Objective: To examine the prognostic value of four important driver gene mutations in patients with radical resection of pancreatic cancer. Methods: The clinical data and follow-up data of pancreatic cancer patients undergoing radical pancreatectomy and targeted sequencing from January 2016 to March 2018 at Department of Hepato-Biliary-Pancreatic Surgery, Changhai Hospital were retrospectively analyzed.There were 159 males and 88 females,aged of (60.8±8.7)years(range:33-83 years) and preoperative CA19-9 of (492.4±496.6)kU/L(range: 2-1 200 kU/L). One hundred and fifty nine cases of tumors were located in the head and 88 cases in the body and tail of the pancreas. After univariate analysis of clinical pathological factors (including gender, age, preoperative CA19-9, tumor location, tumor differentiation, pathological T and N stage, Micr. perineural invasion, Micr. lympho-vascular invasion, resection margin), the variable whose P<0.1 was included in COX regression model with four important driver gene mutations to find which mutation was related to prognosis independently. The number of gene mutations and KRAS subgroups were analyzed by Kaplan-Meier curve. Results: Among 247 patients,the number of KRAS,TP53, SMAD4 and CDKN2A mutations was 212 cases(85.8%), 160 cases(64.8%), 66 cases(26.7%) and 44 cases(17.8%),respectively.KRAS mutation was correlated with the tumor differentiation and pathological T stage (χ(2)=24.570/6.690, P=0.000/0.035), TP53 mutation was correlated with the tumor differentiation and the resected margin(χ(2)=5.500/4.620, P=0.019/0.032), and CDKN2A mutation was correlated with gender(χ(2)=16.574, P=0.000).COX regression model analysis showed that only KRAS mutation was an independent risk factor for disease free survival and overall survival(HR=1.776, 95%CI: 1.079-2.923, P=0.024; HR=1.923, 95%CI: 1.016-3.639, P=0.045); KRAS(G12D) mutation was associated with shorter OS(P=0.007). Conclusion: KRAS and its subgroup KRAS(G12D) mutation can be used as a prognostic index for patients with radical resection of pancreatic cancer.
目的： 探讨4种驱动基因突变状态在评估根治性切除胰腺癌患者预后中的价值。 方法： 回顾性分析2016年1月至2018年3月在海军军医大学附属长海医院胰腺外科接受根治性胰腺切除术并接受靶向测序分析的247例胰腺癌患者的临床资料及随访资料。男性159例，女性88例，年龄（60.8±8.7）岁（范围：33~83岁），术前CA19-9为（492.4±496.6）kU/L（范围：2~1 200 kU/L）；肿瘤位于胰头159例，位于胰体尾88例。对临床病理学因素（性别构成、年龄、术前CA19-9、肿瘤位置、肿瘤分化程度、病理T分期、N分期、神经浸润、脉管浸润、切缘状态）行单因素分析后，将P<0.1的变量纳入包含4种重要驱动突变的COX回归模型分析；采用Kaplan-Meier曲线对基因突变个数亚组和KRAS亚组进行单因素分析。 结果： 247例患者中，KRAS、TP53、SMAD4、CDKN2A的突变频率分别为85.8%（212/247）、64.8%（160/247）、26.7%（66/247）、17.8%（44/247）；KRAS突变与肿瘤分化程度和病理T分期相关（χ(2)=24.570、6.690，P=0.000、0.035），TP53突变与肿瘤分化程度和切缘相关（χ(2)=5.500、4.620，P=0.019、0.032），CDKN2A突变与性别相关（χ(2)=16.574，P=0.000）；COX回归模型分析结果显示，仅KRAS突变是无病生存和总体生存的独立预后因素（HR=1.776，95%CI：1.079~2.923，P=0.024；HR=1.923，95%CI：1.016~3.639，P=0.045）；KRAS(G12D)突变患者的总体生存时间更短（P=0.007）。 结论： KRAS及其亚组KRAS(G12D)突变状态可作为预测根治性切除胰腺癌患者预后的参考指标。.
Keywords: Gene mutation; Pancreatic neoplasms; Prognosis; Radical resection; Survival.