Active surveillance studies can have highly informative results — even when poorly executed!

The Göteborg group in Sweden have just published long-term follow-up data on the management of > 450 men identified with very low-, low-, and intermediate-risk prostate cancer as a consequence of the Gotebörg screening trial and managed on something they call “active surveillance” from as early as January 1, 1995, with up to 20 years of follow-up for the earliest individuals enrolled (through to December 31, 2014).

The patients clearly comprise a cohort with a very mixed set of risk factors. Patients were followed primarily with PSA tests every 3 to 12 months, and they were re-biopsied only when there was a clear case of clinical progression or otherwise every 2 to 3 years. There is no reference to re-biopsies prior to entry into the active surveillance cohort (as recommended and carried out in most other active surveillance cohorts), or multiparametric MRIs, or MRI/TRUS fusion biopsies — and the two latter techniques would not have been available in Sweden (or most other places) at the time of initiation of the management of this cohort.

The authors also state clearly that they had no predefined protocol for the management of these patients, so what actually happened in every individual case was left to the discretion of the treating clinician(s).

Here is what Godtman et al. report as the primary results of their study:

  • The trial enrolled 475 men with an average (median) age of 66.0 years at diagnosis who were initially managed on active surveillance.
  • The average (median) follow-up was 8.0 years.
  • 202/475 men (42.5 percent discontinued active surveillance and initiated treatment of some type.
  • Treatment-free survival for men who remained on active surveillance was
    • 47 percent at 10 years
    • 34 percent at 15 years
  • 54/475 men (11.4 percent) actually failed on active surveillance.
  • 6/475 men (1.2 percent) died of prostate cancer (and none of these men were very low-risk patients).
  • Compared to the risk for patients with very low-risk disease, the relative risk for treatment
    • 40 percent for low-risk patents(hazard ratio [HR] = 1.4)
    • 60 percent for intermediate-risk patients (HR = 1.6)
  • Estimated failure-free survival rates on active surveillance were
    • 87 percent at 10 years for all patients
    • 72 percent at 15 years for all patients
    • 94 percent at 10 years for very low-risk patients
    • 88 percent at 15 years for very low-risk patients
    • 85 percent at 10 years for low-risk patients
    • 77 percent at 15 years for low-risk patients
    • 73 percent at 10 years for intermediate-risk patients
    • 40 percent at 15 years for intermediate-risk patients
  • Again, compared to the very low-risk patients the relative risk for failure on active surveillance was
    • 220 percent for low-risk patients (HR = 2.2)
    • 480 percent for intermediate-risk patients (HR = 4.8).
  • Estimated prostate cancer-specific survival rates were
    • 99.5 percent at 10 years for all  patients
    • 96 percent at 15 years for all patients
    • 100 percent  for very low-risk patients at 10 and at 15 years
    • 100 percent for low-risk patients at 10 years
    • 94 percent for low-risk patients at 15 years
    • 98 percent for intermediate-risk patients at 10 years
    • 90 percent for  intermediate-risk patients at 15 years

Godtman et al. then conclude that active surveillance

… is safe for men with very low-risk [prostate cancer], but for men with low- and intermediate-risk [prostate cancer], active surveillance carries a risk of missing the possibility of being able to cure the cancer. It is questionable whether men who are not in the lowest tumor risk group and who have a long remaining life expectancy are suitable candidates for this strategy.

Worse still, they also state, as a “patient summary”, that

Long-term results from this study indicate that some men will miss their chance of cure with active surveillance and it is questionable whether active surveillance is a suitable strategy for men who are not in the lowest tumor risk group and who have a very long remaining life expectancy.

The “New” Prostate Cancer InfoLink is actually horrified by what we see as a highly misleading set of conclusions on the part of the Göteborg group in this paper. Here’s why:

  1. The research team apparently made no attempt whatsoever to ensure that the patients were good candidates for active surveillance before managing them on what was effectively a modified form of watchful waiting and not active surveillance at all. We acknowledge that they started doing this in 1995, when very few people had made any attempt to define good practice as regards active surveillance, but this was not active surveillance as defined by any other research team.
  2. The individual clinicians were at complete liberty to make recommendations to patients about management, and so almost certainly brought their own personal biases to bear. Again, that is not active surveillance.
  3. The cohort almost certainly contained a significant subset of patients who should never have been on active surveillance in the first place because (a) they had forms of intermediate-risk disease that was not appropriate for this type of management and/or (b) their apparently low-risk disease really wasn’t low-risk, and this could and should have been discovered on repeat biopsy within the initial 12 months after diagnosis.

Despite this, the results are actually astonishingly good for both the very low-risk and the low-risk subsets of patients, and they aren’t terrible for the intermediate-risk patients. Let’s look at those results again:

  • 94 percent failure-free survival at 10 years and 88 percent failure-free survival at 15 years for very low-risk patients
  • 85 percent failure-free survival at 10 years and 77 percent failure-free survival at 15 years for low-risk patients
  • 73 percent failure-free survival at 10 years and 40 percent failure-free survival at 15 years for intermediate-risk patients

No one has ever suggested that active surveillance is intended to be used to manage all patients without the potential for subsequent treatment if and when it becomes necessary. That’s not the point. The point is to defer unnecessary treatment until it becomes necessary (if it ever does) and to do that one must be monitoring the patient actively and assiduously — which did not, apparently, happen to all the patients in this study.

The risks to the men with low-risk and intermediate-risk disease in this study had more to do with the lack of any serious attempt to identify good candidates for active surveillance with accuracy (and then monitor them with care) than their diagnostic risk categories as defined by the Göteborg group!  Placing some of these men on active surveillance was almost certainly poor practice to begin with. On top of that, apparently, some of the patients were being monitored with nothing more than a PSA test every 12 months!

The “New” Prostate Cancer InfoLink sees these data as being highly confirmatory of the value of active surveillance — when the right people are actually placed on this form of monitoring; when the wrong ones are carefully weeded out by the use of confirmatory biopsies within 12 months of diagnosis; and when the treating institution has an appropriate and well-defined set of parameters for how active surveillance is conducted (although that protocol should be flexible enough to include men at relatively higher risk levels should they wish to start on active surveillance).


8 Responses

  1. I sometimes talk to patients who get periodic PSA tests and claim they are on active surveillance. As you point out here, they are putting themselves in danger. Time and again PSA kinetics have been rejected as a sole indicator of progression for very good reasons, mainly (1) PSA is affected by many non-cancer causes, and (2) some of the most virulent prostate cancer cells put out very little PSA. There is no substitute for confirmatory and follow-up biopsies. Thank you for exposing the harms that were unwittingly perpetrated in Göteberg. I know I will refer to this commentary often as a cautionary tale.

  2. But how much room is there for improvement, even from this imperfect design? The median age at entry was 66 years, and at 15 years prostate-specific survival was 96%, when the median age was 81 years. The OECD estimates life expectancy for Swedish males at 65 years at only 18 years, so men in this cohort are much more likely to die of something else than prostate cancer. I’m looking at these data and thinking that contemporary active surveillance protocols will ultimately have very low rates of prostate cancer mortality. This is reassuring for men who make this treatment decision.

  3. Dear Sitemaster:

    Thanks for spotting this paper and providing such a great review! Amen!

    It’s most encouraging for properly done active surveillance!

  4. Dear Jim:

    In the spirit of full disclosure, it was actually Allen Edel that spotted it and brought it to my attention.

    One of the great benefits of having Allen on board as a regular contributor (quite apart from the fact that he is much more knowledgeable about radiation therapy than I am) is that we are both able to keep an eye open for papers that the other might want to write about.

    As usual, two sets of eyes (and ears) are commonly better than one.

  5. Hello Mike,

    I think your comments are too negative on the Goteborg team’s conclusions. I am too cheap to pay for the full paper but in my files I had another (prior) paper by the same authors. This paper is in support of AS as means to reduce over-treatment. I doubt that the Goteborg team was negligent in publishing the initial paper that uses the same (although slightly reduced) cohort of AS men by ignoring those men as qualifying for AS with intent to cure. Granted that there was some mixture and some 20% of patients were not very low- or low-risk patients, but this adds value to the paper considering the study results.

    In defense of the authors I quote from their prior publication:

    “One limitation of this study is the lack of a strict AS protocol. However, patients have been followed mainly in one centre, where the same treatment policy has been applied and maintained during the whole study period. The stable rate of AS after the initial screen is an indication that no major changes in patient selection occurred during the 16-yr study period; in the Goteborg trial, compared to several other centres, AS has been a common treatment strategy since the study started in 1995.”

    In summary, when considering the limitation of biopsies in determining a Gleason with 100% accuracy; knowing how important this score could be in altering the natural course of the disease; knowing that some 30% of low Gleason scores are changed for the worse at surgery and that in AS cohort reports some 30% of patients progress, it seems to me that the caution in their conclusion is warranted and should be common knowledge for those considering AS these days.

  6. Dear Ralph:

    “Caution” is “warranted” whatever one may be thinking about doing as a form of management for localized prostate cancer. And, for men with low-risk and favorable intermediate-risk disease, practicing well-conducted active surveillance is an excellent form of warrantable caution by comparison with immediate treatment that is often unjustifiable on oncologic grounds. No one is suggesting that a significant percentage of men who start out on active surveillance aren’t going to need treatment at some point in the future.

    However, the paper you kindly referred us to still makes it very clear that this group was not practicing “active surveillance” if they failed to give men a careful repeat biopsy prior to placing their patients on the management protocol, men were only being given PSA tests every “3 to 12 months” and men were being “recommended to switch to deferred active treatment in case of a progression” based solely on PSA data, as implied in the paper’s abstract. They were certainly practicing a form of expectant management, but it wasn’t active surveillance as practiced by any other reputable group of researchers. It is the use of the term “active surveillance to describe what they were doing that leads to a thoroughly misleading conclusion.

  7. Mike,

    I still think you are too critical of the study conclusions. Dr. Klotz who is one or the prime pioneers for AS, wrote an editorial in support of the prior study mentioned in my post.

    It might not be AS as currently defined, but the results are good enough to help men decide that AS is the proper choice for early diagnosed prostate cancer. The problem is that currently the atmosphere here is negative to use the PSA test. Without this test many men will be diagnosed with more advanced disease …. Sad story indeed.

  8. Dear Ralph:

    Dr. Klotz is more tactful that I am. You are of course entitled to your opinion. So is Dr. Klotz. And I am going to stick to mine. :O)

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: