Department of Otorhinolaryngology and Head Neck Surgery, The first affiliated hospital of Jinan University, Guangzhou, 510630, Guangdong, China.
Department of oncology, The first affiliated hospital of Jinan University, Guangzhou, 510630, Guangdong, China.
Hereditas. 2022 Jan 8;159(1):3. doi: 10.1186/s41065-021-00214-9.
MicroRNAs (miRNAs) are involved in the prognosis of nasopharyngeal carcinoma (NPC). This study used clinical data and expression data of miRNAs to develop a prognostic survival signature for NPC patients to detect high-risk subject.
We identified 160 differentially expressed miRNAs using RNA-Seq data from the GEO database. Cox regression model consisting of hsa-miR-26a, hsa-let-7e, hsa-miR-647, hsa-miR-30e, and hsa-miR-93 was constructed by the least absolute contraction and selection operator (LASSO) in the training set. All the patients were classified into high-risk or low-risk groups by the optimal cutoff value of the 5-miRNA signature risk score, and the two risk groups demonstrated significant different survival. The 5-miRNA signature showed high predictive and prognostic accuracies. The results were further confirmed in validation and external validation set. Results from multivariate Cox regression analysis validated 5-miRNA signature as an independent prognostic factor. A total of 13 target genes were predicted to be the target genes of miRNA target genes. Both PPI analysis and KEGG analysis networks were closely related to tumor signaling pathways. The prognostic model of mRNAs constructed using data from the dataset GSE102349 had higher AUCs of the target genes and higher immune infiltration scores of the low-risk groups. The mRNA prognostic model also performed well on the independent immunotherapy dataset Imvigor210.
This study constructed a novel 5-miRNA signature for prognostic prediction of the survival of NPC patients and may be useful for individualized treatment of NPC patients.
微小 RNA（miRNA）参与鼻咽癌（NPC）的预后。本研究使用 miRNA 的临床数据和表达数据，为 NPC 患者开发预后生存特征，以检测高危患者。
我们使用 GEO 数据库中的 RNA-Seq 数据鉴定了 160 个差异表达的 miRNA。在训练集中，通过最小绝对值收缩和选择算子（LASSO）构建了包含 hsa-miR-26a、hsa-let-7e、hsa-miR-647、hsa-miR-30e 和 hsa-miR-93 的 Cox 回归模型。通过 5 个 miRNA 特征风险评分的最佳截断值，将所有患者分为高危或低危组，两组患者的生存情况存在显著差异。5 个 miRNA 特征具有较高的预测和预后准确性。在验证集和外部验证集中进一步证实了结果。多变量 Cox 回归分析验证了 5 个 miRNA 特征是独立的预后因素。预测了 13 个靶基因是 miRNA 靶基因的靶基因。PPI 分析和 KEGG 分析网络均与肿瘤信号通路密切相关。使用数据集 GSE102349 中的数据构建的 mRNA 预后模型的靶基因 AUCs 较高，且低危组的免疫浸润评分较高。mRNA 预后模型在独立的免疫治疗数据集 Imvigor210 上也表现良好。
本研究构建了一个 NPC 患者生存预后预测的新型 5 个 miRNA 特征，可以为 NPC 患者的个体化治疗提供帮助。
