Another 63 genetic variants seem to affect risk for diagnosis with prostate cancer

Newly published research in Nature Genetics has reported the identification of yet another 63 new variants in the human genome that increase risk for a diagnosis of prostate cancer in males. This takes the total number of such variants identified to date to something approaching 175.

The actual article (by Schumacher et al.) is based on research by a very large team of internationally based specialists in the genetics of prostate cancer.

But what does this paper actually tell us?

An article published yesterday in Forbes is entitled “New $100 genetic test identifies men who have vastly increased chance of developing prostate cancer” and includes the statement that:

… research published in Nature Genetics indicates “a new DNA ‘spit test’ for prostate cancer can identify men who have an increased risk of developing the disease.”

but is that really accurate?

The answer is of course “Yes and no”.

By comparison, another article, on HealthDay, is entitled “New DNA test may predict prostate cancer risk” and includes the statement that:

According to the new study, the scientists identified 63 new genetic variants associated with increased risk of prostate cancer, and combined them with more than 100 previously identified variants to create the new test.

and goes on to note that this new test

identifies the 1 percent of men who are at highest risk for prostate cancer.

Here is what the actual study tells us, which is that,

  • Prior studies (based on a technique known as genome-wide association analysis) had previously identified “more than 100 prostate cancer … susceptibility loci”
  • The authors of the new study carried out a meta-analysis of data from a new, custom high-density array of
    • 46,939 prostate cancer cases in men of European ancestry
    • 27,910 controls (men without prostate cancer) of European ancestry
  • The authors compared these data to previously genotyped data from
    • 32,255 prostate cancer in men of European ancestry
    • 33,202 controls of European ancestry
  • The analysis identified
    • 62 novel loci associated with risk from  diagnosis with prostate cancer (P < 5.0 × 10−8)
    • 1 novel locus significantly associated with early-onset prostate cancer (in men of ≤ 55 years)

The authors then note that

  • The combination of all loci (the original 100+ and the new 63 loci) captured 28.4 percent of the relative familial risk for prostate cancer.
  • Compared to the overall population, on average, a new “polygenetic” risk score conferred an elevated risk for men in
    • The 90th to 99th percentile (relative risk = 2.69)
    • The first percentile (relative risk = 5.71)

They then conclude that:

These findings improve risk prediction, enhance fine-mapping, and provide insight into the underlying biology of prostate cancer.

This is an accurate conclusion.

But the critical question is this one:

  • Does the existence of a genetic test like this mean that every man should be having a genetic test like this to see if they are at risk for prostate cancer?

In your sitemaster’s very humble opinion, the answer to this question is a resounding “No”, for all of the following reasons:

  • The test “only” identified 28.4 percent of the known, relative familial risk for prostate cancer (i.e., it did not capture 71.6 percent of that relative risk).
  • The test is meaningless for the very large section of the population with any significant amount of non-European ancestry (and here in America that would include a surprisingly large percentage of the population; you don’t have to be “African American” or “Hispanic” or even a “Native” American to have some unexpected genes you may not know about).
  • The data are inclusive of men with any amount of any grade of prostate cancer — much of which is never going to be clinically significant.

Conversely, this test may indeed have utility in the assessment of risk for men of European ancestry who are already known to have a significant familial risk for prostate cancer, but the vast majority of that risk is still associated with only a small number of genes that are already well known.

Your sitemaster would point out that while our ability to generate genetic and genomic data and use such data in the diagnosis and management of all sort of clinical conditions has exploded over the past 10 to 20 years, this does not mean that genetic data provide clinically meaningful information as to the real risk that we get all sorts of disorders that are not, primarily, a consequence of our genetics.

Prostate cancer is a classic case of such a disease. It is a disorder that is primarily associated with how long we live and therefore associated with the risk of accumulated modifications to our individual biological functioning over time. If a specific man was given a test like this when he was 50 years of age, he might have none of the 175 or so genetic variants we now know of. If the test was given to hm again at age 65, he might still have none of the relevant variants. Then, if he was given the test for a third time at age 75, he might have three of the variants and they might or might not significantly increase his risk for diagnosis of prostate cancer — and if he was then diagnosed with low-risk prostate cancer (Gleason 3 + 3 = 6 and a PSA of 4.1) at age 82 and he died of a heart problem at age 87 while still on active surveillance, exactly what would have been accomplished by the genetic testing?


3 Responses

  1. Where can I find more specifics on the 1 novel locus significantly associated with early-onset prostate cancer (in men of ≤ 55 years)?

  2. Dear Mary:

    You’d have to see if you can get a copy of the original paper and/or (probably) talk with one of the authors of this paper. That is a very highly specialized question that I am certainly not competent to answer (and I don’t think most prostate cancer specialists would be either unless they have specific expertise in the genetics of prostate cancer).

  3. This test can be done with spit. In cases without clear indication for prostate biopsy, many might pay for it. I hope these SNP data will help to find important genetic causes and treatments of prostate cancer in the future.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: