Abstract
Background: Clinical trials in IPF are not designed to estimate long-term survival. This analysis used a survival model to predict life expectancy for patients with IPF receiving pirfenidone or BSC.
Methods: Life expectancy was estimated by the area under the curve of parametric survival distributions fit to Kaplan-Meier survival data from clinical studies and IPF registries. Kaplan-Meier survival data for pirfenidone were derived from clinical studies (CAPACITY, ASCEND, RECAP). Kaplan-Meier survival data for BSC were obtained from two independent registries of patients with IPF: the Inova Fairfax Hospital database (n=815) and the National Jewish Health Interstitial Lung Disease (NJH-ILD) database (n=321). The best-fitting distributions were chosen by statistical consideration, visual inspection of the fitted curve and by clinical interpretation. To account for differences between patients enrolled in the clinical trials and the registries, covariate adjustment using propensity scores was used.
Results: Mean life expectancy (95% confidence intervals) was calculated as: 8.7 years (7.7, 10.2) with pirfenidone; 5.9 years (5.1, 6.9) with BSC (Inova); and 6.1 years (5.7, 6.5) with BSC (NJH-ILD). Therefore, pirfenidone improved life expectancy relative to BSC by 2.8 years and 2.6 years as measured by the Inova and NJH-ILD registries, respectively.
Conclusions: The survival model suggests that pirfenidone significantly improves life expectancy compared with BSC by almost 3 years in patients with IPF. Although these findings are based on cross-trial comparisons, they provide support for pirfenidone as an effective treatment option for IPF.
- Copyright ©the authors 2016