Abstract

The log-rank test is most powerful under proportional hazards (PH). In practice, non-PH patterns are often observed in clinical trials, such as in immuno-oncology; therefore, alternative methods are needed to restore the efficiency of statistical testing. Three categories of testing methods were evaluated, including weighted log-rank tests, Kaplan–Meier curve-based tests (including weighted Kaplan–Meier and restricted mean survival time), and combination tests (including Breslow test, Lee’s combo test, and MaxCombo test). Nine scenarios representing the PH and various non-PH patterns were simulated. The power, Type I error, and effect estimate of each method were compared. In general, all tests control Type I error well. There is not a single most powerful test across all scenarios. In the absence of prior knowledge regarding the underlying or non-PH patterns, the MaxCombo test is relatively robust across patterns. Since the treatment effect changes over time under non-PH, the overall profile of the treatment effect may not be represented comprehensively based on a single measure. Thus, multiple measures of the treatment effect should be prespecified as sensitivity analyses to describe the totality of the data. Supplementary materials for this article are available online.

摘要

对数秩检验在比例风险（PH）下最有效。在实践中，通常在临床试验（例如免疫肿瘤学）中观察到非PH模式。因此，需要其他方法来恢复统计测试的效率。评估了三类测试方法，包括加权对数秩检验，基于Kaplan-Meier曲线的检验（包括加权Kaplan-Meier和受限平均生存时间）以及组合检验（包括Breslow检验，Lee组合检验和MaxCombo检验） ）。模拟了代表PH和各种非PH模式的九种情况。比较了每种方法的功效，I类错误和效果估计。通常，所有测试都能很好地控制Type I错误。在所有情况下都没有一个最强大的测试。在没有有关基础或非PH模式的先验知识的情况下，MaxCombo测试在各种模式下都相对可靠。由于在非PH下治疗效果会随时间变化，因此可能无法基于单个度量全面代表治疗效果的整体情况。因此，应预先指定多种治疗效果指标作为敏感性分析，以描述数据的整体。可在线获得本文的补充材料。应该预先指定多种治疗效果的度量作为敏感性分析，以描述数据的整体。可在线获得本文的补充材料。应该预先指定多种治疗效果的度量作为敏感性分析，以描述数据的整体。可在线获得本文的补充材料。