Clinical study on the construction of risk prediction model for postoperative recurrence of advanced epithelial ovarian cancer based on serum HE4,PLR,RLX and KPNA2 / 国际检验医学杂志

ABSTRACT

Objective To study the construction of risk prediction model for postoperative recurrence of ad-vanced epithelial ovarian cancer based on serum human epididymis protein 4(HE4),platelet count/lymphocyte count ratio(PLR),relaxin(RLX),karyopherin α2(KPNA2).Methods 124 patients with advanced epithelial o-varian cancer diagnosed and treated in Suzhou Municipal Hospital(East District)from January 2016 to January 2019 were selected as the study objects,patients with advanced epithelial ovarian cancer were divided into re-currence group and the non-recurrence group based on whether they had recurred or not.The level of HE4 was detected by electrochemical luminescence immunoassay,PLR was calculated according to the blood routine re-sults,and RLX and KPNA2 levels were detected by enzyme-related immunosorbent assay.Multivariate Logis-tic regression analysis was used to analyze the influencing factors of postoperative recurrence in patients with advanced epithelial ovarian cancer,and establish a risk prediction model for postoperative recurrence of ad-vanced epithelial ovarian cancer.Receiver operating characteristic(ROC)curve was used to evaluate the pre-dictive efficacy of the model for postoperative recurrence of advanced epithelial ovarian cancer,and Hosmer-Lemeshow test was used to analyze the fitting of recurrence risk prediction model for patients with advanced epithelial ovarian cancer.Results There was a statistically significant difference in International Federation of Gynecology and Obstetrics(FIGO)staging and serum levels of carbohydrate antigen 125,HE4,PLR,RLX and KPNA2 between the recurrence group and the non-recurrence group(P＜0.05).FIGO staging Ⅳ of cancer and elevated serum HE4,PLR,RLX and KPNA2 were risk factors for postoperative recurrence in patients with advanced epithelial ovarian cancer(P＜0.05).ROC curve analysis showed that,the area under the curve of the recurrence risk prediction model for postoperative recurrence risk of advanced epithelial ovarian cancer was 0.859,which was significantly higher than that single indicator detected by HE4,PLR,RLX and KP-NA2.Hosmer-Lemeshow test showed that the recurrence risk prediction model of advanced epithelial ovarian cancer had a good fitting(x2=7.869,P=0.437).Conclusion The risk prediction model for postoperative re-currence of advanced epithelial ovarian cancer based on serum HE4,PLR,RLX,KPNA2 and FIGO staging of cancer has high predictive value for evaluating postoperative recurrence of advanced epithelial ovarian cancer,and deserves clinical attention.