肺腺癌患者预后预测模型的建立与验证

Establishment and validation of a predictive nomogram for prognosis of lung adenocarcinoma

目的通过探索影响肺腺癌预后、分期和免疫微环境的标志物来预测患者的生存预后。方法利用R语言Limma包筛选从UCSC基因组网站上获得的差异表达基因(DEGs);应用clusterProfiler包进行GO和KEGG通路富集分析。随机拆分样本方法将数据划分为训练集和验证集(比率=0.6∶0.4)。应用LASSO算法和随机森林算法来选择预测因子并构建列线图;研究CACNA2D2和CD79A的免疫浸润与预后的关系并采用ROC曲线和DCA曲线评价列线图的临床应用价值。通过Western印迹法和qRT-PCR实验分析肺腺癌细胞CACNA2D2、CD79A mRNA和蛋白表达水平。采用免疫组织化学染色检测48例肺腺癌患者肿瘤组织CACNA2D2和CD79A表达情况。结果共鉴定出42个上调差异基因和19个下调差异基因,只有CACNA2D2和CD79A与肿瘤分期和生存预后高度相关。其中,CD79A的过表达与免疫反应相关。CACNA2D2和CD79A低表达组的肺腺癌患者出现淋巴结转移和较差预后的风险显著增加。构建了基于关键基因的列线图模型,使用ROC曲线和DCA曲线对模型进行验证。实验结果发现肺腺癌细胞与正常支气管上皮细胞相比,CACNA2D2和CD79A表达较低,并且肺腺癌组织CACNA2D2和CD79A阴性更多出现在Ⅲ~Ⅳ期患者以及有远处转移的患者,而强阳性更多出现在Ⅰ~Ⅱ期患者以及未发生远处转移的患者。结论肿瘤中CACNA2D2和CD79A低表达与肺腺癌患者的肿瘤分期和较差的总生存期相关,CACNA2D2和CD79A可能是肺腺癌的潜在预后标志物和治疗靶点。

Objective To develop and validate a predicting model for prognosis of lung adenocarcinoma(LUAD).Methods R language was utilized for screening differentially expressed protein-coding genes(DEGs) obtained from the UCSC genome website.GO(Gene Ontology) and KEGG(Kyoto Encyclopedia of Genes and Genomes) pathway enrichment evaluations were used.A credible random split-sample method was used to divide data into training and validation dataset(split ratio=0.6∶0.4).Lasso(Least absolute shrinkage and selection operator) Logistic regression and random forest algorithm were applied to select predictors and develop the nomogram.The relationship between immune infiltration and prognosis of CACNA2D2 and CD79A was examined.The clinical usefulness of nomogram was also evaluated with ROC(receiver operating characteristic) and DCA(decision curve analysis) curves.The mRNA and protein expression levels of CACNA2D2 and CD79A in lung adenocarcinoma cells were analyzed by Western blotting and qRT-PCR assays.The expressions of CACNA2D2 and CD79A in 48 LUAD patients were detected by immunohistochemical staining.Results A total of 42 upregulated genes and 19 down regulated genes were identified.Only CACNA2D2 and CD79A were highly correlated with tumor stage and survival prognosis.Among them,the overexpression of CD79A was associated closely with immune response.LUAD patients with CACNA2D2 and CD79A low-expression had a significantly high risk of advanced N stage and worse overall survival(OS).Gene-based nomogram model was constructed,both ROC and DCA curves were used to verify model.Moreover,the expressions of CACNA2D2 and CD79A in lung adenocarcinoma cells were lower than that in bronchial epithelial cells,and negative expressions of CACNA2D2 and CD79A in LUAD tumor tissues were more common in stage Ⅲ-Ⅳ patients and patients with distant metastasis,while strong positive tumors of CACNA2D2 and CD79A were more common in stage Ⅰ-Ⅱ patients and patients without distant metastasis.Conclusion Low-expression CACNA2D2 and CD79A in tumor were associated with tumor stage and unfavorable overall survival in LUAD.CACNA2D2 and CD79A could be a potential predictor and therapeutic target in LUAD.