OBJECTIVES : Treatment costs to US payers are determined by treatment units dispensed, not prescribed. Hemophilia A therapies with more dosage strength options have better dispensing accuracy than those with fewer options, which is key to avoiding product wastage. A model was developed to evaluate the potential treatment unit wastage and impact on cost among patients with severe hemophilia A without inhibitors receiving prophylactic emicizumab (Genentech, Inc., South San Francisco, CA, USA) with 4 dosing strengths, or antihemophilic factor (recombinant) (rAHF; Baxalta US Inc., a Takeda company, Lexington, MA, USA) with 11 dosing strengths in the US.

METHODS : Real-world data on patients with hemophilia A were obtained from specialty pharmacies between January 1, 2017 and July 30, 2018. For each patient, the most recent body weight available was used to estimate units of prescribed and dispensed treatment. Patients received emicizumab loading dose and assigned 1 of 3 dosing regimens (1:1:1): 1.5 mg/kg once/week (Q1W), 3 mg/kg once/2 weeks (Q2W), or 6 mg/kg once/4 weeks (Q4W). Dispensed units were determined using the manufacturer’s emicizumab dosing calculator. Estimations for rAHF were modeled on patients receiving 30 IU/kg every other day. Potential over-dispensed units and annual cost wastage are reported.

RESULTS : This analysis included 1495 patients (939 adult, 556 pediatric) with: severe hemophilia A without inhibitors, ≥1 hemophilia A treatment dispensing record, and available weight data. Median (range) age was 23 (0-88) years and weight 73 (3-179) kg. Estimated mean annual potential wastage in over-dispensed product for patients receiving emicizumab was 7.8% (251 mg, $23,032) (Q1W, 14% [449 mg, $41,317]; Q2W, 6% [197 mg, $18,092]; Q4W, 3% [107 mg, $9800]) versus 1.85% (6944 IU, $7291) for rAHF.

CONCLUSIONS : This model estimated >3-fold greater product wastage and related costs with emicizumab versus rAHF treatment (~2.5-fold at Q2W), which may be due to limited potency options for emicizumab prophylaxis.