To the Editor Based on 2 oxaliplatin-based adjuvant trials, Sinicrope and colleagues¹ recently reported an association between survival after recurrence (SAR) and both DNA mismatch repair (MMR) status and BRAF mutation in patients with stage III colon carcinomas (CC). Given that MMR status and BRAF mutation are strongly associated, they analyzed MMR and BRAF as a combined variable. Among dMMR CC, BRAF-mutated tumors had significantly poorer SAR compared with BRAF wild-type tumors (adjusted hazard ratio [HR], 2.70; 95% CI, 1.23-5.93; P = .01). Some caution is warranted when analyzing these results. When using SAR as an end point, patient treatment at recurrence has an effect on survival but these data were not available. Also, patient numbers were relatively small in the different groups resulting in low statistical power: 33 BRAF-mutated dMMR and 40 BRAF wild-type dMMR. Adjusted median SAR were incalculable owing to small sample size and lack of robustness of the model. In addition, results were adjusted on variables collected at CC diagnosis and not at relapse (ie, performance score), which is not relevant.

致编辑基于 2 项基于奥沙利铂的辅助试验，Sinicrope 及其同事¹最近报告了III 期结肠癌 (CC) 患者的复发后生存 (SAR) 与 DNA 错配修复 (MMR) 状态和BRAF突变之间的关联。鉴于 MMR 状态和BRAF突变密切相关，他们将 MMR 和BRAF作为一个组合变量进行了分析。在 dMMR CC 中，BRAF突变肿瘤的 SAR 显着低于BRAF野生型肿瘤（调整后的风险比 [HR]，2.70；95% CI，1.23-5.93；P = .01)。在分析这些结果时需要谨慎。当使用 SAR 作为终点时，复发时的患者治疗对生存率有影响，但这些数据不可用。此外，不同组中的患者人数相对较少，导致统计功效低：33 名BRAF突变的 dMMR 和 40 名BRAF野生型 dMMR。由于样本量小且模型缺乏稳健性，调整后的中值 SAR 无法计算。此外，根据 CC 诊断时收集的变量而不是复发时收集的变量（即性能评分）调整结果，这不相关。