Frontline treatment with palbociclib demonstrated positive progression-free survival n real-world patients with HR-positive/HER2-negative metastatic breast cancer and may lead to an overall survival benefit , according to data from the pivotal PALOMA-2 trial.
Angela M. DeMichele, MD
Frontline treatment with palbociclib demonstrated positive progression-free survival (PFS) in real-world patients with HR-positive/HER2-negative metastatic breast cancer and may lead to an overall survival (OS) benefit , according to data in the the Flatiron database, whic was used in a retrospective study.
A retrospective database analysis of electronic health records calculated a 48% reduction in the risk of disease progression or death with the addition of palbociclib to letrozole versus letrozole alone in the first-line setting, according to findings presented at the 2019 San Antonio Breast Cancer Symposium (SABCS).1 The median OS was not yet reached in the palbociclib arm compared with 38.1 months (95% CI, 32.3not estimable) in the letrozole-alone group (HR, 0.52; 95% CI, 0.40-0.68; P <.0001).
The real-world data analysis determined that the median PFS was 20 months (95% CI, 17.6-23.7) with the addition of palbociclib compared with 12.1 months (95% CI, 10.3-15.2) with letrozole alone (HR, 0.55; 95% CI, 0.45-0.66; P <.0001).
“The hazard ratio of 0.55 was just about the same as what it was in PALOMA-2,” lead study author Angela DeMichele, the Alan and Jill Miller Professor in Breast Cancer Excellence, coleader of the Breast Cancer Research Program, and director of the Breast Cancer Clinical Trials Unit at the University of Pennsylvania’s Abramson Cancer Center, said in an interview with OncLive at SABCS.
In the pivotal, double-blind phase III PALOMA-2 study, postmenopausal women with ER-positive/HER2-negative advanced breast cancer were randomized in a 2:1 ratio to frontline treatment with palbociclib plus letrozole or placebo plus letrozole. The study, which led to the full FDA approval of the palbociclib combination in this setting, showed a median PFS at 38 months’ follow-up of 27.6 months with the palbociclib combination versus 14.5 months with letrozole alone (HR, 0.56; P <.0001).2
DeMichele and coinvestigators hypothesize that since their real-world analysis of palbociclib produced a similar PFS benefit in a population comparable to PALOMA-2, a similar OS benefit would also have been found when PALOMA-2 OS data are mature if the study had been powered to measure OS.
“The randomized trials of the CDK4/6 inhibitors in the first and subsequent lines were all really not powered for overall survival; they were powered for progression-free survival, but what we didn’t know was whether this translates into people living longer, and we’re never going to be able to know that from those trials, because they weren’t big enough to tell us that. So we said, ‘Maybe we could use real-world data from a big database and see, now that [palbociclib] has been out there for a while, whether there is a survival benefit because we have the advantage of a lot more patients.’”
For their analysis, DeMichele and coinvestigators used data from Flatiron Health Analytic database, which contains electronic health records from more than 280 cancer clinics in the United States. The database includes information on over 2.2 million patients with cancer.
For patient selection, DeMichele said she and her coinvestigators basically used the same eligibility as the PALOMA-2 trial. Individuals selected for the analysis were women aged ≥18 years with HR-positive, HER2-negative metastatic breast cancer treated with frontline palbociclib plus letrozole or letrozole alone between February 2015 and May 2019. Prior treatment with CDK4/6 inhibitors, aromatase inhibitors, tamoxifen, raloxifene, toremifene, or fulvestrant in the metastatic setting was not allowed.
The initial patient selection from the database yielded 772 patients who received frontline palbociclib plus letrozole and 658 patients who received letrozole alone. However, patient characteristics were not all well balanced between the cohorts because it was a real-world, nonrandomized sample with patients selected for treatment based on their characteristics.
“You can imagine with palbociclib, though it’s pretty well tolerated, there is neutropenia, so it may be that patients who were sicker, more frail, and/or older, were more likely to just get letrozole than letrozole plus palbociclib,” said DeMichele.
To address the imbalance, the investigators used a statistical modelling method called propensity score matching (PSM), which accounts for all of the factors that go into a patient’s condition and matches patients between cohorts to level the playing field.
After 1:1 PSM of baseline characteristics, 464 patients remained in each arm. The median age for the PSM-adjusted cohorts was approximately 67 years in each group, with about 25% of patients in each group aged ≥75 years. About 55% of patients in each cohort had stage III/IV disease. Most patients who had available ECOG performance status data had a score of 0 or 1.
Just over one-third of patients in each arm had visceral disease at diagnosis, and about 40% in each group had bone-only disease. Over 90% of patients in each arm did not have brain metastases. More than 80% of patients in each arm had either 1 or 2 metastatic sites. The most common subsequent second-line anticancer treatments received were a CDK 4/6 inhibitor (41.5% in the palbociclib arm vs 60.8% in the control arm), chemotherapy (25.9% versus 9.4%, respectively), and endocrine therapy alone (17.5% vs 22%).
The median follow-up was 23.1 months in the palbociclib group and 24.4 months in the control arm. The 2-year OS rate was 80.1% versus 63.9% in the palbociclib versus control arms, respectively.
“A consistent OS benefit of palbociclib plus letrozole versus letrozole alone was observed generally across all studied subgroups except for race,” noted DeMichele.
Regarding the impact of these data on clinical practice, DeMichele said that, alone, the data are not sufficient to guide treatment as compared with a clinical trial; however, she said the data are corroborative, and an oncologist could use these findings in conjunction with clinical trial data to say, “‘It really does work this way in the real word,’ and it might then push a doctor to give the drug in a situation where he or she might not have otherwise.”
As for next steps with CDK4/6 agents overall, DeMichele noted the push to move this class of agents up to early-stage disease. She is one of the investigators on the phase III PALLAS study (NCT02513394), which is comparing palbociclib plus standard adjuvant endocrine therapy with standard adjuvant endocrine therapy alone in patients with HR-positive/HER2-negative early breast cancer.
“We’ve completed the enrollment of all 5600 patients on the trial and we’re waiting for those results. And there are similar trials going on with the [CDK4/6 inhibitors] ribociclib and abemaciclib.”
The other big focus with these agents, said DeMichele, is overcoming resistance to CDK4/6 inhibitors that eventually develops. Researchers are trying to determine, “Is this like [trastuzumab], where you should stay on the CDK4/6 inhibitor ‘forever’ and just keep adding other [treatments] to target the disease in different ways? Or is it the case that once you progress on these agents, you should move on to something else. To figure that out, we really need to understand what the mechanisms of resistance are,” said DeMichele.