Outcomes, challenges and prospects of emicizumab prophylaxis in Sub-Saharan Africa – Insights from the Tanzanian experience

This cross-sectional study enrolled patients with haemophilia on emicizumab prophylaxis in Tanzania who were undergoing follow-up at Muhimbili National Hospital (MNH). MNH is home to Tanzania's largest haemophilia treatment centre (HTC), catering to both children and adults with haemophilia and other genetic bleeding disorders. Currently, only a limited number of vials are supplied annually, and therefore, emicizumab prophylaxis is strictly initiated at MNH for HA patients with inhibitors, severe disease, poor venous access and patients with history of major bleeding. On request by the HTC in Dar es Salaam, being under five years of age was also considered a standalone criterion in order to prevent early onset joint arthropathy and future disability as seen in the other patients not on prophylaxis. Maintenance doses can, nonetheless, be delivered at regional HTCs or alternative peripheral facilities equipped to provide subcutaneous injections. Patients who are willing and who exhibit proficiency during their follow-up clinics at MNH are encouraged to self-administer at home.

Recruitment of study participants was conducted daily at the HTC. The diagnosis was first verified through the hospital's electronic records platform (JEEVA) before recruitment. Although the country had an overall count of 65 patients receiving emicizumab during the period of the study, only those who consented to participate and had completed their loading dose in the study period (January to March 2024) were enrolled. In addition, invitation to participate in the study was made through phone calls and telephone interviews conducted for individuals who qualified but who did not show up at the clinic. A structured questionnaire administered by the investigators was used to gather both sociodemographic and clinical information. Details regarding clinical characteristics and laboratory investigations pertinent to the study were collected from patients records and JEEVA.

Adherence to emicizumab prophylaxis was evaluated by gathering patient-reported information regarding the vials of emicizumab prescribed and used in the past three months leading up to the interview. This was validated by reviewing the emicizumab dispensing ledger at the clinic alongside pharmacy records. As emicizumab is provided on donation, every vial is tracked by the clinic through the retention of stickers from the vials that include the barcode, which are subsequently affixed to patients’ records in the register for each clinic visit where emicizumab was dispensed and administered. As the donation comprises only the 30mg vial, a formula was applied to determine adherence. This was done by dividing the actual number of vials utilised by the vials dispensed from the MNH pharmacy and thereafter converting it into a percentage. This was performed for each individual month and cumulatively over the three-month period. Patients were subsequently categorized as adherent, sub-optimal, or non-adherent depending on whether they had missed less than 15%, 15–25%, or more than 25% of the recommended dosage, respectively ^[12]. For completion, reasons for skipped doses during these three months were also collected.

In this study, bleeding was characterised as visible bleeding from mucocutaneous areas, bleeding validated by imaging reports for individuals with CNS bleeding and as pain or swelling managed with CFCs or supportive methods (Protection, Rest, Ice, Compression & Elevation) for patients with musculoskeletal bleeds. Collected bleeding variables included the frequency of bleeding episodes, the nature of the bleed (spontaneous versus traumatic), the location of the bleed, and whether there was a resolution of target joints after emicizumab for patients with a noted history of a target joint before prophylaxis. Calculations for an annual bleeding rate (ABR) was done to objectively measure and standardise bleeding rates among patients, regardless of the duration of prophylaxis ^[7]. This was determined for every patient by dividing the total number of bleeding events by number of months on prophylaxis and thereafter multiplying by twelve months. An overall ARB was calculated to involve both traumatic and spontaneous and then separately for each.

An adverse effect was defined as any negative or unintended sign or symptom that arose or intensified from baseline following the commencement of emicizumab prophylaxis. A list of common adverse events was presented to the patient based on reports from HAVEN studies and other real-world reviews, including thrombotic incidents, headaches, diarrhea, reactions at the injection site (such as bruising, discomfort, induration, pain, pruritus, rash, site reactions, swelling, or urticarial formations), arthralgia, myalgia, and fever ^{[6,7,8,11,12,13]}. Participants were permitted to report any additional new symptoms outside those listed that they would associate with the use of emicizumab. The appearance of any of these symptoms at least once after the start of emicizumab was recorded as an adverse event.

Data gathered from the questionnaire was entered into the Open Data Kit (ODK) database and subsequently migrated into a Microsoft Excel sheet, where it was cleaned and analysed utilising Stata version 17. Descriptive statistics were applied for the evaluation of both sociodemographic and clinical information. The Chi-square or Fisher's exact test, as applicable, were employed to evaluate the associations among age, disease severity, adherence patterns, mode of payment, residence, and spontaneous bleeding. A robust Poisson regression model was utilised to assess the predictors of spontaneous bleeding. The Mann Whitney U test was carried out to compare the median occurrences of ABR for spontaneous bleeding before and after emicizumab prophylaxis. A p-value of less than 0.05 was regarded as statistically significant.

Ethical clearance to undertake this study was granted by the Muhimbili University of Health and Allied Sciences (MUHAS) Institutional Review Board (IRB) (Ref No. DA.282/298/01.C/2006) and from the Muhimbili National Hospital Directorate of research and publications (MNH/CRTCU/Perm/2024/080). A letter of permission from MNH was then presented to the head of the MNH HTC. Written consent was obtained from participants who physically attended the clinic while verbal consent was obtained from those who participated via telephone interview.

Our study enrolled a total of 55 child and adult participants with a median age (IQR) of 8 years (4, 14). Although the main HTC providing care is based in Dar es Salaam, nearly half of the participants resided in regions outside the area (49.1%). Most of the participants had severe HA (50.9%), with 40.0% and 7.3% having moderate and mild HA, respectively. Although none of them had quantification of active inhibitors at the point of diagnosis, 23.0% exhibited an inhibitor on mixing study following a record of CFC exposure. Whilst none of the participants had been on a prophylaxis programme previously, the majority had been exposed to CFCs (83.6%) and fresh frozen plasma (72.7%) on demand. Prior to emicizumab prophylaxis, the median ABR (IQR) for spontaneous bleeds was 8 (5,16). For joint health, 34.6% reported the existence of a target joint and 11% had permanent joint destruction in at least one joint causing disability prior to prophylaxis. Table 1 summarises the sociodemographic and clinical characteristics of the study population.

Initiation of emicizumab was founded on a criterion developed by the WFH in partnership with the HTC. Being under five was the most prevalent reason for starting prophylaxis in about 32.7% of participants. History of inhibitors (23.6%), severe disease (20.0%), moderate disease with a background of major bleeding (12.7%) and poor venous access (5.5%), represented additional reasons for initiation. The median length (IQR) of emicizumab prophylaxis in the study group was 12 months (7,32), with the maximum duration recorded at 32 months. For maintenance dosing, the majority of the patients preferred monthly administration at 6mg/kg (67.3%) while the remainder chose bi-monthly (32.7%) at 3mg/kg. Despite being a subcutaneous injection that may be delivered by the patient or a caregiver, only 30.0% of the patients were self-injecting at home. Although around half (49.1%) admitted to having skipped at least one dose of emicizumab within the past three months, the vast majority of study participants were adherent to prophylaxis (78.2%). Based on the monthly adherence pattern, a downward shift in adherence was observed in the last month, where only 56.4% were found to be adherent in contrast to 72.7% during the first month (Figure 1). Among the explanations provided for missed doses, long proximity to the HTC (44.4%), lack of money to pay for consultation (33.3%) and shift to CFC because of scarcity of emicizumab (25.9%) were noted. Other reasons reported included absence of fare funds (11.1%), perceived wellness due to lack of bleeding events (7.4%), and one adult participant felt overwhelmed by the numerous injections, given that only the 30mg vial is available.

Figure 1.

Adherence patterns of study participant on emicizumab prophylaxis

A total of 35 (63.6%) participants experienced at least one bleeding incident throughout the duration of prophylaxis. Among these, 62.9% encountered traumatic bleeding, 17.1% spontaneous bleeding, and 20.0% had both. A total of 77 bleeding events were reported, with most being due to trauma (61%, 41/77) (Figure 2). The overall median annual bleeding rate (ABR)(IQR) for both spontaneous and traumatic bleeds was 0.75 (0,1.7). The ABR (IQR) for spontaneous bleeding by itself significantly decreased to 0 (0,0) for 8 patients before prophylaxis which was statistically significant (p < 0.001). Of the 19 individuals who indicated the existence of a target joint prior prophylaxis, 79.0% indicated resolution of the target joint while on emicizumab. Age was the only factor significantly associated with spontaneous bleeding, where individuals over 18 years old exhibited the highest rate of spontaneous bleeding compared to other age categories (p < 0.001). The association between sociodemographic and clinical characteritics and spontaneous bleeding is shown in Table 2.

Figure 2:

Site and location of bleeds in Haemophilia A patients on emicizumab prophylaxis

Only 5 participants (9.1%) indicated experiencing one or more adverse events during emicizumab prophylaxis. The most frequently reported adverse events were injection site reactions (5.5%) and headaches (3.6%). No thrombotic events were reported. All adverse events were mild and transient.