THE "NEW" PROSTATE CANCER INFOLINK

An  important paper has just been published by a highly respected team of specialists doing research into the role of genetic variations in assessment of risk for prostate cancer.

In this paper, Wiklund et al. point out that we have been very successful at identifying common variants in human DNA sequences and showing that they appear to have an association to risk for prostate cancer. However, they go on to point out that such association is not necessarily clinically important, and (as yet) we have no understanding of whether the presence or absence of such variations in DNA sequences has any prognostic value in assessing clinical risk for prostate cancer.

The authors set out to see if they could confirm any association between prostate cancer susceptibility variants and prostate cancer prognosis. To do this, they used data on 16 established prostate cancer susceptibility variants from a set of  2,875 Swedish prostate cancer patients for whom complete clinical follow-up data were available through January 2008.

During follow-up, 626 men (21.8 percent) died; of those, 440 had prostate cancer classified as their underlying cause of death. However, Wiklund and his colleagues were able to find no association at all between any of the 16 DNA sequence variants that they investigated and prostate cancer-specific mortality. They were also not able to find an association even if they evaluated the cumulative effect of all variants.

The authors conclude that “hitherto established prostate cancer susceptibility variants are not associated with the lethal potential of prostate cancer.”

Now it would be easy to misunderstand the results of this study and decide that genetic variation and risk for aggressive forms of prostate cancer were not associated at all, but that would probably be unwise. What this study tells us is that genetic variation and what that variation actually means in terms of the way our bodies work is complicated. And it re-emphasizes the importance of understanding the difference between an association (Irish people often have red hair) and a prognosis (if you are Irish your children are likely to have red hair).

Wiklund presented these data at the Gentourinary Cancer Symposium in Orlando earlier this year. A point he made then is still true now: it is not enough for us to be publishing all these data that say that people with prostate cancer also seem to have this or that genetic variation. We need to start really focusing on whether this or that genetic variation has any prognostic impact on prostate cancer risk before we waste a lot of money chasing associations that have minimal likelihood of prognostic value.

Filed under: Diagnosis, Risk | Tagged: association, DNA, genetics, prognosis, risk, variation |