Combining vaginal ultrasonography and office endometrial sampling in the diagnosis of endometrial disease in postmenopausal women

Abstract

To investigate the value of the combined use of vaginal ultrasonography and endometrial sampling in the office for the diagnosis of endometrial disease in postmenopausal women. One hundred forty consecutive postmenopausal women presenting with uterine bleeding or endometrial cells on cervical cytology entered the study. Vaginal ultrasonography was used to measure the endometrial thickness, followed by use of the Pipelle endometrial sampler. Pipelle biopsy was not feasible in two patients. The results of hysteroscopy with biopsy or hysterectomy, performed within 6 weeks in all but 12 patients, were considered the final diagnosis. The accuracy of ultrasound and Pipelle was measured against the final diagnosis. The sensitivity of vaginal ultrasonography for endometrial disease was 98.2 and 82.0% if cutoff points for endometrial thickness of 2 and 4 mm, respectively, were used. All six patients with endometrial carcinoma had endometrial thicknesses exceeding 12 mm. Pipelle endometrial sampling had a sensitivity of 44.6% and a specificity of 98.5% for endometrial disease. All cases of endometrial carcinoma were detected by sampling in the office setting. This study illustrates the value of vaginal scanning in the diagnosis of endometrial disease in symptomatic, postmenopausal women. A 4-mm cutoff point for endometrial thickness seemed appropriate. The sensitivity of Pipelle sampling for endometrial carcinoma was excellent, but relatively weak for other endometrial disease because it failed to detect endometrial polyps and submucous myomas. The combined use of ultrasound and Pipelle sampling offers sufficient diagnostic information for most symptomatic postmenopausal women.