What are the right eligibility criteria for active surveillance?

A couple of days ago we mentioned an abstract of a paper we had seen by Suardi et al. that discussed the “misclassification” of patients as being appropriate for active surveillence (AS). We have now had the chance to read this entire paper and are clearer about this analysis.

This is actually a rather interesting paper. What the authors have done is use data from two major academic urological surgery centers in Europe to compare the relative appropriateness of five published sets of guidelines proposed as classification schemes for AS and one even more stringent set. The six sets of guidelines are the following:

  • Those proposed by Hardie et al. (Group A): Clinical stage T1-2; PSA ≤ 20 ng/ml; biopsy Gleason score ≤ 7.
  • Those proposed by Roemeling et al. (Group B): Clinical stage T1c; PSA  ≤ 15 ng/ml; biopsy Gleason score ≤ 7
  • Those proposed by Choo et al. (Group C): Clinical stage T1b-T2b; PSA ≤ 15 ng/ml; biopsy Gleason score ≤ 7
  • Those proposed by Klotz (Group D): Clinical stage T1c-T2a; PSA ≤ 10 ng/ml for patients of age under 70 years and ≤ 15 ng/ml for patients of age over 70 years; biopsy Gleason score ≤ 6 for patients of age under 70 years and ≤ 3 + 4 = 7 for patients of age over 70 years
  • Those proposed by D’Amico and Coleman (Group E): Clinical stage T1c-T2a; PSA ≤ 10 ng/ml; biopsy Gleason score ≤ 6
  • A “most stringent” set of criteria (Group F): Clinical stage T1c; PSA ≤ 4 ng/ml; biopsy Gleason score ≤ 6

Suardi and his colleagues then analyzed how many of the 4,885 patients who received radical prostatectomies at Hamburg University Medical Center (3,337 patients; 68.3 percent) or at the University Vita-Salute San Raffaele in Milan (1,548 patients; 31.7 percent) between 1992 and 2007 actually met one or more of these sets of criteria. Their results are illuminating:

  • For Group A, 4,308/4,885 patients (88.2 percent) met the Hardie et al. criteria pretreatment, but 2,422 of the patients who met the pretreatment criteria for eligibility (56.2 percent) did not conform to the criteria based on post-surgical data.
  • For Group B, 4,047/4,885 patients (82.8 percent) met the Roemeling et al. criteria pretreatment, but 2,241 of the eligible patients (55.4 percent) did not conform to the criteria based on post-surgical data.
  • For Group C, 3,993/4,885 patients (81.7 percent) met the Choo et al. criteria pretreatment, but 1,795 of the eligible patients (45.0 percent) did not conform to the criteria based on post-surgical data.
  • For Group D, 2,455/4,885 patients (50.3 percent) met the Klotz criteria pretreatment, and 1,040 of the eligible patients (42.4 percent) did not conform to the criteria based on post-surgical data.
  • For Group E, 2,345/4,885 patients (48.0 percent) met the D’Amico and Coleman criteria pretreatment, and 925 of the eligible patients (39.4 percent) did not conform to the criteria based on post-surgical data.
  • Finally, for Group F, only 359/4,885 patients (6.9 percent) met the “most stringent” criteria pretreatment, and only 108 of the eligible patients (30.1 percent) did not conform to the criteria based on post-surgical data.

Eligibility was based on post-surgical Gleason scores of 7-10 for the outcomes data given above. If you use Gleason scores of 8-10 to define post-surgical eligibility, then the numbers of patients who did not meet the eligibility critera for each group post-surgery drop considerably, as follows:

  • For Group A, they drop from 2,422/4,308 (56.2 percent) to 1,149/4,308 (26.7 percent).
  • For Group B, they drop from 2,241/4,047 (55.4 percent) to 1,029/4,047 (25.4 percent).
  • For Group C, they drop from 1,795/3,993 (45.0 percent) to 995/3,993 (24.9 percent).
  • For Group D, they drop from 1,040/2,455 (42.4 percent) to 378/2,455 (15.4 percent).
  • For Group E, they drop from 925/2,348 (39.4 percent) to 325/2,348 (13.9 percent).
  • For Group F, they drop from 108/359 (30.1 percent) to 26/359 (7.2 percent).

What is one to make of these data?

Clearly the benefit of active surveillance is that it allows the physician and the patient to make decisions about interventional therapy only when they feel they need to based on their individual interpretations of risk. Arguably, a set of active surveillance criteria in which ≥ 25 percent of patients do not actually meet those criteria post-surgery is not the most acceptable set of criteria. This would imply that only the criteria defined by Klotz (Group D), D’Amico and Coleman (Group E), and the “most stringent” situation (Group F) offer appropriate guidelines for definition of patients suitable for AS at this time, and in fact these only “work” if one accepts Gleason 8-10 post-surgery as post-surgical criteria for eligibility. (In other words, a man with a post-surgical Gleason score of 4 + 3 = 7 would have met the criteria for AS.)

The greatest benefit of this paper, we believe, is that it offers a set of well-characterized guidelines for the development of a large, prospective clinical trial of active surveillance vs. immediate radical surgery for men of 70 years of age and less, who have been stratified by age group and by active surveillance risk group.

We should conclude by noting that the authors state that they cannot be sure that these data are applicable to the US population (because they are based only on Eutropean data). However, this caveat does not in any way detract from the theoretical value of the analysis.

2 Responses

  1. […] And talking of active surveillance (AS) in Europe, a report by van den Bergh et al. discusses the retrospective analysis of potential eligibility for AS of 616 men who were diagnosed with prostate cancer between 1994 and 2007. The authors’ goal was to retrospectively validate the currently used criteria for eligibility for AS. All 616 patients met the following criteria for AS: PSA ≤ 10.0 ng/ml; PSA density < 0.2 ng/ml per ml; clinical stage T1c/T2; biopsy Gleason score ≤ 3+3=6; and ≤ 2 positive biopsy cores. They were initially managed expectantly, with a median follow-up of just under 4 years. The authors report that the calculated 10-year prostate cancer-specific survival (in 21 patients at risk) was 100 percent, which contrasted with the actual 77 percent overall survival. Men still alive showed favorable PSA characteristics. Although the calculated 10-yearr treatment-free survival was only 43 percent, objective signs of progression often did not indicate the shift to radical treatment. The cohort consisted of men on AS and those on watchful waiting (WW). Information on comorbidity or psychological distress was not available. They conclude that AS seems justified in selected men with screen-detected PCa. Prospective protocol-based AS programs are necessary to optimize selection criteria and to find the appropriate trigger points for switching to active therapy. Possible negative psychological reactions with AS against improved quality of life by withholding side-effects from radical treatment should be considered. (Note: It is worth comparing the data reported in this study to the data reported by Suardi et al., on which we have commented previously.) […]

  2. There appears to be a clinical trial being run at the Royal Marsden Hospital in London which is looking at exactly this issue of active surveillance versus immediate treatment.

    (I have a personal interest, having been diagnosed at 43, but could find very little data to help in making a decision.)

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: