U.S. Panel Advises Against Routine Prostate Test
Healthy men should no longer receive a P.S.A. blood test to screen for prostate cancer because the test does not save lives over all and often leads to more tests and treatments that needlessly cause pain, impotence and incontinence in many, a key government health panel has decided.
How can this be? Everyone knows that finding cancer early is important. The PSA test must save lives.
Those of us who study diagnostic tests understand that all tests have probabilities associated. A high PSA level does not mean that a man has prostate cancer. It means that he has an increased probability of prostate cancer. It also does not discriminate between aggressive prostate cancer and slow growing prostate cancer.
What do the studies tell us?
As the P.S.A. test has grown in popularity, the devastating consequences of the biopsies and treatments that often flow from the test have become increasingly apparent. From 1986 through 2005, one million men received surgery, radiation therapy or both who would not have been treated without a P.S.A. test, according to the task force. Among them, at least 5,000 died soon after surgery and 10,000 to 70,000 suffered serious complications. Half had persistent blood in their semen, and 200,000 to 300,000 suffered impotence, incontinence or both. As a result of these complications, the man who developed the test, Dr. Richard J. Ablin, has called its widespread use a “public health disaster.”
Many physicians have understood these results for several years. I have refused prostate cancer screening since turning 50, because I understood two factors – the test has mediocre characteristics and prostate cancer treatment has significant side effects.
Now many have and will decry this result. Many cannot believe that PSA testing could possibly do harm. But when one examines the data, the proper path becomes clear.
{ 7 comments… read them below or add one }
But when one examines the data …
Have you really examined the data? You cite “the test does not save lives” but in the ERSPC trial the number need to treat was 48. While this is not good, it doesn’t mean that lives aren’t saved. Have you examined the PLCO data, or follow up data? Can you explain the decrease in prostate cancer deaths?
If the PSA is elevated, then doctors can help empowered patients make good decisions about a biopsy. If the biopsy is positive, then doctors can help empowered patients make good decisions about treatment. Instead, you are arguing that we just keep patients in the dark to save us all that trouble. Sure that dooms a large number of men (men like me) to die of cancer, but it's so much easier that way.
That is irresponsible, unprofessional, and unethical.
Jan, I haven't examined the data as well as I probably should. But my understanding was that NNT of 48 was for *treatment* of prostate cancer, not screening. Unless I have misunderstood, this would be an apples-to-oranges comparison, and outside the scope of the USPSTF.
The issue is not whether patients should be kept in the dark. Of course they shouldn't. The issue is whether the PSA actually sheds any light. So far, PSA's have provided mostly a lot of confusion. I think the PSA is a very flawed test with a huge number of false positives and false negatives.
I think it's unfair to say that anybody is dooming men to die of prostate cancer, when there is little, if any evidence that routine PSA testing saves any lives. Annual PSA testing also "dooms" many men to incontinence and impotence, who would never have never been killed by prostate cancer.
PSA is an antigen that is released by normal, healthy prostates. Sometimes healthy prostates will release higher-than usual levels. Interpreting such a test will always be fraught with difficulty. Hopefully, some day, people smarter than me will find a screening test better than the PSA. (This may require a revolution in the understanding of the biology of prostate cancer). Until then, as a primary care physician, I have to work with the tools I have. PSA screening is just not a useful tool. It would be unethical and irresponsible to pretend that it is.
"there is little, if any evidence that routine PSA testing saves any lives"
This is the problem. Yes there is. There is research of an effect across whole populations. But more immediately: I am evidence. I know other men like me who had treatment and who — given their age, tumor volume, and Gleason score — will almost certainly live much longer because of that treatment. For men like us, the side effects are worth the treatment effect. The majority of prostate cancer is slow-growing and occurs in older men, but not all of it. If you average us all together then some of us will die.
Right now too many men are treating prostate cancer too aggressively. Many men are getting treatments they don't need and that leads to side effects they shouldn't have to suffer. For other men, though, a deadly prostate cancer is asymptomatic. This is true in a significant minority of cases. To not test IS to doom these men to die of prostate cancer.
I'm glad you're not my primary care physician. Your position on this issue — which is based entirely on statistics and ignores individual differences – would almost certainly have doomed me to die of prostate cancer. How is that ethical and responsible? The PSA test saved my life. How is that not useful?
What we need to do is improve decision-making about when to have a biopsy and when to treat prostate cancer. Just ignoring the information because some people will make the wrong choice is irresponsible.
The American Urological Association states it better than I do:
http://www.auanet.org/content/press/press_releases/article.cfm?articleNo=262
Screening trials with insufficient sample sizes might show a lowering of cancer-specific mortality but not detect increases in all-cause mortality related to screening. Studies of a manageable size have too little discriminatory power and last a long time. Furthermore, results become available decades after trial initiation, by which time they are probably antiquated. Whether screening for prostate cancer is beneficial cannot be assessed in trials, a statement that might also be true for other diseases with low specific mortality…
{ 2 trackbacks }