Pulmonary function and quantitative high-resolution computed tomography measures remained stable over 24 months among patients with rheumatoid arthritis–associated interstitial lung disease, despite an 8% mortality rate.

In the prospective, multicenter study from the BERTHA cohort, researchers described the natural history of predominantly mild rheumatoid arthritis–associated interstitial lung disease (RA-ILD) identified in rheumatology clinics and evaluated conventional and densitometry-based imaging markers over 2 years.

Study Design and Population

The researchers conducted a prospective cohort study across three Brazilian centers, enrolling consecutive adult patients meeting 1987 or 2010 American College of Rheumatology criteria for RA with ILD confirmed on high-resolution computed tomography (CT). The participants were assessed at baseline and at 6, 12, 18, and 24 months. Pulmonary function tests, imaging, 6-minute walk testing, and quality-of-life assessment were performed at baseline, 12, and 24 months.

Among the 105 enrolled patients, 95 had complete baseline data (mean age = 63 years; 81% female). Pulmonary fibrosis, defined by traction bronchiectasis and/or honeycombing on high-resolution CT, was present in 80% of the participants, although just 20% of them were classified as definite or probable usual interstitial pneumonia, with moderate interobserver agreement.

At baseline, the patients with pulmonary fibrosis had lower forced vital capacity (FVC) percent predicted (75% vs 86%) and shorter 6-minute walk distance (361 m vs 423 m) compared with those without fibrosis. The overall cohort profile was consistent with mostly mild RA-ILD.

Functional and Imaging Outcomes

At 24 months, 71% (n = 67) of the patients underwent repeat pulmonary function testing and high-resolution CT. Mean FVC percent predicted remained unchanged (76.4% at baseline vs 76.8% at 24 months). Forced expiratory volume in 1 second was also stable. Stratified analyses by pulmonary fibrosis status showed no statistically significant differences over time.

Qualitative imaging assessment at 24 months showed 12% of the patients improved, 69% remained stable, and 19% worsened. The proportion classified as usual interstitial pneumonia increased numerically from 21% to 28%, largely reflecting reclassification of indeterminate cases rather than new fibrotic involvement.

Densitometry-based quantitative CT measures—including total lung volume, percentage of high-attenuation areas (−600 to −250 Hounsfield units), mean lung density, skewness, and kurtosis—showed no meaningful longitudinal change.

Circulating biomarkers, including matrix metalloproteinase-7, surfactant protein D, chemokine ligand 8, neopterin, and tetrahydrobiopterin, remained stable over time. None discriminated between patients with and without pulmonary fibrosis at baseline.

Mortality and Sensitivity Analyses

Eight patients (8%) died over 24 months; no lung transplantations occurred. Deaths weren't preceded by measurable decline in FVC or quantitative imaging parameters. At least two deaths were unrelated to pulmonary causes, and approximately 50% of them were associated with pulmonary infections.

Sensitivity analyses indicated that a 17% relative reduction in FVC among patients lost to follow-up would have been required to shift overall results toward statistically significant decline.

Interpretation and Limitations

Mortality occurring without preceding FVC decline or imaging progression “challenges the conventional paradigm in ILD where functional decline—particularly loss of FVC—typically precedes death,” wrote lead study author Leticia Kawano-Dourado, of the HCor Research Institute at the Hospital do Coracao as well as the Pulmonary Division at the Heart Institute (InCor) at the Universidade de Sao Paulo in Brazil, and colleagues,. They noted the findings should be interpreted as hypothesis-generating.

Strengths included the study's prospective design, centralized qualitative imaging review, and quantitative imaging analysis. Limitations included its modest sample size, missing follow-up data largely related to the COVID-19 pandemic, and limited mortality events, constraining definitive conclusions about prognostic performance.

The study received research funding from Bristol-Myers Squibb and Boehringer Ingelheim. Leticia Kawano-Dourado reported receiving research funding from Bristol-Myers Squibb and Boehringer Ingelheim. Co–study author Karina Bonfiglioli reported consulting and educational fees from AbbVie, Fresenius Kabi, Johnson & Johnson, Pfizer, and UCB Pharma. Co–study author Licia Maria Henrique da Mota reported a research grant from PANLAR and consulting and educational fees from AbbVie, Fresenius Kabi, Johnson & Johnson, Boehringer Ingelheim, GSK, Livvs, Lilly, Novartis, Pfizer, Roche, Sandoz, and UCB Pharma. Co–study author Alisson Pugliesi reported consulting and educational fees from AbbVie, Fresenius Kabi, Johnson & Johnson, Boehringer Ingelheim, AstraZeneca, Lilly, Pfizer, Sandoz, and UCB Pharma. Co–study author Luis F. Marqueze reported lecture and educational fees from AbbVie, Janssen, Pfizer, Novartis, and UCB Pharma. Co–study author Luciana Lourençoni reported speaker fees from AbbVie and Janssen. The study authors reported no other competing interests.

Source: BMJ Open Respiratory Research