Introduction & Objectives Decline in Forced Vital Capacity (FVC) is predictive of mortality in IPF and is commonly used as an end-point in IPF clinical trials. FVC has not, however, been formally validated as a surrogate end-point for death or hospitalisation, making it difficult to interpret the clinical importance of differences in decline in FVC, and its use controversial. Change in FVC has also been used in other chronic lung diseases as a clinical trial end-point. The aim of this project was to determine if FVC is a surrogate end-point for mortality and/or hospitalisation in IPF, sarcoidosis and bronchiectasis trials. We also aimed to comment on the reporting of FVC in chronic lung disease trials.
Methods A systematic review was conducted. MEDLINE, EMBASE, LILACS and CENTRAL (1990-present) were searched on 25/03/2013. Reference lists and relevant journals (Jan-Apr 2013) were hand-searched, and experts in the field contacted to identify further studies. Only randomised controlled trials of drug interventions in adult participants with IPF, sarcoidosis and bronchiectasis were considered.
Results 42 studies were identified, 37 of which were included in linear regression analysis. All studies reported mortality, and 17 studies reported hospitalisation. No association was found between difference in annual decline in FVC and annual mortality or hospitalisation rate (see graph). It was frequently difficult to combine results across studies due to variation in measures reported, or due to insufficient detail in reports. 32 studies reported change in FVC: 24 studies reported absolute changes (e.g. 7.2% decline from 67.6% to 60.4%), 4 reported relative change (e.g. 7.1% decline from 70.7% to 65.7%), and one study reported both. Clarification of change type from authors was required for 16 studies.
Conclusions 1. Results of this analysis do not support the use of FVC as a surrogate end-point for mortality or hospitalisation in IPF, sarcoidosis or bronchiectasis trials. 2. Reporting of FVC in clinical trials is highly variable making synthesis of results for meta-analysis difficult. We suggest all trials make available baseline, final and change in FVC, both unadjusted and percent predicted, and as relative (% change), at least in an online appendix.