Active surveillance today: a summary of the Sunnybrook experience and related factors

In an important new paper in Current Opinion in Urology (available free as a full-text article), Dr. Laurie Klotz has provided an excellent summary of his perspective on the current application of active surveillance in the management of low-risk prostate cancer, based on the 20-year experience of the Sunnybrook group and other data.

This is a paper that most support group leaders and other advocates will want to print out for their files. Here is a summary of some of Dr. Klotz’s key statements:

  • There is now extensive clinical evidence supporting the near to complete absence of metastatic potential in the vast majority of pure Gleason pattern 3 cancers.
  • There is a truly tiny number of rare case reports of  patients with apparently pure Gleason 3 + 3 disease in which metastasis was observed. (Klotz states that this is just two known cases among tens of thousands now evaluated.)
  • Cardiovascular disease is the commonest cause of death in men with favorable risk prostate cancer and such patients are 10 times more likely to die of causes other than prostate cancer than they are to die of their prostate cancer.
  • Among men initially diagnosed with Gleason 6 prostate cancer and who meet other appropriate criteria for active surveillance
    • PSA testing should be performed every 3 months for 2 years (to establish a baseline of PSA kinetics) and every 6 months indefinitely after that.
    • A confirmatory biopsy is mandatory within 6 to 12 months of the initial diagnostic biopsy.
    • The confirmatory biopsy should target the areas that are typically under-sampled on the initial diagnostic biopsy (the anterior prostate, the prostatic apex, and the prostatic base).
    • If the confirmatory biopsy is negative or confirms microfocal Gleason 3  +  3 disease, subsequent biopsies should be performed every 3 to 5 years until the patient reaches age 80, or has a life expectancy of less than 5 years because of comorbidity.
    • The increasing use of MRI and molecular markers will further reduce biopsy requirements in men with favorable results (i.e., negative MRIs).
  • PSA kinetics are now used by the Sunnybrook group only as a guide to identify patients at a higher risk, but not to drive the decision to recommend treatment.
  • About 25 percent of men on active surveillance will be upgraded on confirmatory or repeat biopsies.
    • The vast majority of such upgrades are a consequence re-sampling (i.e., finding areas of cancer that were not sampled at the time of the original diagnostic biopsy).
    • 98 percent of these patients will be upgraded to Gleason 3  +  4 = 7.
    • Upgrading to Gleason 8 or higher is uncommon.
    • Biological grade progression (i.e., Gleason 3 cells giving rise to Gleason 4 or 5 progeny) does occur over time, but is uncommon.
    • In the Sunnybrook cohort cohort,  biological grade progression occurred at about 1 percent per year from the time of the original biopsy.
  • Low prostate volume (i.e., a high PSA density) is a predictor for risk progression.
  • African Americans on active surveillance have a higher rate of risk reclassification, and PSA failure when treated, than Caucasians — but active surveillance is still an appealing option for African Americans who have been appropriately risk-stratified.
  • Tests such as the Prolaris test, the Oncotype DX test, and others have the potential to interrogating the microfocus of Gleason 6 found on biopsy for accurate information about the presence of higher grade cancer elsewhere in the prostate, or the future likelihood of progression to aggressive disease.
  • Multi-parametric MRI has an important emerging role in the management of men on active surveillance (and Dr. Klotz discusses this at some length).
  • Patients with favorable-risk Gleason 3  +  4  =  7, in which the component of pattern 4 is < 10 percent, have a very similar natural history to those with Gleason 3  +  3 = 6, and Klotz states that such patients “are surveillance candidates”.
  • There is now general consensus on the principles of surveillance but the optimal surveillance strategy and triggers for intervention are still not completely agreed upon.

Klotz also notes that

The results of active surveillance, embodying conservative management with selective delayed intervention for the subset who are reclassified as higher risk over time based on repeat biopsy, imaging or biomarker results, are associated with a 5 percent cancer-specific mortality at 15 years.

What this mean is that out of 1,000 men who are appropriate candidates for active surveillance based on the original Sunnybrook criteria — either true low risk disease (with Gleason 3 + 3 = 6, PSA < 15 ng/ml; and clinical stage T1c to T2a) or favorable intermediate-risk disease (with Gleason 3 + 4 = 7, generally small volume in older patients and the same PSA and clinical staging criteria) — about 25 percent or 250 men will actually progress over time and meet criteria for localized treatment. Of those 250 men whose disease does progress, about 5 percent (i.e., 12 or 13 men) will actually die from prostate cancer within 15 years.

More recent Sunnybrook criteria have placed tighter limits on eligibility for active surveillance among men with Gleason 3 + 4 = 7 to try to reduce the number of men who would actually progress to have metastatic disease. However, even based on the currently available data, what this means is that, at 15 years of follow-up, about 95 percent of men placed on active surveillance and managed according to the Sunnybrook criteria are not going to die of prostate cancer, and about 75 percent will never require treatment at all.

This level of ability to avoid unnecessary treatment and the associated adverse effects of such treatment over time in appropriately identified and well-monitored patients seems to The “New” Prostate Cancer InfoLink to represent extraordinary progress from where we were in the mid-1990s, when immediate treatment was almost de riguer for every man diagnosed with prostate cancer of any Gleason grade. It is also interesting that these data are so very closely correlated with the data first published by Albertsen and his colleagues in the late 1990s; those data had shown that risk of prostate cancer-specific mortality in men with a Gleason score of 6 or less was significantly less than 30 percent if they had no treatment at all until some form of androgen deprivation therapy became necessary.

17 Responses

  1. Wonderful! Made me smile this morning. :-)

  2. What is microfocal Gleason 3 + 3? What numbers are favorable for risk volume?

  3. Dear Paul:

    “Microfocal Gleason 3 + 3” simply means a very small focus of Gleason 3 + 3 = 6 disease. It’s the sort of prostate cancer found very commonly in autopsy specimens of men who died of other causes.

    The volume of low-risk cancer that appears to be “riskier” is a matter of some debate, and I am not aware of any very specific guidelines on this today. The group at Johns Hopkins, which has always used the most stringent criteria for active surveillance in their research protocol, used to state that patients should have no more than three positive biopsy cores with less than 50% of all positive cores exhibiting cancer. They have now removed that restriction, which you can see if you look at their criteria for eligibility. The need for treatment based on volume of cancer in the prostate (if it is all Gleason 3 + 3 = 6) is probably going to have a lot more to do with any effects of the growth of the cancer on the ability to urinate freely (an effect similar to the effect of benign prostatic hyperplasia) than it does with the actual cancer.

  4. An excellent paper! I have a reservation regarding Active Surveillance for African American men in view of the following two papers:

    A December 2014 paper concludes that ‘AA [African American] men with “low risk” prostate cancer, especially those considering active surveillance, should be counseled that their recurrence risks can resemble those of whites in higher risk categories.’ Thus, active surveillance should likely not be considered for these men.

    This is further confirmed by this paper reporting that African American men diagnosed with prostate cancer, no matter at what level, are at “high risk” for more aggressive disease; they should likely opt for a treatment option to address their diagnosis rather than active surveillance.

    The information from this paper by Dr. Klotz combined with that in a recent report posted in the Journal of Urology would provide additional considerations when making a determination as to being a candidate for active surveillance. The report in the Journal of Urology found that, despite a Gleason score of 3 + 3 = 6 from biopsy, pathology of the surgically removed prostate gland was found to be of higher score. Their conclusions were that older men, men with higher PSA levels at diagnosis, a PSA density over 0.15 ng/ml/cm3 (divide PSA level by gland volume), palpable disease on digital rectal examination (DRE), and extent of cancer over 4 mm on biopsy were not appropriate candidates for active surveillance.

  5. Dear Chuck:

    I quote from Dr. Klotz’s review:

    “African-Americans on AS have a higher rate of risk reclassification, and PSA failure when treated, than Caucasian men …. Black men who are surveillance candidates also have a higher rate of large anterior cancers than Caucasians …. However, black patients diagnosed with low-grade prostate cancer include many men who have little or no probability of a prostate cancer related death during their remaining lives, and active surveillance is still an appealing option for those who have been appropriately risk-stratified.”

    There is no doubt that African American patients and their physicians need to exercise greater caution about the decision to consider active surveillance than most white patients. However, to me (and I believe to Dr. Klotz) what this clearly means is that greater care should be exercised early, through the use of (ideally) mpMRI/TRUS-guided re-biopsy and related examination as part of the confirmatory process to ensure that the individual patient is truly an appropriate low-risk patient. And any African American patient on active surveillance needs to be followed with great care until we have more information about the real risk for progression in such patients.

    Also, the Swedish report you refer to in the Journal of Urology is yet another paper based on a retrospective analysis of data from patients who were not given any confirmatory biopsy or any other careful assessment of their suitability for active surveillance prior to their surgery, so the patients included in this study were in no way comparable to the patients that Klotz describes as good candidates for active surveillance. There have been multiple studies like this.

  6. Were I to design a PowerPoint risk stratification presentation from a patient’s perspective (and actually had the data to do so) the format would be: for various begin points, be it PSA level, grade, percentage, doubling time, etc., how do the graph lines of metastases diverge over time between treated and not.

    While retroactive studies tend to group diagnosis into large populations, that is not the critical view from the bottom. How do I fit most closely into a grouping that fits my diagnosis in the terms that are used in my specific case? That is what interests a patient.

    Sorting cases by that format would likely have many individual cases fit into multiple starting points. Scientifically wrong, but from a patient viewpoint the most useful.

    Physicians think in terms of treatment outcomes due to delay. I believe patients, at the time of surveillance or treatment decisions, think in terms of disease progression. Hard to describe the difference, one is mechanical, the other an emotional view.

    I know radical treatment does not reduce risk to zero. What a patient wants to know is: within the terms given to me by my diagnosis, what has happened to those who have journeyed before me, after they made their choice? How does time effect that outcome?

    In terms of any monitoring system, how does delay in treatment effect outcome from any given diagnostic starting point?

    For each starting diagnosis there would be a large group with an overall outcome, then smaller subgroups graphed by treatment delay time periods. Immediate, 6 months, a year, and multiple years of delay. An individual would likely travel through risk groups over time as their diagnosis changes.

  7. Dear Michael:

    Actually, to a large extent, this is exactly how the nomograms constructed by Michael Kattan and colleagues (initially at Baylor, then at MSKCC, and now at the Cleveland Clinic) work. For example, they can tell an individual patient his risk for disease progression after radical prostatectomy at 2, 5, 7, and 10 years after surgery based on initial data that includes the patient’s age, clinical stage, Gleason score, PSA level, etc.

    There is no reason why Dr. Kattan and his colleagues (or others) wouldn’t be able to develop a similar set of nomograms for risk of differing types of progression on active surveillance — given a large enough database. It is my suspicion that, as yet, a sufficiently large, consistent database may not exist. The Sunnybrook data set is about 1,000+ patients, but a large number of those patients opted out because they couldn’t deal with the stress of “living with” a very slowly evolving cancer — even though there was no sign that they needed any treatment. As a consequence, there are probably only about 650 or so patients in the Sunnybrook database who have actually been followed for long enough to provide the data that Dr. Kattan would need to develop such a nomogram. I am planning to be at the AUA meeting in NOLA this year. If I run into some of the right people, I will see whether anyone is planning on the development of such a nomogram.

  8. In agreement with the Sitemaster, researchers at UCSF examined a number of different nomograms and found them inadequate for active surveillance decision making. Sunnybrook has its own risk calculator, but it isn’t based on their active surveillance cohort. I hope the Sitemaster is able to get them interested in constructing one — it is so much more convincing for patients to see it customized to their particular situation than for someone like me to just throw population-based statistics around.

  9. I wonder how many of the 350 regret what they did.

  10. BRAVO (again) to Dr. Klotz! Thanks (again) to Sitemaster!

    Once again we can be grateful to Dr. Klotz and his team for their long-term dedication in pioneering active surveillance. What encouraging long-term results, especially that further solidly based confirmation of the extremely, almost incredibly low rate of Gleason 6 cancer that proves dangerous!

    Here’s a further encouraging perspective on the data for the few patients who are projected to die of the disease in the Klotz paper. The paper said, as stated in Sitemaster’s review above: “… Of those 250 men whose disease does progress, about 5 percent (i.e., 12 or 13 men) will actually die from prostate cancer within 15 years. …” The encouraging perspective is that that projection is based on technology that has been available during the period of the Klotz series; as we know, very significant newer technology has and is emerging that, within 10 to 15 years — the real event horizon for death for the few active surveillance patients who would have died in the past — will no doubt be able to save or extend life beyond 15 years for many of these men.

  11. Sitemaster,

    Thank you for continuing to keep us up to date with the most recent developments in prostate cancer diagnosis and treatment. Your site is indispensible for anyone having to deal with prostate cancer.

    This latest active surveillance recommendation , i.e. mp-MRI and Polaris testing, from Dr. Klotz’s group, is very reasonable and welcomed.

    For those following your site for the past 2 years, this conclusion was quite predictable and would have been what many of us would have insisted upon even before this announcement based on our reading at this site and elsewhere had we ever needed active surveillance.

    The sad thing is that it is taking time, too much in my opinion, for advances in prostate cancer care to be recognized, recommended, and reach the people in need. The reasons are not just scientific, as many of these advanced techniques are available at only a few places and certain procedures/treatments if they reached the “recommended ” status would draw patients away from community centers, so there is politics involved.

    With all the nuances to prostate cancer care brought to light by this site and issues getting recent advances out to people in need, in my opinion, it is really important that prostate cancer patients and their advocates really do their homework and come up with what they would consider to be state of the art care, and compare that with what the doctor they are seeing is offering.

    When one looks at how certain places have mistakenly set up their active surveillance programs, doing repeat biopsies when the PSA doubles or, if not doubling, repeating the biopsy then every 3 years, rather than playing it real safe and biopsing every year or following with mp-MRI then re-biopsy, etc., one realizes just how much this prostate cancer care is on a steep learning curve.

    In this day and age, a given prostate cancer patient may need to do a la carte care, having a local urologist as their “family doctor” and get referrals to other specialty centers for the best mp-MRI, best robot-assisted radical prostatectomy, best active surveillance program, etc.

    Thank you again for helping people navigate this prostate cancer maze. Please keep up this good work.

  12. The article references 13 prospective studies on active surveillance outcome with a table which only includes 12. Wondering why the PIVOT study is not included.

  13. Dear Cliff:

    The PIVOT trial was absolutely not a study of active surveillance (which is a proactive management strategy designed to ensure the option for curative therapy should a low-risk cancer show signs of progressing)). Rather, the PIVOT trial was a study of something much closer to watchful waiting (which is a strategy of deferring all treatment until palliative care becomes necessary, and was an entirely appropriate form of care for the men enrolled in this trial, who were largely older men, often with additional comorbid conditions).

    Active surveillance and watchful waiting are two very different subtypes of the general category known as “expectant management”.

  14. Mike, thanks. Some of this nomenclature is hard to distinguish: active monitoring versus active surveillance.


  15. Cliff:

    The only way one can really distinguish between the different types of expectant management and the uses of the term “monitoring” is to look at the detailed descriptions of the patients being so managed and the precise way they were being monitored and managed. At some point we will get to a more standard language for the different processes — but that’s going to take a while because we don’t even have complete agreement yet on some of the distinctions.

  16. To those of us with Gleason 3+3’s (me included):

    Please take 15 minutes to watch and listen to Dr. Laurence Klotz, the aforementioned pioneer in AS, from a presentation he gave this summer (2016). Yes, he has lots of data from 17-20 years of working with men on AS. And he has closely followed Johns Hopkins AS studies among others. BTW: Dr Peter Pinto (NIH) had almost identical slides showing the important differences between Gleason pattern 3 and Gleason pattern 4 or higher at this conference.

    Know this, pre-2004, if you walked into a urologist’s office with a Gleason score of 5 (2 + 3 or 3 + 2), you were told that you had cancer and needed treatment ASAP. Then in 2005-ish, a worldwide urological vote was taken and poof — no more looking for Gleason 5s; no more treating G5s. Why? Dunno, but a wild guess is that no one was dying from them … then it was on to treating Gleason 3 + 3 = 6 and higher.

    Today, in 2016/17, the reality for those of us with Gleason score of 3 + 3 = 6 is that essentially no one dies from them (< 0.0001%). Moreover, Gleason 3 + 3 = 6s do not progress/metastasize into the deadly Gleason 4 + 3& = 7s (< 0.0001%). And despite popular perception, as the "volume" or size of a Gleason 6 tumor grows, that does not translate into problematic Gleason 4 + 3 = 7. The Klotz video will confirm all this. But do not take my word for it, do your own homework. This site is very good. As is this one at

    Additionally, here is a link to Dr. Bert Vorstman’s paper on the bogus Gleason 6 "cancer". He is a semi-retired urologist, practicing for 35+ years in Coral Springs, Florida. He (and many others) are bemoaning the fear-mongering, cronyism, and likely greed that surrounds the Gleason 3 + 3 = 6 "cancer" diagnosis today — much like there was only 11-12 years ago with "lethal" Gleason 5s. Moreover, he is sick and tired of counseling men that had the da Vinci robot remove their prostate and were left with physical and mental (depression) ailments. Especially for any Gleason 3 + 3 = 6s.

    Bottom line:

    (1) IMHO, if we have Gleason 3 + 3 = 6 (and I do), and any doctor says he will "cure" your "cancer" be wary. How can they possibly "cure" something that will not kill you??

    (2) And more problematic is the understanding that most mortality rates/studies are based on bad input. Curability rates for treating Gleason 5s and Gleason 6s are flawed because they are "treating" something that cannot kill you … hmmm. They could have cut off your right arm, leaving your prostrate intact, and you also would not have died of prostate cancer. …

    (3) Checking the information-gathering box for "family history" is dubious unless your dad, brother, etc., died from prostate cancer. Or, you know the actual Gleason scores of the relative.

    (4) In other words, we do not have a ticking time bomb in our bodies/prostates.

    That said, with Gleason 3 + 3 = 6 and on active surveillance, we do need regular check-ups. From all that I have learned, a targeted, in bore, multiparametric MRI (that is to biopsy only the areas of concern in real time, as opposed to random 12+ samples) is currently the best detection method for finding the truly aggressive, lethal prostate cancers of Gleason 4 + 3 = 7 and higher. PSA tests are a tool, but not fool-proof. Genomic tests are still "to early to tell" as a definitive test/answer, but are another tool. There are, in America alone, over 26,000 men dying from prostate cancer. Just not from the Gleason 3 + 3 = 6 variety.

    Enjoy Dr Klotz’s video … and Dr. Vorstman’s article.

    Best, ScottD

  17. Well said Sir!!!!

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: