Cost-Effectiveness of Long-Term Prophylaxis with Plasma-Derived Vs. Recombinant Von Willebrand Factor in Severe Von Willebrand Disease

Introduction: von Willebrand disease (VWD) is the most common inherited bleeding disorder and exhibits significant variability in bleeding phenotype across patients. International guidelines conditionally recommend long-term prophylaxis to reduce bleeding events for patients with a severe phenotype: all patients with type 3 VWD and a subset of patients with severe phenotypes of types 1 and 2 VWD. At the time of this conditional recommendation, only 10% of patients with severe VWD are treated prophylactically, unlike patients with severe hemophilia (~45%). In the United States, there are currently two approved factor replacement therapies: plasma-derived von Willebrand factor (pdVWF) and recombinant von Willebrand factor (rVWF). The cost-effectiveness of prophylaxis options for patients with VWD in the United States is not known. We sought to fill this gap by determining the cost-effectiveness of pdVWF versus rVWF prophylaxis treatment options for patients with severe VWD.

Methods: In this independent analysis free of industry influence, we constructed a Markov model of adult patients with severe VWD to compare the cost-effectiveness of long-term prophylaxis with pdVWF versus rVWF. Transition probabilities for patients' breakthrough bleeding events while on prophylaxis with each factor replacement product, and product utilization, were derived from an open-label, prospective, phase 3 clinical trial that employed a pdVWF versus rVWF switch design (NCT 02973087). Costs of prophylaxis and costs of VWD-specific breakthrough bleed health resource utilization in the United States (US) accounted for bleed category, severity, and bleed-specific mortality, and were sourced from previously published literature and the Centers for Medicare & Medicaid Services. For effectiveness measured in quality-adjusted life-years (QALY), utilities for patients with severe VWD on prophylaxis were informed from EQ-5D-specific utility data for patients with severe hemophilia on prophylaxis. The primary outcome was the incremental cost-effectiveness ratio (ICER) or the incremental net monetary benefit (iNMB) if the intervention (rVWF) was found to be cost-saving, reported over a lifetime time-horizon across a range of accepted willingness-to-pay (WTP) thresholds in US dollars per QALY. The two secondary outcomes were threshold analyses for 1) the breakthrough bleed rate and 2) frequency of infusions necessary for the non-favored product (i.e., in the base-case) to become favored. In addition, we conducted a scenario analysis assuming the non-significant reduction in bleeding with rVWF (0.55 [95% confidence interval 0.086-3.523]) otherwise nullified in the base-case analysis, was significant. We proceeded with deterministic sensitivity analyses, varying all parameter estimates across the largest of their known 95% confidence intervals or +/-20% of absolute value. We concluded with a probabilistic sensitivity analysis, capturing uncertainty in all parameters simultaneously over 10,000 Monte Carlo simulations.

Results: In the base-case, treatment with prophylactic pdVWF versus prophylactic rVWF yielded discounted lifetime costs of $21.2 million and $19.8 million, with both accruing 20.6 discounted QALYs. The average incremental net monetary benefit across accepted WTP thresholds of $50,000-$150,000/QALY was $1.38 million [95% credible interval $1.10-1.67 million], in favor of rVWF. Deterministic sensitivity analysis showed that the model was most sensitive to the frequency and cost of pdVWF infusions. In all three probabilistic sensitivity analyses across all WTP thresholds, rVWF was favored over pdVWF in 100% of 10,000 Monte Carlo iterations. Threshold analyses revealed that 1) the bleeding rate on pdVWF prophylaxis would have to reduce by at least 43% and, 2) the average frequency of pdVWF infusions would need to decrease from 3 to 2.75 infusions per week for pdVWF to be favored over rVWF prophylaxis. Scenario analysis revealed similar significant improvement in incremental net monetary benefit with rVWF.

Conclusion: Twice-weekly rVWF versus thrice-weekly pdVWF demonstrated similar quality-adjusted life-expectancy, while rVWF was cost-saving due to its longer half-life requiring fewer lifetime infusions. At current pricing in the United States, prophylactic rVWF is a cost-saving option for patients averaging more than 2.75 weekly pdVWF infusions.