PURPOSE: Premature discontinuation from clinical trials is often due to adverse events or the perception of failure to improve. Such discontinuations, without follow-up information, may bias results against effective therapies.
METHODS: We reviewed mortality rates in a 6-month, randomized clinical trial of tiotropium 18 mcg daily vs. usual care (except inhaled anticholinergics) in which tiotropium was shown to decrease COPD exacerbations. Patients were followed for the entire 6 months even if trial medication was prematurely discontinued. Incidence rates (IR) of fatal events = # patients with an event/total patient exposure to trial drug. IRs were calculated for events occurring from: a)randomization to end of trial (0-ET), b)randomization to end of trial drug (0-ED), c)end of trial drug to end of trial (ED-ET).
RESULTS: 1,829 patients participated: mean FEV1=1.04 L [36% predicted], age=68 years, 99% men, approximately 90% had information for the full 6 months. Trial medication was prematurely stopped by 16% (tiotropium) and 27% (placebo). For the entire cohort, the number of fatal events were: all cause=62, cardiac=16, lower respiratory=15, exacerbations=5. IRs of fatal events (per 100 patient-years) increased from 1.9 (0-ED) to 23.0 (ED-ET) with tiotropium and from 1.8 (0-ED) to 19.0 (ED-ET) with placebo. Corresponding increases for cardiac fatal events were 0.7 to 2.8 (tiotropium) and 0.5 to 6.2 (placebo), and for lower respiratory fatal events were 0.7 to 2.8 (tiotropium) and 0.8 to 5.4 (placebo). Rate ratios (tiotropium/placebo) decreased from 1.4 (95%CI 0.2, 8.2) to 0.5 (95%CI 0.1, 1.7) for cardiac fatal events and from 0.9 (95%CI 0.2, 4.5) to 0.5 (95%CI 0.1, 2.0) for lower respiratory fatal events. All fatal exacerbations (n=5) occurred with placebo (4 after trial drug discontinuation).
CONCLUSION: Patients with tiotropium were more likely to continue in clinical trials than those with placebo. Higher IRs of fatal events occurred following premature discontinuation of study medication.
CLINICAL IMPLICATIONS: Failure to account for follow-up in prematurely discontinued patients in trials with efficacious interventions may lead to false conclusions regarding efficacy and safety.
DISCLOSURE: Steven Kesten, Grant monies (from industry related sources) Dr. Niewoehner has received grants from Boehringer Ingelheim and GlaxoSmithKline. Dr. Plautz, Dr. Piquette, and Dr. Habib have received grants from Boehringer Ingelheim; Employee Steven Kesten is an employee of Boehringer Ingelheim; Consultant fee, speaker bureau, advisory committee, etc. Dr. Niewoehner has received honoraria or consultation fees from Boehringer Ingelheim, AstraZeneca, Adams Respiratory Therapeutics, Sanofi Aventis, and Pfizer.