Background: Exposure to NRTIs has been related with mitochondrial toxic effects: neurologic symptoms in infants and lactic acidosis in older patients. Lipodystrophy Syndrome (LDS) has been attributed both to NRTIs and PIs.

Methods: Cohort 1: Prospective observational study in 102 HIV-uninfected children (52 girls) born to HIV-infected women. Clinical symptoms suggesting mitochondrial dysfunction were analyzed in routine clinical follow-up and lactate (L) levels were obtained at 3 wks and 3, 6 and 12 mos of life, together with alanine (Al) plasma levels whenever hyperlactatemia (HL) was observed. HL with hyperAl appears only when mitochondrial injury becomes chronic. Cohort 2: Prospective observational study in 80 HIV-infected pediatric patients (42 girls, mean age: 9 yrs). Fat redistribution was identified by physical examination and parental questionnaire. Fasting blood analysis were obtained every 3–6 mos and included L, Al, triglycerides (TG) and total cholesterol (C).

Results: Cohort 1: Most of the women received ZDV in different regimens during pregnancy and labor. All newborns were treated with ZDV, alone (73/102) or combined with NVP and/or 3TC. HL with hyperAl was detected in 67 children (64%). L peak levels were 8.06, 10.1, 7.28 and 4.48 mmol/l at 6 wks, 3, 6 and 12 mos of life, respectively. In 22 patients (pts), L plasma levels have progressed spontaneously to normality. Three (3) girls presented self-limited axial hypotonia together with L peak levels of 7.28, 4, and 4.6 mmol/l. Cohort 2: Fat redistribution (FR) was identified in 18 pts (22%, 12 females): central fat accumulation (10), peripheral lipoatrophy (3) and combined FR (5). FR-affected pts exhibited a longer duration of HAART (71 vs 54 mos) and higher levels of TG (1.62 vs 1.11 mmol/l) and C (5.08 vs 4.6 mmol/l) (Mann-Whitney test, p = 0.082, p = 0.002 and p = 0.032). HyperTG and hyperC were observed in 26 (32%) and 39 (49%) cases, with fat redistribution in 11 and 13 pts, respectively. Symptom-free HL was identified in 14 pts (17%). L levels correlated positively with both TG and C levels (Pearson test, r = 0.304, p < 0.0001 for TG and r = 0.295, p < 0.0001 for C).

Conclusions: Cohort 1: Benign and self-limited HL affected 64% of our HIV-uninfected infants exposed to NRTIs, affecting neurologic development when symptomatic. Cohort 2: FR and dyslipidemia affected 20% and 60% of our HIV-infected pediatric pts, respectively. The correlation between L and TG and C levels suggests that NRTIs play a role in LDS.