Influence of cytochrome P450 (CYP) 2C8 polymorphisms on the efficacy and tolerability of artesunate-amodiaquine treatment of uncomplicated Plasmodium falciparum malaria in Zanzibar

Background The anti-malarial drug, amodiaquine, a commonly used, long-acting partner drug in artemisinin-based combination therapy, is metabolized to active desethyl-amodiaquine (DEAQ) by cytochrome P450 2C8 (CYP2C8). The CYP2C8 gene carries several polymorphisms including the more frequent minor alleles, CYP2C8*2 and CYP2C8*3. These minor alleles have been associated with decreased enzymatic activity, slowing the amodiaquine biotransformation towards DEAQ. This study aimed to assess the influence of these CYP2C8 polymorphisms on the efficacy and tolerability of artesunate-amodiaquine (AS-AQ) treatment for uncomplicated Plasmodium falciparum malaria in Zanzibar. Methods Dried blood spots on filter paper were collected from 618 children enrolled in two randomized clinical trials comparing AS-AQ and artemether-lumefantrine in 2002-2005 in Zanzibar. Study participant were under five years of age with uncomplicated falciparum malaria. Human CYP2C8*2 and CYP2C8*3 genotype frequencies were determined by PCR-restriction fragment length polymorphism. Statistical associations between CYP2C8*2 and/or CYP2C8*3 allele carriers and treatment outcome or occurrence of adverse events were assessed by Fisher's exact test. Results The allele frequencies of CYP2C8*2 and CYP2C8*3 were 17.5 % (95 % CI 15.4-19.7) and 2.7 % (95 % CI 1.8-3.7), respectively. There was no significant difference in the proportion of subjects carrying either CYP2C8*2 or CYP2C8*3 alleles amongst those with re-infections (44.1 %; 95 % CI 33.8-54.8) or those with recrudescent infections (48.3 %; 95 % CI 29.4-67.5), compared to those with an adequate clinical and parasitological response (36.7 %; 95 % CI 30.0-43.9) (P = 0.25 and P = 0.31, respectively). However, patients carrying either CYP2C8*2 or CYP2C8*3 alleles were significantly associated with an increased occurrence of non-serious adverse events, when compared with CYP2C8 *1/*1 wild type homozygotes (44.9 %; 95 % CI 36.1-54.0 vs. 28.1 %; 95 % CI 21.9-35.0, respectively; P = 0.003). Conclusions CYP2C8 genotypes did not influence treatment efficacy directly, but the tolerability to AS-AQ may be reduced in subjects carrying the CYP2C8*2 and CYP2C8*3 alleles. The importance of this non-negligible association with regard to amodiaquine-based malaria chemotherapy warrants further investigation.