Background:  The MaxCmin1 trial was a phase 4 randomized, open-label, multicenter trial to evaluate the safety and efficacy of indinavir/ritonavir (IDV/RTV; 800/100 mg twice daily) vs saquinavir/ritonavir (SQV/RTV; 1000/100 mg twice daily) in adult HIV-1-infected patients. Comparable antiretroviral efficacy was observed between both arms although a greater number of treatment-limiting adverse events were observed in the Idv/Rtv arm, relative to the Sqv/Rtv arm. In order to investigate whether genetic variation within the MDR1 gene (coding for P-glycoprotein) contributed to treatment endpoints and variability in pharmacokinetics, we genotyped patients for the common single-nucleotide polymorphisms C3435T and G2677T.

Methods:  Of the 306 patients who initiated treatment, 229 patients were genotyped for either C3435T or G2677T or both. Additional blood samples were obtained for analysis of drug concentrations. Genomic DNA was extracted from cell suspensions using the Qiagen genomic DNA isolation kit. Both C3435T and G2677T single-nucleotide polymorphisms were genotyped by taqman allelic discrimination protocol following amplification with specific primers and fluorescent probes. Statistical analyses were then performed using individual single-nucleotide polymorphisms and the haplotypes generated from the two using STATA software (version 7). Logistic regression models were constructed in order to determine whether either genotype or haplotypes were predictors of a viral load>50/400 copies/mL or the time to a CD4 rise of >100 cells/mL from baseline. Similar analyses were carried out to determine whether clinical progression or time to an adverse event were dependent on these single-nucleotide polymorphisms. Finally, regression models were constructed in order to determine whether predicted trough concentrations of SQV, IDV and RTV were dependent on MDR1 genetics.

Results:  No significant differences were observed in any of the regression models for change in CD4 count or VL relative to C3435T, G2677T or haplotype in pooled analyses. No differences were observed in drug levels at week 4 (table) or 48 for MDR1 haplotypes. or individual genotypes.

Conclusions:  Genetic analysis of the MDR1 gene failed to correlate with predicted trough concentrations of SQV, IDV, and RTV in this study. However, his work does not exclude a clinically significant role for P-gp, or a role for other host single-nucleotide polymorphisms in MDR1 and other transporter loci, in influencing clinical progression in response to protease inhibitors.

Keywords: P-glycoprotein; Protease Inhibitors; Genotype