建立一个与 IFNγ 反应相关的特征，用于预测皮肤黑色素瘤的生存情况。

BACKGROUND: The tumor microenvironment (TME) plays a critical role in tumorigenesis, development, and therapeutic efficacy. Major advances have been achieved in the treatment of various cancers through immunotherapy. Nevertheless, only a minority of patients have positive responses to immunotherapy, which is partly due to conditions of the immunosuppressive microenvironment. Therefore, it is essential to identify prognostic biomarkers that reflect heterogeneous landscapes of the TME. METHODS AND MATERIALS: Based upon the ESTIMATE algorithm, we evaluated the infiltrating levels of immune and stromal components derived from patients afflicted by various types of cancer from The Cancer Genome Atlas database (TCGA). According to respective patient immune and stromal scores, we categorized cases into high- and low-scoring subgroups for each cancer type to explore associations between TME and patient prognosis. Gene Set Enrichment Analyses (GSEA) were conducted and genes enriched in IFNγ response signaling pathway were selected to facilitate establishment of a risk model for predicting overall survival (OS). Furthermore, we investigated the associations between the prognostic signature and tumor immune infiltration landscape by using CIBERSORT algorithm and TIMER database. RESULTS: Among the cancers assessed, the immune scores for skin cutaneous melanoma (SKCM) were the most significantly correlated with patients' survival time (P < .0001). We identified and validated a five-IFNγ response-related gene signature (UBE2L6, PARP14, IFIH1, IRF2, and GBP4), which was closely correlated with the prognosis for SKCM afflicted patients. Multivariate Cox regression analysis indicated that this risk model was an independent prognostic factor for SKCM. Tumor-infiltrating lymphocytes and specific immune checkpoint molecules had notably differential levels of expression in high- compared to low-risk samples. CONCLUSION: In this study, we established a novel five-IFNγ response-related gene signature that provided a better and increasingly comprehensive understanding of tumor immune landscape, and which demonstrated good performance in predicting outcomes for patients afflicted by SKCM.

背景：肿瘤微环境（TME）在肿瘤发生、发展和治疗效果中起着关键作用。免疫疗法在治疗各种癌症方面取得了重大进展。然而，只有少数患者对免疫疗法有积极反应，部分原因是免疫抑制微环境的状况。因此，识别反映 TME 异质性景观的预后生物标志物至关重要。

方法和材料：基于 ESTIMATE 算法，我们评估了来自癌症基因组图谱数据库（TCGA）中各种类型癌症患者的免疫和基质成分的浸润水平。根据各自的患者免疫和基质评分，我们将病例分为高评分和低评分亚组，以探索 TME 与患者预后之间的关系。进行了基因集富集分析（GSEA），并选择富集 IFNγ 反应信号通路的基因，以建立预测总生存期（OS）的风险模型。此外，我们使用 CIBERSORT 算法和 TIMER 数据库研究了预后特征与肿瘤免疫浸润景观之间的关系。

结果：在所评估的癌症中，皮肤黑色素瘤（SKCM）的免疫评分与患者的生存时间最显著相关（P<0.0001）。我们鉴定并验证了一个与 SKCM 患者预后密切相关的五个 IFNγ 反应相关基因特征（UBE2L6、PARP14、IFIH1、IRF2 和 GBP4）。多变量 Cox 回归分析表明，该风险模型是 SKCM 的独立预后因素。肿瘤浸润淋巴细胞和特定免疫检查点分子在高风险样本中表达水平明显不同于低风险样本。

结论：在这项研究中，我们建立了一个新的五个 IFNγ 反应相关基因特征，它提供了对肿瘤免疫景观的更好和日益全面的理解，并在预测 SKCM 患者结局方面表现出良好的性能。