Search

Background Many clinical trials focus on the comparison of the treatment effect between two or more groups concerning a rarely occurring event. In this situation, showing a relevant effect with an acceptable power requires the observation of a large number of patients over a long period of time. For feasibility issues, it is therefore often considered to include several event types of interest, non-fatal or fatal, and to combine them within a composite endpoint. Commonly, a composite endpoint is analyzed with standard survival analysis techniques by assessing the time to the first occurring event. This approach neglects that an individual may experience more than one event which leads to a loss of information. As an alternative, composite endpoints could be analyzed by models for recurrent events. There exists a number of such models, e.g. regression models based on count data or Cox-based models such as the approaches of Andersen and Gill, Prentice, Williams and Peterson or, Wei, Lin and Weissfeld. Although some of the methods were already compared within the literature there exists no systematic investigation for the special requirements regarding composite endpoints. Methods Within this work a simulation-based comparison of recurrent event models applied to composite endpoints is provided for different realistic clinical trial scenarios. Results We demonstrate that the Andersen-Gill model and the Prentice- Williams-Petersen models show similar results under various data scenarios whereas the Wei-Lin-Weissfeld model delivers effect estimators which can considerably deviate under commonly met data scenarios. Conclusion Based on the conducted simulation study, this paper helps to understand the pros and cons of the investigated methods in the context of composite endpoints and provides therefore recommendations for an adequate statistical analysis strategy and a meaningful interpretation of results. Electronic supplementary material The online version of this article (10.1186/s12874-017-0462-x) contains supplementary material, which is available to authorized users.

In clinical trials, the choice of an adequate primary endpoint is often difficult. Besides its clinical relevance, the endpoint must be measurable within reasonable time and must allow differentiating between the treatments. Often, the most relevant endpoint is ‘’time-to-death,” but if the overall survival prognosis is good, only a few deaths are observed during the study duration. A possible solution is to use surrogate endpoints instead. However, various examples from the literature demonstrate that surrogates do not always perform as intended. Sometimes, the surrogate effect is smaller than for the original endpoint, or the latter shows a higher effect than anticipated so using the surrogate is not reasonable. In this work, different adaptive design strategies for two candidate endpoints are proposed to solve these problems. The idea is to base the efficacy proof on the significance of at least one endpoint. At an interim analysis, both candidates are evaluated. If it is not possible to stop th...

The resulting treatment effect of a composite endpoint alone is often difficult to interpret as the individual components do not necessarily contribute the same amounts to this net effect. As discussed before, current guidelines therefore recommend to always analyze the components of composite endpoints separately. The current well-established practice is to provide descriptive analyses of the components and the composite in addition to the confirmatory analysis of the composite. However, there exist a number of different methods to evaluate the single components in a descriptive manner. In this chapter, the most common approaches for a descriptive analysis of the individual components will be presented and discussed. Moreover, we will deduce recommendations for a meaningful presentation and interpretation of the component-specific results.

This chapter provides exemplary source code written in the software R implementing different approaches outlined in Part III of this book.

In this chapter, we discuss several original clinical trials for composite binary endpoints along with their designs, results, and possible limitations. The focus of this chapter is to propose and apply alternative planning and analysis strategies which might help to improve the performance of the design or to ease interpretation of results.

Many authors have proposed different approaches to combine multiple endpoints in a univariate outcome measure in the literature. In case of binary or time-to-event variables, composite endpoints, which combine several event types within a single event or time-to-first-event analysis are often used to assess the overall treatment effect. A main drawback of this approach is that the interpretation of the composite effect can be difficult as a negative effect in one component can be masked by a positive effect in another. Recently, some authors proposed more general approaches based on a priority ranking of outcomes, which moreover allow to combine outcome variables of different scale levels. These new combined effect measures assign a higher impact to more important endpoints, which is meant to simplify the interpretation of results. Whereas statistical tests and models for binary and time-to-event variables are well understood, the latter methods have not been investigated in detail so far. In this paper, we will investigate the statistical properties of prioritized combined outcome measures. We will perform a systematical comparison to standard composite measures, such as the all-cause hazard ratio in case of time-to-event variables or the absolute rate difference in case of binary variables, to derive recommendations for different clinical trial scenarios. We will discuss extensions and modifications of the new effect measures, which simplify the clinical interpretation. Moreover, we propose a new method on how to combine the classical composite approach with a priority ranking of outcomes using a multiple testing strategy based on the closed test procedure.

Rationale for adaptive designs with multiple endpoints In clinical trials, the efficacy of a new treatment is usually assessed by a single primary endpoint [1,2]. Other important aspects concerning the new treatment, such as adverse events or quality of life, are taken into account by the definition of secondary endpoints. Usually, only the primary endpoint is analyzed confirmatorily with a statistical test whereas the secondary endpoints are evaluated with descriptive statistical methods. A descriptive analysis, however, does not result in a definite proof and therefore provides only limited additional information. The current research on multiple testing strategies offers many opportunities to incorporate secondary endpoints in the confirmatory analysis of a trial, for example, by a predefined hierarchical testing procedure or enhanced gatekeeping strategies [3–6]. While the nature of secondary endpoints is that they are only of subordinate clinical importance, there also exist many clinical trial applications where the efficacy of a new treatment cannot be adequately described by only one primary endpoint. The EMEA Guideline Points to Consider on Multiplicity Issues states that ‘If, however, a single variable is not sufficient to capture the range of clinically relevant treatment benefits, the use of more than one primary variable may become necessary’ [2]. In some situations, the efficacy claim is therefore based on the significance of several primary endpoints, also referred to as coprimary endpoints. However, even if several primary endpoints are considered, it is not always necessary to demonstrate a significant and relevant effect in all primary endpoints under investigation. Another option is to base the efficacy claim of the new intervention on the significance of at least one endpoint out of a predefined set of primary endpoints considered as clinically relevant. Whenever several endpoints are included in the confirmatory analysis of a clinical trial, a multiple testing problem arises. In order to control the probability to falsely reject at least one of the null hypotheses under investigation at a predefined global significance level α, the local significance levels corresponding to the individual test hypotheses have to be adjusted by an adequate multiple testing procedure. There exist a variety of multiple testing procedures in the statistical literature for many different possible applications [3–8]. When several endpoints are assessed within a multiple test problem, the power of the trial depends on the expected effects of all endpoints and on the correlations between them. The calculation of an adequate sample size therefore is a particular challenge, as the number of required parameter assumptions to be estimated in the planning stage is much larger as compared with a single endpoint test problem. For this reason, adaptive study designs which allow reacting flexibly to these uncertainties in the planning stage play a major role for clinical trials with multiple endpoints.

Composite endpoints combine several events of interest within a single variable. These are often time-to-first-event data, which are analyzed via survival analysis techniques. To demonstrate the significance of an overall clinical benefit, it is sufficient to assess the test problem formulated for the composite. However, the effect observed for the composite does not necessarily reflect the effects for the components. Therefore, it would be desirable that the sample size for clinical trials using composite endpoints provides enough power not only to detect a clinically relevant superiority for the composite but also to address the components in an adequate way. The single components of a composite endpoint assessed as time-to-first-event define competing risks. We consider multiple test problems based on the cause-specific hazards of competing events to address the problem of analyzing both a composite endpoint and its components. Thereby, we use sequentially rejective test procedures to reduce the power loss to a minimum. We show how to calculate the sample size for the given multiple test problem by using a simply applicable simulation tool in SAS. Our ideas are illustrated by two clinical study examples. Copyright © 2013 John Wiley & Sons, Ltd.