How active should “active surveillance” really be? Are we being overly cautious?


There are a couple of brief but interesting new articles about active surveillance by Dr. Oliver Sartor and Laurence Klotz on the CancerNetwork.com web site this week. They deal with differing perceptions about the application of active surveillance.

In his article (“Active surveillance not only reduces morbidity, it saves lives“), Dr. Klotz — a highly respected urologic oncologist who specializes in active surveillance — argues the “conservative” position that men being monitored on active surveillance need close monitoring with regular appropriate biopsies because

  • Some 25 percent of men initially diagnosed with low-risk, Gleason 6 disease actually have higher-rsk disease (although in 90 percent of these cases they are upgraded only to lower intermediate risk disease with a Gleason score of 3+ 4 = 7)
  • About 1 percent of patients who really do have low-risk disease will de-differentiate per year to higher-grade disease. In other words, if we started with 1,000 men with truly low risk prostate cancer, 10 of these men would have higher-risk disease by the end of the first year; another 9 or 10 in the second year; etc.

For those patients (and their doctors) of a cautious disposition, this is a perfectly reasonable argument … and it comes with implications related to regular PSA tests, MRI scans, biopsies, and (arguably) treatment at the first sign of almost any amount of Gleason 3 + 4 = 7 disease.

Dr. Sartor — a respected medical oncologist who specializes in the treatment of men with advanced prostate cancer, in his article (“Surveillance for prostate cancer: are the proceduralists running amok?“), makes a rather more provocative suggestion. He argues that the available clinical data suggest no benefit compared to treatment for 12+ years in the only two trials that have ever compared to observation to treatment for men with clinically diagnosed low-risk disease (PSA < 10 ng/mL and Gleason score of 6, and clinical stage T1c or T2a), and that what benefit there may have been for men with lower intermediate-risk disease (i.e., PSA < 10 ng/mL and Gleason score of 3 + 4 = 7, and clinical stage T1c or T2a) was at best small and not really well-characterized in these two studies. He therefore proposes that actual “observation” as opposed to excessive “”poking, probing, and testing” is entirely appropriate as a means of monitoring men initially diagnosed with low-risk disease, and goes on to say that

… we currently make surveillance too complicated. We do too many biopsies and have unnecessarily complicated algorithms. Too many people are excluded from surveillance in our protocols. Should the vast majority of patients with low-risk prostate cancer simply be followed for symptoms and our procedures held at bay?

For those patients (and their doctors) who tend toward the view that we shouldn’t obsess over every possible chance that we will be dead in 10 years and just want to “get on with their lives”, this is also a perfectly reasonable argument … albeit coming with the implication that a few of those patients may indeed suddenly find that they have incurable progressive prostate cancer. (They could also have died from being hit by a bus.)

The two articles almost perfectly demonstrate two very different mindsets about health care today: Dr. Klotz’s position represents the position that we should make every reasonable attempt to extend life, even if it comes with significant procedural implications and does not exclude the continuing risk for over-diagnosis and over-treatment (albeit less over-treatment than if one just gets one’s prostate removed the minute anyone says the word “cancer”). Dr. Sartor’s position is the one in which we stop using the word “cancer” at all for men with what is currently defined as low-risk, Gleason 6 disease. He would argue that almost none of these men is actually going to die from prostate cancer and if they are carefully observed we will “catch” most of the men that have significant clinical progression and can still give them curative treatment if it is needed, so let’s just “keep a close eye” on them and treat them if they really need it.

Your correspondent should be clear that he tends to fall into the category of patient that is more sympathetic (at 65 years of age and on an entirely personal level) to Dr. Sartor’s perspective, but for patients who are more cautious, Dr. Klotz’s approach is most certainly very reasonable. Either way around, not immediately treating low-risk prostate cancer would seem to be a good idea since, as Dr. Klotz points out:

In fact, patients with surgically confirmed Gleason 6 disease do not die of prostate cancer. This has been confirmed in several large series involving more than 10,000 patients.

It is certainly the case that some active surveillance protocols are so cautious that they pretty much include only men who are almost guaranteed not to have any risk at all for clinically significant prostate cancer. To The “New” Prostate Cancer InfoLink, that does seem to be overly confining and we have difficulty seeing why most such patients would need to be closely monitored at all.

15 Responses

  1. Excellent perspectives!

    It confirms my position of avoiding “going under the knife” as long as possible.

  2. But Dr. Sartor does encourage close “observations” which I see as equal to “active surveillance.” And any such “observations” or “active surveillance” mean much more than just sitting back and waiting until 6 months to a year from diagnosis to check again. In my research and study of what I perceive as more appropriate “active surveillance” I provide patients seeking information regarding active surveillance this guidance, which in my opinion is being reasonably careful.

  3. Dear Chuck:

    With respect, I think that the whole point that Dr. Sartor is making is that for many men — and especially older men with a reasonable life expectancy of (say) 15 years — there is a major distinction between procedure-driven monitoring (the type of “active surveillance” which you appear to be recommending in your article) and pure clinical observation on a regular basis. The latter may well not be appropriate for men like you, but that doesn’t mean it isn’t highly appropriate for other men (such as me).

  4. The problem as I see it is that there is no set protocol. Urologists for the most part are not riding this wagon yet. That hurts men with little prostate cancer education. Biopsy requirement is a hard sale. They need better imaging to monitor occult progression.

    There is no study maturity for younger men. Older men with verified low-risk disease will benefit from avoiding treatment. Age at diagnosis continues to be a significant parameter for disease progression.

  5. Ralph:

    That is completely true … although Klotz’s series does, in fact, include a reasonable number of younger patients.

  6. Mike,

    The latest information I have is that the follow-up on Klotz’s cohort is median of 80 months with a 79% of OS and 97% of CSS. The cohort is 450 men. Probably good results and guidance for older men with low-risk probably. The question remains for younger men and the perspective of periodical biopsies. The need for a reliable follow-up protocol is high: one that has reliable imaging instead of biopsies. At this point in time the patient is left on his own with little knowledge of the basics.

  7. Ralph:

    The last update I am aware of from the Sunnybrook database was based on 862 patients followed for an average of 6.4 years. Click here for the details.

  8. DR. SARTOR’S THOUGHT-PROVOKING QUESTION, BUT INADEQUATE DATA

    Dr. Sartor, who has contributed so much to prostate cancer knowledge, is asking a thought-provoking question here. I hope the issue of the best active surveillance management protocol will receive continued attention, hopefully with the result of less burden on the patient, which is the goal envisioned by Dr. Sartor. Indeed, Dr. Klotz, the example for Dr. Sartor’s thoughts, is himself a strong proponent of fewer biopsies during surveillance coupled with more intense mining of information from the patient’s previous biopsies, imaging, testing, and consideration of the patient’s overall circumstances.

    However, having read Dr. Sartor’s piece, I cannot get around the appearance of his reliance on premature data. He supports his line of argument with two studies, SPCG-4, a European study, and PIVOT, a US, chiefly Veterans Administration study, with his key observation being that there was no significant difference in either overall or prostate cancer-specific survival between the two arms in either study. The critical problem for Dr. Sartor, to cut to the chase, is that median follow-up in SPCG-4 was just 12.8 years, and median follow-up in PIVOT was 10 years, in a modern survival context where true differences in survival would be most unlikely to reveal themselves until substantially more time had passed.

    After all, even as of half a decade ago 5-year survival in the US was within a fraction of a percent of 100%, with survival at 10 years — all comers and all approaches to the disease — at 98%, and 15-year survival still over 90% at 93%! This suggests that at 10 years, for the US trial, we should not have expected much of a difference in survival between the arms of the study. While it is plausible that less aggressive screening or some other factors in Europe may be causing shorter survival than in the US, it is arguable that the difference in survival would not be large compared to the US, and therefore that 12.8 years of follow-up in SPCG-4 would also be quite premature. Sitemaster has observed that far, far longer follow-up is probably needed to fully illuminate at least the screening issue, and similar thinking probably is applicable here.

    If Dr. Sartor were to limit his argument to observation in lieu of surveillance for AS eligible men with a life expectancy with a high probability of NOT exceeding about 10 years, the argument would seem to rest on firmer ground.

    The consensus issue. Dr. Sartor also states, regarding active surveillance criteria for patient eligibility, that “… No consensus is apparent; furthermore, there is little consensus on who has ‘progression’ and who does not ….” Dr. Sartor is making the important point that there is no standard of care yet. However, we need to bear in mind that there is substantial emerging convergence toward a standard among leaders in active surveillance. For instance, the 2007 Active Surveillance Conference led by Dr. Peter Carroll of UCSF voted to accept a consensus eligibility standard that featured six criteria for safe active surveillance at any age (Gleason score < 7; % positive cores < 34%; PSA < 10; PSAV < 2; PSAD < 0.15; and DRE = no nodule. The point here is that the absence of a “standard of care” should not discourage us from using technology that already works quite well.

  9. Dear Jim:

    If you make the assumption that the whole point of active surveillance is almost never to have to treat anyone who is initially diagnosed with (tightly defined forms of) low-risk disease, then you are correct … but that is not the base premise of active surveillance at all.

    The base premise of active surveillance is that we should seek to defer or avoid treatment for as many men with low-risk disease as possible until it is demonstrably necessary in order to optimize patients’ quality of life with minimum risk to their quantity of life. Seen through that lens, I think Dr. Sartor’s proposition is indeed thought-provoking and is actually based on very good data indeed — although better data, which will presumably come from the ProtecT study, will very certainly be welcome.

    For a 60-year-old man, diagnosed tomorrow morning with classic, low-risk prostate cancer, who may well be able to have a good doctor clinically “observe” his cancer for (say) 12 years (without all the prodding and biopsies) and then see some progression (to Gleason 3 + 4 at 72) and then (maybe) have some sophisticated form of focal HIFU (that does not exist as yet) that puts him into remission for another 12 years (until he is 84), such that at the end of his life he needs (say, worst case scenario) a year or two on ADT, this looks like a way better option (to me) than trying to decide now what form of invasive treatment he needs in the next 3 months (complete with the potential side effects for the ensuing 25+ years).

    And with respect to the consensus issue, I believe Dr. Sartor is pointing out that consensus among the “proceduralists” is not necessarily in the best interests of those patients who put a slightly higher value on the quality of their lives compared to the qnatity thereof.

  10. Sitemaster — Thinking hard about your comment of 6:05 PM (and, by the way, thanks very much for this topic)

    I think I’m seeing most of this as you are: same base premise for active surveillance (in fact, I’m encouraged by the success rate of about 60% for long-term surveillance); same appreciation that Dr. Sartor is spot-lighting an important issue that I expect will pull active surveillance to a higher level of impact; and appreciation for your scenario as a highly desirable goal that will likely become achievable for many men.

    But I’m still cannot get past what appears to me to be inadequate follow-up in the two studies, at this point. Maybe a night’s sleep will increase insight.

    Along the same line as the thrust of Dr. Sartor’s thought, I just read a piece by Dr. Charles Myers in which he describes his work toward expanding active surveillance for Gleason 7 patients. He is enthusiastic about the prognostic power that advanced imaging is now providing, in conjunction with other technology.

  11. Jim:

    Please appreciate that monitoring with “observation” (as opposed to procedure-based active surveillance) doesn’t exclude appropriate testing, but is based on the idea that you test when you really need to test (using whatever tools are available, including the newer types of scan) as opposed to testing based on the idea of rigid protocols. What Dr. Sartor is saying is that doctors should treat patients as people first and not be just “servicing them” (like your heating and air-conditioning system, that the servicing company does things to by rote twice a year whether all those things really need to be done or not).

    Also, would I like to have 20-year data to support his proposition? Of course I would … but we won’t have that data for at least another decade, if ever. The PIVOT and SPCG-4 data are certainly not sufficient to “prove” his hypothesis, but arguably they are good enough to justify it because they certainly don’t contradict it.

  12. Hi again Sitemaster,

    I think we are seeing things the same way but using words that have created semantic problems. The kind of case management you are describing as “observation” strikes me as very close to what the Klotz group is actually doing in Toronto. I’m convinced that group is continuously thinking about its protocol and adjusting it to make it easier on the patient where warranted as more cohort data becomes available. I’ve heard Dr. Klotz state that they want that follow-up biopsy in a year, as that, in conjunction with the initial biopsy, gives them a lot of prognostic power, but that subsequent biopsies are timed based on the patient’s circumstances, including those two biopsy results, and tests/scans. They use an actual algorithm for that, cranking in the various biopsy, test, scan, etc. results, and it’s clear they tweak that algorithm as they achieve an ever more mature view of the natural history of patients on active surveillance. Dr. Klotz told us that the third biopsy would typically be spaced several years later, as would subsequent biopsies, and that they ceased biopsies at age 80.

    Dr. Klotz too seems to be in the choir with you and Dr. Sartor. He specifically stated to us his misgivings about active surveillance programs that biopsied each patient every year, for instance. Once source of confusion is that Dr. Sartor emphasized Dr. Klotz in close conjunction with his critique of proceduralists, to coin a word. All that collegial comment aside, I’m thinking Dr. Klotz would disagree sharply with waiting for symptoms, a hallmark of late-stage disease, to appear. Somehow I suspect Dr. Sartor would too, though that does not come across in his article.

  13. Dear Jim:

    If Dr. Sartor was in the degree of agreement with Dr. Klotz that you are “seeing”, then why would he have done this Pro/Con piece at all? I still think you are missing Dr. Sartor’s point, although I would agree with you that Dr. Klotz (among the proceduralists) is moving nearer to Dr. Sartor’s position. Dr. Klotz is not the problem. The problem is that there are many proceduralists (inside and outside of academica) who are making active surveillance so invasive for the patient that they might as well have a radical prostatectomy … and what’s the point of that?

  14. Right!

  15. I see very little evidence, as opposed to anecdotes, that treatment has any benefit at lengthening your life until at least you get a palpable tumor or symptoms. On the the other hand, treatment usually has some side effects. In the USA things are further confused as prostate intervention is very profitable for a lot of people, which means there is a tendency to be proactive. Everyone has to make their own choice — I just gave up on PSA tests, which were through the roof, and decided not to do more biopsies, after one indicating mild cancer. Once you are in your 70s you are not going to live forever anyway. I would prefer not to spend whatever time I have left in hospitals and treatment centers.

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: