Hope S. Rugo, MD, FASCO: Now at this meeting, we saw updated data from the PALOMA-2 trial, the overall survival data, and then we also had seen some data published last year from the PARSIFAL trial looking at giving an aromatase inhibitor versus fulvestrant. We'd seen previous data suggesting there was a subset of patients who might do better with fulvestrant—really a unique group of patients who had no visceral disease, bone-only, had never received prior endocrine therapy. The PARSIFAL trial looked at a more real-world population. You had to be a year from your last AI [aromatase inhibitor] and you were randomized to take palbociclib with either an AI or fulvestrant. Initially, it was a superiority trial and then it changed to an equivalent trial. They didn't quite meet that confidence interval either, but the curves are right dead-on. That doesn't mean they seem to make any difference, which is intriguing. There's another study going on now, looking at getting letrozole first, with or without the CDK4/6 [cyclin-dependent kinases 4 and 6] inhibitor, and then you go to the fulvestrant with or without the CDK4/6 inhibitor, so that you find out whether you do better first or second. My bias on that is that you're going to do better if you get the better drug earlier because that's what we've always seen, but it's called the SONIA trial. With that in mind—and a lot of our trials had used palbociclib first-kid-on-the-block, we saw the long-awaited survival data from PALOMA-2 at ASCO [American Society of Clinical Oncology] 2022 with disappointing results. There was no difference in the intent-to-treat population and overall survival and long follow-up, but there were several unique characteristics of PALOMA-2 that are worth keeping in mind. One is it's the only trial in the first-line setting that included patients who had a disease-free interval from early- to late-stage disease of less than 12 months. That represented about one-third of the population. In that group of patients, when you look at a forest plot and divide the numbers up, they didn't clearly have a survival benefit, but the disease-free interval of greater than 12 months—the actual overall survival numbers are quite close to the other trial—that first-line ribociclib and AI trial in terms of the overall survival in the CDK group being well over 60 months. The first-line fulvestrant trial has the highest overall survival, but this is close to the AI trial. It's interesting that when you look at the population that are similar in the other trial, you still see this benefit. Of course, that wasn't the original plan of the study, so it makes it a little bit complicated. They also lost a lot of data, which is unfortunate and impossible to recoup. A lot of patients withdrew consent for reasons that are unclear, and so they just didn't have survival data on those patients, unfortunately. I'm interested in your guys' take on that. Let’s start with the Aditya; just your thoughts on the survival data from PALOMA-2.

Aditya Bardia, MD, MPH: Absolutely. PALOMA-2 was the first large phase 3 trial evaluating CDK4/6 inhibitor-AI versus AI-placebo as first-line therapy. I think this was the trial—PALOMA-1—that opened the field of CDK4/6 inhibitors. Since then, we've had several studies. Now, in terms of why the results of PALOMA-2 were negative while we are seeing positive results from MONALEESA-2, MONALEESA-3, MONALEESA-7, as well as we've seen positive results with MONARCH-2, I think this could be differences in study populations [or other causes]. As you were saying, the disease-free interval was different. If you look at that subset, the results looked similar. The second possibility is what happened after the patients had disease progression on first-line therapy. That was not controlled, so there could be some imbalances in the placebo versus the palbociclib arm. The third possibility is that there are some differences in the drug; it's just impossible to know without head-to-head trial comparison.

Hope S. Rugo, MD, FASCO: Anne, your thoughts.

Anne O'Dea, MD: My feeling is that if we look at the entire PALOMA series, we see the dramatic efficacy of this drug; 10% of patients continue on palbociclib at 7.5 years. I certainly have a couple of patients in my own clinic that started the drug, really, in February of 2015 and remain on the drug today. Clearly, it's an effective agent. Of course, someone must be first; there's some difference in the populations, as you said. There appeared to be a little bit of an imbalance in the missing data, like a higher proportion of missing data in the placebo arm compared with the [INAUDIBLE] arm. If you look, I believe, Dr Richard S. Finn from the University of California Los Angeles, showed a forest plot of the subgroups where the missing data was balanced, and it favored an overall survival trend for palbociclib. There's a lot of nuances to this. Later we'll be discussing some real-world data with palbociclib that's encouraging as well.

Transcript edited for clarity.