Page 1 of 16

AMERICAN THORACIC SOCIETY

DOCUMENTS

Treatment of Systemic Sclerosis–associated Interstitial Lung

Disease: Evidence-based Recommendations
An Official American Thoracic Society Clinical Practice Guideline
Ganesh Raghu, Sydney B. Montesi, Richard M. Silver, Tanzib Hossain, Madalina Macrea, Derrick Herman,
Hayley Barnes, Ayodeji Adegunsoye, Arata Azuma, Lorinda Chung, Gregory C. Gardner, Kristin B. Highland,
Marie Hudson, Robert J. Kaner, Martin Kolb, Mary Beth Scholand, Virginia Steen, Carey C. Thomson,
Elizabeth R. Volkmann, Fredrick M. Wigley, Dee Burlile, Karen A. Kemper, Shandra L. Knight,
and Marya Ghazipura; on behalf of the American Thoracic Society Assembly on Clinical Problems

THIS OFFICIAL CLINICAL PRACTICE GUIDELINE OF THE AMERICAN THORACIC SOCIETY WAS APPROVED MAY 2023

Abstract equity were all considered in making the recommendations. This

was in accordance with the American Thoracic Society guideline
Background: Interstitial lung disease (ILD) is a significant cause development process, which is in compliance with the Institute of
of morbidity and mortality in patients with systemic sclerosis Medicine standards for trustworthy guidelines.
(SSc). To date, clinical practice guidelines regarding treatment for
patients with SSc-ILD are primarily consensus based. Results: For treatment of patients with SSc-ILD, the committee:
1) recommends the use of mycophenolate; 2) recommends
Methods: An international expert guideline committee further research into the safety and efficacy of (a) pirfenidone
composed of 24 individuals with expertise in rheumatology, SSc, and (b) the combination of pirfenidone plus mycophenolate; and
pulmonology, ILD, or methodology, and with personal 3) suggests the use of (a) cyclophosphamide, (b) rituximab, (c)
experience with SSc-ILD, discussed systematic reviews of the tocilizumab, (d) nintedanib, and (e) the combination of
published evidence assessed using the Grading of nintedanib plus mycophenolate.
Recommendations, Assessment, Development, and Evaluation
approach. Predetermined conflict-of-interest management Conclusions: The recommendations herein provide an
strategies were applied, and recommendations were made for or evidence-based clinical practice guideline for the treatment
against specific treatment interventions exclusively by the of patients with SSc-ILD and are intended to serve as the
nonconflicted panelists. The confidence in effect estimates, basis for informed and shared decision making by clinicians
importance of outcomes studied, balance of desirable and and patients.
undesirable consequences of treatment, cost, feasibility,
acceptability of the intervention, and implications for health Keywords: interstitial lung disease; systemic sclerosis; SSc-ILD

You may print one copy of this document at no charge. However, if you require more than one copy, you must place a reprint order. Domestic
reprint orders: [email protected]; international reprint orders: [email protected].
ORCID IDs: 0000-0001-7506-6643 (G.R.); 0000-0002-5724-6808 (S.B.M.); 0000-0002-2038-3278 (R.M.S.); 0000-0002-1995-7828 (T.H.);
0000-0002-5352-9587 (M.M.); 0000-0002-4209-8128 (D.H.); 0000-0002-7615-4191 (H.B.); 0000-0002-7015-9610 (A. Adegunsoye);
0000-0003-0506-9966 (A. Azuma); 0000-0003-0072-6939 (L.C.); 0000-0003-3481-8591 (G.C.G.); 0000-0001-5072-0725 (K.B.H.);
0000-0003-0527-576X (R.J.K.); 0000-0003-3837-1467 (M.K.); 0000-0001-6513-8893 (M.B.S.); 0000-0002-6333-0133 (V.S.); 0000-0003-3750-6569
(E.R.V.); 0000-0002-9941-809X (K.A.K.); 0000-0002-4404-3833 (S.L.K.); 0000-0003-4328-6822 (M.G.).

Supported by the American Thoracic Society.

An editorial, a clinical practice guideline summary for clinicians, and systematic reviews (https://www.atsjournals.org/doi/abs/10.1513/AnnalsATS.
202301-053OC, https://www.atsjournals.org/doi/abs/10.1513/AnnalsATS.202301-054OC, https://www.atsjournals.org/doi/abs/10.1513/AnnalsATS.
202301-055OC, https://www.atsjournals.org/doi/abs/10.1513/AnnalsATS.202301-056OC, https://www.atsjournals.org/doi/abs/10.1513/AnnalsATS.
202301-081OC) related to this document will be published in upcoming issues of the Annals of the American Thoracic Society.
Correspondence and requests for reprints should be addressed to Ganesh Raghu, M.D., University of Washington Medical Center-Montlake
Campus, Center for Interstitial Lung Diseases Medical Speciality Clinics, 1959 NE Pacific Avenue, 3rd Floor, Box 356175, Seattle, WA 98195.
E-mail: [email protected].
This document has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org.
Am J Respir Crit Care Med [online ahead of print] 29 Sep 2023; DOI: 10.1164/rccm.202306-1113ST.
Copyright © 2023 by the American Thoracic Society
Internet address: www:atsjournals:org

American Thoracic Society Documents

AJRCCM Articles in Press. Published September 29, 2023 as 10.1164/rccm.202306-1113ST

Copyright © 2023 by the American Thoracic Society
 Page 2 of 16

AMERICAN THORACIC SOCIETY DOCUMENTS

Contents Manuscript Preparation Question 5: Should patients with

Overview Evidence-based Recommendations SSc-ILD be treated with
Use of the Guideline for the Treatment of SSc-ILD nintedanib?
SSc-ILD Definition Question 1: Should patients with Question 6: Should patients with
Therapies Assessed and SSc-ILD be treated with SSc-ILD be treated with
Summary of Recommendations cyclophosphamide? nintedanib plus mycophenolate?
Introduction Question 2: Should patients with Question 7: Should patients with
Methods SSc-ILD be treated with SSc-ILD be treated with
Committee Composition mycophenolate? pirfenidone?
Confidentiality Agreement and Question 3: Should patients with Question 8: Should patients with
Conflict-of-Interest Management SSc-ILD be treated with SSc-ILD be treated with
Formulating Clinical Questions rituximab? pirfenidone plus mycophenolate?
Literature Search Question 4: Should patients with Guideline Considerations and
Evidence Review and SSc-ILD be treated with Future Directions
Development of Clinical tocilizumab? Conclusions
Recommendations

Overview remarks accompanying each recommendation came to a consensus that it was an important
are integral and serve to facilitate more point of consideration worth evaluation.
The purpose of this guideline is to provide accurate interpretation. They should never be When opinions and interpretation of the
clinicians with evidence-based treatment omitted when quoting or translating existence and the results of evidence differ
recommendations for patients with recommendations from this guideline. among experts of the science, the matter
interstitial lung disease (ILD) associated with needs to be resolved by consensus reached
systemic sclerosis (SSc), or SSc-ILD, based on SSc-ILD Definition after discussion of all the points among the
an analysis of the evidence available through We defined our patient population as having experts to try to come to a truth on matters,
October 2022. The recommendations are SSc-ILD if both SSc and ILD are present. which is what was done in this guideline
to be implemented within the context of SSc is defined using the former American development process for the subgroups and
individual patient values and preferences. Rheumatology Association 1980 criteria their definitions. The guideline committee
They factor in confidence in effect estimates, (1) or the 2013 American College of came to a consensus to evaluate three
importance of outcomes studied, balance of Rheumatology and European Alliance of different subgroups: initial diagnosis of SSc-
desirable and undesirable consequences of Associations for Rheumatology (formerly ILD, stable SSc-ILD, and progressive SSc-
treatment, cost, feasibility, acceptability of the the European League Against Rheumatism) ILD. Those patients in whom the treatment
intervention, and implications for health classification criteria (2). ILD is defined as the intervention was initiated at the time of the
equity. However, they should not be viewed radiologic presence of reticulation, traction initial diagnosis may respond differently
as unequivocal recommendations and do not bronchiectasis, traction bronchiolectasis, from those who had a stable disease course
propose one treatment regimen over another honeycomb cysts, ground-glass opacities or and from those with progressive disease.
or any particular sequence in therapy. air space consolidation, other interstitial lung Although the definitions of initial diagnosis
abnormalities, or any of the recognized and stable disease are relatively
Use of the Guideline patterns of interstitial pneumonias straightforward, the definition of progressive
This guideline provides the basis for rational (usual interstitial pneumonia [UIP], disease was less clear. Initial diagnosis refers
decisions in the management of SSc-ILD and is probable UIP, indeterminate for UIP, to a new SSc-ILD diagnosis before treatment
not intended to impose a standard of care. nonspecific interstitial pneumonia, initiation. Patients with known SSc-ILD not
Clinicians, patients, third-party payers, organizing pneumonia, lymphoid interstitial meeting criteria for progressive SSc-ILD were
institutional review committees, the courts, pneumonia, pleuroparenchymal fibroelasto- deemed to have stable SSc-ILD.
and/or other stakeholders should not view the sis, or unclassifiable interstitial pneumonias) To define progressive SSc-ILD, the
recommendations herein as dictates. reported in the context of SSc. committee came to a consensus to adapt
No guidelines and recommendations can After careful thought and consideration, the definition of progressive pulmonary
consider all the often compelling unique the committee felt that it was clinically fibrosis (PPF) as defined in the recently
individual clinical circumstances. Therefore, no relevant to define subgroups of the disease published 2022 American Thoracic Society
one charged with evaluating clinicians’ actions state to help understand and appreciate the (ATS) clinical practice guideline for a
should attempt to apply the recommendations potential of differential treatment response non–idiopathic pulmonary fibrosis (non-
from these guidelines by rote or in a blanket to a particular treatment intervention. The IPF) population (3). Although the criteria
fashion. Statements about the underlying committee members acknowledged that the for PPF used in the 2022 guideline were
values and preferences as well as qualifying evidence for subgroups may be limited but extrapolated from the patient population of

American Journal of Respiratory and Critical Care Medicine

AJRCCM Articles in Press. Published September 29, 2023 as 10.1164/rccm.202306-1113ST

Copyright © 2023 by the American Thoracic Society
Page 3 of 16

AMERICAN THORACIC SOCIETY DOCUMENTS

patients with IPF, the committee felt it a. Cyclophosphamide (

䊊䊊, low To improve the quality of care delivered to
appropriate to adapt the criteria to define confidence in effect estimates) patients with SSc-ILD, we provide the
progressive SSc-ILD after eliminating the b. Rituximab (
䊊䊊䊊, very low following evidence-based recommendations
timeline for disease progression. Therefore, confidence in effect estimates) to assist clinicians in the treatment of
progressive SSc-ILD was defined as c. Tocilizumab (
䊊䊊䊊, very low patients with SSc-ILD.
manifesting at least two of the following confidence in effect estimates) Other well-known manifestations of
three criteria during follow-up in patients d. Nintedanib (
䊊䊊䊊, very low SSc, including pulmonary hypertension,
with SSc-ILD: 1) worsening dyspnea or confidence in effect estimates) esophageal dysmotility, and gastroesophageal
cough; 2) physiological evidence of disease e. Nintedanib plus mycophenolate reflux disease, which may affect outcomes in
progression (>5% absolute decline in FVC (
䊊䊊䊊, very low confidence in effect patients with SSc-ILD, were not addressed in
[forced vital capacity] or >10% absolute estimates) this guideline, which is focused on treatment
decline in DLCO [diffusing capacity of the 3. The recommendation for further of SSc-ILD. The standard of care for patients
lung for carbon monoxide] adjusted for research because of insufficient with SSc includes lifestyle modification
hemoglobin); or 3) radiological evidence of evidence was made for the following measures to minimize risks for overt and/or
disease progression (radiological agents for the treatment of SSc-ILD and occult microaspiration, with or without
interpretation of increase in the extent or is strong: treatment with proton pump inhibitors,
severity of ILD features on computed a. Pirfenidone (
䊊䊊䊊, very low H2-blockers, and/or prokinetic agents.
tomography [CT] assessed visually). confidence in effect estimates) Because the question of whether to treat
b. Pirfenidone plus mycophenolate patients with SSc-ILD with targeted
Therapies Assessed and Summary (
䊊䊊䊊, very low confidence in effect treatment for symptomatic or asymptomatic
of Recommendations estimates) gastroesophageal reflux was not addressed in
Six individual therapies and two combination this guideline, we refer providers to existing
therapies were assessed in this guideline guidelines for treatment of this clinical
(Table 1). Each therapy in this guideline was Introduction manifestation of SSc (11). Similarly, we refer
compared with either placebo or standard providers to existing guidelines for treatment
of care. Standard of care for SSc was Interstitial lung disease (ILD) is a significant of pulmonary hypertension (12).
determined a priori by the committee to be cause of morbidity and mortality in patients The guideline committee found it
cyclophosphamide or mycophenolate based with systemic sclerosis (SSc) (6). ILD has important to note that renal crisis in patients
on findings from the SLS I (Scleroderma emerged as the leading cause of SSc-related with SSc has been associated with systemic
Lung Study I) (4) and mycophenolate for all death in patients with SSc, contributing to corticosteroid therapy in some cases,
studies conducted after the completion of the 35% of deaths (7). As such, optimal particularly in those with early diffuse
SLS II (5). Following is the summary of treatment recommendations for the care of cutaneous SSc (13). The likelihood of
recommendations for specific therapies, which patients with SSc-ILD is imperative. Several developing renal crisis is less if the daily dose
were not assessed as a stepwise algorithm: randomized controlled trials (RCTs) have of corticosteroids does not exceed the
been performed demonstrating the efficacy equivalent of 15 mg prednisone (14).
1. The recommendation for the use of the of specific treatments for SSc-ILD (8), and Therefore, the committee issued the
following agent for the treatment of clinical practice guidelines regarding the following best clinical practice statement:
SSc-ILD is strong: treatment approach for patients with SSc-
a. Mycophenolate (
䊊䊊䊊, very low ILD are primarily consensus based (9, 10). ❖ Caution must be taken in using
confidence in effect estimates) There is practice variation in terms of 1) systemic corticosteroids in patients
2. The recommendation for the use of the when ILD-directed treatment is initiated; 2) with SSc with or without SSc-ILD.
following agents for the treatment of what the initial therapeutic agent of choice is; Whenever possible, the daily dose
SSc-ILD is conditional: and 3) when add-on therapy is needed. should not exceed the equivalent of 15
mg prednisone.
Table 1. PICO Questions for Systemic Sclerosis–associated Interstitial Lung
Disease Therapies Assessed
Methods
Question Therapy This guideline follows the ATS guideline
development process and is in compliance
1 Should patients with SSc-ILD be treated with cyclophosphamide? with the National Academy of Medicine
2 Should patients with SSc-ILD be treated with mycophenolate? (formerly Institute of Medicine) standards
3 Should patients with SSc-ILD be treated with rituximab?
4 Should patients with SSc-ILD be treated with tocilizumab? for trustworthy guidelines (15).
5 Should patients with SSc-ILD be treated with nintedanib? Recommendations were informed by
6 Should patients with SSc-ILD be treated with nintedanib plus mycophenolate? systematic reviews of the published
7 Should patients with SSc-ILD be treated with pirfenidone? evidence assessed using the Grading of
8 Should patients with SSc-ILD be treated with pirfenidone plus mycophenolate?
Recommendations, Assessment,
Definition of abbreviation: PICO = population, intervention, comparison, outcome; SSc-ILD = Development, and Evaluation (GRADE)
systemic sclerosis–associated interstitial lung disease. approach (16).

American Thoracic Society Documents

AJRCCM Articles in Press. Published September 29, 2023 as 10.1164/rccm.202306-1113ST

Copyright © 2023 by the American Thoracic Society
 Page 4 of 16

AMERICAN THORACIC SOCIETY DOCUMENTS

Committee Composition Formulating Clinical Questions made when appropriate and additional data
This guideline was developed by an The committee met virtually to discuss the incorporated if critical studies were not
international multidisciplinary committee scope and objectives of the project before the identified by the search. When there was a
that consisted of pulmonologists with literature search. Questions on population paucity of randomized trials available,
recognized ILD expertise (n = 7: G.R., S.B.M., definition, interventions, comparisons, post hoc analyses and observational studies
A. Adegunsoye, A. Azuma, R.J.K., M.K., and and outcomes were formulated, with six were included.
M.B.S.), rheumatologists with recognized therapeutic areas prioritized based on The quality of evidence was determined
SSc expertise (n = 7: R.M.S., L.C., G.C.G., consensus voting by the committee, and an using the GRADE approach, with focus on
M.H., V.S., E.R.V., and F.M.W.), a additional two questions were added based risk of bias, precision, consistency, directness
general pulmonologist (C.C.T.), a on the availability of time. Of note, the of the evidence, risk for publication bias,
pulmonologist/rheumatologist (K.B.H.) committee did consider including presence of dose–effect relationship,
with recognized expertise in SSc and ILD, azathioprine and autologous hematopoietic magnitude of effect, and assessment of the
an information scientist (S.L.K.), and two stem cell transplantation. Although the effect of plausible residual confounding or
patients with SSc-ILD (D.B. and K.A.K.) who committee would have liked to have been bias. The confidence in the effect estimates
were recommended for participation by the able to include these therapeutic modalities, based on the quality of evidence was
National Scleroderma Foundation and were given limits in the number of clinical categorized as high, moderate, low, or very
not under the care of any of the clinical questions that could feasibly be addressed low. In developing recommendations, the
experts on the committee. The committee for the guidelines, the consensus of the quality of evidence played a critical role,
was chaired by G.R. and co-chaired by committee was to exclude them because together with the balance of desirable and
M.G., S.B.M., and R.M.S. of limited data and, for autologous undesirable effects, values and preferences,
The clinical experts worked closely hematopoietic stem cell transplantation feasibility of implementation, and
with five health research methodologists in particular, the fact that it is in general implications for resource use and health
(M.G., T.H., M.M., D.H., and H.B.) with practice a salvage therapy for patients not equity. Incorporating the above, the
expertise in evidence synthesis and guideline responding to more conventional therapy. committee came to final recommendations
development. The methodologists conducted based on consensus voting. All votes
systematic reviews and prepared the Literature Search
required at least 80% participation by
systematic evidence summaries following In collaboration with the lead methodologist
nonconflicted voting members and at least
the GRADE approach (16). (M.G.), an information scientist (S.L.K.)
70% agreement for or against a proposed
designed a search strategy using medical
therapy. Recommendations were either
Confidentiality Agreement and subject heading keywords and text words
“strong” or “conditional” (or “weak”) in
Conflict-of-Interest Management limited to human studies inclusive of
favor of or against each therapy. Strong
Committee members disclosed all potential nonindexed citations and articles in English
recommendations begin with the phrase
conflicts of interest according to the ATS or in any language with English abstracts.
“we recommend,” whereas conditional
policies. All potential conflicts of interest of The Ovid platform was used to search
recommendations begin with the phrase
committee members were reviewed by the MEDLINE, EMBASE, Cochrane Registry
staff of the ATS conflict-of-interest and of Controlled Trials, Health Technology “we suggest.” The implications of the strong
documents units. Assessment, and the Database of Abstracts and conditional recommendations are listed
Six members were considered to have of Reviews of Affects. in Table 2.
disqualifying conflicts of interest based on Methodologists reviewed references
disclosures, whereas an additional four for additional articles and contacted experts Manuscript Preparation
had manageable conflicts of interest. All outside the committee for additional data The writing committee (Co-Chair M.G. and
conflicted members were permitted to when appropriate. After removal of the methodologists T.H., M.M., D.H., and
participate in the discussions of the evidence duplicates, two methodologists screened H.B.; the Chair G.R. and Co-Chairs S.B.M.
with the rest of the committee; however, titles, abstracts, and full texts for inclusion and R.M.S.) drafted the guideline document.
those with disqualifying conflicts were of studies on the basis of predefined The manuscript was then reviewed by the
instructed to abstain from voting on all eligibility criteria. entire committee. Feedback was provided
recommendations, and those with via electronic communication. The
manageable conflicts were instructed to Evidence Review and Development of entire committee (both conflicted and
abstain from voting for the question(s) for Clinical Recommendations nonconflicted members) had the opportunity
which they were considered conflicted. The methodologists extracted data and to correct factual errors, clarify the
This approach was applied to all questions. created a summary of the evidence following presentation of background information
Adherence to these rules was strict, with the GRADE approach (16). Data were or evidence summaries, and suggest changes
voting and discussions monitored by one of pooled, and meta-analysis was performed to the rationale sections to capture the
the co-chairs (M.G.) together with ATS staff where appropriate. The evidence summaries discussion from the virtual meetings.
members participating in the meetings. The were disseminated to committee members However, only the nonconflicted voting
nonconflicted committee members and the and presented in virtual meetings. members were permitted to comment on the
methodologists were allowed unrestricted Committee members provided feedback on recommendations. Conflicted committee
participation (n = 14). the evidence summaries, with corrections members were not permitted to comment

American Journal of Respiratory and Critical Care Medicine

AJRCCM Articles in Press. Published September 29, 2023 as 10.1164/rccm.202306-1113ST

Copyright © 2023 by the American Thoracic Society
Page 5 of 16

AMERICAN THORACIC SOCIETY DOCUMENTS

Table 2. Implications of the Guideline Recommendations for Patients with Systemic Sclerosis–associated Interstitial Lung
Disease

Stakeholder Strong Recommendation Conditional Recommendation

Patients Most individuals in this situation would The majority of individuals in this situation
want the recommended course of action, would want the suggested course of
and only a small proportion would not. action, but some would not.
Clinicians Most individuals should receive the Recognize that different choices will be
intervention. Adherence to this appropriate for individual patients and
recommendation according to the that you must help each patient arrive
guideline could be used as a quality at a management decision consistent
criterion or performance indicator. with their values and preferences.
Formal decision aids are not likely to be Decision aids may be useful in helping
needed to help individuals make individuals to make decisions
decisions consistent with their values consistent with their values and
and preferences. preferences.
Policy makers The recommendation can be adopted as Policy making will require substantial
policy in most situations. debate and involvement of various
stakeholders.

on the recommendations pertinent to the efficacy in certain types of inflammation- 12 months of placebo followed by 12 months
specific conflicted question and restricted driven ILD, including SSc-ILD (23). off therapy. SLS I included participants with
their feedback to the presentation of the However, it is associated with serious side SSc and with active alveolitis on BAL or
evidence and the identification of errors. effects, including bone marrow suppression, ground-glass opacity on high-resolution CT
The wording of recommendations infections, hemorrhagic cystitis, and (HRCT) of the chest and at least moderate
(including strength and direction) was not subsequent risk for bladder cancer (24). dyspnea (4). Hoyles and colleagues reported
altered once finalized by the nonconflicted The committee addressed the question on a multicenter U.K.-based, charitable
members during the virtual meetings. One of “Should patients with SSc-ILD be treated donation–funded RCT including
the nonconflicted co-chairs (M.G.) with cyclophosphamide?” The guideline participants with SSc and evidence of
confirmed that the written version of the committee identified potential heterogeneity pulmonary fibrosis on HRCT or lung biopsy
guideline reflected the recommendations in the intervention, as cyclophosphamide that compared placebo to a regimen of
made by the nonconflicted members. The may be administered intravenously or intravenous cyclophosphamide monthly for
same process was followed for each version orally, with potential differences in efficacy 6 months plus prednisolone 20 mg every
of the document. The final approved version and adverse events between routes of other day followed by azathioprine (25).
was submitted to ATS for peer review. administration. Therefore, this led to One RCT and two case–control studies
dividing the overarching research compared the use of cyclophosphamide to
question into: 1) “Should patients with mycophenolate. SLS II was a multicenter,
Evidence-based SSc-ILD be treated with intravenous U.S.-based, NIH-funded, double-blind
Recommendations for the cyclophosphamide?” and 2) “Should RCT that compared oral cyclophosphamide
Treatment of SSc-ILD patients with SSc-ILD be treated with oral for 12 months followed by placebo for
cyclophosphamide?” Critical outcomes 12 months to oral mycophenolate for
No studies were identified that addressed the included disease progression (including 24 months (5). Shenoy and colleagues (26)
committee’s predefined subgroups of interest FVC, DLCO, and the modified Rodnan Skin and Panopoulos and colleagues (27) were
for disease status (initial manifestation of Score [mRSS], with the latter as an indirect both retrospective, single-center, unfunded
SSc-ILD, stable SSc-ILD, and/or progressive measure for progression of disease used case–control studies that identified patients
SSc-ILD), and thus the treatment primarily for SSc) and mortality. Important with SSc-ILD who had been treated with
recommendations made in this guideline outcomes included quality of life (including intravenous or oral cyclophosphamide and
document are for a heterogenous population the Health Assessment Questionnaire– compared outcomes with patients who were
of patients with SSc-ILD regardless of Disability Index [HAQ-DI]) and treated with oral mycophenolate for 12 or
their disease status. The systematic adverse events. 24 months. There was not enough evidence
reviews informing the committee’s Summary of evidence. A systematic to be able to separate cyclophosphamide
recommendations are being published review of the evidence identified five studies therapy by route of administration.
separately (17–22). (see Table E1 in the online supplement). Two DISEASE PROGRESSION. When compared
RCTs compared cyclophosphamide to with placebo, the mean change in FVC %
Question 1: Should patients with SSc- placebo. SLS I (Scleroderma Lung Study I) predicted at 12 months was 2.8%, favoring
ILD be treated with cyclophosphamide? was a 24-month, multicenter U.S.-based, cyclophosphamide. Treatment
Background. Cyclophosphamide is a NIH-funded RCT that randomized patients with cyclophosphamide was associated with
cytotoxic agent that has demonstrated to 12 months of cyclophosphamide or an improvement in FVC % predicted at

American Thoracic Society Documents

AJRCCM Articles in Press. Published September 29, 2023 as 10.1164/rccm.202306-1113ST

Copyright © 2023 by the American Thoracic Society
 Page 6 of 16

AMERICAN THORACIC SOCIETY DOCUMENTS

12 months in a greater proportion of Quality of evidence. The quality of ILD (QILD) scores, had an increased risk for
participants compared with placebo (49.3% evidence was rated as low for these long-term mortality (29).
vs. 26.4%, respectively). At 24 months, the outcomes, which means there was low Future research opportunities. The
mean values of FVC % predicted were confidence in the estimated effects. committee noted that in clinical practice
similar between treatment and placebo Therefore, the data should be interpreted mycophenolate is used more frequently than
groups. There was no difference in DLCO % with caution. Quality of evidence was cyclophosphamide because of a more
predicted between groups at 12 or reduced in cyclophosphamide compared favorable side effect profile with
24 months. When cyclophosphamide was with placebo because few trials studied this mycophenolate. The committee
compared with mycophenolate, there was a comparison, leading to imprecision, and the acknowledges that this applies to countries
difference in DLCO % predicted favoring intervention in Hoyles and colleagues where both medications are accessible and
mycophenolate at 6 months and 18 months, included azathioprine and prednisolone may not be the case in countries where
but not at 12 months or 24 months. When as well as cyclophosphamide (25). The mycophenolate access is limited. In addition,
cyclophosphamide was compared with quality of evidence was low for the it was noted that most beneficial effects of
mycophenolate, both groups showed an cyclophosphamide versus mycophenolate cyclophosphamide treatment waned 1 year
improvement in FVC % predicted, but there comparison because of imprecision, study after cessation of cyclophosphamide
was no difference between the two groups at design (retrospective case–control studies), treatment (30). Further research is needed to
any time point. The change from baseline at and indirectness of the comparator (multiple determine treatment duration for disease-
12 months for the mRSS was 3.06 better in formulations of mycophenolate). modifying therapies for SSc-ILD. Future
the subset of patients with diffuse SSc. Recommendation 1: We suggest using research should also assess the use of
MORTALITY. When comparing placebo cyclophosphamide to treat patients with cyclophosphamide in patients with an initial
and cyclophosphamide, there was no SSc-ILD (conditional recommendation, low- diagnosis of SSc-ILD, in stable SSc-ILD, and
difference in mortality at 12 or 24 months. quality evidence). The committee vote was in progressive SSc-ILD, as well as by route of
When comparing cyclophosphamide to as follows: strongly in favor to use administration (oral vs. intravenous).
mycophenolate, there was no difference in cyclophosphamide in people with SSc-ILD:
mortality between groups at 24 months. 5 of 17 (29%); conditional recommendation Question 2: Should patients with SSc-
QUALITY OF LIFE. When compared with to use cyclophosphamide in people with ILD be treated with mycophenolate?
placebo, there was a significant improvement SSc-ILD: 12 of 17 (71%); conditional Background. Given that mycophenolate is
in the cyclophosphamide arm for recommendation to not consider an inhibitor of inosine monophosphate
breathlessness and disability according to the cyclophosphamide: 0 of 17 (0%); strong dehydrogenase that impairs T and B cell
HAQ-DI. When comparing recommendation to not consider proliferation (31) and is currently used as a
cyclophosphamide to mycophenolate, cyclophosphamide: 0 of 17 (0%). No standard-of-care treatment for SSc, the
although both arms showed significant guideline participants (0%) abstained committee asked the question, “Should
improvement in quality of life (QoL) from this vote. patients with SSc-ILD be treated with
outcomes such as breathlessness, cough, Justification and implementation mycophenolate?” The committee was
and disability, there was no difference considerations. The committee discussed interested in the evidence base for the
between groups. that although QoL measures were considered different formulations of mycophenolate,
ADVERSE EVENTS. When compared with a priori to be important, rather than critical specifically mycophenolate mofetil and
placebo, there was a 15-fold increased risk of outcomes, the significant improvement in mycophenolic acid. Cyclophosphamide,
hematologic adverse events using QoL measures as shown by the data would which was the previous standard of care
cyclophosphamide at 12 months, including be important to clinicians and patients when since the publication of SLS I (4), and
leukopenia (requiring discontinuation in considering this treatment. In addition, the placebo were both deemed appropriate
seven cyclophosphamide cases) and committee discussed the clinical importance comparators. Critical outcomes included
thrombocytopenia. There was also a of the increased risk of adverse events. For mortality and disease progression (defined as
fourfold increased risk of infections using example, although there was a 38-fold changes in FVC, DLCO, radiologic disease,
cyclophosphamide at 12 months. At increased risk of leukopenia using and the mRSS, with the latter an indirect
24 months, there was an increased risk cyclophosphamide compared with placebo, measure for progression of disease used
of constitutional symptoms using the confidence intervals were wide, reducing primarily for SSc). Important outcomes
cyclophosphamide. There was no increased our confidence in the estimate of the true included quality of life indices (including the
incidence of hematuria or hemorrhagic effect, and, ultimately, cyclophosphamide- Transition Dyspnea Index [TDI]) and
cystitis using cyclophosphamide compared related leukopenia resulted in an actual adverse events.
with placebo (28). When compared with change in treatment for only seven patients. Summary of evidence. A systematic
mycophenolate, participants were 1.7 times It is also worth noting that trial design review of the evidence identified seven total
more likely to prematurely discontinue impacts outcomes, as evidenced by the fact studies (5, 26, 27, 32–35) (Table E2). Two
cyclophosphamide therapy. There was a that the response to cyclophosphamide in were RCTs (5, 33): three were post hoc
sixfold increased risk of leukopenia using SLS II was better than in SLS I. In addition, analyses of RCTs (32, 34, 35), and two were
cyclophosphamide compared with in a post hoc analysis of the SLS I and SLS II observational studies (26, 27). Two studies
mycophenolate, but there was no difference population, patients with radiologic compared mycophenolate to placebo
in any other reported adverse events. progression of ILD, assessed by quantitative (33, 35), and five compared mycophenolate

American Journal of Respiratory and Critical Care Medicine

AJRCCM Articles in Press. Published September 29, 2023 as 10.1164/rccm.202306-1113ST

Copyright © 2023 by the American Thoracic Society
Page 7 of 16

AMERICAN THORACIC SOCIETY DOCUMENTS

to cyclophosphamide (5, 26, 27, 32, 34). Quality of evidence. The quality of and mortality. Important outcomes included
The predominance of data comparing evidence for all outcomes was rated very low, QoL measures (including the Short Form 36
mycophenolate to placebo was from a post meaning the effect estimates should be questionnaire and HAQ-DI) and
hoc study that compared those who received interpreted with caution. The primary adverse events.
mycophenolate in SLS II with those patients reasons were due to the majority of Summary of evidence. A systematic
who received placebo in SLS I (35). The SLS outcomes drawing data from indirect review of the evidence identified three RCTs
II trial provided the majority of evidence for evidence. The main study comparing that enrolled patients with SSc and evaluated
the comparison between mycophenolate and mycophenolate to placebo, for example, was the effects of rituximab compared with
cyclophosphamide (5). All studies, except for post hoc in nature, with significant placebo (38–40) (Table E3). However, two of
one, used mycophenolate mofetil for the differences in baseline characteristics the studies enrolled participants with SSc
drug formulation. between the groups. without a priori confirmation of ILD, thus
DISEASE PROGRESSION. When compared Recommendation 2: We recommend providing only indirect data on the SSc-ILD
with placebo, the mean FVC % predicted using mycophenolate to treat patients with population (38, 40). The sample sizes were
significantly improved from baseline to 12 SSc-ILD (strong recommendation, very low- small, ranging from a total of 14 to 54
and 24 months for mycophenolate, with quality evidence). The committee vote was patients, and two of the studies were
about a 5% difference between the two arms. as follows: strongly in favor to use underpowered for the studied outcomes (38,
In addition, the rate of overall improvement mycophenolate in people with SSc-ILD: 14 of 39). Follow-up for these trials ranged from
in FVC % predicted at 12 and 24 months was 18 (78%); conditional recommendation to 24 to 96 weeks. Patients received rituximab
nearly 2.3-fold higher at both time points in use mycophenolate in people with SSc-ILD: 4 infusion on Days 1 and 15 and at 6 months
the mycophenolate arm compared with of 18 (22%); conditional recommendation to in one study (38), weekly for four doses at
placebo. Similarly, the mean change from not consider mycophenolate: 0 of 18 (0%); baseline and at 6 months in a second study
baseline in DLCO % predicted was .4% less strong recommendation to not consider (39), and only weekly for four doses at
at both 12 and 24 months for the mycophenolate: 0 of 18 (0%). No guideline baseline in a third (40).
mycophenolate arm compared with placebo, participants (0%) abstained from this vote. DISEASE PROGRESSION. Meta-analysis
favoring mycophenolate. There were no Justification and implementation revealed that at 24–48 weeks, rituximab
differences between mycophenolate considerations. Although the quality of the attenuated the decline in FVC % predicted
and cyclophosphamide in mean change evidence was poor by GRADE criteria, the by 3.3% when compared with placebo.
in FVC % predicted or DLCO % predicted committee decided to make a strong Individual study and pooled data analyses
at 12 or 24 months. There were also no recommendation for mycophenolate in showed no differences in the mean change in
differences between mycophenolate and patients with SSc-ILD, given the significant the DLCO % predicted at 24, 24–48, or
cyclophosphamide in several measures of reduction in disease progression and 96 weeks between the rituximab and placebo
radiologic disease, given both treatments led improvement in QoL measures with arms. Two studies found that rituximab
to improvements in radiologic disease minimal adverse events. According to reduced the decline in DLCO (improvement
individually. In addition, between GRADE guidelines, strong in DLCO, 0.7 to 9.7 ml/min/mm Hg), whereas
mycophenolate and placebo, changes in the recommendations can rarely be made despite one found rituximab increased the decline in
mRSS favored mycophenolate. low confidence in effect estimates when there DLCO (23.5 ml/min/mm Hg). There were no
MORTALITY. There was no significant is a possibility of appreciable gain with a low significant differences in mean changes in
difference in mortality at 24 months between incidence of adverse effects (36). several measures of radiographic disease at
mycophenolate and placebo or between Future research opportunities. Future 24 or 48 weeks, but the estimates are based
mycophenolate and cyclophosphamide. research should assess the use of on small sample sizes. Patients with SSc-ILD
QOL. Significant differences in mycophenolate in patients with an initial who received rituximab had larger decline in
breathlessness (measured using the TDI diagnosis of SSc-ILD, in stable SSc-ILD, and the mRSS at 24–48 weeks by 7 points.
score) at all time points, including in progressive SSc-ILD. In addition, specific MORTALITY. There were no significant
24 months, favored mycophenolate over focus should be placed on comparing differences at 24 weeks between the rituximab
placebo. There was no difference in any QoL mycophenolate mofetil to and placebo arms for mortality.
measure between mycophenolate and mycophenolic acid. QOL. Individual study and pooled data
cyclophosphamide, although both showed analyses showed no differences between the
significant improvement separately. Question 3: Should patients with rituximab and placebo arms for the Short
ADVERSE EVENTS. There was a ninefold SSc-ILD be treated with rituximab? Form 36 bodily pain and general health
increased risk of anemia in patients treated Background. Rituximab is a monoclonal question subsets or the HAQ-DI scores.
with mycophenolate versus placebo, but antibody that binds to cell surface proteins ADVERSE EVENTS. No significant differences
there were no differences in premature found on B cells to eliminate them. B cells in adverse events were noted between the
discontinuation, serious adverse events, are thought to play a key part in the rituximab and placebo arms at 24 weeks
hematuria, leukopenia, neutropenia, or pathogenesis of SSc (37). Therefore, the (diarrhea, enterocolitis, gastroesophageal
thrombocytopenia. Compared with guideline committee addressed the question, reflux disease, mucositis, respiratory tract
patients receiving cyclophosphamide, the “Should patients with SSc-ILD be treated infection, arthralgia, decreased neutrophil
mycophenolate arm had a 41% lower risk of with rituximab?” Critical outcomes included count, dermatitis, increased C-reactive
premature discontinuation of therapy for any disease progression (including changes in protein, skin ulcerations and pulmonary
reason and 86% lower risk of leukopenia. FVC, DLCO, radiographic disease, and mRSS) valve disease) or 96 weeks (blood and

American Thoracic Society Documents

AJRCCM Articles in Press. Published September 29, 2023 as 10.1164/rccm.202306-1113ST

Copyright © 2023 by the American Thoracic Society
 Page 8 of 16

AMERICAN THORACIC SOCIETY DOCUMENTS

lymphatic disorders, infections and patients with SSc (41). Tocilizumab is a arm, with a median change of 3.4% less, but
infestations, neoplasm, reproductive and monoclonal antibody that targets the IL-6 at 96 weeks (when the placebo arm was also
breast, or vascular disorders). Similarly, no receptor. Therefore, the committee addressed given tocilizumab) there was no significant
differences at 24 weeks were present for any the question “Should patients with SSc-ILD difference between the tocilizumab and
adverse event, serious adverse event, or be treated with tocilizumab?” Critical placebo arms. The risk of FVC % predicted
serious adverse event leading to outcomes included disease progression decrease .10% by 48 weeks was three times
treatment withdrawal. (including changes in FVC and DLCO, fibrotic less in the tocilizumab arm, whereas the risk
Quality of evidence. The quality of changes on HRCT imaging, and, as an of any increase in the FVC % predicted at
evidence for study outcomes was very low indirect measure, mRSS) and mortality. 48 weeks was nearly twice as much in the
as defined by the GRADE approach, Important outcomes included QoL measures tocilizumab arm compared with placebo. By
because of risk of bias (premature closing in (including the 5-D Itch score, HAQ-DI, 96 weeks (when the placebo arm was also
enrollment), imprecision (limited number Functional Assessment of Chronic Illness given tocilizumab) there were no significant
of participants/studies contributing to the Therapy [FACIT]-Fatigue, and the Patient differences in risk for these parameters. In
findings, different rituximab dosing between Global Visual Analog Scale score) and contrast to the above trends, when evaluating
studies), and indirectness (ILD not adverse events. data from 48 to 96 weeks in the open-label
determined a priori in the participants). Summary of evidence. A systematic extension period, the mean change in the
Recommendation 3: We suggest using review of the evidence identified five studies absolute FVC was 54.9 ml less and the mean
rituximab to treat patients with SSc-ILD for inclusion: the faSScinate trial (42) and its change in FVC % predicted was 1.3% less in
(conditional recommendation, very low- open-label extension (43), the focuSSced trial the placebo arm. The mean change in DLCO
quality evidence). The committee vote was (41) and its open-label extension (44), and a % predicted from baseline to 48 weeks was
as follows: strongly in favor to use rituximab post hoc analysis of data from the focuSSced 1.5% less in the tocilizumab arm, but the
in people with SSc-ILD: 1 of 18 (5.6%); trial (45) (Table E4). The faSScinate trial was difference was not significant at 96 weeks.
conditional recommendation to use a phase 2 RCT that assigned 87 subjects with During the interval from 48 to 96 weeks, the
rituximab in people with SSc-ILD: 16 of 18 SSc across five countries to subcutaneous mean decrease in DLCO % predicted was
(88.9%); conditional recommendation to tocilizumab or placebo over 48 weeks. 5.4% less in the tocilizumab arm. At
not consider rituximab: 0 of 18 (0%); strong The open-label extension was extended to 48 weeks, the change in QILD and QLF
recommendation to not consider rituximab: 96 weeks and gave tocilizumab to 30 subjects scores across all categories favored the
0 of 18 (0%). One guideline participant in the original tocilizumab arm and 31 tocilizumab group. The mRSS change from
(5.6%) abstained from this vote because subjects in the original placebo arm. The baseline at 72 weeks was 4.1 better in the
of insufficient expertise to render a focuSSced trial was a phase 3 RCT that tocilizumab arm when compared with
thoughtful judgment. assigned 210 subjects with SSc across 20 placebo but was 0.8 better in the placebo arm
Justification and implementation countries to subcutaneous tocilizumab or compared with tocilizumab when looking at
considerations. The committee decided to placebo over 48 weeks. The open-label 48–96 weeks in the open-label extension
make a conditional recommendation for extension extended to 96 weeks and gave period when the placebo arm was also given
rituximab in patients with SSc-ILD, tocilizumab to 60 subjects in the original tocilizumab.
balancing the significant reduction in disease tocilizumab arm and 54 subjects in the MORTALITY. There was no significant
progression with no difference in adverse original placebo arm. The post hoc analysis difference in mortality between the
events against the very low-quality evidence. assessed QILD and quantitative lung fibrosis tocilizumab and placebo arms at 24, 48,
Future research opportunities. The (QLF) scores on imaging with QILD or 96 weeks.
committee noted from clinical experience categorized as mild (.5–10%), moderate QOL. At 96 weeks in the open-label study,
that the use of rituximab for SSc-ILD is often (.10–20%), or severe (.20%). For both the the mean change from baseline in the 5-D
as rescue therapy in individuals with faSScinate and focuSSced trials, the presence Itch score, HAQ-DI score, FACIT-Fatigue
evidence of SSc-ILD progression despite of ILD was not an inclusion criterion, and score, and the Patient Global Visual Analog
treatment with mycophenolate. However, the change in mRSS was the primary outcome. Scale score all favored the placebo group that
systematic review did not address the use of But in the focuSSced trial, 136 of the 210 was transitioned to tocilizumab during the
rituximab in this context. Further research is participants (65%) were deemed to have open-label period.
needed to determine the most optimal SSc-ILD based on a visual read of HRCT by ADVERSE EVENTS. At 48 weeks, there were
timing for the use of rituximab in the disease a thoracic radiologist. 3.8 fewer hypersensitivity events, 44 fewer
course of SSc-ILD (in patients with an initial DISEASE PROGRESSION. The differences in overall adverse events, 7.6 fewer adverse
diagnosis of SSc-ILD, in stable SSc-ILD, and mean absolute change from baseline in FVC events leading to treatment discontinuation,
in progressive SSc-ILD). between the tocilizumab and placebo arms 9.1 fewer infectious serious adverse events,
were 118 ml less at 24 weeks, 241 ml less at and 27.4 fewer overall serious adverse events,
Question 4: Should patients with 48 weeks, and 128.7 ml less at 96 weeks (the all per 100 patient-years, for tocilizumab
SSc-ILD be treated with tocilizumab? latter being the open-label period) in favor of compared with the placebo group. In the
Background. Elevated concentrations of IL-6 tocilizumab. Similarly, the difference in mean open-label extension from 48 to 96 weeks,
(interleukin-6) have been associated with change from baseline to 48 weeks in FVC % the arm that received tocilizumab the full
skin fibrosis and development of SSc-ILD in predicted was 6.5% less in the tocilizumab 96 weeks had 96.7 fewer overall adverse

American Journal of Respiratory and Critical Care Medicine

AJRCCM Articles in Press. Published September 29, 2023 as 10.1164/rccm.202306-1113ST

Copyright © 2023 by the American Thoracic Society
Page 9 of 16

AMERICAN THORACIC SOCIETY DOCUMENTS

events, 5.6 fewer infectious serious adverse individuals with SSc-ILD who would not absolute change from baseline in FVC was
events, and 8.6 overall serious adverse events have met the inclusion criteria for the 46.4 ml less for the nintedanib arm, with the
per 100 patient-years. The placebo arm, focuSSced trial (later stage SSc, limited risks of absolute decline from baseline in
however, was found to have 10.2 fewer cutaneous SSc, patients with SSc without FVC of .5% predicted and relative decline
injection site reactions per 100 patient-years elevated levels of acute-phase reactants). in ml of .5%, both about 25% less in the
at 48 weeks and 6.8 fewer hypersensitivity Future research should assess the use of nintedanib arm. When looking at the MCID
events per 100 patient-years from 48 tocilizumab in patients at the initial diagnosis (50), the nintedanib arm had .20% reduction
to 96 weeks. of SSc-ILD, in stable SSc-ILD, and in in risk of FVC decrease >3.3% predicted (the
Quality of evidence. The quality of progressive SSc-ILD and also compare the MCID threshold for worsening FVC) and had
evidence was rated as very low for all use of tocilizumab to other SSc-ILD a 50% increase in risk of FVC increase of
outcomes. Therefore, the effects summarized therapies, such as mycophenolate mofetil. >3.0% predicted (the MCID threshold for
should be interpreted with caution, because improvement in FVC). There was no
the committee had low confidence in the Question 5: Should patients with significant difference in the mRSS.
estimated effects. The overall very low-quality SSc-ILD be treated with nintedanib? MORTALITY. There was no significant
rating is based on the lowest quality of Background. Nintedanib is an oral difference between the nintedanib or placebo
evidence rating among the critical outcomes intracellular tyrosine kinase inhibitor that arms for all-cause mortality, fatal adverse
disease progression and mortality. The blocks pathways involved in fibrogenesis. events, or serious adverse events that
studies included did not a priori document It has been recommended in previous included death. However, for composite
ILD at enrollment and include post hoc and guidelines for the treatment of IPF and PPF outcomes of absolute decline in FVC >10%
open-label extension studies, leading to (3, 46). Therefore, the committee addressed predicted or death at 52 weeks and for
indirectness of evidence and imprecision. the question “Should patients with SSc-ILD absolute decline in FVC >10% predicted or
Recommendation 4: We suggest using be treated with nintedanib?” Critical between 5% and 10% predicted with DLCO
tocilizumab to treat patients with SSc-ILD outcomes included disease progression decline >15% predicted or death at 52 weeks,
(conditional recommendation, very low- (including changes in FVC, DLCO, and, the rate was approximately 40% less in the
quality evidence). The voting by the indirectly, mRSS) and mortality, whereas nintedanib arm compared with placebo.
committee was as follows: strong important outcomes included QoL measures QOL. There was no significant difference
recommendation for tocilizumab: 0 of 16 (including HAQ-DI, FACIT-Dyspnea, and between the nintedanib or placebo arms for
(0%); conditional recommendation for the St. George’s Respiratory Questionnaire absolute change from baseline in the HAQ-
tocilizumab: 16 of 16 (100%); conditional [SGRQ]) and adverse events. DI, FACIT-Dyspnea, or SGRQ scores.
recommendation against tocilizumab: 0 of 16 Summary of evidence. A systematic ADVERSE EVENTS. Nintedanib increased the
(0%); and strong recommendation against review of the evidence identified three studies risk of nausea (2.3 times), vomiting (2.4
tocilizumab: 0 of 16 (0%). No participants for inclusion: the safety and efficacy of times), diarrhea (2.4 times), weight loss (2.8
(0%) abstained from voting. nintedanib in systemic sclerosis (SENSCIS) times), and adverse events leading to
Justification and implementation trial (47), a post hoc analysis of the SENSCIS treatment discontinuation (1.8 times) but
considerations. The committee decided to trial (48), and a post hoc analysis of the decreased the risk of cough as an adverse
make a conditional recommendation for INBUILD trial (49) (Table E5). The event by 35%.
tocilizumab in patients with SSc-ILD, SENSCIS trial was a phase 3 RCT that Quality of evidence. The quality of
balancing the significant reduction in disease assigned 576 subjects with SSc-ILD across evidence was rated as very low for all
progression against the very low-quality 32 countries to nintedanib or placebo over outcomes. Therefore, the effects summarized
evidence. Despite there being randomized 52 weeks. Of note, background therapy with should be interpreted with caution, because
trials with a true placebo arm in which mycophenolate was allowed, with about half the committee had low confidence in the
participants were not exposed to any other of the subjects receiving the therapy. The post estimated effects. The overall very low-
medications, the data gathered were from hoc analysis examined changes in FVC % quality rating is based on the lowest quality
two randomized studies, post hoc analyses, predicted at categorical ranges, including at of evidence rating among the critical
and open-label extensions with the patient 5%, 10%, and by the minimal clinically outcomes disease progression and mortality.
population indirectly related to SSc-ILD important difference (MCID) for Despite the SENSCIS trial being an RCT, the
because of a lack of a priori inclusion of improvement and worsening of FVC (50). overall evidence quality was downgraded
patients with ILD as defined by this The INBUILD trial was an RCT that because the only other studies were post hoc
committee. In addition, the primary assigned 663 subjects with progressive ILD analyses, leading to indirectness of evidence
outcome of change in mRSS was not met in across 15 countries to nintedanib or placebo and imprecision. In addition, patients in the
either the faSScinate or focuSSced trials. over 52 weeks. The post hoc analysis assessed placebo arm of the SENSCIS trial were not
Future research opportunities. The prespecified subgroups based on ILD true placebos, as many were receiving
committee noted that the focuSSced trial diagnosis, from which 39 patients with background immunosuppressive
included individuals with early diffuse SSc-ILD were extracted for data analysis. medications for treatment of SSc-ILD.
cutaneous SSc with elevated levels of acute- DISEASE PROGRESSION. The annual rate of Recommendation 5: We suggest using
phase reactants, who represent a small decline in FVC was 44.5 ml less and the nintedanib to treat patients with SSc-ILD
proportion of patients with SSc-ILD. As decline in FVC % predicted was 1.2% less in (conditional recommendation, very low-
such, further research is needed into the the nintedanib arm compared with placebo, quality evidence). The voting by the
efficacy and effectiveness of tocilizumab for based on data from the SENSCIS trial. The committee was as follows: strong

American Thoracic Society Documents

AJRCCM Articles in Press. Published September 29, 2023 as 10.1164/rccm.202306-1113ST

Copyright © 2023 by the American Thoracic Society
 Page 10 of 16

AMERICAN THORACIC SOCIETY DOCUMENTS

recommendation for nintedanib, 1 of 14 Therefore, the committee addressed the nintedanib plus mycophenolate versus
(7%); conditional recommendation for question, “Should patients with SSc-ILD placebo, mycophenolate only, or
nintedanib, 11 of 14 (79%); conditional be treated with nintedanib plus nintedanib only.
recommendation against nintedanib, 1 of 14 mycophenolate?” The committee remained MORTALITY. There were no differences in
(7%); and strong recommendation against interested in the evidence base for the fatal adverse events comparing combination
nintedanib, 0 of 14 (0%). One participant different formulations of mycophenolate, therapy with nintedanib plus mycophenolate
(7%) abstained from voting, citing specifically mycophenolate mofetil and to placebo, mycophenolate only, or
insufficient expertise to render a mycophenolic acid, but these data were not nintedanib only.
thoughtful judgment. ascertainable. Appropriate comparators to QOL. There were no differences identified
Justification and implementation combination therapy included placebo, in SGRQ scores between combination
considerations. The committee decided to mycophenolate only, or nintedanib only. therapy with nintedanib plus mycophenolate
make a conditional recommendation for Critical outcomes included disease to placebo, mycophenolate only, or
nintedanib in patients with SSc-ILD, progression (including changes in FVC, nintedanib only.
balancing the significant reduction in disease DLCO, or radiographic disease) and mortality. ADVERSE EVENTS. Combination therapy
progression against gastrointestinal adverse Important outcomes included quality of life was associated with a sevenfold higher risk of
events, which can be managed with (using the SGRQ) and adverse events. decreased appetite, more than 2.5-fold higher
medication discontinuation, and low-quality Summary of evidence. A systematic risk of diarrhea, and about threefold higher
evidence attributable to the limited number review of the evidence identified three risk of nausea, vomiting, and/or fatigue
of randomized studies and primarily post hoc studies meeting inclusion criteria (47, 51, 52) compared with placebo. Combination
data analysis. It is worth noting that a recent (Table E6). One, the SENSCIS trial, was a therapy was also associated with nearly twice
guideline suggested the use of nintedanib for study that randomized 576 patients with the risk of diarrhea, nausea, and vomiting
non-IPF PPF, including progressive SSc-ILD SSc-ILD to nintedanib or placebo (as noted compared with mycophenolate only.
(3). Our recommendation broadens that above), but patients who had been on at least Combination therapy was associated with a
suggestion to include all patients with SSc- 6 months of therapy with mycophenolate at a 1.65-fold increase in serious adverse events
ILD, regardless of whether their disease is stable dosage were permitted in the trial (47). (defined as an event that resulted in death,
progressive or stable. This reflects the The second study was a post hoc subgroup was life-threatening, resulted in
observation that relevant trials found a analysis of the SENSCIS trial that examined hospitalization or prolongation of
benefit from nintedanib in heterogeneous the efficacy and safety of patients treated with hospitalization, resulted in persistent or
populations of SSc-ILD. The trials did not mycophenolate and nintedanib (51). This clinically significant disability or incapacity,
distinguish those with progressive disease study reported results for four groups— was a congenital anomaly or birth defect, or
from those with stable disease; therefore, it is combination therapy, mycophenolate plus was deemed to be serious for any other
possible that 1) the modest benefit in all placebo, nintedanib plus placebo, and reason) compared with mycophenolate only.
patients with SSc-ILD is the net effect of a placebo only—and provided the majority of Adverse event data could not be pooled for
large benefit in progressive SSc-ILD and no data for the systematic review. The third trial the comparison between combination
benefit in stable SSc-ILD; or 2) there is a was an open-label extension of the SENSCIS therapy and nintedanib only, but,
benefit in patients with stable SSc-ILD trial, in which all patients were offered interestingly, combination therapy was
through prevention of progression in those 52 weeks of therapy with nintedanib to associated with a 60% lower risk of liver
who are destined to eventually progress. examine safety and efficacy (52). test abnormalities compared with
Future research opportunities. Future DISEASE PROGRESSION. Compared with nintedanib only.
research should assess the use of nintedanib placebo, there was nearly an 80 ml and 2.5% Quality of evidence. The quality of
in patients with an initial diagnosis of SSc- lower annual rate of decline in FVC and FVC evidence for all outcomes was rated as very
ILD, in stable SSc-ILD, and in progressive % predicted, respectively, for combination low, meaning that the committee had very
SSc-ILD to discriminate its effect in stable therapy with nintedanib plus mycophenolate. low confidence in the estimated effects. As
versus progressive disease. Similarly, in the combination therapy arm, a result, the effect estimates should be
the risk of absolute decrease from baseline in interpreted with caution. There were
Question 6: Should patients with FVC of .5% predicted and .10% predicted multiple reasons for the very low quality of
SSc-ILD be treated with nintedanib were 50% and 75% less than the placebo arm, evidence. Each outcome was informed by
plus mycophenolate? respectively. These changes met established only a single study, leading to imprecision.
Background. For SSc-ILD, it is unknown MCID thresholds (50). There were no Furthermore, study design limitations
if combining therapies with different significant differences in the annual rate of downgraded evidence quality, as the majority
mechanisms of action is preferable to decline in FVC or FVC % predicted between of data were informed by a post hoc analysis
individual agents and, if dual therapy is combination therapy and mycophenolate or of an RCT. Finally, although treatment
preferred, what is the combination of choice. combination therapy and nintedanib, but the with nintedanib was randomized, therapy
The use of mycophenolate has demonstrated risk of FVC decrease from baseline by .5% with mycophenolate was not randomized,
improvement in the absolute FVC % was about one-third less in the combination and those patients receiving background
predicted over time, whereas nintedanib has therapy arm when compared with either therapy with mycophenolate had several
been shown to reduce the rate of disease mycophenolate alone or nintedanib alone. differences in demographics compared with
progression compared with placebo in There were no differences identified in patients not on background mycophenolate
patients with SSc-ILD (33, 35, 47, 48). mRSS between combination therapy with therapy (51).

American Journal of Respiratory and Critical Care Medicine

AJRCCM Articles in Press. Published September 29, 2023 as 10.1164/rccm.202306-1113ST

Copyright © 2023 by the American Thoracic Society
Page 11 of 16

AMERICAN THORACIC SOCIETY DOCUMENTS

Recommendation 6: We suggest using In addition, only 6% of the total participants was reached. Instead, a research
the combination of nintedanib plus myco- received the pirfenidone target dose of recommendation is made to further evaluate
phenolate to treat patients with SSc-ILD 2,400 mg/d. A majority of participants were treatment with pirfenidone in SSc-ILD.
(conditional recommendation, very low- receiving background therapy, mostly Future research opportunities. Because
quality evidence). The voting by the mycophenolate mofetil, azathioprine, and of the paucity of studies looking at
committee was as follows: strong prednisolone, which may have confounded pirfenidone for the treatment of patients
recommendation for nintedanib plus the effect of pirfenidone on proposed with SSc-ILD, future research should focus
mycophenolate, 1 of 14 (7%); conditional outcomes. Although SSc-ILD was confirmed on comparing the use of pirfenidone to
recommendation for nintedanib plus before enrollment, the extent of the ILD is placebo with larger sample sizes and evaluate
mycophenolate, 11 of 14 (79%); conditional not known. It is mentioned, however, that administration by disease status (initial
recommendation against nintedanib plus the majority of participants (61.7%) had diagnosis of SSc-ILD, stable SSc-ILD,
mycophenolate, 0 of 14 (0%); and strong nonspecific interstitial pneumonia, with the progressive SSc-ILD).
recommendation against nintedanib plus remaining (32.2%) having a UIP pattern on
mycophenolate, 0 of 14 (0%). Two the HRCT of the chest. Question 8: Should patients with
participants (14%) abstained from voting, DISEASE PROGRESSION. There were no SSc-ILD be treated with pirfenidone
citing insufficient expertise to render a significant differences between pirfenidone plus mycophenolate?
thoughtful judgment. and placebo for change from baseline in FVC Background. For SSc-ILD, it is unknown if
Justification and implementation % predicted, 6-minute-walk distance, combining therapeutic agents with different
considerations. The committee decided to or mRSS. mechanisms of action is preferable to
make a conditional recommendation for MORTALITY. Mortality was not reported in individual agents and, if dual therapy is
the combination of nintedanib plus this study. preferred, what is the combination of choice.
mycophenolate in patients with SSc-ILD, QOL. There was no difference at 24 weeks To evaluate combination therapy with
balancing the significant reduction in disease between pirfenidone and placebo in the pirfenidone and mycophenolate for the
progression against gastrointestinal adverse median change from baseline in the treatment of SSc-ILD, the guideline
events and very low-quality evidence due to TDI scores. committee addressed the question,
primarily post hoc data analysis. The decision ADVERSE EVENTS. There was no difference “Should patients with SSc-ILD be treated
to initiate combination therapy should be at 24 weeks between pirfenidone and placebo with pirfenidone plus mycophenolate?”
informed based on patient values and for any adverse event (including nausea, Critical outcomes included disease
preferences in this situation, given the vomiting, diarrhea, rashes, loss of appetite, progression (including changes in FVC,
increased risk of gastrointestinal side effects. constitutional symptoms, thrombocytopenia, DLCO, radiographic disease, and mRSS)
Future research opportunities. Future or elevation of transaminases). and mortality, whereas important outcomes
research should assess the use of Quality of evidence. The quality of included QoL measures (including TDI and
combination nintedanib plus mycophenolate evidence for both critical and important HAQ-DI scores) and adverse events.
in patients with an initial diagnosis of outcomes was very low as defined by the Summary of evidence. A systematic
SSc-ILD, in stable SSc-ILD, and in GRADE approach, due primarily to study review of the evidence identified one
progressive SSc-ILD. bias (low enrollment numbers owing to lack published study, the LOTUSS trial (54), and
of pirfenidone availability as a study drug) one abstract from the SLS III trial (55) for
Question 7: Should patients with and imprecision (limited number of inclusion (Table E8). The LOTUSS trial (54)
SSc-ILD be treated with pirfenidone? participants/studies contributing to the was an open-label phase 2 study of 63
Background. Pirfenidone is an antifibrotic findings, and lack of uniform distribution of patients with SSc-ILD monitored over 16
agent that has been recommended for use in pirfenidone dosing among the participants). weeks assessing safety and tolerability of
IPF and evaluated in guidelines for PPF Recommendation 7: We recommend pirfenidone. Patients were not randomized
(3, 46). Therefore, the committee addressed further research into the safety and efficacy to mycophenolate, but 63.5% of patients
the question, “Should patients with SSc-ILD of pirfenidone to treat patients with were concomitantly on it, so the data
be treated with pirfenidone?” Critical SSc-ILD. The voting by the committee was analyzed was post hoc. The baseline
outcomes included disease progression as follows: strong recommendation for mycophenolate dose varied between
(including changes in FVC, 6-minute-walk pirfenidone, 0 of 13 (0%); conditional participants, and 20% of patients were
distance, and mRSS) and mortality, whereas recommendation for pirfenidone, 0 of 13 on steroids and other antirheumatic
important outcomes included quality of (0%); conditional recommendation against medications. In addition, changes in lung
life measures (including TDI scores) and pirfenidone, 2 of 13 (15%); and strong function were exploratory outcomes, not
adverse events. recommendation against pirfenidone, 0 of 13 primary. The abstract described the results of
Summary of evidence. A systematic (0%). Eleven participants (85%) abstained the SLS III RCT that compared the treatment
review of the evidence identified one RCT from voting, citing insufficient evidence to with combined pirfenidone and
evaluating the use of pirfenidone in SSc-ILD render a thoughtful judgment. mycophenolate to mycophenolate plus
(53) (Table E7). This study, however, was Justification and implementation placebo, with the primary outcome being
underpowered for the proposed outcomes, considerations. Given the .20% abstention change in lung function at the end of 18
as it enrolled only 53% of the total planned rate citing insufficient evidence to render a months. The study was aborted due to
participants (n = 34) because of limited thoughtful judgment by the voting members inability to enroll the intended sample size
availability of pirfenidone as a study drug. of the committee, an insufficient quorum and had just enrolled 51 of the targeted 150

American Thoracic Society Documents

AJRCCM Articles in Press. Published September 29, 2023 as 10.1164/rccm.202306-1113ST

Copyright © 2023 by the American Thoracic Society
 Page 12 of 16

AMERICAN THORACIC SOCIETY DOCUMENTS

participants, so the results noted in the pirfenidone plus mycophenolate, 0 of 13 included in the systematic reviews. In
abstract were from a very small sample size (0%). Twelve participants (92%) abstained addition, high-quality data were not
and thus the study was underpowered (55). from voting, citing insufficient evidence to available to assess the treatment effect
While the abstract of the SLS III study does render a thoughtful judgment. for mycophenolate and cyclophosphamide
not include many secondary outcomes that Justification and implementation by drug formulation or route of
are anticipated to be published in the full considerations. Given the .20% abstention administration, respectively. Finally, the
report in the near future, the published rate citing insufficient evidence to render a questions addressed by this guideline did not
primary outcomes in the abstract were also thoughtful judgment by the voting members focus on comparing one therapeutic option
our critical outcomes of interest for of the committee, an insufficient quorum against others. Therefore, the committee felt
decision-making. was reached. Instead, a research that a prescriptive decision tree would not be
DISEASE PROGRESSION. No significant recommendation is made for further appropriate for this patient population
difference in FVC % predicted or DLCO % evaluation of combination treatment with with the evidence that was available.
predicted was observed between the pirfenidone plus mycophenolate in SSc-ILD. While an algorithm for different options
combination pirfenidone plus mycophenolate Future research opportunities. Because among the several treatment regimens
arm and pirfenidone alone. There were also of the paucity of studies looking specifically available is what clinicians are looking for in
no differences between the combination at the combination of mycophenolate and a guideline, such an approach is ideal only if
pirfenidone plus mycophenolate arm and the pirfenidone for the treatment of patients with the evidence permits us to provide this
mycophenolate and placebo arms in FVC % SSc-ILD, future research should focus on this approach. Because of the lack of data to
predicted or time duration to .3% combination against placebo and support a preferential and/or stepwise
increase in FVC % predicted at 18 months. mycophenolate alone with larger sample approach to treatment interventions for SSc-
MORTALITY. Mortality was not sizes and by disease status (initial diagnosis ILD, the committee provided the scientific
reported in either the LOTUSS trial or the of SSc-ILD, stable SSc-ILD, progressive information as it unfolded in the careful
SLS III abstract. SSc-ILD). review of the evidence to date to inform the
QOL. The LOTUSS trial found that treatment recommendations. The committee
compared with mycophenolate alone, the refrained from providing an algorithmic
combination of pirfenidone plus Guideline Considerations and approach to treatment of SSc-ILD that
mycophenolate showed a 2-point Future Directions would have been based solely on a consensus
improvement in the TDI score at 16 weeks, of opinions of the committee and not
but there was no significant difference in The focus of this ATS clinical practice evidence based.
HAQ-DI scores. guideline on the treatment of SSc-ILD is to In addition to the limitations in the
ADVERSE EVENTS. The LOTUSS trial did assess the use of separate therapies on their evidence, the committee also acknowledges
not observe any significant differences in own or in combination for mycophenolate the highly variable adverse event profiles,
severe adverse events, withdrawal because of with nintedanib and pirfenidone. The making it less than ideal to provide a
severe adverse events, or infections at 16 summary of our recommendations are noted hierarchical recommendation for these
weeks between combination therapy and the in Figure 1. The committee made a strong medications. Instead, the committee
pirfenidone-only arm. recommendation in favor of mycophenolate evaluated the medications independently, as
Quality of evidence. The quality of and conditional recommendations in favor of some patients may prefer one over another
evidence was very low by the GRADE cyclophosphamide, rituximab, tocilizumab, based on their adverse effect profiles, pill
approach because of bias (premature closure nintedanib, and the combination of burden, routes of administration, and/or
of enrollment), imprecision (limited number nintedanib plus mycophenolate. Research costs. The committee believes the current
of participants/studies contributing to the recommendations were made for the use approach in fact empowers the clinician to
findings, lack of uniform distribution of of pirfenidone and the combination of craft an algorithm tailored to the clinician’s
mycophenolate treatment in the pirfenidone pirfenidone plus mycophenolate. It is opinion in partnership with the well-
and mycophenolate participants), and worth highlighting that among the informed patient.
indirectness of evidence (post hoc analysis conditional recommendations, there was This guideline aims to serve as a starting
of data). a differentiation, with 29% of committee point to highlight gaps in evidence to
Recommendation 8: We recommend members voting strongly in favor of encourage future research into topics and
further research into the safety and efficacy cyclophosphamide, whereas for nintedanib comparisons that can then provide more
of pirfenidone plus mycophenolate combi- and the combination of nintedanib plus prescriptive guidance. Further research is
nation therapy to treat patients with mycophenolate only 7% voted strongly in therefore needed to determine treatment
SSc-ILD. The voting by the committee was favor, for rituximab only 6% voted strongly efficacy, safety, and impact on patient QoL
as follows: strong recommendation for in favor, and for tocilizumab 0% voted by disease status to determine if there exists a
pirfenidone plus mycophenolate, 0 of 13 strongly in favor. differential treatment effect going from initial
(0%); conditional recommendation for Although the committee sought to diagnosis of SSc-ILD to development of
pirfenidone plus mycophenolate, 0 of 13 differentiate SSc-ILD by initial diagnosis, progressive SSc-ILD. Furthermore,
(0%); conditional recommendation against stable SSc-ILD, and progressive SSc-ILD for additional studies looking at combination
pirfenidone plus mycophenolate, 1 of 13 each research question, disease status could therapy and the sequence of initiating each
(8%); and strong recommendation against not be extracted from any of the studies therapy would be of clinical benefit.

American Journal of Respiratory and Critical Care Medicine

AJRCCM Articles in Press. Published September 29, 2023 as 10.1164/rccm.202306-1113ST

Copyright © 2023 by the American Thoracic Society
Page 13 of 16

AMERICAN THORACIC SOCIETY DOCUMENTS

Figure 1. Summary of treatment recommendations for patients with systemic sclerosis–associated interstitial lung disease (SSc-ILD).
The SSc-ILD Guideline Committee:
1) Recommends the use of mycophenolate to treat patients with SSc-ILD (18 votes: 14 strong recommendation for use, 4 conditional
recommendations for use).
2) Suggests the use of cyclophosphamide to treat patients with SSc-ILD (17 votes: 5 strong recommendation for use, 12 conditional
recommendations for use).
3) Suggests the use of rituximab to treat patients with SSc-ILD (18 votes: 1 strong recommendation for use, 16 conditional recommendation
for use, 1 abstention due to insufficient expertise).
4) Suggests the use of tocilizumab to treat patients with SSc-ILD (16 votes: 16 conditional recommendation for use).
5) Suggests the use of nintedanib to treat patients with SSc-ILD (14 votes: 1 strong recommendation for use, 11 conditional recommendation
for use, 1 conditional recommendation against use, 1 abstention due to insufficient expertise).
6) Suggests the use of nintedanib plus mycophenolate to treat patients with SSc-ILD (14 votes: 1 strong recommendation for use,
11 conditional recommendation for use, 2 abstentions due to insufficient expertise).
7) Recommends further research into the efficacy, effectiveness, and safety of pirfenidone to treat patients with SSc-ILD (13 votes:
2 conditional recommendation against use, 11 abstentions due to insufficient evidence).
8) Recommends further research into the efficacy, effectiveness, and safety of pirfenidone plus mycophenolate to treat patients with SSc-ILD
(13 votes: 1 conditional recommendation against use, 12 abstentions due to insufficient evidence).
The above recommendations were not assessed as a stepwise algorithm. Clinicians are encouraged to use these recommendations in
conjunction with shared decision-making with patients, incorporating side effects and personal values and preferences before administration.

Conclusions 1) recommends the use of and safety of these therapies on SSc-ILD

mycophenolate; 2) recommends further subgroups by disease status (initial
The recommendations in this document research into the safety and efficacy of (a) diagnosis, stable SSc-ILD, and progressive
provide an evidence-based clinical pirfenidone and (b) the combination of SSc-ILD) and in various combinations.
practice guideline for the treatment of pirfenidone plus mycophenolate; and 3) This guideline was reviewed by the ATS
patients with SSc-ILD (Figure 1) and are suggests the use of (a) cyclophosphamide, Quality Improvement and
intended to serve as the basis for (b) rituximab, (c) tocilizumab, (d) Implementation Committee. None of the
informed and shared decision making by nintedanib, and (e) the combination of recommendations in this guideline are
clinicians and patients. For treatment of nintedanib plus mycophenolate. Future considered appropriate targets for a
patients with SSc-ILD, the committee: research should investigate the efficacy performance measure. 䊏

American Thoracic Society Documents

AJRCCM Articles in Press. Published September 29, 2023 as 10.1164/rccm.202306-1113ST

Copyright © 2023 by the American Thoracic Society
 Page 14 of 16

AMERICAN THORACIC SOCIETY DOCUMENTS

This official clinical practice guideline was prepared by an ad hoc subcommittee of the ATS Assembly on Clinical Problems.
Members of the subcommittee are as follows: Rheumatology, University of Washington Pulmonary Fibrosis Foundation. A.Azuma
GANESH RAGHU, M.D. (Chair)1,2,3* Medical Center, Seattle, Washington; served as consultant for Boehringer Ingelheim,
MARYA GHAZIPURA, PH.D., M.S., M.PHIL. (Co-Chair)4,5‡ 21
Respiratory Institute, Cleveland Clinic, Nippon Boehringer Ingelheim, Kyorin, Toray;
SYDNEY B. MONTESI, M.D. (Co-Chair)6* Cleveland, Ohio; 22Division of Rheumatology, served on data safety and monitoring board for
RICHARD M. SILVER, M.D. (Co-Chair)7§ Department of Medicine, Jewish General Nippon Boehringer Ingelheim; served as
TANZIB HOSSAIN, M.D., M.S.8‡ Hospital and 23McGill University, Montreal, speaker for Nippon Boehringer Ingelheim;
MADALINA MACREA, M.D., PH.D.9,10‡ Quebec, Canada; 24Weill Cornell Medical received research support from Boehringer
DERRICK HERMAN, M.D.11‡ Center, New York, New York; 25Department of Ingelheim and Taiho. L.C. served as consultant
HAYLEY BARNES, M.B.B.S., PH.D.12,13,14‡ Public Health Sciences, College of Behavioral, for Eicos, Genentech, IgM, Jasper, Kyverna,
AYODEJI ADEGUNSOYE, M.D., M.S.15* Social, and Health Sciences, Clemson Lilly, Mitsubishi Tanabe; received research
ARATA AZUMA, M.D., PH.D.16,17* University, Clemson, South Carolina; 26Strauss support from Acceleron, Boehringer Ingelheim,
DEE BURLILE, M.ED.18jj Health Sciences Library, University of Colorado Genentech, Horizon, Kadmon, Mitsubishi
LORINDA CHUNG, M.D.19§ Anschutz Medical Campus, Aurora, Colorado; Tanabe, Prometheus. G.C.G. served in
GREGORY C. GARDNER, M.D.20§ 27
Division of Respirology, Department of leadership role for American College of
KRISTIN B. HIGHLAND, M.D.21*§ Medicine, McMaster University, Hamilton, Rheumatology; provided expert panel
MARIE HUDSON, M.D., M.P.H.22,23§ Ontario, Canada; 28University of Utah, Salt testimony on SLE case for local attorney;
ROBERT J. KANER, M.D.24* Lake City, Utah; 29Georgetown University served as speaker for American College of
KAREN A. KEMPER, M.S.P.H., PH.D.25jj School of Medicine, Washington, D.C.; Physicians. K.B.H. served as consultant for
SHANDA L. KNIGHT, M.L.S.26¶ 30
Department of Medicine, Mount Auburn Boehringer Ingelheim and Genentech; served
MARTIN KOLB, M.D., PH.D.27* Hospital/Beth Israel Lahey Health, Harvard as speaker for Actelion, Acceleron, Bayer
MARY BETH SCHOLAND, M.D.28* Medical School, Cambridge, Massachusetts; Healthcare, United Therapeutics; served as
VIRGINIA STEEN, M.D.29§ 31
Division of Rheumatology, Department of speaker and received travel support from
CAREY C. THOMSON, M.D., M.P.H.30* Medicine, University of California at Los Boehringer Ingelheim; received research
ELIZABETH R. VOLKMANN, M.D., M.S.31§ Angeles, David Geffen School of Medicine, Los support from Acceleron, Actelion, Boehringer
FREDRICK M. WIGLEY, M.D.32§ Angeles, California; 32Johns Hopkins University Ingelheim, E.I. Lilly, Gossamer Bio, NIH, Reata,
1 School of Medicine, Baltimore, Maryland United Therapeutics, Vela Bio. M.H. served on
Center for Interstitial Lung Diseases, Division
of Pulmonary, Critical Care, and Sleep advisory committee for Alexion, Boehringer
*Pulmonologist
Medicine, 2Department of Medicine, and ‡ Ingelheim, Mallinckrodt; received honoraria
Methodologist
3
Department of Laboratory Medicine and § from Boehringer Ingelheim; received research
Rheumatologist
Pathology, University of Washington, Seattle, jj support from AstraZeneca, Boehringer
Patient representative
Washington; 4ZS Associates, Global Health ¶ Ingelheim, Bristol-Myers Squibb, Octapharma.
Medical information scientist
Economics and Outcomes Research, New R.J.K. served on advisory committee for
York, New York; 5Divisions of Epidemiology Author disclosures: G.R. served on advisory AstraZeneca, Boehringer Ingelheim,
and Biostatistics, Department of Population board for Pulmonary Fibrosis Foundation; Galapagos, Genentech, United Therapeutics;
Health, New York University Langone Health, served as consultant for Boehringer Ingelheim, served as consultant for Boehringer Ingelheim,
New York, New York; 6Division of Pulmonary Bellerophan, Blade Therapeutics, Bristol Myers Galapagos, UpToDate; served on data safety
and Critical Care Medicine, Massachusetts Squibb, Electra Therapeutics, Fibrogen, Nitto, and monitoring board for Pliant and PureTech;
General Hospital, Boston, Massachusetts; Roche-Genentech, United Therapeutics, has financial stake in AirCycleSystems and
7
Department of Medicine, Division of Veracyte, Zambon; served on data safety and Doximity; served in leadership role for
Rheumatology and Immunology, Medical monitoring board for Avalyn; served on steering Pulmonary Wellness Foundation and Stony
University of South Carolina, Charleston, South committee for Bellerophan, Bristol Myers World Foundation; served as medical writer
Carolina; 8Division of Pulmonary, Critical Care, Squibb, Gilead, Fibrogen, Nitto, Novartis, for AstraZeneca, Boehringer Ingelheim,
and Sleep Medicine, Department of Medicine, Veracyte; received research support from Galapagos, Genentech; served as speaker
Grossman School of Medicine, New York Biogen, Boehringer Ingelheim, Fibrogen, for Boehringer Ingelheim, France Foundation,
University Langone Health, New York, New NIH/NHLBI, Novartis; received royalties from Genentech, Vindico; received research
York; 9Division of Pulmonary and Sleep UpToDate. S.B.M. served on advisory support from Bellerophon, Boehringer
Medicine, Salem Veterans Affairs Medical committee for APIE Therapeutics and Pliant; Ingelheim, Chiesi, CSL Behring, Genentech,
Center, Salem, Virginia; 10Department of served as consultant for DevPro, Gilead, NIH, Respivant, Toray, US Department of
Medicine, University of Virginia, Charlottesville, Pieris, Roche; served in leadership role for Defense; received royalties from UpToDate;
Virginia; 11Division of Pulmonary, Critical Care, Massachusetts Pulmonary Society; received received travel support from Boehringer
and Sleep Medicine, Department of Internal research support from Boehringer Ingelheim, Ingelheim and United Therapeutics. M.K.
Medicine, The Ohio State Wexner Medical Merck, National Scleroderma Foundation, NIH/ served on advisory committee for Algernon;
Center, Columbus, Ohio; 12Central Clinical NHLBI, Pliant, Three Lakes Foundation, served as consultant for Abbvie, Algernon,
School and 13Centre for Occupational and United Therapeutics; received royalties from AstraZeneca, Bellerophon, Boehringer
Environmental Health, Monash University, UpToDate; served as speaker for Cowen; Ingelheim, Cipla, CSL Behring, Horizon,
Melbourne, Australia; 14Department of received travel support from DevPro. R.M.S. LabCorp, Roche, Structure Therapeutics,
Respiratory Medicine, Alfred Hospital, served on advisory committee for Actelion/ United Therapeutics; served on data safety
Melbourne, Australia; 15Section of Pulmonary Jansen and Boehringer Ingelheim; served as and monitoring board for LabCorp and United
and Critical Care, The University of Chicago speaker for Boehringer Ingelheim; received Therapeutics; provided expert testimony for
School of Medicine, Chicago, Illinois; 16Mihara research support from Boehringer Ingelheim, Roche; served in leadership role for ERJ;
General Hospital, Pulmonary Medicine and Genentech, Horizon, Kadmon, Prometheus. served as speaker for Boehringer Ingelheim,
Clinical Research Center, Saitama; 17Nippon H.B. received travel support from Boehringer Novartis, Roche; received research support
Medical School, Tokyo, Japan; 18National Ingelheim, Janssen, United Therapeutics. from Boehringer Ingelheim, Structure
Advocacy Committee Northwest Chapter Chair, A.Adegunsoye served as consultant for Therapeutics, United Therapeutics. M.B.S.
Idaho; 19Division of Immunology and Boehringer Ingelheim, Genentech, Inogen; served on advisory committee for Bristol
Rheumatology, Stanford University School of served as speaker for Boehringer Ingelheim; Myers Squibb, Genentech, Polarean, Tvardi,
Medicine, Palo Alto, California; 20Division of received research support from NIH/NHLBI and United Therapeutics, Veracyte; served as

American Journal of Respiratory and Critical Care Medicine

AJRCCM Articles in Press. Published September 29, 2023 as 10.1164/rccm.202306-1113ST

Copyright © 2023 by the American Thoracic Society
Page 15 of 16

AMERICAN THORACIC SOCIETY DOCUMENTS

consultant for Boehringer Ingelheim, Bristol Kadman. C.C.T. served on advisory board and Forbius, Horizon, Prometheus. D.B. served on
Myers Squibb, United Therapeutics, Veracyte; as consultant for Median Technology; served advisory committee and as chair for National
served as speaker for Boehringer Ingelheim, as chair for National Lung Cancer Roundtable Scleroderma Foundation; served as president
Genentech, United Therapeutics, Veracyte; Early Detection Task Group; served as editor for Scleroderma Outreach Northwest. K.A.K.
received research and travel support from for Lung Cancer Screening for Primary Care – served in leadership role for National
United Therapeutics. V.S. served on advisory Springer; received royalties from UpToDate; Scleroderma Foundation; served as speaker
board, as consultant, speaker, and received received research support from American for Boehringer Ingelheim. S.L.K. served as
research support from Boehringer Ingelheim; Cancer Society and NIH. E.R.V. served as consultant for Guideline panel. M.G. employee
served on advisory board and as consultant consultant for Boehringer Ingelheim, CSL of ZS Associates. T.H., M.M., D.H., F.M.W.
for CSL Behring and Eicos; served as speaker Behring, GlaxoSmithKline, Roche; served as reported no commercial or relevant non-
for Johnson & Johnson; received research speaker for Boehringer Ingelheim; received commercial interests from ineligible
support from Corbus, Eicos, Genentech, research support from Boehringer Ingelheim, companies.

References 14. Steen VD, Medsger TA Jr. Case-control study of corticosteroids and other
drugs that either precipitate or protect from the development of
1. Preliminary criteria for the classification of systemic sclerosis scleroderma renal crisis. Arthritis Rheum 1998;41:1613–1619.
(scleroderma): subcommittee for scleroderma criteria of the American 15. Graham R, Mancher M, Miller Wolman D, Greenfield S, Steinberg E,
Rheumatism Association Diagnostic and Therapeutic Criteria editors. Clinical practice guidelines we can trust. Washington, DC:
Committee. Arthritis Rheum 1980;23:581–590. National Academies Press; 2011.
2. van den Hoogen F, Khanna D, Fransen J, Johnson SR, Baron M, 16. Schu €nemann HJ, Jaeschke R, Cook DJ, Bria WF, El-Solh AA, Ernst A,
Tyndall A, et al. 2013 classification criteria for systemic sclerosis: an et al.; ATS Documents Development and Implementation Committee.
American College of Rheumatology/European League against An official ATS statement: grading the quality of evidence and strength
Rheumatism collaborative initiative. Arthritis Rheum 2013;65: of recommendations in ATS guidelines and recommendations. Am J
2737–2747. Respir Crit Care Med 2006;174:605–614.
3. Raghu G, Remy-Jardin M, Richeldi L, Thomson CC, Inoue Y, Johkoh T, 17. Barnes H, Ghazipura M, Herman D, Macrea M, Knight SL, Silver RM,
et al. Idiopathic pulmonary fibrosis (an update) and progressive et al. Cyclophosphamide in patients with systemic sclerosis-associated
pulmonary fibrosis in adults: an official ATS/ERS/JRS/ALAT clinical interstitial lung disease: a systematic review and meta-analysis. Ann
practice guideline. Am J Respir Crit Care Med 2022;205:e18–e47. Am Thorac Soc (In press)
4. Tashkin DP, Elashoff R, Clements PJ, Goldin J, Roth MD, Furst DE, et al.; 18. Herman D, Ghazipura M, Barnes H, Macrea M, Knight SL, Silver RM,
Scleroderma Lung Study Research Group. Cyclophosphamide versus et al. Mycophenolate in patients with systemic sclerosis-associated
placebo in scleroderma lung disease. N Engl J Med 2006;354: interstitial lung disease: a systematic review and meta-analysis. Ann
2655–2666. Am Thorac Soc [online ahead of print] 7 Apr 2023; DOI: 10.1513/
5. Tashkin DP, Roth MD, Clements PJ, Furst DE, Khanna D, Kleerup EC, AnnalsATS.202301-054OC.
et al.; Sclerodema Lung Study II Investigators. Mycophenolate mofetil 19. Macrea M, Ghazipura M, Herman D, Barnes H, Knight SL, Silver RM,
versus oral cyclophosphamide in scleroderma-related interstitial lung et al. Rituximab in patients with systemic sclerosis-associated interstitial
disease (SLS II): a randomised controlled, double-blind, parallel group lung disease: a systematic review and meta-analysis. Ann Am Thorac
trial. Lancet Respir Med 2016;4:708–719. Soc (In press)
6. Hoffmann-Vold AM, Allanore Y, Alves M, Brunborg C, Airo
P, 20. Ghazipura M, Macrea M, Herman D, Barnes H, Knight SL, Silver RM,
Ananieva LP, et al.; EUSTAR collaborators. Progressive interstitial et al. Tocilizumab in patients with systemic sclerosis-associated
lung disease in patients with systemic sclerosis-associated interstitial interstitial lung disease: a systematic review and meta-analysis. Ann
lung disease in the EUSTAR database. Ann Rheum Dis 2021;80: Am Thorac Soc (In press)
219–227. 21. Herman D, Ghazipura M, Barnes H, Macrea M, Knight SL, Silver RM,
7. Tyndall AJ, Bannert B, Vonk M, Airo  P, Cozzi F, Carreira PE, et al. Causes et al. Nintedanib therapy alone and combined with mycophenolate in
and risk factors for death in systemic sclerosis: a study from the EULAR patients with systemic sclerosis-associated interstitial lung disease:
Scleroderma Trials and Research (EUSTAR) database. Ann Rheum Dis systematic reviews and meta-analysis. Ann Am Thorac Soc (In press)
2010;69:1809–1815. 22. Macrea M, Ghazipura M, Herman D, Barnes H, Knight SL, Silver RM,
8. Roofeh D, Lescoat A, Khanna D. Treatment for systemic sclerosis- et al. Pirfenidone therapy alone and combined with mycophenolate in
associated interstitial lung disease. Curr Opin Rheumatol 2021;33: patients with systemic sclerosis-associated interstitial lung disease:
240–248. systematic reviews. Ann Am Thorac Soc (Submitted, Under Review)
9. Khanna D, Lescoat A, Roofeh D, Bernstein EJ, Kazerooni EA, Roth MD, 23. Kowal-Bielecka O, Kowal K, Rojewska J, Bodzenta-Lukaszyk A,
et al. Systemic sclerosis-associated interstitial lung disease: how to Siergiejko Z, Sierakowska M, et al. Cyclophosphamide reduces
incorporate two Food and Drug Administration-approved therapies in neutrophilic alveolitis in patients with scleroderma lung disease: a
clinical practice. Arthritis Rheumatol 2022;74:13–27. retrospective analysis of serial bronchoalveolar lavage investigations.
10. Rahaghi FF, Hsu VM, Kaner RJ, Mayes MD, Rosas IO, Saggar R, et al. Ann Rheum Dis 2005;64:1343–1346.
Expert consensus on the management of systemic sclerosis- 24. Fraiser LH, Kanekal S, Kehrer JP. Cyclophosphamide toxicity.
associated interstitial lung disease. Respir Res 2023;24:6. Characterising and avoiding the problem. Drugs 1991;42:781–795.
11. Katz PO, Dunbar KB, Schnoll-Sussman FH, Greer KB, Yadlapati R, 25. Hoyles RK, Ellis RW, Wellsbury J, Lees B, Newlands P, Goh NS, et al. A
Spechler SJ. ACG clinical guideline for the diagnosis and management multicenter, prospective, randomized, double-blind, placebo-controlled
of gastroesophageal reflux disease. Am J Gastroenterol 2022;117: trial of corticosteroids and intravenous cyclophosphamide followed by
27–56. oral azathioprine for the treatment of pulmonary fibrosis in scleroderma.
12. Humbert M, Kovacs G, Hoeper MM, Badagliacca R, Berger RMF, Arthritis Rheum 2006;54:3962–3970.
Brida M, et al.; ESC/ERS Scientific Document Group. 2022 ESC/ERS 26. Shenoy PD, Bavaliya M, Sashidharan S, Nalianda K, Sreenath S.
guidelines for the diagnosis and treatment of pulmonary hypertension. Cyclophosphamide versus mycophenolate mofetil in scleroderma
Eur Respir J 2023;61:2200879. interstitial lung disease (SSc-ILD) as induction therapy: a single-centre,
13. Trang G, Steele R, Baron M, Hudson M. Corticosteroids and the risk of retrospective analysis. Arthritis Res Ther 2016;18:123.
scleroderma renal crisis: a systematic review. Rheumatol Int 2012;32: 27. Panopoulos ST, Bournia VK, Trakada G, Giavri I, Kostopoulos C,
645–653. Sfikakis PP. Mycophenolate versus cyclophosphamide for progressive

American Thoracic Society Documents

AJRCCM Articles in Press. Published September 29, 2023 as 10.1164/rccm.202306-1113ST

Copyright © 2023 by the American Thoracic Society
 Page 16 of 16

AMERICAN THORACIC SOCIETY DOCUMENTS

interstitial lung disease associated with systemic sclerosis: a 2-year 43. Khanna D, Denton CP, Lin CJF, van Laar JM, Frech TM, Anderson ME,
case control study. Lung 2013;191:483–489. et al. Safety and efficacy of subcutaneous tocilizumab in systemic
28. Furst DE, Tseng CH, Clements PJ, Strange C, Tashkin DP, Roth MD, sclerosis: results from the open-label period of a phase II randomised
et al.; Scleroderma Lung Study. Adverse events during the controlled trial (faSScinate). Ann Rheum Dis 2018;77:212–220.
Scleroderma Lung Study. Am J Med 2011;124:459–467. 44. Khanna D, Lin CJF, Furst DE, Wagner B, Zucchetto M, Raghu G, et al.
29. Volkmann ER, Tashkin DP, Roth MD, Goldin J, Kim GHJ. Early Long-term safety and efficacy of tocilizumab in early systemic sclerosis-
radiographic progression of scleroderma: lung disease predicts interstitial lung disease: open-label extension of a phase 3 randomized
long-term mortality. Chest 2022;161:1310–1319. controlled trial. Am J Respir Crit Care Med 2022;205:674–684.
30. Tashkin DP, Elashoff R, Clements PJ, Roth MD, Furst DE, Silver RM, 45. Roofeh D, Lin CJF, Goldin J, Kim GH, Furst DE, Denton CP, et al.;
et al.; Scleroderma Lung Study Research Group. Effects of 1-year focuSSced Investigators. Tocilizumab prevents progression of early
treatment with cyclophosphamide on outcomes at 2 years in systemic sclerosis-associated interstitial lung disease. Arthritis
scleroderma lung disease. Am J Respir Crit Care Med 2007;176: Rheumatol 2021;73:1301–1310.
1026–1034. 46. Raghu G, Rochwerg B, Zhang Y, Garcia CA, Azuma A, Behr J, et al.;
31. Allison AC, Eugui EM. Mycophenolate mofetil and its mechanisms of American Thoracic Society; European Respiratory society; Japanese
action. Immunopharmacology 2000;47:85–118. Respiratory Society; Latin American Thoracic Association. An official
32. Goldin JG, Kim GHJ, Tseng CH, Volkmann E, Furst D, Clements P, et al. ATS/ERS/JRS/ALAT clinical practice guideline: treatment of idiopathic
Longitudinal changes in quantitative interstitial lung disease on pulmonary fibrosis. An update of the 2011 clinical practice guideline.
computed tomography after immunosuppression in the Scleroderma Am J Respir Crit Care Med 2015;192:e3–e19. [Published erratum
Lung Study II. Ann Am Thorac Soc 2018;15:1286–1295. appears in Am J Respir Crit Care Med 192:644.]
33. Naidu GSRSNK, Sharma SK, Adarsh MB, Dhir V, Sinha A, Dhooria S, 47. Distler O, Highland KB, Gahlemann M, Azuma A, Fischer A, Mayes MD,
et al. Effect of mycophenolate mofetil (MMF) on systemic sclerosis- et al.; SENSCIS Trial Investigators. Nintedanib for systemic sclerosis-
related interstitial lung disease with mildly impaired lung function: a associated interstitial lung disease. N Engl J Med 2019;380:
double-blind, placebo-controlled, randomized trial. Rheumatol Int 2020; 2518–2528.
40:207–216. 48. Maher TM, Mayes MD, Kreuter M, Volkmann ER, Aringer M, Castellvi I,
34. Volkmann ER, Tashkin DP, LeClair H, Roth MD, Kim G, Goldin J, et al. et al.; SENSCIS Trial Investigators. Effect of nintedanib on lung
Treatment with mycophenolate and cyclophosphamide leads to function in patients with systemic sclerosis-associated interstitial lung
clinically meaningful improvements in patient-reported outcomes in disease: further analyses of a randomized, double-blind, placebo-
scleroderma lung disease: results of Scleroderma Lung Study II. ACR controlled trial. Arthritis Rheumatol 2021;73:671–676.
Open Rheumatol 2020;2:362–370. 49. Wells AU, Flaherty KR, Brown KK, Inoue Y, Devaraj A, Richeldi L, et al.;
35. Volkmann ER, Tashkin DP, Li N, Roth MD, Khanna D, Hoffmann-Vold AM, INBUILD trial investigators. Nintedanib in patients with progressive
et al. Mycophenolate mofetil versus placebo for systemic sclerosis- fibrosing interstitial lung diseases-subgroup analyses by interstitial lung
related interstitial lung disease: an analysis of Scleroderma Lung Studies disease diagnosis in the INBUILD trial: a randomised, double-blind,
I and II. Arthritis Rheumatol 2017;69:1451–1460. placebo-controlled, parallel-group trial. Lancet Respir Med 2020;8:
36. Andrews JC, Schu €nemann HJ, Oxman AD, Pottie K, Meerpohl JJ, 453–460.
Coello PA, et al. GRADE guidelines: 15. Going from evidence to 50. Kafaja S, Clements PJ, Wilhalme H, Tseng CH, Furst DE, Kim GH,
recommendation-determinants of a recommendation’s direction and et al. Reliability and minimal clinically important differences of
strength. J Clin Epidemiol 2013;66:726–735. forced vital capacity: results from the Scleroderma Lung Studies
37. Sircar G, Goswami RP, Sircar D, Ghosh A, Ghosh P. Intravenous (SLS-I and SLS-II). Am J Respir Crit Care Med 2018;197:644–652.
cyclophosphamide vs rituximab for the treatment of early diffuse 51. Highland KB, Distler O, Kuwana M, Allanore Y, Assassi S, Azuma A,
scleroderma lung disease: open label, randomized, controlled trial. et al.; SENSCIS trial investigators. Efficacy and safety of nintedanib in
Rheumatology (Oxford) 2018;57:2106–2113. patients with systemic sclerosis-associated interstitial lung disease
38. Boonstra M, Meijs J, Dorje
e AL, Marsan NA, Schouffoer A, Ninaber MK, treated with mycophenolate: a subgroup analysis of the SENSCIS trial.
et al. Rituximab in early systemic sclerosis. RMD Open 2017;3:e000384. Lancet Respir Med 2021;9:96–106.
39. Daoussis D, Liossis SN, Tsamandas AC, Kalogeropoulou C, Kazantzi A, 52. Allanore Y, Vonk MC, Distler O, Azuma A, Mayes MD, Gahlemann M,
Sirinian C, et al. Experience with rituximab in scleroderma: results from et al.; SENSCIS-ON trial investigators. Continued treatment with
a 1-year, proof-of-principle study. Rheumatology (Oxford) 2010;49: nintedanib in patients with systemic sclerosis-associated interstitial
271–280. lung disease: data from SENSCIS-ON. Ann Rheum Dis 2022;81:
40. Ebata S, Yoshizaki A, Oba K, Kashiwabara K, Ueda K, Uemura Y, et al. 1722–1729.
Safety and efficacy of rituximab in systemic sclerosis (DESIRES): a 53. Acharya N, Sharma SK, Mishra D, Dhooria S, Dhir V, Jain S. Efficacy and
double-blind, investigator-initiated, randomised, placebo-controlled trial. safety of pirfenidone in systemic sclerosis-related interstitial lung
Lancet Rheumatol 2021;3:E489–E497. disease-a randomised controlled trial. Rheumatol Int 2020;40:703–710.
41. Khanna D, Lin CJF, Furst DE, Goldin J, Kim G, Kuwana M, et al.; 54. Khanna D, Albera C, Fischer A, Khalidi N, Raghu G, Chung L, et al. An
focuSSced investigators. Tocilizumab in systemic sclerosis: a open-label, phase II study of the safety and tolerability of pirfenidone in
randomised, double-blind, placebo-controlled, phase 3 trial. Lancet patients with scleroderma-associated interstitial lung disease: the
Respir Med 2020;8:963–974. LOTUSS trial. J Rheumatol 2016;43:1672–1679.
42. Khanna D, Denton CP, Jahreis A, van Laar JM, Frech TM, Anderson ME, 55. Khanna D, Spino C, Bernstein E, Goldin J, Tashkin D, Roth M.
et al. Safety and efficacy of subcutaneous tocilizumab in adults with Combination therapy of mycophenolate mofetil and pirfenidone vs.
systemic sclerosis (faSScinate): a phase 2, randomised, controlled trial. mycophenolate alone: results from the Scleroderma Lung Study III.
Lancet 2016;387:2630–2640. Arthritis Rheumatol (In press)

American Journal of Respiratory and Critical Care Medicine

AJRCCM Articles in Press. Published September 29, 2023 as 10.1164/rccm.202306-1113ST

Copyright © 2023 by the American Thoracic Society