Algorithm Calculates Risk of Developing Prostate Cancer Using Results from Two Blood Markers

In the new study, researchers at the University College London (London, UK) developed an algorithm estimating a person’s risk of developing prostate cancer based on age and the levels of two prostate cancer markers, prostate-specific antigen (PSA) and hK2 (human kalliknein peptidase). Prostate cancer is the most common form of cancer in men, although the current best first-line test - a blood test that detects raised levels of the PSA - is not wholly reliable, missing some harmful cancers as well as giving false positives. False positives include not just false alarms where there is no cancer but the discovery of harmless cancers that are unnecessarily treated.

The researchers tested how well their algorithm could predict prostate cancer by comparing blood samples of men who later died after a prostate cancer diagnosis with those who were never diagnosed with the disease. For the study, the researchers looked at data and blood samples from more than 21,000 men recruited into the prospective BUPA study over 40 years ago. They analyzed a number of prostate cancer markers in blood samples of 571 men who later died from or with prostate cancer, comparing these with a control group of 2,169 men who were never diagnosed with the disease. They noted that while hK2 was a relatively weak marker for prostate cancer on its own, it was relatively independent of PSA so the two together yielded a more accurate test.

The researchers categorized the results of the total PSA and hK2 tests based on how far away from average they were according to the participant’s age. They also included age into their assessment of risk. All men who were estimated to have a one in 20 or greater risk of developing prostate cancer in the next five years were counted as “screen positive”. The researchers found that if men aged 55 and over were screened at least five yearly using this risk cut-off, 90% of cancer cases would be detected, with only 1.2% of cases being false positives. If a PSA test had been used to screen for the disease on its own, in one scenario modeled by the researchers, an 86% detection rate would have been accompanied by a false positive rate of 2%. By comparison, if the risk-based approach had been adjusted to have a detection rate of 86%, the false positive rate would have been 0.5% - a reduction of three quarters. The researchers also found that men’s PSA levels were significantly elevated up to 30 years before a prostate cancer diagnosis, suggesting that a cause of prostate cancer plays a role long before it is diagnosed. However, the levels of PSA are not elevated enough to be useful in screening this far ahead of disease diagnosis.

“Our study shows a different screening approach could reduce the number of false positives by three quarters. This would make screening for prostate cancer safer and more accurate, reducing overdiagnosis and overtreatment,” said lead author Professor Sir Nicholas Wald from the UCL Institute of Health Informatics. “The next step is to test the feasibility of this approach in practice with a pilot project inviting healthy men for screening. If the project is successful, we believe this approach ought to be considered as part of a national screening program for all men.”

“The approach is innovative for cancer, as it screens people on the basis of their overall risk rather than the results of a single test. This is the same approach used in screening during pregnancy for certain fetal and maternal health conditions,” added co-author Jonathan Bestwick, from the Queen Mary University of London.

“This is a novel approach which utilizes the levels of two prostate cancer markers, PSA and hK2 (human kallikrein peptidase) to refine prostate cancer screening,” explained Professor Roger Kirby, President of the Royal Society of Medicine and Vice-President of Prostate Cancer UK, who was not involved in the study. “The use of PSA alone has significant drawbacks in terms of screening, but the addition of the hK2 marker in this context carries the genuine promise of significantly reducing the death rate from this most common cancer in men.”

Related Links:
University College London