Suidan术前评估模型在预测晚期上皮性卵巢癌肿瘤细胞减灭术效果中的应用研究

目的 探讨Suidan术前评估模型在预测晚期上皮性卵巢癌患者肿瘤细胞减灭术效果中的应用价值。方法 回顾性队列研究。纳入2019年1月—2020年10月蚌埠医学院第一附属医院妇瘤科手术治疗的晚期上皮性卵巢癌患者152例。患者年龄18~85岁,中位年龄55岁;依据国际妇产科联盟（FIGO）分期,Ⅲ期117例,Ⅳ期35例;采用Suidan术前评估模型评分为0~9分。患者均行卵巢癌肿瘤细胞减灭术,按治疗方案及Suidan评分将患者分为3组：行初始型肿瘤细胞减灭术（PDS）且术前Suidan评分<3分为PDS-A组（42例）,行PDS且术前Suidan评分≥3分为PDS-B组（46例）,行中间型肿瘤细胞减灭术（IDS）且术前Suidan评分≥3分为NACT+IDS组（64例）。比较3组患者的临床基线资料,以及不满意减瘤率、肉眼残留病灶率;比较不同Suidan评分,患者间不满意减瘤率、肉眼残留病灶率的差异;采用受试者操作特征（ROC）曲线评估Suidan术前评估模型对晚期上皮性卵巢癌患者治疗后不满意减瘤、肉眼残留病灶的预测价值。结果 3组患者年龄、病理类型、组织学分级及FIGO分期等比较,差异均无统计学意义（P值均>0.05）。3组患者不满意减瘤率、残留病灶率由高到低依次为PDS-B组[71.7%（33/46）、78.3%（36/46）]、NACT+IDS组[48.4%（31/64）、54.7%（35/64）]、PDS-A组[23.8%（10/42）、33.3%（14/42）],差异均有统计学意义（χ²=20.19、18.05,P值均<0.001）。行PDS患者不同Suidan评分时,患者不满意减瘤率、肉眼残留病灶率随着评分数值的增加逐渐升高。ROC曲线显示：Suidan术前评估模型预测晚期上皮性卵巢癌患者不满意减瘤曲线下面积（AUC）为0.761[95% 可信区间（CI） 0.657~0.864],约登指数为0.478时,最佳评分阈值为3分,灵敏度为76.7%,特异度为71.1%;预测晚期上皮性卵巢癌患者手术残留病灶AUC为0.730（95% CI 0.624~0.836）,约登指数为0.457时,最佳评分阈值为3分,灵敏度为72.0%,特异度为73.7%。结论 Suidan术前评估模型对晚期上皮性卵巢癌患者肿瘤细胞减灭术的效果有较好的预测价值,Suidan评分为3分时,灵敏度及特异度较高。

Importance of a Suidan preoperative evaluation model in predicting the outcome of cell reduction for advanced epithelial ovarian cancer

Objective This study aimed to investigate the importance of a Suidan preoperative evaluation model in predicting the effect of tumor cell reduction in patients with advanced epithelial ovarian cancer. Methods In this study, a retrospective cohort study design was used. A total of 152 patients with advanced epithelial ovarian cancer who underwent surgical treatment in the Gynecologic Oncology Department of the First Affiliated Hospital of Bengbu Medical College from January 2019 to October 2020 were included. Patients' age ranged from 18 to 85 years, with a median age of 55 years. According to the International Union of Obstetrics and Gynecology classified,117 cases as stage Ⅲ and 35 cas es as stage Ⅳ. All patients were evaluated using a Suidan preoperative evaluation model and scored 0-9 points. All patients received ovarian cancer cell reduction, and they were divided into three groups in accordance with the surgical method and Suidan score: patients receiving primary debulking surgery (PDS) with Suidan's preoperative score of <3 were included in the PDS-A group (42 cases); patients receiving PDS with Suidan's preoperative score of ≥3 were included in the PDS-B group (46 cases); patients receiving IDS combined with NACT with Suidan's preoperative score of ≥3 were included in the NACT+IDS group (64 cases). The clinical baseline data, unsatisfactory tumor reduction rate, and gross residual lesion rate of patients in the three groups were compared, and the differences in surgical unsatisfactory tumor reduction rate and gross residual lesion rate among patients with different Suidan scores were compared. Moreover, the predictive value of the Suidan preoperative evaluation model for unsatisfactory tumor reduction and gross residual lesion after treatment in patients with advanced epithelial ovarian cancer was evaluated using the receiver operating characteristic (ROC) curve. Results No significant differences in age, pathological type, histological grade, and FIGO stage were found among the three groups (all P values >0.05). The following unsatisfactory tumor reduction rate and residual lesion rate of patients in the three groups were arranged from high to low: PDS-B group (71.7%[33/46], 78.3%[36/46]), NACT+IDS group (48.4%[31/64], 54.7%[33/64]), and PDS-A group (23.8%[10/42], 33.3%[14/42]). The differences were statistically significant (χ²=20.19, 18.05, all P values <0.001). For patients receiving PDS with different Suidan scores, the rate of surgical unsatisfactory tumor reduction and the rate of gross residual lesions gradually increased with the increase of patient scores. The ROC curve showed that the Suidan preoperative assessment model predicted an area under the tumor reduction curve (AUC) of 0.761 for patients with advanced epithelial ovarian cancer (95%[credibility interval]CI) 0.657-0.864), and the highest Youden index was 0.478. In addition, the optimal scoring threshold, sensitivity, and specificity were 3 points, 76.7%, and 71.1%, respectively. The predictive AUC of surgical residual lesions in patients with advanced epithelial ovarian cancer was 0.730 (95%CI 0.624-0.836), and the highest Yuden index was 0.457. Moreover, the optimal scoring threshold was 3, sensitivity, and specificity were 72.0% and 73.7%. Conclusion The Suidan preoperative evaluation model has good predictive value for tumor cell reduction in patients with advanced epithelial ovarian cancer, with a Suidan score of 3, indicating high sensitivity and specificity.