You are here

  • Home
  • Archive
  • Volume 78, Issue 12
  • Phenotyping by persistent inflammation in systemic sclerosis associated interstitial lung disease: a EUSTAR database analysis

Article Text

Article menu
Interstitial lung disease
Original research
Phenotyping by persistent inflammation in systemic sclerosis associated interstitial lung disease: a EUSTAR database analysis
  1. Sabina Guler1,2,
  2. Adela-Cristina Sarbu3,
  3. Odile Stalder4,
  4. Yannick Allanore5,
  5. Vera Bernardino6,
  6. Joerg Distler7,
  7. Armando Gabrielli8,
  8. Anna-Maria Hoffmann-Vold9,
  9. Marco Matucci-Cerinic10,11,
  10. Ulf Müller-Ladner12,
  11. Vera Ortiz-Santamaria13,
  12. Simona Rednic14,
  13. Valeria Riccieri15,
  14. Vanessa Smith16,
  15. Susanne Ullman17,
  16. Ulrich A Walker18,
  17. Thomas K Geiser1,2,
  18. Oliver Distler19,
  19. Britta Maurer2,3,
  20. Florian Kollert3
  1. 1 Department of Pulmonary Medicine, Inselspital University Hospital Bern, Bern, Switzerland
  2. 2 Department of BioMedical Research, University of Bern, Bern, Switzerland
  3. 3 Department of Rheumatology and Immunology, Inselspital University Hospital Bern, Bern, Switzerland
  4. 4 CTU Bern, University of Bern, Bern, Switzerland
  5. 5 Department of Rheumatology, Cochin Hospital, Paris, France
  6. 6 Internal Medicine Department 7.2, Hospital Curry Cabral, Centro Hospitalar Universitário Lisboa Central, Lisboa, Portugal
  7. 7 Department of Rheumatology and Hiller Research Center, University Hospital Düsseldorf, Medical Faculty, Heinrich Heine University, Düsseldorf, Germany
  8. 8 Fondazione di Medicina Molecolare e Terapia Cellulare, Università Politecnica delle Marche, Ancona, Italy
  9. 9 Rheumatology, Oslo University Hospital Ullevål, Oslo, Norway
  10. 10 Department of Experimental and Clinical Medicine, University of Florence, Firenze, Italy
  11. 11 Unit of Immunology, Rheumatology, Allergy and Rare diseases (UnIRAR), San Raffaele Hospital, Milano, Italy
  12. 12 Department of Rheumatology and Clinical Immunology, University of Giessen, Campus Kerckhoff, Bad Nauheim, Germany
  13. 13 Unidad de Enfermedades Sistémicas, Reumatología, Hospital General de Granollers, Granollers, Spain
  14. 14 Department of Rheumatology, Iuliu Hatieganu University of Medicine and Pharmacy, Cluj-Napoca, Romania
  15. 15 Clinical Medicine and Therapy, Sapienza University of Rome, Roma, Italy
  16. 16 Rheumatology, Ghent University Hospital, Gent, Belgium
  17. 17 Department of Dermatology, Copenhagen University Hospital, Bispebjerg, Denmark
  18. 18 Department of Rheumatology, University Hospital Basel, Basel, Switzerland
  19. 19 Department of Rheumatology, University Hospital Zurich, Zurich, Switzerland
  1. Correspondence to Dr Sabina Guler, Department of Pulmonary Medicine, Inselspital University Hospital Bern, Bern, Bern, Switzerland; Sabina.Guler{at}insel.ch

Abstract

Background Systemic sclerosis (SSc) is a heterogeneous disease with frequently associated interstitial lung disease (SSc-ILD). We aimed to determine the prognostic potential of phenotyping patients with SSc and SSc-ILD by inflammation and to describe disease trajectories stratified by inflammation and immunosuppressive treatment.

Methods Patients from the European Scleroderma Trials and Research (EUSTAR) group cohort were allocated to persistent inflammatory, intermediate and non-inflammatory phenotypes if C-reactive protein (CRP) levels were ≥5 mg/L at ≥80%, at 20–80% and at <20% of visits, respectively. Cox regression models were used to analyse mortality risk and mixed effect models to describe trajectories of FVC and diffusing capacity for carbon monoxide (DLCO) %-predicted stratified by inflammation and immunosuppressive treatment.

Results 2971 patients with SSc and 1171 patients with SSc-ILD had at least three CRP measurements available. Patients with SSc-ILD with a persistent inflammatory phenotype had a 6.7 times higher risk of mortality within 5 years compared with those with a persistent non-inflammatory phenotype (95% CI 3 to 15). In the inflammatory phenotype, FVC %-predicted was declining without (−1.11 (95% CI −2.14 to −0.08)/year), but stable with immunosuppressive treatment (−0.00 (95% CI −0.92 to 0.92)/year). In the non-inflammatory phenotype, patients with and without immunosuppressive treatment had a significant decline in FVC %-predicted, which was more pronounced in those with immunosuppressive treatment (−1.26 (95% CI −1.87 to −0.64) and −0.84 (95% CI −1.35 to −0.33)/year, respectively).

Conclusions Phenotyping by persistent inflammation provides valuable prognostic information, independent of demographics, disease duration, cutaneous subtype, treatment and SSc-ILD severity. The findings from this study support early immunosuppressive treatment in patients with SSc-ILD with persistent inflammation.

  • Rheumatoid lung disease
  • Systemic disease and lungs
  • Interstitial Fibrosis

Data availability statement

Data are available upon reasonable request. Requests to access the data analysed in this project can be addressed to the EUSTAR board.

https://doi.org/10.1136/thorax-2023-220541

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    What is already known on this topic

    • Preliminary studies in patients with systemic sclerosis associated interstitial lung disease (SSc-ILD) suggest that C-reactive protein blood levels might be prognostically important.

    • Several randomised controlled trials demonstrate efficacy of immunosuppressive and antifibrotic drugs in the heterogeneous SSc-ILD population, however, strategies on phenotyping for treatment decisions are lacking.

    What this study adds

    • This study shows that persistent inflammation is an important risk factor for early mortality in SSc-ILD, independent from demographics, time from disease onset, cutaneous phenotype, baseline pulmonary function and immunosuppressive treatment.

    • This study also describes patients with SSc-ILD with an inflammatory phenotype to show stable FVC trajectories with immunosuppression, whereas those with a non-inflammatory phenotype have declining FVC when treated with conventional immunosuppressive drugs.

    How this study might affect research, practice or policy

    • Phenotyping by persistent inflammation is a useful concept for clinical practice and might facilitate risk stratification and potentially treatment decisions in SSc-ILD.

    Introduction

    Systemic sclerosis (SSc) is a heterogeneous disease characterised by vasculopathy, fibrosis and autoimmunity. Affected individuals suffer from severe organ manifestations which impact on patients’ quality of life and survival.1 Interstitial lung disease (ILD) is a frequent and severe SSc complication and approximately 40–50% of patients suffer from a progressive decline in pulmonary function, leading to a loss of physical functionality and early mortality.2

    Management of SSc remains challenging, and several immunosuppressive drugs are used to treat organ manifestations with several randomised controlled trials demonstrating efficacy in patients with SSc-ILD.3–6 Cyclophosphamide, mycophenolate mofetil, azathioprine and methotrexate are frequently used,3 4 7 and autologous stem-cell transplantation and biologicals are emerging treatment options in selected patients.5 6 8–10 Furthermore, nintedanib has been shown to reduce lung function decline and has recently been approved for the treatment of SSc-ILD.11 12

    The availability of immunosuppressive and antifibrotic SSc-ILD treatments emphasises the need for biomarkers to estimate response to therapy and facilitate treatment decisions in the individual patient. It is likely that not all patients with SSc benefit from immunosuppressive therapies, and predictors of treatment response are urgently needed.5 13 14 The concept of phenotyping patients with SSc by serum C-reactive protein (CRP) levels and corresponding inflammation has been suggested previously. Elevated serum CRP levels have been observed more frequently in the diffuse cutaneous subtype and have been associated with higher disease activity, poorer pulmonary function and shorter survival.15–17 To date, it is unclear whether consideration of multiple CRP values over time is more informative than single CRP values15 and whether patients with inflammatory SSc respond better to immunosuppressive therapy than patients with a non-inflammatory phenotype.

    We hypothesised that patients with SSc with persistent inflammation are not only characterised by a more severe phenotype and higher mortality but also respond better to immunosuppressive therapy than patients with a non-inflammatory phenotype. To investigate this, we identified patients with persistent inflammation and assessed phenotype, overall survival and disease course with and without immunosuppressive treatment in the European Scleroderma Trials and Research group (EUSTAR) cohort.18

    Methods

    Study design and patient population

    This multicentre cohort study analysed data from patients with SSc included in the prospective EUSTAR database. The EUSTAR database is an international registry, established in 2004, which follows patients with SSc annually.18 19 At the time of data extraction, 19 115 patients had been included in EUSTAR. Inclusion criteria for this study were fulfilling the ACR/EULAR classification criteria for SSc,20 and availability of CRP measurement at baseline (defined as first visit documented in the EUSTAR database) and at least two follow-up visits. EUSTAR demands entry of regular and not high-sensitivity CRP measurements. Patients with reports of infections and those treated with tocilizumab were excluded due to invalidity of CRP values for inflammatory phenotyping. Presence of SSc-ILD was considered in patients who were clinically labelled as having pulmonary fibrosis and/or had indication of radiological ILD characteristics on chest CT scans, including groundglass opacities, reticulations, traction bronchiectasis and/or honeycombing.

    Definition of the inflammatory phenotypes

    For the persistent inflammatory phenotype, ≥3 CRP values from consecutive visits were included in the evaluation. Patients were stratified as persistent inflammatory, if CRP values were ≥5 mg/L on ≥80% of visits, intermediate, if CRP≥5 mg/L at 20–80% of visits and persistent non-inflammatory if CRP≥5 mg/L at <20% of visits. This classification has been established in a previous study and has shown to be resistant to treatment with immunosuppressive drugs.15

    For a sensitivity analysis, the baseline inflammatory phenotype was defined by the first available (baseline) CRP value. Patients were stratified as baseline inflammatory versus non-inflammatory if CRP values were ≥5 mg/L and<5 mg/L at baseline, respectively.

    Definition of treatment groups

    Patients were stratified to the treatment group if they had received any of the following immunosuppressive treatments within 5 years from baseline: cyclophosphamide, mycophenolate mofetil, methotrexate, azathioprine, rituximab or prednisone >10 mg per day at any visit. The no treatment group consisted of the patients who had not received any immunosuppressive drugs.

    Outcomes

    Overall survival of the patients with inflammatory phenotypes was estimated by analysing time from the first available CRP measurement (baseline) to all-cause death or censoring within 5 years from baseline. Pulmonary function tests were performed in the individual centres using established protocols.

    Statistical analysis

    Patients’ characteristics were described as number (percentage) or median (IQR). Kaplan-Meier curves were used to illustrate survival function for the inflammatory phenotypes. To examine the association between the inflammatory phenotypes and mortality, we used Cox proportional hazards models to estimate HR and its 95% CI. Models were adjusted for prespecified confounders with conceptual importance. Age, sex, ever smoking, years from onset of first non-Raynaud manifestation, diffuse vs limited cutaneous disease and immunosuppressive treatment were the confounders for the SSc-ILD subgroup, with addition of ILD for the entire SSc cohort.15–17 21–23 The proportional hazards assumption was verified using log-log plots of survival.

    In patients with SSc-ILD, FVC %-predicted and diffusing capacity for carbon monoxide (DLCO) %-predicted mean trajectories with 95% CI were estimated for the persistent inflammatory phenotypes using mixed effect models with random intercepts accounting for repeated measurements per patient. An independent variance–covariance structure of the random effects was used and mixed effects models were adjusted for the respective baseline values of FVC and DLCO %-predicted. For the adjusted analyses, the following prespecified confounders were included in the models: age, sex, ever smoking, years from onset of first non-Raynaud manifestation, diffuse versus limited cutaneous disease and immunosuppressive treatments. Multiple imputation was performed using chained equations to impute covariates’ missing values. Imputation models were based on all other covariates as well as the baseline inflammatory phenotype. In total, 50 imputed data sets were generated, which were analysed as described above using Rubin’s rules to combine results across data sets.24 A p<0.05 indicated statistical significance for all comparisons. Data were analysed using Stata V.17.0 (Stata Corporation).

    Results

    Study overview and patient characteristics by inflammatory phenotype

    Of the 19 115 patients in the registry, 8852 with available CRP measurements were included in the study, with follow-up visits available in 5776. For stratification by persistent inflammation, at least three CRP values were needed and 2971 patients with SSc and 1171 with SSc-ILD were considered for these analyses (online supplemental figure S1, table 1). CT scans were available in 810 (69%) of the patients with SSc-ILD, FVC %-predicted in 1131 (97%) and DLCO %-predicted in 1085 (93%).

    Supplemental material

    View this table:
    Table 1

    Patient characteristics stratified by persistent inflammatory phenotypes in all SSc and in patients with SSc-ILD

    Of the entire SSc cohort, 424 patients (14%) had a persistent inflammatory, 1499 (50%) an intermediate and 1048 (35%) a persistent non-inflammatory phenotype. Patients with a persistent inflammatory phenotype were less frequently women (79% vs 88%) and white (74% vs 92%), more frequently smokers (32% vs 26%) and presented more often with a diffuse cutaneous phenotype (34% vs 22%) compared with patients with a persistent non-inflammatory phenotype. The median time from onset of first non-Raynaud’s symptoms was 1 year shorter in persistent inflammatory compared with non-inflammatory patients (median 7.4 vs 8.4 years). The persistent inflammatory phenotype was also associated with more severe skin disease (median (IQR) modified Rodnan Skin Score 6 (2–13) and 4 (1–9) in inflammatory and non-inflammatory patients, respectively). Joint synovitis and digital ulcers were more frequently present in the persistent inflammatory phenotype (34% vs 31% and 43% vs 40%, respectively). Autoantibodies similarly showed a distinct pattern in the inflammatory phenotypes with anti-topoisomerase positivity being more common and anti-centromere positivity less common in persistent inflammatory compared with non-inflammatory patients (36% vs 26% and 29% vs 36%, respectively). Pulmonary hypertension was reported in 11% of the persistent inflammatory and 7% of the persistent non-inflammatory SSc patients. In the entire SSc cohort, persistent inflammatory patients were more often treated with immunosuppressive drugs compared with non-inflammatory patients (any immunosuppression in 44% vs 33%).

    Compared with patients with persistent non-inflammatory SSc-ILD, patients with persistent inflammatory SSc-ILD had poorer lung function (median (IQR) FVC 79 (67–95) vs 90 (79–105) %-predicted and DLCO 55 (40–68) vs 64 (51–75) %-predicted, respectively) and shorter 6 min walk distances (median (IQR) 470 (380–545) vs 510 (435–564) metres). Furthermore, CT scans showed more frequent reticulations and traction bronchiectasis (34% vs 24% and 25% vs 13%). Overall, immunosuppression was more common in the SSc-ILD subgroup compared with the entire SSc population, and among patients with SSc-ILD, those with persistent inflammation were more often treated with immunosuppression compared with those with a non-inflammatory phenotype (54% vs 39%) (table 1).

    Phenotyping by persistent versus baseline inflammation for mortality risk stratification

    Over a median (IQR) follow-up time of 3.5 (2.4–5) years, 373 of the 5776 patients (6.4 %) with SSc and 207 of the 2100 patients with SSc-ILD (9.9 %) died.

    Phenotyping by persistent inflammation showed that the intermediate phenotype had double the risk of mortality compared with the persistent non-inflammatory phenotype with HR 1.99 (95% CI 1.10 to 3.61, p<0.001) in the entire SSc population and HR 1.93 (95% CI 0.86 to 4.32, p=0.11) in patients with SSc-ILD. The persistent inflammatory phenotype was associated with a six times higher risk for mortality in the entire SSc population HR 6.45 (95% CI 3.45 to 12.05) and in the SSc-ILD subpopulation HR 6.74 (95% CI 2.95 to 15.4) compared with the persistent non-inflammatory phenotype. This association remained largely unchanged and statistically significant with adjustment for confounding factors including age, sex, disease duration, diffuse versus limited cutaneous disease, ILD (or FVC % predicted for the SSc-ILD population), and immunosuppressive treatment (figure 1A,B, table 2).

    Figure 1
    Figure 1

    Survival by persistent (A), (B) and baseline (C), (D) inflammatory phenotype in all SSc (A), (C) and SSc-ILD (B), (D) patients. SSc, systemic sclerosis; ILD, interstitial lung disease.

    View this table:
    Table 2

    Mortality risk within 5 years from baseline by persistent inflammatory and baseline inflammatory phenotypes in the entire SSc population and in the subgroup with SSc-ILD

    Phenotyping by the baseline inflammatory phenotype (baseline CRP ≥5 mg/L) showed that patients with SSc with a baseline inflammatory phenotype had a 2.6 times (95% CI 2.11 to 3.18) higher risk of mortality compared with those with a non-inflammatory phenotype. Similarly, in the SSc-ILD subgroup, the baseline inflammatory phenotype showed a significant association with mortality risk HR 2.54 (95% CI 1.93 to 3.34). These findings remained largely unchanged and statistically significant in the adjusted analyses (figure 1C,D, table 2).

    Trajectories of pulmonary function stratified by persistent inflammatory phenotype in SSc-ILD

    The median (IQR) number of FVC %-predicted and DLCO %-predicted follow-up measurements were 4 (3–5) and 3 (3–4) with median (IQR) time intervals of 372 (279–481) and 378 (289–499) days between measurements, respectively.

    Over 5 years from baseline, the persistent inflammatory phenotype had a non-significant annual decline in FVC %-predicted (−0.51 (95% CI −1.19 to 0.18, p=0.148)), whereas average patients with an intermediate phenotype had significant (−0.79 (95% CI −1.14 to −0.44), p<0.001) and patients with a persistent non-inflammatory phenotype larger FVC declines (−1.01 (95% CI −1.40 to −0.62, p<0.001, respectively)).On average, the persistent inflammatory phenotype had a significant annual decline in DLCO %-predicted (−1.06 (95% CI −1.81 to −0.30), p=0.006) and average patients with an intermediate phenotype had also significant but smaller DLCO declines (−0.50 (95% CI −0.88 to −0.12), p=0.01). Patients with a persistent non-inflammatory phenotype did not have a significant decline in DLCO %-predicted (−0.29 (95% CI −0.72 to 0.13, p=0.177) (figure 2, online supplemental table S1).

    Figure 2
    Figure 2

    Trajectories of FVC %-predicted (A) and DLCO %-predicted (B) stratified by persistent inflammatory phenotype in patients with SSc-ILD. Average slopes in FVC and DLCO %-predicted are adjusted for baseline FVC (respectively, DLCO %-predicted), age, sex, smoking status, years from onset of first non-Raynaud manifestation, diffuse vs limited cutaneous disease and immunosuppressive treatment. The grey shaded bands illustrate 95% CIs. The number of patients is indicated for follow-up times ≥0 to 4 years and exactly 5 years, as follow-up time was limited to 5 years. DLCO, diffusing capacity for carbon monoxide; SSc-ILD, systemic sclerosis-associated interstitial lung disease.

    The estimated FVC %-predicted and DLCO %-predicted trajectories remained robust when models were adjusted for differences in age, sex, smoking, disease duration, diffuse versus limited cutaneous disease and immunosuppressive treatment as confounders (online supplemental table S1).

    Trajectories of pulmonary function in SSc-ILD patients with and without immunosuppressive treatment

    Any immunosuppressive treatment was used in 526 patients with SSc-ILD (45%) and 645 (55%) had not received any immunosuppressive treatment. Over a mean (SD) follow-up time of 3.0 (1.6) and 4.2 (1.8) years, patients with a persistent inflammatory and non-inflammatory phenotype were on immunosuppressive treatment for a mean (SD) of 2.1 (1.5) and 2.8 (1.6) years, respectively.

    Among patients with a persistent inflammatory phenotype, those without immunosuppressive treatment had a significant annual decline in FVC %-predicted of −1.11 (95% CI −2.14 to −0.08, p=0.034), whereas those with immunosuppressive treatment had stable FVC %-predicted (−0.00, 95% CI −0.92 to 0.92, p=0.996). Among patients with a persistent non-inflammatory phenotype, those without immunosuppressive treatment also had a significant annual decline in FVC %-predicted of −0.84 (95% CI −1.35 to −0.33, p=0.001), which was more pronounced in those with immunosuppressive treatment (−1.26, 95% CI −1.87 to −0.64, p=0.001). The persistent non-inflammatory patients with immunosuppressive treatment represented the group with the largest mean annual FVC decline, which was significantly larger than the FVC decline in persistent inflammatory patients with immunosuppressive treatment (interaction p=0.027). Interaction terms for immunosuppressive treatment were not statistically significant. After adjustment for age, sex, smoking, disease duration and diffuse versus limited cutaneous disease, these findings remained unchanged (figure 3, table 3).

    Figure 3
    Figure 3

    Trajectories of FVC %-predicted (A), (B) and DLCO %-predicted (C), (D) by immunosuppressive treatment in non-inflammatory (A), (C) and inflammatory (B), (D) SSc-ILD. Average slopes in FVC and DLCO %-predicted are adjusted for baseline FVC (respectively DLCO %-predicted), age, sex, smoking status, years from onset of first non-Raynaud manifestation, and diffuse vs limited cutaneous disease. The grey shaded bands illustrate 95% CIs. DLCO, diffusing capacity for carbon monoxide; SSc-ILD, systemic sclerosis-associated interstitial lung disease.

    View this table:
    Table 3

    Annual change in pulmonary function by immunosuppressive treatment and persistent inflammatory phenotypes in SSc-ILD

    Among patients with a persistent inflammatory phenotype, those without immunosuppressive treatment had a non-significant annual decline in DLCO %-predicted of −0.96 (95% CI −2.11 to 0.19, p=0.10), whereas those with immunosuppressive treatment had a significant decline in DLCO %-predicted (−1.11, 95% CI −2.12 to −0.11, p=0.029). Patients with a persistent non-inflammatory phenotype had stable DLCO %-predicted with (−0.42, 95% CI −1.08 to 0.24, p=0.21) and without (−0.20, 95% CI −0.76 to 0.35, p=0.48) immunosuppressive treatment. With adjustment for age, sex, smoking, disease duration and diffuse versus limited cutaneous disease, these findings remained unchanged (figure 3, table 3).

    Discussion

    Patients with SSc with a persistent inflammatory phenotype characterised by persistently elevated CRP levels showed features of more severe ILD and had a six times higher risk of mortality within 5 years from baseline. With stratification by immunosuppressive treatment, we observed the largest FVC decline in patients with non-inflammatory SSc-ILD who were treated with immunosuppressives drugs, whereas the inflammatory phenotype showed stable FVC %-predicted when treated with immunosuppressive drugs. These findings should stimulate discussions whether traditionally used immunosuppressive therapies are equally beneficial in patients with persistent inflammatory and non-inflammatory SSc-ILD. This simple approach to inflammatory phenotyping demonstrates the benefit of repeatedly measuring CRP values for risk stratification and potentially also for treatment decisions in SSc-ILD.

    This is the largest cohort study ever to show that SSc and specifically patients with SSc-ILD with a persistent inflammatory phenotype had a substantially higher risk of early mortality compared with those with a persistent non-inflammatory phenotype—independent from demographics, time from disease onset, cutaneous phenotype, baseline pulmonary function and immunosuppressive treatment. We also explored stratification by a single CRP measurement at baseline but found that the risk of mortality was roughly three times higher in the persistent versus the baseline inflammatory phenotype. Consequently, stratification by persistent inflammation outperforms consideration of single CRP measurements for the estimation of mortality risk within the next 5 years. A few previous studies have shown that CRP blood levels might be prognostically important in SSc, however, mortality risk was not adjusted for confounders and no dedicated analyses for SSc-ILD were performed.16 23

    While patients in the persistent inflammatory subgroup had a higher risk of mortality and a larger decline in DLCO %-predicted compared with non-inflammatory patients, the decline in FVC %-predicted was smaller in this subgroup. There are several potential reasons for this observation. Although respiratory failure remains the primary cause of death in SSc, it is noteworthy that 80–90% of patients die of non-respiratory causes.25 26 Cardiovascular diseases and cancer contribute significantly to mortality rates,26 and there is a well-studied relationship between inflammation, cardiovascular and cancer mortality.27 In contrast to FVC, DLCO decline is less specific for ILD progression, but can decline also due to cardiovascular disease, particularly pulmonary hypertension. Some studies demonstrate that change in DLCO is the stronger predictor of long-term survival compared with change in FVC.28–30 Furthermore, analyses from the scleroderma lung studies suggest that despite the positive effect of treatment on FVC, there is no difference in long-term survival in patients who had been randomised to receive cyclophosphamide, mycophenolate or placebo.26 The latter leads us to suspect that differences in FVC % trajectories we observe between patients treated with and without immunosuppressive drugs might not have translated to overall survival.

    Furthermore, our analyses showed no impact of immunosuppressive treatment on DLCO trajectories, which might partially be explained by the more frequent pulmonary hypertension in the inflammatory compared with the non-inflammatory phenotype (11% vs 7%), which has been described previously,31 and further links DLCO trajectories to mortality risks in SSc-ILD. The retrospective design of the current study limits our ability to draw definite conclusions regarding the interplay of these factors.

    Our study has several other limitations. Inherent to observational studies, no definite inference on cause and effect can be drawn from the observed associations. For example, we observed that patients with immunosuppression were followed-up slightly more closely. We cannot completely exclude that missingness of data is not at random and that there is residual confounding in the comparison between patients with and without immunosuppression. We choose linear mixed models for our analysis because of their well-established performance in the context of missing data, variable follow-up observations per patients and variable follow-up intervals.32 The survival and repeated measures models were adjusted for prespecified confounders between the relationship of inflammation, immunosuppressive treatment and outcomes. This approach accounts for the most important reasons for medication prescription (eg, patients with more severe ILD are more likely to be treated with immunosuppressive drugs). We did not take treatment duration into account for our analyses, which differed only marginally between patients with the persistent inflammatory and non-inflammatory phenotype. However, overall residual confounding cannot be completely excluded. Pulmonary function testing is dependent on qualifications of technicians and strict adherence to established protocols. While EUSTAR sites are generally recognised as interdisciplinary centres of excellence, there may still be some degree of local variability in the performance of pulmonary function testing. Inflammatory phenotyping was based on a previous study describing a distinct, severe clinical SSc phenotype with repeatedly positive CRP values that were robust to long-term follow-up and unchanged with conventional immunosuppression.15 However, previously other CRP cut-offs have been used,16 and there is a possibility that other CRP cut-offs or CRP as a continuous parameter, with or without splines, might be valid in this and other populations. For this study, 115 patients treated with tocilizumab (2% of the study population) were excluded, due to the effect of tocilizumab on CRP levels that would have interfered with our phenotyping approach. It is unlikely that observations from these few patients would have influenced our findings, however based on a recent clinical trial,5 inclusion of these patients might have led to a more pronounced benefit from immunosuppression in the inflammatory phenotype.

    This study supports the concept of personalised management and phenotyping for monitoring and treatment decisions in SSc. The high mortality and disease progression in the inflammatory phenotype suggest close monitoring and potentially early immunosuppressive therapy in these patients. On the contrary, the largest FVC decline with immunosuppressive treatment observed in the non-inflammatory phenotype questions the effectiveness of immunosuppressive treatment in patients with SSc-ILD without signs of inflammation, particularly if extrapulmonary manifestations are mild or stable. A post hoc analysis of the scleroderma lung study I4 showed a greater treatment response in patients with severe cutaneous disease and pulmonary fibrosis.14 As the extend of skin and lung fibrosis is associated with elevated CRP levels,15–17 a reason for the more favourable response in the severe subgroup might be an underlying inflammatory phenotype. Furthermore, a recent phase 3 study investigating the efficacy of tocilizumab in patients with SSc with a set of inflammatory features showed a large FVC decline in the SSc-ILD placebo group and stabilisation in the tocilizumab group, which confirm the progressive nature and likely favourable immunosuppressive treatment response of the inflammatory SSc-ILD phenotype.5 On the contrary, antifibrotic drugs are effective for the treatment of idiopathic pulmonary fibrosis (IPF),33 other types of progressive ILDs12 and SSc-ILD.11 Since the publication of the PANTHER trial showing a deleterious effect of a treatment regimen including azathioprine in patients with IPF, immunosuppressive treatment is prohibited in IPF.34 Our findings are likely to stimulate further discussions and might encourage clinicians to carefully consider the potential benefits and harms of using immunosuppressive drugs in their patients with SSc-ILD who have minimal extrapulmonary SSc activity and show no signs of an inflammatory phenotype.

    In summary, findings from the large, prospective, multicentre EUSTAR cohort support phenotyping of patients with SSc by the persistent inflammatory phenotype based on its prognostic importance and potential impact for management decisions. While baseline CRP measurements provide some prognostic information, the impact of persistent inflammation on survival is clearly larger. The stable FVC in inflammatory but declining FVC trajectories in the patients with non-inflammatory SSc-ILD who are treated with immunosuppressive drugs might prompt clinical trials to specifically investigate the potential harm of immunosuppression in patients with a non-inflammatory SSc-ILD phenotype. Clinical trials might enrich not only for participants with inflammatory but depending on the drug target also for patients with SSc-ILD with a non-inflammatory phenotype. Overall, phenotyping by persistent inflammation might contribute to improve personalised treatment approaches in SSc-ILD.

    Data availability statement

    Data are available upon reasonable request. Requests to access the data analysed in this project can be addressed to the EUSTAR board.

    Ethics statements

    Patient consent for publication

    Ethics approval

    This study involves human participants and was approved by the Swiss Ethics Committee (BASEC Nr. 010/10) and by the corresponding ethical committees of all EUSTAR centres. Participants gave informed consent to participate in the study before taking part.

    Acknowledgments

    The authors would like to acknowledge the patients contributing their data to the EUSTAR registry. They also thank all EUSTAR centres and the Stiftung Lindenhof, Bern, Switzerland, for supporting this study.

    References

    1. ↵
      2. Elhai M ,
      3. Meune C ,
      4. Avouac J , et al
      . Trends in mortality in patients with systemic sclerosis over 40 years: a systematic review and meta-analysis of cohort studies. Rheumatology (Oxford) 2012;51:1017–26. doi:10.1093/rheumatology/ker269
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      2. Guler SA ,
      3. Winstone TA ,
      4. Murphy D , et al
      . Does systemic sclerosis-associated interstitial lung disease burn out? specific phenotypes of disease progression. Ann Am Thorac Soc 2018;15:1427–33. doi:10.1513/AnnalsATS.201806-362OC
      OpenUrlPubMed
    3. ↵
      2. Tashkin DP ,
      3. Roth MD ,
      4. Clements PJ , et al
      . Mycophenolate mofetil versus oral cyclophosphamide in scleroderma-related interstitial lung disease (SLS II): a randomised controlled, double-blind, parallel group trial. Lancet Respir Med 2016;4:708–19. doi:10.1016/S2213-2600(16)30152-7
      OpenUrl
    4. ↵
      2. Tashkin DP ,
      3. Elashoff R ,
      4. Clements PJ , et al
      . Cyclophosphamide versus placebo in scleroderma lung disease. N Engl J Med 2006;354:2655–66. doi:10.1056/NEJMoa055120
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      2. Khanna D ,
      3. Lin CJF ,
      4. Furst DE , et al
      . Tocilizumab in systemic sclerosis: a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet Respir Med 2020;8:963–74. doi:10.1016/S2213-2600(20)30318-0
      OpenUrl
    6. ↵
      2. Ebata S ,
      3. Yoshizaki A ,
      4. Oba K , et al
      . Safety and efficacy of rituximab in systemic sclerosis (DESIRES): a double-blind, investigator-initiated, randomised, placebo-controlled trial. The Lancet Rheumatology 2021;3:e489–97. doi:10.1016/S2665-9913(21)00107-7
      OpenUrl
    7. ↵
      2. Kowal-Bielecka O ,
      3. Fransen J ,
      4. Avouac J , et al
      . Update of EULAR recommendations for the treatment of systemic sclerosis. Ann Rheum Dis 2017;76:1327–39. doi:10.1136/annrheumdis-2016-209909
      OpenUrlAbstract/FREE Full Text
    8. ↵
      2. Maher TM ,
      3. Tudor VA ,
      4. Saunders P , et al
      . Rituximab versus intravenous cyclophosphamide in patients with connective tissue disease-associated interstitial lung disease in the UK (RECITAL): a double-blind, double-dummy, randomised, controlled, phase 2B trial. Lancet Respir Med 2023;11:45–54. doi:10.1016/S2213-2600(22)00359-9
      OpenUrl
    9. ↵
      2. Zhao Y ,
      3. Gao Y ,
      4. Petnak T , et al
      . Effect size of rituximab on pulmonary function in the treatment of connective-tissue disease-related interstitial lung disease: a systematic review and meta-analysis. Respir Res 2022;23. doi:10.1186/s12931-022-02082-x
    10. ↵
      2. van Laar JM ,
      3. Farge D ,
      4. Sont JK , et al
      . Autologous hematopoietic stem cell transplantation vs intravenous pulse cyclophosphamide in diffuse cutaneous systemic sclerosis: a randomized clinical trial. JAMA 2014;311:2490–8. doi:10.1001/jama.2014.6368
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      2. Distler O ,
      3. Gahlemann M ,
      4. Maher TM
      . Nintedanib for systemic sclerosis-associated interstitial lung disease. N Engl J Med 2019;381:1596–7. doi:10.1056/NEJMc1910735
      OpenUrlPubMed
    12. ↵
      2. Flaherty KR ,
      3. Wells AU ,
      4. Cottin V , et al
      . Nintedanib in progressive fibrosing interstitial lung diseases. N Engl J Med 2019;381:1718–27. doi:10.1056/NEJMoa1908681
      OpenUrlCrossRefPubMed
    13. ↵
      2. Hoyles RK ,
      3. Ellis RW ,
      4. Wellsbury J , et al
      . A multicenter, prospective, randomized, double-blind, placebo-controlled trial of corticosteroids and intravenous cyclophosphamide followed by oral azathioprine for the treatment of pulmonary fibrosis in scleroderma. Arthritis Rheum 2006;54:3962–70. doi:10.1002/art.22204
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      2. Roth MD ,
      3. Tseng C-H ,
      4. Clements PJ , et al
      . Predicting treatment outcomes and responder subsets in scleroderma-related interstitial lung disease. Arthritis Rheum 2011;63:2797–808. doi:10.1002/art.30438
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      2. Mitev A ,
      3. Christ L ,
      4. Feldmann D , et al
      . Inflammatory stays inflammatory: a subgroup of systemic sclerosis characterized by high morbidity and inflammatory resistance to cyclophosphamide. Arthritis Res Ther 2019;21. doi:10.1186/s13075-019-2057-x
    16. ↵
      2. Muangchan C ,
      3. Harding S ,
      4. Khimdas S , et al
      . Association of C-reactive protein with high disease activity in systemic sclerosis: results from the Canadian scleroderma research group. Arthritis Care Res (Hoboken) 2012;64:1405–14. doi:10.1002/acr.21716
      OpenUrl
    17. ↵
      2. Lis-Święty A ,
      3. Widuchowska M ,
      4. Brzezińska-Wcisło L , et al
      . High acute phase protein levels correlate with pulmonary and skin involvement in patients with diffuse systemic sclerosis. J Int Med Res 2018;46:1634–9. doi:10.1177/0300060518760955
      OpenUrl
    18. ↵
      2. Meier FMP ,
      3. Frommer KW ,
      4. Dinser R , et al
      . Update on the profile of the EUSTAR cohort: an analysis of the EULAR scleroderma trials and research group database. Ann Rheum Dis 2012;71:1355–60. doi:10.1136/annrheumdis-2011-200742
      OpenUrlAbstract/FREE Full Text
    19. ↵
      2. Walker UA ,
      3. Tyndall A ,
      4. Czirják L , et al
      . Clinical risk assessment of organ manifestations in systemic sclerosis: a report from the EULAR scleroderma trials and research group database. Ann Rheum Dis 2007;66:754–63. doi:10.1136/ard.2006.062901
      OpenUrlAbstract/FREE Full Text
    20. ↵
      2. van den Hoogen F ,
      3. Khanna D ,
      4. Fransen J , et al
      . Classification criteria for systemic sclerosis: an American college of rheumatology/European league against rheumatism collaborative initiative. Arthritis & Rheumatism 2013;65:2737–47. doi:10.1002/art.38098
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      2. Hoffmann-Vold A-M ,
      3. Allanore Y ,
      4. Alves M , et al
      . Progressive interstitial lung disease in patients with systemic sclerosis-associated interstitial lung disease in the EUSTAR database. Ann Rheum Dis 2021;80:219–27. doi:10.1136/annrheumdis-2020-217455
      OpenUrlAbstract/FREE Full Text
    22. ↵
      2. Elhai M ,
      3. Meune C ,
      4. Boubaya M , et al
      . Mapping and predicting mortality from systemic sclerosis. Ann Rheum Dis 2017;76:1897–905. doi:10.1136/annrheumdis-2017-211448
      OpenUrlAbstract/FREE Full Text
    23. ↵
      2. Liu X ,
      3. Mayes MD ,
      4. Pedroza C , et al
      . Does C-reactive protein predict the long-term progression of interstitial lung disease and survival in patients with early systemic sclerosis Arthritis Care & Research 2013;65:1375–80. doi:10.1002/acr.21968
      OpenUrl
    24. ↵
      2. Rubin DB
      . Multiple imputation for nonresponse in surveys. Hoboken. NJ: John Wiley & Sons, 1987. doi:10.1002/9780470316696
    25. ↵
      2. Tyndall AJ ,
      3. Bannert B ,
      4. Vonk M , et al
      . Causes and risk factors for death in systemic sclerosis: a study from the EULAR scleroderma trials and research (EUSTAR) database. Ann Rheum Dis 2010;69:1809–15. doi:10.1136/ard.2009.114264
      OpenUrlAbstract/FREE Full Text
    26. ↵
      2. Volkmann ER ,
      3. Tashkin DP ,
      4. Sim M , et al
      . Short-term progression of interstitial lung disease in systemic sclerosis predicts long-term survival in two independent clinical trial cohorts. Ann Rheum Dis 2019;78:122–30. doi:10.1136/annrheumdis-2018-213708
      OpenUrlAbstract/FREE Full Text
    27. ↵
      2. Singh-Manoux A ,
      3. Shipley MJ ,
      4. Bell JA , et al
      . Association between inflammatory biomarkers and all-cause, cardiovascular and cancer-related mortality. CMAJ 2017;189:E384–90. doi:10.1503/cmaj.160313
      OpenUrlAbstract/FREE Full Text
    28. ↵
      2. Bouros D ,
      3. Wells AU ,
      4. Nicholson AG , et al
      . Histopathologic subsets of fibrosing alveolitis in patients with systemic sclerosis and their relationship to outcome. Am J Respir Crit Care Med 2002;165:1581–6. doi:10.1164/rccm.2106012
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      2. Goh NS ,
      3. Hoyles RK ,
      4. Denton CP , et al
      . Short-term pulmonary function trends are predictive of mortality in interstitial lung disease associated with systemic sclerosis. Arthritis & Rheumatology 2017;69:1670–8. doi:10.1002/art.40130
      OpenUrl
    30. ↵
      2. Moore OA ,
      3. Proudman SM ,
      4. Goh N , et al
      . Quantifying change in pulmonary function as a prognostic marker in systemic sclerosis-related interstitial lung disease. Clin Exp Rheumatol 2015;33(4 Suppl 91):S111–6.
      OpenUrlPubMed
    31. ↵
      2. Jha M ,
      3. Wang M ,
      4. Steele R , et al
      . NT-proBNP, Hs-cTnT, and CRP predict the risk of cardiopulmonary outcomes in systemic sclerosis: findings from the Canadian scleroderma research group. J Scleroderma Relat Disord 2022;7:62–70. doi:10.1177/23971983211040608
      OpenUrl
    32. ↵
      2. Bates D ,
      3. Mächler M ,
      4. Bolker B , et al
      . Fitting linear mixed-effects models using Lme4. arXiv 1406.5823. [Preprint] 2014. doi:10.48550/arXiv.1406.5823
    33. ↵
      2. Richeldi L ,
      3. du Bois RM ,
      4. Raghu G , et al
      . Efficacy and safety of nintedanib in idiopathic pulmonary fibrosis. N Engl J Med 2014;370:2071–82. doi:10.1056/NEJMoa1402584
      OpenUrlCrossRefPubMedWeb of Science
    34. ↵
      1. Idiopathic Pulmonary Fibrosis Clinical Research Network,
      2. Raghu G ,
      3. Anstrom KJ , et al
      . Prednisone, azathioprine, and N-Acetylcysteine for pulmonary fibrosis. N Engl J Med 2012;366:1968–77. doi:10.1056/NEJMoa1113354
      OpenUrlCrossRefPubMedWeb of Science

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • SG and A-CS are joint first authors.

    • Twitter @SabinaAGuler

    • SG and A-CS contributed equally.

    • Contributors SG is the guarantor of the study and responsible for the overal content of the manuscript. SG, A-CS, BM and FK contributed to the conception and design of the study, and acquisition, analysis and interpretation of the data. OS contributed to analysis of the data. YA, VB, JD, AG, A-MH-V, MM-C, UM-L, VO-S, SR, VR, VS, SU, UAW, TKG and OD contributed to the acquisition and interpretation of the data. All authors revised the manuscript for important intellectual content and provided final approval of the version to be published. SG and A-CS contributed equally to this paper.

    • Funding This study was funded by Stiftung Lindenhof, Bern, Switzerland.

    • Competing interests SG reports grants, contracts, consulting or lecture fees from Roche, MSD, Boehringer Ingelheim, and support for this study from the Stiftung Lindenhof, Bern, Switzerland. YA reports grants, contracts, consulting or lecture fees from Medsenic, Alpine ImmnunoSciences, Boehringer, Astra-Zeneca, Galderma, Prometheus, Abbvie, Chugai, Benevolent. JHWD has received research funding from Anamar, ARXX, BMS, Bayer Pharma, Boehringer Ingelheim, Cantargia, Celgene, CSL Behring, Galapagos, GSK, Inventiva, Kiniksa, Sanofi-Aventis, RedX, UCB. JD is stock owner of 4D Science. JD has consultancy relationships with AbbVie, Active Biotech, Anamar, ARXX, AstraZeneca, Bayer Pharma, Boehringer Ingelheim, Celgene, Galapagos, GSK, Inventiva, Janssen, Novartis, Pfizer, and UCB. Scientific lead of FibroCure. AG reports grants or contracts from Janssen, Boehringer Ingelheim, Roche. A-MH-V reports grants, contracts, consulting or lecture fees from ARXX, Boehringer Ingelheim, Janssen, Medscape, Roche, Genentech, Bayer, Lilly, Merck Sharp & Dohme. MMC reports consulting or lecture fees from Sandoz, Biogen, Boehringer. SR reports grants, contracts, consulting or lecture fees from Janssen, Novartis, Boehringer Ingelheim, Abbvie, Lilly, Sandoz, Ewopharma, Pfizer, Astra Zeneca, Novartis. VR reports consulting or lecture fees from MSD, Boehringer Ingelheim. VS reports grants, contracts, consulting or lecture fees from Research Foundation Flanders, Belgian Fund for Scientific Research, Boehringer Ingelheim, Janssen-Cilag, Galapagos. TKG reports consulting or lecture fees Boehringer Ingelheim, Roche. OD reports grants, contracts, consulting or lecture fees from 4P-Pharma, Abbvie, Acceleron, Alcimed, Altavant, Amgen, AnaMar, Arxx, AstraZeneca, Baecon, Blade, Bayer, Boehringer Ingelheim, Corbus, CSL Behring, Galderma, Galapagos, Glenmark, Gossamer, iQvia, Horizon, Inventiva, Janssen, Kymera, Lupin, Medscape, Merck, Miltenyi Biotec, Mitsubishi Tanabe, Novartis, Prometheus, Redxpharma, Roivant, Sanofi and Topadur. BM reports grants contracts, consulting or lecture fees from AbbVie, Protagen, Novartis Biomedical Research, Novartis, Boehringer Ingelheim, Janssen-Cilag, GSK, Boehringer-Ingelheim, GSK, Novartis, MSD, Medtalk, Pfizer, Roche, Actelion, Mepha, MSD. FK is a shareholder of Roche, was a consultant of Actelion, BMS, Boehringer-Ingelheim, and Pfizer, had grant/research support from Gilead, Pfizer and Roche, and is employed by Roche. All other authors (A-CS, OS, VB, UM-L, VO-S, SU, UAW) declare no competing interests.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

    Linked Articles