Background: Recent reports indicate that certain pre- and intraoperatively determined risk factors are predictive of pelvic lymph node metastases from endometrial cancer, allowing selective pelvic lymph node dissection. The objective of this study was to evaluate the accuracy of pre-, pre-/intra- and postoperatively determined tumor characteristics.
Methods: The study is based on 100 patients treated from 1987-1991 with total abdominal hysterectomy and bilateral salpingo-oophorectomy. In all patients thorough pelvic lymphadenectomies were performed (no sampling). These patients were evaluated according to different macroscopic and histologic tumor characteristics retrospectively in a blind fashion (the lymph node status was later determined separately). Multivariate analysis was applied and the results were compared using receiver operator characteristic curves. In 15 of 100 patients, pelvic lymph node metastases could be histologically demonstrated.
Results: Multivariate analysis of 22 tumor characteristics identified the following as being independent in relation to pelvic lymph node metastases: preoperatively determined characteristics: serous papillary tumor type, invasion of myometrium, and histologic grade (Christopherson); pre-/intraoperatively: serous papillary type, histologic grade (Christopherson), and cervical involvement; and postoperatively: lymphangiosis carcinomatosa and hemangiosis carcinomatosa. Receiver operator characteristic curves show that for pelvic node metastases the postoperatively determined histologic findings are more predictive than all other factors that can be evaluated pre- and/or intraoperatively.
Conclusion: Pre- and intraoperative tumor characteristics can determine the individual risk for pelvic lymph node involvement, but additional studies addressing the therapeutic value of pelvic lymphadenectomy would be necessary to define a probability threshold for lymphadenectomy in a decision analysis.