The Philadelphia VA Hospital prostate debacle should cause us to examine how doctors make clinical decisions.  This is not a diatribe about physician failure, but an examination of the state of our knowledge about prostate disease. There is little doubt that the men treated in Philadelphia needed the treatment. They had documented cancers. It just was not done correctly and the errors not found through the proper channels of peer review.

We must ask ourselves what we know about prostate disease. How do we assess it? Currently, doctors assess the prostate for cancer by rectal digital exam and blood tests for serum prostate-specific antigen (PSA). Positive exams may indicate cancer or non-cancerous disease, or be erroneous.  These tests are done yearly in men over 50 and from age 40 on if a history of disease exists or the man is African-American.

The PSA is a protein substance found circulating in the blood and produced by the prostate gland.  Elevated levels can indicate cancer or hyperplasia of the gland (benign prostate hypertrophy– BPH).  In cancer, PSA levels are usually higher than 4 nanograms per milliliter of blood (ng/ml), but cancer is possible at any level.  Data indicates that if a man has a normal digital examination and a PSA below 4ng/ml, he has a 15% chance of having prostate cancer.  Higher PSA levels yield a higher per-cent age of cancer risk.  If the PSA if over ten ng/ml, the risk is 67%. The trend of the PSA and its time interval are also important. We must also note that most elevated PSA tests are not due to cancer.  In young men, who have smaller prostates, an elevation above 2.5 ng/ml may be meaningful.

In March 2009, a major study was reported in THE NEW ENGLAND JOURNAL OF MEDICINE. Almost 77,000 men in ten medical centers were studied over 7-10 years. In the “screened” half of men, eighty five % had both exams while the other half received “usual care” (of these men only 1/2 had exams and PSA tests).  More cancers were found in the “screened” group while both groups had their cancers found at same stages (II). They received similar treatment for the cancers.  The amazing finding is that there is no detectable difference between the “screened” versus “usual” care groups. Death rates were the same in each group.

In another study, when a PSA of 3 was used as a cutoff level, a 20 % decrease in deaths was found. There is continuing evidence that screening reduces death due to prostate cancer.  However, screening may cause patients to be treated unnecessarily. Some cancers are very lethal even when found early.  We do not understand why. The researchers suggested that we may need to use lower PSA values (2-3 ng/ml) as a cutoff now instead of four.

New avenues for screening are being sought in the face of the uncertainty of the effectiveness of current testing. An ongoing study in 155,000 people is looking at “usual” care versus “screening” for prostate, lung, colorectal and ovarian cancer. Are the biological markers we use in these diseases any better at helping us diagnosing them?  The results will be very interesting. Will “usual care” find as much disease as “screened care” and just as early like we see in the prostate study?  In a nutshell:  we still don’t have good biochemical markers of disease to use for early and specific diagnosis.