Effect of Pirfenidone on cough in patients with Idiopathic Pulmonary Fibrosis

Rationale: In patients with Idiopathic Pulmonary Fibrosis (IPF) cough is a major disabling symptom and an independent predictor of disease progression. Clinical observations showed that pirfenidone, an anti-fibrotic drug, might decrease cough. In this study we aimed to objectively measure the effect of pirfenidone on cough in patients with IPF. Methods: We report the data on the first 30 patients of a prospective, observational, international multicenter study evaluating the effect of pirfenidone on cough at 4 and 12 weeks (www.clinicaltrials.gov, NCT02009293). Patients with IPF with chronic cough, about to start on pirfenidone according to regular practice, were included. The primary end-point was change in 24-hour objective cough measured with the Leicester Cough Monitor (LCM) at 12 weeks compared to baseline. Secondary endpoints were changes in Leicester Cough questionnaire (LCQ) scores and cough Visual Analogue Scale (VAS). LCM recordings were centrally analyzed with automated cough detection software. To assess the effect of pirfenidone on cough during time we used a linear mixed model. Objective cough data were log transformed as they were not normally distributed. To assess the relationship between cough data, pulmonary function tests and cough questionnaires Pearson correlations were used. Results: Demographics are shown in table 1. Pirfenidone decreased objective 24-hour cough by 35% (95% Confidence Interval (CI) 16 – 49; p = 0.002) over 12 weeks. Decrease in cough was 35%, CI 15 – 50; p = 0.003 during the day and 45%, CI 18 – 63, p = 0.005 at night. A decrease of 19% in 24-hour cough was already achieved at 4 weeks (CI 12 – 26; p = 0.03). Subjective measurements showed the same; LCQ scores improved with 2.3 points, (CI 1.2 – 3.5; p < 0.001) over 12 weeks (minimal clinical important difference for chronic cough is 1.3). Also, the cough VAS improved (21 mm, CI 11 – 31; p < 0.001). Objective 24-hour cough correlated well with subjective cough measurements at baseline; LCQ (r = 0.45, p = 0.02), VAS cough (r = 0.52, p = 0.01). No correlations were found between FVC, DLCO and objective 24-hour cough. Tab. 1: Demographics At baseline After 4 weeks After 12 weeks (n = 30) (n = 26) (n = 24) Age, years, mean (SD) 71 (6) - - Male, n (%) 21 (70) - - Smoking status, n (%) Never 7 (23) - - Former 23 (77) - - FVC % predicted, mean (SD) 79 (15) 81 (17) 80 (19) DLCO % predicted, mean (SD) 52 (14) - 52 (16) 24-hour cough, median (range) 520 (91 – 3394) 511 (65 – 1947) 359 (75 – 1746) Coughs per hour, median (range) Daytime 26 (5 – 171) 25 (4 – 106) 18 (4 – 121) Nighttime 7 (1 – 101) 4 (0 – 45) 3 (0 – 54) LCQ total, mean (SD) 12 (3) 14 (3) 15 (4) VAS cough, mean (SD) 68 (14) 49 (21) 46 (28) SD, Standard Deviation; FVC, Forced Vital Capacity; DLCO, diffusing capacity of the lung for carbon monoxide; LCQ, Leicester Cough Questionnaire; VAS, Visual AnalogueScale Conclusion: This observational study showed that in patients with Idiopathic Pulmonary Fibrosis pirfenidone reduced objective 24-hour cough counts. Subjective measurements of cough also showed this positive effect.