THE "NEW" PROSTATE CANCER INFOLINK

One of the problems with prostate cancer is that it can take many years for the disease to develop from a very small group of cells in a man’s prostate into a disorder that has actual clinical significance. So how are we ever going to really find out whether a specific biologic marker (a set of genes, or a set of levels of specific biochemicals in a blood or a urine sample) can be used to definitively identify men who have clinically significant disease compared to those that don’t?

The majority of men with an elevated PSA level (say higher than 3 ng/ml) don’t have prostate cancer. So we biopsy them all in an attempt to discover whether they have prostate cancer or not. Then we obviously find a whole bunch of small, early stage prostate cancers that would never harm these patients during their natural lives (as well, of course, as many that are dangerous and need to be treated).

What is the ideal clinical study to resolve this problem? That’s easy.

We’ll take the 50,000 men in their 40s and 50s in a city like Boston, and take a large blood sample and a urine sample from them all. We can then use these carefully preserved blood samples to test for every marker we discover over time. In the meantime we will follow all these men carefully until they die. We can even monitor their diet and exercise patterns. The only thing we will not do is … we will never give them a repeat test that might indicate risk for prostate cancer unless they have a clinical sign or symptom of prostate cancer such as a positive digital rectal exam.

After about 40 years, we will be able to correlate the data from the original blood sample with the natural history of the development of clinically significant prostate cancer. We will know exactly how many men developed symptomatic prostate cancer that needed first-line treatment; we will know what percentage of those men had progressive disease; and we will know how many men actually died of their prostate cancer. It’s the perfect study! Right?

The only problem is … it’s unethical. We know that we can already find prostate cancer early in a large percentage of these men by giving them serial PSA tests and biopsies. And even if we could find 50,000 men who would volunteer to participate in such a study after a full explanation of the risks (which is conceivable if unlikely) …  it would still be unethical and no medical ethics committee would approve such a trial.

Why, then, are we bringing this idea up at all?

There are two reasons.

In the first place, there are lots of studies in which people are trying to identify new markers for early detection of prostate cancer. Most of these studies are designed to use a biopsy result as an endpoint showing that the marker is or isn’t of value. But biopsy results aren’t accurate. We know that men with negative biopsy results can have a positive result on a second biopsy. We know that men with a positive biopsy result on a first biopsy can have a negative result on a second biopsy. And we know that men with a biopsy Gleason score of 3 + 3 = 6 can often be found to have Gleason 7 disease at surgery. Biopsy results alone (like total PSA results alone) will never allow us to accurately identify markers that can differentiate between risk for clinically significant prostate cancer and indolent prostate cancer that needs no treatment.

The second reason is that this is what makes the study reported a few days ago by Vickers and his colleagues so interesting. Even though it is not a prospective clinical trial, it did in fact allow this research group to “mimic” such a trial and follow the development of prostate cancer in a cohort of 20,000 men who had a single blood test in 1996 and who had no subsequent PSA or other tests for prostate cancer unless they had signs or symptoms of the disease.

Unfortunately, the chances are very small that there is any other really large cohort of men anywhere in the world where we could carry out another such study and model the impact of a single blood test on the ability to predict the development of clinically significant prostate cancer. On the other hand, we may be able to tease further information out of the Malmö patient cohort if there are other promising tests that could be run on the stored blood samples from the men in that cohort.

Filed under: Diagnosis, Risk Tagged: | clinically significant, early stage, ethics, Malmo, prognosis, trial