Zhou Shuwei, Liu Yao, Yin Wanchun, Liao Qianqian, Quan Quan, Mu Xiaoling
Department of Gynecology, The First Affiliated Hospital of Chongqing Medical University, Chongqing 400016, China.
Department of Obstetrics and Gynecology, Chengdu First People Hospital, Chengdu 610000, China.
Transl Cancer Res. 2019 Oct;8(6):2396-2404. doi: 10.21037/tcr.2019.09.53.
Epithelial ovarian cancer (EOC) is the most lethal gynaecological malignancy, and there is a deficiency of information in the literature on the early recognition of short-term survivor (STS). This study aimed to identify the clinicopathological factors associated with STS in late-stage EOC and to establish a predictive model to identify STS.
Selected patients with International Federation of Gynecology and Obstetrics (FIGO) stage III or IV EOC were included in the study, and a retrospective analysis was performed. The characteristics of the patients who survived not more than 2 years (STS) were compared to those who survived at least 2 years (defined as long-term survivors, LTS). Binary logistic regression and receiver operating characteristic curve (ROC) were used to identify the independent prognostic factors associated with EOC and assess the predictive accuracy for STS.
We identified 254 patients with advanced EOC including 57 STS and 197 LTS. A univariate analysis revealed that STS had a tendency to have omental metastasis and larger tumor size, to be platinum resistant, to have non-serous histology, to undergo suboptimal cytoreduction, to have comorbidity, and to undergo primary chemotherapy less than 6 courses. Binary regression analysis revealed that tumor size (P=0.033), platinum resistance (P<0.001), non-serous histology (P=0.048) and number of primary chemotherapy (P=0.028) were significant independent predictors of STS. A developed predictive model using these predictors had an AUC =0.831; platinum resistance alone had an AUC =0.732.
Tumor size, omental metastasis, platinum resistance, non-serous histology, and number of primary chemotherapy are predictors associated with STS when controlling other confounding factors. Tumor size and omental metastasis may be considered novel, important prognostic factors for advanced EOC patients. Platinum resistance was the most important prognosticator for STS; hence, more work is needed for the early identification and treatment of these EOC patients.
上皮性卵巢癌（EOC）是最致命的妇科恶性肿瘤，而关于早期识别短期生存者（STS）的文献资料不足。本研究旨在确定晚期EOC中与STS相关的临床病理因素，并建立一个预测模型来识别STS。
选取国际妇产科联盟（FIGO）III期或IV期EOC患者纳入研究，并进行回顾性分析。将生存期不超过2年的患者（STS）与生存期至少2年的患者（定义为长期生存者，LTS）的特征进行比较。采用二元逻辑回归和受试者工作特征曲线（ROC）来识别与EOC相关的独立预后因素，并评估对STS的预测准确性。
我们纳入了254例晚期EOC患者，其中57例为STS，197例为LTS。单因素分析显示，STS倾向于有网膜转移、肿瘤体积更大、对铂耐药、组织学类型为非浆液性、细胞减灭术不充分、有合并症以及接受的初始化疗疗程少于6个疗程。二元回归分析显示，肿瘤大小（P = 0.033）、铂耐药（P < 0.001）、非浆液性组织学（P = 0.048）和初始化疗疗程数（P = 0.028）是STS的显著独立预测因素。使用这些预测因素建立的预测模型的曲线下面积（AUC）= 0.831；仅铂耐药的AUC = 0.732。
在控制其他混杂因素时，肿瘤大小（网膜转移）、铂耐药、非浆液性组织学和初始化疗疗程数是与STS相关的预测因素。肿瘤大小和网膜转移可能被视为晚期EOC患者新的重要预后因素。铂耐药是STS最重要的预后因素；因此，需要更多工作来早期识别和治疗这些EOC患者。
