertensive
heart disease, hyperlipidemia, coronary heart disease and diabetes,
an AF duration of less than half a year, complications of antic-
oagulation, and re-admission during the follow-up period.
Grouping: First, the patients were divided into two groups
(group MS and group MR) according to the valve lesion, and
each group was divided into three groups according to the purpose
of the study, and anti-structural remodeling drugs used by the
patients: group ACEI (take ACEI), group ARB (take ARB), and
control group (unused any anti-structural remodeling drugs).
There are six groups, as shown in Fig. 1.
2.2 Methods The data of clinical and related laboratory examinations at dis-
charge and follow-up were recorded, including gender, age, dura-
2.2.1 Collection Clinical
tion of atrial fibrillation, electrocardiogram, and heart color
Material
Doppler ultrasound data, etc.
The mitral valve replacement was performed by routine ster-
notomy, median incision, extracorporeal circulation, mechanical
valve replacement, and warfarin anticoagulation for a lifetime,
with a strength of INR maintained at 1.5–2.5.
Radiofrequency catheter ablation of atrial fibrillation during

mechanical mitral valve replacement was performed using the Med-
tronic bipolar radiofrequency ablation clamp in accordance with the
standard COX-IV Operation line.
2.2.2 Medication All patients underwent cardiac ultrasonography at discharge and

and Follow-up were treated with or without ACEI and ARB drugs according to
their condition and doctor’s willingness. ACEI was prescribed
Capoten 12.5 mg, once daily, and ARB was 75 mg, once a day.
Other drugs that do not affect the research purposes can be used
routinely to regulate heart function and cardiac rhythm. The
patients needed to see a doctor at outpatient service after operation
every month to adjust the oral medication and monitor blood
coagulate functions and understand the heart rate and rhythm of
the patient. The dosage of amiodarone, metoprolol, and other
drugs were adjusted according to the ventricular rate and heart
rate of the patients. Clinicians recorded the duration of ACEI and
ARB at outpatient department and observed side effects of medica-
tions such as cough and pruritus to assess whether patients need to
adjust the dosages of ACEI and ARB or discontinue ACEI and
ARB medications.
2.2.3 Assessment Eligible patients were recommended for 24 h Holter, but the
of Sinus Rhythm cardiac rhythm of the patients after 6 months was also assessed
comprehensively by monthly electrocardiogram and cardiac ultra-
sound according to the current situation of clinical work and the
feasibility of large sample data acquisition. AF was diagnosed
according as all the 12 lead ECG duration 10 s of atrial fibrillation
rhythm. Comprehensive evaluation was completed independently
by three different professional physicians from departments of
cardiac ultrasonography, internal medicine-cardiovascular and car-
diac surgery, and only three doctors have determined that the
patient had sinus rhythm to assess the patient’s sinus rhythm.
2.3 Statistical In this study, the data were expressed as mean standard deviation
Analysis (X s), the number of cases, age, AF duration, LAD, LVD, RA,
RVD, LVEF, and LVFS between groups of patients were compared
by t test, comparison of sex between two groups of patients using
X2 test. Statistical analysis was performed with SPSS Ver17.0 statis-
tical software, p < 0.05 was considered statistically significant.
3 Results
3.1 Patient Grouping During the study, west China hospital data of Chinese adult cardiac
Results surgical database, a total of 1334 cases of patients with mitral valve
disease with AF underwent mitral valve replacement concomitant
radiofrequency ablation. During the 6 months of follow-up, there
ALGrawany
were a total of 172 patients of failing to adhere to the completion of

drug treatment, lung infection, and other re-admission, again car-
diac surgery or other surgery, electric cardioversion in 6 months,
death and incomplete 6 months follow-up, etc. 87% of patients, a
total of 1162 cases, were followed up, including 825 cases of MS
group, 337 cases of MR group. Each group was further divided
into three groups according to the use of anti-structural recon-
struction drugs of different types and unused drugs, and the num-
ber of groups and cases in each group was shown in Fig. 1.
3.2 The Comparisons There was no statistically significant difference in age, gender, AF
of General Clinical duration, LAD, LVD, RAD, RVD, LVEF, and LVFS in the three
Data at the Time groups (P > 0.05), as shown in Table 1.
of Discharge and AF
Cardioversion Using
Different
Anti-structural Drugs
of Three Groups
of Patients
in MS Group
3.2.1 The Comparison
of General Clinical Data
of Three Groups of Patients
in MS Group at the Time
of Discharge
3.2.2 The Comparison With ACEI and ARB Group, LAD and LVD were significantly
of Cardiac Color Doppler reduced, the difference was statistically significant (P < 0.05), but
Data Compared there was no statistically significant difference between RAD, RVD,
with the Control Group LVEF, and LVFS (P > 0.05), as shown in Fig. 2.
in Three Groups of Patients Both ACEI and ARB can significantly reduce the size of left
with MS Group After atrial and left ventricle in MS patients, while the improvement of
6 Months Follow-up right atrial, right ventricle, LVEF, and LVFS is not obvious. ACEI
group and ARB group in the improvement of left atrial and left
ventricle were equivalent.
3.2.3 The Comparison Different drugs can improve the cardioversion rate of AF in patients
of Effects of ACEI and ARB with MS, but the degree of improvement is different. ACEI and
on the Rate of AF ARB were able to increase the cardioversion rate from 79.1% in the
Conversion in MS Group control group to 83.7% and 82.8%, respectively (P ¼ 0.03 and
Using Different P ¼ 0.04), and the difference was statistically significant, while
Anti-structural the comparison of the cardioversion of AF between ACEI and
Remodeling Drugs ARB was not statistically significant in 83.7 vs. 82.8% (P ¼ 0.21),
as shown in Fig. 3.
Can be seen in MS patients, ACEI and ARB can significantly
improve the cardioversion rate of AF, while the effect between the
two was no difference.
Table 1
Comparison of general clinical data of three groups of patients in MS group at the time of discharge
(X s)
ACEI group ARB group Control group

(N ¼ 256) (N ¼ 91) (N ¼ 479) P
Age (years) 51.3 11.2 50.9 13.2 52.1 12.6 0.64
Male/female 105/151 42/49 192/287 0.31
AF duration (months) 11.1 4.5 13.6 4.8 12.7 6.1 0.42
LAD(mm) 53.6 12.1 52.9 10.7 52.3 15.4 0.64
LVD(mm) 44.4 6.3 48.6 9.5 46.1 7.8 0.24
RAD(mm) 48.9 14.2 47.6 11.4 49.2 14.9 0.78
RVD(mm) 24.5 5.6 27.3 7.9 25.6 5.7 0.66
LVEF(%) 58.7 9.3 57.6 7.3 56.8 6.2 0.56
LVFS(%) 33.1 4.3 34.6 6.7 35.8 5.8 0.34
60
55
50
45
40
35
30
25
20
LAD (mm) LVD (mm) RAD (mm) RVD (mm) LVEF(%) LVFS(%)
ACEI Group 45.8 40.5 42.1 24.7 53.9 33.5
ARB Group 44.4 40.3 41.3 23.1 53.4 31.9
Control Group 52.9 47.8 44.7 22.8 55.1 34.7
Fig. 2 Comparison of cardiac color Doppler data in three groups of patients with MS Group after 6 months
follow-up (mean)
85.0%
P=0.03
84.0%
P=0.21 P=0.04
83.0%
82.0%
AF Conversion Rate
81.0%
80.0%
83.7%
82.8%
79.0%
78.0%
79.1%
77.0%
76.0%
ACEI Group ARB Group Control Group
AF Conversion Rate 83.7% 82.8% 79.1%
Fig. 3 Comparison of effects of ACEI and ARB on the cardioversion rate of AF in MS group using different anti-
structural remodeling drugs
3.3 The Comparisons There was no statistically significant difference in age, gender, AF
of General Clinical duration, LAD, LVD, RAD, RVD, LVEF, and LVFS in the three
Data at the Time groups (P > 0.05), as shown in Table 2.
of Discharge and AF
Cardioversion Using
Different
Anti-structural Drugs
of Patients
in MR Group
3.3.1 The Comparison
of General Clinical Data
of Three Groups of Patients
in MR Group at the Time
of Discharge
3.3.2 The Comparison Compared with the control group, the differences of LAD, LVD,
of Cardiac Color Doppler RAD, RVD, LVEF, and LVFS were statistically insignificant in
Data Compared patients with ACEI (P > 0.05), while in patients with ARB
with the Control Group group, LAD and LVD were significantly reduced, the difference
in Three Groups of Patients was statistically significant (P < 0.05), but there was no statistically
with MR Group After significant difference between RAD, RVD, LVEF, and LVFS
6 Months Follow-up (P > 0.05). Compared with the patients in ARB group, the LAD
and LVD of patients in ACEI group were significantly reduced and
the difference was statistically significant (P < 0.05), but there was
Table 2
Comparison of general clinical data of three groups of patients in MR group at the time of discharge
(X s)
ACEI group ARB group Control group

(N ¼ 108) (N ¼ 67) (N ¼ 162) P
Age (years) 56.1 12.2 54.4 13.6 55.3 10.7 0.71
Male/female 59/49 36/31 89/73 0.45
AF duration (months) 10.9 6.8 12.6 5.7 11.2 9.2 0.56
LAD(mm) 55.9 14.2 54.7 9.8 55.3 10.7 0.47
LVD(mm) 42.5 9.1 46.3 10.5 44.4 8.4 0.43
RAD(mm) 47.1 9.7 44.7 14.5 46.5 12.6 0.66
RVD(mm) 25.6 6.7 28.5 9.1 23.6 8.8 0.69
LVEF(%) 55.5 10.3 53.6 8.4 54.8 9.7 0.91
LVFS(%) 31.1 5.5 32.5 4.9 34.2 6.7 0.45
60
55
50
45
40
35
30
25
20
LAD(mm) LVD(mm) RAD(mm) RVD(mm) LVEF(%) LVFS(%)
ACEI Group 49.6 46.3 42.2 27.6 55.9 33.5
ARB Group 42.1 40.4 44.1 25.5 57.8 32.1
Control Group 48.7 45.8 43.4 24.7 56.9 35.4
Fig. 4 Comparison of cardiac color Doppler data in three groups of patients with MR group after 6 months
follow-up (mean)
no statistically significant difference between RAD, RVD, LVEF,

and LVFS (P > 0.05), as shown in Fig. 4.
It can be seen that the long-term use of ACEI did not improve
the left atrium, left ventricle, right atrium, and right ventricular of
patients with MR and did not improve LVEF and LVFS. ARB
significantly improved the left atrium and left ventricle of MR
patients, while the right atrium, right ventricle, LVEF, and LVFS
were not improved significantly. The improvement of left atrium
and left ventricular in ARB group was superior to ACEI in patients
83.0%
P=0.62
82.0%
P=0.03 P=0.02
81.0%
80.0%
AF Conversion Rate
79.0%
78.0%
81.6%
77.0%
76.0%
75.0% 77.2%
76.1%
74.0%
73.0% ACEI Group ARB Group Control Group
AF Conversion Rate 77.2% 81.6% 76.1%
Fig. 5 Comparison of effects of ACEI and ARB on the rate of AF conversion in MR group using different anti-
structural remodeling drugs
with MR, but there was no significant difference in other indicators

of cardiac color Doppler ultrasonography.
3.3.3 The Comparison Different drugs can improve the conversion rate of AF in patients
of Effects of ACEI and ARB with MR, but the degree of improvement is also different. ACEI
on the Rate of AF can increase the conversion rate of 76.1% in the control group to
Conversion in MR Group 77.2% and P ¼ 0.62, but the difference was not statistically signifi-
Using Different cant. ARB could increase the conversion rate of 76.1% in the
Anti-structural control group to 81.6%, P ¼ 0.02, the difference was statistically
Remodeling Drugs significant; ACEI group and ARB group between the AF conver-
sion rate was 81.6 vs. 77.2%, P ¼ 0.03 difference was statistically
significant, see Fig. 5.
Obviously, in patients with MR, ACEI can also improve the AF
conversion rate, but there is no statistically difference with the
control group, and ARB can obviously improve the rate of AF
cardioversion.
4 Discussion
At present, more and more patients with AF, the goal of treatment
of AF is to improve the quality of life to prevent complications and
death [6]. AF treatment strategies include heart rate control and
cardiac rhythm control, the former can improve the quality of life of
patients, but cannot improve the effective prevention of complica-
tions and death, while cardiac rhythm control is the high goal of
atrial fibrillation treatment. Recently, cardiac surgery at the same

time radiofrequency ablation of AF is the most important method
of AF treatment to control cardiac rhythm. However, recurrence of
atrial fibrillation after radiofrequency ablation is quite common,
resulting in poor clinical treatment satisfaction of patients. The
recurrence rate of paroxysmal atrial fibrillation with the fastest
response can reach 60% within 1.5 years [13, 14]. In addition,
long-term recurrence is frequent even after initial successful sur-
gery, possibly due to underlying disease progression [15]. Further
improvement of the AF conversion rate is the bottleneck of AF
therapy. Breakthrough current status of using IC drugs in the same
period of pulmonary vein isolation, individualized treatment for
atrial fibrillation occurrence and maintenance mechanism is the
direction of AF treatment in the next 20 years [16, 17].
Heart rate control is an indispensable part of the treatment of
patients with atrial fibrillation, which can usually relieve the symp-
toms related to atrial fibrillation. Compared with stroke prevention
and rhythm control, there is little conclusive evidence of the opti-
mal type and intensity of rate control therapy, with most of the data
coming from short-term crossover trials and observational studies
[18, 19]. Pharmacological control of acute or long-term frequency
can be achieved by using β-blockers, digoxin, calcium channel
blockers diltiazem, and verapamil or combination therapy. Some
antiarrhythmic drugs also reduce heart rate (amiodarone, tenidar-
one, sotalol and, to some extent, propafenone), but they can only
be used in patients who need rhythm control therapy [19]. How-
ever, clinical studies have shown that there is no significant differ-
ence in a series of clinical events, cardiac function classification, and
hospitalization between strict heart rate control and loose heart rate
control. It is also worth noting that many patients with “adequately
controlled heart rate” (resting heart rate 60–100 bpm) have severe
symptoms and need further management [20].
The success of rhythm control therapy depends on a variety of
factors, including the number, type, and severity of underlying
disease, age, sex, compliance with antiarrhythmic drug therapy,
and factors related to the quality of atrial fibrillation ablation sur-
gery [21]. Amiodarone is more effective in maintaining sinus
rhythm than other antiarrhythmic drugs [19], but antiarrhythmic
drugs (AAD) are associated with a variety of potential adverse
reactions. A recent Cochrane collaborative analysis examined the
outcomes of AAD to maintain sinus rhythm after cardioversion.
The results of blind trials showed that the total mortality rate of
AAD tended to be higher, and the adverse event withdrawal rate of
almost all drugs was higher than that of the control group [22].
At present, it is considered that the occurrence and mainte-
nance of AF are closely related to atrial remodeling, which mainly
includes atrial electrical remodeling and atrial structural remodeling
[6]. The short-term atrial electrical remodeling after continuous AF
cardioversion can disappear completely, but its structural remodel-

ing still exists, while AF is still easy to recur at this time
[23, 24]. This phenomenon suggests that atrial structural remodel-
ing rather than atrial electrical remodeling may be the key factor in
the occurrence and maintenance of AF. Atrial structural remodeling
slows localized conduction and increases conduction heterogeneity.
Abnormal conduction provides conditions for unidirectional con-
duction block and reentry, which provide the basis for AF, and even
in a small area of myocardial structural remodeling can occur AF
[25]. Komatsu et al. found that the use of drugs to prevent atrial
remodeling can not only increase the effect of AF electrocardio-
gram, but also prevent the recurrence of AF [26]. Our research
team has noticed the importance of structural remodeling of AF
and took the lead in using drugs for structural remodeling (vitamin
c + captopril + simvastatin) to cardioverter 49 patients with AF after
mitral valve replacement in outpatient clinic. Compared with the
traditional cardioversion therapy, this method reduced the adverse
drug reactions and achieved a certain clinical effect. About 35% of
the patients recovered sinus rhythm, while only 6% of the control
group recovered sinus rhythm, suggesting that blocking structural
remodeling therapy for AF is effective [27].
At home and abroad, our research group took the lead in a
series of studies on pathogenesis and maintenance mechanism of
AF in different types of valvular diseases. Through the clinical
observation, we found that mitral stenosis and mitral regurgitation
with AF has the different incidence, and that mitral stenosis is more
prone to AF, about 64% patients with AF, and that mitral regurgi-
tation is associated with less AF. Through a series of studies, we
confirmed that there is a beautiful circle in the pathogenesis of
valvular disease with AF, as shown in Fig. 6 [9]. The circle mainly
consists of three parts: (1) The left part of the circle proved that
different types of mitral valve disease show different pathogenesis of
AF, MS patients with AF are associated with angiotensin-
converting enzyme (ACE) and angiotensin II (Ang II), whereas
in MR patients with AF only associated with Ang II [10]. (2) The
right part of the circle proved that there is structural remodeling in
valvular disease with AF, and that left atrial structural remodeling is
not the same in patients with different types of mitral valve disease.
Compared with MR patients, the remodeling of left atrial structure
in MS patients is more pronounced [12]. On the basis of the above
two studies, the key step of the research team to complete this circle
is to demonstrate that Ang II and ACE are connected to RAAS and
structural remodeling, and ultimately point out that there is a
beautiful circle in the pathogenesis of valvular disease with AF
from RAAS to structural reconstruction [9]. A series of pathogene-
sis studies suggest that preventing the structural remodeling of AF
may further improve the conversion rate of AF. Different types of
valvular disease need to use different drugs, which is the theoretical
Fig. 6 The beautiful circle in the pathogenesis of valvular disease with AF from RAAS to structural remodeling
basis of this study. To the best of our knowledge, this study is the
largest sample of retrospective studies on the choices of using
structural remodeling drug therapy after AF radiofrequency abla-
tion for the pathogenesis.
The preliminary basic studies determine the type of drug
selected in this study. The studies found that the pathogenesis of
AF was different in different types of mitral valve lesions, and AF
was most related to with ACE and AngII in MS patients, while AF
was only associated with Ang II in MR patients [10]. Therefore,
this study selected ACEI or ARB drugs to intervene AF. The drugs
have been commonly used in cardiovascular clinical drugs because
of the use of safety and less side effects. And follow-up found that
the main side effects are cough and skin itching, etc., which often
can resume medication after suspended for a week, while a small
number of patients need to stop continuing medication. In addi-
tion, this study using ACEI or ARB dose is small, which has less
impact on patients with blood pressure, and there is no case of
stopping medication due to obvious fluctuations in blood pressure.
Because of the low side effects of the drugs in this study and the low
incidence rate and its large sample size, so the study expels the
patients with intermittent medication or terminate the drug.
This study found that ACEI and ARB can significantly improve
the conversion rate of AF in MS patients, while the effect between
the two was no difference, and that in patients with MR, ARB can
obviously improve the rate of AF cardioversion, while ACEI can
also improve the AF conversion rate, but there is no statistically
difference with the control group. The results of this clinical study
coincide with previous basic studies, which further prove that AF is

associated with ACE and AngII in patients with MS, and using
ACEI or ARB can further improve the cardioversion rate of AF, and
that AF only with Ang II in MR patients, and only ARB can further
improve the cardioversion rate of AF, while the effect of ACEI is
not obvious.
The indexes of evaluating structure remodeling were selected
from the heart color Doppler ultrasound, mainly due to previous
studies that found that cardiac structural remodeling mainly man-
ifested as myocardial ultrastructural changes, myocardial type I and
type III fibrillary collagen changes, and changes in cardiac size
[12]. Myocardial ultrastructural changes and fibrillary collagen
changes are invasive, difficult to achieve in clinical follow-up, and
they are not conducive to large sample size study, so this study
selected the heart size provided by cardiac color Doppler ultra-
sound to evaluate structural remodeling. It is suggested by the
cardiac color Doppler ultrasound data that in patients with MS,
whether ACEI or ARB intervention can improve the left atrial and
left ventricular as the representative of the structural remodeling,
whereas in MR patients only ARB intervention can improve the left
atrial and left ventricular structure remodeling, and ACEI interven-
tion could not improve its structural remodeling. The structural
remodeling of different drug treatments is consistent with its
corresponding cardioversion rate of AF, which sufficiently demon-
strates that the series of basic studies of AF structural remodeling
can guide the clinical treatment of AF.
Researches showed that after the successful cardioversion of
radiofrequency ablation of persistent AF, the atrial electrical remo-
deling can be completely stopped in a short time, but the myocar-
dial structural remodeling is still ongoing, and at the same time the
AF is still easy to relapse [24]. It can be seen that the structural
remodeling of the heart is the material basis of AF recurrence, not
the electrical remodeling. Based on the studies of AF pathogenesis,
our research team continues to deepen and think about whether
the use of structural remodeling drugs after the cessation of electri-
cal remodeling in radiofrequency ablation of valvular disease with
AF. Although a study has shown that RAAS blockers can reduce
new onset AF of 21%, while ARB can reduce 22%, but the study is
significantly different from ours because of its too general subjects
and AF patients with the miscellaneous underlying disease without
radiofrequency ablation [16]. Based on the study of the pathogen-
esis of AF, this study only chooses ACEI and ARB in RAAS blockers
for accurate targeted therapy and also carries out fine grouping of
subjects. From the results of the study, combined with the different
underlying diseases and different pathogenesis of AF, further appli-
cation of precision drug continuous therapy of AF after radiofre-
quency ablation is necessary and effective. This scheme, adopting
the individualized “precision treatment“according to the study of
fine pathogenesis of AF, can further improve the cardioversion rate

on the basis of radiofrequency ablation of AF.
In a word, there is still structural remodeling after radiofre-
quency ablation of valvular disease with AF, and different types of
valvular lesion have different pathogenesis of AF, different drugs,
and different effects. Individualized, precise anti-structural remo-
deling drug continuous therapy can further improve the cardiover-
sion rate of AF after radiofrequency ablation.
References
1. Camm AJ, Savelieva I, Potpara T et al (2016) disease. Chin J Clin Thorac Cardiovasc Surg
The changing circumstance of atrial fibrillation 39(11):674–677
- progress towards precision medicine. J Intern 12. Qian Y, Meng J, Tang H et al (2010) Different
Med 279(5):412–427 structural remodelling in atrial fibrillation with
2. Pitt B, Remme W, Zannad F et al (2003) Epler- different types of mitral valvular diseases. Euro-
enone, a selective aldosterone blocker, in pace 12(3):371–377
patients with left ventricular dysfunction after 13. Macle L, Khairy P, Weerasooriya R et al (2015)
myocardial infarction. N Engl J Med 348 Adenosine-guided pulmonary vein isolation for
(14):1309–1321 the treatment of paroxysmal atrial fibrillation:
3. Schaff HV (2015) Surgical ablation of atrial an international, multicentre, randomised
fibrillation--when, why, and how? N Engl J superiority trial. Lancet 386(9994):672–679
Med 372(15):1465–1467 14. Verma A, Jiang CY, Betts TR et al (2015)
4. Gillinov AM, Gelijns AC, Parides MK et al Approaches to catheter ablation for persistent
(2015) Surgical ablation of atrial fibrillation atrial fibrillation. N Engl J Med 372
during mitral-valve surgery. N Engl J Med (19):1812–1822
372(15):1399–1409 15. Schreiber D, Rostock T, Frohlich M et al
5. Zhu X, Li Q, Li Y et al (2016) Analysis of (2015) Five-year follow-up after catheter abla-
bipolar radiofrequency ablation in treatment tion of persistent atrial fibrillation using the
of atrial fibrillation associated with rheumatic stepwise approach and prognostic factors for
heart disease. PLoS One 11(3):e0151248 success. Circ Arrhythm Electrophysiol 8
6. Heijman J, Guichard JB, Dobrev D et al (2):308–317
(2018) Translational challenges in atrial fibril- 16. Khatib R, Joseph P, Briel M et al (2013) Block-
lation. Circ Res 122(5):752–773 ade of the renin-angiotensin-aldosterone sys-
7. Roy D, Talajic M, Nattel S et al (2008) Rhythm tem (RAAS) for primary prevention of
control versus rate control for atrial fibrillation non-valvular atrial fibrillation: a systematic
and heart failure. N Engl J Med 358 review and meta analysis of randomized con-
(25):2667–2677 trolled trials. Int J Cardiol 165(1):17–24
8. Chen L, Xiao Y, Ma R et al (2014) Bipolar 17. Gillis AM, Krahn AD, Skanes AC et al (2013)
radiofrequency ablation is useful for treating Management of atrial fibrillation in the year
atrial fibrillation combined with heart valve dis- 2033: new concepts, tools, and applications
eases. BMC Surg 14:32 leading to personalized medicine. Can J Car-
9. Yongjun Q, Huanzhang S, Wenxia Z et al diol 29(10):1141–1146
(2015) From changes in local RAAS to struc- 18. Al-Khatib SM, Allen LaPointe NM, Chatterjee
tural remodeling of the left atrium: a beautiful R et al (2014) Rate- and rhythm-control thera-
cycle in atrial fibrillation. Herz 40(3):514–520 pies in patients with atrial fibrillation: a system-
10. Qian Y, Liu Y, Tang H et al (2013) Circulating atic review. Ann Intern Med 160(11):760–773
and local renin-angiotensin-aldosterone system 19. Kirchhof P, Benussi S, Kotecha D et al (2016)
express differently in atrial fibrillation patients 2016 ESC Guidelines for the management of
with different types of mitral valvular disease. J atrial fibrillation developed in collaboration
Renin-Angiotensin-Aldosterone Syst 14 with EACTS. Europace 18(11):1609–1678
(3):204–211 20. Kirchhof P, Ammentorp B, Darius H et al
11. Qian Y, Shao H, Luo T et al (2008) Plasma (2014) Management of atrial fibrillation in
angiotensin converting enzyme level and per- seven European countries after the publication
manent atrial fibrillation with mitral valvular of the 2010 ESC Guidelines on atrial
fibrillation: primary results of the PREvention fibrillation: new look at atrial fibrillation. Euro-
oF thromboemolic events—European Registry pace 14(Suppl 5):v121–v124
in Atrial Fibrillation (PREFER in AF). Euro- 25. Roshanali F, Mandegar MH, Yousefnia MA
pace 16(1):6–14 et al (2009) Prevention of atrial fibrillation
21. Kirchhof P, Breithardt G, Aliot E et al (2013) after coronary artery bypass grafting via atrial
Personalized management of atrial fibrillation: electromechanical interval and use of amiodar-
proceedings from the fourth Atrial Fibrillation one prophylaxis. Interact Cardiovasc Thorac
competence NETwork/European Heart Surg 8(4):421–425
Rhythm Association consensus conference. 26. Komatsu T, Tachibana H, Sato Y et al (2009)
Europace 15(11):1540–1556 Long-term efficacy of upstream therapy using
22. Lafuente-Lafuente C, Valembois L, Bergmann angiotensin-converting enzyme inhibitors and
JF et al (2015) Antiarrhythmics for maintain- statins in combination with antiarrhythmic
ing sinus rhythm after cardioversion of atrial agents for the treatment of paroxysmal atrial
fibrillation. Cochrane Database Syst Rev (3): fibrillation. Int Heart J 50(4):465–476
Cd005049 27. Qian YJ, Xiao XJ, Yuan HS et al (2008) Com-
23. Ausma J, van der Velden HM, Lenders MH bination pharmacological cardioversion of per-
et al (2003) Reverse structural and manent atrial fibrillation in post-prosthetic
gap-junctional remodeling after prolonged mitral valve replacement outpatients: a novel
atrial fibrillation in the goat. Circulation 107 approach for the treatment of atrial fibrillation.
(15):2051–2058 J Int Med Res 36(3):537–543
24. Shenasa M, Soleimanieh M, Shenasa F (2012)
Individualized therapy in patients with atrial
Chapter 14
Precision Medicine and Dilated Cardiomyopathy

Xiang Li and Wenyan Zhu
Abstract
As the most common cardiomyopathy, dilated cardiomyopathy (DCM) is currently defined as a heart
muscle disease which is characterized by left ventricular (LV) or biventricular dilation and systolic dysfunc-
tion at the exclusion of either pressure or volume overload or severe coronary artery disease sufficient to
explain the dysfunction. For established DCM patients, treatment is directed at the major clinical manifes-
tations of heart failure and arrhythmias, including pharmacological treatment, device therapies, and heart
transplantation. But this traditional strategy is incompletely effective and untenable for the consistently
high morbidity and mortality of DCM. Implementation of precision medicine in the field of DCM is
expected to greatly improve the prognosis of patients and reduce the cost by shifting the current focus on
disease treatment to prevention and individualized treatment. This chapter intends to summarize the
progress of accurate medical diagnosis and treatment of dilated heart disease.
Key words Dilated cardiomyopathy, Precision medicine, Pathogenic gene, Gene mutation, Genetic
detection
1 Introduction
As the most common cardiomyopathy, dilated cardiomyopathy

(DCM) is currently defined as a heart muscle disease which is
characterized by left ventricular (LV) or biventricular dilation and
systolic dysfunction at the exclusion of either pressure or volume
overload or severe coronary artery disease sufficient to explain the
dysfunction [1]. The clinical manifestations of DCM are heart
failure, arrhythmia, thromboembolism, syncope, and even sudden
death, which make a poor prognosis and impose a heavy social and
economic burden with an estimated prevalence of 40/100000
individuals and an annual incidence of 7/100000 individuals
[2]. For established DCM patients, treatment is directed at the
major clinical manifestations of heart failure and arrhythmias,
including pharmacological treatment, device therapies, and heart
transplantation [3]. But this traditional strategy is incompletely
effective and untenable for the consistently high morbidity and
mortality of DCM.
161
162 Xiang Li and Wenyan Zhu
To investigate the possible reason, we believe that these tradi-

tional treatments come from the guidance of clinical guidelines,
which are based on evidence-based medicine represented by large-
scale randomized controlled trials. However, the existing clinical
trials usually have strict inclusion criteria and withdrawal criteria,
and only a small number of patients who meet the conditions can be
included in the study, which make the race and specific population
are significant limitations. While the guidelines based on clinical
studies lack individual attention on the pathophysiological process,
clinical manifestation, and treatment response of patients. In this
context, the concept of precision medicine was brought forward
and gradually began to be widely used in the diagnosis and treat-
ment of cardiovascular diseases in recent years.
The concept of precision medicine was first put forward by the
United States in 2011, which refers to a large sample population
study of diseases, based on individual genomic information, com-
bined with proteomics [4], metabonomics [5], transcriptome [6],
epigenetics [7], and other combinatorial studies to find biomole-
cule markers related to diseases, so as to accurately find the causes of
diseases and therapeutic targets. Precision medicine aims to design
the best treatment plan for patients, and finally achieve the goal of
personalized and accurate treatment for diseases and specific
patients, and improve the efficiency of disease diagnosis and pre-
vention [8]. However, the brightest spotlight was provided in 2015
by President Obama in his State of the Union address where he laid
out a vision for a national Precision Medicine Initiative in the
United States [9]. Since then, China, the United Kingdom and
many other countries have launched the precision medicine plan
one after another, and precision medicine has become an important
direction of medical treatment in the future [10].
At present, precision medicine has been widely used in the field
of cardiovascular disease and has made gratifying achievements. For
example, in the establishment of a large cohort study, the
CHANCE study led by Professor Wang Yongjun of China verified
the effectiveness of accurate dual antiplatelet therapy after treating
more than 5100 patients [11]. In the aspect of cardiovascular
disease genomics research, scientists have designed small molecule
MYK-461 to successfully inhibit the progress of hypertrophic car-
diomyopathy (HCM) in animal experiments [12], which is
expected to become a targeted drug for the treatment of HCM
and also become a classic example of precision treatment based on
the genetic characteristics of cardiovascular disease. Professor Huo
Yong’s team in China has developed enalapril maleate folate tablets
for patients with “High Hcy hypertension” through cardiovascular
pharmacogenomics research, which is the only innovative drug in
the world that is allowed to control two major risk factors for stroke
(hypertension and high Hcy) at the same time [13]. The achieve-
ment of these results indicates that the implementation of precision
medicine in the field of DCM may also benefit.
Precision Medicine and Dilated Cardiomyopathy 163
Implementation of precision medicine in the field of DCM is

expected to greatly improve the prognosis of patients and reduce
the cost by shifting the current focus on disease treatment to
prevention and individualized treatment. The central principle sup-
porting this emerging area is that a detailed understanding of each
person’s unique genetic variation, environment, and lifestyle factors
will allow for a subtle assessment of disease risk and personalized
interventions. This chapter intends to summarize the progress of
accurate medical diagnosis and treatment of dilated heart disease.
2 Precision Medicine in the Diagnosis and Treatment of DCM
2.1 The Etiology The etiology of DCM can be divided into hereditary or
and Common non-hereditary factors. Idiopathic DCM refers to hereditary car-
Pathogenic Genes diomyopathy, in which familial hereditary cardiomyopathy is the
of DCM main type of idiopathic DCM with accounting for about 50% of
idiopathic DCM [14]. At present, more than 60 pathogenic genes
of idiopathic DCM have been reported, most of which affect sarco-
mere, nucleic acid protein, cardiomyocyte ion channel, cardiac
development, and so on. About 25% of patients with DCM are
caused by genetic factors of gene mutation [15], suggesting that
genetic defects play an important role in the pathogenesis of dilated
cardiomyopathy. At present, the known pathogenic genes of DCM
include coding sarcomere, Z-line, cytoskeleton, mitochondria,
RNA-binding protein, sarcoplasmic reticulum, and nuclear mem-
brane, among which sarcomere and cytoskeleton proteins are the
most common mutation targets in DCM [16]. The following we
mainly introduce several gene mutations that affect sarcomere.
TTN is the most common pathogenic gene of DCM, account-
ing for about 25% of DCM pathogenic genes. Patients usually
develop typical clinical symptoms before the age of 40 [17]. Titin
protein encoded by TTN gene is the largest protein (4200 kDa)
and the third most abundant muscle protein in human body. Myo-
sin plays an important role in assembling sarcomere, senses
mechanical stimulation, and converts it into biochemical signal,
provides passive tension in striated muscle, and mediates active
contractile force of sarcomere [18, 19]. After selective splicing,
TTN produces a variety of skeletal muscle and myocardial subtypes,
among which the heart subtypes include N2B, N2BA, and
NOVEX-3, which are regulated by the RNA-binding motif protein
(RBM) gene [17, 20]. TTN mutations can cause DCM [21]. The
types of TTN mutations recorded in OMIM, GHMD, and LOVD
databases include nonsense mutations, frame shift mutations, mis-
sense mutations, and splice site mutations. Among them, TTN
mutations associated with DCM, including 29 nonsense mutations,
17 frame shift mutations, 18 mutations affecting TTN gene edit-
ing, and 7 missense mutations [22].
RNA-binding motif protein 20 (RBM20) gene is located on

human chromosome 10 and contains 14 exons with a relative
molecular weight of 1,400,000. It is mainly expressed in cardio-
myocytes and skeletal muscle cells, but is not expressed or rarely
expressed in non-muscle tissues. It regulates the splicing of mRNA
more than 30 genes [23, 24]. Current studies have shown that
RBM20 is closely related to familial DCM. It has been reported
that as many as dozens of RBM20 gene mutations can lead to DCM
[25], including RBM20 can regulate the alternative splicing of
TTN gene to make it transform to different subtypes
[26, 27]. RBM20 gene mutation can not only lead to DCM, but
also cause early clinical symptoms in patients with DCM, increase
the risk of early systolic dysfunction and heart failure, further
increase mortality and affect prognosis.
So far, the pathogenic genes related to familial DCM that affect
sarcomere function include MYBPC3, MYH7, TNNC1, TNNI3,
TNNT2, TPM1, LDB3, and TTR [28].
2.2 Precise Different from the application of precision medicine in tumor and
Diagnosis other fields, precision medicine with gene detection as the main
and Treatment of DCM means rarely produce direct diagnosis and treatment effect in the
Under the Guidance field of DCM, such as developing direct targeted drugs to treat
of Genetic Information DCM, changing the existing clinical treatment of patients with
DCM, and so on. But the genetic test results are of great signifi-
cance in assessing the future risk of the disease among asymptom-
atic family members. Family members with positive genotype and
negative phenotype need regular follow-up echocardiography,
which can detect systolic dysfunction early before clinical symptoms
appear [29]. It can be more targeted regular follow-up and even
necessary preventive treatment for those patients. For example,
patients with DCM caused by mutations in the LMNA gene are
often associated with malignant bradycardia or heart failure
[30, 31]. In the LMNA heterozygous deficiency mouse model,
early use of the β-blocker carvedilol attenuated DCM development
[32], and preliminary human data showed a beneficial ventricular
remodeling effect [33]. If genetic tests identify LMNA mutations at
an early stage, this population can obtain lifestyle guidance, such as
avoiding competitive sports [34], and may also benefit from pro-
phylactic implantable cardioverter defibrillator devices and early
heart transplants [30]. In addition, mutations in the SCN5A gene
may also be associated with a particularly severe arrhythmic DCM
[35–38]. Compared with the relative inefficacy of standard heart
failure therapy, taking drugs with sodium channel blocking can
significantly improve ventricular systolic function and reduce the
burden of arrhythmias [35, 37, 38].
In addition, although precision medicine with genetic detec-
tion as the main means has a certain application in patients with
DCM, its yield is relatively low, which also lacks of specific treat-
ment. In most cases, even if there are positive results from genetic
test, it rarely can be fully used and affect the clinical treatment. This
situation also hinders the application and promotion of gene detec-
tion in the diagnosis and treatment of DCM [39]. It can be
expected that if there is a significant impact on clinical practice,
clinicians’ demand for and interest in genetic information will be
greatly increased. Precision medicine based on genetic knowledge
are promising to meet this clinical need. But if this is to be met, the
first challenge is to increase the number of individuals with positive
genetic results. More people need to be tested. In addition, it is
necessary to confirm as many as possible that gene mutations have
functional alterations associated with DCM. This work needs to be
done by functional genomics analysis, in which mice are preferred
as the animal model for the study [40]. In addition, zebrafish can be
selected as a model to evaluate the role of genes in cardiac develop-
ment, cardiomyopathy and arrhythmias [41, 42], but there are
some limitations in its research value because of its important
differences with humans in anatomy and cardiovascular hemody-
namics [43, 44]. At present, transgenic animals are increasingly
used for disease modeling and can be used to evaluate heart func-
tion in adults [45]. Once the functional change effects of new genes
and new variants are determined, feedback can be used to guide
clinical treatment.
The great progress made in noninvasive imaging and other
cardiovascular diagnostic methods can be used as another part of
precision medicine to help early diagnosis and guide the interven-
tion of DCM. For example, the reduction of left ventricular ejec-
tion fraction by transthoracic echocardiography is the most widely
used method for the diagnosis of dilated cardiomyopathy, but it is
usually insensitive, so it can identify diseases relatively accurately.
Cardiac imaging techniques, including the assessment of myocar-
dial strain on speckle-tracking echocardiography, are producing
more sensitive and specific markers of systolic dysfunction [46],
which are expected to predict the prognosis of patients with symp-
tomatic DCM and identify early ventricular dysfunction in asymp-
tomatic variant carriers [47–49]. In addition, in evaluating the
efficacy and prognosis of DCM, blood biomarkers are expected to
be used as alternative indicators of ventricular dysfunction, such as
circulating clones of bone marrow-derived hematopoietic cells con-
taining somatic mutations have recently been added to the list of
cardiac biomarkers, which have been confirmed that it is signifi-
cantly associated with age and progression of heart failure [50].
3 Controversies and Prospect
Precision medicine is a systematic project that integrates research,

prevention, and treatment, which has a broader application pros-
pect than the traditional medical model. With the progress and
development of current science and technology, many gratifying
achievements have been made in the clinical application of precision

medicine. However, there are more limitations or even challenges:
although the current technology has been able to collect individual
genes, transcripts, proteins, and phenotypes on a large scale, the
huge data scale will magnify the bias of bioinformation collection
[51]. More importantly, the existing research experience still lacks
the ability to interpret most of the information of the human
genome, while epigenetics, transcriptome, proteomics, and micro-
biology started even later. More research is needed to explain the
specific significance of the information contained in genomics
[52]. In addition, with the advent of the era of precision medicine,
the collection of patient information will be more intensive. How
to reasonably protect patients’ privacy, avoid information leakage,
avoid genetic discrimination, and so on, which poses a great chal-
lenge to personal privacy and medical ethics on the big data envi-
ronment [53, 54]. While carrying out a large number of
information collection and information system construction, the
corresponding medical expenditure will also increase significantly.
In terms of drug enterprises, accurate medical care means that fewer
people benefit from the same drug, and the cost of drug develop-
ment increases. The higher the unit price of the drug will be
[54]. Therefore, how to reasonably integrate the existing resources,
ensure the practice of precision medicine on the basis of acceptable
expenditure, and further reduce the costs of sequencing, research,
and development are all problems that need to be solved.
At present, it has made a lot of gratifying progress in the field of
diagnosis and treatment of DCM by precision medicine, but there
are still many unsolved problems needed further research. For
example, the potential use of stem cells to improve prognosis of
patients with congestive heart failure and reduced left ventricular
ejection fraction (LVEF) is a topic of considerable interest
[55]. The preclinical studies of stem cell therapies in DCM have
been limited by the small number of suitable experimental models.
In addition, no clinical studies have shown that stem cell therapies
can improve the clinical outcome of patients with DCM [56]. In
addition, although human genome sequence data are now readily
available, it will be necessary to expand the implementation of
genetic testing and combine effective strategies to identify harmful
variants and explain their prognostic significance. More researches
are needed to determine when and how to treat patients and
relatives who carry harmful genetic mutations, as well as how to
monitor the therapeutic efficacy [3]. Many interesting questions
are still related to the genetic basis and clinical treatment of DCM,
and many cases of idiopathic and familial diseases have not been
explained yet. The discovery of new pathogenic gene mutations and
the possible use of whole genome sequencing should help to
improve the diagnosis rate of patients with DCM. It remains to
be seen whether patients with preclinical disease will benefit from
early pharmacological (or any other) interventions to prevent or

delay the onset of clinical cardiomyopathy. Despite the knowledge
gap, precise medicine in cardiology is no longer a theoretical vision,
but a real opportunity to treat patients with DCM in the future.
Despite these challenges, the successful combination of compre-
hensive sequence analysis and detailed phenotypic analysis should
create a new clinical framework for future evidence-based and
personalized DCM therapy.
Cardiovascular disease, represented by DCM, seriously affects
the health of our people, resulting in a heavy economic and social
burden. Thanks to the progress of multi-biomics research repre-
sented by genomics and the rapid progress of data acquisition and
processing technology, precision medicine has initially shown great
application value in the field of cardiovascular diseases. It is applied
to basic research to better understand the biological and environ-
mental factors of the occurrence and development of various car-
diovascular diseases. In clinical practice, precision medicine has
further promoted the individualized progress of cardiovascular
disease risk assessment, improved existing diagnostic strategies,
and expanded cardiovascular disease intervention methods, which
provides an important support for further reducing the occurrence
of cardiovascular disease and improving the prognosis of cardiovas-
cular disease. At present, the application of precision medicine in
the field of cardiovascular diseases is still in its infancy. In the
scientific research and clinical practice of cardiovascular disease,
we need to establish a thinking model of precision medicine and
carry out intensive research on accurate prevention and control
technology, the discovery of molecular markers for diagnosis and
prognosis, as well as clinical accurate treatment by using modern
technical means such as genomics and big data analysis to seek
accurate intervention targets. It aims to achieve early detection,
early diagnosis, and early treatment of cardiovascular diseases,
meanwhile maximize individual and social health benefits with
effective, safe, and economical medical services and promote
China’s medical and health security level to the forefront of the
world.
References
1. Elliott P, Andersson B, Arbustini E, Bilinska Z, 2. Weintraub RG, Semsarian C, Macdonald P
Cecchi F, Charron P, Dubourg O, Kühl U, (2017) Dilated cardiomyopathy. Lancet
Maisch B, McKenna WJ, Monserrat L, (London, England) 390(10092):400–414.
Pankuweit S, Rapezzi C, Seferovic P, https://doi.org/10.1016/s0140-6736(16)
Tavazzi L, Keren A (2008) Classification of 31713-5
the cardiomyopathies: a position statement 3. Fatkin D, Huttner IG, Kovacic JC, Seidman
from the European Society Of Cardiology JG, Seidman CE (2019) Precision medicine in
Working Group on Myocardial and Pericardial the management of dilated cardiomyopathy:
Diseases. Eur Heart J 29(2):270–276. https:// JACC state-of-the-art review. J Am Coll
doi.org/10.1093/eurheartj/ehm342
Cardiol 74(23):2921–2938. https://doi.org/ suppresses hypertrophic cardiomyopathy in

10.1016/j.jacc.2019.10.011 mice. Science (New York, NY) 351
4. Lindsey ML, Mayr M, Gomes AV, Delles C, (6273):617–621. https://doi.org/10.1126/
Arrell DK, Murphy AM, Lange RA, Costello science.aad3456
CE, Jin YF, Laskowitz DT, Sam F, Terzic A, 13. Huo Y, Li J, Qin X, Huang Y, Wang X, Gottes-
Van Eyk J, Srinivas PR (2015) Transformative man RF, Tang G, Wang B, Chen D, He M,
impact of proteomics on cardiovascular health Fu J, Cai Y, Shi X, Zhang Y, Cui Y, Sun N, Li X,
and disease: a scientific statement from the Cheng X, Wang J, Yang X, Yang T, Xiao C,
American Heart Association. Circulation 132 Zhao G, Dong Q, Zhu D, Wang X, Ge J,
(9):852–872. https://doi.org/10.1161/cir. Zhao L, Hu D, Liu L, Hou FF (2015) Efficacy
0000000000000226 of folic acid therapy in primary prevention of
5. Shah SH, Newgard CB (2015) Integrated stroke among adults with hypertension in
metabolomics and genomics: systems China: the CSPPT randomized clinical trial.
approaches to biomarkers and mechanisms of JAMA 313(13):1325–1335. https://doi.org/
cardiovascular disease. Circ Cardiovasc Genet 8 10.1001/jama.2015.2274
(2):410–419. https://doi.org/10.1161/cir 14. de Gonzalo-Calvo D, Quezada M,
cgenetics.114.000223 Campuzano O, Perez-Serra A, Broncano J,
6. Wende AR (2016) Post-translational modifica- Ayala R, Ramos M, Llorente-Cortes V,
tions of the cardiac proteome in diabetes and Blasco-Turrión S, Morales FJ, Gonzalez P,
heart failure. Proteomics Clin Appl 10 Brugada R, Mangas A, Toro R (2017) Familial
(1):25–38. https://doi.org/10.1002/prca. dilated cardiomyopathy: a multidisciplinary
201500052 entity, from basic screening to novel circulating
7. Fox CS, Hall JL, Arnett DK, Ashley EA, biomarkers. Int J Cardiol 228:870–880.
Delles C, Engler MB, Freeman MW, Johnson https://doi.org/10.1016/j.ijcard.2016.11.
JA, Lanfear DE, Liggett SB, Lusis AJ, 045
Loscalzo J, MacRae CA, Musunuru K, Newby 15. Petretta M, Pirozzi F, Sasso L, Paglia A, Bona-
LK, O’Donnell CJ, Rich SS, Terzic A (2015) duce D (2011) Review and metaanalysis of the
Future translational applications from the con- frequency of familial dilated cardiomyopathy.
temporary genomics era: a scientific statement Am J Cardiol 108(8):1171–1176. https://
from the American Heart Association. Circula- doi.org/10.1016/j.amjcard.2011.06.022
tion 131(19):1715–1736. https://doi.org/ 16. Hershberger RE, Siegfried JD (2011) Update
10.1161/cir.0000000000000211 2011: clinical and genetic issues in familial
8. Natarajan P, O’Donnell CJ (2016) Reducing dilated cardiomyopathy. J Am Coll Cardiol 57
cardiovascular risk using genomic information (16):1641–1649. https://doi.org/10.1016/j.
in the era of precision medicine. Circulation jacc.2011.01.015
133(12):1155–1159. https://doi.org/10. 17. Herman DS, Lam L, Taylor MR, Wang L,
1161/circulationaha.116.021765 Teekakirikul P, Christodoulou D, Conner L,
9. Collins FS, Varmus H (2015) A new initiative DePalma SR, McDonough B, Sparks E, Teo-
on precision medicine. N Engl J Med 372 dorescu DL, Cirino AL, Banner NR, Pennell
(9):793–795. https://doi.org/10.1056/ DJ, Graw S, Merlo M, Di Lenarda A,
NEJMp1500523 Sinagra G, Bos JM, Ackerman MJ, Mitchell
10. Siva N (2015) UK gears up to decode 100,000 RN, Murry CE, Lakdawala NK, Ho CY, Bar-
genomes from NHS patients. Lancet (London, ton PJ, Cook SA, Mestroni L, Seidman JG,
England) 385(9963):103–104. https://doi. Seidman CE (2012) Truncations of titin caus-
org/10.1016/s0140-6736(14)62453-3 ing dilated cardiomyopathy. N Engl J Med 366
(7):619–628. https://doi.org/10.1056/
11. Wang Y, Wang Y, Zhao X, Liu L, Wang D, NEJMoa1110186
Wang C, Wang C, Li H, Meng X, Cui L, Jia J,
Dong Q, Xu A, Zeng J, Li Y, Wang Z, Xia H, 18. Voelkel T, Linke WA (2011) Conformation-
Johnston SC (2013) Clopidogrel with aspirin regulated mechanosensory control via titin
in acute minor stroke or transient ischemic domains in cardiac muscle. Pflugers Archiv
attack. N Engl J Med 369(1):11–19. https:// 462(1):143–154. https://doi.org/10.1007/
doi.org/10.1056/NEJMoa1215340 s00424-011-0938-1
12. Green EM, Wakimoto H, Anderson RL, Evan- 19. Tharp CA, Haywood ME, Sbaizero O, Taylor
chik MJ, Gorham JM, Harrison BC, Henze M, MRG, Mestroni L (2019) The giant protein
Kawas R, Oslob JD, Rodriguez HM, Song Y, titin’s role in cardiomyopathy: genetic, tran-
Wan W, Leinwand LA, Spudich JA, McDowell scriptional, and post-translational modifica-
RS, Seidman JG, Seidman CE (2016) A small- tions of TTN and their contribution to
molecule inhibitor of sarcomere contractility
cardiac disease. Front Physiol 10:1436. cardiomyopathy, regulates titin splicing. Nat
https://doi.org/10.3389/fphys.2019.01436 Med 18(5):766–773. https://doi.org/10.
20. Beqqali A, Bollen IA, Rasmussen TB, van den 1038/nm.2693
Hoogenhof MM, van Deutekom HW, 27. Li S, Guo W, Dewey CN, Greaser ML (2013)
Schafer S, Haas J, Meder B, Sørensen KE, van Rbm20 regulates titin alternative splicing as a
Oort RJ, Mogensen J, Hubner N, Creemers splicing repressor. Nucleic Acids Res 41
EE, van der Velden J, Pinto YM (2016) A (4):2659–2672. https://doi.org/10.1093/
mutation in the glutamate-rich region of nar/gks1362
RNA-binding motif protein 20 causes dilated 28. Pérez-Serra A, Toro R, Sarquella-Brugada G,
cardiomyopathy through missplicing of titin de Gonzalo-Calvo D, Cesar S, Carro E,
and impaired Frank-Starling mechanism. Car- Llorente-Cortes V, Iglesias A, Brugada J,
diovasc Res 112(1):452–463. https://doi. Brugada R, Campuzano O (2016) Genetic
org/10.1093/cvr/cvw192 basis of dilated cardiomyopathy. Int J Cardiol
21. Gerull B, Gramlich M, Atherton J, 224:461–472. https://doi.org/10.1016/j.
McNabb M, Trombitás K, Sasse-Klaassen S, ijcard.2016.09.068
Seidman JG, Seidman C, Granzier H, 29. Hershberger RE, Givertz MM, Ho CY, Judge
Labeit S, Frenneaux M, Thierfelder L (2002) DP, Kantor PF, McBride KL, Morales A, Tay-
Mutations of TTN, encoding the giant muscle lor MRG, Vatta M, Ware SM (2018) Genetic
filament titin, cause familial dilated cardiomy- evaluation of cardiomyopathy-A Heart Failure
opathy. Nat Genet 30(2):201–204. https:// Society of America Practice Guideline. J Card
doi.org/10.1038/ng815 Fail 24(5):281–302. https://doi.org/10.
22. Chauveau C, Rowell J, Ferreiro A (2014) A 1016/j.cardfail.2018.03.004
rising titan: TTN review and mutation update. 30. Kumar S, Baldinger SH, Gandjbakhch E,
Hum Mutat 35(9):1046–1059. https://doi. Maury P, Sellal JM, Androulakis AF,
org/10.1002/humu.22611 Waintraub X, Charron P, Rollin A, Richard P,
23. Wells QS, Becker JR, Su YR, Mosley JD, Stevenson WG, Macintyre CJ, Ho CY,
Weeke P, D’Aoust L, Ausborn NL, Ramirez Thompson T, Vohra JK, Kalman JM,
AH, Pfotenhauer JP, Naftilan AJ, Markham L, Zeppenfeld K, Sacher F, Tedrow UB, Lakda-
Exil V, Roden DM, Hong CC (2013) Whole wala NK (2016) Long-term arrhythmic and
exome sequencing identifies a causal RBM20 nonarrhythmic outcomes of lamin A/C muta-
mutation in a large pedigree with familial tion carriers. J Am Coll Cardiol 68
dilated cardiomyopathy. Circ Cardiovasc (21):2299–2307. https://doi.org/10.1016/j.
Genet 6(4):317–326. https://doi.org/10. jacc.2016.08.058
1161/circgenetics.113.000011 31. Fatkin D, MacRae C, Sasaki T, Wolff MR,
24. Guo W, Zhu C, Yin Z, Wang Q, Sun M, Porcu M, Frenneaux M, Atherton J, Vidaillet
Cao H, Greaser ML (2018) Splicing factor HJ Jr, Spudich S, De Girolami U, Seidman JG,
RBM20 regulates transcriptional network of Seidman C, Muntoni F, Müehle G, Johnson W,
titin associated and calcium handling genes in McDonough B (1999) Missense mutations in
the heart. Int J Biol Sci 14(4):369–380. the rod domain of the lamin A/C gene as
https://doi.org/10.7150/ijbs.24117 causes of dilated cardiomyopathy and
25. Ellinor PT, Sasse-Klaassen S, Probst S, conduction-system disease. N Engl J Med
Gerull B, Shin JT, Toeppel A, Heuser A, 341(23):1715–1724. https://doi.org/10.
Michely B, Yoerger DM, Song BS, Pilz B, 1056/nejm199912023412302
Krings G, Coplin B, Lange PE, Dec GW, Hen- 32. Chandar S, Yeo LS, Leimena C, Tan JC, Xiao
nies HC, Thierfelder L, MacRae CA (2006) A XH, Nikolova-Krstevski V, Yasuoka Y,
novel locus for dilated cardiomyopathy, diffuse Gardiner-Garden M, Wu J, Kesteven S,
myocardial fibrosis, and sudden death on chro- Karlsdotter L, Natarajan S, Carlton A,
mosome 10q25-26. J Am Coll Cardiol 48 Rainer S, Feneley MP, Fatkin D (2010) Effects
(1):106–111. https://doi.org/10.1016/j. of mechanical stress and carvedilol in lamin A/
jacc.2006.01.079 C-deficient dilated cardiomyopathy. Circ Res
26. Guo W, Schafer S, Greaser ML, Radke MH, 106(3):573–582. https://doi.org/10.1161/
Liss M, Govindarajan T, Maatz H, Schulz H, circresaha.109.204388
Li S, Parrish AM, Dauksaite V, Vakeel P, 33. Yeoh T, Hayward C, Benson V, Sheu A,
Klaassen S, Gerull B, Thierfelder L, Regitz- Richmond Z, Feneley MP, Keogh AM,
Zagrosek V, Hacker TA, Saupe KW, Dec GW, Macdonald P, Fatkin D (2011) A randomised,
Ellinor PT, MacRae CA, Spallek B, Fischer R, placebo-controlled trial of carvedilol in early
Perrot A, Özcelik C, Saar K, Hubner N, Got- familial dilated cardiomyopathy. Heart Lung
thardt M (2012) RBM20, a gene for hereditary
Circ 20(9):566–573. https://doi.org/10. Judge DP, Le Marec H, McKenna WJ,

1016/j.hlc.2011.06.004 Schulze-Bahr E, Semsarian C, Towbin JA,
34. Pelliccia A, Solberg EE, Papadakis M, Adami Watkins H, Wilde A, Wolpert C, Zipes DP
PE, Biffi A, Caselli S, La Gerche A, Niebauer J, (2011) HRS/EHRA expert consensus state-
Pressler A, Schmied CM, Serratosa L, Halle M, ment on the state of genetic testing for the
Van Buuren F, Borjesson M, Carrè F, channelopathies and cardiomyopathies: this
Panhuyzen-Goedkoop NM, Heidbuchel H, document was developed as a partnership
Olivotto I, Corrado D, Sinagra G, Sharma S between the Heart Rhythm Society (HRS)
(2019) Recommendations for participation in and the European Heart Rhythm Association
competitive and leisure time sport in athletes (EHRA). Europace 13(8):1077–1109.
with cardiomyopathies, myocarditis, and peri- https://doi.org/10.1093/europace/eur245
carditis: position statement of the Sport Cardi- 40. van der Weyden L, White JK, Adams DJ, Logan
ology Section of the European Association of DW (2011) The mouse genetics toolkit: reveal-
Preventive Cardiology (EAPC). Eur Heart J 40 ing function and mechanism. Genome Biol 12
(1):19–33. https://doi.org/10.1093/ (6):224. https://doi.org/10.1186/gb-2011-
eurheartj/ehy730 12-6-224
35. Mann SA, Castro ML, Ohanian M, Guo G, 41. Huttner IG, Wang LW, Santiago CF,
Zodgekar P, Sheu A, Stockhammer K, Horvat C, Johnson R, Cheng D, von
Thompson T, Playford D, Subbiah R, Frieling-Salewsky M, Hillcoat K, Bemand TJ,
Kuchar D, Aggarwal A, Vandenberg JI, Fatkin Trivedi G, Braet F, Hesselson D, Alford K,
D (2012) R222Q SCN5A mutation is asso- Hayward CS, Seidman JG, Seidman CE, Fene-
ciated with reversible ventricular ectopy and ley MP, Linke WA, Fatkin D (2018) A-band
dilated cardiomyopathy. J Am Coll Cardiol 60 titin truncation in zebrafish causes dilated car-
(16):1566–1573. https://doi.org/10.1016/j. diomyopathy and hemodynamic stress intoler-
jacc.2012.05.050 ance. Circ Genom Precis Med 11(8):e002135.
36. McNair WP, Sinagra G, Taylor MR, Di https://doi.org/10.1161/circgen.118.
Lenarda A, Ferguson DA, Salcedo EE, 002135
Slavov D, Zhu X, Caldwell JH, Mestroni L 42. Shi X, Chen R, Zhang Y, Yun J, Brand-
(2011) SCN5A mutations associate with Arzamendi K, Liu X, Wen XY (2018) Zebrafish
arrhythmic dilated cardiomyopathy and com- heart failure models: opportunities and chal-
monly localize to the voltage-sensing mecha- lenges. Amino Acids 50(7):787–798. https://
nism. J Am Coll Cardiol 57(21):2160–2168. doi.org/10.1007/s00726-018-2578-7
https://doi.org/10.1016/j.jacc.2010.09.084 43. Wilkinson RN, Jopling C, van Eeden FJ (2014)
37. Laurent G, Saal S, Amarouch MY, Béziau DM, Zebrafish as a model of cardiac disease. Prog
Marsman RF, Faivre L, Barc J, Dina C, Mol Biol Transl Sci 124:65–91. https://doi.
Bertaux G, Barthez O, Thauvin-Robinet C, org/10.1016/b978-0-12-386930-2.00004-5
Charron P, Fressart V, Maltret A, Villain E, 44. Bakkers J (2011) Zebrafish as a model to study
Baron E, Mérot J, Turpault R, Coudière Y, cardiac development and human cardiac dis-
Charpentier F, Schott JJ, Loussouarn G, ease. Cardiovasc Res 91(2):279–288. https://
Wilde AA, Wolf JE, Baró I, Kyndt F, Probst V doi.org/10.1093/cvr/cvr098
(2012) Multifocal ectopic Purkinje-related pre- 45. Bartlett HL, Escalera RB 2nd, Patel SS, Wede-
mature contractions: a new SCN5A-related meyer EW, Volk KA, Lohr JL, Reinking BE
cardiac channelopathy. J Am Coll Cardiol 60 (2010) Echocardiographic assessment of car-
(2):144–156. https://doi.org/10.1016/j. diac morphology and function in Xenopus.
jacc.2012.02.052 Comp Med 60(2):107–113
38. Zakrzewska-Koperska J, Franaszczyk M, 46. Collier P, Phelan D, Klein A (2017) A test in
Bilińska Z, Truszkowska G, Karczmarz M, Szu- context: myocardial strain measured by
mowski Ł, Zieliński T, Płoski R, Bilińska M speckle-tracking echocardiography. J Am Coll
(2018) Rapid and effective response of the Cardiol 69(8):1043–1056. https://doi.org/
R222Q SCN5A to quinidine treatment in a 10.1016/j.jacc.2016.12.012
patient with Purkinje-related ventricular
arrhythmia and familial dilated cardiomyopa- 47. Potter E, Marwick TH (2018) Assessment of
thy: a case report. BMC Med Genet 19(1):94. left ventricular function by echocardiography:
https://doi.org/10.1186/s12881-018-0599- the case for routinely adding global longitudi-
4 nal strain to ejection fraction. J Am Coll Car-
diol Img 11(2 Pt 1):260–274. https://doi.
39. Ackerman MJ, Priori SG, Willems S, Berul C, org/10.1016/j.jcmg.2017.11.017
Brugada R, Calkins H, Camm AJ, Ellinor PT,
Gollob M, Hamilton R, Hershberger RE,
48. van der Bijl P, Bootsma M, Hiemstra YL, Poulter N, Farrall M, Gansevoort RT,
Ajmone Marsan N, Bax JJ, Delgado V (2019) Padmanabhan S, M€agi R, Stanton A,
Left ventricular 2D speckle tracking echocardi- Connell J, Bakker SJ, Metspalu A, Shields
ography for detection of systolic dysfunction in DC, Thom S, Brown M, Sever P, Esko T,
genetic, dilated cardiomyopathies. Eur Heart J Hayward C, van der Harst P, Saleheen D,
Cardiovasc Imaging 20(6):694–699. https:// Chowdhury R, Chambers JC, Chasman DI,
doi.org/10.1093/ehjci/jey169 Chakravarti A, Newton-Cheh C, Lindgren
49. Verdonschot JAJ, Merken JJ, Brunner-La CM, Levy D, Kooner JS, Keavney B,
Rocca HP, Hazebroek MR, Eurlings C, Tomaszewski M, Samani NJ, Howson JM,
Thijssen E, Wang P, Weerts J, van Empel V, Tobin MD, Munroe PB, Ehret GB, Wain LV
Schummers G, Schreckenberg M, van den (2017) Genome-wide association analysis iden-
Wijngaard A, Lumens J, Brunner HG, Hey- tifies novel blood pressure loci and offers
mans SRB, Krapels IPC, Knackstedt C (2020) biological insights into cardiovascular risk.
Value of speckle tracking-based deformation Nat Genet 49(3):403–415. https://doi.org/
analysis in screening relatives of patients with 10.1038/ng.3768
asymptomatic dilated cardiomyopathy. J Am 53. Liddicoat J, Skopek JM, Liddell K (2017) Pre-
Coll Cardiol Img 13(2 Pt 2):549–558. cision medicine: legal and ethical challenges.
https://doi.org/10.1016/j.jcmg.2019.02. University of Cambridge Faculty of Law
032 Research Paper No 64/2017. doi:https://doi.
50. Dorsheimer L, Assmus B, Rasper T, Ortmann org/10.2139/ssrn.3070388
CA, Ecke A, Abou-El-Ardat K, Schmid T, 54. Joyner MJ, Paneth N (2015) Seven Questions
Brüne B, Wagner S, Serve H, Hoffmann J, for Personalized Medicine. JAMA 314
Seeger F, Dimmeler S, Zeiher AM, Rieger (10):999–1000. https://doi.org/10.1001/
MA (2019) Association of mutations contri- jama.2015.7725
buting to clonal hematopoiesis with prognosis 55. Bolli R, Ghafghazi S (2015) Current status of
in chronic ischemic heart failure. JAMA Car- cell therapy for non-ischaemic cardiomyopa-
diol 4(1):25–33. https://doi.org/10.1001/ thy: a brief overview. Eur Heart J 36
jamacardio.2018.3965 (42):2905–2908. https://doi.org/10.1093/
51. Goldfeder RL, Priest JR, Zook JM, Grove ME, eurheartj/ehv454
Waggott D, Wheeler MT, Salit M, Ashley EA 56. Martino H, Brofman P, Greco O, Bueno R,
(2016) Medical implications of technical accu- Bodanese L, Clausell N, Maldonado JA,
racy in genome sequencing. Genome Med 8 Mill J, Braile D, Moraes J Jr, Silva S, Bozza A,
(1):24. https://doi.org/10.1186/s13073- Santos B, Campos de Carvalho A (2015) Mul-
016-0269-0 ticentre, randomized, double-blind trial of
52. Warren HR, Evangelou E, Cabrera CP, Gao H, intracoronary autologous mononuclear bone
Ren M, Mifsud B, Ntalla I, Surendran P, Liu C, marrow cell injection in non-ischaemic dilated
Cook JP, Kraja AT, Drenos F, Loh M, cardiomyopathy (the dilated cardiomyopathy
Verweij N, Marten J, Karaman I, Lepe MP, arm of the MiHeart study). Eur Heart J 36
O’Reilly PF, Knight J, Snieder H, Kato N, (42):2898–2904. https://doi.org/10.1093/
He J, Tai ES, Said MA, Porteous D, Alver M, eurheartj/ehv477
Chapter 15
Research Progress in Pathogenesis of Total Anomalous

Pulmonary Venous Connection
Xin Shi, Yanan Lu, and Kun Sun
Abstract
Congenital heart defect (CHD) is one of the most common birth defects and the leading course of infant
mortality. Total anomalous pulmonary venous connection (TAPVC) is a rare type of cyanotic which
accounting for approximately 1–3% of congenital heart disease cases. Based on where the anomalous
veins drain, TAPVC can be divided into four subtypes: supracardiac, cardiac, infracardiac, and mixed. In
TAPVC, all pulmonary veins fail to link to the left atrium correctly but make abnormal connections to the
right atrium or systemic venous system. The mortality of TAPVC patients without proper intervention is
nearly 80% in the first year of life and 50% of them died within 3 months after birth. However, the
pathogenesis and mechanism of TAPVC remains elusive. In this chapter, we systematically review the
epidemiology, anatomy, and pathophysiology of TAPVC and give an overview of the research progress of
TAPVC pathogenesis.
Key words Congenital heart disease, Total anomalous pulmonary venous connection, Genetics,
Genome
1 TAPVC Epidemiology
Congenital heart defect (CHD) is one of the most common birth

defects and the leading course of infant mortality [1, 2]. Total
anomalous pulmonary venous connection (TAPVC) is recognized
as a rare and severe cyanotic CHD, affecting about 1–3% of infants
with cardiovascular malformations [3]. In birth defect data from
population-based birth defects surveillance programs across the
United States from 2005 to 2009, researchers estimated that each
year about 1 in every 10,000 infants born with TAPVC [4]. An
international population-based study from 1998 to 2004 showed a
higher incidence of TAPVC than previous studies at 7.1 per
100,000 live births [5]. A large retrospective study reviewed the
medical records of 377 children with TAPVC between 1946 and
2005 [4]. In this study, the anomalous venous connection was
supracardiac in 44%, infracardiac in 26%, cardiac in 21%, and
173
174 Xin Shi et al.
mixed in 9% [6]. Most TAPVC cases exist independently, but some

of them can occur in conjunction with a wide variety of cardiac and
extracardiac anomalies and complicated syndrome, such as Schmid-
Fraccaro syndrome, Holt-Oram syndrome, Asplenia syndrome,
and Fryns syndrome [7, 8]. Despite great care with case identifica-
tion, it is still likely that a small number of cases were missed, and in
some, a diagnosis may never have been made, either in life or death.
2 Development of Pulmonary Veins
At early gestational periods, pulmonary vein emerged from a big

venous plexus located within the splanchnic mesoderm, in contrast,
systemic venous tributaries develop laterally on the junction
between the splanchnic and somatic mesoderm by muscularization
of the mesenchyme that surrounds the common cardinal veins.
After the development of vascular channels within the lung buds,
connection with the pharyngeal mesenchyme will develop as the
portal of entry for the pulmonary vein. However, whether the
pulmonary vein as a branch from the left atrium obtains a connec-
tion to the lung plexus or the pulmonary vein forms as a solitary
vessel in the dorsal mesocardium and is only secondarily
incorporated into the atrium remains controversial.
Previous studies suggest that TAPVC occurs when the mid-
pharyngeal endothelial strand (MES), which is the precursor of the
common pulmonary vein, arrange at the improper location
[9]. According to this suggestion, TAPVC is due to the defects in
the formation or maintenance of the MES [10]. Recently, high-
resolution three-dimensional reconstructions of avian embryos
have helped clarify that the pulmonary vein derives from a greater
vascular plexus within the splanchnic mesoderm [11]. Incomplete
remodeling of this plexus with failure of separation into distinct
pulmonary and systemic vascular zones has been postulated to be a
developmental defect that could result in anomalous pulmonary
venous connection.
In pathology research of TAPVC, no vessel wall was found in
the smooth-walled left atrial body and myocardial layer was struc-
turally unnatural. No myocardial layer was organized around the
pulmonary veins. An open connection to the left atrial is mandatory
for proper development of the left atrial vascular wall and myocar-
dium and the pulmonary venous muscular sleeve. The posterior
heart field is suggested to be responsible for the abnormal memo-
rialization and smooth muscle cell formation of the left atrial dorsal
wall and pulmonary veins in TAPVC [12].
Research Progress in Pathogenesis of Total Anomalous Pulmonary Venous Connection 175
3 TAPVC Anatomy
In TAPVC, none pulmonary veins link to the left atrium correctly

but make abnormal connections to the right atrium or systemic
venous system. The abnormal connection of pulmonary veins deli-
vers oxygen-rich blood to the right side of the heart. To survive, a
patent foramen ovale (PFO) or an atrial septal defect (ASD) always
exists, so that oxygen-rich blood which entered the right atrium
from the pulmonary vein could go across to the left atrium and out
to the body. Different subtypes of TAPVC are recognized depend-
ing on anatomical position of the pulmonary veins drain to the
heart: supracardiac, cardiac, infracardiac, and mixed.
1. Supracardiac form: The pulmonary veins come together and
form an abnormal connection above the heart to the superior
vena cava, which is a main blood vessel that brings oxygen-poor
blood from the upper part of the body to the heart.
2. Cardiac form: The pulmonary veins meet behind the heart and
connect to the right atrium. The coronary sinus, which helps
bring oxygen-poor blood from the heart muscle back to the
heart, helps connect the pulmonary veins to the right atrium.
3. Infracardiac form: The pulmonary veins come together and
form abnormal connections below the heart. Blood returns to
the right atrium from the veins of the liver and the inferior vena
cava, which is the main blood vessel that brings oxygen-poor
blood from the lower part of the body to the heart.
4. Mixed form: This form of TAPVC may consist of any of the
above connections. The repair of mixed type TAPVC involved a
combination of the above approaches as dictated by the specific
anatomy of the lesion.
4 Gene Mutations in TAPVC Patients
Because most TAPVC is sporadic, and its mortality is extremely

high without proper intervention, and other factors, the previous
studies mainly based on clinical data, few researches depend on the
pathogenesis of TAPVC. They were summarized in Table 1.
Bleyl et al. revealed that 4q12 is related to TAPVC using
genetic linkage analysis, and further study showed that the candi-
date genes in this region include vascular endothelial growth factor
receptor 2 (VEGFR2) and platelet-derived growth factor receptor
2 (PDGFR2) [13, 14]. In mouse and chick embryos for both the
PDGFRA receptor and its ligand PDGFRA show temporal and
spatial patterns consistent with a role in pulmonary vein develop-
ment. These data supported a role for PDGF-signaling in
176 Xin Shi et al.
Table 1
Gene mutations in TAPVC patients
Gene Gene name Variants Reference PMID

ACVRL1 Activin A receptor-like type 1 p.T217G; Li et al. Oncotarget (2017) 28412737
p.W217X;
p.G219H
ANKRD1 Ankyrin repeat domain 1 p. Cinquetti et al. Hum Mutat 18273862
Thr116Met (2008)
GJA1 Gap junction protein alpha 1 p.G717A; Huang et al. J Cardiovasc Dis 22135478
p.C555T Res. (2011)
KDR Kinase insert domain receptor / Bleyl et al. Am J Med Genet A 17036341
(2006); 12112663
Gutierrez et al. Hum Mutat
(2002)
NKX2–5 NK2 homeobox 5 / Ye et al. Development. (2015) 26138475
NRP1 Neuropilin 1 / AghajanianJ et al. Biol Chem 24825896
(2014)
PDGFRA Platelet-derived growth factor p.G429R Bleyl et al. Hum Mol Genet 20071345
receptor alpha (2010)
RBP5 Retinol binding protein 5 p.E70Q; Nash et al. Plos One (2015) 26121141
p.Glu70Gln
SEMA3D Semaphorin 3D P.F602L; Degenhardt et al. Nature 23685842
S65P Medecine (2013)
SGCD Sarcoglycan delta p.D239E Li et al. Oncotarget (2017) 28412737
SMAD1 SMAD family member 1 / Fahed et al.Perros et al. 26130118
Circulation (2015)
ZIC3 Zic family member 3 / Circ Res (2013) 23410880
pulmonary veins development and suggest that dysregulation of

the PDGFRA gene confers susceptibility to the occurrence of
TAPVC [15].
Previously reported a TAPVC patient bearing a de novo 10;21
balanced translocation. Cinquetti et al. then cloned both transloca-
tion breakpoints from this patient and mapped the ANKRD1 gene,
encoding a cardiac transcriptional regulator, in situ hybridization
analysis performed on murine embryos showed ANKRD1 expres-
sion in the developing pulmonary veins. in lymphoblastoid cell lines
derived from TAPVR patient, ANKRD1 expression levels were
found to be highly increased. In vitro calpain-mediated degradation
assays, mutation from TAPVC patients enhances both the stability
of the ANKRD1/CARP protein and its transcriptional repression
activity upon the cardiac-specific atrial natriuretic factor (ANF)
promoter. Taken together, these results indicated ANKRD1 as a
possible candidate gene for TAPVR pathogenesis [16–18].
Research Progress in Pathogenesis of Total Anomalous Pulmonary Venous Connection 177
Karl et al. found without semaphorin 3D (SEMA3D), endo-

thelial tubes form in a region that is normally avascular, resulting in
aberrant connections. SEMA3D provides a repulsive cue to endo-
thelial cells in this area, establishing a boundary. Sequencing of
SEMA3D in individuals with anomalous pulmonary veins identified
a phenylalanine-to-leucine substitution that adversely affects
SEMA3D function. They demonstrated SEMA3D to be a crucial
gene in pulmonary venous connection because SEMA3D / mice
displayed the TAPVC or partial APVC (PAPVC) phenotype [19].
Li et al. analyzed WES data from six sporadic TAPVC cases,
providing evidence for ACVRL1 as a known causative gene and for
SGCD as a candidate TAPVC gene [20]. Nash et al. used WGS
analysis to identify a nonsynonymous variant predicted to be dele-
terious and overrepresented in TAPVC. This variant lies in the
shared segment in the retinol binding protein 5 (RBP5) gene [17].
Shi et al. used WES data from 178 TAPVC cases and filtered
three novel candidate genes (SNAI1, HMGA2, and VAV2) which
have not previously been reported in either humans or animals, and
this is the largest series of WES in TAPVC cases reported to date
[21]. Besides, based on the results of CNV discovery in a case-
control cohort Shi et al. found evidence that CNVs of 7 candidate
genes (PCSK7, RRP7A, SERHL, TARP, TTN, SERHL2, NBPF3)
could contribute to the genetic etiology of TAPVC [22].
These candidate genes open new fields of investigation into
TAPVC pathology and provide novel insights into pulmonary
vein development. Up to now, only a few genes have been identified
as candidate genes for TAPVC pathogenesis. These candidate genes
explain only a small fraction of the molecular mechanism underly-
ing TAPVC pathogenesis, and comprehensive genomic data are still
lacking.
5 Conclusions
TAPVC is a rare and severe heart defect, the pathogenesis of

TAPVC remains complicated and undiscovered. Through the past
decades, some copy number variants and gene mutations have been
considered to associated with TAPVC. It has reached a consensus
that the accumulation of copy number variants, gene mutations,
and environmental factors may play an important role in the devel-
opment of pulmonary vein. However, because of the great hetero-
geneity, the genetic molecular mechanism of TAPVC is still unclear.
More information is required to illustrate the relationship between
genotype and phenotype. Furthermore, with the progress in
genetic testing technologies and the use of next-generation
sequencing, candidate gene, and underlying mechanism of
TAPVC will open new fields of investigation into TAPVC pathol-
ogy and provide novel insights into pulmonary vein development.
178 Xin Shi et al.
References
1. Pediatric Cardiac Genomics C, Gelb B, veins: state of the art and correlation with clini-
Brueckner M et al (2013) The congenital cal entities. Int J Cardiol 147(1):13–24
heart disease genetic network study: rationale, 12. Douglas YL, Jongbloed MR, den Hartog WC
design, and early results. Circ Res 112 et al (2009) Pulmonary vein and atrial wall
(4):698–706 pathology in human total anomalous pulmo-
2. Jin SC, Homsy J, Zaidi S et al (2017) Contri- nary venous connection. Int J Cardiol 134
bution of rare inherited and de novo variants in (3):302–312
2,871 congenital heart disease probands. Nat 13. Bleyl SB, Botto LD, Carey JC et al (2006)
Genet 49(11):1593–1601 Analysis of a Scottish founder effect narrows
3. Shi G, Zhu Z, Chen J et al (2017) Total anom- the TAPVR-1 gene interval to chromosome
alous pulmonary venous connection: the cur- 4q12. Am J Med Genet A 140(21):2368–2373
rent management strategies in a pediatric 14. Bleyl S, Nelson L, Odelberg SJ et al (1995) A
cohort of 768 patients. Circulation 135 gene for familial total anomalous pulmonary
(1):48–58 venous return maps to chromosome 4p13-
4. Mai CT, Riehle-Colarusso T, O’Halloran A q12. Am J Hum Genet 56(2):408–415
et al (2012) Selected birth defects data from 15. Bleyl SB, Saijoh Y, Bax NA et al (2010) Dysre-
population-based birth defects surveillance gulation of the PDGFRA gene causes inflow
programs in the United States, 2005-2009: tract anomalies including TAPVR: integrating
featuring critical congenital heart defects tar- evidence from human genetics and model
geted for pulse oximetry screening. Birth organisms. Hum Mol Genet 19(7):1286–1301
Defects Res A Clin Mol Teratol 94 16. Cinquetti R, Badi I, Campione M et al (2008)
(12):970–983 Transcriptional deregulation and a missense
5. Seale AN, Uemura H, Webber SA et al (2010) mutation define ANKRD1 as a candidate gene
Total anomalous pulmonary venous connec- for total anomalous pulmonary venous return.
tion: morphology and outcome from an inter- Hum Mutat 29(4):468–474
national population-based study. Circulation 17. Nash D, Arrington CB, Kennedy BJ et al
122(25):2718–2726 (2015) Shared segment analysis and next-
6. Nurkalem Z, Gorgulu S, Eren M, Bilal MS generation sequencing implicates the retinoic
(2006) Total anomalous pulmonary venous acid signaling pathway in total anomalous pul-
return in the fourth decade. Int J Cardiol 113 monary venous return (TAPVR). PLoS One 10
(1):124–126 (6):e0131514
7. Correa-Villasenor A, Ferencz C, Boughman 18. Arimura T, Bos JM, Sato A et al (2009) Cardiac
JA, Neill CA (1991) Total anomalous pulmo- ankyrin repeat protein gene (ANKRD1) muta-
nary venous return: familial and environmental tions in hypertrophic cardiomyopathy. J Am
factors. The Baltimore-Washington Infant Coll Cardiol 54(4):334–342
Study Group. Teratology 44(4):415–428 19. Degenhardt K, Singh MK, Aghajanian H et al
8. Siddharth CB, Yadav M, Bhoje A, Hote MP (2013) Semaphorin 3d signaling defects are
(2018) Dual drainage total anomalous pulmo- associated with anomalous pulmonary venous
nary venous connection: a rare mixed variant. connections. Nat Med 19(6):760–765
Asian Cardiovasc Thorac Ann 26(4):305–307 20. Li J, Yang S, Pu Z et al (2017) Whole-exome
9. van den Berg G, Moorman AF (2011) Devel- sequencing identifies SGCD and ACVRL1
opment of the pulmonary vein and the systemic mutations associated with total anomalous pul-
venous sinus: an interactive 3D overview. PLoS monary venous return (TAPVR) in Chinese
One 6(7):e22055 population. Oncotarget 8(17):27812–27819
10. DeRuiter MC, Gittenberger-De Groot AC, 21. Shi X, Huang T, Wang J et al (2018) Next-
Wenink AC, Poelmann RE, Mentink MM generation sequencing identifies novel genes
(1995) In normal development pulmonary with rare variants in total anomalous pulmo-
veins are connected to the sinus venosus seg- nary venous connection. EBioMedicine
ment in the left atrium. Anat Rec 243 38:217–227
(1):84–92 22. Shi X, Cheng L, Jiao X et al (2018) Rare copy
11. Douglas YL, Jongbloed MR, Deruiter MC, number variants identify novel genes in spo-
Gittenberger-de Groot AC (2011) Normal radic total anomalous pulmonary vein connec-
and abnormal development of pulmonary tion. Front Genet 9:559
Part IV
Precision Medicine of Other Complex Diseases

Chapter 16
Airway Inflammation Biomarker for Precise Management

of Neutrophil-Predominant COPD
Xue Liang, Ting Liu, Zhiming Zhang, and Ziyu Yu
Abstract
Chronic obstructive pulmonary disease (COPD) course can be divided into stable stage and acute exacer-
bation. Deepen the understanding to the function and role of airway inflammatory cells in stable COPD is
important for developing new therapies to restore airway dysfunction and preventing stable stage COPD
progress to acute exacerbation COPD. Neutrophil is a feature of lower airways and lung inflammation in
majority COPD patients at stable stage and increased neutrophils usually means COPD patients are in a
more serious stage. Neutrophil-predominant COPD always accompanied by increased numbers of macro-
phages, lymphocytes, and dendritic cells. The composition proportion of different inflammatory cells are
changed with disease severity. Recently, neutrophilic inflammation has been proved to be correlated with
the disturbance of airway resident microbiota, which promote neutrophil influx and exacerbates inflamma-
tion. Consequently, understanding the details of increased neutrophils and dysbacteriosis in COPD is
necessary for making precise management strategy against neutrophil-associated COPD.
Key words Neutrophil, COPD, Precision medicine, Inflammation, Diagnosis, Resident microbiota,
Dysbacteriosis
1 Introduction
Chronic obstructive pulmonary disease (COPD) is a common dis-

ease with high morbidity and mortality in the world, mainly caused
by air pollution or cigarette smoking. Airway inflammation is a
consistent feature of COPD, but still have heterogeneity. COPD
patients according to different airway inflammation phenotyping
can be divided into four patterns, including inflammasome-
neutrophil-predominant bacterial-associated COPD, T2-eosino-
phil-predominant COPD, T1-viral-associated COPD, and
pro-inflammatory bacterial-associated COPD [1]. Neutrophil-
associated COPD is the most common inflammatory phenotype
in COPD (Fig. 1), which is characterized by the activation of the
inflammation. However, more evidence suggests that neutrophilic
inflammation as well as the concomitant inflammasome activation
181
182 Xue Liang et al.
Fig. 1 The inflammatory phenotype pattern of COPD
are probably driven by the dysbiosis of airway colony microbiota.

Eosinophil-associated COPD probably occupy 10–40% of COPD,
which is characterized by increased eosinophilic inflammation in
the sputum or blood, in asthma-COPD overlap patients, the con-
comitant elevated neutrophils and eosinophils are together con-
tributed to a greater future risk of severe exacerbation [2].
Airway inflammation, one main symptom of COPD airway, is
implicated in the pathogenesis of COPD. While, COPD with dif-
ferent phenotype shows different inflammation heterogeneity, and
for neutrophil-predominant COPD, activation of the inflamma-
some is its main feature [3]. Unfortunately, the therapies target
neutrophilic inflammation or inflammasome have been proved
ineffective. Therefore, ascertaining the causing of dominant neu-
trophil features of airway is essential to developing an effective and
accurate management. Recent studies show the neutrophil-related
inflammation may induced by dysbacteriosis, which may result in
failure in treatment directly target neutrophilic inflammation
[3, 4]. For instance, smoking as a risk factor of COPD can promote
neutrophilic inflammation, and also impair its antibacterial defense,
leading to the disturbance of airway resident microbiota, instead,
dysbacteriosis in turn promotes neutrophil influx and exacerbates
inflammation [5, 6]. In neutrophil-predominant COPD patients,
Airway Inflammation Biomarker for Precise Management of Neutrophil. . . 183
there is an increase in the abundance of the bacterial phylum

Proteobacteria and H. Influenzae, so that the ration of gamma-
proteobacteria to firmicutes (γP:F) increase [7–9].
Lung microbiome can vary even in stable stage; therefore, the
microbial changes from stable stage to exacerbations contains two
sides: the regular temporal perturbations of lung resident micro-
biota, and the disease-associated disruption of lung resident micro-
biota. Therefore, examining the baseline variability should be the
first step to more precisely assess the dysbiosis during different
disease stage. Here, we reviewed the inflammatory and immuno-
logical features of neutrophil-predominant COPD, and how these
immune events crosstalk with the resident microbiota, aim to find a
way out to COPD precise management.
2 Inflammatory and Immunological Features of COPD
In COPD, chronic inflammation process is triggered by external

causes and internal causes, afterwards, exists throughout every
stage of COPD [10]. Distinguishing the inflammatory phenotypes
of different stage COPD is vital to promoting precise clinical man-
agement. The progression of COPD is a long course with progres-
sive and irreversible airflow limitation, according to the Global
Initiative for Chronic Obstructive Lung Disease (GOLD), airflow
limitation of COPD should be comply with a forced expiratory
volume in the first second (FEV1) to forced vital capacity (FVC)
ratio of less than 0.7 [11]. COPD can be categorized into four
stages based on the different extent of airflow limitation, symptoms
like shortness of breath, and frequency of exacerbation
[12]. Immune response is also correlated with COPD progression
[13, 14]. Neutrophil numbers in sputum and bronchoalveolar
lavage fluid (BALF) of COPD patients is positively correlated
with the disease severity. This is especially true in process of COPD
exacerbation, which is usually complicated by respiratory infections
when increasing amounts of neutrophils migrating into airway or
lung [15–17]. COPD patients at GOLD stage I, who usually
have sustained airflow limitation, however, the majority do not
progress to advanced stages [18]. Chronic bronchitis and asthma
are also risk factors for pulmonary emphysema, thus it can be
included into this stage (Fig. 2). The initial steps that activating
and spreading of the innate and adaptive immune responses in early
COPD are still not very clear. At first, it was thought that innate
immune inflammation was predominant in the mild stage of
COPD. But now, it is believed that innate immune and adaptive
immune may act simultaneously since individual susceptibility to
several risk factors is different, that is the reason that young popu-
lation also could develop airflow obstruction [19, 20]. Both innate
Fig. 2 The symptoms and inflammatory phenotype in different COPD stage
immunity and adaptive immunity events are involved in mild

COPD. Innate lymphoid cells (ILCs), alveolar macrophages, neu-
trophils, dendritic cells, and CD8+ T cells are the original immuno-
logic events induced by external triggers (Fig. 2). All these
immunity responses belong to Th1 type immune response and
protect lungs against bacteria, viruses, or other intracellular
microbes. Th1 cells express chemokine receptors such as CCR5
and CXCR3 and are responsible for secreting chemokine such as
IL-2, IFN-γ, and also [21, 22]. Emphysema is the symptom of III
or IV stage COPD, contributing to the loss of lung function.
Patients with emphysema were proved to have high percent
of CD4+ and CD8+ T lymphocytes which expressed the T helper
1 cell (Th1) marker such as CCR5, CXCR3, and secreted more
interferon gamma (IFNγ), interferon-inducible protein 10 (IP-10),
and monokine induced by interferon gamma (MIG), which means
Th1 type immune response was predominant to regulate migration
of immune cells to specific loci [13, 23]. In addition to Th1 type
immune response, Th17 type immune response is also important in
COPD. Th17 type immune cells including CD4+ Th17 cells, CD8
+ Tc17 cells, and group 3 innate lymphoid cell (ILC3s) are stimu-
lated by IL-6 and IL-23 and differentiated from Th0 cells. Th17
cells can secrete IL-17, TNF-α, and IL-22, which can collectively
recruit and activate neutrophils [24] and protect lung against extra-
cellular bacteria and fungi [25].
3 Inflammatory and Immunological Features of Neutrophil-Predominant COPD
The main pathologic events during COPD process includes airway

lumens remodeling, airway epithelial junction impairation and loss
of lung function. Airway damage leads to a pro-inflammatory when
various pro-inflammatory mediators including IL-6, IL-8, tumor
necrosis factor (TNF)-α, tumor necrosis factor (TNF), and vascular
endothelial growth factor (VEGF) are released by damaged airway
epithelial cells. Immune cells like neutrophils were first recruited to
release more pro-inflammation chemokines to attract more innate
cells and trigger subsequent adaptive immunity. Neutrophil num-
bers and neutrophil-derived proteins like neutrophil MPO increase
depending on the inflammatory mediator levels such as IL-6,
IL-1β, TNFα, GM-CSF, and so on. A cohort study showed that a
significant positive correlation was found between the sputum
neutrophil numbers and faster FEV1%Pred decline in patients
with moderate-to-severe COPD, while no significant correlation
was found between macrophage or lymphocyte numbers and rapid
lung function decline in the same patients [26]. Therefore, neu-
trophils rather than other immune cells in COPD are suitable for
assessing the COPD status. In neutrophil-predominant COPD
patients, neutrophil counts have been increased in both blood
and sputum. COPD in mild stage sustains a low level of slightly
increased neutrophils and macrophages, while COPD patients in a
more serious stage are usually accompanied by abundant neutro-
phils and B lymphocytes infiltration [27]. In neutrophil-
predominant COPD, inflammatory mediators such as IL-6,
TNFα, IL-1β, CXCL8, GM-CSF, and CRP are responsible for
neutrophil priming and subsequently activating more immune
cells in circulating systerm. In peripheral blood of COPD patients,
several abnormal alterations in neutrophils function have been
identified, including expression of cell surface receptors, degranu-
lation, phagocytosis, and chemotaxis. In COPD, neutrophils show
enhanced chemotaxis [28], increased degranulation as evidenced
by increased MPO and neutrophil elastase [16, 29]. Thus, the
neutrophils in COPD appear to be functionally primed and
immune response enhanced, suggesting that chronic neutrophil
activation increases its effector responses as well as worsen systemic
inflammation, both of which contribute to the COPD
pathogenesis.
Various inflammatory cells also have been proved to express
distinct chemokine receptors by which they can gather around
invading pathogens, regulating immune responses. Neutrophils
are firstly recruited to damaged sites of where its immune phagocy-
tosis is triggered, then various antimicrobial products are released
to recruit more neutrophils and other immune cells [30]. Neutro-
phils express several kinds of surface receptors to recognize invaded
pathogen, including G-protein coupled chemokine and chemoat-

tractant receptors, Fc-receptors, innate immune receptors like Toll-
like receptors and C-type lectins, as well as cytokine receptors
[31]. G-protein coupled receptors expressed on neutrophils
include formyl-peptide receptors such as FPR1, FPR2, and FPR3,
classical chemoattractant receptors such as BLT1, BLT2, PAFR,
and C5aR, and chemokine receptors such as CXCR1, CXCR2,
CCR1, and CCR2. NOD-like receptors such as NOD2 and
NLRP3 and RIG-like receptors such as RIG-1 MDA5 are also
expressed on neutrophils. In addition, although some chemokine
receptors like CXCR4 does not express on neutrophils, it can still
play roles in neutrophils mobilization [32]. The G-protein coupled
formyl peptide receptors on neutrophils can directly mediate its
phagocytosis, by which neutrophils catch and engulf bacteria
[33]. In COPD airway, immunologic events such as innate immune
deficit in bacterial phagocytosis and progressive decrease in lung
microbial diversity also occurred [34]. Recent study proved that
this was related to the formation of neutrophil extracellular traps
(NETs) and increased NET complexes will reduce airway neutro-
phil phagocytosis [35]. Formation of neutrophil extracellular traps
(NETs) was first discovered in 2004. During NETs formation,
numerous proteins such as myeloperoxidase (MPO), neutrophil
elastase (NE), and histones were released, all of which attribute
antimicrobial properties to NETs [36, 37]. In patients of severe
stage COPD, NETs are more abundant and are associated with
reduced microbiota diversity and increased abundance of Haemo-
philus species [35].
4 Crosstalk Between Bacteria and Neutrophil in COPD
Accumulating evidence links microbiota to pulmonary disease.

Bacterial pathogens can reach and colonize lower airways of
COPD or asthma patients through respiration [38, 39]. Microbiota
plays a vital part in pathogen-associated molecular pattern (PAMP)
shaping immune function and response to pathogenic stress
[40]. Insights into airway inflammation and the reduced micro-
biome diversity have been derived from lung specimens and spu-
tum. When compared to healthy individuals, individuals with
COPD show decreased pulmonary bacterial diversity and gene
richness [35]. Airways of COPD patients are colonized by com-
mensal bacterial microbiota containing gram-negative pathogenic
Proteobacteria, Bacteroidetes, Actinobacteria, Firmicutes, Streptococ-
cus, Prevotella, Moraxella, Haemophilus, Acinetobacter, Fusobacter-
ium, Veillonella, and Neisseria [39, 41, 42]. Among these
microbes, H. influenzae was the most prevalent and the only path-
ogen persistently exists in patients with stable COPD [7].
Neutrophilic inflammation is the most common inflammatory

phenotype in COPD. Following exposure to the external triggers
of COPD such as smoking, there is airway damage and onset of
damage-associated molecular patterns (DAMPs). Murine cigarette
smoke-induced COPD models show that H. influenzae enhances
airway inflammation characterized by producing pro-inflammatory
mediators TNF-α, IL-6, and IL-1β, and inducing airway neutro-
philia infiltration [43]. Prevotella species are common colonizers of
airway, which have weak innate stimulatory properties and their col-
onization in airway is tolerable by the respiratory immune system
[44]. Proteobacteria are pathogenic bacteria of airway and have
innate stimulatory capacity to drive COPD progression [44].
Bacteria contain several compounds including the microbe-
associated molecular pattern (MAMP), which can activate immune
response. LPS is a cell membrane constituent of Gram-negative
bacteria and is considered to be a potent MAMP. Toll-like receptor
is the major innate receptor of LPS, and most pro-inflammatory
innate response of human leucocytes to pathogenic gram-negative
bacteria is also TLR4 and TLR2 dependent [45]. Different bacteria
trigger different innate respiratory immune responses.
Non-typeable Haemophilus influenza induced severe Toll-like
receptor 2 (TLR2)-independent inflammation which is character-
ized by predominant airway neutrophil and neutrophilic cytokine/
chemokine expression [44]. While Prevotella nanceiensis, one of the
most common bacteria in airway, can diminish neutrophilic airway
inflammation via TLR2 [41, 44]. Haemophilus influenzae can trig-
ger innate immune response as well as mobilizing intrapulmonary
neutrophils via TLR4 [46]. Moraxella catarrhalis, another com-
mon gram-negative bacteria which can as a cause of respiratory tract
infections, can trigger inflammatory immune response via TLR2
[47]. Gram-negative anaerobic Prevotella species were reduced in
COPD and asthma, most Prevotella species rarely cause respiratory
infections and as we have mentioned above, Prevotella nanceiensis
could diminish neutrophilic airway inflammation [41, 48]. How-
ever, Proteobacteria, which has a increasing relative abundance in
COPD, was considered as a co-drivers of COPD besides Prevotella
species. Also, it could potently mediated airway neutrophilia pro-
duction (Fig. 3). Meanwhile, metabolites from activated immune
cells also could promote growth of Proteobacteria [49]. Veillonella
species are also associated with increasing host inflammation during
COPD progress to later stages. Haemophilus species, one impor-
tant specific Proteobacteria in COPD, whose relative abundance was
positively associated with an increased volume fraction of neutro-
phils in lung [50]. Moraxella species, another specific Proteobac-
teria in COPD, whose abundance in airway is also associated with
increased airway neutrophils [51].
Fig. 3 Microbes of neutrophil-predominant COPD
The role of bacterial metabolites within the lung has not yet
been studied well, but there is some indication that unique meta-
bolic milieu which was created by lung microbiota will promote
neutrophil-mediated inflammation and Th17 type immune
response (Fig. 3) [52]. However, there are still bioactive bacterial
metabolites like glycolic acid and indol-3-acetate, both of which
have anti-inflammatory effects [53]. These metabolites provides
alternative options to alleviate the inflammation of COPD through
remeding dysbacteriosis.
5 Summary
Neutrophil-predominant COPD is a borderless stage between

moderate and extremely severe stage. Neutrophil in COPD also
works as a link point between DAMP, PAMP, and MAMP. Lung
microbiota is associated with the phagocytosis, degranulation, and
chemotaxis of neutrophil, and holds the potential to distinguish the
endotypes of COPD as new biomarker. Considering the changes of

multiple genera, colonization of airway could no doubt add addi-
tional accuracy to disease progression risk assessment. With the
development of technology, detecting the microbiome profiles by
amplifying bacterial 16S rRNA genes and sequencing them is very
convenient and costless. Also sputum qPCR has proved to be a
sensitive way to measure airway bacteria and identify neutrophilic
inflammation in stable COPD patients [7]. Therefore, comprehen-
sive consideration of the counts of neutrophil in peripheral blood
and sputum, as well as the resident microbiota profiles in sputum or
BALF, developing a personalized anti-dysbacteriosis strategy may
play a role in COPD management.
References
1. Bafadhel M et al (2011) Acute exacerbations of 10. Singh D et al (2019) Global strategy for the
chronic obstructive pulmonary disease: identi- diagnosis, management, and prevention of
fication of biologic clusters and their biomar- chronic obstructive lung disease: the GOLD
kers. Am J Respir Crit Care Med 184 science committee report 2019. Eur Respir J
(6):662–671 53(5):1900164
2. Bafadhel M et al (2018) Predictors of exacer- 11. Menezes AM et al (2017) The PLATINO
bation risk and response to budesonide in study: description of the distribution, stability,
patients with chronic obstructive pulmonary and mortality according to the Global Initiative
disease: a post-hoc analysis of three randomised for Chronic Obstructive Lung Disease classifi-
trials. Lancet Respir Med 6(2):117–126 cation from 2007 to 2017. Int J Chron
3. Brightling C, Greening N (2019) Airway Obstruct Pulmon Dis 12:1491–1501
inflammation in COPD: progress to precision 12. Han MK et al (2013) GOLD 2011 disease
medicine. Eur Respir J 54(2):1900651 severity classification in COPDGene: a pro-
4. Ghebre MA et al (2018) Biological exacerba- spective cohort study. Lancet Respir Med 1
tion clusters demonstrate asthma and chronic (1):43–50
obstructive pulmonary disease overlap with dis- 13. Grumelli S et al (2004) An immune basis for
tinct mediator and microbiome profiles. J lung parenchymal destruction in chronic
Allergy Clin Immunol 141(6):2027–2036.e12 obstructive pulmonary disease and emphy-
5. Willemse BWM et al (2005) Effect of 1-year sema. PLoS Med 1(1):e8–e8
smoking cessation on airway inflammation in 14. Baines KJ, Simpson JL, Gibson PG (2011)
COPD and asymptomatic smokers. Eur Respir Innate immune responses are increased in
J 26(5):835–845 chronic obstructive pulmonary disease. PLoS
6. Morris A et al (2013) Comparison of the respi- One 6(3):e18426–e18426
ratory microbiome in healthy nonsmokers and 15. Barnes PJ (2008) Immunology of asthma and
smokers. Am J Respir Crit Care Med 187 chronic obstructive pulmonary disease. Nat
(10):1067–1075 Rev Immunol 8(3):183–192
7. Bafadhel M et al (2015) Airway bacteria 16. Quint JK, Wedzicha JA (2007) The neutrophil
measured by quantitative polymerase chain in chronic obstructive pulmonary disease. J
reaction and culture in patients with stable Allergy Clin Immunol 119(5):1065–1071
COPD: relationship with neutrophilic airway 17. Stănescu D et al (1996) Airways obstruction,
inflammation, exacerbation frequency, and chronic expectoration, and rapid decline of
lung function. Int J Chron Obstruct Pulmon FEV1 in smokers are associated with increased
Dis 10:1075–1083 levels of sputum neutrophils. Thorax 51
8. Hussain M et al (2018) Enzalutamide in men (3):267–271
with nonmetastatic, castration-resistant pros- 18. Vestbo J, Lange P (2002) Can GOLD Stage
tate cancer. N Engl J Med 378(26):2465–2474 0 provide information of prognostic value in
9. Wang Z et al (2016) Lung microbiome dynam- chronic obstructive pulmonary disease? Am J
ics in COPD exacerbations. Eur Respir J 47 Respir Crit Care Med 166(3):329–332
(4):1082–1092
19. Suzuki M et al (2017) The cellular and molec- 35. Dicker AJ et al (2018) Neutrophil extracellular
ular determinants of emphysematous destruc- traps are associated with disease severity and
tion in COPD. Sci Rep 7(1):9562–9562 microbiota diversity in patients with chronic
20. Martinez FD (2016) Early-Life Origins of obstructive pulmonary disease. J Allergy Clin
Chronic Obstructive Pulmonary Disease. N Immunol 141(1):117–127
Engl J Med 375(9):871–878 36. Brinkmann V et al (2004) Neutrophil extracel-
21. Luther SA, Cyster JG (2001) Chemokines as lular traps kill bacteria. Science (New York, N.
regulators of T cell differentiation. Nat Immu- Y.) 303(5663):1532–1535
nol 2(2):102–107 37. Remijsen Q et al (2011) Dying for a cause:
22. Loetscher P et al (1998) CCR5 is characteristic NETosis, mechanisms behind an antimicrobial
of Th1 lymphocytes. Nature 391 cell death modality. Cell Death Differ 18
(6665):344–345 (4):581–588
23. Panina-Bordignon P et al (2001) The C-C 38. Erb-Downward JR et al (2011) Analysis of the
chemokine receptors CCR4 and CCR8 identify lung microbiome in the “healthy” smoker and
airway T cells of allergen-challenged atopic in COPD. PLoS One 6(2):e16384–e16384
asthmatics. J Clin Invest 107(11):1357–1364 39. Wang L et al (2017) Role of the lung micro-
24. Pedraza-Zamora CP et al (2017) Th17 cells biome in the pathogenesis of chronic obstruc-
and neutrophils: close collaborators in chronic tive pulmonary disease. Chin Med J 130
Leishmania mexicana infections leading to dis- (17):2107–2111
ease severity. Parasite Immunol 39(4). https:// 40. Maynard CL et al (2012) Reciprocal interac-
doi.org/10.1111/pim.12420 tions of the intestinal microbiota and immune
25. Annunziato F, Romagnani C, Romagnani S system. Nature 489(7415):231–241
(2015) The 3 major types of innate and adap- 41. Cabrera-Rubio R et al (2012) Microbiome
tive cell-mediated effector immunity. J Allergy diversity in the bronchial tracts of patients
Clin Immunol 135(3):626–635 with chronic obstructive pulmonary disease. J
26. Donaldson GC et al (2005) Airway and sys- Clin Microbiol 50(11):3562–3568
temic inflammation and decline in lung func- 42. Caverly LJ, Huang YJ, Sze MA (2019) Past,
tion in patients with COPD. Chest 128 present, and future research on the lung micro-
(4):1995–2004 biome in inflammatory airway disease. Chest
27. Caramori G et al (2016) COPD immunopa- 156(2):376–382
thology. Semin Immunopathol 38(4):497–515 43. Moghaddam SJ et al (2008) Haemophilus
28. Burnett D et al (1987) Neutrophils from sub- influenzae lysate induces aspects of the chronic
jects with chronic obstructive lung disease obstructive pulmonary disease phenotype. Am
show enhanced chemotaxis and extracellular J Respir Cell Mol Biol 38(6):629–638
proteolysis. Lancet (London, England) 2 44. Larsen JM et al (2015) Chronic obstructive
(8567):1043–1046 pulmonary disease and asthma-associated Pro-
29. Noguera A et al (2001) Enhanced neutrophil teobacteria, but not commensal Prevotella
response in chronic obstructive pulmonary dis- spp., promote Toll-like receptor
ease. Thorax 56(6):432–437 2-independent lung inflammation and pathol-
30. Kolaczkowska E, Kubes P (2013) Neutrophil ogy. Immunology 144(2):333–342
recruitment and function in health and inflam- 45. Elson G et al (2007) Contribution of Toll-like
mation. Nat Rev Immunol 13(3):159–175 receptors to the innate immune response to
31. Futosi K, Fodor S, Mócsai A (2013) Neutro- Gram-negative and Gram-positive bacteria.
phil cell surface receptors and their intracellular Blood 109(4):1574–1583
signal transduction pathways. Int Immuno- 46. Wang X et al (2002) Toll-like receptor 4 med-
pharmacol 17(3):638–650 iates innate immune responses to Haemophilus
32. De Filippo K, Rankin SM (2018) CXCR4, the influenzae infection in mouse lung. J Immunol
master regulator of neutrophil trafficking in 168(2):810–815
homeostasis and disease. Eur J Clin Invest 48 47. Slevogt H et al (2007) Moraxella catarrhalis is
(Suppl 2):e12949–e12949 internalized in respiratory epithelial cells by a
33. Wen X et al (2019) G-protein-coupled formyl trigger-like mechanism and initiates a TLR2-
peptide receptors play a dual role in neutrophil and partly NOD1-dependent inflammatory
chemotaxis and bacterial phagocytosis. Mol immune response. Cell Microbiol 9
Biol Cell 30(3):346–356 (3):694–707
34. Soriano JB, Polverino F, Cosio BG (2018) 48. Nagy E (2010) Anaerobic infections: update
What is early COPD and why is it important? on treatment considerations. Drugs 70
Eur Respir J 52(6):1801448 (7):841–858
49. Scales BS, Dickson RP, Huffnagle GB (2016) A phenotype in severe preschool wheeze. J
tale of two sites: how inflammation can reshape Allergy Clin Immunol 143(4):1607–1610.e3
the microbiomes of the gut and lungs. J Leu- 52. Segal LN et al (2016) Enrichment of the lung
koc Biol 100(5):943–950 microbiome with oral taxa is associated with
50. Sze MA et al (2015) Host response to the lung lung inflammation of a Th17 phenotype. Nat
microbiome in chronic obstructive pulmonary Microbiol 1:16031–16031
disease. Am J Respir Crit Care Med 192 53. Segal LN et al (2017) Randomised, double-
(4):438–445 blind, placebo-controlled trial with azithromy-
51. Robinson PFM et al (2019) Lower airway cin selects for anti-inflammatory microbial
microbiota associates with inflammatory metabolites in the emphysematous lung. Tho-
rax 72(1):13–22
Chapter 17
Genome Variation and Precision Medicine in Systemic

Lupus Erythematosus
Ru Yang, Yaqi Hu, and Lin Bo
Abstract
Systemic lupus erythematosus (SLE) is a complex autoimmune disease which is facing the difficulties in
treatment. Genetics play an important role in SLE. Several studies have shown that genetic factors not only
affect the development of SLE, but also affect its clinical progress. In this review article, we focus on
exploring the influence of genetics on different aspects of SLE pathogenesis, clinical course, and treatment
and will provide some references in further precision medicine for SLE patients. The coming era of precision
medicine, SLE patients will be stratified by genetic profiling. This will enable us to make more effective and
precise choices of treatment plan.
Key words SLE, Genome variation, GWAS, Precision medicine
1 Introduction
Systemic lupus erythematosus (SLE) is a complex autoimmune

disease, which is characterized by the destruction of immune toler-
ance, which promotes the formation of auto-reactive B and T cells,
the abnormal production of cytokines, and the subsequent produc-
tion of autoantibodies against DNA- and RNA-based self-antigens
[1, 2].
It has been proposed treat to target in immunotherapy of SLE,
with the goal of clinical remission or low disease activity that pre-
dicts favorable long-term outcomes [3]. However, only a small
number of patients can achieve the goal of prolonged remission
or low disease activity [4, 5]. Therefore, immunotherapy is facing
the difficulties of unpredictable and frequent relapses of active SLE,
flares and maintenance of remission.
Previous studies have proved that genetics play an important
role in SLE. Several studies have shown that genetic factors not only
Ru Yang and Yaqi Hu contributed equally to this work.
193
194 Ru Yang et al.
affect the development of SLE, but also affect its clinical progress.
Especially, the research of genome-wide association (GWAS) has
increased dramatically, and more than 100 loci with significant
GWAS with SLE have been identified [6]. In this review article,
we focus on exploring the influence of genetics on different aspects
of SLE pathogenesis, clinical course, and treatment and will provide
some references in further precision medicine for SLE patients.
2 Genetic Pathogenesis of SLE
In the past few years, humans have made great progress in the
pathophysiological identification of SLE, but there are still many
problems. It is generally believed that SLE is caused by the recog-
nition of nuclear antigen by the immune system. Some environ-
mental factors, such as ultraviolet (UV) light, toxins, and infection,
lead to cellular apoptosis and damage in clearance of apoptotic
bodies, leading them to be recognized by both innate and adaptive
immune system [7]. Autoantibodies against these nuclear antigens
lead to the construction of immune complex (IC), which deposits
on susceptible organs in susceptible organs (such as glomerulus) in
SLE patients because of the defective IC clearance mechanism. The
interaction of immune cells with these ICs leads to clinical mani-
festations and further tissue damage in SLE patients. But there is
still little evidence that can link the pathophysiology to the clinical
course of the disease. It enhances the difficulties in diagnosis and
treatment of SLE.
According to recent studies, the differences in the prevalence of
SLE in different ethnicities support the important role of genetics
in this disease [8, 9]. Consistent with this, a high prevalence of SLE
in affected individuals and monozygotic twins further supports this
issue [10]. According to a population-based study of 23 million
participants in Taiwan, the relative risk of SLE was 315.94 for twins
and 23.68 for siblings [11]. These results suggest that genetics play
an important role in SLE susceptibility.
2.1 Monogenic SLE Identification of causal mutations in monogenic SLE or lupus-like

autoimmune syndrome provides important clues for the pathogen-
esis of SLE. Lupus-prone families with primary defects of classical
complement pathways (C1q, C1r/s, C2, C4A and C4B) reveal the
importance of this pathway in the pathogenesis of lupus [12]. Com-
plement is essential for the regulation and clearance of immune
complex and apoptotic cells.
In addition, mutations in genes involved in DNA processing
during apoptosis lead to lupus-like systemic autoimmunity [13]. In
20 patients with SLE, there were 2 cases with decreased DNASE1
activity due to heterozygous nonsense DNASE1 mutation
[14]. DNASE1L3 mutation resulted in familial SLE with positive
ANA, anti-dsDNA antibody, anti-neutrophil cytoplasmic antibody
(ANCA) and low C3/C4 [15].
Genome Variation and Precision Medicine in Systemic Lupus Erythematosus 195
Furthermore, it was reported that a homozygous missense

mutation of PRKCD homozygote encoding protein kinase (PKC)
in a juvenile-onset SLE family [16]. PKC is a serine/threonine
kinase, which is related to the negative selection of B cells and
control of cell proliferation [17, 18]. These insights highlight the
central role of complement pathway, nucleic acid metabolism, and
self-reactive B cells in the pathogenesis of human SLE.
2.2 Polygenic SLE The GWAS study includes the screening of the association between
loci and common multifactor diseases (such as SLE). More than
100 SNPs have been proved to be closely related to SLE, most of
which are located in the noncoding region and affect gene expres-
sion through transcription or epigenetic modification
[19, 20]. Some reported genes are related to abnormal recognition
of self-nucleic acids (Ncf1, Ncf2, FCGR2A, ITGAM, etc.), type I
IFN overproduction/TLR signaling (IFIH1, IRF5, TNFAIP3,
etc.), and defective immune cell signaling (BLK, TNFSF13B,
etc.). Human leukocyte antigen (HLA) was involved in antigen
presentation, complement components C2 and C4, and cytokines
TNF-α [6, 19]. Most of the genes in this region are related to
immunity. In different ethnic populations, HLA-DR and—DQ
loci have the same correlation with SLE [20–22]. GWAS results
also strongly support the participation of non-HLA III region
genes in SLE. SLE-GWAS SNPs are enriched in B cell- and T cell-
specific gene expression and epigenetic enhancer markers [22,
23]. Additionally, SLE is a kind of clinical heterogeneous disease.
Some phenotype-related loci have been reported, such as PDGRFA
in lupus nephritis and ITGAM in arthritis [24, 25]. However, the
genetic structure of SLE subtypes has not been fully elucidated.
The further analysis of GWAS and clinical subphenotypes can iden-
tify new sites of association.
2.3 SLE-GWAS Here, we introduce some SLE-GWAS candidate gene. The con-
Candidate Gene firmed loci are organized by the dominant pathway they are
involved in.
2.3.1 MHC Gene Human leukocyte antigen (HLA) is involved in the process of T cell
antigen presentation. Therefore, HLA polymorphism is related to a
variety of autoimmune diseases including SLE. Genetic analysis
supports the association between HLA class II polymorphism and
SLE [26]. Several studies showed that HLA-DRB1*0301 allele,
HLA-DR3-DQ2 haplotype, HLA-DR2, HLA-DR3, DR9, DR15,
HLA-DRB1*0301, HLA-DRB1*0801, and HLA-DQA1*0102
alleles were related to disease susceptibility, while
HLA-DRB1*1101allele, HLA-DR4, HLA-DR5, DR11, DR14
alleles had protective effects [27]. In addition, HLA-DR4 and
DR11 alleles are protective variants of lupus nephritis (LN), while
HLA-DR3 and DR15 alleles are related to renal involvement.
196 Ru Yang et al.
The relationship between MHC gene polymorphism and SLE

susceptibility was also found in GWAS. According to the European
studies, the relationship between HLA-B*0801,
HLA-DQA1*0501, HLA-DQB1*0201, HLA-DRB1*0301,
HLA-DRB3*01, HLA-DRB3*02 alleles, DR17 (broad antigen
DR3), B8 antigen, and SLE was reported. However,
HLADQB1*0301, DQ7 (broad antigen DQ3) seem to have pro-
tective effects [22, 28–30].
The relationship between MHC gene polymorphism and clini-
cal manifestations of SLE has not been studied in detail, but there is
limited evidence that these genetic variations can also affect the
progress of SLE. A study of Chinese Han population showed that
rs3077 and rs9277535 in HLA-DP gene were related to the risk of
SLE. In addition, rs3077 polymorphism was associated with cuta-
neous vasculitis, serum IL17, and INF-γ levels. Another study in
Hungarian patients with SLE showed that the HLA-DRB1*03 and
DRB1*07 alleles are associated with LN, while HLA-DRB1*1501
appears to be a protective allele [31]. In addition, HLA-DRB1*07
was positively correlated with serositis, severe renal, and cardiopul-
monary damage, and seemed to be related to fatal manifestations.
In addition, HLA DRB1*04 and DRB1*1112 alleles were asso-
ciated with drug-resistant leukopenia and discoid lupus erythema-
tosus (DLE), respectively. In another study of patients with Arabian
SLE, HLA-DRB1*10, DRB1*11, DQB1*03, and DRB1*15
alleles were found to be related to hematology, neurology, skin
and kidney diseases, respectively [32]. In addition, serositis is
related to HLA-DRB3 and DRB1*11 alleles. In Portuguese SLE
patients, the risk of nervous system involvement in HLA-DRB1*08
carriers was almost four times higher, and HLA DRB1*03 allele
was positively correlated with LN [33]. In addition,
HLA-DRB1*01 alleles have been reported to be significantly over-
expressed in patients with SLE neurological disease [34]. In gen-
eral, the association between MHC gene variation and SLE clinical
manifestations has been reported almost all over the world, but
these associations seem to be very different in the population.
2.3.2 Non-MHC Gene IRF5 interferon regulatory factor 5 (IRF5) is one of the most
important genes related to SLE. The increase of IFN-α and its
IRF5 gene transcripts in blood cells are related to the severity of the
disease. The IRF5 gene encodes a transcription factor that induces
the expression of IFN-related proteins and type I interferon
[35]. The single nucleotide of IRF5 gene was studied in four
independent SLE case-control cohorts [36]. It has been reported
that T allele of the rs2004640 SNP of IRF5 haplotypes contribute
to the expression of multiple IRF5 subtypes, increased expression
of IRF5, and increased risk of SLE susceptibility [36]. Another
study found that multiple over transmitted haplotypes may be
associated with the involvement of IRF5 in SLE [37]. Further case-

control analysis showed that single nucleotide polymorphisms of
IRF5 gene, such as rs2070197, were strongly associated with SLE
risk, especially in the Latin American population [38]. Several years
later, 3230 common variants of IRF5-TNPO3 were studied in
more than 8000 SLE patients and 7000 healthy controls of differ-
ent races supports the hypothesis that risk variation of IRF5 gene
increases gene expression [39]. A recent study has shown that
different functional SNPs contribute to the establishment of dan-
gerous haplotypes and have a cumulative effect on the risk of
SLE [40].
STAT4 STAT4 transcription 4 (STAT4) gene signaling and activators were

first proposed in 2007 to be associated with SLE [41]. Further
studies confirmed that STAT4 rs3821236, rs3024866, rs7574865,
rs3024896, and rs7601754 SNPs were associated with SLE. In
addition, STAT4 rs7582694 CC and CG genotypes are associated
with SLE in adolescents in the Egyptian population [42]. Another
study in patients of European descent suggested that STAT4 poly-
morphism might be associated with LN [43], another group exam-
ined the association between STAT4 gene and LN in two Swedish
cohorts by GWAS [44]. In addition, STAT4 rs7582694 SNPs were
associated with severe renal complications [44]. The study also
showed that in SLE patients, STAT4 susceptible sites are closely
related to the presence of anti-dsDNA [43]. Based on these find-
ings, the association of STAT4 rs7582694 CC genotype with skin
manifestations, proteinuria, Ana, and anti-dsDNA positivity was
reported in Egyptian adolescents with SLE [42]. Further study of
the animal model may help to elucidate the role of STAT4 and its
signaling pathway in the pathogenesis of SLE.
ITGAM There was a significant correlation between polymorphisms of

integrin alpha M (ITGAM) and SLE. The alleles rs1143678,
rs4548893, rs9888739, rs1143679, and rs1143683 were identi-
fied by GWAS in European women with SLE, which were related to
disease susceptibility [45]. There is limited data on the association
of ITGAM polymorphism with clinical manifestations of SLE. A
study of patients of European descent showed that the secondary
allele of rs1143679 polymorphism was positively correlated with
LN, discoid rash, and immune performance (including anti-
dsDNA and anti-ribonucleoprotein antibody (anti-nRNP)
[46]. In another study of Asian patients, 13 SNPs, including
rs1143679, were confirmed to be associated with LN. On the
contrary, they were negatively correlated with discoid rash, but
anti-SM antibody was positive. In Hong Kong and Thailand SLE
patients, the small alleles of rs1143679 and rs1143663 polymor-
phism were linked with LN, and the small alleles of rs1143680 and
198 Ru Yang et al.
rs1143678 SNP were positively correlated with neural invasion

[47]. Therefore, these polymorphisms have an important impact
on the development of LN.
BANK1 The relationship between B cell scaffold protein with ankyrin

repeats (bank1) and SLE susceptibility was also studied. Nine
SNPs were identified in Swedish SLE patients: rs4522865 (A),
rs4572885 (T), rs10516487 (G), rs10516486 (C), rs17200824
(A), rs6849308 (C), rs10516482 (C), rs10516483 (C), and
rs2631271 (G) [48]. The association of rs10516487,
rs17266594, and rs3733197 was replicated in Argentina, Ger-
many, Italy, and Spain, with a combined odds ratio of 1.38, 1.42,
and 1.23, respectively. The association of rs10028805 with SLE
was previously reported in another GWAS for European
patients [49].
The clinical significance of BANK1 gene polymorphism has not
been studied in detail, but there is limited evidence, suggesting that
the single nucleotide polymorphism of BANK1 may affect the
disease subtype. A European cohort study of SLE reported that
different BANK1 SNPs were associated with different SLE-specific
autoantibodies and clinical manifestations [50]. In the SNPs men-
tioned earlier, rs17266594 and rs10516487 were negatively corre-
lated with immune disorders and kidney involvement. In addition,
risk variants of BANK1 are described to be associated with anti-
dsDNA-positive SLE cases, but not with their anti-dsDNA-nega-
tive counterparts [51].
3 Potential Clinical Value of Genetic in SLE
It is still a long way in the era of using personalized treatments to

cure SLE patients. Nevertheless, some cases have been reported
that this approach has made gratifying progress. For example, a
recent study reported a 4-year-old SLE patient with Malayan rash,
arthritis, high ANA titer, anti-dsDNA, anticardiolipin antibody,
leukopenia, and Coombs-positive anemia at 3 years old [52]. At
4 years old, she developed right hemiplegia and was diagnosed with
SLE vasculitis/encephalitis based on irregular medium-size vessels
in MRA. It was identified an R97H homozygous mutation in three
prime repair exonuclease 1 (TREX1) gene. TREX1 is an extracellu-
lar cleavage enzyme, which can increase serum IFN-α level. There-
fore, this patient is suitable for the treatment with anti-IFN-α
antibody, and genetic technology is helpful to find a suitable thera-
peutic target.
Recent studies proved that genetic polymorphism can well
predict the clinical progression of SLE [53]. According to the
meta-analysis, the polymorphism of immune complex clearance
pathway-related genes was significantly associated with
neuropsychiatric lupus erythematosus (NPSLE) [54]. In addition,

individuals with FCGR3A 158ff, Fcgr3B Na1/2, and
itgamrs1143679 HH genotypes were more likely to develop neu-
ropsychiatric symptoms of their disease [55]. Patients with the wild
type rs7925662 TT of TRCP6 gene had a greater risk of NPSLE
[55]. In vivo and in vitro studies have shown that anti-NMDA
antibodies can cause neuronal apoptosis and remodeling in SLE
patients, leading to neuropsychiatric symptoms [56]. These
insights imply that this pathway may be a novel candidate treatment
strategy for NPSLE.
Genetic polymorphism can also predict the outcome of treat-
ment in SLE patients. The relationship between HSP90AA1 gene
polymorphism and clinical response to glucocorticoid therapy in
SLE patients was studied [57]. The results showed that rs7160651,
rs10873531, and rs2298877 polymorphisms were related to glu-
cocorticoid response. A study also showed that CCCGAA-
CATCCC haplotypes of HSP90B1 gene were associated with
glucocorticoid efficacy [58]. In addition, it had been proved that
the patients with the AA/AG genotype in the 1082 position of
the IL10 gene and the AA/GA genotype in the 308 of the TNFA
gene had the best response to antimalarial drugs [59]. These SNPs
are located in the promoter regulatory region of related genes and
can regulate the basal concentrations of IL-10 and TNF-α.
The relationship between genetic polymorphism and LN treat-
ment outcome was also studied. In one study, it was found that the
CT genotype of glutathione S-transferase (GST) A1 gene during
cyclophosphamide therapy had a lower LN remission rate com-
pared to CC carriers [60]. In line with previous studies, ile105val
genotype of GSTP1 gene is an independent factor of renal dysfunc-
tion in LN patients treated with cyclophosphamide, and the null
genotype of GSTM1 gene is related to adverse drug reactions
[61]. Another study showed that the genetic polymorphism of Fc
fragment of IgG receptor (FCGR) gene cluster had an effect on
renal outcome after cyclophosphamide treatment. The results
showed that the carriers of rs6697139, rs10917686, and
rs10917688 alleles were located between FCGR2B and Fc receptor
like a (FCRLA) genes and had a low response to treatment
[62]. Therefore, drug therapy may be affected by gene mutation.
4 Future Perspectives and Conclusion
Since the first description, SLE has been recognized as an autoim-

mune disease that could affect nearly everybody organ. Its physio-
pathology is complex. Environmental factors, such as ultraviolet
light, can induce cell apoptosis and subsequent nuclear autoantigen
exposure to the immune system. The destruction of apoptotic body
clearing mechanism leads to its accumulation, which leads to the
200 Ru Yang et al.
destruction of tolerance of autoantibodies secreted by self-reactive

B cells. The formation of IC and its poor clearance lead to their
deposition in organs of anatomically susceptible people. ICs can
further activate adaptive and innate immune pathways, leading to
tissue damage and organ failure. More and more evidences have
suggested the footprint of genetic polymorphism is evident in all of
the above pathways. In addition, with the help of GWAS, some
SNPs are considered play impression roles in the pathogenesis of
SLE.
SLE has an extremely unpredictable clinical process. Every
physician cannot avoid facing life-threatening cases or treatment-
resistant cases although not every patient presents this type of
aggressive manifestation. Therefore, treatment of SLE is still facing
severe challenges and difficulties. However, genetic variation and
genetic pathway have contributed to a clearer understanding of the
role of these SNPs in the pathogenesis, clinical course, and treat-
ment of SLE and bring new chances. As mentioned before, some
SNPs have been detected that can predict the therapeutic response
to some SLE drugs. In addition, the association of these poly-
morphisms with specific clinical and serologic manifestations has
also been reported. Although all the studies are generally proved in
theory, and the conclusions have little practical application at
patient bedside, we still hope that future genetic research will
eventually become a tool that can better diagnosis and differential
diagnosis SLE patients. The coming era of precision medicine, SLE
patients will be stratified by genetic profiling. Each patient will be
longitudinally evaluated and predict the natural course of the dis-
ease. This will enable us to make more effective and precise choices
of treatment plan.
References
1. Rahman A, Isenberg DA (1976) Systemic 6. Cui Y, Sheng Y, Zhang X (2013) Genetic sus-
lupus erythematosus. Ryoikibetsu Shokogun ceptibility to SLE: recent progress from GWAS.
Shirizu 358(9281):585–585 J Autoimmun 41(Complete):25–33
2. Domeier PP, Schell SL, Rahman ZS (2017) 7. Tsokos GC, Lo MS, Reis PC et al (2016) New
Spontaneous germinal centers and autoimmu- insights into the immunopathogenesis of sys-
nity. Autoimmunity 50(1):4–18 temic lupus erythematosus. Nat Rev Rheuma-
3. Van VRF, Mosca M, Bertsias G et al (2014) tol 12(12):716–730
Treat-to-target in systemic lupus erythemato- 8. Rees F, Doherty M, Grainge MJ et al (2017)
sus: recommendations from an international The worldwide incidence and prevalence of
task force. Ann Rheum Dis 73(6):958–967 systemic lupus erythematosus: a systematic
4. Zen M, Iaccarino L, Gatto M et al (2015) review of epidemiological studies. Rheumatol-
Prolonged remission in Caucasian patients ogy 56:1945
with SLE: prevalence and outcomes. Ann 9. Lim SS, Bayakly AR, Helmick CG et al (2014)
Rheumatic Dis 74:2117–2122. https://doi. The incidence and prevalence of systemic lupus
org/10.1136/annrheumdis-2015-207347 erythematosus, 2002-2004: the Georgia
5. Urowitz MB, Feletar M, Bruce IN et al (2005) Lupus Registry. Arthritis Rheumatol 66
Prolonged remission in systemic lupus erythe- (2):357–368
matosus. J Rheumatol 32(8):1467–1472
10. Aslani S, Rezaei R, Jamshidi A et al (2018) autoimmune disease variants. Nature 518
Genetic and epigenetic etiology of autoim- (7539):337–343
mune diseases: lessons from twin studies. 24. Chung SA, Brown EE, Williams AH et al
Rheumatol Res 3:45–57 (2014) Lupus nephritis susceptibility loci in
11. Kuo CF, Grainge MJ, Valdes AM et al (2015) women with systemic lupus erythematosus. J
Familial aggregation of systemic lupus erythe- Am Soc Nephrol 25(12):2859–2870
matosus and coaggregation of autoimmune 25. Taylor KE, Chung SA, Graham RR et al (2011)
diseases in affected families. JAMA Intern Risk alleles for systemic lupus erythematosus in
Med 175(9):1518–1526 a large case-control collection and associations
12. Ghodke-Puranik Y, Niewold TB (2015) with clinical subphenotypes. PLoS Genet 7(2):
Immunogenetics of systemic lupus erythema- e1001311
tosus: a comprehensive review. J Autoimmun 26. Relle M, Schwarting A (2012) Role of
64:125–136 MHC-linked susceptibility genes in the patho-
13. Saeed M (2017) Lupus pathobiology based on genesis of human and murine lupus. Clin Dev
genomics. Immunogenetics 69(1):1–12 Immunol 2012(2):584374
14. Yasutomo K, Horiuchi T, Kagami S et al 27. Castano-Rodriguez N, Diaz-Gallo LM,
(2001) Mutation of DNASE1 in people with Pineda-Tamayo R et al (2008) Meta-analysis
systemic lupus erythematosus. Nat Genet 28 of HLA-DRB1 and HLA-DQB1 polymorph-
(4):313–314 isms in Latin American patients with systemic
15. Al-Mayouf SM, Sunker A, Abdwani R et al lupus erythematosus. Autoimmun Rev 7
(2011) Loss-of-function variant in DNA- (4):322–330
SE1L3 causes a familial form of systemic lupus 28. Niu Z, Zhang P, Tong Y (2014) Value of
erythematosus. Nat Genet 43(12):1186–1188 HLA-DR genotype in systemic lupus erythe-
16. Belot A, Kasher PR, Trotter EW et al (2013) matosus and lupus nephritis: a meta-analysis.
Protein kinase Cdelta deficiency causes mende- Int J Rheum Dis 18(1):17–28
lian systemic lupus erythematosus with B cell- 29. Morris DL, Taylor KE, Fernando MM et al
defective apoptosis and hyperproliferation. Unraveling multiple MHC gene associations
Arthritis Rheum 65(8):2161–2171 with systemic lupus erythematosus: model
17. Mecklenbrauker I, Saijo K, Zheng NY et al choice indicates a role for HLA alleles and
(2002) Protein kinase Cdelta controls self- non-HLA genes in Europeans. Am J Hum
antigen-induced B-cell tolerance. Nature 416 Genet 91(5):778–793
(6883):860–865 30. Zhang J, Zhan W, Yang B et al (2017) Genetic
18. Miyamoto A, Nakayama K, Imaki H et al polymorphisms of rs3077 and rs9277535 in
(2002) Increased proliferation of B cells and HLA-DP associated with systemic lupus
auto-immunity in mice lacking protein kinase erythematosus in a Chinese population. Sci
Cdelta. Nature 416(6883):865–869 Rep 7:39757
19. Deng Y, Tsao BP (2017) Updates in lupus 31. Endreffy E (2003) HLA class II allele polymor-
genetics. Curr Rheumatol Rep 19(11):68 phism in Hungarian patients with systemic
20. Bentham J, Morris DL, Graham C, Deborah S lupus erythematosus. Ann Rheum Dis 62
et al (2015) Genetic association analyses impli- (10):1017–1018
cate aberrant regulation of innate and adaptive 32. Wadi W, Elhefny NEAM, Mahgoub EH et al
immunity genes in the pathogenesis of systemic (2014) Relation between HLA typing and clin-
lupus erythematosus. Nat Genet 47 ical presentations in systemic lupus erythema-
(12):1457–1464 tosus patients in Al-Qassim region, Saudi
21. Sun C, Molineros JE, Looger LL et al (2016) Arabia. Int J Health Sci 8(2):159–165
High-density genotyping of immune-related 33. Vasconcelos C, Carvalho C, Leal B et al (2009)
loci identifies new SLE risk variants in indivi- HLA in portuguese systemic lupus erythema-
duals with Asian ancestry. Nat Genet 48 tosus patients and their relation to clinical fea-
(3):323–330 tures. Ann N Y Acad Sci 1173(1):575–580
22. Armstrong DL, Zidovetzki R, Alarcón- 34. Hachicha H, Kammoun A, Mahfoudh N et al
Riquelme ME et al (2014) GWAS identifies (2018) Human leukocyte antigens-DRB1*03
novel SLE susceptibility genes and explains is associated with systemic lupus erythematosus
the association of the HLA region. Genes and anti-SSB production in South Tunisia. Int
Immun 15(6):347–354 J Health Sci 12(1):21–27
23. Farh KH, Marson A, Zhu J et al (2014) 35. Rönnblom L, Eloranta ML, Alm GV (2006)
Genetic and epigenetic fine mapping of causal The type I interferon system in systemic lupus
202 Ru Yang et al.
erythematosus. Arthritis Rheum 54 47. Yang W, Zhao M, Hirankarn N et al (2009)

(2):408–420 ITGAM is associated with disease susceptibility
36. Graham RR, Kozyrev SV, Baechler EC et al and renal nephritis of systemic lupus erythema-
(2006) A common haplotype of interferon reg- tosus in Hong Kong Chinese and Thai. Hum
ulatory factor 5 (IRF5) regulates splicing and Mol Genet 18(11):2063–2070
expression and is associated with increased risk 48. Kozyrev SV, Abelson AK, Wojcik J et al (2008)
of systemic lupus erythematosus. Nat Genet 38 Functional variants in the B-cell gene BANK1
(5):550–555 are associated with systemic lupus erythemato-
37. Graham DSC, Manku H, Wagner S et al sus. Nat Genet 40(2):211–216
(2006) Association of IRF5 in UK SLE families 49. Bentham J, Morris DL, Cunninghame Graham
identifies a variant involved in polyadenylation. DS et al (2015) Genetic association analyses
Hum Mol Genet 16(6):579–591 implicate aberrant regulation of innate and
38. Reddy MVPL, Velázquez-Cruz R, Baca V et al adaptive immunity genes in the pathogenesis
(2007) Genetic association ofIRF5with SLE in of systemic lupus erythematosus. Nat Genet
Mexicans: higher frequency of the risk haplo- 47(12):1457–1464
type and its homozygozity than Europeans. 50. Guo L, Deshmukh H, Lu R et al (2009) Repli-
Hum Genet 121(6):721–727 cation of the BANK1 genetic association with
39. Kottyan LC, Zoller EE, Bene J et al (2015) The systemic lupus erythematosus in a European-
IRF5-TNPO3 association with systemic lupus derived population. Genes Immun 10
erythematosus has two components that other (5):531–538
autoimmune disorders variably share. Hum 51. Chung SA, Taylor KE, Graham RR et al (2011)
Mol Genet 24(2):582–596 Differential genetic associations for systemic
40. Raj P, Rai E, Song R et al (2016) Regulatory lupus erythematosus based on anti–dsDNA
polymorphisms modulate the expression of Autoantibody Production. PLoS Genet 7(3):
HLA class II molecules and promote autoim- e1001323
munity. eLife 5:e12089 52. Ellyard JI, Jerjen R, Martin JL et al (2016)
41. Remmers EF, Plenge RM, Lee AT et al (2007) Identification of a pathogenic variant in
STAT4 and the risk of rheumatoid arthritis and TREX1 in early-onset cerebral SLE by whole-
systemic lupus erythematosus. N Engl J Med exome sequencing. Pathology 48:S47
357(10):977–986 53. Sarah MG, Stewart W, Joanna W et al (2018)
42. Nageeb RS, Omran AA, Nageeb GS (et al, Neurological disease in lupus: toward a perso-
2018) STAT4 gene polymorphism in two nalized medicine approach. Front Immunol
major autoimmune diseases (multiple sclerosis 9:1146
and juvenile-onset systemic lupus erythemato- 54. Ho RC, Ong H, Thiaghu C, Lu Y et al (2016)
sus) and its relation to disease severity. Egypt J Genetic variants that are associated with neu-
Neurol Psychiatr Neurosurg 54(1):16 ropsychiatric systemic lupus erythematosus. J
43. Chung SA, Criswell LA (2008) PTPN22: its Rheumatol 43(3):541–551
role in SLE and autoimmunity. Autoimmunity 55. Ramirez GA, Lanzani C, Bozzolo EP et al
40(8):582–590 (2015) TRPC6 gene variants and neuropsy-
44. Bolin K, Sandling JK, Zickert A et al (2013) chiatric lupus. J Neuroimmunol 288:21–24
Association of STAT4 polymorphism with 56. Arinuma Y (2018) Antibodies and the brain:
severe renal insufficiency in lupus nephritis. anti-N-methyl-D-aspartate receptor antibody
PLoS One 8(12):e84450 and the clinical effects in patients with systemic
45. Harley JB, Alarcón-Riquelme ME, Criswell LA lupus erythematosus. Curr Opin Neurol
et al (2008) Genome-wide association scan in 31:294
women with systemic lupus erythematosus 57. Zou YF, Xu JH, Gu YY et al (2016) Single
identifies susceptibility variants in ITGAM, nucleotide polymorphisms of HSP90AA1
PXK, KIAA1542 and other loci. Nat Genet gene influence response of SLE patients to glu-
40(2):204–210 cocorticoids treatment. Springerplus 5(1):222
46. Kimhoward X, Maiti AK, Anaya JM et al 58. Sun XX, Li SS, Zhang M et al (2017) Associa-
(2010) ITGAM coding variant (rs1143679) tion of HSP90B1 genetic polymorphisms with
influences the risk of renal disease, discoid efficacy of glucocorticoids and improvement of
rash and immunological manifestations in HRQoL in systemic lupus erythematosus
patients with systemic lupus erythematosus patients from Anhui Province. Am J Clin Exp
with European ancestry. Ann Rheum Dis 69 Immunol 7(2):27–39
(7):1329
59. López P, Gómez J, Mozo L et al (2006) Cyto- polymorphisms are independent prognostic
kine polymorphisms influence treatment out- factors in lupus nephritis treated with cyclo-
comes in SLE patients treated with antimalarial phosphamide. PLoS One 11(3):e0151696
drugs. Arthritis Res Ther 8(2):R42 62. Kim K, Bang SY, Joo YB et al (2016) Response
60. Wang HN, Zhu XY, Zhu Y et al (2015) The to intravenous cyclophosphamide treatment
GSTA1 polymorphism and cyclophosphamide for lupus nephritis associated with polymorph-
therapy outcomes in lupus nephritis patients. isms in the FCGR2B-FCRLA locus. J Rheuma-
Clin Immunol 160(2):342–348 tol 43(6):1045–1049
61. Audemard-Verger A, Silva NM, Verstuyft C
et al (2016) Glutathione S transferases
Part V
Engineering and Surgical Developments in Precision

Medicine
Chapter 18
Precision Medicine in Tissue Engineering on Bone

Bingkun Zhao, Qian Peng, Rong Zhou, Haixia Liu, Shengcai Qi,
and Raorao Wang
Abstract
With the rapidly development of clinical treatments, precision medicine has come to people eyes with the
requirement according to different people and different disease situation. So precision medicine is called
personalized medicine which is a new frontier of healthcare. Bone tissue engineering developed from
traditional bone graft to precise medicine era. So scientists seek approaches to harness stem cells, scaffolds,
growth factors, and extracellular matrix to promise enhanced and more reliable bone formation. This review
provides an overview of novel developments on precision medicine in tissue engineering of bone hoping it
can open new perspectives of strategies on bone treatment.
Key words Precision medicine, Tissue engineering, Bone
1 Introduction
Tissue engineering precision medicine needs to combine a variety

of fields. This area aims to treat patients on an individual basis
according to their specific characters and disease state. Scientists
defined precise therapies as those in which the therapeutic strategy
are individualized to patient’s needs [1]. Since biomaterial have
been developed in various applications about scaffolds, drug deliv-
ery, and immunomodulation. However, what makes these material
limited on theory level lies in their ignorance in realistic disease
mode making it lack specific target. Different people always at
different age, different basic disease, and tissue-specific microenvir-
onments which raise the requirement of tissue engineering devel-
oping into more precisely times.
At present, more than half a million patients in the USA have
received functional bone implants, with a total cost higher than
$2.5 billion in 2012, which is expected to double by 2020
Bingkun Zhao and Qian Peng contributed equally to this work.
207
208 Bingkun Zhao et al.
[2]. Although bone tissue engineering has developed for a long

period, there remain many challenges on it and lack a perfect way to
treat bone injury. So different kinds of strategies spring out and was
invented by scientists aimed to improve bone healing. In this study,
we cited some novel research and classified them in order to present
the precision medicine views in bone tissue engineering.
2 Traditional Consideration on Bone Precise Medicine
2.1 Traditional Human bone has their nature physical property and different part
Physical Property has their own mechanical characters. Long bone (tensile testing:
ultimate strength ¼ 133 MPa, Young’s moduli ¼ 17GPa, Com-
pressive testing: ultimate strength ¼ 193 MPa, Young’s mod-
uli ¼ 18.2GPa) need to bear the gravity of body, and they
perform better in tensile and compressive tests compared to cranial
bone (tensile testing: ultimate strength ¼ 43.4 MPa, Young’s
moduli ¼ 5.4GPa, Compressive testing: ultimate
strength ¼ 96.5 MPa, Young’s moduli ¼ 5.6GPa) [3]. But
mouse bone are much weaker than human because of the body
weight and activity behavior (Cranial bone Young’s modulus
1.26 0.29 GPa), which also need to be taken into consideration
in designing the model [4]. So in future, we need to take this into
bone material design thinking the different situation between ani-
mal model and human. The work should satisfy this basic mechani-
cal property.
2.2 Stimulus There are some stimuli factors that scientists used to trigger struc-
Sensitivity tural or chemical proper changes in designed biomaterials to meet
different task. So proper stimulus is indispensable factor in design.
There are some stimuli that are often used in design bone tissue
biomaterial like: pH [5], temperature [6], magnetic fields [7], light
[8], and even self-remodeling material [9]. Among them FDA has
approved the use of a thermos-sensitive liposome with
hyperthermia-triggered release of chemotherapeutics for prostate
cancer treatment [10].
2.3 Biocompatibility However, some research point out that material properties cannot
accurately correspond to the steps of the material–host response
due to its complex biological nature [11]. In bone healing, it
consists of hematoma formation stage, inflammation stage, callus
formation stage, and bone remodeling stage [12]. These steps
conduct orderly and many scientists begin focus their material
targeting one or two steps in contribution to the bone healing.
Precision Medicine in Tissue Engineering on Bone 209
3 Bone Tissue Engineering Target Different Healing Stage
3.1 Strategies on Acute inflammatory stage peaks at 24 h [13]. This stage is a key step
Inflammation Stage to create a proper microenvironment to initiate the tissue repair.
But exacerbated and/or chronic inflammation will do negative
function on bone healing [14]. So how to moderate inflammation
stage according to different people in different disease situation
attract scientist’s attention. Hassan Rammal’s work built bioactive
and osteoinductive calcium phosphate/chitosan/hyaluronic acid
substrate (CaP-CHI-HA) system to regulate inflammation factor
(TNF-α, MCP-1, IL-6, IL-8, and IL-10) and growth factor (VEGF
and TGF-β) manipulate macrophage between pro-inflammatory
(M1) phenotype and anti-inflammatory (M2) phenotype, which
give out a good result in bone healing [15]. In order to control
the inflammation harmony and prevent overload of inflammation
factors, Basu S established a shear-thinning DNA two-dimensional
silicate nanodisks (nSi) hydrogel aiming modulate the release of a
model osteogenic drug dexamethasone (Dex) and can own the
properties of rapid self-healing [16]. This method can control the
drug delivery and moderate inflammation response more precisely.
Won JE engineered hierarchically structured microchanneled scaf-
folds [17]. This scaffold was made of biocompatible polymer poly-
caprolactone and designed on a 3D printable platform. Besides
regulating immune/inflammatory responses, the highlight of this
scaffold lies in the ability in contribution angiogenesis, and stem
cell recruitment. Old people always suffer from osteoporosis which
delayed bone formation because of chronic low levels of inflamma-
tion [18]. Therefore, precision medicine of tissue engineering also
puts much efforts on the older group. Alhamdi JR created a biomi-
metic calcium phosphate coating to serve as a highly localized
delivery system to guide macrophage phenotype transitions which
may get a good results in helping bone formation in the elder
[19]. More remarkable, inflammation response, which plays a sig-
nificant role in the process of bone injury healing, needs to be
thoroughly investigated in the future in order for future clinical
application and theory based on inflammation of bone healing still
remain the major limitation currently.
3.2 Strategies Callus formation and bone remodeling stage happen simulta-
on Callus Formation neously, and these steps need a balance of operation of two stage
Stage and Bone causing the formation and reconstruction of bone. Wang T made a
Remodeling Stage novel layer-by-layer engineering platform (PCL/collagen/HAp
composite nanofibrous mesh) for construction of a versatile biomi-
metic periosteum, enabling further assembly of a multicomponent
and multifunctional periosteum replacement for bone defect repair
and reconstruction [20]. They point out that this multiple layer
structure can combine with various natural and synthetic structural
bone graft materials to repair defects of any size and shape and
further permits controlled insertion of multiple functional compo-
nents. Jiang N created a Ti-inplants (Micro/Nanoscaled Hierarchical
Ti Phosphate/Ti Oxide Hybrid Coating) which exhibit osseointe-
gration performance and have a strong influence on the cell beha-
viors, such as proliferation, adhesion, and differentiation [21].
Growth factors act throughout the two stages [22]. So scien-
tists always focus on it. Fujioka-Kobayashi M delivered bone heal-
ing growth factor BMP2 and FGF18 by using cholesteryl group-
and acryloyl group-bearing pullulan (CHPOA) nanogels and this
hydrogel strongly enhanced and stabilized the BMP2-dependent
bone repair, inducing osteoprogenitor cell infiltration reach an
efficient bone tissue engineering [23]. Seeherman HJ improved
retention of BV256, a BMP-2/BMP-6/activin A chimera, by A
composite matrix (CM) which contains calcium-deficient hydroxy-
apatite granules suspended in a macroporous, fenestrated, polymer
mesh-reinforced recombinant human type I collagen matrix
[24]. This research paved a better way to apply BMP-2 in clinical
trials which is far more efficient than single BMP-2 application.
4 Tissue Location-Specific Consideration
Injury always occurs at the interface between bone and other tissues
like bone–cartilage structure. Mature articular cartilage is an avas-
cular tissue which is different from bone tissue. So in osteochondral
system, this special structure always add much difficulties in healing
and always accompanied by the poor prognosis due to the limited
ability to heal itself [25]. So more and more scientists focus their
attention on these challenges and make bone tissue developed into
precision medicine times. Based on this special structure, researcher
provide multiple different layers to deal with this structure. Jiang j
created two layers with different material to facilitate both bone
and cartilage region regeneration [26]. After that professor
Nukavarapu pointed out this structure’s short comings that the
interface of two different materials may cause the delayed healing
in the future because the barrier will impede the interactions of
cells in two scaffolds and further lead to the poor interfacial inte-
gration between the new formed cartilage and bone [27]. In order
to overcome this disadvantages, Deng T used a novel three-
dimensional (3D) heterogeneous/bilayered scaffold [28]. This
scaffold has to distinct materials, but they are integrated which
corresponds to the interface of cartilage and bone interface. A silk
fibroin (SF) layer and a silk-nano calcium phosphate (silk-nanoCaP)
layer was created by Le-Ping Yan [29]. Differ from previous study,
this scaffolds realized fully integrated and homogeneous porosity
distribution. Based on widely used growth factor in tissue engineer-
ing, BMP-2 was also applied in osteochondral for more precise
tissue engineering [30]. Recently, Yanlun Zhu created an injectable

continuous structurally and functionally biomimetic sodium alginate
(SA)/bioglass (BG) composite hydrogel [31], which has proper
swelling properties and not only stimulate healing of an entire
osteochondral unit but also promote the integration between the
newly formed tissues and the host tissue.
Another important interface of bone structure is bone–tendon
structure which differ from bone–cartilage structure. This structure
consists compositionally and mechanically graded structure with
bone and tendon properties. In another words, bone–tendon inter-
face is also functionally and compositionally graded structure
[32]. In mechanical level, it needs to be considered more about
stability in cycles of physiological tensile loading with the repair site
over the long term. Chen CH tried to use periosteum to envelop
the injured tendon to improve tendon–bone healing [33]. This
research outlines the tendon–bone incorporation repair. Some
scientists also apply growth factor to increase the self-healing of
bone–tendon interface [34, 35]. To increase the firm tendon–bone
anchoring and with more stability in knee, Mutsuzaki H use cal-
cium phosphate (CaP)-hybridized attach to the tendon graft which
successfully simulated the nature state of tendon-bone connec-
tion [36]. Since Li D pointed out that nanofibers can be easily
assembled into a range of arrays or hierarchically structured films
by manipulating their alignment, stacking, and/or folding
[37]. This special property was regarded as an ideal material to
contribute to tendon–bone interface repair. Liu W combined poly
(lactic acid) (PLA) (a material that has been approved by the FDA
for tissue engineering applications) with nanofibers, which exhibit
good biocompatibility and inherent biodegradability [38]. Ker D
created functionally graded material called QHM polymer (quadrol
(Q), hexamethylene diisocyanate (H), and methacrylic anhydride
(M)) which is crosslinked by UV. This study finds a new way to
mimic tendon–bone interface structure and improved a lot in
mechanical property compared to prior nanofiber material
[39]. From the above research, they performed some strategies
contributing to bone–tendon conjunction repair. But tendon is a
unique tissue that not only need to bear the gravity but also need to
bear tensile-loose cycle. This require the material to have high
standards about mechanical property which is still the main prob-
lem and challenge of tissue engineering on bone–tendon interface
repair.
5 3D Printing-Based Strategy
3D printing recently is an eye-catching technology to mimic the

human tissue and recently scientists have successfully print the heart
and pulmonary which has shocked the world making the 3D
printing an ideal option on tissue engineering [40]. Therefore,

scientists also try different methods on bone tissue engineering
research, especially making the material more precise for different
situation. So many scientists put much efforts on improving and
polishing the 3D printing technology making it more suitable for
bone tissue engineering. The Silvan Gantenbein and their team
established a three-dimensional (3D) printing approach to generate
recyclable lightweight structures with hierarchical architectures,
complex geometries, and unprecedented stiffness and toughness
which can be applied as lightweight biological materials such as
bone [41]. Notably, as microfluidics tissue engineering can control
the chemical environment on a micrometer scale within a macro-
scopic scaffold could aid in engineering complex tissues
[42]. Marco Costantini combined the 3D printing and microflui-
dics by using valve-based flow-focusing chip to establish a prede-
fined 3D geometry and a controlled, spatially varying internal
porous architecture, such as a model of a bone [43].
Besides the breakthrough of the structure simulation of bone,
the microenvironments of ECM (extracellular matrix) of bone
tissues simulation also regarded as an important step in more
precise tissue engineering. A leading strategy in tissue engineering
is the design of biomimetic scaffolds that stimulate the body’s
repair mechanisms through the recruitment of endogenous stem
cells to sites of injury [44]. The extracellular matrix (ECM) plays a
crucial role in governing organ branching during tissue regenera
tion. It not only provides structural integrity for tissue elasticity
[45], but also many ECM components are locally synthesized and
continuously reorganized to modulate diverse cellular processes
[46]. So some scientists try to pave the way on the combination
of 3D printing and ECM. AlaaMansour coated synthetic dicalcium
phosphate (DCP) bioceramics with the bone ECM and proved that
to enhance the integration of synthetic biomaterials and improve
their ability to regenerate bone probably by modulating the host
immune reaction [47]. Linxiao Wu’s work used 3D printing to
remodel elastomer nanohybrid to mimic ECM and regulate chon-
drogenesis and osteogenesis [46]. Chen et al. pointed out that
autogenous ECM is biocompatible, bioactive, and bio-safe, making
it an optimal choice for future clinical applications of 3D printing
and they combined ECM and 3D printing together successfully to
repair skull defect [48]. Shi W used silk fibroin with gelatin in
combination with BMSC-specific affinity peptide using 3D printing
(3DP) technology to establish a suitable 3D microenvironment for
BMSC proliferation, differentiation, and extracellular matrix pro-
duction which is a promising biomaterial for knee cartilage repair
[49]. Pati et al. enhanced the osteogenic potential of 3D-printed
PCL/PLGA/β-TCP scaffolds by using human nasal inferior turbi-
nate tissue-derived mesenchymal stromal cells to deposit bone-like
ECM [50]. This is the application of autologous cell
transplantation to form the osteogenetic environments, but the

method has still limited that the patients’ situation is proper.
Demirci U teamwork established a 3D with neural crest cells in
3-D microenvironments [51]. In their study, Col II expression
(chondrogenesis marker) did not increase in osteogenic conditions
and aggrecan expression did not alter significantly. These findings
indicate that there is no chondrogenic tissue formation and calcium
deposition was all from the intramembranous. This method enables
precise control over the microenvironment which is more suitable
for primary embryonic craniofacial bone disorders. From the above
research, we can see that there is a trend that takes both advantages
of ECM and 3D printing. It is obvious that 3D printing has its own
special advantages on simulation of the bone structure. But it is
important to make it specific and proper to different situation
which needs other complements.
6 Summary
From developmental view, precise tissue engineering strategy has its

indispensable advantages that it can treat the diseases according to
different functional requirement, different place, and different per-
sonal situation. Bone tissue engineering moving from bulk material
to more precise and personalized material is still in urgent need.
Acknowledgements
Bingkun Zhao and Qian Peng contributed equally to this work.
References
1. Oliva N, Unterman S, Zhang Y, Conde J, Song 5. Barros J, Melo L, Silva RA, Ferraz MP,
HS, Artzi N (2015) Personalizing biomaterials Azeredo J, Pinheiro V, Colaco B, Fernandes
for precision nanomedicine considering the MH, Gomes PS, Monteiro FJ (2019)
local tissue microenvironment. Adv Healthc Encapsulated bacteriophages in alginate-
Mater 4:1584 nanohydroxyapatite hydrogel as a novel deliv-
2. Bose S, Sarkar N (2019) Natural medicinal ery system to prevent orthopedic implant-
compounds in bone tissue engineering. Trends associated infections. Nanomedicine-Uk
Biotechnol 24:102145
3. Zhao J, Shen G, Liu C, Wang S, Zhang W, 6. Wasupalli GK, Verma D (2020) Injectable and
Zhang X, Zhang X, Ye D, Wei J, Zhang Z, thermosensitive nanofibrous hydrogel for bone
Jiang X (2012) Enhanced healing of rat calvar- tissue engineering. Mater Sci Eng C Mater Biol
ial defects with sulfated chitosan-coated cal- Appl 107:110343
cium-deficient hydroxyapatite/bone 7. Bianchi M, Boi M, Sartori M, Giavaresi G,
morphogenetic protein 2 scaffolds. Tissue Lopomo N, Fini M, Dediu A, Tampieri A,
Eng Part A 18:185 Marcacci M, Russo A (2015) Nanomechanical
4. Schenk MPAP (2014) Determining mechani- mapping of bone tissue regenerated by mag-
cal properties of adult murine calvaria. 60th netic scaffolds. J Mater Sci Mater Med 26:5363
Annual Meeting of the Orthopaedic Research 8. Celikkin N, Mastrogiacomo S, Jaroszewicz J,
Society 2014 Walboomers XF, Swieszkowski W (2018)
Gelatin methacrylate scaffold for bone tissue for bone repair and reconstruction. Biomater-
engineering: the influence of polymer concen- ials 182:279
tration. J Biomed Mater Res A 106:201 21. Jiang N, Guo Z, Sun D, Li Y, Yang Y, Chen C,
9. Meng L, Shao C, Cui C, Xu F, Lei J, Yang J Zhang L, Zhu S (2018) Promoting osseointe-
(2020) Autonomous self-healing silk fibroin gration of Ti implants through micro/nanos-
injectable hydrogels formed via surfactant-free caled hierarchical Ti phosphate/Ti oxide
hydrophobic association. ACS Appl Mater hybrid coating. ACS Nano 12:7883
Interfaces 12:1628 22. Siddiqui JA, Partridge NC (2016) Physiologi-
10. Arias JL, Clares B, Morales ME, Gallardo V, cal bone remodeling: systemic regulation and
Ruiz MA (2011) Lipid-based drug delivery growth factor involvement. Physiology
systems for cancer treatment. Curr Drug Tar- (Bethesda) 31:233
gets 12:1151 23. Fujioka-Kobayashi M, Ota MS, Shimoda A,
11. Cooper DK, Novitzky D, Wicomb WN (1988) Nakahama K, Akiyoshi K, Miyamoto Y, Iseki
Hormonal therapy in the brain-dead experi- S (2012) Cholesteryl group- and acryloyl
mental animal. Transplant Proc 20:51 group-bearing pullulan nanogel to deliver
12. Ono T, Takayanagi H (2017) Osteoimmunol- BMP2 and FGF18 for bone tissue engineering.
ogy in bone fracture healing. Curr Osteoporos Biomaterials 33:7613
Rep 15:367 24. Seeherman HJ, Berasi SP, Brown CT, Martinez
13. Einhorn TA, Gerstenfeld LC (2015) Fracture RX, Juo ZS, Jelinsky S, Cain MJ, Grode J,
healing: mechanisms and interventions. Nat Tumelty KE, Bohner M, Grinberg O, Orr N,
Rev Rheumatol 11:45 Shoseyov O, Eyckmans J, Chen C, Morales PR,
14. Mountziaris PM, Mikos AG (2008) Modula- Wilson CG, Vanderploeg EJ, Wozney JM
tion of the inflammatory response for enhanced (2019, 11) A BMP/activin A chimera is supe-
bone tissue regeneration. Tissue Eng Part B rior to native BMPs and induces bone repair in
Rev 14:179 nonhuman primates when delivered in a com-
posite matrix. Sci Transl Med:eaar4953
15. Rammal H, Bour C, Dubus M, Entz L,
Aubert L, Gangloff SC, Audonnet S, Bercu 25. Huang BJ, Hu JC, Athanasiou KA (2016) Cell-
NB, Boulmedais F, Mauprivez C, Kerdjoudj based tissue engineering strategies used in the
H (2018) Combining calcium phosphates clinical repair of articular cartilage. Biomaterials
with polysaccharides: a bone-inspired material 98:1
modulating monocyte/macrophage early 26. Jiang J, Tang A, Ateshian GA, Guo XE, Hung
inflammatory response. Int J Mol Sci 19:3458 CT, Lu HH (2010) Bioactive stratified poly-
16. Basu S, Pacelli S, Feng Y, Lu Q, Wang J, Paul A mer ceramic-hydrogel scaffold for integrative
(2018) Harnessing the noncovalent interac- osteochondral repair. Ann Biomed Eng
tions of DNA backbone with 2D silicate nano- 38:2183
disks to fabricate injectable therapeutic 27. Nukavarapu SP, Dorcemus DL (2013) Osteo-
hydrogels. ACS Nano 12:9866 chondral tissue engineering: current strategies
17. Won JE, Lee YS, Park JH, Lee JH, Shin YS, and challenges. Biotechnol Adv 31:706
Kim CH, Knowles JC, Kim HW (2020) Hier- 28. Deng T, Lv J, Pang J, Liu B, Ke J (2014)
archical microchanneled scaffolds modulate Construction of tissue-engineered osteochon-
multiple tissue-regenerative processes of dral composites and repair of large joint defects
immune-responses, angiogenesis, and stem in rabbit. J Tissue Eng Regen Med 8:546
cell homing. Biomaterials 227:119548 29. Yan LP, Silva-Correia J, Oliveira MB, Vilela C,
18. Franceschi C, Bonafe M, Valensin S, Olivieri F, Pereira H, Sousa RA, Mano JF, Oliveira AL,
De Luca M, Ottaviani E, De Benedictis G Oliveira JM, Reis RL (2015) Bilayered silk/
(2000) Inflamm-aging. An evolutionary per- silk-nanoCaP scaffolds for osteochondral tissue
spective on immunosenescence. Ann N Y engineering: in vitro and in vivo assessment of
Acad Sci 908:244 biological performance. Acta Biomater 12:227
19. Alhamdi JR, Peng T, Al-Naggar IM, Hawley 30. Vayas R, Reyes R, Rodriguez-Evora M, Del
KL, Spiller KL, Kuhn LT (2019) Controlled RC, Delgado A, Evora C (2017) Evaluation
M1-to-M2 transition of aged macrophages by of the effectiveness of a bMSC and BMP-2
calcium phosphate coatings. Biomaterials polymeric trilayer system in cartilage repair.
196:90 Biomed Mater 12:45001
20. Wang T, Zhai Y, Nuzzo M, Yang X, Yang Y, 31. Zhu Y, Kong L, Farhadi F, Xia W, Chang J,
Zhang X (2018) Layer-by-layer nanofiber- He Y, Li H (2019) An injectable continuous
enabled engineering of biomimetic periosteum stratified structurally and functionally
biomimetic construct for enhancing osteo- 42. Choi NW, Cabodi M, Held B, Gleghorn JP,
chondral regeneration. Biomaterials 192:149 Bonassar LJ, Stroock AD (2007) Microfluidic
32. Lu HH, Thomopoulos S (2013) Functional scaffolds for tissue engineering. Nat Mater
attachment of soft tissues to bone: develop- 6:908
ment, healing, and tissue engineering. Annu 43. Costantini M, Jaroszewicz J, Kozon L,
Rev Biomed Eng 15:201 Szlazak K, Swieszkowski W, Garstecki P,
33. Chen CH, Chen WJ, Shih CH, Chou SW Stubenrauch C, Barbetta A, Guzowski J
(2004) Arthroscopic anterior cruciate ligament (2019) 3D-printing of functionally graded
reconstruction with periosteum-enveloping porous materials using on-demand reconfigur-
hamstring tendon graft. Knee Surg Sports able microfluidics. Angew Chem Int Ed Engl
Traumatol Arthrosc 12:398 58:7620
34. Yamazaki S, Yasuda K, Tomita F, Tohyama H, 44. Pacelli S, Basu S, Whitlow J, Chakravarti A,
Minami A (2005) The effect of transforming Acosta F, Varshney A, Modaresi S,
growth factor-beta1 on intraosseous healing of Berkland C, Paul A (2017) Strategies to
flexor tendon autograft replacement of anterior develop endogenous stem cell-recruiting bio-
cruciate ligament in dogs. Arthroscopy active materials for tissue repair and regenera-
21:1034 tion. Adv Drug Deliv Rev 120:50
35. Chen CH, Liu HW, Tsai CL, Yu CM, Lin IH, 45. Cox TR, Erler JT (2011) Remodeling and
Hsiue GH (2008) Photoencapsulation of bone homeostasis of the extracellular matrix: impli-
morphogenetic protein-2 and periosteal pro- cations for fibrotic diseases and cancer. Dis
genitor cells improve tendon graft healing in a Model Mech 4:165
bone tunnel. Am J Sports Med 36:461 46. Wu L, Magaz A, Wang T, Liu C, Darbyshire A,
36. Mutsuzaki H, Sakane M, Hattori S, Loizidou M, Emberton M, Birchall M, Song W
Kobayashi H, Ochiai N (2009) Firm anchoring (2018) Stiffness memory of indirectly
between a calcium phosphate-hybridized ten- 3D-printed elastomer nanohybrid regulates
don and bone for anterior cruciate ligament chondrogenesis and osteogenesis of human
reconstruction in a goat model. Biomed mesenchymal stem cells. Biomaterials 186:64
Mater 4:45013 47. Mansour A, Abu-Nada L, Al-Waeli H, Mezour
37. Li D, Ouyang G, McCann JT, Xia Y (2005) MA, Abdallah MN, Kinsella JM, Kort-
Collecting electrospun nanofibers with pat- Mascort J, Henderson JE, Ramirez-Garcialuna
terned electrodes. Nano Lett 5:913 JL, Tran SD, Elkashty OA, Mousa A, El-Hadad
38. Liu W, Lipner J, Moran CH, Feng L, Li X, AA, Taqi D, Al-Hamad F, Alageel O, Kaartinen
Thomopoulos S, Xia Y (2015) Generation of MT, Tamimi F (2019) Bone extracts immuno-
electrospun nanofibers with controllable modulate and enhance the regenerative perfor-
degrees of crimping through a simple, mance of dicalcium phosphates bioceramics.
plasticizer-based treatment. Adv Mater Acta Biomater 89:343
27:2583 48. Chen H, Zhang J, Li X, Liu L, Zhang X,
39. Ker D, Wang D, Behn AW, Wang E, Zhang X, Ren D, Ma C, Zhang L, Fei Z, Xu T (2019)
Zhou BY, Mercado-Pagan AE, Kim S, Multi-level customized 3D printing for autog-
Kleimeyer J, Gharaibeh B, Shanjani Y, enous implants in skull tissue engineering. Bio-
Nelson D, Safran M, Cheung E, Campbell P, fabrication 11:45007
Yang YP (2018) Functionally graded, bone- 49. Shi W, Sun M, Hu X, Ren B, Cheng J, Li C,
and tendon-like polyurethane for rotator cuff Duan X, Fu X, Zhang J, Chen H, Ao Y (2017)
repair. Adv Funct Mater 28:1707107 Structurally and functionally optimized silk-
40. Grigoryan B, Paulsen SJ, Corbett DC, Sazer fibroin-gelatin scaffold using 3d printing to
DW, Fortin CL, Zaita AJ, Greenfield PT, Cala- repair cartilage injury in vitro and in vivo. Adv
fat NJ, Gounley JP, Ta AH, Johansson F, Mater 29:1701089
Randles A, Rosenkrantz JE, Louis-Rosenberg 50. Pati F, Song TH, Rijal G, Jang J, Kim SW, Cho
JD, Galie PA, Stevens KR, Miller JS (2019) DW (2015) Ornamenting 3D printed scaffolds
Multivascular networks and functional intra- with cell-laid extracellular matrix for bone tis-
vascular topologies within biocompatible sue regeneration. Biomaterials 37:230
hydrogels. Science 364:458 51. Namkoong B, Guven S, Ramesan S,
41. Gantenbein S, Masania K, Woigk W, Sesseg J, Liaudanskaya V, Abzhanov A, Demirci U
Tervoort TA, Studart AR (2018) Three- (2016) Recapitulating cranial osteogenesis
dimensional printing of hierarchical liquid- with neural crest cells in 3-D microenviron-
crystal-polymer structures. Nature 561:226 ments. Acta Biomater 31:301
Chapter 19
Progress of Clinical Application for Ex Vivo Lung Perfusion

(EVLP) in Lung Transplantation
Chang Gu, Xufeng Pan, and Jianxin Shi
Abstract
In recent years, medical advances make lung transplantation become a standard treatment for terminal lung
diseases (such as emphysema, pulmonary fibrosis, pulmonary cystic fibrosis, and pulmonary arterial hyper-
tension) that cannot be cured by drugs or surgery (Lund et al., J Heart Lung Transplant 34:1244, 2015).
However, the current number of donor lungs that meet the transplant criteria is no longer sufficient for
transplanting, causing some patients to die while waiting for a suitable lung. Current methods for
improving the situation of shortage of lung transplant donors include the use of donation after cardiac
death (DCD) donors, smoker donors, and Ex Vivo Lung Perfusion (EVLP). Among them, EVLP is a
technique for extending lung preservation time and repairing lung injury in the field of lung transplanta-
tion. By continuously assessing and improving the function of marginal donor lungs, EVLP increases the
number of lungs that meet the transplant criteria and, to some extent, alleviates the current situation of
shortage of donor lungs. This chapter reviews the clinical application and research progress of EVLP in the
field of lung transplantation.
Key words Ex vivo lung perfusion, Lung transplantation, Emphysema, Pulmonary fibrosis, Pulmo-
nary cystic fibrosis, Pulmonary arterial hypertension
1 The Current Situation of Lung Transplant Donor Shortage
Since Hardy successfully implemented the first human lung trans-

plant in 1963, the medical level has been increasing and lung
transplantation has become the standard treatment for terminal-
stage lung disease that cannot be treated by drugs or surgery. The
current ideal donor lung criteria include: (1) age < 55 years of age;
(2) smoking index <20; (3) no chest trauma; (4) clear chest image;
(5) oxygenation index 300 mmHg; and (6) there is no purulent
secretion in the respiratory tract and the secretion is negative for
Gram stain [1]. However, donor lungs often fail to meet transplant
criteria due to factors such as neurogenic pulmonary edema, severe
pneumonia, and acute respiratory distress syndrome (ARDS).
Although in recent years the management and preservation
217
218 Chang Gu et al.
techniques for donated organs have improved significantly, only

15–20% of the donor lungs meet the transplant criteria [2, 3]. More-
over, in order to avoid the primary graft dysfunction (PGD) caused
by the ischemia reperfusion injury (IRI) during lung transplanta-
tion, the transplant team is more selective on the choice of lungs
[4]. For the above reasons, there is a shortage of donor organs that
meet the transplantation criteria, lung transplantation could not be
widely carried out and a considerable number of patients dying
because they cannot get the donor lungs in time [5]. Since January
1, 2015, China has completely stopped using the organ donation
from prisoners sentenced to death. Voluntary donations from citi-
zens have become the sole source of organ transplant donors,
which has led to a decline in the number and quality of legally
donor lungs in China.
2 Measures to Alleviate the Shortage of Lung Transplant Donors
At present, donors using donation after cardiac death (DCD) and

EVLP have been used to expand the criteria for donor lung selec-
tion, thereby alleviating the current shortage of lung transplant
donors. Schiavon et al. [6] reviewed 10 studies on “marginal
donor lungs” from 1993 to 2010 and found that although there
were no significant differences in medium- and long-term survival
rates, four of them showed “marginal donor lung” receptors expe-
rienced more early postoperative adverse events (30-day and 90-day
mortality, ICU time, hospital stay, etc.). Sommer et al. [7] found
that the transplantation of “marginal donor lungs” to low-risk
patients can achieve better clinical results, which can alleviate the
current situation of lung shortage to some extent.
The first successful lung transplant came from a DCD donor,
and many related clinical studies have emphasized the use of DCD
donor lungs had the potential to alleviate donor lung shortage
[8, 9]. However, since DCD donor lungs are susceptible to dam-
age, especially during hot and cold ischemia, and a portion of the
lungs are from uncontrolled DCD donors (uncontrolled lungs), it
is more prudent to use DCD for lung transplantation. De Antonio
et al. [10] found that 29 lung transplantation patients who received
uncontrolled lungs had good prognosis (3-month survival rate of
82%, 1-year survival rate of 69%, and 3-year survival rate of 58%).
The donor lungs from smokers has been proven to be used for
lung transplantation in recent years and is a method to expand the
source of donor lungs [11, 12]. Bonser et al. [11] studied 1295
patients (including 510 donors with a history of smoking) who had
lung transplantation, patients received donors with a history of
smoking for lung transplantation had a poor 3-year survival rate
while relatively better postoperative survival was found in terminal
interstitial lung diseases (ILD) patients.
Progress of Clinical Application for Ex Vivo Lung Perfusion (EVLP) in Lung. . . 219
In recent years, EVLP has increased the number of donor lungs

through evaluating and improving the function of “marginal donor
lungs,” and to some extent, improves the current situation of
donor shortage of lung transplantation [1].
3 Ex Vivo Lung Perfusion (EVLP)
3.1 EVLP The idea of organ perfusion was first proposed by Carrel and
Development History Lindbergh in the 1930s, and they also performed perfusion experi-
ments on organs such as the heart, kidneys, ovaries, spleen, and
adrenal glands [13]. In the 1990s, lung perfusion techniques were
used to study lung physiology. The first clinical application of
pulmonary perfusion was performed by Steen et al. [14]. In
2001, Steen et al. reported the evaluation of lung function in a
54-year-old patient with myocardial infarction by EVLP, using
Perfadex© fluid perfusion to allow local cooling of the lung before
being removed, and perfusion for 65 min by EVLP before lung
transplantation, which establishing a foundation for EVLP clinical
application. Subsequently, Steen et al. [3] applied short-term EVLP
to improve the function of “marginal donor lungs,” and 6 cases of
“marginal donor lungs” were successfully transplanted after EVLP
perfusion (61–121 min). Then, the Toronto General Hospital
applied and expanded EVLP, not only to re-evaluate the function
of “marginal donor lungs,” but also to provide a new way to repair
the lungs at room temperature [15]. The basis of the application of
the Toronto EVLP strategy is: (1) gradually rewarming to normal
temperature; (2) gradually increasing the flow of perfusion blood as
the lung rewarming process, until 40% of the donor’s estimated
cardiac output; (3) protective lung ventilation; (4) cell-free perfu-
sion with high colloid osmotic pressure.
Current EVLP perfusion indications for brain death or cardiac
death donation include: (1) optimal oxygenation index
<300 mmHg; (2) X-ray or physical examination revealed signs of
pulmonary edema; (3) low lung compliance was found in the
process of dissect lungs; (4) high-risk clinical history, such as history
of >10 units of blood transfusion or history of suspected aspiration;
(5) time span from cardiac arrest to withdrawal of life support
system >60 min in cardiac death donation [16].
The criteria for inclusion of “marginal donor lungs” after 4–6 h
of clinical EVLP were: (1) oxygenation index >400 mmHg, (2) sta-
ble or improved pulmonary artery pressure, (3) stable or improved
airway pressure, (4) stable or improved lung compliance; exclusion
criteria were: (1) oxygenation index <400 mmHg, (2) pulmonary
artery pressure worsened by more than 15%, (3) airway pressure/
pulmonary compliance worsened by more than 15% [16].
There are many platform systems that support EVLP, such as
the XPS (XVIVO Perfusion) system that complies with the Toronto
220 Chang Gu et al.
strategy, the LS1 (Vivoline Medical) system that complies with the
Steen strategy, and the portable OCS Lung (Transmedics) system.
3.2 Evaluation The current method of graft preservation used in most transplant
of the Effects of EVLP centers is the “Cold Static Preservation” (CSP), which is perfused
with cold transplanted perfusate (Perfadex©, etc.) and meanwhile
administers pulmonary ventilation, maintaining lung a static expan-
sion at 4 C [17]. Low temperature can maintain the lung function
in the ischemic state for a period of time by reducing the metabolic
rate. Currently, CSP is still the main approach of preservation for
the lung [18]. In 2003, Steen et al. [19] applied the normal
temperature perfusion technique to the field of lung transplanta-
tion. A major advantage of lung perfusion at room temperature is
to maintain the active metabolic function of the cells, thereby
facilitating the continuous assessment of lung function during
in vitro preservation [17].
The currently used EVLP perfusion device simulates the perfu-
sion mode under physiological conditions and provides support
means such as tracheal intubation and mechanical ventilation. Vari-
ous sensors are used to measure various parameters in the EVLP
system, including pulmonary arteriovenous partial pressure, pH,
temperature, flow, perfusion pressure, ventilation parameters, etc.
In addition, techniques such as histology and imaging can also be
applied to the EVLP process to assess lung function. These mea-
surements can be used to obtain real-time data at any time during
the EVLP process to quantify lung function indicators and contrib-
ute to the assessment of lung function [20]. This continuous,
multi-parameter approach allows the transplant team to be more
reliable in determining whether the “marginal donor lungs” meet
the transplant criteria and implement the transplant decision. The
observed indicators in EVLP are as follows:
1. Perfusate gas and hemodynamics.
Partial oxygen pressure (PaO2) and partial pressure of
carbon dioxide (PaCO2), oxygenation index (PaO2/FiO2),
pulmonary vascular resistance (PVR) [PVR ¼ PAP/pulmonary
artery flow 80 (dynes/s/cm5)], and lung compliance [tidal
volume/airway platform pressure PEEP (mL/cmH2O)] is a
direct indicator of lung function during EVLP.
2. Pulmonary dry weight ratio (W/D lung weight ratio).
The pulmonary dry and wet ratio has become an indirect
but reliable indicator of evaluation. By comparing the pulmo-
nary dry and wet ratio of the specimens taken at different time
periods, the lung preservation quality can be judged whether
the lungs have pulmonary edema, reperfusion injury, alveolar
damage, etc. [21–23].
3. Glucose consumption.
By monitoring glucose consumption during EVLP, lung
metabolism can be estimated, and lung function can be
inferred. Valenza et al. [24] studied the EVLP pig model and
found that the worse the lung function, the more the lung
glucose consumption during EVLP, and the glucose consump-
tion was related to pulmonary edema.
4. Histological evaluation.
Observing the pathological sections of the donor lung at
different stages of EVLP can directly detect signs of lung injury,
such as edema and hemorrhage [25]. Medeiros et al. [26]
found that the pulmonary edema signs were reduced and the
average number of apoptotic cells decreased after EVLP
(16/mm2 vs. EVLP) after 16 cases of lung-free lung transplan-
tation. 20/mm2, p ¼ 0.063).
5. Imaging evaluation.
The application of imaging techniques such as conven-
tional X-ray and spiral CT can be used to confirm clinical
diagnosis [2]. Such as changes in signs of pulmonary edema
before and after EVLP.
6. Expression of inflammatory factors.
During ischemia-reperfusion injury, cytokines play an
important role in mediating inflammatory damage leading to
permanent lung dysfunction [27]. Sadaria et al. [28] found that
the expression of proinflammatory cytokines IL-8, IL-6,
MCP-1, and G-CSF increased after 7 h of EVLP in 7 lung
donors that did not meet the transplant criteria. CSF expres-
sion was reduced while the anti-inflammatory cytokine IL-10
was not detected during EVLP.
3.3 Repair In recent years, the role of EVLP in improving and repairing lung
and treatment Effects function has got further attention and recognition.
of EVLP
3.3.1 Pulmonary Edema The Steen solution used for perfusion in EVLP has physiological
osmotic pressure and high colloid osmotic pressure, which can
remove excess water from pulmonary interstitial and alveolar dur-
ing perfusion, stabilize cell connection, and ensure the integrity of
alveolar capillary membrane [14, 19, 29].
3.3.2 Aspiration In the lungs that did not meet the transplant conditions, 14% were
of Gastric Contents caused by inhalation of gastric contents [25]. On the one hand,
gastric acid and pepsin in the stomach contents cause chemical
damage to the lung tissue, and on the other hand, food particles
in the stomach contents cause airway obstruction. Meers et al. [25]
used pig lung model and found that lungs that have been damaged
by gastric contents no longer deteriorate lung function during
222 Chang Gu et al.
EVLP, but EVLP alone cannot repair lung tissue that has been
damaged by gastric contents. Inci et al. [22] added airway surfac-
tant to EVLP perfusate to repair lung tissue damaged by gastric
contents, thereby improving lung function.
3.3.3 Pulmonary Part of the reason for donor lung with lower pulmonary function
Embolism fail to meet the transplant conditions is due to undetected potential
pulmonary embolism [30]. Inci et al. [31] found that the addition
of urokinase to EVLP perfusate can improve lung function by
reducing pulmonary vascular resistance and improving pulmonary
oxygenation. In recent years, Inci et al. [32] applied this study to
the clinic, and used EVLP as a thrombolytic drug carrier to suc-
cessfully perform double lung transplantation after repairing a mar-
ginal donor lung.
3.3.4 Pulmonary A significant number of donor lungs with lower pulmonary func-
Infection tion are caused by infections that occur during the process of brain
death or in the hospital. Andreasson et al. [33] added high-dose,
broad-spectrum antibiotics to EVLP perfusate and studied
18 donor lungs that did not meet the transplant conditions. The
results confirmed the bacterial load in the bacteriological tissue of
the lung was significantly reduced.
3.3.5 Other Interventions Certain specific drugs are added to the EVLP perfusate to improve
lung function, thereby achieving the purpose of affecting the prog-
nosis of the transplant. Kondo et al. [34] found that the addition of
β2 adrenergic receptor agonists to perfusate can reduce lung dam-
age. Noda et al. [35] confirmed that inhalation of H2 during EVLP
plays an important role in improving lung function by reducing
inflammation and promoting mitochondrial production. Harada
et al. [36] studied a pig lung model with warm ischemia for 2 h
and found that granulocyte elastase inhibitors used in EVLP can
reduce granulocyte elastase ( p < 0.01), pulmonary dry and wet
ratio ( p < 0.05), pulmonary vascular resistance ( p < 0.01), and
improve lung compliance ( p < 0.05) at the same time. In addition
to the purpose of improving lung function in the EVLP process,
lung function can be improved by means of gene or stem cell
therapy. Cypel et al. [37] showed that adenovirus-mediated IL-10
can be expressed in lung cells during EVLP, which plays an impor-
tant role in the repair of lung function. Lee et al. [38] confirmed
that allogeneic human mesenchymal stem cells can repair
endotoxin-mediated acute lung injury during EVLP.
3.4 Clinical Effects The prognosis of patients who had not previously met the trans-
of EVLP plantation criteria for lung transplantation after EVLP was not
different from that of standard lung transplantation patients.
Cypel et al. [16] performed a controlled study of 20 patients with
EVLP lung transplantation and 116 patients with conventional
lung transplantation and found that there was no difference

between the two groups in primary graft dysfunction within 72 h
after transplantation (EVLP group vs. routine group 15 vs. 30%,
p ¼ 0.11), and serious postoperative adverse events were not asso-
ciated with EVLP. Similarly, Valenza et al. [39] through a con-
trolled study of 7 patients with EVLP before lung transplantation
and 29 patients with standard lung transplantation and found that
there was no significant difference in primary graft dysfunction
within 72 h (EVLP vs. standard group 32 vs. 28%, p ¼ 1), 30-day
mortality (EVLP vs. standard group 0% vs. 0%, p ¼ 1), and overall
survival (EVLP vs. standard group 71 vs. 86%, p ¼ 0.27) between
the two groups.
4 Summary
In summary, by continuously assessing and improving the function

of marginal donor lungs, EVLP increases the number of lungs that
meet the transplant criteria and, to some extent, alleviates the
current situation of shortage of donor lungs. In addition, lung
injury can be repaired by adding drugs to the EVLP process to
further improve lung function. With the deepening of research, the
clinical application value of EVLP in the field of lung transplanta-
tion will get further attention and recognition.
References
1. Pierre AF, Sekine Y, Hutcheon MA et al (2002) 7. Sommer W, Kühn C, Tudorache I et al (2013)

Marginal donor lungs: a reassessment. J Thorac Extended criteria donor lungs and clinical out-
Cardiovasc Surg 123(3):421–428 come: results of an alternative allocation algo-
2. Egan TM, Haithcock JA, Nicotra WA et al rithm. J Heart Lung Transplant 32
(2006) Ex vivo evaluation of human lungs for (11):1065–1072
transplant suitability. Ann Thorac Surg 81 8. Cypel M, Sato M, Yildirim E et al (2009) Initial
(4):1205–1213 experience with lung donation after cardiocir-
3. Ingemansson R, Eyjolfsson A, Mared L et al culatory death in Canada. J Heart Lung Trans-
(2009) Clinical transplantation of initially plant 28(8):753–758
rejected donor lungs after reconditioning 9. Levvey BJ, Harkess M, Hopkins P et al (2012)
ex vivo. Ann Thorac Surg 87(1):255–260 Excellent clinical outcomes from a national
4. Aigner C, Slama A, Hötzenecker K et al (2012) donation after determination of cardiac death
Clinical ex vivo lung perfusion—pushing the lung transplant collaborative. Am J Transplant
limits. Am J Transplant 12(7):1839–1847 12(9):2406–2413
5. Yusen RD (2009) Lung transplantation out- 10. de Antonio DG, Marcos R, Laporta R et al
comes: the importance and inadequacies of (2007) Results of clinical lung transplant from
assessing survival. Am J Transplant 9 uncontrolled non-heart-beating donors. J
(7):1493–1494 Heart Lung Transplant 26(5):529–534
6. Schiavon M, Falcoz PE, Santelmo N et al 11. Bonser RS, Taylor R, Collett D et al (2012)
(2012) Does the use of extended criteria Effect of donor smoking on survival after lung
donors influence early and long-term results transplantation: a cohort study of a prospective
of lung transplantation? Interact Cardiovasc registry. Lancet 380(9843):747–755
Thorac Surg 14(2):183–187 12. Sabashnikov A, Patil NP, Mohite PN et al
(2014) Influence of donor smoking on
224 Chang Gu et al.
midterm outcomes after lung transplantation. perfusion. J Heart Lung Transplant 31

Ann Thorac Surg 97(3):1015–1021 (3):305–309
13. Carrel A, Lindbergh CA (1935) The culture of 27. De Perrot M, Sekine Y, Fischer S et al (2002)
whole organs. Science 81(2112):621–623 Interleukin-8 release during early reperfusion
14. Steen S, Sjöberg T, Pierre L et al (2001) Trans- predicts graft function in human lung trans-
plantation of lungs from a non-heart-beating plantation. Am J Respir Crit Care Med 165
donor. Lancet 357(9259):825–829 (2):211–215
15. Cypel M, Yeung JC, Hirayama S et al (2008) 28. Sadaria MR, Smith PD, Fullerton DA et al
Technique for prolonged normothermic (2011) Cytokine expression profile in human
ex vivo lung perfusion. J Heart Lung Trans- lungs undergoing normothermic ex-vivo lung
plant 27(12):1319–1325 perfusion. Ann Thorac Surg 92(2):478–484
16. Cypel M, Yeung JC, Liu M et al (2011) Nor- 29. Trebbia G, Sage E, Le Guen M et al (2019)
mothermic ex vivo lung perfusion in clinical Assessment of lung edema during ex-vivo lung
lung transplantation. N Engl J Med 364 perfusion by single transpulmonary thermodi-
(15):1431–1440 lution: a preliminary study in humans. J Heart
17. Cypel M, Keshavjee S (2011) Extracorporeal Lung Transplant 38(1):83–91
lung perfusion. Curr Opin Organ Transplant 30. Ware LB, Wang Y, Fang X et al (2002) Assess-
16(5):469–475 ment of lungs rejected for transplantation and
18. Fischer S, Matte-Martyn A, de Perrot M et al implications for donor selection. Lancet 360
(2001) Low-potassium dextran preservation (9333):619–620
solution improves lung function after human 31. Inci I, Zhai W, Arni S et al (2007) Fibrinolytic
lung transplantation. J Thorac Cardiovasc Surg treatment improves the quality of lungs
121(3):594–596 retrieved from non-heart-beating donors. J
19. Steen S, Liao Q, Wierup PN et al (2003) Trans- Heart Lung Transplant 26(10):1054–1060
plantation of lungs from non–heart-beating 32. Inci I, Yamada Y, Hillinger S et al (2014) Suc-
donors after functional assessment ex vivo. cessful lung transplantation after donor lung
Ann Thorac Surg 76(1):244–252 reconditioning with urokinase in ex vivo lung
20. Dı́ez FJM, de Ugarte AV (2010) Ex-vivo per- perfusion system. Ann Thorac Surg 98
fusion: assessment, recovery and optimisation (5):1837–1838
of human lungs for transplant. Arch Bronco- 33. Andreasson A, Karamanou DM, Perry JD et al
neumol 46(5):213–214 (2014) The effect of ex vivo lung perfusion on
21. Oka T, Puskas JD, Mayer E et al (1991) microbial load in human donor lungs. J Heart
Low-potassium UW solution for lung preser- Lung Transplant 33(9):910–916
vation: comparison with regular UW, LPD, and 34. Kondo T, Chen F, Ohsumi A et al (2015) β
euro-Collins solutions. Transplantation 52 2-adrenoreceptor agonist inhalation during
(6):984–988 ex vivo lung perfusion attenuates lung injury.
22. Inci I, Ampollini L, Arni S et al (2008) Ex vivo Ann Thorac Surg 100(2):480–486
reconditioning of marginal donor lungs injured 35. Noda K, Shigemura N, Tanaka Y et al (2014)
by acid aspiration. J Heart Lung Transplant 27 Hydrogen preconditioning during ex vivo lung
(11):1229–1236 perfusion improves the quality of lung grafts in
23. Andrade CF, Martins LK, Tonietto TA et al rats. Transplantation 98(5):499–506
(2004) Partial liquid ventilation with perfluor- 36. Harada M, Oto T, Otani S et al (2015) A
odecalin following unilateral canine lung allo- neutrophil elastase inhibitor improves lung
transplantation in non–heart-heating donors. J function during ex vivo lung perfusion. Gen
Heart Lung Transplant 23(2):242–251 Thorac Cardiovasc Surg 18(9):1–7
24. Valenza F, Rosso L, Pizzocri M et al (2011) 37. Cypel M, Liu M, Rubacha M et al (2009)
The consumption of glucose during ex vivo Functional repair of human donor lungs by
lung perfusion correlates with lung edema IL-10 gene therapy. Sci Transl Med 1(4):1–9
[C]. Transplant Proc Elsevier 43(4):993–996 38. Lee JW, Fang X, Gupta N et al (2009) Alloge-
25. Meers CM, Tsagkaropoulos S, Wauters S et al neic human mesenchymal stem cells for treat-
(2011) A model of ex vivo perfusion of porcine ment of E. coli endotoxin-induced acute lung
donor lungs injured by gastric aspiration: a step injury in the ex vivo perfused human lung. Proc
towards pretransplant reconditioning. J Surg Natl Acad Sci U S A 106(38):16357–16362
Res 170(1):159–167 39. Valenza F, Rosso L, Coppola S et al (2014) Ex
26. Medeiros IL, Pêgo-Fernandes PM, Mariani vivo lung perfusion to improve donor lung
AW et al (2012) Histologic and functional eval- function and increase the number of organs
uation of lungs reconditioned by ex vivo lung available for transplantation. Transpl Int 27
(6):553–561
INDEX
A G
Androgen deprived treatment (ADT) .....................75, 80 Gene expression profiling ..............................55, 112, 113
Angiotensin-converting enzyme inhibitor Genome variation................................................. 193–200
(ACEI) ................... 147–149, 151, 152, 156, 157 Genome-wide association (GWAS).............194–198, 200
Angiotensin II-receptor blocker (ARB)............. 147–149, Genomics .................................................. 4–6, 13, 14, 19,
151, 152, 156, 157 20, 35, 36, 42, 44, 47, 53, 56, 69, 78, 92, 124,
Atrial fibrillation (AF) ......................... 123–130, 145–158 162, 165, 167, 177
Atrial septal defect (ASD) ............................................. 175 Guilt by association............................................ 40, 43, 46
B H
Bayesian .....................................................................39, 40 Heat diffusion .................................................... 40, 43, 44
Biomedical imaging ..................................................63–70 Homologous recombination deficiency (HRD) .......... 76,
Breast cancer...........................53–59, 63–70, 77, 79, 111 78, 82
Hypermethylated promoters ........................................ 126
C
I
Cancers of unknown primary origin ............................ 109
Cardiac death (DCD) .......................................... 218, 219 Immunohistochemistry ................. 34, 77, 111, 112, 117
Cardiovascular disease (CVD) ............133–135, 162, 167 Inflammation ..................... 126, 181–189, 208, 209, 222
Chronic obstructive pulmonary disease
(COPD) .................................................... 181–189 L
Circulating EBV DNA...................................99–101, 103
Low-density lipoprotein cholesterol
Co-expression............................................................39, 40 (LDL-C) ................................................... 133–141
Colorectal cancer (CRC) ..........................................33–37 Low-dose computed tomography (LDCT) .................. 91
Computer-aided diagnosis (CAD) ......................... 63, 64, Lung cancer ............................................... 36, 91–96, 111
134, 135
Lymphocyte-activation gene (LAG-3) .......................... 36
Congenital heart defect (CHD)................................... 173
Cytotoxic T lymphocyte-associated antigen-4 M
(CTLA-4).......................................................36, 37
Magnetic resonance imaging (MRI)................ 63, 65–67,
D 69, 100, 102, 110, 111
Major histocompatibility complex
Dilated cardiomyopathy (DCM)......................... 161–167 (MHC).................................................. 3, 195–198
DNA damage response (DDR) ............ 22, 76–80, 83, 86 Metastasis................................ 55, 57, 78, 91, 95, 96, 99,
DNA methylation ................................... 91–96, 123–130 100, 102, 104, 110, 111, 115, 116
Drug targets .................................................................... 18
Metastatic castration-resistant prostate cancer
Dysbacteriosis ................................................................ 182 (mCRPC).......................................................75–86
Microsatellite instability (MSI) ................................33–37
E
Microsatellite stable (MSS).......................................34–37
Early diagnosis...................................... 56–57, 63–70, 93, Mismatch repair (MMR) ................................... 33–37, 76
96, 99, 100, 106, 165, 167 Modules .............................................................. 25, 40–42
Ex Vivo Lung Perfusion (EVLP) ........................ 217–223 Molecular diagnostics .......................................... 110, 115
Multi-omics ..................................................................... 47
Tao Huang (ed.), Precision Medicine, Methods in Molecular Biology, vol. 2204, https://doi.org/10.1007/978-1-0716-0904-0,
225
PRECISION MEDICINE
226 Index
N R
Nanopore sequencing ...............................................13–31 Radiomics ........................................................................ 69
Nasopharyngeal carcinoma (NPC) ........................99–106 Randomized clinical trials (RCTs) ..............134, 136–140
Network embeddings ........................................ 40, 44–47 Random walks ............................................. 40, 43, 45, 46
Neutrophils........................................................... 182–189 Regression .............................................................. 41, 141
Next-generation sequencing (NGS) ...................... 14, 18, Resident microbiota .................................... 182, 183, 189
19, 23, 56, 79, 124, 177
Node2vec...................................................................44–46 S
Shortest paths............................................................40, 43
P
Structure variation (SV).................................................. 14
Point of care testing (POCT)......................................... 14 Systemic lupus erythematosus (SLE).................. 193–200
Precise drug sequential therapy........................... 145–158
Precision medicine ..................................... 53–59, 75, 76, T
86, 101, 106, 123–130, 162–167, 193–200,
T cell receptors (TCRs) .................................................... 3
207–210 Tissue engineering ............................................... 207–213
Prevention ..............................................54, 59, 139, 152, Total anomalous pulmonary venous connection
154, 163, 165, 167
(TAPVC)................................................... 173–178
Programmed death-1 (PD-1)............................ 36, 37, 76 Treatments............................................ 14, 18, 33–37, 46,
Programmed death-1 ligand 1 (PD-L1) .................36, 76 53–59, 63, 64, 68, 70, 75, 78–81, 85, 86, 95, 96,
Proprotein convertase subtilisin/kexin type 9
99–106, 115–117, 123, 124, 126–128, 130, 134,
(PCSK9).................................................... 134–141 136, 137, 139–141, 145–147, 149, 152, 154,
Prostate cancer (PCa) .......................................75, 76, 78, 157, 161–167, 182, 194, 198–200, 217, 221–222
86, 115, 208 Tumor suppressor genes (TSGs)................ 14, 54, 55, 95

Precision Medicine 2020

Uploaded by

Copyright:

Available Formats

Precision Medicine 2020

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Precision Medicine 2020

Uploaded by

Copyright:

Available Formats

Methods in

Molecular Biology 2204

Tao Huang Editor

METHODS IN MOLECULAR BIOLOGY

For further volumes:

ISSN 1064-3745 ISSN 1940-6029 (electronic)

© Springer Science+Business Media, LLC, part of Springer Nature 2020

With the development of omics technologies, especially next-generation sequencing, disease

Shanghai, China Tao Huang

PART I EMERGING EXPERIMENTAL AND BIOINFORMATICS TECHNOLOGIES

1 T Cell Receptor Repertoire Sequencing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

PART II PRECISION MEDICINE OF CANCERS

5 Diagnosis and Treatment of Breast Cancer in the Precision

10 A Review on Cancer of Unknown Primary Origin: The Role

PART III PRECISION MEDICINE OF CARDIOVASCULAR DISEASES

PART IV PRECISION MEDICINE OF OTHER COMPLEX DISEASES

16 Airway Inflammation Biomarker for Precise Management

PART V ENGINEERING AND SURGICAL DEVELOPMENTS

18 Precision Medicine in Tissue Engineering on Bone . . . . . . . . . . . . . . . . . . . . . . . . . 207

LIN BO • Department of Rheumatology, The Second Affiliated Hospital of Soochow

LINJING LI • Department of Clinical Laboratory Center, The Second Hospital of Lanzhou

Emerging Experimental and Bioinformatics Technologies

T Cell Receptor Repertoire Sequencing

T lymphocytes play essential roles in the adaptive immune system. A

After DNA random combining, the diversity is up to a mathemati-

Unless indicated, the reagents with analytical grade and ultrapure

T Cell Receptor Repertoire Sequencing 5

9. Resuspend the pellet with 180 μL Buffer ATL. Add 20 μL

T Cell Receptor Repertoire Sequencing 7

2. Place the EP tube on thermocycler (95 C for 2 min; 30 cycles

3.2.2 End Preparation 1. Prepare end-preparation reaction mix including 10 μL End

3.2.4 Clean Up 1. Add 56 μL (ratio 0.8) of Agencourt™ Ampure XP beads and

T Cell Receptor Repertoire Sequencing 9

java –jar mixcr.jar analyze amplicon -s <species> --

The parameter “species” refers to the organism, and the

1. Install R language and load the R package “tcR”:

2. Input data and data manipulation:

immdata1 <- parse.file("~/mixcr_data/immdata1.txt",

The parameter “immdata1.txt” is the output documents of

After analysis, a summary of counts of nucleotide and

repDiversity(immdata1, ’div’, ’read.count’)

It computes the Inverse Simpson Index, the command is as

repDiversity(immdata1, ’inv.simp’, ’read.prop’)

T Cell Receptor Repertoire Sequencing 11

It computes the Gini-Simpson index, the command is as

repDiversity(immdata1, ’gini.simp’, ’read.prop’)

It computes diversity of repertoire using Chao index, the

repDiversity(immdata1, ’chao1’, ’read.count’)

Nanopore Sequencing and Its Clinical Applications

Nanopore sequencing technology integrates nanopores with

14 Xue Sun et al.

turn-around time, portable and real-time data analysis, etc. also

2 Identification of Cancer-Related Structure Variations Using Nanopore Sequencing

2. In a 0.2 mL thin-walled PCR tube, mix the following in

Reagent Volume (μL)

Reagent Volume (μL)

7. Wash the beads by adding either 250 μL Long Fragment Buffer

Reagent Volume (μL)

Fig. 1 The bioinformatic pipeline to analyze nanopore sequencing data for

8. Add 75 μL mixed sample to the SpotON sample port in a