Transvaginal ultrasound is an effective method of
screening for endometrial cancer among postmenopausal women, according to data
from the United Kingdom Collaborative Trial of Ovarian Cancer Screening. The
sensitivity and specificity of the test was between 80% and 90%.
“In the general population, there has been limited
enthusiasm to explore the usefulness of screening for endometrial cancer
because patients have a good prognosis relative to other cancers,” the
researchers wrote. “However, in view of the rising incidence of
endometrial cancer, coupled with increasing life expectancy, there is now a
need to revisit screening.”
Using data from the United Kingdom Collaborative Trial
of Ovarian Cancer Screening (UKCTOCS), a prospective trial of ovarian cancer
screening, researchers performed a nested case-control study of 48,230
postmenopausal women who underwent transvaginal ultrasound. At the time of the
UKCTOCS, researchers recorded data on endometrial thickness and endometrial
abnormalities and documented the diagnosis of endometrial cancer using national
registries and a mailed questionnaire. The primary outcome measure was
endometrial cancer and atypical endometrial hyperplasia, according to the
researchers.
Within 1 year of transvaginal ultrasound, they also
calculated performance characteristics of endometrial thickness and
abnormalities for the detection of endometrial cancer. A logistic regression
model was developed, and a screening strategy for women at higher risk was
assessed. Median follow-up was 5.11 years.
Patients with hysterectomy (n=9,078) and those whose
endometrial thickness was not measured (n=2,271) were excluded from the study;
however, 157 of these women had an abnormality on transvaginal ultrasound and
were included. The primary analysis included 136 women with endometrial cancer
or atypical endometrial hyperplasia within 1 year of transvaginal ultrasound.
The optimum endometrial thickness cutoff for endometrial
cancer or atypical endometrial hyperplasia was 5.15 mm; transvaginal ultrasound
had 80.5% sensitivity (95% CI, 72.7-86.8) and 86.2% specificity (95% CI,
85.8-86.6). For a cutoff of at least 5 mm, sensitivity was 80.5% (95% CI,
72.7-86.8) and specificity was 85.7% (95% CI, 85.4-86.6); for women with a
cutoff of at least 5 mm plus endometrial abnormalities, the sensitivity was
85.3% (95% CI, 78.2-90.8) and specificity was 80.4% (95% CI, 80-80.8).
Sensitivity was 54.1% (95% CI, 45.3-62.8) and specificity was 97.2% (95% CI,
97.0-97.4) for a cutoff of at least 10 mm.
Ninety-six women had endometrial cancer or atypical
endometrial hyperplasia with no symptoms of postmenopausal bleeding before
diagnosis. Among these women, a cutoff of 5 mm had sensitivity of 77.1% (95%
CI, 67.8-84.3) and specificity of 85.8% (95% CI, 85.7-85.9). According to the
logistic regression model, 25% of patients were at high risk, and 39.5% of
those with endometrial cancer or atypical endometrial hyperplasia were included
in this high-risk group. A cutoff of 6.75 mm had 84.3% (95% CI, 71.4-93)
sensitivity and 89.9% (95% CI, 89.3-90.5) specificity in this population.
“The rising incidence of endometrial cancer and the
difficulty of doing a randomized controlled trial large enough to specify
mortality as an endpoint suggest that the decision to introduce screening for
all or a subgroup of asymptomatic women will rest on surrogate criteria such as
the performance characteristics described in our report and future studies of
acceptability, health economics and risk stratification,” the researchers
wrote. “We do not advocate population screening for endometrial cancer
until further data are available from these studies.”