Shoulderology

Searching for “Whys” Related

to the Shoulder
Eiji Itoi

123
Shoulderology
Eiji Itoi

Shoulderology
Searching for “Whys” Related to the Shoulder
Eiji Itoi
Department of Orthopaedic Surgery
Tohoku Rosai Hospital
Sendai, Miyagi, Japan

ISBN 978-981-99-0344-3    ISBN 978-981-99-0345-0 (eBook)

https://doi.org/10.1007/978-981-99-0345-0

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore
Pte Ltd. 2023
The book was originally published in Japanese language by Igaku-Shoin Ltd. This English edition was
published as a co-edition with its original Japanese language edition, Katagaku - Rinsho no “Naze” to
sono Tsuikyu, copyright © 2021 by Igaku-Shoin Ltd., Tokyo Japan.
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of 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 Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Foreword

Looking back on history, we took care of many patients with whiplash injuries
caused by traffic accident in the 1970s, who complained of neck and shoulder pain.
We mainly focused on neck injuries rather than shoulder injuries because neck inju-
ries could be more serious. Then, we started to focus on patients with only shoulder
pain. They often complained of severe night pain, which made it difficult to fall
asleep. Arthrography was the only imaging modality useful for making a diagnosis
of shoulder diseases in those days. As the shoulder cases increased in number, I was
able to pile up the knowledge and experience of diagnosing and treating shoulder
patients. I asked then-professor Minoru Sakurai to send Dr. Itoi to my institute so
that I could teach him all the knowledge and experiences I had about taking care of
shoulder patients. Professor Sakurai accepted my request, and Dr. Itoi came to our
institute as a shoulder fellow right after he finished his thesis.
He spent only 1 year with me, but during his stay, he dramatically improved his
skills in making a diagnosis and treating shoulder patients. In addition, he analyzed
the patients’ data one after another, and published five articles in international jour-
nals. During this process, he became interested in shoulder function and biome-
chanics, which made him determine to go to Mayo Clinic for further research.
I was impressed by his eagerness to search for the meaning of what we were
doing routinely, like shoulder arthrography. He continued joint distension procedure
for patients with frozen shoulder until he thought that he understood the mechanism
how this treatment worked. When you look at the photographs of his examinations
in this book, you will understand his enthusiasm in taking care of patients and
searching for “whys.”
Following are the examples of his searching for whys. He provided evidence
through basic research and clinical trials that immobilization in external rotation
after first-time dislocation of the shoulder was better than immobilization in internal
rotation in reducing the risk of recurrence. When he was a chairman of the ISAKOS
Shoulder Committee, he picked up frozen shoulder as a topic for consensus meet-
ing. Later, this ISAKOS classification was accepted worldwide.

v
vi Foreword

In this book, he put a light on various functions and pathologies of the shoulder
and provided evidence to have a better understanding of the shoulder. His experi-
ence in studying abroad would encourage young doctors who would want to broaden
their perspectives.

Iwaki Central Hospital Shiro Tabata

Iwaki, Fukushima, Japan
December 2020
Foreword

The past three decades since 1990 have been a real blessing to have such a good
friend professionally and personally as Dr. Eiji Itoi. We collaborated in the biome-
chanical studies of the shoulder joint. We developed instruments and devices for
experimental assessment to understand the static and dynamic constraints of the
glenohumeral joint. During his time in the laboratory at Mayo, we discovered the
importance of the long head of the biceps in stabilization and movement of the
shoulder joint. Later on, he continuously encouraged and sent his colleagues to our
laboratory to further understand the importance of bony constraints to joint instabil-
ity. In the meantime, we were interested in understanding the biomechanical prin-
ciples associated with the etiology, diagnosis, and surgical treatment of the rotator
cuff tear. Each question we explored was clinically relevant, and thus made great
impact in the field of shoulder surgery and rehabilitation. All these achievements
from our collaboration reflect Dr. Itoi’s characteristics as stated in his recent article
in the Journal of Shoulder and Elbow Surgery entitled “What I’ve Learned.” He is a
surgeon of “curiosity,” asking “why” all the time. He was able to learn as “inflow”
from his mentors and then provide his own research and observations to share with
others as “outflow.” He has made a tremendous impact in the field of shoulder sur-
gery globally and has earned respect from his peers.
Specifically, he made several important contributions to the treatment of shoul-
der joint disorders, including joint instability, frozen shoulder, and rotator cuff tear.
His laboratory studies led to the identification of the stabilizing function of the
biceps in stable and unstable shoulders. Furthermore, he comprehensively studied
the mechanism of bony defects responsible for shoulder joint instability. His team
introduced the concept of the glenoid track, which helped us understand the dynamic
interaction between bony lesions on the glenoid and humeral head and facilitate
clinical diagnosis and treatment planning. Through experimental and analytical
studies, he explored the biomechanical factors associated with the etiology of rota-
tor cuff tear. Proper placement of the suture anchor based on the “Deadman Theory”
was revisited with his rigorous biomechanical analyses. These findings were all
evaluated in well-designed clinical trials and ultimately translated into patient care.

vii
viii Foreword

I am pleased that Dr. Itoi systematically compiled all these exciting and impor-
tant findings into this book entitled “Shoulderology: Searching for Whys Related to
the Shoulder.” The information in this book will be very useful for clinicians in
treating patients with shoulder pathology. It will also be beneficial to investigators
and junior surgeons to enjoy the outstanding models of translational research
between bench and bedside. I thank our Lord for giving the wisdom and gifts that
enable Dr. Itoi to make such a great contribution to the field, and also Dr. Itoi’s will-
ingness and stamina in educating the next generation, including through the publi-
cation of this wonderful book.

Mayo Clinic Kai-Nan An

Rochester, MN, USA
December 2020
Preface

I retired from Tohoku University in March 2021 after being an orthopedic surgeon
for 40 years and a shoulder surgeon for 34 years. I came across so many people
through my career and got a hint from them to answer some of the clinical questions
that I had. On my retirement, I was offered an opportunity from Igaku-Shoin to
write a monograph summarizing my career. Now that I am retiring from the front
line of clinical and research works, I would like to send a message to the next gen-
eration and the generation after next. In 2018, Bill Mallon, Editor of J Shoulder
Elbow Surg (JSES), asked me to write a short article entitled “What I’ve Learned”
to send a message to the young. In this JSES article, I picked up three topics I found
interesting to share. In each topic, I found various whys, and after confirming that
no studies on the topic had been found in the literature, I performed basic research
and clinical studies to find out the answer. This was how I approached my “whys.”
In the Introduction of this monograph, these three topics are introduced, showing
the importance of searching for whys and the process of solving the clinical ques-
tions. In Part I “Shoulder Examination,” I show how I performed physical examina-
tions of the shoulder using many photographs I took during the examinations so that
the readers can easily follow the text. In Parts II–IV “Main Shoulder Diseases,” I
have selected very common shoulder diseases such as rotator cuff tears, shoulder
instability, frozen shoulder, long head of the biceps tendon pathologies, and base-
ball shoulders. Especially, shoulder instability is the one I found most exciting, and
as a result, so many research works have been done. I have tried to systematically
construct the contents of these topics in this book. In Part V “Biomechanical
Research Related to Suture Anchors,” I show our research studies to clarify the
misunderstanding regarding the appropriate insertion angle of a suture anchor and
the safe distance between two suture anchors. In Part VI “Rare Cases,” I show two
cases, which were rare but quite interesting in terms of restoring the shoulder func-
tion. In the Front Matter “Message to the Young,” I have talked about the importance
of making friends in and outside of your own country, studying English hard, and
keeping on searching for whys. This book is a compilation of my own experience
and research works, so it does not cover the whole spectrum of shoulder diseases.
The basic philosophy of this book is that it is not the outcome or solution but the

ix
x Preface

process to solve a problem that matters. I hope that the readers enjoy this book and
would be persuaded to be involved in various research works. This English edition
was published as a co-edition with its original Japanese language edition, Kata
Gaku Rinsho No “Naze” to Sono Tsuikyu, first edition, copyright ©2021 by Igaku-­
Shoin Ltd., Tokyo, Japan. I would like to thank the staff members of Igaku-Shoin
and Springer for their support.

Sendai, Miyagi, Japan Eiji Itoi

March 2022
Message to the Young

Make Friends in and Outside of Your Country

In Tohoku University School of Medicine, a small group of fifth-year medical stu-

dents come to our department for 2-week clinical exposure every other week. I
always ask them if they had any experience of visiting foreign countries. Roughly
speaking, 20% of them said yes about 10 years ago. However, very few students say
yes these days. I am afraid they are getting more and more introverted. This is true
not only for the medical students, but also for the young generation in Japan. If you
look at the number of students studying abroad in Asian countries, the number is
increasing in almost all Asian countries, especially remarkably in India and China.
However, in Japan, it is steadily declining. This introversion seems to be a common
trend among the young generation in Japan. They probably think that they can get
anything they want and learn anything they want in Japan. Why bother to go abroad?
Obtaining information and knowledge can be achieved on the internet. That is true,
but it is a one-way learning. Going abroad is totally different. You meet the people
there with different culture and lifestyle, have conversation with them, get to know
each other, and make friends. You can have a wider perspective. When you attend an
international congress, it is a one-way learning if you just sit and listen to the pre-
sentations. If you find some interesting presentation which is close to what you are
doing for example, then after the presentation, talk to the presenter what you are
doing and what you felt about the presentation. Then, you may have an in-depth
discussion on the topic both of you are interested in. You exchange the name card or
email address so that you can communicate in the future. Keep doing this when you
attend an international congress. Some of them may become your lifetime friends. I
would like to say, “Be extroverted and look outside. Go abroad.”
Most of our works cannot be completed by us alone. During surgery, we need an
assistant. During research, we need support from others who set up the instruments,
take care of animals, and prepare cadavers. We always need support from others.
When you are a junior resident, you learn from the senior residents and staff mem-
bers. When you decide your own subspecialty, advice from them is helpful. So is it

xi
xii Message to the Young

when you do your research and write your thesis. When I visited Mayo Clinic as a
research fellow, I worked on eight research projects with one resident in the first
year and four research projects with two other residents in the second year. They
were not only friendly co-workers, but also excellent English teachers for me. After
coming back to Japan, I spent 12 years in Akita University and then 15 years in
Tohoku University. During this time, I received great support and collaboration
from my colleagues, who made it possible for me to fulfill my assignment as a
researcher, clinician, and teacher. A bright idea may pop into my mind during an
ordinary conversation or during surgery, which reminds me that being in contact
with other people is essential. At other times, when I was bogged down with a dif-
ficult problem, it was my colleagues who helped me and encouraged me. Looking
back on 40 years of my career, I was so blessed to have great colleagues and friends.
I am truly thankful to them, and I hope that my juniors would aim high, discipline
themselves, and develop through friendly competition.

Importance of English

To make friends worldwide, you need to learn English, an international tool neces-
sary to communicate with people from other countries. You may think it is too late
to start learning English, but I believe that each and every one of you has room for
growth. Especially I encourage those in the young generation, who have potency to
become a world-renowned surgeon or researcher. You do not need to be fluent in
English, but at least you need to convey what you want to say in English.
I will show you one example how the world expands if you give a talk in English.
It has been 10 years since the Great East Japan Earthquake and Tsunami. I could
learn various things from this disaster like what to do and what to prepare for a
future disaster. In 2017, I was the Congress President of the 90th Annual Meeting of
the Japanese Orthopaedic Association held in Sendai. I wanted to make this oppor-
tunity to share our experiences with the participants. We had close to 10,000 partici-
pants including 150 participants from abroad. Thus, I delivered Presidential Lecture
in English. Among the audience was Dr. Ondor Aydingoz from Turkey, who was
going to organize the Biennial Congress of the Asia Pacific Orthopaedic Association
in 2018. After my presentation, he came up to me and asked me if I could deliver the
same talk in APOA congress in Antalya, Turkey, the following year. Of course, I
said yes without hesitation. I gave this presentation in Antalya in 2018. Then, among
the audience was Dr. Per Kjærsgaard-Andersen from Denmark, the Congress
President of the 2019 Annual Meeting of the European Federation of National
Associations of Orthopaedics and Traumatology (EFORT) in Lisbon. I accepted
this invitation. In 2019, I gave this talk not only in Lisbon, but also in New York at
the Annual Meeting of the American Shoulder and Elbow Surgeons, and in Rennes
at the Annual Meeting of the French Arthroscopy Society. I am very grateful that I
could share my experience and thoughts through these presentations to different
people in different countries, especially the following issues: (1) We need to be
Message to the Young xiii

prepared for future disasters. They may come anytime anywhere. With preparation,
we cannot avoid the disaster, but we can minimize the damages. (2) We need to
think which direction we should go regarding energy. From the Fukushima nuclear
disaster, it is obvious to rely on renewable energy, not on nuclear energy. I could
share these issues with the audience simply because I could deliver the message in
English, an international language.
How can we improve our English skills? I want you to abandon the idea that
studying abroad is the closest way. In other words, you can improve your English
skills in your own country. I know some people who have been in English-speaking
countries for a long time but do not have excellent skills, and I know some people
who have never studied abroad but have excellent skills. My first trip abroad was to
attend the Annual Meeting of the Western Pacific Orthopaedic Association in Perth,
Australia. I was a second-year resident. My boss’s abstract had been accepted for
oral presentation, but since he was not good at speaking English, he invited me to
give a presentation instead of him. I was excited to attend the meeting with him.
When we arrived at Sydney Airport, a tour guide was waiting for us, a Japanese
woman who had been married to an Australian and living in Sydney for many years.
I was shocked to hear her English with a strong Japanese accent. Since she had been
married to an Australian and living in Sydney for many years, it was an ideal envi-
ronment to learn English. I thought she should have been able to speak English like
a native speaker. However, that was not true. This experience crashed my idea that
going and staying in English-speaking countries would make us speak English flu-
ently without effort. I want you to understand this. Babies can absorb English or any
language without effort, but once you become an adult, you must make efforts to
learn English.
After this experience in Sydney, I made more efforts to learn English. Then, I had
an opportunity to study abroad at Mayo Clinic. I was very careful about following a
rule that I made. I decided to ask native speakers to correct my English if I said
something weird or incorrect. They would never correct my English unless I asked
them to. Because I knew that in Japan, if I saw a foreigner who spoke a little bit of
Japanese, I would pay respect to the effort of learning Japanese and would not cor-
rect any mistakes unless I was asked to. Learning English was one of my main
purposes of studying abroad, and I decided to make it clear. Thus, when I first met
with a resident I was going to work with, I asked him to correct my English. And he
did. He advised me by saying, “We usually don’t say so,” or “We never say so,”
something like that. I wrote it down each time he gave me an advice or correction.
Next year, I did the same with other residents.
Also, I intentionally made various opportunities to learn English. I asked my par-
ent back in Japan to send me the cassette of radio English conversation program
every month. I made a point to read the front page of a local newspaper, mark all the
words I could not understand, look them up in the dictionary, and write them down
in a notebook. Chances will not come to you by themselves. You must make a
chance by yourself with all efforts.
xiv Message to the Young

I will explain how to learn English a little bit more from my experience. I think
there are many ways to do it. First thing you need to do is to know what your weak
point is. Is it listening or is it speaking or both?
There are three reasons why you are not good at listening:
1. You may not be familiar with the phonetics. The phonetics of your own mother
tongue may be quite different from English. You must become familiar with the
English sound. Practice makes perfect.
2. Your vocabulary may be limited. If you do not know the meaning of a word, you
cannot understand it regardless of how many times you listen to it. You must look
it up in the dictionary and memorize it at your desk.
3. You need to know the meaning of idioms. Idioms are commonly used in daily
conversation, but you seldom have a chance to learn them at school. This is also
something you must learn at your desk.
If you are not good at speaking English, the following reasons may be considered:
1. When you write a letter, you make a sentence in your mind and then write it
down. However, conversation is different. You already have various sentences in
the storeroom in your mind and reflexively take them out in response to some-
thing your conversation partner said. It is like a ping-pong game: as soon as you
receive the ball, you must hit it back. If you do not have enough sentences in
your storeroom, you need to make a sentence each time you reply. It takes time,
and the conversation does not go smoothly. You must keep various expressions
in your storeroom.
2. If you are keen of hearing, you can differentiate various sounds of pronunciation.
Then you can make various sounds by yourself. English vowel sounds are cre-
ated deep in the throat, whereas Japanese vowel sounds are created more super-
ficially. You must be familiar with different sounds in different languages.
3. You may be a born poor talker. This may not be easy to control, but there are
various ways available to improve your talking skills.
Points (2) and (3) are related to your born ability and thus may be difficult to
handle, but (1) is something you can definitely improve by your efforts. I encourage
you to keep working on it.
There is no royal road to learning. Spare no efforts in learning English.

Always Search for Whys

Looking back on my career, even major research that has established a new concept
started from a simple “why.” The question “Why does the long head of the biceps
(LHB) tendon run in contact with the humeral head?” initiated a series of research
to clarify the function of the LHB as a stabilizer and a mover. The question “Why
doesn’t a long and rigid immobilization prevent recurrent dislocations?” started a
series of basic and clinical research, which clarified the benefit of immobilization in
Message to the Young xv

external rotation. The question “Why do we use a nonscientific surgical indication

like 1/3 of the glenoid bone loss?” resulted in establishing the glenoid track concept
and clinical application of on-track/off-track lesions. You may have your own “why”
during your daily patient care. Cherish it and do the literature search to find any
clues or answers to your why. Most of the time, you can find an answer. However,
if you cannot, go ahead and search for the way how to answer your why. The pro-
cess to approach the goal may be painstaking and time consuming, but you can go
forward step-by-step. Finally, you will be able to find a new world, where no one
has ever reached. This excitement is really rewarding and blows away all the hard-
ships you have had. This is the real joy of research, and the real joy of shoulderol-
ogy!! As Albert Einstein said, “The important thing is not to stop questioning.” I
hope you keep questioning your whys and eventually delve into the mysteries of
shoulderology.
Acknowledgments

I express my sincere gratitude to my mentors, Dr. Shiro Tabata and Prof. Kai-Nan
An, for their kind instruction and guidance through my career. I am also grateful to
late Prof. Emeritus Minoru Sakurai for opening a gate to the shoulderology. I would
like to thank my colleagues who helped me to complete the research projects and
shared the joy of accomplishment together.
Those related to Tohoku University (1980–1990, 2006–Present):
Toshitake Aizawa, MD, PhD; Yoshiyuki Abe, MD, PhD; Akira Ando, MD, PhD;
Hirotada Ishii, MD, PhD; Hiroaki Ishikawa, PT, PhD; Koji Ishibashi, MD, PhD;
Ryogo Ichinose, MD, PhD; Yoshiaki Itoigawa, MD, PhD; Toshimitsu Etoh, MD,
PhD; Rei Omi, MD, PhD; Yasushi Omori, MD, PhD; Kenji Kanazawa, MD, PhD;
Jun Kawakami, MD, PhD; Atsuko Kanno, MD, PhD; Koshi Kishimoto, MD, PhD;
Jun Kamagai, MD, PhD; Daisuke Kurokawa, MD, PhD; Yoshiyuki Kuwahara, MD,
PhD; Yoichi Koike, MD, PhD; Masashi Koide, MD, PhD; Yoshimasa Sakoma, MD,
PhD; Katsumi Sato, MD, PhD; Hirotaka Sano, MD, PhD; Yuki Shiota, MD, PhD;
Kiyotsugu Shinagawa, MD, PhD; Nobuhisa Shinozaki, MD, PhD; Takuya
Sekiguchi, MD, PhD; Hiroyuki Takahashi, MD, PhD; Minoru Tanaka, MD, PhD;
Hideaki Nagamoto, MD, PhD; Yoshihiro Hagiwara, MD, PhD; Taku Hatta, MD,
PhD; Takashi Hayakawa, MD, PhD; Shin Hitachi, MD, PhD; Ko Himori, MD, PhD;
Ichiro Hirose, MD, PhD; Gaku Matsuzawa, MD, PhD; Mitsuyoshi Mineta, MD,
PhD; Takayuki Muraki, PT, PhD; Nobuyuki Yamamoto, MD, PhD
Those related to Akita University (1994–2006):
Toshiaki Aizawa, MD, PhD; Hidekazu Abe, MD, PhD; Masakazu Urayama,
MD, PhD; Tatsushi Kamata, MD, PhD; Kazuma Kikuchi, MD, PhD; Hiroaki
Kijima, MD, PhD; Tadato Kido, MD, PhD; Takashi Kobayashi, MD, PhD; Moto
Kobayashi, MD, PhD; Norikazu Konno, MD, PhD; Hidetomo Saito, MD, PhD;
Ryuji Sashi, MD, PhD; Takeshi Sato, MD, PhD; Akihisa Sano, MD, PhD; Togo
Shimizu, MD, PhD; Nobutoshi Seki, MD, PhD; Shu Takahashi, MD, PhD; Tsunehisa
Tsuchida, MD, PhD; Tatsuru Tomioka, MD, PhD; Yuji Hatakeyama, MD, PhD;
Hiroshi Matsuura, MD, PhD; Hiroshi Minagawa, MD, PhD; Naohisa Miyakoshi,
MD, PhD; Nobuyuki Yamamoto, MD, PhD; Ikuko Wakabayashi, MD, PhD;
Rabindra L. Pradhan, MD, PhD; Yilihamu Tuoheti, MD, PhD

xvii
xviii Acknowledgments

Those related to Mayo Clinic (1990–1993, 1997–1998):

Kai-Nan An, PhD; Bernard F. Morrey, MD; Stephen W. Carmichael, PhD; Robert
H. Cofield, MD; David K. Kuechle, MD; Neil E. Motzkin, MD; Stephen R. Newman,
MD; John W. Sperling, MD; Scott P. Steinmann, MD; Kimberly K. Amrami, MD;
Lawrence J. Berglund, BS
Finally, I would like to thank my wife, Mika, for supporting my work and family
without complaining of my coming home late at night due to research or being away
from home for almost half a month on overseas business trips.
Contents

1 Introduction����������������������������������������������������������������������������������������������    1

Part I Shoulder Examination

2 Physical Examination of the Shoulder ��������������������������������������������������    7

Part II Main Shoulder Diseases—Rotator Cuff Tear

3 Rotator Cuff Tear������������������������������������������������������������������������������������   29

Part III Main Shoulder Diseases—Shoulder Instability

4 Stabilizing Mechanisms ��������������������������������������������������������������������������   75
5 Traumatic Dislocation�����������������������������������������������������������������������������   89
6 Multidirectional Instability (Loose Shoulder)�������������������������������������� 135

Part IV Main Shoulder Diseases—Others

7 Frozen Shoulder �������������������������������������������������������������������������������������� 147
8 Long Head of the Biceps�������������������������������������������������������������������������� 167
9 Throwing Shoulder���������������������������������������������������������������������������������� 187

Part V Biomechanical Research Related to Suture Anchors

10 Anchor Insertion Angle: Deadman Theory and Its Interpretation ���� 207
11 
Minimal Safe Distance Between Suture Anchors: Relationship Between
the Anchor Distance and Pullout Strength�������������������������������������������� 217

xix
xx Contents

Part VI Rare Shoulder Diseases

12 Facioscapulohumeral Muscular Dystrophy (FSHD)���������������������������� 223
13 Trapezius Paralysis���������������������������������������������������������������������������������� 237

Index������������������������������������������������������������������������������������������������������������������ 245
Chapter 1
Introduction

I would like to talk a little bit about the title of this book Shoulderology first. This is
a coined word by blending “shoulder” and “ology” which means a branch of knowl-
edge. This word seems to be first used in a book entitled Shoulderology! Bustology!
published in 1896, written by William Henry Sladdin, a seamster in London. He
mentioned the importance of shoulder for men and bust for women when tailoring
clothes. Professor Hiroaki Fukuda, late professor of Tokai University, often used
this term, such as in a book title An Anthology of Shoulderology commemorating his
retirement from the university in 2001 and also in his article “Trends in modern
shoulder surgery: personal observations” in J Orthop Sci in 2007 [1]. A summary of
Charles S. Neer, II, a world renowned shoulder surgeon, known as the giant of the
shoulder, was published in J Shoulder Elbow Surg in 2009. In this article, Dr. Neer
was introduced as “Dr. Shoulderology” [2]. A direct translation of shoulderology
into Japanese language is “Katagaku,” an unsourced term used here and there.
Recently, Dr. Jun Kumagai, Congress President of the 37th Annual Meeting of
Japan Shoulder Society in 2010 picked up this word “Katagaku” as a theme of the
meeting and presented a mug with the logo “Katagaku” to all the participants. I am
sure some of the shoulder surgeons use this mug to drink coffee in the morning,
contemplating on the shoulder. Different from shoulder science, shoulderology
implies not only the meaning of knowledge and science, but also the meaning of
attractivity, charm, or mystery of the shoulder. This book is a culmination of all the
works I have done as a shoulder researcher and shoulder surgeon. However, differ-
ent from other textbooks, I put emphasis on the process, not the outcome. All the
new findings started from asking whys. Once you have whys, start literature search,
and if no studies have been done regarding your whys, design your own study to find
the answer and perform it. During this process, new whys may appear one after
another. Try to do the same to these new whys as well. Your research will expand
enormously into a wide spectrum. From this book, I would like to convey the mes-
sage how important and how exciting this process of finding an answer is. I hope the
readers would also start enjoying their own search for whys.

© The Author(s), under exclusive license to Springer Nature Singapore Pte 1

Ltd. 2023
E. Itoi, Shoulderology, https://doi.org/10.1007/978-981-99-0345-0_1
2 1 Introduction

Oftentimes, I am asked why I became a shoulder surgeon. I graduated from

Tohoku University School of Medicine in March 1980, and I became a resident at
National Mito Hospital in June. During the residency program of 3 years, I started
to be interested in knee surgery. Dr. Masanao Shimazaki, a grand disciple of Masaki
Watanabe, the Father of Arthroscopy, came to our hospital every week to perform
knee arthroscopy. I learned knee arthroscopy from him. In addition, I visited Dr.
Hiroshi Ikeuchi, a disciple of Masaki Watanabe, at Tokyo Teishin Hospital in Tokyo
every day for 3 months before I finished my training at National Mito Hospital. In
those days, we used Watanabe No. 21 arthroscope for knee examination, not for
knee surgery. If there was a meniscus tear, then we opened the joint and removed the
torn meniscus. Because of my experience in knee arthroscopy, I started to be inter-
ested in knee surgery. However, when I returned to Tohoku University in 1983, then
Professor Eikichi Wakamatsu assigned me to the bone metabolism team. I took care
of patients with osteoporosis and I was given a topic of my thesis related to osteopo-
rosis, a research on a relationship between the posture and standing balance in osteo-
porotic patients [3]. Then, Professor Minoru Sakurai took over Professor Wakamatsu
in 1986. He was to be a congress president of the 14th Annual Meeting of Japan
Shoulder Society in 1987. His specialty was peripheral nerve injuries, and his
method of detecting the peripheral nerve injury by assessing the function of sweat
glands called bromophenol blue printing method is well known worldwide. He had
done many reconstructive surgeries for paralyzed shoulder and elbow in patients
with brachial plexus injury and birth palsy. This was the reason why he was elected
to be the congress president of Japan Shoulder Society. On this occasion, he wanted
to establish a shoulder team in our department to perform clinical and research
works at the university. I was invited to join the team by the professor, and that was
why I started my career as a shoulder surgeon. The first thing I did was to follow up
the patients treated with trapezius transfer (Bateman procedure). The reconstructed
function of the shoulder varied greatly among the patients depending upon the loca-
tion of the transfer as well as the remnant function of the shoulder. This experience
stimulated my interest toward the function of the shoulder and its reconstruction.
The number of shoulder cases was quite limited at the university hospital.
Therefore, Dr. Shiro Tabata at Iwaki Kyoritsu Hospital, one of the most well-known
shoulder surgeons in Japan, invited me as a shoulder fellow for 1 year. He taught me
almost everything he had from shoulder examination to surgical techniques. He also
allowed me to analyze the data of patients piled up in his office. These analyzed data
were published one by one in international journals [4–9]. This fellowship aroused
many clinical questions in my mind, which persuaded me to study abroad to find
answers. First, I wanted to go to Professor Hans K. Uhthoff at the University of
Ottawa, but there was no position available. Professor Sakurai told me to contact a
Mayo doctor, Dr. Donald C. Campbell (foot and ankle surgeon), who visited us in
Sendai in 1986 to attend International Geriatric Orthopaedic Seminar. He recom-
mended that a research fellow at Orthopedic Biomechanics Laboratory would be the
best option if I could stay more than 1 year at Mayo Clinic. In April 1990, I had an
interview with Dr. Edmund Y.S. Chao, then director of the Lab, who asked me what
I wanted to do in his Lab. I talked to him about the various clinical questions that I
1 Introduction 3

had. Then he said that as several of my clinical questions could be solved in his Lab,
I was permitted to be a research fellow starting from October 1990. Dr. Kai-Nan An,
then the vice director of the Lab, was specializing in the biomechanics of upper
extremity. Under his guidance, I started shoulder research using the method of bio-
mechanics. It was fun and exciting to find an answer to each of my clinical questions
one after another. During my stay in the Lab for two and a half years, I completed
many biomechanical studies, some of which expanded later in the clinical field. I
would like to introduce three major studies out of them.
1. Function of the long head of the biceps (see the details in Chap. 8)
The long head of the biceps is well known as a flexor and supinator of the
forearm, but its function to the shoulder was not well described. Its anatomy is
unique in that it runs inside the shoulder joint, in close contact with the humeral
head. Why does the long head of the biceps tendon have unique anatomy of run-
ning in close contact with the humeral head? This was my first question. My
co-workers and I did several research studies regarding the stabilizing function
and motor function of the biceps. We found that it had a stabilizing function of
the shoulder, which was more prominent with the arm in external rotation. We
also found that it was a flexor, abductor, and horizonal flexor of the shoulder,
which was also affected by the rotation of the arm.
2. Immobilization in external rotation (see details in Chap. 5)
Immobilization in internal rotation using a sling for 3 weeks has long been
performed after reduction of shoulder dislocation. As a result, approximately
50% of the patients have recurrent dislocation after a first-time dislocation.
Longer and more rigid immobilization does not reduce the recurrence rate. Why?
If the lesion is well coapted, it should heal. If it does not heal, the lesion may not
be coapted in the conventional position of internal rotation. A cadaveric study
and an MRI study showed that the lesion was wide open with the arm in internal
rotation, whereas it was coapted with the arm in external rotation. A prospective
randomized clinical trial demonstrated that immobilization in external rotation
reduced the risk of recurrence.
3. Glenoid bone loss (see details in Chap. 5)
The textbooks used to say that bone graft should be considered when a gle-
noid bone loss is more than one third of the glenoid. However, nobody can tell
exactly that a lesion is one third of the glenoid. Using this eyeballing method, we
cannot precisely determine the size of bone loss, and we cannot define the criti-
cal size necessary for bone grafting. First, I quantitated the size of bone loss and
found the critical size. Next, we introduced a new concept of the glenoid track to
assess the risk of Hill–Sachs lesion. This concept is used worldwide these days.

As shown in these examples, all these research studies started from asking why.
After confirming that no studies had been performed through the literature search,
we started research studies to find out an answer. The first study of the long head of
the biceps was all done at Mayo Clinic and we received the Neer Award (best basic
science study) from the American Shoulder and Elbow Surgeons in 1992 [10]. The
4 1 Introduction

second study was mainly performed when I was in Akita University after coming
back from the USA. We again received the Neer Award (best clinical study) in 2003
[11]. The third study which was started at Mayo Clinic and continued through Akita
University and Tohoku University. We received the Most Downloaded Publication
Award and the Most Cited Publication Award from the Journal of Arthroscopy in
2019 [12]. Looking back my career, it started when Professor Sakurai invited me to
join the shoulder team, followed by Dr. Tabata who taught me all about clinical
aspects of the shoulder, and Dr. An who taught me all about the biomechanics of the
shoulder. Dr. Tabata and Dr. An are my clinical and basic research mentors. I owe
what I am now to these great mentors. I would like to express my sincere apprecia-
tion to them, who opened my eyes to this intriguing world of shoulderology.
Our lives are influenced by the people we encounter. Meeting with people you
admire as a mentor has a significant effect on your life. I am blessed that I was
allowed to learn about the shoulder, how to do patient care, how to perform research,
and how to publish the data, all from my mentors. I would like to emphasize not
only the excitement of finding out something new out of your research but also the
importance of meeting with your lifelong mentor. Best wishes to you all.

References

1. Fukuda H, Mikasa M. Trends in modern shoulder surgery: personal observations. J Orthop

Sci. 2007;12(1):4–13.
2. Bigliani LU, et al. Charles Neer: on the giant of the shoulder. J Shoulder Elb Surg.
2009;18(3):333–8.
3. Itoi E. Roentgenographic analysis of posture in spinal osteoporotics. Spine (Phila Pa 1976).
1991;16(7):750–6.
4. Itoi E, et al. Isokinetic strength after tears of the supraspinatus tendon. J Bone Jt Surg Br.
1997;79(1):77–82.
5. Itoi E, Tabata S. Conservative treatment of rotator cuff tears. Clin Orthop Relat Res.
1992;275:165–73.
6. Itoi E, Tabata S. Incomplete rotator cuff tears. Results of operative treatment. Clin Orthop
Relat Res. 1992;284:128–35.
7. Itoi E, Tabata S. Rotator cuff tears in the adolescent. Orthopedics. 1993;16(1):78–81.
8. Itoi E, Tabata S. Range of motion and arthrography in the frozen shoulder. J Shoulder Elb Surg.
1992;1(2):106–12.
9. Itoi E, Tabata S. Rotator cuff tears in anterior dislocation of the shoulder. Int Orthop.
1992;16(3):240–4.
10. Itoi E, et al. Stabilizing function of the long head of the biceps in the hanging arm position. J
Shoulder Elb Surg. 1994;3(3):135–42.
11. Itoi E, et al. A new method of immobilization after traumatic anterior dislocation of the shoul-
der: a preliminary study. J Shoulder Elb Surg. 2003;12(5):413–5.
12. Di Giacomo G, Itoi E, Burkhart SS. Evolving concept of bipolar bone loss and the Hill–Sachs
lesion: from “engaging/non-engaging” lesion to “on-track/off-track” lesion. Arthroscopy.
2014;30(1):90–8.
 Part I
Shoulder Examination
Chapter 2
Physical Examination of the Shoulder

Physical examination of the shoulder already starts when a patient comes into the
examination room. If a patient is holding one arm with another, you can tell without
asking the patient that he/she is holding the arm to avoid any movement of the
shoulder because a subtle movement can cause shoulder pain. It could be a shoulder
dislocation, calcifying tendinitis, etc. By observing the walking pattern such as
using a cane or using a wheelchair, you can roughly estimate how much external
load is applied to the shoulder during the activities of daily living.

1 Inspection

First, we observe the patient from behind to examine the location and degree of
atrophy of the shoulder muscles. The most common cases are those with muscle
atrophy of the supraspinatus and infraspinatus in patients with a massive rotator
cuff tear and isolated atrophy of the infraspinatus in patients with suprascapular
nerve palsy caused by a ganglion (Fig. 2.1). Among the shoulder muscles, deltoid
atrophy and infraspinatus atrophy are easy to detect because these muscles are
located just beneath the skin. Atrophy pattern caused by a nerve palsy is quite
specific to each nerve such as deltoid atrophy caused by axillary nerve palsy and
trapezius atrophy caused by accessory nerve palsy. You also need to pay attention
to the height of the scapula. The affected scapula is located higher in patients with
Sprengel deformity, whereas it is located lower and more lateral in patients with
accessory nerve palsy (Fig. 2.2). When the proximal tendon of the long head of
the biceps is ruptured, the muscle belly shifts distally, creating a hump of the
muscle, which looks like the arm of Popeye (Popeye sign) (Fig. 2.3). In case with
a rupture of the pectoralis major, a dimple corresponding to the location of the
rupture can be observed. In case with anterior shoulder dislocation, the humeral
head shifts anteriorly and inferiorly, creating an empty space beneath the

© The Author(s), under exclusive license to Springer Nature Singapore Pte 7

Ltd. 2023
E. Itoi, Shoulderology, https://doi.org/10.1007/978-981-99-0345-0_2
8 2 Physical Examination of the Shoulder

Fig. 2.1 Left

suprascapular nerve palsy.
A ganglion at the
spinoglenoid notch
compressed the
suprascapular nerve and
caused isolated atrophy of
the infraspinatus

Fig. 2.2 Right trapezius

palsy. The right scapula is
shifted laterally and
inferiorly compared to the
contralateral side

Fig. 2.3 Rupture of the

proximal tendon of the left
long head of the biceps. A
typical Popeye sign is
observed
2 Palpation 9

acromion. The humeral head seldom dislocates inferiorly, but if it does, the patient
comes to the clinic with the arm elevated overhead because the arm is locked in
this position due to inferior dislocation. Thus, it is called “luxatio erecta” (Latin
for erect dislocation). An acromioclavicular dislocation is clearly detectable with
the distal end of the clavicle protruding upward. In patients with multidirectional
instability, a sulcus is often observed between the acromion and the humeral head
because the head is inferiorly subluxated.

2 Palpation

A tear of the supraspinatus tendon is difficult to detect by observation because it

is covered by the large deltoid muscle and the acromion. However, it is easy to
palpate the tear with the shoulder in extension (Fig. 2.4). A muscle tear can also
be palpated if it is located subcutaneously. Tenderness is also an important phys-
ical finding. There is a localized tenderness along the bicipital groove when
there is a tendinitis of the long head of the biceps. Pain is suspected to be coming
from the acromioclavicular joint when there is localized joint tenderness together
with pain induction with the arm in horizontal flexion or abduction-internal
rotation.

Fig. 2.4 Palpation of a

delle (defect). With the
shoulder in slight
extension, the
supraspinatus tendon
comes out anteriorly from
under the acromion. In this
position, a full-thickness
tear of the supraspinatus
tendon can easily be
palpated as a delle
10 2 Physical Examination of the Shoulder

3 Shoulder Motion

Ask the patient to elevate/depress both arms together and observe the motion of the
arm and the scapula to see if there is any right and left difference. Pay special atten-
tion to the scapular motion. A marked scapular winging is observed with uncon-
trollable scapular motion due to serratus anterior palsy or trapezius palsy (Fig. 2.5).
A mild scapular winging may be seen during arm depression in case with a painful
impingement between the acromion and the rotator cuff tendon. Range of motion
of the shoulder is measured first with active motion followed by passive motion. If
both active and passive motions are restricted, it is a stiff shoulder. If the active
motion is restricted but not the passive one, it is likely to be caused either by
muscle weakness or by active motion pain. Elimination of pain after injection of
local anesthetics into the subacromial space enables us to differentiate these
two causes.

Fig. 2.5 Trapezius palsy.

During arm elevation, the
scapula rotates downward,
and the medial margin of
the scapula is protruded
posteriorly, showing a
typical scapular winging
5 Physical Examinations of Rotator Cuff Tear Patient 11

4 Muscle Strength

Muscle strength is usually detected with use of the manual muscle testing (MMT).
It is because this method is available anywhere without any special device. The
muscle strength varies greatly among patients depending upon age, gender, occupa-
tion, physical activity level, and so on. Thus, it is important to compare the involved
and uninvolved shoulders, and if you detect an obvious difference between them,
you are sure that there is muscle weakness. Next, you need to differentiate whether
the weakness comes from force production failure (muscle atrophy or muscle paral-
ysis) or force transmission failure (tendon rupture). Muscle weakness caused by a
transient nerve palsy recovers gradually with various speed. Using a dynamometer,
the recovery pattern of muscle strength can objectively be demonstrated at various
time points.

5 Physical Examination of Rotator Cuff Tear Patient

5.1 Pain Induction

Pain is often induced by motion in patients with a rotator cuff tear. When a pain is
induced during active elevation or depression of the arm, it is called a positive
“painful arc sign.” This is caused by mechanical impingement between the acro-
mion and the rotator cuff tendon. On the other hand, if a pain is induced during
passive elevation or depression of the arm, it is called a positive “impingement
sign.” While holding the scapula to avoid scapular rotation, the arm is passively
elevated forward and depressed downward. If a pain appears around 90° of arm
elevation, it is called a positive “Neer impingement sign.” This maneuver can be
performed with the arm in internal rotation and in external rotation (Fig. 2.6). As the
site of impingement changes with different arm rotations, it is helpful to narrow
down the location of impingement. Another maneuver is called Hawkins impinge-
ment sign. In this maneuver, the humerus is internally rotated with the arm kept at
90° of flexion or 90° of abduction (Fig. 2.7). If an injection of local anesthetics into
the subacromial bursa eliminates the pain and converts the impingement sign from
positive to negative, it is called a positive “impingement test.”

5.2 Supraspinatus Tear

The supraspinatus tendon is the most common site of tear. You can palpate a
tear beneath the deltoid muscle. This tendon is located just beneath the acro-
mion with the arm in the hanging position. With the shoulder in extension, the
12 2 Physical Examination of the Shoulder

a b

Fig. 2.6 Neer impingement sign. With one hand holding the scapula to avoid rotation, the arm is
passively elevated forward and depressed downward. It can be done with the arm in internal rota-
tion (a) or in external rotation (b)

tendon comes out anteriorly from under the acromion. In this position, the
supraspinatus tendon can be easily palpated just anterior to the acromion.
Rotation of the arm makes it even easier to palpate a delle (defect) of the tendon
(Fig. 2.4). The diagnostic accuracy of palpation to detect a full-thickness tear
of the supraspinatus tendon is reported to be equivalent to that of MRI or
ultrasonography.
Next, you perform a site-specific test for a supraspinatus tear. Ask a patient to
elevate both arms up to 90° in the scapular plane with the thumb pointing downward
and ask the patient to resist downward force applied to both arms (Fig. 2.8). If
muscle weakness is detected during this procedure (positive supraspinatus test), it
indicates a tear of the supraspinatus. This test is also called “empty can test” because
the thumb-down position is similar to a position to empty a can. This position, how-
ever, is apt to cause subacromial pain, which also affects the muscle strength. To
avoid pain provocation during this test, a modified test is proposed. This test is basi-
cally the same as the empty can test except that it is performed in thumb-up
5 Physical Examinations of Rotator Cuff Tear Patient 13

a b

Fig. 2.7 Hawkins impingement sign. The original maneuver is to passively rotate the arm inter-
nally while keeping the shoulder in 90° of flexion (a). Modified method is to do the same proce-
dure with the arm in 90° of abduction (b) [1]

a b

Fig. 2.8 Empty can test. While keeping the arm at 90° of elevation in the scapular plane with the
thumb pointing downward (internal rotation of the humerus), an examiner applies downward force
to bilateral arms (a). If a patient cannot keep the elevated position due to muscle weakness (b), the
empty can test is positive
14 2 Physical Examination of the Shoulder

a b

Fig. 2.9 Full can test. The same test as the empty can test except that the thumb is pointing upward
is called the full can test (a). If a patient cannot keep the elevated position due to muscle weakness,
the test is positive (b)

position, which is like a position to hold a full can (Fig. 2.9). Because of this posi-
tion, this modified supraspinatus test is also called “full can test.” The diagnostic
accuracy of the full can test is 75% and that of the empty can test is 70% without a
significant difference [2].

5.3 Infraspinatus/Teres Minor Tear

The second most common site of tear is the infraspinatus. As it is an important

external rotator of the shoulder, muscle weakness in external rotation is obvious
once it is torn. The best position to assess the muscle strength of the infraspinatus is
the neutral rotation in the hanging arm position. If the MMT for external rotators is
weaker than Grade 2 (Poor), a patient cannot keep the position of passive maximum
external rotation, resulting in less external rotation of the arm. This phenomenon is
called an “external rotation lag sign” (Figs. 2.10 and 2.11). If a tear extends into the
teres minor tendon, external rotation strength in abduction is also decreased.
Bringing your hand to your mouth requires the external rotators to keep the arm in
neutral rotation in abducted position. When both the infraspinatus and teres minor
are torn, the patient has to elevate the internally rotated arm with the shoulder
abducted and the wrist hyperextended to touch the mouth. Since this position is
similar to the arm position of a hornblower, this is called a hornblower’s sign
(Fig. 2.12). The teres minor is also required to keep the arm in abduction and exter-
nal rotation. A patient with a tear of the infraspinatus and teres minor cannot keep
the arm in abduction and external rotation, resulting in internal rotation of the arm.
5 Physical Examinations of Rotator Cuff Tear Patient 15

a b

Fig. 2.10 External rotation lag sign (uninvolved side). The left shoulder, uninvolved side, is in
passive maximum external rotation (a). This position can be actively maintained (b), meaning that
the external rotation lag sign is negative

a b

Fig. 2.11 External rotation lag sign (involved side). On the right side with an infraspinatus tear,
the arm is kept in passive maximum external rotation (a), but the patient cannot keep the arm in
this position and it returns to a neutral rotation (b). The external rotation lag sign is positive, and
the lag is about 60°

Fig. 2.12 Hornblower’s

sign. This patient brings
the right hand to the mouth
by elevating the elbow
high and hyperextending
the wrist (a positive
hornblower’s sign). This is
because the shoulder is in
internal rotation and the
forearm is horizontal to the
floor due to a tear of
infraspinatus and teres
minor. He can bring the
left hand to the mouth
without elevating the
elbow because he can keep
the forearm pointing
upward due to the function
of the intact external
rotators
16 2 Physical Examination of the Shoulder

This is called a dropping sign (Figs. 2.13 and 2.14). Clinically, it is diagnosed as an
infraspinatus tear if an external rotation lag (a difference between the passive and
active ranges of external rotation) is less than 40°, negative hornblower’s sign, and
negative dropping sign, whereas it is diagnosed as a tear involving both the infraspi-
natus and teres minor if an external rotation lag is greater than 40°, positive horn-
blower’s sign, and positive dropping sign [3].

a b

Fig. 2.13 Dropping sign (intact side). With the arm in abduction and external rotation (a), the
examiner lets the arm go and it remains in the same position because the teres minor is intact (b)

a b

Fig. 2.14 Dropping sign (involved side). The passive position of abduction and external rotation
(a) cannot be maintained actively, and the forearm drops down to neutral rotation (horizontal to the
floor) (b). This positive dropping sign indicates that the teres minor is not functioning
5 Physical Examinations of Rotator Cuff Tear Patient 17

5.4 Subscapularis Tear

The subscapularis is one of the internal rotators. Because there are other internal
rotators like the pectoralis major, latissimus dorsi, and teres major, it is difficult to
detect muscle weakness of the subscapularis with the arm in adduction. In a specific
position close to the maximum internal rotation, all the other internal rotators can-
not function anymore, and it is the subscapularis that works as an internal rotator in
this position. This phenomenon can be tested with the hand placed on the back at the
lumbar area. In this position, all the internal rotators except the subscapularis do not
function anymore. Further internal rotation such as lifting the hand off the back can
be achieved by the function of the subscapularis. If the subscapularis is torn, the
hand on the back cannot be lifted off. This is called a positive lift-off test (Fig. 2.15).
During the lift-off test, we observe the active motion. To observe the passive motion
is more demonstrable (internal rotation lag sign). While keeping the hand passively
elevated in a lift-off position, the examiner asks the patient to keep this position and
let the hand go. If the hand remains in the same position, internal rotation lag sign
is negative with the intact function of the subscapularis (Fig. 2.16). If the hand drops
to the back, the sign is positive with a functional deficit of the subscapularis
(Fig. 2.17). If the hand cannot be placed on the back due to limited range of internal
rotation, the belly press test can be used instead. Ask the patient to press the belly
while keeping the hand and elbow in the same frontal plane. This motion can be
done only when the subscapularis is intact. If the subscapularis is weak, the patient
cannot press the belly and tries to do so with the shoulder in extension and the wrist
in flexion (Fig. 2.18). This is a positive belly press test. Another method to detect a
functional deficit of the subscapularis is the bear-hug test. The palm is placed on top
of the opposite shoulder and ask the patient to keep this position while the examiner

a b

Fig. 2.15 Lift-off test. The left hand (intact side) can be lifted off the back due to the function of
the subscapularis (a). The right hand (involved side) cannot be lifted off the back due to the sub-
scapularis weakness (b). The lift-off test is positive on the right side
18 2 Physical Examination of the Shoulder

a b

Fig. 2.16 Internal rotation lag sign (intact side). The left hand is passively lifted off the back to
keep the shoulder in maximum internal rotation (a). Ask the patient to keep this position and let the
hand go. If the subscapularis functions normally, the patient can keep this position of maximum
internal rotation (negative internal rotation lag sign) (b)

a b

Fig. 2.17 Internal rotation lag sign (involved side). On the affected side, the hand is passively
lifted off the back (a), but the patient cannot keep this position due to a functional deficit of the
subscapularis (b). The internal rotation lag sign is positive on the right side
5 Physical Examinations of Rotator Cuff Tear Patient 19

a b

Fig. 2.18 Belly press test. On the intact side, the patient can press the belly while keeping the hand
and elbow in the same frontal plane (a). On the affected side, the patient cannot press the belly
while keeping the hand and elbow in the same plane (positive belly press test). He tries to press the
belly with the shoulder extended and the wrist flexed to compensate for the deficit of the subscapu-
laris, but in vain (b)

a b

Fig. 2.19 Bear-hug test. On the intact side (left shoulder), the patient can press the opposite shoul-
der strongly with his left hand so that the examiner cannot elevate the hand from the shoulder (a).
On the affected side (right shoulder), he cannot firmly press the left shoulder with his right hand
because of the subscapularis tear so that the examiner can easily elevate the hand from the shoulder
(b). The bear-hug test is positive on the right side

tries to pull the hand from the shoulder. If the subscapularis function is normal, the
examiner cannot pull the patient’s hand. If the subscapularis is weak, the examiner
can easily pull the hand from the shoulder (Fig. 2.19). This is a positive bear-­
hug test.
20 2 Physical Examination of the Shoulder

6 Physical Examination of Shoulder Instability Patient

6.1 Traumatic Instability

An anterior dislocation accounts for 97% of first-time or recurrent dislocations.

Sometimes, the translation of the humeral head may be incomplete, and the humeral
head reduces spontaneously. It is called a subluxation. However, a complete disloca-
tion may be reduced spontaneously if the shoulder joint is lax. This is also called
subluxation in the clinical setting. Thus, it is sometimes difficult to clearly differen-
tiate a dislocation (the humeral head completely out of the glenoid socket) from a
subluxation (the head halfway out of the socket). Furthermore, it is not necessary to
differentiate them when selecting a treatment option. Because of these reasons, the
term “instability” is more commonly used these days including both dislocation and
subluxation. Most patients with anterior instability feel apprehension (fear of dislo-
cation) when the arm is placed in abduction, external rotation, and horizontal exten-
sion. This is called the anterior apprehension test (Fig. 2.20). In this position, the
anteroinferior glenohumeral ligament (AIGHL) is supposed to become tight and
prevent the anterior translation of the humeral head. However, this stabilizing mech-
anism is deficient in patients with recurrent anterior instability, and as a result, the
humeral head shifts anteriorly and causes apprehensive feeling in this arm position.
Since there is a risk of causing an anterior dislocation during this test, I grasp the
humeral head to avoid a dislocation in patients with a strong apprehension. The
standard anterior apprehension test is performed with the arm at 90° of abduction,
90° of external rotation, and forced horizontal extension. This position reflects the
IGHL injury around 4–5 o’clock position, which is the most common site of injury.
If the lesion extends more superiorly (2–3 o’clock), lower angle of arm abduction
causes anterior apprehension, whereas the lesion extending more inferiorly (6–7
o’clock) would cause anterior apprehension at higher angle of abduction. Therefore,
I make it a rule to check the anterior apprehension at three different angles of abduc-
tion: 60°, 90°, and 120°.
A posterior instability accounts for only 3% of shoulder instability. Patients with
recurrent posterior instability have a deficiency of the posteroinferior glenohumeral
ligament (PIGHL) to prevent a posterior translation of the humeral head with the
arm in flexion and internal rotation. Thus, placing the patient’s arm in flexion and
internal rotation causes an apprehensive feeling for patients with posterior instabil-
ity. This is called the posterior apprehension test (Fig. 2.21).

6.2 Multidirectional Instability

Abovementioned recurrent anterior and posterior instabilities occur as a result of

traumatic dislocation. Another group of instability known as multidirectional insta-
bility occurs without any major trauma, so it is also called atraumatic instability.
6 Physical Examinations for Shoulder Instability 21

a b

Fig. 2.20 Anterior apprehension test. The most common position of the arm is 90° of abduction,
90° of external rotation, and horizontal extension (b). I always check the apprehension at 60° of
abduction (a) and 120° of abduction (c) to assess the extent of IGHL injury

Hisao Endo was the first who described this pathology in 1971. When Dr. Neer was
invited to the third Annual Meeting of the Japan Shoulder Society in 1976, he had
an opportunity to listen to Dr. Endo’s lifework regarding inferior subluxation in
hyperlax shoulders (he called it loose shoulders). Dr. Neer was very impressed with
Dr. Endo’s report. When Dr. Neer published his first paper regarding multidirec-
tional instability in 1980, he cited Dr. Endo’s report at the Annual Meeting of the
Japan Shoulder Society as a reference. Dr. Neer’s paper popularized the concept of
multidirectional instability throughout the world. Years later, the concepts of laxity
and instability were clarified, and it is now considered that in most cases of MDI,
the pathology is categorized as unidirectional instability with multidirectional lax-
ity. However, since some cases have multidirectional instability with multidirec-
tional laxity and the term MDI is used comprehensively, I use the term MDI covering
all instabilities related to multidirectional laxity in this book.
22 2 Physical Examination of the Shoulder

Fig. 2.21 Posterior

apprehension test. With the
arm at 90° of flexion and
internal rotation, the
examiner applies a
posterior force through the
elbow. If the humeral head
shifts posteriorly and
causes an apprehensive
feeling, it is assessed as a
positive posterior
apprehension test

There is no normal value for laxity. Someone may have a great laxity to be able
to subluxate or even dislocate the shoulder, but if the laxity accompanies no symp-
toms, it is considered physiologic and no treatment is necessary. If the laxity
accompanies symptoms such as pain, it is pathologic and called instability. During
examination of a patient with MDI, check the laxity of the shoulder in various
directions. Anterior and posterior laxity can be detected by the anterior and poste-
rior drawer tests (also known as the load and shift test) (Fig. 2.22), and inferior
laxity by a sulcus sign, which appears below the acromion due to the inferior trans-
lation of the humeral head (Fig. 2.23). At the same time, ask the patient whether the
induced translation of the humeral head causes any symptoms. If it does, confirm
whether it is the same symptom as the patient usually feels. MDI patients also show
inferior translation of the humeral head with the arm in abduction. This inferior
translation can be demonstrated using the abduction inferior stability test (ABIS
test) (Fig. 2.24) [4]. By applying downward force to the proximal humerus with the
arm in abduction, the humeral head shifts inferiorly, creating a step on the skin
surface. The hyperabduction test also becomes positive in patients with MDI
(Fig. 2.25) [5]. This test is to check the length or tension of the IGHL by passively
abducting the arm while pressing the scapula downward to avoid scapular rotation
during this maneuver. Passive abduction is less than 90° in normal shoulders due to
the tension of the IGHL. However, if it is greater than 105°, it means either the
IGHL is elongated by birth (hyperlaxity) or ruptured by a trauma (traumatic
dislocation).
6 Physical Examinations for Shoulder Instability 23

a b

Fig. 2.22 Laxity test. Posterior drawer test (a), neutral position (b), and anterior drawer test (c).
This maneuver is also known as the load and shift test

a b

Fig. 2.23 Sulcus sign. With the arm hanging freely (a), apply a downward force to the arm. Then
the humeral head translates inferiorly according to the degree of shoulder laxity and a sulcus
appears under the acromion (b). This is called a sulcus sign
24 2 Physical Examination of the Shoulder

a b

Fig. 2.24 Abduction inferior stability test (ABIS test). With the arm in abduction (a), apply down-
ward force to the proximal arm. In a patient with shoulder hyperlaxity, the humeral head shifts
inferiorly, creating a step on the skin surface (b). The ABIS test is positive in this case

a b

Fig. 2.25 Hyperabduction test. In this case, passive abduction angle is more than 105° on both
sides (a, b). Since the patient has no history of trauma and does not have any symptoms, this posi-
tive test indicates that he has elongated IGHL by birth or shoulder hyperlaxity
References 25

References

1. Itoi E. Rotator cuff tear: physical examination and conservative treatment. J Orthop Sci.
2013;18(2):197–204.
2. Itoi E, et al. Which is more useful, the “full can test” or the “empty can test,” in detecting the
torn supraspinatus tendon? Am J Sports Med. 1999;27(1):65–8.
3. Collin P, et al. What is the best clinical test for assessment of the teres minor in massive rotator
cuff tears? Clin Orthop Relat Res. 2015;473(9):2959–66.
4. Itoi E, et al. Scapular inclination and inferior stability of the shoulder. J Shoulder Elb Surg.
1992;1(3):131–9.
5. Gagey OJ, Gagey N. The hyperabduction test. J Bone Jt Surg Br. 2001;83(1):69–74.
 Part II
Main Shoulder Diseases—Rotator Cuff
Tear
Chapter 3
Rotator Cuff Tear

1 Anatomy

The rotator cuff is the name for a group of four muscles and their tendons attaching
to the scapula as origins and the proximal humerus as insertions. They are the supra-
spinatus, infraspinatus, teres minor, and subscapularis. The tendinous portions of
these four muscles unite and cover the entire humeral head except for a gap between
the supraspinatus and the subscapularis, which is called the rotator interval.
Gray’s Anatomy describes that the insertion of the supraspinatus is the superior
facet, that of the infraspinatus is the middle facet, and that of the teres minor is the
inferior facet of the greater tuberosity of the humerus. However, a meticulous exam-
ination of these tendons revealed that the supraspinatus tendon attaches not only to
the superior facet, but also to the upper half of the middle facet. This posterior half
of the supraspinatus tendon is covered from posterior by the infraspinatus tendon
(Fig. 3.1) [1]. The anterior half of the supraspinatus tendon with an average width
of 12.6 mm is composed of a single tendon of the supraspinatus, but the posterior
half with an average width of 9.8 mm is a double-layer structure with the supraspi-
natus tendon in deep layer and the infraspinatus tendon in superficial layer. The
anteroposterior width of the supraspinatus tendon is 22 mm on average. A 20-mm
tear in width used to be thought a single tendon tear of the supraspinatus, but it
contains both the supraspinatus and infraspinatus tendons. We think that this ana-
tomical complexity would explain why a patient with a supraspinatus tear often has
an infraspinatus atrophy. This anatomical relationship was further studied from a
different group [2].
The supraspinatus muscle looks like a fusiform muscle, but there is an intramus-
cular tendon inside the muscle belly (Fig. 3.2). The muscle fibers run obliquely and
attach to the intramuscular tendon. Different from a bipennate muscle which is a
planar muscle, the supraspinatus has the muscle fibers all around the intramuscular
tendon. Thus, this muscle is called a circumpennate muscle. After removing all the
muscle fibers, the intramuscular tendon is clearly visible, which is located not in the

© The Author(s), under exclusive license to Springer Nature Singapore Pte 29

Ltd. 2023
E. Itoi, Shoulderology, https://doi.org/10.1007/978-981-99-0345-0_3
30 3 Rotator Cuff Tear

a b

Fig. 3.1 Surface view of the supraspinatus and infraspinatus. (a) The first layer of the tendon cov-
ers both supraspinatus and infraspinatus tendons together, so the tendon proper (the second layer)
cannot be observed. (b) Removing the first layer reveals the second layer of these tendons. The
posterior half of the supraspinatus tendon is covered by the infraspinatus tendon from behind. SSP
supraspinatus, ISP infraspinatus. (Reproduced with modification from Minagawa et al. [1])

a b

Fig. 3.2 The supraspinatus. (a) The muscle looks like a fusiform muscle from outward appear-
ance. (b) Separating the muscle belly, the intramuscular tendon appears inside the belly. The mus-
cle fibers run obliquely, parallel to each other, and attach to the intramuscular tendon

center but in the anterior part of the muscle belly (Fig. 3.3) [3]. Distally, the intra-
muscular tendon extends into the anterior 1/3 of the extramuscular tendon.
Approximately 70% of the muscle fibers attach to the intramuscular tendon, whereas
the rest 30% attach directly to the extramuscular tendon. Similarly, the infraspinatus
(Fig. 3.4) and the teres minor (Fig. 3.5) have a single intramuscular tendon. These
muscles with a single intramuscular tendon are called uni-circumpennate muscle.
On the other hand, the subscapularis, which has 5 or 6 independent intramuscular
tendons, is called a multi-circumpennate muscle (Fig. 3.6). The circumpennate
muscles have shorter muscle fibers but greater in numbers compared to the fusiform
muscles. This means that the excursion of the circumpennate muscle is short, but
they can produce greater force compared to the fusiform muscle. Arm elevation is
composed of the humeral motion relative to the scapula (scapulohumeral motion)
and scapular motion relative to the trunk (scapulothoracic motion). Because of this
motion sharing, not a great excursion but a great force against the deltoid is required
for the rotator cuff muscles. The circumpennate muscles are very reasonable in
shape to achieve this purpose.
1 Anatomy 31

Fig. 3.3 The

intramuscular tendon of
the supraspinatus. The
intramuscular tendon is
located not in the center
but in the anterior 1/3 of
the muscle belly. This
tendon extends into the
anterior 1/3 of the
extramuscular tendon.
(Reproduced with
modification from
Minagawa et al. [3])

Fig. 3.4 The intramuscular tendon of the infraspinatus. There is a single intramuscular tendon in
the infraspinatus, which consists of two parts: horizontal tendon fiber bundles (a) and the oblique
tendon fiber bundles (b). Distally, the oblique part goes under the horizontal part and both parts
blend and attach to the greater tuberosity. (a) View from the dorsal side, (b) view from the ventral
side. (Reproduced with modification from Minagawa et al. [3])
32 3 Rotator Cuff Tear

Fig. 3.5 The intramuscular tendon of the teres minor. There is a single intramuscular tendon with
the tendon fiber bundles running parallel. (Reproduced with modification from Minagawa et al. [3])

Fig. 3.6 The intramuscular tendons of the subscapularis. There are several independent intramus-
cular tendons, which merge to form the extramuscular tendon. (Reproduced with modification
from Minagawa et al. [3])
2 Function of the Rotator Cuff 33

2 Function of the Rotator Cuff

There are various methods to determine the muscle function: (1) estimate the func-
tion based on the origin and insertion of the muscle, (2) stimulate the muscle to
observe how the muscle contraction affects the joint motion, and (3) measure the
moment arm of the muscle. The moment arm is the length of the perpendicular line
from the center of rotation to the line of action. The textbooks describe the function
of the rotator cuff muscles as follows: the supraspinatus is an abductor, the infraspi-
natus and teres minor are external rotators, and the subscapularis is an internal rota-
tor. These basic functions of the rotator cuff muscles have been confirmed by
electromyographic activities of these muscles during shoulder motion in live
patients. We investigated the functions of these muscles in cadaveric shoulders by
measuring the kinematic moment arms during various motions of the shoulder [4–
7]. These research studies were performed in the Orthopedic Biomechanics
Laboratory, Mayo Clinic.
Based on these studies, we found that an abductor function of the supraspinatus
was much less than it used to be thought, but instead the subscapularis had much
greater function as an abductor [4]. This was also confirmed by the study using posi-
tron emission tomography (PET) [8]. In addition, the following findings were dem-
onstrated in our moment arm study of shoulder muscles other than the rotator cuff.
The anterior deltoid functioned mainly as a flexor, and the middle deltoid as an
abductor. Adduction of the shoulder was achieved by the pectoralis major, latissi-
mus dorsi, and teres major. Horizontal flexion was mainly achieved by the pectoralis
major, whereas horizontal extension by the posterior deltoid and infraspinatus. The
external rotators were the infraspinatus, teres minor, and posterior deltoid in the
order of strength both in adduction and in abduction, whereas the main internal rota-
tors were the subscapularis and pectoralis major [5]. Furthermore, we evaluated the
function of arm muscles to the shoulder. The long head of the biceps was a flexor
and abductor, while the short head of the biceps and coracobrachialis were adduc-
tors, and the long head of the triceps was an extensor and adductor [7]. The long
head of the biceps changed its rotation function with the change of arm position. It
was an internal rotator with the arm in flexion, but an external rotator with the arm
in abduction.
From these data, we can calculate how much each muscle contribute to the
strength during various motions of the shoulder. The torque each muscle produces
can be calculated by the following equation,

Torque = PCSA × 4.7 kg / cm 2 × MA × %MVC,

where PCSA is physiological cross-sectional area (cm2), MA is moment arm,

and %MVC is percent maximum voluntary contraction. With the torques of all the
muscles, we can estimate the percentage contribution of each muscle to shoulder
torque production (Table 3.1).
34 3 Rotator Cuff Tear

Table 3.1 Contribution of shoulder muscles to shoulder motion

Abduction External rotation Internal rotation
Deltoid 76% 11% –
Supraspinatus 10% 5% –
Infraspinatus 1% 66% –
Teres minor – 18% –
Subscapularis 13% – 52%
Pectoralis major – – 39%
Teres major – – 5%
Latissimus dorsi – – 3%
Contribution was calculated from the moment arms [4, 5], physiological cross-sectional areas [9],
and muscle activities [10, 11]

The kinematic moment arms obtained from cadavers give us precise data about
the function of shoulder muscles. However, the kinematic moment arms are obtained
during passive motion of the shoulder. Muscle activities during active motion have
been studied mainly by using electromyography. Disadvantages of electromyogra-
phy using surface electrodes are that it can record activities of only the superficial
muscles and that it is difficult to select a single muscle. On the other hand, using
needle electrodes, we can select a single muscle or a deep muscle easily. However,
normal physiological muscle activities may be disturbed by the presence of pain at
the site of needle electrodes. Recently, several research studies assessed the muscle
activities during motion with use of positron emission tomography (PET). The prin-
ciple of this method is as follows. When a muscle contracts, it takes glucose into the
muscle fibers as a source of energy. When glucose labeled with positron-emitting
radionuclide 18F is injected before a series of muscle exercise, distribution of the
radionuclide is observed only in the muscle fibers that were activated during the
exercise. Simply put, we can visualize the muscle activities. Using this method, we
detected the distribution of 18F-fluorodeoxyglucose (FDG) after repetitive abduction
exercise. FDG was predominantly seen in the upper 1/3 of the subscapularis muscle
(Fig. 3.7) [8]. As this muscle has a wide expansion in the craniocaudal direction, the
superior 1/3 is located above the center of rotation of the humeral head and func-
tions as an abductor, whereas the inferior 2/3 are silent because they are located
below the center of rotation. This finding supports the data of above-mentioned
moment arm study.
We further investigated the external and internal rotators using the same method.
The infraspinatus and teres minor are well known as the external rotators. We com-
pared their functions in adduction and in abduction [12]. In both positions, the infra-
spinatus and teres minor showed much higher FDG uptake than the other rotator
cuff muscles. The infraspinatus showed the highest uptake in adduction, whereas
the teres minor showed the highest uptake in abduction (Fig. 3.8). In other words,
the infraspinatus is the main external rotator in adduction and the teres minor in
abduction. It is clinically known that patients with a tear of the infraspinatus show a
negative hornblower’s sign, but those with a tear involves both the infraspinatus and
2 Function of the Rotator Cuff 35

Fig. 3.7 FDG uptake after

abduction exercise. This
PET scan shows a
localized high uptake of
FDG in the superior 1/3 of
the subscapularis muscle
(white arrow). (Reprinted
from Fig. 6, Omi et al. [8])

a b

Fig. 3.8 FDG uptake after external rotation exercise. In adduction (a), the infraspinatus showed
the highest uptake of FDG (white arrowheads). In abduction (b), the teres minor showed the high-
est uptake (black arrowheads). (Reprinted from Fig. 4, Kurokawa et al. [12])

teres minor show a positive sign. The data obtained in this PET study gives support
to these clinical findings (also refer to Chap. 2 Physical Examination).
Why does the main mover change in adduction and abduction? In adduction, the
muscle fibers in the upper half of the infraspinatus run almost perpendicular to the
long axis of the humerus (Fig. 3.9a). These muscle fibers are in the most efficient
orientation in producing a torque on the humeral shaft. On the other hand, the teres
minor is in the most efficient orientation with the arm in abduction (Fig. 3.9b).
36 3 Rotator Cuff Tear

a b

Fig. 3.9 Perpendicular portion of the external rotators to the humerus. In adduction (a), the mus-
cle fibers of the upper half of the infraspinatus are almost perpendicular to the long axis of the
humeral shaft, and accordingly more beneficial in producing a torque on the humeral shaft than the
teres minor. In abduction (b), however, the teres minor runs almost perpendicular to the humeral
shaft and functions more efficiently as an external rotator than the infraspinatus. ISP infraspinatus,
TM teres minor

These anatomical aspects seem to explain why the infraspinatus worked most in
adduction and the teres minor in abduction.
Similarly, we performed a study investigating the function of internal rotators in
adduction and abduction. This study revealed that the middle and inferior 2/3 of the
subscapularis mainly worked in adduction and the inferior 1/3 mainly worked in
abduction (Fig. 3.10) [13].
This shift of functional portion of the muscle in different arm positions may also
be explained by the anatomical relationship to the humerus. The middle 1/3 of the
subscapularis is perpendicular to the humeral shaft with the arm in adduction,
whereas the inferior 1/3 is perpendicular to the humeral shaft with the arm in abduc-
tion (Fig. 3.11). However, the middle 1/3 showed lower uptake than the inferior 1/3
with the arm in adduction. Why is this? There are many internal rotators such as the
pectoralis major, teres major, and latissimus dorsi, sharing the function with the
subscapularis. Among these internal rotators, the pectoralis major is a large muscle,
and its middle portion is perpendicular to the humeral shaft with the arm in adduc-
tion. Because of the contribution of the pectoralis major, the middle 1/3 of the sub-
scapularis might be less active. Whatever the reason, it is interesting that the main
internal and external rotators change with the change of the arm position.
Another method to estimate the function of the muscle is to measure the muscle
strength of a patient with a single tendon tear of the rotator cuff with use of a
2 Function of the Rotator Cuff 37

a b

Fig. 3.10 FDG uptake after internal rotation exercise. In adduction (a), the inferior 1/3 of the
subscapularis showed the highest uptake among the three portions of the muscle. In abduction (b),
the inferior 1/3 further increased the uptake, followed by the middle 1/3, which also showed higher
uptake than in adduction. The superior 1/3 showed constant uptake regardless of arm position. S
superior 1/3, M middle 1/3, I inferior 1/3. (Reprinted from Fig. 5, Matsuzawa et al. [13])

a b

Fig. 3.11 Perpendicular portion of the subscapularis to the humerus. In adduction (a), the middle
1/3 of the subscapularis is perpendicular to the humeral shaft. In abduction (b), the inferior 1/3 is
perpendicular to the humeral shaft. SSC subscapularis
38 3 Rotator Cuff Tear

dynamometer. Comparing with the intact side, we can estimate the contribution of
the muscle with a torn tendon to the measured motion. We measured the abduction
strength of patients with an isolated tear of the supraspinatus with and without
local anesthetics. Without pain relief, the abduction strength was approximately
60% of the contralateral shoulder. However, after pain relief, it increased up to
80% [14]. This means that the pain decreased abduction strength by 20% and the
force transmission failure caused by a tendon tear also decreased the
strength by 20%.

3 Epidemiology

When I was working at Akita University, we performed an epidemiological

study at Kamikoani village in the mountain area. A man who was born and bred
in this village entered Akita University School of Medicine for the first time in
the village history. The villagers enjoyed a grand display of fireworks celebrat-
ing his entrance to the medical school. This man became an orthopedic surgeon
and became a member of our department. This was the main reason we selected
this village for our survey, and everyone was cooperative to us, thanks to his
presence. We asked the history of shoulder symptoms if any, took physical find-
ings of bilateral shoulders, and performed ultrasonography to detect a full-
thickness rotator cuff tear in 1328 shoulders of 664 residents aged older than
20 years. We reported the data at the 79th Annual Meeting of the Japanese
Orthopaedic Association in 2006 [15]. According to this survey, the prevalence
of full-thickness rotator cuff tear increased with age, such as 0% in the 20–40 s,
10.7% in the 50 s, 15.2% in the 60 s, 26.5% in the 70 s, and 36.6% in the 80 s
and above. Interestingly, only one third of those with a rotator cuff tear had
symptoms such as pain, and the rest two third were symptom free. Although a
publication of this study was delayed [16], this was the first epidemiological
study in Japan. Similar data regarding the prevalence of rotator cuff tear were
reported later in 2010, from a survey of 1366 shoulders of 683 residents in a
village in Gunma prefecture [17].

4 Mechanism of Tear

There are two main theories: intrinsic theory attributing a tear to intrinsic factors
versus extrinsic theory attributing a tear to extrinsic factors. Degeneration of the
cuff tendon is believed to be the main intrinsic factor, which is modified and aggra-
vated by aging, stress overloading, smoking, etc. On the other hand, a contact or
impingement between the cuff tendon and the coracoacromial arch is the main
extrinsic factor.
4 Mechanism of Tear 39

4.1 Degeneration of the Tendon

It has been well known that a tendon degeneration advances with age [18], whereas
the mechanical strength decreases with age [19, 20]. These findings suggest that
degenerative changes of the tendon occurring with age are the cause of a degenera-
tive tear. We performed histological examination of tendon specimens obtained
from 5 patients with rotator cuff tendinitis with a positive impingement sign and
compared them with the intact tendon specimens from 10 shoulders of 10 autopsy
cases as controls [21]. In tendinitis cases, disturbance of collagen fiber orientation
and angiogenesis were observed with a significantly greater number of apoptotic
cells compared to the control shoulders. The apoptotic index expressed as the num-
ber of apoptotic cells among 500 cells was 7% in the normal tendon, 25% in the
tendinitis group. The apoptotic index was reported to be 34% in specimens with a
rotator cuff tear [22]. Considering these numbers, it may be legitimate to presume
the cuff tendinitis as a preceding stage of rotator cuff tears.

4.2 Stress Concentration

We investigated the stress distribution inside the supraspinatus tendon, simulating

elevation and depression of the arm in the scapular plane with a 1-kg weight in the
hand [23]. From this study, we found that there was a stress concentration on the
articular surface of the supraspinatus tendon near its insertion to the greater tuberos-
ity (Fig. 3.12). This area of stress concentration coincides with an onset area of
articular-side tears (Fig. 3.13). This mechanical load seems to be associated with a
degeneration and tear of the rotator cuff tendon.

a b c

Fig. 3.12 Tensile stress concentration in the supraspinatus tendon. In the hanging arm position
(a), a slight stress concentration was observed on the articular side of the tendon approximately
3 cm from the insertion site. As the arm was elevated in the scapular plane to 30° (b) and 60° (c),
the area of stress concentration became more focused and shifted distally close to the tendon inser-
tion site. (Reprinted from Wakabayashi et al. [23])
40 3 Rotator Cuff Tear

Fig. 3.13 MRI of an

articular-side partial-­
thickness tear. Detention of
the joint fluid indicates the
location of an articular-side
partial-thickness tear
(white arrows). This site
coincides well with the site
of stress concentration in
Fig. 3.12

4.3 Impingement Between the Tendon

and the Coracoacromial Arch

The supraspinatus tendon which is located most superiorly among the cuff tendons
is just beneath the acromion. With the motion of the arm, the supraspinatus tendon
slides underneath the acromion in close contact with each other. It is easy to imagine
that anything in this narrow space such as subacromial spur or tendon swelling
caused by an inflammation would lead to a serious conflict between them. First, we
wanted to know how the supraspinatus tendon would slide under the acromion in
the physiological status. Using cadaveric shoulders with intact rotator cuff tendons,
we evaluated the contact pattern between the cuff tendon and the coracoacromial
arch [24]. We found that the contact pressure increased during the activities of daily
living such as flexion, abduction, and horizontal extension. Especially during hori-
zontal extension, the contact pressure under the acromion increased up to five times
greater than that under the coracoacromial ligament. As shown in this study, there
are various degrees of contact between the cuff tendon and the coracoacromial arch
in physiological status. Later, we found that the contact pressure decreased with an
anterior tilt of the scapula [25] and increased with joint contracture [26].
Charles S. Neer was the first to propose a hypothesis that this repetitive contact
or impingement between the cuff tendon and the coracoacromial arch would cause
inflammation and an eventual tear of the tendon [27–29]. He called a series of
pathologic changes of the cuff tendon an “impingement syndrome” with three
stages: stage I, a reversible change characterized by edema and bleeding; stage II, a
combination of subacromial bursal fibrillation and cuff tendinitis; and stage III, a
tendon tear. He reported that impingement sign was useful in making a diagnosis of
4 Mechanism of Tear 41

impingement syndrome. While holding the scapula with one hand to prevent scapu-
lar rotation, the examiner elevates the patient’s arm passively. This maneuver, called
Neer impingement sign, induces pain coming from mechanical impingement
between the cuff tendon and the coracoacromial arch (see more details in Chap. 2.
Physical Examination) [28]. If this pain disappears after an injection of local anes-
thetics into the subacromial bursa, it is called a positive impingement test [29].
Another impingement sign is Hawkins impingement sign, which induces pain by
rotating the arm internally while keeping the arm at 90° of flexion [30]. Are these
signs demonstrating the same pathology? In order to answer this question, we per-
formed a cadaveric study to clarify which parts of the tendon and the coracoacro-
mial arch would be in contact with each other during these maneuvers [31]. On the
coracoacromial arch side, both maneuvers showed a similar contact area, i.e., the
coracoacromial ligament and its attachment to the acromion, the lateral border of
the acromion. On the cuff tendon side, the supraspinatus tendon was in contact with
the coracoacromial arch during the Neer impingement sign, whereas the subscapu-
laris was during the Hawkins impingement sign. In other words, the Neer and
Hawkins impingement signs reveal different location of the cuff tendon pathology.
Since the first report by Neer, many studies focused on the relationship between
the acromion and a rotator cuff tear. Among them, the one by Bigliani et al. is well
known. They classified the shape of acromion into three types: type I, flat; type II,
curved; and type III, hooked. According to them, a rotator cuff tear was seldom seen
in those with type I acromion, whereas it was seen in 20–30% of those with type II
acromion and 70–80% of those with type III acromion [32]. This concept that a
hooked acromion causes a rotator cuff tear has dominated the world ever since
(Fig. 3.14) [32, 33].
It is understandable that a bursal-side tear may be caused by the hook of the
acromion by continuous friction between them. However, it is difficult to accept that
the hooked acromion causes an intra-tendinous tear or even an articular-side tear,
which is on the other side of the hook. In addition, further anatomical studies
showed that the ratio of acromial types was different from the original report.
Bigliani et al. reported that the ratios of type I, II, III were 17%, 43%, 39%, or
approximately 2:4:4 [32]. However, studies using cadavers or dry bones revealed
that the ratios were 21–23% for type I, 68–72% for type II, and 7–9% for type III or
approximately 2:7:1 [34, 35]. Type III, a possible cause of a rotator cuff tear, was far
less common than the original report. Furthermore, if you look at the dry bone acro-
mion from various directions, you will notice that the same acromion may be clas-
sified as type I, type II, or type III depending upon the direction you look at
(Fig. 3.15).

a b c

Fig. 3.14 Acromial shape. (a) Type I (flat), (b) type II (curved), (c) type III (hooked)
42 3 Rotator Cuff Tear

a b c

Fig. 3.15 Acromial morphology. The same dry bone acromion may look flat (a), curved (b), or
hooked (c) depending upon the viewing direction

As shown above, the typing of acromial morphology is not reproducible, and

accordingly, it is considered unreasonable to use acromial morphology in discuss-
ing etiology of rotator cuff tears. Some investigators reported that there was no
relationship between the acromial morphology and rotator cuff tears [36]. If the
hooked acromion causes a cuff tear, only the hooked portion would be in contact
with the remnant cuff tendon (Fig. 3.14c). We hypothesized that the contact area
would be narrower in shoulders with a cuff tear compared to those without. We
investigated the contact area between the acromion and the rotator cuff tendon using
cadaveric shoulders [37]. This study showed that an anteroposterior dimension of
the contact area was the same between those with and without a tear, but a medial–
lateral dimension was significantly greater in those with a tear. This finding was
opposite to our hypothesis and provided negative data regarding the acromial mor-
phology as a cause of a cuff tear.
Because of these findings, the classification of type I, II, and III are seldom used
these days. Instead of acromial morphology, more focus has been put to (1) the
subacromial spur and (2) the location of the acromion relative to the humeral head
in the sagittal and coronal planes. Regarding the osteophytes, there are various mor-
phological types, but the heel type is reported to be associated with a cuff tear [36].
Zuckerman et al. were the first who paid attention to the location of the acromion
in the sagittal plane [38]. They measured the location of the tip of the acromion in
the sagittal plane using cadaveric shoulders and found that the acromial tip of those
with a rotator cuff tear was located an average 3.8 mm more anterior than those
without. We also performed a similar study by taking sagittal-view photographs of
cadaveric shoulders with and without a cuff tear. We measured an anterior acromial
projection angle defined as an angle between the vertical line passing through the
center of the humeral head and the line connecting the acromial tip and the center of
humeral head (Fig. 3.16). This angle was −2° on average in normal shoulders and
6° on average in shoulders with a cuff tear [39].
The location of the acromion relative to the humeral head in the sagittal plane is
clearly visible on the scapular Y view of plain X-rays (Fig. 3.17). In normal
4 Mechanism of Tear 43

a b

Fig. 3.16 Anterior acromial projection angle. This angle is positive when the acromial tip is ante-
rior to the vertical line and negative when it is posterior to the vertical line. Normal shoulders (a)
showed an average −2°, whereas shoulders with a cuff tear (b) showed an average 6° [39]

shoulders (Fig. 3.17a), the acromion is located posterosuperior to the humeral head.
The coracoacromial ligament (orange line) is just above the supraspinatus tendon
(yellow line). On the other hand, in shoulders with a rotator cuff tear, the acromion
projects more anteriorly and is located above the humeral head. The supraspinatus
tendon (yellow line) is always under the hard bone (acromion, while dotted line). As
described above, the pressure under the acromion is much higher than that under the
coracoacromial ligament. This anatomical relationship seems to put the supraspina-
tus tendon at risk of a tear.
More recently, the location of acromion relative to the humeral head in the coro-
nal plane has been paid attention to. Nyffeler et al. were the first to notice that in cuff
tear shoulders, the acromion extended more laterally, covering a large portion of the
humeral head on the coronal view of plain X-rays [40]. They measured the acro-
mion-to-humeral head ratio and called it the “acromion index” (Fig. 3.18).
According to their report, the average acromion index was 0.73 in cuff tear
shoulders, 0.64 in normal shoulders, and 0.60 in osteoarthritis shoulders. Their
impression that the acromion was projected more laterally in cuff tear shoulders was
proved by this study, showing that the acromion index was significantly greater in
cuff tear shoulders than in normal shoulders.
The question is why the lateral projection of the acromion is related to cuff tears.
According to their explanation, the force vector of the deltoid was more vertical
with a lateral projection of the acromion (large acromion index). They speculated
that the vertical component of this vector would push the humerus upward, increase
44 3 Rotator Cuff Tear

a b

Fig. 3.17 The acromion and the humeral head on the scapular Y view. In normal shoulders (a), the
acromion is located posterosuperior to the humeral head. The coracoacromial ligament (orange
line) is on top of the supraspinatus tendon (yellow line). In shoulders with a cuff tear (b), the acro-
mion projects more anteriorly and the supraspinatus tendon (yellow line) is just beneath the acro-
mion (white dotted line)

Fig. 3.18 Acromion index.

A distance between a line
connecting the superior
point (P1) and inferior
point (P2) of the glenoid
and a line parallel to the
first line and passing
through the lateral margin
of the acromion (P3) is
defined as “A”. Another
distance between the P1–P2
line and a line parallel to
this line and passing
through the lateral margin
of the greater tuberosity
(P4) is defined as “H”. The
ratio A/H is defined as the
“acromion index”
4 Mechanism of Tear 45

a b

Fig. 3.19 Force vector of the deltoid. With a small acromion index (a), the resultant force F of the
deltoid is pointing more medially, with a small vertical component FV and a large horizontal com-
ponent FH. The small FV is less likely to push the head upward and the large FH stabilizes the
humeral head into the glenoid socket. On the other hand, with a large acromion index (b), the verti-
cal component FV is much greater, which pushes the head upward against the acromion

the contact pressure under the acromion, and possibly advance degenerative process
and an eventual tear of the tendon (Fig. 3.19).
In their original paper, Nyffeler et al. showed that the acromion index was 0.60 in
osteoarthritis shoulders. Osteoarthritis makes the joint space narrower and makes
the distance “H” in Fig. 3.18 shorter than before the osteoarthritis has developed,
which in turn makes the acromion index larger than it truly is. To assess the cover-
age by the acromion more accurately regardless of the location of the humeral head,
they introduced another parameter, the “critical shoulder angle” (Fig. 3.20). This is
an angle between the line P1–P2 and the line P2–P3.
The critical shoulder angle is determined by the three points on the scapula, so it
is not affected by the narrowing of the glenohumeral joint space or the location of
the humeral head. They reported that the average critical shoulder angle was 33.1°
in the normal shoulders, 38.0° in cuff tear shoulders, and 28.1° in osteoarthritis
shoulders [41]. If the critical shoulder angle was less than 30°, 93% of them had
osteoarthritis. If it was equal to or greater than 35°, 84% had a cuff tear. As Spiegl
et al. also reported similar findings [42], it is widely accepted in Europe and North
America that the normal critical shoulder angle ranges 30°–35°. More than 35°, a
risk of cuff tear increases, and less than 30°, a risk of osteoarthritis increases.
46 3 Rotator Cuff Tear

Fig. 3.20 Critical shoulder

angle. An angle between
the line P1–P2 and the line
P2–P3 is defined as the
critical shoulder angle

When the critical shoulder angle is large, the vertical component of the deltoid
resultant force increases and pushes the head upward against the acromion. This
may advance degeneration of the tendon. Furthermore, the small horizontal compo-
nent cannot stabilize the humeral head in the glenoid socket during arm elevation.
To support centralization of the humeral head, the rotator cuff muscles are required
to pull the humeral head into the glenoid socket. This requirement of producing
large forces acting through the cuff tendons may also aggravate the degenerative
process of the tendon. Gerber et al. proved this hypothesis in a cadaveric study.
When they increased the critical angle from 33° to 38°, a greater force (13–33%
increase) was required for the supraspinatus muscle to elevate the arm at the initial
phase of elevation [43]. As shown in this study, a greater contraction force is passing
through the supraspinatus tendon whenever elevating the arm, which may be related
degeneration and a tear of the tendon.
The acromion index and the critical shoulder angle may vary among the races.
We had a research fellow from Brazil in our department, and I asked him to compare
the acromion index between the Japanese and Brazilian populations. The average
acromion index in the Japanese was 0.67 in normal shoulders and 0.68 in cuff tear
shoulders; there was no significant difference [44]. On the other hand, the acromion
index in the Brazilians was 0.72 in cuff tear shoulders, which was significantly
greater than 0.67 in normal shoulders. We also compared the critical shoulder angle
in the Japanese with the previous reports from other countries. The average critical
shoulder angle in the Japanese was 33.9° in cuff tear shoulders, which was signifi-
cantly greater than 32.3° in normal shoulders [45]. Although the critical shoulder
angle in cuff tear shoulders was significantly greater than that in normal shoulders
4 Mechanism of Tear 47

in the Japanese, it was within the normal range (30°–35°) in Europe and North
America. The critical shoulder angle in osteoarthritis shoulders, on the other hand,
was 33.0° in the Japanese, which was also within the normal range and showed no
significant difference from the normal values. It was rare to find those with the criti-
cal shoulder angle less than 30° among the Japanese population. This might explain
why the primary osteoarthritis of the shoulder is so uncommon in the Asian coun-
tries than in other areas.

4.4 Smoking

Smoking has long been linked to adverse effects on the respiratory and cardiovascu-
lar systems. Recently, its negative effects on the musculoskeletal system have
attracted attention such as low bone mineral density, risk of fracture, delayed bone
union, low back pain, delayed wound healing, and high risk of surgical site infec-
tion. The abstracts of the tenth Annual Meeting of the American Shoulder and
Elbow Surgeons were published in J Shoulder Elbow Surg in 1995. I found an inter-
esting abstract by Mallon et al. in which they reported that (1) the smoking rate
among 57 consecutive patients who underwent cuff repair surgery was 82%, which
was significantly higher than that of the US population (27%), and (2) the surgical
outcomes were worse in smokers than in non-smokers. As far as I know, this was the
first study to shed light on a link between smoking and cuff tear. Encouraged by this
abstract, we did a retrospective survey of 104 consecutive patients who underwent
cuff repair surgery. We found that the smoking rate of cuff tear patients was 38%,
which was almost the same as that of general population in Japan (36%). However,
the smoking index (the number of cigarettes smoked per day multiplied by years
smoked) and the size of cuff tear showed a significant positive correlation, indicat-
ing that smoking might be a risk factor of tear propagation. This study was pre-
sented at the 22nd Annual Meeting of the Japan Shoulder Society in 1995. This
study was very well accepted as a novel and unique study. However, some senior
smokers in the audience who were influential in the Society were not happy and
asked me why smoking would cause a tear. I could have said some plausible but
boring answer such as “Nicotine is related to vasoconstriction and deteriorate blood
circulation, which might be related to a tear.” Instead, I answered with a serious
face, “When you smoke, your arm comes to a position of impingement. Daily rep-
etition of this extremely high-risk motion would drive your cuff tendon to a tear.”
This answer provoked a laugh among the audience, and the senior smokers did not
argue with me anymore. This study was published in Japanese in Katakansetsu, an
official journal of the Society [46]. Later, we investigated the risk factors of cuff tear
propagation and found that smoking was one of the risk factors of tear propagation
[47]. An epidemiological study from the USA demonstrated that smoking was a risk
factor not only for tear propagation but also for the onset of cuff tears [48]. We per-
formed further studies to clarify the mechanism of cuff tear by smoking. Low-dose
or high-dose nicotine was continuously injected into the subcutaneous tissue of the
48 3 Rotator Cuff Tear

rats for 12 weeks and biomechanical properties of the supraspinatus tendon was
measured [49]. Although no differences were observed in the failure load and maxi-
mum stress, the modulus of elasticity, regardless of nicotine dosage, was signifi-
cantly higher than in the control group. This change in biomechanical properties
may be related to cuff tear. Furthermore, we investigated the effects of nicotine on
primary porcine tenocytes exposed to cyclic stretch [50]. Tenocytes exposed to nic-
otine represented decreased gene expression of MMP-9 and TIMP-3 under the
cyclic stretch. Downregulation of these gene expression might deteriorate normal
metabolism of the tendon.

5 Tear Propagation

As mentioned above, the stress concentration on the articular side of the supraspi-
natus tendon might be related to the onset of articular-side tears. How does the
stress concentration pattern change once a tear occurs? Is there any relationship
between the site of initial tear and how it expands? These clinical questions led us
to perform a study to clarify the stress concentration with and without a tear using a
2D finite element model analysis [51]. This study revealed that when a tear was on
the articular side, the stress concentration was observed at the bottom of the tear,
which seemed to further make the tear deeper and larger (Fig. 3.21a). However,
when a tear was on the bursal side, there was no stress concentration neither at the
bottom of nor around the tear. This finding strongly suggests that the stress distribu-
tion in the tendon has nothing to do with the occurrence or propagation of bursal-­
side tear. It would rather be related to an acromial spur (Fig. 3.21b).
We were also interested in how a full-thickness tear would propagate. In a 3D
finite element model analysis, we found that a stress concentration was observed at
the anterior and posterior edges of a tear when an abduction force was applied to the

a b

Fig. 3.21 Stress distribution with partial-thickness tears. With an articular-side tear (a), a stress
concentration is observed at the bottom of the tear. However, with a bursal-side tear (b), no stress
concentration is observed at the bottom of or around the tear. (Reprinted from Figs. 7A and 8A,
Sano et al. [51])
6 Muscle Changes with Cuff Tear 49

Fig. 3.22 Stress distribution with three different sizes of tears. From left to right: a small-sized
tear (1 cm), a medium-sized tear (2 cm) and a large-sized tear (3 cm). As a tear became larger, the
stress increased from yellow to red and the stress concentration area became wider at the anterior
and posterior edges of the tear. (Reprinted from Fig. 3A, Sano et al. [52])

tendons (Fig. 3.22) [52]. The larger the tear, the greater the stress and the wider the
stress concentration area. This seems to explain why a tear tends to progress in the
anteroposterior direction like undoing a zipper known as “zipper phenomenon.”
Based on these biomechanical data, we followed up 174 shoulders of 171 patients
with a cuff tear for an average 19 months to observe the change in tear size [47].
This study showed that tear propagation was observed in 82 shoulders (47%) with
the speed of 3.8 mm per year in length and 2.0 mm per year in width. The risk fac-
tors of tear progression were (1) a full-thickness tear, (2) a medium-sized tear, and
(3) smoking.

6 Muscle Changes with Cuff Tear

6.1 Mechanism of Fatty Change

It is well known that the cuff muscles undergo atrophy with a cuff tear. Muscle
atrophy usually accompanies fat accumulation in the muscle. Goutallier’s grading
of fatty change of the muscle using CT is widely used [53], but MRI is more com-
monly used these days instead of CT to avoid radiation. What is the mechanism of
fat accumulation? In the murine myogenic cell line, inhibition of Wnt10b upregu-
lated adipogenic marker gene expression and induced adipocytes [54]. Also in the
murine rotator cuff tear model, fatty change and its molecular events were remark-
ably seen in the distal 1/3 of the detached supraspinatus muscle. We concluded that
Wnt10b was a key signal in fat accumulation in the muscle with a cuff tear. Similarly,
in the murine myogenic cell line treated by hypoxia, adipogenic differentiation of
myoblasts was observed and adipogenic marker gene expression was increased
[55]. However, in this experiment, Wnt signaling was not decreased but increased
opposite to our expectation. Thus, adipogenic differentiation observed in hypoxia
50 3 Rotator Cuff Tear

seems to be different from that triggered by downregulation of Wnt signaling. We

further investigated the effect of cuff repair on the fatty changes in the leporine cuff
tear model [56]. The leporine supraspinatus tendon was resected and repaired at
1 week, 2 weeks, and 3 weeks. Five weeks after the repair, adipocytes were stained
and gene expressions of adipogenic differentiation markers and Wnt10b were
assessed in the supraspinatus muscle. The adipocytes in the muscle increased sig-
nificantly after 3 weeks of tendon resection. Wnt10b was downregulated at 1 week
after the resection, whereas adipogenic differentiation markers were increased at 2
and 3 weeks of resection and remained increased at 5 weeks after the repair. When
the tendon was repaired 1 week after the resection, no fatty changes were observed.
However, tendon repair after 2 and 3 weeks of resection, the adipogenic differentia-
tion markers exhibited a significant increase, and as a result, the repair could not
prevent the fatty changes of the muscle. Since the downregulation of Wnt10b, which
is the trigger of adipogenic differentiation, occurs as early as 1 week after the ten-
don resection, it seems desirable to repair a cuff tear as soon as possible to avoid
fatty changes of the muscle. As described here, the fatty changes of the muscle are
caused by the differentiation of myoblast to adipocyte instead of myotube. Therefore,
we believe that it would be more appropriate to use the term “fatty change” rather
than “fatty degeneration” because there is no degeneration or “fatty infiltration”
because the fat tissues do not infiltrate into the muscle from outside but it is pro-
duced in the muscle by turning the myoblasts into adipocytes.

6.2 Extensibility of the Muscle

Whether the torn edge of the rotator cuff tendon can be pulled out and reattached to
the greater tuberosity depends on the extensibility (flexibility) of the muscle. The
flexibility is defined as displacement divided by the force. For example, when a
displacement (elongation) of 3 cm is caused by a tensile force of 2 kg, flexibility is
3/2 = 1.5 cm/kg. The larger the flexibility, the easier to extend. The stiffness is the
opposite of the flexibility, and in this example, it is 2/3 = 0.67 kg/cm. The larger the
stiffness, the harder to extend. We know that the muscle with fatty change is difficult
to extend (large stiffness). First, we looked at the relationship between the stiffness
of the cuff muscles and the MR signal intensity assessed by the MR number
(0–1024) [57]. This study showed that the highest positive correlation (r = 0.740)
was observed between the stiffness and T2-weighted MR number of the infraspina-
tus muscle, followed by that of the supraspinatus muscle (r = 0.582). The normal
muscle looks almost black on T2-weighted image with a low MR number, but as
fatty change occurs, it looks brighter with increased MR number. The brighter the
muscle on T2-weighted MR image, the greater the stiffness and the more difficult to
repair. Next, using a more recent technology known as a shear wave elastography,
we assessed the stiffness of the supraspinatus muscle. As a preliminary study, we
explored the potential feasibility of a shear wave elastography combined with
B-mode ultrasound to measure in vivo stiffness of the supraspinatus muscle in three
7 Pain 51

cadaveric shoulders and three shoulders of healthy volunteers [58]. We confirmed

that by optimizing the probe position in accordance with the muscle fiber orienta-
tion, it was feasible to measure the in vivo stiffness of the muscle. Next, using this
method, we compared the stiffness obtained by the shear wave elastography with
the stiffness obtained by biomechanical testing [59]. This study showed that the
stiffness obtained by shear wave elastography and that obtained by biomechanical
testing were moderately correlated with each other with the correlation coefficient
ranging between 0.57 and 0.63. In vivo measurements of these parameters during
surgery revealed that the highest correlation (r = 0.69) was observed in posterior
deep region of the supraspinatus muscle [60]. Using MRI or shear wave elastogra-
phy, we are able to estimate the extensibility of the cuff muscles to a certain extent
before surgery.

7 Pain

A rotator cuff tear causes pain in some people but not in the other. According to our
epidemiological study, two thirds of those with a full-thickness rotator cuff tear
were without pain and one third were with pain [16]. What makes this difference?
This is a very interesting question. First, we investigated the location of pain. We
divided the patients based on the location of a tear: superior group (supraspinatus
tear), posterosuperior group (supraspinatus and infraspinatus tear), and anterosupe-
rior group (supraspinatus and subscapularis tear) [61]. Interestingly, the most com-
mon site of pain was anterolateral area of the shoulder regardless of a tear location.
This region coincides with the region of pain caused by injecting hypertonic saline
solution into the subacromial bursa [62]. These findings suggests that wherever a
tear is, it induces subacromial bursitis and causes a pain in the anterolateral area of
the shoulder. Kinematic studies showed that there were no differences in shoulder
kinematics between those with and without pain. If subacromial bursitis exists in
painful cuff tear shoulders, inflammation may extend from the bursa to the sur-
rounding tissues including the bone. Using bone scintigraphy, we compared the
uptake of radioisotope (99mTc-HMDP) in patients with symptomatic cuff tears
(symptomatic group), asymptomatic cuff tears (asymptomatic group), and the intact
rotator cuff (control group) [63]. The symptomatic group showed significantly
higher uptake than the other groups, but there was no significant difference between
the asymptomatic group and the control group. An increased uptake of radioisotope
on bone scintigraphy indicates an increase in bone metabolism. From this study, we
cannot tell why the bone metabolism increased in symptomatic patients. Bone
metabolism might be triggered by subacromial bursitis, but it might also be related
to glenohumeral synovitis, cuff tendinitis, cuff enthesopathy, and glenohumeral
osteoarthritis. These relationships need to be clarified in the future studies.
Furthermore, we compared the muscle function between symptomatic and asymp-
tomatic patients using positron emission tomography [64]. This study showed that
during abduction motion, the deltoid was the prime mover in asymptomatic patients,
52 3 Rotator Cuff Tear

but in symptomatic patients, the parascapular muscles such as the trapezius and
levator scapulae were the main movers (Figs. 3.23 and 3.24). These differences in
functioning muscles may be the cause of pain but may also be the result of pain. In
order to avoid pain coming from the subacromial bursa, the parascapular muscles
might have been used more actively instead of the deltoid to achieve arm abduction
with less glenohumeral motion.
Shoulder pain can be classified into rest pain, motion pain, and night pain. At
initial stage, motion pain is common caused by mechanical impingement between
the coracoacromial arch and the rotator cuff tendon. At more advanced stage with
aggravated subacromial bursitis, rest pain appears. Night pain is also very common
among cuff tear patients. They do not have pain during the daytime, but night pain
often awakens the patients during the sleep. Why is the shoulder silent during the

a b

Fig. 3.23 Glucose uptake after abduction exercise (coronal PET images). Asymptomatic shoulder
(a) showed high uptake in the deltoid (red arrows) and little uptake in the trapezius (yellow arrows).
Symptomatic shoulder (b) showed lower uptake in the deltoid (red arrows) and higher uptake in
the trapezius (yellow arrows). (Reprinted from Fig. 3, Shinozaki et al. [64])

a b

Fig. 3.24 Glucose uptake after abduction exercise (axial PET images). Asymptomatic shoulder
(a) showed little uptake in the trapezius (red arrows) and levator scapulae (yellow arrows).
Symptomatic shoulder (b) showed high uptakes in both the trapezius (red arrows) and levator
scapulae (yellow arrows). (Reprinted from Fig. 3, Shinozaki et al. [64])
7 Pain 53

daytime but painful during the night? There are several explanations about the
mechanism of night pain. First, you might have an experience of not being able to
sleep because of toothache, which is not too bad during the daytime. At night, you
may concentrate on your shoulder pain coming from inside your body because there
is little distraction from outside such as light or sound. This is one explanation of
night pain. Second, the shoulder is a hanging joint. The humerus is constantly pulled
downward by the weight of the arm, which is about 3–4 kg, in standing and sitting
positions. However, when you lie down, this downward traction force disappears,
and as a result, the humerus is relatively pushed upward, which may aggravate the
pain at the subacromial bursa. In fact, when cuff tear patients feel night pain in lying
position, sitting up on the bed makes them feel better. Third explanation is the pain
threshold. There is a strong relationship between body temperature and pain thresh-
old. For example, when you feel muscle pain after performing unusual exercise,
taking a hot bath makes you feel much better. Warming up your body temperature
also raises the pain threshold. Before warming up, the stimulation was above the
threshold, and it was painful. After warming up, the same stimulation is now below
the pain threshold, and it is no more painful. Stimulated by this relationship between
body temperature and pain threshold, we measured the circadian rhythm of the
shoulder skin temperature in cuff tear patients, cuff tendinitis patients, and normal
control [65]. In normal shoulders, the skin temperature was constantly around 35 °C
during the daytime, but it gradually declined after 9 p.m., came to the bottom of
34 °C between 1 a.m. and 4 a.m., and then gradually rose (Fig. 3.25). This circadian
rhythm was also confirmed in cuff tear patients. The cuff tendinitis patients, on the
other hand, showed no temperature drop during the night. In addition, we looked at
the relationship between the shoulder skin temperature and bone scintigraphy find-
ings [66]. We found that the cuff tear patients showed high uptake of radioisotope
on the affected side alone, but the skin temperature showed similar circadian rhythm
on both sides. From these results, we assume that the circadian rhythm of shoulder
temperature lowers the pain threshold during the night, which in turn converts a

a b a
a a a a 36
a a a b b b a
36 b a bc b a a a a c b a
Skin temperature (ºC)

b c b d b c c a
c a b b b b b
Skin temperature (ºC)

c d c d d b C
d d C B C
a C a a 35 A A A A A
35 C B
B B B B D
A

34 34

33 33

1 3 5 7 9 11 1 3 5 7 9 1 3 5 7 9 11 1 3 5 7 9
PM AM PM AM
Time Time

Fig. 3.25 Circadian rhythm of shoulder skin temperature. Both the normal shoulder (a) and cuff
tear shoulder (b) showed high shoulder skin temperature during the daytime and low temperature
during the night. There were significant differences between a—A, b—B, c—C, and d—D. (Reprinted
from Figs. 4 and 5, Miyakoshi et al. [65])
54 3 Rotator Cuff Tear

stimulation below the threshold during the daytime to a stimulation above the
threshold during the night. This may be related to an onset of night pain in cuff tear
patients.

8 Diagnosis

Diagnosis of a rotator cuff tear is based on imaging examinations in combination

with clinical findings. Refer to Chap. 2 for Physical Examination. Imaging modali-
ties most commonly used are plain X-rays, ultrasound, and MRI.

8.1 Plain X-rays

A rotator cuff tear is a lesion of the soft tissue, not the bony structures. Thus, plain
X-rays often do not show any abnormality of bony structures. A large critical
shoulder angle on anterior-posterior view indicates a risk of cuff tear. A bony spur
under the acromion also suggests the presence of a cuff tear. In long-standing
cases, the superior migration of the humeral head is a typical X-ray finding with a
decrease in the acromiohumeral interval and disruption of the axillary line
(Shenton’s line of the shoulder) (Fig. 3.26). Further advanced cases may show cuff
tear arthropathy with narrowing of the glenohumeral joint space, osteophyte and
osteosclerosis of the greater tuberosity, acetabularization of the acromion, and
destruction of the humeral head as a result of contact between the acromion and the
humeral head.

8.2 Ultrasonography

Ultrasonography has become a popular imaging method in the field of orthope-

dics these days. First of all, it can be used in the office of a clinic during the
examination without making reservations. We can share the real-time image with
the patient. Image resolutions in recent ultrasound machines are even higher
than MRI. Disadvantages are the ultrasound cannot reach a deep area from the
skin and an area behind a hard tissue like the bone. Ultrasound can be used not
only for making a diagnosis, but also for aspirating a ganglion compressing the
suprascapular nerve and injecting into the subacromial bursa or bicipital ten-
don sheath.
To visualize the supraspinatus tendon, patient’s hand is at the waist and the
shoulder is in extension in the sitting position. In this position, the supraspinatus
8 Diagnosis 55

a b

Fig. 3.26 Plain X-ray AP view. (a) In normal shoulders, the acromiohumeral interval (space
between the yellow lines) is ≥8 mm and the axillary line (red dotted line) is continuous and smooth.
(b) In cuff tear shoulders, the acromiohumeral interval is narrowed and the axillary line is dis-
rupted due to the superior migration of the humeral head. (Reproduced with modification from
Figs. 37–31 in Standard Textbook Series: Orthopedics, 14th ed. [67])

Fig. 3.27 Ultrasonogram.

Between the supraspinatus
tendon (SSP) and the
greater tuberosity (GT),
there is hypoechoic
discontinuity (*), which
indicates a tear of the
tendon. DL deltoid, GT
greater tuberosity, HH
humeral head, SSP
supraspinatus tendon
(Reproduced with
modification from
Figs. 37–14 in Standard
Textbook Series:
Orthopedics, 14th ed. [67])

tendon is located just anterior to the acromion. Putting a probe right in front of the
acromion depicts a clear image of the supraspinatus tendon (Fig. 3.27). The normal
ultrasonogram reveals the deltoid, subdeltoid/subacromial bursa, supraspinatus ten-
don, and the humeral head and greater tuberosity. A tear can easily be detected as
hypoechoic discontinuity.
56 3 Rotator Cuff Tear

8.3 MRI

This imaging modality is most widely and commonly used for a diagnosis of rota-
tor cuff tear. On T2-weighted image, the tendon is in low signal intensity and the
joint fluid is in high signal intensity. A tendon tear is clearly depicted in clear con-
trast because the tear (defect in the tendon) is filled with joint fluid (Fig. 3.28).
High resolution MRI also visualizes muscle atrophy and fatty change. A degenera-
tive tear commonly observed in the elderly is usually a full-thickness tear and is
visualized clearly on MRI. However, a partial-thickness tear, especially a small
articular side tear often seen in baseball players is difficult to be visualized because
of its size and the lesion is not filled with the joint fluid. In such cases, MR arthrog-
raphy using gadolinium as a contrast material is useful because the gadolinium
infiltrates into a small partial-thickness tear and depicts its location and shape very
clearly (Fig. 3.29). As gadolinium appears white on T1-weighted image and black
on T2-weighted image, T1-weighted images are used for diagnosis using MR
arthrography.

Fig. 3.28 MRI (coronal

oblique T2-weighted
image). There is a high
signal intensity area (*)
between the supraspinatus
tendon (SSP) and the
greater tuberosity (GT),
which indicates a
full-thickness tear of the
supraspinatus tendon.
T2-weighted MR image
depicts the joint fluid white
and the tendon black,
making a clear contrast
between them. DL deltoid,
GT greater tuberosity, HH
humeral head, SSP
supraspinatus tendon.
(Reproduced with
modification from
Fig. 37–32 in Standard
Textbook Series:
Orthopedics, 14th ed. [67])
9 Conservative Treatment 57

Fig. 3.29 MR arthrogram

(coronal oblique
T1-weighted image). MRI
is taken after injecting
gadolinium into the
glenohumeral joint.
T1-weighted image depicts
gadolinium as white and
the tendon as black,
making a clear contrast
between them. This image
shows an articular side
partial-thickness tear of the
supraspinatus tendon
(white arrows)

9 Conservative Treatment

9.1 Pain Relief

In our experience, 99% of cuff tear patients come to see us because of pain.
They want us to take away their pain. As I previously mentioned, only one third
of residents with a cuff tear had pain and the rest two thirds were asymptomatic.
Conservative treatment only means to convert a painful cuff tear to an asymp-
tomatic cuff tear without repairing the tear. When we select a conservative treat-
ment, we always need to keep this in mind. Young cuff tear patients are supposed
to use their shoulders for many years to come. We know that 50–60% of cuff
tears become larger with time, and a large tear ≥3 cm is known to have a high
risk of retear with poor outcomes. These facts should be taken into consider-
ation in selecting treatment options.
For pain relief, non-steroidal anti-inflammatory drugs (NSAIDs) are commonly
used orally, in patch, or in ointment. Also, an injection of steroid or hyaluronic acid
into the subacromial bursa or the glenohumeral joint is commonly used. The steroid
injection is limited to three times at the most with an interval of 3–4 weeks. If three
shots of steroid do not work, then we switch to hyaluronic acid and continue the
periodic injection until the pain subsides.
Regarding the ADL, we advise the patients to avoid painful activities. Medications
for neuropathic pain may be effective for night pain. Beach chair position during the
night is recommended to reduce night pain. A shoulder supporter to keep the shoul-
der warm may also be effective for night pain.
How effective is the conservative treatment? Our data show that 82% of 54
patients who underwent conservative treatment were satisfied with the outcomes
58 3 Rotator Cuff Tear

after an average 3.4 years [68]. Comparing those with satisfactory outcomes and
those without, those with well-preserved strength of abduction and external rota-
tion showed better outcomes than those with weak abductors and external rota-
tors. Furthermore, we wanted to know whether there were any factors relevant to
successful outcome of conservative treatment. In our series of 123 shoulders of
118 patients, we compared the clinical findings on their initial visit of those with
successful conservative treatment and those with failed conservative treatment
[69]. There were four factors predictive of successful outcome after conservative
treatment: (1) little or no atrophy of the supraspinatus muscle, (2) negative
impingement sign, (3) preserved range of external rotation, and (4) the integrity
of the intramuscular tendon on MRI. If a patient had at least three of these four
factors, the success rate was 87%. In other words, those with preserved function
of the supraspinatus and infraspinatus are likely to respond well to conservative
treatment.

9.2 Modalities

Modalities refer to the administration of thermal, mechanical, electromagnetic, and

light energies for therapeutic purposes such as to reduce pain, improve circulation,
and achieve muscle relaxation. Applying heat would increase pain threshold and
reduce muscle stiffness so that patient would feel less pain, which is beneficial for
the following range of motion and strengthening exercises.

9.3 Exercises

Cuff tear patients often exhibit scapular winging during arm depression to avoid
pain. Correcting abnormal activities of the parascapular muscles and educating the
intact muscles to function effectively would compensate for a deficit caused by a
tendon tear. A physical therapist performs muscle shaking and stretching to obtain
muscle relaxation before the range of motion exercise. Supine position is the best to
perform these physical therapy procedures because it is easier to achieve shoulder
muscle relaxation than the standing or sitting position. Once a sufficient range of
motion is obtained, muscle strengthening exercises are started. Best position for
each muscle for training using a rubber band (TheraBand) needs to be instructed to
patients: arm elevation in the scapular plane for the supraspinatus, external rotation
in adduction for the infraspinatus, external rotation in abduction for the teres minor,
and internal rotation in slight abduction for the subscapularis. For strengthening the
parascapular muscles, shoulder shrug for scapular elevation, wall pushup for scapu-
lar abduction, and scapular retraction in prone position for scapular adduction are
instructed to patients.
10 Surgical Treatment 59

10 Surgical Treatment

10.1 Basic Research for Tendon Repair

10.1.1 Tendon Healing

Growth factors are known to appear during wound healing. We determined the
expression of various growth factors during the healing process of acute rotator cuff
tears in the rabbit [70]. Basic fibroblast growth factor appeared with its peak on days
7 and 9, insulin-like growth factor 1 appeared with its peak on day 5, platelet-derived
growth factor appeared with a mild expression between days 7 and 14, and trans-
forming growth factor β appeared with constant mild expression throughout the
observation period. It is likely that each of these growth factors plays a role in the
early phase of healing of the supraspinatus tendon in rabbits.
Octacalcium phosphate (OCP) is known to be a precursor of hydroxyapatite, the
primary constituent of bone inorganic components. Professor Osamu Suzuki of
Tohoku University Graduate School of Dentistry has provided various data showing
biodegradability and osteoconductivity of OCP. Prof. Suzuki and I have worked
together in the research of OCP and demonstrated that it promotes differentiation of
osteoblast and formation of osteoclast. Due to this biological activity of OCP, we
wanted to know if OCP had any effect on the healing process of cuff tendon to the
bone. We created an infraspinatus tendon tear in the rabbit model to determine the
effect of OCP on tendon-bone healing [71]. Treatment with OCP increased the col-
lagen fibers and the Sharpey’s fibers at the tendon-to-bone insertion compared to the
control. OCP might be beneficial for the healing of rotator cuff tendon to the bone.

10.1.2 Repair Method

Using cadaveric shoulders, we created a supraspinatus tendon tear, repaired it by (1)

single-row, (2) double-row, or (3) transosseous technique, and compared the contact
area and contact pressure between the tendon and the bone using a pressure-­sensitive
film [72]. Repair technique and pressure distribution pattern of each technique are
shown in Fig. 3.30. The contact area by the double-row technique was the greatest,
whereas the contact pressure by the double-row and single-row techniques were
greater than that by the transosseous technique. These results suggest that the
double-­row technique provides a better environment for tendon healing.
Using a finite element model analysis comparing these three repair techniques,
we found that transosseous technique created less stress concentration in the tendon
compared to the other two techniques [73]. As high stress concentration in the ten-
don is one of the risks of retear, transosseous technique seems to have an advantage
over the other repair techniques using suture anchors.
Furthermore, we investigated the stress distribution in the tendon and in the bone
using a finite element model analysis. We compared the data using three different
60 3 Rotator Cuff Tear

a b

Fuji Film

c d

Fuji Film

f
e

Fuji Film

Fig. 3.30 Repair techniques and contact area/pressure. (a) Single-row technique, (b) pressure
distribution of single-row technique, (c) double-row technique, (d) pressure distribution of double-­
row technique, (e) transosseous technique, (f) pressure distribution of transosseous technique.
(Reprinted from Figs. 1, 2, and 3, Tuoheti et al. [72])

types of suture anchors: a titanium screw-type anchor (TwinFix Ti, 5 mm, Smith &
Nephew), absorbable coil-type anchor (HEALICOIL RG, 4.75 mm, Smith &
Nephew), and non-absorbable coil-type anchor (HEALICOIL PK, 4.5 mm, Smith
& Nephew) [74]. We found that the maximum stress concentration in the bone was
just beneath the cortical bone when a screw-type anchor was used, but it was at the
tip of the screw when a coil-type anchor was used. The maximum pullout strength
was observed with the absorbable coil-type anchor. As the screw-type anchor
10 Surgical Treatment 61

created a stress concentration beneath the cortex at the anchor entrance, preparing
the bony bed at the footprint of the rotator cuff tendon by removing the cortical bone
would weaken the pullout strength of the anchor. Also, the entrance hole itself
seems to weaken the pullout strength of the anchor. These things need to be consid-
ered when using screw-type anchors for cuff repair. From this study, we concluded
that the absorbable coil-type anchor provided the best healing environment for ten-
don healing among these three types of anchors.
Recently, the most common repair technique has become transosseous-­equivalent
(TOE) technique, which simulates the transosseous technique, but using suture
anchors. In this technique, the sutures from the medial anchors are passed through
the tendon, pulled laterally, and fixed to the lateral cortex of the greater tuberosity
with the lateral anchors. The medial sutures may or may not be tied over the tendon
before pulling laterally. Some studies showed that tying the sutures of medial
anchors provided better biomechanical environment, but other studies showed that
there was no difference between them. We measured the strain of various portions
of the infraspinatus tendon repaired with and without medial row suture tying in a
cadaver study [75]. The strain distal to the medial row was significantly reduced
when the tendon tear was repaired using TOE technique regardless of suture tying
compared to the normal tendon. This was probably because the distal tendon was
compressed to the bone by the overlying sutures. On the other hand, the strain of the
tendon at the medial suture level was significantly greater when the medial sutures
were tied compared with those untied. As a result, when the medial sutures were
tied, the strain increased medial to the medial sutures and decreased lateral to the
medial sutures. This great discrepancy in the strain at the medial row level seems to
create a great stress concentration, which has been confirmed by a finite element
model analysis [76]. The clinical data that 70% of retear after TOE technique were
observed at the medial row level also support our stress concentration theory. Based
on these findings, we avoid tying the medial sutures when we use the TOE technique.

10.2 Indication for Surgical Repair

Since the conservative treatment is effective in 70–80% of the patients, it should be

the first-line treatment. However, as previously described, those with poor function
of the supraspinatus and infraspinatus are less likely to respond to conservative
treatment. In those cases, surgery needs to be considered without wasting too much
time for conservative treatment. Also, conservative treatment is to convert a painful
cuff tear to an asymptomatic tear while a tear remains as it is. The risk of tear propa-
gation is 50–60%, which often accompanies pain. With repeated recurrence of pain,
a small tear may become a massive irreparable tear in the long run. According to our
data, retear rate after surgery was 0% for small tears, 19% for medium-sized tears,
41% for large tears, and 75% for massive tears. Considering all these issues, those
who are young or middle aged, who are engaged in sport activities or heavy labor,
do well to undergo surgical treatment while the tear is still small. There is no
62 3 Rotator Cuff Tear

question that treatment option should be chosen by patients. We need to provide

necessary information about the treatment with regard to age, gender, occupation,
sports activities, and lifestyle to help the patients make a final decision and to obtain
informed consent from them.

10.3 Subacromial Decompression

Since Neer introduced a theory that impingement between the acromion and the
rotator cuff tendon was a cause of cuff tear (impingement syndrome) and recom-
mended a resection procedure known as acromioplasty to remove the undersurface
of the anterior 1/3 of the acromion, this open procedure had been routinely used in
the treatment of rotator cuff repair. Then arthroscopic acromioplasty took the place
of open procedure and was called arthroscopic subacromial decompression (ASD).
However, misunderstandings spread that (1) impingement syndrome was equivalent
to shoulder pain and (2) ASD was a treatment for shoulder pain. These misunder-
standings led to an abuse of ASD. For example, the number of ASD in the UK dur-
ing 2009–2010 was 7.5 times more than that during 2000–2001. In order to stop this
abuse, first we need to clearly define shoulder pathologies such as rotator cuff tear
and rotator cuff tendinitis instead of using this ambiguous term “impingement syn-
drome.” In 2018, ISAKOS Shoulder Committee members had a consensus meeting
in Munich. During that meeting, we had an in-depth discussion about ASD. One of
the committee members, Giuseppe Milano, and his colleagues performed a random-
ized clinical trial to clarify the necessity of ASD [77]. They randomly assigned 20
cuff tear patients to ASD + cuff repair and another 20 patients to cuff repair alone.
There were no differences in clinical outcomes and subjective symptoms at 2-year
follow-up. A systematic review in 2012 also concluded that there was no benefit of
adding ASD to cuff repair. However, one randomized clinical trial cited in this
review showed a higher retear rate in those without ASD. Another member of our
committee, Geoffrey Abrams and his colleagues performed a randomized clinical
trial, which demonstrated that the retear rate was 1.9% (1/52) in those with ASD and
9.3% (4/43) in those without (p = 0.11) [78]. The difference was not statistically
significant, but it would surely be significant if the number of patients increases.
Further large-scale high-quality studies are necessary. I prefer to decide whether
ASD is necessary or not during the surgery. If fibrillation or eburnation is found
underneath the acromion, that is a clear sign that there has been mechanical impinge-
ment underneath the acromion. In addition, I perform dynamic examination to con-
firm the existence of impingement between the rotator cuff tendon and the
coracoacromial arch. In such cases, I perform ASD. On the other hand, if the under-
surface of the acromion is smooth without fibrillation, and there is no impingement
during dynamic examination, I do not perform ASD. Based on this indication,
almost all the middle-aged and elderly cases with a degenerative cuff tear require
ASD, whereas young cases with a traumatic tear may not require ASD. During the
committee meeting, I asked the members whether they perform ASD during cuff
10 Surgical Treatment 63

repair. Half of the members mainly from North America said yes, and the rest half
mainly from Europe said no. Even in the committee, opinions split. Again, more
studies are required on the necessity of ASD [79].

10.4 Repair Technique

Although there are various techniques in open cuff repair, the transosseous (TO)
technique had been most commonly used in Japan. When I learnt shoulder surgery
from Dr. Tabata in Iwaki Kyoritsu General Hospital in 1989, arthroscopy was just
introduced in shoulder surgery. In Iwaki, we started to use the Morgan technique,
the very initial arthroscopic procedure for Bankart repair. However, arthroscopic
cuff repair was not established yet in those days. As a result, all the cuff repairs were
performed either by TO technique or fascial patch graft (Fig. 3.31).
Later, the suture anchors were developed and started to be used in cuff repair.
Based on the number of rows of suture anchors, single-row, double-row, and triple-­
row techniques have been reported. The double-row technique is further divided
into standard double-row technique, TOE double-row technique, and diamondback
double-­row technique. So far, TOE technique is most commonly used (Fig. 3.32).
There are many studies comparing different repair techniques. We performed
a systematic review of the literature and reported a meta-analysis comparing
single-­row, double-row, and triple-row techniques [80]. We extracted 580 cases
in 10 studies. There were no significant differences in the clinical outcomes
between the single-row and double-row techniques. There were very limited
number of triple-­row studies in the literature, which made it impossible to

a b

Fig. 3.31 Open repair techniques. (a) A small tear repaired by transosseous technique. (b) A large
tear repaired using an autologous fascia lata graft. A tear defect is completely covered with the
patch graft
64 3 Rotator Cuff Tear

a b

Fig. 3.32 Arthroscopic rotator cuff repair. (a) A medium-sized tear, (b) repaired with TOE double-­
row technique. (Reprinted from Fig. 37–33 in Standard Textbook Series: Orthopedics, 14th
ed. [67])

compare the clinical outcomes of this technique with the others. The retear rate
of double-row repairs (19.3%) was significantly lower than that of single-row
repairs (30.3%), whereas the retear rate of triple-row repairs (23.5%) could not
be compared with the others due to a lack of power. Subgroup analysis compar-
ing the standard versus TOE double-­row repairs showed that the difference in
the retear rate of the standard double-row repairs (20%) and the TOE double-
row repairs (9.5%) did not reach a significant level due to a small number of
TOE repairs. In conclusion, there was no difference in clinical outcomes among
these three types of repairs although the retear rate of double-row repairs was
lower than that of single-row repairs.
There repair techniques can be used only when a tear is reparable. We release
the coracohumeral ligament from the base of the coracoid process, release the
joint capsule to pull out the retracted tendon. If the torn edge of the tendon can-
not be pulled out to the footprint regardless of these releasing procedures, we
must consider other types of treatment such as fascial patch graft, medialization
of the footprint, partial repair, superior capsular reconstruction, small-sized
humeral head replacement, etc. Dr. Tabata showed me many irreparable cases
treated with an autologous fascia lata graft. In this technique, a tear defect is
covered with the fascial patch graft without applying tension to the remnant cuff
tendons. This method is effective not only in reducing pain, but also in transmit-
ting force through the tendon [81]. I am going to share our two studies related to
fascial patch graft. One study was performed at Iwaki General Hospital, show-
ing the surgical outcomes of 38 shoulders of 36 patients with partial-thickness
cuff tears [82]. There were 12 bursal-side tears, 3 intra-tendinous tears, and 23
articular side tears. All the partial-­thickness tears had tear completion to become
full thickness and repaired either by side-to-side suture (13 cases), TO repairs
10 Surgical Treatment 65

(8), fascial patch graft (17), or TO + fascial patch graft (1). Mid-term results at
4.9-year follow-up demonstrated that 82% were satisfactory regardless of tear
location or repair techniques. Another study was performed at Tohoku University
Hospital, showing the surgical outcomes of patch graft using the long head of
the biceps tendon to cover an irreparable tear [83]. After tenodesis of the long
head of the biceps tendon, the intra-articular portion of the tendon was resected
and split into two layers or filleted into three. A tear defect was covered by this
biceps tendon patch graft. In this study, 14 patients were enrolled. The size of
the biceps graft was 3.2 cm × 1.6 cm on average. At an average 28-month fol-
low-up, active flexion improved from 69° to 149°, shoulder function assessment
score of the Japanese Orthopaedic Association (JOA score) improved from 54.7
points to 83.1 points. On postop MRI, 13 out of 14 cases (93%) showed no retear
of the patch graft. The long head of the biceps tendon is available in patients
tenodesis or tenotomy is indicated. The resected tendon may be used as a patch
graft without any additional skin incision and donor-site comorbidities.
The greatest news in the field of shoulder in recent years in Japan is the reverse
shoulder arthroplasty, which was introduced to Japan in 2014. The first generation
was developed in France in 1987, followed by the second generation in 1991, and
the third generation became available in Europe in 1994. It is stunning that the
device lag was 20 years between Europe and Japan. This delay was due to a strict
regulation by the Japanese government. Thanks to the gradual deregulation, reverse
shoulder prostheses finally came into Japan. The design of this prosthesis is that the
head is set on the glenoid side and the socket is on the humeral side. The head-­socket
relationship is reversed compared with the anatomic total shoulder arthroplasty or
the normal anatomy. This is the reason why it is called the “reverse” total shoulder
arthroplasty. The best candidates are the elderly ≥70 years of age with cuff tear
arthropathy, who has difficulty in elevating the arm. The greatest advantage of
reverse total shoulder arthroplasty is that the center of rotation for shoulder abduc-
tion is displaced medially and inferiorly, which makes the deltoid function most
effectively due to an increase in the moment arm and an increase in the deltoid ten-
sion. Because of these advantages, we often see the patients who have been unable
to elevate the arm due to cuff tear arthropathy can elevate the arm after the reverse
total shoulder arthroplasty, which allows this operation to be called “miracle opera-
tion” (Figs. 3.33 and 3.34). Before the introduction of reverse total shoulder arthro-
plasty, only anatomic total shoulder was available in Japan. As a result, the main
indication had been totally limited to a primary osteoarthritis of the shoulder with
intact rotator cuff tendons, which was quite rare in Japan. After the introduction of
reverse total shoulder, the number of total shoulder arthroplasty has increased dra-
matically. Three years after the introduction of reverse total shoulder, 75% of the
entire shoulder arthroplasties were reverse. Of course, the reverse total shoulder is
not versatile. Internal and external rotations become limited after the surgery.
Especially, internal rotation is often required in ADL, and the prosthetic design and
surgical techniques to preserve the functions of internal and external rotators are
under way.
66 3 Rotator Cuff Tear

a b

Fig. 3.33 Reverse total shoulder arthroplasty (before surgery). (a) Arm elevation was almost
impossible. When she tried to elevate the arm, the humeral head migrated superiorly and the center
of rotation was lost, which made her unable to elevate the arm. She looked as if she had paralysis
of her shoulder, but no neurological deficits. This phenomenon is called “pseudoparalysis.” (b) The
humeral head is slightly migrated superiorly with bone sclerosis of the acromion, osteophyte for-
mation of the humeral head, and joint space narrowing

a b

Fig. 3.34 Reverse total shoulder arthroplasty (after surgery). (a) She can fully elevate the arm
3 months after the surgery. (b) Plain X-ray taken 3 months after the surgery. The reverse prosthesis
is in place and the arm approximately 3 cm elongated
10 Surgical Treatment 67

10.5 Research on the Position of Postoperative Immobilization

After the cuff repair surgery, we use an abduction brace to keep the shoulder at 30°
of abduction (Fig. 3.35). Some surgeons prefer not to use an abduction brace and
use a sling instead. We performed a biomechanical study, which showed that the
arm should be kept in 30° of abduction after the cuff repair [84]. Using cadavers, we
created a rotator cuff tear with 2 cm in width and 1.5 cm in length and repaired it to
the greater tuberosity under a 3-kg tensile force with the arm in adduction. Then we
measured the strain of the repaired tendon in 50 different arm positions: elevation at
0°, 15°, 30°, and 45° in the coronal, sagittal, and scapular planes, in five different
rotations (60° IR, 30° IR, neutral, 30° ER, 60° ER). This study revealed that when
the arm was elevated ≥30° in the coronal plane or scapular plane (≥30° of abduc-
tion or scaption), the strain significantly decreased compared to 0° of abduction.
Also, between neutral rotation and 60° of external rotation at 30° of abduction, no
significant increase in the strain was observed. On the other hand, internal rotation
and flexion significantly increased the strain, and thus should be avoided in early
postoperative phase. Based on this research, we use an abduction brace postopera-
tively and the passive range of motion exercise is allowed only in the coronal or
scapular plane elevation above 30° and external rotation up to 60°.
Active arm elevation is started at 4 weeks postoperatively, followed by brace
removal at 6 weeks, TheraBand exercise and car driving at 8 weeks, desk work at
3 months, and sports/heavy labor at 6 months postoperatively.

Fig. 3.35 Abduction

brace. The shoulder is kept
at 30° of abduction
68 3 Rotator Cuff Tear

References

1. Minagawa H, et al. Humeral attachment of the supraspinatus and infraspinatus tendons: an

anatomic study. Arthroscopy. 1998;14(3):302–6.
2. Mochizuki T, et al. Humeral insertion of the supraspinatus and infraspinatus. New anatomi-
cal findings regarding the footprint of the rotator cuff. J Bone Jt Surg Am. 2008;90(5):962–9.
3. Minagawa H, Itoi E, Sato T, Konno N, Hongo M, Sato K. Morphology of the transitional zone of
intramuscular to extramuscular tendons of the rotator cuff. Katakansetsu. 1996;20(1):103–10.
4. Kuechle DK, et al. Shoulder muscle moment arms during horizontal flexion and elevation. J
Shoulder Elb Surg. 1997;6(5):429–39.
5. Kuechle DK, et al. The relevance of the moment arm of shoulder muscles with respect to
axial rotation of the glenohumeral joint in four positions. Clin Biomech (Bristol, Avon).
2000;15(5):322–9.
6. Chang YW, et al. Prediction of muscle force involved in shoulder internal rotation. J Shoulder
Elb Surg. 2000;9(3):188–95.
7. Itoi E, et al. Moment arms of the arm muscles at the glenohumeral joint using the tendon
excursion method. J Musculoskelet Res. 2008;11(2):45–53.
8. Omi R, et al. Function of the shoulder muscles during arm elevation: an assessment using
positron emission tomography. J Anat. 2010;216(5):643–9.
9. Veeger HE, et al. Inertia and muscle contraction parameters for musculoskeletal modelling of
the shoulder mechanism. J Biomech. 1991;24(7):615–29.
10. Townsend H, et al. Electromyographic analysis of the glenohumeral muscles during a baseball
rehabilitation program. Am J Sports Med. 1991;19(3):264–72.
11. Greis PE, et al. Validation of the lift-off test and analysis of subscapularis activity during maxi-
mal internal rotation. Am J Sports Med. 1996;24(5):589–93.
12. Kurokawa D, et al. Muscle activity pattern of the shoulder external rotators differs in adduc-
tion and abduction: an analysis using positron emission tomography. J Shoulder Elb Surg.
2014;23(5):658–64.
13. Matsuzawa G, et al. Muscle activity pattern of the shoulder internal rotators: an assessment
using positron emission tomography. In: Read at 2016 Annual meeting of ORS, Orlando,
March 5–8, 2016, 2016.
14. Itoi E, et al. Isokinetic strength after tears of the supraspinatus tendon. J Bone Jt Surg Br.
1997;79(1):77–82.
15. Minagawa H, Itoi E, Abe H, Fukuta M, Yamamoto N, Seki N, Kikuchi K. Epidemiology of
rotator cuff tears. J Jpn Orthop Assoc. 2006;80(3):S217.
16. Minagawa H, et al. Prevalence of symptomatic and asymptomatic rotator cuff tears in the gen-
eral population: from mass-screening in one village. J Orthop. 2013;10(1):8–12.
17. Yamamoto A, et al. Prevalence and risk factors of a rotator cuff tear in the general population.
J Shoulder Elb Surg. 2010;19(1):116–20.
18. Codman EA. The shoulder: rupture of the supraspinatus tendon and other lesions in and around
the subacromial bursa. Boston: Thomas Todd Company; 1934.
19. Itoi E, et al. Tensile properties of the supraspinatus tendon. J Orthop Res. 1995;13(4):578–84.
20. Sano H, et al. Degeneration at the insertion weakens the tensile strength of the supraspinatus
tendon: a comparative mechanical and histologic study of the bone-tendon complex. J Orthop
Res. 1997;15(5):719–26.
21. Tuoheti Y, et al. Apoptosis in the supraspinatus tendon with stage II subacromial impingement.
J Shoulder Elb Surg. 2005;14(5):535–41.
22. Yuan J, et al. Apoptosis in rotator cuff tendonopathy. J Orthop Res. 2002;20(6):1372–9.
23. Wakabayashi I, et al. Mechanical environment of the supraspinatus tendon: a two-dimensional
finite element model analysis. J Shoulder Elb Surg. 2003;12(6):612–7.
24. Yamamoto N, et al. Contact between the coracoacromial arch and the rotator cuff tendons in
nonpathologic situations: a cadaveric study. J Shoulder Elb Surg. 2010;19(5):681–7.
References 69

25. Muraki T, et al. The effect of scapular position on subacromial contact behavior: a cadaver
study. J Shoulder Elb Surg. 2017;26(5):861–9.
26. Shiota Y, et al. Contact pressure of the coracoacromial arch in shoulders with joint contracture:
a cadaveric study. J Orthop Sci. 2017;22(6):1031–41.
27. Neer CS 2nd. Anterior acromioplasty for the chronic impingement syndrome in the shoulder:
a preliminary report. J Bone Jt Surg Am. 1972;54(1):41–50.
28. Neer CS 2nd, Welsh RP. The shoulder in sports. Orthop Clin North Am. 1977;8(3):583–91.
29. Neer CS 2nd. Impingement lesions. Clin Orthop Relat Res. 1983;173:70–7.
30. Hawkins RJ, Kennedy JC. Impingement syndrome in athletes. Am J Sports Med.
1980;8(3):151–8.
31. Yamamoto N, et al. Impingement mechanisms of the Neer and Hawkins signs. J Shoulder Elb
Surg. 2009;18(6):942–7.
32. Bigliani LU, Morrison DS, April EW. The morphology of the acromion and its relationship to
rotator cuff tears. Orthop Trans. 1986;10:228.
33. Bigliani LU, et al. The relationship of acromial architecture to rotator cuff disease. Clin Sports
Med. 1991;10(4):823–38.
34. Getz JD, et al. Acromial morphology: relation to sex, age, symmetry, and subacromial enthe-
sophytes. Radiology. 1996;199(3):737–42.
35. Gallino M, Santamaria E, Doro T. Anthropometry of the scapula: clinical and surgical consid-
erations. J Shoulder Elb Surg. 1998;7(3):284–91.
36. Oh JH, et al. Classification and clinical significance of acromial spur in rotator cuff tear: heel-­
type spur and rotator cuff tear. Clin Orthop Relat Res. 2010;468(6):1542–50.
37. Lee SB, et al. Contact geometry at the undersurface of the acromion with and without a rotator
cuff tear. Arthroscopy. 2001;17(4):365–72.
38. Zuckerman JD, et al. The influence of coracoacromial arch anatomy on rotator cuff tears. J
Shoulder Elb Surg. 1992;1(1):4–14.
39. Sakoma Y, et al. Coverage of the humeral head by the coracoacromial arch: relationship with
rotator cuff tears. Acta Med Okayama. 2013;67(6):377–83.
40. Nyffeler RW, et al. Association of a large lateral extension of the acromion with rotator cuff
tears. J Bone Jt Surg Am. 2006;88(4):800–5.
41. Moor BK, et al. Is there an association between the individual anatomy of the scapula and the
development of rotator cuff tears or osteoarthritis of the glenohumeral joint? A radiological
study of the critical shoulder angle. Bone Jt J. 2013;95-b(7):935–41.
42. Spiegl UJ, et al. The critical shoulder angle is associated with rotator cuff tears and shoulder
osteoarthritis and is better assessed with radiographs over MRI. Knee Surg Sports Traumatol
Arthrosc. 2016;24(7):2244–51.
43. Gerber C, et al. Supraspinatus tendon load during abduction is dependent on the size of the
critical shoulder angle: a biomechanical analysis. J Orthop Res. 2014;32(7):952–7.
44. Miyazaki AN, et al. Comparison between the acromion index and rotator cuff tears in the
Brazilian and Japanese populations. J Shoulder Elb Surg. 2011;20(7):1082–6.
45. Shinagawa K, et al. Critical shoulder angle in an East Asian population: correlation to
the incidence of rotator cuff tear and glenohumeral osteoarthritis. J Shoulder Elb Surg.
2018;27(9):1602–6.
46. Itoi E, Minagawa H, Konno N, Kobayashi T, Sato T, Sato K, Nishi T. Smoking habits in
patients with rotator cuff tears. Katakansetsu. 1996;20(1):209–12.
47. Yamamoto N, et al. Risk factors for tear progression in symptomatic rotator cuff tears: a pro-
spective study of 174 shoulders. Am J Sports Med. 2017;45(11):2524–31.
48. Baumgarten KM, et al. Cigarette smoking increases the risk for rotator cuff tears. Clin Orthop
Relat Res. 2010;468(6):1534–41.
49. Ichinose R, et al. Alteration of the material properties of the normal supraspinatus tendon by
nicotine treatment in a rat model. Acta Orthop. 2010;81(5):634–8.
50. Hatta T, et al. Nicotine reduced MMP-9 expression in the primary porcine tenocytes exposed
to cyclic stretch. J Orthop Res. 2013;31(4):645–50.
70 3 Rotator Cuff Tear

51. Sano H, Wakabayashi I, Itoi E. Stress distribution in the supraspinatus tendon with partial-­
thickness tears: an analysis using two-dimensional finite element model. J Shoulder Elb Surg.
2006;15(1):100–5.
52. Sano H, et al. Stress distribution within rotator cuff tendons with a crescent-shaped and an
L-shaped tear. Am J Sports Med. 2013;41(10):2262–9.
53. Goutallier D, et al. Fatty muscle degeneration in cuff ruptures. Pre- and postoperative evalua-
tion by CT scan. Clin Orthop Relat Res. 1994;304:78–83.
54. Itoigawa Y, et al. Molecular mechanism of fatty degeneration in rotator cuff muscle with ten-
don rupture. J Orthop Res. 2011;29(6):861–6.
55. Itoigawa Y, et al. Hypoxia induces adipogenic differentiation of myoblastic cell lines. Biochem
Biophys Res Commun. 2010;399(4):721–6.
56. Kuwahara Y, et al. Fatty degeneration and wnt10b expression in the supraspinatus
muscle after surgical repair of torn rotator cuff tendon. J Orthop Surg (Hong Kong).
2019;27(3):2309499019864817.
57. Itoi E, Shimizu T, Minato T, Yamada S, Watanabe W, Sato K. Stiffness of the rotator cuff
muscles: measurement and prediction. J East Jpn Orthop Traumatol. 2000;12(1):61–3.
58. Itoigawa Y, et al. Feasibility assessment of shear wave elastography to rotator cuff muscle.
Clin Anat. 2015;28(2):213–8.
59. Hatta T, et al. Quantifying extensibility of rotator cuff muscle with tendon rupture using shear
wave elastography: a cadaveric study. J Biomech. 2017;61:131–6.
60. Itoigawa Y, et al. Shear wave elastography can predict passive stiffness of supraspina-
tus musculotendinous unit during arthroscopic rotator cuff repair for presurgical planning.
Arthroscopy. 2018;34(8):2276–84.
61. Itoi E, et al. Are pain location and physical examinations useful in locating a tear site of the
rotator cuff? Am J Sports Med. 2006;34(2):256–64.
62. Gerber C, Galantay RV, Hersche O. The pattern of pain produced by irritation of the acromio-
clavicular joint and the subacromial space. J Shoulder Elb Surg. 1998;7(4):352–5.
63. Koike Y, et al. Symptomatic rotator cuff tears show higher radioisotope uptake on bone scintig-
raphy compared with asymptomatic tears. Am J Sports Med. 2013;41(9):2028–33.
64. Shinozaki N, et al. Differences in muscle activities during shoulder elevation in patients with
symptomatic and asymptomatic rotator cuff tears: analysis by positron emission tomography.
J Shoulder Elb Surg. 2014;23(3):e61–7.
65. Miyakoshi N, et al. Skin temperature of the shoulder: circadian rhythms in normal and patho-
logic shoulders. J Shoulder Elb Surg. 1998;7(6):625–8.
66. Koike Y, et al. Shoulder surface temperature and bone scintigraphy findings in patients with
rotator cuff tears. Ups J Med Sci. 2011;116(2):142–7.
67. Itoi E. Chapter 27: The shoulder. In: Itoi E, Yoshikawa H, Tsumura H, Tanaka S, Takagi M,
editors. Standard textbook series: orthopedics. 14th ed. Tokyo: Igaku-Shoin; 2020. p. 426–50.
68. Itoi E, Tabata S. Conservative treatment of rotator cuff tears. Clin Orthop Relat Res.
1992;275:165–73.
69. Tanaka M, et al. Factors related to successful outcome of conservative treatment for rotator
cuff tears. Ups J Med Sci. 2010;115(3):193–200.
70. Kobayashi M, et al. Expression of growth factors in the early phase of supraspinatus tendon
healing in rabbits. J Shoulder Elb Surg. 2006;15(3):371–7.
71. Itoigawa Y, et al. The role of an octacalcium phosphate in the re-formation of infraspinatus
tendon insertion. J Shoulder Elb Surg. 2015;24(7):e175–84.
72. Tuoheti Y, et al. Contact area, contact pressure, and pressure patterns of the tendon-bone inter-
face after rotator cuff repair. Am J Sports Med. 2005;33(12):1869–74.
73. Sano H, et al. Stress distribution in the supraspinatus tendon after tendon repair: suture anchors
versus transosseous suture fixation. Am J Sports Med. 2007;35(4):542–6.
74. Sano H, et al. Comparison of fixation properties between coil-type and screw-type anchors
for rotator cuff repair: a virtual pullout testing using 3-dimensional finite element method. J
Orthop Sci. 2016;21(4):452–7.
References 71

75. Nagamoto H, et al. Transosseous-equivalent repair with and without medial row suture
tying: a cadaveric study of infraspinatus tendon strain measurement. JSES Open Access.
2017;1(2):104–8.
76. Sano H, et al. Tight medial knot tying may increase retearing risk after transosseous equivalent
repair of rotator cuff tendon. Biomed Mater Eng. 2017;28(3):267–77.
77. Milano G, et al. Arthroscopic rotator cuff repair with and without subacromial decompression:
a prospective randomized study. Arthroscopy. 2007;23(1):81–8.
78. Abrams GD, et al. Arthroscopic repair of full-thickness rotator cuff tears with and with-
out acromioplasty: randomized prospective trial with 2-year follow-up. Am J Sports Med.
2014;42(6):1296–303.
79. Itoi E, Yamamoto N. Definition and classification of different forms of impingement. In:
Imhoff AB, Savoie III FH, editors. Rotator cuff across the life span—ISAKOS Consensus
Book. Berlin: Springer; 2019. p. 83–8.
80. Rush LN, Savoie FH 3rd, Itoi E. Double-row rotator cuff repair yields improved tendon struc-
tural integrity, but no difference in clinical outcomes compared with single-row and triple-row
repair: a systematic review. J ISAKOS. 2017;2:260–8.
81. Mura N, et al. Biomechanical effect of patch graft for large rotator cuff tears: a cadaver study.
Clin Orthop Relat Res. 2003;415:131–8.
82. Itoi E, Tabata S. Incomplete rotator cuff tears. Results of operative treatment. Clin Orthop
Relat Res. 1992;284:128–35.
83. Sano H, et al. Tendon patch grafting using the long head of the biceps for irreparable massive
rotator cuff tears. J Orthop Sci. 2010;15(3):310–6.
84. Hatakeyama Y, et al. Effect of arm elevation and rotation on the strain in the repaired rotator
cuff tendon. A cadaveric study. Am J Sports Med. 2001;29(6):788–94.
 Part III
Main Shoulder Diseases—Shoulder
Instability
Chapter 4
Stabilizing Mechanisms

1 Anatomy

The shoulder joint has the greatest range of motion among all the joints in our body.
At the same time, it is the easiest joint to dislocate. Approximately 45% of all the
traumatic dislocations occur at the shoulder joint. The bony structures will tell us
why the shoulder joint has these two extreme features. It is a ball-and-socket joint
with a large hemisphere of the humeral head (Fig. 4.1a) and a small kidney-shaped
socket of the glenoid (Fig. 4.1b). The articular surface of the glenoid is only one
fourth of that of the humeral head, indicating that there is little bony constraint in
the shoulder joint. The axillary view of the shoulder is often compared to a golf ball
sitting on a tree (Fig. 4.1c).
The frontal view of the shoulder shows a large amount of the humeral head
uncovered by the glenoid (Fig. 4.2a). This is quite advantageous for the shoulder to
have a wide range of motion. However, little bony constraint also means that little
stability is provided by the bony structures, making the shoulder very susceptible to
dislocation. In other words, the shoulder gains its wide range of motion at the sacri-
fice of stability. The hip joint, which is also a ball-and-socket joint, is totally differ-
ent (Fig. 4.2b). The femoral head is almost entirely covered by the acetabulum so
that the articular surface of the femoral head is almost invisible. This provides a
great bony stability to the hip joint at the expense of range of motion. Being bipedal,
it is reasonable for us to have stable hip joints for standing, walking, and running
while the shoulder with great mobility allows the hand to come to any point on the
hemisphere within our reach.
The glenohumeral joint is formed by the articulation of the humeral head and the
glenoid, which is also called the “shoulder joint in a narrow sense.” Adding two
more functional joints (the subacromial gliding mechanism known as the second
shoulder joint and the scapulothoracic joint) and two more anatomic joints (the
acromioclavicular joint and the sternoclavicular joint), a complex of these five joints
is called the “shoulder joint in a broad sense” or “shoulder complex.”

© The Author(s), under exclusive license to Springer Nature Singapore Pte 75

Ltd. 2023
E. Itoi, Shoulderology, https://doi.org/10.1007/978-981-99-0345-0_4
76 4 Stabilizing Mechanisms

a b c

Fig. 4.1 Skeletal anatomy of the shoulder. The humeral head is a large hemisphere (a), whereas
its counterpart, the glenoid socket (b) is small and occupies only one fourth of the humeral head
articular surface. The axillary view of the shoulder (c) is often compared to a golf ball sitting
on a tee

a b

Fig. 4.2 Bony structures of the shoulder and hip. (a) The shoulder has a large hemisphere of the
humeral head and the small, shallow glenoid fossa facing with each other. Most of the humeral
head can be seen, indicating that there is little bony stability. (b) The hip has the femoral head
almost entirely covered by the acetabulum. This structure provides a great bony stability to the hip

The glenohumeral joint is entirely surrounded by the joint capsule with three
thickenings in the anterior capsule known as the glenohumeral ligaments
(Fig. 4.3).
The inferior glenohumeral ligament (IGHL) has long been thought as a single
ligament. However, detailed histological studies have revealed that there is a thick-
ening from the anteroinferior (4:00 clockface position of the right shoulder), infe-
rior (6:00), to posteroinferior (8:00) part of the capsule. The anterior margin of this
thick portion of the capsule is called the anterior band (AB) of the IGHL or the
1 Anatomy 77

Fig. 4.3 Glenohumeral

ligaments. Intra-articular
view from posterior
opening shows three
discrete thickenings of the
anterior capsule: the
superior glenohumeral
ligament (SGHL), middle
glenohumeral ligament
(MGHL), and inferior
glenohumeral ligament
(IGHL). LHBT long head
of the biceps tendon, GL
glenoid

Fig. 4.4 Soft tissue components of the shoulder joint. CHL coracohumeral ligament, SSP supra-
spinatus, LHBT long head of the biceps tendon, ISP infraspinatus, TM teres minor, IGHL (PB)
inferior glenohumeral ligament (posterior band), IGHL complex inferior glenohumeral ligament
complex, IGHL (AB) inferior glenohumeral ligament (anterior band), MGHL middle glenohumeral
ligament, SSC subscapularis, SGHL superior glenohumeral ligament, RIC rotator interval capsule,
CP coracoid process (Reproduced with modification from Fig. 4, Itoi E [1])

anteroinferior glenohumeral ligament (AIGHL) and the posterior margin is called

the posterior band (PB) of the IGHL or posteroinferior glenohumeral ligament
(PIGHL) (Fig. 4.4). The thick inferior capsule including the anterior and posterior
bands is called the IGHL complex.
78 4 Stabilizing Mechanisms

One thing we need to be aware of about the IGHL complex is that the ante-
rior band is a discrete structure with localized thickening of the capsule
(Fig. 4.5a, b), whereas the posterior band is the margin between the thick infe-
rior capsule and the thin posterior capsule without any band-like structure.
Under arthroscopic view, the posterior band appears as a band-like structure
with the arm in abduction and internal rotation where the posterior band
becomes tight (Fig. 4.5c), but it disappears completely with the arm in neutral
or external rotation (Fig. 4.5d). The posterior band is more of a functional struc-
ture than an anatomic structure, which can only be detected in a certain position

a b

c d

Fig. 4.5 IGHL complex. The anterior band of IGHL (while arrowheads) can be clearly identified
both in internal rotation (a) and external rotation (b). On the other hand, the posterior band (black
arrowheads) appears as a discrete band in abduction and internal rotation where it becomes tight
(c) but disappears completely in neutral rotation due to a loss of tension (d). Different from the
anterior band, the posterior band is more of a functional structure than an anatomic structure. L
anterior labrum
2 End-Range Stabilizers 79

a b

Fig. 4.6 Attachment of AIGHL-labrum complex to the glenoid. (a) The complex attaches to both
the articular cartilage and the bone. (b) The complex attaches only to the bone. (Reprinted from
Fig. 4, Itoigawa et al. [2])

during arthroscopy. This is the reason why there is no description about the
posterior band in the classic anatomy textbooks. For a long time, the term
“IGHL” had been used to refer to the anterior band of IGHL because there was
no posterior band identified. Even today, the term IGHL is sometimes used to
refer to the anterior band of IGHL.
The IGHL complex attaches to the glenoid rim through the labrum. We examined
the attachment pattern of AIGHL in cadaveric shoulders [2]. We found that the
AIGHL attached to both the articular cartilage and the glenoid neck bone in 88% of
the shoulders and only to the bone in 12% (Fig. 4.6). The cartilage is mechanically
far less strong than the bone. Therefore, although there are two patterns of attach-
ment, the strength of the AIGHL is thought to be provided by the bony attachment,
not the cartilage attachment. This finding provides evidence in favor of repairing the
Bankart lesion to the bony surface of the glenoid neck.

2 End-Range Stabilizers

The shoulder joint has a wide range of motion, but whichever direction you
move your arm, it comes to a stop. That is the end range of motion. Your hand
at various stops with the arm stretched creates a large circle around the shoul-
der. The area within this circle is called the mid range of motion. Different
80 4 Stabilizing Mechanisms

a b

Fig. 4.7 AIGHL. In the hanging arm position (a), the AIGHL (*) is lax. With the arm in abduction
and external rotation (b), the AIGHL becomes tight and prevents further motion of the arm. This
is the end range of motion, where the AIGHL works as a stabilizer

stabilizers are at work at the end range of motion and in the mid range of
motion. First, let’s look at the end-range stabilizers. What prevents further
motion of the arm at the end range of motion? It is the capsule with the gleno-
humeral ligaments. The AIGHL is lax in the hanging arm position (Fig. 4.7a),
but it becomes tight in abduction and external rotation and prevents further
movement of the arm (Fig. 4.7b). This is the end range of motion, where the
AIGHL works as a stabilizer.
The three glenohumeral ligaments that exist in the anterior capsule have exten-
sively been studied, and we know that the anterior stability is provided by the SGHL
in the hanging arm position, by the MGHL at 45° of abduction and external rotation,
and by the AIGHL at 90° of abduction and external rotation. We performed a bio-
mechanical study using cadaveric shoulders to determine the stabilizing function of
the IGHL complex including the anterior band, axillary pouch, and the posterior
band. We measured the strain of each portion of the IGHL complex using strain
gauges in various positions of the arm [3]. This study demonstrated that the anterior
band and the axillary pouch showed the greatest strain with the arm at abduction and
external rotation, whereas the posterior band showed the greatest strain with the arm
in flexion and internal rotation. These results indicate that the anterior band and the
axillary pouch work as anterior stabilizers and the posterior band as a posterior
stabilizer. Because of this stabilizing function of the IGHL complex, it is said that
releasing and firmly fixing the inferior portion of the Bankart lesion at 6:00 is
important when repairing it.
3 Mid-Range Stabilizers 81

3 Mid-Range Stabilizers

In the mid range of motion, stabilizers other than the glenohumeral ligaments should
be at work since all these ligaments are lax. There are two groups of mid-range
stabilizers. The one group works when the muscles are silent (at rest) and the other
works when the muscles are active (during motion).

3.1 At Rest

In the sitting or standing position with the arm hanging freely, the humeral head
looks as if it is sticking to the vertical cliff of the glenoid (Fig. 4.8). I used to wonder
why the humeral head would not fall.
The first biomechanical study that I did at Mayo Clinic was to clarify this mecha-
nism using cadaveric shoulders [4]. The medial border of the scapula was fixed to a
jig and the arm resected in the middle was hanging freely. Under this condition, the
humeral head did not subluxate inferiorly (Fig. 4.9a). After removal of the skin and
subcutaneous tissues, the head still remained in the glenoid socket (Fig. 4.9b). Next,
we removed the deltoid, but nothing happened (Fig. 4.9c). The glenohumeral joint
was now covered only by the capsule and the rotator cuff.
A needle was then inserted into the glenohumeral joint through the posterior
capsule, but nothing happened (Fig. 4.10a). Here, a three-way cock attached to the
needle was opened, and suddenly, the air was sucked into the joint with a hissing
noise, followed by an inferior dislocation of the humeral head (Fig. 4.10b) [5]. As

Fig. 4.8 The humeral head

and the glenoid socket. The
humeral head is sticking to
the vertical cliff of the
glenoid socket. It is
inexplicable why the
humeral head does not fall
82 4 Stabilizing Mechanisms

a b c

Fig. 4.9 Mid-range stabilizers. The humeral head remained in the glenoid socket with everything
intact (a), after removal of the skin and subcutaneous tissues (b), and after removal of the deltoid
(c). It never dislocated inferiorly under these conditions

a b

Fig. 4.10 Experiment of intra-articular pressure. (a) A needle was inserted into the joint through
the posterior capsule while the rotator cuff and the capsule were kept intact. (b) A three-way cock
was opened and suddenly, the air was sucked into the joint and the humeral head dislocated inferi-
orly. (Reproduced with modification from Fig. 2, Itoi E et al. [5])

the shoulder joint has a closed space entirely covered by the capsule, a downward
pull of the humerus with the weight of the arm creates a negative pressure inside the
joint, which keeps the humeral head in the glenoid socket. This pressure is called the
intra-articular pressure. In live shoulders, the intra-articular pressure is approxi-
mately −40 cmH2O with the arm hanging freely. This negative intra-­articular pres-
sure is the main mid-range stabilizer at resting position.
Intra-articular pressure is determined by three factors: load, joint volume, and
material properties of the capsule [6]. The first factor is the load. The larger the load
applied to the humerus, the greater the negative value of the pressure [5]. The nega-
tive value increases in a linear fashion with an increase of the load (Fig. 4.11).
3 Mid-Range Stabilizers 83

–100

#3
–200
Pressure (cmH2O)

–300 #4

–400
#1

–500

#2
–600
0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
Load (kg)

Fig. 4.11 Intra-articular pressure and load. The negative values of four specimens (#1–#4)
increased linearly with an increase of the load although the inclination was different in each speci-
men. (Reprinted from Fig. 4, Itoi et al. [5])

The second factor is the joint volume [7]. As the joint volume was decreased
stepwise by anterior and posterior capsular plication, the negative value of the intra-­
articular pressure increased with a decrease in the joint volume (Fig. 4.12).
The third factor is the material properties of the capsule [6]. With the same mate-
rial properties, the thicker the capsule, the easier to create the negative value of the
intra-articular pressure. Similarly, with the same thickness of the capsule, the greater
the elastic modulus, the easier to create the negative pressure value. A
2D-mathematical model shows that the intra-articular pressure P is defined as:

P = Et ( 2θ / h − 1) = Ct / R,

where E is the elastic modulus; t is the capsular thickness; C is the capsular ten-
sion; R is the radius of curvature; h is the joint space; and θ is the opening angle
(Fig. 4.13) [8].
From this equation, it is clear that the intra-articular pressure is proportional to
the elastic modulus of the capsule. It may be difficult to understand, but just imagine
that you cover the tip of the syringe with the rubber membrane and try to pull the
plunger. You can probably understand in an intuitive way that it would be easier to
pull the plunger if the rubber membrane is thin or less elastic (softer), whereas it
would be much harder if it is thick or more elastic (hard). As shown in this example,
84 4 Stabilizing Mechanisms

Load (kg)
0 0.5 1.0
0

–100

–200

–300 Volume
Pressure
(cmH2O)
37 ml
–400

26 ml
–500

15 ml
–600

–700

Fig. 4.12 Intra-articular pressure and joint volume. With a decrease of the joint volume from 37
to 15 mL, the intra-articular pressure created by a 1.0-kg load changed from −366 to −548 cmH2O,
creating 50% greater negative value of the intra-articular pressure. As shown in this experiment,
the smaller the joint volume, the easier to create the negative value of the intra-articular pressure.
(Reproduced with modification from Fig. 3, Yamamoto N, et al. [7])

(BONE: FIXED)

(CAPSULE)
Vº ,P †

h
q
C C
R

(BONE: MOVING)

Fig. 4.13 2D-mathematical model of intra-articular pressure. Intra-articular pressure P is defined

as P = Et(2θ/h − 1) = Ct/R. V0 joint volume, P intra-articular pressure, E elastic modulus of the
capsule, t capsular thickness, C capsular tension, R radius of curvature, h joint space, θ opening
angle (Reprinted from Fig. 4, Itoi et al. [8])
3 Mid-Range Stabilizers 85

a b

Fig. 4.14 Inferior capsular shift procedure. The anterior capsule is opened in T shape (a), and the
inferior capsule is shifted up and the superior capsule is shifted down (b). This procedure reduces
the joint volume and doubles the anterior capsular thickness, both of which are advantageous in
stabilizing the shoulder in the mid range of motion. (Reprinted from Fig. 4.40, Neer [9])

the intra-articular pressure is well maintained and the shoulder is stable when the
capsule is thick and more elastic, and the joint volume is small.
In shoulders with multidirectional instability, the joint volume is enlarged with
the thin capsule, which makes it more difficult to maintain the intra-articular pres-
sure, leading to an inferior subluxation or dislocation of the shoulder in the hanging
arm position. The capsular shift procedure (Fig. 4.14) is known to be effective
because it creates the steadfast intra-articular pressure by reducing the joint volume
(second factor) and overlapping the anterior capsule (third factor).
The negative value of the intra-articular pressure is created by the weight of the
arm. However, when the muscle contraction occurs to move the shoulder, the humeral
head is compressed against the glenoid socket by the muscle force. This compression
force turns the intra-articular pressure from negative to positive. The pressure becomes
especially high at the end range of motion such as full elevation, maximum external
rotation, and maximum internal rotation [10]. During passive motion, the pressure
also becomes positive at the end range of motion because of the tight capsule.

3.2 During Motion

The shoulder muscles and parascapular muscles are firing during arm motion, which
create a resultant force to push the humeral head against the glenoid. The glenoid
has a deeper concavity in the superior-inferior direction and a shallower concavity
in the anterior-posterior direction (Fig. 4.15).
86 4 Stabilizing Mechanisms

a b

Fig. 4.15 Glenoid concavity. (a) The anterior view of the glenoid shows that the glenoid has a
deeper concavity in the superior-inferior direction. (b) The axillary view shows that the concavity
is much shallower in the anterior-posterior direction. (Reproduced with modification from Fig. 4,
Itoi [6])

If the humeral head is compressed against the glenoid socket (compression

force), a certain amount of force (dislocation force) proportional to the compression
force is needed to move the humeral head out of the glenoid socket.

Dislocation force = Compression force × k.

This proportional constant k is solely determined by the depth of the glenoid and
is called the stability ratio (Fig. 4.16).
The stability ratio is determined by the depth of the glenoid. As described previ-
ously, the glenoid has a deeper concavity in the superior-inferior direction. In addi-
tion, the labrum is attached to the entire glenoid rim. This labrum increases the
depth of the glenoid. With the labrum intact, the stability ratio is 55–60% in the
superior-inferior direction and 30–35% in the anterior-posterior direction. Removal
of the labrum decreases the stability ratio by approximately 20% in each direction.
In cases of shoulder dislocation with a glenoid bone loss, the mid-range stability is
reduced due to a decrease in the glenoid rim height. The stability provided by the
compression force to the humeral head against the glenoid concavity is called the
concavity-compression effect. The deeper the glenoid socket and the greater the
compression force, the greater the stability [11].
Although the stability ratio is determined by the depth of the glenoid, the depth
itself changes with the inclination of the scapula. We looked at the relationship
between the scapular inclination and the stability ratio [12]. When the humeral head
dislocates inferiorly, it is the height of the inferior glenoid rim that the humeral head
3 Mid-Range Stabilizers 87

Fig. 4.16 Stability ratio.

When the humeral head is
compressed against the
glenoid (compression
force), a certain amount of
force (dislocation force)
proportional to the
compression force is
needed to dislocation the
humeral head. The
proportional constant k is
called the stability ratio

140
130
120
110
100
90
Stability ratio, %

80
70
60
50
40
h
30
20 h
10
0
–20° –15° –10° –5° 0° 5° 10° 15° 20°
downward rotation upward rotation

Fig. 4.17 Scapular inclination and stability ratio. With the scapula in neutral position (vertical
position), the average inferior stability ratio is 55%. With the scapula rotated upward, the height of
the inferior glenoid rim (h) increases, resulting in an increase in the inferior stability ratio. With the
scapula rotated downward, the height of the inferior glenoid rim (h) decreases with a decrease in
the inferior stability ratio. (Reproduced with modification from Fig. 3, Kikuchi et al. [12])

has to override. The inferior stability ratio with the scapula in neutral position is
55% on average (Fig. 4.17). When the scapula is rotated upward, the glenoid is fac-
ing upward and the height of the inferior glenoid rim (h) increases with an increase
in the inferior stability ratio up to more than 100%. On the other hand, when the
scapula is rotated downward, the glenoid is facing downward and the height of the
inferior glenoid rim (h) decreases with a decrease in the inferior stability ratio down
to 20%. The superior stability ratio is the opposite: it increases with downward rota-
tion of the scapula and decreases with upward rotation of the scapula. Similarly, the
anterior stability ratio increases with external rotation of the scapula and decreases
with internal rotation of the scapula, while the posterior stability ratio is the
88 4 Stabilizing Mechanisms

opposite. The shoulder stability is intimately related to the position of the scapula.
In patients with multidirectional instability, the scapula is often rotated downward,
which decreases the inferior stability ratio. Strengthening the parascapular muscles
is effective in correcting the downward rotation of the scapula, which improves the
inferior stability.

References

1. Itoi E. Functional anatomy and pathophysiology of the shoulder. In: Matsui N, Yoneda M, edi-
tors. Shoulder arthroscopy. Tokyo: Medical View; 1999. p. 17–24.
2. Itoigawa Y, et al. Attachment of the anteroinferior glenohumeral ligament-labrum complex to
the glenoid: an anatomic study. Arthroscopy. 2012;28(11):1628–33.
3. Urayama M, et al. Function of the 3 portions of the inferior glenohumeral ligament: a cadav-
eric study. J Shoulder Elb Surg. 2001;10(6):589–94.
4. Motzkin NE, et al. Contribution of passive bulk tissues and deltoid to static inferior glenohu-
meral stability. J Shoulder Elb Surg. 1994;3(5):313–9.
5. Itoi E, et al. Intraarticular pressure of the shoulder. Arthroscopy. 1993;9(4):406–13.
6. Itoi E. Pathophysiology and treatment of atraumatic instability of the shoulder. J Orthop Sci.
2004;9(2):208–13.
7. Yamamoto N, et al. The effect of the inferior capsular shift on shoulder intra-articular pressure:
a cadaveric study. Am J Sports Med. 2006;34(6):939–44.
8. Itoi E, et al. Biomechanics of loose shoulder. MB Orthop. 2002;15(5):11–6.
9. Neer CS 2nd. Dislocations. In: Neer CS, editor. Shoulder reconstruction. 2nd ed. Philadelphia:
W.B. Saunders; 1990. p. 273–362.
10. Shibata Y, Takeshita M, Takagishi N. Intra-articular pressure of the gleno-humeral joint during
movement. Katakansetsu. 1987;11(1):4–8.
11. Itoi E, Morrey BF, An KN. Biomechanics of the shoulder. In: Rockwood CAJ, et al., editors.
The shoulder. 4th ed. Philadelphia: W.B. Saunders; 2009. p. 213–65.
12. Kikuchi K, et al. Scapular inclination and glenohumeral joint stability: a cadaveric study. J
Orthop Sci. 2008;13(1):72–7.
Chapter 5
Traumatic Dislocation

1 Epidemiology

Shoulder dislocation is classified into four types based on the direction of disloca-
tion: anterior, posterior, inferior (luxatio erecta), and superior dislocation. Inferior
dislocation is so rare that it is reported as a rare case. We published a case report of
bilateral luxatio erecta [1]. Superior dislocation does not occur unless the acromion
is fractured and displaced. Thus, shoulder dislocation usually refers to either ante-
rior or posterior dislocation. The anterior dislocation accounts for 98% of shoulder
dislocation, far more common than the posterior one [2]. Since my personal experi-
ence is very limited regarding the posterior dislocation, I would like to focus on
anterior dislocation in this chapter.
According to the previous reports from North America and Europe, the incidence of
traumatic anterior dislocation of the shoulder shows two peaks, one in the young and
the other in the middle to the aged. A report in 1950s showed two peaks in the teens
and in the 50s, [2] whereas reports in 1980s and 1990s showed that the peaks were in
the 20s and 60s [3, 4]. We performed a randomized clinical trial of first-time disloca-
tors in 2007. There were 198 patients enrolled in this study. The age distribution shows
a large peak in the 20s and a small peak in the 70s (Fig. 5.1) [5]. As the population gets
older with time, so the peaks might have shifted toward the elderly. This trend is espe-
cially true in Japan, where we are already in the super-aged society. The interesting
thing is that 87% of patients in the first peak were male and 76% in the second peak
were female. Accordingly, the onsets were different: many young male patients in the
first peak were injured during collision/contact sports such as rugby and judo, whereas
many in the second peak were elderly females, suffered a dislocation by a subtle injury
such as falling on the outstretched arm from the standing position.
Simonet et al. reported that the incidence of first-time shoulder dislocation was
8.2/100,000 person-years and that of all shoulder dislocations was 11.2/100,000
person-years in Olmsted County, Minnesota, USA [6]. Based on these data, they
estimated that 19,000 people suffered first-time shoulder dislocations and 25,000

© The Author(s), under exclusive license to Springer Nature Singapore Pte 89

Ltd. 2023
E. Itoi, Shoulderology, https://doi.org/10.1007/978-981-99-0345-0_5
90 5 Traumatic Dislocation

90

Male
80 Female

70

60
Number of cases

50

40

30

20

10

0
< 20 20 30 40 50 60 70 80 <
Age

Fig. 5.1 Age distribution of initial dislocators (n = 198). There is a biphasic pattern with a large
peak in the 20s and a small peak in the 70s. Those in the large peak are mainly male and those in
the small peak are mainly female. (Produced from the data, Itoi et al. [5])

people suffered all shoulder dislocations every year in the USA. Using the database
of all injuries presenting to emergency departments in the USA, Zacchilli and
Owens estimated the incidence of all shoulder dislocations to be 23.9/100,000
person-­years [7]. Trojan et al. focused on the incidence of shoulder dislocation
among collision athletes. Using the National Collegiate Athletic Association Injury
Surveillance Program, they calculated the incidence of all shoulder instability inju-
ries (both dislocation and subluxation) to be 31.3/100,000 athlete-exposures [8].

2 Pathophysiology

2.1 Arm Position at the Time of Dislocation and Location

of the Lesion

An anterior shoulder dislocation usually occurs with a posterior directed force on

the arm in abduction and external rotation, similar to the position of the arm when
cocking back to throw a ball. This is called a “typical position” of the arm when a
2 Pathophysiology 91

dislocation occurs. However, a recent video analysis revealed that among profes-
sional rugby players, 80% of shoulder dislocations occurred when the arm was
forcefully abducted at 90° or more, which was close to the typical arm position, but
the rest 20% occurred by a direct hit when falling on the shoulder while holding a
rugby ball, which was close to adducted position of the arm [9]. In baseball, on the
other hand, head sliding with both arms in full elevation is the most common mech-
anism of dislocation. Thus, the typical position of the arm when a shoulder disloca-
tion occurs may vary according to the sports.
When an anterior dislocation occurs, the anterior band and the axillary pouch of
the IGHL complex (3:00 to 5:00 on the clockface of right shoulder) are often
detached from the glenoid known as a Bankart lesion, which is found in 96–97% of
the cases after a first-time dislocation [10, 11]. Less commonly, detachment may
occur on the humeral side known as HAGL lesion (Humeral Avulsion of the
Glenohumeral Ligament), or in the mid-substance of IGHL complex, or a combi-
nation of the Bankart and HAGL lesions known as the floating IGHL. As previ-
ously described, the SGHL functions in adduction, the MGHL at 45° of abduction,
and the AIGHL at 90° of abduction. When a dislocation occurs with the arm in less
than 90° of abduction, the lesion tends to extend superiorly. Similarly, when a dis-
location occurs with the arm abducted more than 90°, the lesion tends to extend
inferiorly. Based on this concept, I prefer to perform the anterior apprehension test
at three different angles of abduction to estimate the location and extent of the
lesion (Also refer to Chap. 2 Physical Examination).
At the end range of motion, the AIGHL becomes tight and stabilizes the
humeral head. If a force exceeds the structural property of the IGHL complex, a
failure of the IGHL complex occurs either at the glenoid attachment and/or the
humeral attachment or in the mid-substance. In the laboratory, we can measure the
tensile load of the IGHL complex using a universal testing machine. The load-
displacement curve shows a linear relationship between the load and the displace-
ment. During this linear phase, the IGHL complex returns to the original length
once the tensile load is removed (Fig. 5.2). Thus, this deformation is elastic and
reversible. However, with an increase of the tensile load, the relationship is not
linear and the deformation becomes plastic and irreversible. The load when the
deformation becomes plastic is called the yield load. Further loading of the speci-
men will result in a failure of the specimen (ultimate load). Therefore, a plastic
deformation is always present when the IGHL complex is torn after the first-time
dislocation.
The same thing happens when a shoulder dislocation occurs. By the time a
Bankart lesion is created by over tensioning the IGHL complex, it already has irre-
versible deformation. How much deformation does it have? We measured the length
of the capsule at the level of AIGHL, axillary pouch, and PIGHL of bilateral shoul-
ders of patients with unilateral recurrent shoulder dislocations using MR arthrogra-
phy [12]. This study showed that AIGHL was elongated by 6.5 mm (19% of the
contralateral shoulder), axillary pouch by 2.6 mm (12%), and PIGHL by 4.3 mm
(12%). Elongation of the AIGHL and axillary pouch was significantly greater than
the contralateral shoulder, but the PIGHL did not show a significant difference.
92 5 Traumatic Dislocation

Fig. 5.2 Load-­ Load (N)

displacement curve. There 600
is a linear relationship
between the load and
displacement. In this linear 500
portion, the deformation is Ultimate load
elastic and reversible. With 400
an increase of the load, the Yield load
relationship becomes
nonlinear and eventually a 300
failure occurs at the
ultimate load. The load 200
between the linear and
nonlinear portions is called
100
the yield load. In this
nonlinear portion, the
deformation is plastic and 0
irreversible 0 5 10 15 20
Displacement (mm)

Based on this study, when we perform a Bankart repair, we do well to put the repair
stitches about 10 mm from the torn edge so that the capsule will be imbricated to
correct the plastic deformation.

2.2 Bone Loss

The shoulder often has a glenoid bone loss and a compression fracture of the
humeral head known as a Hill–Sachs lesion at the time of dislocation. In our series
of recurrent shoulder dislocations, 86% of the patients had a glenoid bone loss and
94% had a Hill–Sachs lesion. These two lesions are often concurrent and called a
bipolar lesion. The prevalence of bipolar lesion was 81% in our series [13]. Thus,
four out of five patients with recurrent anterior shoulder dislocation have both the
glenoid bone loss and Hill–Sachs lesion together.

2.2.1 Glenoid Bone Loss

Necessity to Quantify a Bone Loss

It used to be said that a bone loss greater than one third of the glenoid should be
treated surgically. However, no one can tell precisely how large the one third of the
glenoid is. We cannot determine either a defect size or the surgical indication if we
use one third or one fourth to describe a defect size because it is just a visual estima-
tion, not the precise measurement of defect size. First, we need to quantify a defect
size using a method which is reliable and available for anyone. Area is not suitable
2 Pathophysiology 93

because we cannot measure the area unless area calculation software is available. It
is too complicated. On the other hand, length is easy to measure, and anyone can
measure it using a ruler. Thus, we decided to use the length measurement to quan-
tify the defect. Next, we need to determine which length to measure. As the margin
of bony defect of the glenoid is usually straight, the easiest length we can measure
is the distance between the margin line of the defect and a line parallel to the defect
line and tangential to the glenoid rim. The measured value needs to be normalized
by the diameter of outer fitting circle (glenoid length) or that of inner fitting circle
(glenoid width).

Bone Loss and Instability

In our first study, we created a bony defect stepwise at the anteroinferior portion of
the glenoid and normalized the width of each defect using the glenoid length (length
of the long axis of the glenoid = diameter of the outer fitting circle) (Fig. 5.3).
This study showed that shoulder stability could not be restored even after the
Bankart repair if a glenoid defect was equal to or greater than 21% of the glenoid
length (osteotomy line #2 in Fig. 5.3).
We further investigated the effect of bone graft to this critical bone loss to bony
stability of the shoulder [15]. We confirmed that if a grafted bone created the same
height of the original glenoid, the bony stability was fully restored (Fig. 5.4).
We know that an anterior shoulder dislocation occurs in the anteroinferior direc-
tion. However, our later study showed that the scapula was anteriorly tilted average
39° relative to the trunk [16]. This means that the humeral head dislocates in the

Fig. 5.3 Bony defects of osteotomy lines

anteroinferior glenoid. #4
Stepwise osteotomy was A
performed along the #3
osteotomy lines #1–#4. W1 #2
Line AB is the diameter of
W1 #1
the outer fitting circle,
which corresponds to the W1
W1
glenoid length. The widths
of the defect W1 and W2 C W2
were divided by the
glenoid length AB to
normalize the data.
W1 = 12.5% of AB,
W2 = 9% of AB, C: center
of the outer fitting circle. B
(Reprinted from Fig. 2, Itoi
et al. [14])
94 5 Traumatic Dislocation

a b

Fig. 5.4 Restoration of bony stability by bone graft. (a) Creation of the anteroinferior glenoid
defect with the width of 21% of the glenoid length (dotted line). (b) A bone graft secured to the
defect with use of two screws. (Reproduced with modification from Fig. 1, Montgomery et al. [15])

Fig. 5.5 Location of the

dislocated humeral head.
The humeral head
dislocates in the
anteroinferior direction
relative to the coordinate
axes of the trunk (white
arrows). As the scapula is
tilted average 39°
anteriorly, the direction of
dislocation is almost
anterior relative to the
coordinate axes of the
scapula (yellow arrows)

anteroinferior direction relative to the trunk, but it dislocates almost in the anterior
direction relative to the scapula (Fig. 5.5) [17].
In our first study, we created anteroinferior glenoid defects. Now that we know
the defects are located almost anterior to the glenoid, we need to look at the effect
of anterior glenoid defects on shoulder stability. In our next study, we created an
anterior glenoid defect stepwise with a 2-mm increment of the defect width. Under
2 Pathophysiology 95

Fig. 5.6 Anterior glenoid Y

defects. In our previous
study, we created
anteroinferior glenoid
defects (gray area). In the
present study, we created
an anterior glenoid defect
stepwise with a 2-mm
increment of the defect
width up to 8 mm (dotted
lines). (Reprinted from
Fig. 2, Yamamoto et al.
[18]) X
C

6-mm 2-mm

8-mm 4-mm

these conditions, we assessed the shoulder stability using the stability ratio
(Fig. 5.6) [18].
This study showed that with an increase of the glenoid defect, the stability gradu-
ally decreased. The stability ratio with a glenoid defect of 6 mm or greater decreased
significantly compared to that without a defect. A 6-mm defect corresponded to
20% of the glenoid length or 26% of the glenoid width. We just looked at the rela-
tionship between bony stability and anterior glenoid defect. In the next study, we
did exactly the same as the first study except the location of the glenoid defect to see
the effect of anterior glenoid defect on shoulder stability after the Bankart repair
[19]. Similar to the previous study, this study showed that a glenoid defect of 6 mm
or greater significantly decreased the shoulder stability even after the Bankart repair
and that the stability was fully restored after the bone graft. This study was per-
formed using cadaveric shoulders at Mayo Clinic: the size of the glenoid obtained
in the USA was greater than that obtained in Japan. As a result, a 6-mm defect was
equivalent to 19% of the glenoid length or 25% of the glenoid width.
These serial studies on quantifying a defect and defining the critical size trig-
gered a huge amount of research and clinical studies in this field. Burkhart et al.
called the glenoid with a large defect an “inverted-pear” glenoid because it
looked like an inverted pear [20]. The recurrence rate after arthroscopic Bankart
repair was 67% in those with an inverted-pear glenoid versus 4% in those with-
out [20]. Later, they clarified that the defect size of an inverted-pear glenoid was
96 5 Traumatic Dislocation

between 25 and 45% of the glenoid width [21]. Based on these research and
clinical studies, it has been widely accepted that the critical size of the glenoid
defect is 25% of the glenoid width. A defect equal to or greater than the critical
size needs to be treated by bone graft such as coracoid transfer or iliac crest
bone graft.

Mechanism of Instability by a Bone Loss

What is the mechanism of instability caused by a glenoid bone loss? In the normal
shoulder, the humeral head is stabilized in the glenoid socket and no anterior trans-
lation occurs at the end-range of motion because the joint capsule is tight at the
end-range of motion such as abduction and external rotation (Fig. 5.7a). Even with
a large glenoid defect, the humeral head is kept in place by the tense capsule if the
capsule is repaired (Fig. 5.7b). In the mid-range of motion, however, the capsule is
lax and does not keep the humeral head in the glenoid socket (Fig. 5.7c). The stabil-
ity in this position solely depends on the depth of the glenoid socket. As the glenoid
socket is shallow without the anterior rim due to the defect, the humeral head cannot
stay in the socket and it easily slides out of the socket (Fig. 5.7d). Therefore, the
glenoid defect is related to the mid-range instability.

a b c d

Fig. 5.7 Mechanism of instability caused by a glenoid bone loss. In the normal shoulder, the
humeral head is stabilized in the glenoid socket by the tight capsule at the end-range of motion (a).
Even with a large glenoid defect, the humeral head is stabilized by the tight repaired capsule at the
end-range of motion (b). However, in the mid-range, the capsule becomes lax and cannot hold the
humeral head in the glenoid (c). In this position, the stability solely depends on the glenoid depth.
Due to the presence of a defect, the glenoid is shallow and the humeral head easily comes out of
the glenoid (d). (Reprinted from Figs. 2 and 3, Itoi [22])
2 Pathophysiology 97

Critical Bone Loss, Subcritical Bone Loss

As shown above, the critical size of the glenoid defect is 25% of the glenoid width.
Then the question is what about 24%. Is it safe? What about 23%? Shaha, Tokish
and colleagues of Tripler Army Medical Center, Hawaii, USA, are the ones who put
a new light on this issue. They evaluated the effect of glenoid bone loss on func-
tional outcomes and failure rates after primary arthroscopic stabilization [23].
Regarding the failure rate, it was 27.8% in those who had a bone loss ≥20%, which
was significantly higher than 7.3% in those with a bone loss <20%. Regarding the
functional outcomes, the Western Ontario Shoulder Instability Index (WOSI) score
was significantly worse in those with a bone loss between 13.5 and 20% than those
with a bone loss <13.5% even though there was no significant difference in the
recurrence rate between them. They called 20% as the critical bone loss which was
related to high recurrence rate, and 13.5–20% as the “subcritical bone loss” where
WOSI score was unsatisfactory, even though no difference in the recurrence rate.
The subcritical bone loss is a transitional zone or gray zone between the critical
bone loss and the safe bone loss. We evaluated our patients whether the subcritical
bone loss was also found in our patients of general population [24]. This study
revealed that they also had the subcritical bone loss, which ranged between 17 and
25%. Our values were a little greater than those reported by Shaha et al. This was
probably because their patient cohort was military patients with high demand,
whereas our patients were general population. A difference in the risk of the cohorts
may explain this difference in subcritical defect size. The subcritical defect size
may be related to risk factors such as age, gender, activity level, sports, and
occupation.

2.2.2 Hill–Sachs Lesion

Hill–Sachs lesion is a compression fracture of the posterolateral aspect of the

humeral head. When the humeral head dislocates anteriorly, the posterior humeral
head and the anterior glenoid rim are contacting each other. Due to a strong force
created by the shoulder muscles and soft tissues, the anterior rim of the glenoid cre-
ates a compression fracture on the humeral head. This lesion was first reported by a
French surgeon, Joseph-François Malgaigne. In his book “Fracture and Dislocation”
published in 1855, a longitudinal compression fracture of the posterolateral aspect
of the humeral head was clearly depicted, which was observed in a person who had
had recurrent dislocations (Fig. 5.8) [25]. Since he was the first reporter, this lesion
is called “Malgaigne lesion” (lésion de Malgaigne) in France.
Since his first report, there were several case reports of this lesion. In 1940, two
American radiologists, Harold A. Hill and Maurice D. Sachs, published a compre-
hensive article in Radiology, reviewing the previous case reports combined with 119
cases of shoulder dislocation they experienced at San Francisco Hospital between
98 5 Traumatic Dislocation

Fig. 5.8 Compression

fracture of the humeral
head. (Reprinted from
Fig. 1, PL. XXI,
Malgaigne [25])

1930 and 1940 [26]. This lesion had been known as the “typical defect” seen in
dislocated shoulders before their publication, but they determined that this lesion
was a compression fracture of the humeral head created by the glenoid rim.
Thereafter, this lesion has been called a “Hill–Sachs lesion” I use the term “Hill–
Sachs lesion” in this book because it is most widely and commonly used.

Mechanism of Instability by Hill–Sachs Lesion

How does a Hill–Sachs lesion cause shoulder instability? As the Hill–Sachs lesion
is located on the posterolateral aspect of the humeral head, it has nothing to do with
shoulder instability in the mid-range of motion, where it does not come in contact
with the glenoid (Fig. 5.9a). At the end-range of motion, the lesion comes in contact
with the glenoid, but if it is still completely covered by the glenoid, the shoulder is
stable (Fig. 5.9b). However, if the medial margin of the Hill–Sachs lesion is located
more anterior than the anterior rim of the glenoid (Fig. 5.9c), the lesion can override
the glenoid and dislocation occurs (Fig. 5.9d). Therefore, the Hill–Sachs lesion is
related to the end-range instability.
2 Pathophysiology 99

a b c d

Fig. 5.9 Instability caused by Hill–Sachs lesion. In the mid-range of motion (a), Hill–Sachs lesion
has nothing to do with shoulder instability. At the end-range of motion, the shoulder is stable if the
Hill–Sachs lesion is completely covered by the glenoid (b). However, if the Hill–Sachs lesion
extends more medially over the anterior glenoid rim (c), the Hill–Sachs lesion overrides the ante-
rior glenoid rim and dislocation occurs (d). (Reprinted from Figs. 11–13, Itoi [22])

How to Assess the Risk of Hill–Sachs Lesion

It is important to understand that risk assessment of Hill–Sachs lesion is much

more complex than the glenoid defect. As mentioned above, instability risk of the
glenoid defect is solely determined by the defect size. However, a Hill–Sachs
lesion causes instability when it overrides the glenoid. A Hill–Sachs lesion which
does not override the intact glenoid (Fig. 5.10a) may override the glenoid with a
defect (Fig. 5.10b). The same Hill–Sachs lesion can be stable or unstable depend-
ing upon the glenoid size. Therefore, instability risk of Hill–Sachs lesion needs to
be assessed not only by the location of Hill–Sachs lesion, but also by the size of the
glenoid.

2.2.3 Risk Assessment of Bipolar Lesion

To our surprise, risk assessment of Hill–Sachs lesion had focused on the size of
Hill–Sachs lesion alone. None of the previous studies had paid attention to the gle-
noid. With the assumption that there is no glenoid defect, it makes sense to look at
only the Hill–Sachs lesion. However, 80–90% of the cases with shoulder instability
have a combined glenoid defect. It is absolutely necessary to look at the glenoid as
well as the Hill–Sachs lesion for risk assessment of instability.
100 5 Traumatic Dislocation

Fig. 5.10 Hill–Sachs a b

lesion and the glenoid with
and without a bone loss.
This Hill–Sachs lesion (a)
is completely covered by
the glenoid at the
end-range of motion. So,
this lesion does not cause
instability. However, the
same-sized Hill–Sachs
lesion may override the
glenoid and cause a
dislocation (b) if the Hill–
Sachs lesion cannot be
completely covered by the
glenoid due to a defect.
(Reprinted from Fig. 14,
Itoi [22])

Dynamic Examination

How can we assess both lesions at the same time? One method is dynamic exami-
nation. During arthroscopic surgery, we assess the closeness of Hill–Sachs lesion
to the anterior rim of the glenoid with the arm in abduction, external rotation, and
horizontal extension. If the Hill–Sachs lesion comes very close to or overrides the
anterior rim of the glenoid (dislocation), the lesion is called an engaging Hill–
Sachs lesion, which is thought to be a good indication for remplissage. However,
there is a large pitfall here. If you perform dynamic examination before the Bankart
repair, the humeral head easily slides anteriorly and the Hill–Sachs lesion
approaches to the anterior rim of the glenoid (Fig. 5.11a). If you add a little more
force, the head will dislocate anteriorly. However, engagement of the humeral
head (dislocation) before the Bankart repair does not necessarily mean that it also
engages after the Bankart repair. After the Bankart repair, the humeral head is well
stabilized in the center of the glenoid and it does not easily slide anteriorly
(Fig. 5.11b). As a result, the Hill–Sachs lesion which engaged before the Bankart
repair is now located far posteriorly at the level of the posterior rim of the glenoid,
and accordingly it does not engage anymore after the Bankart repair (Fig. 5.11c).
If an engagement still occurs after the Bankart repair, it is a true indication for
remplissage. Unfortunately, many doctors perform dynamic examination before
the Bankart repair and determine the indication of remplissage based on this
dynamic examination. There is no question that this is an over-indication of
remplissage.
2 Pathophysiology 101

a b c

Fig. 5.11 Dynamic examination during surgery. (a) Before the Bankart repair, the Hill–Sachs
lesion approached very close to the anterior rim of the glenoid with the arm in abduction and exter-
nal rotation (pre-repair dynamic examination). (b) After the Bankart repair, the humeral head was
well stabilized in the glenoid and it did not slide anteriorly with the arm in abduction and external
rotation (post-repair dynamic examination). (c) In this position, the Hill–Sachs lesion was located
far posteriorly at the level of the posterior glenoid rim (c). Suture anchors were inserted in the
Hill–Sachs lesion because remplissage had been indicated based on the pre-repair dynamic exami-
nation. GA glenoid anterior rim, GP glenoid posterior rim, HSL Hill–Sachs lesion, L labrum

Post-repair dynamic examination is the correct way to perform. However, there

are a couple of problems. One problem is if you perform dynamic examination after
the Bankart repair, you put a stress on the repair site, which may damage the repair.
Another problem is that remplissage after the Bankart repair is extremely difficult
to perform. The humeral head is already stabilized and it does not easily slide ante-
riorly, so the bony bed preparation and insertion of suture anchors into the Hill–
Sachs lesion is very difficult to perform. If we are to perform remplissage, we need
to do it before the Bankart repair. For this purpose, dynamic examination is useless.

Glenoid Track: A New Concept

This is the reason why we introduced a new concept the “glenoid track” to assess
the risk of Hill–Sachs lesion preoperatively. First, we looked at the relationship
between the glenoid and the humeral head when the arm was moved along the pos-
terior end-range of motion. With the arm in full external rotation and full horizontal
extension, the arm was elevated in the coronal plane [27]. The glenoid moved along
the posterior articular margin of the humeral head from inferomedial to superolat-
eral direction, creating a track on the humeral head (Fig. 5.12). We called this track
the “glenoid track.”
Next, we looked at the relationship between the glenoid track and the Hill–Sachs
lesion (Fig. 5.13). As Hill–Sachs lesion “A” is located inside the glenoid track, there
is no chance that the anterior glenoid rim (medial margin of the glenoid track)
engages with “A” lesion. So, this Hill–Sachs lesion is safe with no risks of instabil-
ity. We call it an “on-track lesion” [22, 28]. On the other hand, Hill–Sachs lesion
102 5 Traumatic Dislocation

Fig. 5.12 Glenoid track.

The arm was elevated from
0°, 30°, up to 60° of
abduction while keeping
the arm in maximum
external rotation and
horizontal extension.
During this arm motion,
the glenoid moved along
the posterior rim of the
articular surface of the
humeral head from
inferomedial to
superolateral direction.
This track created by the
glenoid is called the
“glenoid track.”
(Reproduced with
modification from Fig. 4A,
Yamamoto et al. [27])

Fig. 5.13 Glenoid track

and Hill–Sachs lesion.
Hill–Sachs lesion “A” is
located inside the glenoid
track, so there is no risk
that the anterior glenoid
rim engages with “A”
lesion. “A” lesion is called
an “on-track lesion.” On
the other hand, “B” lesion
extends more medially
over the medial margin of
the glenoid track. In this
case, there is a risk of
engagement (dislocation)
when the arm comes to a
position where the medial
margin of the glenoid track
overlaps with “B” lesion
(red dotted line). “B”
lesion is called an
“off-track lesion”

“B” extends more medially over the medial margin of the glenoid track. When the
arm comes to a position where the medial margin of the glenoid track and “B”
lesion overlap with each other (red dotted line), there is a risk of engagement (dis-
location). We call it an “off-track lesion.” If the Hill–Sachs lesion is off track, the
Bankart repair alone is not sufficient to achieve stability of the shoulder [22, 28].
2 Pathophysiology 103

“A” lesion used to be called a “non-engaging Hill–Sachs lesion” and “B” lesion an
“engaging Hill–Sachs lesion.” However, as mentioned above, these terms engaging/
non-engaging are misused based on the pre-repair dynamic examination, which is
not correct. In order to avoid this misunderstanding and confusion, Stephen
S. Burkhart (a namer of engaging/non-engaging lesion), Giovanni Di Giacomo
(a missionary of glenoid track concept), and myself sat together and discussed this
issue. Since the glenoid track concept has been accepted worldwide, we should
incorporate this concept into the new terminology. We came to a consensus that we
use “on-track/off-track lesion” instead of “engaging/non-engaging lesion” [28].

Measurement of Glenoid Track Width

Where is the medial margin of the glenoid track located? First, we need to define the
medial and lateral margins of the glenoid track. The medial margin is the track cre-
ated by the anterior glenoid rim. However, the lateral margin is not clear because the
posterior glenoid rim overrides the cuff insertion site (Fig. 5.14). Thus, we defined
the lateral margin of the glenoid track as the lateral margin of the articular cartilage
of the humeral head, which is equivalent to the medial margin of the footprint of the
rotator cuff. With this definition, the average width of the glenoid track was 84% of
the glenoid width in cadaveric shoulders because 16% of the posterior glenoid over-
lapped the footprint of the rotator cuff.
Next, we measured the glenoid track width in live shoulders using healthy volun-
teers (Fig. 5.15) [29]. The arm was gradually elevated in the positions of anterior
apprehension test (a, b, c). Simulating these positions, we took MRI of the shoulder

Fig. 5.14 Glenoid track

width (cadaveric shoulder).
As the posterior rim of the
glenoid overlaps the cuff
footprint, the average
width of the glenoid track
is 84%, a little narrower
than the glenoid width
(100%)
104 5 Traumatic Dislocation

a b c

d e f

g h i

Fig. 5.15 Glenoid track in live shoulder. The arm was passively elevated while keeping the arm in
maximum external rotation and horizontal extension. Three positions are shown with the arm
abducted at 60° (a), 90° (b), and 120° (c). Simulating these positions, MRI was taken in 7 positions
and 3D skeletal images were created (d–f). Using these skeletal models, the humeral head was
fixed and the glenoid motion on the humeral head was observed (g–i)

with the arm abducted at 0°, 30°, 60°, 90°, 120°, 150°, and maximum while keeping
the arm in maximum external rotation and horizontal extension. The 3D skeletal
images were extracted from these MR images (d, e, f). Finally, the humeral head
was fixed and the glenoid motion on the humeral head was observed (g, h, i).
Using this method, we visualized the glenoid track in live shoulders (Fig. 5.16)
[29]. The most medial point of the anterior glenoid (blue point) creates the medial
margin of the glenoid track. The medial margin of the cuff footprint is displayed
in black dotted line. From this image, we calculated the average width of the
2 Pathophysiology 105

Fig. 5.16 Glenoid track width (live shoulder). Similar to what we observed in cadaveric shoul-
ders, the glenoid moved along the posterior rim of the articular cartilage of the humeral head. The
blue line represents the medial margin of the glenoid track. The black dotted line indicates the
medial margin of the footprint of the rotator cuff, which is defined as the lateral margin of the
glenoid track. In live shoulders, the average glenoid track width was 83% of the glenoid width

glenoid track to be 83% of the glenoid width. This was quite close to the value
obtained from cadaveric shoulders. We decided to use this 83% value for clinical
application.

Clinical Application of Glenoid Track

How can we use the value 83% in the clinical setting? At present, 3D-CT is the most
accurate and reliable method of assessment. The important thing when we order
3D-CT is that we ask the radiologist to create 3D images of not only the involved
shoulder but also the contralateral shoulder. When CT scan is being performed,
bilateral shoulders are in the CT gantry. That means the CT data of bilateral shoul-
ders are obtained. No need to scan twice. All we need to do is just ask the radiologist
to make 3D images of bilateral shoulders. With these bilateral images, I will explain
how to determine the location of the glenoid track (Fig. 5.17) [22]. First, the glenoid
width D is measured in the contralateral shoulder and 83% of D (0.83D) is the gle-
noid track width (Fig. 5.17a). The glenoid width of the involved shoulder is a little
narrower than the intact glenoid, and the difference between them (d) is the defect
width (Fig. 5.17b). 0.83D is the width of the glenoid track without a glenoid defect.
In this case, the defect size needs to be subtracted from 0.83D to obtain the glenoid
track width (0.83D − d). In the posterior 3D image of the humeral head (Fig. 5.17c),
the dotted line R represents the medial margin of the cuff footprint. The dotted line
106 5 Traumatic Dislocation

a b c

Fig. 5.17 How to depict the glenoid track. (a) En face view of the contralateral glenoid. The width
D is measured to obtain 0.83D, the glenoid track width of the intact glenoid. (b) En face view of
the involved glenoid. The difference of the width between the intact glenoid and the glenoid with
a bony defect is the defect width (d). The glenoid track width on this involved side is determined
by subtracting d from 0.83D. The white dotted line indicates the shape of the original glenoid,
obtained as a mirror image of the intact glenoid. (c) Posterior view of the humeral head. R is the
medial margin of the cuff footprint. G, the medial margin of the glenoid track, is obtained by
parallel-­shifting R by the glenoid track width (0.83D − d). This Hill–Sachs lesion is on track.
(Reproduced with modification from Fig. 18, Itoi [22])

G, which is a parallel shift of R by the width of the glenoid track, is the medial mar-
gin of the glenoid track. In this case, the Hill–Sachs lesion stays on the glenoid track
(on-track lesion).

Factors Affecting the Glenoid Track Width

The glenoid track width, which is 83% of the glenoid width, is an average value of
the healthy adults. This value changes on how much the glenoid overrides the cuff
footprint. For example, in patients with a frozen shoulder, the range of shoulder
motion is limited in all directions including horizontal extension. Then, there is no
overriding of the posterior glenoid on the cuff footprint, indicating that the glenoid
track width increases. On the contrary, in patients with hyperlaxity, horizontal
extension is much greater, and as a result, there is a greater overlapping of the gle-
noid and the cuff footprint. This means the width of the glenoid track decreases in
hyperlax shoulders. If we can clarify this relationship between the range of motion
and the glenoid track width, we can provide the individualized glenoid track width
for each patient we take care. This seems to be much better than using a fixed
value of 83%.
First, we determined the relationship between the shoulder range of motion and
the glenoid track width using cadaveric shoulders [30]. This study showed that there
were negative correlations between external rotation angle and the glenoid track
width and between horizontal extension angle and the glenoid track width. The
greater these angles, the narrower the glenoid track. Next, we examined this rela-
tionship in healthy volunteers [31]. We measured the ranges of various motions,
2 Pathophysiology 107

both active and passive, and both in the sitting and supine positions. The highest
correlation was observed between active horizontal extension in the sitting position
and the glenoid track width. The relationship is as follows:

Y = −0.49 X + 90,

where X (°) is the active horizontal extension angle in the sitting position and Y is
the glenoid track width (%). For example, if the active horizontal extension angle is
10° (normal value), the glenoid track width is calculated to be 85%. If it is a stiff
shoulder with 0° of horizontal extension, the glenoid track width is 90%. If it is a
hyperlax shoulder with 20° of horizontal extension, the glenoid track width is 80%.
By measuring the active horizontal extension angle of the contralateral shoulder in
the sitting position, we can provide a more precise estimate of the glenoid track
width each individual has.
Furthermore, it has become clear that the location of the Hill–Sachs lesion within
the glenoid track affects the outcome of surgical stabilization. We divided the gle-
noid track into four zones with equal width. Then the patients were divided into four
groups based on the zone where the medial margin of the Hill–Sachs lesion was
located. We compared the WOSI score of patients among the four groups [32]. This
study demonstrated that those whose Hill–Sachs lesion was located in the lateral 3/4
of the glenoid track showed no difference in WOSI score, but those in the most
medial 1/4 of the glenoid track showed significantly worse WOSI score than the
others (Fig. 5.18). We call the most medial 1/4 of the glenoid track the “peripheral
track” and the rest 3/4 the “central track.” If the Hill–Sachs lesion is a peripheral-­
track lesion, we have to select surgical procedures based on the risks of the patient.

Fig. 5.18 Peripheral track

and central track. The most
medial 1/4 of the glenoid
track (orange zone) is
called the peripheral track,
whereas the rest 3/4 is
called the central track.
Those with a peripheral-­
track lesion showed
significantly worse WOSI
score than those with a
central-track lesion.
(Reprinted from Fig. 4, Itoi
[33])
108 5 Traumatic Dislocation

If the patient belongs to a low-risk group, it is fine to take it as an on-track lesion.

However, if the patient belongs to a high-risk group such as contact/collision ath-
letes, we should select surgical procedures assuming that it is an off-track lesion.

2.2.4 Mechanism of Bone Loss

There are several theories about the mechanism of bone loss, but there have been no
research studies demonstrating in cadaveric shoulders how a bony defect is created.
Regarding the Hill–Sachs lesion, it is obvious from X-rays and CT images that it is
a compression fracture of the humeral head caused by the anterior glenoid rim
(Figs. 5.19 and 5.20).
However, it is still unknown when the Hill–Sachs lesion is created. Is it immedi-
ately after the dislocation? Or is it sometime later when the arm is lowered down?
From 3D-CT images of patients with recurrent shoulder dislocation, we recon-
structed 3D images of Hill–Sachs lesion and the glenoid. By fitting these images,
we determined the arm position when the Hill–Sachs lesion engaged with the gle-
noid rim [34]. The average position of the arm when these lesions fit well was 74°
of abduction, 27° of external rotation, and 3° of horizontal flexion, which was quite
different from the typical position of dislocation, i.e., 90° of abduction and 90° of
external rotation. This study implies that regardless of the arm position at the time
of dislocation, the Hill–Sachs lesion is likely to be created sometime later when the
arm is lowered down.

a b

Fig. 5.19 X-ray before and after reduction. (a) The humeral head is dislocated anteriorly, and the
anterior glenoid rim and a Hill–Sachs lesion are engaging with each other. (b) After reduction, the
Hill–Sachs lesion is clearly seen on the humeral head with the shape of glenoid rim
2 Pathophysiology 109

Fig. 5.20 CT image with

the shoulder dislocated.
The shape of Hill–Sachs
lesion perfectly fits with
the shape of anterior
glenoid rim. This image
tells that the Hill–Sachs
lesion is a compression
fracture caused by the
anterior glenoid rim

Next, we performed a study to clarify when, how, and which type of bony defects
were created during experimental dislocation in cadaveric shoulders [35]. This
experiment was performed in two stages: stage 1, anterior dislocation; and stage 2,
compression of the humeral head to the glenoid. In stage 1, the shoulder was
abducted at 60° to the scapula (90° to the trunk) and maximum external rotation,
and then horizontal extension force of average 352 N was applied until an anterior
dislocation occurred. All 15 shoulders dislocated anteriorly, and avulsion fracture of
the anterior glenoid was observed under fluoroscopy in four shoulders (27%)
(fragment-­type bone loss) (Fig. 5.21).
In stage 2, the humeral head was first reduced to the glenoid socket. Then, while
keeping the arm in 60° of abduction and maximum external rotation with use of
25-N force each, 800-N force was applied to the humeral head in line with the pec-
toralis major muscle. This caused an anterior dislocation, and the arm came to equi-
librium. The equilibrium position of the arm was 40° of flexion, 45° of abduction,
and 27° of external rotation relative to the scapula. During this step, a Hill–Sachs
lesion was created in 6 of 15 shoulders (40%) (Fig. 5.22) and a glenoid bone loss
was created in 7 of 15 (46%) (erosion-type bone loss).
Micro CT and histologic examination performed in a single specimen with an
erosion-type bone loss showed a flattened anterior glenoid rim with collapse of
trabecular bone in addition to a fracture line of the subchondral bone (Fig. 5.23).
From this experiment, we concluded that the fragment-type glenoid bone loss was
created at the time of dislocation whereas the erosion-type bone loss (compression
fracture) and Hill–Sachs lesions were created when the arm came to equilibrium in
the mid-range of motion.
110 5 Traumatic Dislocation

Fig. 5.21 Fragment-type

bone loss. An avulsion
fracture (arrow) of the
anterior glenoid rim was
created at the time of
dislocation. G glenoid, HH
humeral head (Reprinted
from Fig. 5, Etoh et al.
[35])

Fig. 5.22 Hill–Sachs

lesion. A compression
fracture (arrowheads) of
the humeral head
(Hill–Sachs lesion) was
created during equilibrium
by the anterior glenoid rim.
(Reproduced with
modification from Fig. 6,
Etoh et al. [35])

The equilibrium position of the arm in this experiment (40° of flexion, 45° of
abduction, and 27° of external rotation relative to the scapula) was 68° of elevation
in the scapular plane and 27° of external rotation relative to the trunk. This is very
similar to the position of the arm where the 3D reconstructed images of Hill–Sachs
lesion and the glenoid engaged with each other (74° of abduction, 27° of external
rotation, 3° of horizontal flexion). This similarity seems to support the validity of
2 Pathophysiology 111

Fig. 5.23 Erosion-type bone loss (hematoxylin–eosin stain, × 8). The anterior rim of the glenoid
was compressed and flattened (black arrow). A fracture line (white arrow) was observed on the
glenoid surface with intact articular cartilage. (Reproduced with modification from Fig. 8, Etoh
et al. [35])

our experimental dislocation model. It is noteworthy that the Hill–Sachs lesion was
created not at the time of dislocation but later when the arm was in the equilibrium
position. You may think that a patient with shoulder dislocation comes to your clinic
with the arm in almost adduction and internal rotation, not in abduction and external
rotation. This is because it is easier for the patient to keep the arm in adduction and
internal rotation than in abduction and external rotation. Although the scapula and
the humerus are locked in the engaged position of abduction and external rotation,
the patient tries to lower the arm by rotating the scapula downward and internally.
Due to this compensatory motion of the scapula, the arm looks as if it is in adduc-
tion and internal rotation (Fig. 5.24).

2.3 Transition to Recurrent Instability

It is well known since early times that shoulder dislocation is very easy to recur. In
the era of Hippocrates, recurrent dislocation was common and many physicians
burnt the shoulder at the top or at the anterior part to avoid dislocations, but
Hippocrates recommended to burn the armpit skin using a red-hot iron [36]. An
armpit burn healed with hard scar tissues, which probably helped prevent further
dislocations. More than 2000 years have passed since Hippocrates, but we still have
the same problem: recurrent shoulder dislocation. Age is one important factor: the
younger the patient, the higher the recurrence. According to Rowe and Sakellarides,
the recurrence rate was 100% for those younger than 10 years, 94% for those aged
11–20 years, 79% for those aged 21–30 years, 50% for those aged 31–40 years,
24% for those aged 41–50 years, and less than 16% for those older than 51 years
[37]. They reported that other than age, those with epilepsy and a large Hill–Sachs
lesion were more likely to have recurrence, whereas those with a greater tuberosity
112 5 Traumatic Dislocation

Fig. 5.24 A first-time

dislocator at the first visit.
The arm looks in slight
abduction (30°) and
internal rotation. However,
it appears so by the
compensatory scapular
motion of downward
rotation, abduction, and
internal rotation

fracture were less likely to. A recent meta-analysis has also demonstrated that young
age and male are the risk factors of recurrence, whereas a greater tuberosity fracture
is a negative risk factor [38].

2.4 Healing of Bankart Lesion

Using a rabbit model, we created a Bankart lesion and observed the histological
healing processes and return of the mechanical strength every week up to 5 weeks
postoperatively [39]. We observed the inflammatory cell infiltration at 1 week and
completion of histological healing at 3 weeks with the gap filled with collagen-rich
connective tissues. Mechanical strength decreased at 1 week and returned to normal
at 4 weeks. Thus, it took 1 month for the histological and biomechanical healing of
the Bankart lesion in a rabbit model. There are some differences in the healing
3 Diagnosis 113

process between rabbits and humans. For example, the healing period of tibial frac-
ture is reported to be 6 weeks in rabbits and 18 weeks in humans. We do not know
how much difference there is in the healing process of the Bankart lesion, but
extrapolating the difference in fracture healing, the Bankart healing in humans is
likely to take approximately 3 months. Return to play after 3 months of surgery,
which is commonly performed in the clinical setting, seems to be legitimate.

3 Diagnosis

3.1 Plain X-rays

When the humeral head is dislocated and engaged with the glenoid, it is difficult to
abduct the arm. Thus, plain X-rays are taken in the anteroposterior view and scapu-
lar Y-view, but not the axillary view. Attention needs to be paid to pathologies other
than a dislocation because sometimes the dislocation is combined with fractures of
the glenoid rim, greater tuberosity, and other parts of the shoulder components.
There are some plain X-ray views specific to a certain pathology such as the West-­
Point view and Bernageau view for a glenoid rim fracture and the Stryker view for
a Hill–Sachs lesion (Fig. 5.25a). If an anterior glenoid rim line disappears on

a b

Fig. 5.25 Plain X-rays. (a) The Stryker view depicts a Hill–Sachs lesion clearly. (b) On antero-
posterior view, the anterior glenoid rim line is clearly seen on the upper half of the glenoid (white
arrows), but it disappears in the lower half (black arrows). This indicates a bone loss of the anterior
glenoid rim
114 5 Traumatic Dislocation

anteroposterior view, this indicates a glenoid bone loss (Fig. 5.25b). Using plain
X-rays, we can make a diagnosis of bony lesions, but it is impossible to quantify the
bony lesions.

3.2 CT

Recently, the use of 3D-CT spreads rapidly, and it is the most appropriate imaging
modality to quantify the bony lesions. For the assessment of glenoid bone loss, we
need to estimate the original glenoid rim in the following two methods: (1) to use a
best-fitting circle (best-fit circle method, Fig. 5.26a) or (2) to use the contralateral
glenoid as a control (contralateral comparison method, Fig. 5.26b). We compared
the reliability of these two methods [40]. Both the intraobserver and interobserver
reliabilities were higher when the contralateral comparison method was used. We
concluded that the contralateral comparison method was more reliable than the
best-fit circle method for quantifying the amount of glenoid bone loss. One reason
is that the size of the best-fit circle varies greatly among the observers. The other
reason is that some normal glenoid does not fit well to a circle, and the best-fit circle
method makes the intact glenoid look as if there is an erosion-type bone loss
(Fig. 5.26c). In such a case, if we use the best-fit circle method, we overestimate the
size of bone loss. Therefore, the contralateral comparison method is better than the
best-fit circle method except for cases of bilateral shoulder involvement in which
the contralateral glenoid cannot be a control. In such cases, we must use the best-fit
circle method.

a b c

Fig. 5.26 Quantification of bony defect. (a) The best-fit circle method. (b) The contralateral com-
parison method. A and B are the same glenoid but the defect looks larger using the best-fit circle
method (a) than using the contralateral comparison method (b). (c) The contralateral glenoid. In
this case, the best-fit circle does not fit well, and it looks as if there is an erosion-type bone loss
although this is an intact glenoid
3 Diagnosis 115

3.3 MRI

MRI is very good at depicting the soft tissues. Detachment of the labrum-IGHL
complex (Bankart lesion) or a capsular tear is clearly seen on MR images. However,
if there is not sufficient joint fluid or effusion, the torn edges may be attached and
the lesion may not be visualized by plain MRI. In order to clearly visualize the
lesion, Gd-based contrast medium is injected into the shoulder joint to fill the defect
with the liquid to have better visualization, called MR arthrography (Fig. 5.27). The
mechanism of visualization is similar to filling the water tank with the water. The
water plants in the tank will spread and are better visualized with water inside com-
pared to the plants sticking to the tank floor without any water.
If Gd is used without dilution, it appears low signal intensity on both T1- and
T2-weighted images. On the other hand, between 100-fold and 500-fold dilution, it
appears high signal intensity on T1-weighted image (Fig. 5.28). Because of this
relationship, we use 200-fold dilution of Gd for MR arthrography in our institute.
Contrary to soft tissues, bones appear as signal void on MR images, so bone on
MR image is not so clear as that on CT scan. CT is superior in bone imaging and
MRI is superior in soft tissue imaging. Due to technical advancement, a recent
meta-analysis has shown that both CT and MRI accurately measure glenoid bone
loss [42].

a b

Fig. 5.27 MR arthrograms of shoulder with recurrent dislocation. (a) T1-weighted coronal
oblique image, (b) T1-weighted axial image. The glenohumeral joint is filled with Gd contrast
agent (200-fold dilution), which appears white on T1-weighted image. The labrum-IGHL complex
(yellow arrow) is detached from the glenoid and displaced inferiorly (a) and anteriorly (b).
(Reproduced with modification from Fig. 27–23, Itoi [41])
116 5 Traumatic Dislocation

Fig. 5.28 Serial dilution

of Gd and MR signal
intensity. High signal
intensity is obtained on
T1-weighted image with
use of Gd contrast agent
diluted between 100 and
500 fold. (By courtesy of
Dr. Shin Hitachi,
Department of Diagnostic
Radiology, Tohoku
University Hospital)

4 Treatment

4.1 Position of Immobilization After Reduction

4.1.1 Basic Research

The greatest mystery of shoulder dislocation is why it recurs so often. It has been
well known that young patients are highly prone to recurrence. The most common
lesion created by shoulder dislocation is a Bankart lesion, detachment of the labrum-­
IGHL complex from the glenoid. Since it is traumatically detached from the gle-
noid, it is supposed to heal if it is back to the original position. There is no question
that the lesion has healing ability since the lesion heals after surgical stabilization.
The question is why it does not heal and results in recurrent dislocation. The highest
peak of incidence of first-time dislocation is in the teens and twenties. One explana-
tion is that the young are very active, but the soft tissues are still immature com-
pared to the adult. The soft tissues might not be strong enough to resist high
activities, resulting in first-time and recurrent dislocations. However, we also know
that how long and how rigid we immobilize the shoulder, we still cannot get rid of
recurrence. Rowe looked at the effect of length of immobilization (1-week increment
4 Treatment 117

from 1 to 6 weeks) on recurrence rate, but there was no difference among them [2].
He also looked at the effect of rigidity (no immobilization, sling, strapping, cast) on
recurrence, but again there was no difference [2]. What do these results indicate?
The Bankart lesion is expected to heal once it is reduced to be in contact with the
glenoid and stay in that position for a certain period as is always the case in other
soft tissue lesions. If it does not heal despite the long and rigid immobilization, the
only possible explanation is that the lesion is not well reduced and coapted. For
example, when you treat an Achilles tendon rupture using a cast, you immobilize
the ankle in plantar flexion so that the ruptured edges of the tendon would be
coapted. If you immobilize the ankle in dorsiflexion, it never heals. I thought the
same thing might be going on in the shoulder.

Arm Position and Bankart Reduction

I was surprised to find out that there had been no studies to examine the reduction of
the Bankart lesion in the conventional position of immobilization, i.e., adduction and
internal rotation. This was the very beginning of the serial studies on this topic. First,
we created the Bankart lesion in cadaveric shoulders with all the soft tissues removed
except the capsule, and observed opening and closing of the lesion in reference to the
arm position [43]. This study revealed that the Bankart lesion was well coapted within
the range of maximum internal rotation to 30° of external rotation in the hanging arm
position. At 30° of abduction, it was coapted from maximum internal rotation to 0° of
external rotation. However, at 45° and 60° of abduction, it was not coapted anymore.
This study clarified that the Bankart lesion was coapted in certain ranges of motion,
but the limitation of this study was that all the muscles including the rotator cuff were
removed. The arm rotation seems to affect the tension of the muscles around the
shoulder, which is also likely to affect the coaptation of the lesion.
The next study to clarify the relationship between the arm position and coapta-
tion of the Bankart lesion was performed in patients [44]. Six shoulders had had a
first-time dislocation and 13 shoulders had had recurrent dislocation. Plain MRI was
taken when the dislocation had occurred less than 2 weeks earlier, and MR arthrog-
raphy was taken when the dislocation had occurred more than 2 weeks earlier. The
MRI was taken with the arm in two different positions: adduction-internal rotation
(mean 29°) and adduction-external rotation (mean 35°). This study demonstrated
that the labrum and the capsule were separated from the glenoid with the arm in the
conventional position of internal rotation (Fig. 5.29a), but they were better reduced
with the arm in external rotation (Fig. 5.29b).

4.1.2 Clinical Studies

Based on these findings, we hypothesized that immobilization in external rotation

would achieve better reduction of the Bankart lesion and result in lower recurrence
rate. To prove our hypothesis, we planned a prospective randomized clinical trial.
118 5 Traumatic Dislocation

a b

Fig. 5.29 Reduction of Bankart lesion. (a) With the arm in adduction and internal rotation, the
labrum (orange arrow) was displaced medially, and the capsule (yellow arrow) was separated from
the glenoid. (b) With the arm in adduction and external rotation, both the labrum (orange arrow)
and the capsule (yellow arrow) were better reduced. (Reproduced with modification from Fig. 4,
Itoi et al. [44])

We enrolled 198 patients with a first-time shoulder dislocation from 15 institutes.

The patients were randomly assigned to either the internal rotation group with sling
fixation (94 shoulders) or the external rotation group with a brace (104 shoulders)
(Fig. 5.30). After 3 weeks of immobilization, they were free to move their arms and
allowed to return to sports after 3 months.
We analyzed data of 159 patients (80%) available for 2-year follow-up. In
both groups, the patients were instructed to wear the brace (external rotation
group) or a sling and swathe (internal rotation group) continuously, except
when they took a shower, for 3 weeks. The compliance rate of immobilization
was significantly higher in the external rotation group (72%) than in the inter-
nal rotation group (53%) (p = 0.013). The recurrence rate was significantly
lower in the external rotation group (26%) than in the internal rotation group
(42%) (p = 0.033). The relative risk reduction was 38%. Return to sports was
72% in the external rotation group and 63% in the internal rotation group, but
return to pre-injury sports level was 37% and 20%, respectively, with no sig-
nificant difference.
We further investigated how long the healing Bankart lesion should be protected.
We made a shoulder motion restriction band connecting the trunk and the elbow to
avoid stretching the healing IGHL complex (Fig. 5.31). All the patients were immo-
bilized in external rotation for 3 weeks. Following the immobilization, they were
randomly assigned to one of the three groups: no restriction band (36 patients),
restriction band for 3 weeks (37 patients), or restriction band for 6 weeks (36
patients) [45]. This study showed no significant difference in recurrence rate or
return to sports among these three groups.
4 Treatment 119

a b

c d

Fig. 5.30 Braces for immobilization in ER. First, we used external rotation immobilizer consist-
ing of a splint and a stockinette (a, b). In November 2003, we started to use a prototype of the
external rotation immobilization brace (Alcare, Tokyo, Japan) (c, d). (Reprinted from Figs. 1 and
2, Itoi et al. [5])

Searching for a Better Position

We further searched for a better position of immobilization if there is any. Hart and
Kelly reported that during arthroscopic examination of shoulders after first-time
dislocation, 92% of the Bankart lesions were reduced in adduction and external
rotation. However, the best reduction was achieved with the arm in 30° of abduction
and 60° of external rotation. The Bankart lesion displaces in anteroinferior direction
relative to the trunk. External rotation tightens the anterior soft tissues, which pushes
the lesion posteriorly. Similarly, abduction tightens the inferior soft tissues, which
pushes the lesion superiorly. In terms of reducing the Bankart lesion, it seems quite
120 5 Traumatic Dislocation

b c d
a

e f g h i

Fig. 5.31 Shoulder motion restriction band. The band is composed of three portions: the elbow
portion, the trunk portion, and the connecting portion (a). Resting position with the band applied
(b). With the arm in adduction (c, d), there is no restriction in shoulder rotation. However, arm
elevation is restricted at 60° (e–g). Arm rotation is not restricted at 60° of abduction (h, i).
(Reprinted from Fig. 2, Itoi et al. [45])

Fig. 5.32 Reduction of

Bankart lesion. External
rotation tightens the
anterior soft tissues, which
pushes the Bankart lesion
posteriorly. Similarly,
abduction tightens the
inferior soft tissues, which
pushes the lesion
superiorly. Abd abduction,
ER external rotation

reasonable to add abduction to external rotation to achieve a better reduction of the

lesion (Fig. 5.32).
In order to determine whether we should add abduction, we performed the next
study. We took MRI of first-time dislocators in four different arm positions: (1)
adduction and internal rotation, (2) adduction and external rotation, (3) 30° of
abduction and 30° of external rotation, (4) 30° of abduction and 60° of external rota-
tion. Since the shoulder joint space was filled with hematoma after the dislocation,
plain MRI was sufficient to assess the displacement of the labrum and capsule from
the glenoid. This study showed that the anterior part of the Bankart lesion was
reduced regardless of adduction or abduction as long as the arm was in external
4 Treatment 121

rotation. However, the inferior part of the Bankart lesion was better reduced in
abduction than in adduction. The best reduction was achieved in 30° of abduction
and 60° of external rotation, followed by 30° of abduction and 30° of external rota-
tion. Before applying these findings to a clinical trial, we need to assess patient
compliance with treatment using these immobilization positions. If keeping these
arm positions is too uncomfortable or even painful, it cannot be used in the clinical
setting. To clarify this point, we performed the next study.

Acceptability of the Brace

Healthy volunteers were enrolled in this study to wear a brace on the dominant arm
in four different positions mentioned above for 24 h each in random orders [46].
After completion of each 24-h immobilization, they were asked to assess the dis-
comfort of bracing for individual activities using a visual analog scale. The results
showed that the position of 30° of abduction and 60° of external rotation was signifi-
cantly more uncomfortable than adducted positions. The position of 30° of abduc-
tion and 30° of external rotation did not show any significant differences compared
to the other arm positions. This study provided two options: option #1, to use the
best position for reduction of the Bankart lesion even though it is uncomfortable or
option #2, to use the second-best position to put a higher value on compliance. After
careful consideration, we decided to go with option #2, immobilization in 30° of
abduction and 30° of external rotation.
This new position of immobilization was convenient for us because we already
had a brace (Shoulder Brace ER, Alcare, Tokyo, Japan) to immobilize the shoulder
in adduction and external rotation. By applying a certain attachment, we were able
to convert this brace to a brace to immobilize the shoulder in 30° of abduction and
30° of external rotation (Fig. 5.33b). Using this brace, we started a multicenter

a b

Fig. 5.33 Randomized clinical trial. A randomized clinical trial is ongoing to compare the efficacy
of immobilization in adduction-internal rotation (a) and in abduction-external rotation (b)
122 5 Traumatic Dislocation

randomized clinical trial comparing adduction-internal rotation (Fig. 5.33a) and

abduction-­external rotation (Fig. 5.33b) in February 2020, which is currently in
progress.
In addition, we compared the performance, comfort, and user-friendliness of
four external rotation immobilizers commercially available [47]. These braces were
UltraSling (Donjoy), Shoulder Brace ER (Alcare), SmartSling (Össur), and Omo
Immobil (Otto Bock) (Fig. 5.34). The former two braces were for immobilization in
adduction and external rotation and the latter two were for immobilization in abduc-
tion and external rotation. We measured the angle of external rotation at the time of
brace application, after several activities of daily living, and after re-attachment of
the brace. None of these braces showed significant differences among these three
measurements. The average initial angle was significantly greater in Shoulder Brace
ER (15.2°) than in UltraSling (10.9°). There was no difference in difficulty of reap-
plication, but abduction-external rotation braces were less comfortable than
adduction-­external rotation braces. All these four braces seemed to be suitable for
clinical use.

a b

c d

Fig. 5.34 Four braces used for external rotation immobilization. (a) UltraSling ER (Donjoy, Vista,
USA), (b) Shoulder Brace ER (Alcare, Tokyo, Japan), (c) SmartSling (Össur, Reykjavik, Iceland),
(d) Omo Immobil (Otto Bock, Duderstadt, Germany). (Reprinted from Fig. 1, Hatta et al. [47])
4 Treatment 123

4.1.3 Meta-Analysis

Since we reported a first randomized clinical trial comparing immobilization in

internal rotation and in external rotation in 2007, many randomized clinical trials
have been reported. Also, there are several meta-analyses comparing the outcomes
of these trials, but their conclusions are not consistent. Some analyses concluded
that external rotation was better, but other analyses concluded that there was no dif-
ference between the two. Thus, we conducted our own meta-analysis including the
most recent trials available [48]. Using the electronic database as of January 2019,
we searched for articles using the following texts: [shoulder dislocation OR gleno-
humeral dislocation] AND [primary OR first time] AND [immobilization] AND
[external rotation]. A total of 9 studies (817 patients) were extracted from 189 arti-
cles obtained from the search. The recurrence rate was 35% (130/373) in the inter-
nal rotation group and 22% (84/390) in the external rotation group. Immobilization
in external rotation significantly reduced the recurrence rate with the relative risk
reduction of 44% (p = 0.007) (Fig. 5.35). In the subgroup analyses, those immobi-
lized full-time and those aged 20–40 years showed a significant effectiveness of
immobilization in external rotation.

4.2 Surgical Treatment

4.2.1 Surgical Indications for First-Time Dislocators

Since the recurrence rate was high, especially in collision athletes, an idea of surgi-
cal treatment for first-time dislocators arose from military doctors in the USA. As
the military patients were in high demand of physical activities, they persuaded
surgical stabilization and reported excellent outcomes. Pushed by this movement, a
trend of treating initial dislocators surgically has gained popularity. However, a
recent meta-analysis raised an alarm on this trend. Kavaja et al. warned that if you

External Rotation Internal Rotation Risk Ratio Risk Ratio

Study or Subgroup Events Total Events Total Weight M-H, Random, 95% CI Year M-H, Random, 95% CI
Itoi et al. 2007 22 85 31 74 17.2% 0.62 [0.39, 0.97] 2007
Finestone et al. 2009 10 27 10 24 13.7% 0.89 [0.45, 1.76] 2009
Taskoparan et al. 2010 1 16 5 17 3.6% 0.21 [0.03, 1.63] 2010
Nagaraj et al. 2010 3 22 9 20 8.2% 0.30 [0.10, 0.96] 2010
Liavaag et al. 2011 28 91 23 93 16.9% 1.24 [0.78, 1.99] 2011
Wang et al. 2011 5 47 16 49 10.6% 0.31 [0.12, 0.77] 2011
Heidari et al. 2014 2 51 17 51 6.3% 0.12 [0.03, 0.48] 2014
Whelan et al. 2014 6 27 8 25 10.8% 0.69 [0.28, 1.72] 2014
Murray et al. 2018 7 24 11 23 12.7% 0.61 [0.29, 1.30] 2018

Total (95% CI) 390 373 100.0% 0.56 [0.37, 0.86]

Total events 84 130
2 2 2
Heterogeneity: Tau = 0.22; Chi = 19.87, df = 8 (P = 0.01); I = 60%
0.01 0.1 1 10 100
Test for overall effect: Z = 2.68 ( p =0.007) External Rotation Internal Rotation

Fig. 5.35 Forest plot of the recurrence rate. The recurrence rate was 22% in the external rotation
group and 35% in the internal rotation group. Immobilization in external rotation decreased the
recurrence rate with the relative risk reduction of 44% (p = 0.007). (Reprinted from Fig. 3,
Shinagawa et al. [48])
124 5 Traumatic Dislocation

routinely perform stabilization surgery for all the initial dislocators, the risk of
doing unnecessary surgery would be 50–79%. At the first visit, you never know
whether a patient in front of you with an initial dislocation will have recurrent dis-
location (chronic instability) in the future. The surgery performed to that patient is
a preventive surgery against chronic instability. Surgery as a prevention seems to be
too invasive. This is one concern. Another concern is that surgery for initial disloca-
tors has high risks of doing unnecessary surgery. On the other hand, if you wait until
the second dislocation occurs, the patient already has chronic instability. Surgery is
a treatment, not a prevention, for chronic instability. Also, there is no risk of doing
unnecessary surgery. Surgery for patients with chronic instability is 100% war-
ranted. That is why I wait to perform surgery until the second dislocation occurs. If
the patient agrees, immobilization in external rotation is a good indication. If
patients are in-season athletes, they usually want to go back to sports as soon as
possible. In that case, we let them go back to sports with a protective brace. After
the season, if instability remains, surgery should be indicated. Does it cause any
harm to wait for the second dislocation to occur? There is no direct answer to this
question. Dickens et al. let 39 intercollegiate contact athletes select a treatment
option after an in-season first-time dislocation. Ten athletes selected conservative
treatment and 29 selected surgical treatment. Among the 29 who selected surgical
treatment, 9 underwent surgery right after the dislocation. The rest 20 continued to
play until the end of the season and underwent surgery after the season. They com-
pared these groups with immediate surgery versus off-season surgery. Return to
play was 89% (immediate surgery) versus 90% (off-season surgery), and recurrence
rate was 11% versus 10%, respectively. There were no differences between these
two groups. They did not specify how many additional dislocations these 20 athletes
suffered between the initial dislocation and off-season surgery. In other reports from
the same group of authors, the average number of in-season dislocation was 1.4–2.2.
From these values, we can say that there is no difference in recurrence rate or return
to play whether contact athletes are treated surgically after an initial dislocation or
after the season with one or two additional dislocations at most. Therefore, I make
it a rule to wait to perform surgery until the second dislocation occurs except for
cases with large bony defects, which is an absolute indication for surgery (Also
refer to Sect. 4.3 Management of Bone Loss in this chapter).

4.2.2 Bankart Repair

Transition of Surgical Procedures

The primary goal of surgery is to regain stability by repairing the lesion created by
the dislocation. To achieve this goal, the Bankart repair is considered the gold stan-
dard for the surgical treatment. Open procedure was the only option before the
introduction of arthroscopy. There were two methods to open the glenohumeral
joint: horizontal capsulotomy and vertical capsulotomy. We compared the range of
motion after surgery between groups used these two methods of capsulotomy [49].
4 Treatment 125

Fig. 5.36 Arthroscopic

Bankart repair. The
anterior labrum (L) is
firmly fixed to the glenoid
(G) by the sutures attached
to the suture anchors
inserted to the anterior
glenoid rim. H
humeral head

We found that the horizontal capsulotomy was better in terms of regaining the range
of motion after the surgery. Then, arthroscopic procedures were introduced into the
field of shoulder stabilization in late 1980s. At first, since no suture anchors were
available, sutures put through the Bankart lesion were pulled backward through the
glenoid and tied over the infraspinatus muscle such as the Morgan procedure and
the Caspari procedure. As no rigid fixation of the Bankart lesion was achieved by
these techniques, the recurrence rate was almost as high as 50%. Introduction of
suture anchors enabled us to achieve firm fixation of the Bankart lesion to the gle-
noid. Accordingly, arthroscopic Bankart repair has been accepted widely and has
become the standard procedure these days (Fig. 5.36).

Location of Suture Anchor Insertion

It is controversial where to insert the suture anchors. We investigated the effect of

anchor location on stability in cadaveric shoulders [50]. The anchors were inserted
in three different locations: (1) the glenoid surface, (2) the glenoid rim, and (3) the
glenoid neck. Those whose anchors were inserted on the glenoid surface or at the
glenoid rim showed no significant difference in the outcomes, but those with the
anchors inserted to the glenoid neck showed significantly lower stability.
Considering that no healing is expected on the articular surface, inserting the anchor
to the glenoid rim seems to be the best. However, the lowest anchor at 5 o’clock
position of the right shoulder was difficult to insert at the glenoid rim through the
standard anterior portal and usually inserted on the glenoid surface instead. Thanks
to a curved-drill guide system, we can insert the 5 o’clock suture anchor to the gle-
noid rim [51].
126 5 Traumatic Dislocation

Remove Cartilage or Leave It On

Some surgeons recommend removing the articular cartilage from the anteroinferior
rim of the glenoid to obtain better healing of the detached Bankart lesion to the
bone. The major concerns of this procedure are as follows. First, if the peripheral
articular cartilage is removed from the anteroinferior glenoid rim, the contact area
of the glenoid surface decreases, which in turn increases the stress on the articular
cartilage. The stress increase may eventually result in osteoarthritis in the long term.
Second, the narrowing of the glenoid width has been reported in cases with cartilage
removal. This is probably the result of loss of contact between the subchondral bone
of the glenoid rim and the humeral head, which may have resulted in resorption of
the subchondral bone without the cartilage. Third, there are two patterns of labrum-­
IGHL complex attachment to the glenoid: (1) it attaches to both the articular carti-
lage and the glenoid neck and (2) it attaches only to the glenoid neck [52]. Whichever
the attachment pattern, there is no question that the major mechanical strength
depends on the attachment of the soft tissues to the bone, not to the cartilage, con-
sidering the difference in mechanical strength of the bone and cartilage. In our
series of patients treated with arthroscopic Bankart repair without removing the
articular cartilage, we found no change in the glenoid width 2 years after arthroscopic
Bankart repair [53]. In other words, we confirmed that there was no subchondral
bone resorption with the cartilage preserved. From these observations, I strongly
recommend that the articular cartilage be kept intact.
We have been consistent in not removing the cartilage during arthroscopic
Bankart repair. The clinical outcomes of our series showed that the recurrence rate
was 5% after open Bankart repair and 4% after arthroscopic Bankart repair. There
was no difference between them. However, subgroup analysis showed that the
recurrence rate of contact/collision athletes was 10% in open procedure and 14% in
arthroscopic procedure [54]. Based on these data, we currently perform open
Latarjet procedure for contact/collision athletes.

4.3 Management of Bone Loss

4.3.1 Selection of Surgical Procedures

Our surgical treatment paradigm for shoulder instability with various sizes of bone
loss is in Table 5.1.
As shown in the table, patients with on-track Hill–Sachs lesion and glenoid
defect <25% can successfully be treated with arthroscopic Bankart repair. In those
with on-track Hill–Sachs lesion and glenoid defect ≥25%, glenoid defect needs to
be treated with bone graft or coracoid transfer. I prefer to do Latarjet procedure in
such cases. If Hill–Sachs lesion is off-track and glenoid defect <25%, either of the
following procedures works: (1) arthroscopic Bankart repair plus remplissage (fill-
ing the Hill–Sachs lesion with the infraspinatus tendon) so that the glenoid cannot
4 Treatment 127

Table 5.1 Treatment paradigm

Hill–Sachs lesion
On-track lesion Off-track lesion
Glenoid <25% A/S Bankart A/S Bankart repair + remplissage or Latarjet
defect repair
≥25% Latarjet Latarjet (+ remplissage or bone graft to Hill–Sachs
lesion) or ICBG
A/S arthroscopic, ICBG iliac crest bone graft

fall into the pitfall of Hill–Sachs lesion or (2) Latarjet procedure to widen the gle-
noid track to convert off-track lesion to on-track lesion. I prefer to use Latarjet
procedure to avoid limited range of motion caused by remplissage. However, in
patients who do not want to have anatomy-changing operation like the Latarjet, I
select Bankart repair plus remplissage. In those with off-track lesion and glenoid
defect ≥25%, the glenoid defect always needs to be fixed. Usually, the Latarjet pro-
cedure is sufficient, but if the Hill–Sachs lesion remains off-track even after the
Latarjet procedure, either adding remplissage or iliac crest bone graft with much
larger bone graft than the coracoid process should be selected.
We analyzed the treatment outcome of 92 patients (81 shoulders with on-track
lesion, 11 shoulders with off-track lesion) treated according to this paradigm [55].
All cases with on-track lesion had glenoid defect <25% so they were treated with
arthroscopic Bankart repair. Among 11 shoulders with off-track lesion, 6 cases had
glenoid defect <25% and 5 cases ≥25%. One patient with glenoid defect <25% was
treated with arthroscopic Bankart repair plus remplissage, and the rest 10 patients
were treated with open Latarjet procedure. The recurrence rate after surgery was 5%
(4/81) in those with on-track lesion and 0% in those with off-track lesion. We were
able to obtain satisfactory results from our treatment strategy. This study supports
the validity of our treatment strategy. It is also supported by recent clinical studies
from other institutes showing that shoulders with off-track lesion showed poor out-
come if treated with arthroscopic Bankart repair.
More recent studies have revealed that there is a subcritical bone loss of the gle-
noid and there are peripheral track and central track within the glenoid track. These
borderline bony lesions need to be considered as critical bone loss and off-track
lesion if the patient belongs to a high-risk group.

4.3.2 Coracoid Transfer

The Latarjet procedure is getting more and more popular these days. Michel
Latarjet was a surgeon in Lyon, and he reported his surgical procedure in Lyon
Chirurgical in 1954. In his original technique, the subscapularis tendon was cut
vertically and the coracoid process was fixed to the glenoid with one screw. Later,
his original procedure was modified to split the subscapularis horizontally instead
of cutting the tendon vertically and to use two screws instead of one to fix the
coracoid process to the glenoid (Fig. 5.37a). The Bristow procedure, which is
128 5 Traumatic Dislocation

a b

Fig. 5.37 Coracoid transfer techniques. (a) Latarjet procedure. The coracoid process is lying and
fixed with two screws. (b) Bristow procedure. The coracoid process is standing on the glenoid and
fixed with one screw

similar to the Latarjet but the coracoid process is standing, was started by a British
orthopedic surgeon, W. Rowley Bristow. Although he did not write an article
about his procedure, his protege, Arthur J. Helfet from South Africa, reported this
procedure 10 years after the death of Bristow. Since Helfet reported this proce-
dure in JBJS-Br in 1958 in memory of his mentor Bristow, this procedure has been
called the Bristow procedure with various modifications. The main difference
between the Bristow procedure and the Latarjet procedure is that the coracoid
process is lying in the Latarjet, but it is standing in the Bristow (Fig. 5.37b).
Sometimes, the term “Bristow-­Latarjet procedure” is used to refer to the coracoid
transfer with its modifications.

From Bristow to Latarjet

The Bristow procedure was widely used in Japan with excellent outcomes. However,
in the USA, it started to be replaced by the Bankart repair in 1990s because of high
complication rate related to screws. It was even called an “abandoned operation.” In
1990, when I first visited Mayo Clinic, the Bristow procedure was very popular in
Japan with low complication rates. Thus, stabilization mechanism of the Bristow
procedure was one of my research topics I had planned to do at Mayo. However,
during the research meeting at the biomechanics laboratory, Bernard F. Morrey, the
4 Treatment 129

Chair of the Department of Orthopedics, commented on this research topic. He had

worked at NASA as an aerospace engineer for 2 years before he attended medical
school at the University of Texas Medical Branch. Because of his career, he was the
consultant of the biomechanics laboratory and attended research meeting every
week. He said that since the Bristow procedure was an abandoned operation in the
USA, nobody would care about the research on this procedure. His comment dis-
suaded me from this research topic. Later, however, glenoid bone loss became the
main topic of interest because the Bankart repair could not provide sufficient stabil-
ity in cases with a large glenoid bone loss. The Latarjet procedure had been widely
performed in France regardless of bone loss, and its outcome was excellent even in
cases with a large glenoid bone loss. As a result, the Latarjet procedure, which was
similar to the “abandoned” Bristow procedure, received the center stage spotlight.
Many doctors from around the world visited France to learn the Latarjet procedure,
which promoted widespread use of this procedure. When I visited Gilles Walch in
1994 as a traveling fellow from the Japan Shoulder Society to the European Society
for Surgery of the Shoulder and Elbow, he showed me the Latarjet. I still have a
vivid recollection of his performing the Latarjet with skillful and beautiful tech-
nique. These days, off-track lesions are mainly treated with the Latarjet procedure
(Fig. 5.38).
As mentioned above, I could not perform research on the stabilization mecha-
nism of the Bristow procedure, but later in 2007, my junior colleague, Nobuyuki
Yamamoto, visited Mayo as a research fellow and completed the project not on the
“abandoned” Bristow but on the “spotlighted” Latarjet.

Fig. 5.38 Latarjet

procedure. The coracoid
process (CP) is fixed to a
bone loss of the glenoid
(G). The conjoint tendon
(CT) originates from the
tip of the coracoid process.
Three suture anchors are
inserted along the anterior
rim of the glenoid for
capsular repair
130 5 Traumatic Dislocation

Stabilizing Mechanism of Latarjet

Didier Patte should take credit for widespread use of Latarjet procedure in France.
He speculated that there were three mechanisms of stabilization by the Latarjet: (1)
transferred coracoid process (bone block effect), (2) the conjoint tendon and the
subscapularis (muscle effect or sling effect), and (3) the capsule (ligament effect).
He called these stabilizing mechanisms the “triple blocking” effect. This “triple
blocking” effect was only his speculation, not proved by him. We performed a
biomechanical study to clarify the stabilizing mechanism of Latarjet procedure.
Using cadaveric shoulders, a 6-mm glenoid defect was created, the coracoid pro-
cess was transferred and fixed to this defect, and the capsule was repaired to the
remnant coracoacromial ligament attached to the coracoid process. Then, stability
was assessed with and without various forces to the subscapularis and conjoint
tendon, followed by stepwise release of the muscles and capsule [56]. We were
able to confirm that there was the “triple blocking” effect as described by Patte.
The contribution of each component relative to the entire stability is summarized
in Table 5.2. The muscle contribution depended on the muscle forces. It was the
least with 0-N muscle force at the end-range of motion, but it increased with an
increase of muscle force up to 62% with a 30-N force. The contributions of the
ligament and of the bone block were maximum, 39% and 56% each, at the end-
range of motion without muscle force, and they decreased with an increase of the
muscle force.
We really do not know how much force is produced by the shoulder muscles at
the time of dislocation. In this study, we calculated the force produced by the sub-
scapularis muscle with the arm in abduction and neutral rotation to be 58 N based
on the data from an EMG study [57] and physiological cross-sectional area study
[58]. However, in order to achieve reproducibility and avoid any damage to the
muscle, we selected the force between 0 and 30 N to the subscapularis muscle and
between 0 and 7.5 N to the conjoint tendon. In reality, much greater force may be
produced by the muscles, or the muscles may be overstretched at the time of dislo-
cation and may not be able to produce any contraction force. The contributions of
these components vary with different muscle forces.

Table 5.2 Contribution of each component to stability

Muscle loada Position Muscle (%) Ligament (%) Bone (%)
30 N + 7.5 N Mid-range 62 6 32
End-range 61 16 23
10 N + 2.5 N Mid-range 51 8 41
End-range 60 16 24
0N+0N Mid-range 43 9 48
End-range 5 39 56
Produced from the data in Yamamoto et al. [56]
a
Muscle load refers to the load applied to the subscapularis + the load applied to the conjoint tendon
4 Treatment 131

Capsular Repair in Latarjet

During the Latarjet procedure, the capsule can be repaired in two ways. One method
is to repair the capsule to the remnant coracoacromial ligament attached to the cora-
coid process and the other one is to repair the capsule to the native glenoid using
suture anchors (Fig. 5.39).
French surgeons usually repair the capsule to the coracoid process. In that case,
the coracoid process is located inside the joint, and accordingly there is a risk of
osteoarthritis in the future. On the other hand, if the capsule is repaired to the gle-
noid rim, the capsule attaches more medially than its original site of attachment.
This means the coracoid process is located outside the joint capsule, which lowers
the risk of osteoarthritis, but instead there is a risk of limiting the range of external
rotation. We assessed the effect of these two repair techniques on stability and range
of motion using cadaveric shoulders [59]. As we expected, the glenoid group
(repairing to the glenoid) showed limited range of external rotation by 13° in adduc-
tion and 15° in abduction compared to the coracoid group (repairing to the coracoid
process). There were no limited motions other than external rotation. On the other
hand, mid-range stability was significantly reduced in the coracoid group compared
to the glenoid group although no significant difference was observed in end-range
stability. From these results, it is clear that we have to choose either stability or the
range of motion. It is a difficult decision, but I prefer to repair the capsule to the
glenoid because of the following reasons: (1) achieving stability is the most impor-
tant goal of this surgery, (2) we would like to avoid risks of osteoarthritis, and (3) if
we repair the capsule to the coracoid process, we need to preserve the coracoid
attachment of the coracoacromial ligament, which prevents a wide view of the gle-
noid surface and makes it more difficult to align the coracoid process flush with the
glenoid surface.

a b

Fig. 5.39 Capsular repair techniques. (a) Repairing to the coracoid process (coracoid group). (b)
Repairing to the glenoid rim (glenoid group). (Reproduced with modification from Fig. 1, Itoigawa
et al. [59])
132 5 Traumatic Dislocation

References

1. Tsuchida T, et al. Luxatio erecta of bilateral shoulders. J Shoulder Elb Surg. 2001;10(6):595–7.
2. Rowe CR. Prognosis in dislocations of the shoulder. J Bone Jt Surg Am. 1956;38-a(5):957–77.
3. Kiviluoto O, et al. Immobilization after primary dislocation of the shoulder. Acta Orthop
Scand. 1980;51(6):915–9.
4. Hoelen MA, Burgers AM, Rozing PM. Prognosis of primary anterior shoulder dislocation in
young adults. Arch Orthop Trauma Surg. 1990;110(1):51–4.
5. Itoi E, et al. Immobilization in external rotation after shoulder dislocation reduces the risk of
recurrence. A randomized controlled trial. J Bone Jt Surg Am. 2007;89(10):2124–31.
6. Simonet WT, et al. Incidence of anterior shoulder dislocation in Olmsted County, Minnesota.
Clin Orthop Relat Res. 1984;186:186–91.
7. Zacchilli MA, Owens BD. Epidemiology of shoulder dislocations presenting to emergency
departments in the United States. J Bone Jt Surg Am. 2010;92(3):542–9.
8. Trojan JD, et al. Epidemiology of shoulder instability injuries in collision collegiate sports
from 2009 to 2014. Arthroscopy. 2020;36(1):36–43.
9. Montgomery C, et al. Video analysis of shoulder dislocations in Rugby: insights into the dis-
locating mechanisms. Am J Sports Med. 2019;47(14):3469–75.
10. Owens BD, et al. Pathoanatomy of first-time, traumatic, anterior glenohumeral subluxation
events. J Bone Jt Surg Am. 2010;92(7):1605–11.
11. Taylor DC, Arciero RA. Pathologic changes associated with shoulder dislocations. Arthroscopic
and physical examination findings in first-time, traumatic anterior dislocations. Am J Sports
Med. 1997;25(3):306–11.
12. Urayama M, et al. Capsular elongation in shoulders with recurrent anterior disloca-
tion. Quantitative assessment with magnetic resonance arthrography. Am J Sports Med.
2003;31(1):64–7.
13. Kurokawa D, et al. The prevalence of a large Hill–Sachs lesion that needs to be treated. J
Shoulder Elb Surg. 2013;22(9):1285–9.
14. Itoi E, et al. The effect of a glenoid defect on anteroinferior stability of the shoulder after
Bankart repair: a cadaveric study. J Bone Jt Surg Am. 2000;82(1):35–46.
15. Montgomery WH Jr, et al. Anteroinferior bone-grafting can restore stability in osseous glenoid
defects. J Bone Jt Surg Am. 2005;87(9):1972–7.
16. Kurokawa D, et al. Anterior tilt of the glenoid relative to the trunk: measurement using 3D-CT
images. In: Presented at 39th annual meeting. Tokyo: Japan Shoulder Society; 2012.
17. Saito H, et al. Location of the glenoid defect in shoulders with recurrent anterior dislocation.
Am J Sports Med. 2005;33(6):889–93.
18. Yamamoto N, et al. Effect of an anterior glenoid defect on anterior shoulder stability: a cadav-
eric study. Am J Sports Med. 2009;37(5):949–54.
19. Yamamoto N, et al. Stabilizing mechanism in bone-grafting of a large glenoid defect. J Bone
Jt Surg Am. 2010;92(11):2059–66.
20. Burkhart SS, De Beer JF. Traumatic glenohumeral bone defects and their relationship to fail-
ure of arthroscopic Bankart repairs: significance of the inverted-pear glenoid and the humeral
engaging Hill–Sachs lesion. Arthroscopy. 2000;16(7):677–94.
21. Lo IK, Parten PM, Burkhart SS. The inverted pear glenoid: an indicator of significant glenoid
bone loss. Arthroscopy. 2004;20(2):169–74.
22. Itoi E. ‘On-track’ and ‘off-track’ shoulder lesions. EFORT Open Rev. 2017;2(8):343–51.
23. Shaha JS, et al. Redefining “critical” bone loss in shoulder instability: functional outcomes
worsen with “subcritical” bone loss. Am J Sports Med. 2015;43(7):1719–25.
24. Yamamoto N, et al. Effect of subcritical glenoid bone loss on activities of daily living in
patients with anterior shoulder instability. Orthop Traumatol Surg Res. 2019;105(8):1467–70.
25. Malgaigne JF. Traité des fractures et luxations/Atlas. Paris: JB Baillière; 1855.
26. Hill HA, Sachs MD. The grooved defect of the humeral head: a frequently unrecognized com-
plication of dislocations of the shoulder joint. Radiology. 1940;35(6):690–700.
References 133

27. Yamamoto N, et al. Contact between the glenoid and the humeral head in abduction, exter-
nal rotation, and horizontal extension: a new concept of glenoid track. J Shoulder Elb Surg.
2007;16(5):649–56.
28. Di Giacomo G, Itoi E, Burkhart SS. Evolving concept of bipolar bone loss and the Hill–Sachs
lesion: from “engaging/non-engaging” lesion to “on-track/off-track” lesion. Arthroscopy.
2014;30(1):90–8.
29. Omori Y, et al. Measurement of the glenoid track in vivo as investigated by 3-dimensional
motion analysis using open MRI. Am J Sports Med. 2014;42(6):1290–5.
30. Yamamoto N, et al. The relationship between the glenoid track and the range of shoulder
motion: a cadaver study. Orthop Traumatol Surg Res. 2018;104(6):793–6.
31. Kawakami J, et al. In vivo glenoid track width can be better predicted with the use of shoulder
horizontal extension angle. Am J Sports Med. 2019;47(4):922–7.
32. Yamamoto N, et al. Peripheral-track and central-track Hill–Sachs lesions: a new concept of
assessing an on-track lesion. Am J Sports Med. 2020;48(1):33–8.
33. Itoi E. Opinion editorial—first-time shoulder dislocation: my approach. In: Matsen FA, et al.,
editors. Rockwood and Matsen’s the shoulder. 16th ed. Philadelphia: Elsevier; 2022. p. 635–8.
34. Kawakami J, et al. In which arm position is a Hill–Sachs lesion created? Am J Sports Med.
2019;47(10):2464–8.
35. Etoh T, et al. Mechanism and patterns of bone loss in patients with anterior shoulder disloca-
tion. J Shoulder Elb Surg. 2020;29(10):1974–80.
36. Hippocrates. On the articulations, the genuine works of Hippocrates. Clin Orthop Relat Res.
2002;400:19–25.
37. Rowe CR, Sakellarides HT. Factors related to recurrences of anterior dislocations of the shoul-
der. Clin Orthop. 1961;20:40–8.
38. Wasserstein DN, et al. The true recurrence rate and factors predicting recurrent instability
after nonsurgical management of traumatic primary anterior shoulder dislocation: a systematic
review. Arthroscopy. 2016;32(12):2616–25.
39. Abe H, et al. Healing processes of the glenoid labral lesion in a rabbit model of shoulder dis-
location. Tohoku J Exp Med. 2012;228(2):103–8.
40. Kuberakani K, et al. Comparison of best-fit circle versus contralateral comparison methods to
quantify glenoid bone defect. J Shoulder Elb Surg. 2020;29(3):502–7.
41. Itoi E. Chapter 27: The shoulder. In: Itoi E, Yoshikawa H, Tsumura H, Tanaka S, Takagi M,
editors. Standard textbook series: Orthopedics. 14th ed. Tokyo: Igaku-Shoin; 2020. p. 426–50.
42. Walter WR, et al. Imaging quantification of glenoid bone loss in patients with glenohumeral
instability: a systematic review. AJR Am J Roentgenol. 2019;212:1096–105.
43. Itoi E, et al. Position of immobilization after dislocation of the shoulder. A cadaveric study. J
Bone Jt Surg Am. 1999;81(3):385–90.
44. Itoi E, et al. Position of immobilization after dislocation of the glenohumeral joint. A study
with use of magnetic resonance imaging. J Bone Jt Surg Am. 2001;83-a(5):661–7.
45. Itoi E, et al. Is protecting the healing ligament beneficial after immobilization in external rota-
tion for an initial shoulder dislocation? Am J Sports Med. 2013;41(5):1126–32.
46. Hatta T, et al. Comfort and acceptability of various immobilization positions using a shoulder
external rotation and abduction brace. J Orthop Sci. 2017;22(2):285–8.
47. Hatta T, et al. Immobilizing performances, comfort, and user-friendliness of the shoulder
abduction-external rotation braces. Ups J Med Sci. 2013;118(2):105–10.
48. Shinagawa K, et al. Immobilization in external rotation reduces the risk of recur-
rence after primary anterior shoulder dislocation: a meta-analysis. Orthop J Sports Med.
2020;8(6):2325967120925694.
49. Itoi E, et al. Range of motion after Bankart repair. Vertical compared with horizontal capsu-
lotomy. Am J Sports Med. 2001;29(4):441–5.
50. Yamamoto N, et al. Does the “bumper” created during Bankart repair contribute to shoulder
stability? J Shoulder Elb Surg. 2013;22(6):828–34.
134 5 Traumatic Dislocation

51. Liu T, et al. Curved-guide system is useful in achieving optimized trajectory for the most infe-
rior suture anchor during arthroscopic Bankart repair. J Shoulder Elb Surg. 2019;28(9):1692–8.
52. Itoigawa Y, et al. Attachment of the anteroinferior glenohumeral ligament-labrum complex to
the glenoid: an anatomic study. Arthroscopy. 2012;28(11):1628–33.
53. Nagaya LH, et al. Does glenoid remodeling occur with an erosion-type bone loss after
arthroscopic Bankart repair? JSES Int. 2020;4(4):814–7.
54. Yamamoto N, et al. Outcome of Bankart repair in contact versus non-contact athletes. Orthop
Traumatol Surg Res. 2015;101(4):415–9.
55. Hatta T, et al. Surgical decision making based on the on-track/off-track concept for anterior
shoulder instability: a case-control study. JSES Open Access. 2019;3(1):25–8.
56. Yamamoto N, et al. The stabilizing mechanism of the Latarjet procedure: a cadaveric study. J
Bone Jt Surg Am. 2013;95(15):1390–7.
57. Kronberg M, Nemeth G, Brostrom LA. Muscle activity and coordination in the normal shoul-
der. An electromyographic study. Clin Orthop Relat Res. 1990;257:76–85.
58. Bassett RW, et al. Glenohumeral muscle force and moment mechanics in a position of shoulder
instability. J Biomech. 1990;23(5):405–15.
59. Itoigawa Y, et al. Repairing the capsule to the transferred coracoid preserves external rotation
in the modified Latarjet procedure. J Bone Jt Surg Am. 2016;98(17):1484–9.
Chapter 6
Multidirectional Instability (Loose
Shoulder)

1 Pathophysiology

Multidirectional instability (MDI) was a new concept of shoulder disease, first

reported by Hisao Endo in 1971 [1]. He reported this disease using the German term
“sog. Schulterschlottergelenk,” which means the so-called loose shoulder. His
patients had no history of trauma, had hyperlaxity of the shoulder, and showed infe-
rior dislocation or subluxation in the hanging arm position at rest. In addition to dull
pain at rest, they often complained motion pain of the shoulder and a sense of fatigue
around the shoulder girdle. Due to pain, it was difficult for them to elevate the arm.
He performed motion analysis of 8 patients and found that upward rotation of the
scapula was disturbed during the early phase of arm elevation, which made them dif-
ficult to elevate the arm. By manually supporting the scapula to rotate upward, infe-
rior instability disappeared, and they were able to fully elevate the arm. He thought
this scapular dyskinesis was the essential pathophysiology of this disease and per-
formed the pectoralis major transfer to the inferior angle of the scapula to promote
upward rotation of the scapula in one patient after failed Bankart repair. Four months
after the surgery, this patient showed no inferior subluxation at rest and could fully
elevate the arm without pain or difficulty. This report was published in the Central
Japan Journal of Orthopaedic Surgery & Traumatology in 1971. It was a short report,
but I was very impressed by the logical and organized contents of this paper. I still
recollect a vivid sensation of excitement when I first read this paper. He searched for
the why (pathophysiology) of this disease, performed a surgery to correct it, and
reported the outcome of his surgery. This was one of the best articles I have ever
encountered. Most unfortunately, however, this paper was written in Japanese. No
one outside of Japan had paid attention to this paper. In 1976, Charles Neer was
invited to the third Annual Meeting of the Japan Shoulder Society and had a chance
to listen to Endo’s presentation for the first time. Neer seemed to be very interested
in and stimulated by Endo’s presentation. Four years later, he published his famous
paper about MDI in JBJS-Am in 1980. This was exactly what Endo reported in 1971.

© The Author(s), under exclusive license to Springer Nature Singapore Pte 135
Ltd. 2023
E. Itoi, Shoulderology, https://doi.org/10.1007/978-981-99-0345-0_6
136 6 Multidirectional Instability (Loose Shoulder)

In Neer’s paper, he referred to Endo’s presentation at JSS meeting, but not the paper
in 1971. Therefore, most people who read his paper believed that Neer was the first
to report the concept of MDI. This was the reason why we wanted to introduce
Endo’s work to the world. I sent his original article with English translation to
William J. Mallon, Editor of JSES, saying as follows. “There are many excellent
papers in the world but most of them are embedded in its own country because it is
written in its own language. People have no access to them unless they were written
in English. As the top journal in this field, I think JSES should have a role to intro-
duce these excellent papers to the world as ‘Classic Articles’.” Bill Mallon and
Jonathan B. Ticker, Editor of compilation of history of American Shoulder and Elbow
Surgeons, both agreed to publish this paper. In 2012, Endo’s paper was introduced to
the world as the first “Classic Article” in JSES [2]. Our dream has come true. Forty-
one years have passed since his original publication. We feel sorry that Dr. Endo
passed away just 1 year before his work appeared in JSES. We believe that his con-
tribution would be remembered by all the shoulder surgeons who read his paper.
The pathophysiology of MDI can be summarized into dysfunction of the mid-­
range stabilizers, i.e., the intra-articular pressure and the concavity-compression
effect. Patients have hyperlaxity of the shoulder with increased range of motion.
The joint cavity is expanded with the thin capsule surrounding the joint, which is
not suitable to create sufficient negative value of intra-articular pressure (Figs. 6.1
and 6.2). This makes the humeral head easily shift downward, causing an inferior
subluxation or even dislocation in the hanging arm position. Patients have muscle

Fig. 6.1 Arthrogram.

Iodinated contrast agent
(looks white on plain
X-ray) is in the joint cavity,
showing a large extent of
the joint cavity
2 Diagnosis 137

a b

Fig. 6.2 MR arthrogram. T1-weighted fat-saturated coronal oblique image (a) and T1-weighted
fat saturated sagittal oblique image (b). The axillary pouch is expanded (arrows) and the rotator
interval is wide open (arrowheads)

imbalance of relatively weak internal rotators, which causes scapular dyskinesis.

The glenoid is sometimes dysplastic, which causes less efficient concavity-­
compression effect. The clinical pictures are formed by various combinations of
these pathologies.

2 Diagnosis

2.1 Physical Findings

Patients have hyperlaxity of the shoulder, but usually not in other joints. It is less com-
mon to have generalized joint laxity. In the affected shoulder, the scapula is rotated
downward, and the humeral head is often subluxated inferiorly. In severe cases, the
humeral head is dislocated inferiorly at rest. Downward pull of the arm reveals a posi-
tive sulcus sign. Anterior and posterior drawer tests show a great laxity in each direc-
tion, so does the abduction inferior stability (ABIS) test. Hyperabduction test is also
positive due to the enlarged joint capsule. Refer to Chap. 2 for further details.

2.2 Imaging

2.2.1 Plain X-rays

On anteroposterior view, the scapula may be rotated downward, and the humeral head
may be inferiorly displaced. Traction X-ray taken with a 3-kg weight around the wrist
clearly shows an inferior subluxation of the humeral head (Fig. 6.3). Severity of the dis-
ease is assessed by the displacement of the humeral head on traction X-ray.
138 6 Multidirectional Instability (Loose Shoulder)

a b

Fig. 6.3 Plain X-ray with downward traction. (a) No traction. (b) 3-kg weight applied to the arm
(the humeral head is subluxated inferiorly)

If there is a dysplasia of the posteroinferior rim of the glenoid or an increased

retroversion of the glenoid, the humeral head slips posteroinferiorly with the arm in
elevation. It is called a “slipping,” which indicates bony abnormalities such as dys-
plasia or increased retroversion (Fig. 6.4).

2.2.2 CT, MRI

CT is suitable for detecting bony abnormalities such as glenoid dysplasia and

increased retroversion of the glenoid. MRI is good at depicting the soft tissues.
Plain MRI may be able to visualize the extent of axillary pouch, but MR arthrog-
raphy more clearly and consistently demonstrates the extent of the enlarged joint
cavity (Fig. 6.2). An MRI study by Inui et al. clearly showed the slipping of the
humeral head with the arm in elevation [3]. MRI taken with the arm in elevation
showed that the humeral head was well centered within the glenoid in normal
shoulders, but it was scattered and shifted posteroinferiorly in MDI shoulders. In
some shoulders, the humeral head was completely dislocated posteriorly or infe-
riorly (Fig. 6.5).
2 Diagnosis 139

Fig. 6.4 Slipping of the

humeral head. The humeral
head is subluxated in the
posteroinferior direction
with the arm in full
elevation. This is called a
“slipping”

Fig. 6.5 Center of the

humeral head with the arm
in elevation. The centers of –0.6 –0.4 –0.2 0 0.2 0.4
the humeral head in normal
shoulders (yellow circles)
were all inside the glenoid
socket, whereas those in
MDI patients (red circles)
shifted posteriorly and
inferiorly, making most of
them outside of the glenoid
socket. (Reproduced with
modification from Fig. 5,
Inui et al. [3])
140 6 Multidirectional Instability (Loose Shoulder)

3 Treatment

3.1 Conservative Treatment

Since there is muscle imbalance, it is important to strengthen the weak muscles.

Rehabilitation exercise should be strengthened step by step. For example, for
strengthening the parascapular muscles, active motion of the scapular abduction/
adduction, upward rotation/downward rotation, internal rotation/external rotation,
and elevation/depression is the first step, followed by push-up against the wall in
standing straight, push-up against the desk, and push-up against the floor as the final
step. Strengthening of the deltoid and the cuff muscles are also important. A resis-
tance band can be used by increasing the resistance stepwise. In approximately 80%
of the patients with MDI, symptoms improve after 3–6 months of rehabilitation. A
scapular brace to keep the scapula in upward rotation is also useful. A combination
of rehabilitation and the scapular brace is reported to be effective in 85% of the
patients.

3.2 Mechanism of Capsular Plication and Its

Clinical Outcome

In recalcitrant cases to conservative treatment, we need to select suitable surgi-

cal options to fix the pathologies each patient has. The most common pathol-
ogy is an enlarged joint cavity surrounded by the thin joint capsule. This
pathology can be corrected by capsular plication, first reported by Neer as
inferior capsular shift. If the main direction of instability is inferior or antero-
inferior, the anterior capsule is exposed through an anterior approach, the cap-
sule is opened in T-shape, the superior flap is shifted downward and the inferior
flap is shifted upward to reduce the joint volume and increase the thickness of
the anterior capsule (Fig. 6.6). Both reducing the joint volume and increasing
the capsular thickness are effective in creating a greater negative value of intra-
articular pressure (see Chap. 4 for detail) (Fig. 6.7). If the main direction of
instability is posteroinferior, the posterior capsule is shifted the same manner
through the posterior approach.
Both open and arthroscopic procedures are available, and many clinical studies
have shown that capsular shift procedure is effective in 86–94% using an open pro-
cedure and in 94–96% using an arthroscopic procedure. A meta-analysis showed
that there was no difference in patient satisfaction, but less limitation in external
rotation was observed in arthroscopic group [4].
3 Treatment 141

a b

c d

Fig. 6.6 Open capsular shift procedure. (a) T-shaped capsulotomy of the anterior capsule. (b)
Downward shift of the superior capsular flap. (c) Upward shift of the inferior capsular flap. (d)
Completion of the procedure

3.3 Pectoralis Major Transfer

If scapular downward rotation remains after the conservative treatment and a patient
voluntarily subluxates or dislocates the shoulder inferiorly, the pectoralis major
transfer is indicated. The pectoralis major is released from the humeral shaft and
transferred to the inferior angle of the scapula. This muscle pulls the inferior angle
anteriorly to keep the scapula more rotated upward, which is effective in preventing
inferior instability (Fig. 6.8). This is especially effective in patients with volun-
tarism because the pectoralis major, which induces voluntary dislocation before
surgery, is converted to a stabilizer after surgery.
142 6 Multidirectional Instability (Loose Shoulder)

a b

Fig. 6.7 Traction X-rays before and after the capsular shift. (a) Pre-operative traction X-ray
showed an inferior subluxation of the humeral head. The negative value of the intra-articular pres-
sure increased by 8% with 3-kg weight. (b) After the capsular shift procedure, traction X-ray
showed no inferior translation of the humeral head. The negative value of the intra-articular pres-
sure increased by 72% after the surgery, which showed that reducing the joint volume was effective
in creating the strong intra-articular pressure to keep the head in the glenoid socket

The pectoralis major transfer is commonly performed when the subscapularis is

irreparable. It is also used in cases with serratus anterior palsy. However, application
of this procedure to MDI is only reported in Japan. According to the clinical reports
from Japan, this procedure is effective in 90–100% of MDI cases.

3.4 Glenoid Osteotomy

Glenoid osteotomy is indicated in cases with glenoid dysplasia or increased retro-

version of the glenoid. These cases can be detected on plain X-rays. A slipping of
the humeral head on X-ray taken with the arm in elevation indicates that there is a
bony abnormality. Patients with a mild dysplasia of the glenoid may respond well to
the conservative treatment. However, if it does not work, glenoid osteotomy is to be
indicated. The posterior capsule and the glenoid neck are exposed through the pos-
terior approach. Before osteotomy, slipping of the humeral head is confirmed with
the arm in elevation. An opening wedge osteotomy is performed at the posterior
glenoid neck 1 cm medial from the joint line using a chisel. Before completion of
3 Treatment 143

a b

Fig. 6.8 Pectoralis major transfer. (a) The pectoralis major (PM) was released from the humeral
shaft. (b) The inferior angle (IA) of the scapula was approached through a vertical skin incision
while the scapula was held in most upward rotated position. The released pectoralis major (PM)
was pulled out to this incision. (c) The pectoralis major (PM) was tightly fixed to the inferior angle
(IA) of the scapula to keep the scapula in upward rotation

the osteotomy, the osteotomy line is opened by elevating the chisel, and the degree
of slipping is once again checked. If there is no more slipping, the amount of open-
ing seems to be sufficient. Finally, the iliac crest bone graft is inserted into the
wedge osteotomy space. Post-surgical CT shows a conversion from retroversion to
anteversion of the glenoid (Fig. 6.9). The posterior capsular imbrication may be
added in cases with severe capsular redundancy.
The clinical outcomes of glenoid osteotomy have been reported from outside of
Japan, and 81–92% of the patients showed excellent to good outcomes after
surgery.
144 6 Multidirectional Instability (Loose Shoulder)

a b

Fig. 6.9 Glenoid osteotomy. (a) 4°-retroversion (yellow arrow) of the glenoid before surgery.
There was a severe dysplasia of the posteroinferior glenoid. (b) After glenoid osteotomy with an
iliac crest bone graft (white arrow), the glenoid showed 17°-anteversion (yellow arrow)

References

1. Endo H, et al. Diagnosis and treatment of Sog Schulterschlottergelenk. Cent Jpn J Orthop Surg
Traumatol. 1971;14(4):630–2.
2. Endo H, et al. Loose shoulder: diagnosis and treatment: [Translated by Eiji Itoi, MD, PhD from
the original paper in Japanese published in central Jpn J Orthop Surg Traumatol 1971;14:630-2].
J Shoulder Elb Surg. 2012;21(12):1782–4.
3. Inui H, et al. Three-dimensional relationship of the glenohumeral joint in the elevated position
in shoulders with multidirectional instability. J Shoulder Elb Surg. 2002;11(5):510–5.
4. Chen D, et al. Effects of surgical management on multidirectional instability of the shoulder: a
meta-analysis. Knee Surg Sports Traumatol Arthrosc. 2016;24(2):630–9.
 Part IV
Main Shoulder Diseases—Others
Chapter 7
Frozen Shoulder

1 Definition and Classification

1.1 Definition

The term “frozen shoulder” may not be familiar to the physicians in Japan. The term
“fifties shoulder” has long been used in Japan to refer to a painful shoulder with an
insidious onset often seen in the 50s. This term was already used in the Edo period
(1603–1868) at the latest because it was in a dictionary of common names
“Rigenshuran” in the Edo period (Fig. 7.1). It says “Those around the age of 50
sometimes have arm or shoulder pain. It soon disappears without medication.
People call it fifties arm, fifties shoulder, or longevity disease.”
Western medicine came into Japan in the Meiji period (1868–1912), and the first
textbook of orthopedic surgery “Jinnaka’s Orthopaedic Surgery” was published in
1939 by Prof. Seiichi Jinnaka. In this textbook, Prof. Jinnaka wrote as follows in the
chapter of fifties shoulder. “So-called fifties shoulder is in the same disease entity as
painful stiff shoulder, frozen shoulder, and periarthritis of the glenohumeral joint.
As pathophysiology is still unknown due to a lack of pathological and anatomical
studies, I would like to use the common name ‘fifties shoulder’ for a while to refer
to this disease.” Prof. Isaharu Miki made a comprehensive report on fifties shoulder
at the 19th Annual Meeting of the Japanese Orthopaedic Association (JOA) in 1946.
He wrote in his book Fifties Shoulder that pathophysiology of the so-called fifties
shoulder needed to be clarified by research studies to find an evidence-based appro-
priate name for this disease because the term “fifties shoulder” was just a common
name. He described the features of the disease as painful stiff shoulder often seen in
the elderly without a trauma and X-ray showed no specific findings [1]. Later, imag-
ing modalities developed, which clarified that fifties shoulder included various
shoulder pathologies such as rotator cuff tendinitis, rotator cuff tear, calcific tendi-
nitis, and tendinitis of the long head of the biceps. After excluding these

© The Author(s), under exclusive license to Springer Nature Singapore Pte 147
Ltd. 2023
E. Itoi, Shoulderology, https://doi.org/10.1007/978-981-99-0345-0_7
148 7 Frozen Shoulder

Fig. 7.1 Edo-period dictionary of common names “Rigenshuran” (held by the National Diet
Library). Red box says “Fifties arm: those around the age of 50 sometimes have arm or shoulder
pain. It soon disappears without medication. People call it fifties arm, fifties shoulder, or longevity
disease”

pathologies, there still remained a certain group of painful stiff shoulders without
known pathologies. The JOA decided to use the term “fifties shoulder” to refer to
this disease entity. In other words, they decided to use “fifties shoulder” as a medical
term. Now that the term “fifties shoulder” started to be used as a common name and
as a medical term, it caused great confusion. Motohiko Mikasa wrote in Personal
View column of the Clinical Journal of Orthopaedic Surgery in 2018, “We have
been in bondage to Prof. Jinnaka’s statement to use the term fifties shoulder. Now
that 80 years have passed, we should be free from this bondage.” Internationally,
there is also a movement in terminology of this disease. Duplay was thought to be
the first to report this disease as péri-arthrite scapulo-humérale in 1872. Later,
Codman used the term “frozen shoulder” in 1934, and Neviaser “adhesive capsuli-
tis” in 1945. In 1993, the American Academy of Orthopaedic Surgeons (AAOS)
defined frozen shoulder as a condition of uncertain etiology characterized by sig-
nificant restriction of both active and passive shoulder motion that occurs in the
absence of a known intrinsic shoulder disorder [2]. In addition, the AAOS classified
frozen shoulder into primary (idiopathic) and secondary divisions; the former is not
identified with any etiology, and the latter associates with known disorders or trauma
1 Definition and Classification 149

[2]. The problem of the AAOS definition and classification is that on the one hand,
frozen shoulder is defined after excluding known shoulder disorders, whereas on the
other hand, frozen shoulder is classified into primary without known shoulder dis-
order and secondary with known shoulder disorders. It is obvious that the definition
and the classification are contradictory to each other.
I have been a member of the ISAKOS Upper Extremity Committee since 2007
(renamed to Shoulder Committee in 2015). I served as a Chair of the Committee
from 2013 to 2015. Every 2 years, we have a closed consensus meeting on a certain
topic. When I was a Chair, I wanted to pick up frozen shoulder as a topic because it
was a mysterious disease with very limited knowledge and evidence. I proposed to
the committee members about my plan, but as I expected, most of them were against
my plan. Rotator cuff tear and shoulder instability seemed to be much more exciting
as a theme of in-depth discussion. I was almost ready to abandon my plan, when one
of the members, Ron Dierks, said, “Eiji, I agree with you. We should do it as a pro-
fessional team because if we don’t do it, who will do it?” His word encouraged me
and eventually we decided to have our consensus meeting in Amsterdam to discuss
all aspects of frozen shoulder in 2014 (Fig. 7.2).
During this meeting, we focused on the contradiction between the definition and
classification of frozen shoulder proposed by the AAOS, and we came to a

Fig. 7.2 Consensus Meeting of the ISAKOS Upper Extremity Committee (Amsterdam 2014).
During the 2-day meeting, we had an in-depth discussion on frozen shoulder and other stiff
shoulders
150 7 Frozen Shoulder

consensus that the term frozen shoulder should be used exclusively for primary
idiopathic stiff shoulder [3, 4]. There is no primary or secondary frozen shoulder
because it is always primary. Furthermore, we proposed which terminology we
should use among all the names given to this disease. Since there is no inflammation
around the shoulder, the term periarthritis is not appropriate. Since there is no adhe-
sion between the capsule and the articular surface of the humeral head, the term
adhesive capsulitis is not appropriate. As neither of these commonly used terms
were appropriate, we concluded that we should use the term frozen shoulder. The
definition of frozen shoulder proposed by the ISAKOS is exactly the same as what
is diagnosed as fifties shoulder in Japan. As mentioned in the next section, we per-
formed questionnaire to the members of the Japan Shoulder Society (JSS), dis-
cussed this issue in the research committee, and decided to adopt the proposal by the
ISAKOS and use the term frozen shoulder instead of fifties shoulder as an official
diagnostic name.

1.2 Classification

There are many classifications of stiff shoulder, but the ISAKOS Upper Extremity
Committee proposed the following classification in 2014 (Fig. 7.3) [4].
According to the ISAKOS classification, stiff shoulder is classified into primary
and secondary, and primary stiff shoulder is exclusively called frozen shoulder. Stiff
shoulders following a trauma or various shoulder disorders are called secondary
stiff shoulders. The secondary stiff shoulders are further divided into four groups
based on the location of pathologies: intra-articular, capsular, extra-articular, and
neurological. As the risk factors of stiff shoulder such as diabetes mellitus, thyroid
disease, and smoking have not been proven to have direct causal relationship with
stiff shoulder, primary stiff shoulders seen in patients with these risk factors should

Stiff Shoulder

Frozen Shoulder
2° Stiff Shoulder
(1° Stiff Shoulder)

Intra-articular Capsular Extra-articular Neurological

Fig. 7.3 ISAKOS classification of stiff shoulder. Stiff shoulder is classified into primary stiff
shoulder and secondary stiff shoulder following a trauma or various shoulder pathologies. The
term frozen shoulder is used exclusively for primary idiopathic stiff shoulder. The secondary stiff
shoulder is further classified to intra-articular, capsular, extra-articular, and neurological based on
the location of the pathology. (Reproduced with modification from Fig. 1.7, Bain et al. [3])
2 Pathophysiology 151

be called the frozen shoulder. There is no primary or secondary frozen shoulder

because frozen shoulder is always primary. Stiff shoulder seen in a patient with
diabetes should be called frozen shoulder if there is neither trauma nor other shoul-
der disorders. This classification clearly separates frozen shoulder from stiff shoul-
ders with other shoulder disorders.
As the ISAKOS classification was released in 2015, the Japan Shoulder Society
(JSS) performed a survey regarding stiff shoulders among the society members [5].
The purposes of this survey were (1) to know the opinions of the JSS members regard-
ing definition and classification of frozen shoulder proposed by the AAOS [6] and (2)
to compare the survey results with those obtained among the members of the American
Shoulder and Elbow Surgeons (ASES). Regarding the definition of frozen shoulder,
67% of JSS members agreed (85% of ASES members). However, less people agreed
about dividing frozen shoulder into primary and secondary (53% agreed in JSS, 82%
agreed in ASES) and about including risk factors into the secondary frozen shoulder
(agreed 54% in JSS, 66% in ASES). Among all the names used to refer to this disease,
frozen shoulder is most commonly used in Japan. These results tell us that JSS mem-
bers have closer idea about the ISAKOS definition and classification than ASES
members do. Based on these results and in-depth discussion, the JSS decided to offi-
cially adopt the ISAKOS definition and classification of stiff shoulder [5].

2 Pathophysiology

2.1 Rat Knee Contracture Model

Animal models are useful in clarifying pathophysiology. Rats are easy to handle,
but in order to create a shoulder contracture, the humerus must be firmly fixed to the
scapula as the shoulder is a ball-and-socket joint with great mobility. However, the
scapula is paper thin, which makes it very difficult to obtain a rigid fixation of the
shoulder. Therefore, we first selected the knee joint, a hinge joint, which was much
easier to immobilize than the shoulder. We created a rat knee contracture model by
immobilizing the rat knee joints at 150° of flexion using a plate and screws. We
analyzed the expression of transforming growth factor (TGF)-β1 and connective
tissue growth factor (CTGF) at 3 days, 1, 2, 4, 8, and 16 weeks after immobilization
[7]. The in situ hybridization demonstrated that the mRNAs of both TGF-β1 and
CTGF increased continuously during the first 2 weeks after immobilization and
then decreased. The response was relatively higher in the posterior capsule than in
the anterior one. In contrast, the immunoreactivity of both TGF-β1 and CTGF
increased gradually with time. The response was much stronger in the posterior
capsule than in the anterior one. We concluded that the capsule had a potency to
produce TGF-β1 and CTGF after immobilization, and these growth factors might
play a role in causing and maintaining capsular fibrosis mainly in the posterior cap-
sule, which might have caused limited motion in extension in this rat knee contrac-
ture model.
152 7 Frozen Shoulder

Next, we analyzed the expression of type I collagen, a major component of the

joint capsule, in a rat knee contracture model [8]. mRNA rapidly decreased after
immobilization, but the immunoreactivity remained unchanged at any time points in
both the immobilized and control groups. We also analyzed the expression of type
III collagen, an important player in wound healing and repair [9]. The expression of
mRNA and protein levels of collagen type III were not increased after immobiliza-
tion. These studies indicated that accumulation of these two types of collagens was
not the etiology of joint contracture.

2.2 Rat Shoulder Contracture Model [10]

After a series of these studies using a rat knee contracture model, we challenged to
make a rat shoulder contracture model. The most difficult part of this study was how
to firmly fix the humerus to the thin, fragile scapula. By holding the scapula between
two plates, we managed to develop a rat shoulder contracture model without break-
ing the scapula. The shoulder was immobilized at 60° of abduction (similar to hang-
ing arm position in human). After 8 weeks of immobilization, we measured the
range of motion. Abduction was 6° in the immobilization group versus 19° in the
control group and total rotation was 17° versus 43°. The shoulder contracture model
was successfully developed. To confirm which structure contributed to the contrac-
ture, we released the soft tissues one by one in the order of (1) the muscles other
than the rotator cuff muscles, (2) the rotator cuff muscles, and (3) the capsule. Both
abduction and external rotation improved significantly after releasing the capsule,
suggesting that the capsular shortening played a role in limiting the range of motion.
In fact, we assessed the capsular length by measuring the length of synovial intima,
which was 2.11 mm in the immobilized group versus 7.10 mm in the control group.
The length of synovial intima in shoulder contracture model was less than one third
of the normal joint. Histologically, the immobilized group showed obliteration of
the axillary pouch and thickening of the capsule (Fig. 7.4).
This rat shoulder contracture model was created by immobilizing the joint. In
other words, it was a secondary stiff shoulder caused by immobilization. However,
the histology was demonstrated to be quite similar to that of frozen shoulder (see the
next section).

2.3 Specimens from Patients with Frozen Shoulder

Now we shift from the shoulder contracture model in rats to frozen shoulder in
humans. We collected samples of the capsule from 12 patients with frozen shoulder
during surgery and compared them with the samples from 16 patients with a rotator
cuff tear without contracture [11]. This study showed that the collagen bundles were
dense with less space between the bundles, and the number of fibroblast-like cells
was significantly higher in frozen shoulders (Fig. 7.5).
2 Pathophysiology 153

a b

c d

Fig. 7.4 Histology of the capsule (hematoxylin–eosin stain). The axillary pouch (arrow) is obliter-
ated, and the capsule is thickened in the immobilized group (a), whereas the axillary pouch is wide
open (arrow) and the capsule is thin in the sham group (b). (a, b: white bar = 500 μm; original
magnification, × 4). At higher magnification, the synovium surface is flat (arrows) in the immobi-
lized group (c), whereas it has many folds (arrows) in the sham group (d). (c, d: white bar = 50 μm;
original magnification, × 40). (Reprinted from Fig. 7, Hagiwara et al. [10])

We also found that the material properties of the capsule had been changed.
Scanning acoustic microscopy (SAM) is a device to assess the tissue elasticity by
measuring the ultrasound speed in the tissue based on the theory that the sound
speed in the tissue is proportional to the tissue elasticity. SAM gradation color
images showed that the sound speed was faster (red to orange in Fig. 7.6a) in the
samples obtained from frozen shoulders, which meant that the tissue elasticity was
high, or the tissue was stiff in frozen shoulders. On the other hand, the samples from
cuff tear shoulders showed slower sound speed (blue in Fig. 7.6b), which meant that
the tissue elasticity was low, or the tissue was soft.
A strong alcian blue staining was observed in the capsule from frozen shoulders,
which indicated that proteoglycans, the major component of cartilage matrix, were
rich in the capsule (Fig. 7.7a). The samples from cuff tear shoulders showed almost
no staining of alcian blue (Fig. 7.7b).
Gene expressions related to fibrosis, inflammation, and chondrogenesis were sig-
nificantly higher in frozen shoulders compared to cuff tear shoulders.
We further performed proteome analysis to elucidate pathogenesis of frozen
shoulder based on protein expression levels in the capsule [12]. Biological
154 7 Frozen Shoulder

a b

Fig. 7.5 Histology of the capsule (hematoxylin–eosin stain). (a) In frozen shoulders, the collagen
bundles were dense with less space between the bundles. The cells were densely populated. (b) In
cuff tear shoulders, the collagen bundles had more spaces between them with sparsely populated
cells. (By courtesy of Associate Professor Yoshihiro Hagiwara, Department of Reconstruction in
Sports Activity and Motor Function, Tohoku University Graduate School of Medicine)

a b

Fig. 7.6 SAM gradation color image of the capsule. The anteroinferior capsule from frozen shoul-
ders (a) shows mainly red to orange color, meaning high sound speed and high tissue elasticity
(stiff capsule). The capsule from cuff tear shoulders (b) shows mainly blue color, which indicates
slow sound speed and low tissue elasticity (soft capsule). (Reprinted from Fig. 1, Hagiwara
et al. [11])

processes related to tissue repair, collagen metabolism and fibrillation, cell–cell and
cell–matrix adhesion, blood coagulation, and immune response were upregulated in
the upper part (rotator interval and MGHL) of the capsule, whereas processes
related to phagocytosis, glutathione metabolism, retinoid metabolism, and lipid
metabolism were downregulated in the lower part (IGHL) of the capsule of frozen
shoulders. Taken together, these results show that the pathophysiology of frozen
shoulder differs between the upper and lower parts of the shoulder. From a series of
these studies, we can say that the main pathophysiology of frozen shoulder is fibro-
sis, inflammation, and chondrogenesis.
3 Diagnosis 155

a b

Fig. 7.7 Alcian blue staining of the capsule. Strong alcian blue staining was observed in the cap-
sule from frozen shoulders (a). The capsule from cuff tear shoulders (b) showed very weak alcian
blue staining. (By courtesy of Associate Professor Yoshihiro Hagiwara, Department of
Reconstruction in Sports Activity and Motor Function, Tohoku University Graduate School of
Medicine)

2.4 Posture and Blood Circulation of the Shoulder

There is a study reporting that physical therapy for patients with frozen shoulder
and kyphoscoliosis resulted in not only recovery of shoulder range of motion but
also improvement of posture. This study led us to think there might be a relationship
between posture and frozen shoulder. As an increase in thoracic kyphosis rotates the
scapula internally, we investigated the relationship between scapular rotation and
blood circulation of the shoulder [13]. We took MRI of 113 healthy volunteers with
the scapula in internal rotation and in external rotation. Blood flow of the anterior
humeral circumflex artery decreased with scapular internal rotation, but no change
was observed in blood flow of the posterior humeral circumflex artery. With age,
thoracic kyphosis increases, and people become more susceptible to decrease in
blood flow to the shoulder due to increased scapular internal rotation. This decrease
in blood flow might have a role in pathophysiology of frozen shoulder.

3 Diagnosis

3.1 Clinical Pictures in Each Phase

Symptoms of frozen shoulder are divided into three phases: freezing phase, frozen
phase, and thawing phase. Freezing phase is the initial phase of frozen shoulder. It
starts with an insidious onset of pain without any trauma. It gradually aggravates.
Initially, the pain is seen only during motion, but later, advances to rest pain and
night pain. In the freezing phase, shoulder motion may aggravate the symptoms, so
no rehabilitation should be prescribed. It lasts for 2–9 months. Then, the pain gradu-
ally decreases with an appearance of limited range of motion. This is the frozen
156 7 Frozen Shoulder

phase. At the maximum of the frozen phase, the range of motion is limited in all the
directions (global motion loss; flexion <100°, external rotation <10°, internal rota-
tion <L5). The frozen phase lasts for 4–12 months. After this phase, the thawing
phase comes with the range of motion gradually recovering to normal. During the
thawing phase, patients complain very little or no pain. It lasts for 5–26 months. In
total, it takes 1–4 years to recover from this disease. Can we expect full recovery? A
long-term follow-up study demonstrated that more than half the patients still had
some pain and/or limited range of motion at 3–11-year follow-up [14]. Thus, if
severely limited range of motion still remains after 6 months of physical therapy,
manipulation under anesthesia or arthroscopic capsular release is recommended.

3.2 Physical Findings

During arm elevation, the humerus and the scapula share the motion with a certain
ratio in normal shoulders. The ratio of scapulohumeral motion to scapulothoracic
motion is approximately 2:1. For example, with the arm in full abduction of 180°,
the humerus moves 120° relative to the scapula and the scapula moves 60° relative
to the trunk. This ratio is called the scapulohumeral rhythm. If the scapula is held in
place to avoid any scapular motion, the humerus can be passively elevated up to
120° (Fig. 7.8).
However, in frozen shoulders, the glenohumeral motion is severely restricted so
that the arm elevation can be achieved almost only by the scapulothoracic motion.
In a frozen shoulder patient with maximum elevation of 80°, holding the scapula to
avoid scapular motion allows the examiner to passively elevate the arm to 20° only
(Fig. 7.9). This tells us that in 80° of arm elevation, 60° was achieved by the scapu-
lar motion, and only 20° was achieved by the glenohumeral motion.
During the frozen phase, the range of motion is limited in all directions (global
motion loss). Diagnostic criteria of global motion loss by the ISAKOS Upper
Extremity Committee are flexion <100°, external rotation <10°, and internal rota-
tion <L5 [4].

a b c

Fig. 7.8 Normal motion of the scapula and humerus. The normal shoulder in full elevation (a). By
stabilizing the scapula in place to avoid any scapular motion (b), the humerus can be passively
elevated up to 120° (c)
3 Diagnosis 157

a b c

Fig. 7.9 Motion of the scapula and humerus in frozen shoulder. Active flexion is limited to 80°
(a). When the scapula is stabilized to avoid its motion (b), passive arm elevation is only 20° (c)

It is relatively easy to make a diagnosis of frozen shoulder during the frozen

phase. However, during the freezing phase, the only symptom is pain without lim-
ited range of motion, and thus, it is difficult to differentiate frozen shoulder from
other painful shoulders. Whichever phase a patient is in, imaging diagnosis is essen-
tial to exclude other pathologies causing a painful shoulder. After differentiating all
other known pathologies, a diagnosis of frozen shoulder can be made.

3.3 Imaging

There are no specific X-ray findings of frozen shoulder. If X-ray shows a calcified
deposit, the diagnosis is calcific tendinitis, and frozen shoulder should be excluded.
An arthrogram shows narrowing of the joint cavity, especially in the axillary pouch,
and obliteration of the tendon sheath of the long head of the biceps and subscapu-
laris bursa (Fig. 7.10). These findings are specific to frozen shoulder.
However, as there is no correlation between these characteristic findings and the
range of motion, these findings cannot be used to predict the severity of contracture
to come [15]. During arthrography, increased intra-articular pressure may open the
obliteration and the contrast agent may come into the tendon sheath of the long head
of the biceps or the subscapularis bursa. Also, during injection of the contrast agent
into the shoulder joint or during motion after injection, the subscapularis bursa may
rupture due to high intra-articular pressure. Since these findings on arthrograms can
also be detected on MRI, MRI has taken the place of arthrography. The only occa-
sion to use arthrography is to perform joint distension procedure as a treatment. The
most characteristic finding on MRI is narrowing of the axillary pouch with thicken-
ing and shortening of the axillary capsule (Fig. 7.11).
In normal shoulders, the subcoracoid fat triangle is clearly seen on sagittal
oblique MR images (Fig. 7.12) [16]. In frozen shoulders, this triangle is completely
(32%) or partially obliterated (45%), and the coracohumeral ligament is often
thicker than 4 mm (Fig. 7.13) [17].
158 7 Frozen Shoulder

Fig. 7.10 Arthrogram of

frozen shoulder. The joint
cavity is narrow with
reduced axillary pouch
(arrow). The tendon sheath
of the long head of the
biceps is not visualized
(white arrowheads). The
subscapularis bursa (black
arrowheads) is open in
this case

a b

Fig. 7.11 MRI of frozen shoulder (T2-weighted coronal oblique). (a) In normal shoulders, the
capsule which surrounds the wide axillary pouch is thin (dotted line). (b) In frozen shoulders, the
axillary pouch is obliterated, and the capsule is thick and shortened (solid line)
3 Diagnosis 159

Fig. 7.12 Subcoracoid fat triangle. On sagittal oblique MR images, a triangle filled with fat is seen
just below the coracoid process. This is called the subcoracoid fat triangle. Borders of this triangle
are defined anterosuperiorly by the coracoid process (C), superiorly by the coracohumeral liga-
ment (black arrowhead), and posteroinferiorly by the rotator interval capsule (white arrowhead).
SSP supraspinatus. (Reproduced with modification from Fig. 1, Kanazawa et al. [16])

Fig. 7.13 Obliteration of

the subcoracoid fat
triangle. In patient with
frozen shoulder, the
subcoracoid fat triangle is
filled with the fibrous
tissue (arrow) and the
coracohumeral ligament is
thickened (arrowheads)
160 7 Frozen Shoulder

4 Treatment

4.1 Conservative Treatment

The mainstay of treatment of frozen shoulders is conservative treatment.

Approximately 95% of patients respond well to conservative treatment with reduced
pain and improved range of motion.

4.1.1 Freezing Phase

The chief complaint during the freezing phase is pain caused by inflammation.
Thus, anti-inflammatory agents such as oral medication of NSAIDs, intra-articular
or intra-bursal injection of steroid or hyaluronic acid are commonly used. Since
motion exercise can aggravate the inflammation during the freezing phase, forceful
motion is a contraindication during this phase. Rather, it is important to keep the
arm in a sling to reduce pain and calm down inflammation. It is a good idea to
advise patients to keep the arm in a comfortable position while sleeping using a pil-
low or bath towel. The anti-inflammatory effect of steroid injection is well known.
Regarding hyaluronic acid injection, there is not sufficient evidence. Using a rat
knee contracture model described above, we compared those with weekly injections
of hyaluronic acid (HA group) versus saline (control group). The HA group showed
the increased synovial membrane length, lower gene expressions of IL-6, IL-1β,
TGF-β, CTGF, COL1a1, COL3a1, SPARC, and HIF1-α, and decreased sound speed
of the capsule compared to the control group. This study demonstrated that HA
injections suppressed inflammatory, fibrotic, and hypoxic conditions observed in
the immobilized joint capsule [18].

4.1.2 Frozen and Thawing Phases

During the frozen and thawing phases, pain gradually subsides, and contracture
instead becomes the chief complaint. In order to speed up the recovery process of
contracture, physical therapy including modalities and exercises are important.
Modalities such as hot pack and microwave diathermy are used to increase the pain
threshold and reduce pain to allow the following exercises easier to perform.
Therapeutic exercise performed by a physical therapist starts with muscle stretching
to reduce muscle stiffness, followed by mobilization of the scapulothoracic joint,
and then the glenohumeral joint. A well-known at-home exercise is the stooping
exercise proposed by Codman. In the standing position with both arms hanging
freely, you bend forward with the arms still hanging freely. The arms stay in almost
vertical position while the trunk is bent forward. The more you bend, the more the
arms are elevated relatively to the trunk. This stooping exercise is a passive eleva-
tion exercise. This exercise is also known as Codman exercise or pendulum
4 Treatment 161

exercise. Shoulder table glides are commonly prescribed as well. While sitting next
to a table with your hand on the table, gently slide your arm forward or sideways
while bending forward or laterally. You may place a towel under your hand and
forearm to reduce friction. Hold the position for 3–5 s and return to the starting posi-
tion. This passive motion of flexion and abduction should be repeated 10–20 times.
You may use a bar to assist your rotational exercise. In the supine position, exter-
nally rotate your arm passively using a bar. In the standing position with your hand
on your back, use the bar to assist internal rotation of the involved shoulder. In
Connolly exercise, you grab something stable at the highest level you can reach and
while grabbing it, lower your body by bending your knees. This exercise passively
elevates your arm. Once a week, you visit the physical therapist to undergo muscle
stretching and passive range of motion exercise. Other than that, you perform exer-
cises at home every day, desirably three to four times a day. It used to be said, “No
pain, no gain.” However, this is not true. One of the ISAKOS Upper Extremity
Committee members, Ron Diercks, performed a randomized clinical trial, compar-
ing two groups of patients with frozen shoulder: those treated with intensive exer-
cise by a physical therapist beyond the pain threshold (32 patients) and those with
gentle at-home exercise within painless range (45 patients). After 2 years of treat-
ment, the gentle exercise group showed better outcomes than the intensive exercise
group [19]. Painful exercise seems to delay the recovery process.

4.1.3 Joint Distension

Joint distension was commonly performed at the time of arthrography when it was
routinely used as a diagnostic modality. When the joint cavity was reduced with
obliteration of the subscapularis bursa and the tendon sheath of the long head of the
biceps, additional injection of saline mixed with local anesthetics was used to
mechanically expand the joint cavity. Joint distension is not as commonly used these
days as it used to be. This is indicated for cases with severely limited range of
motion, who have not responded to conservative treatment. The joint cavity is often
as narrow as 10 mL or less. When the saline with local anesthetics is injected into the
joint up to its limit, the injection resistance increases together with the pain. Within
this range, injection and aspiration are repeated several times. During this procedure,
the subscapularis bursa, the weakest point of the capsule, may rupture and the saline
leaks out of the joint. Suddenly, the pain goes away due to a loss of intra-articular
high pressure. The whole procedure is completed with an injection of steroid. In
short term, this procedure is effective in reducing pain and improving function.
However, high quality evidence is lacking [20]. We still do not know whether the
efficacy of joint distension depends on the amount of saline used, combined with
steroid, or frequency of the procedure. We also do not know the mechanism of joint
distension: mechanical expansion effect or avoidance of high intra-articular pressure
after a rupture of subscapularis bursa. Some reports say that the second joint disten-
sion after 2 weeks of the first procedure does not have an additional effect, or that
there is no difference in the outcomes compared with manipulation under anesthesia.
162 7 Frozen Shoulder

Conservative treatment mentioned above works in 93–95% of the cases. The rest
5–7% are an indication of manipulation under anesthesia or arthroscopic capsular
release. When should this decision be made? There is no evidence, but generally it
is accepted that these interventions should be considered after 6–12 months of failed
conservative treatment. According to our study, the risk factors of poor response to
conservative treatment were comorbidity of diabetes and severe contracture (<90°
of flexion, <0° of external rotation) at the first visit [21].

4.2 Manipulation Under Anesthesia

This is a procedure to passively move the shoulder under interscalene block or gen-
eral anesthesia. With this procedure, the stiff and thickened joint capsule which has
restricted the shoulder motion is torn and the motion regains. You might want to
avoid this procedure during the freezing phase. If you use the interscalene block,
this procedure can be done at the outpatient clinic. The recurrence rate of contrac-
ture is 5–20%. Complications are reported in 3–5% of the cases, such as fracture,
dislocation, brachial plexus palsy, rotator cuff tear, and tendon rupture of the long
head of the biceps. These complications are more commonly seen when the elbow
is held during the manipulation procedure using a long lever arm or when passive
motion is forced with relatively high speed. To avoid these complications, (1) make
the lever arm as short as possible (hold the humerus close to the shoulder or hold the
wrist so that the elbow can function as a force absorber) and (2) move the arm as
gentle and slow as possible. Following these instructions, the complication rate can
be reduced as low as less than 1%.
The manipulation is usually performed in the following order: (1) flexion in the
supine position to release the anteroinferior capsule, (2) external rotation while
slowly depressing the arm to release the entire anterior capsule, (3) horizontal flex-
ion and flexion-internal rotation to release the posteroinferior capsule, and (4)
extension-internal rotation to release the posterosuperior capsule to finish with.
Usually, a typical cracking sound is accompanied with manipulation, indicating a
rupture of the capsule. However, the cracking sound may not always be accompa-
nied. Don’t manipulate the shoulder until you hear a cracking. We should be aware
that there is no difference in the efficacy of manipulation with and without the
cracking sound.

4.3 Surgical Treatment

Arthroscopic capsular release is indicated to a recalcitrant frozen shoulder. Under

general anesthesia, the entire capsule is released around the clock. Different from
manipulation under anesthesia, we can release the capsule alone without damaging
4 Treatment 163

a b

Fig. 7.14 Arthroscopic capsular release. (a) Viewing through the anterior portal, the posteroinfe-
rior capsule is released. (b) The axillary capsule is completely released, and the red muscle belly
of the subscapularis is clearly seen

other structures. Thus, it is safer and effective, but disadvantages are it requires
admission to the hospital, general anesthesia, and surgery which is more invasive
than the manipulation. It is up to the surgeon’s discretion which procedure to choose.
I prefer to perform arthroscopic capsular release. Viewing through the standard pos-
terior portal, the rotator interval capsule, anterior to anteroinferior capsule, and the
coracohumeral ligament are released with use of a punch and electrocautery through
the anterior portal. Complete release of the coracohumeral ligament from the lateral
base of the coracoid process is important to gain motion. Next, switching to the
anterior viewing portal, the entire posterior capsule is released to complete the pro-
cedure (Fig. 7.14).
Make sure that the range of motion is well restored after the capsular release
(Fig. 7.15).
Postoperatively, the arm is held in full elevation with 1-kg traction while in bed,
whereas the arm is in a sling when out of bed. Encourage the patient to use the arm
as soon as possible within the comfortable range. Although the full range of motion
is achieved during the surgery, postoperative rehabilitation is important to avoid
recurrence of contracture. On postop day 1, we start muscle stretching and range of
motion exercise by a physical therapist, and passive range of motion exercise by the
patients using a pulley and a bar. Once the patients can perform all the exercises by
themselves, they are discharged from the hospital. Arthroscopic capsular release
can provide early recovery of the range of motion.
We reported that the thicker the coracohumeral ligament, the severer the contrac-
ture [16]. We also reported that after releasing the entire capsule, the range of motion
further increased by adding a release of the coracohumeral ligament [22]. Based on
these reports, we have emphasized the importance of releasing the coracohumeral
ligament together with the entire capsule [23].
164 7 Frozen Shoulder

a b

c d

e f

Fig. 7.15 Range of motion before and after the capsular release. Flexion: 80° preop (a) to 170°
postop (b) external rotation in adduction: 10° preop (c) to 60° postop (d) external rotation in
abduction: 30° preop (e) to 90° postop (f) internal rotation in abduction: 40° preop (g) to 80°
postop (h)
References 165

g h

Fig. 7.15 (continued)

There are many reports on the outcomes of manipulation under anesthesia and
arthroscopic capsular release. A meta-analysis showed that there was little differ-
ence in clinical outcomes between these two procedures [24]. However, the quality
of evidence available is low. High-quality studies are required to definitively evalu-
ate the relative benefits of these procedures.

References

1. Miki I. Fifties shoulder. Tokyo: Nihon Igaku Shuppan; 1947.

2. Zuckerman JD, Cuomo FC. Frozen shoulder. In: Matsen 3rd FA, Fu FH, Hawkins RJ, editors.
The shoulder: a balance of mobility and stability. Rosemont: AAOS; 1993. p. 253–67.
3. Bain G, Clitherow HDS. The pathogenesis and classification of shoulder stiffness. In: Itoi E,
et al., editors. Shoulder stiffness: current concepts and concerns. Heidelberg: Springer; 2015.
p. 3–19.
4. Itoi E, et al. Shoulder stiffness: current concepts and concerns. Arthroscopy.
2016;32(7):1402–14.
5. Kobayashi T, et al. Representative survey of frozen shoulder questionnaire responses from
the Japan Shoulder Society: what are the appropriate diagnostic terms for primary idiopathic
frozen shoulder, stiff shoulder or frozen shoulder? J Orthop Sci. 2019;24(4):631–5.
6. Zuckerman JD, Rokito A. Frozen shoulder: a consensus definition. J Shoulder Elb Surg.
2011;20(2):322–5.
7. Hagiwara Y, et al. Expression of transforming growth factor-beta1 and connective tis-
sue growth factor in the capsule in a rat immobilized knee model. Ups J Med Sci.
2008;113(2):221–34.
8. Hagiwara Y, et al. Expression of type I collagen in the capsule of a contracture knee in a rat
model. Ups J Med Sci. 2007;112(3):356–65.
9. Hagiwara Y, et al. Expression patterns of collagen types I and III in the capsule of a rat knee
contracture model. J Orthop Res. 2010;28(3):315–21.
10. Kanno A, Sano H, Itoi E. Development of a shoulder contracture model in rats. J Shoulder Elb
Surg. 2010;19(5):700–8.
11. Hagiwara Y, et al. Coexistence of fibrotic and chondrogenic process in the capsule of idio-
pathic frozen shoulders. Osteoarthr Cartil. 2012;20(3):241–9.
166 7 Frozen Shoulder

12. Hagiwara Y, et al. Comparative proteome analysis of the capsule from patients with frozen
shoulder. J Shoulder Elb Surg. 2018;27(10):1770–8.
13. Hagiwara Y, et al. Blood flow changes of the anterior humeral circumflex artery decrease with
the scapula in internal rotation. Knee Surg Sports Traumatol Arthrosc. 2015;23(5):1467–72.
14. Shaffer B, Tibone JE, Kerlan RK. Frozen shoulder. A long-term follow-up. J Bone Jt Surg Am.
1992;74(5):738–46.
15. Itoi E, Tabata S. Range of motion and arthrography in the frozen shoulder. J Shoulder Elb Surg.
1992;1(2):106–12.
16. Kanazawa K, et al. Correlations of coracohumeral ligament and range of motion restriction
in patients with recurrent anterior glenohumeral instability evaluated by magnetic resonance
arthrography. J Shoulder Elb Surg. 2017;26(2):233–40.
17. Mengiardi B, et al. Frozen shoulder: MR arthrographic findings. Radiology.
2004;233(2):486–92.
18. Kanazawa K, et al. Preventing effects of joint contracture by high molecular weight hyaluro-
nan injections in a rat immobilized knee model. Int J Clin Exp Pathol. 2015;8(4):3426–40.
19. Diercks RL, Stevens M. Gentle thawing of the frozen shoulder: a prospective study of super-
vised neglect versus intensive physical therapy in 77 patients with frozen shoulder syndrome
followed up for 2 years. J Shoulder Elb Surg. 2004;13(5):499–502.
20. Buchbinder R, et al. Arthrographic distension for adhesive capsulitis (frozen shoulder).
Cochrane Database Syst Rev. 2008;1:Cd007005.
21. Ando A, et al. Identification of prognostic factors for the nonoperative treatment of stiff shoul-
der. Int Orthop. 2013;37(5):859–64.
22. Hagiwara Y, et al. Effects of arthroscopic coracohumeral ligament release on range of motion
for patients with frozen shoulder. Open Orthop J. 2018;12:373–9.
23. Hagiwara Y, et al. Effects of joint capsular release on range of motion in patients with frozen
shoulder. J Shoulder Elb Surg. 2020;29(9):1836–42.
24. Grant JA, et al. Comparison of manipulation and arthroscopic capsular release for adhesive
capsulitis: a systematic review. J Shoulder Elb Surg. 2013;22(8):1135–45.
Chapter 8
Long Head of the Biceps

1 Anatomy and Function

1.1 Anatomy

The biceps brachii muscle has two heads: the long head which originates from the
supraglenoid tubercle and the short head which originates from the coracoid pro-
cess. Both heads merge to form a distal muscle belly and the main portion inserts to
the radial tuberosity of the radius and the rest to the deep fascia of the forearm called
bicipital aponeurosis. The biceps brachii is a two-joint muscle crossing the shoulder
and the elbow. The long head of the biceps (LHB) tendon runs inside the shoulder
joint in contact with the humeral head. Its proximity to other intra-articular struc-
tures causes inflammation and rupture combined with other shoulder disorders like
a rotator cuff tear. It is also said to be related to pathogenesis of SLAP lesion seen
in overhead athletes.
The LHB tendon with its origin at the supraglenoid tubercle runs inside the joint
over the humeral head and exits the joint through the bicipital groove. As the loca-
tion of the bicipital groove changes with the rotation of the arm, its anatomical
relationship to the humeral head and the glenoid also changes with arm rotation
(Fig. 8.1). It is important to realize this anatomical relationship to understand the
stabilizing function of the LHB.

1.1.1 Widening of the LHB Tendon in Cuff Tear Patients

It is known that the ratio of cross-sectional areas of a muscle and its tendon is con-
stant [1]. The muscle increases its volume with exercise, but at the same time, the
tendon increases its volume as well. If the muscle becomes hypertrophic without an
increase in tendon size, the muscle strength may cause a rupture of its own tendon.

© The Author(s), under exclusive license to Springer Nature Singapore Pte 167
Ltd. 2023
E. Itoi, Shoulderology, https://doi.org/10.1007/978-981-99-0345-0_8
168 8 Long Head of the Biceps

a b

Fig. 8.1 The long head of the biceps (LHB) tendon and the humeral head. (a) With the arm in
external rotation (ER), the bicipital groove (*) is located opposite to the glenoid (G), and the LHB
tendon (L) runs perpendicular to the glenoid, dividing the humeral head into almost the same size.
(b) With the arm in internal rotation (IR), the bicipital groove (*) comes anteriorly, and the LHB
tendon (L) grips only the anterior portion of the humeral head

To void this catastrophe, hypertrophy of the tendon proportional to that of the mus-
cle occurs. You may call it physiological adaptation. In shoulders with a rotator cuff
tear, we often find that the LHB tendon is widened. Does this hypertrophy of the
biceps tendon accompany that of the biceps muscle? Or is this a localized widening
of the tendon independent of the muscle volume? There were no answers to this
question. Therefore, we measured the cross-sectional areas of the biceps muscle and
biceps tendon in 14 embalmed shoulders [2]. There were seven shoulders with a
rotator cuff tear and seven shoulders with intact rotator cuff. The intra-articular por-
tion of the LHB tendon was cut into four segments, and the portion in the bicipital
groove was cut into five segments. We measured the cross-sectional area of these
tendon segments. We also measured the physiological cross-sectional area of the
biceps muscle by dividing the muscle volume by the length of the muscle fibers
(Fig. 8.2). In normal shoulders, the biceps tendon was the widest at its origin at the
glenoid, gradually decreased its cross-sectional area, and stayed the same size in the
bicipital groove. However, in cuff tear shoulders, the cross-sectional area increased
from its origin to the entrance to the bicipital groove and gradually decreased in the
bicipital groove. It was the widest at the entrance to the groove and four slices
including this widest portion showed significantly greater cross-sectional areas
compared to the biceps tendon in normal shoulders. However, there were no differ-
ences in the physiological cross-sectional areas of the long and short heads of the
biceps. This study suggests that widening of the biceps tendon is likely to be a local
structural change.
1 Anatomy and Function 169

Fig. 8.2 Cross-sectional

area of the LHB tendon. Rotator cuff tear
shoulders
The cross-sectional area of
the LHB tendon in cuff 40
tear shoulders (red marks)
Rotator cuff tear shoulders
was significantly greater 30

Cross sectional
than that of normal Normal shoulders

ares (mm )
2
shoulders (green marks) at 20
the entrance to the bicipital
groove and three more 10
slices distally. *p < 0.05.
GL glenoid origin, BG 0
bicipital groove entrance,
GL BG MTJ
MTJ musculotendinous
junction (Reproduced with Normal
modification from Fig. 2, shoulders
Toshiaki et al. [2])

a b c

Fig. 8.3 Entirely posterior type. (a) Macroscopically, all the biceps tendon attachment was to the
posterior labrum. (b) Histologic examination with hematoxylin–eosin staining showed that the
labrum attachment of the biceps was totally posterior (circle). The IGHL was connected to the
superior glenohumeral ligament through a connecting band (CB) (arrow). (c) The polarized micro-
scopic examination of the circled area in (b) showed that the fiber orientation of the biceps was
totally posterior. SL superior labrum, LHB long head of the biceps, CB connecting band, SGHL
superior glenohumeral ligament, MGHL middle glenohumeral ligament, IGHL inferior glenohu-
meral ligament. (Reprinted from Fig. 1, Tuoheti et al. [3])

1.1.2 LHB Tendon and IGHL (Inferior Glenohumeral Ligament)

Next, we assessed the attachment types of the LHB tendon to the glenoid labrum
and their relationships with the glenohumeral ligaments macroscopically and histo-
logically in 101 cadaveric shoulders [3]. We found that the glenoid attachment pat-
terns were classified into three types: entirely posterior type in 28 shoulders (28%)
(Fig. 8.3), posterior-dominant type in 56 shoulders (55%) (Fig. 8.4), and equal type
in 17 shoulders (17%) (Fig. 8.5). In the posterior-dominant and equal types, the
IGHL attached relatively high (average 2:49 o’clock for the posterior-dominant
170 8 Long Head of the Biceps

a b c

Fig. 8.4 Posterior-dominant type. (a) Macroscopically, most of the tendon attachment was to the
posterior labrum, but some to the anterior labrum. The IGHL attachment was high. (b) Histologic
examination with hematoxylin–eosin staining showed that all the tendon attachment was to the
posterior labrum. (c) The polarized microscopic examination of the circled area in (b) showed that
the fiber orientation of the biceps tendon was totally posterior. SL superior labrum, LHB long head
of the biceps, SGHL superior glenohumeral ligament, IGHL inferior glenohumeral ligament.
(Reprinted from Fig. 2, Tuoheti et al. [3])

a b c

Fig. 8.5 Equal type. (a) Macroscopically, the tendon attachment was equally distributed to the
posterior and anterior labrum. (b) Histologic examination with hematoxylin–eosin staining showed
that the tendon attachment was entirely posterior. The IGHL and MGHL linked to the SGHL on
the superior labrum. (c) The polarized microscopic examination of the circled area in (b) showed
that the fiber orientation of the biceps tendon was totally posterior. SL superior labrum, LHB long
head of the biceps, SGHL superior glenohumeral ligament, MGHL middle glenohumeral ligament,
IGHL inferior glenohumeral ligament. (Reprinted from Fig. 3, Tuoheti et al. [3])

type and 1:49 o’clock for the equal type), whereas in the entirely posterior type, the
IGHL attached low (average 4:32 o’clock). Histologically, in the posterior-­dominant
and equal types, the fibers of the IGHL directly linked to the fibers of the biceps
tendon. In the entirely posterior type, no such direct connections were observed. In
all types, the fibers of the biceps tendon attached to the posterior labrum. By assess-
ing the biceps tendon attachment type, we can estimate the original attachment site
of the IGHL during surgical repair of the detached IGHL.
1 Anatomy and Function 171

1.2 Function

1.2.1 Function as a Stabilizer

When I started to learn shoulder surgery, I noticed that the LHB tendon was running
right on top of the humeral head. I thought the anatomical closeness of the tendon
and the head might suggest functional closeness. The LHB tendon looked as if it
stabilized the humeral head against the glenoid. I wanted to do this biomechanical
research on the function of the LHB tendon, but unfortunately, there was no biome-
chanical laboratory in our department in Sendai back in those days. Thus, when I
visited Mayo Clinic as a biomechanical research fellow in 1990, I started this proj-
ect using cadaveric shoulders.

Stabilizing Function of LHB

With the arm hanging freely, various loads were applied to the LHB tendon using a
spring and the translation of the humeral head was measured in the anterior, poste-
rior, and inferior directions to observe how the loading would affect the translation
of the humeral head (Fig. 8.6) [4].
This study showed that (1) loading the biceps tendon decreased the anterior
and posterior translations of the humeral head regardless of arm rotation, (2) the

Fig. 8.6 Experimental

setup of stability research
(adduction). The humeral
head was translated in
three directions using a
weight, and the translation
was measured with use of
an electromagnetic
tracking device. The load F
was applied to the biceps
tendon with a spring.
(Reproduced with
modification from Fig. 1,
Itoi et al. [4])
172 8 Long Head of the Biceps

inferior translation was decreased only when the arm was in external rotation,
and (3) the stabilizing function of the biceps tendon was most prominent with the
arm in external rotation. This is easy to understand from anatomical relationship
of the biceps tendon and the humeral head (Fig. 8.1). Clinical relevance of this
study is that as the LHB functions as a stabilizer of the humeral head, it may be
used in shoulder instability patients as prevention and treatment measures. We
presented this study at the Open Meeting of American Shoulder and Elbow
Surgeons in 1992 and were honored to receive the Neer Award for best scientific
research.
In our next cadaveric experiment, we evaluated the stabilizing function of the
long and short heads of the biceps with the arm in abduction and external rotation,
the so-called dislocation position [5]. Using 13 fresh cadaveric shoulders, the
specimen was fixed with the arm in 90° of abduction relative to the trunk and in
three different external rotations (60°, 90°, 120°). While three loads of 0 kg,
1.5 kg, and 3 kg were applied to each of the long and short heads of the biceps,
anterior translation of the humeral head by applying a 1.5-kg translation force
was measured with use of an electromagnetic tracking device. The capsular con-
ditions were (1) intact, (2) vented, and (3) with a Bankart lesion. The shoulder
was stabilized anteriorly by loading either the long or short head of the biceps in
both 60° and 90° of external rotation. With the arm in 120° of external rotation,
the stabilizing function was observed only after creating a Bankart lesion. From
this study, we concluded that both the long and short heads of the biceps had sta-
bilizing function, and it became more prominent in shoulders with instability.
Strengthening the biceps could be used in conservative treatment for shoulder
instability.
Furthermore, we compared the stabilizing function of the biceps and the rotator
cuff muscles (Fig. 8.7) [6]. The forces applied to the muscles were determined

Fig. 8.7 Experimental setup of stability research (abduction-external rotation). In three different
arm rotations and with three different stability conditions (intact, vented, Bankart lesion created),
each muscle was loaded, and the anterior translation of the humeral head was measured. LHB long
head of the biceps, SHB short head of the biceps, SSP supraspinatus, ISP/TM infraspinatus/teres
minor, SSC subscapularis (Reproduced with modification from Fig. 1, Itoi et al. [6])
1 Anatomy and Function 173

based on the physiological cross-sectional areas (PCSA), which were obtained by

dividing the muscle volume by the muscle fiber length. The maximum contraction
force of each muscle can be obtained by multiplying the PCSA by the constant.
Thus, we applied muscle forces proportional to their maximum contraction force.
The results showed that in stable shoulders, the subscapularis was the least impor-
tant anterior stabilizer and as the shoulder became unstable, the biceps was the most
important anterior stabilizer among all the muscles tested.
These are the basic research studies about the biceps brachii as a stabilizer.

Muscle Activity of LHB During Arm Elevation

Next, I will introduce our clinical studies about the biceps. First, we paid attention
to the most common shoulder disease, a rotator cuff tear. It is well known that the
humeral head migrates superiorly in shoulders with a longstanding tear of the
rotator cuff. The biceps tendon may function as a superior stabilizer. In order to
function as a superior stabilizer, the biceps must contact. We compared electro-
myographic (EMG) activities of the biceps during arm elevation in the scapular
plane in 40 patients with a rotator cuff tear and in 40 healthy volunteers [7]. In
normal shoulders, the EMG activity of the biceps was always less than 10% of the
maximum voluntary contraction (MVC). In cuff tear shoulders, 14 patients (35%)
showed increased EMG activities between 10 and 30% of MVC (Figs. 8.8 and
8.9). The EMG activity increased as the angle of elevation increased and with a
1-kg weight in the hand. This study suggested a possibility that the biceps activity
increased either to stabilize the humeral head superiorly or to assist arm elevation
as a mover.

%MVC (%) No load %MVC (%) 1-kg load

30 30

20 20

10 10

0 0
30 60 90 120 30 60 90 120
Elevation angle (º) Elevation angle (º)

Fig. 8.8 EMG activity of the biceps during arm elevation (normal shoulders). In normal shoul-
ders, the %MVC of the biceps was less than 10% regardless of elevation angle. (Reproduced with
modification from Fig. 1, Kido et al. [7])
174 8 Long Head of the Biceps

%MVC (%) No load %MVC (%) 1-kg load

30 30

20 20

10 10

0 0
30 60 90 120 30 60 90 120
Elevation angle (º) Elevation angle (º)

Fig. 8.9 EMG activity of the biceps during arm elevation (cuff tear shoulders). In cuff tear shoul-
ders, the %MVC increased more than 10% in 14 of 40 shoulders (35%). It increased at higher
angles of elevation and with load application. (Reproduced with modification from Fig. 2, Kido
et al. [7])

LHB as a Superior Stabilizer in Cuff Tear Shoulders

In our next clinical study, we observed how the superior migration of the humeral
head would be prevented by actively contracting the biceps in cuff tear shoulders.
The mechanism of superior migration of the humeral head in cuff tear shoulders is
as follows. During arm elevation, the humeral head needs to be stabilized in the
glenoid for the deltoid to function as an elevator. Otherwise, the deltoid only pushes
the humerus upward. The rotator cuff muscles pull the humeral head into the gle-
noid socket so that a center of rotation for arm elevation is firmly created. In cuff
tear shoulders, however, this function of the cuff muscles becomes weak due to a
tear, and accordingly the deltoid force pushes the humerus upward. This is how the
superior migration of the humerus is seen in cuff tear shoulders. The superior migra-
tion may be fixed in longstanding cuff tears. We investigated the function of the
biceps on superior migration of the humeral head in 19 cuff tear patients and 19
normal controls [8]. A brace was used to maintain contraction of the biceps during
elevation (Fig. 8.10). The force to pull a rubber band was approximately 1.5 kg,
which was 10–12% of MVC. X-rays were taken with the arm elevated at 0°, 45°,
and 90° with and without biceps contraction.
We found that in cuff tear shoulders, there was significantly greater superior
migration of the humeral head without biceps contraction, but it was depressed sig-
nificantly when the biceps was loaded (Fig. 8.11). We concluded that the biceps was
an active depressor of the humeral head in cuff tear shoulders.
As mentioned above, one third of cuff tear patients showed increased activity of
the biceps during arm elevation. However, we still do not know whether it worked
as a stabilizer to prevent superior migration or as a mover to elevate the arm. In addi-
tion, the rest two third of patients showed no increase in biceps activity. Further
investigation is necessary to clarify the biceps function in cuff tear shoulders.
1 Anatomy and Function 175

Fig. 8.10 Brace to

maintain biceps
contraction. By pulling a
rubber band during
elevation, the biceps
produces a force equivalent
to 10–12% of
MVC. (Reproduced with
modification from Fig. 1,
Kido et al. [8])

4 4
w/out biceps load w/out biceps load
w/ biceps load w/ biceps load
3 3

2 2
Head position

1 1
(mm)

p = 0.0012
0 0

–1 –1

–2 –2

–3 –3
0º 45º 90º 0º 45º 90º
Normal shoulders Rotator cuff tear shoulders

Fig. 8.11 Center of the humeral head relative to the glenoid. In normal shoulders, the humeral
head was well centered in the glenoid regardless of biceps contraction. In cuff tear shoulders, the
humeral head migrated superiorly with arm elevation. Biceps contraction significantly depressed
the superior migration, making the head position almost equal to that in normal shoulders.
(Reproduced with modification from Fig. 3, Kido et al. [8])

1.2.2 Function as a Mover

It is well known that the biceps is an elbow flexor and forearm supinator, but not
much is known about the biceps function to the shoulder.
There are various methods to determine muscle function. Direct observation of
joint motion by pulling the muscle tendon or electrically stimulating the muscle,
assessment of functional deficit caused by a tear of a certain muscle in clinical
176 8 Long Head of the Biceps

a b

Fig. 8.12 Cross-section of the arm. (a) The four muscles are identified. LHB long head of the
biceps, SHB short head of the biceps, CBR coracobrachialis, LHT long head of the triceps. (b) The
center of the cross-section of each muscle was approximated, and the screws with eyelet were
inserted so that the eyelet came to the approximated center of each muscle belly. (Reproduced with
modification from Fig. 1, Itoi et al. [9])

cases, moment arm measurement in cadaveric shoulders, etc. We measured the

moment arm of the arm muscles that crossed the shoulder joint: the long and short
heads of the biceps, the long head of the triceps, and coracobrachialis [9]. While the
scapula was fixed, the humerus was passively moved in the coronal plane (abduc-
tion), scapular plane (scaption), sagittal plane (flexion), and the horizontal plane
(horizontal flexion) by one radian (180°/π = 57.3°) and the tendon excursion of each
muscle was measured. This tendon excursion was the moment arm of each muscle
during that motion. The positive value means shortening or agonistic movement and
the negative value means extension or antagonistic movement. The cable attached to
each muscle passed through the approximated center of the cross-section of the
muscle (Fig. 8.12).
The LHB was a flexor, horizontal flexor, and abductor (Fig. 8.13). These func-
tions were affected by arm rotation. With the arm in external rotation, the moment
arms of abduction and scaption significantly increased, whereas with the arm in
internal rotation, the moment arm of horizontal flexion increased significantly.
The SHB was an adductor in the coronal plane, opposite to the LHB (Fig. 8.14).
It had little function in the scapular plane and showed the function of flexion and
horizontal flexion similar to the LHB. The function of the SHB was also affected by
the arm rotation: adduction function increased in internal rotation, and flexion func-
tion increased in external rotation. The coracobrachialis had similar function to
the SHB.
The LHT was an extensor and adductor, opposite to the function of the LHB. The
LHT function was not affected by arm rotation (Fig. 8.15). This study clarified the
individual function of the arm muscles at the glenohumeral joint.
1 Anatomy and Function 177

40
ER
30 NR
p<0.0001 p<0.0001 p=0.37 p<0.0001 IR
20
Moment Arm (mm)

10

–10

–20

–30

–40 Abduction Scaption Flexion Horiz. Flexion

Fig. 8.13 Function of the LHB. The LHB was a flexor, horizontal flexor, and abductor. The func-
tion of abduction and scaption increased in external rotation, whereas the function of horizontal
flexion increased in internal rotation. (Reproduced with modification from Fig. 3, Itoi et al. [9])

40 p=0.0635
ER
30 p<0.0001 NR

20 IR
Moment Arm (mm)

10 p<0.0001 p<0.0001

–10

–20

–30

–40
Abduction Scaption Flexion Horiz. Flexion

Fig. 8.14 Function of the SHB. The SHB was an adductor, flexor, and horizontal flexor. The
adductor function increased in internal rotation, whereas flexor function increased in external rota-
tion. (Reproduced with modification from Fig. 4, Itoi et al. [9])
178 8 Long Head of the Biceps

40
ER
30 NR

20 IR
Moment Arm (mm)

p=0.0001
10
p=0.55 p=0.0721 p=0.0490
0

–10

–20

–30

–40 Abduction Scaption Flexion Horiz. Flexion

Fig. 8.15 Function of the LHT. The LHT was an extensor and adductor, opposite to the function
of the LHB. The LHT function was not affected by arm rotation. (Reproduced with modification
from Fig. 6, Itoi et al. [9])

How much strength is reduced when the LHB tendon is ruptured? Some
report said that 20% reduction in elbow flexion and 17% reduction in shoulder
abduction were observed in acute phase, but no reduction observed in chronic
phase [10], but other report said that even in chronic phase, there were 29%
reduction in elbow flexion, 23% reduction in forearm supination, and 28%
reduction in forearm pronation [11]. Whether this amount of strength loss
affects patient’s lifestyle is one of the key factors to be considered in selecting
treatment option in acute phase.

2 Pathophysiology

2.1 LHB Tendinitis

LHB tendinitis can be induced by various causes. It is caused by works or sports that
put repetitive stress on the LHB tendon, spread from subacromial bursitis accompa-
nied by a rotator cuff tear (Fig. 8.16), mechanical stimulation by subacromial
impingement, and so on. In case with a tear of the subscapularis tendon, the LHB
tendon may dislocate out of the bicipital groove and reduced back to the groove.
This repetitive motion may cause irritation and inflammation of the LHB tendon.
2 Pathophysiology 179

Fig. 8.16 LHB tendinitis.

This tendinitis is
accompanied by a rotator
cuff tear. Hyperemia is
seen on the superior
surface of the LHB tendon

LHB tendinitis is reported to accompany 60% of patients with a rotator cuff tear. If
the LHB tendon dislocates in young patients, it may be caused by a dysplasia of the
bicipital groove.

2.2 Tear of the LHB Tendon

Tear of the LHB tendon is divided into three types based on its location: proximal,
musculotendinous junction, and distal. As most LHB tendon tears occur at the prox-
imal tendon, I would like to focus on the proximal tear. The proximal tear of the
LHB tendon is further divided into two groups: traumatic tear and degenerative tear.
A traumatic tear is common among the athletes. A degenerative tear is usually seen
in patients older than 50 years of age, especially those with a rotator cuff tear. An
osteophyte in the bicipital groove could be a cause of LHB tear. Most degenerative
tears occur at the proximal end of the tendon, the attachment site to the superior
labrum, because this area is known to have low vascularity, which makes this area
more susceptible to degeneration and a tear. The mechanisms of traumatic tears are
to pull the arm downward in adduction, to fall on the outstretched arm, repetitive
stress by throwing motion, etc. Anterosuperior impingement between the subscapu-
laris and the anterosuperior labrum with the arm in flexion and internal rotation is
thought to be one of the causes of LHB tear. In patients with a supraspinatus tear,
the LHB tendon may be sandwiched between the acromion and the humeral head,
which may lead to a partial or complete tear of the LHB tendon (Fig. 8.17). It is
reported that 6% of patients with full-thickness tears of the rotator cuff have a
comorbid tear of the LHB tendon. Patients on dialysis often have shoulder pain due
180 8 Long Head of the Biceps

Fig. 8.17 Partial tear of

the LHB tendon. The LHB
tendon runs horizontally in
this view. The inferior part
of the tendon is fibrillated
and torn. A part of the
humeral head is seen on
the left

to amyloid deposition, which is called dialysis arthropathy. In dialysis arthropathy

of the shoulder, the amyloid deposition causes inflammation, thickening, and tear of
the LHB tendon [12].

3 Diagnosis

3.1 LHB Tendinitis

The common physical findings of LHB tendinitis are tenderness on the bicipital
groove, positive Yergason test (Fig. 8.18), and positive Speed test (Fig. 8.19). If an
injection of local anesthetics into the bicipital groove eliminates pain during these
tests, we are sure that the pain comes from LHB tendinitis.

3.2 Tear of LHB Tendon

It is obvious on inspection that the contour of the biceps muscle belly shifts distally
and bulges prominently in the distal aspect of the biceps region. Due to a rupture of
the proximal tendon, the muscle belly loses its tension and migrates distally to
3 Diagnosis 181

Fig. 8.18 Yergason test.

With the patient’s elbow
flexed at 90° and the
forearm in pronation, ask
the patient to supinate the
forearm against resistance.
If this maneuver induces
pain at the bicipital groove,
the Yergason test is positive

Fig. 8.19 Speed test. With

the patient’s elbow
extended and the forearm
supinated, ask the patient
to elevate the arm forward
against resistance. If this
maneuver induces pain at
the bicipital groove, the
Speed test is positive

create a mass. This bulging muscle makes the arm appear like that of muscular car-
toon character Popeye so that this bulging is called “Popeye sign” (Fig. 8.20).
In an obese patient, the “Popeye sign” may not be clearly seen. In that case, ask
the patient to flex the elbow and supinate the forearm. Active contraction of the
biceps makes the bulging more prominent. After the tear, subcutaneous hematoma
is often seen. It may extend down to the elbow area (Fig. 8.21).
182 8 Long Head of the Biceps

Fig. 8.20 Popeye sign.

The entire muscle belly of
the LHB migrates distally
with a torn proximal LHB
tendon and creates a
bulging. This bulging
makes the arm appear like
that of muscular cartoon
character Popeye so that it
is called “Popeye sign”

a b

Fig. 8.21 Subcutaneous hematoma. On the next day of injury, an extensive subcutaneous hema-
toma is observed from the lateral aspect of the arm (a) to the medial aspect of the elbow (b)
4 Diagnosis 183

4 Diagnosis

There are few image findings that are specific to LHB tendinitis. Effusion around
the LHB tendon on ultrasound and MRI indicates that there is abundant effusion in
the glenohumeral joint, caused by either LHB tendinitis, synovitis of the biceps
tendon sheath, or synovitis of the glenohumeral joint. LHB tendon rupture, on the
other hand, shows unique findings.

4.1 Arthrography

Arthrography is not so commonly performed these days as it used to be. Contrast

agent injected into the glenohumeral joint spreads and fills the biceps tendon sheath
in the groove. If there is a tear of the LHB tendon, the tendon sheath may also be
torn and the contrast agent leaks distally into the surrounding soft tissues (Fig. 8.22).

Fig. 8.22 Arthrogram of

LHB tendon tear. Contrast
agent in the glenohumeral
joint leaks from the biceps
tendon sheath (white
arrow) distally along the
muscle belly (white
arrowheads). Distally
migrated LHB tendon is
also seen (black
arrowheads)
184 8 Long Head of the Biceps

4.2 MRI

MRI is the most used imaging modality these days because it reveals not only a
tendon tear but also tendon thickening, a partial tear, and effusion around the ten-
don. In cases with a tear of the LHB tendon, the torn tendon may migrate out of the
joint to the midportion of the arm with hematoma around (Fig. 8.23). The swollen
muscle belly is also visible.

Fig. 8.23 T2-weighted

MRI of LHB tendon tear.
The torn tendon of the
LHB is pulled distally and
the tip is tightly wound like
the fiddlehead fern (white
arrow). Distal to the tendon
is a mass created by the lax
muscle belly of the biceps
(white arrowheads)
5 Treatment 185

5 Treatment

5.1 LHB Tendinitis

Conservative treatment should come first. If it is a pure inflammation, steroid

injection is effective. At the same time, physical therapy to improve the cuff mus-
cle function is important. If tendinitis is accompanied by a rotator cuff tear with
or without a dislocation of the LHB tendon, physical therapy to improve the
motion of scapulothoracic joint together with a steroid injection is effective. If the
conservative treatment fails, repair of the subscapularis or supraspinatus should
be considered. During arthroscopic rotator cuff repair, we often see LHB tendini-
tis. Cuff repair is just sufficient in these cases unless there is a partial tear of the
LHB tendon. LHB tendinitis caused by a rotator cuff tear will go away once a cuff
tear is repaired. In recalcitrant tendinitis cases and those with a partial tear of the
tendon, tenodesis is indicated. In the elderly patients, tenotomy may be indicated,
but I prefer to perform tenodesis to preserve the function of the biceps to the
elbow. When I visited Gilles Walch in Lyon in 1994 as a Traveling Fellow, he
showed me several cases of painful rotator cuff tears to be treated with LHB
tenotomy alone. He said the LHB tendon was a pain generator and just cutting the
tendon would relieve the pain in cuff tear patients. I was shocked by his comment
and by his action because I had been deeply involved in the research of the biceps.
I could not help calling him a “biceps killer.” This spontaneous naming seemed to
intrigue him. He laughed aloud and said, “I honorably accept your naming.” Since
then, I have often noticed that he uses the term “biceps killer” to introduce him-
self to the audience. I am sure he loves this naming. Twenty-five years have passed
since I first called him a biceps killer, and biceps tenotomy seems to have citizen-
ship these days. However, tenodesis still seems to be more preferable than tenot-
omy among most surgeons except in France. The reasons are (1) both functions of
the biceps to the shoulder and to the elbow will be lost by tenotomy, whereas
elbow function is preserved by tenodesis to the humerus and both shoulder and
elbow functions are more or less preserved by tenodesis to the coracoid process
or conjoint tendon, (2) muscle cramping pain is a problem after tenotomy, and (3)
the Popeye sign may not be acceptable especially for thin patients. Today, most
doctors seem to perform tenodesis except for the elderly patients with low demand
of the shoulder, who will be successfully treated with tenotomy. Tenodesis is
divided into three groups based on the location of tenodesis: (1) proximal tenode-
sis (fix the tendon to the bicipital groove), (2) suprapectoral tenodesis (fix the
tendon distal to the bicipital groove and just proximal to the upper border of the
pectoralis major), and (3) subpectoral tenodesis (fix the tendon behind the pecto-
ralis major tendon using a mini-open technique). There are advantages and disad-
vantages in each technique. I prefer to perform proximal tenodesis using suture
anchors.
186 8 Long Head of the Biceps

5.2 Tear of the LHB Tendon

Immediately after a tear, arm immobilization in a sling, cooling, and NSAIDs are
used for pain relief. If it is a degenerative tear in the middle-aged to the elderly
patients, conservative treatment is expected to reduce pain, while the functional
deficit is minimal with muscle weakness of about 20% in elbow flexion and forearm
supination. This amount of strength reduction causes little problem for the patients
unless they are involved in occupation or sports to perform strenuous repetitive
motion of supination and elbow flexion. If the patient is involved in these types of
occupation or sports, tenodesis should be considered soon after the injury. In young
patients with a traumatic tear, surgery is also indicated.

References

1. Itoi E, et al. Morphology of the torn rotator cuff. J Anat. 1995;186(Pt 2):429–34.
2. Toshiaki A, et al. Cross-sectional area of the tendon and the muscle of the biceps bra-
chii in shoulders with rotator cuff tears: a study of 14 cadaveric shoulders. Acta Orthop.
2005;76(4):509–12.
3. Tuoheti Y, et al. Attachment types of the long head of the biceps tendon to the glenoid labrum
and their relationships with the glenohumeral ligaments. Arthroscopy. 2005;21(10):1242–9.
4. Itoi E, et al. Stabilizing function of the long head of the biceps in the hanging arm position. J
Shoulder Elb Surg. 1994;3(3):135–42.
5. Itoi E, et al. Stabilising function of the biceps in stable and unstable shoulders. J Bone Jt Surg
Br. 1993;75(4):546–50.
6. Itoi E, et al. Dynamic anterior stabilisers of the shoulder with the arm in abduction. J Bone Jt
Surg Br. 1994;76(5):834–6.
7. Kido T, et al. Electromyographic activities of the biceps during arm elevation in shoulders with
rotator cuff tears. Acta Orthop Scand. 1998;69(6):575–9.
8. Kido T, et al. The depressor function of biceps on the head of the humerus in shoulders with
tears of the rotator cuff. J Bone Jt Surg Br. 2000;82(3):416–9.
9. Itoi E, et al. Moment arms of the arm muscles at the glenohumeral joint using the tendon
excursion method. J Musculoskelet Res. 2008;11(2):45–53.
10. Soto-Hall R, Stroot JH. Treatment of ruptures of the long head of biceps brachii. Am J Orthop.
1960;2:192–3.
11. Deutch SR, et al. Permanent disabilities in the displaced muscle from rupture of the long head
tendon of the biceps. Scand J Med Sci Sports. 2005;15(3):159–62.
12. Ando A, et al. Magnetic resonance imaging classification of haemodialysis-related amyloi-
dosis of the shoulder: risk factors and arthroscopic treatment. Knee Surg Sports Traumatol
Arthrosc. 2017;25(7):2217–24.
Chapter 9
Throwing Shoulder

1 Throwing Shoulder and Throwing Motion

The terms “pitch” and “throw” in baseball have different meanings. “Throwing
shoulder” refers to injuries related to both pitching motion and throwing motion in
various sports such as baseball, softball, volleyball, handball, javelin/hammer/dis-
cus throw, serving in tennis/badminton, and volleyball spike. Repeating these over-
head motions may cause various injuries of the shoulder, which is called throwing
shoulder or thrower’s shoulder. When pain aggravates, not only throwing motion,
but also the activities of daily living may be affected by pain.

1.1 Baseball Pitching and Pitching Phase

Baseball is very popular in Japan, and people in all generations enjoy baseball. It is
one of the national sports. Among baseball players, pitchers are required to pitch
fastballs and breaking balls with high accuracy. To fulfill these requirements, vari-
ous pitching forms have been developed. Pitching motion is divided into six phases:
(1) wind-up, (2) early cocking, (3) late cocking, (4) acceleration, (5) deceleration,
and (6) follow-through (Fig. 9.1).
During the wind-up phase, the pitcher (right-handed) elevates the left leg
while keeping a balance on the right leg. The early cocking phase starts from the
time the right hand with a ball exits the left-hand glove and ends with the left foot
landing on the mound. The shoulder is elevated in the scapular plane above 90°
and externally rotated to approximately 45°. During the late cocking phase, the
shoulder comes to the maximum external rotation of 170°. Most injured pitchers
experience pain during this phase. In the acceleration phase, the shoulder

© The Author(s), under exclusive license to Springer Nature Singapore Pte 187
Ltd. 2023
E. Itoi, Shoulderology, https://doi.org/10.1007/978-981-99-0345-0_9
188 9 Throwing Shoulder

Wind-up Late Deceler- Follow-

Early cocking Acceleration ation
cocking through
Start Hands Foot Maximum Ball Finish
apart down external release
rotation

Fig. 9.1 Pitching motion. Pitching motion is divided into six phases: wind-up, early cocking, late
cocking, acceleration, deceleration, and follow-through. (Reprinted from Pink and Perry [1]. From
Fig. 25-1 The Shoulder, third Ed. Rockwood CA Jr., Matsen FA III, Wirth MA, Lippitt SB, eds.
Philadelphia: WB Saunders, 2004, p.1280)

internally rotates 80°, reaching peak angular velocity of 7000°/s. It is said that
this high-velocity rotation is the fastest motion our body can make. Within 0.05 s,
the ball is released with the speed of 140–150 km/h. During the deceleration
phase, the infraspinatus, teres minor, and parascapular muscles contract eccentri-
cally to decelerate the quick motion of internal rotation. Also, the biceps con-
tracts eccentrically to decelerate elbow extension. During this phase, a centrifugal
force of 1.0–1.5 times body weight is applied to the arm. The arm is lowered and
adducted during the follow-through phase with continued eccentric contraction
of the parascapular muscles, while a strong traction force is applied to the poste-
rior capsule.

1.2 Injuries in Each Pitching Phase

There are injuries characteristic to each pitching phase. Anterior instability caused
by excessive stretching of the anterior capsule is commonly seen during the late
cocking phase. Rotator cuff and labral injuries related to rapid internal rotation are
characteristic to the acceleration phase. Superior labrum anterior and posterior
(SLAP) lesion, Bennet lesion, and subacromial impingement caused by centrifugal
force are often seen during the follow-through phase. In young baseball pitchers
before the growth plate closes, repetitive pitching motion puts too much stress on
the growth plate to result in growth plate separation (epiphyseolysis). This is called
“Little Leaguer’s shoulder.”
2 Pathophysiology 189

2 Pathophysiology

2.1 Glenohumeral Internal Rotation Deficit (GIRD)

In 1960s, a decrease in abduction-internal rotation and an increase in abduction-­

external rotation were first noted in tennis players. Similar findings were also
observed in throwing athletes such as baseball and softball players. This limitation
in internal rotation with the arm in abduction seen in throwing athletes is called
“glenohumeral internal rotation deficit” (GIRD).

2.1.1 Physiologic GIRD and Pathologic GIRD

Is GIRD something that should be avoided? According to Wilk et al. (2002), asymp-
tomatic dominant shoulders of 372 professional baseball players showed a 7° increase
in abduction-external rotation and a 7° decrease in abduction-internal rotation. As
seen in this report, the entire arc of rotation shifts toward external rotation without
changing the range of arc. This shift of rotation arc may be skeletal or soft tissue
adaptation to throwing motion. Some say this asymptomatic shift of rotation arc or
GIRD is physiologic and should be differentiated from pathologic GIRD. The ques-
tion is how much GIRD is physiologic and how much GIRD is pathologic. Pathologic
GIRD used to be defined as GIRD >20° or >25°. However, the Disabled Throwing
Shoulder Summit in 2013 defined pathologic GIRD as either GIRD >18° or decrease
in entire rotation arc >5° [2]. The presence of pathologic GIRD increases the risk of
various pathologies such as articular-side partial-thickness rotator cuff tears, labral
injuries, internal impingement, subacromial impingement, and scapular dyskinesis.

2.1.2 Causes of GIRD

Why does GIRD occur? The main possible causes are (1) increased retroversion of
the humeral head, (2) tightness of the posterior capsule, and (3) thixotropy of the
muscles (increased muscle stiffness after repetitive stretching stresses). The retro-
version of the humeral head is discussed in detail in the next section, but it is 65° in
infants, gradually decreases with age and ceases at 20°–30° in adults. Repetitive
pitching motion suppresses the derotation process and as a result retroversion angle
remains large. This suppression seems to be our body’s adaptation because it is
reported that those with larger retroversion of the humeral head is less likely to have
throwing injuries. Thus, GIRD caused by an increase in retroversion seems to be
physiologic. On the other hand, the tightness of the posterior capsule and rotator
cuff muscles is thought to be caused by stretching stress on the capsule or eccentric
contraction of the cuff muscles during the deceleration and follow-through phases.
Eccentric contraction increases muscle stiffness as an acute reaction and GIRD
comes to a peak 4–18 h after pitching [2]. This reversible change of muscle is called
“thixotropy” [3]. When this reversible change becomes irreversible, it causes
190 9 Throwing Shoulder

pathologic GIRD. Thus, it may not be appropriate to determine one fixed value to
define pathologic GIRD, but rather it should be determined taking the angle of
humeral retroversion into consideration. In fact, a recent meta-analysis showed that
pathologic GIRD changed with age with the definition >11.4° for adolescents and
>13.8° for adults [4]. They concluded that higher quality prospective research was
needed to clarify the role that GIRD played in future injuries to overhead athletes of
various ages.

2.2 Changes of Humeral Retroversion with Age

As mentioned above, one of the causes of increased external rotation and decreased
internal rotation in abduction is retroversion of the humeral head. In normal non-­
throwing shoulders, the retroversion ranges between 20° and 30°, but in throwing
shoulders, it ranges between 30° and 40° with an increase of 10°. When does a dif-
ference in retroversion between the throwing and non-throwing shoulders begin?

2.2.1 Measurement of Retroversion (First to Eighth Graders)

First, we measured the humeral retroversion in 66 young baseball players from the
first to the eighth graders [5]. Using ultrasonography, we measured an angle between
a line passing through the bicipital groove and the center of the humeral head and a
line parallel to the axis of the forearm with the elbow flexed at 90° (bicipital-­forearm
angle: BF angle) (Fig. 9.2). With the subject lying supine with 90° of shoulder

a b

Fig. 9.2 How to measure the BF (bicipital forearm) angle. In supine position with 90° of shoulder
abduction and elbow flexion, short-axis image of the bicipital groove was obtained with the ultra-
sound probe (P) in front of the shoulder (a). Putting the bicipital groove (BG) in the center of the
image, the humerus was rotated so that the apexes of the greater tuberosity (GT) and lesser tuberos-
ity (LT) aligned horizontally (b). While holding this arm position, the angle between the vertical
line and the forearm axis line was measured by a digital inclinometer (I), which was defined as the
BF angle. (Reproduced with modification from Fig. 1, Kurokawa et al. [6])
2 Pathophysiology 191

abduction and elbow flexion, we put the ultrasound probe on the anterior surface of
the shoulder. Looking at the short-axis image of the bicipital groove, we rotated the
arm so that the apexes of the greater and lesser tuberosities forming the groove
aligned horizontally in this image. Then, we measure the angle between the vertical
line and the forearm axis line, which was defined as the BF angle. The larger the BF
angle, the smaller the humeral retroversion.
This study showed that the BF angle was smaller (greater retroversion) on the
throwing side and the BF angle showed a moderate positive correlation with the age
(r = 0.427 on the throwing side, r = 0.472 on the non-throwing side). In other words,
the humeral retroversion decreases with age, but this decrease seems to be sup-
pressed on the throwing side.

2.2.2 Measurement of Retroversion (First to Sixth Graders)

As we wanted to clarify the relationship between the humeral retroversion and the
age, we studied in a larger sample size of 172 baseball players from the first to the
sixth graders [6]. This study clearly demonstrated that the BF angle increased (retro-
version decreased) with age on the non-throwing side but it remained almost the
same on the throwing side, resulting in significant side-to-side differences at the
fourth grader and older (Fig. 9.3). From this study, we conclude that (1) the humeral
retroversion decreases with age (physiological derotation) and (2) the physiological
derotation is suppressed on the throwing side. We assume that repetitive throwing
motion is the key factor that suppresses the physiological humeral derotation process.
Edelson measured the humeral retroversion in 180 shoulders of 90 complete
skeletons of young people aged from 4 months to 19 years [7]. He reported that the
humeral retroversion was the greatest between 4 months and 4 years with an aver-
age value of 65°. Most of the derotation process took place by the age of 8 and then
it continued slowly until 16–19 years of age. This anatomical study coincides with
our data showing the physiological derotation observed in non-throwing side of
young baseball players.

Fig. 9.3 Change in BF angle (º)

humeral retroversion with 40

age. The BF (bicipital-­ 35

forearm) angle of the
30
non-dominant shoulder Non-Dominant
increased (retroversion 25
∗ ∗
decreased) with age, but it 20
remained unchanged in the ∗ Dominant
15
dominant shoulder. The
side-to-side differences 10
were significant (*) in the 5
fourth to sixth grades.
0
(Produced from Table 2, 1
st
2
nd rd
3
th
4 5
th
6
th
Kurokawa et al. [6]) Grader
192 9 Throwing Shoulder

2.3 Tightness of the Posterior Capsule

Tightness of the posterior capsule is considered one of the causes of GIRD in throw-
ing athletes. During pitching motion, especially during follow-through, strong ten-
sion is applied to the posteroinferior capsule, which causes fibrosis, thickening, and
contracture of the capsule confirmed by MRI as well as histological examination of
the capsule obtained during surgery.
Tension of the posteroinferior capsule is also related to subacromial impinge-
ment. We measured the contact pressure between the coracoacromial arch and the
rotator cuff tendons using cadaveric shoulders in positions simulating each phase of
pitching motion [8]. In normal shoulders, the contact pressure was low during the
early and late cocking phases, whereas it increased gradually from the acceleration
phase to the follow-through phase. The contact pressure in the follow-through phase
increased even more if the posteroinferior capsule was imbricated. This study sug-
gests a possibility that posterior capsular tightness induces subacromial impinge-
ment, which in turn induces rotator cuff injuries.
A Bennett lesion is another type of throwing injury. First, this lesion was thought
to be an osteophyte formation caused by a traction of the long head of the triceps.
Now, we believe that it is an osteophyte formation along the posteroinferior border
of the glenoid caused by a traction of the posteroinferior capsule, especially the
posteroinferior glenohumeral ligament, during the follow-through phase. In gen-
eral, it is symptom free, but comorbid posterior instability may cause a labral injury
or a partial-thickness tear of the rotator cuff, which accompanies pain.

2.4 Injury of the Labrum: Traction by the LHB Tendon

and Labral Strain

During pitching motion, very fast movement of internal rotation occurs with a speed
of 7000°/s. How much force is applied to the labrum during this quick motion? One
of the causes of superior labrum anterior and posterior (SLAP) lesion is a traction
by the LHB tendon. Thus, we set cadaveric shoulders in several simulated pitching
positions and measured the strain of the superior labrum while applying a force to
the LHB tendon [9]. Using 10 cadaveric shoulders, strain gauges were attached to
the anterosuperior labrum (1 o’clock in the right shoulder), posterosuperior labrum
(11 o’clock), and the LHB tendon (Fig. 9.4).
The strain data of the anterosuperior and posterosuperior labrum are shown in
Fig. 9.5. Both portions of the labrum showed the greatest strain in the late cocking
phase, which was significantly greater than in the other phases. Also, the strain of
the posterosuperior labrum was significantly greater than that of the anterosupe-
rior labrum.
We know from the anatomical study that the LHB tendon fibers are continuous to
the posterosuperior labrum [10]. Therefore, it is understandable that a large amount
2 Pathophysiology 193

Fig. 9.4 Strain

measurement. Strain
gauges were attached to
the anterosuperior labrum
(SG-1), posterosuperior
labrum (SG-2), and the
LHB tendon (SG-3). GL
glenoid, HH humeral head,
LHB long head of the
biceps

Fig. 9.5 Labral strain Strain (%)

during pitching motion. 5 *
Both the anterosuperior
and posterosuperior **
4.5
portions of the labrum Anterosperior
showed significantly Posterosuperior
4
greater strain in the late **
cocking phase than in the 3.5
other phases
(**P < 0.0001). Also, the 3
strain of the
posterosuperior labrum 2.5
was significantly greater
than that of the 2
anterosuperior labrum
(*P = 0.0009). 1.5
(Reproduced with
modification from Fig. 2, 1
Pradhan et al. [9])
0.5

0
Early Late Acceleration Follow-
cocking cocking Deceleration through
Pitching Motion
194 9 Throwing Shoulder

of strain occurs in the posterosuperior labrum when the LHB tendon is loaded. Also,
we know from the moment arm study that the SHB (short head of the biceps) works
as a horizontal flexor with the arm in abduction-external rotation, but with the arm
in abduction-internal rotation, the LHB works as a horizontal flexor [11]. The sud-
den change of moment arm of the LHB from external to internal rotation may be
related to injuries at the junction between the LHB tendon and the superior labrum.
This study suggests that the greatest strain of the posterosuperior labrum during the
late cocking phase may have an association with the onset of a SLAP lesion.

2.5 Pain from the Acromioclavicular Joint

Most throwing injuries occur at the glenohumeral joint, but sometimes the acromio-
clavicular (AC) joint is also involved. This is the reason why we investigated the
prevalence of AC joint pain in baseball players [12]. We examined 150 baseball
players ranging from the first to the eighth graders. We diagnosed that the pain came
from the AC joint when there was tenderness on the AC joint and the pain provoca-
tion tests were positive. Ultrasonography was performed with the hand on the oppo-
site shoulder (horizontal flexion) (Fig. 9.6).
The prevalence of AC joint pain increased with age: 1.4% in the 1st–3rd graders,
13% in the 4th–6th graders, and 15% in the 7th–8th graders. The difference in superior
capsular bulging between the throwing and non-throwing sides was significantly
greater in players with AC joint pain (1.6 mm ± 1.2 mm) than in those without
(0.2 mm ± 0.8 mm). The superior capsular bulging was supposed to reflect increased
joint effusion due to synovitis of the AC joint. It had already been reported that AC joint
pain was a part of throwing injuries among the professional and collegiate baseball
players, but we were surprised to find in this study that AC joint pain already started
among the elementary baseball players. Based on these findings, restriction of pitch
numbers should be considered in elementary and junior high school baseball players.

a b

Fig. 9.6 Ultrasonogram of the AC joint (long-axis image). We defined the superior capsular bulg-
ing as a distance (D) between an extended line along the superior margin of the distal clavicle (c)
(solid line) and a parallel line (dotted line) tangent to the top of the hypoechoic area. In asymptom-
atic shoulders (a), the distance was very small, but in shoulders with AC joint pain (b), a large
distance was observed. ac acromion (Reprinted from Fig. 1, Hatta et al. [12])
3 Epidemiological Survey 195

3 Epidemiological Survey

We performed several surveys of adolescent baseball players regarding injuries

related to baseball with the help of the Miyagi Amateur Sports Association and
Miyagi Junior Sports-club Association.

3.1 Shoulder/Elbow Pain and Pain from Other Body Parts

Our first study was aimed to explore the association between trunk/lower extremity
pain with shoulder/elbow pain in young throwing athletes [13]. The study popula-
tion comprised 2215 young throwing athletes who were the members of the Miyagi
Amateur Sports Association. They played baseball, softball, volleyball, handball,
tennis, and badminton with an average age of 11 years (range 6–15). We asked their
physical status using a questionnaire. Multiple logistic regression analysis revealed
the higher prevalence of shoulder/elbow pain in athletes with back pain (odds ratio
[OR] = 5.52), hip pain (OR = 6.13), knee pain (OR = 2.28), and foot pain (OR = 3.03),
compared with those without pain. There are possibilities that insufficient support
by the trunk and lower extremity might induce upper extremity injuries and that
intervention to the trunk and lower extremity muscles such as stretching and
strengthening might be beneficial in preventing shoulder/elbow pain in young
throwing athletes.
Next study was a questionnaire survey of 1582 youth baseball players belonging
to the Miyagi Amateur Sports Association [14]. A total of 24.8% (n = 381) had
shoulder/elbow pain, whereas 8.5% (n = 130) had low back pain and 13.1%
(n = 201) had knee pain. The prevalence of shoulder/elbow pain with concomitant
low back and knee pain was 61.2% (n = 82) and 51.9% (n = 108) (p < 0.001),
respectively. The presence of low back pain (OR = 4.31) and knee pain (OR = 2.92)
was significantly associated with the prevalence of shoulder/elbow pain among
youth baseball players. We should check the complaints of the trunk and lower
extremities as well as those of the shoulder and elbow for preventing severe injuries
in youth baseball players.
So far, we clarified that shoulder/elbow pain was associated with pain of the
trunk and lower extremities. What about shoulder/elbow pain and joint flexibility?
We examined the association of shoulder/elbow pain with flexibility of the trunk
and lower extremities [15]. We analyzed baseball players who participated in the
3-day National Junior Sports Clubs Baseball Games. Range of motion in external
rotation and internal rotation of the hip, as well as the finger-to-floor distance and
heel-to-buttock distance, was measured. The straight-leg-raise test was also con-
ducted. Of 210 players surveyed, 177 without shoulder/elbow pain were included in
the analysis. Of these 177 participants, 16 (9.0%) reported having shoulder/elbow
pain during the 3-day tournament. Participants with the incidence of shoulder/elbow
pain had a significant restriction in hip internal rotation of the stride leg compared
196 9 Throwing Shoulder

to those without pain (35.8° vs. 43.7°, P = 0.022). There were no significant associa-
tions of other joint flexibilities with shoulder/elbow pain. This study suggested that
flexibility of the trunk and extremities was the intrinsic risk factors of throwing
injuries and that keeping the trunk and extremities flexible might prevent upper
extremity injuries.

3.2 Shoulder/Elbow Pain and Videogames

Among the extrinsic risk factors, not only the factors directly associated with play-
ing baseball such as position, pitch count, and pitch type, but also lifestyle behaviors
are important. One of them was playing video games. We performed a survey to
examine the association of video game playing and television viewing with shoul-
der/elbow pain among 200 elite young baseball players participated in the National
Junior Sports Clubs Baseball Games [16]. At the time of survey, 30% of them had
shoulder/elbow pain. Playing video games ≥3 h/day was significantly associated
with shoulder/elbow pain versus spending <1 h/day playing video games
(OR = 5.59). However, there was no association of watching TV and shoulder/elbow
pain. Why is playing video games associated with shoulder/elbow pain? There are
two possible explanations. First, watching TV is a passive activity with very little
energy consumption almost equal to that of sitting quietly, whereas playing video
games is an active behavior, which requires concentration in the game, gazing at the
screen with rapid movements of the fingers and upper extremities. This causes
increased load on the muscles of the arm, shoulder, and neck, which leads to muscle
fatigue and stiffness, especially the trapezius. The muscles with increased stiffness
may lead to scapular dysfunction and throwing injuries. A second possible explana-
tion is the posture. Because handheld game consoles and tablets are widely used,
video game playing typically involves gazing downward and flexing the head for-
ward, whereas TV viewing can be performed while maintaining a forward gaze.
Poor posture involving forward neck flexion and thoracic kyphosis leads to scapular
dysfunction in the overhead athletes, which may increase the risk of throwing inju-
ries. Whatever the cause, lifestyle habits such as playing video games are modifiable
risk factors. This study highlights the need to take action so that throwing injuries
may be prevented in young baseball players. This study had such a great impact that
the Wall Street Journal picked up this study in an article entitled “The latest threat to
pitchers’ arms could be ‘Fortnite’,” published on August 25–26 edition, 2018 (From
THE WALL STREET JOURNAL, Vol. CCLXXII, No. 47, August 25–26, 2018).
When I first read this article, I did not know what “Fortnite” was because I did
not play video games. Fortnite is a free-to-play, online video game, which is the
world’s most popular video game of 2021. In May 2018, Fortune magazine pub-
lished an article entitled “Fortnite addiction is becoming a problem for Major
League Baseball.” People paid more and more attention to video game addiction in
baseball players. This could be the background of the above-mentioned article in the
Wall Street Journal. Major League Baseball players have plenty of time between
4 Diagnosis 197

baseball games and travel, and it is easy for them to play Fortnite whether they are
in the clubhouse, their hotel room, or on the bus. Although we do not know if there
is any association of playing video games and throwing injuries in the Major League
Baseball players, we need to pay attention to the lifestyles of young baseball players
to prevent throwing injuries.

3.3 Verbal and Physical Abuse by the Coaches

Verbal and physical abuse has become a social problem in recent years. We per-
formed a cross-sectional study on youth sports team coaches (n = 380) using a self-­
reported questionnaire regarding verbal and physical abuse [17]. The prevalence of
verbal abuse was 75% and that of physical abuse 6.6%. Verbal abuse was signifi-
cantly associated with personal experience of verbal abuse by former coaches
(OR = 2.00), and acceptance of verbal/physical abuse (OR = 2.61). Physical abuse
was significantly associated with greater than 10 years of coaching experience
(OR = 7.16), personal experience with physical abuse by former coaches
(OR = 4.25), and acceptance of verbal/physical abuse (OR = 3.85). Coaches’ expe-
riences with verbal/physical abuse by former coaches negatively affected young
athletes. Breaking this vicious cycle is necessary to keep athletes physically and
mentally healthy and interested in playing baseball.

4 Diagnosis

In pitchers with throwing injuries, pain at the time of pitching is characteristic. The
phase in which they feel pain gives us an idea what types of injuries they are likely
to have (see 1–2. Injuries in each pitching phase).
In physical examination, we need to pay attention to muscle atrophy caused by a
rotator cuff tear or nerve injuries, scapular dyskinesis, and pathologic GIRD. Possible
pathologies behind these physical findings are rotator cuff tendinitis, rotator cuff
tear, LHB tendinitis, and labral lesion including SLAP lesion. There are many diag-
nostic tests for SLAP lesion reported in the literature, but diagnostic accuracy is
almost the same. I prefer to use the biceps tension test (same maneuver as the Speed
test, but induces pain not in the bicipital groove, but inside the glenohumeral joint).
Imaging of each pathology is described in each section. To avoid overlap, I would
like to show an X-ray of the Bennett lesion (Fig. 9.7) and MR arthrogram of type II
SLAP lesion (Fig. 9.8). Sometimes, multiple lesions can be found. Then, we need to
determine the responsible lesion for the symptom. Pitching test is useful for narrow-
ing down the responsible lesion. After an injection of local anesthetics into a sus-
pected lesion, ask the patient to perform pitching. Comparing the symptoms before
and after the pain block at each suspected lesion, we can determine the responsi-
ble lesion.
198 9 Throwing Shoulder

Fig. 9.7 Bennett lesion on

plain X-ray (arm elevated).
Osteophyte formation
(white arrow) at the
posteroinferior glenoid is
clearly visible

Fig. 9.8 Type II SLAP

lesion on MR arthrogram.
The superior labrum is
detached from the glenoid
and the contrast agent is in
the space between the
labrum and the glenoid
(white arrow)
5 Treatment 199

5 Treatment

5.1 Conservative Treatment

Throwing injuries are caused by repetitive pitching and throwing. If players feel
pain, they need to rest their arms. For the prevention of throwing injuries, limitation
in pitch numbers, warming up and cooling down, and stretching are beneficial.
Pitching form is also important. Poor form is related to various throwing injuries.
Correction of pitching form can be done step by step by a team of physician, physi-
cal therapist, and trainer who are well versed to the pitching forms. Together with
the form correction, strengthening of shoulder muscles and parascapular muscles as
well as stretching the muscles and joints should be performed. Recently, attention is
being paid to scapular dyskinesis.

5.1.1 Scapular Dyskinesis

First, we need to know the terminology of scapular motion, which is unique and
different from the glenohumeral motion. The scapula moves three dimensionally on
the thoracic cage so its motion is quite complex and decomposed to several motions.
The terms we use to describe the scapular motion is listed in Fig. 9.9 [18]. These
terms are officially adopted by the Japan Shoulder Society. Superior/inferior move-
ment of the scapula is called elevation/depression, lateral/medial movement is
called abduction/adduction (protraction/retraction is also used), rotation in the hori-
zontal plane is called internal/external rotation, rotation in the coronal plane is

Fig. 9.9 Scapular motion. Terminology of scapular motion (Reproduced with modification from
Fig. 1, Hamada et al. [18])
200 9 Throwing Shoulder

called upward/downward rotation, and rotation in the sagittal plane is called ante-
rior/posterior tilt.
The scapular dyskinesis falls into three categories: Type I, II, and III [19]. Type
I is characterized by prominence of the inferior medial scapular border. The scap-
ula is anteriorly tilted. Type II is characterized by prominence of the entire medial
scapular border and represents abnormal internal rotation around a vertical axis.
Type III is characterized by superior translation of the entire scapula and promi-
nence of the superior medial scapular border. Types I and II are commonly
observed in patients with subacromial impingement, and Type III is commonly
observed in patients with internal impingement (posterosuperior impingement). In
Type I, due to anterior tilt of the scapula, the pectoralis minor is shortened.
Stretching the pectoralis minor and strengthening the lower trapezius and serratus
anterior are recommended. In Type II with strong internal rotation of the scapula,
strengthening the serratus anterior is recommended. In Type III where the scapula
is elevated and rotated upward, mobilization and stretching of the upper trapezius
is important.

5.1.2 Muscle Strengthening and Stretching

Stretching of the pectoralis minor can be performed by a physical therapist with a

patient in the supine position. Pushing the humerus upward along the axis of the
humerus with the elbow in full flexion and the shoulder in 30° of flexion widens the
distance between the origin and insertion of the pectoralis minor muscle and accord-
ingly stretches the muscle (Fig. 9.10).
To strengthen the lower trapezius, anti-gravity shoulder flexion and external rota-
tion in the prone position is recommended. To strengthen the serratus anterior, push-
­up exercise used to be recommended, but open-kinetic chain exercise is also

Fig. 9.10 Stretching for

the pectoralis minor. With
a patient in the supine
position, a physical
therapist pushes the
humerus upward with the
elbow in full flexion and
the shoulder in 30° of
flexion. This maneuver
stretches the tight
pectoralis minor.
(Reprinted from Fig. 1,
Muraki et al. [20])
5 Treatment 201

Fig. 9.11 Strength exercise for the serratus anterior. In the standing position, ask the patient to flex
both arms at 90° and hold a rubber band with the involved hand (left hand) while the band is pulled
backward by the physical therapist. Next, ask the patient to retract both scapulae and then push
both arms forward against the resistance of the rubber band. Both scapulae are abducted (pro-
tracted) by the contraction of the serratus anterior. (Reproduced with modification from Fig. 2,
Muraki et al. [20])

recommended (Fig. 9.11) because open-kinetic chain exercise requires greater mus-
cle strength than closed-kinetic chain exercise like push-up.
To correct increased upward rotation and elevation of the scapula, scapulotho-
racic exercise with minimal activation of the upper trapezius are recommended.
Side-lying flexion, side-lying external rotation, and prone horizontal extension with
external rotation can promote the activity of the lower trapezius while minimizing
the activity of the upper trapezius.
Once the scapular muscles are successfully activated, a dynamic exercise such as
lawnmower exercise is added (Fig. 9.12). This exercise requires large joint motion
at multiple segments, leading to greater activity of the lower trapezius compared to
isometric exercise.
Sleeper stretch and cross-body stretch are commonly performed for contracture
of the thickened posterior capsule and increased stiffness of the posterior cuff mus-
cles which cause pathologic GIRD. A meta-analysis comparing these two stretches
showed that the cross-body stretch obtained more internal rotation immediately
after the stretch than the sleeper stretch.
In a poor posture with increased thoracic kyphosis, the scapula is elevated and
anteriorly tilted, which makes it easier to have subacromial impingement. For play-
ers with increased thoracic kyphosis, bending and extending exercise is effective to
obtain flexibility and back muscle strengthening exercise is effective in correcting
kyphosis [21].
202 9 Throwing Shoulder

a b

Fig. 9.12 Lawnmower exercise. (a) Ask the patient to step forward with the contralateral leg and
bend the trunk forward to the contralateral side with the involved hand in front of the stepped leg.
(b) Then ask the patient to rotate and extend the trunk while retracting the scapula. (Reprinted from
Fig. 3, Muraki et al. [20])

5.2 Surgical Treatment

As the conservative treatment is the mainstay of the management for throwing ath-
letes, indication for surgical treatment is quite limited. One reason is that different
from trauma related to sports, sports injuries caused by repetitive motion is very
responsive to conservative treatment. Also, return to play after surgery is another
issue to be considered. According to the data of professional baseball players in the
USA, return to play after shoulder surgery is 63% among all position players, but
among the pitchers, it is only 45% [22]. This is another reason to avoid surgical
References 203

intervention to the throwing athletes. When the symptoms do not disappear after
conservative treatment, or when the pain is bothering even in activities of daily liv-
ing, surgery may be indicated. The most common surgical procedure is labral repair.
In cases of rotator cuff injuries, debridement is usually selected because return to
play after rotator cuff repair is unacceptably low. For the Bennett lesion, removal of
the osteophyte combined with labral repair and capsular plication is commonly
performed.
There are two methods to manage type II SLAP lesion: SLAP repair and biceps
tenodesis. So far, SLAP repair is said to be better in functional recovery and consid-
ered the standard procedure for SLAP lesion. However, as persistent pain and joint
contracture are common complications and re-operation rate is high in the middle
aged or elderly patients, people started to think that SLAP repair should be indi-
cated only to the young patients and biceps tenodesis for the rest. Recent studies
comparing these two surgical procedures demonstrated that there was no significant
difference in clinical outcomes, patient’s satisfaction, and return to play. Comparison
between these procedures among the pitchers, there was no difference in muscle
activities during pitching, but there was a difference in trunk rotation. The maxi-
mum trunk rotation was observed during late cocking to early acceleration in unin-
jured controls and pitchers with biceps tenodesis, but it was observed during the late
acceleration phase in pitchers with SLAP repair [23]. To avoid alteration of pitching
biomechanics, biceps tenodesis is gradually preferred in the young players includ-
ing pitchers.

References

1. Pink M, Perry J. Biomechanics. In: Jobe FW, et al., editors. Operative techniques in upper
extremity sports injuries. St. Louis: Mosby-Year Book; 1996.
2. Kibler WB, et al. The disabled throwing shoulder: spectrum of pathology-10-year update.
Arthroscopy. 2013;29(1):141–161.e26.
3. Proske U, Morgan DL, Gregory JE. Thixotropy in skeletal muscle and in muscle spindles: a
review. Prog Neurobiol. 1993;41(6):705–21.
4. Johnson JE, et al. Glenohumeral internal rotation deficit and injuries: a systematic review and
meta-analysis. Orthop J Sports Med. 2018;6(5):2325967118773322.
5. Yamamoto N, et al. Why is the humeral retroversion of throwing athletes greater in dominant
shoulders than in nondominant shoulders? J Shoulder Elb Surg. 2006;15(5):571–5.
6. Kurokawa D, et al. Differences in humeral retroversion in dominant and nondominant sides of
young baseball players. J Shoulder Elb Surg. 2017;26(6):1083–7.
7. Edelson G. The development of humeral head retroversion. J Shoulder Elb Surg.
2000;9(4):316–8.
8. Muraki T, et al. Effect of posteroinferior capsule tightness on contact pressure and area beneath
the coracoacromial arch during pitching motion. Am J Sports Med. 2010;38(3):600–7.
9. Pradhan RL, et al. Superior labral strain during the throwing motion. A cadaveric study. Am J
Sports Med. 2001;29(4):488–92.
10. Tuoheti Y, et al. Attachment types of the long head of the biceps tendon to the glenoid labrum
and their relationships with the glenohumeral ligaments. Arthroscopy. 2005;21(10):1242–9.
204 9 Throwing Shoulder

11. Itoi E, et al. Moment arms of the arm muscles at the glenohumeral joint using the tendon
excursion method. J Musculoskelet Res. 2008;11(2):45–53.
12. Hatta T, et al. Association between acromioclavicular joint pain and capsular bulging in ado-
lescent baseball players. Knee Surg Sports Traumatol Arthrosc. 2016;24(12):3750–5.
13. Sekiguchi T, et al. Coexistence of trunk or lower extremity pain with elbow and/or shoul-
der pain among young overhead athletes: a cross-sectional study. Tohoku J Exp Med.
2017;243(3):173–8.
14. Sekiguchi T, et al. Youth baseball players with elbow and shoulder pain have both low back and
knee pain: a cross-sectional study. Knee Surg Sports Traumatol Arthrosc. 2018;26(7):1927–35.
15. Sekiguchi T, et al. Restriction in the hip internal rotation of the stride leg is associated with elbow
and shoulder pain in elite young baseball players. J Shoulder Elb Surg. 2020;29(1):139–45.
16. Sekiguchi T, et al. Excessive game playing is associated with musculoskeletal pain among
youth athletes: a cross-sectional study in Miyagi prefecture. J Sports Sci. 2018;36(16):1801–7.
17. Hagiwara Y, et al. The characteristics of coaches that verbally and physically abuse young
baseball players using a self-reported questionnaire. J Orthop Sci. 2019;24(2):361–7.
18. Hamada J, et al. Evaluation of passive scapular motions. Katakansetsu. 2012;36(2):303–5.
19. Kibler WB, McMullen J. Scapular dyskinesis and its relation to shoulder pain. J Am Acad
Orthop Surg. 2003;11(2):142–51.
20. Muraki T, Itoi E. Scapulothoracic disorders and nonsurgical management of the impingement.
In: Imhoff AB, Savoie III FH, editors. Rotator cuff across the life span—ISAKOS Consensus
Book. Berlin: Springer; 2019. p. 113–21.
21. Itoi E, Sinaki M. Effect of back-strengthening exercise on posture in healthy women 49–65
years of age. Mayo Clin Proc. 1994;69(11):1054–9.
22. Chalmers PN, et al. Epidemiology of shoulder surgery among professional baseball players.
Am J Sports Med. 2019;47(5):1068–73.
23. Chalmers PN, et al. Postoperative restoration of upper extremity motion and neuromuscular
control during the overhand pitch: evaluation of tenodesis and repair for superior labral ante-
rior–posterior tears. Am J Sports Med. 2014;42(12):2825–36.
 Part V
Biomechanical Research Related
to Suture Anchors
Chapter 10
Anchor Insertion Angle: Deadman Theory
and Its Interpretation

1 Deadman Theory

The deadman theory proposed by Burkhart is well known regarding the appropriate
insertion angle of suture anchors (Fig. 10.1) [1].
When Burkhart was watching a new wire fence being constructed on a South
Texas ranch, he recognized the analogy of a suture anchor to a deadman. At the
corner fence post, they put a large rock called a deadman buried under the ground
and a deadman wire attached to the rock and the fence post prevented the post from
falling. He applied this structural concept of a deadman preventing a fall of the
fence post to rotator cuff repair using suture anchors. Based on the deadman–fence
post mechanics, he proposed ideal inclination angles of suture anchor and suture.
This is called the deadman theory. According to this theory, both inclination of the
suture (θ2) and inclination of the suture anchor (θ1) are equal to or less than 45° to
achieve an ideal fixation (Fig. 10.2). The smaller these angles, the closer the con-
struct gets to the ideal.
Let us first take a look at the inclination of the suture (deadman wire) (θ2). If the
deadman is too close to the fence post, θ2 is larger than 45° and cannot produce suf-
ficient horizontal force to counterbalance the pull by the fence wire. As a result, the
fence post inclines away from the deadman until the horizontal force of the dead-
man wire is equal to that of the fence wire. In this equilibrium, the inclination of the
deadman wire (θ2) is 45° (Fig. 10.3).
Similarly, during rotator cuff repair, if θ2 is larger than 45°, the pull by the rotator
cuff tendon is greater than the horizontal force of the suture tension and the tendon
retracts proximally until θ2 is 45° to counterbalance the pull by the tendon. If θ2 is
smaller than 45°, the tension in the suture is even smaller, which lowers the risk of
suture breakage or anchor pullout (Fig. 10.4).
However, to make θ2 small, you must insert a suture anchor far more distally
from the tendon edge, exposing a certain length of suture outside of the tendon,

© The Author(s), under exclusive license to Springer Nature Singapore Pte 207
Ltd. 2023
E. Itoi, Shoulderology, https://doi.org/10.1007/978-981-99-0345-0_10
208 10 Anchor Insertion Angle: Deadman Theory and Its Interpretation

Fig. 10.1 Anchoring effect

of a corner fence post by a
deadman. The deadman is
a large rock buried under
the ground and attached to
the fence post by a wire at
an angle θ to the ground.
The deadman wire pulls
the fence post (TX) against
the pull by the fence wire
(WX). (Reprinted from
Fig. 1, Burkhart [1])

Fig. 10.2 Deadman

theory. Ideally, both θ1 and
θ2 should be equal to or
less than 45°. (Reprinted
from Fig. 7, Burkhart [1])
ø1
ø2
1 Deadman Theory 209

Fig. 10.3 A deadman too close to the fence post. The left figure shows a deadman which is too
close to the fence post (θ2 = 60°). As it cannot provide sufficient horizontal force to counterbalance
the pull by the fence wire, the fence post inclines away from the deadman until the horizontal force
of the deadman wire becomes equivalent to the fence wire force (right figure). In this equilibrium,
the inclination of the deadman wire (θ2) is 45°

Fig. 10.4 Balance of suture tension and tendon pull. When θ2 = 45°, the horizontal force (TX) of
the suture tension is equal to the tendon pull (WX). If θ2 = 10°, the suture tension (T′) to create the
same horizontal force (TX) is smaller than the suture tension (T) when θ2 = 45°. Less tension on the
suture means less chance of suture breakage or anchor pullout

which increases a risk of suture breakage. In addition, the footprints of the rotator
cuff of the greater and lesser tuberosities do not extend laterally so that it is techni-
cally impossible to make θ2 much smaller than 45°. Thus, practically suture anchors
are inserted close to the torn edge of the rotator cuff tendon to make θ2 approxi-
mately 45°.
Next, let us look at the relationship between the anchor insertion angle and the
suture. When the suture runs 45° inclined in the tendon, it is generally believed that
the smaller the angle between the suture line and the line perpendicular to the axis
of suture anchor (θ1), the stronger the pullout strength. Therefore, it is believed that
suture anchors provide best performance when they are inserted perpendicular to
the suture line or 45° to the bone surface (Fig. 10.5).
210 10 Anchor Insertion Angle: Deadman Theory and Its Interpretation

Fig. 10.5 Suture and suture anchor. The deadman theory is interpreted as follows. When the
suture is inclined 45° to the bone surface (θ2 = 45°), the maximum pullout strength of the anchor
is obtained with the anchor inserted perpendicular to the suture line or 45° to the bone surface
(θ1 = 0°)

2 Discrepancy Between the Theory and Experiments

After the deadman theory was introduced, many biomechanical studies were per-
formed to prove the hypothesis that a suture anchor inserted at 45° to the bone sur-
face provided the greatest pullout strength [2–5]. However, none of these studies
could prove this hypothesis. On the contrary, these biomechanical studies showed
that pullout strength of anchors inserted at 45° was equal to or even less than that
inserted at 90° in synthetic bones and cadaveric bones. As soon as each of these
studies was published, Burkhart offered a counterargument to each of them in
“Letter to Editor” submitted to the Journal of Arthroscopy [6–8]. These arguments
continued for 20 years, and the Editors of the Journal of Arthroscopy welcomed
these arguments in Editorial, citing the word by Lyman Beecher, “No great advance
has ever been made in science, politics, or religion, without controversy” [9].
However, this comment did not help us understand why there were discrepancies
between the deadman theory and the biomechanical studies.
The theory and reality are not always consistent. In Fig. 6 of the original article
by Burkhart, he compared the pullout strength of a vertically inserted anchor pulled
either vertically (θ = 90°) or obliquely (θ = 45°) (Fig. 10.6).
Based on the trigonometric function, he said that pulling the anchor obliquely at
45° required 41% greater force than pulling it vertically. It is correct theoretically,
but in reality, the pullout strength is different from the theoretical value. We com-
pared the pullout strength of anchors pulled vertically versus 45° obliquely using
synthetic bones and porcine bones [10]. The 45° oblique pull required 22% greater
force in the low-density bone and 12% greater force in the medium density bone
compared to the vertical pull. However, the 45° oblique pull required 19% less force
2 Discrepancy Between the Theory and Experiments 211

Tx T
a b

Ty
ø
ø

Fig. 10.6 Inclination of the pull. When you pull the anchor obliquely (θ = 45°) (a), you need 41%
more force than pulling it vertically (b), calculated from the trigonometric function. (Reprinted
from Fig. 6, Burkhart [1])

than the vertical pull in the high-density bone (Fig. 10.7). In porcine bones, the
oblique pull required 7% less force than the vertical pull.
Here is another example of difference between theory and reality: pulling the
anchor perpendicular to its axis (Fig. 10.8).
If the anchor is pulled perpendicular to its axis, pullout strength is theoretically
infinite, which means you cannot pull it out. However, in reality, you can. Pulling
the anchor perpendicular to its axis requires even 1–4% less force depending on the
bone density than pulling the anchor parallel to its axis [11]. These examples clearly
tell us that theory and reality are different. The reason is the theory assumes that
rigid body does not deform. The anchor comes out only in the direction of insertion.
On this assumption, the anchor does not move by pulling it perpendicular to its axis.
In reality, however, the bone deforms. As the anchor is pulled perpendicularly, it
starts to break the surrounding bone, rotates in the bone, and finally comes out of the
bone in the direction of pull. In this case, the pullout strength depends on the strength
of the bone surrounding the anchor and the friction between the anchor and the
bone. The pullout strength never becomes infinite as calculated by the theory.
212 10 Anchor Insertion Angle: Deadman Theory and Its Interpretation

Fig. 10.7 Theoretical and measured values of pullout strength. Theoretically, 45° oblique pull
requires 41% greater force compared to the vertical pull. In reality, however, the 45° oblique pull
required 19% less force than the vertical pull in the high-density synthetic bone

Fig. 10.8 Theoretical and measured values of pullout strength. If the anchor is pulled perpendicu-
lar to its axis, pullout strength is theoretically infinite. It never comes out! However, our pullout
study showed that the pulling the anchor perpendicular to its axis required 4% less force than pull-
ing it parallel to its axis in the high-density bone

3 Friction Around the Anchor Affects the Theory

The clinical question is which angle a suture anchor should be inserted. We investi-
gated the relationship between the anchor insertion angle and the pullout strength in
different bone density and different anchor-bone friction [12]. In this experiment,
the suture inclination was fixed at 45° (θ2 = 45°). To make a low-friction anchor, we
removed the threads from a commercially available threaded anchor. We inserted
this thread-less anchor into the synthetic bone and compared the pullout strength
with different pulling directions relative to the anchor (Fig. 10.9).
3 Friction Around the Anchor Affects the Theory 213

Fig. 10.9 Pullout strength of thread-less anchor. The greatest pullout strength was obtained when
the anchor was pulled perpendicular to its axis, which was 77% greater than when it was pulled
parallel to its axis. Using a 2.5-mm drill hole, the friction became even lower, and the effect of
perpendicular pull became greater (173% increase). * p < 0.0001

Using a 2.0-mm drill hole for anchor insertion, the perpendicular pull on the
anchor required 77% greater force compared to the parallel pull. Using a 2.5-mm
drill hole to make the friction lower, the perpendicular pull required 173% greater
force than the parallel pull. From this study, we can say that when the friction is low
like a tent peg, inserting the peg to the ground perpendicular to the tent rope or 45°
to the ground would provide the best performance. However, when the friction is
high like a suture anchor, that is another story. Using a threaded suture anchor, the
greatest pullout strength was obtained when it was inserted perpendicular to the
bone surface. Oblique insertion at 45° required 19% less pullout force in the low-­
density bone and 16% less force in the high-density bone compared to the perpen-
dicular insertion (Fig. 10.10) [11]. Our finite element model analysis also supports
the data of these biomechanical tests [13].
This study clearly demonstrated that using a high-friction anchor, it should be
inserted perpendicular to the bone surface to achieve the best anchoring effect. With
a low-friction peg, θ1 = 0° provides the strongest construct as generally believed by
the people, whereas with a high-friction anchor, θ1 = 45° provides the strongest
construct. In both conditions, θ1 is equal to or less than 45°, just as the deadman
theory says.
214 10 Anchor Insertion Angle: Deadman Theory and Its Interpretation

Fig. 10.10 Pullout strength of threaded anchor. Using a threaded anchor, the greatest pullout
strength was obtained when it was inserted perpendicular to the bone surface both in the low-­
density and high-density bones. * p < 0.0005

4 Conclusions

In conclusions, the deadman theory that both the suture inclination and anchor incli-
nation relative to the suture should be equal to or less than 45° to obtain the stron-
gest construct is correct. However, interpretation to insert the anchor at 45° to the
bone surface is not correct because the optimal θ1 changes between 0° and 45°
depending upon the friction. A low-friction tent peg should be inserted at 45° to the
ground, but a high-friction suture anchor should be inserted at 90° to the bone sur-
face. In summary, the deadman theory is correct, but its interpretation is wrong.

References

1. Burkhart SS. The Deadman theory of suture anchors: observations along a South Texas fence
line. Arthroscopy. 1995;11(1):119–23.
2. Liporace FA, et al. The mechanical effects of suture anchor insertion angle for rotator cuff
repair. Orthopedics. 2002;25(4):399–402.
References 215

3. Strauss E, et al. The effect of the angle of suture anchor insertion on fixation failure at the
tendon-suture interface after rotator cuff repair: Deadman’s angle revisited. Arthroscopy.
2009;25(6):597–602.
4. Clevenger TA, et al. The effect of insertion angle on the pullout strength of threaded suture
anchors: a validation of the Deadman theory. Arthroscopy. 2014;30(8):900–5.
5. Green RN, et al. Biomechanical study: determining the optimum insertion angle for screw-in
suture anchors-is Deadman’s angle correct? Arthroscopy. 2014;30(12):1535–9.
6. Burkhart SS. Suture anchor insertion angle and the Deadman theory. Arthroscopy.
2009;25(12):1365.
7. Burkhart SS. The Deadman theory is alive and well. Arthroscopy. 2014;30(9):1049–50.
8. Burkhart SS. Can the Deadman be killed? Arthroscopy. 2015;31(2):181–2.
9. Rossi MJ, et al. A cavalcade of shoulder controversies: Deadman angle revisited... diver-
gent anchor angles and depths, and rotator cuff vectors confuse comparison... and more.
Arthroscopy. 2014;30(12):1529–32.
10. Nagamoto H, et al. A biomechanical study on suture anchor insertion angle: which is better,
90° or 45°? J Orthop Sci. 2017;22(1):56–62.
11. Itoi E, et al. Deadman theory revisited 12. Biomed Mater Eng. 2016;27(2–3):171–81.
12. Nagamoto H, Yamamoto N, Itoi E. Effect of anchor threads on the pullout strength: a biome-
chanical study. J Orthop. 2018;15(3):878–81.
13. Sano H, et al. Stress distribution inside bone after suture anchor insertion: simulation
using a three-dimensional finite element method. Knee Surg Sports Traumatol Arthrosc.
2013;21(8):1777–82.
Chapter 11
Minimal Safe Distance Between Suture
Anchors: Relationship Between the Anchor
Distance and Pullout Strength

The pullout strength of a suture anchor depends on the strength of the surrounding
bone. If two anchors are too close to each other, there may not be sufficient bone
around each anchor to hold it, and as a result, the pullout strength may decrease. It
is generally recommended that two anchors be separated at least 1 cm apart.
However, we could not find any evidence to support this recommendation. Thus, we
performed this biomechanical study to determine the safe distance between two
anchors [1]. We used metal screw-type anchors (TwinFix 5.0 Ti; Smith & Nephew,
Andover, MA) and PEEK (polyether ether ketone) coil-type anchors (HEALICOIL
PK 4.5; Smith & Nephew, Andover, MA) (Fig. 11.1).
Two same-type suture anchors were placed into the synthetic bones with the
center-to-center distance set at 4 mm, 6 mm, 8 mm, and 10 mm. The results showed
that the distance of 4 mm significantly decreased the pullout strength compared to
those with the distance of 6 mm or above (Fig. 11.2). With two anchors inserted at
the distance of 4 mm, the anchor threads overlapped 1.3 mm with the metal anchors
and 0.3 mm with the PEEK anchors. Due to overlapping of anchor threads, there
was not sufficient bone around each anchor, which seemed to make the pullout
strength significantly weaker. For the two tested anchors, the minimum distance
between the anchors without decreasing the pullout strength was 6 mm regardless
of bone density in this biomechanical study. When inserting two suture anchors,
keeping the center-to-center distance at least 6 mm is important to achieve maxi-
mum strength of the anchors.

© The Author(s), under exclusive license to Springer Nature Singapore Pte 217
Ltd. 2023
E. Itoi, Shoulderology, https://doi.org/10.1007/978-981-99-0345-0_11
218 11 Minimal Safe Distance Between Suture Anchors: Relationship Between the Anchor…

Fig. 11.1 Suture anchors. b

(a) Metal screw-type
anchor. (b) PEEK coil-type
anchor. (Reproduced with
modification from Fig. 1,
Kawakami et al. [1]) a
Reference 219

Polyurethane foam Polyurethane foam

Porcine bone
0.08 mg/cm3 0.16 mg/cm3 0.24 mg/cm3 0.08 mg/cm3
900 *
*
* P < 0.05 *
800
Ultimate load to failure (N)

700 *
*
*
600
500
*
400 *
* *
*
300 *
*
*
200 *
100
0
4 6 8 10 4 6 8 10 4 6 8 10 4 6 8 10 4 6 8 10
Distance of anchors (mm)
PEEK coil-
Metal screw-type anchor
type anchor

Fig. 11.2 Pullout strength. The pullout strength with the distance of 4 mm was significantly
weaker than those with 6 mm or greater in both types of anchors. (Reprinted from Fig. 7, Kawakami
et al. [1])

Reference

1. Kawakami J, et al. Minimum distance of suture anchors used for rotator cuff repair without
decreasing the pullout strength: a biomechanical study. Arthroscopy. 2018;34(2):377–85.
 Part VI
Rare Shoulder Diseases
Chapter 12
Facioscapulohumeral Muscular Dystrophy
(FSHD)

For shoulder surgeons, rotator cuff tears and tendinitis, frozen shoulder, and anterior
shoulder instability are common diseases. On the other hand, you may encounter
some rare shoulder diseases. I would like to pick up two cases of rare shoulder dis-
eases: facioscapulohumeral muscular dystrophy and trapezius palsy. Both cases
were treated successfully with functional recovery after surgery.

1 Facioscapulohumeral Muscular Dystrophy (FSHD)

Facioscapulohumeral muscular dystrophy (FSHD) is an autosomal dominant dis-

ease first reported by Landouzy and Dejerine in 1884. Although autosomal domi-
nant, 10–30% arise from spontaneous mutation. The muscles of face and shoulder
girdle are mainly involved. In 1990, it was reported that the genetic defect was
located on chromosome 4q35 (loss of D4Z4 repetition), which is called FSHD type
1 [1]. Later, another type FSHD related to a mutation in the SMCHD1 gene on
chromosome 18 was reported in 2012, which is called FSHD type 2 [2]. Overall,
95% are FSHD type 1.
Typically, individuals with FSHD become symptomatic during teen years, but
age of onset is variable. More than 90% of individuals with FSHD notice problems
by age 20. The symptoms are slowly progressive, but become stationary in adult-
hood, and the life expectancy is normal. The symptoms depend on the muscles
involved, but most commonly involved muscles are parascapular muscles such as
the serratus anterior, trapezius, rhomboids, and latissimus dorsi, and the facial mus-
cles such as the orbicularis oculi and oris. The deltoid is well preserved. Most
patients visit an orthopedic clinic because of difficulty in elevating the arm
(Fig. 12.1).

© The Author(s), under exclusive license to Springer Nature Singapore Pte 223
Ltd. 2023
E. Itoi, Shoulderology, https://doi.org/10.1007/978-981-99-0345-0_12
224 12 Facioscapulohumeral Muscular Dystrophy (FSHD)

Fig. 12.1 Active arm

elevation. During active
arm elevation, marked
scapular winging appears
and elevation is limited
to 70°

a b c

Fig. 12.2 Mechanism of scapular winging. (a) At rest with the arm in the hanging position. (b) In
normal shoulders, the scapula is stabilized to the thorax, and the deltoid contraction is transmitted
to the humerus to initiate abduction. (c) In patients with FSHD, the parascapular muscles are weak
and cannot stabilize the scapula to the thorax. Due to this scapular instability, deltoid contraction
rotates the scapula downward rather than elevates the arm upward, causing scapular winging

Let us consider the mechanism of scapular winging in patients with FSHD. When
you elevate your arm from the hanging arm position (Fig. 12.2a), the deltoid con-
tracts while the scapula is stabilized by the parascapular muscles. As the foundation
is stable, the deltoid can fully function to elevate the arm (Fig. 12.2b). In patients
with FSHD, the parascapular muscles are weak and cannot stabilize the scapula to
the thorax. When the deltoid contracts while the scapula is unstable, it is easier for
the deltoid to rotate the light scapula downward than elevate the heavy arm upward
1 Facioscapulohumeral Muscular Dystrophy (FSHD) 225

(Fig. 12.2c). This is the mechanism of scapular winging, and as a result, arm eleva-
tion is limited to 60°–70°.
If unstable scapula is the main cause of scapular winging and the inability to
elevate the arm, you may wonder what if we stabilize the scapula. In fact, if we
stabilize the scapula by holding it firmly on the thorax, scapular winging does not
appear, and active arm elevation improves up to 120° or so. This maneuver is called
the scapular compression test or Horwitz maneuver (Fig. 12.3) [3–5]. Positive test
means the glenohumeral muscles are functioning, but scapulothoracic muscles are
weak and non-functioning.
Patients with FSHD have a limited arm elevation due to scapular winging. This
is a functional deficit of this disease, but at the same time, unsightly appearance of
scapular winging is another reason patients want to have surgery (Fig. 12.4).
Facial muscle weakness causes symptoms such as difficulty in pursing the lips
like whistling, but they seldom complain of these symptoms unless you ask them. If
a teenager comes to your clinic complaining of difficulty in elevating the arm with
scapular winging, you should ask if whistling is possible.

Fig. 12.3 Scapular

compression test (Horwitz
maneuver). The examiner
stabilizes the scapula to
prevent scapular winging,
which makes the patient
able to elevate the arm
to 130°
226 12 Facioscapulohumeral Muscular Dystrophy (FSHD)

a b

Fig. 12.4 Scapular winging. With the arm hanging freely (a), winging is not obvious. However,
during active arm elevation (b), scapular winging becomes prominent

2 Diagnosis and Treatment

Diagnosis is based on the symptoms (weakness of facial muscles and shoulder gir-
dle muscles), physical findings, positive family history, slightly elevated serum cre-
atine kinase, myogenic pattern on electromyogram, and nonspecific myopathic
changes in biopsy specimen such as variation in muscle fiber size and degenerating/
regenerating fibers. Definitive diagnosis is made by genetic testing to find the con-
traction of D4Z4 repeats or a mutation in the SMCHD1 gene.
If a patient needs to elevate the arm, the scapula must be stabilized. It is difficult
to strengthen the weakened muscles as they do not respond to exercises. Using a
brace to prevent scapular winging may be tried, but it is not practical to constantly
wear a brace that strongly compresses the thorax. Considering the limitation of
conservative treatment, surgical stabilization of the scapula on the thorax is the
treatment of choice. Historically, there are two surgical procedures: scapulopexy
and scapulothoracic arthrodesis. Scapulopexy is to stabilize the scapula to the tho-
rax using fascia lata graft. There remains a slight motion between the scapula and
the thorax. The advantages are (1) as you do not need to wait for bone healing,
patients can move their arms soon after the operation, (2) without immobilization,
muscle wasting may be prevented, and (3) there is no decrease in vital capacity
because rib motion is preserved. The disadvantages are (1) the fascial graft may be
worn and torn in the long term due to residual motion between the scapula and the
thorax and (2) the clinical outcome gradually declines with time. Average 10-year
follow-up of patients treated with scapulopexy by using metal wires without per-
forming arthrodesis showed that this procedure was safe, effective, and reproduc-
ible with low complication rates [6].
The scapulothoracic arthrodesis is to stabilize the scapula on the rib cage using
screws and wires. Once bone healing is completed, the scapula is rigidly stabilized
on the rib cage so that arm elevation is reliably achieved, and it does not decline over
time. Disadvantages are (1) a brace must be used for 3 months postoperatively, (2)
which always accompanies muscle atrophy and weakness, and (3) limited rib
3 Case Presentation 227

motion results in a decrease in vital capacity. According to Bunch who first reported
bilateral scapulothoracic arthrodesis, a 10% decrease in vital capacity was observed
in unilateral arthrodesis and a 23% decrease in bilateral arthrodesis, which did not
affect patient’s activities of daily living [7]. These days, scapulothoracic arthrodesis
is preferred because of its stable long-term outcome.

3 Case Presentation

A 17-year-old male started to feel difficulty in elevating his right arm when he was
in the first year of high school. Because of this symptom, he gave up playing base-
ball. His symptom gradually aggravated in the second year of high school and vis-
ited a nearby clinic. Because of his limited arm elevation and scapular winging, he
was referred to our institute for further examination and treatment. The active eleva-
tion of his right arm was 80° with marked scapular winging. The scapular compres-
sion test was positive. He noticed he had been unable to whistle. His mother had
similar symptoms. Thus, FSHD was strongly suspected. However, he did not want
further examination and treatment because he had to prepare for college entrance
examination. We prescribed a brace to hold the scapula on the thorax, but he did not
come back afterwards.
After college, he started to work as a journalist. However, because of his scapular
winging, he could not enjoy chatting and dining with his friends, and he even tried
to avoid the company of others. His scapular winging was always on top of his mind.
He said he spent 80% of his time worrying about the scapula. To make it worse, he
started to feel the same symptoms on his left shoulder as well. He was so distressed
that he decided to quit his job and come back to us for treatment at the age of 24.
We confirmed the diagnosis of FSHD type 1 by gene testing, which revealed
contraction of D4Z4 repeats on chromosome 4q35. Active elevation of his right arm
was limited to 70° (Fig. 12.5), and the scapular compression test was positive. On

Fig. 12.5 Active elevation.

Active forward elevation
was limited to 70° with
marked scapular winging
228 12 Facioscapulohumeral Muscular Dystrophy (FSHD)

the other hand, his left shoulder symptoms were much milder. Active elevation of
his left arm was 120° with mild scapular winging. Although his bilateral shoulders
were involved, the right side was more severely involved, so we talked to the patient
about scapulothoracic arthrodesis on the right side to improve arm elevation and
stop scapular winging. We also talked to him that if his left side would aggravate
and if he would be happy with the surgical outcome of his right side, we would do
the same procedure for the left side. We obtained informed consent from the patient
and his family.
There are several procedures in scapulothoracic arthrodesis. Since the scapula is
a thin bone, the point is how to rigidly stabilize the thin scapula to the rib cage with-
out breaking the bone. To dissipate the stress on the scapular body, a plate or tibial
bone graft is used with screws or cables. Not only did I collect all the available
information from the literature, but also I asked for advice from my overseas friends
who had been familiar with this procedure. Eventually, we decided to use a metal
plate for stress dissipation with several cables instead of screws, which had a risk of
pleural or lung damage.

3.1 Surgery

Surgery was performed in prone position under general anesthesia. Bilateral arms
were abducted, and the right arm was held in a position so that the scapula was 15° of
upward rotation. A 20-cm longitudinal skin incision was made connecting the mid-
points between the spinous processes and the medial border of the scapula (Fig. 12.6).
Subcutaneous fat tissue was sharply dissected until the trapezius fascia appeared,
and along the fascia, the blunt dissection was continued. Once the atrophied trape-
zius appeared, it was cut along the medial border of the scapula to expose the infra-
spinatus and supraspinatus muscles. These muscles were released from the medial
border of the scapula and elevated laterally from the infraspinous and supraspinous
fossae for 4–5 cm. Next, elevating the medial border of the scapula with a bone
forceps to approach the ventral side of the scapula, the medial portion of the sub-
scapularis was released and elevated from the scapula. On the ventral side of the
scapula, there were the rhomboid major and rhomboid minor medially and the sub-
scapularis and serratus anterior laterally. Along the muscular septum, the lateral
muscles were separated bluntly from the medial muscles until the subscapular bursa
appeared on the rib cage. The bursa was rich in vascularity and carefully removed
to expose the rib cage. The third to sixth ribs were subperiosteally exposed starting
from the inferior border, then the superior border, and finally the ventral surface
(Fig. 12.7). Protecting the pleura with a retractor, bony beds were prepared on the
ribs using a diamond burr. The ventral surface of the scapular body was also slightly
abraded. The medial portion of the subscapularis (6 cm in length and 3 cm in width)
which was interposed between the ribs and the scapula was resected. At this point,
the wound was filled with saline to confirm no air leakage had occurred from
the lung.
3 Case Presentation 229

Fig. 12.6 Skin incision.

The right arm was held
with the scapula in 15° of
upward rotation. A 20-cm
longitudinal skin incision
was made connecting the
midpoints between the
spinous processes and the
medial border of the
scapula

Fig. 12.7 Exposure of the

scapula and ribs. The third
to sixth ribs were exposed
subperiosteally (3–6). The
medial portion of the
scapula was also
exposed (S)
230 12 Facioscapulohumeral Muscular Dystrophy (FSHD)

Fig. 12.8 Plating and

cerclage cabling. The
positioning pins were set in
the third and fifth holes of
the plate, receiving the
cables from the fifth and
sixth ribs, respectively

A 5-cm skin incision was made from the posterior superior iliac spine along the
iliac crest. The gluteus maximus was partly detached from the iliac crest and can-
cellous bone chips were harvested with a bone gouge. A 6-hole small reconstruc-
tion plate 3.5-­mm was bent to fit the contour of the dorsal aspect of the medial
scapular body. Then, 2.5-mm drill holes at first, third, and fifth plate holes were
made through the medial scapular body below the scapular spine. Additional two
2.5-mm drill holes were made above the scapular spine. 1.7-mm cables were
passed under the 3rd—6th ribs. Great care was taken to protect the pleura from
injury. The cable passing under the fourth rib was passed through the holes just
above and below the scapular spine (the strongest part of the scapula), and then
passed through the first hole of the plate. Next, the cable passing under the third
rib was passed through the highest drill hole of the scapula. The cables passed
under the fifth and sixth ribs were passed through the corresponding drill holes of
the scapular body, then passed through the third and fifth holes of the plate using
the positioning pins (Fig. 12.8).
Gradually, the cables were tightened to narrow the space between the scapula
and the ribs until it was about 1 cm. Then, the bone chips from the iliac crest were
inserted into this space (Fig. 12.9).
3 Case Presentation 231

Fig. 12.9 Iliac crest bone

graft. Bone chips from the
iliac crest were inserted to
the space between the
scapula and the ribs,
followed by tightening the
cables

Then, the cables were tightened using a cable tensioner with a force of 20 kg.
The most superior cable was tightened with a little less force to avoid fracture
because there was no plate there. After tightening all the cables, the scapula was
firmly stabilized on the rib cage. The scapula was positioned so that the distance
between the spinous process and the medial border of the scapula was 8 cm at the
scapular spine and 11 cm at the inferior angle of the scapula. The inclination of the
medial border of the scapula to the median line was 15° as we planned. The remnant
bone chips were placed in the available space between the scapula and the rib cage
(Fig. 12.10). The wound was once again filled with saline to check the integrity of
the pleura and lung. The rhomboids, infraspinatus, and supraspinatus were tied to
the cables to reduce the dead space.
After the wound closure, a 45° abduction brace prepared before surgery was
attached, and the patient was turned over to the supine position. The surgical time
was 4 h, and the bleeding was 420 mL.
Postoperative X-ray showed that the scapula was stabilized to the rib cage with
four cables as we planned without any signs of pneumothorax (Fig. 12.11).
232 12 Facioscapulohumeral Muscular Dystrophy (FSHD)

Fig. 12.10 Completion of

cable tightening. After all
the cables were completely
tightened, the remnant
bone chips were placed in
available space between
the scapula and the
rib cage

3.2 Post-operative Treatment

Postoperatively, the 45° abduction brace was worn full time (Fig. 12.12). On post-
­op day 1, passive range of motion exercise was started by a physical therapist. At
6 weeks post-op, active motion of the right shoulder was allowed after confirming
scapular stabilization on plain X-ray.
At 2 months post-op, abduction angle of the brace was lowered down to 30°. At
3 months, CT images were taken. After confirming bone union of the scapula to the
rib cage on CT images, the brace was removed (Fig. 12.13).
After brace removal, active range of motion exercise and muscle strengthening
exercise were started. At 6 months post-op, he could elevate the arm to 130°
(Fig. 12.14).
Since the scapula was fixed in slight upward rotation, the right shoulder looked
square compared to the left (Fig. 12.15). No complaints on this issue were heard
from the patient.
At 1-year post-op, CT showed complete bone union between the scapula and the
ribs (Fig. 12.16). He had no complaint of pain, fatiguability, or discomfort. The
range of motion was flexion 120°/120° (R/L), abduction 170°/150°, extension
40°/70°, external rotation (ER) 40°/50°, internal rotation (IR) T12/T3, abduction-
­ER 90°/90°, and abduction-IR 10°/60°. IR was slightly limited on the right side
3 Case Presentation 233

Fig. 12.11 Post-op X-ray.

The scapula was stabilized
to the rib cage with the
plate and four cables

a b

Fig. 12.12 Immobilization after surgery. The position of immobilization was 45° of abduction,
40° of flexion, and neutral rotation. This brace was used for 3 months post-op. (a) frontal view, (b)
lateral view
234 12 Facioscapulohumeral Muscular Dystrophy (FSHD)

Fig. 12.13 3D-CT

3 months post-op. Bone
union was confirmed on
this 3D-CT image, and the
brace was removed

Fig. 12.14 6 months

post-op. At 6 months
post-op, he could elevate
the arm to 130°
3 Case Presentation 235

Fig. 12.15 6 months

post-op. Since the right
scapula was fixed in
upward rotation and
abduction, the right
shoulder looked square
compared to the left sloped
shoulder in the hanging
arm position

Fig. 12.16 3D-CT

12 months post-op.
Complete bone union was
observed with abundant
callus formation bridging
over the ribs
236 12 Facioscapulohumeral Muscular Dystrophy (FSHD)

because the scapula was fixed in upward rotation so that the motions that required
downward rotation of the scapula such as IR were limited. Especially, it was impos-
sible to reach high on the back. Regarding the left shoulder, as the symptoms were
milder on the left and he did not feel great disturbance in activities of daily living,
he said he would like to wait and see. After the surgery, he found another job. He
looked much happier and more active than before the surgery. He said he was very
satisfied with the outcome.

References

1. Wijmenga C, et al. Location of facioscapulohumeral muscular dystrophy gene on chromosome

4. Lancet. 1990;336(8716):651–3.
2. Lemmers RJ, et al. Digenic inheritance of an SMCHD1 mutation and an FSHD-permissive D4Z4
allele causes facioscapulohumeral muscular dystrophy type 2. Nat Genet. 2012;44(12):1370–4.
3. Horwitz MT, Tocantins LM. Isolated paralysis of the serratus anterior (magnus) muscle. J Bone
Jt Surg Am. 1938;20(3):720–5.
4. Copeland SA, et al. The shoulder in patients with muscular dystrophy. Clin Orthop Relat Res.
1999;368:80–91.
5. Van Tongel A, et al. Medium to long-term outcome of thoracoscapular arthrodesis with screw
fixation for facioscapulohumeral muscular dystrophy. J Bone Jt Surg Am. 2013;95(15):1404–8.
6. Giannini S, et al. Fixation of winged scapula in facioscapulohumeral muscular dystrophy. Clin
Med Res. 2007;5(3):155–62.
7. Bunch WH. Scapulo-thoracic fusion. Minn Med. 1973;56(5):391–4.
Chapter 13
Trapezius Paralysis

The trapezius is innervated by the accessory nerve. The accessory nerve transverses
the neck in the superficial location, which makes it vulnerable to injury. The scapula
is connected to the trunk only through the clavicle so that its stability on the thorax
depends mainly on the parascapular muscles, especially the trapezius. When trape-
zius paralysis occurs, it is difficult to maintain the position of the scapula and it
shifts downward and laterally (Fig. 13.1).

a b

Fig. 13.1 Scapular position in trapezius paralysis. With trapezius paralysis, the right shoulder
girdle is pulled downward and laterally by the weight of the arm. (a) The right shoulder is lower
than the healthy left one. (b) Viewed from behind, the right scapula is shifted downward and away
from the spine

© The Author(s), under exclusive license to Springer Nature Singapore Pte 237
Ltd. 2023
E. Itoi, Shoulderology, https://doi.org/10.1007/978-981-99-0345-0_13
238 13 Trapezius Paralysis

1 Diagnosis and Treatment

In case with complete paralysis of the trapezius, scapular upward rotation which is
necessary for arm elevation cannot be performed by the trapezius, and as a result,
scapular winging occurs as seen in patient with FSHD with restricted arm elevation.
Both active flexion and active abduction are restricted to 70°–80° due to downward
rotation of the scapula (Fig. 13.2).

a b

Fig. 13.2 Active arm elevation. Active flexion (a) and abduction (b) of the right arm were limited
to 70°–80° with marked scapular winging due to trapezius paralysis
2 Case Presentation 239

2 Case Presentation

A 22-year-old male had lymph node swelling of the neck and visited a nearby ENT
clinic, where he underwent a biopsy of the cervical lymph node. The diagnosis was
necrotizing lymphadenitis. After the biopsy, he felt weakness in his right shoulder
and arm elevation was limited to approximately 60°. With a diagnosis of accessory
nerve injury, he had muscle exercise and motion training for 10 months without
improvement. Then he was referred to a hospital, where he underwent neurorrhaphy
of the injured accessory nerve. Although slight regeneration pattern of the upper
trapezius appeared on electromyogram, his arm elevation remained the same level.
Therefore, he was referred to our institute 10 months after the neurorrhaphy and
20 months after the biopsy. His physical findings are shown in Figs. 13.1 and 13.2.
As both conservative treatment and neurorrhaphy did not work, we decided to
perform Eden-Lange procedure (Fig. 13.3). This procedure was first reported by
Eden in 1924, who performed the procedure in two cases. Later, Lange reported
excellent outcome of this procedure in 1951. After this report, this procedure became
widespread and was called Eden-Lange procedure. In this procedure, the levator

a b

Fig. 13.3 Eden-Lange procedure. (a) The scapula is shifted downward and laterally due to trape-
zius paralysis. (b) The levator scapulae (LS) is transferred to the scapular spine and the rhomboid
major (RMj) and rhomboid minor (RMn) are transferred to the center of the infraspinous fossa to
compensate for the lost function of the trapezius by these muscles. (Reproduced with modification
from Fig. 2, Bigliani et al. [2])
240 13 Trapezius Paralysis

scapulae is released from the superior medial border of the scapula and transferred
to the scapular spine to replace the upper trapezius function to rotate the scapula
upward, and the rhomboid major and minor are released from the medial border of
the scapula and advanced to the center of the infraspinous fossa to replace the mid-
dle trapezius function to adduct the scapula (scapular retraction). A modified proce-
dure was reported to transfer the rhomboid major to the infraspinous fossa and the
rhomboid minor to the supraspinous fossa [1].

2.1 Surgery

Under general anesthesia, the patient was in the left lateral decubitus position with
the right arm flexed in a position of holding a pillow. A 20-cm longitudinal skin
incision was made midway between the spinous processes and the medial scapular
border, from cephalad to the superior angle to the inferior angle of the scapula
(Fig. 13.4).
The trapezius, which was thin due to muscle atrophy, was transected along the
medial border of the scapula to expose the levator scapulae and the rhomboids
(Fig. 13.5).
The infraspinatus muscle was elevated from the infraspinous fossa for 5 cm from
the medial scapular border. The supraspinatus was also elevated from the supraspi-
nous fossa just to expose the insertion site of the levator scapulae muscle. Next, with
use of an osteotome, the levator scapulae and the rhomboids were detached from the
medial scapular border with a thin portion of bone (bone sliver) (Fig. 13.6). During
this process, care should be taken to avoid damage to the serratus anterior, which
was on the ventral side of the scapula.
The detached rhomboids were easily pulled laterally to the center of the infraspi-
nous fossa. Using a 1.8-mm Kirschner wire, seven drill holes were made 4 cm lat-
eral from the medial scapular border. No. 2 polyethylene sutures were passed

Fig. 13.4 Skin incision. In

the left lateral decubitus
position, a 20-cm skin
incision was made parallel
to the medial border of the
scapula
2 Case Presentation 241

Fig. 13.5 Exposure of

donor muscles. The donor
muscles inserted into the
medial border of the
scapula (scap) were
exposed. LS levator
scapulae, RMn rhomboid
minor, RMj rhomboid
major

Fig. 13.6 Release of the

levator scapulae and the
rhomboids. These donor
muscles were detached
from the medial scapular
border (Scap) with a thin
bone sliver (Bone)

through these holes and through the rhomboids just medial to the bone sliver
(Fig. 13.7).
By tying these sutures, the rhomboids were firmly fixed to the center of the scap-
ular body (Fig. 13.8).
Since the levator scapulae had little mobility, proximal mobilization for at least
5 cm was necessary to facilitate transfer of the muscle. Once the sufficient mobiliza-
tion was achieved, a 4-cm transverse skin incision was made starting from the acro-
mial angle and extending medially along the scapular spine. The trapezius and the
deltoid were carefully released from the scapular spine and the bony bed was pre-
pared on the superior aspect of the scapular spine using a cutting burr. A tunnel
connecting the medial and lateral wounds were made beneath the trapezius. The
levator scapulae was passed through this tunnel and was secured to the scapular
spine with use of three suture anchors (Figs. 13.9 and 13.10).
242 13 Trapezius Paralysis

Fig. 13.7 Preparation of

rhomboid fixation. The
sutures passing through
seven drill holes in the
infraspinous fossa (*) were
passed through the
rhomboids just medial to
the bone sliver (Bone)

Fig. 13.8 Fixation of the

rhomboids. The sutures
were tied over the bone
sliver so that the
transferred rhomboids
(RMn, RMj) were firmly
attached to the center of
the scapular body

Fig. 13.9 Preparation of

bony bed on the scapular
spine. Bony bed was
prepared on the scapular
spine, and three suture
anchors were inserted
2 Case Presentation 243

Fig. 13.10 Fixation of the

transferred levator
scapulae. The levator
scapulae (LS) was passed
through the tunnel
connecting the medial and
lateral wounds and
securely fixed to the
scapular spine

Fig. 13.11 6 months

post-op. Active arm
elevation was fully
recovered after the surgery

2.2 Post-operative Treatment

Post-operatively, a 60° abduction brace was worn. Passive range of motion exercises
by a physical therapist were started immediately after the surgery to avoid contrac-
ture. Also, isometric exercises of the cuff muscles and the deltoid were added. At
6 weeks post-op, active and active-assistive arm elevation was started. At the same
time, the brace was gradually lowered to prepare for removal, and it was removed at
8 weeks post-op.
Active arm elevation was 125° at 2 months and 175° at 3 months post-op. Then,
the patient was instructed to perform push up, shoulder shrug, and scapular adduc-
tion. At 6 months post-op, although he felt slight weakness of his right shoulder, the
range of motion was normal, and he was highly satisfied (Fig. 13.11).
244 13 Trapezius Paralysis

References

1. Bigliani LU, et al. Transfer of the levator scapulae, rhomboid major, and rhomboid minor for
paralysis of the trapezius. J Bone Jt Surg Am. 1996;78(10):1534–40.
2. Bigliani LU, Perez-Sanz JR, Wolfe IN. Treatment of trapezius paralysis. J Bone Jt Surg Am.
1985;67(6):871–7.
Index

A Bicipital-forearm angle (BF angle), 190, 191

Acromial spur, 48 Bipolar lesion, 92, 99–108
Acromioclavicular joint, 9, 75, 194 Bone loss, 3, 86, 92–111, 113–115, 126–131
Acromioclavicular joint dislocation, 9 Bone scintigraphy, 51, 53
Acromiohumeral interval, 54, 55 Bony defect, 93, 108, 109, 114, 124
Acromion, 9–12, 22, 23, 40–46, 54, 55, 62, 66, Bristow procedure, 127–129
179, 194 Bursal-side tear, 41
Acromion index, 43–46
Arthrography, 56, 157, 161, 183
Arthroscopic Bankart repair, 95, 125–127 C
Arthroscopic capsular release, 156, 162, Capsular plication, 83, 140, 203
163, 165 Capsular shift, 85, 140–142
Arthroscopic rotator cuff repair, 64, 185 Capsular tear, 115
Arthroscopic subacromial decompression Caspari procedure, 125
(ASD), 62, 63 Central track, 107, 127
Articular-side tear, 39, 41 Circumpennate muscle, 29, 30
Axillary pouch, 80, 91, 137, 138, 152, 153, Codman exercise, 160
157, 158 Complete tear, 179
Compression fracture, 92, 97, 98, 108–110
Concavity-compression effect, 86, 136, 137
B Connolly exercise, 161
Bankart lesion, 79, 80, 91, 112–113, 115–121, Contralateral comparison method, 114
125, 126, 172 Coracoacromial arch, 38, 40–47, 52, 62, 192
Bankart repair, 63, 92, 93, 95, 100–102, Coracobrachialis muscle, 33, 176
124–128, 135 Coracohumeral ligament, 64, 77, 157, 159, 163
Bernageau view, 113 Coracoid transfer, 96, 126–131
Best-fit circle method, 114 Critical shoulder angle, 45–47, 54
Biceps brachii muscle Cross-body stretch, 201
long head of the biceps (LHB), 167–171 Cuff tear arthropathy, 54, 65
stabilizing function, 167, 171–173
tear, 179–186, 197
tendinitis, 178–180, 183, 185, 197 D
tendon, 167–171, 179–186 Deadman theory, 207, 208, 210, 213, 214
short head of the biceps (SHB), 33, 172, Delle, 9, 12
176, 177, 194 Dialysis arthropathy, 180
stabilizing function, 3 Dropping sign, 16

© The Editor(s) (if applicable) and The Author(s), under exclusive license to 245
Springer Nature Singapore Pte Ltd. 2023
E. Itoi, Shoulderology, https://doi.org/10.1007/978-981-99-0345-0
246 Index

E I
Eden-Lange procedure, 239 Inferior glenohumeral ligament (IGHL)
Elastography, 50, 51 complex, 77–80, 91, 115, 116,
End-range of motion, 101, 130 118, 126
End-range stabilizer, 79–80 Immobilization, 3, 67, 116–124, 151, 152,
Engaging lesion, 103 186, 226, 233
External rotation immobilization, 119, 122 position, 67, 116–123, 233
External rotation lag sign, 14, 15 Impingement
anterosuperior impingement, 179
internal impingement, 189, 200
F posterosuperior impingement, 200
Facioscapulohumeral muscular dystrophy sign, 11, 39–41, 58
(FSHD), 223–236, 238 subacromial impingement, 178
Floating IGHL, 91 subcoracoid impingement, 159
Freezing phase, 155, 157, 160, 162 syndrome, 40, 41, 62
Frozen phase, 155–157 test, 11, 41
Frozen shoulder, 223 Incomplete tear, 179
Full-thickness tear, 9 Inferior glenohumeral ligament (IGHL), 76,
77, 169–171
Infraspinatus muscle, 50, 125, 240
G Instability
Gadolinium, 56, 57 atraumatic instability, 20
Ganglion, 7, 8, 54 chronic instability, 124
Glenohumeral internal rotation deficit (GIRD), multidirectional instability (MDI), 9,
189–190, 192, 197, 201 20–25, 85, 88, 135
Glenohumeral joint, 45, 54, 57, 75, 76, 81, 115, recurrent instability, 111–112
124, 147, 160, 176, 183, 194, 197 traumatic instability, 20
Glenohumeral ligament, 20, 76, 77, 80, 81, Internal rotation lag sign, 17, 18
169–171, 192 Intra-articular pressure, 83–85, 140
Glenoid bone loss Intramuscular tendon, 29–32, 58
erosion-type bone loss, 109, 114 Inverted pear glenoid, 95
fragment-type bone loss, 109
Glenoid defect, 93–97, 99, 105, 126, 127, 130
Glenoid osteotomy, 142–144 J
Glenoid track, 3, 101–108, 127 Joint distension, 157, 161–162
Graft
fascial patch graft, 63–65
iliac crest bone graft, 96, 127, 143, 144, 231 L
Labral injury, 192
Labral repair, 203
H Labrum, 78, 79, 86, 101, 115–118, 120, 125,
Hawkins impingement sign, 11, 13, 41 126, 169, 170, 179, 192–194, 198
Hill–Sachs lesion, 92, 97–102, 106–111, 113, Latarjet procedure, 126–131
126, 127 Lawnmower exercise, 201, 202
Hornblower's sign, 14–16, 34 Lésion de Malgaigne, 97
Horwitz maneuver, 225 Loose shoulder, 21, 135
Humeral Avulsion of the Glenohumeral
Ligament (HAGL) lesion, 91
Humeral head, 3, 7, 9, 20, 22–24, 29, 34, M
42–46, 54–56, 64, 66, 75, 76, 81, Manipulation under anesthesia, 156, 161,
82, 85–87, 91–94, 96–98, 100–106, 162, 165
108–110, 113, 126, 136–139, 142, Manual muscle testing (MMT), 11, 14
150, 167, 168, 171–175, 179, 180, Middle glenohumeral ligament (MGHL), 77,
189, 190, 193 80, 91, 154, 169, 170
Humeral retroversion, 190–191 Mid-range of motion, 109
Index 247

Mid-range stabilizer, 81–88, 136 Rotator cuff

Moment arm, 33, 34, 65, 176, 194 muscle, 30, 33, 34, 46, 117, 152, 172,
Morgan procedure, 125 174, 189
MR arthrography, 56, 115, 117, 138 tear
Multi-circumpennate muscle, 30 asymptomatic, 51, 57, 61
Muscle degenerative, 39, 56, 62, 179, 186
atrophy, 7, 11, 49, 56, 197, 226, 240 symptomatic, 51
extensibility, 50–51 traumatic, 62, 179, 186
fatty change, 49–50, 56 tendinitis, 39, 62, 147, 197
fatty infiltration, 50 tendon, 39, 40, 42, 50, 52, 59, 61, 62, 65,
flexibility, 50, 195, 196, 201 192, 207, 209

N S
Neer impingement sign, 11, 12, 41 Safe distance between suture anchors, 217
NSAID, 57, 160, 186 Scapular
brace, 140
dyskinesis, 135, 137, 189, 197, 199, 200
O inclination, 86, 87
Off-track lesion, 102, 103, 108, 127, 129 motion, 10, 30, 112, 140, 156, 157, 199
On-track lesion, 101, 106, 108, 127 plane, 12, 13, 39, 58, 67, 110, 173,
Osteoarthritis, 43, 45, 47, 51, 65, 126, 131 176, 187
winging scapula, 10, 58, 224–228, 238
Y-view, 113
P Scapulohumeral rhythm, 156
Pain Scapulopexy, 226
motion, 10, 11, 52, 135 Scapulothoracic arthrodesis, 226–228
neuropathic, 57 Scapulothoracic joint, 75, 160, 185
night, 52–54, 57, 155 Serratus anterior muscle, 10, 142, 200, 201,
rest, 38, 52, 57 223, 228, 240
Painful arc, 11 Shaking, 58
Partial-thickness tear, 40, 48, 56, 57, Shoulder
64, 192 blood circulation, 155
Pectoralis major muscle, 7, 17, 33, 34, 36, dislocation
109, 135, 143, 185 first-time, 116–120, 123, 124
transfer, 135, 141–143 primary, 123, 124
Pectoralis minor muscle, 200 recurrent, 20, 92, 97, 108, 111–112,
Pendulum exercise, 160–161 115–117, 124
Periarthritis, 147, 150 examination, 2, 41
Peripheral track, 107, 127 frozen shoulder, 106, 147–162
Physical therapy, 58, 155, 156, 160, 185 instability, 20–25, 75, 89, 90, 96–99, 126,
Pitching motion, 187–189, 192, 193 149, 172
Popeye sign, 7, 8, 181, 182, 185 Little Leaguer's shoulder, 188
Positron emission tomography (PET), motion, 10–11, 33, 34, 52, 75, 79, 85, 99,
33–35, 51, 52 120, 148, 155
motion restriction band, 118, 120
pain, 7, 38, 52, 53, 62, 147, 148, 157,
R 179, 195
Recurrence rate, 3, 95, 97, 111, 117, 118, stiff shoulder, 10, 107, 147, 148, 150–152
123–127, 162 thrower's shoulder, 187
Recurrent dislocation, 3, 20, 97, 111, throwing shoulder, 187–190
115–117, 124 SLAP repair, 203
Remplissage, 100, 101, 126, 127 Sleeper stretch, 201
Reverse shoulder arthroplasty, 65 Slipping, 138, 139, 142
248 Index

Smoking, 38, 47–49, 150 anterior drawer test, 23

Sprengel deformity, 7 bear hug test, 17, 19
Stability ratio, 86, 87, 95 belly press test, 17, 19
Sternoclavicular joint, 75 biceps tension test, 197
Stooping exercise, 160 empty can test, 12–14
Stretching, 58, 118, 160, 161, 163, 188, 189, full can test, 14
195, 199, 200 hyperabduction test, 22, 24, 137
Stryker view, 113 impingement test, 11, 41
Subacromial bursitis, 51, 52, 178 lift-off test, 17
Subacromial decompression, 62–63 posterior apprehension test, 20, 22
Subacromial impingement, 178, 188, 189, posterior drawer test, 22, 23, 137
192, 200, 201 scapular compression test, 225, 227
Subcoracoid fat triangle, 157, 159 Speed test, 180, 181, 197
Subscapularis muscle, 34, 35, 130 supraspinatus test, 12, 14
Subscapularis tear, 17–20, 51 Yergason test, 180, 181
Sulcus sign, 22, 23, 137 Thawing phase, 155, 156, 160–161
Superior glenohumeral ligament (SGHL), 77, Thixotropy, 189
169, 170 Throwing motion, 187, 189, 191
Superior labrum anterior and posterior lesion Transosseous-equivalent repair (TOE repair),
(SLAP lesion), 188, 192 61, 63, 64
Suprapectoral tenodesis, 185 Transosseous repair, 59, 60, 63
Supraspinatus muscle, 29, 46, 49–51, 58, 228 Trapezius
Supraspinatus tendon, 9, 11, 12, 29, 30, paralysis, 237–244
39–41, 43, 44, 46, 48, 50, 54–57, 59 strengthening, 200
Suture anchor insertion angle, 59–61, 63, 101, Traumatic dislocation, 20, 22, 75, 89
125, 129, 131, 185, 207, 209, 210, Triceps brachii muscle, 33, 176, 192
212, 213, 217, 218, 241, 242 long head of, 192

T U
Tenodesis Uni-circumpennate muscle, 30
proximal tenodesis, 185
subpectoral tenodesis, 185
suprapectoral tenodesis, 185 W
Tenotomy, 65, 185 Western Ontario Shoulder Instability Index
Teres minor muscle, 14–16, 29, 30, 32, (WOSI), 97, 107
34, 35, 188 West-Point view, 113
Teres minor tear, 14–17
Test
abduction inferior stability test (ABIS test), Z
22, 24, 137 Zipper phenomenon, 49
anterior apprehension test, 20, 21, 91, 103