The Predictive Value of Endometrial Thickness in Patients Receiving Tamoxifen

Abstract

Tamoxifen is a nonsteroidal antiestrogen frequently given to patients with breast cancer as a long-term adjuvant therapy. It is known that it may also promote changes in the uterus that resemble endometrial neoplasia. This study was designed to investigate the value and cost-effectiveness of ultrasound measurement of endometrial thickness in predicting endometrial pathology. From a review of the medical charts of patients being treated with tamoxifen at Lehigh Valley Hospital from 1994 to 1997, 56 patients who had also received transvaginal ultrasonography with subsequent histological evaluation of endometrial samples were identified. The records were evaluated for patient age, duration of tamoxifen therapy, endometrial thickness, descriptive ultrasonographic findings, and the final histologic diagnosis. The 56 patients ranged in age from 31 to 82 years (mean = 62 years). They had been receiving tamoxifen therapy for an average of 18.2 months. In 49 women (87.5 percent), endometrial thickness was 5 mm or less, which is the accepted minimum for normal endometrium in asymptomatic postmenopausal women. Histological examination of the endometrial sample in these women showed proliferative changes in 13 (26.5 percent), atrophy in 8 (16.3 percent), polypoid changes in 9 (18.4 percent), adenomatous hyperplasia in 4 (8.2 percent), and adenocarcinoma in 1 (2.0 percent). The endometrium of this patient was more than 50 mm thick. Fourteen patients had insufficient tissue for evaluation. Of the seven women with a measured endometrial thickness greater than 5 mm, five had insufficient tissue for evaluation. The two remaining patients had atrophic or proliferative endometrium. The predictive values of different endometrial thicknesses are shown in Table 1. The sensitivity and negative predictive values decrease and specificity and positive predictive values increase as the threshold for endometrial thickness increases. In this study, maximum efficacy for ultrasound alone as a predictor for endometrial pathology was reached at a cutoff value of 9 mm for endometrial thickness. At this point, the sensitivity and specificity were 77 and 59 percent, respectively, with positive predictive value of 37 percent and a negative predictive value of 89 percent. Using 9 mm or less of endometrial thickness as a threshold for further study would have resulted in 15 fewer endometrial biopsies with no omission of significant pathology. Using conservative measures, a savings of $525 per patient or $7875 was estimated for the 9-mm threshold.Table 1: Sensitivity, specificity, and positive and negative predictive values of endometrial thickness for detection of pathologically significant histologyJ Pelvic Surg 2000;6:16–19