Background/Objectives:Immune checkpoint inhibitors (ICIs) have revolutionized advanced melanoma treatment, yet many patients fail to achieve sustained clinical benefit. Several biomarkers, including tumor microenvironment (TME) signature, PD-1/PD-L1 expression, and IFN-γ signaling, have been proposed. However, robust predictive markers remain elusive. This study aimed to identify molecular markers of response by analyzing tumor and peripheral immune signatures.Methods:This study analyzed 21 advanced melanoma patients treated with ICIs. Formalin-fixed, paraffin-embedded tumors underwent RNA-sequencing targeting 1392 immuno-oncology probes. Genes significantly associated with progression-free survival (PFS) by log-rank test underwent hierarchical clustering analysis (HCA). Differential expression and xCell analyses were then performed on the resulting clusters. Cox multivariate analysis was applied to identify independent PFS predictors. Pre-treatment peripheral blood mononuclear cells were analyzed by mass cytometry, followed by FlowSOM and UMAP clustering.Results:Fifty-five genes significantly associated with PFS identified two molecular clusters via HCA. Cluster A demonstrated prolonged PFS (59.4 vs. 2.4 months,p= 0.0004), while Cluster B was characterized by downregulated IFN-γ signaling, antigen presentation pathways, and reduced immune score. Multivariate Cox analysis confirmed molecular cluster as an independent PFS predictor (p< 0.001). Mass cytometry revealed higher frequencies of circulating PD-1+ CD4+ effector memory (EM) T subpopulations among responders.Conclusions:This study highlights the potential role of molecular and immune profiling in predicting ICI response in advanced melanoma. The identification of distinct molecular clusters underscores significant TME heterogeneity, with immune-cold tumor clusters associated with poorer outcomes. Furthermore, circulating PD-1+ T subpopulations emerged as potential markers of ICI response, suggesting their value in improving patient stratification.
背景/目的：免疫检查点抑制剂（ICIs）已彻底改变了晚期黑色素瘤的治疗，但许多患者未能获得持续的临床获益。已有多种生物标志物被提出，包括肿瘤微环境（TME）特征、PD-1/PD-L1表达以及IFN-γ信号通路。然而，可靠的预测标志物仍然难以确定。本研究旨在通过分析肿瘤和外周免疫特征，以识别与治疗反应相关的分子标志物。 方法：本研究分析了21例接受ICIs治疗的晚期黑色素瘤患者。对福尔马林固定、石蜡包埋的肿瘤组织进行了靶向1392个免疫肿瘤学探针的RNA测序。通过时序检验确定与无进展生存期（PFS）显著相关的基因，并进行层次聚类分析（HCA）。随后对所得聚类进行差异表达分析和xCell分析。应用Cox多变量分析以确定独立的PFS预测因子。通过质谱流式细胞术分析治疗前外周血单个核细胞，并进行FlowSOM和UMAP聚类。 结果：55个与PFS显著相关的基因通过HCA识别出两个分子聚类。聚类A表现出更长的PFS（59.4个月 vs. 2.4个月，p=0.0004），而聚类B的特征是IFN-γ信号通路和抗原呈递通路下调，以及免疫评分降低。多变量Cox分析证实分子聚类是独立的PFS预测因子（p<0.001）。质谱流式细胞术显示，在治疗应答者中，循环PD-1+ CD4+效应记忆（EM）T细胞亚群的频率更高。 结论：本研究强调了分子和免疫特征分析在预测晚期黑色素瘤ICI反应中的潜在作用。不同分子聚类的识别凸显了TME的显著异质性，其中免疫"冷"肿瘤聚类与较差的预后相关。此外，循环PD-1+ T细胞亚群作为潜在的ICI反应标志物出现，表明其在改善患者分层方面具有价值。
肿瘤(癌症)患者之家