OC 24.2 - Estimated Factor VIII Activity Levels at the Time of Bleeding Events in Individuals with Hemophilia A without Inhibitors

Sunday, June 25, 2023

15:00 – 15:15 ET

Room: 520

Session Available As: On-Demand

Presenting Author(s)

PC

Pierre Chelle, PhD

Research Associate
University of Waterloo
Waterloo, Ontario, Canada

Co-Author(s)

DH

Dagmar Hajducek

Statistician/Research Associate
University of Waterloo
Waterloo, Ontario, Canada
ND

Nabil Daoud

Director, Biostatistics and Data Science
American Thrombosis and Hemostasis Network
Rochester, New York, United States
EI

Emma Iserman

Statistical Analyst
Department of Health Research Methods, Evidence, and Impact, McMaster University, Hamilton, ON
Hamilton, Ontario, Canada
CG

Christine Gerber

Senior Project Manager
American Thrombosis and Hemostasis Network
Rochester, New York, United States
Federico Germini, MD, MSc, PhD

Assistant Professor
Department of Medicine, McMaster University, Hamilton, ON
Hamilton, Ontario, Canada
AI

Alfonso Iorio

Professor
Department of Health Research Methods, Evidence, and Impact, McMaster University, Hamilton, ON
Hamilton, Ontario, Canada
AE

Andrea Edginton

Professor of Pharmacy
University of Waterloo
Waterloo, Ontario, Canada
Michael Recht, MD, PhD, MBA (he/him/his)

Chief Medical and Scientific Officer
National Hemophilia Foundation
New York, New York, United States

Background: People with hemophilia A (PwHA) lack clotting factor VIII (FVIII) resulting in bleeding events (BEs) which may be prevented by maintaining a factor level above a target threshold.

Aims: Assess the relationship between FVIII levels and bleeding risk in PwHA using real world data.

Methods: Real world data was retrospectively collected from the American Thrombosis and Hemostasis Network dataset, the Web Accessible Population Pharmacokinetics Service – Hemophilia (WAPPS-Hemo) database and the Canadian Bleeding Disorders Registry, limited to participants with PK, BEs and dose information. BEs were classified according to cause (trauma or spontaneous) and location (joint or non-joint). PK parameters were obtained by Bayesian estimation using WAPPS-Hemo population PK models. FVIII levels were simulated using PwHA’s dose records and estimated PK parameters. FVIII – bleeding risk relationship was assessed by developing and evaluating a repeated time to event (RTTE) model.

Results: 2862 BEs from 427 PwHA aged 1 to 66 years with median of 21 years were identified; including 1434 BEs related to trauma, 1428 spontaneous and 1984 joint bleeding events. Observation period ranged from 3 weeks to 11 years with median of 2.28 years (Table 1). Simulated FVIII was below 1 IU/dL for 864 (30.2%) bleeding events and between 1 and 10 IU/dL for 1295 (45.2%) bleeding events (Figure 1). The RTTE model included Hill and a Weibull models in the hazard function describing the effects of FVIII levels and observation time on bleeding risk, and between-subject variability (BSV) on base hazard (CV=79%), bleed cause probability (CV=59%) and bleed location probability (CV=100%). Agreement between observed vs estimated bleed count was R2=0.95.

Conclusion(s): Simulations performed on virtual subjects demonstrated higher FVIII levels lead to lower bleeding risk. Trough levels were well correlated with bleeding risk independent from dosing. Funding was provided by Sanofi and Sobi.