Abstract
Objective:
Anti-RNA-polymerase-III (ARA) and anti-topoisomerase-I (ATA) autoantibodies are associated with diffuse cutaneous SSc (dcSSc). ARA is associated with rapidly progressive skin thickening, scleroderma renal crisis (SRC), and gastric antral vascular ectasia, while ATA is associated with severe interstitial lung disease (ILD). However, these associations were derived from heterogenous SSc cohorts. As SSc is now often diagnosed earlier, we aimed to ascertain whether these relationships hold true in early dcSSc and to examine autoantibody associations with organ involvement in early dcSSc.
Methods:
The Prospective Registry of Early Systemic Sclerosis (PRESS) includes adults with dcSSc ≤2 years from the first non-Raynaud’s phenomenon symptom who met 2013 ACR/EULAR classification criteria for SSc. Participants were enrolled at 12 U.S. academic centers and evaluated every 6 months. Those with dual-positivity were excluded. Differences across autoantibody groups were assessed using appropriate parametric and nonparametric tests.
Results:
Of 303 enrolled, 233 participants were included (108 ARA+, 68 ATA+, 57 double-negative). Mean age was 51 years; 68% were female. Mean disease duration was 1.2 years. ATA+ participants had lower baseline FVC% predicted (76.0% vs 86.0% ARA+, p=0.0011) and higher prevalence of ILD (69.6% vs 44.3% ARA+, p=0.0029). ARA+ participants had higher baseline mRSS (25.0 vs 17.9 ATA+, p<0.0001) and higher prevalence of SRC (14.4% vs 0.0% ATA+, p=0.0024). Only the double-negative group had improvement in FVC% predicted over time (+1.39% per year, p=0.03).
Conclusions:
Autoantibody-specific differences in organ involvement were evident early. The double-negative group uniquely showed improvement in lung function over time.
Keywords: systemic sclerosis, diffuse cutaneous systemic sclerosis, early disease, autoantibody subsets, anti–RNA polymerase III antibodies, anti–topoisomerase I antibodies
1. Introduction
Systemic sclerosis (SSc) is a systemic autoimmune rheumatic disease associated with significant morbidity and mortality, with up to four times higher mortality in SSc patients compared to controls [1, 2]. Interstitial lung disease (ILD) and pulmonary arterial hypertension (PAH) are the leading and second leading causes of death, respectively, in these patients [3, 4]. With improved evaluation and testing, it is now recognized that organ involvement begins early in the course of disease; thus, identifying biomarkers that correlate with risk of internal organ manifestations is critically important for improving early intervention [5, 6].
Several autoantibodies have been linked to various disease manifestations of SSc. For example, anti-RNA-polymerase-III autoantibodies (ARA) are associated with diffuse cutaneous (dc)SSc, an increased risk of scleroderma renal crisis (SRC), rapidly progressive skin thickening, gastric antral vascular ectasia (GAVE), and malignancy [7–9]. Anti-topoisomerase-I autoantibodies (ATA) are associated with dcSSc as well as with a higher risk of severe and progressive ILD [10, 11]. In the Pittsburgh cohort, 23% of 318 ATA positive (+) patients had severe pulmonary fibrosis with a mean forced vital capacity (FVC) of 68% of predicted [12]. ILD may develop at any time in SSc patients, however the risk is greatest early in the course of the disease[13]. Anti-centromere autoantibodies are associated with limited cutaneous SSc and an increased risk of PAH [12, 14, 15].
The autoantibody associations described in the literature have largely resulted from heterogenous studies, which included patients with both limited and diffuse cutaneous subtypes of varying disease durations. It is unknown whether the same associations hold true for patients with early dcSSc. The aim of this study was to examine the clinical SSc manifestations associated with ATA and ARA positivity among adults with early dcSSc. We hypothesized that among participants with early dcSSc, those who are ATA+ would have a higher prevalence of ILD and a greater decline in FVC over time, and those who are ARA+ would have higher baseline and follow-up skin scores, and a higher prevalence of SRC and GAVE.
2. Methods
2.1. Participants and methods
The Prospective Registry of Early Systemic Sclerosis (PRESS) is a multi-center cohort of adults with early dcSSc with disease duration ≤ 2 years from onset of the first non-Raynaud’s symptom. All participants met 2013 ACR/EULAR classification criteria for SSc [16]. Participants were recruited from 12 academic medical centers in the United States from April 2012 to July 2020. The Institutional Review Boards at each of the 12 participating institutions approved this study. All participants provided written informed consent. Autoantibody testing was performed locally as part of routine clinical care through commercial laboratories. Participants who were positive for both ATA and ARA autoantibodies were excluded, as were participants who were missing either of these autoantibodies.
Data were captured at PRESS enrollment and every 6 months (±3 months) thereafter during routine follow-up visits with participants’ treating rheumatologists. Data captured in the PRESS cohort and analyzed as part of this study include demographics, smoking status, immunosuppressive medication use, pulmonary function tests (PFTs), high resolution computed tomography scan of the chest (HRCT), modified Rodnan skin score (mRSS), and gastrointestinal, renal, vascular, and cardiopulmonary assessments.
2.2. Statistical analysis
Baseline demographic characteristics, clinical manifestations, PFT measurements, and medication use were summarized for the overall cohort and stratified by autoantibody groups: ARA+, ATA+, and double-negative (negative for both ATA and ARA). We reported mean with standard deviation (SD) or median (interquartile range [IQR]), as appropriate. We summarized categorical variables as counts (percentages). PFT measures, change in PFT values, mRSS, and change in mRSS were summarized at 6-, 12-, and 24-months. ILD status, renal crisis and GAVE were assessed at any time during follow-up.
In our primary approach, we compared differences between ARA+ and ATA+ by performing Student’s T-tests or Wilcoxon rank sum tests for continuous variables, and Chi-square tests or Fisher exact tests for categorical variables, as appropriate. In our secondary approach, we compared differences among ARA+, ATA+, and double-negative groups by performing ANOVA or Kruskal-Wallis tests for continuous variables, and Chi-square tests or Monte Carlo simulation for exact tests for categorical variables, as appropriate.
In additional analyses, we fit linear mixed-effects models with random intercepts to evaluate longitudinal changes in FVC% predicted and mRSS over time by autoantibody group. Models were performed in the full cohort and in the subset of participants with ILD for FVC% predicted, while mRSS models were fit in the full cohort only. mRSS models were adjusted for age, sex, race, smoking status, and immunosuppressive medication use. As FVC% predicted was already indexed to age, sex, and race, models for this outcome were additionally adjusted for smoking status and immunosuppressive medication use only.
Statistical significance was defined using a two-sided alpha of 0.05. All analyses were conducted in SAS version 9.4 (SAS Institute Inc, Cary, NC, USA). Figures were generated in R version 4.4.3 (R Foundation for Statistical Computing, Vienna, Austria).
3. Results
3.1. Baseline Characteristics
Three-hundred three participants were enrolled in the PRESS cohort. The 66 excluded due to missing data were more likely to have never smoked, to be ANA negative, and to have no history of renal crisis compared to included participants (Supplementary Table 1). Four participants were positive for both ARA and ATA and were therefore excluded from further analysis (Supplementary Table 2).
Two-hundred thirty-three participants were included in this study – 108 were ARA+, 68 were ATA+, and 57 were double-negative (Table 1). Among the double-negative group, 1 (1.8%) was centromere positive, 1 (1.8%) was fibrillarin positive, and 1 (1.8%) was Th/To positive. In the remainder of the double-negative group, the SSc-specific autoantibodies were negative or unknown, as extended panels including Pm/Scl, Th/To, and fibrillarin were not part of routine clinical workup at the time of data collection (Supplementary Table 3). Eighty percent (44/55) of the double-negative group was ANA positive. The mean disease duration was 1.2 (SD 0.7) years from the first non-Raynaud’s symptom. The median follow-up time was 629.0 (IQR 189.0, 1040.0) days and was significantly shorter in the ARA+ group compared to the ATA+ and double-negative groups (441.5 vs. 769.5 vs. 701.1 days, respectively; p=0.029).
Table 1 –
Baseline Characteristics
| Total (n=233) | ARA (+) (n=108) |
ATA(+) (n=68) |
ARA (−) and ATA(−) (n=57) |
P value [ARA (+) vs. ATA(+)] |
P value (three groups) |
|
|---|---|---|---|---|---|---|
| Mean (SD) age, years | 51.0 (13.6) n=233 |
56.6 (10.7) n=108 |
43.4 (14.6) n=68 |
49.6 (12.6) n=57 |
<0.0001a | <0.0001e |
| Female sex, no. (%) | 159 (68.2%) | 70 (64.8%) | 47 (69.1%) | 42 (73.7%) | 0.5560b | 0.4995b |
| Race, no. (%) | n=229 | n=107 | n=66 | n=56 | ||
| White | 174 (76.0%) | 94 (87.9%) | 44 (66.7%) | 36 (64.3%) | 0.0003c | 0.0007f |
| African American | 33 (14.4%) | 5 (4.7%) | 14 (21.2%) | 14 (25.0%) | ||
| Asian | 9 (3.9%) | 1 (0.9%) | 5 (7.6%) | 3 (5.4%) | ||
| Other | 13 (5.7%) | 7 (6.5%) | 3 (4.5%) | 3 (5.4%) | ||
| Hispanic ethnicity, no. (%) | 29 (12.6%) n=231 |
12 (11.2%) n=107 |
7 (10.4%) n=67 |
10 (17.5%) n=57 |
0.8745b | 0.4193b |
| Smoking status, no. (%) | ||||||
| Never | 135 (57.9%) | 49 (45.4%) | 47 (69.1%) | 39 (68.4%) | 0.0021b | 0.0015b |
| Former/Current | 98 (42.1%) | 59 (54.6%) | 21 (30.9%) | 18 (31.6%) | ||
| Disease duration (years), mean (±SD) since first non-Raynaud’s phenomenon symptom | 1.2 (0.7) | 1.1 (0.7) | 1.2 (0.6) | 1.2 (0.6) | 0.4932d | 0.2495g |
| Disease duration (years), mean (±SD), median (IQR) since onset of Raynaud’s phenomenon | 2.5 (4.7) 1.2 (0.7, 2.0) n=219 |
2.3 (5.1) 1.0 (0.6, 2.0) n=102 |
2.6 (4.7) 1.3 (0.9, 2.0) n=64 |
2.5 (3.8) 1.2 (0.9, 2.2) n=53 |
0.0727d | 0.1229g |
| Disease duration less than 6 months, n (%) | 25 (10.7%) | 12 (11.1%) | 8 (11.8%) | 5 (8.8%) | 0.8942b | 0.8519b |
| First scleroderma symptom, n (%) | ||||||
| Puffy hands or fingers | 103 (44.2%) | 47 (43.5%) | 34 (50.0%) | 22 (38.6%) | 0.6328c | 0.8663f |
| Dyspnea | 12 (5.2%) | 3 (2.8%) | 4 (5.9%) | 5 (8.8%) | ||
| Arthritis | 16 (6.9%) | 7 (6.5%) | 4 (5.9%) | 5 (8.8%) | ||
| Reflux | 4 (1.7%) | 1 (0.9%) | 2 (2.9%) | 1 (1.8%) | ||
| Raynaud’s phenomenon | 47 (20.2%) | 22 (20.4%) | 13 (19.1%) | 12 (21.1%) | ||
| Skin tightening | 30 (12.9%) | 17 (15.7%) | 6 (8.8%) | 7 (12.3%) | ||
| Other* | 21 (9.0%) | 11 (10.2%) | 5 (7.4%) | 5 (8.8%) | ||
| ANA | 202 (91.8%) n=220 |
94 (93.1%) n=101 |
64 (100.0%) n=64 |
44 (80.0%) n=55 |
0.0437c | 0.0003b |
| Baseline mRSS, mean (±SD) | 21.5 (10.2) n=228 |
25.0 (10.1) n=106 |
17.9 (11.0) n=67 |
19.0 (6.9) n=55 |
<0.0001a | <0.0001e |
| PFTs | ||||||
| FVC, liters, mean (±SD) | 3.0 (0.8) n=201 |
3.2 (0.8) n=93 |
3.0 (0.9) n=62 |
2.8 (0.7) n=46 |
0.0553a | 0.0115e |
| FVC, % predicted, mean (±SD) | 80.8 (18.8) n=201 |
86.0 (16.4) n=93 |
76.0 (21.1) n=63 |
76.8 (17.8) n=45 |
0.0019a | 0.0011e |
| TLC, liters, mean (±SD) | 4.8 (1.2) n=141 |
5.2 (1.2) n=63 |
4.4 (1.2) n=46 |
4.4 (0.9) n=32 |
0.0004a | 0.0001e |
| TLC, % predicted, mean (±SD) | 88.0 (20.1) n=137 |
96.1 (18.3) n=60 |
80.6 (19.9) n=46 |
83.3 (18.7) n=31 |
<0.0001a | 0.0001e |
| DLCO, mL/min/mmHg, mean (±SD) | 17.4 (6.3) n=188 |
18.3 (6.3) n=89 |
16.2 (6.2) n=57 |
17.0 (6.3) n=42 |
0.0553a | 0.1404e |
| DLCO, % predicted, mean (±SD) | 69.9 (24.7) n=191 |
75.3 (24.3) n=91 |
62.7 (24.6) n=58 |
68.1 (23.2) n=42 |
0.0025a | 0.0080e |
| Interstitial lung disease on high resolution chest CT, no. (%) | 101 (53.7%) n=188 |
39 (44.3%) n=88 |
39 (69.6%) n=56 |
23 (52.3%) n=44 |
0.0029b | 0.0118b |
| Renal crisis history, no. (%) | 16 (7.5%) n=214 |
15 (14.4%) n=104 |
0 (0.0%) n=58 |
1 (1.9%) n=52 |
0.0024b | 0.0005c |
| GAVE history, no. (%) | 22 (10.7%) n=205 |
14 (14.6%) n=96 |
4 (6.7%) n=60 |
4 (8.2%) n=49 |
0.1321b | 0.2394b |
| Treatments | ||||||
| Mycophenolate mofetil, no. (%) | 93 (39.9%) | 38 (35.2%) | 27 (39.7%) | 28 (49.1%) | 0.5451b | 0.2205b |
| Prednisone, no. (%) | 68 (29.2%) | 32 (29.6%) | 23 (33.8%) | 13 (22.8%) | 0.5589b | 0.3985b |
| Methotrexate, no. (%) | 31 (13.3%) | 19 (17.6%) | 11 (16.2%) | 1 (1.8%) | 0.8078b | 0.0123b |
| Hydroxychloroquine, no. (%) | 25 (10.7%) | 11 (10.2%) | 9 (13.2%) | 5 (8.8%) | 0.5347b | 0.7021b |
| Cyclophosphamide, no. (%) | 4 (1.7%) | 2 (1.9%) | 0 (0.0%) | 2 (3.5%) | 0.5231c | 0.2745c |
| D-penicillamine, no. (%) | 5 (2.1%) | 4 (3.7%) | 1 (1.5%) | 0 (0.0%) | 0.6500c | 0.4478c |
| Azathioprine, no. (%) | 4 (1.7%) | 3 (2.8%) | 1 (1.5%) | 0 (0.0%) | 1.0000c | 0.6957c |
| Investigational drug, no./no. assessed (%) | 7 (3.0%) n=231 |
4 (3.8%) n=106 |
2 (2.9%) n=68 |
1 (1.8%) n=57 |
1.0000c | 0.8896c |
| Supplemental oxygen use, no./no. assessed (%) | 3 (1.3%) n=232 |
1 (0.9%) n=107 |
0 (0.0%) n=68 |
2 (3.5%) n=57 |
1.0000c | 0.2378c |
| Median Follow-Up Time (days), median (IQR) | 629.0 (189.0, 1040.0) | 441.5 (130.5, 922.5) | 769.5 (271.0, 1382.0) | 701.0 (372.0, 1097.0) | 0.0323d | 0.0285g |
Other first scleroderma symptom includes lower extremity swelling, telangiectasias, ankle/wrist inflammation, joint pain, fatigue, myalgias, carpal tunnel, cold/numbness in extremities, pruritis, hypo/hyper-pigmentation, hypertension, cough, and gastrointestinal discomfort
T-test
Chi-square test
Fisher exact test
Mann-Whitney U test
ANOVA test
Monte-Carlo simulation for exact test
Kruskal-Wallis test
ANA, antinuclear antibody; ARA, anti-RNA polymerase III; ATA, anti-topoisomerase I; CT, computed tomography; DLCO, diffusing capacity of the lungs for carbon monoxide; GAVE, gastric antral vascular ectasia; HRCT, high resolution computed tomography; ILD, interstitial lung disease; IQR, interquartile range; mL/min/mmHg, milliliters per minute per millimeter of mercury; mRSS, modified Rodnan skin score; PFT, pulmonary function test; SD, standard deviation; TLC, total lung capacity.
ATA+ participants were younger than ARA+ and double-negative participants (mean 43.4 [SD 14.6] vs. mean 56.6 [SD 10.7] vs. mean 49.6 [SD 12.6] years, p< 0.001). Self-identified race differed between the three groups (p=0.0007). Most participants self-identified as White, including 87.9% (94/107) of the ARA+ group, 66.7% (44/66) of the ATA+ group, and 64.3% (36/56) of the double-negative group. Only 4.7% (5/107) of ARA+ participants self-identified as African American compared to 21.2% (14/66) of ATA+ participants and 25% (14/56) of double-negative participants. A greater proportion of participants in the ARA+ group were ever smokers than those in the ATA+ and double-negative groups (54.6% vs. 30.9% vs. 31.6%, p=0.0015). Mean baseline mRSS was higher in the ARA+ group (25.0 [SD 10.1]) than in the ATA+ (17.9 [SD 11.0]) and double-negative (19.0 [SD 6.9]) groups (p<0.001). The mean FVC% predicted was statistically significantly lower in the ATA+ (76.0 [SD 21.1]) and double-negative groups (76.8 [SD 17.8]) compared to the ARA+ group (86.0 [SD 16.4]) (p=0.0011). Mean total lung capacity (TLC)% predicted was also statistically significantly lower in the ATA+ (80.6 [SD 19.9]) and double-negative groups (83.3 [18.7]) compared to the ARA+ group (96.1 [SD 18.3]) (p=0.0001). Diffusion capacity for carbon monoxide (DLCO)% predicted was lowest in the ATA+ group. A greater proportion of ATA+ participants had ILD on HRCT (69.6%) compared to ARA+ (44.3%) and double-negative (52.3%) participants (p=0.012).
A statistically significantly greater proportion of participants in the ARA+ group had a history of SRC at baseline (14.4%) compared to 0.0% in the ATA+ group and 1.9% in the double-negative group (p=0.0005). A greater proportion of participants in the ARA+ group than in the ATA+ and double-negative groups had a history of GAVE at baseline, although this difference was not statistically significant (Table 1).
At baseline, a statistically significantly greater proportion of participants in the ATA+ and ARA+ groups were taking methotrexate compared to the double-negative group (17.6% vs. 16.2% vs. 1.8%, p=0.012) with no other significant differences in baseline medication use.
3.2. Medication Use at Follow-up
Mycophenolate use increased substantially across all groups over the 24-month follow-up period (Supplementary Table 4), with the largest gains occurring by 6 months. Mycophenolate use nearly doubled in the ATA+ and double-negative groups (39.7% to 78.6% and 49.1% to 84.6%, respectively) from baseline to 24 months. Methotrexate use remained relatively stable over follow-up, while cyclophosphamide use was low across all groups and nearly absent by 12 months. Notably, there were declining sample sizes at later timepoints. Data on rituximab and nintedanib use was not collected in PRESS.
3.3. Interstitial Lung Disease at Follow-up
There was no statistically significant difference between groups in FVC, FVC% predicted, TLC, TLC% predicted, DLCO, or DLCO% predicted at 6 months (Table 2). At 12 months, the FVC% predicted, TLC, TLC% predicted, and DLCO% predicted were all statistically significantly lower in the ATA+ group compared to the ARA+ group (Table 2).
Table 2 –
Pulmonary function tests and high-resolution chest computed tomography findings at 6, 12, and 24 months
| Total (n=233) |
ARA (+) (n=108) |
ATA(+) (n=68) |
ARA (−) and ATA(−) (n=57) |
P value [ARA (+) vs. ATA(+)] |
P value (three groups) |
|
|---|---|---|---|---|---|---|
| 6 months | ||||||
| PFTs | ||||||
| FVC, liters, mean (±SD) | 3.0 (0.8) n=89 |
3.1 (0.6) n=32 |
3.1 (1.0) n=28 |
2.9 (0.8) n=29 |
0.8708a | 0.4298d |
| FVC, % predicted, mean (±SD) | 79.6 (18.4) n=89 |
85.1 (18.6) n=32 |
77.2 (19.8) n=28 |
76.0 (15.7) n=29 |
0.1164a | 0.1063d |
| ΔFVC (FVC at follow-up – initial FVC), mean (±SD) | 0.1 (11.4) n=79 |
2.0 (8.5) n=29 |
−1.0 (7.1) n=25 |
−0.9 (16.7) n=25 |
0.1761a | 0.5526d |
| TLC, liters, mean (±SD) | 4.6 (1.0) n=38 |
4.8 (0.5) n=12 |
4.5 (1.3) n=13 |
4.7 (1.0) n=13 |
0.0770b | 0.7502d |
| TLC, % predicted, mean (±SD) | 86.3 (22.4) n=36 |
91.6 (13.3) n=11 |
81.0 (30.3) n=12 |
86.8 (20.5) n=13 |
0.0904b | 0.5310d |
| DLCO, mL/min/mmHg, mean (±SD) | 17.3 (6.5) n=84 |
18.0 (6.5) n=32 |
17.3 (7.4) n=26 |
16.3 (5.5) n=26 |
0.7213a | 0.6296d |
| DLCO, % predicted, mean (±SD) | 68.0 (23.3) n=85 |
73.7 (23.2) n=32 |
63.4 (25.3) n=27 |
65.8 (20.6) n=26 |
0.1098a | 0.2061d |
| Interstitial lung disease on high resolution CT chest, no. (%) | 105 (54.7%) n=192 |
39 (44.3%) n=88 |
42 (75.0%) n=56 |
24 (50.0%) n=48 |
0.0003c | 0.0011c |
| 12 months | ||||||
| PFTs | ||||||
| FVC, liters, mean (±SD) | 3.0 (0.9) n=92 |
3.2 (0.9) n=41 |
2.9 (0.9) n=28 |
2.8 (0.9) n=23 |
0.1253a | 0.1479a |
| FVC, % predicted, mean (±SD) | 80.9 (20.2) n=92 |
87.9 (19.5) n=41 |
73.5 (21.5) n=28 |
77.7 (16.3) n=23 |
0.0052a | 0.0088a |
| ΔFVC (FVC at follow-up – initial FVC), mean (±SD) | 1.1 (8.9) n=86 |
1.1 (10.0) n=38 |
1.1 (8.6) n=27 |
1.0 (7.3) n=21 |
0.9965a | 0.9991a |
| TLC, liters, mean (±SD) | 4.7 (1.2) n=43 |
5.3 (1.3) n=20 |
4.2 (0.9) n=13 |
4.3 (0.8) n=10 |
0.0143a | 0.0141a |
| TLC, % predicted, mean (±SD) | 85.0 (18.8) n=42 |
94.5 (19.6) n=20 |
74.4 (15.7) n=13 |
79.3 (9.0) n=9 |
0.0041a | 0.0041a |
| DLCO, mL/min/mmHg, mean (±SD) | 17.7 (6.1) n=86 |
17.9 (6.3) n=37 |
17.2 (6.5) n=28 |
17.9 (5.1) n=21 |
0.6328a | 0.8609a |
| DLCO, % predicted, mean (±SD) | 68.5 (21.0) n=86 |
72.1 (19.8) n=37 |
61.2 (21.4) n=28 |
71.9 (21.2) n=21 |
0.0386a | 0.0830a |
| Interstitial lung disease on high resolution CT chest, no. (%) | 111 (56.1%) N=198 |
41 (45.1%) n=91 |
43 (75.4%) n=57 |
27 (54.0%) N=50 |
0.0003c | 0.0013c |
| 24 months | ||||||
| PFTs | ||||||
| FVC, liters, mean (±SD) | 3.0 (0.8) n=54 |
3.1 (0.9) n=21 |
3.0 (0.8) n=15 |
2.9 (0.6) n=18 |
0.8821a | 0.8173d |
| FVC, % predicted, mean (±SD) | 81.5 (17.8) n=54 |
88.3 (15.6) n=21 |
72.1 (19.8) n=15 |
81.3 (15.5) n=18 |
0.0093a | 0.0224d |
| ΔFVC (FVC at follow-up – initial FVC), mean (±SD) | 1.6 (10.1) n=49 |
2.8 (10.9) n=20 |
−2.1 (10.8) n=15 |
3.6 (7.6) n=14 |
0.1972a | 0.2481d |
| TLC, liters, mean (±SD) | 5.2 (1.2) n=21 |
5.3 (1.7) n=8 |
4.6 (1.2) n=4 |
5.3 (0.5) n=9 |
0.5522b | 0.6286e |
| TLC, % predicted, mean (±SD) | 92.5 (18.4) n=21 |
99.4 (25.0) n=8 |
88.0 (11.0) n=5 |
88.5 (13.5) n=8 |
0.4195b | 0.4881e |
| DLCO, mL/min/mmHg, mean (±SD) | 17.2 (5.6) n=51 |
17.0 (6.3) n=21 |
17.8 (5.9) n=13 |
16.9 (4.5) n=17 |
0.7116a | 0.9030d |
| DLCO, % predicted, mean (±SD) | 70.0 (20.7) n=54 |
73.9 (19.2) n=21 |
64.9 (23.6) n=15 |
69.7 (20.0) n=18 |
0.2166a | 0.4456d |
| Interstitial lung disease on high resolution CT chest, no. (%) | 118 (57.6%) n=205 |
45 (47.9%) n=94 |
45 (76.3%) n=59 |
28 (53.8%) n=52 |
0.0005c | 0.0021c |
T-test
Mann-Whitney U test
Chi-square test
ANOVA test
Kruskal-Wallis test
ARA, anti-RNA polymerase III; ATA, anti-topoisomerase I; CT, computed tomography; DLCO, diffusing capacity of the lungs for carbon monoxide; ΔFVC, change in forced vital capacity; FVC, forced vital capacity; ILD, interstitial lung disease; mL/min/mmHg, milliliters per minute per millimeter of mercury; PFT, pulmonary function test; SD, standard deviation; TLC, total lung capacity
At 24 months, FVC% predicted remained statistically significantly lower in the ATA+ group compared to the ARA+ group (p=0.0093). At baseline and at 24 months, the double-negative group had FVC% predicted and DLCO% predicted intermediate between the ARA+ and ATA+ groups. At baseline, these were statistically significant differences, while at 24 months only FVC% predicted was statistically significantly different across groups (p=0.022). The change in FVC (ΔFVC) at 6-, 12-, and 24 months was not statistically significantly different between the groups (Table 2).
At baseline or at any point during the follow-up period, 58.8% (124/211) of participants had or developed ILD on HRCT: 75.8% (47/62) in the ATA+ group, 50.0% (48/96) in the ARA+ group, and 54.7% (29/53) in the double-negative group (p=0.004) (Table 3).
Table 3 –
Prevalence of organ involvement at baseline or at any time during follow-up
| Total (n=233) |
ARA (+) (n=108) |
ATA(+) (n=68) |
ARA (−) and ATA(−) (n=57) |
P value [ARA (+) vs. ATA(+)]a |
P value (three groups)a |
|
|---|---|---|---|---|---|---|
| Interstitial lung disease on high resolution CT chest, no. (%) | 124 (58.8%) n=211 |
48 (50.0%) n=96 |
47 (75.8%) n=62 |
29 (54.7%) n=53 |
0.0012 | 0.0044 |
| Renal crisis, no. (%) | 27 (11.9%) n=226 |
20 (18.9%) n=106 |
4 (6.3%) n=64 |
3 (5.4%) n=56 |
0.0221 | 0.0105 |
| GAVE, no. (%) | 44 (19.7%) n=223 |
25 (24.3%) n=103 |
11 (16.7%) n=66 |
8 (14.8%) n=54 |
0.2388 | 0.2785 |
Chi-square test
ARA, anti-RNA polymerase III; ATA, anti-topoisomerase I; CT, computed tomography; GAVE, gastric antral vascular ectasia; ILD, interstitial lung disease; no., number
Among participants with ILD, baseline FVC %predicted was similar in the ATA+ (74.7%) and double-negative (71.4%) groups and higher in the ARA+ group (82.9%; Supplementary Table 5). Over the first 12 months, FVC% predicted remained stable in the ARA+ (82.9% to 85.9%) and ATA+ groups (74.7% to 73.8%) and improved in the double-negative group (71.4% to 80.1%). Of note, fewer participants in the ILD subgroup had PFTs available at 12 months across all autoantibody groups. PFT availability was further reduced at 24 months, precluding reliable trajectory assessment at that time point.
In a linear mixed effects model of the change in FVC by autoantibody group in the entire cohort, there was no statistically significant change in the FVC %predicted (ΔFVC%) per year in the ARA+ or ATA+ groups (Figure 1, Supplementary Table 6). However, the double-negative group had a statistically significant increase in the ΔFVC% per year of 1.39 (95% confidence interval [CI] 0.11 to 2.67, p=0.03). The difference of ΔFVC% per year between the ARA+ group and the double-negative group was −1.73 (95% CI −3.46 to 0.003, p=0.05). There was no statistically significant difference in the ΔFVC% per year when comparing the ATA+ group to the double-negative group nor when comparing the ARA+ to the ATA+ group.
Figure 1.
Change in FVC% predicted over time by autoantibody group in the full cohort (panel A) and in participants with SSc-ILD (panel B). The red dotted line represents ARA+ participants, the blue dashed line represents ATA+ participants, and the green solid line represents participants who were negative for both ARA and ATA (i.e., “double-negative”). P-values indicate whether the estimated slope of FVC% predicted over time differs significantly from zero within each autoantibody group.
FVC, forced vital capacity; SSc, systemic sclerosis; ILD, interstitial lung disease; ARA, anti-RNA-polymerase-III antibody; ATA, anti-topoisomerase-I antibody
Similarly, when applying the linear mixed effects model only to participants with ILD, there was no statistically significant ΔFVC% per year in the ARA+ or ATA+ groups (Figure 1, Supplementary Table 7). However, the double-negative group had a statistically significant increase in the ΔFVC% per year of 1.61 (95% CI 0.27 to 2.95, p=0.019). The difference in ΔFVC% between the ATA+ and double-negative group was −2.09 (95% CI −3.93 to −0.25, p=0.026). Similarly, the difference between the ARA+ group and the double-negative group was −2.38 (95% CI −4.22 to −0.54, p=0.012). There was no statistically significant difference in the ΔFVC% per year when comparing the ATA+ and ARA+ groups.
3.4. Scleroderma Renal Crisis and Gastric Antral Vascular Ectasia
During the study period, 11.9% (27/226) of participants had or developed SRC during follow up, including 18.9% (20/106) of the ARA+ group, 6.3% (4/64) of the ATA+ group, and 5.4% (3/56) of the double-negative group (p=0.011). In addition, 19.7% (44/223) of participants had or developed GAVE during the follow up period, including 24.3% (25/103) of the ARA+ group, 16.7% (11/66) of the ATA+ group, and 14.8% (8/54) of the double-negative group. The difference between groups was not statistically significant (Table 3).
3.5. Trends in mRSS
While baseline mRSS was statistically significantly higher in the ARA+ group, there was no statistically significant difference in skin scores between groups at the 6-, 12-, or 24-month follow-up time points (Table 4). However, there was a statistically significant difference in the change in the mRSS between groups at 12 months. The mRSS declined by a mean of 6.0 (SD 8.1) units in the ARA+ group and by a mean of 5.7 (SD 4.9) in the double-negative group compared to a mean decline of only 1.9 (SD 8.3) in the ATA+ group (p=0.03).
Table 4 –
Modified Rodnan Skin Score trends
| Total (n=233) |
ARA(+) (n=108) |
ATA(+) (n=68) |
ARA(−) and ATA(−) (n=57) |
P value [ARA(+) vs. ATA(+)]a |
P value (three groups)b |
|
|---|---|---|---|---|---|---|
| mRSS at enrollment, mean (±SD) | 21.5 (10.2) n=228 |
25.0 (10.1) n=106 |
17.9 (11.0) n=67 |
19.0 (6.9) n=55 |
<0.0001 | <0.0001 |
| mRSS at 6-month follow-up, mean (±SD) | 19.1 (9.9) n=143 |
20.8 (10.7) n=65 |
17.0 (10.1) n=39 |
18.2 (7.8) n=39 |
0.0758 | 0.1314 |
| ΔmRSS score from baseline to 6 months, mean (±SD) | −2.2 (6.1) n=140 |
−3.1 (7.1) n=63 |
−0.8 (5.3) n=38 |
−2.0 (4.7) n=39 |
0.0758 | 0.1581 |
| mRSS at 12-month follow-up, mean (±SD) | 15.5 (9.0) n=130 |
17.6 (9.5) n=56 |
15.4 (9.4) n=36 |
12.7 (7.0) n=38 |
0.2830 | 0.0377 |
| ΔmRSS from baseline to 12 months, mean (±SD) | −4.8 (7.5) n=127 |
−6.0 (8.1) n=54 |
−1.9 (8.3) n=35 |
−5.7 (4.9) n=38 |
0.0264 | 0.0300 |
| mRSS at 24-month follow-up, mean (±SD) | 11.7 (8.3) n=75 |
14.7 (9.6) n=29 |
11.2 (7.1) n=23 |
8.3 (6.5) n=23 |
0.1442 | 0.0191 |
| ΔmRSS from baseline to 24 months, mean (±SD) | −10.6 (9.6) n=74 |
−12.4 (10.8) n=29 |
−8.3 (9.8) n=22 |
−10.4 (7.4) n=23 |
0.1665 | 0.3152 |
all comparisons are conducted via T-test.
all comparisons are conducted via ANOVA test.
ANOVA, analysis of variance; ARA, anti-RNA polymerase III; ATA, anti-topoisomerase I; mRSS, modified Rodnan skin score; SD, standard deviation; ΔmRSS, change in modified Rodnan skin score.
In a linear mixed effects model of the change in mRSS by autoantibody group in the entire cohort, there were significant changes in mRSS (ΔmRSS) per year in all three antibody groups (Figure 2, Supplementary Table 8). The ARA+ group had the greatest magnitude of mRSS decline at −4.46 (95% CI −5.13 to −3.79, p<0.0001) followed by the double-negative group at −3.16 (95% CI −3.93 to −2.38, p<0.0001) and the ATA+ group at −1.60 (95% CI −2.26 to −0.94, p<0.0001). The difference of ΔmRSS per year was −1.30 (95% CI −2.33 to −0.28, p=0.013) between the ARA+ group and the double-negative group, 1.56 (95% CI 0.54 to 2.57, p=0.0028) between the ATA+ group and the double-negative group, and −2.86 (95% CI −3.80 to −1.12, p<0.0001) between the ARA+ and ATA+ groups.
Figure 2.
Change in mRSS over time by autoantibody group in the full cohort. The red dotted line represents ARA+ participants, the blue dashed line represents ATA+ participants, and the green solid line represents participants who were negative for both ARA and ATA (i.e., “double-negative”). P-values indicate whether the estimated slope of mRSS over time differs significantly from zero within each autoantibody group.
mRSS, modified Rodnan Skin Score; ARA, anti-RNA-polymerase-III antibody; ATA, anti-topoisomerase-I antibody
4. Discussion
In this prospective cohort of early diffuse SSc, we observed clear autoantibody-specific associations with baseline organ involvement and disease trajectory that mirror patterns reported in broader SSc cohorts, while also revealing autoantibody-specific patterns not previously described in early dcSSc. Consistent with prior heterogenous cohorts, ATA+ participants had a higher burden of ILD and lower baseline FVC% predicted and DLCO % predicted, and ARA+ participants had higher mRSS and an increased prevalence of SRC [5]. The double-negative group showed an improvement in FVC% predicted over time, despite having a higher baseline ILD prevalence compared to the ARA+ group. Moreover, a substantial portion of participants had evidence of major organ involvement (restrictive PFTs, presence of ILD on HRCT, GAVE, and/or SRC) at their baseline visit, which is consistent with previous reports on development of major organ involvement early in the natural history of SSc [5]. Notably, in this early dcSSc cohort, participants only had a mean disease duration of 1.2 (SD 0.7) years at the baseline visit, which means that major organ involvement may occur and can be clinically identified very early in the disease course.
In the PRESS cohort, the mean baseline FVC was 80.8% predicted and 53.7% had ILD on HRCT. These values are comparable to the EUSTAR multi-national prospective cohort, where baseline FVC was 84% predicted among dcSSc participants and 53.4% had ILD on baseline HRCT[17]. In the PRESS cohort, some relationships between autoantibody status and pulmonary function were also consistent with observations from the EUSTAR cohort. ATA+ participants in PRESS had statistically significantly lower FVC, TLC, and DLCO% predicted at baseline and statistically significantly lower FVC at 12- and 24- month follow-up when compared to ARA+ and double-negative participants. Similarly, in an early SSc cohort from EUSTAR, the ATA+ group was more likely to have an FVC<80% predicted compared to ARA+ and other autoantibody groups [5].
A notable finding was the longitudinal trajectory of the double-negative group. In linear mixed effects models, the double-negative group demonstrated a statistically significant annual increase in FVC% predicted, both among participants with ILD and among the entire cohort, while no significant FVC trend was observed in the ARA+ or ATA+ groups. Among participants with ILD, the annual ΔFVC% predicted differed between the double-negative group and both the ARA+ and ATA+ groups. Although immunosuppressive use was similar between the double-negative and ATA+ groups, a differential FVC trajectory was still observed. However, the basis for this differential FVC trajectory remains unclear. It is possible that unmeasured differences in treatment adherence may have contributed, although we cannot draw firm conclusions from the available data. This finding has not been previously reported and suggests that patients with early dcSSc who are negative for ATA and ARA may have a more favorable pulmonary course, potentially reflecting differences in the natural history of dcSSc or greater responsiveness to immunosuppression. Further studies are necessary to clarify this.
ATA+ participants also had a higher prevalence of ILD at baseline and at all follow-up time points than ARA+ and double-negative participants, which is consistent with the known association between ATA positivity and greater ILD prevalence and severity derived from studies of heterogenous SSc populations [5, 18]. Interestingly, ILD prevalence in the double-negative group was intermediate between the ARA+ and ATA+ groups. This suggests that those without ARA or ATA autoantibodies, or those who harbor other SSc-related autoantibodies, may have a higher risk of developing ILD than the ARA+ group. Further study in these populations is necessary to clarify the significance of this finding.
The baseline mRSS was highest in the ARA+ group, which is consistent with this autoantibody’s known association with rapidly progressive skin thickening [18]. There was greater magnitude of improvement in mRSS at follow-up in the ARA+ group compared to the ATA+ group, even though there was no difference in immunosuppressive use between these groups. This may be because rapidly progressive skin thickening happens very early in the disease course and may have already occurred by the time of enrollment in the PRESS cohort. Notably, the mean disease duration of participants in the ARA+ group was 1.1 years, and skin thickening typically peaks within the first 1–3 years [18]. The higher baseline skin scores may represent enrollment during the peak phase of skin thickening, which commonly occurs within the first 1–2 years, before subsequent softening [5, 18, 19]. The higher baseline and greater decline in skin score in the ARA+ group compared to the ATA+ group has been observed in other cohorts [15]. Although baseline mRSS was similar between the double-negative and ATA+ groups, the double-negative group demonstrated a greater improvement in mRSS over time. This difference may reflect variation in disease course between groups and should be explored further in future studies.
While there was a higher prevalence of GAVE in the ARA+ group compared to the ATA+ and double-negative groups, this was not a statistically significant difference. Prevalence of GAVE may be underestimated in our cohort and in the literature, likely because esophagogastroduodenoscopies (EGD) are not routinely performed to screen for GAVE in SSc. Thus, occult GAVE cases may not be captured. One study of SSc participants who underwent EGDs demonstrated that 8.4% had GAVE, and that those with GAVE were more likely to be ARA+ [20]. A greater proportion of participants had GAVE at follow up in our study (22%) compared to other studies (approximately 8–10%), and the ATA+ group had a higher than expected prevalence of GAVE compared to previous reports [21, 22]. This may reflect more frequent screening or diagnostic EGDs over time or may indicate a genuinely higher prevalence of GAVE in early dcSSc compared with historical, more heterogenous SSc populations.
There was a statistically significantly higher proportion of SRC in participants with ARA positivity, consistent with the literature describing higher incidence of renal crisis in ARA+ patients [5]. In our cohort, 18.9% of ARA+ participants had or developed SRC during the follow-up period, which is lower than the 28% prevalence of SRC in the historical Pittsburgh Scleroderma Databank [23]. This change may reflect the overall declining incidence of SRC over time due to preventative changes in practice patterns, such as corticosteroid minimization, or may represent differences in in disease severity between the Pittsburgh and PRESS cohorts.
There are some limitations of this study. The median follow-up time was shorter in the ARA+ group (441.5 days) than in the ATA+ group (769.5 days), which limits comparisons between groups at later timepoints and introduces potential time-dependent bias. Additionally, PFT and mRSS data availability decreased over follow-up; PFT data at 24 months were particularly sparse, reducing statistical power for longitudinal analyses beyond 12 months – especially within-ILD comparisons of FVC trajectory between autoantibody groups. The relatively short follow-up further limited detection of later-emerging organ manifestations. Centralized reads of participants’ HRCTs were not performed. However, participants’ HRCT images were interpreted locally at each site by thoracic radiologists according to American Thoracic Society/European Respiratory Society standards [24, 25]. Furthermore, autoantibody testing was performed locally through commercial laboratories, therefore harmonization across sites was not possible. Given that multiple statistical tests were performed without correction for multiple comparisons, borderline results should be interpreted cautiously and considered hypothesis-generating, while other primary findings were largely concordant with existing literature [4]. Although nailfold capillaroscopy was performed at all participating sites, nailfold videocapillaroscopy equipment was not uniformly available, precluding meaningful analysis of capillaroscopic patterns by autoantibody group. Finally, since all participating sites were tertiary and quaternary care centers with expert scleroderma programs, this cohort may overrepresent individuals with severe organ involvement and rapidly progressive skin disease compared with patients followed in community settings, which may limit generalizability.
There are several strengths of our study. It is the largest multi-center cohort of participants with early diffuse SSc within 2 years of the first non-Raynaud’s symptom, a disease subgroup in which severe organ involvement often evolves rapidly. The focus on a narrowly defined early-disease population reduces heterogeneity and enhances internal validity, allowing more precise evaluation of autoantibody-associated disease patterns. The longitudinal design with standardized timepoints allowed for assessment of disease progression over time to better capture early phenotypic difference between serologic subgroups. Data were collected by experienced scleroderma specialists at tertiary and quaternary referral centers, ensuring diagnostic accuracy and robust collection of disease severity and outcome measures.
In conclusion, this study confirms the association of early major organ involvement in diffuse SSc within the first two years of the first non-Raynaud’s symptom. Consistent with previous reports, autoantibodies are valuable biomarkers for specific disease manifestations and may assist in prognostication even within this narrow early-disease window. ATA+ participants had a higher prevalence of ILD and lower baseline FVC% predicted, though FVC% predicted did not decline significantly over time in either the ATA+ or ARA+ groups. The double-negative group uniquely demonstrated a significant annual improvement in FVC% predicted, a novel finding that warrants cautious interpretation given the heterogeneity of this group and the exploratory nature of these analyses. Those who were ARA+ had higher baseline skin scores and higher prevalence of SRC. Our results highlight the importance of early screening for major organ involvement in diffuse SSc. Treatment of underlying organ involvement should begin promptly at recognition. Further studies are needed to characterize and subtype the double-negative group and their clinical course.
Supplementary Material
Highlights:
Organ involvement often occurs in the first two years of diffuse scleroderma onset.
Autoantibodies predict early organ involvement in diffuse systemic sclerosis.
Forced vital capacity improved over time in some scleroderma subsets.
Funding:
A.A.’s work was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (grant T32-AR-076953). E.J.B.’s work was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (grant K23-AR-075112), the National Heart, Lung, and Blood Institute (grant R01-HL-164758), and the Department of Defense (grant W81XWH2210163). AAS is supported by NIH/NIAMS K24 AR080217.
Disclosure Statement:
EJB reports consulting fees from Boehringer Ingelheim and Synthekine; and fees for contracted research from AstraZeneca, aTyr Pharma, Boehringer Ingelheim, Cabaletta Bio, Bristol Meyers Squibb, and Synthekine. RTD reports consulting for AstraZeneca, Boehringer Ingelheim, and Aisa Pharma. JKG reports consulting for Boehringer Ingelheim and Regeneron. MH reports consulting for Merck, AbbVie and Boehringer Ingelheim and has received funding for investigator-initiated grants from Kadmon Pharmaceuticals and Boehringer Ingelheim. DK has consultancy with Abbvie, Amgen, Argenx, Astra Zeneca, Boehringer Ingelheim, BMS, Cabaletta, Mirador, NKarta, Novartis, Pharmosa, ZuraBio. AS is an investigator for Argenx clinical trial.
Footnotes
Disclaimer: The work described in this manuscript was completed while Dr. Victoria Shanmugam was employed at The George Washington University Medical Faculty Associates. The opinions expressed in this article are the author’s own and do not reflect the view of the National Institutes of Health, the Department of Health and Human Services, or the United States government.
CREdIT Statement:
AA: Conceptualization; methodology; data interpretation; writing – original draft; writing – review & editing.
DH: Conceptualization; investigation; writing – original draft; writing – review & editing.
SH: Formal analysis; statistical methodology; data interpretation; writing – review & editing.
FVC, SA, RTD, TMF, JG, FH, MH, AS, VKS, VDS, DK: Data acquisition; investigation; writing – review & editing.
DK: Data acquisition; investigation; methodology; supervision; writing – review & editing.
EJB: Conceptualization; supervision; methodology; data interpretation; writing – review & editing; senior oversight.
References
- 1.Distler O, Assassi S, Cottin V, Cutolo M, Danoff SK, Denton CP, et al. Predictors of progression in systemic sclerosis patients with interstitial lung disease. European Respiratory Journal. 2020;55(5). doi: 10.1183/13993003.02026-2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Elhai M, Meune C, Avouac J, Kahan A, Allanore Y. Trends in mortality in patients with systemic sclerosis over 40 years: a systematic review and meta-analysis of cohort studies. RHEUMATOLOGY. 2012;51(6):1017–26. doi: 10.1093/rheumatology/ker269. [DOI] [PubMed] [Google Scholar]
- 3.Denton CP, Khanna D. Systemic sclerosis. Lancet. 2017;390(10103):1685–99. doi: 10.1016/s0140-6736(17)30933-9. [DOI] [PubMed] [Google Scholar]
- 4.Tyndall AJ, Bannert B, Vonk M, Airò P, Cozzi F, Carreira PE, et al. Causes and risk factors for death in systemic sclerosis: a study from the EULAR Scleroderma Trials and Research (EUSTAR) database. Annals of the Rheumatic Diseases. 2010;69(10):1809–15. doi: 10.1136/ard.2009.114264. [DOI] [PubMed] [Google Scholar]
- 5.Jaeger VK, Wirz EG, Allanore Y, Rossbach P, Riemekasten G, Hachulla E, et al. Incidences and Risk Factors of Organ Manifestations in the Early Course of Systemic Sclerosis: A Longitudinal EUSTAR Study. Plos One. 2016;11(10). doi: 10.1371/journal.pone.0163894. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Steen VD, Medsger TA. Severe organ involvement in systemic sclerosis with diffuse scleroderma. Arthritis and Rheumatism. 2000;43(11):2437–44. doi: 10.1002/1529-0131(200011)43:11<2437::aid-anr10>3.0.co;2-u. [DOI] [PubMed] [Google Scholar]
- 7.Penn H, Howie AJ, Kingdon EJ, Bunn CC, Stratton RJ, Black CM, et al. Scleroderma renal crisis: patient characteristics and long-term outcomes. Qjm-an International Journal of Medicine. 2007;100(8):485–94. doi: 10.1093/qjmed/hcm052. [DOI] [PubMed] [Google Scholar]
- 8.Gordon SM, Stitt RS, Nee R, Bailey WT, Little DJ, Knight KR, et al. Risk Factors for Future Scleroderma Renal Crisis at Systemic Sclerosis Diagnosis. Journal of Rheumatology. 2019;46(1):85–92. doi: 10.3899/jrheum.171186. [DOI] [PubMed] [Google Scholar]
- 9.Nguyen B, Assassi S, Arnett FC, Mayes MD. Association of RNA Polymerase III Antibodies with Scleroderma Renal Crisis. Journal of Rheumatology. 2010;37(5):1068–9. doi: 10.3899/jrheum.091048. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Reveille JD, Solomon DH, Amer Coll Rheumatology Ad Hoc C. Evidence-based guidelines for the use of immunologic tests: Anticentromere, Scl-70, and nucleolar antibodies. Arthritis & Rheumatism-Arthritis Care & Research. 2003;49(3):399–412. doi: 10.1002/art.11113. [DOI] [PubMed] [Google Scholar]
- 11.Steen VD, Powell DL, Medsger TA. CLINICAL CORRELATIONS AND PROGNOSIS BASED ON SERUM AUTOANTIBODIES IN PATIENTS WITH SYSTEMIC-SCLEROSIS. Arthritis and Rheumatism. 1988;31(2):196–203. doi: 10.1002/art.1780310207. [DOI] [PubMed] [Google Scholar]
- 12.Steen VD. Autoantibodies in systemic sclerosis. Seminars in Arthritis and Rheumatism. 2005;35(1):35–42. doi: 10.1016/j.semarthrit.2005.03.005. [DOI] [PubMed] [Google Scholar]
- 13.Cottin V, Brown KK. Interstitial lung disease associated with systemic sclerosis (SSc-ILD). Respiratory Research. 2019;20. doi: 10.1186/s12931-019-0980-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Walker UA, Tyndall A, Czirják L, Denton C, Farge-Bancel D, Kowal-Bielecka O, et al. Clinical risk assessment of organ manifestations in systemic sclerosis:: a report from the EULAR Scleroderma Trials And Research group database. Annals of the Rheumatic Diseases. 2007;66(6):754–63. doi: 10.1136/ard.2006.062901. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Nihtyanova SI, Sari A, Harvey JC, Leslie A, Derrett-Smith EC, Fonseca C, et al. Using Autoantibodies and Cutaneous Subset to Develop Outcome-Based Disease Classification in Systemic Sclerosis. Arthritis & Rheumatology. 2020;72(3):465–76. doi: 10.1002/art.41153. [DOI] [PubMed] [Google Scholar]
- 16.van den Hoogen F, Khanna D, Fransen J, Johnson SR, Baron M, Tyndall A, et al. 2013 classification criteria for systemic sclerosis: an American college of rheumatology/European league against rheumatism collaborative initiative. Annals of the Rheumatic Diseases. 2013;72(11):1747–55. doi: 10.1136/annrheumdis-2013-204424. [DOI] [PubMed] [Google Scholar]
- 17.Meier FMP, Frommer KW, Dinser R, Walker UA, Czirjak L, Denton CP, et al. Update on the profile of the EUSTAR cohort: an analysis of the EULAR Scleroderma Trials and Research group database. Annals of the Rheumatic Diseases. 2012;71(8):1355–60. doi: 10.1136/annrheumdis-2011-200742. [DOI] [PubMed] [Google Scholar]
- 18.Wirz EG, Jaeger VK, Allanore Y, Riemekasten G, Hachulla E, Distler O, et al. Incidence and predictors of cutaneous manifestations during the early course of systemic sclerosis: a 10-year longitudinal study from the EUSTAR database. Annals of the Rheumatic Diseases. 2016;75(7):1285–92. doi: 10.1136/annrheumdis-2015-207271. [DOI] [PubMed] [Google Scholar]
- 19.Amjadi S, Maranian P, Furst DE, Clements PJ, Wong WK, Postlethwaite AE, et al. Course of the Modified Rodnan Skin Thickness Score in Systemic Sclerosis Clinical Trials Analysis of Three Large Multicenter, Double-Blind, Randomized Controlled Trials. Arthritis and Rheumatism. 2009;60(8):2490–8. doi: 10.1002/art.24681. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Serling-Boyd N, Chung MPS, Li SF, Becker L, Fernandez-Becker N, Clarke J, et al. Gastric antral vascular ectasia in systemic sclerosis: Association with anti-RNA polymerase III and negative anti-nuclear antibodies. Seminars in Arthritis and Rheumatism. 2020;50(5):938–42. doi: 10.1016/j.semarthrit.2020.06.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Meyer O, De Chaisemartin L, Nicaise-Roland P, Cabane J, Tubach F, Dieude P, et al. Anti-RNA Polymerase III Antibody Prevalence and Associated Clinical Manifestations in a Large Series of French Patients with Systemic Sclerosis: A Cross-sectional Study. Journal of Rheumatology. 2010;37(1):125–30. doi: 10.3899/jrheum.090677. [DOI] [PubMed] [Google Scholar]
- 22.Castelino FV, VanBuren JM, Startup E, Assassi S, Bernstein EJ, Chung L, et al. Baseline characteristics of systemic sclerosis patients with restrictive lung disease in a multi-center US-based longitudinal registry. International Journal of Rheumatic Diseases. 2022;25(2):163–74. doi: 10.1111/1756-185x.14253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Steen V Changes in causes of death in systemic sclerosis over the past 30 years. Arthritis and Rheumatism. 2004;50(9):S422–S. [Google Scholar]
- 24.Travis W, Costabel U, Hansell D, King T, Lynch D, Nicholson A, et al. An Official American Thoracic Society/European Respiratory Society Statement: Update of the International Multidisciplinary Classification of the Idiopathic Interstitial Pneumonias. AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE. 2013;188(6):733–48. doi: 10.1164/rccm.201308-1483ST. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Agustí C American Thoracic Society/European Respiratory Society International Multidisciplinary Consensus Classification of the Idiopathic Interstitial Pneumonias (vol 165, pg 277, 2002). AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE. 2002;166(3):426-. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.