| Objective:Ovarian cancer has the highest mortality rate among gynecologic malignancies,and 70%of patients are already clinically advanced at their first visit.At present,the main treatment mode is surgery plus platinum-based combination chemotherapy.Patients in the early stage(stage Ⅰ and Ⅱ)and patients in the late stage(stage Ⅲ and Ⅳ)underwent cytoreductive surgery.The size of residual lesions after surgery is the most important independent prognostic factor for patients with advanced ovarian cancer[1].Preoperative evaluation of whether surgery achieves no gross residual is the key point in deciding whether to perform primary debulking surgery.This study retrospectively analyzed the clinicopathological data of 80 patients with advanced ovarian cancer treated surgically,combined with CT and CT-based cine rendering CR,serum HE4,CA125 and the risk of ovarian malignancy.Algorithm,ROMA)and other indicators,by validating the commonly used but not popular preoperative imaging evaluation model of ovarian cancer-Suidan model(2017),to explore the clinical value of CT,CR technology and serum HE4,CA125 and ROMA index in predicting whether patients with advanced epithelial ovarian cancer can achieve optimal debulking surgery.Methods:1.The clinical and pathological data of 80 patients with ovarian cancer and fallopian tube cancer who were diagnosed as stage Ⅲ-Ⅳ by primary cytoreductive surgery in the Third Affiliated Hospital of Kunming Medical University from September 2017 to October 2019 were collected and analyzed retrospectively.All patients were limited to two fixed senior gynecologic oncologists for surgical evaluation and operation.With the patient’s surgical results as the gold standard,the classification of surgical satisfaction is defined according to the existing industry standards[2-3].Optimal debulking surgery(ODS)is defined as no macroscopic residual lesions or no macroscopic residual lesions<1 cm;Suboptimal debulking surgery(SODS)is defined as macroscopic residual lesions>1 cm.2.Serum HE4 and CA125 values were measured in all 80 ovarian cancer patients before surgery,and ROMA index values before and after menopause were calculated automatically.(Since most ovarian cancer patients were postmenopausal,we only analyzed the ROMA index after menopause for statistical analysis).Using the surgical results as the gold standard,we drew the ROC curves of HE4 and CA125,ROMA index and combined detection of the three,respectively.The optimal cut-off value and AUC area were calculated,and their positive predictive value(PPV),negative predictive value(NPV)and accuracy were compared to analyze the relationship between preoperative serum HE4 value,CA125 value,ROMA index level and postoperative tumor survival,so as to explore the clinical value of each index in predicting unsatisfactory tumor reduction surgery.3.Suidan et al.established the preoperative evaluation model of ovarian cancer based on CT in 2017,and the clinical and imaging scoring criteria were as follows:1.The scoring criteria were all 1 point:age 60 years,CA125 600 U/ml,ASA 3,splenic hilum/ligament lesions,retroperitoneal lymph node metastasis above renal hilum(including supraphrenic lymph nodes),gastrohepatic ligament/hilar lesions,diffuse small intestinal adhesion/thickening;2.Indicators scored as 2 points:gallbladder fossa/hepatic interganglionic fissure lesions,moderate to severe ascites,small cystic lesions(>1 cm);3.Indicators scored as 4 points:lesions in the root of superior mesenteric artery.Because the interpretation of CT may be influenced by the personal factors of the imaging physician,in order to validate this model,we again asked two senior imaging physicians to interpret the CT data of all patients,according to the above Suidan 2017 scoring criteria,to carry out image analysis and measurement,and calculate the predictive score of all cases.Finally,the AUC area obtained from the ROC curve was plotted according to the predictive score and surgical results,and the clinical value of this model for predicting suboptimal debulking surgery in patients with advanced ovarian cancer was evaluated.4.For all collected cases,on the basis of enhanced CT,a new image post-processing technology,namely CR technology,was applied to model,image analysis and measurement were performed again,the predictive score of all cases was calculated,and the AUC area obtained by ROC curve was drawn according to the predictive score and surgical results,and the possibility of using CR technology instead of CT-introduced model to predict suboptimal debulking surgery in advanced ovarian cancer patients was evaluated.Linearity.To compare the accuracy of CT and CR in predicting surgical outcomes in the Suidan modelResults:1.The median preoperative serum HE4 value was 516.6 pmol/L in all patients,and the optimal cut-off value for predicting suboptimal debulking surgeryoutcome in patients with advanced epithelial ovarian cancer was 431.55 pmol/L.The sensitivity,specificity,positive predictive value,negative predictive value,and accuracy of serum HE4 value in predicting unsatisfactory tumor reduction were 80%,54%,51.1%,81.8%,63.75%,and the area under the ROC curve was 0.705(95%CI:0.585-0.824).The median preoperative serum CA125 of all patients was 1048 IU/ml,and the optimal cut-off value for predicting suboptimal debulking surgery outcome in patients with advanced epithelial ovarian cancer was 1707 U/ml.The sensitivity,specificity,positive predictive value and negative predictive value of serum CA125 for predicting suboptimal debulking surgery were 46.7%,76%,54%,70%and 65%,respectively.The area under the ROC curve was 0.622(95%CI:0.495-0.749).The median ROMA index before and after menopause was 0.98,and the optimal cut-off value for predicting suboptimal debulking surgery outcome in patients with advanced epithelial ovarian cancer was 0.622(95%CI:0.495-0.749).0.89,the sensitivity,specificity,positive predictive value,negative predictive value and accuracy of predicting unsatisfactory tumor reduction by ROMA index after menopause were 93.3%,38%,47.5%,90.5%and 58.8%,respectively,and the area under the ROC curve was 0.658(95%CI:0.538-0.779);the sensitivity,specificity,62%and negative predictive value of combined detection of the three methods for predicting unsatisfactory tumor reduction were 60%,77%and 62%,respectively.63.6%,accuracy 71.3%,area under ROC curve 0.721(95%CI:0.605-0.837).2.Divide all patients into optimal debulking surgery group and suboptimal debulking surgery group according to whether the surgical results are satisfactory or not.Divide HE4,CA125 and ROMA into groups with cut-off values and carry out two paired chi-square test with them respectively.It shows that the distribution frequency of unsatisfactory tumor reduction surgery in the group with HE4 greater than 431.5 pmol/L is significantly higher than that in the group with HE4 less than 431.5 pmol/L,P=0.005(OR=4.313;95%CI:1.504-12.365),the difference is significant.There was statistical significance.The distribution frequency of suboptimal debulking surgery was significantly higher in the group with CA125 greater than 1707 pmol/L than in the group with CA125 less than 1707 pmol/L,P=0.036(OR=2.771;95%CI:1.053-7.29),and the difference was statistically significant.The distribution frequency of suboptimal debulking surgery in the group with ROMA index greater than 0.89 was significantly higher than that in the group with ROMA index less than 0.89,P=0.002(OR=8.581;95%CI:1.832-40.187),and the difference was statistically significant.3.80 cases of advanced epithelial ovarian cancer,30 cases were unsatisfactory with cytoreductive surgery,accounting for 37.5%,50 cases were satisfied with cytoreductive surgery,accounting for 62.5%.Age(>60 years)and CA125(>600 U/m)were not included.For the three clinical indicators of L and ASA(>3),only by CT imaging score,the predicted scores(1-2、3-4、5-6、7-8、9-10)were 46,19;12,4,and 2,respectively.Among them,23,12,8,3,and 2 patients were unsatisfied with tumor reduction surgery,and their incidence rates were 50%,63.2%,66.7%,75%,100%,respectively.When the ROC curve area was 0.694;Combined with three clinical indicators to evaluate the suboptimal debulking surgery in patients with ovarian cancer,the predicted scores were 68,29,14,7,and 3 patients with 1-2、3-4、5-6、7-8、9-10 points,respectively.Among them,27,17,8,4,and 3 patients with suboptimal debulking surgery had an incidence of 41.2%,58.6%,57.1%,57.1%,and 100%.The ROC curve area was 0.702.4.Only by CR imaging score,48,17,10,4,2 patients with predicted score(1-2、3-4、5-6、7-8、9-10)were included,among which 25,11,6,3,2 patients with unsatisfactory reduction surgery had an incidence rate of 52%,64%,60%,75%,100%,and the ROC curve area was 0.718.When the three clinical criteria of age 60 years,CA125 600 U/m L and ASA 3 were included in the study,66,33,12,6 and 2 patients scored 1-2、3-4、5-6、7-8、9-10,among which 29,18,6,3 and 2 patients were unsatisfied with tumor reduction surgery,the incidence was 43.9%,54.5%,50%,50%,100%,and the ROC curve area was 0.719.Conclusion:1.Serum CA125,HE4 values and postmenopausal ROMA index have certain predictive value for unsatisfactory tumor reduction in advanced ovarian cancer patients.In the separate test,HE4 predicts unsatisfactory tumor reduction better than ROMA is better than CA125,and the combined test of three indicators predicts unsatisfactory tumor reduction higher than the single test.2.The use of serum CA125,HE4 value and postmenopausal ROMA index as three quantitative indicators may help to more intuitively and accurately predict the likelihood of satisfactory surgery for advanced ovarian cancer.3.Based on the Suidan scoring system,we found that the higher the predictive score,the higher the incidence of unsatisfactory cytoreductive surgery.Preoperative CT scan combined with three clinical indicators has a good agreement between the prediction of surgical satisfaction of patients with advanced ovarian cancer and the actual surgical outcome.Suidan model has a certain clinical value in the prediction of surgical evaluation of advanced ovarian cancer,which can be widely used in clinical practice.4.CR instead of CT entered the Suidan prediction model,and the ROC curve area for unsatisfactory cytoreductive surgery was 0.719,(the ROC curve area for CT was 0.702).CR seems to have a weak advantage over CT in predicting the accuracy of surgery for advanced ovarian cancer.However,because CR requires CT post-processing technology,which increases the workload,it is not suitable to replace CT in the widespread clinical use. |
PDF Full Text Request