843    

Background:  Atherogenic dyslipidemia is common in ART-treated patients, particularly those on protease inhibitors (PI). We sought to identify single nucleotide polymorphisms (SNP) in candidate lipoprotein, adipose, cytokine, and drug transport genes that predict dyslipidemia in subjects on PI-containing ART compared with those on non PI-containing ART.

Methods:  HIV subjects (n = 626, mean age 42 years, 89% male, 60% Caucasian) with fasting lipids measured while on stable PI-containing ART (n = 471; 39% injecting drug users, 32% nelfinavir (NFV), 5% ritonavir (RTV); median exposure 973 days (IQR = 232 to 1240)) or non PI- containing ART (n = 155) were studied. The following SNP were examined:  apolipoprotein C-III (Apo C-III -455C/T and -482C/T); sterol regulatory element binding protein-1c (SREBP-1c 322C/G); tumor necrosis factor-a (TNF-a-238G/A); and multidrug-resistance transporter (MDR1 3435C/T). Primary lipid outcomes were triglycerides and HDL-cholesterol.

Results:  In unadjusted analysis (analysis of variance of natural log-transformed triglycerides and HDL-cholesterol), there was a significant interaction between PI-use and the Apo C-III -455 CC (n = 142) and -482 TT (n = 99) genotypes on triglycerides (p = 0.048, p = 0.015, respectively) and HDL-C (p = 0.037, p = 0.021). Among patients with Apo C-III -455 CC and -482 TT genotypes, PI-use was associated with significantly higher triglycerides (p = 0.037, p = 0.007, respectively) and lower HDL-cholesterol (p = 0.019, p = 0.018) compared with non-PI use. However, controlling for race attenuated the interaction between PI-use and both Apo C-III -455 CC and -482 TT on triglycerides (p = 0.13, p = 0.047, respectively) and HDL-cholesterol (p = 0.22, p = 0.20). The interaction effect further diminished after additionally adjusting for age, sex, study, body mass index, and type 2 diabetes. No interactions were detected between SREBP1c, TNF-a, or MDR1 genotypes and PI-containing ART on lipids.

Conclusions:  Although, genetic variations in Apo C-III in combination with PI use may modulate the development of atherogenic dyslipidemia in HIV subjects, racial admixture and population characteristics may account for this observation in our study. Additional analysis of haplotype effects may provide greater insight into the association of PI, candidate genes, and atherogenic dylipidemia in HIV.

Keywords: Dyslipidemia ; Apolipoprotein C-III Polymorphisms ; Protease Inhibitors