Background:  Human orosomucoid (ORM) is a major binding protein for protease inhibitors such as atazanavir (ATV). The polymorphism of ORM could be responsible for interindividual variation in the plasma binding of ATV, which might influence the disposition and virological efficacy of ATV. The objective of this study was to describe the ORM polymorphism in HIV-infected patients and to determine its possible influence on the ATV pharmacokinetics.

Methods:  The study was conducted in 52 highly HIV-1-infected patients (50 males, 41 [29 to 62] year old) included in the ANRS 107 trial. They had failed previous HAART therapy and were treated by ritonavir boosted ATV (ATV/r), tenofovir, and nucleoside reverse transcriptase. The phenotyping of ORM variants was determined in sera after desialylation by isoelectric focusing (Phastsystem) on polyacrylamide gels following by immunoblotting. Total ORM plasma concentrations were measured by nephelemetry (BN Prospec) at week 6. ATV plasma concentrations were measured at week 6 prior to drug intake (42 patients) and during the dosing interval (10 patients). ATV pharmacokinetic parameters were determined using a population approach (WinNonMix). Relations between ORM phenotypes, total ORM concentrations, and ATV pharmacokinetic parameters were tested using a Kruskal-Wallis test.

Results:  Median total ORM concentrations were 1.0 g/L (0.6 to 1.5). ORM concentration of 14 patients was elevated with respect to their sex and age. The frequencies of the 3 phenotypes observed ORM1*F1-ORM2, ORM1*F1S-ORM2, and ORM1*S-ORM2 were 34.6%, 55.8%, and 9.6%, respectively. The allele ORM1*F2 was not observed in this population. All patients expressed a monomorphic allele on ORM2 locus. Total plasma concentrations of ORM were not found to be significantly dependant on ORM phenotypes. Median predicted trough ATV concentrations were 468 ng/mL (89 to 4978) and were not correlated to ORM phenotype. Median ATV volume of distribution was significantly higher (p = 0.008) in patients presenting the ORM1*S-ORM2 phenotype than the ORM1*F1S-ORM2 or ORM1*F1-ORM2 (87L vs 77L and 80L). In contrast, ORM phenotype did not significantly impair ATV clearance.

Conclusions:  Plasma ORM concentrations was elevated in 27% of these highly pretreated HIV-infected patients. Phenotype frequencies were close to those previously published. The different ORM variants affect ATV volume of distribution. This could be the consequence of different capacities for ATV binding which warrant further studies.