HER2+ BC: A Year in Review and Updates from SABCS - Episode 3

Predictive Biomarkers: HER2-Enriched Subtypes

Hatem Soliman, MD, discusses data presented at the 2020 SABCS for patients with HER2-enriched subtypes that showed benefit from CDK4/6 inhibition.

Hatem Soliman, MD: There are some interesting data discussed by [Aleix] Prat, [MD, PhD,] about biomarkers in HER2-positive breast cancer that are notable to mention. In this presentation, which was in the first general session, there was an analysis that was done across the MONALEESA phase 3 studies, involving ribociclib in combination with endocrine therapy, with various partners and different lines. One of the standouts was in looking at the PAM50 [Prediction Analysis of Microarray 50] subtypes of the patients enrolled in the studies, not just looking at the immunohistochemistry markers of them being hormone receptor-positive, HER2-negative, which was the eligibility criteria, but rather looking at their molecular subtype using the PAM50 luminal A, luminal B, HER2-enriched basal subtypes and normal. They were able to demonstrate some interesting signals there, regarding the benefit from the addition of ribociclib to endocrine therapy, compared to the placebo group. In particular, the basal-like tumors fared quite poorly and didn’t appear to derive much benefit, because we know they don’t have a lot of estrogen signaling. But looking at this HER2-enriched subtype, showing a pretty significant benefit in adjusted hazard ratios for the addition of ribociclib compared to the placebo arm. The PFS [progression-free survival] hazard ratio adjusted was about 1.76 for the combination versus 3.47 for the placebo. When looking at that difference, it was surprising there was that subtype embedded within the hormone receptor-positive group that was really benefitting from CDK4/6 [cyclin-dependent kinases 4 and 6] inhibition.

Andrew Seidman, MD: There is recognition of the idea that triple-negative breast cancer isn’t 1 disease, there are meaningful biological subtypes, and we’re going to develop targeted therapeutics based on those distinctions. We’re beginning to start to see meaningful differences in the ER-positive, HER2-negative group, the largest group. You can split these patients by many different biomarkers. Previous work, looking to see if the most common mutation PI3 kinase would predict benefit to CDK4/6 inhibition, actually did not show any impact of PIK3CA mutation.

It is interesting that another classifier, PAM50, does separate those patients into groups who seem to derive differential benefit, with all of the appropriate caveats because this is called an exploratory retrospective analysis. It’s not that surprising that the more proliferative subtypes here, luminal Bs and HER2-enriched patients, seem to get a greater incremental bang for their ribociclib buck, in terms of PFS, and the response rate data from 9% to 40% has gained attention.

I have not used PAM50 outside of a clinical trial setting. It’s available out there, and prospective validation of these findings could lead to some differentiation, but for right now we just have to think about ribociclib in the clinic based on the population of patients treated in the MONALEESA trials.

Hatem Soliman, MD: The idea of looking at some of these molecular phenotypes across different trials could allow for the generation of new hypotheses to test in finding predictive biomarkers that could help guide treatment. If these patients didn’t get anti-HER2-directed therapy, they wouldn’t have been prescribed that based on their normal clinical characteristics. That they could still derive a significant benefit from that could provide additional rationale for combining these agents and these kinds of HER2-enriched, or HER2-signaling patients, in the future.

Andrew, Seidman, MD: The HER2-enriched group was about 13% of these patients. Hormone receptor-positive, HER2- breast cancer is about 60% of all breast cancer, so 13% of that is still a significant group of patients. This is an observation that deserves prospective evaluation. By the same token, the difference in response rate in the luminal A group was incredibly modest. There was 35% in the placebo group, and 39% in the ribociclib group.

Transcript edited for clarity.