The Performance of Endometrial Biopsy and Transvaginal Sonogram in Endometrial Polyp Diagnosis [12Q

Abstract

INTRODUCTION: To assess and compare the performance of endometrial biopsy (EMB) and transvaginal ultrasound (TVS) in endometrial polyp detection in postmenopausal women. METHODS: Retrospective study of 1808 women, aged 55 and older, who underwent EMB between 1/1997 and 12/2008. Data was collected until 2015 for long term surveillance. We assessed EMB and TVS utilization and performance in diagnosis of endometrial polyps. A cutoff of ≥5 mm was used for “thickened endometrial echo (EE).” RESULTS: 1808 women underwent EMB in the study period and 676/1808 (37.4%) had final diagnosis by surgery. 195/676 (28.8%) had polyp as final diagnosis and 98/195 (50.3%) cases had concurrent TVS assessment. In 23/195 (11.8%) EMB failed to obtain tissue compared to 1/98 (1.0%) of concurrent TVS that failed to visualize the EE adequately (p=0.001). Endometrial polyp histology was diagnosed in 17/195 (8.7%) of EMB and 12/98 (12.2%) of TVS. Thick EE was diagnosed in 84/98 (85.7%) of TVS. EMB had a sensitivity of 8.7%, specificity of 96.7%, NPV of 72.3% and PPV of 51.5% in diagnosis of polyps. TVS had a sensitivity of 12.2%, specificity of 92.9%, NPV of 64.8% and PPV of 50% in diagnosis of polyps. Thick EE seen in TVS had significantly higher sensitivity (85.7% vs 8.7%, p<0.0001) but lower specificity (28.2% vs 96.7%, p<0.0001) than EMB for polyp detection. CONCLUSION: Both EMB and TVS had similarly high specificity, but very low sensitivity for polyp diagnosis. The high sensitivity of TVS endometrial thickening may be essential in polyp detection.