CYP2C9 and VKORC1 polymorphisms are differently distributed in the Brazilian population according to self-declared ethnicity or genetic ancestry.

BACKGROUND

Warfarin-dosing pharmacogenetic algorithms have presented different performances across ethnicities, and the impact in admixed populations is not fully known.

AIMS

To evaluate the CYP2C9 and VKORC1 polymorphisms and warfarin-predicted metabolic phenotypes according to both self-declared ethnicity and genetic ancestry in a Brazilian general population plus Amerindian groups.

METHODS

Two hundred twenty-two Amerindians (Tupinikin and Guarani) were enrolled and 1038 individuals from the Brazilian general population who were self-declared as White, Intermediate (Brown, Pardo in Portuguese), or Black. Samples of 274 Brazilian subjects from Sao Paulo were analyzed for genetic ancestry using an Affymetrix 6.0(®) genotyping platform. The CYP2C92 (rs1799853), CYP2C93 (rs1057910), and VKORC1 g.-1639G>A (rs9923231) polymorphisms were genotyped in all studied individuals.

RESULTS

The allelic frequency for the VKORC1 polymorphism was differently distributed according to self-declared ethnicity: White (50.5%), Intermediate (46.0%), Black (39.3%), Tupinikin (40.1%), and Guarani (37.3%) (p<0.001), respectively. The frequency of intermediate plus poor metabolizers (IM+PM) was higher in White (28.3%) than in Intermediate (22.7%), Black (20.5%), Tupinikin (12.9%), and Guarani (5.3%), (p<0.001). For the samples with determined ancestry, subjects carrying the GG genotype for the VKORC1 had higher African ancestry and lower European ancestry (0.14±0.02 and 0.62±0.02) than in subjects carrying AA (0.05±0.01 and 0.73±0.03) (p=0.009 and 0.03, respectively). Subjects classified as IM+PM had lower African ancestry (0.08±0.01) than extensive metabolizers (0.12±0.01) (p=0.02).

CONCLUSIONS

The CYP2C9 and VKORC1 polymorphisms are differently distributed according to self-declared ethnicity or genetic ancestry in the Brazilian general population plus Amerindians. This information is an initial step toward clinical pharmacogenetic implementation, and it could be very useful in strategic planning aiming at an individual therapeutic approach and an adverse drug effect profile prediction in an admixed population.