Is it time to digitally re-examine the PSA test?

Blood Test QuestionsThe following is a guest blog post by Alan Cassels.

A urologist got an opinion piece, “Bring Back Prostate Screening,” published in the New York Times last week. The author writes: “men should not wait for a government agency to tell them what’s best. My own strongest recommendation is that men insist on a baseline PSA test while in their 40s.” He could have also written, “Men should not wait for what the American Urological Association guideline panel recommends, since that panel “does not recommend routine screening in men between ages 40 to 54 years at average risk.”

That “government agency” rhetoric gets real tired after all this time, especially when it’s one-sidedly incomplete.

Yes, in May 2012, the United States Preventive Services Task Force gave the PSA test a definitive thumbs down due largely to the growing recognition of overdiagnosis.   The USPSTF said that while the “benefits of the test were small”, the harms of this simple blood test which often begins a cascade of interventions that can end in “erectile dysfunction, urinary incontinence, bowel dysfunction, and a small risk for premature death.”   This recommendation probably contributed to the significant drop in the numbers of men being tested for prostate cancer after the publication of the much publicized US Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial, which found no statistically significant effect of PSA screening on prostate cancer deaths after 13 years.

About the same time as the New York Times published the urologist’s commentary, other articles pointed out to other recent fallout from, and followup on, the concerns about overtreatment of prostate cancer.

MedPage Today reported on a study in the Journal of the American Medical Association that found that “Three out of four older men with low-risk prostate cancer opted for active surveillance during a 4-year period ending in 2013, according a large prostate cancer database.” Active surveillance is the new preferred term for used to be called watchful waiting.  The researchers wrote:

“In this analysis of a longstanding national registry, we found that after years of overtreatment for patients with low-risk prostate cancer, rates of active surveillance/watchful waiting for low-risk disease increased sharply in 2010 through 2013,” Matthew R. Cooperberg, MD, and Peter R. Carroll, MD, of the University of California San Francisco, wrote in conclusion.

“The magnitude and speed of the changes suggest a genuine change in the management of patients with prostate cancer in the United States, which could accelerate as more clinicians begin to participate in registry efforts,” they added. “Given that overtreatment of low-risk disease is a major driver of arguments against prostate cancer screening efforts, those observations may help inform a renewed discussion regarding early detection policy in the United States.”

The MPT story had this noteworthy ending:

Perhaps overlooked in the focus on active surveillance, use of radiation therapy decreased during the later years of the study period. The finding is notable, said Willie Underwood, MD, of Roswell Park Cancer Institute in Buffalo, N.Y., because it coincided with a trend toward urologist-owned radiation therapy centers, which caused some concern about financially motivated overuse of radiation therapy.

Let’s not forget the big elephant in the room. Many who recommend PSA testing tend to generally earn their living removing or radiating prostates. And all this treatment creates a huge economic burden where the total cost of care in the US was estimated at about $12 billion in 2010. As this article in the American Journal of Managed Care indicates, high-tech treatments like intensity-modulated radiotherapy (IMRT) and robotic surgery are likely being overused “because many healthcare providers are still operating in a fee-for-service environment.”

Meantime, a story in Urology Times reported on an analysis in CA: A Cancer Journal for Clinicians (2015; 65:264-82). Excerpt:

“Despite increasing use of active surveillance and watchful waiting in men with low-risk, localized prostate cancer, expectant management remains underused. …

Most low-risk patients still undergo surgery or radiation with attendant risks of long-term side effects such as erectile dysfunction and impaired urinary function, they note, even though expectant management has been shown to result in excellent long-term cancer-specific survival with minimal morbidity.

As many as 40% of patients may currently be overtreated, said senior author Mark Litwin, MD, MPH, of UCLA.

… “Active surveillance is a viable approach for most men with low-risk prostate cancer, and its broader adoption has the potential to stop the overtreatment of men with indolent lesions and redirect resources to men with more serious cancers,” the authors concluded.”

Perhaps it’s good to say that physicians are now more open to alternative ways to treat whatever it is they find with your high PSA reading and that active surveillance/watchful waiting is being offered. But really ask yourself, if it is clear that the PSA test turns many healthy men into patients, and places a sword of Damocles over their heads, where fear always looms, does PSA screening even fit the “First do no harm” credo of medical professionals?

In my mind, you can’t separate over-treatment from over-diagnosis.  As the CA journal article states, “In the PSA screening era, it is estimated that as many as 4 in 10 men are overdiagnosed with—and often are over treated for—nonlethal prostate tumors.”  That’s a lot of harm.

You might also like

Comments (5)

Please note, comments are no longer published through this website. All previously made comments are still archived and available for viewing through select posts.

Tim Bartik

July 13, 2015 at 3:15 pm

I don’t think this is a balanced commentary. It is true that men are overdiagnosed and treated for prostate cancer that would never cause them problems, and in many cases are harmed by suffering serious side-effects from the treatment. It is also true that there is a case to be made for the proposition that the benefits of prostate cancer screening exceed these harms, at some ages at least, and at some frequenquencies and some cutoffs for biopsies, and with some reasonable protocols for when diagnosis requires definitive treatment rather than active surveillance. For example, if one believes the results of the European screening study, various simulations done by people such as Ruth Etzioni and others suggest that optimal screening strategies for prostate cancer can reduce the “number needed to detect” to save one prostate cancer death to less than 5 to 1. (Gulati, Gore, & Etzioni, Annals of Internal Medicine, 2013). At those NNDs, if half the men with prostate cancer detected are harmed, then the ratio of harms to deaths is 2 or 3 to 1 — but on the other hand, many men would judge impotence to be preferable to an earlier death from prostate cancer. . Finally, if in fact we succeed in shifting more treatment towards active surveillance, then the case for PSA screening actually becomes stronger, as this will reduce the ratio of harms to deaths resulting from screening.

    Gary Schwitzer

    July 13, 2015 at 3:38 pm


    Thanks for your note. This one was written by Alan Cassels in response to a New York Times commentary that he considered particularly imbalanced.

    So we are in the marketplace of ideas, in which the hope is that “truth may emerge from the competition of ideas in free, transparent public discourse.”

    I will ask Alan to respond to you.

    July 13, 2015 at 4:55 pm

    If I am to read your comment correctly then you are saying that some men, sometimes, under some circumstances and under some reasoned protocols may have some benefit to being screened for PSA. I think that might be a reasonable statement. The issue I would make is that any man who is told their PSA test is “abnormal” becomes a patient, and many of those men will carry the dire psychological (and sometimes physical) consequences of that test for the rest of their lives without ever seeing any benefit. I would argue that in the vast majority of men the PSA test has no proven benefit and violates the ‘first do no harm’ principle of medicine.

      Tim Bartik

      July 14, 2015 at 1:43 pm

      Mr. Cassels: You say the following: “in the vast majority of men the PSA test has no proven benefit and violates the ‘first do no harm’ principle of medicine.” First, my point is that if you believe the Gulati et al. article I cited, the first part of the statement you make is empirically inconsistent with their simulation results, which rely on the European screening study. If NND is 4 to 1, then perhaps 3 men are overdiagnosed for every man whose life is saved. That is not a “vast” majority, in my view.

      Second, in debating screening, or for that matter many complicated health policy issues, we cannot adhere strictly to a “first do no harm” principle. Whether we screen or don’t screen, if the screening has ANY benefits in saving lives, and if the screening has ANY chance of overdiagnois, it is IMPOSSIBLE, regardless of which choice you make, of not harming at least some individual men. The issue is: which policy has the highest net benefits after allowing for the harms of either course of action.

July 15, 2015 at 11:55 am

In the spirit of Mark Twain who said: “Be careful of reading health books, you may die of a misprint,” let me say, “Be careful of arithmetic simulations as you may die of a math error.” I am not trying to be cheeky but just want to express my distrust of simulations. Real data are always better than models and the best data obtainable from decades of PSA testing clearly demonstrate that very, very few, if any, lives are saved and many, many lives are compromised, over diagnosed, and disfigured unnecessarily. And that’s just the physical consequences. Are there any models that measure mental anguish, dread, horror, alcoholism and suicide experienced by men who are handed a false death sentence?

As for the second point, it made me think of the oft-cited “Collateral Damage” explanation, a kind of moral escape clause that says it’s ok to kill civilians on the way to victory. I am guessing that you and I would probably disagree with what constitutes “highest net benefit”. It is probably best to agree to disagree.