焦亡模式通过独特的肿瘤微环境浸润特征描绘胃癌图谱。

BACKGROUND: The potential role of pyroptosis in tumor microenvironment (TME) reprogramming and immunotherapy has received increasing attention. As most studies have concentrated on a single TME cell type or a single pyroptosis regulator (PR), the overall TME cell-infiltrating characteristics mediated by the integrated roles of multiple PRs have not been comprehensively recognized. METHODS: This study curated 33 PRs and conducted consensus clustering to identify distinct pyroptosis patterns in gastric cancer (GC) patients. A single-sample gene set enrichment analysis algorithm was used to quantify the infiltration density of TME immune cells and the enrichment scores of well-defined biological signatures. The pyroptosis patterns of individuals were quantified using a principal component analysis algorithm called the pyroptosis score (PS). RESULTS: Three distinct pyroptosis patterns with significant survival differences were identified from 1422 GC samples; these patterns were closely associated with three TME cell-infiltrating landscapes-namely, the immune-inflamed, immune-excluded, and immune-desert phenotypes. The PS model generated on the basis of the pyroptosis pattern-related signature genes could accurately predict the TME status, existing molecular subtypes, genetic variation, therapeutic response, and clinical outcome; among which, a relatively high PS was highly consistent with immune activation, molecular subtypes with survival advantages, high tumor mutation burden, high microsatellite instability, and other favorable characteristics. In particular, from the Cancer Genome Atlas database, the PS model exhibited significant prognostic relevance in a pan-cancer analysis, and patients with a relatively high PS exhibited durable therapeutic advantages and better prognostic benefits in anti-PD1/L1 therapy. CONCLUSIONS: This study demonstrates that pyroptosis is prominently correlated with TME diversity and complexity, and quantification of the pyroptosis patterns of individuals will enhance our cognition of TME infiltration landscapes and help in formulating more effective immunotherapeutic strategies.

背景：细胞焦亡在肿瘤微环境（TME）重编程和免疫治疗中的潜在作用受到越来越多的关注。由于大多数研究集中于单一 TME 细胞类型或单一细胞焦亡调控因子（PR），因此尚未全面认识到多种 PR 综合作用介导的整体 TME 细胞浸润特征。

方法：本研究整理了 33 个 PR，并通过共识聚类鉴定了胃癌（GC）患者中不同的细胞焦亡模式。使用单样本基因集富集分析算法来量化 TME 免疫细胞的浸润密度和明确定义的生物学特征的富集分数。个体的细胞焦亡模式使用称为细胞焦亡评分（PS）的主成分分析算法进行量化。

结果：从 1422 个 GC 样本中鉴定出三种具有显著生存差异的不同细胞焦亡模式；这些模式与三种 TME 细胞浸润景观密切相关，即免疫激活、免疫排斥和免疫荒漠表型。基于与细胞焦亡模式相关的特征基因生成的 PS 模型可以准确预测 TME 状态、现有分子亚型、遗传变异、治疗反应和临床结局；其中，相对较高的 PS 与免疫激活、具有生存优势的分子亚型、高肿瘤突变负荷、高微卫星不稳定性和其他有利特征高度一致。特别是在癌症基因组图谱数据库中，PS 模型在泛癌分析中表现出显著的预后相关性，且 PS 相对较高的患者在抗 PD1/L1 治疗中表现出持久的治疗优势和更好的预后获益。

结论：本研究表明，细胞焦亡与 TME 的多样性和复杂性密切相关，个体细胞焦亡模式的量化将增强我们对 TME 浸润景观的认识，并有助于制定更有效的免疫治疗策略。