Safety and Efficacy of Emicizumab in Persons with Hemophilia a with or without FVIII Inhibitors: Pooled Data from Four Phase III Studies (HAVEN 1–4)

Introduction: Emicizumab—a subcutaneously administered, bispecific, humanized, monoclonal antibody—promotes effective hemostasis in people with hemophilia A (PwHA). The primary efficacy and safety of emicizumab were reported previously, but long-term data are limited. Here, data from a wide age-range of PwHA with/without factor (F)VIII inhibitors enrolled in the Phase III HAVEN 1 (NCT02622321), HAVEN 2 (NCT02795767), HAVEN 3 (NCT02847637), and HAVEN 4 (NCT03020160) studies are pooled to establish the durable efficacy and safety of emicizumab.

Methods: The studies enrolled pediatric and adult PwHA with/without FVIII inhibitors. Participants received emicizumab prophylaxis 1.5 mg/kg weekly, 3 mg/kg every 2 weeks, or 6 mg/kg every 4 weeks. All participants assigned to receive emicizumab (including those assigned to control arms who later switched) are included in this analysis.

Participants and/or caregivers recorded outcomes of bleeding events via the Bleed and Medication Questionnaire (BMQ). Data from HAVEN 1–4 were pooled for an aggregate analysis of emicizumab efficacy and safety. Efficacy endpoints include calculated mean annualized bleed rates (ABRs; discrete, consecutive 24-week treatment intervals), model-based ABRs (calculated via negative binomial regression for full study period), percentage of participants with zero and 1–3 treated bleeds, and annualized cumulative dose of coagulation factor (ACD). Safety endpoints include incidence of adverse events (AEs) and AEs of special interest.

Results: Overall, 400 PwHA in HAVEN 1, 2, 3 and 4 (n=113, 88, 151, and 48, respectively) are included in the efficacy analysis for a total of 970.3 patient years (cutoff: 15 May 2020). The safety population comprises 399 PwHA who received ≥1 dose of emicizumab (1 PwHA was randomized to receive emicizumab but did not start treatment). The median age at baseline was 28.5 (range 1–77) years. The majority of participants were White (66.8%) or Asian (18.8%); 52.3% had FVIII inhibitors. In the 24 weeks prior to study entry, 60.9% of participants had target joints. The median duration of efficacy period was 120.4 (interquartile range 89.0–164.4) weeks; 85.0% of participants had an efficacy period of ≥74 weeks; 11 participants (2.8%) discontinued study treatment. Across all 4 studies, 90.9–94.8% of the observation period was covered by completed BMQs.

Across all studies, the model-based treated bleed ABR was 1.4 (95% confidence interval 1.1–1.7); treated bleed ABRs remained low throughout, and were seen to decrease with successive 24-week treatment intervals (Table 1). During Weeks 121–144 (n=170), 82.4% of participants had zero treated bleeds, and 15.3% of participants had 1–3 treated bleeds. During the same period, 91.8% and 90.0% had zero treated spontaneous/joint bleeds respectively (Figure 1). The proportion of participants with target joints reduced from 60.9% prior to study entry to 4.6% at Weeks 1–24, then <1.5% in all subsequent treatment intervals.

ACD of FVIII (Table 2), activated prothrombin complex concentrate (aPCC) and activated recombinant FVII (rFVIIa, Table 3) generally decreased across each 24-week treatment interval.

Emicizumab was well tolerated (Table 4), and no participants discontinued due to AEs beyond the five previously described (Oldenburg et al. N Engl J Med 2017; Young et al. Blood 2019; Mahlangu et al. N Engl J Med 2018; Pipe et al. Lancet Haem 2019). At data cut, 1 fatality, 3 thrombotic microangiopathies (TMAs), and 4 thromboembolic events (TEs) have been reported; all but 1 occurred in HAVEN 1. All TMAs and 2 of 4 TEs were associated with concomitant aPCC use. The percentage of participants with ≥1 drug-related AE in Weeks 1–24, 25–48 and 49–72 were 28.8%, 6.8%, and 3.0% respectively; over the same intervals, injection site reactions were observed in 23.3%, 4.8%, and 2.5% of participants.

Conclusions: With nearly 3 years of follow-up, emicizumab maintained low bleed rates in PwHA of all ages, with/without FVIII inhibitors. ABRs continued to decrease and the proportion of participants with zero treated bleeds increased with each consecutive 24-week period; the trend was the same for the proportion of participants with zero joint bleeds and almost all target joints resolved. The ACDs of FVIII, aPCC, and rFVIIa decreased with successive treatment intervals. Emicizumab remains well tolerated over long-term follow-up, and no new safety concerns have been identified to date.