Department of Radiology, First Hospital of Qinhuangdao, Wenhua Road 258, Qinhuangdao, 066000, Hebei, China.
Department of Otolaryngology, The First Hospital of Qinhuangdao, Qinhuangdao, Hebei, China.
Eur Arch Otorhinolaryngol. 2021 Apr;278(4):1129-1138. doi: 10.1007/s00405-020-06444-3. Epub 2020 Oct 27.
Despite advances in the development of treatments for laryngeal cancer (LC), including surgical treatments and radio-chemotherapy, the survival rate of LC remains low. Therefore, novel metabolic signatures are urgently needed to evaluate the prognosis of LC patients.
Differentially expressed metabolic genes were extracted via bioinformatics analysis from the raw data of The Cancer Genome Atlas and Gene Expression Omnibus (GEO) databases. Univariate Cox regression and LASSO analyses were performed to identify metabolic genes that were significantly correlated with overall survival (OS). Using the Kaplan-Meier analysis and receiver operating characteristics, the prognostic power of candidate signatures was evaluated in the two databases. Gene Set Enrichment Analysis (GSEA) was performed to explore significant signaling pathways and underlying mechanisms in the high- and low-risk groups.
Thirteen metabolism genes showed superior ability to predict OS for LC when compared to clinical variables, and patients in the high-risk group showed significantly poorer OS than those in the low-risk group. The area under the curve of receiver operating curves for 5- and 3-year OS was 0.929 and 0.899, respectively, which were better than the OS obtained with clinicopathological variables. Similar results obtained in the GEO cohort indicated that this gene signature could differentiate between LC patients with and without recurrence.
To our knowledge, this study is the first to report that the 13 metabolic genes could serve as an independent biomarker for LC, which could provide vital prognostic information and prediction for personalized treatment of LC.
尽管喉癌 (LC) 的治疗方法（包括手术治疗和放化疗）取得了进展，但 LC 的生存率仍然较低。因此，迫切需要新的代谢特征来评估 LC 患者的预后。
通过对来自癌症基因组图谱和基因表达综合数据库 (GEO) 的原始数据进行生物信息学分析，提取差异表达的代谢基因。采用单变量 Cox 回归和 LASSO 分析鉴定与总生存期 (OS) 显著相关的代谢基因。使用 Kaplan-Meier 分析和受试者工作特征曲线评估候选特征在两个数据库中的预后能力。进行基因集富集分析 (GSEA) 以探讨高低风险组中显著的信号通路和潜在机制。
与临床变量相比，13 个代谢基因在预测 LC 的 OS 方面具有更好的能力，并且高风险组的患者 OS 明显比低风险组差。5 年和 3 年 OS 的受试者工作特征曲线下面积分别为 0.929 和 0.899，均优于临床病理变量获得的 OS。在 GEO 队列中获得的类似结果表明，该基因特征可区分有无复发的 LC 患者。
据我们所知，这项研究首次报道，这 13 个代谢基因可作为 LC 的独立生物标志物，可为 LC 的个性化治疗提供重要的预后信息和预测。
