Combination antiretroviral therapy and chronic HIV infection affect serum retinoid concentrations: longitudinal and cross-sectional assessments

BACKGROUND Several lines of evidence suggest that retinoids (retinol-ROL or vitamin A, and its active metabolites, retinoic acids-RAs) play important pathogenic roles in HIV infection and combination antiretroviral therapy (cART)-related events. We previously reported that antiretrovirals alter RAs synthesis in vitro. We hypothesised that in vivo serum retinoid concentrations are affected by both cART and HIV infection. This might explain several clinical and laboratory abnormalities reported in HIV-infected patients receiving cART. METHODS The effects of optimal cART and chronic HIV on serum retinoids were firstly assessed longitudinally in 10 HIV-infected adults (group1 = G1): twice while on optimal cART (first, during long-term and second, during short term cART) and twice during 2 cART interruptions when HIV viral load (VL) was detectable. Retinoid concentrations during optimal long term cART in G1 were compared with cross-sectional results from 12 patients (G2) with suboptimal cART (detectable VL) and from 28 healthy adults (G3). Serum retinoids were measured by HPLC with ultraviolet detection. Retinoid concentrations were correlated with VL, CD4+ T- cell count and percentages, CD8+38+ fluorescence, triglycerides, cholesterol and C-peptide serum levels. RESULTS During optimal cART, G1 participants had drastically reduced RAs (0.5 ± 0.3 μg/dL; P < 0.01) but the highest ROL (82 ± 3.0 μg/dL) concentrations. During cART interruptions in these patients, RAs slightly increased whereas ROL levels diminished significantly (P < 0.05). G3 had the highest RAs levels (7.2 ± 1.1 μg/dL) and serum ROL comparable to values in North Americans. Serum ROL was decreased in G2 (37.7 ± 3.2 μg/dL; P < 0.01). No correlations were noted between RA and ROL levels or between retinoid concentrations and CD4+ T- cell count, CD8+38+ fluorescence, VL. ROL correlated with triglycerides and cholesterol in G1 (rs = 0.8; P = 0.01). CONCLUSIONS Serum RAs levels are significantly diminished by cART, whereas ROL concentrations significantly decreased during uncontrolled HIV infection but augmented with optimal cART. These alterations in retinoid concentrations may affect the expression of retinoid-responsive genes involved in metabolic, hormonal and immune processes and be responsible for some adverse events observed in HIV-infected persons treated with antiretrovirals. Further studies should assess concomitant serum and intracellular retinoid levels in different clinical situations in larger, homogenous populations.