When should therapeutic intervention take place for men on active surveillance?

The fact of the matter is that, as yet, we don’t have a good answer for this question, and it may be many years before we do. The reasons for this are numerous, but we are beginning to get some of the information we need to clarify the situation.

In the first place, there is no consensus as yet regarding who is an appropriate candidate for active surveillance. The most restrictive cohort is the group of > 1,100 men with very low-risk disease enrolled at Johns Hopkins in Baltimore. The least restrictive (but longest studied) cohort is generally considered to be the one enrolled by Klotz et al. at Sunnybrook, Ontario, Canada. The largest cohort is thought to be the multi-center Prostate Cancer Research International Active Surveillance or PRIAS cohort of nearly 2,500 men.

The table below shows the diagnostic criteria required for entry into each of these three study cohorts:


It will immediately be apparent that there are significant differences between these three cohorts of patients.

Now we know from a variety of data published to date that anywhere between about 10 and 40 percent of patients in different active surveillance studies will undergo reclassification over time for any one of a number of reasons, including the following:

  • There is a finding of Gleason 7 disease — or higher — on repeat biopsy.
  • The patient’s PSA rises to a worrisome level (i.e., > 10 or > 15 ng/ml depending on the study protocol).
  • The patient’s PSA doubling time is too short (e.g., < 3 years).
  • There are findings of new areas of cancer (e.g., more than two positive cores) on repeat biopsy.
  • There are clear clinical symptoms of disease progression.

And of course these criteria do not take account of the men who decide, of their own volition, to stop active surveillance and have some form of treatment even though they do not have any clinical indication that this is necessary. Active surveillance is simply difficult for some men to deal with.

In the Sunnybrook study, patients are all regularly monitored using PSA tests and clinical re-evaluation; repeat biopsies are given as considered appropriate on the basis of PSA and other clinical data; and there are three conditions under which a patient is necessarily advised that interventional treatment may be appropriate:

  • A PSA doubling time of < 3 years
  • Gleason score progression
  • Unequivocal clinical progression

For the Johns Hopkins study, all patients are monitored with semi-annual PSA measurements (total PSA and %free PSA) and DRE, as well as annual 12- to 14-core biopsies; indications for curative intervention include any disease reclassification such that enrollment criteria are no longer met:

  • A PSA density >0.15 ng/ml
  • A Gleason score ≥ 7
  • More than two positive cores on biopsy or > 50 percent cancer involvement of any one core

For the PRIAS study, all patients are given scheduled PSA measurements every 3 months for the first 2 years and every 6 months thereafter; scheduled biopsies occur after years 1, 4, and 7; a PSA doubling time of between 3 and 10 years prompts immediate repeat biopsy; and curative treatment is recommended

  • On biopsy reclassification
  • For any man with a PSA doubling time of < 3 years

Again, it is apparent that these are three very different sets of criteria under which patients are managed and upon which interventional treatment will be recommended.

However, we also know, from the most recently published data from the Sunnybrook study, that

  • 15/993 patients (1.5 percent) on AS had died of prostate cancer
  • 13/978 additional patients (1.3 percent) who did not die of prostate cancer had developed metastatic disease.

In other words, only 2.8 percent of the 993 men enrolled by Klotz et al. who have been followed for up to 20 years have actually gone on to have progressive disease that has or might have been expected to lead to their deaths. Since the Johns Hopkins and the PRIAS studies have more restrictive study entry criteria and more aggressive recommendations for initiation of treatment, we can certainly expect that

  • The 10-year prostate cancer-specific mortality rates in the PRIAS and Johns Hopkins studies will be lower than those reported by Klotz et al.
  • The 10-year rates of prostate cancer metastasis in the PRIAS and Johns Hopkins studies will be lower than those reported by Klotz et al.

In addition, it is arguable that

  • The rates of over-treatment of patients who would never actually have died of their prostate cancer may be higher in the Johns Hopkins and the PRIAS studies than in the Sunnybrook study.

All this comes as a preamble to a recent report by Kates et al. in BJU International, which offers the following findings by application of the PRIAS criteria for therapeutic intervention to the Johns Hopkins patient cohort:

  • At a median follow-up of 2.1 years from study entry,
    • 427/1,125 men (38 percent) in the Johns Hopkins AS cohort had biopsy reclassification based on the Johns Hopkins criteria.
    • 226/427 men (62 percent) who had biopsy reclassification based on the Johns Hopkins criteria were detected at biopsy intervals corresponding to the PRIAS criteria.
    • 106/427 men (16 percent) who had biopsy reclassification based on the Johns Hopkins criteria were detected in between scheduled PRIAS biopsies, resulting in a median delay in detection of 1.9 years.
  • Of the 202 men with > 5 years of follow-up in the Johns Hopkins cohort, 22 (11 percent) were found to have biopsy reclassification after it would have been identified in the PRIAS protocol, resulting in a median delay of 4.7 years to reclassification.
  • 135/1,125 men (12 percent) who would have undergone immediate intervention under the PRIAS criteria  due to abnormal PSA kinetics would never have undergone reclassification on the Johns Hopkins protocol and thus would not have undergone definitive intervention.

The as yet unanswered question about active surveillance is whether we can find a near-to-100 percent protocol that would allow for no cases of prostate cancer-specific mortality and no cases of over-treatment of men who are never going to get metastatic prostate cancer and will die from some other cause. It appears that the protocol being used by Klotz and his colleagues at Sunnybrook may be “not quite good enough”, but is also seems likely that other current protocols may be “too good” at finding reasons to intervene when intervention may not actually be necessary.

To their credit, Kates et al. conclude only that:

There are clear differences between PSA kinetics-based AS programs and biopsy based programs. Further studies should address whether and how the differences in timing of intervention impact subsequent disease progression and prostate cancer mortality.

It should also be noted that none of the three protocols described has formally adopted the used of multi-parametric MRI scanning as a routine component of monitoring men on active surveillance (although Klotz has certainly stated his opinion that this is a potentially very important opportunity). And then there are all sorts of other biomarkers either available or in development that could also impact the ways in which we can monitor men on AS protocols in the not too distant future.

This is all, quite definitely “a work in progress.” Sadly, however, out in the “real world”, active surveillance is still being significantly under-utilized in the routine first-line management of men with low- and very low-risk prostate cancer. Far too many patients never receive a clear message at diagnosis that careful monitoring is a perfectly reasonable option when compared to immediate treatment — most especially for men of 65 years and older with a life expectancy of 10 to 15 years or less.

8 Responses

  1. Excellent post.

  2. Annual 12-point biopsies would be a deterrent for me. Biopsy is not a totally benign procedure. Risk of infection and the possibility of track seeding or otherwise spreading the cancer are possible concerns. AS continues until intervention, perhaps for many years, so cumulative biopsies could add up.

    Seems to be no mention of advanced imaging, which might be a useful addition to biopsy and PSA kinetics. Perhaps that is being studied separately?

    Correction, you did mention multiparametric MRI. Seems a good addition to the protocols.

  3. You summarized that 2.8% of the Klotz low-risk cohort have gone on to experience progressive disease or mortality. It seem appropriate to compare this rate to the several percent (pick your study) of low-risk patients who undergo immediate aggressive treatment yet then also experience progressive disease or mortality.

    There simply seems to be a very small baseline of what initially appears to be a low-risk case but in reality is a rare variety of prostate cancer which cannot be stopped. Shouldn’t the measuring stick of whether an AS criteria is “good enough” be compared to that baseline?

  4. Sitemaster,

    Thank you for writing this review and posting the links so viewers can read the entire articles themselves.

    The visual comparison of the three different active surveillance studies was particularly good and helped clarify them.

    One thing that has crossed my mind is whether there is a way to number each article that is done on the site, e.g., using the date 021715 and original article author name, and use this number/author name to reference a piece by piece story of answers to ongoing critical questions (e.g., if one is electing AS, what monitoring criteria would be best for your doctor to use?) that are then saved in say the newbie question section of this site.

    The idea would be to add new information to this question as it becomes available with the date/author tag (that the reader can go back to and read more) and would make it easier for someone new to the site to get up to speed quickly on a subject rather than have to read a year’s worth of articles to piece everything together.

    Just a thought. Don’t mean to make more work for the sitemaster. If done at the same time each review is done, it might not seem as much added work.

  5. Dear Casey:

    Arguably, yes, AS should be compared to the baseline you describe. However, since, at present, we don’t know how to identify, sufficiently early, these men with apparently low-risk disease at diagnosis who go on to have metastasis and die of prostate cancer, the important thing (at this time) is to make it abundantly clear to both the urology community and the patient community that active surveillance is a perfectly reasonable way to manage the majority of men with low-risk disease.

  6. Dear George:

    If you want to try to develop such a system and build pages that contain that information, please feel able to produce some samples! Alas your sitemaster already has the equivalent of 2.5 full time jobs and making time to talk to his wife is rather higher on his list of priorities!

    The other thing that you should be aware of is that we are selective about the articles we comment on. There are many thousands of new articles on prostate cancer each year, and we absolutely do not try to “cover” all of these. It is no longer even a reasonable possibility.

  7. I am employing mpMRI as a biopsy surrogate in my AS patients providing their DRE and PSA density and velocity are not worrisome. … Very high patient satisfaction and no patients progressing.

  8. Sitemaster,

    Thank you for everything you are doing. You are right that we do not want to overtax you, because you are already doing a tremendous amount. I don’t know how you do it all, to be honest.

    The consolation is that people all over the world are truly benefiting from your efforts, as there is no website quite like yours on prostate cancer. You are also to be applauded for your selection of important articles to review.

    There are no easy answers on this prostate cancer subject. Thank you again for all your efforts to help us understand it better.

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 )

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: