Department of Gynecologic oncology, Guangxi Medical University Cancer Hospital, 71 Hedi Road, Nanning, Guangxi, 530021, People's Republic of China.
J Ovarian Res. 2020 Feb 12;13(1):17. doi: 10.1186/s13048-020-0614-1.
The aim of this study is to establish a noninvasive preoperative model for predicting primary optimal cytoreduction in advanced epithelial ovarian cancer by HE4 and CA125 combined with clinicopathological parameters.
Clinical data including preoperative serum HE4 and CA125 level of 83 patients with advanced epithelial ovarian cancer were collected. The sensitivity, specificity, positive predictive value, negative predictive value and overall accuracy of each clinical parameter were calculated. The Predictive Index score model and the logistic model were constructed to predict the primary optimal cytoreduction.
Optimal surgical cytoreduction was achieved in 62.65% (52/83) patients. Cutoff values of preoperative serum HE4 and CA125 were 777.10 pmol/L and 313.60 U/ml. (1) Patients with PIV ≥ 6 may not be able to achieve optimal surgical cytoreduction. The diagnostic accuracy, NPV, PPV and specificity for diagnosing suboptimal cytoreduction were 71, 100, 68, and 100%, respectively. (2) The logistic model was: logit p = 0.12 age - 2.38 preoperative serum CA125 level - 1.86 preoperative serum HE4 level-2.74 histological type-3.37. AUC of the logistic model in the validation group was 0.71(95%CI 0.54-0.88, P = 0.025). Sensitivity and specificity were 1.00 and 0.44, respectively.
Age, preoperative serum CA125 level and preoperative serum HE4 level are important non-invasive predictors of primary optimal surgical cytoreduction in advanced epithelial ovarian cancer. Our PIV and logistic model can be used for assessment before expensive and complex predictive methods including laparoscopy and diagnostic imaging. Further future clinical validation is needed.
本研究旨在通过 HE4 和 CA125 联合临床病理参数建立一种非侵入性的术前模型，以预测晚期上皮性卵巢癌的初次理想肿瘤细胞减灭术。
收集 83 例晚期上皮性卵巢癌患者的术前血清 HE4 和 CA125 水平等临床资料。计算各临床参数的灵敏度、特异度、阳性预测值、阴性预测值和总准确率。构建预测指数评分模型和逻辑回归模型来预测初次理想肿瘤细胞减灭术。
62.65%（52/83）的患者达到了最佳的手术减瘤效果。术前血清 HE4 和 CA125 的截断值分别为 777.10 pmol/L 和 313.60 U/ml。（1）PIV≥6 的患者可能无法达到最佳的手术减瘤效果。诊断亚理想肿瘤细胞减灭术的诊断准确性、NPV、PPV 和特异性分别为 71%、100%、68%和 100%。（2）逻辑回归模型为：logit p=0.12 age-2.38 术前血清 CA125 水平-1.86 术前血清 HE4 水平-2.74 组织学类型-3.37。验证组逻辑模型的 AUC 为 0.71（95%CI 0.54-0.88，P=0.025）。灵敏度和特异度分别为 1.00 和 0.44。
年龄、术前血清 CA125 水平和术前血清 HE4 水平是预测晚期上皮性卵巢癌初次理想肿瘤细胞减灭术的重要非侵入性预测指标。我们的 PIV 和逻辑模型可用于评估腹腔镜和诊断性影像学等昂贵和复杂的预测方法之前的情况。需要进一步的临床验证。
