More modeling of outcomes for men with low-risk prostate cancer on active surveillance


A paper just published on line in Clinical Cancer Research has attempted to project the prostate cancer-specific mortality rates of men managed by active surveillance (followed by radical prostatectomy if and when necessary) as compared to the disease-specific mortality rates of men treated with an immediate radical prostatectomy.

Xia et al. created a mathematical “simulation model” that combined (a) data on time from diagnosis to treatment under active surveillance and associated disease progression (from the Johns Hopkins active surveillance cohort of 769 patients); (b) data on time from radical prostatectomy to recurrence from the 3,470 cases in the CaPSURE database with clinical T stage ≤ T2a; and (c) data on time from recurrence to prostate cancer death from men with a T stage ≤ T2a among a Johns Hopkins cohort of 963 patients whose disease recurred after radical prostatectomy.

The authors then used these data to project outcomes for a hypothetical cohort of men aged between 40 and 90 years with low-risk prostate cancer (clinical stage ≤ T2a, Gleason score ≤ 6, and PSA level ≤ 10 ng/ml).

Here are their findings:

  • 2.8 percent of men on active surveillance in the hypothetical cohort would die of prostate cancer within 20 years after diagnosis.
  • 1.6 percent of men given an immediate radical prostatectomy would die of prostate cancer within 20 years after diagnosis.
  • The average projected increase in life expectancy associated with immediate radical prostatectomy (as opposed to initial active surveillance) was 1.8 months.
  • Men on active surveillance would — on average — remain free of treatment for an additional 6.4 years relative to men treated immediately.

Xia et al. conclude that,

Active surveillance is likely to produce a very modest decline in prostate cancer-specific survival among men diagnosed with low-risk prostate cancer but could lead to significant benefits in terms of quality of life.

In a media release issued by the American Association for Cancer Research, the publisher of Clinical Cancer Research, Dr. Ruth Etzioni (the senior author of the study) is quoted as follows:

We are now diagnosing many more men with low-risk prostate cancer, a large fraction of whom would never have known they had disease in the absence of screening. These men have cancers that may not have caused them harm if they had not been detected through screening, and we are faced with the dilemma that not all of these men will benefit from treatment.

She goes on to state that:

Although this is not a new result, it is confirmation of what we expected and it substantiates data from previous studies looking at watchful waiting. Very few men with low-risk disease die from prostate cancer regardless, and the difference between treatments appears to be very modest.

Dr. Etzioni also acknowledges that — to date — we know relatively little about the differences between the quality of life between these two groups of patients. We obviously know that immediate treatment is associated with short- and long-term side effects, including impotence and incontinence. However, active surveillance might also have an effect on patients’ quality of life.

Says Etzioni:

That 6-year treatment-free interval means that men who choose active surveillance will not have to endure treatment side effects during that time, but whether that is replaced by a negative impact on quality of life because of anxiety or repeat biopsies is not well known.

14 Responses

  1. Good news for those of us on active surveillance. Just hope to die with prostate cancer and not from it.

    Al H., Alberta, Canada

  2. I understand what Al means; however, I have been on AS for 4 years with urinary retention and impotence, and I have endured three biopsies along with the anxiety.

    Sometimes I think my quality of life would be better if all I had to deal with was incontinence. It is encouraging to know these issues are being considered.

  3. At age 70, my Gleason 6 was downgraded to PIN by John Hopkins with a second opinion. I have chosen active surveillance so I will not have to endure the treatment side effects for now, and I don’t think about the negative impact on quality of life because of anxiety or repeat biopsies because I am doing something about it. I have started a prostate friendly diet and exercise program that is also benefiting my general health and giving me hope that I now have control of my life and my future health.

  4. The mortality rate in this simulation seems to contradict actual results from two other studies I have found.

    The first of these studies is “Predicting 15-year prostate cancer specific mortality after radical prostatectomy” by Eggener et al. This study had follow-up of 11,521 patients and of them 9,557 had organ-confined Gleason 6. The mortality rate in this study was only 3 men after 15 years. That is a 99.97% survival rate.

    The next study I would like to refer to is “A pathological reassessment of organ-confined, Gleason score 6 prostatic adenocarcinomas that progress after radical prostatectomy“, which was discussed on this site in September 2009. It was from Johns Hopkins and was based on data from 2,551 patients. All were pathologically low-risk patients. In this study patient follow-up ranged from 1 to 22 years. No patient died of prostate cancer and only 11 had biochemical recurrence and did not seem to be progressing.

    In the study described above, the authors are saying that if you are classified as low risk via “clinical stage” and you choose AS, 2.8% of men would have died after 20 years. If you choose “immediate surgery” 1.6% of men would have died after 20 years. A 1.2% difference. First of all, for a cancer patient that is a large number no matter what the authors think.

    The projected mortality rate seems very high based on the above two studies I included. Both of my studies are based on “low risk” patients, although they are pathological low-risk, and are actual outcomes vs. projections. So how could the projection be so far off of real world data?

    Also, Group C included 936 patients that had “low risk” clinical stage, had surgery, and then went on to die of prostate cancer. I know that after surgery approximately 30 to 40% of men are upgraded from low risk to a higher risk stage. Let’s err on the high side and say 40% were upgraded. That leaves 578 pathological low risk patients that progressed to prostate cancer death. That number is way off of both of my cited studies. Not to mention the fact that the second study I quoted was from Johns Hopkins as well. Where did these 578 “low risk” deaths at Johns Hopkins come from in this study?

    I am missing something.

  5. Dear Chris:

    First … You cannot accurately compare the results of the two studies that were based on a retrospective analysis of data from selected patients at John Hopkins to the results developed by this model. You are comparing apples to oranges. The details of the assumptions being made in the model (based on clinical low risk) and the details of the patients at Hopkins based on pathological low risk are radically different.

    Second … We don’t know enough about the details in any of these cases to be able to make accurate comparisons anyway.

    The messages that the authors of the current study are sending are not, in any case, the ones you are reading into their paper. The messages that I believe they are sending are as follows:

    (a) The estimated difference in 20-year mortality between men with clinically low-risk prostate cancer who start out on active surveillance and those who have immediate surgery is small (1.2%).

    (b) We really do not know much about the long-term quality of life of men on long-term active surveillance (as compared to the long-term quality of life on men who have immediate surgery), and without that information there is a lot of speculation.

    (c) What happens in academic centers like Johns Hopkins and what happens in the “real world” (which is where data from the CaPSURE initiative are collected) is very different.

    At the end of the day, we do not have the right data to make these types of comparisons accurately yet — which is exactly why I emphasize that this is all a mathematical model. A single inaccurate aspect of that model might be throwing the data way off. However, it is also fair to point out that the men being treated at Johns Hopkins over the past 20 years have not exactly been representative of the real world.

    You are basically trying to read far too much into all these data.

    I would also point out that the 1.2% difference may not seem like enough to justify active surveillance to you (or to some other men). I can most certainly assure you, however, that that 1.2% difference would be trivial to others, who would consider it exactly what they needed to hear to justify active surveillance as opposed to surgery — particularly if they were older patients whose likelihood of 20-year all-cause survival was, in any case, limited.

  6. I was a candidate for AS, but opted for immediate RP. From the post-op pathology, my clinical stage was upgraded to one that would not have made me a candidate for AS. I wonder if the study deals with this situation.

    Although I am stuggling to cope with the impotence, I am sure that I would dread the biopsies, having to endure the continuing urinary flow problems, and living with the general worry of having cancer inside my body.

    So, for me, no regrets about RP rather than AS.

  7. It is very important to appreciate that the choice between active surveillance and immediate treatment for men with low-risk prostate cacner is not a “right/wrong” issue. It is about newly diagnosed men with low-risk disease knowing that they have choices … and making sure that they understand the full range and the implications of those choices before they decide what they want to do.

  8. Here is the part I don’t get. This projection was designed to compare mortality rates of “clinically low risk” patients on AS vs “clinical low risk” patients that had “immediate treatment.” I am assuming that everyone in this study who stayed on AS over 20 years remained clinically low risk or they would have been moved off. Also, we all know that some of the “immediate treatment” group would have been reclassified after surgery to a higher risk. So, did they throw these people out because they technically were never low risk to begin with? I know we don’t know. Assuming that they did throw them out that still leaves the “projected mortality” from low-risk immediate treatment people dying of prostate cancer much higher than any of the studies I have read that followed real world patients.

    I guess the apples to apples comparison to me would be comparing men on AS who remain “clinically low risk” to men who had “immediate treatment” and were still deemed to have “low risk” PCa. It appears to me that this is the projection the authors are claiming their study is making. If that is true the projected mortality rate for the immediate surgery group seems way higher than the studies I have seen that actually followed a low-risk surgery group.

    Sometimes I feel like I have a metal plate in my head …

  9. “a) The estimated difference in 20-year mortality between men with clinically low-risk prostate cancer who start out on active surveillance and those who have immediate surgery is small (1.2%).”

    Please ignore my previous post. I just re-read your response and the above statement from you makes sense. After reading the actual study I feel the title is kind of misleading. The devil is in the details …

  10. Chris:

    I think you are making the wrong assumption. I think that the model is based on the outcomes of all the low-risk men from day 1, regardless of wehat happened to then over time (otherwise it would be a pointless model). No one gets “thrown out” for any reason. This is the “real” clinical world comparison. In other words, what the authors are projecting is that if you take all men initially diagnosed with low-risk disease, then according to their model, whether they are (a) initially assigned to active surveillance and stay on it for 20 years; (b) initially assigned to active surveillance and subsequently have treatment (after any period upo to 20 years); (c) have immediate surgery; or (d) need other forms of treatment over time because their disease progresses, the difference between the rates of prostate cancer-specific mortality is small.

    You certainly could also do the comparison you suggest, but that is just a subset of the “real” clinical world, but it is not the one that I believe that authors model is addressing at all. If it was, I think that they would say that … and they clearly don’t.

  11. And re your third post … Oh well … I had already answered the second post before I saw the third one!

  12. I wonder how many could go through life and just die naturally without ever knowing that the prostate gland had cancer — if a PSA test had never been done. … Boy, it sure creates a lot of stress. …

  13. We know the answer to than one too, Nancy — but it’s not a very good answer!

    The best current estimate is that somewhere between about 20 and 35% of all the men who have a positive biopsy after a PSA test indicating risk for prostate cancer will, in fact, never have serious clinical consequences from their diagnosis and would probably never benefit from treatment. However, …

    Unfortunately we are not good enough yet at being able to accurately differentiate the “apparently at risk but really quite safe” from the “definitely at risk who need treatment a.s.a.p.” … But we are awful lot better at it than we were 20 or so years ago, of that I can assure you. That’s why we have started to learn how to defer treatment by active monitoring (not “watchful waiting”) and can still implement treatment in a timely manner if risk levels change.

  14. A problem with these studies is that they don’t mention the drop out rates of patients followed after treatment or AS. These cases could have died from cancer.

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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: