Post‑Marketing Surveillance of the Safety and Effectiveness of Emicizumab in Japanese Patients With Congenital Haemophilia A With Inhibitors

ABSTRACT

Introduction

The bispecific monoclonal antibody emicizumab was approved for prophylactic treatment of congenital haemophilia A (HA) in Japan in 2018.

Aim

To monitor long‐term safety and effectiveness of emicizumab, including appropriate concomitant use of bypassing agents (BPAs), in Japanese patients with congenital HA with inhibitors who initiated emicizumab within 1 year of availability.

Methods

This all‐case post‐marketing surveillance (PMS) study was conducted between May 2018 and January 2023, in patients of all ages and HA severities. Patients were registered retrospectively after emicizumab initiation or prospectively at the time of initiation. Adverse events (AEs) and adverse drug reactions (ADRs), including thromboembolic events (TEs) and thrombotic microangiopathy (TMA), and their association with BPA use, were examined. FVIII inhibitor levels and annualised bleeding rates (ABRs) were evaluated.

Results

In total, 134 patients were included in the analysis. Mean (standard deviation) duration of emicizumab treatment was 145.5 (34.8) weeks. Overall, 112 AEs occurred in 47 patients (35.1%) and 22 ADRs occurred in 12 patients (9.0%); 8 ADRs (36.4%) were serious. No TEs/TMAs associated with concomitant use of BPAs occurred. FVIII inhibitor levels remained stable or decreased for 34 patients (55.7%). Sixty‐eight patients (50.7%) received BPAs; none received activated prothrombin complex concentrate. Mean (95% confidence interval) model‐based all bleed ABR and treated bleed ABR were 1.4 (0.9–2.3) and 1.3 (0.8–2.1), respectively.

Conclusion

This study comprises 3 years’ PMS for emicizumab. No TEs/TMA associated with BPAs occurred. Results further support the safety and effectiveness of emicizumab in Japanese patients with congenital HA with inhibitors.

1. Introduction

Congenital haemophilia A (HA) is a hereditary bleeding disorder caused by blood coagulation factor (F)VIII deficiencies, characterised by recurrent bleeding into joints and muscles [1]. The standard of care for HA is prophylaxis to prevent/reduce bleeding [2]; however, approximately 30% of patients with severe HA receiving FVIII replacement therapy develop inhibitors [3], limiting the effectiveness of treatment and posing an additional therapeutic challenge. Consequently, non‐factor therapies for HA were developed to mitigate inhibitor development risk and provide effective prophylaxis options that are unaffected by FVIII inhibitors [2].

Emicizumab, a recombinant, humanised, bispecific monoclonal antibody that mimics the cofactor function of activated FVIII and facilitates haemostasis [4], is approved for use in people with HA in multiple countries [4, 5, 6, 7]. Efficacy and safety of emicizumab have been demonstrated in clinical trials involving patients with HA with and without inhibitors, many with long‐term follow‐up periods [8, 9, 10, 11, 12, 13]. In a pooled analysis of the HAVEN 1–4 studies [14], emicizumab maintained low annualised bleeding rates (ABRs) of <1 through 24‐week treatment intervals in patients with HA with and without inhibitors. Over a median exposure of 130.3 weeks, 381/399 (95.5%) participants in the safety population reported ≥1 adverse event (AE); the most common treatment‐related AE was injection‐site reaction (n = 107, 26.8%), and most were mild. Longer follow‐up of the HAVEN 3 and 4 studies showed a decrease in mean treated bleed ABR (95% confidence interval [CI]) from 2.0 (0.23–7.15) at Weeks 1–24 to 0.9 (0.01–5.28) at Weeks 217–240 [15]. There were no thromboembolic events (TEs) or thrombotic microangiopathy (TMA) associated with activated prothrombin complex concentrate (aPCC) in clinical trials beyond those previously reported in HAVEN 1. The STASEY study provided further confirmatory evidence on the safety profile of emicizumab in patients with HA and FVIII inhibitors [16].

Investigations into long‐term safety and efficacy of emicizumab in Japan include a Phase 1/2 study of patients with severe HA with and without inhibitors receiving emicizumab prophylaxis for up to 5.8 years [17]. Treated bleed ABRs decreased for all patients and remained low at 1.25, 0.83 and 0.22 for patients receiving emicizumab at 0.3, 1 and 3 mg/kg once weekly (QW), respectively. Overall, 226 AEs were reported; most (89.4%) were mild. In a multicentre, open‐label study involving Japanese patients aged <12 years with severe HA without inhibitors (HOHOEMI) [18], after ≥24 weeks of emicizumab treatment, treated bleed ABRs were low, and many patients (53.8%) reported no treated bleeds. One treatment‐related AE was reported, and no TEs/TMA occurred.

Emicizumab was approved in Japan in 2018 for routine prophylaxis in people with HA following the standard dosing regimens used in global Phase 3 studies, that is, 3 mg/kg QW for 4 weeks followed by maintenance doses of 1.5 mg/kg QW, 3 mg/kg every 2 weeks (Q2W) or 6 mg/kg every 4 weeks (Q4W) [6]. The main safety concern of emicizumab is the risk of TE/TMA development related to concomitant use of aPCC at cumulative doses of >100 U/kg/24 h for ≥24 h [8, 19]. Global safety database reports for emicizumab have allowed for continued pharmacovigilance of TEs/TMA [20]. In Japan, the regulatory authority requires that post‐marketing surveillance (PMS) is conducted to evaluate drug safety profiles in real‐world clinical settings. Some studies may analyse the safety information of all patients who have used the drug, whilst others may also assess its effectiveness. This PMS study (UMIN000032515) monitored long‐term safety and effectiveness of emicizumab, and appropriate concomitant use of bypassing agents (BPAs), including recombinant activated factor VII (rFVIIa), plasma‐derived activated factor VII (FVIIa) and FX mixtures, and aPCC, among Japanese patients with congenital HA with FVIII inhibitors.

2. Materials and Methods

2.1. Study Design

This observational all‐case PMS study was conducted in patients with congenital HA with FVIII inhibitors receiving subcutaneous injections of emicizumab across 96 medical institutions in Japan. All treatment centres that used emicizumab for the treatment of patients with congenital HA during the study registration period from May 2018 to June 2022 were included. Patients were registered between May 2018 and June 2022, and were eligible regardless of age, sex and HA severity. Patients initiated emicizumab between the launch of the drug in Japan in May 2018 and June 2019. Those who had previously received emicizumab in clinical trials were included and registered from the date of switch to the marketed product following its launch.

The surveillance period was May 2018–January 2023, and patients were monitored for up to 3 years from baseline, which was defined as the start of emicizumab treatment.

2.2. Study Objectives

Primary objectives were to evaluate the incidence of AEs and adverse drug reactions (ADRs) during emicizumab treatment, including TEs/TMA and any association with BPA use, and any incidence of inappropriate bleed control from misinterpretation of coagulation test results due to emicizumab interference with activated partial thromboplastin time (aPTT); and bleeding events and associated treatment with haemostatic agents. Secondary objectives were to measure FVIII inhibitor titres over time and calculate ABRs following emicizumab initiation.

The following endpoints were also evaluated: emicizumab treatment duration and dose; presence and site of target joints; previous prophylactic haemostatic agent use; number of bleeding events ≤6 months prior to baseline; use of coagulation factor products for bleed treatment ≤6 months prior to baseline; FVIII inhibitor titre 6 months prior to baseline and at baseline; and any complications, including renal impairment and hepatic dysfunction.

2.3. Data Collection

A centralised registration system was used, whereby physicians completed a registration form on behalf of their patients and faxed it to the registration centre. Patients who received emicizumab prior to this study were registered retrospectively. Information was collected using surveillance forms at 6 and 36 months after baseline. FVIII inhibitor titre measurements were conducted at study sites/independent commercial laboratories (SRL Diagnostics). Information collected after the observation period was also included in the analysis. Routine testing for anti‐drug antibodies (ADAs) was not conducted in this study. However, plasma emicizumab concentration measurements were available for cases where ADA development was suspected, such as in patients with prolonged aPTT or increased bleeding frequency during treatment.

2.4. Data Analyses

All registered patients deemed eligible formed the safety analysis group (SAG), since pharmaceutical authorities must evaluate the safety of all patients who have received a drug after its marketing launch as part of Japanese PMS. Patients who had newly initiated emicizumab formed the effectiveness analysis group (EAG); patients who had continued from clinical studies, and had already received emicizumab for a long time, were excluded from the evaluation of its effectiveness. ADRs were defined as events where a causal relationship with emicizumab could not be ruled out, which is standard for post‐marketing data in Japan. Bleeding events were recorded for all SAG patients; however, to allow for real‐world application of results, ABRs were only calculated for the EAG. To reflect real‐world clinical practice, the definition of a bleed was at the discretion of the physician; no confirmatory tests were mandated by the study protocol. Bleeding caused by congenital HA was not considered an AE. Treated and untreated bleed ABRs (bleeds/year) and 95% CIs were calculated using a negative binomial regression model, with total emicizumab treatment duration (years) as an offset variable and number of bleeding events (<9/≥9) in the 6 months prior to baseline as a covariate. As information on bleeding events and the use of coagulation factors was collected in the 6 months prior to baseline, model‐based ABRs are based upon this information, which is limited to this time period. Basic statistics (number of bleeds after the start date of first emicizumab treatment/total duration of emicizumab treatment [days] × 365.25) annualised the number of bleeds. FVIII inhibitor titre was defined as the starting FVIII inhibitor titre measured from 0 to 30 days prior to baseline.

3. Results

3.1. Patient Characteristics

Overall, 138 patients were registered across 92 medical institutions; 134 formed the SAG and 101 formed the EAG (Figure 1). Overall, 126 patients were registered after initiating emicizumab, and eight were registered at the time of initiation. Thirty‐three patients were excluded from the EAG due to previous emicizumab treatment in clinical trials. All SAG patients were male, and the median (min–max) age at the start of emicizumab treatment was 17.0 (0–78) years (Table 1). Most had severe congenital HA (n = 116; 86.6%); nine patients (6.7%) each had mild/moderate congenital HA. Mean (standard deviation) treatment duration was 145.5 (34.8) weeks; median (min–max) treatment duration was 157.0 (5–189) weeks. A total of 70 (50.7%) participants had data collected for more than 3 years. In total, 61 (45.5%) patients had complications, which were defined as conditions or symptoms present at baseline.

FIGURE 1.

Patient attrition.

TABLE 1.

Patient characteristics.

Safety analysis group	N = 134
Male, n (%)	134 (100)
Total treatment duration (weeks)
Median (Q1–Q3)	157.0 (156–157)
Age (years)
Median (Q1–Q3)	17.0 (4.3–48.8)
Age groupings (years), n (%)
<2	21 (15.7)
≥2–<18	48 (35.8)
≥18–<65	50 (37.3)
≥65	15 (11.2)
Weight (kg)
Number of patients, n (%)	132 (98.5)
Median (Q1–Q3)	50.2 (16.7–66.0)
Unknown/unanswered, n (%)	2 (1.5)
Severity of HA, n (%)
Mild	9 (6.7)
Moderate	9 (6.7)
Severe	116 (86.6)
Measurement of FVIII inhibitor titre within 6 months prior to emicizumab treatment, n (%)
No	42 (31.3)
Yes	79 (59.0)
Unknown/unanswered	13 (9.7)
Measurement of FVIII inhibitor titre after initiating emicizumab treatment, n (%)
No	12 (9.0)
Yes	120 (89.6)
Unknown/unanswered	2 (1.5)
Presence of target joints, n (%) a
No	99 (73.9)
Yes	35 (26.1)
Previous treatment with prophylactic BPAs, n (%)
No	83 (61.9)
Yes	51 (38.1)
Emicizumab treatment continued from a clinical trial, n (%)
No	101 (75.4)
Yes	33 (24.6)
All bleeding (≤6 months before emicizumab treatment initiation), n (%)
None	53 (40.0)
Yes	77 (57.5)
Use of coagulation factor product to treat bleeding	71 (92.2) b
Unknown/unanswered	4 (3.0)
Effectiveness analysis group	N = 101
ITI treatment received during the study in the effectiveness analysis group, n (%)
No	85 (84.2)
Yes	16 (15.8)

Abbreviations: BPA, bypassing agent; F, factor; HA, haemophilia A; ITI, immune tolerance induction; Q, quartile.

^a Multiple answers were allowed.

^b Percentage of 77 patients reporting bleeds.

Of the 101 EAG patients, most (n = 83) were on a maintenance regimen of 1.5 mg/kg QW, 15 were on 3 mg/kg Q2W and three were on 6 mg/kg Q4W (Figure S1). Of the 33 patients who were previous clinical trial participants, 20 remained on the same regimen of 1.5 mg/kg QW. The other 13 patients changed regimen: nine to another approved emicizumab regimen, 3 mg/kg Q2W (n = 8) or 6 mg/kg Q4W (n = 1) (Figure S1).

At survey form completion, 116 patients in the SAG (86.6%) were still on emicizumab; 18 (13.4%) discontinued treatment due to reasons including non‐fatal AEs (n = 1; 0.7%), lack of efficacy (n = 1; 0.7%), death (n = 11; 8.2%) and other (n = 4; 3.4%). One patient (0.7%) withdrew from the study due to failure to visit and a change in medical institution.

3.2. Safety

3.2.1. Incidence of AEs and ADRs, Including TEs and TMA

Overall, 112 AEs occurred in 47 patients (35.1%) in the SAG; 51 were serious and 61 were non‐serious (Table 2). The 51 serious AEs were reported in 28 (20.9%) patients, and included cellulitis (n = 2), aspiration pneumonia (n = 1), seizure (n = 2), multiple organ dysfunction syndrome (n = 1) and ganglioneuroblastoma (n = 1) (Table S1).

TABLE 2.

Occurrence of AEs and ADRs in the safety analysis group (N = 134).

	N = 134
Number of patients with ≥1 AE, n (%)	47 (35.1)
Total number of AEs	112
Number of patients with ≥1 SAE, n (%)	28 (20.9)
Total number of SAEs	51
Total number of AEs related to use of BPAs a
Thromboembolic event	0
Thrombotic microangiopathy	0
Significant bleeding associated with inappropriate control of bleeding due to the interference of emicizumab with coagulation test results	0
Number of patients with ≥1 ADR, n (%)	12 (8.95)
Total number of ADRs	22
Number of patients with ≥1 serious ADR, n (%)	4 (2.99)
Total number of serious ADRs	8
Number of patients with positive test for ADA, n (%)	1 (0.75)

Abbreviations: ADA, anti‐drug antibody; AE, adverse event; ADR, adverse drug reaction; aPCC, activated prothrombin complex concentrate; BPA, bypassing agent; F, factor; SAE, serious adverse event.

^a Includes recombinant activated FVII, plasma‐derived activated FVII and FX mixtures and aPCC.

Twenty‐two ADRs were reported in 12/134 patients (9.0%; Table 2). ADR incidence was highest in patients aged ≥65 years (n = 15); three patients (20.0%) in this group reported four ADRs. Fourteen non‐serious ADRs occurred in nine patients (6.7%), including dizziness, head discomfort, hypertension, injection‐site erythema, one case of aPTT prolonged and one case of aPTT shortened. In the case of prolonged aPTT, the aPTT value at the time of the ADR occurring was 40 s and subsequently decreased to 30 s. Emicizumab treatment continued, although this patient's overall treatment adherence is unknown. Four injection‐site reactions (4/134 [3.0%]) were reported as non‐serious ADRs. Eight serious ADRs occurred in four patients (3.0%), including ganglioneuroblastoma, diabetes insipidus, hypopituitarism, obesity, death, lacunar infarction and drug‐specific antibody present (Table S2). In the case of the presence of a drug‐specific antibody, plasma emicizumab concentration was measured, as aPTT tended to be prolonged in this patient (approximately 100 s at the time of discontinuation of emicizumab) and the frequency of bleeding increased during emicizumab treatment. Since the emicizumab concentration was found to be below the detection limit, emicizumab was discontinued, and treatment was changed to BPAs. Thereafter, to confirm the presence/absence of ADA expression, ADA was measured as joint research between the manufacturer and the medical institution, and a positive result was determined. Regarding the death, no information, such as causal relationship, onset date, outcome date or treatment, could be obtained. A total of 11 deaths were reported; in two cases, there were no AEs leading to death reported, and investigator assessments of these deaths were not available. In the remaining nine deaths, a causal relationship with emicizumab was ruled out in all but one death, for which insufficient information was available (Table S3).

No TEs/TMA were associated with concomitant BPA use (rFVIIa or plasma‐derived FVIIa and FX mixtures). Two TEs (lacunar infarction) were reported in one patient aged 73 years; this patient did not receive concomitant BPAs and had prior risk factors, including hypertension, diabetes mellitus, advanced age, drinking alcohol and smoking, but the investigator determined a causal relationship with emicizumab could not be excluded.

3.3. Bleeding Events

3.3.1. Bleeds and Associated Treatment

Seventy‐nine SAG patients (59.0%) reported 319 bleeding events following emicizumab initiation (Figure S2); 288 were treated and 31 were untreated. Most traumatic bleeds were subcutaneous (65 events), and most non‐traumatic bleeds were intraarticular (60 events) (Further information on non‐traumatic bleeds is included in Table S4). Fifty‐three SAG patients (39.6%) reported no bleeds.

Within the SAG, 68 patients (50.7%) received BPAs for bleeding while on emicizumab: 68 received rFVIIa, five received plasma‐derived FVIIa and FX mixtures, and none received aPCC (Figure S2). Twenty‐nine patients (21.6%) received haemostatic agents other than BPAs for bleeding, predominantly tranexamic acid (n = 23) and carbazochrome sodium sulfonate hydrate (n = 8).

Overall, 36 patients (26.9%) received BPAs (rFVIIa or plasma‐derived FVIIa and FX mixtures) for events other than bleeding: single‐use prophylaxis (n = 9) and surgical intervention (n = 26). Thirty patients (22.4%) received haemostatic agents other than BPAs for events other than bleeding: single‐use prophylaxis (n = 12), surgical intervention (n = 28) and immune tolerance induction (ITI) therapy (n = 16). aPTT and use of coagulation factor products were evaluated for all reported bleeding events, and it was confirmed that there was no instance of misinterpretation.

3.3.2. ABR in the EAG

In the 101 EAG patients, the mean (95% CI) model‐based ABR for all bleeds was 1.4 (0.9–2.3). Mean (95% CI) model‐based treated bleed ABRs were 1.3 (0.8–2.1) (Table 3).

TABLE 3.

ABR in patients forming the effectiveness analysis group (n = 101).

	Model‐based a ABR (95% CI)
All bleeds	1.4 (0.9–2.3)
Traumatic bleeds	0.6 (0.3–1.1)
Spontaneous bleeds	0.8 (0.4–1.5)
All treated bleeds b	1.3 (0.8–2.1)
Traumatic bleeds	0.6 (0.3–1.1)
Spontaneous bleeds	0.8 (0.4–1.5)
Bleeds treated with coagulation factor product c	1.2 (0.7–2.1)
Bleeds treated with supportive haemostatic product	1.2 (0.7–2.1)

Abbreviations: ABR, annualised bleeding rate; BPA, bypassing agent; CI, confidence interval; F, factor.

^a Estimation was performed using a negative binomial regression model with the total duration of emicizumab treatment as an offset variable, and the number of bleeds (<9 bleeds, ≥9 bleeds) during the period from 6 months before the start of emicizumab treatment to the start of emicizumab treatment as a covariate.

^b Includes those treated with coagulation factor product and those treated with supportive therapy (carbazochrome sodium sulphonate hydrate, thrombin, tranexamic acid, FXIII, rurioctocog alfa, efraloctocog alfa or octocog beta).

^c Includes BPAs, FVIII and other products containing coagulation factors; excludes other supportive therapy.

3.4. Surgery

Thirty patients (22.4%) received surgical intervention for events other than bleeding (Table S5): 19 patients had one surgical procedure, three had two surgical procedures, six had three surgical procedures and two patients underwent ≥4 surgical procedures. Major surgeries performed were total knee arthroplasty, knee prosthesis removal, removal of maxillary sinus tumour mass, definite repair of the incarcerated inguinal hernia and removal of intraperitoneal abscess. Most minor surgeries were the removal of a central venous access device (Table S5). The use of haemostatic products during surgical procedures is also described in Table S5.

3.5. FVIII Inhibitor Levels

Sixty‐one patients had FVIII inhibitor levels available at baseline and at last measurement. At baseline, the median (Q1–Q3) FVIII inhibitor level for these patients was 14.0 (4.8–44.0) BU/mL; at the last measurement, it was 2.1 (0.0–6.4) BU/mL (Table S6). Eighteen (29.5%) of the 61 patients were inhibitor negative, defined as a titre <0.5 BU/mL, at their last measurement. Of these, nine patients received rFVIIa after achieving inhibitor negativity, while one received a FVII/FX combination product and two received recombinant FVIII. A total of 34 (55.7%) of the 61 patients were noted to have stable and decreasing FVIII inhibitor levels (Figure S3).

Changes in FVIII inhibitor test values are shown in Figure S4. One patient had a transient inhibitor level elevation immediately after initiation of efraloctocog alfa (efmoroctocog alfa) for ITI therapy, but FVIII inhibitor levels ultimately decreased.

3.6. ITI Treatment

Overall, 16 patients received ITI treatment during the study (Table 1). The mean model‐based ABR (95% CI) for patients who received ITI therapy was 2.2 (1.1–4.5). Additional data regarding haemostatic products that patients who underwent ITI treatment received are described in Table S7.

4. Discussion

This study assessed the safety and effectiveness of long‐term emicizumab use among Japanese patients with congenital HA with FVIII inhibitors. Most patients received emicizumab at 1.5 mg/kg QW maintenance dose. After a mean treatment duration of almost 3 years, 12 patients (9.0%) reported 22 ADRs; eight were serious.

The real‐world safety data collected in this PMS study further substantiate those from the STASEY safety study and the long‐term Phase 1/2 trial in Japanese patients [16, 17], as well as Phase 3 global clinical trials [14]. The Phase 3b, multicentre, single‐arm STASEY study evaluated safety (primary endpoint) and efficacy of emicizumab in patients aged ≥12 years with HA with FVIII inhibitors [16]. In total, 193 patients received emicizumab at 1.5 mg/kg QW maintenance dose for 2 years. Emicizumab was well tolerated after a median (min–max) treatment duration of 103.1 (1.1–108.3) weeks. Overall, 163 (84.5%) patients experienced 800 AEs; 35 (18.1%) reported emicizumab‐related AEs; all but two events (catheter‐site abscess and postprocedural haematoma) were non‐serious. No new or unexpected safety concerns were identified in the present study, which included patients of any age, any disease severity, and those being treated with any of the three approved dosing regimens. Most AEs were minor, and no serious events related to concomitant administration of BPAs (rFVIIa or plasma‐derived FVIIa and FX mixtures) occurred. Incidence of AEs tended to be greater in patients aged ≥65 years (n = 12/15), which is consistent with the previous emicizumab global safety database report [20]. Compared with previous studies, including STASEY, where 2 (1%) participants died, a higher number of deaths occurred (11 [8%]); however, a causal relationship to emicizumab was ruled out in 8/11 cases. Two possible contributing factors include the high number of registered elderly patients aged >65 years (15/134 [11.2%]), and those with complications (61 [45.5%]), including hepatic dysfunction (15 [11.2%]), who are typically excluded from clinical trials due to potential interference with study conduct or result interpretation [8, 9, 16]. The treatment adherence of the patient who experienced the prolonged aPTT ADR was not recorded during the study; therefore, this patient may have been non‐adherent to treatment, potentially contributing to this event. The evaluation of aPTT and coagulation factor use meant there was no misinterpretation of bleeding events in this study.

The guidance recommended that physicians administer the lowest effective dose of BPAs, avoiding aPCC unless all alternatives were unsuccessful [21]. Single‐use BPA prophylaxis was intended for administration prior to exercise. Although the specific reason for each administration was not recorded, it is presumed that all instances were intended to mitigate potential traumatic bleeding. Incidence of TEs/TMA in this PMS study was similar to many global clinical trials following this guidance [8, 10, 11, 16, 17], as no patients received concomitant treatment with aPCC; however, two TEs were noted in one patient (lacunar infarction). This patient did not receive concomitant treatment with BPAs, but had many risk factors for cerebrovascular disease, including hypertension and diabetes mellitus, in addition to old age. Although a causal relationship could not be excluded, it is likely that these events were independent from emicizumab treatment. Nevertheless, the low incidence of TEs/TMA confirms that the benefit–risk profile of emicizumab in Japanese patients is similar to other patient populations, provided that guidance on aPCC use is followed.

In a pooled HAVEN 1–4 analysis [14] conducted in patients with HA with and without inhibitors over a median (min–max) efficacy period of 120.4 weeks (89–164), the mean (95% CI) model‐based treated bleed ABR was 1.40 (1.10–1.70), which is similar to this study (1.31 [0.81–2.12]). While the HAVEN data are from a clinical trial setting, real‐world data from an analysis of emicizumab‐treated patients with HA with FVIII inhibitors by the UK Haemophilia Centre Doctors' Organisation showed a mean (95% CI) treated bleed ABR of 0.32 (0.18–0.58) over a median follow‐up of 168 weeks [22]. Additionally, data from patients with HA enrolled in the Canadian Hemophilia Bleeding Disorders Registry, many (55.5%) of whom had current/historical FVIII inhibitors, found that emicizumab initiation was associated with a significant decrease in mean (95% CI) ABR, from 5.31 (3.79–7.45) to 0.87 (0.58–1.32) (p < 0.0001). Taken together, these data demonstrate no major differences in ABRs for patients on emicizumab in this PMS study, and previous clinical trials and real‐world studies indicate that emicizumab is effective in maintaining low ABR in routine Japanese clinical practice.

In the aforementioned long‐term Japanese Phase 1/2 study, FVIII inhibitor titres decreased in most patients receiving emicizumab [16, 21]; one patient had a transient increase after receiving repeated administrations of aPCC, which may contain small amounts of FVIII [17]. This study confirms these results, as 18/61 (29.5%) patients who had FVIII inhibitor measurements were inhibitor negative at their final measurement, and 34/61 (55.7%) patients were noted to have stable or decreasing FVIII inhibitor levels. This was due to the use of rFVIIa as the initial BPA during emicizumab treatment rather than aPCC. One of the 16 patients who received a FVIII product for ITI experienced an increase in FVIII inhibitor titre following administration of efraloctocog alfa, but this ultimately decreased.

Complementary to this study, which primarily focuses on safety and effectiveness, several ongoing studies are providing data on additional outcomes related to emicizumab treatment. AOZORA is a multicentre, open‐label Phase 4 clinical trial evaluating long‐term safety of emicizumab and effects on joint health in Japanese paediatric patients with severe HA without inhibitors [23]. A 3‐year interim analysis of AOZORA demonstrated efficacy and safety of emicizumab, with maintained/improved joint health in all 10 patients previously enrolled in HOHOEMI [24]. In the final analysis of the TSUBASA study, emicizumab showed efficacy in maintaining bleed control during physical activity among Japanese patients with HA, with 97.3% of patients experiencing no bleeds associated with exercise [25]. Future findings from these trials will supplement safety and efficacy data accumulated from previous studies and post‐marketing analyses of emicizumab in the wider Japanese HA population.

4.1. Strengths and Limitations

A key strength of this study is the large single ethnic group patient population, who were registered regardless of age, HA severity and overall health. However, the study lacked a control group and could not account for treatment adherence and consistency of data collection by participating physicians. The nature of the study also meant that patients could only be monitored until the end of the observation period. Furthermore, in routine clinical practice, patients may not report minor bleeding events; therefore, compared with trials using electronic Patient‐Reported Outcome (ePRO) applications to record bleeding events, these results may underestimate the number of bleeds and ultimately ABR. Additionally, FVIII inhibitor levels were not available for all patients in the study during the observation period, as measurements were taken as per usual clinical practice; conclusions drawn from the FVIII inhibitor data may therefore be limited. Finally, since information on the status of bleeding and the presence or absence of coagulation factor administration was collected from 6 months prior until initiation of emicizumab, model‐based ABRs are based upon this information, which is limited to this time period.

Nevertheless, this study reports clinical data, encompassing all patients who initiated emicizumab treatment within 1 year of its launch in Japan, with a mean follow‐up time of almost 3 years, allowing greater representation of routine clinical practice and real‐world effectiveness.

5. Conclusion

This 3‐year PMS emicizumab study was conducted in an entirely Japanese population within the first year of market launch in Japan. No TEs/TMAs associated with concomitant BPA use (rFVIIa or plasma‐derived FVIIa and FX mixtures) occurred during the surveillance period. Results of this final analysis suggest that emicizumab is well tolerated in Japanese patients with congenital HA with FVIII inhibitors, with similar safety and effectiveness reported in previous clinical trials and real‐world studies.

Author Contributions

Study design: Midori Shima, Katsuyuki Fukutake, Masanori Matsumoto, Tadashi Matsushita, Yoshihiko Sugimura, Chiaki Sugita and Lyu Ji. Study conduct: Midori Shima, Katsuyuki Fukutake, Masanori Matsumoto, Tadashi Matsushita, Yoshihiko Sugimura, Chiaki Sugita and Lyu Ji. Recruitment and follow‐up of patients: Midori Shima, Katsuyuki Fukutake, Masanori Matsumoto and Tadashi Matsushita. Data collection: Midori Shima, Katsuyuki Fukutake, Masanori Matsumoto, Tadashi Matsushita, Lyu Ji, Yoshihiko Sugimura and Chiaki Sugita. Data analysis/interpretation: Chiaki Sugita, Yoshihiko Sugimura, Midori Shima, Katsuyuki Fukutake, Masanori Matsumoto, Lyu Ji, Mika Kawano and Tadashi Matsushita. All authors have reviewed and approved the manuscript prior to submission and agree to be accountable for all aspects of the work.

Ethics Statement

This study was conducted in accordance with relevant regulations in Japan (Ministerial Ordinance on Good Post‐Marketing Study Practice [GPSP], Ministry of Health, Labour and Welfare Ordinance Number 38, 23 March, 2005). The study protocol was reviewed and approved by the Japanese Pharmaceuticals and Medical Devices Agency (PMDA) prior to study initiation. The study did not undergo review by the ethics committees of the participating medical institutions, as this is not required for PMS studies according to Japanese regulations of the Act on Securing Quality, Efficacy and Safety of Products Including Pharmaceuticals and Medical Devices.

Consent

To fulfil the requirements for an all‐case PMS study set out by the Japanese regulatory authority at the time of emicizumab approval, all patients within the specified timeframe who were treated with emicizumab were registered in this study. In line with Japanese regulations of the Act on Securing Quality, Efficacy and Safety of Products Including Pharmaceuticals and Medical Devices, informed consent was not collected.

Conflicts of Interest

Midori Shima has received research funding from CSL Behring, Chugai Pharmaceutical Co., Ltd; honoraria from Chugai Pharmaceutical Co., Ltd, Sanofi, Novo Nordisk, CSL Behring; and speaker's bureau from Chugai Pharmaceutical Co., Ltd, Sanofi, Novo Nordisk and CSL Behring. Katsuyuki Fukutake has received consultancy fees from CellGenTech Inc.; research funding from Chugai Pharmaceutical Co., Ltd, Novo Nordisk, Pfizer Japan Inc., Sanofi K.K; honoraria from Bayer Yakuhin Ltd, Chugai Pharmaceutical Co., Ltd, Eiken Chemical, KM Biologics, KYORIN Pharmaceutical Co., Ltd, Sanofi K.K., Sekisui Medical, Takeda; speaker's bureau fees from Chugai Pharmaceutical Co., Ltd, CSL Behring K.K., Fujimoto, Japan Blood Products Organization, KM Biologics, Novo Nordisk, Pfizer Japan Inc., Sanofi K.K., Sekisui Medical; and reports previous employment at Tokyo Medical University. Masanori Matsumoto has received consultancy and honoraria fees from Sanofi, Takeda, Alexion; and patents and royalties from Alfresa Pharma. Yoshihiko Sugimura, Chiaki Sugita and Lyu Ji are employees of Chugai Pharmaceutical Co., Ltd. Mika Kawano is an employee and shareholder of Chugai Pharmaceutical Co., Ltd. Tadashi Matsushita has received consultancy fees from Takeda, Bayer, Novo Nordisk, Chugai Pharmaceutical Co., Ltd, Pfizer; research funding from Chugai Pharmaceutical Co., Ltd, Novo Nordisk; and honoraria fees from Takeda, Bayer, Sanofi, Chugai Pharmaceutical Co., Ltd, CSL Behring, JB, KMB, Novo Nordisk and Sysmex.

Supporting information

Table S1: Occurrence of all reported serious AEs and any non‐serious AEs that occurred in more than one patient in the safety analysis group.

Table S2: List of ADRs.

Table S3: List of AEs leading to death in the safety analysis group.

Table S4: Non‐traumatic bleeds (intraarticular and gastrointestinal) by age group.

Table S5: Content and details of surgical intervention (Physician's free text on CRF).

Table S6: FVIII inhibitor levels for patients with data available.

Table S7: Details of haemostatic products (other than bypassing haemostatic products) given to ITI treatment patients (n = 16).

Figure S1: Emicizumab dose regimens.

Figure S2: Use of bypassing agents for bleeding events.

Figure S3: Changes in FVIII inhibitor titre before and after initiation of emicizumab treatment (n = 61).

Figure S4: Change in FVIII inhibitor titre over time in patients within the effectiveness analysis group with A) FVIII inhibitor levels <40 [BU/mL] at initiation of emicizumab (n = 42) and B) FVIII inhibitor levels ≥40 [BU/mL] at initiation of emicizumab (n = 16).

Shima M., Fukutake K., Matsumoto M., et al. “Post‑Marketing Surveillance of the Safety and Effectiveness of Emicizumab in Japanese Patients With Congenital Haemophilia A With Inhibitors.” Haemophilia 32, no. 1 (2026): 135–142. 10.1111/hae.70162

Data Availability Statement

The datasets generated and/or analysed during this PMS study are not publicly available as informed patient consent was not obtained, according to Japanese regulations.