Objective: To explore the prognostic factors of epithelial ovarian carcinoma (EOC), construct a nomogram model, and evaluate the prognosis of EOC patients. Methods: A retrospective analysis was performed on clinicopathological data of 208 cases of EOC patients who received initial treatment in the First Affiliated Hospital of Army Medical University from August 11, 2016 to July 11, 2018, including age, preoperative ascites, preoperative neoadjuvant chemotherapy, surgical method, pathological type, pathological differentiation degree, surgical pathology stage, preoperative and post-chemotherapy serum cancer antigen 125 (CA125) level, human epididymal protein 4 (HE4) level, platelet count and platelet/lymphocyte number ratio (PLR). The univariate and multivariate Cox risk ratio models were used to analyze the related factors affecting progression free survival (PFS) in EOC patients, and the prediction nomogram of PFS in EOC patients was established to evaluate its efficacy in predicting PFS. Results: Univariate analysis showed that preoperative neoadjuvant chemotherapy, pathological type, pathological differentiation degree, surgical pathology stage, serum CA125 and HE4 level before operation and after chemotherapy, platelet count and PLR before operation and after chemotherapy were significantly correlated with PFS in EOC patients (all P<0.05). Multivariate analysis showed that surgical pathology stage, preoperative PLR, serum CA125 and HE4 level after chemotherapy were independent prognostic factors affecting PFS of EOC patients (all P<0.01). The index coefficient of the prediction model for the prognosis of EOC patients established by this method was 0.749 (95% CI: 0.699-0.798), which had good prediction ability, and could help clinicians to more accurately evaluate the prognosis of EOC patients. Conclusion: The nomogram model constructed based on surgical pathology stage, preoperative PLR, serum CA125 and HE4 level after chemotherapy could effectively predict the PFS of EOC patients after initial treatment, could help clinicians to screen high-risk patients, provide individualized treatment, and improve the prognosis of EOC patients.
目的： 探讨影响卵巢上皮性癌（卵巢癌）患者预后的相关因素，构建列线图预测模型，对卵巢癌患者的预后进行评估。 方法： 回顾性分析2016年8月11日至2018年7月11日在陆军军医大学第一附属医院进行初始治疗的208例卵巢癌患者的临床病理资料，包括年龄、术前有无腹水、有无新辅助化疗、手术方式、病理类型、病理分化程度、手术病理分期以及术前和化疗后血清癌相关抗原125（CA125）水平、人附睾蛋白4（HE4）水平、血小板计数、血小板计数/淋巴细胞计数比值（PLR）。对影响卵巢癌患者无进展生存时间（PFS）的相关因素进行单因素和多因素Cox比例风险模型分析，并以此建立卵巢癌患者PFS的列线图预测模型，评估其预测卵巢癌患者PFS的效能。 结果： 单因素分析显示，有无新辅助化疗、病理类型、病理分化程度、手术病理分期以及术前和化疗后血清CA125、HE4水平、血小板计数、PLR均与卵巢癌患者PFS预后显著相关（P均<0.05）；多因素分析显示，手术病理分期、术前PLR、化疗后血清CA125水平、化疗后血清HE4水平均为影响卵巢癌患者PFS的独立因素（P均<0.01），以此建立的卵巢癌患者预后的列线图预测模型的一致性系数为0.749（95%CI为0.699~0.798），该模型具有显著的预测效能。 结论： 基于手术病理分期、术前PLR、化疗后血清CA125水平和化疗后血清HE4水平这4个临床病理指标构建的列线图预测模型，有效预测卵巢癌患者初始治疗后的无进展生存情况，可帮助临床医师筛查高危患者并提供个体化治疗，以改善卵巢癌患者的预后。.