A comparison of estimation methods adjusting for selection bias in adaptive enrichment designs with time-to-event endpoints.

Adaptive enrichment designs in clinical trials have been developed to enhance drug developments. They permit, at interim analyses during the trial, to select the sub-populations that benefits the most from the treatment. Because of this selection, the naive maximum likelihood estimation of the treatment effect, commonly used in classical randomized controlled trials, is biased. In the literature, several methods have been proposed to obtain a better estimation of the treatments' effects in such contexts. To date, most of the works have focused on normally distributed endpoints, and some estimators have been proposed for time-to-event endpoints but they have not all been compared side-by-side. In this work, we conduct an extensive simulation study, inspired by a real case-study in heart failure, to compare the maximum-likelihood estimator (MLE) with an unbiased estimator, shrinkage estimators, and bias-adjusted estimators for the estimation of the treatment effect with time-to-event data. The performances of the estimators are evaluated in terms of bias, variance, and mean squared error. Based on the results, along with the MLE, we recommend to provide the unbiased estimator and the single-iteration bias-adjusted estimator: the former completely eradicates the selection bias, but is highly variable with respect to a naive estimator; the latter is less biased than the MLE estimator and only slightly more variable.

Efficacy and safety of trimetazidine after percutaneous coronary intervention (ATPCI): a randomised, double-blind, placebo-controlled trial.

BACKGROUND: Angina might persist or reoccur despite successful revascularisation with percutaneous coronary intervention (PCI) and antianginal therapy. Additionally, PCI in stable patients has not been shown to improve survival compared with optimal medical therapy. Trimetazidine is an antianginal agent that improves energy metabolism of the ischaemic myocardium and might improve outcomes and symptoms of patients who recently had a PCI. In this study, we aimed to assess the long-term potential benefits and safety of trimetazidine added to standard evidence-based medical treatment in patients who had a recent successful PCI. METHODS: We did a randomised, double-blind, placebo-controlled, event-driven trial of trimetazidine added to standard background therapy in patients who had undergone successful PCI at 365 centres in 27 countries across Europe, South America, Asia, and north Africa. Eligible patients were aged 21-85 years and had had either elective PCI for stable angina or urgent PCI for unstable angina or non-ST segment elevation myocardial infarction less than 30 days before randomisation. Patients were randomly assigned by an interactive web response system to oral trimetazidine 35 mg modified-release twice daily or matching placebo. Participants, study investigators, and all study staff were masked to treatment allocation. The primary efficacy endpoint was a composite of cardiac death; hospital admission for a cardiac event; recurrence or persistence of angina requiring an addition, switch, or increase of the dose of at least one antianginal drug; or recurrence or persistence of angina requiring a coronary angiography. Efficacy analyses were done according to the intention-to-treat principle. Safety was assessed in all patients who had at least one dose of study drug. This study is registered with the EU Clinical Trials Register (EudraCT 2010-022134-89). FINDINGS: From Sept 17, 2014, to June 15, 2016, 6007 patients were enrolled and randomly assigned to receive either trimetazidine (n=2998) or placebo (n=3009). After a median follow-up of 47·5 months (IQR 42·3-53·3), incidence of primary endpoint events was not significantly different between the trimetazidine group (700 [23·3%] patients) and the placebo group (714 [23·7%]; hazard ratio 0·98 [95% CI 0·88-1·09], p=0·73). When analysed individually, there were no significant differences in the incidence of the components of the primary endpoint between the treatment groups. Similar results were obtained when patients were categorised according to whether they had an elective or urgent PCI. 1219 (40·9%) of 2983 patients in the trimetazidine group and 1230 (41·1%) of 2990 patients in the placebo group had serious treatment-emergent adverse events. Frequencies of adverse events of interest were similar between the groups. INTERPRETATION: Our results show that the routine use of oral trimetazidine 35 mg twice daily over several years in patients receiving optimal medical therapy, after successful PCI, does not influence the recurrence of angina or the outcome; these findings should be taken into account when considering the place of trimetazidine in clinical practice. However, the long-term prescription of this treatment does not appear to be associated with any statistically significant safety concerns in the population studied. FUNDING: Servier.