Articles

Unique and shared systemic biomarkers for

emphysema in Alpha-1 Antitrypsin deficiency and
chronic obstructive pulmonary disease
K.A. Serban,a,b* K.A. Pratte,a C. Strange,c R.A. Sandhaus,a A.M. Turner,f T. Beiko,c D.A. Spittle,g L. Maier,a,b N. Hamzeh,h
E.K. Silverman,d,e B.D. Hobbs,d,e C.P. Hersh,d,e D.L. DeMeo,d,e M.H. Cho,d,e and R.P. Bowler a,b*
a
Department of Medicine, Division of Pulmonary, Critical Care, and Sleep Medicine, National Jewish Health, Denver, CO,
United States
b
Department of Medicine, Division of Pulmonary Sciences and Critical Care Medicine, University of Colorado, Aurora, CO,
United States
c
Pulmonary, Critical Care, Allergy and Sleep Medicine, Medical University of South Carolina, Charleston, SC, United States
d
Channing Division of Network Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA, United
States
e
Division of Pulmonary and Critical Care Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA,
United States
f
Institute for Applied Health Research, University of Birmingham, Birmingham, UK
g
Institute of Inflammation and Aging, University of Birmingham, UK
h
Pulmonary, Critical Care, Allergy and Sleep Medicine, University of Iowa, Iowa City, IA, United States

Summary
Background Alpha-1 Antitrypsin (AAT) deficiency (AATD), the most common genetic cause of emphysema eBioMedicine 2022;84:
presents with unexplained phenotypic heterogeneity in affected subjects. Our objectives to identify unique and 104262
shared AATD plasma biomarkers with chronic obstructive pulmonary disease (COPD) may explain AATD pheno- Published online 22 Sep-
tember 2022
typic heterogeneity.
https://doi.org/10.1016/j.
ebiom.2022.104262
Methods The plasma or serum of 5,924 subjects from four AATD and COPD cohorts were analyzed on SomaScan
V4.0 platform. Using multivariable linear regression, inverse variance random-effects meta-analysis, and Least
Absolute Shrinkage and Selection Operator (LASSO) regression we tested the association between 4,720 individual
proteins or combined in a protein score with emphysema measured by 15th percentile lung density (PD15) or diffu-
sion capacity (DLCO) in distinct AATD genotypes (Pi*ZZ, Pi*SZ, Pi*MZ) and non-AATD, PiMM COPD subjects.
AAT SOMAmer accuracy for identifying AATD was tested using receiver operating characteristic curve analysis.

Findings In PiZZ AATD subjects, 2 unique proteins were associated with PD15 and 98 proteins with DLCO. Of
those, 68 were also associated with DLCO in COPD also and enriched for three cellular component pathways: insu-
lin-like growth factor, lipid droplet, and myosin complex. PiMZ AATD subjects shared similar proteins associated
with DLCO as COPD subjects. Our emphysema protein score included 262 SOMAmers and predicted emphysema
in AATD and COPD subjects. SOMAmer AAT level <7.99 relative fluorescence unit (RFU) had 100% sensitivity
and specificity for identifying Pi*ZZ, but it was lower for other AATD genotypes.

Interpretation Using SomaScan, we identified unique and shared plasma biomarkers between AATD and COPD
subjects and generated a protein score that strongly associates with emphysema in COPD and AATD. Furthermore,
we discovered unique biomarkers associated with DLCO and emphysema in PiZZ AATD.

Funding This work was supported by a grant from the Alpha-1 Foundation to RPB. COPDGene was supported by
Award U01 HL089897 and U01 HL089856 from the National Heart, Lung, and Blood Institute. Proteomics for
COPDGene was supported by NIH 1R01HL137995. GRADS was supported by Award U01HL112707, U01

Abbreviations: AATD, Alpha-1 Antitrypsin Deficiency; COPD, Chronic Obstructive Pulmonary Disease; PD15, 15th percentile lung
density; DLCO, diffusion capacity; GRADS, The Genomic Research in Alpha-1 Antitrypsin Deficiency and Sarcoidosis; QUAN-
TUM-1, QUANTitative Chest Computed Tomography UnMasking Emphysema Progression in Alpha-1 Antitrypsin Deficiency;
COPDGene, Genetic Epidemiology of COPD
*Corresponding authors.
E-mail addresses: [email protected] (K.A. Serban), [email protected] (R.P. Bowler).
One-sentence summary: Systemic biomarkers predict emphysema in AATD and COPD.

www.thelancet.com Vol 84 October, 2022 1

 Articles

HL112695 from the National Heart, Lung, and Blood Institute, and UL1TRR002535 to CCTSI; QUANTUM-1 was
supported by the National Heart Lung and Blood Institute, the Office of Rare Diseases through the Rare Lung Dis-
ease Clinical Research Network (1 U54 RR019498-01, Trapnell PI), and the Alpha-1 Foundation. COPDGene is also
supported by the COPD Foundation through contributions made to an Industry Advisory Board that has included
AstraZeneca, Bayer Pharmaceuticals, Boehringer-Ingelheim, Genentech, GlaxoSmithKline, Novartis, Pfizer, and
Sunovion.

Copyright Ó 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Keywords: SomaScan; Protein score; Emphysema; Plasma biomarker; Alpha-1 antitrypsin deficiency

Implications of all the available evidence

Research in context Our study supports the use of proteomic platforms for
detection of early emphysema diagnosis and severity
Evidence before this study stratification and also for screening populations for
Alpha-1 Antitrypsin deficiency (AATD) is typically diag- severe- and intermediate-deficient AAT genotypes. Fur-
thermore, SomaScan has excellent diagnostic character-
nosed by Alpha-1 Antitrypsin (AAT) protein measure-
ments and genetic sequencing in individuals istics for severe- and intermediate-deficient AAT
presenting with emphsyema on chest computer tomog- genotypes. Further work is needed to determine
whether these individual biomarkers and/or the protein
raphy (CT). AATD carriers (PiMZ), may develop emphy-
sema if secondary risk factors, like cigarette smoking are score are useful to assess progression of emphysema
present, however most individuals with deficient geno- and airflow obstruction as well as other comorbidities,
types, PiZZ or PiSZ with mild emphysema remain undi- such as exacerbations in AATD individuals.
agnosed or have delayed diagnosis due to near normal
spirometry or diffusion capacity (DLCO). While CT is
diagnostic of emphysema it is associated with radiation
exposure which limits ist use as a biomarker. Hence, Introduction
there is a critical need to develop non-invasive and Alpha-1 Antitrypsin deficiency (AATD), a genetic dis-
accessible biomarkers, e.g. plasma biomarkers, that ease that accounts for 1-2% of chronic obstructive pul-
detect and predict emphysema in AATD at risk individu- monary disease (COPD),1,2 is caused by mutations in
als. Furthermore, there is little known regarding Serpin Family A Member 1 (SERPINA1) gene. The nor-
the plasma proteome in AATD emphsyema and mal SERPINA1 allele is Pi*M, with the most common
whether there is overlap with non-AATD, cigarette disease variants referred to as Pi*Z and Pi*S alleles,
smoke-induced chronic obstructive pulmonary disease
characterized by low AAT serum levels. The mean allele
(COPD)/emphysema.
frequencies in Europe are 0-3% (mean=1.5%) for Pi*Z
Added value of this study and 1-13% for Pi*S (mean=3.3%).3 Mainly the Pi*ZZ
and Pi*SZ severe-deficient genotypes account for severe
Our report comprehensively evaluates the plasma pro- AATD, that presents with clinical significant emphy-
teome using SomaScan platform in three large AATD sema and requires Alpha-1 Antitrypsin (AAT) augmen-
cohorts and one non-AATD, COPD cohort. We identified
tation therapy to prevent emphysema progression.1
two proteins, betacellulin (BTC) and Cyclin Dependent
Kinase-2-associated protein-1 (CDK2AP1) which are Although the genotype predicts the emphysema risk,
uniquely associated with the emphysema in PiZZ indi- AAT levels do not correlate with disease onset or pro-
viduals. We also created an emphysema severity score gression within the genotypes, especially in the interme-
using 262 SomaScan proteins. The score predicts diate-deficient, PiMZ and PiMS subjects.4,5 Thus,
emphysema similarly in AATD and COPD individuals. additional AATD modifiers explain disease heterogene-
While the protein score performed better in AATD and ity and biomarkers are needed to stratify AATD individ-
COPD individuals with less severe emphysema, the pro- ual at risk for emphysema development.
tein score was not superior to DLCO at predicting Current clinical measurements of emphysema,
emphysema in AATD inviduals with advanced emphy- including spirometry [forced expiratory volume at 1 sec-
sema or PiZZ individuals treated with augmentation
ond (FEV1)], diffusion capacity (DLCO), and computer
therapy. Lastly, we report that SomaScan is both sensi-
tive and specific for AATD individuals and carriers and tomography (CT) imaging [visual emphysema score and
would be suitable for identifying previously undiag- the 15th percentile lung density (PD15)] typically identify
nosed subjects. subjects with severe disease. Within the same genotype,

2 www.thelancet.com Vol 84 October, 2022

 Articles

these clinical measurements vary widely,6,7 are not spe- Participant baseline characteristics are presented in
cific to AATD, and only abnormal when AATD is clini- Table 1.
cally advanced. Even emerging systemic biomarkers for COPDGene is a multi-center, longitudinal cohort
COPD, such as C reactive protein (CRP), fibrinogen, funded by the National Heart, Lung, and Blood Institute
interleukin-6 (IL-6) and interleukin-8 (IL-8), and solu- (NHLBI) which enrolled >10,000 non-Hispanic White
ble receptor for advanced glycation end products and African Americans adults with a smoking history of
(sRAGE)8 10 may not be shared with AATD-associated >10 pack-years and either with and without COPD.16
emphysema, since AATD subjects were frequently The cohort specifically recruited for genome-wide asso-
excluded from COPD biomarker studies. Dedicated ciation studies, but other significant clinical, functional,
studies that investigated plasma biomarkers in AATD laboratory, and radiological data were collected at 21 cen-
subjects, using Myriad Discovery MAP and Bio-Rad/ ters across the USA.
Bio-Plex 200 platforms, included small cohorts, and GRADS is a multi-center, cross-sectional cohort of
lacked replication or PiMM COPD groups.11,12 adults older than age 35 years with PiZZ or PiMZ
In the last decade we witnessed rapid advance in Alpha-1 Antitrypsin genotypes. It was designed to con-
unbiased high-throughput proteomic approaches that duct state-of-art genomic, microbiomics and phenotypic
measure thousands of proteins in a single assay. For studies to better understand AATD.
example, SomaScan, an aptamer-affinity-based technol- QUANTUM-1 is multicenter, longitudinal cohort
ogy, allows for multiplexed measurements of >5,000 funded by the National Institutes of Health Office of
proteins.13,14 Since multiple publications have shown Rare Diseases and NHLBI to study radiographic emphy-
that SomaScan is reproducible14 and that many SomaS- sema progression in adults with PiZZ AATD and normal
can biomarkers are highly correlated with traditional lung function. The cohort recruited 51 AATD patients
antibody biomarker-based assays,15 this platform is with normal Forced Expiratory Volume in the first second
increasingly used in clinical biomarker discovery. (FEV1)
80% predicted to determine whether baseline
In this study we measured SomaScan profiles in four CT measurements of emphysema, e.g. lung density, pre-
independent cohorts including a large non-AATD popu- dicted a more rapid decline in FEV1.
lation [COPDGene]16 and three AATD populations [The The Birmingham Alpha-1 Antitrypsin registry
Genomic Research in Alpha-1 Antitrypsin Deficiency and enrolls patients with all AATD phenotypes that undergo
Sarcoidosis (GRADS),17 QUANTitative Chest Computed clincially-indicated pulmonary function, chest CT, and
Tomography UnMasking Emphysema Progression in laboratory tests in Birmingham, England.
Alpha-1 Antitrypsin Deficiency (QUANTUM-1),11 and Bir-
mingham Alpha-1 Antitrypsin registry (Birmingham)]. Clinical data and definitions
We hypothesized that our unbiased proteomic approach COPD was defined using spirometric evidence of air-
would identify new plasma AATD biomarkers for flow obstruction: i.e., post-bronchodilator FEV1/ forced
emphysema that correlate with clinical outcomes. vital capacity (FVC) < 0.70. FEV1 and FVC maneuvers
were recorded per ATS/ERS standards for spirometry in
COPDGene, GRADS, and QUANTUM-1. The severity
Methods of COPD was based on the Global Initiative for Chronic
Obstructive Lung Disease (GOLD) guidelines using
Study populations post-bronchodilator FEV1% as follows: GOLD 1 (

The study cohorts included 5,924 subjects enrolled in 80%); GOLD 2 (
50% and <80%); GOLD 3 (
30%
the COPDGene16 (N= 5,607), the Alpha-1 Antitrypsin- and <50%); or GOLD 4 (<30%). Subjects with FEV1/
Deficiency and Sarcoidosis17 (GRADS) (N=133), the FVC
0.70 and FEV1% < 80% predicted were defined
QUANTitative Chest Computed Tomography UnMask- as having Preserved Ratio Impaired Spirometry
ing Emphysema Progression in Alpha-1 Antitrypsin (PRISm).18 Subjects with FEV1/FVC
0.70 were
Deficiency11 (QUANTUM-1) studies (N=38), and in the defined as controls (GOLD 0). DLCO percent predicted
Birmingham Alpha-1 Antitrypsin registry (N=146), was based on the GLI (traditional units/min/mmHg)
Figure 1. All subjects had AAT genotyping, at least one and adjusted for altitude, age, height, and sex.
spirometry measurement and one plasma or serum Emphysema was reported as the 15th percentile
SomaScan proteomic assay. The respective local Institu- adjusted lung density (PD15, g/L) measured as the
tional Review Boards (IRB) approved all study protocols Hounsfield units (HU) below which the 15% of voxels
and informed consent was obtained from all partici- with the lowest density are distributed, adjusted for the
pants [COPDGene: HS 1883 National Jewish Health; race-corrected total lung capacity and multiplied by
GRADS: Pro00024143 Medical University of South 1000 to be expressed as PD15, as previously
Carolina; QUANTUM-1: HR17301 - Medical University described.19 In addition, percent emphysema was
of South Carolina; Birmingham Alpha-1 Antitrypsin defined as the percentage of lung voxels  -950 HU (%
registry: 18/SC/0541 South Central Oxford C low attenuation areas, % LAA) on the full inspiratory
Research Ethics Committee]. scans.

www.thelancet.com Vol 84 October, 2022 3

 Articles

Figure 1. Study consort. Diagram representation of the individuals recruited in the COPDGene, GRADS, QUANTUM, and Birming-
ham cohorts but excluded in the current study due to absent SomaScan data and exclusion genotypes. Genotypes not included:
COPDGene Pi*FZ, Pi*SZ, Pi*SS; QUANTUM-1 - Pi*MheerlenZ and Pi*SZ; GRADS- discordant genotype / phenotype measurements;
Birmingham - Pi*MMaltonNull, Pi*MprocidaZ, Pi*SS, Pi*ZNull, Pi*FZ, Pi*IZ, Pi*MmaltonZ, Pi*SPLowell.

Visual emphysema severity recorded in the elec- Amendments (CLIA) certified laboratory. Subjects
tronic medical record was used in the Birmingham were included if their serum AAT was less than
cohort. For the purpose of this manuscript visual 11uM or 80mg/dL. {https://www.clinicaltrials.gov/
emphysema was graded as present or absent. ct2/show/NCT00532805}. Two subjects were
excluded because they were liver transplant recipi-
ents, 2 subjects had exclusion genotypes Pi*SZ and
Alpha-1 Antitrypsin testing
Pi*MHeerlenZ, and 7 subjects were dropped from the
COPDGene subjects were genotyped for SERPINA1 analysis because the plasma sample did not meet
Z (rs28929474) and S (rs17580) alleles using 50 to the quality control criteria on the SomaScan
30 exonuclease assays (TaqMan assay, Applied Bio- (Figure 1).
systems, Foster City, CA) and known Pi*ZZ, Birmingham subjects with clinical testing using a
Pi*MZ, and Pi*MS (Coriell Institute for Medical combination of isoelectric focusing and sequencing
Research, Camden, NJ) control samples. Pi Protein were included if they were Pi*ZZ or Pi*SZ pheno-
Phenotyping: Isoelectric focusing was performed type, independent of the AAT serum levels. In the
on plasma samples from subjects with the Pi*ZZ, Birmingham SomaScan group 1 subject was not
Pi*SZ, and Pi*MZ genotypes as described.20 Dis- considered because was receiving augmentation
cordance between the genotyping and protein phe- therapy, 5 subjects were excluded because of >
notyping was resolved by medication review for 1year difference between plasma samples and spi-
AAT augmentation therapy as well as by repeat gen- rometry or CT measurements, 8 subjects had exclu-
otyping and protein phenotyping. sion genotypes Pi*MMaltonNull, Pi*MprocidaZ,
GRADS subjects used clinical tests positive for Pi*SS, Pi*ZNull, Pi*FZ, Pi*IZ, Pi*MmaltonZ, Pi*S-
Pi*MZ or Pi*ZZ genotypes.17 AAT augmentation PLowell, and 4 subjects were dropped from the analy-
therapy was documented during the medication rec- sis because the plasma sample did not meet the
onciliation at the time of study enrollment. Two quality control criteria on the SomaScan. (Figure 1).
subjects were excluded because had discordant
genotype phenotype measurements and 4 sub-
jects were dropped from the analysis because the Proteomics platform
plasma sample did not meet the quality control cri- SomaScan v4.0 uses 4,979 different SOMAmers
teria on the SomaScan (Figure 1) (aptamers) to quantify 4,776 unique human proteins
QUANTUM-1 subjects had the Pi*ZZ or Pi*Znull with 4,720 unique Uniprot numbers.15 SomaScan sig-
genotype confirmed by gene probe analysis. Previ- nal normalization included plate hybridization to con-
ous AAT genotype results were acceptable if docu- trol for variability across array signals, median signal
mented from a Clinical Laboratory Improvement normalization to control for technical variability of

4 www.thelancet.com Vol 84 October, 2022

www.thelancet.com Vol 84 October, 2022

Table 1. Total Population characteristics. COPDGene GRADS QUANTUM-1 Birmingham

MM MS MZ MZ ZZ ZZ SZ ZZ
N = 5101 N=347 N = 159 N = 56 N = 31 N = 46 N = 32 N=6 N=24 N = 122

Augmentation Therapy No No No No No Yes No Yes No No

AAT level
SOMAscan RFU (Natural Log) 10.63 (0.23) 10.37 (0.21) 9.94 (0.23) 10.08 (0.26) 6.63 (0.22) 10.66(0.55) 6.7 (0.24) 10.9 (0.36) 9.10 (0.16) 6.13 (0.26)
AAT µmol/L 13.0 (3.25) 10.4 (2.0) 14.35 (2.78) 4.56 (1.60)
Demographics
Age in years mean (SD) 65.0 (8.8) 66.2 (9.0) 67.7 (8.6) 53.8 (10.9) 52.3 (11.2) 60.5 (9.4) 51.4 (8.3) 56.5 (10.7) 52.7 (12.4) 53.1 (12.2)
Male N (%) 2565 (50.3%) 165 (47.6%) 85 (53.5%) 17 (30.4%) 14 (45.1%) 21 (45.7%) 9 (28.1%) 4 (66.7%) 6 (25.0%) 58 (47.5%)
AA N (%) 1548 (30.3%) 37 (10.7%) 11 (6.9%) 0% 0% 0% 0% 0% 0 (%) 0 (%)
NHW N (%) 3553 (69.7%) 310 (89.3%) 148 (93.1%) 56 (100%) 30 (96.8%) 46 (100%) 32 (100%) 6 (100%) 24 (100%) 121 (99.2%)
BMI (kg/m2) 28.9 (6.3) 28.6 (6.3) 29.4 (6.2) 28.2 (6.0) 26.4 (3.9) 27.8 (5.4) 29.5 (7.8) 25.5 (3.4) 26.5 (5.2) 26.9 (5.3)
Never Smokers N (%) 335 (6.6%) 24 (6.9%) 10 (6.3%) 30 (53.6%) 25 (80.7%) 19 (41.3%) 20 (62.5%) 1 (16.7%) 15 (62.5%) 50 (41.0%)
Current smoking status N (%) 1876 (36.8%) 100 (28.8%) 36 (22.6%) 3 (5.4%) 1 (3.2%) 1 (2.2%) 0% 0% 2 (8.3%) 3 (2.5%)
ATS Pack-years median (IQR) 38.5 (29.0) 39.0 (27.0) 40.0 (32.0) 0.00 (2.5) 0.00 (0.00) 2.1 (18.5) 0.0 (1.8) 6.0 (8.5) 0.00 (17.5) 4.5 (15.0)
Spirometry
COPD GOLD PRISm 603 (11.8%) 37 (10.7%) 13 (8.2%) 7 (12.7%) 2 (6.5%) 2 (4.4%) 1 (3.1%) 0% 0 (0%) 2 (1.6%)
GOLD 0 2335 (45.8%) 152 (43.8%) 67 (42.1%) 35 (62.5%) 17 (54.8%) 4 (8.7%) 24 (75.0%) 5 (83.3%) 15 (62.5%) 28 (23.0%)
GOLD 1 473 (9.3%) 31 (8.9%) 9 (5.7%) 6 (10.7%) 3 (9.7%) 2 (4.4%) 7 (21.9%) 1 (16.7%) 3 (12.5%) 14 (11.5%)
GOLD 2 951 (18.6%) 63 (18.2%) 29 (18.2%) 6 (10.7%) 6 (19.4%) 20 (43.5%) 0% 0% 2 (8.3%) 33 (27.1%)
GOLD 3 487 (9.6%) 41 (11.8%) 26 (16.4%) 0% 2 (6.5%) 16 (34.8%) 0% 0% 2 (8.3%) 30 (24.6%)
GOLD 4 195 (3.8%) 19 (5.2%) 12 (7.6%) 1 (1.8%) 0% 2 (4.4%) 0% 0% 2 (8.3%) 14 (11.5%)
FEV1 Percent Predicted mean (SD) 79.7 (24.6) 77.9 (26.4) 73.1 (27.8) 91.3 (20.6) 87.9 (21.5) 59.7 (21.0) 99.3 (11.8) 97.5 (9.0) 87.5 (32.7) 66.5 (30.2)
FEV1 post BD (Liter) mean (SD) 2.2 (0.9) 2.2 (0.9) 2.1 (1.0) 2.9 (0.9) 3.0 (1.0) 1.9 (0.8) 3.2 (0.8) 3.3 (0.8) 2.6 (1.1) 2.2 (1.1)
FEV1/FVC post BD mean (SD) 0.68 (0.14) 0.66 (0.16) 0.63 (0.18) 0.74 (0.13) 0.70 (0.16) 0.50 (0.15) 0.77 (0.10) 0.70 (0.10) 0.70 (0.19) 0.53 (0.20)
DLCO percent predicted mean (SD) 78.6 (23.0) 80.7 (24.7) 79.6 (24.2) 99.5 (25.3) 90.3 (23.6) 67.3 (25.5) 96.0 (17.6) 71.0 (8.1) 103.5 (24.5) 74.4 (23.8)
CT Emphysema
% emphysema (-950 Hu), total lung, median (IQR) 1.5 (4.7) 1.9 (5.5) 3.7 (9.2) 1.3 (3.1) 5.3 (10.0) 17.4 (19.4) 9.5 (8.1) 16.0 (16.1) NA NA
Adjusted lung density (g/l) mean (SD) 86.2 (25.0) 84.2 (26.9) 76.3 (26.3) 77.5 (20.2) 68.1 (17.2) 51.6 (19.9) 68.9 (20.8) 44.5 (14.2) NA NA
Visual (Yes/No) N (%) 4 (16.7%) 67 (54.9%)
Charlson index 3.4 (1.9) 3.6 (2.0) 3.6 (1.6) NA NA NA 1.6 (1.2) 3.0 (2.2) 1.8 (1.6) 2.1 (1.3)

Table 1: Patient characteristics. Data presented as mean § SD or median § IQR.

Abbreviations: COPDGene = Genetic Epidemiology of COPD; GRADS = Genomic Research in Alpha-1 Antitrypsin Deficiency and Sarcoidosis; QUANTUM = QUANTitative Chest Computed Tomography UnMasking
Emphysema Progression in Alpha-1 Antitrypsin Deficiency; AAT = Alpha-1 Antitrypsin; FEV1 = Forced expiratory volume in 1 second; FEV1/FVC = Forced expiratory volume in 1 second to forced vital capacity;
Dlco = Diffusing capacity for carbon monoxide, post BD = post bronchodilator; SD = standard deviation; IQR = interquartile range; GOLD = Global initiative for chronic Obstructive Lung Disease; PRISm = Preserved
Ratio Impaired SpiroMetry; Hu = Hounsfield units.

Articles
5
 Articles

replicates within a run, and plate scaling and calibration complete data for SomaScan and adjusted lung density
of SOMAmers to control for inter-assay variation (PD15). This dataset was randomly split 70/30% to cre-
between analytes and batch differences between plates. ate the training (n=3324) and testing (n=1421) datasets.
Finally, median normalization to a reference using Using the training dataset, a 10-fold cross-validation on
adaptive normalization by maximum likelihood was standardized variables was used to estimate the model.
applied within the dilution group to quality control rep- Using the test dataset’s mean squared errors (MSE), R2,
licates and individual samples to remove edge effects and Pearson correlations (between observed and pre-
and technical variance. Orthogonal data supporting dicted adjusted lung density), we evaluated whether to
aptamer specificity for target proteins has been previ- use the LASSO model based on the minimum mean
ously published.15 cross-validated error lambda or at 1-standard error from
the minimum. We used the lambda based on 1-standard
Statistical analysis error because it provided a more parsimonious model
with minimal effect on model fit. Predicted values for
Cross-sectional analysis of proteins COPD phenotype adjusted lung density were calculated for COPDGene
associations. Natural log transformed SomaScan pro- PiMZ; GRADS PiMZ, GRADS PiZZ not on and -on
teins, pulmonary function [FEV1, FEV1 percent pre- therapy; QUANTUM-1 PiZZ not on therapy; and Bir-
dicted (FEV1pp), FEV1 /FVC, DLCO], and emphysema mingham PiZZ. MSE, R-squared (R2), and Pearson cor-
measurements [visual emphysema and PD15] were relations between observed and predicted values were
treated as continuous variables. Multivariable linear calculated for each group, with the exception of the Bir-
regression was used to identify proteins significantly mingham group. Birmingham PiZZ had visual emphy-
associated with pulmonary function and emphysema. sema (Yes/No) data only, and logistic regression was
The regression model for FEV1 was adjusted for age, used to calculate the AUC for the predicted adjusted
age2, height, height2, sex, BMI, pack-years, current lung density association with visual emphysema. To
smoking and never smoked status. The FEV1pp model evaluate the emphysema score against another emphy-
included pack-years, and current smoking or never sema biomarker, e.g. DLCO we calculated the R2, Pear-
smoking status. The FEV1/FVC model included age, son correlation between DLCO and measured PD15.
sex, BMI, pack-years, current smoking and never smok-
ing status. The DLCO model was adjusted for BMI, Pathway enrichment analysis
pack-years, and current smoking and never smoking Using Gene Ontology enRIchment anaLysis and visuaLi-
status. In the PD15 model we controlled for age, sex, zAtion tool (GOrilla), we conducted a pathway enrich-
BMI, pack-years, current smoking and never smoking ment analysis on N=98 and N=1,306 SomaScan proteins
status. In the COPDGene cohort analyses, we further significantly associated with DLCO in the PiZZ not on
adjusted for race and included clinical center as a ran- therapy and the PiMM groups, respectively. A hypergeo-
dom effect. Because the ZZ cohorts were predominantly metric test was performed to determine significant
never smokers with very few current smokers, we only enrichment of cellular component GO terms. A color-
included never smoking status in their analysis. QUAN- coded trimmed directed acyclic graph (DAG) and a list of
TUM-1 ZZ on therapy with a small sample size of 6 was all significantly enriched GO terms was generated.21
not included in the analysis. Results across cohorts
were combined by genotype and treatment into an SomaScan diagnostic accuracy. Receiver operating
inverse variance random-effects meta-analysis. STROBE characteristic curve (ROC) analysis, plots and area
guidelines for cohort studies reporting were followed. under the curves (AUC) were estimated for aptamer
The PiZZ group on augmentation therapy (N=46, sequence ID 3580-25 (AAT) using logistic regression in
GRADS) was compared with a 3:1 matched PiMM group PiZZ not on therapy, PiSZ, PiMZ, PiMS, and PiMM
(N=138, COPDGene). Matching was done using SAS individuals. We excluded PiZZ on therapy because they
surveyselct procedure; for never smoker status, sex, and had AAT levels approaching PiMM AAT serum levels.
age category. The Youden J-index was used to select the optimal pre-
dicted cut-point for AAT. Due to complete separation
Predictive modeling of PD15. We used supervised reg- between the tested genotypes the cut-point was deter-
ularization methods to select SomaScan proteins and to mined by the midpoint of the separation range. STARD
derive an emphysema predictive score using PD15. Two guidelines for diagnostic studies reporting were followed.
methods, Least Absolute Shrinkage and Selection Oper- Regression analyses, ROC, histograms, and match-
ator (LASSO) and elastic net were evaluated for the best ing were performed with SAS 9.4 (SAS/STAT 15.1).
fit on a training dataset. We found no significant differ- Meta-analysis and forest plots (metafor V3.0-2); LASSO
ence between models, therefore, we used LASSO (glmnet V4.1-2); beeswarm, scatter and volcano plots
because it provided a more parsimonious model, by (ggplot2 V3.3.5) were generated with R (V4.1.0). A false
retaining only one of the collinear variables. In the discovery rate adjusting for 4,979 SOMAmers (FDR 
COPDGene PiMM dataset 4745 observations had 0.05) was considered significant.22

6 www.thelancet.com Vol 84 October, 2022

 Articles

Role of funders mannose receptor-1 (MRC1), Gremlin-2 (GREM2), CUB

The study design, data collection, data analysis, inter- domain-containing protein-1 (CDCP1), and Fatty acid-
pretation, and writing of report are solely the responsi- binding protein, adipocyte (FABP4) (Table S2). Of 98
bility of the authors and do not necessarily represent the proteins significantly associated with DLCO in PiZZ
official views of the industry, the Foundations, National subjects, 68 were shared proteins associated with
Heart, Lung, and Blood Institute or the National Insti- DLCO in PiMM subjects as well (Figure 2b and S2), but
tutes of Health. importantly, 30 proteins were uniquely associated with
DLCO only in PiZZ off therapy subjects (Figure 2b).
The top 5 unique proteins were: keratin 1, myomesin-2,
Results insulin receptor, insulin-like growth factor, and hydrox-
ymethylbilane synthase; they are depicted in the table
There was significant demographic heterogeneity of
inserts (Figure 2b).
cohorts (Figure 1, Tables 1 and S1), with AATD cohorts
We then considered the 671 and 1305 nominally-sig-
being younger and including more never-smokers than
nificant proteins associated with DLCO in PiZZ off ther-
COPDGene (p<0.0001, ANOVA 1-way). In COPDGene
apy and PiMM individuals, respectively for pathway
we identified 11 and 37 African American subjects with
analysis. Using GOrilla we found three significantly
intermediate-deficient, Pi*MZ and Pi*MS, respectively;
enriched cellular components pathways in PiZZ off
other cohorts recruited predominantly non-Hispanic
therapy: insulin-like growth factor complex, lipid drop-
whites. The GRADS and Birmingham cohorts recruited
let, and myosin complex (Figure 3a, Table S3), as
more subjects with COPD GOLD stages 3 and 4
expected, considering the top nominal proteins we iden-
(Table 1). The PiZZ subjects on AAT therapy in GRADS
tified above are growth factors (e.g. IGF-1, GDF15), lipid
and QUANTUM-1 had more emphysema compared to
transporters or receptors (e.g. FABP4, MRC1), or pro-
those off therapy (p<0.001, ANOVA 1-way), while the
teins derived from skeletal muscle (e.g. myomesin-2,
PiZZ subjects in general had more emphysema and
hemojuvelin). These cellular component pathways did
lower DLCO than their PiMM counterparts (Table 1). A
not overlap with those enriched in PiMM, of which the
third of PiMZ subjects in COPDGene were active smok-
top 3 were: extracellular matrix, extracellular organelle /
ers, with a corresponding higher number of pack-years,
vesicles, and intracellular endoplasmic reticulum lumen
higher emphysema (p<0.006, Kruskal-Wallis), and
(Figure 3b, Table S4).
lower DLCO (p<0.001, Kruskal-Wallis) than the PiMZ
in GRADS, suggesting, as expected, that cigarette smok-
ing is an additive risk factor for disease severity (Table 1).
Biomarkers associated with airflow obstruction, DLCO,
PiMS subjects in COPDGene shared similar demo-
and emphysema in AATD subjects on augmentation
graphics, tobacco exposure, and functional characteris-
therapy
tics with PiMM individuals. Interestingly, PiSZ subjects
We analyzed the AATD subjects on augmentation ther-
in Birmingham cohort, despite significantly lower AAT
apy separately from those off therapy because those on
levels than PiMM and PiMZ subjects, presented with
therapy have already established severe emphysema,
nearly normal FEV1, DLCO, and GOLD severity at simi-
even though they now have normal levels of AAT
lar age as PiZZ subjects. Lastly, COPDGene subjects
achieved through weekly augmentation therapy infu-
had a higher Charlson comorbidity index.
sion. AATD subjects on therapy (N=46) were matched
on age, sex, smoking history, and GOLD severity with
Biomarkers associated with airflow obstruction and 138 PiMM subjects (COPDGene) 1:3 (Table S5). In the
DLCO in AATD patients off augmentation therapy PiZZ on therapy group we identified 373, 255, 111, 100,
We identified 177, 169, and 216 proteins significantly and 137 proteins nominally associated with PD15,
associated with FEV1, FEV1pp, and FEV1/FVC ratio, DLCO, FEV1pp, FEV1/FVC, and FEV1 respectively
respectively (nominal p  0.05, but none with FDR  (Table S6). Two proteins, endothelin-2 and sterol carrier
0.05, multivariable linear regression, Table S2). There protein-2 sterol-binding domain-containing protein-1
were 671 proteins associated with DLCO, 98 of which (SCP2D1) were significantly associated with FEV1 (p =
were either positively or negatively associated (FDR  0.0003 and p = 0.0004, FDR = 0.08, multivariable lin-
0.05, multivariable linear regression) with DLCO ear regression). In the matched PiMM COPDGene sub-
(Figure 2a). Of those positively associated with DLCO, group we identified 265, 517, 227, 341, and 220 proteins
the top SOMAmers were: Cerebellin-4 (CBLN4), Immu- significantly associated with PD15, DLCO, FEV1pp,
noglobulin superfamily DCC subclass member-4 FEV1/FVC, and FEV1 respectively, but only 21 proteins
(IGDC4), Hemojuvelin (HFE2), and Insulin-like growth associated with DLCO (FDR  0.05, multivariable lin-
factor-1 (IGF-1). Of those negatively associated with ear regression, Table S7). The top-50 ranked proteins
DLCO, the top SOMAmers were: Macrophage scaven- associated with DLCO for PiZZ subjects on therapy
ger receptor type-1 (MSR1), Transgelin (TAGL), were rather different than similar to the PiMM sub-
Growth/differentiation factor-15 (GDF15), Macrophage group, suggesting that PiZZ on therapy plasma

www.thelancet.com Vol 84 October, 2022 7

 Articles

Figure 2. SomaScan proteins association with DLCO in PiZZ subjects off augmentation therapy. a. Volcano plot of SomaScan
proteins positively and negatively associated with DLCO (N=98) for PiZZ subjects (N=185) off augmentation therapy from the Bir-
mingham, GRADS, QUANTUM cohorts. SomaScan proteins, labeled with their gene abbreviations, are shown in black if significantly
(FDR  0.05, multivariable linear regression) associated with DLCO; non-significant (FDR >0.05, multivariable linear regression)
SomaScan proteins are shown in grey. b. Euler plot of SomaScan proteins associated with DLCO that are specific (magenta circle,
N=30) and shared (purple circle, N=68) between PiZZ off therapy and PiMM (blue circle, N=1237) subjects. The top five unique pro-
teins keratin 1 (KRT1), myomesin-2 (MYOM2), insulin receptor (INSR), insulin-like growth factor (IGF-1), and hydroxymethylbilane syn-
thase (HMBS) and shared proteins macrophage scavenger receptor types I and II (MSR1), growth/differentiation factor 15 (GDF15),
fatty acid-binding protein, adipocyte (FABP4), tissue-type plasminogen activator (PLAT), intelectin-1 (ITLN1) between PiZZ off ther-
apy and PiMM subjects are shown in the insert tables.

8 www.thelancet.com Vol 84 October, 2022

 Articles

Figure 3. Cellular component pathways enriched in the SomaScan proteins associated with DLCO in PiZZ off augmentation
therapy and PiM subjects. a. Cellular component pathways enriched within the 671 nominally-significant SomaScan proteins asso-
ciated with DLCO in PiZZ subjects off therapy. b. Cellular component pathways enriched within the 1305 nominally-significant
SomaScan proteins associated with DLCO in PiMM subjects. Individual cellular component pathways are color-coded based on the
significance of enrichment p-values depicted at the bottom, light yellow p<0.00005, light orange p<0.000005, dark orange
p<0.00000005, red p<0.000000005 (Pathway enrichment analysis, GOrilla). The graphical representation of Directed acyclic graph
(DAG) was created using GOrilla.

www.thelancet.com Vol 84 October, 2022 9

 Articles

proteomic signature is different from the COPD signa- We used LASSO, the regression analysis that allows
ture, even after accounting for demographics and dis- feature selection and regularization to develop a protein
ease severity (Figure S2). The 2 proteins associated with score for emphysema (PD15). The protein score was
FEV1 in PiZZ on therapy subgroup and the 21 proteins trained and tested on COPDGene PiMM subjects
associated with DLCO in PiMM subgroup were not because of its larger sample size. The best risk score
among the top-50 proteins of the other’s subgroup, had 262 proteins combined in an emphysema (PD15)
highlighting the differences in plasma proteomics protein score showing strong correlations with mea-
between PiZZ and COPD individuals, even after aug- sured emphysema in PiMM subjects from the testing
mentation therapy was instituted (Tables S6 and S7). cohort (R2=0.44, rho=0.66, LASSO, Figure 5, Table 2).
The calculated emphysema protein score was also accu-
rate in PiZZ and PiMZ subjects from AATD cohorts
Biomarkers associated with airflow obstruction, DLCO, (rho from 0.47 to 0.70, LASSO, Figure 5b-d, Table 2).
and emphysema in PiMZ AATD subjects The lower rho values were associated with GRADS
Using individuals in two cohorts, COPDGene and PiMZ and PiZZ -subjects on AAT therapy, which had
GRADS, we identified 10 proteins that were signifi- the mildest and the most severe emphysema, respec-
cantly associated with FEV1; one, CRP, was significant tively (Figure 5d). The correlations between emphysema
(FDR  0.05, multivariable linear regression, protein score and measured emphysema were compara-
Figure 4a). There were 142, 164, and 145 proteins nomi- ble with DLCO correlations with measured emphysema
nally associated with FEV1pp, FEV1/FVC, and PD15 (p  (R2 from 0.20 to 0.76, Pearson correlation, Table 2),
0.05, but none at FDR  0.05, multivariable linear with the highest correlation seen in PiMZ COPDGene
regression, Table S8). There were 280 proteins associ- individuals with mild-moderate emphysema. Quantita-
ated with DLCO; 10 proteins were significant (FDR  tive CT measurements were not available in the Bir-
0.05, multivariable linear regression, Figure 4b). We mingham cohort; however, the emphysema protein
had one SomaScan protein positively associated with score was associated with measured visual emphysema
DLCO, Cyclic AMP-dependent transcription factor ATF- (AUC=0.74, logistic regression).
6 alpha (ATF6), while the other 9 proteins were nega- Although we used three AATD independent cohorts
tively associated: Retinoic acid receptor responder pro- to identify several unique protein - PD15 associations,
tein-2 (RARRES2), Chordin-like protein-1 (CHRDL-1), there were insufficient AATD subjects in any one cohort
R-spondin-1 (RSPO-1), Fibroblast growth factor-binding to develop a unique AATD protein risk score.
protein-1 (FGFBP1), Pleiotrophin (PTN), Spondin-1
(SPO-1), C-C motif chemokine-16 (CCL16), and Colla-
gen alpha-1(XXVIII) chain (COL28A1). Most of these SomaScan identifies AATD subjects
associations were also significant in PiMM subjects Severe-deficient PiZZ subjects are easily identified
(Figures S3 S4, Table S9). based on a low AAT serum level. However, many AATD
subjects intermediate-deficient, i.e. PiSZ, PiMS, and
PiMZ remain unidentified, despite being at increased
Emphysema protein score to boost explanation of risk for COPD,5,23,24 unless AAT genotyping is per-
measured emphysema variance and assess biomarker formed. We found that the SomaScan could readily
similarity between AATD and COPD identify PiZZ subjects not on AAT therapy (Figure 6a-
Using GRADS and QUANTUM-1 cohorts (PiZZ AATD b). The receiver operating characteristic (ROC) curve for
subjects) that had PD15 measured, we identified two diagnosis of PiZZ versus PiSZ, PiMM, PiMS, and
proteins, betacellulin (BTC) and Cyclin Dependent PiMZ, as well as PiSZ versus PIMZ, and PIMM subjects
Kinase-2-associated protein-1 (CDK2AP1) positively were perfect (AUC 1.00, logistic regression, Figure 6c).
associated with PD15 at (FDR  0.05, multivariable lin- PiSZ, PiMS, and PiMZ subjects were also well distin-
ear regression, Figure 5a-b). guished (AUCs 0.8-0.9, logistic regression, Figure 6b-
There were more FDR and nominal significant pro- c). Thus, a cut-off value <7.99 (RFU, natural log) had a
teins associated with markers of emphysema (DLCO, 100% sensitivity/specificity for the diagnosis of PiZZ
PD15) compared to markers of airflow obstruction and values above 10.52 were likely PiMM subjects.
(FEV1, FEV1/FVC) in the AATD cohorts (Figure 2a,
Figure 5a-b, Table S2). Although, in PiMM subjects
there were many more proteins (N=665) nominally sig- Discussion
nificant associated with PD15, in both PiZZ and PiMM Since AATD is only a risk factor for emphysema,6 not
the individual proteins had small effect size. Therefore, all subjects have clinically significant disease. Although
we investigated whether a protein score performs better environmental exposures are important, proteomic dis-
than individual SomaScan proteins at explaining the ease-modifiers may be able to explain part of the emphy-
clinical variance of measured PD15, used as an emphy- sema heterogeneity in AATD and COPD.9,11 Our study
sema marker. has comprehensively tested the AATD proteome

10 www.thelancet.com Vol 84 October, 2022

 Articles

Figure 4. Somascan proteins association with DLCO and spirometry in PiMZ subjects. a. Volcano plot of SomaScan proteins
positively and negatively associated with DLCO (N=10) in PiMZ subjects (N=215) from two cohorts, COPDGene and GRADS studies.
SomaScan proteins, labeled with their gene abbreviations, are shown in black if significantly (FDR  0.05, multivariable linear regres-
sion) associated with DLCO; non-significant (FDR > 0.05, multivariable linear regression) SomaScan proteins are shown in grey. b.
Euler plot of SomaScan proteins associated with DLCO that are shared (green circle, N=10) between PiMZ and PiMM (blue circle,
N=1295) subjects. The top five shared proteins between PiMZ and PiMM subjects are Retinoic acid receptor responder protein 2
(RARRES2), Chordin-like protein 1 (CHRDL1), R-spondin-1 (RSPO1), Fibroblast growth factor-binding protein 1 (FGFP1), and pleiotro-
phin (PTN). c. Forest plot of natural log CRP (RFU, natural log) negative association with FEV1 in PiMZ subjects (N=215) from two
cohorts, COPDGene and GRADS studies.

www.thelancet.com Vol 84 October, 2022 11

 Articles

Figure 5. Somascan individual proteins and emphysema protein score association with adjusted lung density. a-b. Forest
plots of cyclin-dependent kinase 2-associated protein 1 (CDK2AP1) and betacellulin (RFU, natural log) positive association with
adjusted lung density (PD15, g/L) in PiZZ individuals (N=63) off augmentation therapy from two, GRADS and QUANTUM study
cohorts. c-f. Scatter plots showing the association between measured adjusted lung density (PD15, y axis) vs. PD15 predicted by the
emphysema protein score (x axis) in c) COPDGene PiMM (R2=0.44, rho=0.66, LASSO); d) COPDGene PiMZ (R2=0.49, rho=0.70, LASSO);
e) GRADS PiZZ off therapy (R2=0.37, rho=0.61, LASSO); and f) PiZZ on therapy individuals (R2=0.23, rho=0.48, LASSO). Emphysema
protein score (N=262 proteins) was developed using LASSO in the COPDGene PiMM subjects and the other genotypes were used
for validation.

12 www.thelancet.com Vol 84 October, 2022

 Articles

Table 2. Measured adjusted Lung Density (PD15) vs. predicted PD15 by LASSO or vs. measured DLCO
Protein Score (by LASSO) DLCO
2 2
Model Training Testing MSE R Rhop Visual R RhoP Visual
Sample Sample Testing Emphysema Emphysema
Size Size Yes/No AUC Yes/No AUC

COPDGene PiMM 3324 1421 360.9 0.44 0.66 NA 0.15 0.39 NA

Validation
COPDGene PiMZ 146 409.2 0.49 0.70 NA 0.36 0.60 NA
GRADS PiMZ 54 456.9 0.22 0.47 NA 0.06 0.24 NA
GRADS PiZZ off therapy 30 452.1 0.37 0.61 NA 0.33 0.56 NA
GRADS PiZZ on therapy 44 1234.2 0.23 0.48 NA 0.57 0.76 NA
QUANTUM-1 PiZZ off therapy 30 809.9 0.24 0.49 NA 0.04 0.20 NA
Birmingham 100 NA NA NA 0.74 NA NA 0.78

Table 2: Measured adjusted lung density (PD15) vs. predicted PD15 by LASSO emphysema protein score or vs. measured DLCO in the
testing cohort, COPDGene (PiMM) and the validation cohorts: COPDGene (PiMZ), GRADS (PiMZ), GRADS (PiZZ off therapy), GRADS (PiZZ
on therapy), QUANTUM (PiZZ off therapy). In the Birmingham cohort we report measured visual emphysema vs. predicted emphysema
by LASSO protein score. R2 and Pearson correlation (Rho) were calculated in R. AUC was calculated using logistic regression in R.
Abbreviations: MSE: mean square error, AUC: area under the curve.

association with markers of emphysema in PiZZ and off therapy, and not in PiZZ on therapy or in PiMM
PiMZ subjects enrolled in multiple independent with advanced emphysema.
cohorts, including a large PiMM, COPD reference pop- Betacellulin (BTC), a second biomarker specific to
ulation. For both PiMM and PiZZ subjects there were PiZZ subjects off therapy, is a ligand of the epidermal
many different proteins associated with functional and growth factor (EGF) superfamily. Similar to transform-
radiologic markers of emphysema. ing growth factor-a, heparin-binding EGF-like growth
Like other recent publications in non-pulmonary factor and amphiregulin, EGF mediates BTC down-
cohorts, we found that multiple proteins in combination stream signaling and results in airway epithelial reprog-
explained a much higher percentage of the variance of a ramming or epithelial to mesenchymal transition
clinical phenotype compared to individual proteins.9,25 (EMT), mucus hypersecretion and airway obliteration,
Interestingly, patients at risk for cardio-vascular events, and possible malignant transformation of large and
early death in congestive heart failure, and kidney fail- small airways.33,34 BTC ranked 14th in a support vector
ure in diabetic subjects were easily identified by SomaS- machine classifier used to diagnose and endotype
can protein scores.26 28 In fact, a combination of 262 COPD individuals35 and it was higher in ex-smokers
proteins in our emphysema protein score explained with COPD than without COPD.36 Our study reports
more of measured PD15 variance than DLCO in COPD, that BTC is significantly associated with emphysema in
PiMZ, and PiZZ subjects off augmentation therapy, AATD. The positive association between BTC and PD15
outperforming DLCO as a known functional marker of suggests that, in AATD patients off augmentation ther-
emphysema. The emphysema protein score explained apy, emphysema is characterized by an active repair pro-
less of measured PD15 variance in PiZZ on augmenta- cess involving epithelial airway remodeling. More
tion therapy, possibly because the protein score was targeted work needs to be done in AATD pre-clinical
developed in the COPD patients, where the SomaScan models to determine BTC’s role in EMT, mucus hyper-
proteins associated with PD15 were rather different than secretion, and remodeling in AATD emphysema.
similar to those in PiZZ on therapy. Our study confirms previous nominal protein -
We identify CDK2AP1 as unique biomarker associ- emphysema associations (e.g. CRP, FABP4) reported in
ated with PD15 in PiZZ off augmentation therapy sub- 31 AATD patients enrolled in QUANTUM-1 cohort and
jects. CDK2AP1 is the only known inhibitor of cyclin- measured on Myriad discovery panel.11 Other associa-
dependent kinase-2 and a master regulator of cell divi- tions were not confirmed, like leptin, gesolin, and met-
sion cycle. CDK2AP1 is down-regulated in various alloproteinase-3, proteins that were associated with
malignancies29 31 and upregulated during embryonic emphysema at baseline and with its progression when
development.32 Its positive association with PD15 sug- measured on the Myriad panel.11 We do describe the
gests tighter control of cell cycle check-points in PiZZ fatty acid binding protein (FABP4) association with
subjects off augmentation therapy and with early emphysema seen also in QUANTUM-1 cohort,11 but our
emphysema. It remains to be determined in prospective meta-analysis showed stronger association between
AATD cohorts whether CDK2AP1 is a signal of efficient FABP4 and lung function (FEV1 and DLCO). FABP-4
repair, because we only detect CDK2AP1 association association with DLCO was found in all individual PiZZ
with PD15 in PiZZ subjects with early emphysema and cohorts and in the meta-analysis. Our findings suggest

www.thelancet.com Vol 84 October, 2022 13

 Articles

Figure 6. SomaScan Alpha-1 Antitrypsin relative levels in various Alpha-1 genotypes. a. Beeswarm plot of AAT relative fluores-
cence units (RFU, natural log, y axis) by genotype (x axis). AAT relative levels in pooled PiMM (10.63 §0.23, N=5101), PiMS (10.37 §
0.21, N=347), PiMZ (9.98 § 0.25, N=215), PiSZ (9.10 § 0.16, N=24), PiZZ off therapy (6.32 § 0.36, N=185), and PiZZ on AAT augmenta-
tion therapy (10.69 § 0.54, N=52) subjects from all 4 study cohorts. b. Histogram showing the percent distribution of PiZZ (N = 185),
PiSZ (N=24), PiMZ (N=215), PiMS (N=347), and PiMM (N=5101) subjects from all 4 study cohorts. c. Receiving operator curves (ROC)
of AAT (RFU, natural log) for the following genotype comparisons are shown: PiSZ vs. PiMZ, PiMZ vs. PiMS, PiMZ vs. PiMM, and PiMS
vs. PiMM. The ROCs characteristics for the genotype comparisons not graphically depicted are shown in the table insert. *If AUC is
equal to 1.0 it is a midpoint of the range of complete separation. If AUC less than 1.0 the value is the cut-point determined by the
Youden Index.

that FABP-4 is a promising biomarker of AATD severity. and members of the spondin family identified in our
Our study was not powered to investigate FABP-4 associ- PiMZ cohort are previously reported biomarkers of airflow
ation with PD15 or DLCO in response to augmentation obstruction and emphysema in PiMM subjects.8 This sug-
therapy, as we did not see an association with PD15 and gests that plasma proteome of PiMZ is similar to PiMM
DLCO in the PiZZ on augmentation therapy group. subjects and might argue that PiMZ are similar enough to
Although we did find unique protein associations in PiMM subjects to be included in general COPD clinical
PiZZ subjects, many of the biomarkers for PiZZ off AAT studies; however, often they are excluded.23,24
therapy and PiMZ were similar to PiMM subjects. For There were differences in our cohorts which may
instance, CRP, retinoic acid receptor responder protein-2, explain our unique biomarker - emphysema associations.

14 www.thelancet.com Vol 84 October, 2022

 Articles

Most of the biomarkers we observed in AATD cohorts (Pi*MS) appears lower because they present with AAT
were for emphysema measurements (DLCO or PD15) levels similar to Pi*MM genotype. The SomaScan assay
whereas in COPDGene there were more biomarkers asso- can also inform on the ability of augmentation therapy to
ciated with airflow limitation (FEV1 and FEV1/FVC). This raise AAT to near-physiologic levels as evidenced by AAT
is likely because the three AATD cohorts included in this levels within PiMM range in PiZZ subjects on therapy.
study had predominantly subjects with normal-to-low One disadvantage of SomaScan, i.e., that it only reports
FEV1, but evidence of emphysema on CT scan. relative fluorescent, not absolute AAT “units”, doesn’t
The strength of our data relies on representative appear to be a limitation for identifying clinically-signifi-
number of never smokers PiZZ individuals form North cant genotypes. We can’t exclude an inclusion bias that
America and Europe with various degrees of airflow resulted in high sensitivity and specificity in identifying
limitation but significant emphysema who demon- clinically-significant genotypes, as the subjects included
strated unique and similar protein biomarkers to a large in PiZZ cohorts were not selected from the general popu-
cigarette smoke-related COPD cohort. Overall the lation, nevertheless COPDGene subjects were. However,
emphysema phenotype was better predicted by a com- the characteristics of the assay suggest utility in identify-
mon protein score. This score outperformed the more ing undiagnosised AATD subjects.
commonly used DLCO biomarker in all individual Limitations to the SomaScan proteomics include the
PiZZ and PiMZ cohorts. These findings strongly sug- lack of SOMAmers for small molecules such as desmo-
gest that our unique plasma biomarkers and emphy- sine,40 fibrinogen degradation product (Aa-Val360, a spe-
sema protein score could be useful in both AATD and cific product generated by elastase cleavage of
COPD research. Indeed, the identification of emphy- fibrinogen),41 and sphingomyelin,42 which have been sug-
sema in patients without airflow obstruction may be the gested to be emphysema biomarkers in other studies,43
most important aspect of disease prevention, because While our study is the largest biomarker study in AATD
AATD subjects with early emphysema may be the most subjects, none of the cohorts had enough subjects with
likely to progress and have the most at-risk healthy very rare genotypes (Pi*IZ or Pi*SPLowell) to achieve ade-
lung. The generalizability of our newly described quate power. Finally, only the GRADS cohort had match-
plasma biomarkers and emphysema protein score can ing plasma bronchoalveolar lavage samples, therefore
be further validated on more longitudinal cohorts to our study has not investigated lung-specific protein bio-
assess their ability to prospectively predict emphysema markers. There are also geographic differences such as
diagnosis and progression. Additionally, the protein the large number of subjects in US-based GRADS and
score could address the lack a non-invasive, easy to col- QUANTUM-1 on augmentation therapy versus no
lect instrument that is both diagnostic and prognostic of patients on augmentation in the Birmingham cohort.
early emphysema. The instrument would be most use- GRADS subjects on therapy tended to have much worse
ful in higher risk subjects (smokers or never-smokers emphysema, but we were able to identify two candidate
with abnormal pulmonary function test or hypoxemia) proteins in endothelin-2 and SCP2D1 associated with
or patients with abnormal Alpha-1 genotypes, but nor- FEV1, suggesting that these plasma proteins might be
mal spirometry. Additionally the instrument, might used as biomarkers of disease progression or response to
play a useful role in therapeutic clinical trials to study therapy rather than diagnostic biomarkers.
emphysema progression in response to therapy. In summary, we demonstrate that the SomaScan
This study also demonstrates a potential role for proteomic platform helps risk stratify AATD carriers
identifying clinically missed AATD subjects because and subjects with early disease who might be at higher
many AATD subjects are under-diagnosed and may risk for emphysema. Also, the SomaScan emphysema
have unappreciated emphysema. Current AATD testing protein score enhances our ability to predict emphy-
is a two-step procedure, initially AAT serum levels are sema. Furthermore, SomaScan has excellent diagnostic
measured by radial immunodiffusion or nephelometry characteristics for severe- and intermediate-deficient
with the low levels confirmed by a second test, like geno- AAT genotypes, which are the main clinically actionable
typing or protein electrophoresis.37 Unfortunately, this AATD phenotypes. Further work is needed to deter-
strategy misses subjects in the general population with mine whether these same biomarkers are useful to
low clinical suspicion for AATD and subjects with at- assess progression of emphysema and airflow obstruc-
risk, intermediate-deficient genotypes, Pi*MZ and tion as well as other COPD comorbidities such as exac-
Pi*SZ, who may develop emphysema.38,39 We report erbations in AATD subjects.
that SomaScan identifies the most common clinically-
significant AATD genotypes, the severe and intermedi-
ate-deficient Pi*ZZ, Pi*SZ, and Pi*MZ genotypes with Contributors
excellent sensitivity and specificity. This is relevant KAS - data curation, investigation, methodology, data
because SomaScan is frequently used in population- interpretation, visualization, writing - original draft.
based studies and could be useful for diagnosing KAP - data curation, investigation, formal analysis, valida-
AATD. The ability to detect less severe genotypes tion, methodology, visualization, writing -original draft.

www.thelancet.com Vol 84 October, 2022 15

 Articles

CS - conceptualization, funding acquisition, investigation, from Vertex, NIH/NCATS, Alpha-1 Foundation, con-
methodology, data interpretation, project administration, sulting fees from Grifols, CSL Behring, Vertex, Intellia,
supervision, visualization, writing - original draft. Inhibrx, Takeda, Evolve Biologics, has received travel
RAS- conceptualization, funding acquisition, project support from CSL Behring; has served on the Advisory
administration. Board of Arrowhead Pharmaceuticals, and serves as the
AMT- conceptualization, funding acquisition, investiga- Medical Director for AlphaNet and on the board of
tion, methodology, data interpretation, project administra- directors for Global Implementation Solutions, Osteo-
tion, supervision, visualization, writing - original draft. genesis Imperfecta Foundation, Alpha-1 Foundation,
TB - investigation, visualization. and AlphaNet; AMT reports grants or contracts from
DAS - data curation, investigation, visualization. Vertex, Grifols, and CSl Behring; has received consult-
LM - conceptualization, funding acquisition, data inter- ing fees from CSL Behring, Inhibrix, Z-factor, and
pretation, project administration, visualization, writing - Takeda, has received honoraria for lectures from Glaxo
original draft. Smith Kline and AstraZeneca; TB does not report any
NH - conceptualization, funding acquisition, project potential conflict of interest; DAS does not report any
administration, visualization. potential conflict of interest; LM does not report any
EKS- conceptualization, funding acquisition, investigation, potential conflict of interest; NH does not report any
methodology, data interpretation, project administration, potential conflict of interest; EKS reports grants or con-
supervision, visualization, writing - original draft. tracts from Glaxo Smith Kline, Bayer; BDH does not
BDH - investigation, methodology, data interpretation, report any potential conflict of interest; CPH reports
visualization, writing - original draft. grants or contracts from Alpha-1 Foundation, Bayer,
CPH - data interpretation, visualization, writing - origi- Boehringer-Ingelheim, and Vertex, consulting fees
nal draft. from AstraZeneca and Takeda; DLD has received hono-
DLD - data interpretation, visualization, writing - origi- raria for lectures from Novartis and financial support
nal draft. from Bayer towards the institution; MHC reports grants
MHC - investigation, methodology, data interpretation, or contracts from Glaxo Smith Kline, Bayer, consulting
visualization, writing - original draft. fees from AstraZeneca and Genentech, honoraria for
RPB- conceptualization, funding acquisition, investigation, presentations from Illumina, RPB does not report any
methodology, data interpretation, project administration, potential conflict of interest.
supervision, visualization, writing - original draft.
KAS, KAP, CS, AMT, DAS, EKS, CPH, DLD, and RPB
Acknowledgements
have verified the underlying data.
SomaLogic, Inc. as the provider of the proteomic data
All authors have reviewed and approved the final ver-
measured using the modified aptamer-based SomaScanÒ
sion of the manuscript.
Assay. SomaScanÒ and SOMAmerÒ are registered trade-
marks of SomaLogic, Inc. and are used under license.
Data sharing statement
Funding: This work was supported by a grant from the
Data underlying the results from COPDGene and
Alpha-1 Foundation to RPB. COPDGene was supported
Quantum cohorts are deposited and will be available
by Award U01 HL089897 and U01 HL089856 from the
per request in dbGaP database (phs000179.v6.p2 for
National Heart, Lung, and Blood Institute. Proteomics
COPDGene, phs000698.v1.p1 for QUANTUM). Data
for COPDGene was supported by NIH 1R01HL137995.
underlying the results from the Birmingham Alpha-1
GRADS was supported by Award U01HL112707, U01
Antitrypsin registry will be available per request, contact
HL112695 from the National Heart, Lung, and Blood
Dr. Alice M Turner.
Institute, and UL1TRR002535 to CCTSI; QUANTUM-1
was supported by the National Heart Lung and Blood
Declaration of interests
Institute, the Office of Rare Diseases through the Rare
KAS serves on Alpha-1 Foundation Grant Advisory
Lung Disease Clinical Research Network (1 U54
Committee, Alpha-1 Foundation Medical Advisory and
RR019498-01, Trapnell PI), and the Alpha-1 Foundation.
Scientific Committee, ATS - RCMB Website Commit-
COPDGene is also supported by the COPD Foundation
tee, National Jewish Health IBC committee (all unpaid);
through contributions made to an Industry Advisory
KAP does not report any potential conflict of interest;
Board that has included AstraZeneca, Bayer Pharmaceut-
CS reports grants or contracts from Adverum, Arrow-
icals, Boehringer-Ingelheim, Genentech, GlaxoSmithK-
head, AstraZeneca, CSA Medical, Grifols, Nuvaira,
line, Novartis, Pfizer, and Sunovion.
Takeda, Vertex; consulting fees from AstraZenca,
Dicerna, Glaxo Smith Kline, Inhibrx, Morair, UpTo-
Date, Vertex; has received honoraria for presentations Supplementary materials
from the American Thoracic Society; has received sup- Supplementary material associated with this article can
port for travel from CSL Behring; serves as the Medical be found in the online version at doi:10.1016/j.
Director for AlphaNet; RAS reports grants or contracts ebiom.2022.104262.

16 www.thelancet.com Vol 84 October, 2022

 Articles

References 23 Lipson DA, Criner GJ, Lomas DA. Single-Inhaler triple versus dual
1 Silverman EK, Sandhaus RA. Clinical practice. Alpha1-Antitrypsin therapy in patients with COPD. N Engl J Med. 2018;379(6):592–593.
deficiency. N Engl J Med. 2009;360(26):2749–2757. 24 Strange C, Herth FJ, Kovitz KL, et al. Design of the endobronchial
2 Brode SK, Ling SC, Chapman KR. Alpha-1 Antitrypsin deficiency: a Valve for Emphysema Palliation Trial (VENT): a non-surgical
commonly overlooked cause of lung disease. CMAJ. 2012;184 method of lung volume reduction. BMC Pulm Med. 2007;7:10.
(12):1365–1371. 25 Mastej E, Gillenwater L, Zhuang Y, Pratte KA, Bowler RP, Kechris
3 Blanco I, de Serres FJ, Fernandez-Bustillo E, Lara B, Miravitlles M. K. Identifying protein-metabolite networks associated with COPD
Estimated numbers and prevalence of Pi*S and Pi*Z alleles of phenotypes. Metabolites. 2020;10(4):124. https://doi.org/10.3390/
alpha1-antitrypsin deficiency in European countries. Eur Respir J. metabo10040124.
2006;27(1):77–84. 26 Niewczas MA, Pavkov ME, Skupien J, et al. A signature of circulat-
4 Franciosi AN, Hobbs BD, McElvaney OJ, et al. Clarifying the risk of ing inflammatory proteins and development of end-stage renal dis-
lung disease in SZ Alpha-1 Antitrypsin deficiency. Am J Respir Crit ease in diabetes. Nat Med. 2019;25(5):805–813.
Care Med. 2020;202(1):73–82. 27 Cuvelliez M, Vandewalle V, Brunin M, et al. Circulating proteomic
5 Dahl M, Tybjaerg-Hansen A, Lange P, Vestbo J, Nordestgaard BG. signature of early death in heart failure patients with reduced ejec-
Change in lung function and morbidity from chronic obstructive tion fraction. Sci Rep. 2019;9(1):19202.
pulmonary disease in alpha1-antitrypsin MZ heterozygotes: a longi- 28 Tanaka T, Biancotto A, Moaddel R, et al. Plasma proteomic signa-
tudinal study of the general population. Ann Intern Med. 2002;136 ture of age in healthy humans. Aging Cell. 2018;17(5):e12799.
(4):270–279. 29 Choi MG, Sohn TS, Park SB, et al. Decreased expression of p12 is
6 Silverman EK, Province MA, Campbell EJ, Pierce JA, Rao DC. associated with more advanced tumor invasion in human gastric
Family study of alpha 1-antitrypsin deficiency: effects of cigarette cancer tissues. Eur Surg Res. 2009;42(4):223–229.
smoking, measured genotype, and their interaction on pulmonary 30 Figueiredo ML, Kim Y, St John MA, Wong DT. p12CDK2-AP1 gene
function and biochemical traits. Genet Epidemiol. 1992;9:317–331. therapy strategy inhibits tumor growth in an in vivo mouse model
7 Ghosh AJ, Hobbs BD, Moll M, et al. Alpha-1 Antitrypsin MZ het- of head and neck cancer. Clin Cancer Res. 2005;11(10):3939–3948.
erozygosity is an endotype of chronic obstructive pulmonary dis- 31 Zolochevska O, Figueiredo ML. Cell-cycle regulators cdk2ap1 and
ease. Am J Respir Crit Care Med. 2022;205(3):313–323. bicalutamide suppress malignant biological interactions between
8 Serban KA, Pratte KA, Bowler RP. Protein biomarkers for COPD prostate cancer and bone cells. Prostate. 2011;71(4):353–367.
outcomes. Chest. 2021;159(6):2244–2253. 32 Kim Y, McBride J, Kimlin L, Pae EK, Deshpande A, Wong DT. Tar-
9 Zemans RL, Jacobson S, Keene J, et al. Multiple biomarkers predict geted inactivation of p12, CDK2 associating protein 1, leads to early
disease severity, progression and mortality in COPD. Respir Res. embryonic lethality. PLoS One. 2009;4(2):e4518.
2017;18(1):117. 33 Leikauf GD, Borchers MT, Prows DR, Simpson LG. Mucin apopro-
10 Pratte KA, Curtis JL, Kechris K, et al. Soluble receptor for advanced tein expression in COPD. Chest. 2002;121(5 suppl):166S–182S.
glycation end products (sRAGE) as a biomarker of COPD. Respir 34 Vallath S, Hynds RE, Succony L, Janes SM, Giangreco A. Targeting
Res. 2021;22(1):127. EGFR signalling in chronic lung disease: therapeutic challenges
11 Beiko T, Janech MG, Alekseyenko AV, et al. Serum proteins associ- and opportunities. Eur Respir J. 2014;44(2):513–522.
ated with emphysema progression in severe Alpha-1 Antitrypsin 35 Yao Y, Gu Y, Yang M, Cao D, Wu F. The gene expression bio-
deficiency. Chronic Obstr Pulm Dis. 2017;4(3):204–216. markers for chronic obstructive pulmonary disease and interstitial
12 Campos MA, Geraghty P, Holt G, et al. The biological effects of lung disease. Front Genet. 2019;10:1154.
double-dose Alpha-1 Antitrypsin augmentation therapy. A pilot 36 de Boer WI, Hau CM, van Schadewijk A, Stolk J, van Krieken JH,
clinical trial. Am J Respir Crit Care Med. 2019;200(3):318–326. Hiemstra PS. Expression of epidermal growth factors and their
13 Kim CH, Tworoger SS, Stampfer MJ, et al. Stability and reproduc- receptors in the bronchial epithelium of subjects with chronic
ibility of proteomic profiles measured with an aptamer-based plat- obstructive pulmonary disease. Am J Clin Pathol. 2006;125(2):184–
form. Sci Rep. 2018;8(1):8382. 192.
14 Candia J, Cheung F, Kotliarov Y, et al. Assessment of variability in 37 Sandhaus RA, Turino G, Brantly ML, et al. The diagnosis and man-
the SOMAscan assay. Sci Rep. 2017;7(1):14248. agement of Alpha-1 Antitrypsin deficiency in the adult. Chronic
15 Raffield LM, Dang H, Pratte KA, et al. Comparison of proteomic Obstr Pulm Dis. 2016;3(3):668–682.
assessment methods in multiple cohort studies. Proteomics. 38 McElvaney GN, Sandhaus RA, Miravitlles M, et al. Clinical considera-
2020;20(12):e1900278. tions in individuals with alpha1-antitrypsin Pi*SZ genotype. Eur Respir
16 Regan EA, Hokanson JE, Murphy JR, et al. Genetic epidemiology J. 2020;55(6):1902410. https://doi.org/10.1183/13993003.02410-
of COPD (COPDGene) study design. COPD. 2010;7(1):32–43. 2019.
17 Strange C, Senior RM, Sciurba F, et al. Rationale and design of the 39 Molloy K, Hersh CP, Morris VB, et al. Clarification of the risk of
genomic research in Alpha-1 Antitrypsin deficiency and Sarcoidosis chronic obstructive pulmonary disease in alpha1-antitrypsin defi-
study. Alpha-1 protocol. Ann Am Thorac Soc. 2015;12(10):1551–1560. ciency PiMZ heterozygotes. Am J Respir Crit Care Med. 2014;189
18 Wan ES, Castaldi PJ, Cho MH, et al. Epidemiology, genetics, and (4):419–427.
subtyping of preserved ratio impaired spirometry (PRISm) in 40 Cantor J, Ochoa A, Ma S, Liu X, Turino G. Free desmosine is a sen-
COPDGene. Respir Res. 2014;15:89. sitive marker of smoke-induced emphysema. Lung. 2018;196
19 Parr DG, Dirksen A, Piitulainen E, Deng C, Wencker M, Stockley (6):659–663.
RA. Exploring the optimum approach to the use of CT densitome- 41 Manon-Jensen T, Langholm LL, Ronnow SR, et al. End-product of
try in a randomised placebo-controlled study of augmentation ther- fibrinogen is elevated in emphysematous chronic obstructive pul-
apy in alpha 1-antitrypsin deficiency. Respir Res. 2009;10:75. monary disease and is predictive of mortality in the ECLIPSE
20 Foreman MG, Wilson C, DeMeo DL, et al. Alpha-1 Antitrypsin cohort. Respir Med. 2019;160:105814.
PiMZ genotype is associated with chronic obstructive pulmonary 42 Bowler RP, Jacobson S, Cruickshank C, et al. Plasma sphingolipids
disease in two racial groups. Ann Am Thorac Soc. 2017;14(8):1280– associated with chronic obstructive pulmonary disease phenotypes.
1287. Am J Respir Crit Care Med. 2015;191(3):275–284.
21 Eden E, Navon R, Steinfeld I, Lipson D, Yakhini Z. GOrilla: a tool 43 Rabinovich RA, Miller BE, Wrobel K, et al. Circulating desmosine
for discovery and visualization of enriched GO terms in ranked levels do not predict emphysema progression but are associated
gene lists. BMC Bioinf. 2009;10:48. with cardiovascular risk and mortality in COPD. Eur Respir J.
22 Benjamini Y, Drai D, Elmer G, Kafkafi N, Golani I. Controlling the 2016;47(5):1365–1373.
false discovery rate in behavior genetics research. Behav Brain Res.
2001;125(1-2):279–284.

www.thelancet.com Vol 84 October, 2022 17