News Release

Fine-tuned test predicts risk of ovarian cancer with great precision

Peer-Reviewed Publication

KU Leuven

Researchers from KU Leuven, Belgium, have improved a test for ultrasound diagnosis of ovarian tumours. Professors Dirk Timmerman and Ben Van Calster collaborated with scientists from Imperial College London and Lund University.

Ovarian tumours or cysts have either benign or malignant characteristics. Many women with a benign ovarian tumour only need minimally invasive surgery or even no surgery at all. Patients with a malignant tumour, by contrast, have to undergo more radical surgery to remove the tumour. This type of surgery takes much longer and comes with a greater risk of complications. To increase patients' chances of survival, women with a malignant tumour should be treated by a gynaecological oncologist as soon as possible. Therefore, ahead of the surgical intervention, a quick and correct classification of the tumour is very important.

In practice, ovarian tumours are classified as benign or malignant on ultrasound, by means of the popular 'Simple Rules' test. "Until recently, this test was inconclusive for 20-25% of the patients", lead author Professor Timmerman explains. "Our team was able to fine-tune this test. From now on, every patient can get an accurate diagnosis. The new test even provides the exact risk of the tumour being benign or malignant."

Professor Timmerman and the International Ovarian Tumour Analysis team analysed the data of 5,000 patients from 22 different countries. The data were mainly obtained from ultrasound examination of the tumour. This technique allows doctors to, among other features, see how much solid tissue is present in the tumour and how much blood is flowing in it. Based on these data, Professor Timmerman was able to identify the combinations of characteristics that reveal whether a tumour is benign or malignant.

The fine-tuned test is ready for clinical use.

###


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.