DataSheet_1_Identification of tumor mutation burden-associated molecular and clinical features in cancer by analyzing multi-omics data.pdf

BackgroundTumor mutation burden (TMB) has been recognized as a predictive biomarker for immunotherapy response in cancer. Systematic identification of molecular features correlated with TMB is significant, although such investigation remains insufficient.MethodsWe analyzed associations of somatic mutations, pathways, protein expression, microRNAs (miRNAs), long non-coding RNAs (lncRNAs), competing endogenous RNA (ceRNA) antitumor immune signatures, and clinical features with TMB in various cancers using multi-omics datasets from The Cancer Genome Atlas (TCGA) program and datasets for cancer cohorts receiving the immune checkpoint blockade therapy.ResultsAmong the 32 TCGA cancer types, melanoma harbored the highest percentage of high-TMB (≥ 10/Mb) cancers (49.4%), followed by lung adenocarcinoma (36.9%) and lung squamous cell carcinoma (28.1%). Three hundred seventy-six genes had significant correlations of their mutations with increased TMB in various cancers, including 11 genes (ARID1A, ARID1B, BRIP1, NOTCH2, NOTCH4, EPHA5, ROS1, FAT1, SPEN, NSD1,and PTPRT) with the characteristic of their mutations associated with a favorable response to immunotherapy. Based on the mutation profiles in three genes (ROS1, SPEN, and PTPRT), we defined the TMB prognostic score that could predict cancer survival prognosis in the immunotherapy setting but not in the non-immunotherapy setting. It suggests that the TMB prognostic score’s ability to predict cancer prognosis is associated with the positive correlation between immunotherapy response and TMB. Nine cancer-associated pathways correlated positively with TMB in various cancers, including nucleotide excision repair, DNA replication, homologous recombination, base excision repair, mismatch repair, cell cycle, spliceosome, proteasome, and RNA degradation. In contrast, seven pathways correlated inversely with TMB in multiple cancers, including Wnt, Hedgehog, PI3K-AKT, MAPK, neurotrophin, axon guidance, and pathways in cancer. High-TMB cancers displayed higher levels of antitumor immune signatures and PD-L1 expression than low-TMB cancers in diverse cancers. The association between TMB and survival prognosis was positive in bladder, gastric, and endometrial cancers and negative in liver and head and neck cancers. TMB also showed significant associations with age, gender, height, weight, smoking, and race in certain cohorts.ConclusionsThe molecular and clinical features significantly associated with TMB could be valuable predictors for TMB and immunotherapy response and therefore have potential clinical values for cancer management.

背景：肿瘤突变负荷（TMB）已被确认为癌症免疫治疗反应的预测性生物标志物。系统性地识别与TMB相关的分子特征具有重要意义，尽管此类研究仍显不足。方法：本研究利用来自癌症基因组图谱（TCGA）项目及接受免疫检查点阻断治疗的癌症队列的多组学数据，分析了体细胞突变、通路、蛋白质表达、microRNA（miRNA）、长非编码RNA（lncRNA）、竞争性内源RNA（ceRNA）抗肿瘤免疫特征和临床特征与TMB在多种癌症中的关联。结果：在32种TCGA癌症类型中，黑色素瘤携带最高比例的高TMB（≥10/Mb）癌症（49.4%），其次是肺腺癌（36.9%）和肺鳞状细胞癌（28.1%）。有376个基因的突变与其在多种癌症中TMB的增加显著相关，其中11个基因（ARID1A、ARID1B、BRIP1、NOTCH2、NOTCH4、EPHA5、ROS1、FAT1、SPEN、NSD1和PTPRT）的突变特征与其对免疫治疗的良好反应相关。基于三个基因（ROS1、SPEN和PTPRT）的突变谱，我们定义了TMB预后评分，该评分能够预测免疫治疗环境下的癌症生存预后，但不能预测非免疫治疗环境下的预后。这表明，TMB预后评分预测癌症预后的能力与免疫治疗反应与TMB之间的正相关相关。在多种癌症中，有9个与TMB正相关的癌症相关通路，包括核苷酸切除修复、DNA复制、同源重组、碱基切除修复、错配修复、细胞周期、剪接体、蛋白酶体和RNA降解。相反，在多种癌症中，有7个与TMB负相关的通路，包括Wnt、Hedgehog、PI3K-AKT、MAPK、神经营养因子、轴突导向和癌症相关通路。与低TMB癌症相比，高TMB癌症在多种癌症中表现出更高的抗肿瘤免疫特征和PD-L1表达水平。在膀胱癌、胃癌和子宫内膜癌中，TMB与生存预后的相关性为正，而在肝癌和头颈癌中为负。TMB还与某些队列中的年龄、性别、身高、体重、吸烟和种族显示出显著关联。结论：与TMB显著相关的分子和临床特征可作为TMB和免疫治疗反应的有价值预测指标，因此对于癌症管理具有潜在的诊疗价值。