Immune checkpoint inhibitors (ICIs) have revolutionized the landscape of therapeutic options for many cancers. These treatments have demonstrated improved efficacy and often a more favourable toxicity profile compared to standard cytotoxic chemotherapy. There are considerable differences among responders, with some patients experiencing durable long-term disease control and even remission. Given this variability, determining a proper biomarker to select patients for ICI therapy has become increasingly important. The only biomarker proven to be predictive of overall survival benefit with ICI therapy is PD-L1 expression level measured by immunohistochemistry. Several attempts have been made to identify different predictive biomarkers. One of the most intriguing and divisive is tumor mutational burden (TMB). TMB represents the number of mutations per megabase (Mut/Mb) of DNA that were sequenced in a specific cancer. With a higher number of mutations detected, and consequentially an increase in the number neo-epitopes, then it is more likely that one or more of those neo-antigens could be immunogenic and trigger a T cell response.

Initially, TMB was identified as a biomarker for ICIs in melanoma and subsequent studies suggested a possible clinical role for TMB in non-small cell lung cancer. The initial data were not confirmed in a prospective study assessing OS as the primary endpoint. Recently, the FDA has approved pembrolizumab in all cancers with a TMB > 10Mut/Mb[12] based on findings from the phase 2 KEYNOTE-158. Much criticism has emerged about this pan-cancer approval, in particular about the use of TMB as biomarker to select patients.

Here we review the data about the importance and role of TMB as possible pan-cancer one-size-fits-all biomarker. We highlight the strengths and intrinsic limitations of such a complex biomarker and its adoption in the daily practice.

免疫检查点抑制剂 (ICI) 彻底改变了许多癌症的治疗方案。与标准的细胞毒性化疗相比，这些治疗已证明疗效更好，并且通常具有更有利的毒性特征。反应者之间存在相当大的差异，一些患者经历了持久的长期疾病控制甚至缓解。鉴于这种可变性，确定合适的生物标志物来选择 ICI 治疗的患者变得越来越重要。唯一被证明可预测 ICI 治疗总生存获益的生物标志物是通过免疫组织化学测量的 PD-L1 表达水平。已经进行了多种尝试来识别不同的预测性生物标志物。最有趣和最具争议的问题之一是肿瘤突变负荷 (TMB)。TMB 表示在特定癌症中测序的 DNA 每兆碱基 (Mut/Mb) 的突变数。随着检测到的突变数量增多，新表位数量随之增加，那么一种或多种新抗原更有可能具有免疫原性并触发 T 细胞反应。

最初，TMB 被确定为黑色素瘤 ICI 的生物标志物，随后的研究表明 TMB 在非小细胞肺癌中可能具有临床作用。一项以 OS 为主要终点的前瞻性研究未证实初始数据。最近，基于 KEYNOTE-158 2 期的研究结果，FDA 已批准 pembrolizumab 用于 TMB > 10Mut/Mb[12] 的所有癌症。关于这一泛癌批准出现了很多批评，特别是关于使用 TMB 作为生物标志物来选择患者的问题。

在这里，我们回顾了有关 TMB 作为可能的泛癌一刀切生物标志物的重要性和作用的数据。我们强调了这种复杂生物标志物的优势和内在局限性及其在日常实践中的采用。