Background:  HAART-related long-term toxicities are being increasingly reported in HIV-infected children, many of which have been ascribed to direct mitochondrial (mt) toxicity of the nucleoside analogues. HIV infection per se can also cause mt damage.

Methods:  We present a case series of 8 perinatally HIV-infected pediatric patients (6 girls, median age 10.5 years) with optimal long-term response to a first-line HAART regimen who underwent planned treatment interruption. mtDNA content from peripheral blood mononuclear cells (PBMC) was assessed by means of a real-time polymerase chain reaction technique at the time of HAART interruption and 12 months later, while off therapy, and expressed as a mt/nuclear DNA ratio. Lactate plasmatic levels (normal values:  0.55 to 1.77 mM/L) were also analyzed. Non-parametric tests were used as appropriate.

Results:  At the time of HAART interruption, patients had remained a median time of 4.8 years and 4.5 years on therapy (4 of 8 on a protease inhibitor [PI] -based regimen) and with complete suppression of viral replication, respectively. One month after treatment interruption, a blip in HIV plasmatic viral load up to a median value of 4.6 log copies /mL (range 2.9 to 5.6) was observed in all cases; HIV viral load stabilized thereafter. One year after treatment interruption, a median increase of 0.99 (range –0.59 to 2.57) in mtDNA/nuclearDNA ratio was observed (from 1.37 to 1.82, median values; Wilcoxon rank test p = 0.093), together with a decrease in lactate plasmatic levels (from 1.3 to 0.69; p = 0.012). None of the patients showed symptoms consistent with hyperlactatemia, neither while on HAART nor during treatment interruption.

Conclusions:  In this series, treatment interruption leads to a partial restoration of mtDNA levels and to a significant decrease in lactatemia. Our results suggest that laboratory parameters consistent with mitochondrial toxicity in HIV-infected children are rather due to the use of nucleoside analogues than to HIV infection itself.