Background: At present, no approved pharmacotherapies are available for unclassifiable interstitial lung disease (ILD), which is characterised by progressive fibrosis of the lung. We aimed to assess the efficacy and safety of pirfenidone in patients with progressive fibrosing unclassifiable ILD., Methods: We did a multicentre, double-blind, randomised, placebo-controlled phase 2 trial at 70 centres in Australia, Belgium, Canada, Czech Republic, Denmark, Germany, Greece, Ireland, Israel, Italy, Poland, Portugal, Spain, and the UK. Eligible patients (aged ≥18-85 years) had progressive fibrosing unclassifiable ILD, a percent predicted forced vital capacity (FVC) of 45% or higher and percent predicted carbon monoxide diffusing capacity (DLco) of 30% or higher, more than 10% fibrosis on high-resolution CT, and a high-resolution CT from the previous 12 months. Patients were randomly assigned (1:1) to 2403 mg oral pirfenidone daily or placebo using a central validated interactive voice or web-based response system, stratified by concomitant mycophenolate mofetil use and presence or absence of interstitial pneumonia with autoimmune features. Investigators, site personnel, and patients were masked to treatment assignment. The primary endpoint was mean predicted change in FVC from baseline over 24 weeks, measured by daily home spirometry. Secondary endpoints were change in FVC measured by site spirometry, proportion of patients who had a more than 5% or more than 10% absolute or relative decline in percent predicted FVC measured by clinic-based spirometry, change in percent predicted DLco, change in 6-min walk distance (6MWD), change in University of California San Diego-Shortness of Breath Questionnaire (UCSD-SOBQ) score, change in Leicester Cough Questionnaire score, change in cough visual analogue scale, and changes in total and subscores of the St George's Respiratory Questionnaire (SGRQ), all of which were compared with baseline. Additional secondary endpoints included proportion of patients who had non-elective hospitalisation (respiratory and all-cause) and acute exacerbations, and progression-free survival. Efficacy was analysed in the intention-to-treat (ITT) population, which included all randomly assigned patients. Safety was assessed in the safety analysis set, which included all randomly assigned patients who received at least one dose of study drug. This trial is registered with ClinicalTrials.gov, NCT03099187, and is no longer recruiting., Findings: Between May 15, 2017, and June 5, 2018, 253 patients were randomly assigned to receive 2403 mg pirfenidone (n=127) or placebo (n=126) and were included in the ITT analysis set. Analysis of the primary endpoint was affected by intraindividual variability in home spirometry values, which prevented application of the prespecified statistical model. Over 24 weeks, predicted median change in FVC measured by home spirometry was -87·7 mL (Q1-Q3 -338·1 to 148·6) in the pirfenidone group versus -157·1 mL (-370·9 to 70·1) in the placebo group. Over 24 weeks, predicted mean change in FVC measured by site spirometry was lower in patients given pirfenidone than placebo (treatment difference 95·3 mL [95% CI 35·9 to 154·6], p=0·002). Compared with the placebo group, patients in the pirfenidone group were less likely to have a decline in FVC of more than 5% (odds ratio [OR] 0·42 [95% CI 0·25 to 0·69], p=0·001) or more than 10% (OR 0·44 [0·23 to 0·84], p=0·011). At week 24, mean change in DLco from baseline was -0·7% (SD 7·1) for the pirfenidone group and -2·5% (8·8) for the placebo group, and mean change in 6MWD from baseline was -2·0 m (68·1) for the pirfenidone group and -26·7 m (79·3) for the placebo group. Changes from baseline in UCSD-SOBQ, Leicester Cough Questionnaire score, cough visual analogue scale, and SGRQ scores were similar between the pirfenidone and placebo groups at week 24. Analysis of acute exacerbations, hospital admissions, and time to death from respiratory causes during the study yielded no meaningful results due to a small number of events. No differences in progression-free survival were identified between the pirfenidone and placebo groups, irrespective of the definition of progression-free survival used. Treatment-emergent adverse events were reported in 120 (94%) of 127 patients in the pirfenidone group and 101 (81%) of 124 patients in the placebo group. Serious treatment-emergent adverse events were reported in 18 (14%) patients in the pirfenidone group and 20 (16%) patients in the placebo group. The most common treatment-related treatment-emergent adverse events were gastrointestinal disorders (60 [47%] in the pirfenidone group vs 32 [26%] in the placebo group), fatigue (16 [13%] vs 12 [10%]), and rash (13 [10%] vs nine [7%])., Interpretation: Although the planned statistical model could not be applied to the primary endpoint data, analysis of key secondary endpoints suggests that patients with progressive fibrosing unclassifiable ILD could benefit from pirfenidone treatment, which has an acceptable safety and tolerability profile. These findings support further investigation of pirfenidone as an effective treatment for patients with progressive fibrotic unclassifiable ILD., Funding: F Hoffmann-La Roche., (Copyright © 2020 Elsevier Ltd. All rights reserved.)