Predictive value of preoperative DWI-PCI and MRI morphological features for postoperative residual disease and prognosis in advanced high-grade serous ovarian cancer
Predictive value of preoperative DWI-PCI and MRI morphological features for postoperative residual disease and prognosis in advanced high-grade serous ovarian cancer
Objective:To explore the predictive value of preoperative DWI-PCI and MRI morphological features for post-operative residual disease and prognosis in advanced high-grade serous ovarian cancer (HGSOC). Methods:The clinical and imaging data of 96 patients with advanced HGSOC were retrospectively analyzed. Binary Logistic regression was used to identi-fy independent influencing factors predicting postoperative residual disease,and the predictive ability of independent influenc-ing factors was evaluated using ROC curve. The relationship between factors such as DWI-PCI and progression-free survival (PFS) was analyzed using Kaplan-Meier survival curves and Cox proportional hazards regression models. Results:DWI-PCI>11 (OR=4.864,95％CI:1.087～21.756,P=0.038) and infiltrative type on MRI morphology (OR=4.227,95％CI:1.226～14.569,P=0.022) were independent predictors of surgical residual disease. The AUC of DWI-PCI for predicting surgical residual disease was 0.796,with a sensitivity and specificity of 59.1％ and 88.5％,respectively. The AUC of DWI-PCI combined with MRI morphological features for predicting surgical residual disease was 0.827,with a sensitivity and specificity of 74.6％ and 75.9％,respectively. DWI-PCI score(HR=1.086,95％CI:1.015～1.161,P=0.017) and postoperative residual disease(HR=3.187,95％CI:1.499～6.777,P=0.003) were independent risk factors affecting progression-free survival in HGSOC patients. Conclu-sion:DWI-PCI and MRI morphological features are independent predictors of postoperative residual disease in advanced HG-SOC patients,and their combination has high predictive value for surgical outcomes. DWI-PCI is an independent risk factor affecting progression-free survival in patients with HGSOC.
维普
万方数据