Background:  The independent role of hepatitis C virus genotype 3 (HCV-3) in lipid elevations following HAART is unexplored.

Methods:  Patients were selected if baseline and 3-month determinations of serum lipid levels were available after new HAART was initiated from January 2001 to November 2004, ranked by HCV infection (HCV-3, HCV-non-3 genotypes or HCV-negative). On-treatment analysis was conducted to account for possible ongoing influence of prescribed drugs. Univariate and multivariate GEE model was used to identify factors correlated with lipid serum levels across the entire follow-up as coefficient (i.e., monthly mean variation, Coef.). Logistic regression analysis was used to identify predictors of hyper-trigliceridemia (hyper-TG either abnormal: >180 mg/dL or relevant: >500 mg/dL) and hyper-cholesterolemia (hyper-CHOL either abnormal: >200 mg/dL or relevant >240 mg/dL) after 3-month follow-up. LDL- and non-HDL-CHOL were also analyzed where available. 

Results:  Analysis of 415 HIV mono-infected and 307 HIV/HCV co-infected patients (12.7% HCV-3 and 29.8% HCV-non-3) was conducted over a mean (SD) of 380 (281), 396 (308), and 417 (307) days in the 3 groups, respectively. HCV-3 correlated with lower TG (Coef. = –38.51; p = 0.017), independently from male gender (Coef. = 36.14; p = 0.001), baseline CD4⁺ T cell count (Coef. = 0.07; p = 0.017), baseline total CHOL (Coef. = 0.47; p <0.001), use of stavudine (d4T) (Coef. = 35.61; p = 0.002) or lopinavir/ritonavir (LPV/r) (Coef. = 69.49; p <0.001), and lowest recorded CD4⁺ T-cell count (Coef. = –0.13; p = 0.004). Among ART-experienced patients, longer exposure to ART was an adjunctive independent risk factor for TG increase (Coef. = 0.014; p = 0.014). HCV infection per se appeared to be protective, however HCV-3 in particular was independently associated with lower CHOL during the follow-up (Coef. = –45.94; p <0.001). With respect to HIV alone, HCV-3 was also associated with protection from abnormal (OR 0.25, 95%CI 0.09 to 0.7; p = 0.008) or relevant (OR: 0.12, 95%CI 0.01 to 1.1; p = 0.06) hyper-TG, and from abnormal (OR 0.24, 95%CI 0.09 to 0.65; p = 0.005) or relevant hyper-CHOL (OR 0.12, 95%CI 0.01 to 0.95; p = 0.045), independently from the other factors studied.

Conclusions:  Our data confirm that HCV co-infection is associated with lower risk of hyper-CHOL on HAART. However this effect is driven by HCV-3, which is the unique HCV genotype associated with protection from hyper-TG. These findings could have important epidemiological and clinical implications.