A Test You Don't Have to Study For!

News article:
Kolata, G. (2008, January 17). $300 to learn risk of prostate cancer. The New York Times.

Primary scientific article:
Zheng, S.L. et al. (2008) Cumulative association of five genetic variants with prostate cancer. New Engl. J. Med. 10.1056/NEJM0a075819.

This article has not yet been cited.

After skin cancer, prostate cancer (PC) is the most common cancer to inflict men. It is also the second deadliest cancer, for men, after lung cancer. This form of cancer, like many others, becomes more common as men age and currently, men over the age of 50 are regularly screened for prostate cancer using the PSA test. This test looks for increased amounts of prostate-specific antigen (PSA), which could be a sign of abnormal (and thus, potentially cancerous) prostate cells.

As with any cancer, however, it is better to “catch” the cancer at the earliest possible stage. With recent advances in genome sequencing (that is, getting the entire DNA sequence of an individual), many cancer researchers are wondering if a person’s DNA can tell doctors whether that person is likely to get cancer or not. For example, it is already known that mutations within certain genes involved in regulating cell growth can contribute towards cancer. The question posed by scientists was: Are there genetic differences between people who are more likely to get PC compared to those who aren’t?

Earlier this month, a group of researchers published a study that suggests that there are 5 single nucleotide polymorphisms (SNPs) that can indicate, each independently, the likelihood of a man developing PC. A SNP is just a small difference (a single DNA base change) between two individuals’ genomes. There are probably 1000s of SNPs in the human genome, but prior to this study, there were already 5 regions of the genome with SNPs that had been correlated with PC – that is, it seemed that the presence/absence of these SNPs might be good indicators for risk of developing PC.

In this new study, Zheng and colleagues investigated these 5 regions of the human genome in 5000 patients, both those that had and did not have PC. The research suggests that each of these 5 SNPs can reveal the chances of getting PC. The more of these SNPs one has, moreover, the more likely one is to get PC. In fact, if one has 4 or 5 of these specific SNPs, and a family history of PC, then one’s chances of getting PC are 9x more likely than someone who has none of these SNPs. The corresponding New York Times (NYT) article about this study reveals that the researchers have developed a test, that will cost about $300, which men can take and will tell them if they have any of the 5 SNPs associated with PC. The researchers suggest that men as young as 35 years old can take this test, and therefore, decide whether they want to begin PSA testing (or other screening) before they are 50.

What the NYT says that the research article doesn’t say is that it is unclear whether taking these sort of tests is a good idea for the patients. It is unclear, for example, what the test will really tell the patient. It isn’t 100% accurate (only 46% of the patients with PC had any of the 5 SNPs) and it doesn’t tell the patient anything about the aggressiveness of the possible cancer, or whether additional early screening and tests will actually extend the patients’ lives. No one is questioning the science that was done by the researchers – it al seems very sound. But people are definitely questioning what should be done with the information garnered from this study.