Debu Tripathy, MD: Let’s move on to the area of neoadjuvant therapy. Clearly, there have been a lot of data coming out from using immunotherapy in that setting with different checkpoint inhibitors. Hope, do you want to comment on the data that were presented here at the 2019 San Antonio Breast Cancer Symposium in the context of the existing data?

Hope S. Rugo, MD: It’s actually fascinating because now we’re in another quandary. But the biggest trial is KEYNOTE-522, and that trial is really important because it shows our change in thinking. We want to do trials in the neoadjuvant setting evaluating new drugs that are powered by event-free survival [EFS], so it has almost 1200 patients. That’s really a new generation—it should be our standard. The pCR [pathologic complete response] end point was the first end point for that trial that added pembrolizumab to paclitaxel and CARBO [carboplatin] followed by AC [doxorubicin, cyclophosphamide], and everybody got paclitaxel—CARBO because you couldn’t have a non-CARBO arm. It would have been too big of a trial and would have taken forever.

The interesting thing is they could get to that end point at 600 patients. They saw more than a 13% improvement in pCR in those 600 patients that is clinically important to all of us. It is 1 of the highest pCR rates ever reported in a triple-negative breast cancer population and in a large population. They looked at the CPS [combined positive score] with that 22C3 antibody for PD-L1 [programmed death-ligand 1], which showed no difference. Really nothing predicted any difference. The only thing that they showed at this meeting was that if you got weekly CARBO, it was better than receiving carboplatin every 3 weeks, which doesn’t make a whole lot of sense.

Adam M. Brufsky, MD, PhD: I thought if your CPS was higher, you have something around a 71% pCR rate.

Hope S. Rugo, MD: The pCR rates were higher, but the differential was about the same.

Adam M. Brufsky, MD, PhD: The absolute number was higher.

Hope S. Rugo, MD: There may be a prognostic… The trouble is, when you look at CPS 20, there are only a small number of patients. We’re in a difficult situation knowing, but there has been a suggestion that it’s prognostic from a metastatic trial also. So, they’ll be able to look at that more with event-free survival. But we’re waiting for EFS. The toxicity was modest—exactly what you would expect. The EFS looks encouraging with a tiny bit of events, but we’re still waiting for it.

But then here we saw the NeoTRIPaPDL1 Michelangelo trial, where they gave atezolizumab with just a taxane and then went to surgery, taxane—CARBO and went to surgery. Basically, the issue is we think to maybe not give an anthracycline before surgery because the pCR rates were actually not better with adding atezolizumab, which was surprising given this big difference in KEYNOTE-522 trial. Again, I have the bias that it’s not agent-specific, but it is more that you actually generate a lot of immunity from the host by giving an anthracycline with cyclophosphamide—maybe you don’t need the cyclophosphamide; who knows.

Ian E. Krop, MD, PhD: You’re the expert in this by far. One thing that occurred to me when we were seeing those data was, is it possible that part of the difference between the 2 trials was the duration?

Hope S. Rugo, MD: People brought that up in the discussion also.

Ian E. Krop, MD, PhD: Did they? OK.

Adam M. Brufsky, MD, PhD: Duration of what?

Hope S. Rugo, MD: The duration was longer.

Ian E. Krop, MD, PhD: Total therapy.

Adam M. Brufsky, MD, PhD: It was 4 cycles versus 6, which is 8.

Ian E. Krop, MD, PhD: In the Italian study, did they have enough time to generate response?

Hope S. Rugo, MD: What’s interesting is that maybe that is why in the I-SPY 2 trial we only gave 4 doses of PEMBRO [pembrolizumab], but everybody got AC afterward. But it was only 2 months.

Adam M. Brufsky, MD, PhD: The answer to that would be B-59, which is ATEZO [atezolizumab].

Hope S. Rugo, MD: No. In I-SPY, what I’m mentioning is that there were only 4 doses of PEMBRO. It does bring up the question of whether it really has anything to do with duration. Lastly, the SAFIR02-IMMUNO trial, which is a fascinating trial in France, looked at durvalumab versus continuing maintenance chemotherapy. As we’ve seen, if you take an unselected group of patients getting maintenance chemotherapy versus nothing, it results in a better outcome, and we know that. That’s a lot of trials. It turns out durvalumab wasn’t much in this situation. It was an unselected group of patients, and maybe we need to give it in a different way or select the right group of patients. For example, if you took the responders to checkpoint inhibition, maybe it would have made a difference. That trial needs some follow-up, but I think it’s a little hard to interpret.

Ian E. Krop, MD, PhD: Or the immunotherapy has to have chemotherapy with it.

Hope S. Rugo, MD: It has to get a partner first.

Joyce A. O’Shaughnessy, MD: The triple-negative population, the PD-L1—positive population…

Hope S. Rugo, MD: There was a suggestion of benefit.

Joyce A. O’Shaughnessy, MD: Yeah. It was significant for survival. One was significant, 1 was not—small numbers but very encouraging for the triple-negative and for the PD-L1-positive, I thought.

Adam M. Brufsky, MD, PhD: Isn’t the IMpassion030 trial a longer design than the other IMpassion studies, the IMpassion with ATEZO? B-59 is a longer design.

Hope S. Rugo, MD: All the neoadjuvant to adjuvant trials are continuing except some of the HER2 [human epidermal growth factor receptor 2] trials are continuing the immunotherapy for…trastuzumab.

Adam M. Brufsky, MD, PhD: But what I’m asking is the amount of immunotherapy you get and chemotherapy you get before surgery.

Hope S. Rugo, MD: They’re all giving everything up front.

Joyce A. O’Shaughnessy, MD: Yeah.

Adam M. Brufsky, MD, PhD: Right. But NeoTRIPaPDL1 Michelangelo we said is 4 cycles. This is 8 cycles.

Joyce A. O’Shaughnessy, MD: It was 8 cycles, and each cycle was 3 weeks, so it was 6 months. It was just 2 agents, but it was 6 months.

Debu Tripathy, MD: I think this area is going to be really complicated by the fact that we’ve got a lot of trials, and the events are going to be small. With the exception of some of the large studies, they aren’t really powered to look at event-free survival. Here’s where maybe some of our newer diagnostics with minimal residual burden might be informative. We’re hoping that that area might allow us to shrink our trials. But clearly immunotherapy has many different possible applications with many different agents. The good news is that we have a lot of trials in this space. There’s been a lot of interest from the community in subscribing patients to these trials.

Transcript Edited for Clarity