Expert: Gleason pattern 3 is not a predictor of prostate cancer risk

In this video, Andrew J. Vickers, PhD, discusses the background and key findings from the study, “Amount of Gleason Pattern 3 Is Not Predictive of Risk in Grade Group 2–4 Prostate Cancer,” for which he served as the lead author. Vickers is an attending research methodologist in the department of epidemiology and biostatistics at Memorial Sloan Kettering Cancer Center in New York, New York.

Video Transcript:

Could you describe the background for this analysis?

The most important determinant of treatment for localized prostate cancer is your Gleason score. As everyone is aware, this has recently turned into what's called the Gleason grade groups from 1 to 5. Grade group 1 means you only have pattern 3. Grade group 5 means you have some pattern 5. We know what to do with very group 1; there's active surveillance. Grade group 5 needs very aggressive treatment. The real bread and butter decision making is for this group of patients in the middle, where we have to make some really tough choices. That's grade group 2, 3 and 4. These are all patients with some pattern 4, but no pattern 5. The way that we have split them up into these different grade groups, and therefore how we treat them, is based on the ratio of pattern 3 to pattern 4. If you have more pattern 3 than pattern 4, you are in grade group 2. If you have more pattern 4 than pattern 3, you're in grade group 3. If you have pattern 4 and no significant amount of 3 at all, then you're in grade group 4.

The whole idea of ratios is very unusual in cancer medicine. We always talk about tumor size. In breast cancer, we talk about how big the lump is, not the ratio of benign to cancerous tissue in the breast. Or in a melanoma, we look at the diameter of all the depth of the tumor. We don't look at how much is abnormal lead vs how much is is the actual cancerous melanoma. It's a little unusual approach we take in prostate cancer, and it also doesn't make any sense when you think about it. You can easily come up with scenarios, and we actually show one in the paper of men with a very large amount of 4, which is the aggressive cancer, and a very large amount of 3, which can very easily turn into aggressive cancer, we believe, but because there's slightly more 3 than 4, we put those in grade group 2. Then you find men with just a very small amount of pattern 4, and no pattern 3, and we put them in great group 4. So, the men with a smaller amount of pattern 4, that's the aggressive type, and much less cancer overall, are treated much more aggressively. That just doesn't make any sense.

What were the key findings from this analysis?

What we're essentially doing when we're making these decisions about how to treat men in grade group 2, 3, and 4 is to look at this ratio between pattern 3 and pattern 4. In a very well known paper that was by Ross et al and had Jonathan Epstein, who's one of the premier pathologists in prostate cancer. They said the pattern 3 does not metastasize; you need pattern 4 and pattern 5 to metastasize. So, it seems odd to condition our treatment on the amount of pattern 3. We actually asked the question about whether once you know how much pattern 4 a patient has, is it helpful to know how much pattern 3 you have? We looked at 2 very different datasets. The first data set was close to 500 men with grade group 2 prostate cancer. This was on biopsy, and we looked at how much pattern 4 they had and how much pattern 3 they had on the cores from their biopsy. We used as an end point did they have aggressive pathology on radical prostatectomy? Had the cancer spread outside the prostate? That's extra prosthetic extension, seminal vesicle invasion, or lymph node invasion. What we found in that paper was that once you knew how much 4 the patient had, the amount of pattern 3 on biopsy wasn't even statistically associated with your risk of having cancer that has grown outside the prostate. It's not was it clinically of interest; it just wasn't even significant at all.

Then we looked at quite separate data set. In this data set, we looked at men with grade group 2, 3, and 4 prostate cancer on radical prostatectomy. Our end point there was PSA as a general marker of oncologic aggressiveness. Then we found that if you knew how much 4 was in the prostate, so you've taken out the entire prostate, you can actually work out how much pattern 4 there is, in terms of CCs and pattern 4. Once you knew that, the amount of pattern 3, there was a slight statistical association with your PSA level, but it was completely not meaningful. In fact, it was less than the amount of benign disease. Then if you actually looked at the grade groups 2 and 3, we excluded the 4s, the amount of 3 wasn't even statistically associated with your PSA level. Our conclusion was that the amount of pattern 3 in men with grade groups 2, 3, and 4 is not predictive of oncologic outcome once you know the amount of 4, and therefore our treatment decision making and the prognoses we give to patients should not depend on the amount of pattern 3, which is what we do right now. Therefore, what we're doing right now is problematic and needs to be revised.

This transcription has been edited for clarity.