Tong Xiaoxia, Qu Xiaofei, Wang Mengyun
Cancer Institute, Fudan University Shanghai Cancer Center, Shanghai, China.
Department of Oncology, Shanghai Medical College, Fudan University, Shanghai, China.
Front Oncol. 2021 Mar 24;11:639874. doi: 10.3389/fonc.2021.639874. eCollection 2021.
Cutaneous melanoma (CM) is one of the most aggressive cancers with highly metastatic ability. To make things worse, there are limited effective therapies to treat advanced CM. Our study aimed to investigate new biomarkers for CM prognosis and establish a novel risk score system in CM.
Gene expression data of CM from Gene Expression Omnibus (GEO) datasets were downloaded and analyzed to identify differentially expressed genes (DEGs). The overlapped DEGs were then verified for prognosis analysis by univariate and multivariate COX regression in The Cancer Genome Atlas (TCGA) datasets. Based on the gene signature of multiple survival associated DEGs, a risk score model was established, and its prognostic and predictive role was estimated through Kaplan-Meier (K-M) analysis and log-rank test. Furthermore, the correlations between prognosis related genes expression and immune infiltrates were analyzed Tumor Immune Estimation Resource (TIMER) site.
A total of 103 DEGs were obtained based on GEO cohorts, and four genes were verified in TCGA datasets. Subsequently, four genes (, and ) model was developed by univariate and multivariate Cox regression analyses. The K-M plots showed that the high-risk group was associated with shortened survival than that in the low-risk group ( < 0.0001). Multivariate analysis suggested that the model was an independent prognostic factor (high-risk vs. low-risk, HR= 2.06, < 0.001). Meanwhile, the high-risk group was prone to have larger breslow depth (< 0.001) and ulceration (< 0.001).
The four-gene risk score model functions well in predicting the prognosis and treatment response in CM and will be useful for guiding therapeutic strategies for CM patients. Additional clinical trials are needed to verify our findings.
皮肤黑色素瘤（CM）是最具侵袭性且转移能力极强的癌症之一。更糟糕的是，治疗晚期CM的有效疗法有限。我们的研究旨在探究CM预后的新生物标志物，并建立一种新的CM风险评分系统。
从基因表达综合数据库（GEO）下载并分析CM的基因表达数据，以鉴定差异表达基因（DEG）。然后，在癌症基因组图谱（TCGA）数据集中，通过单变量和多变量COX回归对重叠的DEG进行预后分析验证。基于多个与生存相关的DEG的基因特征，建立风险评分模型，并通过Kaplan-Meier（K-M）分析和对数秩检验评估其预后和预测作用。此外，在肿瘤免疫估计资源（TIMER）网站分析预后相关基因表达与免疫浸润之间的相关性。
基于GEO队列共获得103个DEG，在TCGA数据集中验证了4个基因。随后，通过单变量和多变量Cox回归分析建立了由4个基因（、和）组成的模型。K-M图显示，高风险组的生存期比低风险组缩短（<0.0001）。多变量分析表明，该模型是一个独立的预后因素（高风险与低风险，HR = 2.06，<0.001）。同时，高风险组更容易出现更大的 Breslow深度（<0.001）和溃疡（<0.001）。
四基因风险评分模型在预测CM的预后和治疗反应方面表现良好，将有助于指导CM患者的治疗策略。需要更多的临床试验来验证我们的发现。
