Pharmacogenetic-guided warfarin dosing algorithm in African American

Sameh Al-Zubiedi, Mohammad Saleh. The University of Jordan, Amman, Jordan

African Americans are underrepresented in warfarin dosing studies. Under representation of African American contributes to the poor performance of various algorithms when applied to African American (1). VKORC1 and CYP2C9 genotypes explain less variability in individuals of African descent than in those of European or Asian origin (2). A recent genome wide association study a novel CYP2C single nucleotide polymorphism (rs12777823) that affect warfarin dose in African Americans (3)

Aims: To develop warfarin dosing algorithm for African American that include demographics information, other genetic variants (i.e. CYP2C9, CYP4F2, VKORC1 genes, and CYP2C rs12777823), health status factors, and concomitant medications

Methods: Demographic, clinical, and genetic data from a previously collected cohort of 104 African American patients with a stable warfarin dose who were able to achieve an observed international normalized ratio of 2-3. We explored two approaches to develop the algorithm: multiple linear regression and artificial neural network (ANN) approach. The clinical significance of the two dosing algorithms was evaluated by calculating the percentage of patients whose predicted dose of warfarin was within 20% of the actual stable therapeutic dose. The clinical significance was also compared to a previously published dosing algorithm

Results: Both the linear regression and the ANN models explained 49% of the dose variability observed among patients. The mean absolute error using linear regression model was estimated to be 9.8 mg compared to 10.1 mg using ANN. Linear regression model and ANN model predicted the ideal dose in 57% and 54% of the patients, respectively. Linear regression and ANN models identified several predictors of warfarin dose including: age, weight, CYP2C9 genotype 1/1, VKORC1 genotype GG, rs12777823 genotype GG, rs2108622 genotype AG and GG, congestive heart failure, and aspirin use

Conclusion: We developed a warfarin dosing algorithm for African American using linear regression model. The proposed dosing algorithm has the potential to recommend warfarin doses that are close to the appropriate doses. The use of more sophisticated ANN approach did not result in improved predictive performance of the dosing algorithm

(1) International Warfarin Pharmacogenetics (2009) N Engl J Med 360: 753-64.

(2) Gage BF et al. (2008) Clin Pharmacol Therap 84: 326-31.

(3) Limdi NA et al. (2008) Pharmacogenomics 9: 511-526.

(4) Perera et al. (2013) Lancet 382: 790-6.