Abstract Number: PB1066
Meeting: ISTH 2020 Congress
Background: Inhibitors are the most important treatment-related complication among persons with hemophilia. Currently, the only validated risk prediction tool relies on information about prior product exposure, limiting its utility to predict inhibitor risk in untreated or minimally-treated patients. A risk prediction tool based on patient-related risk factors that could be available before initiation of treatment could be more useful clinically.
Aims: This investigation evaluates six risk prediction tools based on genetic risk factors for inhibitors among persons with hemophilia A. The tools combine information on genotypes for variants in genes in the immune response pathway and the hemophilia genotype.
Methods: Variants in genes in the immune response pathway were weighted based on 1) estimates from a meta-analysis or 2) results of a prior investigation in this population and combined with 3 different hemophilia genotype categorization schemes, for a total of 6 tools evaluated. The ability of each tool to correctly predict inhibitor status was evaluated in 558 White, non-Hispanic participants in the Hemophilia Inhibitor Research Study (HIRS) with complete genetic information and informed consent. Tool performance was evaluated using the area under the receiver operating curve (AUC) after cross-validation.
Results: Among 450 inhibitor-negative and 108 inhibitor-positive HIRS participants, the prediction tool that combined information on immune-response genes previously-found to be associated with inhibitors in this population and categorized the hemophilia genotype based on previously-reported estimates of effect performed best (AUC=0.75 and AUC=0.62 among persons with mild/moderate and severe disease, respectively) (Table 1).
Conclusions: The best-performing tool did not perform as well as a previously-validated tool among persons with severe disease but performed well among persons with mild/moderate disease. Although cross-validation methods were used to reduce bias introduced by using the same data to simultaneously fit and evaluate the models, validation of these findings in other populations is warranted.
|All (AUC)||Mild/Moderate (AUC)||Severe (AUC)|
|Validated Inhibitor Prediction Tool (ter Avest et al. 2008)||0.74|
|Hemophilia Variant Categorization||Immune Response Variant Estimates|
|(1)Pathogenicity-based: hemophilia-causing variant weighted based on American College of Medical Genetics and Genomics pathogenicity criteria (2)Functionally-based: hemophilia-causing variant weighted based on predicted functional impact on factor VIII protein (3)Evidence-based: hemophilia-causing variant weighted based on previously-published estimates of effect (Mild/Moderate: Eckhardt et al. 2013; Severe: Gouw et al. 2012) (4)Meta-analysis: immune response gene variant genotypes weighted based on meta-analysis results (5)HIRS-based: immune response gene variant genotypes weighted based on estimates of effect derived from analysis in Hemophilia Inhibitor Research Study (HIRS) cohort|
[Table 1: Comparison of area under receiver operating curve (AUC) for inhibitor risk prediction tools]
To cite this abstract in AMA style:Payne AB, Miller CH, Ellingsen D, Driggers J, Bean CJ, Mulle JG, Soucie JM, Hemophilia Inhibitor Research Study Investigators . Evaluation of Inhibitor Risk Prediction Tools Based on Genetic Risk Factors in Persons with Hemophilia A [abstract]. Res Pract Thromb Haemost. 2020; 4 (Suppl 1). https://abstracts.isth.org/abstract/evaluation-of-inhibitor-risk-prediction-tools-based-on-genetic-risk-factors-in-persons-with-hemophilia-a/. Accessed September 29, 2023.
« Back to ISTH 2020 Congress
ISTH Congress Abstracts - https://abstracts.isth.org/abstract/evaluation-of-inhibitor-risk-prediction-tools-based-on-genetic-risk-factors-in-persons-with-hemophilia-a/