Feasibility of routine lymphadenectomy in clinical stage-I endometrial cancer.

BACKGROUND: To determine the accuracy of several preoperative tests in predicting lymph node (LN) metastases and the feasibility of doing a routine lymphadenectomy in clinical stage-I endometrial cancer.

MATERIAL/METHODS: We reviewed 132 patients with clinical stage-I endometrial cancer. The preoperative tests used to predict LN metastases were serum CA-125 level, histologic type and grade, LN status assessed by pelvic magnetic resonance image (MRI) or computed tomography (CT), and depth of myometrial invasion assessed only by pelvic MRI. The cutoff value of the serum CA-125 level was determined using receiver operating characteristic curves. Multivariate logistic regression analyses were used to determine which tests are good predictors of LN metastases.

RESULTS: Of 132 patients, 13 (9.8%) had LN metastases. On univariate logistic regression analysis, a high CA-125 level and preoperative LN evaluation by pelvic MRI or CT were significant predictors for LN metastases (OR=17.41, 95% CI: 4.36-69.56 and OR=14.30, 95% CI: 4.02-50.63, respectively). However, on multivariate logistic regression analysis adjusted for age and all preoperative tests, a high CA-125 level was the most significant predictor (OR=13.73, 95% CI: 2.03-92.73). Among the 97 patients with no significant predictor of LN metastases, pelvic LN metastases were observed in 3 patients (3.1%) and para-aortic LN metastases were observed in 1 patient (1.1%). Surgical complications were mild (lymphocele, n=9; lymphedema, n=2; wound problem, n=2).

CONCLUSIONS: Considering the importance of LN metastases as an indicator of prognosis, and the relatively low surgical risk of lymphadenectomy, clinicians should cautiously consider routine lymphadenectomy in patients with clinical stage-I endometrial cancer.