Transcript:Adam M. Brufsky, MD: So TDM-1 is a great drug. You presented great data several years ago suggesting it as a first-line agent, we all got really excited and then MARIANNE came out. So what happened?

Sara A. Hurvitz, MD: It’s a lesson to us all about phase II data.

Adam M. Brufsky, MD: So we should probably explain to everybody what MARIANNE was before we continue.

Sara A. Hurvitz, MD: So MARIANNE was a phase III randomized trial. It was a three-arm study comparing taxane, whichever one the investigator chose, plus trastuzumab. So there’s no arm of THP. It was taxane/trastuzumab versus TDM-1 plus placebo or TDM-1 plus pertuzumab. So it was begging this question of will the addition of pertuzumab and trastuzumab in the TDM-1 component be better than TDM-1. But it was the big question, is there a role for TDM-1 in the first-line HER2-positive setting. Is it going to beat docetaxel and trastuzumab or paclitaxel and trastuzumab? The TDM4450 study that you just were speaking about was a smaller phase II randomized study, non-blinded, that looked at docetaxel/trastuzumab versus TDM-1, and there was a very nice five month progression-free survival benefit with the use of TDM-1 and about half the level of grade 3/4 toxicity. So we were all expecting that the MARIANNE study was going to show similar data.

Now, MARIANNE was set up as a noninferiority design where they compared TDM-1 versus the standard of care of taxane/trastuzumab. And what they showed was that, in the TDM-1 arms, neither of them were noninferior to the taxane/trastuzumab arm. So they met that part of the equation, but they were not superior either. So everyone has interpreted this as a negative study because it didn’t show that benefit over first-line trastuzumab and taxane. There was less toxicity. Patients did stay on TDM-1 longer when you look at the duration of therapy in the different arms. However, with the survival data that we have for THP with CLEOPATRA, I don’t think any of us are really using TDM-1 in the frontline setting because we aren’t seeing that sort of result in this study. And pertuzumab added to the TDM-1 didn’t appear to add much. That was the other kind of surprising feature of this study.

Adam M. Brufsky, MD: When you look at the subsets, numerically, they clearly showed some benefit, especially in the pretreated population Again, I know we’re using subset analysis. You can throw things at me if you want. But, seriously, when you look at the subset of the pretreated patients, there clearly were benefits to the TDM-1.

Hope S. Rugo, MD: In the taxane pretreated patients who in my clinical practice I don’t see as much. But I think it’s hard. Those are small numbers.

Adam M. Brufsky, MD: I know they’re small numbers.

Hope S. Rugo, MD: It’s quite intriguing.

Adam M. Brufsky, MD: It has to be considered in the study, I agree.

Hope S. Rugo, MD: But, you know, I do give TDM-1 first-line sometimes.

Sara A. Hurvitz, MD: Older patients, patients who don’t want to lose their hair.

Hope S. Rugo, MD: Don’t want to lose their hair, you got it, yeah.

Adam M. Brufsky, MD: Right. So, clearly, for whatever reason, TDM-1 and it was a disappointment because we all were looking really forward to that being the regimen of choice. But, okay, so we use it second-line and beyond. What about using of everolimus? What about BOLERO-1?

Sara A. Hurvitz, MD: Yeah. So we presented the BOLERO-1 study looking at inhibiting mTOR in patients with HER2-positive disease. There are fairly good preclinical data, some from our own lab and a lot from other people’s labs, looking at PI3 kinase/AKT/mTOR inhibition in patients with HER2-positive disease, showing that you can circumvent resistance pathways. There were good phase I and phase II data that were promising. Ruth O’Regan presented and published second- and third-line data looking at everolimus with vinorelbine and trastuzumab showing a five week benefit. Again, we get back to toxicity in the era where we have all these lesser toxic therapies.

So we looked at paclitaxel/trastuzumab with or without vinorelbine, a 2:1 design, and there were two primary endpoints. One was in the full population, progression-free survival, and the other one was in the patients who had ER-negative and HER2-positive breast cancer. So, for the full population there was no benefit with the addition of everolimus. In the patients who had hormone receptor-negative HER2-positive breast cancer, there was a seven-month delta in progression-free survival, favoring the everolimus therapy. However, it didn’t meet that prespecified, very stringent P-value that had been set forth when we split the alpha to do that analysis. And that was an analysis that was added before the study was unblinded, but well after it closed to enrollment. So it wasn’t really powered to look at these two things. So it’s interesting and hypothesis generating. Should we be looking at inhibitors of the PI3 kinase pathway in hormone receptor-negative HER2-positive breast cancer? Possibly. Should we be looking at endocrine therapy plus HER2-targeted therapy, plus PI3 kinase inhibition?

Hope S. Rugo, MD: Or CDK inhibitors.

Adam M. Brufsky, MD: Or CDK inhibitors, yeah.

Sara A. Hurvitz, MD: Definitely CDK inhibitors, definitely. So it’s not something we’re using prime-time I don’t think in today’s setting. Are any of you using everolimus?

Adam M. Brufsky, MD: No, everolimus, no.

Hope S. Rugo, MD: I loved your presentation and paper, but what I thought was a very intriguing analysis that came out of BOLERO-1 and BOLERO-3 was Dennis Slamon’s analysis of patients whose tumors had activating mutations in PI3 kinase, and that included a lot of archival tumor samples.

Sara A. Hurvitz, MD: Yes.

Hope S. Rugo, MD: So it wasn’t all metastatic samples which, you know, PI3 kinase may not be altered as much over time as some other things. And what it suggested was that if the patients didn’t have an activating mutation in the pathway, so that’s not just looking at PI3 kinase, it’s looking at anything that causes activation of the pathway which is I think a critical part of where we need to be going at looking at these predictive markers. And in that they found a big difference in the population combined of BOLERO-1 and -3, so first- and then the second- or third-line patients. So that I hope will translate into something we can learn more about with the new generation of PI3 kinase inhibitors as we move forward.

Sara A. Hurvitz, MD: I agree. That’s a great point.

Adam M. Brufsky, MD: Yeah. But, for now, I think that as a clinical practice changing thing, probably not. This is more unfortunate. There’s just more to kind of drive the next series of research projects forward.

Transcript Edited for Clarity