Discrepancy between preoperative endometrial sampling and hysterectomy diagnosis in endometrial cancer

Abstract

BACKGROUND: A study was conducted to determine the accuracy of preoperative endometrial sampling histology type and tumour grade results compared with the final postoperative diagnosis. METHODS: This was a retrospective chart audit of patients with endometrial cancer and atypical hyperplasia admitted to Grey’s Hospital in Pietermaritzburg, South Africa, from January 2013 to December 2017. RESULTS: Sixty patients met the inclusion criteria. For endometrial cancer histological types, the accuracy of preoperative endometrial sampling was 94.7% (36/38) for endometrioid adenocarcinoma, 42.9% (3/7) for serous papillary carcinoma, 85.7% (6/7) for carcinosarcoma and 75% (9/12) for atypical hyperplasia. A kappa value of 0.825 was obtained with a p-value of 0.000 for agreement between preoperative endometrial sampling and the final postoperative diagnosis. For endometrioid adenocarcinoma tumour grading 1–3 (G1–3), only 16/38 (42.1%) patients met the criteria to compare the pre- and postoperative results, which were as follows: of the eight patients with grade 1 tumour on preoperative sampling one patient (1.25%) was upgraded to grade 2 tumour postoperatively. There were no changes in tumour grading for grade 2 and 3 tumours, 3/3 and 5/5 respectively. CONCLUSION: Our study results for endometrioid adenocarcinoma are comparable to previous literature. However, there were significant discrepancies for non-endometrioid adenocarcinoma. Deficiencies that need to be addressed by laboratories in order to improve both preoperative surgical staging and postoperative adjuvant therapy planning were also highlighted.