Purpose: We sought to examine whether a preoperative assessment with usual means, available in most hospitals (preoperative histology, pelvic MRI, serum CA-125) can confidently exclude from a full staging surgical procedure low-risk endometrial carcinoma (EC) patients according to ESMO-ESTRO-ESGO criteria (stage I endometrioid EC, grade 1 or 2, myometrial invasion <50% and negative lymphovascular space invasion).
Methods: We retrospectively identified all EC patients that underwent total hysterectomy with bilateral salpingo-oophorectomy (TH-BSO) plus lymph node dissection (LND) as primary treatment for endometrioid tumors from January, 2000 to December, 2010. Extensive review was made through patients' medical records. Having set the final pathology report as the "gold standard", we applied the ESMO-ESGO-ESTRO criteria to classify patients into risk categories (low-risk and non-low risk). We also evaluated preoperative risk status using combined data from preoperative biopsy, pelvic MRI and serum CA-125. We classified patients according to the following criteria: grade 1 or 2 on preoperative histology, myometrial invasion on MRI <50% and serum CA-125 <35 IU/ml, in low risk group. Receiver operating characteristic (ROC) curves were plotted. The area under the ROC curve (AUC), quantifying the overall ability of the combined preoperative assessment to discriminate between patients at low and non-low risk, was the primary outcome of our study. False negative rate was the secondary outcome.
Results: Preoperative data on histology, MRI and CA-125 levels were available for 292 patients. The sensitivity and specificity of combined preoperative assessment to discriminate between low- and non-low risk EC patients according to ESMO-ESTRO-ESGO criteria were 96.1% and 73.6% respectively. AUC of the corresponding ROC curve was 0.849. False negative rate was 3.8% (9/235). Among the 9 patients falsely classified as low-risk, one patient had nodal metastasis (1/9, 11.1%) after full staging.
Conclusion: A selective LND strategy for EC patients based on preoperative assessment is possible and would probably be cost-effective, while not jeopardizing patients' survival or patient quality of life (QoL).