As discussed above, analytical (structural and functional) characterization is the first and foundational step in the determination of biosimilarity and begins with the identification of critical quality attributes that may influence the PK, efficacy, safety, and immunogenicity of the proposed biosimilar [5, 8]. A comprehensive analytical characterization of ABP 980 was conducted using state-of-the-art techniques to identify any potential differences that may exist in structure and function between ABP 980 and trastuzumab RP , and demonstrated high analytical similarity of the two in physicochemical characteristics (primary structure/amino acid sequence, post-translational modifications, higher-order structure, protein content, and product-related impurities and substances) [9, 10]. ABP 980 has the same amino acid sequence and molecular mass (consistent with theoretical mass), and the same N-linked glycosylation at N300 in the same glycopeptide, as trastuzumab RP. No new peaks or missing peaks were noted in the chromatograms or peptide maps of ABP 980 versus trastuzumab RP. A comparison of the glycan profiles of ABP 980 and trastuzumab RP indicated that while the levels of specific species of glycans (e.g., A1G0F) in each may differ, the overall levels of key glycan groups (e.g., afucosylated, galactosylated, sialylated, and high mannose species) that have the ability to influence binding to the FcγRIIIa receptor and affect antibody-dependent cellular toxicity (ADCC) were similar in both. Moreover, the predicted NK92 ADCC activity for any given glycan combination was found to be similar for ABP 980 and trastuzumab RP. There were no new peaks in the ABP 980 glycan map compared with that of trastuzumab RP.
Likewise, functional similarity evaluations of ABP 980 and trastuzumab RP assessed Fab-mediated, Fc-mediated, and combined Fab- and Fc-mediated activities. Since the biological activity of trastuzumab is primarily mediated via binding to the extracellular domain of HER2 and subsequent internalization and activation of downstream ligand-independent signaling events, evaluation of these parameters showed that ABP 980 exhibited similar binding to the extracellular domain of HER2, similar potency, and similar antibody-dependent cellular cytotoxicity activities in experimental cell-based systems [9,10,11].
ABP 980 Mechanism of Action Animation (MPG 2115141kb)
ABP 980 and trastuzumab RP exhibited similar on-rate, off-rate, and relative affinity for the extracellular domain of the HER2 receptor as measured with surface plasmon resonance, with all values for ABP 980 being contained within the ranges of the RP. A competitive cell binding assay using SK-BR-3 breast cancer cells was conducted comparing ABP 980 and trastuzumab RP which showed that the mean relative HER2 cell binding of all lots and products tested was approximately 100%, confirming similar binding activity of the two. Post-binding HER2 internalization documented by flow cytometric analysis of HER2-expressing SK-BR-3 cells treated with ABP 980 and trastuzumab RP confirmed similar levels of internalization of the two antibodies. Finally, similar inhibition of proliferation with ABP 980 and RP in NCI-N87 gastric cancer cells expressing high HER2 levels and similar lack of inhibition of proliferation in MCF7 breast cancer cells expressing low levels of HER2 confirmed biosimilarity of ABP 980 with trastuzumab RP as well as ABP 980 activity in gastric cancer cells and specificity of ABP 980 for HER2-overexpressing cells. Further, similar synergistic activity in the presence of the chemotherapeutic agent docetaxel was demonstrated in combination studies using gastric cancer cells in vitro . Nonclinical pharmacology evaluations have established that ABP 980 and trastuzumab RP have similar dose-dependent antitumor activity in tumor xenograft models (breast or gastric cancer cells) that overexpressed HER2, and multiple-dose toxicology studies in cynomolgus monkeys showed similar in vivo drug exposure and dose accumulation [11, 13].
The clinical pharmacology of ABP 980 was evaluated in a randomized, single-blind, single-dose PK equivalence study in healthy subjects. This study confirmed the PK similarity between ABP 980 and trastuzumab RP (Table 2) . Additionally, the percentage of treatment-related adverse events (AEs) of all grades was similar in both groups, and immunogenicity evaluations revealed that no patients developed antidrug antibodies during the study.
Finally, clinical safety and efficacy of ABP 980 were demonstrated in a randomized, multicenter, double-blind, comparative clinical study (LILAC) conducted in patients with early breast cancer (EBC) [15, 16]. This patient population is considered to be a sensitive population in which to evaluate a trastuzumab biosimilar because of its homogeneity and minimal confounding factors such as prior treatments that may confound immunogenicity evaluation, line of therapy, disease burden, comorbidities, and risk of secondary cancers [17,18,19]. These characteristics make it more suitable for an equivalence study between two similar products such as a biosimilar and its RP.
A total of 725 patients were randomized in the LILAC study to receive ABP 980 (n = 364) or trastuzumab RP (n = 361) plus paclitaxel (175 mg/m2 every 3 weeks or 80 mg/m2 once weekly for 12 cycles if that was the standard of care in that region) followed by surgery. The co-primary efficacy endpoints in this study were risk difference (RD) and risk ratio (RR) of pathological complete response (pCR) in breast tissue and axillary lymph nodes assessed at a local laboratory [15, 16]. A central laboratory performed a sensitivity analysis of tumor samples in order to reduce inter-pathologist variability in response determination. Based on local evaluation, pCR was achieved in 172/358 assessable patients (48%, 90% CI 43–53%) in the ABP 980 group and 137/338 patients (41%, 90% CI 35–46%) in the trastuzumab RP group (RD 7.3%, 90% CI 1.2–13.4; RR 1.188, 90% CI 1.033–1.366); the upper bounds of CIs exceeded predefined equivalence margins of 13% for RD and of 1.318 for RR (Fig. 2). The pCR was noted in 162/339 (48%) patients in the ABP 980 group at baseline and 138/330 (42%) patients in the trastuzumab RP group at baseline (RD 5.8%, 90% CI – 0.5 to 12.0, and RR 1.142, 90% CI 0.993–1.312) in central lab assessment. Based on these sensitivity analyses, similar efficacy of ABP 980 and trastuzumab RP was confirmed, as the RR and RD estimates for the two groups were contained within the predefined equivalence margins of the study.
In the LILAC study, after neoadjuvant treatment and surgery, patients received adjuvant treatment with ABP 980 or RP every 3 weeks for up to 1 year from start of study treatment. During the adjuvant phase of the study, patients who had previously received ABP 980 continued with ABP 980 (6 mg/kg); patients who had previously received trastuzumab RP either continued to receive trastuzumab RP or were switched to receive ABP 980. The overall type, frequency, and severity of AEs and AEs of interest were similar between the ABP 980 and trastuzumab RP during both the neoadjuvant and adjuvant phases of the clinical study (Fig. 3) . There was no difference in the incidence of AEs among patients who were switched from the RP to ABP 980 and those who continued to receive trastuzumab RP in the adjuvant setting. Median left ventricular ejection fraction (LVEF), a measure of cardiotoxicity, did not change in any treatment group over the course of the study, and switching had no impact on LVEF (Fig. 4). The overall frequency of cardiac disorders was low throughout the study [13, 15, 20]. The frequency of immunogenicity was found to be low and similar between groups, and no patient tested positive for neutralizing antibodies .
Based on the sum of these results, the TOE for ABP 980 confirms that it is highly similar to trastuzumab RP, thereby supporting extrapolation to all approved indications of the RP .