Interim analysis incorporating short‐ and long‐term binary endpoints

Abstract

Designs incorporating more than one endpoint have become popular in drug development. One of such designs allows for incorporation of short‐term information in an interim analysis if the long‐term primary endpoint has not been yet observed for some of the patients. At first we consider a two‐stage design with binary endpoints allowing for futility stopping only based on conditional power under both fixed and observed effects. Design characteristics of three estimators: using primary long‐term endpoint only, short‐term endpoint only, and combining data from both are compared. For each approach, equivalent cut‐off point values for fixed and observed effect conditional power calculations can be derived resulting in the same overall power. While in trials stopping for futility the type I error rate cannot get inflated (it usually decreases), there is loss of power. In this study, we consider different scenarios, including different thresholds for conditional power, different amount of information available at the interim, different correlations and probabilities of success. We further extend the methods to adaptive designs with unblinded sample size reassessments based on conditional power with inverse normal method as the combination function. Two different futility stopping rules are considered: one based on the conditional power, and one from P‐values based on Z‐statistics of the estimators. Average sample size, probability to stop for futility and overall power of the trial are compared and the influence of the choice of weights is investigated.

Evaluating futility of a binary clinical endpoint using early read‐outs

“…First, we compare the performance of the proposed method to the standard method and the method introduced by Niewczas, Kunz, and König (referred to as method NKK). In addition, we investigated different strengths of association between both endpoints and the influence of the information fraction on the cutoff value and the “power” of a correct futility decision.…”

“…We generally found our proposal to perform better than the standard conditional power method and the conditional power method introduced by Niewczas et al The larger amount of information, obtained by incorporating short‐term endpoint data as well as historical data, resulted in a higher overall probability to stop for true futility.…”

“…In this article, we aim to improve performance in settings where the asymptotic theory, as used for the single known decision procedure for this problem, may fail. In particular, when the number of patients with complete data at the interim assessment is small, or the considered event is rare, the variance estimator of Marschner and Becker may perform poorly.…”

Improving interim decisions in randomized trials by exploiting information on short‐term endpoints and prognostic baseline covariates

“…In view of the above, recent research has focused on incorporating predictive baseline covariates and short-term measurements in interim analyses, see for example Stallard, 4 Hampson and Jennison, 5 Qian and Rosenblum, 6 Niewczas et al, 7 and Van Lancker et al 8 This brings several challenges, which make currently available methods limiting. First, the additional information accrued between the interim analysis and the final analysis may involve observations of subjects who already contributed information to the interim analysis.…”

Response to comments on Jaki et al., A proposal for a new PhD level curriculum on quantitative methods for drug development. Pharm Stat17(5):593‐606, Sep/Oct 2018., DOI: https://doi.org/10.1002/pst.1873

“…How to best train young researchers in evolving statistical methodology such as adaptive designs, dose finding, and analysis methods? Traditionally, universities play a central role in the conception and implementation of students training. Academic institutions are the most experienced to propose and run well‐designed teaching programmes.…”