The PSA Story: The Triumph of Science Over Common Sense

My father, who celebrated his 79th birthday yesterday, is many things: eccentric, nearly deaf (though he attributes this to my mom being a “low talker”), hilarious, crotchety, and a true mensch. But he’s neither incontinent nor impotent.

It might have been otherwise. Like most men over 50, he began having yearly PSAs checked about 20 years ago, and, in the late-1990s, one came back in the 8 ng/mL range (normal is below 4). The usual stuff ensued: a transrectal ultrasound, multiple biopsies, and a small focus of prostate cancer with a lowish Gleason score. So now what?

Although a local urologist recommended surgery, he received second (and third and fourth – courtesy of “my son, the doctor”) opinions from three superb academic urologists – Peter Carroll of UCSF, Peter Scardino at Memorial Sloan-Kettering, and later, Mark Soloway of the University of Miami. All three told him the same thing: there is no evidence that surgery or radiation improves outcomes in this situation, the treatments can cause substantial morbidity (including incontinence and impotence), and there’s a decent chance that without screening, you would have lived with this till you were 90, when the indolent cancer would have been discovered at autopsy. They all recommended that his approach to the cancer be “watch and wait.”

But each of them, at one time or another, told me something like, “I give this advice to all the men in your dad’s situation, and 90% of them end up getting surgery or XRT. Sometimes, it’s because the clinical situation changes. But far more often, it’s because they just can’t stand the waiting – and the peer pressure – any longer.”

My dad’s a pretty tough cookie, and he decided to wait. Despite a couple of scares through the years, he has held out. And a recent re-biopsy, 11 years after the original, found no evidence of cancer.

I am proud of him for waiting, particularly in the face of all the well-meaning advice he received from the tennis court and cocktail party crowds: “Murray, how can you stand it?”  “Did you hear about the new radiation machine they just got at the hospital?” “Mur, you hear about Joe – he’s dying of prostate cancer. How can you let yours just sit there!”?

It does seem like plain-old common sense that a program of screening for a tumor, finding it early, and treating it aggressively would be the right thing to do. After all, this is CANCER we’re talking about!

Well, this commonsensical approach is probably wrong. This week’s New England Journal includes the two most rigorous studies to date addressing the value of PSA screening.  One, a European study, found that PSA screening led to little benefit; one person in 50 in whom cancer was found through screening was apparently cured through early treatment (an absolute reduction of 7 prostate cancer deaths for every 10,000 men screened). Unfortunately, this small benefit came at the cost of significant treatment-related morbidity (and even mortality) for the other 49 patients.

And the American study found that, after 7 years of follow-up, the unscreened group had a 13% lower mortality rate than the screened group, with not a whiff of benefit.

The accompanying editorial by Michael Barry of Mass General summed things up this way:

Serial PSA screening has at best a modest effect on prostate-cancer mortality during the first decade of follow-up. This benefit comes at the cost of substantial overdiagnosis and overtreatment. It is important to remember that the key question is not whether PSA screening is effective but whether it does more good than harm.

Remember our recent discussion about comparative effectiveness? The question I raised was whether we were mature enough to act on data like these. Let’s say that the two additional huge PSA screening trials currently in progress confirm the findings of this week’s American study: no benefit. Will payers have the guts and Kevlar to stop paying for PSA screening? And even if they do, the issue is far more complex than simply withholding payment for PSA testing: the cost of the PSA assay (about $100) is trivia compared with the cost of follow-up testing, biopsies, and ultimately of treatments such as surgery and radiation therapy. Even if one envisioned a system in which insurers didn’t cover PSA testing, it is hard to imagine one in which they wouldn’t cover subsequent costs of diagnosis and therapy.

Returning to the present, I see these NEJM studies as a triumph of science over common sense. Common sense tells us that PSA screening should be effective, and early observational studies and mortality trends over the past 15 years hinted at benefit. But like estrogen before it (and rapid response teams, for that matter), not everything that should work does, an argument for applying scientific methods, particularly when the test or treatment at issue is morbid or expensive.

A tip of the hat to Gil Welch, the Dartmouth internist / epidemiologist who has been a lone voice in the woods for the past decade, arguing that PSA screening (and perhaps screening for other cancers) might lead to more harm than benefit. He has pointed out, for example, that PSA screening could generate apparent improvements in 5-year survival for prostate cancer through lead-time bias (patients would die at the same time as they would have without screening, but the earlier diagnosis makes it appear that they lived longer). And in his superb 2006 book, Should I Be Tested for Cancer?: Maybe Not and Here’s Why, he wrote:

I object to the emerging mindset that patients should be persuaded, frightened, and coerced into undergoing [mammography, PSA tests, fecal occult blood testing]. There is today a certain “medical correctness” about screening – making patients feel guilty if they choose not to pursue testing. This is wrong.

Turning more specifically to PSA screening and prostate cancer, Welch added,

… In the final analysis, most of these one million extra cases of prostate cancer [diagnosed through PSA screening] represent nothing more than pseudodisease: disease that would never progress enough to cause symptoms—or flat-out would never progress at all… I should be clear that none of this should be interpreted as saying that PSA testing will never work. In fact, while many men may have been hurt by the ensuing treatment, a few others may have been helped. And it is possible that the PSA test can be modified such that it finds much less cancer and therefore makes much more sense. But right now we need to be clearer about the primary effect of the way it has been used so far: it has been making men sick.

A heartfelt thanks to my dad’s urologists, who gave him what proved to be sage advice when all of the pressures – social, psychological, and financial – might have steered them toward more aggressive recommendations.

So happy birthday, dad. So glad that you’re well, and that all the plumbing still works.

10 Responses to “The PSA Story: The Triumph of Science Over Common Sense”

  1. mort schulman March 21, 2009 at 10:05 pm #

    Everthing  you said about Murray is true and a real Mench !! From a tennis partner on the deck.    Mort Schulman

  2. BOB CRANE March 22, 2009 at 2:39 pm #

    MURRAY.

    MY PSA HAS ALWAYS BEEN BETWEEN 5.0 AND 7.0. AT MY AGE 86.

    DOCTOR SAYS FORGET IT.

    NO MORE PSA’S

    DIE FROM NATURAL CAUSES , NOT CANCER.

    KEEP ‘EM LAUGHING , MAKES US ALL SEEM MORE ALIVE

    ella and bob

  3. menoalittle March 23, 2009 at 5:33 am #

    Bob,

    Your presentation of how another ideological approach to medical care has gone by the wayside after being methodologically studied is timely. Another “common sense” based costly strategy of an “open artery” (coronary, that is) after heart attack was also recently dismissed by a scientific study by a low talker, fellow New Yorker, cardiology researcher, J. Hochman.

    Using your language expressing the sentinel medical science principle that not everything that should work does, with a few relevant word changes…”common sense tells us (and them, Obama, Tullman, DeParle, Blumentha, Hamburg, Sibelius) that CPOE administered medical care should be effective, and early observational studies and handwriting trends over the past 15 years hinted at benefit…not everything that should work does.”

    So where is the scientific method determining safety and efficacy of CPOE machines prior to our government spending (wasting) $billions and endangering many?

    There has got to be more than one Gil Welch think alike out there on CPOE and other HIT. Are you there?

    Best regards and go for par, Murray,

    Menoalittle

  4. jnmed March 26, 2009 at 3:03 pm #

    Screening, early detection and prevention are so intellectually and emotionally appealing, that it has been easy for us to justify elaborate and expensive cost centers just to meet these goals. But how do we stop the train at this point? Whole institutions survive on inappropriate angiography, endoscopy, and (some) cancer screening-

    Even much of the lay press has been reluctant to say what you have said above- there is still a lot of hedging about the conclusion.

    Maybe we should also focus on preventing other trains from leaving the station (cardiac CT, some PET scanning, dynamic nuclear imaging of coronary plaques)- until we have some serious outcomes analysis.

    Like menoalittle, I am also skeptical about the rush to the national EMR/CPOE. Of course it seems to be a no-brainer (who could be against nice handwriting), but the reality of the current products, the difficult transition, the ability to integrate linear/algorithmic technologies into our chaotic and often narrative clinical world- this is another expensive area where common sense may have too strong a position

    Welch has been wonderfully ahead of this curve. Thanks for giving him a little press.

  5. davisliumd April 6, 2009 at 3:58 pm #

    Glad to hear Dad is well. Gutsy call. The peer pressure alone from friends and other doctors, who all mean well, to stand up against the “war on cancer” and instead opting for watchful waiting can give moments of doubt and uncertainty one is doing the right thing.

    Shannon Brownlee’s Reader’s Digest article is right on the money, although the quote of complications from colonoscopy might be a little misleading. The complication rate of 5 in 1000 was based on a retrospective study on patients who had colonoscopies for any indication, rather than those who had the procedure done simply for screening.

    Davis Liu, M.D.
    Author of Stay Heatlhy, Live Longer, Spend Wisely – Making Intelligent Choices in America’s Healthcare System

  6. WRS April 19, 2009 at 5:45 pm #

    From the decision of the Merenstein case, it seems foolhardy to travel down this road–even if it is a best practice. Pretty scary stuff.

    http://www.annfammed.org/cgi/reprint/5/2/120
    http://web.ebscohost.com/ehost/pdf?vid=4&hid=117&sid=7121a4ff-417e-4a84-9811-12cca4098e94%40sessionmgr104

  7. bladder weakness products August 16, 2009 at 12:53 pm #

    Early screening has its ups and downs. More for the ups because the early the better. But sometimes, to early learn the sickness may give more anxiety to the patient because he might self-pity or feel down. An example is bladder weakness. This is an embarrassment although this is but normal especially to our seniors. But because of the wrong perception, instead of curing the illness, it may become worse if not properly address to. Good for you because you are always by your dad’s side.
    http://www.dryforlife.co.uk

  8. Robert August 21, 2009 at 6:20 pm #

    My recent psa was 6.6, up from 1.3 a year before. I am 57 years old, no fam history, DRE normal, some BPH noted. I was immediately referred to a urologist who specializes in robotic prostate removal. He wanted to schedule a biopsy which I refused, opting for another test in 6 weeks. He plan is to biopsy if not under 3.0. If negative, another biopsy in 3 months. If negative, another one in 3 months. keep looking until you find something I guess. I have the impresion that he will not be happy until he has every man in the states prostate in a jar on his desk. I will find another urologist before the 6 week test.

  9. David April 17, 2012 at 4:57 pm #

    My PSA rose from 2-4 over 11 years; on biopsy all samples were postive for low grade CA; had robotic prostatectomy, doing well. I know i shall die of course but hopefully not from very painful debilitating prostate carcinoma.

  10. Randi December 26, 2012 at 5:20 pm #

    I need to to thank you for this wonderful read!

    ! I absolutely enjoyed every little bit of it.
    I’ve got you bookmarked to check out new stuff you post…

Leave a Reply