晚期上皮性卵巢癌铂耐药复发预测模型的建立和验证  

Establishment and Validation of a Platinum Resistance Recurrence Prediction Model for Advanced Epithelial Ovarian Cancer

作　　者：蒋亚萍 汪昊涵 宁显灵 杨祖娇 王文艳 刘洲梅 杨谢兰 JIANG Yaping;WANG Haohan;NING Xianling(Department of Gynecology,The 3rd Affiliated Hospital of Kunming Medical University,Yunnan Cancer Hospital,Kunming Yunnan 650118,China)

机构地区：[1]云南省肿瘤医院昆明医科大学第三附属医院妇科,云南昆明650118

出　　处：《实用妇产科杂志》2024年第5期374-380,共7页Journal of Practical Obstetrics and Gynecology

摘　　要：目的:分析影响晚期上皮性卵巢癌(AEOC)铂耐药复发的相关因素,建立预测AEOC铂耐药复发的列线图模型并内部验证。方法:收集2013年6月1日至2021年12月31日云南省肿瘤医院妇科经初始治疗后均达到完全缓解的577例AEOC患者的临床病理资料,根据无铂间期(PFI)是否<6个月,将患者分为铂耐药复发组(130例)与非铂耐药组(447例,包括铂敏感复发、随访满6个月未复发的患者),采用多因素Logistic回归分析筛选影响患者铂耐药复发的独立危险因素,基于独立危险因素建立列线图预测模型,采用Bootstrap法内部验证,以受试者工作特征(ROC)曲线下面积(AUC)、校准曲线、决策曲线(DCA)评估模型效能。结果:①铂耐药复发组与非铂耐药组患者的年龄、肿瘤是否双侧卵巢受侵、国际妇产科联盟(FIGO)分期、绝经、行新辅助化疗(NACT)、化疗时间间隔(TTC)、血小板计数(PLT)、PLT/淋巴细胞计数(L)比值(PLR)、纤维蛋白原/L比值(FLR)、预后营养指数(PNI)、白蛋白(ALB)、糖类抗原125(CA 125)水平、腹水量、残留病灶、围术期化疗次数、CA 125半衰期比较,差异均有统计学意义(P<0.05)。②多因素Logistic回归分析示,肿瘤双侧卵巢受侵、FIGO分期Ⅳ期、TTC>16天、初诊时腹水量>1000 ml、围术期化疗次数>9次、手术为R2切除、CA 125半衰期>52天是AEOC患者铂耐药复发的独立危险因素(OR>1,P<0.05);③基于以上7项指标构建的列线图模型,AUC为0.791(95%CI 0.747~0.835),校准曲线和理想曲线拟合良好,DCA显示模型的净获益区间为0.037~0.800。结论:基于独立危险因素建立的AEOC患者铂耐药复发的列线图预测模型具有较好的区分度、校准度与临床适用度,能够在AEOC患者初治结束后对其铂耐药复发风险进行较好预测。Objective:To analyze the influencing factors of platinum resistant recurrence in advanced epithelial ovarian cancer(AEOC),establish a nomogram model to predict platinum-resistant recurrence of AEOC,and internally validate it.Methods:The clinicopathological data of 577 AEOC patients who achieved complete remission after initial treatment in the Department of Gynecology,Yunnan Cancer Hospital from June 1,2013 to December 31,2021 were collected.According to whether the platinum free interval(PFI)was less than 6 months,the patients were divided into platinum-resistant recurrence group(130 cases)and non-platinum-resistant group(447 cases,including patients with platinum-sensitive recurrence and no recurrence after 6 months of follow-up).Multivariate Logistic regression analysis was used to screen for the independent risk factors affecting the recurrence of platinum-resistant patients.Based on the independent risk factors,a nomogram prediction model was established,and Bootstrap method was used for internal verification.The area under the ROC curve(AUC),calibration curve and decision curve(DCA)were used to evaluate the performance of the model.Results:①There were statistically significant differences in age,bilateral ovarian invasion,FIGO staging,menopause,neoadjuvant chemotherapy(NACT),chemotherapy interval(TTC),platelet count(PLT),platelet count/lymphocyte count ratio(PLR),fibrinogen/lymphocyte count ratio(FLR),prognostic nutritional index(PNI),albumin(ALB),CA 125 level,ascites volume,residual lesions,perioperative chemotherapy frequency,and CA 125 half-life between the two groups(P<0.05).②Multivariate Logistic regression analysis showed that bilateral ovarian invasion,FIGO stage IV,TTC>16 days,initial ascites volume>1000ml,perioperative chemotherapy frequency>9 times,surgery with R2 resection,CA 125 half-life>52 days were independent risk factors for recurrence of platinum-resistant AEOC patients(OR>1,P<0.05).③The AUC of the nomogram model constructed based on the above 7 indicators was 0.791(95%CI 0.747-0.835

关 键 词：晚期上皮性卵巢癌 铂耐药复发 预测模型 

分 类 号：R737.31[医药卫生—肿瘤]