Predictive ability of maximal tumor diameter on MRI for high-risk endometrial cancer.

Abstract

To investigate the predictive ability of tumor size for deep myometrial invasion (≥50%) and metastatic lymphadenopathy, on maximal tumor diameter (MRI) of endometrial cancer. Our study population consisted of 105 patients (mean age: 59.8years) with histologically confirmed endometrial cancer. All patients underwent preoperative pelvic MRI. Tumor maximal diameter (size) was calculated on multiple sequences, and the largest value was recorded. Logistic regression analysis was performed to investigate the association of maximal tumor diameter (MRI) with the depth of myometrial invasion and the presence of pelvic nodal metastases (histology); optimal tumor size cut-off for the prediction of deep myometrial involvement and nodal metastases was calculated using ROC analysis. Surgicopathological specimen examination was the standard of reference. Tumor size on MRI, independently predicted deep myometrial invasion. Optimal maximal tumor diameter cut-off for the prediction of deep myometrial invasion was 2cm (SE 90%, SP 50.9%). When tumor size was used as a categorical variable in the multiple logistic regression model, tumor size >2cm had 10.04 times greater odds of deep myometrial invasion (95% CI 3.34-30.17, p<0.001). Optimal tumor size cut-off for prediction of nodal metastases was 4cm (SE 60%, SP 76.9%). Multiple logistic regression analysis with nodal metastases as a dependent variable showed that tumor size >4cm had 4.79 times greater odds for malignant dissemination to the lymph nodes (95% CI 1.00-23.09, p=0.047). Maximal tumor diameter on preoperative MRI may be yet another prognosticator for deep myometrial invasion and metastatic lymphadenopathy in patients with endometrial carcinoma.