A post-hoc PREDMETH analysis presented at ERS 2026 links lower baseline FVC and shorter disease duration with greater FVC gains on first-line therapy.
Lower baseline forced vital capacity (FVC) and shorter disease duration were associated with a stronger treatment response to first-line prednisone or methotrexate in patients with
The randomized PREDMETH trial, published in the New England Journal of Medicine in 2025, found methotrexate noninferior to prednisone as first-line treatment for pulmonary sarcoidosis, with mean FVC gains of 6.11 and 6.75 percentage points at week 24, respectively (adjusted difference, −1.17 percentage points; 95% CI, −4.27 to 1.93).² A recent Delphi consensus of international sarcoidosis experts endorsed an improvement of 10% predicted or more in FVC as a clinical trial endpoint.³ Whether study inclusion criteria influence the likelihood of reaching this threshold had not been established.¹
"Patients with a lower FVC were likely to respond better than patients with a higher FVC," said Vivienne Kahlmann, a PhD candidate at Erasmus MC and a pulmonologist specializing in interstitial lung diseases at Leiden University Medical Center, in an interview with HCPLive. "This has implications for future trial design. When we design trials, we may use baseline FVC, for example, as a stratification factor to divide patients equally across both groups."
PREDMETH randomized 138 patients with treatment-naive pulmonary sarcoidosis at centers across the Netherlands to receive prednisone (n = 70) or methotrexate (n = 68).² In the post-hoc analysis, Kahlmann and colleagues evaluated predictors of FVC change at week 24 using univariate and multivariate linear regression, with data available for 135 patients.¹ Responder analyses stratified patients by baseline FVC at the median of 77% predicted and defined response as improvements of ≥ 5% and ≥ 10%.¹
In univariate analysis, baseline FVC, age, and disease duration were each associated with FVC change at week 24.¹ Treatment assignment to prednisone or methotrexate showed no association with FVC change.¹ In the multivariate model, greater baseline FVC % predicted (P <.001) and longer disease duration in years (P =.022) were both associated with smaller FVC improvement, and age did not reach significance (P =.138).¹
Patients with baseline FVC < 77% predicted were more likely to reach the 5% response threshold (odds ratio [OR], 2.80; P =.004) and the 10% threshold (OR, 2.34; P =.021) compared with patients at or above the median.¹ Kahlmann noted many Delphi panelists also supported a 5% threshold, with consensus ultimately forming around 10%.
"[When] looking at endpoints in clinical trials, it's better not to look binary at whether patients reach the 10% threshold or not, but more at continuous data," said Kahlmann. Her team is now examining optimal FVC cutoff points alongside secondary measures, including diffusing capacity and patient-reported outcomes.
Kahlmann described the disease duration signal as modest. She said the association was small and is not yet relevant for clinical decision-making. Identifying patients most likely to respond to a given therapy remains the broader goal of ongoing work in the Dutch cohort.
Colleagues are evaluating blood biomarkers as potential predictors of response, Kahlmann said. For methotrexate, the team is measuring blood drug levels to assess any link with treatment response. For prednisone, investigators are studying glucocorticoid receptor expression, hypothesizing patients with greater receptor expression respond better to corticosteroid therapy.
Kahlmann recently received a grant from the Foundation for Sarcoidosis Research to investigate whether baseline CT imaging phenotypes predict treatment response. The work builds on a prior Delphi consensus defining 7 radiological phenotypes of sarcoidosis, and her team is assessing whether response differs across these phenotypes. She identified biomarkers as a major unmet need in sarcoidosis, with research underway internationally.
“There is a lot of research going on [about] biomarkers in sarcoidosis because I think that's also a major unmet need yet, and hopefully we will define in the future better biomarkers,” Kahlmann said.
Kahlmann has no reported disclosures.
References
