The primary endpoint in oncology is usually overall survival, where differences between therapies may only be observable after many years. To avoid withholding of a promising therapy, preliminary approval based on a surrogate endpoint is possible. The approval can be confirmed later by assessing overall survival within the same study. In these trials, the correlation between surrogate endpoint and overall survival has to be taken into account for sample size calculation and analysis. For a binary surrogate endpoint, this relation can be modeled by means of the responder stratified exponential survival (RSES) model proposed by Xia, Cui, and Yang (2014). We derive properties of the model and confidence intervals based on Maximum Likelihood estimators. Furthermore, we present an approximate and an exact test for survival difference. Type I error rate, power, and required sample size for both newly developed tests are determined exactly. These characteristics are compared to those of the logrank test. We show that the exact test performs best. The power of the logrank test is considerably lower in some situations. We conclude that the logrank test should not be used within the RSES model. The proposed method for sample size calculation works well. The interpretability of our proposed methods is discussed.
翻译:肿瘤学中的首要终点通常是总体生存, 治疗方法之间的差别只能在多年后才能观察到。 为了避免预留有希望的治疗方法, 初步批准是可能的, 基于替代端点的初步批准是可能的。 批准可以稍后在同一研究中评估总体生存情况。 在这些试验中, 样本大小的计算和分析必须考虑到替代端点与总体生存的相互关系。 对于二进代端点, 对于二进代端点, 可以通过 Xia、 Cui 和 Yang (2014) 提议的响应器分分批指数生存模型( RSES) 模型来模拟这种关系。 我们根据最大相似度估计值来得出模型和信任间隔的属性。 此外, 我们提出了一个大致和精确的生存差异测试。 类型I 误差率、 功率和新开发的测试要求的样本大小是准确的。 这些特征与日志测试的特征比较。 我们证明精确的测试效果是最佳的。 日志测试在某些情况下能力要低得多。 我们的结论是, 圆点测试不应该在 RSES 模型中使用。 拟议的样本大小计算方法是很好的解释。