Abstract
Objective We aimed to compare the progression of systemic sclerosis–associated interstitial lung disease (SSc-ILD) based on serological status.
Methods In a posthoc analysis of the SENSCIS trial (nintedanib vs placebo in SSc-ILD; ClinicalTrials.gov: NCT02597933), we analyzed the rate of decline in forced vital capacity (FVC) over 52 weeks in 3 subsets: (1) positive for anticentromere antibody (ACA), (2) positive for anti-RNA polymerase III antibody (ARA), and (3) negative for ACA, ARA, and antitopoisomerase I antibody (ATA).
Results Among study participants who underwent baseline serological evaluation, 32/549 (5.8%) were ACA positive, 98/528 (18.6%) were ARA positive, and 127/526 (24.1%) were negative for ACA, ARA, and ATA. Among the serological subsets of interest, in the placebo arm, the adjusted rate (standard error) of decline in FVC was −31.2 (41.5) mL/year among participants who were positive for ACA and −64.7 (35.1) mL/year among participants who were positive for ARA, numerically lower than in the overall SENSCIS trial population (−93.3 [13.5] mL/yr). However, participants who were negative for ACA, ARA, and ATA experienced a numerically greater rate of decline in FVC than the overall trial population, both in those randomized to placebo (−115.6 [35.4] mL/yr vs −93.3 [13.5] mL/yr) and those randomized to nintedanib (−91.8 [34.3] mL/yr vs −52.4 [13.8] mL/yr).
Conclusion These analyses of data from the SENSCIS trial suggest that patients with SSc-ILD who are ACA positive or ARA positive can experience progression of SSc-ILD. Patients negative for ACA, ARA, and ATA had a higher rate of progression than the overall trial population and should be monitored closely.
Several observational studies in patients with systemic sclerosis–associated interstitial lung disease (SSc-ILD) have demonstrated that the presence of specific autoantibodies affects outcomes. For example, antitopoisomerase I antibody (ATA) positivity has been associated with an increased risk of developing clinically significant ILD,1-3 as well as progressive ILD.4-6 In contrast, anticentromere antibody (ACA) positivity is associated with a lower risk of developing clinically significant SSc-ILD.1-3,7 A limitation of these studies is a lack of control for confounding factors, such as treatment, that may affect outcomes. Clinical trial cohorts offer a unique opportunity to investigate outcomes based on serological profiling among patients receiving standardized therapy with regular follow-up assessments.
The SENSCIS trial (ClinicalTrials.gov: NCT02597933) was conducted in patients with SSc-ILD.8 The primary outcome showed that treatment with nintedanib was associated with a significant reduction vs placebo in the rate of decline in forced vital capacity (FVC; mL/yr) over 52 weeks,8 leading to the regulatory approval of nintedanib as a treatment for SSc-ILD and to conditional recommendations for its use in treatment guidelines for SSc-ILD.9,10 The present study used data from this trial to determine whether the presence of specific autoantibodies affected the rate of progression of SSc-ILD. Whereas a prior analysis addressed outcomes based on the presence or absence of ATA,11 the current analyses evaluated the rate of decline of FVC and the effects of nintedanib in the following 3 subsets: (1) positive for ACA, (2) positive for anti-RNA polymerase III antibody (ARA), and (3) negative for ACA, ARA, and ATA.
METHODS
Patient population. The SENSCIS trial was a randomized, double-blind, placebo-controlled trial conducted in 32 countries.8 The entry criteria have been published.8 Key inclusion criteria included: (1) first non-Raynaud phenomenon symptom in the prior ≤ 7 years, (2) extent of fibrotic ILD on high-resolution computed tomography scan of the chest ≥ 10% (based on assessment of the whole lung), (3) FVC ≥ 40% predicted, and (4) diffusing lung capacity for carbon monoxide 30-89% predicted. Patients on prednisone ≤ 10 mg/day and/or stable therapy with mycophenolate (MMF) or methotrexate for ≥ 6 months were permitted to participate. Patients were randomized 1:1 to receive nintedanib 150 mg twice daily or placebo, stratified by the presence of ATA.
Serological profiles. ATA status was determined based on historical information (local laboratory data) or, if this was not available, central laboratory data obtained using a BioPlex 2200 System bead assay (Bio-Rad). ACA status was determined in a central laboratory using a BioPlex 2200 System bead assay (Bio-Rad). The cut-off to be classified as ACA negative was an antibody index < 1.0. ARA status was determined in a central laboratory based on ELISA (QUANTA Lite plate reader, Inova Diagnostics). The cut-off to be classified as ARA negative was < 20 U/mL.
Outcomes. In these posthoc analyses, we analyzed the rate of decline in FVC (mL/yr) over 52 weeks in patients who were positive for ACA at baseline; patients who were positive for ARA at baseline; patients who were negative for ATA, ACA, and ARA at baseline; and in subgroups based on ACA and ARA status at baseline. We also analyzed the proportions of patients who experienced an absolute decline in FVC > 5% predicted at week 52 in patients who were negative for ATA, ACA, and ARA at baseline and in subgroups by ACA and ARA status.
Statistical approach. In patients who were negative for ATA, ACA, and ARA at baseline, the rate of decline in FVC (mL/yr) over 52 weeks was analyzed using a random coefficient regression model with a fixed categorical effect of sex; fixed continuous effects of time, baseline FVC (mL), age, and height; and including treatment-by-time and baseline-by-time interaction terms. In patients who were positive for ACA or positive for ARA at baseline, the rate of decline in FVC (mL/yr) over 52 weeks was analyzed using a random coefficient regression model (with random slopes and intercepts) with fixed categorical effects of sex and treatment; fixed continuous effects of baseline FVC (mL), age, and height; and including treatment-by-time and baseline-by-time interaction terms. We also used models that included ATA status (positive vs negative) or MMF use at baseline as an additional fixed categorical effect.
In subgroups by ACA or ARA status (positive vs negative) at baseline, the rate of decline in FVC (mL/yr) was analyzed using a random coefficient regression model, as described above, but without the fixed categorical effect of treatment and including baseline-by-time, treatment-by-subgroup, and treatment-by-subgroup-by-time interaction terms instead. We also used models that included ATA status (positive vs negative) or MMF use (yes or no) at baseline as an additional fixed categorical effect. Exploratory interaction P values were calculated to assess potential heterogeneity in the effect of nintedanib vs placebo between the subgroups by ACA or ARA status. These analyses were based on all measurements taken over 52 weeks, including those from patients who discontinued trial medication and in patients who received ≥ 1 dose of trial medication and had a baseline and ≥ 1 postbaseline FVC measurement.
Among patients who were negative for ATA, ACA, and ARA, and in the overall trial population, the proportions of patients with an absolute decline in FVC > 5% predicted at week 52 were compared between treatment groups using a Cochran-Mantel-Haenszel test. In subgroups by ACA or ARA status at baseline, this outcome was analyzed using a logistic regression model with terms for treatment, ATA status, subgroup, and treatment-by-subgroup interaction, and odds ratios were estimated for the effect of treatment within each subgroup. These analyses were conducted in patients who received ≥ 1 dose of trial medication. Missing values were imputed using a worst value carried forward approach.
RESULTS
Patient characteristics. Of the 576 patients enrolled in the SENSCIS trial, of whom 350 (60.8%) were ATA positive, 549 and 528 underwent testing for ACA and ARA, respectively. Of these patients, 32 (5.8%) were ACA positive and 98 (18.6%) were ARA positive. Compared with patients who were ARA positive, patients who were ACA positive had a lower mean modified Rodnan skin score, higher mean FVC% predicted, and a lower percentage were male, had the diffuse cutaneous subtype of SSc, or were receiving MMF (Table 1). Compared with the overall trial population, among patients who were negative for ACA, ARA, and ATA, a lower percentage were female or had diffuse cutaneous SSc (Table 1).
Baseline characteristics by ACA, ARA, ATA status in the SENSCIS trial.
Rate of decline in FVC over 52 weeks. The rate of decline in FVC (mL/yr) over 52 weeks was numerically greater in patients who were negative for ATA, ACA, and ARA compared with the overall trial population in the placebo group (−115.6 mL/yr vs −93.3 mL/yr) and in the nintedanib group (−91.8 mL/yr vs −52.4 mL/yr; Figure 1). The effect of nintedanib vs placebo in reducing the rate of decline in FVC (mL/yr) over 52 weeks was numerically smaller in patients who were negative for ATA, ACA, and ARA (difference: 23.8 mL/yr [95% CI −74.4 to 122.1]) than in the overall trial population (difference: 41.0 mL/yr [95% CI 2.9-79.0]; Figure 1).
Rate of decline in FVC (mL/yr) over 52 weeks in patients who were ACA positive, ARA positive, or ATA, ACA, and ARA negative at baseline in the SENSCIS trial. * Based on random coefficient regression with fixed categorical effects of treatment, ATA status, and sex; fixed continuous effects of time, baseline FVC (mL), age, and height; and including treatment-by-time and baseline-by-time interaction terms. † Based on random coefficient regression with fixed categorical effects of treatment and sex; fixed continuous effects of baseline FVC (mL), age, and height; and including treatment-by-time and baseline-by-time interaction terms. ‡ Based on random coefficient regression with fixed categorical effect of sex; fixed continuous effects of time, baseline FVC (mL), age, and height; and including treatment-by-time and baseline-by-time interaction terms. ACA: anticentromere antibody; ARA: anti-RNA polymerase III antibody; ATA: antitopoisomerase I antibody; FVC: forced vital capacity; SE: standard error.
Among patients who were ACA positive or ARA positive, the rate of decline in FVC (mL/yr) over 52 weeks in the placebo arm (−31.2 mL/yr and −64.7 mL/yr, respectively) was numerically lower than in the overall SENSCIS trial population (−93.3 mL/yr) and in patients who were negative for ATA, ACA, and ARA (−115.6 mL/yr; Figure 1). Results were similar in models that adjusted for ATA status or MMF use at baseline (Supplementary Table S1, available with the online version of this article).
In analysis of subgroups by ACA or ARA status at baseline, in the placebo arm, the adjusted rate (standard error [SE]) of decline in FVC (mL/yr) over 52 weeks was numerically lower in patients who were ACA positive than ACA negative (−61.6 [54.5] mL/yr vs −94.8 [14.3] mL/yr) and numerically lower in those who were ARA positive than ARA negative (−62.7 [31.6] mL/yr vs −99.1 [15.9] mL/yr; Supplementary Table S2, available with the online version of this article). In the nintedanib arm, the adjusted rate (SE) of change in FVC (mL/yr) over 52 weeks was 35.1 (62.1) mL/yr and −60.1 (14.3) mL/yr in patients who were ACA positive and negative, respectively, and 5.8 (36.1) mL/yr and −72.2 (15.6) mL/yr in patients who were ARA positive and negative, respectively (Supplementary Table S2). The effect of nintedanib vs placebo on reducing the rate of decline in FVC over 52 weeks was numerically more pronounced in patients who were ACA positive than ACA negative (difference: 96.6 mL/yr [95% CI −65.7 to 259.0] vs 34.8 mL/yr [95% CI −5.1 to 74.6]) and in patients who were ARA positive than ARA negative (difference: 68.5 mL/yr [95% CI −25.9 to 162.8] vs 26.9 mL/yr [95% CI −16.8 to 70.7]), but the interaction P values (0.47 and 0.43, respectively) did not indicate heterogeneity in the effect of nintedanib between these subgroups. Results were similar in models that adjusted for ATA status or MMF use at baseline (Supplementary Table S2).
Absolute decline in FVC > 5% predicted at week 52. Among patients who were negative for ATA, ACA, and ARA, the proportion of patients with an absolute decline in FVC > 5% predicted at week 52 (35.8% and 35% in the nintedanib and placebo groups, respectively) was numerically greater than in the overall SENSCIS trial population (20.6% and 28.5% in the nintedanib and placebo groups, respectively; Figure 2).
Proportions of patients with an absolute decline in FVC > 5% predicted at week 52 in patients negative for ATA, ACA, and ARA at baseline and in the overall SENSCIS trial population. ACA: anticentromere antibody; ARA: anti-RNA polymerase III antibody; ATA: antitopoisomerase I antibody; FVC: forced vital capacity; OR: odds ratio.
The percentage of patients with an absolute decline in FVC > 5% predicted at week 52 was numerically lower in patients who were ACA positive than ACA negative in the placebo group (16.7% vs 29.2%) and in the nintedanib group (7.1% vs 21.7%; Supplementary Figure S1A, available with the online version of this article). The proportion of patients with this outcome was numerically lower in patients who were ARA positive than ARA negative in the nintedanib group (11.1% vs 23.4%), but similar in patients who were ARA positive and ARA negative in the placebo group (28.3% vs 28%, respectively, Supplementary Figure S1B). Interaction P values did not indicate heterogeneity in the effect of nintedanib between these subgroups (Supplementary Figure S1A-B).
DISCUSSION
With close to 600 participants, the SENSCIS trial is the largest randomized controlled trial conducted in patients with SSc-ILD, providing an unparalleled opportunity to understand the course of SSc-ILD within patient subsets. The present posthoc analyses investigated FVC outcomes in this trial based on serological profiles, including in patients who were negative for 3 of the most commonly tested SSc-specific autoantibodies (ACA, ARA, and ATA). We found that patients randomized to placebo or nintedanib who were negative for ACA, ARA, and ATA experienced a numerically greater rate of decline in FVC over 52 weeks compared with the overall trial population. However, patients randomized to placebo who were ACA or ARA positive still experienced a decline in FVC over 52 weeks. The effect of nintedanib on slowing FVC decline was numerically more pronounced in patients who were ACA or ARA positive than in patients who were negative for these autoantibodies. Results were similar between analyses that did and did not adjust for MMF use at baseline.
Among 91 patients with SSc-ILD in 1 prospective study, the presence of ACA or ARA was not significantly associated with change in FVC% predicted after 1 year.5 In a retrospective analysis of data from patients with SSc-ILD at a single center, neither ACA nor ARA status was significantly associated with change in FVC% predicted per year, although, as in the present study, the decline in FVC was numerically smaller in patients who were ACA positive vs negative or ARA positive vs negative.6 In the Scleroderma Lung Study II, which compared MMF to cyclophosphamide for the treatment of SSc-ILD,12 the presence of ACA or ARA was not associated with FVC course, but the numbers of patients who were ACA or ARA positive were small.13 Little is known about the course of SSc-ILD in patients who do not possess SSc-specific autoantibodies. Our finding that patients in the SENSCIS trial who were negative for ACA, ARA, and ATA experienced a numerically greater rate of decline in FVC than the overall trial population is noteworthy, as such patients may not receive the same degree of close monitoring and/or aggressive therapy as patients who possess ATA, for example. Such “triple negative” patients may also experience delays in diagnosis of SSc, particularly those with limited cutaneous disease, prolonging the time to initiation of therapy. The present results suggest that patients who are negative for ACA, ARA, and ATA still require treatment and close monitoring for SSc-ILD progression.
Interestingly, patients in the SENSCIS trial who were positive for ACA or ARA had a numerically greater response to nintedanib than patients who were negative for ACA or ARA. Patients in the nintedanib group who were positive for ACA or ARA actually experienced an improvement in their FVC over 52 weeks, most notably in the ACA positive group. Although the interaction P values did not indicate heterogeneity in the effect of nintedanib between these subgroups, the analyses may have been underpowered to detect a difference.
Our study has important limitations. First, there was a relatively small number of patients who were ACA or ARA positive in the SENSCIS trial. Second, given that patients needed to have at least 10% radiological fibrosis at study entry, it is unclear whether the present findings are generalizable to patients with a lower burden of fibrosis. An additional limitation is the lack of confirmatory central assessment of ATA status in patients who underwent ATA testing in a local laboratory. Finally, although ACA, ATA, and ARA are the most widely tested SSc-specific autoantibodies, other autoantibodies known to affect SSc-ILD progression (eg, Th/To, U3RNP, and Ro52) were not evaluated in the present study.
In conclusion, these analyses of data from the SENSCIS trial demonstrate that patients who were negative for ACA, ATA, and ARA were at heightened risk for progression of SSc-ILD, but that patients with SSc-ILD who possessed ACA or ARA were still at risk for progression of SSc-ILD. Future studies are needed to determine whether the presence or absence of specific autoantibodies portends an improved response to therapies, as this may facilitate the application of personalized medicine to the treatment of SSc-ILD.
ACKNOWLEDGMENT
We thank the patients who participated in the SENSCIS trial. The SENSCIS trial was supported by Boehringer Ingelheim. The authors did not receive payment for development of this manuscript. Writing assistance was provided by Elizabeth Ng and Wendy Morris of Fleishman-Hillard, London, UK, which was contracted and funded by Boehringer Ingelheim. Boehringer Ingelheim was given the opportunity to review the manuscript for medical and scientific accuracy as well as intellectual property considerations.
Footnotes
CONTRIBUTIONS
ERV: conceptualization, visualization, writing – original draft. SA: investigation, writing – review & editing. CPD: investigation, writing – review & editing. RS: formal analysis, methodology, visualization, writing – review & editing. SS: visualization, writing – review & editing. MA: methodology, supervision, visualization, writing – review & editing. EJB: conceptualization, visualization, writing – review & editing.
FUNDING
The SENSCIS trial was supported by Boehringer Ingelheim.
COMPETING INTERESTS
ERV reports research support from Boehringer Ingelheim, GSK, Horizon, Kadmon, National Heart, Lung, and Blood Institute, and Prometheus; consulting fees from AbbVie, Boehringer Ingelheim, and GSK; and speaker fees from Boehringer Ingelheim. SA reports research support from Boehringer Ingelheim, Janssen, and aTyr; and consulting fees from AstraZeneca, aTyr, AbbVie, Merck, Mitsubishi Tanabe, Takeda, Boehringer Ingelheim, CSL Behring, and TeneoFour. CPD reports research support from AbbVie, Arxx Therapeutics, CSL Behring, GSK, Horizon, and Servier; consulting fees from AbbVie, Acceleron, Arxx Therapeutics, Bayer, Boehringer Ingelheim, Certa, Corbus, CSL Behring, Galapagos, GSK, Horizon, Inventiva, Janssen, Lilly, Novartis, Roche, Sanofi-Aventis, and Zurabio; and speaker fees from Boehringer Ingelheim, Corbus, and Janssen. RS is an employee of mainanalytics GmbH, Sulzbach (Taunus), Germany, which was contracted by Boehringer Ingelheim to perform analyses presented in this paper. SS and MA are employees of Boehringer Ingelheim. EJB reports research support from Boehringer Ingelheim, BMS, and aTyr; and consulting fees from Boehringer Ingelheim and Cabaletta.
ETHICS AND PATIENT CONSENT
The SENSCIS trial was performed in 32 countries and was approved by an independent ethics committee or institutional review board at every site. All patients provided written informed consent before trial entry. The trial was conducted in accordance with the trial protocol, the principles of the Declaration of Helsinki, and the Harmonized Tripartite Guideline for Good Clinical Practice from the International Conference on Harmonization and was approved by local authorities.
DATA AVAILABILITY
To ensure independent interpretation of clinical study results and enable authors to fulfill their role and obligations under the International Committee of Medical Journal Editors criteria, Boehringer Ingelheim grants all external authors access to relevant clinical study data. In adherence with the Boehringer Ingelheim policy on Transparency and Publication of Clinical Study Data, scientific and medical researchers can request access to clinical study data after publication of the primary manuscript in a peer-reviewed journal, regulatory activities are complete, and other criteria are met. Researchers should use https://vivli.org/ to request access to study data and visit https://www.mystudywindow.com/msw/ datasharing for further information.
- Accepted for publication March 25, 2025.
- Copyright © 2025 by the Journal of Rheumatology
This is an Open Access article, which permits use, distribution, and reproduction, without modification, provided the original article is correctly cited and is not used for commercial purposes.
REFERENCES
SUPPLEMENTARY DATA
Supplementary material accompanies the online version of this article.
