Immunogenicity of Emicizumab in People with Hemophilia A (PwHA): Results from the HAVEN 1-4 Studies

Introduction

Emicizumab is a bispecific humanized monoclonal antibody given subcutaneously, which bridges FIXa and FX to restore the function of missing FVIIIa in PwHA. It is approved for routine prophylaxis in PwHA with inhibitors of all ages, and its efficacy and safety in PwHA without inhibitors was recently described (Mahlangu et al. N Engl J Med 2018, in press; Pipe et al. WFH 2018). Anti-drug antibody (ADA) formation is commonly observed with biologics (e.g. monoclonal antibodies). Presence of ADAs may alter the pharmacokinetics (PK) and pharmacodynamics (PD), and can affect efficacy and safety. Therefore, characterization of the rate and clinical significance of anti-emicizumab antibodies is of importance. Four patients (4/18) tested positive for ADAs in the phase I/II study of emicizumab (Shima et al. Blood Adv 2017). We optimized the immunoassay sensitivity and report here the updated analysis of the immunogenicity of emicizumab in four phase III studies: HAVEN 1, 2, 3 and 4.

Methods

All PwHA enrolled in HAVEN 1 (NCT02622321), HAVEN 2 (NCT02795767), HAVEN 3 (NCT02847637) or HAVEN 4 (NCT03020160) who had at least one ADA measurement after exposure to emicizumab were included in the analysis. Blood samples for detection of anti-emicizumab antibodies were collected at baseline and at regular intervals (initially monthly and later every 12 weeks) during emicizumab treatment or 24 weeks after the last dose for discontinuations. Samples were screened for the presence of ADAs with a sandwich enzyme-linked immunosorbent assay (ELISA) method. Positive samples underwent confirmatory analysis in the presence of unlabeled emicizumab to ensure binding specificity. The assay cut-point for ADA positivity was generated from all study patients' samples before emicizumab exposure (n=398). The confirmatory cut-point was generated from randomly selected baseline samples from these patients (n=100). ADA positivity was assessed according to the recommendations of the American Association of Pharmaceutical Scientists Therapeutic Protein Immunogenicity Focus Group (Shankar at al. AAPS J 2014). In absence of a sensitive neutralizing antibody assay, ADAs associated with consistent decline of emicizumab exposure were considered as having a neutralizing potential; associated loss of PD effect (e.g. decreased FVIII activity, decreased generation of thrombin and prolonged activated partial thromboplastin time) was reviewed as supportive information. The association between ADAs and clinical response or adverse events was also examined.

Results

Of 398 HAVEN patients included in this analysis, 14 (3.5%) tested positive for ADA (2/111, 4/88, 6/151 and 2/48 in HAVEN 1, 2, 3 and 4, respectively). For 7 patients, ADAs were transient, i.e. detected on a single occasion. Three patients (n=1 HAVEN 1; n=2 HAVEN 2), who tested positive for ADA and displayed decline of PK and reduced PD effect, were classified as having ADAs with neutralizing potential, i.e. 0.75% of the overall population. Of those, 1 patient discontinued emicizumab due to a loss of efficacy and resumed pre-study treatment without complication. The presence of ADAs without neutralizing potential was not associated with decreased efficacy. No cases of anaphylaxis or hypersensitivity were reported in ADA-positive patients. Patients who tested positive for ADA did not have a safety profile that differed from that of the overall population. Specifically, there was no trend towards increased frequency or severity of injection site reactions after patients tested positive for ADAs.

Conclusions

Emicizumab treatment is associated with a low rate of patients who tested positive for ADA (3.5%), similar to other humanized monoclonal antibodies. ADAs with a neutralizing potential were observed in <1% of patients. Importantly, the presence of ADAs was not associated with impact on safety. Based on this analysis, the rate of ADA with risk of clinical consequences in patients treated with emicizumab is very low, suggesting that routine monitoring is not warranted.