A real-world evidence analysis of the impact of switching from factor VIII to emicizumab prophylaxis in patients with hemophilia A without inhibitors

ABSTRACT

Background

This study retrospectively compared annualized billed bleed rates (ABR_b) in people with hemophilia A (PwHA) without inhibitors who switched from factor VIII (FVIII) prophylaxis to emicizumab.

Research design and methods

A real-world comparison study was performed on the effect of switching from FVIII to emicizumab prophylaxis in male, non-inhibitor patients on ABR_b using an all-payer claims database (APCD) dataset from 1 January 2014, to 31 March 2021. The identification period was from 1 November 2017, to 30 September 2020.

Results

One hundred and thirty-one patients were included with a total of 82 and 45 bleeds in the pre- and post-switch periods, respectively. The average follow-up period pre-switch was 978.37 days (SD 555.03), whereas the average follow-up period post-switch was 522.26 days (SD 191.36). No significant differences in mean ABR_b were observed pre-/post-switch (0.25 and 0.20, respectively; P = 0.4456).

Conclusions

The results of this study demonstrate no significant reduction in ABR_b, suggesting that switching from FVIII to emicizumab may not deliver incremental benefits to PwHA receiving prophylactic care.

KEYWORDS:

• Hemophilia A
• billed annualized bleed rate
• factor VIII
• emicizumab
• real-world evidence
• retrospective analysis
• prophylaxis
• treatment switching

1. Introduction

Hemophilia A (HA) is a rare recessive X-linked genetic bleeding disorder caused by a deficiency of clotting factor VIII (FVIII), resulting in disruption of normal hemostasis [1,2]. In the United States, HA affects approximately 1 in 5000 male births [1,3]. Spontaneous bleeding episodes characteristically occur in joints; this can lead to permanent joint damage, resulting in chronic joint pain, reduced range of motion, and decreased overall quality of life for people with HA (PwHA). Potentially fatal intracranial bleeding can also occur [4,5].

Prophylaxis is important for the prevention of bleeding episodes and mitigation of long-term tissue damage in bleed sites [5–7]. FVIII replacement via intravenous infusion of FVIII products is administered for prophylactic or on-demand therapy and has a long history of safe and effective use in children and adults [6, 8, 9]. Emicizumab (Hemlibra®; Genentech Inc., South San Francisco, CA) is a non-factor therapy indicated for the prophylactic treatment of PwHA [10–13]. Emicizumab is a bispecific antibody that functions by bringing together factors IXa and X to restore the blood clotting process, regardless of whether FVIII inhibitors are present [11,14].

Recent studies have begun to shed light on the outcomes of switching from FVIII to emicizumab prophylaxis in PwHA without inhibitors; however, many of these analyses relied on closed claims databases [11,14]. To address this knowledge gap, this study utilized an all-payer claims database (APCD), to compare billed annualized bleed rates (ABR_b) retrospectively in PwHA without inhibitors before and after switching from FVIII prophylaxis to emicizumab.

2. Patients and methods

This is a retrospective study of deidentified data from existing claims databases without any direct enrollment of patients. Personal identifiable data is not collected in this study. Informed Consent was therefore not obtained. Data complies with the patient confidentiality requirements of the Health Insurance Portability and Accountability Act.

2.1. Study design

This retrospective, observational study compared clinical outcomes among male PwHA without inhibitors before and after switching from FVIII to emicizumab prophylaxis (Figure 1) using an APCD population licensed by STATinMED (STATinMED LLC, Dallas, TX). The APCD provides insight into ~ 80% of US healthcare systems and has large-scale databases that systematically collect medical claims, pharmacy claims, and eligibility and provider files from private and public payers [15,16]. The study period was from 1 January 2014, to 31 March 2021, and the identification period, defined as the period during which the pre-defined inclusion and exclusion criteria were applied, was from 1 November 2017, to 30 September 2020.

Figure 1. Study design

The index date represents the switch date to emicizumab from FVIII prophylaxis during the identification period. The post-switch follow-up period was the period from the index date up to the end of study period (31 March 2021). The pre-switch baseline period began from the start of FVIII prophylaxis up until 1 day prior to the index date.

Patient identification included male PwHA without inhibitors with both medical and pharmacy health benefit plans from the United States, who switched from FVIII to emicizumab prophylaxis during the identification period.

2.1.1. Participant selection criteria

Males (all ages) with ≥ 1 outpatient/pharmacy claim for emicizumab in the identification period and FVIII prophylaxis in the baseline period, prior switching to emicizumab, were included. To identify patients using FVIII for prophylaxis, utilization and pattern criteria were established. The utilization criterion was an annual FVIII dose of > 2340 IU/kg, excluding units consumed 7 days prior to bleed diagnosis, and was applied to each year in the baseline period prior to switching to emicizumab. The pattern criterion was continuous FVIII billing patterns before the index date, determined using the following requirements: ≥1 FVIII claim in each of the 6-month intervals, to be counted as part of the patient’s prophylaxis treatment; the earliest date of a FVIII claim in the earliest 6-month interval for the patient’s continuous prophylaxis treatment duration was considered the prophylaxis start date. Continuous capture of information for ≥ 6 months pre-index and ≥ 6 months post-index was also required. Inclusion into the study was allowed for participants on standard half-life or extended half-life FVIII prophylaxis as long as the participant selection criteria were met.

Exclusion criteria included select patients who did not meet the utilization criterion (annual FVIII dose > 2340 IU/kg) but met the pattern criterion (continuous FVIII billing patterns before index date) based on manual review (possibly due to small sample size). PwHA with inhibitors in the baseline period (identified using ≥ 1 claim of the bypassing agents FEIBA-aPCC, rFVIIa, or Novoseven RT [Novo Nordisk Inc., Plainsboro, NJ] or immune tolerance induction therapy) were excluded. Immune tolerance induction therapy was defined as ≥ 1 claim(s) for Bethesda/Nijmegen assays in the study period and a high dose of FVIII dispensed in the complete baseline period; this was defined as > 3× the median IU factors dispensed for patients in the same age group for more than five consecutive 28-day intervals, allowing one interval gap. PwHA with ≥ 2 claims in any diagnosis position ≥ 30 days apart within the study period for hemophilia B, von Willebrand disease, acquired hemophilia A, or any other bleeding disorder were excluded. PwHA with < 180 days of FVIII prophylaxis treatment in the baseline period before index date were also excluded.

2.2. Outcomes

Patients were stratified using the Charlson Comorbidity Index (CCI). Data were reported as ABR_b during prophylaxis treatment. ABR_b was calculated as the number of billed bleeding events divided by the number of months in the observation period. Bleed events in this study were those for which patients visited an in-patient or out-patient facility, and as such, the medical encounter was processed and reimbursed by a payer. All bleeding-related medical claims were extracted using previously finalized diagnosis codes and grouped by location. Bleeding events were captured using diagnosis codes in the primary diagnosis position, and any claims in the same anatomical location within 7 days were grouped as a single event. Claims for bleeds in different locations were treated as separate events. ABR_b was estimated using all billed bleed events at the patient level and stratified by bleed type (spontaneous or traumatic). Within bleed types, ABR_b was further stratified by location including joint, soft tissue and muscle, intracranial, gastrointestinal, genitourinary, intracavity, perinatal, and facial. Average follow-up (in days) was calculated as the time from index date to date of the last emicizumab treatment.

2.3. Statistical methods

Descriptive statistics were used to summarize ABR_b outcomes. Counts and percentages were used for categorical variables (e.g. demographics). Mean, standard deviation (SD), median (interquartile range [IQR]), and range (min and max) were calculated for continuous variables (e.g. ABR_b). The robustness of these findings was assessed by identifying bleeding events using data from all diagnosis positions as opposed to just the primary diagnosis on medical claims, which was used in the main analysis. Correlations between the pre- and post-switch outcomes were captured using the following equation:

where, r is the correlation coefficient, SD_pre and SD_post are the standard deviations of the outcomes in the pre- and post-switch periods, respectively, and SD_D is the standard deviation of the difference in outcomes in the post- and pre-switch periods [17].

3. Results

3.1. Demographics

Eight hundred and fifty-three patients with ≥ 1 outpatient/pharmacy claim for emicizumab were identified during the identification period. After applying the selection criteria, 131 male patients who switched from FVIII prophylaxis to emicizumab were included in the study. All demographic data are reported in Table 1. The mean age was 28.9 years (median, 25 years; range, 28 years, minimum and maximum, 3 and 83 years). The largest age group was adults older than 26 years (age 3–6, 8.4%; 7–12, 16%; 13–17, 10%; 18–26, 17%; > 26, 49%); 53% (n = 70), 38% (n = 50), and 8% (n = 11) were enrolled in commercial, Medicaid, and Medicare payer channels, respectively.

Table 1. Demographic characteristics.

	Participants who switched from FVIII to emicizumab prophylaxis N = 131
Index year, years (%)
2017	0 (0.0)
2018	19 (14.5)
2019	80 (61.1)
2020	32 (24.4)
Age, years
Mean (SD)	28.9 (18.6)
Median	25
Range	28
Minimum	3
Maximum	83
IQR	13–41
Age group, n (%)
0–2 years	0 (0.0)
3–6 years	11 (8.4)
7–12 years	21 (16.0)
13–17 years	13 (9.9)
18–26 years	22 (16.8)
> 26 years	64 (48.9)
Sex, n (%)
Male	131 (100)
Female	0 (0.0)
Payer channel, n (%)
Commercial	70 (53.4)
Medicaid	50 (38.2)
Medicare	11 (8.4)
VA/Other	0 (0.0)
Geographic region, n (%)
Northeast	26 (19.8)
Midwest	43 (32.8)
South	33 (25.2)
West	29 (22.1)
Other	0 (0.0)

FVIII: factor VIII; SD: standard deviation.

3.2. Clinical characteristics

The mean CCI score was 0.23. A total of 47% of patients had joint health-related (including arthritis and other arthropathies) individual comorbidities and 34% of patients had musculoskeletal pain (including back pain); 27% of patients had soft-tissue disorders (in the muscles, synovium, and tendons). In all, 18% of patients had hepatitis C/B and 8% of patients had HIV. The average pre-index annual FVIII dose utilization was 4386 IU/kg.

3.3. Billed annualized bleed rates

ABR_b data are reported in Table 2. There was a total of 82 and 45 bleed events in the pre- and post-switch periods, respectively (P = 0.9491) (mean ABR_b = 0.25 pre-switch vs 0.20 post-switch; P = 0.4456). The correlation coefficient (a statistical measure of the strength of a linear relationship between two variables) between pre-switch and post-switch ABR_b was 0.2860. PwHA in all age groups demonstrated a numerically lower, but not statistically significant, mean change in ABR_b post-switch. The mean ABR_b for spontaneous bleeds was 0.23 and 0.18 in the pre- and post-switch periods, respectively (P = 0.5173). The correlation coefficient between pre-switch and post-switch spontaneous ABR_b was 0.2862. The mean ABR_b for traumatic bleeds was 0.02 and 0.01 in the pre- and post-switch periods, respectively (P = 0.4388). The correlation coefficient between pre-switch and post-switch traumatic ABR_b was −0.0278 (all correlation coefficient data are summarized in Table 3). PwHA with commercial insurance had numerically higher mean ABR_b in the post-switch period, but the difference was not found to be statistically significant (0.14 pre-switch vs 0.26 post-switch, P = 0.1824; Table 2). Medicare and/or Medicaid-insured PwHA had numerically lower mean ABR_b in the post-switch period, but the difference was not statistically significant (Medicaid: 0.43 pre-switch vs 0.16 post-switch, P = 0.0560; Medicare: 0.16 pre-switch vs 0.00 post-switch, P = 0.2045). When stratified by CCI score, post-switch mean ABR_b was numerically lower, but not statistically significant (CCI = 0, 1, and 2: 0.25 pre-switch vs 0.20 post-switch, P = 0.5850; 0.14 pre-switch vs 0.00 post-switch, P = 0.1079; 1.01 pre-switch vs 0.86 post-switch, P = 0.8773, respectively), whereas the mean ABR_b for CCI score 3+ was the same (0.17 pre-switch vs 0.17, P = 0.9956).

Table 2. ABR_b outcomes.

	Pre-switch N = 131	Post-switch N = 131	P-value
Total number of bleed events, n	82	45
ABRb, mean (SD)	0.25 (0.69)	0.20 (0.72)	0.4456
ABRb by bleed type, mean (SD)
Spontaneous bleeds	0.23 (0.68)	0.18 (0.71)	0.5173
Spontaneous bleeds: joint	0.18 (0.60)	0.18 (0.71)	0.9630
Traumatic bleeds	0.02 (0.09)	0.01 (0.09)	0.4388
Traumatic bleeds: joint	0.00 (0.03)	0.01 (0.06)	0.4758
ABRb by age group, mean (SD)
0–2 years	N/A	N/A	N/A
3–6 years	0.11 (0.21)	0 (0.00)	0.1060
7–12 years	0.42 (1.00)	0.37 (1.42)	0.8739
13–17 years	0.15 (0.29)	0.05 (0.17)	0.1709
18–26 years	0.35 (0.75)	0.28 (0.73)	0.7320
> 26 years	0.21 (1.00)	0.17 (0.46)	0.6927
ABRb by payer channel, mean (SD)
Commercial	0.14 (0.34)	0.26 (0.90)	0.1824
Medicaid	0.43 (1.00)	0.16 (0.45)	0.0560
Medicare	0.16 (0.39)	0 (0.00)	0.2045
ABRb by CCI, mean (SD)
0	0.25 (0.72)	0.20 (0.74)	0.5850
1	0.14 (0.29)	0 (0.00)	0.1079
2	1.01 (0.26)	0.86 (1.49)	0.8773
3+	0.17 (0.16)	0.17 (0.30)	0.9956

ABR_b: billed annualized bleed rates; CCI: Charlson Comorbidity Index; N/A; not applicable; SD: standard deviation.

Table 3. Correlation analysis of pre- versus post-ABR_b.

	SD pre-switch	SD post-switch	SDd	Correlation
ABRb	0.6874	0.7166	0.8392	0.2860
Spontaneous ABRb	0.6824	0.7137	0.8344	0.2862
Traumatic ABRb	0.0885	0.0917	0.1292	−0.0278
Spontaneous joint ABRb	0.5966	0.7137	0.8261	0.2147
Traumatic joint ABRb	0.0282	0.0647	0.0710	−0.0152

ABR_b: billed annualized bleed rate; SD_d: standard deviation of the difference.

The majority of the post-COVID period (after March 2020) was captured post-switch, with an average follow-up of 293.7 days compared to average 12.2 days in the pre-switch period (P < 0.0001) (Table 4).

Table 4. Average follow-up days for study patients and estimate of follow-up time after March 2020 between pre- and post-switch patients.

Exploratory outcomes	Pre-switch N = 131					Post-switch N = 131					P-value
	Mean/N	SD	Median	Range (IQR)	Min, max	Mean/N	SD	Median	Range (IQR)	Min, max
Average follow-up days for study patients	978.37	555.03	930	867 (481, 1348)	205, 2308	522.26	191.36	537	303.5 (370, 674)	209, 884	<0.0001
Average follow-up days after March 2020	12.2	34.76	0	0 (0, 0)	0, 156	293.72	118.43	364	78.5 (287, 365)	0, 365	<0.0001

IQR: interquartile range; max: maximum; min: minimum; SD: standard deviation.

There were 73 (89%) and 42 (93%) spontaneous bleeds identified in the pre- and post-switch periods, respectively (Table 5). Of the 73 spontaneous bleeds in the pre-switch period, the majority (64 bleeds [88%]) were related to joint bleeds, while 100% of the spontaneous bleeds identified in the post-switch period were joint related. The correlation coefficient between pre-switch and post-switch spontaneous joint ABR was 0.2147. Of the bleeds identified in the pre- and post-switch periods, 9 (11%) and 3 (7%) were traumatic, respectively (Table 5). The correlation coefficient between pre-switch and post-switch traumatic joint ABR was −0.0152 (Table 3). There were 2 (3%) spontaneous intracranial bleeds pre-switch and zero post-switch; there were no traumatic intracranial bleeds before or after switch (Table 5).

Table 5. Bleeding events by type and site.

	Pre-switch N = 131		Post-switch N = 131		P-value
	Number (spontaneous and traumatic – primary)	% of bleeds	Number (spontaneous and traumatic – primary)	% of bleeds
All bleeds, n	82	N/A	45	N/A	0.9491
Spontaneous bleeds, n
All	73	100%	42	100%	0.9505
Joint	64	88%	42	100%	0.9105
Soft tissue (muscles)	1	1%	0	0%	N/A
Intracranial	2	3%	0	0%	N/A
Gastrointestinal	2	3%	0	0%	N/A
Genitourinary	0	0%	0	0%	N/A
Intracavity	0	0%	0	0%	N/A
Perinatal	0	0%	0	0%	N/A
Facial	4	5%	0	0%	N/A
Other sites	0	0%	0	0%	N/A
Traumatic bleeds, n
All	9	100%	3	100%	N/A
Joint	2	22%	2	67%	N/A
Soft tissue (muscles)	2	22%	0	0%	N/A
Intracranial	0	0%	0	0%	N/A
Gastrointestinal	0	0%	0	0%	N/A
Genitourinary	0	0%	0	0%	N/A
Intracavity	1	11%	0	0%	N/A
Perinatal	0	0%	0	0%	N/A
Facial	1	11%	0	0%	N/A
Other sites	3	33%	1	33%	N/A

N/A: not applicable.

4. Discussion

This study evaluated the baseline and clinical characteristics and change in ABR_b among male PwHA who switched from FVIII prophylaxis to emicizumab treatment through a retrospective analysis of claims data from an APCD.

Although several real-world evidence studies have examined outcomes in patients treated with emicizumab, information on the effects of switching from FVIII to emicizumab prophylaxis in the non-inhibitor population is limited [14,18–21]. However, in a recent observational retrospective study of 673 people from the United Kingdom with severe HA without inhibitors, the switch from FVIII to emicizumab prophylaxis was associated with significantly improved bleed control in all age groups (P < 0.001) [22]. In our study (in the United States), no significant differences were observed for mean ABR_b values among all age groups and payer channels, suggesting that switching from FVIII to emicizumab may not provide any incremental benefit to patients receiving prophylactic care for HA. These studies, however, are not directly comparable due to differences in patient demographics and clinical characteristics, such as age and bleed sites.

Switching from FVIII to emicizumab indicated improvements in ABR_b in all age groups, though this difference was non-significant. Interestingly, a recent report demonstrated that the median age of adults receiving emicizumab prophylaxis who experienced spontaneous bleeding was older compared with other adults in the study, suggesting that the frequency of bleed type may change with increasing age [14]. Furthermore, a recent longitudinal prospective study of 70 PwHA (aged 1 month to 74.9 years, with and without inhibitors) compared the occurrence of breakthrough bleeding in these patients, from emicizumab initiation, and reported that the odds of bleeding while on emicizumab increased by a factor of 1.029 (P = 0.034) for every 1 year of age [18].

ABR_b worsened post-switch in the commercial channel and downwards post-switch in both the Medicare and Medicaid channels, although the observed differences in this study were not statistically significant, suggesting that switching might not confer additional benefit to either group. It is noteworthy, however, that commercially insured patients may have better access to healthcare resources and could therefore be more likely to seek care for minor bleeds (not including patients with access to hemophilia treatment centers) [23]. Additionally, measures of access to care for Medicaid patients worsened in several states during the COVID-19 pandemic, which may have contributed to a reduction in Medicaid patients with HA seeking care for less severe bleeds [24]. Further research is warranted to better understand the potential differences in bleed events in these populations before and after switching from FVIII to emicizumab prophylaxis.

For PwHA, joint bleeds are among the most clinically significant bleed types, as they can lead to long-term arthropathy, which can persist in adults despite adequate hemostasis [5,14,25]. Most bleeds experienced by PwHA are into joints [8,26], and our results corroborate these findings (Table 5). Furthermore, hemarthrosis (hemorrhage into a joint) is the most common site for bleeding in ambulatory patients, representing up to 80% of hemorrhages [27]. A recent study reported that the risk of bleeding persists despite emicizumab therapy, especially in PwHA with target joints and in older patients [18]. In our study, 88% of spontaneous bleeds prior to switching to emicizumab were joint bleeds (64/73). Although 100% (42/42) of spontaneous bleeds post-switch were joint bleeds, the difference was not significant (P = 0.9105) (Table 5).

A retrospective review that analyzed the rate and management of breakthrough bleeding events in pediatric PwHA with and without inhibitors receiving emicizumab prophylaxis reported a total of 29% (15/51) of minor bleeds that resolved spontaneously or with antifibrinolytics [28]. As such, it is possible that bleeds that may have occurred and been resolved spontaneously in the PwHA included in our study might not have been reported and thus not captured in the emicizumab group by the database. It is also possible that the different routes of administration for emicizumab and FVIII (subcutaneous vs intravenous), the ability to self-administer, and bleed severity may have factored into individual patient’s decisions to seek medical care for bleeding events. Additionally, due to inherent differences in the pharmacokinetics of emicizumab and FVIII, patients on emicizumab may be advised to wait and see if minor bleeds resolve, whereas patients on FVIII prophylaxis who experience a bleed are usually advised to seek treatment immediately due to the shorter half-life and faster coagulatory effects of FVIII compared to the slower, longer-term effects of emicizumab.

Differences in pre- and post-switch ABR_b (mean ABR_b = 0.25 pre-switch vs 0.20 post-switch; P = 0.4456) were likely attributable to differences in follow-up times: the pre-switch follow-up period was longer on average than the post-switch follow-up period (978 vs 522 days, respectively). Therefore, the respective ABR_b was similar. As bleeding events are known to increase in frequency with age in people with HA, it is possible that the young age of participants in this study (the majority of patients were under the age of 27 years and the mean age was 28.9 years) may account for the relatively low bleed rates seen in this study. Moreover, participants in this study may have been reluctant to seek medical care for minor/non-emergency bleeds during periods of high COVID-19 prevalence; this may have affected the post-switch groups more, as evidenced by the lower mean follow-up days during the post-COVID period as compared to the overall post-switch period (328.97 and 529.6 days, respectively; Table 4). Furthermore, it is possible that participants in this study had recurring bleeds in target joints post-switch that may resolve with time, and longer-term analyses of post-emicizumab bleed rates are warranted as more long-term data become available.

These results are in line with previous findings, which utilized data from the IQVIA PharMetrics® Plus US administrative claims database, where there were also no significant changes in ABR_b pre- and post-switch. Although there were no significant changes in ABR_b pre- and post-switch in this analysis, previous studies, which utilized Bayesian inferences to assess the difference in ABR_b and total cost of care, estimated a higher probability of increased ABR_b and a significant increase in total cost of care of prophylaxis after switching from FVIII to emicizumab (note: the database used for this study did not provide cost data) [29].

The limitations of this study reflect those inherent in claims analyses: lack of detail regarding whether medications were used as dispensed, and no records for medications administered in hospitals. Variation in coding and billing practices may have led to inconsistencies in diagnosis and procedure codes. Additionally, it was unknown in this study how many patients were receiving standard half-life or extended half-life FVIII prophylaxis in the pre-switch period. There may be a risk of misclassification bias due to lack of disease-specific and clinical parameters in administrative claims, such as inhibitor status, HA severity, FVIII trough target, weight, and differentiation between prophylaxis and on-demand treatment. While on-demand patients were excluded during the identification period, it is additionally possible that participants receiving frequent on-demand FVIII therapy for bleeding who then switched to emicizumab prophylaxis may have met the pattern criterion for continuous billing. In the absence of diagnosis codes, inhibitor patients were operationally defined by the usage of Bethesda/Nijmegen assays (for the detection of FVIII inhibitors), bypassing agents, or immune tolerance induction. Some low-titer inhibitor patients could have been treated by higher concentration of FVIII and might not have been identified as inhibitor patients in this analysis.

Potential biases should be acknowledged; for example, differences in approval times between FVIII and emicizumab products may lead to history, maturation, and selection bias between pre- and post-switch cohorts. Moreover, the use of administrative claims data also gives rise to inherent information bias. Regression to mean bias could affect some of the outcomes of interest. Outliers, censoring, and unmeasured covariates could also complicate interpretation of results. It should also be noted that the COVID-19 virus was more prevalent during the post-switch period, potentially confounding post-switch ABR_b as patients may have been reluctant to seek in-person care for non-emergency bleeds. Lastly, differences in individual participant adherence to FVIII prophylaxis pre-switch or emicizumab post-switch could not be accounted for in this study. Annual usage and ‘reason for switching’ as a metric for compliance should be considered in future studies.

5. Conclusions

These findings suggest similar rates of bleeding after switching from FVIII to emicizumab prophylaxis in non-inhibitor PwHA. Importantly, this study highlights the importance of continuous monitoring of bleed rates in all age groups after switching patients from FVIII treatments and of implementing individualized treatment regimens. Further analyses of ABR_b by age and payer channel are also warranted as, in contrast to FVIII therapy, long-term data on the prevention of HA sequelae for emicizumab, including the development of joint disease in children, are not available. A more comprehensive understanding of the impact of switching treatment will help develop improved individualized strategies for managing different PwHA populations.

Declaration of Interests

M Escobar receives honoraria from Bayer, BioMarin, CSL Behring, Genentech/Roche, Hemabiologics/LFB, Kedrion, NHF, Novo Nordisk, Pfizer, Sanofi, Takeda, and uniQure. The University of Texas participates in clinical trials sponsored by CSL Behring, Bayer, Genentech/Roche, Novo Nordisk, Sanofi, Takeda, and uniQure. M Bullano, AG Mokdad, J Caicedo, BG Schultz, and Q Fan are employees and stockholders of Takeda Pharmaceuticals U.S.A., Inc., Lexington, MA. S Verma is an employees of STATinMED, which was contracted by Takeda for work on this study. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Reviewer Disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author contributions

Conception and design: M Bullano, J Caicedo, AG Mokdad, BG Schultz, S Verma. Analysis/interpretation of data: J Caicedo, M Bullano, BG Schultz, Q Fan, AG Mokdad, S Verma, M Escobar. Drafting the manuscript: Q Fan, AG Mokdad, S Verma. Revising the manuscript critically for intellectual content: J Caicedo, M Bullano, BG Schultz, Q Fan, AG Mokdad, S Verma, M Escobar.

Acknowledgments

Writing and editorial assistance for the development of this manuscript was provided by Ari Simenauer and Zela Keuylian (employees of Parexel), funded by Takeda Pharmaceuticals U.S.A., Inc., Lexington, MA.

Data availability statement

The data underlying this article were provided by STATinMED via confidential license agreement, which has legally enforceable usability restrictions that preclude our making the data available publicly.

Additional information

Funding

This work was supported by Takeda Pharmaceuticals U.S.A., Inc.