Lactate levels in children with HIV/AIDS on highly active antiretroviral therapy

A total of 251 venous lactate levels were obtained from 127 children with HIV/AIDS; 104 on highly active antiretroviral therapy, and 23 not on therapy. Asymptomatic hyperlactatemia (> 2 mmol/l) was found in 41 children; no hepatic dysfunction or life-threatening lactic acidosis occurred. Asymptomatic hyperlactatemia is associated with treatment with nucleoside reverse transcriptase inhibitors or protease inhibitors and with undetectable viral loads regardless of treatment regimen, suggesting that elevated lactate levels are useful in evaluating adherence. Children with HIV/AIDS live longer and healthier lives because of treatment advances such as highly active antiretroviral therapy (HAART) [1–4]. Elevated lactate levels and life-threatening hepatic dysfunction have been described in adults, especially with stavudine therapy [5–12], but no reports of elevated lactate levels in children were found on an AIDSline or Medline search. In this Institutional Review Board approved chart review, we correlated venous lactate levels in children with HIV/AIDS on HAART with their antiretroviral regimen, clinical and virological markers. A total of 127 children (1–17 years, 61 male) with perinatal HIV infection were included. The viral load (HIV-RNA quantitative polymerase chain reaction, range 400–750 000 copies/ml; Roche Amplicor 1.5 Monitor, Roche Molecular Systems, Somerville, NJ, USA) was measured every 3 months. Free flowing venous blood collected in sodium fluoride–potassium oxalate tubes was analysed for lactate (range 0.9–2.0 mmol/l). A total of 251 lactate readings were obtained, 204 being drawn in conjunction with the viral load. An additional 47 measurements were noted on chart review. For these, viral loads obtained within 6 weeks were used for statistical analysis. The SAS statistical software package (SAS Institute, Cary, NC, USA) was used. To account for the correlation among lactate scores taken from the same child, mixed linear models were used for analysis, with therapy as a fixed effect and subject as a random effect. Lactate values were raised to the power of −0.5 to correct for skewed distribution. The effect of treatment with the three pharmaceutical classes of HAART on lactate levels was analysed. The association of lactate with the viral load was also analysed. A total of 104 out of 127 patients (81.8%) were on therapy at the time of sampling. Of these, 20 (19%) were on a non-nucleoside reverse transcriptase inhibitor (NNRTI), 102 (98%) were on one or more nucleoside reverse transcriptase inhibitor (NRTI), and 95 (91%) were on one or more protease inhibitor (PI). Overall, 98 were on at least a three-drug HAART regimen involving at least two of these classes. The most common regimen was two NRTI and one PI in 57 patients (55%). Stavudine and ritonavir were most often used. Fifty-one patients (49%) were on regimens including stavudine, 38 (36.5%) included ritonavir, whereas 17 (16.3%) were on stavudine plus ritonavir-based therapy. Stavudine, ritonavir, and stavudine plus ritonavir combination regimens were analysed separately for correlation with lactate levels. The mean lactate level was 1.70 mmol/l (range 0.6–4.4, SD 0.81). Forty-one out of 127 children (32.3%) had at least one value higher than 2 mmol/l. Of the total of 251 lactate readings, 56 (25.6%) were above 2 mmol/l. Only one patient was symptomatic, with abdominal pain/nausea and vomiting, which resolved on the temporary discontinuation of medications. There were no cases of life-threatening lactic acidosis or elevated liver transaminases. Of 204 viral load readings, 85 (41.6%) were undetectable (< 400 copies/ml). Elevated lactate levels were associated with undetectable viral loads, irrespective of the treatment regimen (P = 0.01). Elevated lactate levels were also associated with therapy with NRTI (P = 0.01) or PI (P = 0.04) drugs when considered as a class. NNRTI therapy, therapy with stavudine or ritonavir alone or in combination was not associated with elevated lactate levels (Table 1).Table 1: Association of lactate with highlyhly active antiretroviral therapy and viral load.The 32% prevalence of asymptomatic hyperlactatemia in our study is similar to that (8–35%) reported in the adult literature [13,14]. A recent study suggested that chronic hyperlactatemia is common in adults on HAART, but has a poor positive predictive value for symptomatic lactic acidosis/hepatic steatosis [14]. The accompanying editorial recommended against routine lactate measurements in adults on NRTI [15]. In adults, an initial increase in lactate occurs with the initiation of therapy, followed by a prolonged plateau of lactate in a mildly elevated range [14]. This suggests that an asymptomatic and stable elevation of venous lactate results in patients who are well adherent to HAART. The utility of lactate measurements could extend beyond the evaluation of hepatic dysfunction. Significantly, in our study, elevated lactate levels were associated with an undetectable viral load irrespective of the treatment regimen. Asymptomatic hyperlactatemia may serve as a surrogate marker for adherence in children on HAART with no clinical or laboratory evidence of hepatic toxicity/disease. Lactate measurement may be especially valuable in the setting of a rising viral load, when it is unclear whether the viral load is elevated as a result of emerging drug resistance or non-adherence. Resistance testing and therapeutic drug monitoring are not widely available in many countries, and cost and access are limiting factors in their clinical use. Our findings suggest that clinicians may be able to distinguish between non-adherence and drug resistance on the basis of changes in lactate levels from a previously established baseline for an individual patient. In this context, serial lactate measurement may prove to be clinically useful. Nucleoside analogue-induced mitochondrial DNA depletion results in mitochondrial dysfunction and impaired oxidative phosphorylation [16]. The reduction in the mitochondrial to nuclear DNA ratio is correlated with symptomatic hyperlactatemia in adults [17]. We found that therapy with the NRTI or PI classes of drugs, but not NNRTI, was associated with a significant elevation in lactate levels in children. In contrast with published reports in adults, stavudine and ritonavir, alone or in combination, were not associated with elevated lactate levels. However, NRTI or PI drugs when considered as a class were associated with higher lactate levels. PI have not been directly implicated in causing hyperlactatemia. The incidence of symptomatic hyperlactatemia in adults has shown an increase over the years, and it has been speculated that either an intensification of antiretroviral therapy or the use of NRTI such as stavudine may be to blame [18,19]. Our data raise the novel possibility that therapy with PI may contribute to elevated lactate levels. This finding needs further corroboration with in-vitro and clinical studies, although the individual contribution of the concurrent agents in HAART may be difficult to determine. Our study shows that asymptomatic elevation of lactate occurred in 32% of children on HAART, but no severe hepatic dysfunction/lactic acidosis was recorded. Elevated lactate levels are associated with NRTI or PI therapy, and with an undetectable viral load, regardless of treatment regimen. Elevated lactate may be a useful correlate of adherence, especially in the setting of drug resistance when viral load is no longer a useful marker.