Aligning sample size calculations with estimands in clinical trials with time-to-event outcomes

The ICH E9(R1) guidance recommended a framework to align planning, design, conduct, analysis, and interpretation of any clincial trial with its objective and estimand. How to handle intercurrent events (ICEs) is one of the five attributes of an estimand and sample size calculation is a key step in the trial planning and design. Therefore, sample size calculation should be aligned with the estimand and, in particular, with how the ICEs are handled. ICH E9(R1) summarized five strategies for handling ICEs, and five approaches have been proposed in the literature for sample size calculation when planning trials with quantitative and binary outcomes. In this paper, we discuss how to apply the five strategies to deal with ICEs in clinical trials with time-to-event outcomes and propose five approaches for sample size calculation that are aligned with the five strategies, respectively.