Background/Aim: Accurate genotyping of CYP2D6 is challenging due to its inherent genetic variation, copy number variation (duplications and deletions) and hybrid formation with highly homologous pseudogenes. Because a relatively high percentage (∼25%) of clinically prescribed drugs are substrates for this enzyme, accurate determination of its genotype for phenotype prediction is essential. Methods: A cohort of 365 patient samples was genotyped for CYP2D6 using Sanger sequencing (as the gold standard), hydrolysis probe assays or pyrosequencing. Results: A discrepant result between the three genotyping methods for the loss of function CYP2D6*3 (g.2549delA, rs35742686) genetic variant was found in one of the samples. This sample also contained the CYP2D6 g.2470T>C (rs17002852) variation, which had an allele frequency of 2.47% in our cohort. Redesign of the CYP2D6*3 pyrosequencing and hydrolysis probe assays to avoid CYP2D6 g.2470 corrected the anomaly. Conclusion: To evidence allele drop out and increase the accuracy of genotyping, intra-patient validation of the same genetic variation with at least two separate methods should be considered.

No abstract