For years, urologist H. Ballentine Carter, M.D., has been studying PSA, looking for ways to make it a more accurate crystal ball. Among his many breakthroughs are the concept of PSA velocity — looking for changes in a man's PSA levels over time — and the development of guidelines for when and how often men should be tested for prostate cancer. Carter has spent much of his career poring over the records of men in the Baltimore Longitudinal Study of Aging (BLSA) a huge treasure trove that includes, among other data, blood samples taken over decades from men — many of whom eventually developed prostate cancer. Carter has identified changes over time in those men as they headed toward cancer.
In a recent study, published in the Journal of Urology, Carter and urology resident Stacy Loeb, M.D., together with colleagues at the BLSA, asked a different question, based on work led by Hopkins scientists showing links between certain genetic variations and a man's risk of prostate cancer. Some of these genetic variations are located near the PSA gene, and are also associated with the PSA level. Would it be possible, Carter and Loeb wondered, to use these genetic tests — which help identify men at higher risk of developing prostate cancer — to help interpret the PSA level?
Men with specific genetic
landmarks had a 28-percent
higher risk of developing
prostate cancer for every one unit
rise in PSA — say, from 2.5
to 3.5. But for men without these
suspicious genes, using this
equation, prostate cancer risk
Carter and Loeb studied 505 men from the BLSA; 61 of these men were diagnosed with prostate cancer. They found that for each one-unit increase in PSA — from 2.5 to 3.5 ng/ml, for example — the men's risk of developing prostate cancer increased by 18 percent.
Then they plugged in the genetic data, and "the equation changed considerably," says Loeb. Men with specific genetic landmarks — or particular variations, identified at chromosomes 10 and 19 — had a 28-percent higher risk of developing prostate cancer for every one-unit rise in PSA. Interestingly, while the men with the suspicious genes had a much higher-than-average risk using this genetic equation, prostate cancer risk was correspondingly lower for men without this genetic profile. "Men who did not have these genetic variants had only a 10-percent increase in risk for each one unit increase in PSA," says Loeb, who also notes that the relationship between these variants and prostate cancer risk depends on the PSA level.
Overall, these results suggest a couple of things, adds Carter. One is that "genetics seem to influence a man's risk of having prostate cancer at a given PSA level." Another is that one day, future prostate cancer screening might include a man's genetic profile along with the PSA measurement, "giving him a custom-tailored PSA result, and providing an even more accurate assessment of his risk for cancer."