The over-treatment of low-risk disease in the Medicare community

So there’s a lot of media coverage this morning about an article in the Journal of the American Medical Association suggesting that, here in the USA, more men with low-risk forms of localized prostate cancer are getting more expensive types of treatment for their cancer (and that much of this treatment may be of dubious value).

The full text of this article by Jacobs et al. (“Use of advanced treatment technologies among men at low risk of dying from prostate cancer“) is freely available on line, and, to be frank, it doesn’t really tell us anything new if one has any good degree of familiarity with the complexities of the diagnosis and management of prostate cancer. What it does do, on the other hand, is give us some more concrete data about the over-treatment of low-risk disease in the period from 2004 to 2009. We would encourage all prostate cancer support group leaders and educators to have a close look at this article.

For those who are interested in other media “takes” on this article, it is covered in stories on the web sites of Reuters, MedPage Today, HealthDay, and Medscape (to single out just four).

What Jacobs et al. did was conduct a retrospective analysis of data available from the SEER-Medicare database on the management of men diagnosed with prostate cancer in the 6-year period beginning in 2004. They looked very specifically at the patients’ diagnostic risk level and at their risk of dying from their prostate cancer within 10 years as well as the type of management implemented. It is important to recognize, however, that all these patients were 66 years or older at the time of their initial diagnosis and treatment.

Here are the core findings of the study:

  • The study encompassed a total of 83,191 patients for whom follow-up data was available through December 31, 2010.
  • The authors excluded a total of 27,244 patients, including
    • Men who had metastatic disease at diagnosis (n = 2,261)
    • Men who had other non-designated forms of treatment (n = 21,898), e.g., all men treated with proton beam radiation therapy or brachytherapy
    • Men who had important clinical (n = 1,647) or demographic (n = 1,078) data missing
    • Men who were ineligible for some other reason (n = 360)
  • Specifically, the authors looked at detailed data for 55,947 men treated with
    • Observation (n = 16,384)
    • Intensity-modulated radiation therapy or IMRT (n = 23,633)
    • Traditional external beam radiation therapy or EBRT (n = 3,926)
    • Robot-assisted laparoscopic radical prostatectomy or RALP (n = 5,881)
    • Traditional (“open”) radical prostatectomy (n = 6,123)
  • The projected 10-year mortality rates of these men were:
    • 19 percent were at low-risk of death within 10 years
    • 61 percent were at intermediate-risk of death within 10 years
    • 91 percent were at high-risk of death within 10 years

Now the actual paper contains a vast amount of detailed data broken down by patient age, Gleason score, D’Amico risk classification, where in the country they were treated, etc. However, here are the primary results provided by the authors:

  • Between 2004 and 2009, compared to the use of observation, EBRT, and open radical prostatectomy,
    • The use of IMRT or RALP increased from 32 to 44 percent among men with low-risk prostate cancer
    • The use of IMRT or RALP increased from 36 to 57 percent among men at high risk for non-cancer-specific death.
    • The use of IMRT or RALP increased from 25 to 34 percent among men with low-risk prostate cancer and a high risk for non-cancer-specific death.

The bottom line seems pretty straightforward. We were over-treating about a quarter of these patients in 2004 and by 2009 we were over-treating more like a third of them with more expensive forms of care!

Having said that, it is actually quite gratifying to see what a relatively small percentage of these patients of 66 and older received surgical treatment during the time period (just 12 percent) and what a relatively high percentage were managed on observation of some type (nearly 30 percent). However, the proportion of all patients who received observation as their first-line management during the 6-year period remained stable at about 16-18 percent of patients per year.

One of the key conclusions of this paper is worth quoting in full here:

Continued efforts to differentiate indolent from aggressive disease and to improve the prediction of patient life expectancy may help to reduce the use of advanced treatment technologies in this patient population.

This is a conclusion with which The “New” Prostate Cancer InfoLink is in complete agreement.

We would also point out that the data in this paper is not in any way telling patients that they shouldn’t receive treatment just because they don’t need it. The take home message is that the medical and the advocacy communities need to do a much better job of explaining the risks and benefits of treatment as compared to monitoring … most specifically for men of 65 and older who are diagnosed with low-risk disease and who have a life expectancy of 10 years or less.

13 Responses

  1. The figures looked really strange — far too high, so I checked the paper itself. These are the percentages for what seems to be co-morbidity terciles for their 10-year survival calculation, rather than figures for D’Amico risk categories.

    I’m including a couple of excerpts that say this, though not as clearly as could be said:

    “Using this sample, we modeled the probability of dying within 10 years based on age, race, comorbidity, socioeconomic class, area of residence (eg, rural vs urban), and SEER region. This model was internally cross-validated using a bootstrap aggregating technique. The discrimination of the prediction model was high (C index = 0.90). We applied this model to men diagnosed with prostate cancer as their first and only cancer to estimate their probability of dying within 10 years. We sorted patients into terciles based on this mortality probability.”

    “When we applied our predictive model to our cohort of prostate cancer patients, the predicted mortality rates at 10 years were 19%, 61%, and 91% for patients in low, intermediate, and high terciles, respectively ….”




    I’m trying to get a handle on the prostate cancer risk level context for this study. My initial approach was that the study would involve typical risk levels for a population. However, it appears that risk level was unusually skewed toward the higher level, and I’m not sure why that is the case. Here’s one key fact I noticed: “poorly differentiated or undifferentiated tumors”, I’m thinking Gleason 8-10, but possibly this means 7-10, account for about two-thirds each of the IMRT, EBRT, robotic, and open groups, therefore all definitely on the higher than average side, and even accounts for about a third of the observation group (also unusually high for observation). This is from the table on the demographic and clinical characteristics of this population.

    I’m picking up that this study appears to be strong on co-morbidity factors and how they should influence treatment decision making, with age perhaps effectively acting as a proxy for co-morbidity. (I can sure identify with that at age 70!) I’m thinking that age is perhaps a key factor altering the risk picture, skewing the study population toward higher prostate cancer risk, as only men aged 66 or older were included.


    Once again the PLCO and ERSPC screening studies, and the US Preventive Services Task Force recommendation essentially against screening, were all mentioned as if they provided substantial help and evidence in decision making on screening and treatment. Clearly, some clues can be gleaned from the two studies, but they were mentioned with no caveats about critical flaws that severely impact interpretation. I look forward to the day when papers put those studies in their proper place, and ignore the profoundly flawed recommendation of the Task Force.

  4. Jim … I’ll have another look at this. … Parts of the paper are very poorly written (in my personal opinion).


    Here are the problematic sentences in the “Discussion” section of the paper:

    “… Research and policy change represent 2 avenues that could affect the current treatment patterns for prostate cancer. The Surveillance Therapy Against Radical Treatment (START) trial is a randomized controlled trial that will provide valuable information regarding the effectiveness of radiation, surgery, and active surveillance for low-risk prostate cancer. However, because of the protracted natural history of prostate cancer, results from this trial are not expected for another 5 to 10 years. …”

    The paper does provide a convenient hot link to the trial, but unfortunately, that link reveals the START trial was terminated due to failure to accrue enough patients. The trial overview site was last updated in January, and the paper’s authors visited on May 13, well after the termination notice was posted. Jeez guys! — a bit sloppy; another instance of what Sitemaster noted.

    Moreover, it may be just as well that the START trial is not going forward as it would likely have just added more misinformation to the treatment vs. surveillance picture. That’s because the primary outcome measure was disease-specific survival with a time frame of 10 years. That time frame would have guaranteed the results would have been premature, in view of what is currently routinely achievable with prostate cancer in the US, especially low-risk prostate cancer. Do we not know, based on statistics of a few years ago, that 10-year survival for all comers (all risk levels) is 98%, and even 93% at 15 years (ACS statistics)? That means that a likely maximum of only 2% survival is in play at all with 10-year follow-up, and even that is almost surely among higher risk (or very poorly treated or very unlucky) patients, such as the small fraction of a percent who die during or shortly after surgery. And, not to pile on (but I can’t help it), that 98% figure is based on results from a time frame a few years ago, with even better results likely achievable for patients who would have been treated or observed during the period of the START trial.

    Those of us who are on panels reviewing trial proposals need to get tough with trials whose protocols guarantee to produce premature results.


    The context is that the paper establishes that the more advanced modes of care were used with impressively increasing frequency over the period covered by the study.

    The paper addresses likely reasons in the discussion section, making a number of thoughtful points, but also noting in part that “… Second, a perceived improvement in outcomes compared with prior alternatives may make these advanced treatment technologies seem more palatable. However, comparative studies have shown that the advantages of these newer treatments are marginal at best” (with two reference, nos. 3 and 4).

    Reference 3, dated 2009, compares robotic with open RP side effect and complication outcomes, and reference 4, dated 2012, does the same for EBRT (specifically 3D-CRT) versus IMRT. I’ve read both abstracts, and have a fair amount of background knowledge from other sources, having had to choose a therapy myself in the past year. My impression is that robotic surgery was getting great press until fairly recently, when enthusiasm somewhat declined. Even reference 3, in my eyes, presents a balance of side effects/complications that tips clearly in favor of robotic. Regarding radiation, technology has been steadily evolving, and my impression is that IMRT, even a few years ago, was demonstrating a substantially lower incidence of side effects and complications coupled with the ability to deliver a dose level (especially 78 Gy and somewhat higher) shown by research to be more effective. Even reference 4 did note superiority in side effects/complications for IMRT for higher-risk men who did not get ADT (though non-superiority for others). Moreover, the expanding use of advanced targeting and positioning technology has only occurred, I believe, in recent years, essentially beyond the ref. 4 study period of 2001-07; that improved technology, while not yet proven conclusively by research, shows solid promise of further driving down complications and side effects. On a personal note, I am quite pleased with my own situation at the 2-month point post-TomoTherapy that included pelvic radiation and on-going ADT for a challenging case, now considered likely curable.

    My point is that the advantages of the new therapies are more than marginal as far as side effects/complications go, contrary to the thrust of the paper’s discussion. However, the paper’s main focus is the issue whether too many men unlikely to benefit from any therapy are still opting for therapy. That’s another question. Despite faults in the paper, I’m thinking the authors have added evidence that too many men are choosing therapy over observation.

  7. Dear Jim:

    I think you are probably correct about the risk tertiles being risk of 10-year mortality and I have adjusted the text above accordingly. However, I still have a hard time following some of the detail in this paper.

  8. Can someone please tell me what we do for younger guys? I’m 46. I find very little information/guidance on what to do if you’re on the younger end of this disease.

    I’m Gleason 6 (3 + 3); clinical stage T1c; 2/12 cores with prostate cancer (20%; left base); PSA = 2.8; negative DRE; asymptomatic.

    Healthy and very active. Lots of exercise. I have improved my diet: no red meat, no eggs, very little dairy, very little processed food. Eat tomoato products and cruciferous veggies. Drink green tea. Take Pomi-T tablets.

  9. Mac:

    No one can tell you what to do with absolute certainty. However, you have a lot of options, and if you join our social network we will be happy to discuss those options with you in as much detail as you like.

  10. Thank you. I will join now. I’m looking for advice/direction. It seems (and rightfully so) that all of the testing and studies are done on older men. We have all of these drugs that are prescribed, but it seems they are only prescribed when the prostate cancer has metastasized. Why not take them now, while it’s still in the prostate? Why wait for it to spread and then start taking them. Thanks. Good health to all.

  11. How did such a paper, with obvious conclusions (poorly expressed) make it into the AMA journal, such a rigorous, top journal? A statement of the obvious?

  12. Jim correctly said: “My point is that the advantages of the new therapies are more than marginal as far as side effects/complications go, contrary to the thrust of the paper’s discussion. However, the paper’s main focus is the issue whether too many men unlikely to benefit from any therapy are still opting for therapy. That’s another question. Despite faults in the paper, I’m thinking the authors have added evidence that too many men are choosing therapy over observation.”

    The peculiar thing is that the cohort of men selected in the study were mostly intermediate- or high-risk patients. Men with low-risk disease were observed in their majority.

    For all the talk of low-risk disease being diagnosed and treated these days, the study reported a treatment cohort (all treatments new and old) of 39,563 men. Of those, 29,119 (73.6%) were reported to have intermediate- or high-risk disease (intermediate-risk disease defined as clinical stage T2b and/or Gleason 7 and/or a PSA 10.1 to 20.0 ng/ml; high-risk disease defined as clinical stage of T2c or higher and/or Gleason 8 to 10 and/or a PSA > 20 ng/ml).

    Also, of the 39,563 treated patients, some 65% were reported to have poorly or undifferentiated disease.

    Seems to me that the data showed that most of these patients did require treatment. Where are those 80%+ over-diagnosed and over-treated patients we hear about all the time…

    BTW: Tarhoosier it is a hard to read paper.

  13. Dear Ralph:

    All these men were > 65 years of age, so the study has excluded all men aged between 50 and 65, who are way more likely to be diagnosed with early stage, D’Amico low-risk disease. In addition, I don’t think that anyone has been suggesting that 80+% of patients are being over-diagnosed and over-treated.

    The percentage of men with D’Amico low-risk disease who were observed was only 8,871/19,315 or 45.9 percent (regardless of their life expectancy). By contrast, and correctly, the proportion of men at low risk of non-prostate cancer-specific mortality who were given treatment was (appropriately) high at 13,906/18,019 or 77.2 percent.

    Also, 13,035/20,329 men at high risk of non-prostate cancer-specific mortality within 10 years (64.1 percent) and another 14,058/19,035 men at intermediate-risk of non-prostate cancer-specific death within 10 years (73.9 percent) received definitive first-line treatment that had a relatively low probability of affecting their risk for metastasis or overall survival within that time frame. Isn’t that over-treatment? Most of those men probably got little to no survival benefit from their treatment compared to observation and treatment with ADT if and when they progressed (potentially for a couple of years).

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: