Another new article validates proposed new IUSP grading system

Another new article (to be published in European Urology) has used data from > 25,000 patients to validate the proposed IUSP grading system, scheduled to replace the Gleason grading system over the next few years.

The new paper by Epstein et al. (see also this media release from Johns Hopkins Medicine) discusses a careful re-analysis of outcomes data — based on biochemical recurrence — from 20,845 consecutive men treated by radical prostatectomy at five, major academic institutions and an additional 5,501 men were treated by radiation therapy at two, major academic institutions.

The authors report the following:

  • Large differences in the biochemical recurrence rates between surgical patients with
    • Gleason scores of 3 + 4 = 7 as opposed to 4 + 3 = 7
    • Gleason scores of 8 as opposed to 9
  • Compared to men originally classified as having Gleason 6 disease, the hazard ratios (HRs) for biochemical recurrence were
    • 1.9 for men with Gleason 3 + 4 = 7 disease
    • 5.1 for men with Gleason 4 + 3 = 7 disease
    • 8.0 for men with Gleason 8 disease
    • 11.7 for men with Gleason 9 and 10 disease
  • The differences were less evident in the radiotherapy cohort as a whole (because of use of adjuvant or neoadjuvant androgen deprivation therapy among many patients with high-grade disease) but were clearly evident in patients undergoing radiotherapy only.
  • The new IUSP five-grade group system had the highest prognostic discrimination for all cohorts on univariable and on multivariable analysis.
  • The major limitation of the study was the unavoidable use of PSA-related biochemical recurrence rates as an end-point as opposed to cancer-related death.

The authors conclude that the proposed ISUP grading system has meaningful benefits over even the most recent updates to the Gleason grading system, specifically including:

  • Greater accuracy of grade stratification
  • The simplified 5-grade grouping (from grouped grades 1 through 5)
  • The potential to help reduce over-treatment of prostate cancer

While The “New” Prostate Cancer InfoLink understands that many physicians, patients, and advocates are concerned by changes that will be needed as we adapt to this new ISUP grading system for prostate cancer over the next few years, we concur with the increasing evidence that supports this proposed change away from the historic Gleason  grading system, and we believe (as do many pathologists and other specialist clinicians) that use of the new system will improve management of prostate cancer over time.

We also acknowledge, however, that it will take us time to adapt to this new system, and so we are highly supportive of the intent to switch to the new system carefully and slowly, with full adoption to be expected in something like 2018 to 2020 rather than immediately.

10 Responses

  1. It is a mistake. Way too generalized. If your practitioner does not know the difference between Gleason 7 = 3 + 4 and Gleason 4 + 3 you have a problem. Why group Gleason 4 + 4 with Gleason 5 + 5? This new system assumes practitioners are ill informed. That is a scary situation. If so they need to spend an hour doing some research. It is a step backward not forward.

  2. Dear Susan:

    (1) The new system does not group Gleason 4 + 4 = 8 with Gleason 9 and 10 at all. It very specifically separates current Gleason 8 from current Gleason 9 and 10. Currently, they are often grouped together.

    (2) There are other subtleties to the ISUP system that are explained in an earlier commentary (see here). It is not quite as simple as just replacing Gleason scores of 3 + 3 with 1, 3 + 4 with 2, 4 + 3 with 3, 8 with 4, and 9 and 10 with 5.

    (3) It eliminates any need to worry about tertiary Gleason pattern tissues.

    I do not agree with you at all that this is a step backward. It is a different way to provide clearer guidance about prostate cancer grading and aggressiveness.

  3. I admit I don’t understand any of this. Wasn’t it established long ago — possibly by Gleason himself, that recurrence risk goes up (and prognosis goes down) in the order: 3 + 3, 3 + 4, 4 + 3, 4 + 4, 3 + 5, 5 + 3, 4 + 5, 5 + 4, and 5 + 5? I can understand that 4 + 5, 5 + 4, and 5 + 5 behave so similarly and there is no currently known therapy that treats any better than the other, so they are logically grouped. But what is the value of grouping 3 + 5 and 5 + 3 with the 4 + 4s? D’Amico has shown they have different prognoses. The presence of any amount of Grade 5 might be worth noting separately. This system provides patients with less information, not more information.

    I think that Epstein may be confusing two separate ideas: (1) factors that determine risk (e.g., PSA, Gleason score, stage, cancer volume, PSA density, age, etc.) and (2) risk stratification or categories (e.g., very low risk, low risk, favorable intermediate risk, unfavorable intermediate risk, high risk, and very high risk — the current NCCN categories). The risk factors are grouped for inclusion into risk categories. For example, NCCN has already recognized Gleason 3 + 4 with < 50% positive cores and PSA < 20 and Stage < 3 as a noteworthy risk category useful for determining active surveillance eligibility. What impact could it make to call the grade component Group 2 instead of 3 + 4 within that risk category?

    I don't understand what patient benefit there is to expressing it as say, Gleason Grade Group 3 rather than Gleason Score 4 + 3. So far, I've only heard confusion from patients about this hodge-podge of yet another set of new numbers. I also think that evidence-based medicine demands that the nomenclature change be tested among patients.

  4. Dear Allen (and others):

    The ISUP system deals with one thing and one thing only — the interpretation of tissue histology under a microscope. This is also what the Gleason grading system dealt with. However, we have changed the Gleason grading system multiple times over the years because it didn’t work well for all sorts of reasons. Patients often think they know what Gleason scores imply but they really don’t because they don’t understand in detail what the actual grades relate to (the appearance of cells when stained in specific ways and viewed under a microscope).

    As I have tried to keep stating, each time this issue comes up, the ISUP system does not actually “convert” Gleason scores into ISUP grades at all. If you want to understand ISUP grading, you at least need to go and look at the summary table in this article. But even that is a massive simplification. The reason that people are interpreting ISUP grading as a simple conversion is because it is true that in many cases a man with (for example) a Gleason score of 3 + 4 = 7 would be found to have an ISUP grade of 3. But this is because of the histological appearance of the tissue sample, not because of what his Gleason score was.

    So long as people see this new system as a conversion, they will be confused. The fact that there is a rough correlation is useful, but it isn’t a conversion. For example, I can’t tell you that a specific man who is accurately or inaccurately assigned a Gleason score of 5 + 3 = 8 or 3 + 5 = 8 will be either a 4 or a 5 under the ISUP system. That would depend on the details of his individual histology — and some of those 3 + 5s would probably be 5s. Others might be 4s.

    The benefit of this system is that (a) it is histologically more accurate than the Gleason grading system; (b) to date it actually appears to be prognostically more accurate that the Gleason system (when combined with all the other relevant data — PSA, stage, etc.); (c) it eliminates the confusions related to all the oddities of Gleason grading (like 3 + 5) which are both rare and often found to be inaccurate when re-examined by actual experts; and a whole bunch of other stuff.

    People are currently working in great detail on the ways in which the ISUP system will need to be integrated into things like the NCCN risk categories. I find it rather funny that you think that Dr. Epstein might be confused about the factors that determine risk and risk categories. Epstein’s work over the past 25 years has been both fundamental and critical to the understanding of which factors determined levels of risk and thus the development of those risk categories.

    The idea that we should “test” the understanding of the ISUP system with patients would first require that patients understand how both a Gleason score and an ISUP grade get assigned. I have yet to meet more than a couple of patients who really do understand that. After all, very few patients have ever been given a detailed lecture on prostate anatomy and histology. And let me be very clear … Even though I do know quite a lot about this, I certainly don’t have the skill to look at a set of slides from a prostate biopsy and assign either a Gleason score or an ISUP grade.

    The Gleason scoring system and the ISUP grading system are tools. As I said, people often find change difficult. Look at the way with urology community and the radiation oncology community often resist change. The ISUP grading system is a better tool than the Gleason system. It is going to be adopted, and we are all going to need to get used to it.

  5. Mike,

    Thank you for the explanation. I admit I don’t understand the details of the grading system, but also I have never looked at prostate cancer cells under a microscope.

    What I do understand and have learned over the years that we need better classification of this disease on the front end. Especially in areas where a Gleason score was a little vague, such as my case, for which the pathology was 7 (4 + 3) with small amounts of Gleason pattern 5 on both sides of the gland. There was also small amounts of Gleason pattern 5 in my biopsy but not including in my staging. My understanding is that we are beginning to understand that there are different sub-types of this disease and that a move towards better classification may help patients and clinicians to seek out more specific treatments as they become available.

    Just my layman’s opinion. Thank you again for all your good summaries of important information. It is appreciated.


  6. Happy new year Bill.

    Chances are that if Dr. Epstein looked at your original biopsy slides today, he would tell you that you were actually an ISUP 4 or 5 as opposed to a Gleason 4 + 3, which would also help to explain the progression of your disease over time.

  7. Sitemaster,

    You are quite right in identifying the issue here. Is it a change in histological grading and scoring, or is it a simple conversion?

    The reason I think of it as a simple conversion is because that is the way Epstein describes it on his own website.

    According to Epstein, there are and have been since 2005, three histologically determined patterns of tissue architecture:

    Pattern 3: Discrete well-formed glands (pattern 1 and 2 not reported since 2005)
    Pattern 4: Cribriform/Poorly-formed/Fused Glands (cribriform was shifted up to pattern 4 in 2005)
    Pattern 5: Sheets/Cords/Single Cells/Solid Nests/Necrosis

    None of those are changing, as far as I can tell.

    Also not changing, as far as I can tell, are the ways the patterns are combined in the X + Y score. X is still the predominant grade, Y, if different from X, is the minor component, 5%.

    The exception set up in 2005 was that if any amount of a biopsy core is high grade — pattern 4 or pattern 5, it is reported as the second component. On pathology, it would be reported as a “tertiary” pattern instead. The only change, as far as I can tell, is that biopsy cores and pathology will be reported the same way.

    The rare scores you note, 3 + 5 and 5 + 3, are still very much part of the ISUP system (within Group 4), and are still histologically identified. This has not changed in the new system.

    I think you are quite right that this change in reporting WILL happen, but if I am confused by what this is about, imagine what it will cause in the average patient. I’ve already fielded patient questions stemming from their confusion.

  8. What I meant by my comment about testing among patients stems from my understanding of this as only a nomenclature change designed to prevent some of the anxiety the patient feels when presented with the numbers in the old way. I agree that cancer anxiety leads many patients to write-off active surveillance — but will this revised nomenclature change that? Epstein presents it as the goal. As Epstein puts it, “the potential to reduce overtreatment of indolent prostate cancer.” That is a testable hypothesis.

    There are alternatives to throwing yet more numbers at patients. What about A, B, C, D, and E? Or Lowest, Low, Middle, High and Highest? The ISUP Grade Group and maybe one of those alternatives can be tested among patients at Johns Hopkins quite easily: Report the ISUP Grade Group to one randomized set of patients, the old score (expressed as X + Y) to another set, and maybe one of the non-numeric alternatives to a third set. Then, see if those patient-sets differ in their treatment selection, time to treatment, and number of doctors seen. Did those low-risk men who saw the ISUP Grade Groups choose active surveillance more, less, or the same? Did they take more time to make an informed decision? You can follow up with a questionnaire probing their understanding or confusion. This would provide real medical evidence for Epstein’s hypothesis, and would seem to be fairly easy to do at Johns Hopkins.

  9. I am wrong in my reply above where I implied that patterns 1 and 2 will continue to not be reported. His new system reports Gleason scores less than or equal to 6 in his Group 1. This is a major change since 2005, and a disastrous one, in my opinion. Millions of more men will now be told they have cancer.

  10. Dear Allen:

    What we have in all of this is not a cancer problem. It is a temporary communication problem. And it is not going to be a communication problem to any man who gets diagnosed using the ISUP system once it is fully in place because they will never know their Gleason score. I disagree with you that Epstein thinks that this is a “conversion”. Note his comment that, “As a result of significant differences in criteria and reporting compared to the Gleason’s original grading system, we have regarded the newly proposed grade groups as a new grading system.”

    You also are incorrect in your assumption that men whose cancers look like pattern 1 or pattern 2 are currently not being told they have cancer. What is actually happening is that any visible cancer of pattern 1, 2, or 3 on biopsy is currently “rounded up” to 3 + 3 = 6, so changing all those cancers to ISUP Grade 1 won’t make any difference at all, except that there will be a clear associated statement that such ISUP Grade 1 cancers rarely need treatment.

    Could ISUP have used A, B. C, etc., instead of 1, 2, 3 … I suppose so. Would it have made any testable difference? I doubt it. All these patients are still going to be diagnosed with “cancer”. The major change that is going to take time is the change in the mindset of men and women to start to understand that most cancers being diagnosed today (as compared to most cancers being diagnosed in 1960) are actually not clinically significant at all. Incidentally, the new NCCN guidelines are now recommending active surveillance as initial management for favorable intermediate-risk prostate cancer … which will help.

    Do I think that ISUP could have done a better job of explaining why the revised grading system is important? Sure I do. But this is not a marketing exercise. It doesn’t have to be “sold”. And in my view, because it is going to happen whether people like it or not, people like you and I have an obligation to help to explain it simply and accurately.

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 )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: