Abstract
To investigate the feasibility of synthetic magnetic resonance imaging (syMRI) in predicting the lymphatic vascular space invasion (LVSI) status of early-stage cervical cancer, and its added value to morphological MRI. A total of 72 patients with pathology-confirmed early-stage cervical cancer were enrolled, and classified into LVSI- positive (n=41) and LVSI- negative (n=31) groups. Together with morphological parameters including gross tumor volume (GTV) and maximum tumor diameter (MTD), the T1, T2, and proton density (PD) values of the tumors were also measured and compared between two groups. Binary logistic regression analysis was used to identify the independent variable associated with LVSI. Receiver operating characteristic curve analyses and DeLong tests were used to evaluate and compare the performances of significant parameters or their combination in predicting LVSI. LVSI- positive group showed significantly higher GTV (P=0.008) and MTD (P=0.019), and lower T1 (P<0.001) and PD values (P=0.041) than LVSI- negative group. However, no statistical significance was observed regarding the T2 values (P=0.331). Binary logistic regression indicated that T1 value (odds ratio [OR]=0.993; P=0.001) and MTD (OR=1.903, P=0.027) were independent variables associated with LVSI in early cervical cancer. Optimal performance could be achieved [area under ROC curve (AUC)=0.784; cut-off value=0.56; sensitivity=80.5%; specificity=71.0%] when combining T1 and MTD for predicting LVSI. Its performance was significantly better than that of MTD alone (AUC, 0.784 vs 0.662, P=0.035). syMRI might be a feasible approach, and it can provide added value to morphological MRI in predicting the LVSI status of early-stage cervical cancer.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call