Area under the curve: Comparing the value of factor VIII replacement therapies in haemophilia A

Authors: Persson, S; Berndt, C; Engstrand, S; Trinczek, A; Carlsson, KS; Berntorp, E

Affiliations: Swedish Institute for Health Economics, Lund, Sweden. Department of Clinical Sciences, Lund University, Lund, Malmö, Sweden. Bayer AB, Solna, Sweden. Bayer A/S, Copenhagen, Denmark. Department of Translational Medicine, Lund University, Lund, Malmö, Sweden.

Publication: Haemophilia; 2022

Abstract: INTRODUCTION: In factor VIII (FVIII) prophylaxis for haemophilia A, cost comparisons have used price per international unit (IU) based on the once reasonable assumption of equivalent outcome per IU. Now, with several extended half-life (EHL) products available, new outcome-oriented ways to compare products are needed. Area under the curve (AUC) quantifies FVIII levels over time after infusion providing comparable data. AIM: To develop a decision analytical model for making indirect comparisons of FVIII replacement products based on AUC. METHODS: A literature search identified 11 crossover studies with relevant pharmacokinetic data. A common comparator FVIII level curve was calculated using pooled data from selected studies. Absolute curves for other products were estimated based on relative differences to the common comparator (% difference vs the anchor). Three scenarios were investigated: (1) Kogenate(®) versus Kovaltry(®) and Jivi(®) ; (2) Advate(®) versus Elocta(®) , NovoEight(®) , Kovaltry, Adynovate(®) , Afstyla(®) , and ReFacto(®) ; and (3) Jivi versus Elocta, Adynovate, and Kogenate. Sensitivity analyses investigated effects of assay type and dose. RESULTS: In scenario 1, Jivi (+50%) and Kovaltry (+14%) showed larger AUCs versus Kogenate. In scenario 2, EHL products, Elocta and Adynovate, had the largest AUC (+64% and +58%, respectively) versus Advate. Compared with all other products in scenario 3, Jivi had the largest AUC by +13%-28%. CONCLUSION: This analysis concludes that EHL products differ in relative AUC, have a larger AUC compared with standard half-life, and thus, different FVIII levels over time after infusion. This model may aid decision makers in the absence of head-to-head data.