Phase 3 Study Confirms Clinical Equivalence for ABP 980 and Reference Trastuzumab

Following the previous release of top-line results, complete phase 3 data show that a biosimilar almost beat reference trastuzumab (Herceptin) for the primary endpoint of pathologic complete response (pCR). Detailed findings from the phase 3 trial with the proposed biosimilar, ABP 980, were presented at the 2017 annual meeting of the European Society for Medical Oncology (ESMO) in Madrid, Spain.

In the head-to-head trial, Lilac, the relative advantage of the biosimilar for the primary endpoint of pathologic complete response (pCR) reached statistical significance on local (but not on central) review, reported Serafin Morales, MD, a coauthor of the study and medical oncologist at the University Hospital Arnau de Vilanova, Lledia, Spain. However, a demonstration of superiority was not the goal of the trial. The key study conclusion is, says Dr. Morales, that “ABP 980 and trastuzumab show great biosimilarity.”

The phase 3 study was the last step in testing ABP 980 after extensive laboratory and early clinical phase studies confirmed similar pharmacokinetics and biologic activity. In this double-blind, multicenter study, 725 women with HER2-positive early breast cancer were randomized to receive ABP 980 or the reference trastuzumab. After 4 cycles of chemotherapy (epirubicin and cyclophosphamide), patients received a neoadjuvant course of their assigned therapy along with paclitaxel for 4 cycles.

After surgery, patients received ABP 980 or reference trastuzumab every 3 weeks for up to 1 year, but the efficacy analysis was conducted from tissue collected at surgery. The study’s primary endpoints were risk differences (RD) and risk ratio (RR) of pCR in breast tissue and axillary lymph nodes by local laboratory and central pathology evaluation. Secondary endpoints were incidence of treatment-emergent adverse events, changes in left ventricular ejection fraction (LVEF), incidence of anti-drug antibodies (ADSs), and on-study event-free survival (EFS).

The proportion of patients who received all 4 doses of their assigned neoadjuvant therapy was similar, reaching 98.1% for ABP 980 and 97.5% for reference trastuzumab. The weight-based cumulative dose of the 2 drugs was identical at 25.8 mg/kg.

By local review, the primary pCR endpoint was achieved in 48.0% of those randomized to ABP 980 versus 40.5% of those randomized to reference trastuzumab. The advantage for the biosimilar exceeded the prespecified 13% margin for bioequivalence by RD. While the lower boundary of the 95% confidence interval (CI) for RD exceeded 0% (1.2%), the upper boundary surpassed 13% (13.4%), which met the criterion of a significant advantage. The RR on local review analysis showed the same result.

However, the prespecified margins were not exceeded in the central review. Although the median relative RD was greater for ABP 980 than for reference trastuzumab (5.8%), the margins were below 0% (—0.5%) and below 13% (12%). The central review of the RR margins also fell short of statistical significance.

Richard Markus, MD, PhD, vice president for global development at Amgen, said “We cannot absolutely conclude that ABP 980 is not superior.” However, he emphasized that superiority was not a study hypothesis, and that the most prudent summary from the data is that there is “no clinically meaningful difference.”

The safety profile also supports equivalence. When adverse events (AE) were compared collectively and individually, the rates were similar for ABP 980 and reference trastuzumab over the course of neoadjuvant treatment. This included, respectively, any treatment-emergent AE (80.2%, 79.5%), any grade 3 or higher treatment-emergent AE (14.8%, 14.1%), and any treatment-emergent AE leading to discontinuation (0.8%, 0.6%).

There was a higher rate of serious treatment-emergent AEs in the ABP 980 group than in the reference trastuzumab group (4.9%, 1.4%), but there was no difference in the pattern of individuals AEs. The most common AEs for ABP 980 and reference trastuzumab, respectively, were peripheral neuropathy (13.7%, 11.9%), arthralgia (17.3%, 15.2%), asthenia (14.8%, 16.3%), alopecia (5.2%, 6.4%), nausea (6.3%, 5.3%), and bone pain (3.3%, 8.0%) The most common grade 3 or higher AE was neutropenia (8.6%, 8.0%). No other grade 3 or higher event occurred in more than 2% of patients. The proportion of patients with a 10% of greater decline in LVEF was uncommon and similar for ABP 980 and reference trastuzumab (0.3%, 0.9%, respectively).

The emergence of anti-drug antibodies was also uncommon in the 2 study arms, occurring in 0.6% of those randomized to ABP 980 and 1.4% of those randomized to reference trastuzumab.

A review of 11 trastuzumab studies published since 2007 in similar study populations produced a range in pCR rates from a low of approximately 30% to a high of nearly 60%, but most fell between 45% and 55%. The average was about 48%, which is consistent with the result of this study, according to Dr. Morales, who concluded that the data “add to the totality of evidence demonstrating similarity between ABP 980 and trastuzumab.”

ABP 980 is part of a larger biosimilar program being pursued by Amgen and Allergan. Earlier this year, the FDA ODAC unanimously recommended FDA approval of ABP 215, a biosimilar to bevacizumab (Avastin), which was later approved. Most of the other monoclonal antibody biosimilars in development are for indications outside oncology, but a biosimilar to rituximab (Rituxan), ABP 798, is being developed for treatment of both hematological malignancies and non-oncologic applications.

“Biosimilars are going to be a tremendous option for patients who have limited access to biologics,” said Dr. Markus, who emphasized that Amgen has devoted a “significant financial investment” in bringing these products forward.

Dr. Markus also emphasized that “no two biosimilars are the same,” and as a result, “Stakeholders need to consider the reliability of the biosimilar as well as the price.” Both the FDA and the European Medicines Agency have established criteria for the designation of a biosimilar, but he emphasized that “Biologics are complex drugs. We think that consistent quality in the manufacturing network will also be an important variable for clinical benefit,” Dr. Markus said.