European Multidisciplinary Cancer Congress (EMCC) 2011, 23-27 September, Stockholm

Cost effective ways to predict prostate cancer 2/2

Dr David Orsted – Copenhagen University Hospital, Denmark


The title of this project was A Prostate Specific Antigen or PSA for Long-Term Prediction of Prostate Cancer Incidence and Mortality. Now PSA is used by doctors to diagnose and monitor prostate cancer but it’s unknown if PSA can be used to predict long-term risk of prostate cancer and even dying from prostate cancer. So if we take a blood sample and measure PSA in a man from the general population, can we say something about the long-term risks? This is important because prostate cancer is a major cancer disease in Europe, there were 340,000 cases in 2008 and 70,000 deaths from prostate cancer in Europe in 2008.

To study this question we used the Copenhagen City Heart Study which is a general population study, prospective study, going back to the ‘80s. We used 4,500 participants that had blood samples collected in 1981-1983 and these samples were then stored for thirty years, not used for anything else until we decided to measure PSA in 2010. We followed these men from 1981 in the Danish Health Register, saw if they had a diagnosis of prostate cancer or died from prostate cancer. If we look first at prostate cancer, what we found was here you have the cumulative incidence of prostate cancer and age of the participants and then we divided PSA into six groups. What’s important to note here is that about 90% of our participants had a PSA below 2.

We could see that with increasing PSA levels we had an increasing incidence of prostate cancer. When we looked at prostate cancer death we saw the same picture, here you have again the PSA categories and risk of prostate cancer death, note that this is a logarithmical scale with a doubling each time you move one step. We could see that each time you moved up one category in PSA value, you had an increased risk of dying from prostate cancer and this was highly significant.

We decided also to look at absolute risk because we attempted to create a model that could be used clinically and what we looked at was ten year risk of prostate cancer by PSA category. This figure is also divided by age, so you have men less than 50 years, 50-60 years, 60-70 and above 70 years of age. Again we saw an increased risk, a ten year absolute risk, with an increase in PSA but what’s important to note on this figure is that for the large majority, this was about 90% of our participants with a PSA below 2, there was actually a very low ten year absolute risk. This can be used to reassure these men and maybe reduce one of the major problems in prostate cancer which is over-diagnosis or at least two many PSA tests and maybe also unnecessary prostate biopsies.

So, to sum up, the main conclusion is that elevated prostate specific antigen in men from the general population, the first time they are tested, can be used to predict their long-term risk of having prostate cancer or dying from prostate cancer. And for prostate cancer it was a 3 to 44 fold increased risk and a 2 to 12 fold increased risk of prostate cancer death.

Thank you.