Does PSA Screening Find Cancers That Could Be Better Left Alone?
by William J. Catalona, M.D.
Some material in recent articles suggest that PSA screening finds prostate cancers that are then treated when they don't need to be or when the treatment carries more risk than the presence of the cancer.
In other words, PSA screening is finding cancers, which might or could be left alone. I do not think this situation happens very often.
PSA screening does not detect many of the tiny, low-grade prostate cancers that are frequently found at autopsy in men who died of other causes. We know that about 30-50% of men over age 50 have such findings.
However, PSA screening detects cancer in only 3-5% of men over age 50 who are screened. Thus, PSA testing detects only about 10% of the cancers, and the great majority of these have the potential to cause disability or death during the patient's lifetime.
Nobody knows for certain which cancers are "potentially harmless". About 20-25% of men who are diagnosed with prostate cancer die of the disease. That does not mean that the discovery and treatment of the cancer in the remaining 75-80% was unnecessary. In many of these patients, treatment with surgery and radiotherapy cures them and prevents future disability and death; in others, hormonal therapy slows the progression sufficiently that they die from other causes.
Nevertheless, many patients do suffer and die from prostate cancer - it is the second-leading cause of death from cancer in men in the US.
Experienced clinicians can identify most men who do not need immediate treatment. These men usually have a low-volume cancer that has a low Gleason grade and, therefore, is likely to progress slowly.
If these men have a life expectancy of less than 10 years, their tumor can be monitored or treated conservatively with hormonal therapy and it is unlikely that they will experience disability or death from the cancer. But it is important to know that the cancer is there so they can be monitored carefully with serial PSA levels - the most accurate indicator of disease progression.
Using current screening techniques and varying definitions of possibly harmless cancer, only 7-25% of cancers detected through PSA testing would meet the criteria for cancers that may not need immediate treatment.
Younger patients who have a long life expectancy (10-40 years) should be treated immediately, because prostate cancer does progress relentlessly. If given sufficient time, the cancer will silently spread and become incurable.
Even if PSA testing does cause some very small number of patients to be treated for prostate cancer who may not have died from the disease, that situation does not take away from the life-saving value of PSA testing.
In the absence of large prospectively randomized clinical trials in which some people have early detection and effective treatment and other patients do not (virtually impossible to perform), it is hard to prove that PSA testing saves lives.
The declining mortality rates that we have observed in the US since the widespread use of PSA testing certainly suggest PSA screening saves lives. However, it is not hard to prove that treatment causes side effects in some patients. This latter data is used against prostate cancer screening by medical organizations that have an anti-screening bias.
Finally, the gene expression profiling studies that we are doing now may be a great help in the future to identify, by the pattern of genes that are turned on and genes that are turned off, which tumors are dangerous and need treatment, which are potentially harmless, and perhaps which are so aggressive that they may be beyond treatment with surgery or radiation.