Efficacy
In HAVEN 3, the primary outcome related to the ABR ratio for treated bleeds demonstrated a statistically significant reduction in bleeding for both 1.5 mg/kg emicizumab weekly and 3.0 mg/kg emicizumab every 2 weeks compared to no prophylaxis (i.e., episodic FVIII) for patients previously treated with episodic FVIII. This reduction was also considered to be clinically meaningful, according to the clinical experts consulted by CADTH for this review. Similar efficacy findings were reported for secondary bleeding outcomes (i.e., ABR ratio for all bleeds, treated joint bleeds, treated spontaneous bleeds). The data for all bleeding outcomes in HAVEN 3 were robust, given the consistency of the primary results with various sensitivity analyses. HAVEN 3 assessed treated target joint bleeds as an “other” outcome outside the statistical testing hierarchy. Treated target joint bleeds showed consistent results with the primary and secondary bleeding outcomes. This reduction was also considered to be clinically meaningful, according to the clinical experts consulted by CADTH for this review.
The only data available for patients previously treated with FVIII prophylaxis are from a small sample (n = 48) in the HAVEN 3 study. In an intra-patient assessment, patients treated with FVIII prophylaxis in the NIS entered HAVEN 3, where they received maintenance treatment with 1.5 mg/kg emicizumab weekly. The ABR ratio was statistically significant and considered clinically meaningful by the clinical experts for the bleeding outcomes assessed (i.e., all bleeds, treated bleeds).
In the single-arm HAVEN 4 study, ABRs for patients treated with 6.0 mg/kg emicizumab every 4 weeks were generally aligned with the reductions in bleeding reported in HAVEN 3 (i.e., ABR for treated bleeds, all bleeds, treated joint bleeds, treated spontaneous bleeds, treated target joint bleeds). HAVEN 4 is the only study included in this review that evaluated the efficacy and safety of 6.0 mg/kg emicizumab every 4 weeks, which is one of the doses approved for use in adolescents and adults greater than or equal to 40 kg. Although all analyses were descriptive, and no formal hypothesis testing was performed, Health Canada concluded that “…the model based ABR (in HAVEN 4) was consistent with HAVEN 3, although slightly higher but limited by cross-study comparisons. The HAVEN 4 results are consistent with the results of HAVEN 3 and HAVEN 1, indicating the 6.0 mg/kg Q4W [every 4 weeks] dose is an acceptable maintenance dose in this patient population (greater than or equal to 12 years of age and greater than 40 kg).”29
HAVEN 3 and HAVEN 4 included several pre-specified subgroup analyses that were identified as relevant in the CADTH review protocol (bleed rate in the 24 weeks prior to enrolment [less than 9, greater than or equal to 9], age category at baseline, presence or absence of target joints, previous treatment regimen [episodic, prophylactic], and FVIII inhibitor status). For the subgroups with sufficient data, there were no clear differences observed in ABR. However, due to small sample sizes, most these subgroups could not be appropriately analyzed. This was acknowledged by the sponsor, who wrote: “…due to the small sample size, all subgroup analyses are highly sensitive to variability caused by individual patients and should be interpreted with caution.”10
The current standard of care in Canada for patients with severe hemophilia A is treatment with FVIII prophylaxis. Although some Canadian patients are treated with episodic FVIII (generally related to patient preference in adults), the use of episodic FVIII only, as required by HAVEN 3 in arms A, B, and C, is not generally reflective of Canadian clinical practice. Additionally, patients previously treated with episodic FVIII were required to have 5 or more bleeds in the 24 weeks prior to study entry. This inclusion criteria creates a study population with more uncontrolled bleeding than would be observed in the Canadian clinical population. It is possible that the efficacy results observed in the trials may be exaggerated compared to what would be seen in clinical practice in Canada. This highlights a gap in evidence, given that the true magnitude of the treatment effect in patients with better control than those included in the trials (i.e., consistent with the Canadian clinical population) is unknown. In the clinical setting, it is also important to consider that patients with moderate hemophilia A may require prophylaxis; however, data from HAVEN 3 and HAVEN 4 were based primarily on severe patients, illustrating another gap in evidence.
The magnitude of the effect of emicizumab in patients previously treated with episodic FVIII in HAVEN 3 may be overestimated based on these limitations to external validity. However, differences in the magnitude of the efficacy results are not expected to invalidate the clinical significance of the bleeding outcomes in HAVEN 3, which are expected to address a gap in therapy experienced by patients currently treated with FVIII.
Based on the patient group input received for this review, HRQoL was an outcome of importance. The Haem-A-QoL physical health subscore was a secondary outcome (within the statistical testing hierarchy) in HAVEN 3 that did not show a statistically significant improvement for the arm receiving 1.5 mg/kg emicizumab weekly, which stopped statistical testing based on the pre-specified statistical testing hierarchy. Results from the Haem-A-QoL total score, EQ-5D-5L, and EQ VAS were outside the statistical testing hierarchy specified in HAVEN 3, and the results associated with each dosing regimen were inconsistent, although the 24-week evaluation period may not have been of sufficient duration to observe an effect. Additionally, the open-label study design rendered the interpretation of these results questionable. Despite the similar magnitude of ABRs for the different bleeding types in both emicizumab treatment regimens, the inconsistent HRQoL results prevent conclusions from being drawn regarding the impact of emicizumab on HRQoL. Additionally, minimal data assessing the effect of emicizumab on HRQoL in patients aged less than 18 years of age were available at the time of this review.
Input from the patient group and clinical experts consulted for this review highlights the importance for patients of being able to attend school and work. Based on descriptive analysis, patients in HAVEN 3 who were treated with emicizumab (1.5 mg/kg weekly, 3.0 mg/kg every 2 weeks) did not report missing any days of school, while those in the no-prophylaxis arm reported being away for an average of 10 days. In HAVEN 4, patients treated with 6.0 mg/kg emicizumab every 4 weeks were away from school for an average of less than 1 day. Based on descriptive analyses, patients were away from work for an average of less than 1 day. Data on days away from school and work were assessed based on a period of 4 weeks prior to baseline and prior to week 25. The effect of treatment for the remaining duration is unclear.
HAVEN 3 and HAVEN 4 provided descriptive results of patient preference assessed through the EmiPref and SQ-ISHI. The results showed a consistent patient preference for treatment with emicizumab compared to their previous treatment (FVIII). While this finding is consistent with the expectations of the patients and clinicians consulted for this review, it should be noted that EmiPref was a sponsor-developed scale; as well, neither scale was validated and neither had a recognized MID. Based on patient group input, the main unmet need for patients with severe hemophilia A is a user-friendly treatment, given that the IV route of administration for FVIII concentrates is challenging, complex, time-consuming, and less convenient than other routes of administration. Patients and clinicians anticipate that the SC route of administration associated with emicizumab could increase adherence in patients transitioning from pediatric to adolescent or adult care as well as in patients for whom venous access is difficult or inconvenient (e.g., geriatric patients). Patients and clinicians expect that the availability of a weekly or less frequent SC treatment (compared to frequent IV treatment with FVIII prophylaxis) will simplify treatment administration and may promote greater adherence. While adherence was assessed in both trials, the level of adherence in the real-world clinical population remains unknown.
Data from the NMA submitted by the sponsor suggested that emicizumab prophylaxis was associated with a reduction of bleed rates compared with FVIII product prophylaxis in the treatment of patients with severe hemophilia A without inhibitors. Limitations included: the small number of trials with small sample size included in the NMA; the high degree of heterogeneity across the included studies in disease severity; the use of different comparator FVIII products in different trials; inconsistent or unclear definitions of bleed outcomes; variable outcome estimation time points across trials; and differences in study design. The ICER ITC findings indicated uncertainty in terms of whether there was a difference in the reduction of treated bleeds between patients receiving emicizumab prophylaxis (combination of 1.5 mg/kg weekly and 3.0 mg/kg every 2 week) and those receiving FVIII prophylaxis. Overall, no robust conclusions can be drawn regarding the comparative clinical efficacy and safety profile of emicizumab prophylaxis versus rFVIII prophylaxis in patients with hemophilia without inhibitors.
Patients under 12 years of age were not included in HAVEN 3 or HAVEN 4. However, this is unlikely to affect the study results’ generalizability to patients under 12 years of age, given the mechanism of action of emicizumab, which functions independently of inhibitor status or age.8 The clinical efficacy of emicizumab in children with inhibitors was assessed in a previous submission by the National Advisory Committee on Blood and Blood Products (not reviewed in this report). The only data available in pediatric patients with hemophilia A for this review were from the HOHOEMI study. Results showed that emicizumab prophylaxis was associated with a reduction in bleeding events when comparing pre- and post-treatment ABRs. The model-based ABR for treated bleeds was 8.9 bleeds per year (95% CI, 4.43 to 17.80) for the pre-treatment period (FVIII prophylaxis) and 1.1 bleeds per year (95% CI, 0.65 to 1.97) for the treatment period (3.0 mg/kg every 2 weeks emicizumab). Generally, these data are consistent with the intra-patient analysis in HAVEN 3. Data from the HOHOEMI study showed that all caregivers of pediatric patients preferred emicizumab prophylaxis over previous hemophilia treatments (FVIII prophylaxis). The main reasons provided were lower treatment frequency (38%) and fewer effects on daily activities and social interactions (23.1%). However, only 6 patients were included in the treatment group of interest to this review, and no efficacy conclusions can be drawn regarding the efficacy of emicizumab, given that no statistical testing was conducted. Overall, the evidence from the HOHOEMI study on the efficacy and safety of emicizumab in pediatric patients is limited by concerns regarding internal validity and generalizability to the Canadian population. Despite the lack of high-quality evidence available for pediatric patients, the clinical experts consulted for this review did not express any concerns with using emicizumab in this population.
Harms
In HAVEN 3, AEs occurred in 94.4% of patients receiving 1.5 mg/kg emicizumab weekly, 85.7% of patients receiving 3.0 mg/kg emicizumab every 2 weeks emicizumab, 50.0% of patents in the no-prophylaxis arm, and 87.3% of patients in the previous-FVIII-prophylaxis arm (1.5 mg/kg emicizumab weekly). In HAVEN 4, 73.2% of patients treated with 6.0 mg/kg emicizumab every 4 weeks experienced an AE. The most common AEs in both studies were injection-site reactions. In HAVEN 3, injection-site reactions occurred in 25.0% of patients receiving 1.5 mg/kg emicizumab weekly, 20.0% of patients receiving 3.0 mg/kg emicizumab every 2 weeks, 12.5% of patients in the no-prophylaxis arm, and 31.7% of patients in the previous-FVIII-prophylaxis arm (1.5 mg/kg emicizumab weekly). In HAVEN 4, 22.0% of patients treated with 6.0 mg/kg emicizumab every 4 weeks experienced an injection-site reaction.
Throughout HAVEN 3 and HAVEN 4, there were no instances of de novo inhibitor development detected in patients who tested negative for inhibitors (titre less than 0.6 CBU/mL) at baseline.
In HAVEN 3, SAEs occurred in 2.8% of patients receiving 1.5 mg/kg emicizumab weekly, 8.6% of patients receiving 3.0 mg/kg emicizumab every 2 weeks, 0% of patients in the no-prophylaxis arm, and 12.7% of patients in the previous-FVIII-prophylaxis arm (1.5 mg/kg emicizumab weekly). In HAVEN 4, 2.4% of patients treated with 6.0 mg/kg emicizumab every 4 weeks experienced an SAE. No patients died in the studies.
The 24-week assessment period of HAVEN 3 and HAVEN 4 was determined to be of sufficient duration by the clinical experts consulted in this review. Overall, the safety profiles were likely comparable between the 2 emicizumab treatment regimens (1.5 mg/kg weekly and 3.0 mg/kg every 2 weeks); however, due to the small sample size, further study is warranted.
The safety profile of emicizumab compared to FVIII prophylaxis remains unknown due to the absence of direct, comparative evidence. An assessment of harms data was not conducted in the NMA submitted by the sponsor.