1). One patient was lost to follow-up from week 8. Five participants modified their uridine dose to one sachet daily on 30 days per month because of diarrhoea. Overall adherence

(including patients who discontinued therapy or were lost to follow-up), as measured by pill and sachet count, was 91% for uridine recipients and 88% for pravastatin recipients. ART remained unchanged during the study follow-up period for 40 find more patients (89%); five patients (11%) were switched from LPV/r between week 4 and week 12, mainly for gastrointestinal disturbances (four to atazanavir/ritonavir and one to fosamprenavir/ritonavir). Limb fat increased modestly in all groups (Table 2). At week 24, the difference in the mean change in limb fat mass between those

who received uridine and those who did not receive uridine was 0.03 kg [95% confidence interval (CI) −0.35, +0.28; P=0.79] (Fig. 2;table 2). The difference between EPZ015666 datasheet those who received pravastatin and those who did not receive pravastatin was −0.03 kg (95% CI −0.29, + 0.34; P=0.84). Uridine increased mean lean mass (without weight change) by 1.0 kg (P=0.03) and there was a nonsignificant increase in lean mass in patients assigned to pravastatin (0.9 kg; P=0.07). Total visceral adipose tissue, subcutaneous adipose tissue and truncal fat did not change significantly with either intervention. In a post hoc analysis, a nonsignificant change in limb fat mass with uridine was also observed when the analysis was confined to the 15 participants who only received uridine at the planned dose of 36 g tid for 10 consecutive days each month (mean limb fat difference for the 16 uridine recipients vs. uridine controls, −0.08 kg; 95% CI −0.44, +0.29; P=0.68). The relationship between changes in cholesterol and changes in limb fat in pravastatin recipients was not significant (ρ=0.17; P=0.50). In a post hoc analysis, we stratified patients by baseline limb fat mass, body mass index (BMI) (<25% percentile, between 25 and 75%, and >75%), age and tNRTI duration (using median values to define categories).

We found no effect of any of these variables on the magnitude of limb fat change with either intervention (data not about shown). Pravastatin decreased total cholesterol (mean relative decline 0.4 mmol/L for pravastatin recipients vs. pravastatin controls; P=0.099) and serum bicarbonate (mean relative decline 2 mmol/L; P=0.005), but had no effect on other metabolic parameters, including HDL and estimated low-density lipoprotein (LDL) cholesterol (Table 3). Uridine caused a significant but modest decrease in serum potassium (mean change −0.2 mmol/L; P=0.05). There was no loss of virological control with either intervention. Five participants (11%) developed sustained grade 3 or 4 hypertriglyceridaemia (four of whom had initiated LPV/r at screening), three patients developed a grade 3 or 4 elevation in creatine kinase and one patient developed grade 3 thrombocytopenia.