In Reply We thank these authors for their interest in our article. It is true that this definition of 20% or more concurrent overlap might have led to an apparent decrease in capecitabine-proton pump inhibitor (PPI) effect, but we were more willing to accept finding a PPI effect and underestimating the interaction than to miss the interaction entirely, for 2 reasons. First, we do not believe that taking a PPI for 3 days (<20%) out of 14 days of capecitabine administration would have clinically meaningful effects. Furthermore, for brevity in the original article, we did not specify that almost all patients identified as having less than 20% concurrent PPI and capecitabine exposure had in fact never taken a PPI. Second, we have successfully used this definition of concurrent use for other PPI-focused oncology studies^1- 3—for continuity, we continued this definition in this study. We found similar concerning results with patients receiving concurrent PPIs and adjuvant capecitabine in patients with colon cancer, for whom 5-year recurrence-free survival was 74% vs 83% in those not receiving a PPI (P = .03).³

作为回复，我们感谢这些作者对我们文章的兴趣。确实，这种 20% 或更多同时重叠的定义可能导致卡培他滨质子泵抑制剂 (PPI) 效应明显降低，但我们更愿意接受发现 PPI 效应并低估相互作用而不是错过完全交互，原因有两个。首先，我们认为在卡培他滨给药的 14 天内服用 PPI 3 天（<20%）不会产生有临床意义的效果。此外，为简洁起见，在原始文章中，我们没有具体说明几乎所有被确定为同时使用低于 20% 的 PPI 和卡培他滨暴露的患者实际上从未服用过 PPI。其次，我们已经成功地将这种同时使用的定义用于其他以 PPI 为重点的肿瘤学研究^{1 -} 3——为了连续性，我们在本研究中继续这个定义。我们发现结肠癌患者同时接受 PPI 和辅助卡培他滨的患者的相关结果相似，5 年无复发生存率为 74%，而未接受 PPI 的患者为 83% ( P  = .03)。³