How many men are being over-treated?

Are we over-treating 50,000 men each year? Could it be even more?

There is absolutely no doubt in the mind of this writer that we “over-treat” prostate cancer. By that we mean that a significant percentage of the men who are newly diagnosed each year with localized disease, and who go on to get some invasive form of treatment (surgery, external beam radiation, brachytherapy, cryotherapy, HIFU, whatever) will never actually “benefit” clinically from that treatment because their cancer is actually relatively indolent and is never going to have a significant clinical impact on their lives. It is also likely that the older the patient, the more accurate that this probability becomes simply because the patient’s life expectancy is shorter and so the chance that he will have disease that progresses to clinical significance is smaller.

An article by Welch and Albertsen in this week’s issue of the Journal of the National Cancer Institute has now provided an estimate of just how much “over-treatment” occurred during the 20-year period from 1986 and 2005. It is their belief that it is of the order of 1 million men (an average of about 50,000 per year) or higher — and we have no reason to disagree with them.

We aren’t going to get into the pros and cons of the process used by Drs. Welch and Albertsen to make this specific estimate. They themselves discuss the potential problems with their estimating process in a very reasonable manner. The point that they are trying to make is that, of the approximately 200,000 newly diagnosed prostate cancer patients in America every year, about 25 percent (and maybe many more) will be subjected to the risks and adverse consequences of forms of treatment but will have no possible benefit from the treatment because they were never ever going to have clinically significant disease.

Prostate cancer advocates may not like this sort of computation and these sorts of data, but they are certainly true in principle and very probably relatively accurate, and we are all going to need to learn to live with them.

By contrast, as we reported last week, we also know that — over exactly the same time frame — the use of the PSA test has massively reduced the likelihood that a newly diagnosed patient will be diagnosed with advanced as opposed to localize disease. We noted at the time the implication that this at least allowed the patient to be aware of his risk and the treating physician to work with the patient to manage the risk of progression as opposed to progressive disease itself.

The problem, of course, is that (on an individual basis) we are utterly unable to distinguish between the patients who are at minimal risk for clinically significant disease and those at notably higher risk for clinically significant disease based on any known test available today (including the DRE, the PSA test and its derivatives, and even a biopsy result). The consequence, not unnaturally, is that men who are told they are at some degree of risk, and who have a PSA test, and a “positive” indicator of risk based on that PSA are highly likely to have a biopsy. And if that biopsy is positive, they are highly likely to have treatment. And this is a very reasonable behavior, because as humans we fear the potential ramifications of cancer.

In his editorial accompanying the Welch-Albertsen article, Otis Brawley, who is not an advocate for mass, population-based, prostate cancer screening, finally makes the same clear point that The “New” Prostate Cancer InfoLink has been making since we came back on line early last year. So that we are very clear about what Dr. Brawley wrote, the following is the precise quote from his editorial:

We desperately need the ability to predict which patient has a localized cancer that is going to metastasize and cause suffering and death and which patient has a cancer that is destined to stay in the patient’s prostate for the remainder of his life.

Until such a test is available, all of the discussion about whether widespread screening is justified is simply noise, because prostate cancer kills 30,000 men ever year, and the only way we know to prevent its progression is to treat it early. Since we have the ability to detect its existence, we will inevitably continue to treat it early — except in the cases of those men who decide that they will risk tossing the dice and not get early PSA tests. And based on  the available data, this decision is just as justifiable as getting a diagnosis as a consequence 0f a PSA test and a biopsy and then tossing the dice about the outcome of early treatment.

Now if economics was the only factor under consideration it might be possible to justify the refusal to cover PSA testing as a means to suggest risk for prostate cancer. But economics isn’t the only indicator, and as Grace Lu-Yao and her colleagues very clearly demonstrated last week, the use of the PSA test has significantly shifted the stage at which prostate cancer is diagnosed. If we can’t come up with a better way to differentiate between high risk and indolent prostate cancer (and quickly), then we need to come up with a better way to manage prostate cancer based on the information currently available. The academic discussion is of intellectual interest, but is not a basis for clinical decision making. We need a way to help men and their doctors make the right decision about management once a diagnosis is made (as opposed to telling them that diagnosing prostate cancer may be  the wrong thing to do).

Is early prostate cancer over-aggressively treated? Absolutely. So let’s focus on the real problem, which is how to manage the expectations of the newly diagnosed patient in a constructive manner. Not giving men PSA tests is simply going to lead to a gold rush for lawyers.

If this is a teachable moment (as Dr. Brawley clearly hopes it is), the point to be taught is that prostate cancer is a complicated and difficult disease to manage, and we need to do a great deal better than we have to date. Abandoning the PSA test because it is far from perfect will solve nothing.

22 Responses

  1. Terrific analysis, Mike. Very timely here in New Zealand, where a Parliamentary inquiry into this question will begin its hearing of submissions on September 16.
    I’ll be reporting that, of course.

    Fascinating that the chairman of the committee, Dr Paul Hutchison, MP, has published an article today, in our second-largest newspaper, making his views on the issue all too plain.

  2. I’ve always said that over a million men have been over-treated, a fact that has been vehemently denied in the past by those who see any such statement as an attack on the staus quo.

    It is as difficult to measure the potential number of lives saved, but this seems on the highest estimate to be in a ratio of about 8:1.

    Is that a reasonable estimate and shouldn’t this kind of information be included in the discussions with men when they have their PSA test?

    I have always found it interesting that in a country like South Africa, doctors cannot do a test for HIV unless they have been through a prescribed counselling session to tell the people what the llikely oucomes and options might be.

    Surely, until the mythical better test is developed, the least that could be done is to ensure that better informaiton is provided with the test?

  3. This is a fine article. I believe every word that is written. When I look back over the last 2 years and the 4 doctors and the biopsy, cryosurgery, radiation treatments, and all the pain, I truly believe that I would have had a better quality of life doing nothing. Especially at age 71 when I was diagnosed. Doctors tend to overreact to prostate cancer in the name of the “lesser of two evils.” There must be a better way. Especially for older patients.

  4. The Welch-Albertsen paper suggests that the number of lives saved may be of the order of 1 for every 50 men over-diagnosed. However, there are a lot of assumptions being made in creation of this estimate. It could be a lot higher.

  5. Over-treatment wouldn’t be so bad if the treatments most men are given didn’t have such high morbidity.

    It’s one thing to be given a treatment you didn’t need, another when that treatment causes ED, incontinence, or bowel problems.

    And just yesterday, I received a flier in the mail from the preeminent local hospital: “A prostate screening will take only 30 minutes out of your Saturday, but it could add years to your life. The American Cancer Society and the Maine Urologic Association recommend an annual screening for men between the ages of 50 and 74. African-American men and all men with a family history of prostate cancer should consider screening annually beginning at age 40.”

    “…years to your life” — at what price?

  6. I estimated from [information at] just at one website as many as 30% were over treated…. Sounds line with this post. But when I saw the use of the term “over-diagnosed,” it still makes my skin crawl to hear that…. No such thing as “over-diagnosed.” “Over-treated,” yes, but not over diagnosed. If 50,000 a year are over-treated, then that means 150,000 are not?

  7. I will continue to maintain that until a test comes along that is more conclusive than the PSA and DRE to identify developing cancer, we need the PSA and DRE. In my opinion, the term “over-treatment” only applies if the physician diagnosing the existence of cancer encourages immediate treatment whether necessary or not. The onus lies on the physicians. If a man is diagnosed with earliest development of prostate cancer and chooses to be treated despite likely having years, if not a lifetime, with treatment not necessary, that is NOT over-treatment. That is patient choice. The AUA needs to pressure (police?) its members to be fully explanatory to patients of their options, with emphasis on active surveillance (or whatever one chooses to call it) when patient diagnostics indicate. Even with all the current attention on over-treatment, patients are reporting that active surveillance is not even mentioned as an option.

  8. Mike,
    I also have no doubt that PCa is over-treated. So are many other diseases. The key is to what extent is really the disease over-treated. There is a tendency to treat diseases aggressively. Part of it is that this is a safeguard against malpractice. Physicians should be able to police themselves and weed out the bad apples…. Same for lawyers!

    That said, with all the uncertainty present in current diagnostics I am hard pressed to understand how you know (and the study authors and chorus) that the over-treated patients will never benefit. Studies have demonstrated that prostate cancer and other diseases lay dormant and recur many years later. How can Welch, Albertsen, Brawley, and Scott have such certainty? You do mention age in passing, but this is something that has a tremendous impact in PCa mortality.

    You said “we have no reason to disagree” in reference to the magnitude of over-treatment during the 1986 to 2005 period. I do not see how those numbers could be accurate. The period of 1986 to 1994 yielded many diagnoses of more advanced disease. Treatment of those high-risk cases had a high treatment failure rate. Those men had significant clinical stages of prostate cancer. The peak for incidence was during those early 1990s and so was the peak for treatment. The shift in stage started after the pool of undiagnosed disease was reduced. This coincided with a reduction in PCa mortality that continues today.

    The question we need to answer is if over-treatment of indolent disease is as prevalent as they estimate, why the current drop in mortality? A few basic questions pertinent to this issue are:

    1. How indolent really is prostate cancer?
    2. What is the natural history of the untreated disease?
    3. How is age significant for the development of clinical disease and disease-specific mortality?
    4. How can treatment of indolent disease impact the mortality rate?
    5. Is prostate cancer a progressive disease of variable growth potential?
    6. Why is it so hard to admit that early detection with an imperfect marker and effective treatment is related to reducing PCa deaths? It has been like extracting water from a rock…

    Admitting that this imperfect marker has reduced the risk of an advanced disease diagnosis is a difficult enough issue for those that deny the value of PSA. It is then that the issue of over-diagnosis and over-treatment is brought to the forefront.

    The PSA test is not the problem. Overdiagnosis is not the problem. The problem is that no one currently knows how to accurately differentiate between “indolent” disease and disease that can ultimately kill the patient given enough time. Somehow the authors seem to have solved the problem … maybe a crystal ball? Nah, that would be too simple!

  9. Dear Ralph:

    You lost me.

    I don’t think that Welch and Albertsen have “solved” any problem. I think that have made an estimate of the order of magnitude of the number of men who are over-treated for prostate cancer. And I think that estimate (when evaluated on the basis of order of magnitude) is probably about right. In other words, it’s nearer to 50,000 men a year than 5,000 or 500,000.

    I entirely agree with most of your other statements (and pretty much did so in the original article). However, stage shifting very certainly started in 1990, when Prostate Cancer Awareness Week was initiated. How do I know? I helped to implement it.

    I also think that the number over-treated is not a factor in the reduction in mortality. Those that weren’t going to die from prostate cancer anyway didn’t (whether they were over-treated or not). By contrast, Welch and Albertsen agree that at least some men who were treated might have died if they hadn’t been — but they can’t tell how many, any more than we can.

  10. Mike,

    The point I am trying to get across is that the estimate has to be affected by the high grade cancer in the first 9 years.

    Almost no PSA use the first 3 or 4 years and then more use in the early ’90s.

    For an average of 50,000 over-treated men a year to be accurate at present, the last 11 years of the study range the over-treated men need to be close to 100,000 men a year (or about 50% of those diagnosed). This is the over-treatment figure these days. I believe that is an overstatement …

  11. So let’s be very clear. Welch and Albertsen constructed a mathematical model and used it to estimate a total number of men “over-treated” at approximately 1 million since 1986. Yours truly simply noted that “on average” that was 50,000 patients per annum over that time period in order to give people an “order of magnitude” for the number of men who might be considered to be “over-treated” annually on the basis of that estimate. Welch and Albertson made no specific representation about what the actual rate of overtreatment might have been at any specific point in time over the 20-year period, and nor am I. We just don’t know. But we do know there is over-treatment. Welch and Albertsen would probably argue that this has clinical and economic implications, and in that I would agree with them. I would add that it has emotional implications too.

    Whether the number of men being “over-treated” today is 25,000 a year or 100,000 a year, we don’t know. The point is only that it is likely to be somewhere between the two numbers.

    I don’t know if you have been able to read the actual paper, but I think Albertsen and Welch would argue that actually over-treatment could easily have hit 50,000 per annum or higher by 1992, based on their statement that “Overall prostate cancer incidence rose rapidly after 1986, at about 12% per year, until it peaked in 1992 (at 237.2 per 100 000 men).”

  12. It is difficult for me to believe that that level of over-treatment existed in 1992 based on my own experience. I had my first PSA at age 58 and it was 13.6 ng/ml. That meant nothing to me but I already had metastatic disease. If they were over-treating men in 1992 at the tune of 50,000 a year, then they are over-treating 100K to 150K these days …

    Over-treatment is being used to negate the value of PSA. Is over-treatment really caused by PSA testing? I feel that the problem is caused by the uncertainty that exists at diagnosis induced by the heterogeneity and multifocal nature of prostate cancer. Biopsy material is not always a match to pathology at surgery. There is a tendency to under-grade. I feel this is highly correlated to treatment failure.

    The very definition of over-treatment is disturbing. Worse, there are several and some difficult to interpret.

    How is that they are able to know (with a degree of certainty) that these patients would never progress to clinical stages of the disease? It seems to me that there is dependency on other comorbidities doing the patients in before their time.

    Do we need better and more specific markers of disease aggressivity. No question, but until something better comes along, telling men to avoid knowing is shameful. That is my very personal opinion.

  13. I am disturbed that the attention that this article has been given. In my opinion very biased about the overall prostate cancer experience. The emphasis is on over-treatment, but we have a problem still with under-treatment. We also had over 28,000 men die of prostate cancer, some very young and some very quickly. Maybe if we pull out the population that dies of prostate cancer we would see that the loss to society is greater than it needs to be? Also maybe if we pulled out the men who died under the age of 65, we would see that the negative impact on their families and on society if they die while they are in their working years is greater than this article makes it appear to be?

    One thing that we have to remember is that this article looks at the situation from an epidemiological basis and does not take into account the reality of the individual. If we stop at the screening stage and do not encourage men to get involved at an appropriate clinical level, we are saving some men from over-treatment but we are damning some men at the clinical level. There may be a significant number of men who are over-treated but we still have the problem of the significant number of men who die from prostate cancer. Personally I like the AUA recommendations because I see a major change by emphasizing patient education prior to biopsy rather than just prior to PAS/DRE or after diagnosis. It is a reasonable balance. Prior to testing it is all about epidemiology and after diagnosis it is difficult for men to hear or learn because they are faced with the reality of a cancer diagnosis. If we want more men to consider AS seriously then this might be the best time to educate them about what the treatments are and how each treatment fits based on the variables of the clinical data.

    I guess the problem I have is that this article emphasizes only the situation of the total population and ignores the issues that the individual faces. That is the bias that I see in this presentation. Unfortunately we are in a time where health care has become politicized, and if this is the only viewpoint that is in the news then decisions may be made from an unbalanced perspective.

  14. It is easy from a “statistically significant ” standpoint to talk about over-treatment and the need for more expectant management …. However, I am a 62-year-old with a new diagnosis, of a Gleason 6 lesion with no apparent spread. Everyone in my family, without exception, has lived into their mid-80s. I have no hypertension, no sugar issues, and only take Lipitor for hypercholesterolemia, which it controls. It is difficult for me to wait until this is outside the capsule and I’m “on the clock.” Yes , I may live 5-8 years after secondary treatment, but I don’t want to live that as a chemical or anatomic eunuch. I’m leaning toward surgery.


  15. Dear Jason:

    No one — repeat, no one — is saying these decisions are easy, or that one specific form of management is “right” or “wrong” for every individual.

    Like every other man with a diagnosis of low-risk prostate cancer, you will have to make a treatment decision that works for you, and given that you are saying you have a reasonable life expectancy of another 20+ years, it is reasonable for you to opt for active treatment. Just make sure that (assuming you do indeed decide on surgery) you find yourself a surgeon who is highly skilled and highly experienced, so that you minimize your risks for the adverse effects of radical prostatectomy.

  16. If statistics have no significance, what do you calculate your chances of having the following side effects from surgery — and how do you make the calculation:

    1. Erectile dysfunction
    2. Bladder incontinence
    3. Climacturia
    4. Peyronie’s disease
    5. Loss of size of penis

    Of course not every man has these side effects — and some have more of them for the 20+ years they live.

    And some of the men who live 20+ years with no treatment may develop similar problems as they age.

    But, as I say, if statistics have no significance in individual cases, how can anyone make any decisions?

    We each have to do what we believe is right for us.

  17. Hello Jason,

    When someone makes the choice of a treatment option with no coercion by a physician and when that person’s diagnostic data indicate he is also a candidate for active surveillance, that is not over-treatment. Over-treatment is when, despite data indicating very early prostate cancer development where active surveillance is certainly an option, a physician encourages the patient to early/immediate active treatment (whether it be surgery, radiation, or any other forms of active intervention) — particularly when that physician fails to explain active surveillance as being a reasonable option.

  18. Terry:

    If we could answer that question, we might not need this site (or Yananow either!).

    The core problem in all of this is our inability to answer the first fundamental issue that every man at real risk for prostate cancer (and his doctor) has to deal with: “Am I at risk for clinically significant prostate cancer that really needs treatment as compared to clinically indolent prostate cancer that doesn’t?”

    If we could answer that question successfully, 30-40 percent of the men getting treated today wouldn’t need to get treated, which would put a significant number of the less good prostate cancer treaters (surgeons, radiation oncologists, and others) out of business pretty fast! Then we could focus on improving the quality of care for those who really do need treatment!

  19. Mike,

    You say the question that cannot be answered is: “Am I at risk for clinically significant prostate cancer that really needs treatment as compared to clinically indolent prostate cancer that doesn’t?”

    When a leading surgeon like Dr Scardino, who seems to have supported surgery as the ‘gold standard’ for many years, is reported as saying that we can predict men at risk from prostate cancer with a 95% certainty, then there is no excuse for the current agenda of so many parties who focus on frightening rather than enlightening.

    But to get back to your question. Who can answer these two any more accurately than the 95% quoted by the good doctor:

    Am I at risk for heart failure or thrombosis in the next 20 years ?

    Am I at risk for an auto accident in the next 20 years that kill me or only for one that will injure me badly?

    Life is full of uncertainties — to demand certainty in conjunction with the vagaries of current diagnostic procedures, the principle one of which (the grading of material in a needle biopsy) is absolutely subjective, borders on the ridiculous.

  20. Terry:

    No one asks you after your heart attack and before they treat you whether you want treatment. No one asks you 30 seconds before you have a car crash how you would like to be treated afterwards!

    Every prostate cancer patient, by comparison, is asked to make decisions like this with very limited information. Dr. Scardino, who I have known for 20 years, is talking in general numbers, and yes, in general, men with certain prostate cancer characteristics have a 95% probability of certain types of clinical outcome over time if they receive treatment. However, the whole point of this discussion is not about the prostate cancer outcome. It is about whether you need the treatment at all.

    What neither Dr. Scardino nor anyone else I am aware of can tell you at present is, “I don’t need to treat you. I am 95% certain that you have indolent disease that will never have a significant clinical consequence” (or conversely, “I am absolutely sure I need to treat you because I am 95% sure you will have progressive disease in 5 years time if I don’t”).

    If a doctor could do that, with accuracy, do you not believe that it would make it a great deal easier for men to consider the opportunity to accept some form of expectant monitoring (when appropriate) than decide to risk active treatment?

  21. Maybe I am particularly stupid this morning, but I really can’t undertssnd the fine nuance that this discussion is not about prostate cancer outcome. Surely the probable outcome of a diagnosis is what drives the decision as to the appropriate level of treatment?

    When I had my heart failure episode (not a heart attack) the cardiologist discussed my condition fully and frankly, and set out the options for dealing with my condition. He could give me no guarantee whatsoever as to how my condition would progress, regress or fail in the next 20 years. Because he couldn’t.

    One of the options for treating my condition is a heart transplant, but guess what? We didn’t discuss that as an option because my condition didn’t warrant it.

    Show me the man with a prostate cancer diagnosis that is clearly very likely to be an indolent form of the disease who is not advised to have surgery, and the sooner the better.

    Enlighten, don’t frighten — that’s my new watchword.

  22. Ah but Terry … your perspective is not “average.” If it was, you would not have done what you did so many years ago!

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 )

Google photo

You are commenting using your Google 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: