Breast Cancer: A Continuum of Care - Episode 3
Transcript:Adam M. Brufsky, MD, PhD: Just to follow up on the IHC4. We developed something very similar called the “Magee Equations.” We’re a big place, like Memorial Sloan Kettering Cancer Center and you guys. We have a lot of cases of breast cancer, a lot of 21-gene assays done on them in probably the 2005-to-2008 range. We took them and we did a regression equation based on ER/PR, HER2, grade, Ki-67, and tumor size. We put that together—again, we have 3 different equations that use different parameters, so they’re not all using every one—and we were able to predict the Oncotype DX score, or the 21-gene score, with that equation pretty well. And so, what some of us have done is started to use that actually, to tell us whether someone should get a genomic test or not. Because we found if you were low by the equation, you were never high. What do you guys think of those pre-screening assays to use? Kim hates it, I can tell.
Kimberly L. Blackwell, MD: No, no. I was going to say that I think lots of groups have shown that, but we all work in an environment where I think we trust our pathologists. I just want to point that out.
Adam M. Brufsky, MD, PhD: Absolutely.
Kimberly L. Blackwell, MD: And we give lots of second opinions at all of our institutions. We’ve all sat through tumor boards where the pathology was read out as grade 2 or grade 3, and then it’s presented at a tumor board and our pathologist says, “No, that’s not what I would have called it.” So, I think if you’re going to utilize something outside of a large industrial commercial genomic assay, you have to have a relationship with your pathologist and you have to put a lot of trust in your pathologist. Unfortunately, there’s just not that many breast-dedicated pathologists that are out in community hospitals. I hope I’m not offending the pathologists in the audience, but in some of these large central pathology labs, I don’t necessarily think the grading and the pathology report is necessarily as good as what we’re getting in our institution. That’s my plug.
Adam M. Brufsky, MD, PhD: Agreed. And I think that’s how these tests were developed. They’re developed not for our pathologists. It’s a matter of seeing a lot. If you’re seeing 300 to 500 or 1000 cases a year, you can pretty much develop your grading. If you’re seeing 15 or 20, that’s a little bit different.
Mark E. Robson, MD: But I do think the general point here that came up earlier, and that Aditya also is making here, is that the score itself is so precise.
Adam M. Brufsky, MD, PhD: Right, excellent point.
Mark E. Robson, MD: The number is the number. It’s not because it is derived from a group of patients who may or may not be similar to the person who’s sitting in front of you. You take other clinical factors into account. We already have RSPC (recurrence score pathology-clinical) for the recurrence score. But the one that we haven’t talked about is the additional information from other prognostic signatures in TransATAC. So, clearly, there’s other information that comes on board. And so, when you’re using the information to give prognostic information to a patient, you have to qualify it because there may be a difference between somebody who has a 0.5-cm tumor with grade 1 and a recurrence score of 10 and somebody who has a 3-cm tumor with 1 positive node and a recurrence score of 10—10 is unfair because, take an intermediate score, the low scores probably are generic.
Adam M. Brufsky, MD, PhD: Just for our audience: RSPC is the Oncotype DX with various clinical parameters put in. It’s available on the Genomic Heath website, so you can actually calculate that. And I used to use it before I started switching away a little bit from the 21-gene assay. But I think it’s very good. These are good points.
Transcript Edited for Clarity