Adjuvant Palbociclib Fails in Early, High-Risk Breast Cancer

Adding a year of CDK4/6 inhibition to adjuvant endocrine therapy failed to improve invasive disease-free survival (iDFS) in early hormone receptor (HR)-positive breast cancer patients who had residual disease following neoadjuvant chemotherapy, a phase III trial showed.

In the PENELOPE-B study of over 1,200 patients, no difference in the primary endpoint of iDFS was observed between the group receiving palbociclib (Ibrance) and those assigned to placebo, after a median follow-up of 43 months (stratified HR 0.93, 95% CI 0.74-1.17, P=0.525), reported Sibylle Loibl, MD, PhD, of the German Breast Group.

"The addition of 1 year of palbociclib to endocrine therapy in patients with hormone receptor-positive/HER2-negative breast cancer and at high risk of relapse after neoadjuvant chemotherapy does not improve invasive disease-free survival," she said during her presentation at the virtual San Antonio Breast Cancer Symposium (SABCS). "No group could be identified with a higher benefit from palbociclib."

Landmark analyses of iDFS with palbociclib versus placebo, respectively, favored CDK4/6 inhibition early on:

• 2-year iDFS: 88.3% vs 84.0%
• 3-year iDFS: 81.2% vs 77.7%
• 4-year iDFS: 73.0% vs 72.4%

"At 2 years you can see that there was a benefit for palbociclib in the range of 4.3% compared to placebo," noted SABCS invited discussant Ruth O'Regan, MD, of the University of Wisconsin in Madison. At 3 years, the 3.5% difference in iDFS was still meaningful, she said, but by 4 years the curves had completely come together.

Recently reported data from the monarchE trial found an iDFS benefit when 2 years of abemaciclib (Verzenio), another CDK4/6 inhibitor, was added to adjuvant endocrine therapy in a similarly high-risk patient population (2-year iDFS: 92.3% vs 89.3% with placebo), but the trial had shorter follow-up than the current study. Another negative trial, PALLAS, tested 2 years of palbociclib.

"It's not definitive why the trials have different outcomes," said O'Regan. "If the monarchE results hold up, I think we have to assume that it's either the different CDK inhibitor, definition of high risk, or adherence, although the issue of adherence is not really supported by the PENELOPE-B results."

Loibl pointed out that PENELOPE-B is the first study showing mature iDFS results for a CDK4/6 inhibitor as part of adjuvant therapy in this high-risk population, and that long-term follow-up is critical for all of these trials to reveal if metastasis is simply being delayed rather than curing patients at risk for relapse.

Both Loibl and O'Regan suggested that the duration of CDK4/6 inhibition with palbociclib could have impacted the outcomes.

"It is possible PENELOPE-B might have been positive if palbociclib had been given for longer," said O'Regan. "And clearly we need to await the results on NATALEE that uses 3 years of ribociclib [Kisqali] for high-risk patients, as this will likely shed some light on this question."

Another factor, said O'Regan, could be that biologically high-risk cancers -- such as luminal B cancers -- benefit more from abemaciclib. "This hypothesis is supported by monarchE data showing greater benefit for abemaciclib in patients with high Ki67 tumors compared to the intent-to-treat population," she said.

Patients who fail to achieve a pathological complete response (pCR) following neoadjuvant chemotherapy have an inferior prognosis compared with those who achieve a pCR. For identifying those at high risk of relapse, PENELOPE-B used the clinical, pathological stage-estrogen/grade (CPS-EG) staging system, which can separate out patients with a very poor prognosis from those with a very good prognosis. Patients included in the trial had a CPS-EG score of ≥3 (59.4% of the population) or a score of ≥2 with nodal involvement (40.6%).

An interim analysis of overall survival (OS) in the trial showed no significant difference between the palbociclib and placebo arms, respectively (stratified HR 0.87, 95% CI 0.61-1.22, P=0.420):

• 2-year OS: 96.3% vs 94.5%
• 3-year OS: 93.6% vs 90.5%
• 4-year OS: 90.4% vs 87.3%

Overall, patients were more likely to discontinue palbociclib than placebo (19.5% vs 15.5%), with 5.2% and 0.8% due to toxicity.

Grade 3/4 adverse events (AEs) were four times more frequent in the study arm compared to the placebo arm (79.6% vs 20.1%, respectively), largely driven by hematologic events (73.1% vs 1.3%). Non-hematologic grade 3/4 AEs were similar, at about 20% in each arm. Serious AEs occurred in 9.3% of the study group and 8.7% of the placebo group.

Dose reductions were more common in the palbociclib arm, and by the last cycle about half of patients remained on the full dose of palbociclib.

Study Details

Starting in 2013, PENELOPE-B enrolled 1,244 patients with HR-positive/HER2-negative breast cancer who failed to achieve a pCR following neoadjuvant chemotherapy and surgery with or without radiation therapy. Patients received endocrine therapy based on local standards and were randomized to 1 year of daily oral palbociclib (125 mg) or placebo.

Study participants had a median age of 49 years, 50% had ypN0-1 lymph node involvement, 50% had yPN2-3 involvement, and one-fourth had Ki67 >15%. Stratification factors included age (≤50 years vs >50 years), region (Asian vs non), lymph node status (ypN0-1 vs ypN2-3), Ki67 (>15% vs ≤15%), and CPS-EG score (≥3 or ≥2 plus nodal involvement).

For treatment adherence, 88.6% of patients in the palbociclib arm completed at least seven cycles of treatment versus 90.3% in the placebo arm, with 80.5% and 84.5%, respectively, completing all 13 planned cycles. Patients on palbociclib had a median treatment exposure of 52.6 weeks and median dose intensity of 82.1%.

In 74% of patients, iDFS events were distant recurrences, and in 16% relapses were locoregional. The remaining 10% consisted of contralateral breast cancer events, second primary noninvasive disease, or death.

Comment