Ultrasonography capable of detecting endometrial cancer before symptoms are evident in postmenopausal women

13 Dec 2010

Transvaginal ultrasound screening (TVS) has good (80–90%) sensitivity and specificity at detecting endometrial cancer before symptoms appear in postmenopausal women, according to the first large-scale study to assess the performance of TVS screening for endometrial cancer.

However, the findings published Online First in The Lancet Oncology, report that further studies to examine acceptability, health economics, and risk are needed before routine population screening or targeted screening for endometrial cancer can be advocated.

Endometrial cancer is the most common gynaecological cancer. But because of the good prognosis of patients with endometrial cancer compared with other cancers, few studies have been done to assess the benefits of screening in asymptomatic women (without bleeding). However, increasing obesity, falling fertility, and an aging population suggest that incidence will continue to rise.

To date, no screening test to detect early stage endometrial cancer in asymptomatic women has been shown to be effective. Measuring endometrial thickness with TVS (a technique commonly used to assess symptomatic women with abnormal vaginal bleeding) has been suggested as a possible screening method.

To investigate the performance of TVS screening on a large scale, a team led by Ian Jacobs from University College, London, UK, analysed data from the United Kingdom Collaborative Trial of Ovarian Cancer Screening (UKCTOCS). As part of the trial, TVS was performed on 37 038 postmenopausal women.

The researchers calculated the performance characteristics of endometrial thickness (ET) and endometrial abnormalities at detecting endometrial cancer within 1 year of TVS screening, and examined the sensitivity and specificity of TVS for different risk groups. Modelling using epidemiological variables was done to assess a screening strategy for women at higher risk.

In the 133 women diagnosed with endometrial cancer or atypical endometrial hyperplasia (AEH) within 1 year of screening, 107 (81%) had an ET of 5mm or greater. Most of the 36 731 women who did not have cancer had an ET of less than 5mm.

At the optimum ET cutoff of 5.15mm, TVS would detect 80.5% of cancer cases (sensitivity) and would have a 14.3% false-positive rate (85.7% specificity). Using this threshold, if the entire population were screened, one case of endometrial cancer would be detected for every 47.7 women.

The authors point out that although the number of false positives could be reduced substantially by increasing the cutoff for ET, this would result in a corresponding fall in sensitivity. An increased cutoff of 10mm would detect 54.1% of cancer cases with 97.2% specificity, and 17 women would undergo further procedures to detect each case of endometrial cancer.

When the analysis was restricted to the 96 women with cancer who reported no symptoms of postmenopausal bleeding at the scan, an ET cutoff of 5mm was able to detect 77.1% of women who developed cancer with a 14.2% false-positive rate.

Additionally, the modelling analysis identified 25% of the population as at high risk, and this high risk group included 40% of endometrial cancer or AEH cases. In this group, the optimum ET cutoff of 6.75mm would detect 84.3% of cancer cases and achieve a specificity of 89.9%. According to the researchers, "this would reduce the burden of screening to 25% of the population with detection of about 40% of the cases."

The authors say: "A targeted screening approach might help reduce the overall number of false-positive findings while maintaining a high sensitivity."

They conclude: "Although the role of population screening for endometrial cancer remains uncertain, the findings are of immediate value in the management of increased endometrial thickness in postmenopausal women undergoing pelvic scans for reasons other than vaginal bleeding...Our findings provide the basis for further studies to assess the acceptability, health economics, and risk stratification."

In a Comment, Ignace Vergote and colleagues from University Hospitals Leuven, Leuven, Belgium caution that the findings do not prove "a benefit for screening for endometrial cancer because of an absence of survival data."

They add: "The study...provides important ultrasonographic endometrial findings in asymptomatic postmenopausal women, but screening for endometrial cancer with ultrasonography has not been sufficiently proven to be beneficial because of the high cost, the prevalence of cancers that are not clinically relevant or hyperplasia, the number of unnecessary surgical interventions and associated morbidity, the presence of clinically detected symptoms such as bleeding that happen at an early stage in many women with endometrial cancer, and the absence of prospective randomised trial data that show a survival benefit in screened patients."

Source: The Lancet