A recent French study nicely illustrates the very real problem associated with the selection of the best candidates for active surveillance and the exclusion from active surveillance of patients who probably would be better served by immediate treatment.
Ploussard et al. conducted a retrospective analysis of data from their series of patients to identify those who met the following, very rigorous, criteria:
- They had received a 21-core biopsy.
- Their PSA level was ≤ 10 ng/ml.
- They had clinical stage T1-T2a disease.
- They had a Gleason score ≤ 6.
- They had only 1 or 2 positive biopsy cores.
- The tumor length in each positive biopsy core was < 3 mm.
In other words, these were patients who, at least in theory, would be very strong candidates for active surveillance (especially if their life expectancy was < 20 years).
Among this cohort of patients, 96 were actually given an immediate radical prostatectomy (RP) and they had also received a prostate MRI prior to their surgery.
The primary goal of the study was to discover whether an MRI prior to surgery was capable of accurately identifying the who had “unfavorable disease features” at RP. In other words, was it possible to accurately identify the patients who were the best candidates for active surveillance as opposed to treatment based on the MRI data? “Unfavorable disease features” were defined as a pathologic stage of pT3 or pT4 and/or a Gleason score ≥ 4 + 3 = 7.
Here are the basic results of this study:
- The average (mean) age of the patients was 62.4 years.
- The average (mean) PSA level was 6.1 ng/ml.
- 17.7 percent of the patients had a pathologic stage of pT3 at surgery.
- 24.0 percent of the patients had one or more “unfavorable disease features” at surgery.
- MRI suggested that 28/96 patients (29.2 percent) had T3 disease.
- MRI results were not a significant predictor of the presence of pT3 disease in RP specimens, the rate of unfavorable disease, the presence of positive surgical margins, or Gleason upgrading.
- In fact, no preoperative parameter was an independent predictor of unfavorable disease in the RP specimen (based on a statistical model).
- After a mean follow-up of 29 months, biochemical recurrence-free survival (bRFS) was statistically equivalent between men with T3 on MRI and those with T1-T2 disease on MRI.
The authors carefully conclude that, “when the selection of patients for [active surveillance] is based on an extended 21-core biopsy scheme, and uses the most stringent inclusion criteria, MRI does not improve the prediction of high-risk and/or non-organ-confined disease” in a radical prostatectomy specimen.
However, there is a bigger probem. This problem is that about a quarter of the patients in this study appear to be poor candidates for active surveillance based on their post-surgical pathology data despite the fact that they meet very rigorous criteria for eligibility for active surveillance. (This is regardless of the MRI data.) Of course this series of patients has to include many with a life expectancy of 20 or more years (based on the average age of the patients in the study). If one excluded any patient in this study who had a life expectancy of > 15 years, one might be able to argue that there was no issue here. But one of the real values of active surveillance in the long term would be our ability to use it with great skill in men well under 60 years of age, so that we could avoid unnecessary invasive therapy in more men who still enjoyed a highly active sex life.
What this study tells The “New” Prostate Cancer InfoLink is that there is a way to go yet in being able to accurately identify patients who are truly the best candidates for active surveillance — despite the fact that we are getting better at this.
Filed under: Diagnosis, Management, Risk | Tagged: active surveillance, criteria, prognosis, selection |
THE "NEW" PROSTATE CANCER INFOLINK 
I agree with your conclusion, but to put it more strongly, the study seems to indicate that given the uncovered uncertainty, it seems fair to state that AS is mostly a decision to defer treatment.
Reuven:
I think that AS is only ever a decision to defer the need for treatment. However, the better we can be at deciding who the best candidates for AS are, the greater the likleihood that one can defer treatment for long periods of time (and even forever). What worries me most about the study discussed above is that patients who already have unidentified T4 disease at diagnosis and initial “work up” are clearly not candidates for AS at any point in time. Their treatment should never even be deferred (unless for other reasons they have a relatively brief life expectancy).
Mike,
If the amount of over-treatment is as real and high (30% to 50%) as they claim, why those men that were never going to be affected by prostate cancer (and still treated) are so difficult to select. There is more to that smoke screen …
Ralph:
It’s easy to suggest that most clear cases of over-treatment can be laid at the feet of physicians. Unfortunately, that is not entirely the case. I am personally aware of many, many men who, with very clear indications of disease that would best be managed by active surveillance (based on their PSA, biopsy data, life expectancy, etc.), and who had been so advised by at least one physician, still insisted on having active treatment.
Now I am not suggesting that there aren’t physicians who believe that if you find prostate cancer you should remove it. (Dr. Catalona argues against active surveillance regularly at major medical meetings.) However, what I am suggesting is that there is a very strong cultural bias in the U.S. toward immediate treatment of any and all forms of cancer — however insignificant. This has been built up over the past 60 or so years based on the evidence (from back in the 1950s) which showed that nearly all cancer was terminal. And at that time it was, because most cancers were only diagnosed when they had become clinically significant. Even after you were diagnosed — nearly 20 years ago now — people like Terry Herbert and the late Tom Feeney (as well as your truly) would be regularly “shot at” by the patient community for suggesting that not everyone needed to rush to treatment.
Over-treatment of prostate cancer is not (in my very humble opinion) only a physician issue. It is a cultural issue too. That cultural perspective took 50 years to build. It may take the same amount of time to tear down. Do I like that? No, of course not. But at a time when the NCCN guidelines clearly state that active surveillance should be used as first-line management of a large subset of men diagnosed with low- and very low-risk prostate cancer, I don’t think we can lay all of the blame on the medical community any more.
I don’t follow: “If one excluded any patient in this study who had a life expectancy of > 15 years, one might be able to argue that there was no issue here. ” Seems like that is refusing to look at the circumstance that matters most (consideration of AS for patients with long life expectancy).
Cultural biases notwithstanding, doesn’t the study say 1/4 of carefully scrutinized AS patients would have had a good chance of benefiting from treatment intervention? That’s a significant enough risk of AS to make any patient leery of AS.
Dave:
The NCCN guidelines currently recommend active surveillance only for men with low-risk prostate cancer and a life expectancy of < 15 years or for men with very low-risk prostate cancer and a life expectancy of < 20 years.
We don't know (from the abstract; and I haven't seen the full paper) how many of the men in the above study met one or other of those criteria. Maybe none of them did — but I doubt that.
However, you are absolutely correct. A key issue in the most beneficial use of active surveillance is whether it can be used (at least for a while) in men with longer life expectancies. And the whole point of the commentary is that if about 25% of men who seem to meet pretty rigorous criteria for active surveillance (excluding their life expectancies) may actually have disease that needs early treatment, then, as you say, "That’s a significant enough risk … to make any patient leery of AS."
Bottom line: Current criteria for AS are better than nothing … but they still need improvement.