• Overall, patients with HA in Australia had a reduced rate of bleeding after changing to emicizumab for prophylaxis.

  • However, bleeding rate remained unchanged in certain subgroups including adults and those on regular prophylaxis before emicizumab.

Abstract

Emicizumab became routinely available in Australia in November 2020 as regular prophylaxis for certain patients with hemophilia A (HA). We performed an intraindividual comparison of bleeding outcomes in Australian patients with HA before and after commencement of emicizumab. Data regarding demographics, severity, treatment, inhibitors, and number and type of intraindividual treated bleeds before and after starting emicizumab in patients with HA were extracted from the Australian Bleeding Disorders Registry. As of April 2022, there were 459 eligible patients with HA on emicizumab in Australia, 397 of 459 (86%) of whom had severe disease. Overall, 59 of 459 (13%) had a current inhibitor. Adults (aged ≥18 years) composed 49% (223/459) of the population. The proportion of patients with zero bleeds increased from 54% to 63% after commencement of emicizumab (relative risk [RR], 1.24; 95% confidence interval [CI], 1.09-1.41; P < .01). RR for zero treated bleeds after commencement was significant in subgroups including pediatric patients (RR, 1.34; 95% CI, 1.13-1.59; P < .01) and those not on regular prophylaxis prior (RR, 1.75; 95% CI, 1.22-2.52; P < .01). There was no significant difference in zero-bleed prevalence in the adult, standard half-life, and extended half-life subgroups. Spontaneous bleeding was reduced (RR, 1.69; 95% CI, 1.34-2.13; P < .01), whereas provoked bleeding was not (P = .15). Real-world data from Australia shows a reduction in bleeding events with emicizumab prophylaxis for the overall population of patients with HA, although not in all subgroups. This reduction appears to be most pronounced in spontaneous bleeds within the pediatric population, and in those on on-demand therapy before switching.

Hemophilia A (HA) is an X-linked bleeding disorder characterized by recurrent bleeding events due to a deficiency of factor VIII (FVIII).1 Severity is defined by the degree of FVIII deficiency (mild, >5% and <40% of normal FVIII activity; moderate, 1%-5%; and severe, <1%).2 The prevalence of HA in Australia is 1 in 6000 to 10 000 males.3 In Australia, HA has predominantly been treated with recombinant FVIII (rFVIII) infusions, typically as prophylaxis among patients with moderate to severe disease.3 A costly complication of this treatment in ∼30% of patients with HA is the development of alloantibodies (inhibitors) directed toward the recombinant product,4 with resulting reduced quality of life and increased morbidity because of bleeds with a reduced response to rFVIII.5-7 

Emicizumab (Hemlibra; Chugai Pharmaceutical Co, Ltd, Tokyo, Japan) is a recombinant, humanized, bispecific monoclonal antibody, which mimics the activity of FVIII. Emicizumab allows the necessary interaction between activated FIX and FX to occur in the absence of endogenous FVIII and irrespective of the presence of FVIII inhibitors.8,9 It is administered as a subcutaneous injection on a weekly, 2-weekly or 4-weekly schedule.

The HAVEN-1 and -2 trials demonstrated efficacy of emicizumab in the prophylactic setting for adult (aged ≥18 years) and pediatric (aged <18 years) patients with HA with inhibitors,10,11 providing a treatment avenue to a cohort that historically have had limited and expensive prophylaxis options.12 A subsequent trial (HAVEN-3)13 demonstrated efficacy of emicizumab prophylaxis in adult and adolescent patients with HA without inhibitors.14 Although this data confirms the efficacy of emicizumab in the context of a clinical trial for adults and children alike, current published literature reporting real-world efficacy of this therapy is from limited patient populations. Real-world data are critical in confirming the generalizability of trial data to routine clinical practice. The limited real-world data available correlate with trial data with respect to low bleeding rates.15-18 However, studies with large numbers and intraindividual comparisons of bleeding events before commencement vs after commencement of emicizumab are lacking. In particular, robust comparison of emicizumab with FVIII prophylaxis has not been well described in large cohorts. This study is therefore critical in accurately evaluating the efficacy and safety of this therapy in the real-world setting, particularly in comparison with the long-standing gold standard treatment of prophylactic factor replacement.

Objectives

Funded emicizumab was made available to eligible patients with severe or moderate HA without inhibitors and all patients with HA with inhibitors in Australia through the governance of hemophilia treatment centers (HTCs) in November 2020.19 Data on bleeding events and factor usage are recorded in the Australian Bleeding Disorder Registry (ABDR).20 

We aimed to use the ABDR data to evaluate the rates of treated bleeding events in eligible Australian patients with HA (both those with and those without current inhibitors) who have commenced emicizumab prophylaxis. Comparison between annualized bleeding rates (ABR) and zero-bleed prevalence before and after commencement of emicizumab was performed, including subgroup analysis in those with and without current inhibitors, and according to age group, bleed site, bleed cause, and prior prophylaxis agent.

We conducted a multicenter retrospective cohort study of all Australian patients with HA on emicizumab registered in the ABDR as of 6 April 2022, with a minimum of 6 months post–emicizumab commencement data collected for all patients analyzed. The ABDR contains data on registered patients with a bleeding disorder from all HTCs around Australia.20 Research projects using the ABDR are conducted with human research ethics committee approval (project no. 252/15) and in accordance with local state/territory compliance requirements. The following nonidentifiable data points were extracted from the ABDR on consented patients: demographic data collected included age, weight, severity of disease, and inhibitor status; data on emicizumab included dose (in mg/kg), dosing frequency, and date of drug commencement.

Recorded bleeds were a combination of patient-reported (recorded within the myABDR section of the ABDR) and HTC-reported treated events. Bleeding data included only treated bleeds because only treated events are recorded in the ABDR. Bleeds recorded within 7 days after the start date of emicizumab therapy were recorded as occurring before commencement of emicizumab. Hospitalizations for a bleed included emergency department presentation or inpatient admission. ABR is a measure of the number of bleeds over 1 year. The precommencement ABR was an exact measure whereas the postcommencement ABR was derived using the duration on emicizumab divided by 365 days, multiplied by the total number of bleeds since commencement of emicizumab, given that the duration on emicizumab varied between patients. Our focus in this work, however, was on the proportions of patients with zero treated bleeds before and after commencement of emicizumab, given that we believe this offers the most accurate representation of emicizumab effect on bleeding risk with the data available, for the following reasons: the data is nonparametric and therefore, not accurately described using mean ABR and standard deviation. Additionally, the data produced low median ABRs (ie, overall median ABR of 0 both before commencement and after commencement of emicizumab) given the high frequency of zero bleeds both before and after commencement; such ABRs fail to usefully quantify the true effect of emicizumab.

The median and interquartile range (IQR) of ABR between precommencement and postcommencement groups were compared using Wilcoxon signed rank test. Risk ratios were derived from the McNemar test to compare proportions of patients with zero treated bleeds before and after commencement (given use of paired nominal data). Two-tailed P values < .05 were considered significant. Statistical calculations were performed using Stata version 12.1 (StataCorp LLC, College Station, TX). Figures were created with Microsoft Excel version 16.77.1.

Common terms used in the description of HA and its management are provided in Table 1.

Table 1.

Common terms and their definitions used in the description and management of HA

TermDefinition
Prophylaxis Factor administration in the absence of bleeding 
On-demand Factor given as treatment for an acute bleed or before surgery 
Immune-tolerance induction Frequent and regular administration of rFVIII to eradicate an inhibitor 
SHL rFVIII rFVIII products with a half-life of 8-12 h requiring dosing 3-4 times per week 
EHL rFVIII rFVIII products with a half-life approaching 16 h requiring dosing twice per week 
Inhibitor Alloantibody directed toward rFVIII product 
High-responding inhibitor Inhibitor with a level of ≥5 BU 
Low-responding inhibitor Inhibitor with a level of <5 BU 
Inhibitor status Absence of inhibitor, presence of a low-responding inhibitor, or presence of a high-responding inhibitor 
Pediatric Participant aged <18 y 
Adult Participant aged ≥18 y 
Provoked bleed A bleed that does not occur spontaneously, that is, typically caused by trauma or surgery 
Spontaneous bleed A bleed that occurs without a provoking factor 
Funded emicizumab Emicizumab made available to eligible patients with severe or moderate HA without inhibitors, and all patients with HA with inhibitors in Australia through the governance of HTCs in November 2020 
Trial emicizumab Emicizumab administered as part of a research trial including the HAVEN trials 
TermDefinition
Prophylaxis Factor administration in the absence of bleeding 
On-demand Factor given as treatment for an acute bleed or before surgery 
Immune-tolerance induction Frequent and regular administration of rFVIII to eradicate an inhibitor 
SHL rFVIII rFVIII products with a half-life of 8-12 h requiring dosing 3-4 times per week 
EHL rFVIII rFVIII products with a half-life approaching 16 h requiring dosing twice per week 
Inhibitor Alloantibody directed toward rFVIII product 
High-responding inhibitor Inhibitor with a level of ≥5 BU 
Low-responding inhibitor Inhibitor with a level of <5 BU 
Inhibitor status Absence of inhibitor, presence of a low-responding inhibitor, or presence of a high-responding inhibitor 
Pediatric Participant aged <18 y 
Adult Participant aged ≥18 y 
Provoked bleed A bleed that does not occur spontaneously, that is, typically caused by trauma or surgery 
Spontaneous bleed A bleed that occurs without a provoking factor 
Funded emicizumab Emicizumab made available to eligible patients with severe or moderate HA without inhibitors, and all patients with HA with inhibitors in Australia through the governance of HTCs in November 2020 
Trial emicizumab Emicizumab administered as part of a research trial including the HAVEN trials 

BU, Bethesda units.

There were 459 eligible Australian patients with HA on emicizumab prophylaxis recorded in the ABDR. Eligibility included having an indication for funded emicizumab as outlined in “Objectives.” Clinical characteristics of the total population and pediatric, adult, current inhibitor and no current inhibitor subgroups are described in Table 2. The majority of this cohort (397/459 [86%]) had severe disease. The 4 patients with mild HA had a current inhibitor. In total, 59 of 459 (13%) had a current inhibitor. Adults (aged ≥18 years) composed 49% (223/459) of the population.

Table 2.

Characteristics of Australian patients with HA on emicizumab included in analysis, split by age group and current inhibitor status

Patient characteristicsPediatric population (n = 236)Adult population (n = 223)Current inhibitor (n = 59)No current inhibitor (n = 400)Total population (N = 459)
Age (y), median (IQR) 9 (5-14) 36 (26-50) 20 (9-41) 17 (9-34) 17 (9-35) 
Pediatric (aged <18 y) NA NA 7.5 (5-10) 9.5 (5-14) 9 (5-14) 
Adult (aged ≥18 y) NA NA 38 (27-49) 36 (26-50) 36 (26-50) 
Diagnosis severity, n, (% of N)      
Severe 205 (87) 192 (86) 52 (88) 345 (86) 397 (86) 
Moderate 31 (13) 27 (12) 3 (5) 55 (14) 58 (13) 
Mild 0 (0) 4 (2) 4 (7) 0 (0) 4 (1) 
Inhibitor response, n, (% of N)      
No current inhibitor 210 (89) 190 (85) NA 400 (100) 400 (87) 
High responding 22 (9) 24 (11) 46 (78) NA 46 (10) 
Low responding 4 (2) 9 (4) 13 (22) NA 13 (3) 
Prior prophylaxis agent, n (% of N)      
SHL 52 (22) 110 (49) 16 (27) 146 (37) 162 (35) 
EHL 130 (55) 72 (32) 9 (15) 193 (48) 202 (44) 
On-demand 54 (23) 41 (18) 34 (58) 61 (15) 95 (21) 
Duration on emicizumab in weeks, median (IQR) 81.2 (61.8-96.5) 81.9 (66.4-91.3) 97.8 (84.4-100.5) 78.1 (59.5-91.1) 81.1 (62.4-95.1) 
Emicizumab dosing frequency, n (% of N)      
Weekly 50 (21) 87 (39) 21 (36) 116 (29) 137 (30) 
Every 2 wk 166 (70) 95 (43) 33 (56) 228 (57) 261 (57) 
Every 4 wk 20 (8) 41 (18) 5 (8) 56 (14) 61 (13) 
Patient characteristicsPediatric population (n = 236)Adult population (n = 223)Current inhibitor (n = 59)No current inhibitor (n = 400)Total population (N = 459)
Age (y), median (IQR) 9 (5-14) 36 (26-50) 20 (9-41) 17 (9-34) 17 (9-35) 
Pediatric (aged <18 y) NA NA 7.5 (5-10) 9.5 (5-14) 9 (5-14) 
Adult (aged ≥18 y) NA NA 38 (27-49) 36 (26-50) 36 (26-50) 
Diagnosis severity, n, (% of N)      
Severe 205 (87) 192 (86) 52 (88) 345 (86) 397 (86) 
Moderate 31 (13) 27 (12) 3 (5) 55 (14) 58 (13) 
Mild 0 (0) 4 (2) 4 (7) 0 (0) 4 (1) 
Inhibitor response, n, (% of N)      
No current inhibitor 210 (89) 190 (85) NA 400 (100) 400 (87) 
High responding 22 (9) 24 (11) 46 (78) NA 46 (10) 
Low responding 4 (2) 9 (4) 13 (22) NA 13 (3) 
Prior prophylaxis agent, n (% of N)      
SHL 52 (22) 110 (49) 16 (27) 146 (37) 162 (35) 
EHL 130 (55) 72 (32) 9 (15) 193 (48) 202 (44) 
On-demand 54 (23) 41 (18) 34 (58) 61 (15) 95 (21) 
Duration on emicizumab in weeks, median (IQR) 81.2 (61.8-96.5) 81.9 (66.4-91.3) 97.8 (84.4-100.5) 78.1 (59.5-91.1) 81.1 (62.4-95.1) 
Emicizumab dosing frequency, n (% of N)      
Weekly 50 (21) 87 (39) 21 (36) 116 (29) 137 (30) 
Every 2 wk 166 (70) 95 (43) 33 (56) 228 (57) 261 (57) 
Every 4 wk 20 (8) 41 (18) 5 (8) 56 (14) 61 (13) 

The median length of time on emicizumab was 81.1 weeks (IQR, 62.4-95.1). No patient discontinued emicizumab or reverted back to a prior prophylactic agent. Prophylaxis use before emicizumab varied between standard half-life (SHL; 35%) and extended half-life (EHL; 44%) rFVIII products, and on-demand therapy without regular prophylaxis (21%). In total, 25 patients with HA with a current inhibitor were documented as receiving either SHL or EHL products before commencing emicizumab, predicted to be on these products for immune tolerance induction. Most patients (57%) were prescribed emicizumab every 2 weeks. Apart from a change from the initial 4 weeks of loading dose to the ongoing maintenance dose, 129 of 459 (28%) patients had a dose escalation of emicizumab recorded, 9 of 459 (2%) patients had a change in schedule but no overall change in dose recorded, and 2 of 459 (<1%) had a reduction in emicizumab dose recorded. Reasons for dose changes are not recorded in the ABDR, however 111 of 129 (86%) occurred in the pediatric population and thus presumed to reflect dose adjustments based on age-related weight changes related to growth. No adverse events were recorded in the Australian Haemophilia Safety Surveillance System for emicizumab during this period.

The total number of bleeds in the 365 days before emicizumab commencement was 749 bleeds in 209 patients. The bleeding site was not specified in 526 of 749 (70%) bleeds, most of which were bleeds in adult patients (301/526; 57%) and in patients without an inhibitor (506/526; 96%). The cause (ie, provoked vs spontaneous) was not specified in 63 of 749 (8%) of these precommencement bleeds, most of which were bleeds in pediatric patients (43/63, 68%) and in patients without an inhibitor (50/63; 79%). The total number of bleeds since commencement for each patient on emicizumab, at the time of data collection on 6 October 2022, was 394 bleeds in 168 patients (noting the duration on emicizumab for each of these patients varied from 26.3 to 100.6 weeks). The site was not specified in 273 of 394 (69%) of these postcommencement bleeds, most of which were bleeds in adult patients (164/273; 60%) and in patients without an inhibitor (248/273; 91%). The cause was not specified in 21 of 394 (5%), just over half of which were bleeds in adults (11/21; 52%) and mostly in patients without an inhibitor (17/21; 81%). Additional data on the number of specified bleeds within the pediatric and adult populations are included in supplemental Tables 1 and 2. Median overall ABR in both the precommencement and postcommencement groups was 0, with an IQR of 0 to 2 before commencement of emicizumab, and an IQR of 0 to 0.54 after commencement, with a statistically significant difference (P < .01) between the 2 groups.

Zero-bleed prevalence in the total population and pediatric, adult, current inhibitor, and absent–current inhibitor subgroups are presented in Table 3. In the total cohort, 250 of 459 (54%) patients had no treated bleeding events before receiving emicizumab compared with 291 of 459 (63%) patients who had no treated bleeding events after commencement of emicizumab (relative risk [RR], 1.24; 95% confidence interval [CI], 1.09-1.41; P < .01; Figures 1 and 2; Table 3). This improvement in zero-bleed prevalence remained statistically significant in those without a current inhibitor (1.23; 95% CI, 1.07-1.41; P < .01) and in those with a current high-responding inhibitor (RR, 1.46; 95% CI, 1.05-2.28; P < .01). It also remained significant in patient-reported bleeding events (RR, 1.41; 95% CI, 1.12-1.78; P < .01) and spontaneous bleeding events (RR, 1.69; 95% CI, 1.34-2.13; P < .01). However, no significant difference was observed in HTC-recorded bleeds, provoked bleeds, and bleeds that required hospitalization before vs after commencement of emicizumab (Figures 1A and 2). In terms of age, increase in zero-bleed prevalence was significant within the pediatric population (48% before vs 61% after emicizumab commencement; RR, 1.34; 95% CI, 1.13-1.59; P < .01), although not significant in the adult population (61% before vs 65% after emicizumab; RR, 1.13; 95% CI, 0.92-1.38; P = .29).

Table 3.

RRs for zero-bleed prevalence after commencement of emicizumab compared with before commencement of emicizumab, in various subgroups and bleed subtypes

Subgroup or bleed subtypePediatric (n = 236)
RR (95% CI), P value
Adult (n = 223)
RR (95% CI), P value
Current inhibitor (n = 59)
RR (95% CI), P value
No current inhibitor (n = 400)
RR (95% CI), P value
Total population (n = 459)
RR (95% CI), P value
Overall 1.34 (1.13-1.59), P < .01 1.13 (0.92-1.38), P = .29 1.37 (0.91-2.05), P = .19 1.23 (1.07-1.41), P < .01 1.24 (1.09-1.41), P < .01 
Cause      
Spontaneous 2.14 (1.49-3.8), P < .01 1.40 (1.03-1.89), P = .04 1.78 (0.87-3.62), P = .17 1.68 (1.31-2.14), P < .01 1.69 (1.34-2.13), P < .01 
Provoked 1.14 (0.92-1.40), P = .28 1.13 (0.87-1.46), P = .43 1.00 (0.60-1.66), P = 1.00 1.15 (0.97-1.36), P = .14 1.13 (0.96-1.34), P = .15 
Bleed site      
Joint 1.79 (1.14-2.82), P = .01 0.90 (0.51-1.60), P = .86 2.00 (0.74-5.43), P = .27 1.30 (0.89-1.89), P = .21 1.38 (0.97-1.96), P = .09 
Muscle 1.4 (0.62-3.15), P = .54 3.25 (1.13-9.31), P = .04 P = .06 1.57 (0.82-3.01), P = .23 1.93 (1.03-3.62), P = .05 
Soft tissue 1.86 (0.78-4.44), P = .24 3.00 (0.61-14.86), P = .29 1.5 (0.38-6.00), P = 1.00 2.29 (0.94-5.56), P = .09 2.11 (0.98-4.53), P = .08 
Epistaxis 1.00 (0.40-2.52), P = 1.00 1.00 (0.06-15.99), P = 1.00 P = 1.00 0.75 (0.28-2.00), P = 1.00 1.00 (0.37-2.66), P = 1.00 
Gastrointestinal P = 1.00 1.00 (0.38-2.61), P = 1.00 2.00 (0.50-8.00), P = 1.00 1.00 (0.30-3.32), P = 1.00 1.2 (0.47-3.09), P = 1.00 
Intracranial hemorrhage 0.5 (0.05-5.51), P = 1.00 P = .13 P = 1.00 2.00 (0.37-10.92), P = .69 2.5 (0.49-12.89), P = .45 
Bleed recorder      
Patient recorded (myABDR) 1.63 (1.27-2.09), P < .01 1.22 (0.96-1.56), P = 14 0.88 (0.44-1.75), P = 1.00 1.46 (1.22-1.75), P < .01 1.41 (1.19-1.68), <0.01 
HTC recorded 1.05 (0.79-1.38), P = .82 1.03 (0.70-1.52), P = 1.00 1.43 (0.85-2.41), P = .26 0.98 (0.76-1.26), P = .93 1.04 (0.83-1.31), P = .80 
Bleeds requiring hospitalization 1.09 (0.81-1.48), P = .65 1.43 (0.89-2.30), P = .19 1.75 (1.02-3.00), P = .06 1.08 (0.81-1.45), P = .68 1.19 (0.92-1.53), P = .22 
Treatment before emicizumab      
On demand 2.13 (1.32-3.42), P < .01 1.25 (0.71-2.20), P = .61 1.67 (0.89-3.13), P = .18 1.79 (1.15-2.79), P = .01 1.75 (1.22-2.52), P < .01 
SHL prophylaxis 1.05 (0.80-1.36), P = 1.00 1.22 (0.93-1.58), P = .20 1.17 (0.69-1.97), P = 1.00 1.15 (0.94-1.42), P = .24 1.15 (0.95-1.40), P = .20 
EHL prophylaxis 1.23 (0.98-1.53), P = .10 0.96 (0.67-1.40), P = 1.00 1.00 (0.38-2.66), P = 1.00 1.14 (0.94-1.39), P = .22 1.14 (0.94-1.38), P = .24 
Emicizumab schedule      
Weekly 1.92 (1.21-3.03), P < .01 1.36 (0.98-1.89), P = .10 1.33 (0.55-3.26), P = .75 1.56 (1.19-2.05), P < .01 1.53 (1.17-1.99), P < .01 
Every 2 wk 1.15 (0.95-1.39), P = .19 0.86 (0.63-1.18), P = .46 1.36 (0.84-2.21), P = .34 1.02 (0.86-1.21), P = .91 1.05 (0.90-1.24), P = .59 
Every 4 wk 2.80 (1.29-6.09), P = .01 1.38 (0.87-2.20), P = .27 1.50 (0.67-3.34), P = 1.00 1.81 (1.17-2.80), P = .01 1.78 (1.19-2.65), P < .01 
Subgroup or bleed subtypePediatric (n = 236)
RR (95% CI), P value
Adult (n = 223)
RR (95% CI), P value
Current inhibitor (n = 59)
RR (95% CI), P value
No current inhibitor (n = 400)
RR (95% CI), P value
Total population (n = 459)
RR (95% CI), P value
Overall 1.34 (1.13-1.59), P < .01 1.13 (0.92-1.38), P = .29 1.37 (0.91-2.05), P = .19 1.23 (1.07-1.41), P < .01 1.24 (1.09-1.41), P < .01 
Cause      
Spontaneous 2.14 (1.49-3.8), P < .01 1.40 (1.03-1.89), P = .04 1.78 (0.87-3.62), P = .17 1.68 (1.31-2.14), P < .01 1.69 (1.34-2.13), P < .01 
Provoked 1.14 (0.92-1.40), P = .28 1.13 (0.87-1.46), P = .43 1.00 (0.60-1.66), P = 1.00 1.15 (0.97-1.36), P = .14 1.13 (0.96-1.34), P = .15 
Bleed site      
Joint 1.79 (1.14-2.82), P = .01 0.90 (0.51-1.60), P = .86 2.00 (0.74-5.43), P = .27 1.30 (0.89-1.89), P = .21 1.38 (0.97-1.96), P = .09 
Muscle 1.4 (0.62-3.15), P = .54 3.25 (1.13-9.31), P = .04 P = .06 1.57 (0.82-3.01), P = .23 1.93 (1.03-3.62), P = .05 
Soft tissue 1.86 (0.78-4.44), P = .24 3.00 (0.61-14.86), P = .29 1.5 (0.38-6.00), P = 1.00 2.29 (0.94-5.56), P = .09 2.11 (0.98-4.53), P = .08 
Epistaxis 1.00 (0.40-2.52), P = 1.00 1.00 (0.06-15.99), P = 1.00 P = 1.00 0.75 (0.28-2.00), P = 1.00 1.00 (0.37-2.66), P = 1.00 
Gastrointestinal P = 1.00 1.00 (0.38-2.61), P = 1.00 2.00 (0.50-8.00), P = 1.00 1.00 (0.30-3.32), P = 1.00 1.2 (0.47-3.09), P = 1.00 
Intracranial hemorrhage 0.5 (0.05-5.51), P = 1.00 P = .13 P = 1.00 2.00 (0.37-10.92), P = .69 2.5 (0.49-12.89), P = .45 
Bleed recorder      
Patient recorded (myABDR) 1.63 (1.27-2.09), P < .01 1.22 (0.96-1.56), P = 14 0.88 (0.44-1.75), P = 1.00 1.46 (1.22-1.75), P < .01 1.41 (1.19-1.68), <0.01 
HTC recorded 1.05 (0.79-1.38), P = .82 1.03 (0.70-1.52), P = 1.00 1.43 (0.85-2.41), P = .26 0.98 (0.76-1.26), P = .93 1.04 (0.83-1.31), P = .80 
Bleeds requiring hospitalization 1.09 (0.81-1.48), P = .65 1.43 (0.89-2.30), P = .19 1.75 (1.02-3.00), P = .06 1.08 (0.81-1.45), P = .68 1.19 (0.92-1.53), P = .22 
Treatment before emicizumab      
On demand 2.13 (1.32-3.42), P < .01 1.25 (0.71-2.20), P = .61 1.67 (0.89-3.13), P = .18 1.79 (1.15-2.79), P = .01 1.75 (1.22-2.52), P < .01 
SHL prophylaxis 1.05 (0.80-1.36), P = 1.00 1.22 (0.93-1.58), P = .20 1.17 (0.69-1.97), P = 1.00 1.15 (0.94-1.42), P = .24 1.15 (0.95-1.40), P = .20 
EHL prophylaxis 1.23 (0.98-1.53), P = .10 0.96 (0.67-1.40), P = 1.00 1.00 (0.38-2.66), P = 1.00 1.14 (0.94-1.39), P = .22 1.14 (0.94-1.38), P = .24 
Emicizumab schedule      
Weekly 1.92 (1.21-3.03), P < .01 1.36 (0.98-1.89), P = .10 1.33 (0.55-3.26), P = .75 1.56 (1.19-2.05), P < .01 1.53 (1.17-1.99), P < .01 
Every 2 wk 1.15 (0.95-1.39), P = .19 0.86 (0.63-1.18), P = .46 1.36 (0.84-2.21), P = .34 1.02 (0.86-1.21), P = .91 1.05 (0.90-1.24), P = .59 
Every 4 wk 2.80 (1.29-6.09), P = .01 1.38 (0.87-2.20), P = .27 1.50 (0.67-3.34), P = 1.00 1.81 (1.17-2.80), P = .01 1.78 (1.19-2.65), P < .01 

RRs with significant P value ≤ .5 are shown in bold.

For bleed subtypes where no RR or 95% CI is included, the RR and 95% CI were unable to be calculated due to small subsample population.

Figure 1.

Proportions of patients with zero bleeds before and after commencement of emicizumab by various subgroups. (A) Proportion of zero bleeds in various bleed types. (B) Proportion of zero bleeds in various bleed sites. The asterisk highlights a significance level (P < .05) indicating that a difference exists between zero-bleed prevalence before commencement of emicizumab and zero-bleed prevalence after commencement of emicizumab. ICH, intracranial hemorrhage.

Figure 1.

Proportions of patients with zero bleeds before and after commencement of emicizumab by various subgroups. (A) Proportion of zero bleeds in various bleed types. (B) Proportion of zero bleeds in various bleed sites. The asterisk highlights a significance level (P < .05) indicating that a difference exists between zero-bleed prevalence before commencement of emicizumab and zero-bleed prevalence after commencement of emicizumab. ICH, intracranial hemorrhage.

Close modal
Figure 2.

Forest plot demonstrating RRs for achieving zero bleeds after commencement of emicizumab in various subgroups. Red, total bleeds; blue, spontaneous bleeds; black, provoked bleeds. Inhibitor refers to patients with a current inhibitor. P values, used to test the hypothesis that zero-bleed prevalence is different before and after commencement of emicizumab, have been included if significant (P < .05).

Figure 2.

Forest plot demonstrating RRs for achieving zero bleeds after commencement of emicizumab in various subgroups. Red, total bleeds; blue, spontaneous bleeds; black, provoked bleeds. Inhibitor refers to patients with a current inhibitor. P values, used to test the hypothesis that zero-bleed prevalence is different before and after commencement of emicizumab, have been included if significant (P < .05).

Close modal

In terms of bleed sites, difference in zero-bleed prevalence before and after commencement was only significant in muscle bleeds within the total population (RR, 1.93; 95% CI, 1.03-3.62; P = .05; Figure 1B; Table 3). Overall, zero-bleed prevalence did not significantly increase in joint bleeds (RR, 1.38; 95% CI, 0.97-1.96; P = .09). However, absence of joint bleeding significantly increased within the pediatric subgroup (RR, 1.79; 95% CI, 1.14-2.82; P = .01) and absence of muscle bleeding significantly increased within the adult subgroup (RR, 3.25; 95% CI, 1.13-9.31; P = .04). It should be noted that the bleed-site field had a large amount of missing data, as described earlier.

A total of 98 patients had a documented joint bleed either before commencement of emicizumab, after commencement, or both. The bubble plot (Figure 3) illustrates the numerical distribution of joint ABR before and after emicizumab commencement for these 98 patients, with the greatest proportion of patients who had a bleeding event either before or after emicizumab commencement being reduced from a joint ABR of 1 before commencement to a joint ABR of 0 after emicizumab commencement (41/98; 42%). However, this figure also demonstrates that 39% (38/98) of this cohort had an increase in joint bleeds after commencing emicizumab. Most of the bleeds in this cohort were provoked bleeds, that is, 39 of 56 (70%) joint bleeds in the 38 patients who experienced an increase in joint ABR after starting emicizumab were provoked. These 38 patients were relatively evenly distributed with regards to age; 18 of 38 (47%) were pediatric patients, and 20 of 38 (53%) were adults.

Figure 3.

Joint bleeds before and after commencement of emicizumab. The bubble plot demonstrates the numerical distribution of patients by joint ABR, before and after commencement of emicizumab in those who had a joint bleed before commencement of emicizumab, after commencement of emicizumab, or both (ie, does not include those who had 0 bleeds both before commencement and after commencement of emicizumab). The postcommencement joint ABR has been rounded up to the nearest whole number given that the ABR was a derived calculated rate resulting in values with decimal places. One outlier has been removed to improve readability of the figure; this was a patient who had 6 bleeds before commencement of emicizumab and 31 bleeds after commencement. The size of the bubble indicates the number of patients.

Figure 3.

Joint bleeds before and after commencement of emicizumab. The bubble plot demonstrates the numerical distribution of patients by joint ABR, before and after commencement of emicizumab in those who had a joint bleed before commencement of emicizumab, after commencement of emicizumab, or both (ie, does not include those who had 0 bleeds both before commencement and after commencement of emicizumab). The postcommencement joint ABR has been rounded up to the nearest whole number given that the ABR was a derived calculated rate resulting in values with decimal places. One outlier has been removed to improve readability of the figure; this was a patient who had 6 bleeds before commencement of emicizumab and 31 bleeds after commencement. The size of the bubble indicates the number of patients.

Close modal

In those receiving on-demand therapy, there was a statistically significant increase in zero-bleed prevalence from before emicizumab commencement to after commencement (RR, 1.78; 95% CI, 1.24-2.56; P < .01). This difference was not observed in those on either SHL or EHL rFVIII products before emicizumab (Figure 2).

Notably, when spontaneous bleeds alone were evaluated, the difference in zero-bleed prevalence became significant for all subgroups except for patients with current low-responding inhibitor and in the total current inhibitor population (Figure 2). This suggests that the lack of significance in the adult, and SHL and EHL groups was driven predominantly by a lack of significant difference in provoked bleeding events. In the adult population (n = 223), the RR for zero spontaneous bleeds was 1.40 (95% CI, 1.03-1.89; P = .04) while the RR for zero provoked bleeds was not significant at 1.13 (95% CI, 0.87-1.46; P = .43; Table 3). Within the pediatric population (n = 236), the RR for zero spontaneous bleeds was 2.14 (95% CI, 1.49-3.08; P < .01) and for zero provoked bleeds was not significant at 1.14 (95% CI, 0.92-1.40; P = .28). Data on the number of bleeds within the pediatric and adult populations are included in supplemental Tables 1 and 2. For bleed causes that were specified, the proportion of provoked bleeds increased in the pediatric population after commencement of emicizumab (58% to 73%). The adult population demonstrated a more modest increase (46% to 53%).

We performed subgroup analysis of the 400 patients who did not have a current inhibitor (Table 3; supplemental Table 3). In this subgroup, there was an increase in those who had zero bleeds after commencement of emicizumab (63%) compared with before commencement (54%; RR, 1.23; 95% CI, 1.07-1.41; P < .01). An increase in the number of pediatric patients with HA without any bleeds (50% before vs 61% after emicizumab commencement; RR, 1.30; 95% CI, 1.08-1.56; P < .01) was observed, but there was no significant increase in the number of adults with zero bleeds (59% before vs 58% after emicizumab commencement; RR, 1.15; 95% CI, 0.93-1.42; P = .25). There was an increase in the number of patients on on-demand therapy with zero bleeds (RR, 1.79; 95% CI, 1.15-2.79; P = .01) but no difference in zero-bleed prevalence in the SHL (RR, 1.15; 95% CI, 0.94-1.42; P = .24) and EHL (RR, 1.14; 95% CI, 0.94-1.39; P = .22) populations. These results are concordant with those from the overall population.

There was no difference in 0-bleed prevalence within the current inhibitor population (56% before vs 68% after emicizumab commencement; RR, 1.37; 95% CI, 0.91-2.05; P = .19). This may be attributable to a small subgroup (n = 59), underpowered to detect significant differences. The zero-bleed prevalence in the adult current inhibitor population was 73%, both before and after emicizumab commencement. In the current inhibitor pediatric population, the precommencement zero-bleed percentage was 35% compared with 62% after emicizumab commencement, however this difference was not statistically significant (RR, 1.7; 95% CI, 1.03-2.80; P = .07).

Our data demonstrate favorable zero-bleeding rates, which are similar to those previously published in both trial and real-world settings.10,11,13-18,21,22 This work examines, to our knowledge, the largest cohort of patients on emicizumab described in the published literature to date, and is one of few real-world data sets to compare intraindividual bleeding events before commencement of emicizumab with those after commencement. Additionally, the follow-up time of this cohort (median duration on emicizumab, 81.1 weeks) was significantly longer than that of other published real-world studies,15-18,22 allowing interpretations of both the early- and medium-term impact on bleeding outcomes of this therapy.

Notably, no significant difference in the rate of zero all-cause bleeds was observed in adults, or patients on either SHL or EHL before changing to emicizumab, although the increase in zero spontaneous bleeds for these groups was significant. This analysis suggests that the lack of reduction in provoked bleeds for several subgroups drives the lack of all-cause–bleed reduction for these patients. For bleeds with specified cause, the proportion of bleeds that were provoked increased after commencement (231/373; 62%) compared with before emicizumab commencement (353/686; 51%), suggesting that the phenotype becomes milder and characterized by fewer spontaneous bleeds. Additionally, the increase in zero-bleed prevalence in the pediatric population which was absent in the adult population, suggests that the overall improvement in bleeding was mainly driven by reduced bleeding in younger patients. The high percentage of adult patients with zero bleeds before commencing emicizumab (61%) also suggests that this cohort already had a high level of hemostatic control on prior therapy, which may reflect better adherence and reduced risk-taking activity generally, although these parameters were not measured in this study. Studies that have incorporated measures of activity have demonstrated that the majority of participants on emicizumab report an increase in physical activity since commencement of the agent.15,23 This supports the postulation that although emicizumab reduces the risk of bleeding overall, bleeding can still occur and tends to be trauma induced in a patient cohort with a greater degree of physical confidence.

Real-world data evaluating the efficacy and safety of emicizumab have been published, although these data are limited and from studies without particularly large sample sizes or robust intraindividual comparisons. Data for 40 children with HA both with and without inhibitors from an Israeli pediatric center demonstrated that 50% of the cohort experienced no bleeds over a median follow-up of 45 weeks, and all bleeds reported were trauma provoked.18 Another study on the real-world use of emicizumab in a predominantly pediatric population (N = 93) by McCary et al demonstrated an ABR of 0.4 in both inhibitor and noninhibitor groups, with a zero-bleeding prevalence of 66%.17 Similar ABRs (0.5-0.9) and prevalence of zero-bleeding events (63%-67%) have been demonstrated in adults in the real-world setting.15,16,22 

In terms of data assessing intraindividual comparisons of bleeding before and after commencement of emicizumab, a minority of small data sets have been published. The HAVEN-3 trial of adults and adolescents included 48 participants who had received previous FVIII prophylaxis before commencing the trial arm of emicizumab, with an increase in zero-bleed prevalence from 40% to 54%,13 percentages that were lower than observed in our total population. It should be noted that only 66% of the aforementioned participants with adherence assessed (defined as administration of ≥80% of prescribed doses) were adherent to their FVIII prophylaxis in the HAVEN-3 trial. Adherence in our study could not be measured, because not every participant documented every prophylactic dose in the myABDR self-documentation section of the ABDR. High levels of adherence and compliance with therapy (both factor replacement and emicizumab) may account for the high occurrence of zero-bleed percentages and smaller differences before and after commencement of emicizumab compared with other intraindividual data published.

McCary et al17 analyzed their 93-patient real-world cohort before and after switch to emicizumab.17 The proportion of patients experiencing zero bleeding events during the initial 6 months of emicizumab initiation saw a significant increase when compared with the 6 months preceding the commencement of emicizumab in both the inhibitor group (increasing from 42% to 95%; P < .01) and the non-inhibitor group (increasing from 59% to 89%; P < .01). Note is made of the small sample size and shorter length of follow-up both before and after commencement of emicizumab. In contrast, work by Escobar et al, published more recently in 2023,24 examining billed bleeding events, as identified from a claims database, in patients with HA without inhibitors on rFVIII prophylaxis before and after switching to emicizumab. They demonstrated no significant difference in billed bleeding events between these two 2 periods. This is in keeping with our results demonstrating a lack of difference in zero-bleed prevalence in those on SHL or EHL rFVIII. However, the method of data collection (using a claims database) in this study biases toward a certain demographic and may not be as accurate as a medical registry in terms of recording of bleeding outcomes. Nonetheless, this study, in conjunction with our work, is relevant in highlighting discrepancies that can occur in translation between clinical trial and real-world data.

The overall numbers of bleeds in each bleed site were reduced in both pediatric and adult populations (supplemental Tables 1 and 2). However, analysis of zero-bleed prevalence did not result in significant differences except in muscle bleeds in the total population, joint bleeds in the pediatric population, and muscle bleeds in the adult population. It is possible that the lack of difference in joint bleeding in the adult population is driven by target joints. Levy-Mendelovich et al demonstrated ongoing high rates of spontaneous bleeding into target joints in their cohort of 70 patients on emicizumab.25 Ongoing target joint bleeds coupled with ongoing, or even increased, activity in more adventurous patients may account for this lack of reduction in joint bleed risk. Moreover, it is difficult to accurately interpret these findings because of the large volume of missing bleed-site data. Both before commencement and after commencement of emicizumab, 69% to 70% of bleeds did not have a bleed site specified in the ABDR. This highlights an area for future improvement within the registry.

Patients with HA with inhibitors within adult (63%) and pediatric (77%) populations in the HAVEN-1 and HAVEN-2 trials,10,11 respectively, had zero-bleed percentages after commencing emicizumab that were as similarly high as as in our adult (73%) and pediatric (62%) cohorts. Curiously, however, no significant differences in zero-bleed prevalence before emicizumab commencement compared with after emicizumab commencement were present in our study within the overall current inhibitor group, although a reduction in bleeding was observed within the high-responding inhibitor group. There were also high numbers of adult patients with current inhibitor with zero bleeds (73%) both before and after commencement of emicizumab. This surprisingly low bleeding rate in a high-risk population may be explained by some of these patients being on emicizumab through a clinical trial before changing to funded emicizumab. However, this cannot be verified because that information was not available from the ABDR data extract. Greater sample sizes in the real-world setting are required to interrogate this lack of difference further.

Several limitations are present in this study. Incomplete recording of data in the registry curtails in-depth analysis of various subgroups, as does the possibility of ascertainment bias given the reliance on patient-reported bleeding events. Notably, data on target joints, arthropathy, and synovitis are not available in the ABDR. Perioperative data have historically not been entered reliably in a standardized manner in the ABDR, and, thus, were not able to be retrospectively collected and analyzed for inclusion in this study. However, these data are currently being prospectively collected for future analysis. Secondly, the follow-up after commencement of emicizumab varied from patient to patient. However, most patients were followed-up for >365 days (392/460 patients, 85%), rendering any inaccuracy to be likely minimal and more likely to result in an underestimation of zero-bleed prevalence after emicizumab commencement. Thirdly, although compliance with the Australian Haemophilia Safety Surveillance System reporting system is believed to be high, there is a small, unquantifiable risk that not all adverse reports are reported, because this is the responsibility of each individual HTC. In the real-world setting, minor adverse events are less likely to be reported than under the more stringent setting of a clinical trial. Additionally, the earliest possible starting date for funded emicizumab (November 2020) commenced after the publications of the HAVEN-1 to HAVEN-4 trials. These trials had highlighted the signal of thrombotic risk when the combination of activated prothrombin complex concentrate and emicizumab was used,10,11,13,14 enabling clinicians to avoid this combination. We believe that the avoidance of this known risk has reduced the real-world occurrence of thrombotic events, which was the primary adverse signal that arose from trial data. Less severe reactions, such as localized injection site reactions known to be relatively common at ∼15% to 25% of patients11,13,15 may have been less likely to be reported because of the minor nature of the reaction.

Strengths of this study include a large data set (ie, the entire Australian HA population currently on emicizumab), a long duration of follow-up, and intraindividual comparisons, which suggest that the difference in overall zero-bleed prevalence before and after commencement calculated is a highly accurate estimate of the changes in the bleeding outcome landscape since rollout of this therapy in the real-world setting. Given that only data on funded emicizumab was collected (because of trial proprietary reasons), the postcommencement outcomes may, in fact, be an underestimation, given that some patients had been on trial emicizumab before changing to funded treatment.

Conclusion

These real-world results demonstrate a reduction in bleeding events with the use of emicizumab prophylaxis in Australian patients with HA. The zero-bleed prevalence increased after commencement of emicizumab in the overall population and in several subgroups including those with a current high-responding inhibitor, those without a current inhibitor, and pediatric patients. There was a lack of difference in the all-cause zero-bleed prevalence in the adult, and SHL and EHL subgroups. This appears to be driven by a lack of difference in provoked bleeding events, given that there was a significant difference in the zero–spontaneous bleed prevalence. One of the most significant reductions in bleeding observed was in the zero spontaneous bleed prevalence within the pediatric population, reflective of a shift toward a phenotype characteristic of mild HA, and raising the possibility of increased physical activity accounting for the lack of difference in provoked bleeding. With, to our knowledge, the largest study population evaluating real-world bleeding outcomes of emicizumab in patients with HA to date, including diversity in age groups, inhibitor status, dosing frequency, and prior prophylaxis agents, these results support the efficacy of emicizumab in reducing bleeding events in this cohort.

The authors thank the National Blood Authority, AHCDO Executive Committee, and all hemophilia treatment center directors and staff for their support and guidance. The authors also thank the ABDR Data Managers Group for maintaining and validating Australian Bleeding Disorder Registry data.

Contribution: R.R., S.P., J.D.M., and H.A.T. contributed to study design; all authors participated in project meetings to develop research methods and for data interpretation; R.R., S.P., and L.L.A. analyzed the data and created the tables/figures; R.R. wrote the manuscript; and all authors critically revised the manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Huyen A. Tran, Australian Centre for Blood Diseases, Monash University Alfred Health, 55 Commercial Rd, Melbourne VIC 3004, Australia; email: huyen.tran@monash.edu.

1.
Hoyer
LW.
.
Hemophilia A
.
N Engl J Med
.
1994
;
330
(
1
):
38
-
47
.
2.
Blanchette
VS
,
Key
NS
,
Ljung
LR
,
Manco-Johnson
MJ
,
van den Berg
HM
,
Srivastava
A
;
Subcommittee on Factor VIII, Factor IX and Rare Coagulation Disorders of the Scientific and Standardization Committee of the International Society on Thrombosis and Hemostasis
.
Definitions in hemophilia: communication from the SSC of the ISTH
.
J Thromb Haemost
.
2014
;
12
(
11
):
1935
-
1939
.
3.
National Blood Authority
.
Australian Bleeding Disorders Registry Annual Report 2019-2020
. Accessed 4 April 2023. https://www.blood.gov.au/sites/default/files/ABDR-Annual-Report-2019-20-FINAL.pdf.
4.
Astermark
J
.
Why do inhibitors develop? Principles of and factors influencing the risk for inhibitor development in haemophilia
.
Haemophilia
.
2006
;
12
(
suppl 3
):
52
-
60
.
5.
Mahlangu
J
,
Oldenburg
J
,
Callaghan
MU
, et al
.
Health-related quality of life and health status in persons with haemophilia A with inhibitors: a prospective, multicentre, non-interventional study (NIS)
.
Haemophilia
.
2019
;
25
(
3
):
382
-
391
.
6.
Walsh
CE
,
Soucie
JM
,
Miller
CH
;
United States Hemophilia Treatment Center Network
.
Impact of inhibitors on hemophilia A mortality in the United States
.
Am J Hematol
.
2015
;
90
(
5
):
400
-
405
.
7.
Morfini
M
.
Articular status of haemophilia patients with inhibitors
.
Haemophilia
.
2008
;
14
(
suppl 6
):
20
-
22
.
8.
Kitazawa
T
,
Igawa
T
,
Sampei
Z
, et al
.
A bispecific antibody to factors IXa and X restores factor VIII hemostatic activity in a hemophilia A model
.
Nat Med
.
2012
;
18
(
10
):
1570
-
1574
.
9.
Shima
M
,
Hanabusa
H
,
Taki
M
, et al
.
Factor VIII–mimetic function of humanized bispecific antibody in hemophilia A
.
N Engl J Med
.
2016
;
374
(
21
):
2044
-
2053
.
10.
Young
G
,
Liesner
R
,
Chang
T
, et al
.
A multicenter, open-label phase 3 study of emicizumab prophylaxis in children with hemophilia A with inhibitors
.
Blood
.
2019
;
134
(
24
):
2127
-
2138
.
11.
Oldenburg
J
,
Mahlangu
JN
,
Kim
B
, et al
.
Emicizumab prophylaxis in hemophilia A with inhibitors
.
N Engl J Med
.
2017
;
377
(
9
):
809
-
818
.
12.
Chai-Adisaksopha
C
,
Nevitt
SJ
,
Simpson
ML
,
Janbain
M
,
Konkle
BA
.
Bypassing agent prophylaxis in people with hemophilia A or B with inhibitors
.
Cochrane Database Syst Rev
.
2017
;
9
(
9
):
CD011441
.
13.
Mahlangu
J
,
Oldenburg
J
,
Paz-Priel
I
, et al
.
Emicizumab prophylaxis in patients who have hemophilia A without inhibitors
.
N Engl J Med
.
2018
;
379
(
9
):
811
-
822
.
14.
Pipe
SW
,
Shima
M
,
Lehle
M
, et al
.
Efficacy, safety, and pharmacokinetics of emicizumab prophylaxis given every 4 weeks in people with haemophilia A (HAVEN 4): a multicentre, open-label, non-randomised phase 3 study
.
Lancet Haematol
.
2019
;
6
(
6
):
e295
-
e305
.
15.
Warren
BB
,
Chan
A
,
Manco-Johnson
M
, et al
.
Emicizumab initiation and bleeding outcomes in people with hemophilia A with and without inhibitors: a single-center report
.
Res Pract Thromb Haemost
.
2021
;
5
(
5
):
e12571
.
16.
Ebbert
PT
,
Xavier
F
,
Seaman
CD
,
Ragni
MV
.
Emicizumab prophylaxis in patients with haemophilia A with and without inhibitors
.
Haemophilia
.
2020
;
26
(
1
):
41
-
46
.
17.
McCary
I
,
Guelcher
C
,
Kuhn
J
, et al
.
Real-world use of emicizumab in patients with haemophilia A: bleeding outcomes and surgical procedures
.
Haemophilia
.
2020
;
26
(
4
):
631
-
636
.
18.
Barg
AA
,
Livnat
T
,
Budnik
I
, et al
.
Emicizumab treatment and monitoring in a paediatric cohort: real-world data
.
Br J Haematol
.
2020
;
191
(
2
):
282
-
290
.
19.
National Blood Authority
.
National Supply Arrangements for Hemlibra (Emicizumab)
. Accessed 4 April 2023. https://www.blood.gov.au/national-supply-arrangements-hemlibra-emicizumab.
20.
National Blood Authority
.
Australian Bleeding Disorders Registry (ABDR)
. Accessed 4 April 2023. https://www.blood.gov.au/abdr.
21.
Callaghan
MU
,
Negrier
C
,
Paz-Priel
I
, et al
.
Long-term outcomes with emicizumab prophylaxis for hemophilia A with or without FVIII inhibitors from the HAVEN 1-4 studies
.
Blood
.
2021
;
137
(
16
):
2231
-
2242
.
22.
Arcudi
S
,
Gualtierotti
R
,
Marino
S
, et al
.
Real-world data on emicizumab prophylaxis in the Milan cohort
.
Haemophilia
.
2022
;
28
(
5
):
e141
-
e144
.
23.
Négrier
C
,
Mahlangu
J
,
Lehle
M
, et al
.
Emicizumab in people with moderate or mild haemophilia A (HAVEN 6): a multicentre, open-label, single-arm, phase 3 study
.
Lancet Haematol
.
2023
;
10
(
3
):
e168
-
e177
.
24.
Escobar
M
,
Bullano
M
,
Mokdad
AG
, et al
.
A real-world evidence analysis of the impact of switching from factor VIII to emicizumab prophylaxis in patients with hemophilia A without inhibitors
.
Expert Rev Hematol
.
2023
;
16
(
6
):
467
-
474
.
25.
Levy-Mendelovich
S
,
Brutman-Barazani
T
,
Budnik
I
, et al
.
Real-world data on bleeding patterns of hemophilia A patients treated with emicizumab
.
J Clin Med
.
2021
;
10
(
19
):
4303
.

Author notes

Data are available on request from the corresponding author, Huyen A. Tran (huyen.tran@monash.edu).

The online version of this article contains a data supplement.

Supplemental data