It has long been understood that men with a Gleason score (GS) of 7 could be divided into two groups: those with Gleason 4 + 3 disease (in which Gleason pattern 4 was more common or “dominant”) and those with Gleason 3 + 4 disease (in which Gleason pattern 3 was dominant).
It was first suggested many years ago that Gleason 4 + 3 = 7 disease was associated with a higher risk of progression over time than Gleason 3 + 4 = 7 disease, and clinicians have been taking account of this factor in their prognostic and management decisions ever since. However, it is gratifying to see that this trend continues to be true even in a much more modern series of patients.
Alenda et al. have reported data on the impact of the primary Gleason pattern on biochemical progression-free survival (bPFS) after radical prostatectomy (RP) in a series 2,239 patients all treated for localized prostate cancer at a single center in France between 1998 and 2008. Biochemical recurrence in these patients was defined as a PSA level > 0.2 ng/ml at any time after surgery.
The results of their study show that:
- 1,248/2,239 patients had an actual GS of 7 based on their post-surgical pathology.
- Median postoperative follow-up was 23.4 months.
- 721/1,248 patients (57.8 percent) had a final GS of 3 + 4
- 527/1,248 patients (42.2 percent) had a final GS of 4 + 3.
- Patients with GS 4 + 3 had a significantly higher risk of biochemical progression than those with GS 3 + 4 (P < 0.001).
- The 3- and 5-year bPFS for patients with GS 3 + 4 cancers were 84.6 and 76.4 percent, respectively
- The 3- and 5-year bPFS for patients with GS 4 + 3 cancers were 69.9 and 61.1 percent, respectively.
- When analysis was stratified by both pathologic stage and margin status
- The predictive value of the primary Gleason was significant in men with pT2N0, pT3-4N0, and pT3-4N1 cancers.
- Survival curves were not statistically different for men with pT2N1 cancers.
The authors conclude that in their modern series of patients, a primary Gleason pattern of 4 after radical prostatectomy remains a significant predictor for biochemical progression within 2 years of their surgery, compared to a primary Gleason pattern of 3 in men with GS 7 disease.
Filed under: Diagnosis, Management, Risk, Treatment | Tagged: biochemical, Gleason, pattern, progression, risk, score |
THE "NEW" PROSTATE CANCER INFOLINK 
I have always been puzzled that the Gleason grade does not take into account the percentage of each type of cells. Thus, a 90% Gleason 4 and 10% Gleason 3 are reported the same as a 51% Gleason 4 and 49% Gleason 3. Does anybody have an idea for the reason behind that?
Reuven:
I believe that there are two simple reasons for this.
The first is that that is simply the way that Gleason grading was originally structured by Dr. Gleason himself, back in the 1960s. He developed a system that was based on “the most” (the primary) and “the second most” (the secondary) amounts of tumor tissue of specific types.
The second (which perhaps explains why Dr. Gleason structured his system the way he did) is perhaps because the subtleties of interpretation of whether a particular piece of tissue, when seen under a microscope, is Gleason pattern 3 or Gleason pattern 4 can be difficult even for a highly experienced pathologist to interpret. To ask him or her to then make a sub-interpretation of whether it is a 60/40 ratio of pattern 4 to pattern 3 or a 70/30 ratio may simply be asking for an unreasonable level of accuracy for something that is — even at a first stage — based on the personal opinion of the pathologist. Have a look at the actual slides showing Gleason pattern 4 and Gleason pattern 3, and remember that these were John McNeal’s presentation slides that he used specifically to try to show the differences between the patterns to urologists. Then look at the graphic in the section on Gleason grading elsewhere on this site. Look at how the appearance of Gleason pattern 3 gradually gets closer and closer to that of Gleason pattern 4.
Thanks. I realize assigning Gleason scores is more of an art than a science. That being said, the same difficulty applies to deciding whether the samples are 3 + 4 or 4 + 3. I am sure almost nobody can decide whether the grade 4 cells are 51% or 55%, but I believe they can probably have a finer categorization and I have a feeling that 90%/10% is different from 60%/40%. I just wonder if it won’t be worth adding an additional number to the Gleason score and research what the implication on mortality and biochemical failure might be.
On a different subject, but also related to statistics. Most studies that I read on results of primary and salvage treatments don’t account for the age and health condition of the patients. The only one that differentiated between age cohorts is, I believe, the Goteborg study. This study showed marked differences between cohorts. I wish researchers addressed this issue.
Reuven:
The ability to provide data broken out by age group depends on all sorts of things from actual study structure and size to whether other factors are more significant than age. As Cooperberg and his colleagues in San Francisco keep pointing out, chronological age may be much less significant than co-morbidities and “physiological” age. I understand your frustration. “Young” Tony Crispino has the same aggravation. The problem with ideally strucutured studies is that it may not be possible to actually do them. I suspect the reason we haven’t seen breakdowns by age for studies like the PLCO and the ESRPC is the data was statistically meaningless. They certainly had all of the relevant data.
The report supports what I have been advising patients for several years. Granted I became aware of the importance of this differentiation from medical oncologist Stephen Strum.