Introduction HAVEN 3 was a phase 3 study investigating the use of emicizumab as prophylaxis in adult and adolescent (≥12 years old) persons with hemophilia A (PwHA) without factor VIII (FVIII) inhibitors (NCT02847637; Mahlangu et al. 2018). HAVEN 3 demonstrated that emicizumab prophylaxis once weekly or every two weeks was safe and highly effective in bleed prevention. The primary analysis of HAVEN 3 included an intrapatient comparison of 48 participants who received FVIII prophylaxis in a non-interventional study (NIS) prior to enrollment in HAVEN 3. Compared with emicizumab prophylaxis during the HAVEN 3 study, emicizumab prophylaxis resulted in an annualized bleed rate that was 68% lower than the rate with previous FVIII prophylaxis (1.5 vs 4.8, p<0.001). No dosing guidance was provided regarding the use of on-demand FVIII in HAVEN 3, and investigators prescribed FVIII at their own discretion. In this subsequent analysis, we characterize the dose and frequency of replacement FVIII used for the treatment of breakthrough bleeding in these 48 participants. Methods The primary comparisons in our analyses are focused on on-demand FVIII use for breakthrough bleeding while participants were on FVIII prophylaxis during the NIS versus its use while on emicizumab prophylaxis during HAVEN 3. Any use of on demand FVIII other than to manage breakthrough bleeding (e.g. prior to activity) was not included in our analyses. Given that, collectively, the total exposure time to emicizumab during HAVEN 3 was more than twice the exposure time to FVIII prophylaxis during the NIS (75.8 vs 28.6 years respectively), any treatment comparisons are drawn on an annualized basis. Annualized on-demand FVIII use was calculated by dividing by the number of days in the efficacy period and multiplying the resulting daily consumption by 365.25 days. The number of infusions and cumulative doses of on-demand FVIII use are described at the participant level as well as at the individual bleed level and are presented descriptively for both the NIS and HAVEN 3 exposure periods. No formal statistical inferences (i.e. calculation of p-values) have been conducted. All analyses were based on an October 2018 data cutoff. Results A total of 48 participants who were treated with FVIII prophylaxis during the NIS were then treated with emicizumab prophylaxis during HAVEN 3 and thus make up the total cohort for our analyses. Annualized infusion rates of on-demand FVIII per participant and cumulative doses of on-demand FVIII (in international units [IU] per kilogram) per participant were higher during the FVIII prophylaxis period when compared with the emicizumab exposure period (mean 15.3 vs 7.2; median 3.6 vs 0.6 annual infusions per participant and mean 602.4 IU/kg vs 209.0 IU/kg; median 75.5 IU/kg vs 19.1 IU/kg, respectively). At the individual bleed level, FVIII infusions and total cumulative dose suggested that participants were administered a similar amount of medication to treat bleeds during both the NIS and HAVEN 3 study periods: median number of infusions per bleed were 1.0 (interquartile range [IQR]=1.0) versus 2.0 (IQR=3.0) and median cumulative doses were 43.5 (IQR=35.1) versus 50.0 (IQR=72.7) IU/kg, respectively (Table 1). Conclusions This analysis revealed a lower annualized infusion rate and a correspondingly lower annualized cumulative dose of FVIII for treatment of breakthrough bleeds during emicizumab prophylaxis compared with FVIII prophylaxis. At the individual bleed level, the amount of on-demand FVIII used per bleeding episode was comparable between NIS and HAVEN 3 exposure periods. Thus, based on this single analysis, it appears that patients received less on-demand FVIII during emicizumab prophylaxis compared with FVIII prophylaxis, as a result of overall reduction of bleed frequency, while the treatment of individual bleeds appeared similar regardless of the prophylaxis therapy administered. Disclosures Callaghan: Octapharma: Consultancy; Novonordisk: Consultancy, Speakers Bureau; Global Blood Therapeutics: Consultancy; Sanofi: Consultancy; Takeda: Consultancy, Research Funding; Bayer: Consultancy, Speakers Bureau; Alnylum: Equity Ownership; Biomarin, Bioverativ, Grifols, Kedrion, Pfizer, Roche/Genentech, Shire, and Spark Therapeutics: Consultancy; Roche/Genentech: Speakers Bureau; Shire/Takeda: Speakers Bureau; Pfizer: Research Funding; Roche: Research Funding. Trzaskoma:Genentech: Employment, Equity Ownership. Ko:Genentech, Inc.: Employment. Lee:Genentech, Inc.: Employment. Patel:Genentech: Employment; Roche/Genentech: Equity Ownership. Tzeng:Genentech, Inc.: Employment. Shah:Genentech: Employment. Chang:Genentech, Inc.: Employment; Genentech/Roche: Equity Ownership. Niggli:F. Hoffmann-La Roche Ltd: Employment. Dhalluin:F. Hoffmann-La Roche Ltd: Employment. Mahlangu:Sanofi: Research Funding; Pfizer: Research Funding; Novartis: Research Funding; Biomarin: Research Funding; Spark: Consultancy, Honoraria, Research Funding, Speakers Bureau; Roche: Consultancy, Honoraria, Research Funding, Speakers Bureau; Novo Nordisk: Consultancy, Honoraria, Research Funding, Speakers Bureau; Catalyst Biosciences: Consultancy, Honoraria, Research Funding, Speakers Bureau; CSL Behring: Consultancy, Honoraria, Research Funding, Speakers Bureau; Baxalta: Consultancy, Honoraria, Research Funding, Speakers Bureau; Unique: Research Funding.