Experiences with Emicizumab: A Retrospective Review of Patients on Emicizumab Therapy at the North Texas Hemophilia Treatment Center

Background: The HAVEN 1-4 trials demonstrated Emicizumab's success in lowering annualized bleeding rates (ABR) among patients with congenital Hemophilia A with and without inhibitors. Multiple real world experiences have continued to demonstrated efficacy and a tolerable adverse effect profile. Owing to higher efficacy and ease of administration compared to prophylactic factor VIII infusions, many centers now consider emicizumab as standard of care for prophylaxis of bleeding. We sought to evaluate efficacy, side effects, and barriers to therapy among patients receiving Emicizumab at the North Texas Hemophilia Treatment Center.

Methods: We conducted a retrospective review of patients >18 years who were prescribed Emicizumab for prophylaxis of bleeding at Parkland Hospital, a public hospital funded by Dallas County, and Clements University Hospital (CUH) affiliated to University of Texas Southwestern. Patients were taught how to administer emicizumab in clinic and monitored for adverse effects after the first dose by clinic staff. Patients reported bleeding and adverse events by entering information into an event log, informing the clinic via telephone or MyChart, or by recall during clinic visits. Data was collected by chart review from date of initiation of Emicizumab.

Results: 18 patients at Parkland and 42 patients at CUH were reviewed. 2 other patients were excluded due to insufficient documentation. The median age was 36, with majority males (n=57, 98%), all of whom were diagnosed with Hemophilia A. One female patient was prescribed Emicizumab for von Willebrand disease. Racial demographics included White (55%), followed by Hispanic (23%), African American (13%) and Asian (7%) patients. Mean ABRs across both sites were 0.65 (95% CI 0.57-0.96). No significant differences were seen in ABRs between the county (mean ABR 0.72, 95% CI 0.38-1.04) and university hospital (mean ABR 0.61, 95% CI 0.66-1.04). Arthralgias were the most common side effect (n=35, 55%), followed by headaches (n=13, 22%), myalgias (n=12, 20%), diarrhea (n=4, 6.7%), and fever (n=2, 3%). Post-injection adverse effects included bruising (n=1), fatigue and nausea (n=1), and a panic attack (n=1) after injection, none of which were seen with subsequent injections. One patient was found to have an unprovoked deep vein thrombus in the right brachial vein during evaluation for unexplained fever and leukocytosis; no other thrombotic events (TE) were seen. No patients were diagnosed with a thrombotic microangiopathy (TMA). Four patients died (6.6%), two due to intracranial bleeds and two due to causes unrelated to hemophilia. One patient suffered an out-of-hospital cardiac arrest and developed anoxic brain injury. Another patient died due to uncontrolled shock and acute hypoxemic respiratory failure. Two patients discontinued treatment, one due to perceived lack of efficacy and one due to loss of insurance. Four patients had interruptions in therapy, three due to non-compliance and one due to a pharmacy error in delivery of the medication. 78% of patients were on a preferred provider organization (PPO) plan and 13% were on federal health insurance (FHI). Average time between decision to start medication to administration of first dose (time to treatment) was 132 days (95% CI 96-168 days) on a PPO plan and 172 days on an FHI plan (95% CI 0-391 days).

Discussion: Our cohort had significant differences compared to HAVEN 1-4 participants, including a higher median age due to exclusion of patients below 18 years and a larger Hispanic population largely attributable to patient demographics in the county hospital. We had an increased incidence of arthralgias, lower rate of TEs and no TMAs. Our ABR was lower than that of HAVEN participants' 1.4 (95% CI 1.1-1.7). Rate of self-discontinuation of therapy was 3% among our group, comparable to 2.5% among HAVEN participants. Time to treatment was higher for patients on an FHI plan compared to a PPO plan, although statistical significance could not be determined due to low sample size of the FHI subgroup. Limitations include smaller sample size (n=60) than HAVEN (n=400) and underestimated ABR given recall bias due to the retrospective nature of our study contrary to real-time monitoring in prospective studies and clinical trials.

No relevant conflicts of interest to declare.

One patient received off-label emicizumab for von Willebrand disease.