Search

8044 Background: ABP 798* is being developed as a biosimilar to rituximab, a CD20-directed cytolytic antibody. A randomized, double-blind, active-controlled study compared the efficacy, safety, and immunogenicity of ABP 798 with rituximab reference product (RP) in subjects with CD20-positive NHL; results of the final analysis are presented here. Methods: Adult subjects with grade 1, 2, or 3a follicular B-cell NHL and low tumor burden were randomized to receive intravenous ABP 798 or RP (375 mg/m2) once weekly for 4 weeks, then at weeks 12 and 20. The primary endpoint was risk difference (RD) of overall response rate (ORR) by week 28. Secondary endpoints included RD of ORR at week 12, pharmacokinetics, pharmacodynamics, safety, and immunogenicity. Results: 254/256 randomized subjects were treated with at least one infusion of ABP 798 (n = 128) or RP (n = 126); ORR by week 28, based on independent central blinded assessment of the modified full analysis set, was comparable between the ABP 798 and RP groups (78% vs. 70%, respectively). The 2-sided 90% confidence interval of RD of ORR (-1.4%; 16.8%) was within the pre-specified margin (-15%; 35.5%) thereby establishing clinical equivalence between ABP 798 and RP. This result was supported by analyses of the secondary efficacy endpoint of RD of ORR at week 12. In the two groups, the geometric least squares means for serum concentrations over time (e.g., week 12, pre-dose: ABP 798, 21.89 vs. RP, 20.57; week 12 post-dose: ABP 798, 201.30 vs. RP, 203.52) and the extent of B-cell depletion from day 1 to day 8 (ABP 798, 98.3% vs. RP, 98.3%) were similar. Frequency, type, and severity of adverse events (AEs) were comparable between ABP 798 and RP groups; grade ≥3 AEs were reported in 10.9% and 10.3% of subjects and serious AEs in 3.9% and 4.0%, respectively. Most common AEs were headache, fatigue, and nausea; the most common AE of interest was infusion reactions. No new or unexpected safety signals were observed. Binding and neutralizing anti-drug antibodies were comparable between groups. Conclusions: Results of this study demonstrated clinical similarity between ABP 798 and rituximab RP in subjects with CD20-positive NHL. *At the time of this submission, ABP 798 had not been approved by the FDA or any relevant regulatory agency and the indications are yet undetermined. Please consult ABP 798’s later approved label in the relevant country for information regarding the approved uses for ABP 798.

e20708 Background: ABP 215 (MVASI™ [bevacizumab]) is the first biosimilar to Avastin (bevacizumab) approved in the US and EU. This phase 3, double-blind study compared efficacy of ABP 215 with bevacizumab reference product (RP) in patients with advanced non-squamous NSCLC. Here we report the totality of evidence supporting similarity between ABP 215 and RP, including preclinical VEGF-A isoform binding relevant to the mechanism of action across indications, and clinical results of central and investigator evaluation of tumor response. Methods: VEGF-A kinetic parameters were compared for common isoforms 121, 165 and 189. Patients were randomized 1:1 to ABP 215 or RP 15 mg/kg Q3Wx6 cycles. All patients received carboplatin and paclitaxel Q3W for up to 6 cycles. Disease assessments (CT/MRI) were performed at screening, weeks 7, 13, 19, and Q9W thereafter by the investigator and, independently, by central, blinded radiologists. Efficacy was based on objective tumor assessments according to RECIST 1.1. Copies of all radiographs were submitted for central analysis. The primary efficacy endpoint was risk ratio of objective response rate (ORR); clinical equivalence was confirmed if the 2-sided 90% CI of the risk ratio was within the margin of 0.67-1.5. Additional efficacy analyses included duration of response (DOR) and progression-free survival (PFS). Results: Binding to all 3 isoforms was similar for ABP 215 and RP. In the MAPLE study, 642 patients (ABP 215, 328; RP, 314) were randomized. Based on central analysis, ORR was achieved in 128 (39.0%) patients in the ABP 215 and 131 (41.7%) in the RP groups, (ORR risk ratio: 0.93 [90% CI: 0.80, 1.09]). Based on investigator analysis, ORR was achieved in 157 (47.9%) patients in the ABP 215 and 151 (48.1%) in the RP groups (ORR risk ratio: 1.01 [90% CI: 0.88, 1.16]). Hazard ratio (HR, ABP 215 vs RP) for DOR was 1.08 (95% CI, 0.76, 1.54) and 0.76 (95% CI, 0.48, 1.23) per investigator and central assessment. PFS HR was 1.10 (90% CI, 0.92, 1.33) and 1.03 (90% CI, 0.83, 1.29) per investigator and central assessment. Conclusions: These results further confirm the similarity of ABP 215 and RP and support extrapolation to all available bevacizumab indications. Clinical trial information: NCT01966003.

557 Background: Although trastuzumab is generally well-tolerated, cardiotoxicity is the main limitation in its use, leading to a severe heart failure in 2-4% of patients in adjuvant trials. In the phase 3 LILAC trial, trastuzumab biosimilar ABP 980 demonstrated similar efficacy, safety, and immunogenicity to trastuzumab reference product (RP) in women with HER2-positive early breast cancer. Here we report analyses comparing cardiac safety of ABP 980 and RP. Methods: In the neoadjuvant phase, all 725 patients received 4 cycles of chemotherapy with epirubicin + cyclophosphamide Q3W and were then randomized 1:1 to ABP 980 (n = 364) or RP (n = 361) with a loading dose of 8 mg/kg, then 6 mg/kg IV Q3W for 3 cycles plus paclitaxel Q3W (4 cycles) or QW (12 cycles). After surgery, patients received investigational product (IP) Q3W for up to 1 year; ABP 980-treated patients continued ABP 980, and RP-treated patients either continued RP (n = 190) or switched to ABP 980 (RP/ABP 980; n = 171). AEs were assessed every 3 weeks and cardiac safety every 3 months. Cardiac safety was monitored by computerized 12-lead ECG; LVEF was assessed by 2D ECHO. LVEF decline was defined as LVEF value decrease from study baseline by ≥10 percentage points and to < 50%. Results: Treatment groups were well balanced with regard to IP disposition. Over the entire study, 22 (3.1%) patients had LVEF decline by ≥10 percentage points compared to baseline and to < 50%; no meaningful between-group differences were observed (ABP 980:10/359 [2.8%], RP: 6/184 [3.3%], RP/ABP 980: 6/171 [3.5%]). The incidence of cardiac AEs was low and comparable in treatment groups. One grade 3 cardiac failure event was reported in the RP/ABP 980 arm; another in the RP arm was coincident with LVEF decline. No patient discontinued IP during adjuvant phase because of cardiac failure. Conclusions: These pre-specified analyses confirm the tolerability of ABP 980 and demonstrate clinical similarity of ABP 980 and RP with respect to cardiac safety. The incidence of LVEF decline was consistent with the known cardiac safety profile of the RP. No new cardiac safety signals were observed whether patients were on ABP 980 or switched from RP to ABP 980. Clinical trial information: NCT01901146.

583Background: The phase 3 LILAC Study compared ABP 980 with trastuzumab (TRAS) on pathologic complete response (pCR) in women with HER2+ early breast cancer. The primary efficacy results, based on...

9050 Background: Approval of biosimilars is based on totality of evidence (TOE) that includes demonstration of analytical, functional, and pharmacokinetic (PK) similarity, and comparable clinical efficacy, safety, and immunogenicity. The TOE supports scientific justification for extrapolation across other indications that a share common mechanism of action (MOA). The proposed biosimilar ABP 215 is similar to bevacizumab (BEV) with respect to inhibition of vascular endothelial growth factor (VEGF)-induced signaling; PK in healthy adults; and clinical efficacy, safety, and immunogenicity in patients with non-small cell lung cancer (NSCLC). The PK results from the NSCLC study are reported here. Methods: In this double-blind study, adults with non-squamous NSCLC receiving first-line chemotherapy with carboplatin and paclitaxel were randomized 1:1 to ABP 215 or BEV (15 mg/kg IV Q3W for 6 cycles). Trough (pre-dose) PK samples were collected at baseline, at weeks 4, 7, and 13, and at end of study. Clinical evaluations included efficacy, safety, and immunogenicity. Results: 642 patients (ABP 215, n = 328; BEV, n = 314) were randomized. Demographic and baseline characteristics were well balanced. The primary endpoint of risk ratio for objective response rate was 0.93 (2-sided 90% CI, 0.80-1.09), which was contained within the pre-specified equivalence margin. The incidence of adverse events was comparable between ABP 215 and BEV. During the study, 4 (1.4%) and 7 (2.5%) patients in the ABP 215 and BEV group, respectively, developed binding antibodies; no patient tested positive for neutralizing antibodies. Steady-state trough serum concentrations for both ABP 215 and BEV were reached by week 13 (median 124.9 vs 124.4 µg/mL, respectively) and were consistent between ABP 215 and BEV throughout the study. Conclusions: ABP 215 is similar to BEV in multiple analytical, functional, and clinical assessments. PK results in patients presented here provide further evidence for similarity between ABP 215 and BEV and contribute to the TOE supporting extrapolation of clinical similarity across BEV indications based on its common MOA. Clinical trial information: NCT01966003.

9095Background: ABP 215 is a biosimilar candidate highly similar to BEV, a VEGF inhibitor, in analytical and functional comparisons. Pharmacokinetic similarity between ABP 215 and BEV has been demo...