This study summary is an excerpt from the book 2 Minute Medicine’s The Classics in Medicine: Summaries of the Landmark Trials

1. Endovaginal ultrasound (EVUS) revealing an endometrial stripe of 5 mm or greater was 96% sensitive for endometrial cancer and 92% sensitive for other endometrial disease, allowing for noninvasive screening and subsequent determination of candidacy for endometrial biopsy.

2. EVUS was less specific among postmenopausal women utilizing hormone replacement therapy (HRT) than among those who refrained from HRT, and this overall low sensitivity necessitated endometrial biopsy in all women with an abnormal EVUS.

Original Date of Publication: November 1998

Study Rundown: Endometrial cancer is the third most common cancer of the female reproductive tract after ovarian and cervical cancers, but has a low mortality rate if brought to early medical attention. It typically presents with painless, abnormal uterine bleeding among postmenopausal women, particularly those who are obese or utilizing HRT without progesterone supplementation, as unopposed estrogens promote endometrial hyperplasia. Prior to the widespread use of EVUS in screening symptomatic women for endometrial abnormalities, endometrial biopsy was the only diagnostic tool used in the workup of at-risk women with abnormal uterine bleeding. However, endometrial biopsy is invasive, painful, and can be nondiagnostic or inaccurate in over a quarter of attempts. The referenced study reviewed the literature from 1966 to 1996 regarding the use of EVUS as the first-line screening tool among women with suspicion of endometrial cancer in an effort to determine if a negative screening EVUS could obviate the need for endometrial biopsy. Utilizing an endometrial thickness threshold of 5 mm for positivity on any given EVUS examination, the authors found a pooled sensitivity of 96% for the detection of endometrial cancer and 92% for other endometrial disease, including atypical endometrial hyperplasia or endometrial polyps. Additionally, no significant sensitivity difference was found between women using HRT and those not. However, specificity for the detection of endometrial cancer varied significantly between women who did and did not use HRT, finding almost three times as many women on HRT to have an abnormal EVUS despite normal histology on biopsy. Of note, this analysis did not include studies involving women using tamoxifen, which may markedly thicken the endometrium, and also did not evaluate the use of EVUS as a screening tool in asymptomatic women due to the rarity of endometrial cancer. The variability and overall low-to-moderate specificity of EVUS demonstrated it to be an improper tool for the ultimate diagnosis of an endometrial abnormality, but it displayed excellent sensitivity characteristics in the initial workup of at-risk women presenting with abnormal uterine bleeding concerning for malignancy. A negative EVUS result was shown to reduce the pretest odds of cancer by roughly 90% regardless of HRT status, therefore effectively ruling out cancer and preventing unnecessary invasive biopsies.

Click to read the study in JAMA

In-Depth [systematic review and meta-analysis]: This systematic review and meta-analysis pooled data from 35 studies published between 1966 and 1996 examining the use of EVUS in the diagnostic workup of primarily postmenopausal patients with suspected endometrial cancer, all of which were confirmed by endometrial biopsy. A total of 5892 women (mean age 61 years; 471 women using HRT) were included, with 94% presenting with vaginal bleeding. The sensitivity, specificity, and positive and negative predictive values were calculated for EVUS in the detection of both endometrial cancer and other endometrial abnormalities (polyps or atypical hyperplasia), and among women who were and were not utilizing postmenopausal HRT. Within the sampled group, the prevalence of endometrial cancer was calculated to be 13%, while that of endometrial polyps or hyperplasia was 40%. Women with normal endometrial histology displayed a mean endometrial thickness of 4±1 mm, while women with endometrial polyps, hyperplasia and cancer displayed mean thicknesses of 10±3 mm, 14±4 mm and 20±6 mm, respectively. A 5 mm threshold for EVUS positivity was selected as it displayed the best overall sensitivity and specificity characteristics. Regardless of HRT status, the sensitivity of EVUS in testing for endometrial cancer was 96% (95%CI 94-98%), while it was 92% (95%CI 90-93%) in testing for other endometrial abnormalities. This lack of variation despite HRT status and high sensitivity demonstrated that EVUS is highly accurate in the exclusion of endometrial disease, with a negative likelihood ratio of 0.05-0.12. Therefore, in a woman with a 10% pretest probability of endometrial disease, typical for that of postmenopausal women presenting with abnormal uterine bleeding, a negative EVUS reduced her risk of disease to 1%. Regarding specificity, the study found that EVUS cannot accurately distinguish between endometrial cancers, polyps or hyperplasia, necessitating a subsequent endometrial biopsy in all women with an abnormal EVUS. This study definitively established EVUS as an effective first-line tool in the diagnostic workup of women with suspected endometrial cancer while providing the baseline, evidence-based threshold for examination positivity used in current practice.

Smith-Bindman R, Kerlikowske K, Feldstein V, Subak L, Scheidler J, Segal M, Brand R, Grady D. Endovaginal ultrasound to exclude endometrial cancer and other endometrial abnormalities. Journal of the American Medical Association. 1998 Nov 4;280(17):1510–17.

©2022 2 Minute Medicine, Inc. All rights reserved. No works may be reproduced without expressed written consent from 2 Minute Medicine, Inc. Inquire about licensing here. No article should be construed as medical advice and is not intended as such by the authors or by 2 Minute Medicine, Inc.