Partial success for womb cancer test

Behind the Headlines

Monday December 13 2010

The study used ultrasound to measure the lining of the womb

Doctors have developed a test that can detect womb cancer months before any symptoms, the Daily Mail has reported.

The news is based on research in 37,000 postmenopausal women. It found that an ultrasound technique called transvaginal ultrasound (TVS) can detect about 80% of cases of endometrial cancer (cancer of the womb lining) before symptoms appear.

This well-conducted research was the first large study to look at the accuracy of TVS for endometrial cancer. It found that TVS, which measures the thickness of the womb lining (endometrium), had a relatively high accuracy in predicting cases of endometrial cancer and ruling out the presence of cancer. However, the accuracy of the test varied depending on the endometrial thickness that was considered abnormal, as well as risk factors such as bleeding and family history of the disease.

Importantly, this study did not look at whether screening led to improved survival rates from the disease. In addition, the potential benefits of screening need to outweigh possible problems, such as false diagnoses and unnecessary treatment. Further research is now needed to identify specific groups that TVS womb screening would clearly benefit.


Where did the story come from?

The study was carried out by researchers from University College London, Barts and the London NHS Trust, Nottingham City Hospital, St Mary’s Hospital Manchester, Cardiff University, Derby City Hospital, Queen Elizabeth Hospital Gateshead, Liverpool Women’s Hospital and the Royal Cornwall Hospitals Trust. It was funded by a number of organisations, including the UK Medical Research Council, Cancer Research UK and the UK Department of Health.

The study was published in the peer-reviewed medical journal The Lancet Oncology.

The study was reported accurately in the Daily Mail, although the paper did not mention the study’s limitations, such as the absence of survival data, or discuss the possible disadvantages of screening.


What kind of research was this?

The researchers point out that cases of endometrial cancer are increasing in Europe, partly due to the rise in rates of obesity (a known risk factor for the cancer) and a fall in fertility rates (pregnancy is a known protective factor because it temporarily halts the uterus’ exposure to oestrogen). The disease is often detected through early symptoms, and has a good prognosis rate compared with other cancers.

Mass screening for endometrial cancer is not currently carried out, although it is used in women with a rare genetic disorder that makes them more likely to get the disease. It involves transvaginal ultrasound (TVS) to look for unusual thickening of the womb lining, and is a process commonly used to assess women with symptoms such as abnormal vaginal bleeding. The use of TVS has been suggested as a possible screening method for endometrial cancer.

The researchers set out to assess the accuracy of TVS as a screening method for detecting early-stage endometrial cancer in postmenopausal women without symptoms. They used a nested case control design within a group of more than 37,000 women who were taking part in the United Kingdom Collaborative Trial of Ovarian Cancer Screening (UKCTOCS), a separate study that looked at TVS screening for ovarian cancer. A nested case control study identifies cases of a disease that occur in a defined group of women and selects a specified number of matched controls from the same group who have not developed the disease.


What did the research involve?

As part of the original trial for ovarian cancer screening, more than 48,000 women, recruited between 2001 and 2005, were given a yearly TVS, carried out by experienced technicians at 13 trial centres across the UK. The TVS was used to measure the thickness of the endometrial tissue lining the womb, and women were also asked about any symptoms of postmenopausal bleeding. Women found to have endometrial thickness above a certain level (5mm), or with irregular bleeding, were advised to see a family doctor or clinician.

On average, researchers followed up all the participants for around five years, using national cancer registries and postal questionnaires to document cases of endometrial cancer or a pre-cancerous condition called atypical endometrial hyperplasia (AEH), which is abnormal overgrowth of the endometrium.

They used standard statistical methods to look at endometrial thickness and abnormalities found using TVS, both in women who did and did not develop endometrial cancer in the year after screening. These were then used to calculate the screening’s accuracy in predicting cancer.


What were the basic results?

Of the 36,867 women included in the final analysis, 136 were diagnosed with endometrial cancer or AEH within a year of screening. Of these, 107 had an endometrial thickness of 5mm or greater. Most of the 36,731 women who did not have cancer had an endometrial thickness of less than 5mm.

  • The researchers calculated that using a thickness “cut-off point” of 5mm would have a test sensitivity of 80.5% (95% confidence interval [CI] 72.7 to 86.8), which means that it would correctly detect 80.5% of women with cancer.
  • A cut-off level of 5mm would give a specificity of 85.7% (95% CI 85.8 to 86.6), which means that it would correctly exclude cancer in 85.7% of women without the disease. This would give the test a 14.3% rate of false positives, which are instances where a positive screening result was not caused by endometrial cancer.
  • The researchers obtained similar results when they performed further calculations using different womb thickness cut-off points and different risk groups.
  • When analysis was restricted to the 96 women with cancer who had reported no symptoms of abnormal bleeding at the scan, the test’s accuracy in identifying women with cancer fell. Among these women a cut-off point of 5mm would detect 77.1% (95% CI 67.8 to 84.3) of cancers, with a false positive rate of 14.2% (specificity 85.8%, 85.7 to 85.9).
  • The researchers also say their model indicated 25% of the population to be at highest risk (when also considering other risk factors including pregnancy and oral contraceptive use), with 39.5% of endometrial cancer or AEH cases occurring in women with these highest risk features. In this population, a cut-off at 6.75mm would identify 84.3% of cancers, with a false positive rate of 10.1% (specificity 89.9%, 89.3 to 90.5).


How did the researchers interpret the results?

According to the researchers, TVS screening for endometrial cancer has good accuracy in postmenopausal women. False positive results (those incorrectly suggesting the presence of cancer) could be reduced, they argue, by limiting screening to higher-risk women. While the role of screening for the general population remains uncertain, the researchers say their findings are of immediate value for clinicians and women undergoing TVS for reasons other than vaginal bleeding.



The accuracy of this test varies depending on the endometrial thickness value used as a cut-off point and adjustment for other risk factors for the condition, such as a family history of endometrial cancer. The researchers considered TVS screening to have the highest accuracy when using a cut-off thickness of 6.5mm (rather than using 5 or 10mm), and when the woman was in the highest-risk category due to the presence of other risk factors.

This large, well-designed study found that, depending on the cut-off point used for endometrial thickness and adjustment for symptoms and other risk factors, TVS could identify around 80% of endometrial cancers that occurred among postmenopausal women. However, when considering the use of TVS for cancer screening, it's important to remember the following:

  • The study did not look at the possible effect of TVS on survival rates. While it has suggested a benefit in detection, this study provides no evidence that this will lead to improvements in treatment or survival.
  • The original study, from which this research has been adapted, looked at the effects of TVS on ovarian cancer, not endometrial cancer. There is now a need for data from a randomised controlled trial which looks at the effects of TVS screening on diagnosis of and survival from endometrial cancer, in addition to the possible adverse effects.
  • It is uncertain if reports about postmenopausal bleeding were accurate, which introduces the possibility of bias. For example, if women were found to have greater endometrial thicknesses, it is possible that the presence of any postmenopausal bleeding might have then been more adequately assessed and recorded.
  • The study findings highlight that a substantial number of “false positives” would occur. These would result in unnecessary surgical interventions in women without cancer. 

As the authors of this research themselves point out, further research is needed to look at the acceptability and cost effectiveness of screening and to assess which specific groups of women might benefit most.

Analysis by Bazian

Edited by NHS Choices

Links to the headlines

Scan detects womb cancer even before symptoms appear. Daily Mail, December 13 2010

Links to the science

Jacobs I, Gentry-Maharaj A, Burnell M et al. Sensitivity of transvaginal ultrasound screening for endometrial cancer in postmenopausal women: a case-control study within the UKCTOCS cohort. The Lancet Oncology, December 13 2010 (early online publication)


Vergote I, Amant F, Timmerman D et alShould we screen for endometrial cancer? The Lancet Oncology, December 13 2010 (early online publication)


How helpful is this page?

Average rating

Based on 10 ratings

All ratings

9  ratings
1  ratings
0  ratings
0  ratings
0  ratings

Add your rating