Purpose: Papillary renal cell carcinoma (PRCC), the second most common kidney cancer, is morphologically, genetically, and molecularly heterogeneous with diverse clinical manifestations. Genetic variations of PRCC and their association with survival are not yet well-understood. This study aimed to identify and validate survival-specific genes in PRCC and explore their clinical utility. Materials and Methods: Using machine learning, 293 patients from the Cancer Genome Atlas-Kidney Renal Papillary Cell Carcinoma (TCGA-KIRP) database were analyzed to derive genes associated with survival. To validate these genes, DNAs were extracted from the tissues of 60 Korean PRCC patients. Next generation sequencing was conducted using a customized PRCC gene panel of 202 genes, including 171 survival-specific genes. Kaplan–Meier and Log-rank tests were used for survival analysis. Fisher’s exact test was performed to assess the clinical utility of variant genes. Results: A total of 40 survival-specific genes were identified in the TCGA-KIRP database through machine learning and statistical analysis. Of them, 10 (BAP1,BRAF,CFDP1,EGFR,ITM2B,JAK1,NODAL,PCSK2,SPATA13, andSYT5) were validated in the Korean-KIRP database. Among these survival gene signatures, three genes (BAP1,PCSK2, andSPATA13) showed survival specificity in both overall survival (OS) (p= 0.00004,p= 1.38 × 10−7, andp= 0.026, respectively) and disease-free survival (DFS) (p= 0.00002,p= 1.21 × 10−7, andp= 0.036, respectively). Notably, the PCSK2 mutation demonstrated survival specificity uniquely in both the TCGA-KIRP (OS:p= 0.010 and DFS:p= 0.301) and Korean-KIRP (OS:p= 1.38 × 10−7and DFS:p= 1.21 × 10−7) databases. Conclusions: We discovered and verified genes specific for the survival of PRCC patients in the TCGA-KIRP and Korean-KIRP databases. The survival gene signature, including PCSK2 commonly obtained from the 40 gene signature of TCGA and the 10 gene signature of the Korean database, is expected to provide insight into predicting the survival of PRCC patients and developing new treatment.
目的：乳头状肾细胞癌（PRCC）是第二常见的肾癌，在形态学、遗传学和分子水平上具有异质性，临床表现多样。目前对PRCC的遗传变异及其与生存期的关联尚不明确。本研究旨在识别并验证PRCC中与生存相关的特异性基因，并探讨其临床应用价值。 材料与方法：利用机器学习方法，对癌症基因组图谱-肾乳头状细胞癌（TCGA-KIRP）数据库中293例患者数据进行分析，筛选出与生存相关的基因。为验证这些基因，我们从60例韩国PRCC患者的组织中提取DNA，使用包含202个基因（其中171个为生存相关基因）的定制PRCC基因面板进行二代测序。采用Kaplan-Meier法和Log-rank检验进行生存分析，并使用Fisher精确检验评估变异基因的临床效用。 结果：通过机器学习和统计分析，在TCGA-KIRP数据库中鉴定出40个生存特异性基因。其中10个基因（BAP1、BRAF、CFDP1、EGFR、ITM2B、JAK1、NODAL、PCSK2、SPATA13和SYT5）在韩国KIRP数据库中得到验证。在这些生存特征基因中，三个基因（BAP1、PCSK2和SPATA13）在总生存期（OS）（p值分别为0.00004、1.38×10⁻⁷和0.026）和无病生存期（DFS）（p值分别为0.00002、1.21×10⁻⁷和0.036）均表现出生存特异性。值得注意的是，PCSK2突变在TCGA-KIRP（OS：p=0.010，DFS：p=0.301）和韩国KIRP（OS：p=1.38×10⁻⁷，DFS：p=1.21×10⁻⁷）数据库中均显示出独特的生存特异性。 结论：我们在TCGA-KIRP和韩国KIRP数据库中发现并验证了PRCC患者生存相关的特异性基因。从TCGA的40个基因特征和韩国数据库的10个基因特征中共同获得的生存特征基因（包括PCSK2），有望为预测PRCC患者生存期和开发新疗法提供重要依据。
肿瘤(癌症)患者之家