Clinical Rationale for PI3K Inhibitors in BC

Transcript:

Joyce A. O’Shaughnessy, MD: One of the things I want most in my practice is an inhibitor of PIK3CA mutations, given how common they are in breast cancer. At ASCO, we also heard an update on some of the PI3-kinase inhibitor trials, and we have some other data that we’ve heard before, from buparlisib. Hope, can you give us your take on where we’re at with the inhibitors of the PI3-kinase mutations?

Hope S. Rugo, MD: I’m very, very interested in having PI3-kinase inhibitors in the clinic. I think we’re moving closer to that with data presented this year, although we still need more. As you pointed out, it’s the most common mutation that is seen in HR—positive breast cancer. Close to 50% of patients have a PI3-kinase mutation. Now we’re seeing that’s a change, over time, as we talked about before. We see these mutations emerge during therapy, so it’s actually quite interesting. Maybe we’ll be able to screen more with blood as we move forward…? The first PI3-kinase inhibitor that reported data was buparlisib, which is a pan–PI3-kinase inhibitor. There was another one called pictilisib, which had a negative phase II trial.

Buparlisib was the subject of 2 phase III studies. These phase III studies included either patients who’d had prior mTOR inhibition or patients who did not. The trials were not set up prospectively to look just at the

PI3-kinase—mutant population as a primary endpoint. Buparlisib, as a pan–PI3-kinase inhibitor, is quite toxic. It’s got a marked increase in transaminases in a subset of patients, and it has a half-life of 40 hours. You’ve got to wait a long time for those liver enzymes to get better. People get depressed, and the depression can be quite severe. It’s not just the normal depression that we see in our patients, so there are side effects that are concerning with buparlisib, which I think make it less useful.

The really cool thing about these 2 different trials, BELLE-2 and BELLE-3, is that they looked at patients who had PI3-kinase mutations, looking at cell-free DNA or situations where they also had tissue. One of the studies actually showed a very nice correlation between the mutations in blood and the mutations in archived tissue, which is really important. It’s not perfect, but it’s pretty good. They also showed that the progression-free survival benefit, even in the positive trial, was pretty minor. With the toxicity, we wouldn’t consider that to be a clinically meaningful difference—1.9 months. But what they did show was that the difference was much greater in the PI3-kinase—mutant population. Again, they’re not definitive data, but the fact that both BELLE-2 and BELLE-3 showed that was very encouraging. That led to 2 randomized phase III trials using more alpha-specific PI3-kinase inhibitors, with a primary endpoint that looked at this in the PI3-kinase–mutant population. That’s really important.

At ASCO this year, we saw taselisib, which is a PI3-kinase inhibitor that has both alpha and delta activity. That trial is called the SANDPIPER trial, which was presented by Jose Baselga. It included over 500 patients who had known PI3-kinase mutations. In that group of patients, there was an improvement in progression-free survival, although it wasn’t what we’ve seen with CDK4/6 inhibitors. It was just 2 months. And, in fact, in that mutant population, the progression-free survival was relatively short, which was kind of intriguing because the partner was fulvestrant, where we’re used to seeing these much longer progression-free survivals. And we saw them in the CDK4/6 inhibitor second-line trials. This is clearly a subpopulation that we’re looking at here. The hazard ratio was 0.7.

They also looked at a small population who had PI3-kinase wild-type—about 100 patients. It was very difficult to interpret data in that group. The biggest issue with taselisib is the toxicity. A lot of patients dose-reduced, or held. Some patients discontinued due to toxicity, and that toxicity was diarrhea because it causes colitis. Sixty percent of patients had some degree of diarrhea. They also saw rash and hyperglycemia, which is a class effect of the PI3-kinase inhibitors. I think that we’ve moved from understanding that this may be a selection criteria for trying to find a good inhibitor that causes less toxicity. We’ve moved 1 step closer.

Joyce A. O’Shaughnessy, MD: Very interesting. And, like you said, this is a positive trial. It raises questions, because very few patients had prior CDK4/6 inhibitors…

Hope S. Rugo, MD: Less than 2%.

Joyce A. O’Shaughnessy, MD: That’s the question we really need to know, isn’t it? The PI3 kinase can be a mechanism of escape.

Debu Tripathy, MD: Also because the driving mutations in breast cancer are in the alpha subunit, PIK3CA. And so, theoretically, the more alpha-specific, the better. And alpelisib—we’re still waiting for the readout of that randomized trial. This one is a little more selective for the alpha isoform and may have fewer toxicities based on the data so far. We’ll see, of course, when the phase III data are out. So we are still looking for this class of drugs to make a difference.

Hope S. Rugo, MD: The SOLAR-1 trial is, in fact, a phase III trial looking at alpelisib with fulvestrant. This has a very similar design, and it’s powered to look at the PI3-kinase—mutant population. We hope to see that data later this year. So, 2018 will, hopefully, be the year when we see both of these studies get reported out.

And there’s an ongoing trial called BYLieve, which is looking at alpelisib along with 2 different hormone therapies in patients who have a PI3-kinase mutation, whose last treatment included a CDK4/6 inhibitor. This is very frustrating in the clinical trial setting, but it will give us the information we want.

Joyce A. O’Shaughnessy, MD: Yes. So, we have to stay tuned. There is some real proof of concept emerging, but we have to stay tuned.

Transcript Edited for Clarity