The mechanisms of peripheral neuropathy caused by HIV and neurotoxic antiretroviral therapy (ARV) are unknown . Carnitine is a critical substrate in the production of energy during the beta-oxidation of free fatty acids. Famularo et al.  reported that serum acetyl carnitine levels are depressed in HIV-infected patients with neuropathy associated with didioxynucleoside (`d-drug') ARV. We report an analysis of plasma carnitine levels in subjects enrolled in a controlled trial of recombinant human nerve growth factor in the treatment of peripheral neuropathy associated with HIV and d-drug therapy.
AIDS Clinical Trials Group Protocol 291 was a randomized, placebo-controlled trial of recombinant human nerve growth factor in the treatment of patients with HIV-associated painful peripheral neuropathy . Plasma levels of free, acetyl and total carnitine were assayed using tandem mass spectrometry in a central laboratory (D.M.), using methods as described previously . We assessed the association of baseline carnitine levels with several measures of neuropathy, including the subject's rating of pain (Gracely scale), global pain assessment, as rated by subject and examiner, neurological examination (pin perception in feet), epidermal skin biopsy nerve fiber density results , plasma HIV-RNA levels (Roche Amplicor, Roche Diagnostic Systems, Inc., Branchburg, NJ, USA), CD4 lymphocyte counts and quantitative sensory testing (QST). Vibration and cooling detection thresholds were measured using the computer-assisted sensory evaluator IV . An abnormal QST score was defined as a score higher than the 95th percentile compared to normative controls.
The association of carnitine levels with markers of neuropathy was measured by the Spearman correlation coefficient. Mean carnitine levels were compared between subgroups determined by the severity of neuropathy using the non-parametric Wilcoxon two-sample and Kruskal–Wallis tests. Mean levels of the other neuropathic markers were compared between groups with abnormal and normal carnitine.
A total of 232 out of the 270 subjects enrolled had carnitine samples available for analysis at baseline. Abnormal levels for total, free and acetyl carnitine were calculated as those being below the bottom fifth percentile compared with healthy controls (1.645 standard deviations below the mean). The cut-off point for abnormal levels of total carnitine was 29.16 μM (48.9 − 1.645 × 12.00), for free carnitine it was 25.08 μM (41.20 − 1.645 × 9.80), and for acetyl carnitine it was 2.01 μM (5.88 − 1.645 × 2.35).
There were 20 subjects (8.6%) with abnormal total carnitine, 32 subjects (13.7%) with abnormal free carnitine and six subjects (2.6%) with abnormal acetyl carnitine. The percentage of subjects with abnormal total and free carnitine levels was significantly higher among study subjects (one-sided z-test P values of 0.0060 and < 0.0001, respectively). There was no significant difference in the percentage of subjects with abnormal acetyl carnitine levels compared with normal controls (P = 0.995).
There were also no correlations between total, free or acetyl carnitine levels and pain, QST measures, epidermal nerve fiber density, plasma HIV-RNA levels or CD4 lymphocyte counts. Conversely, no difference in total, free or acetyl carnitine levels was detected between subgroups of subjects defined by baseline pain levels or QST scores (Table 1).
Patients were stratified based on exposure to d-drug ARV as current use (N = 86), discontinuation 8–26 weeks before study entry (N = 21), and discontinuation 26 weeks before study entry (N = 125). No differences were found in total, free or acetyl carnitine levels between these groups on the basis of d-drug ARV exposure. There were also no differences in any of the neuropathy indices between subjects with normal and abnormal carnitine levels.
These results do not support an association between plasma carnitine levels and the severity of HIV-associated peripheral neuropathy, as measured by subjective and objective indices. There is no evidence that carnitine levels differ between patients with neuropathy, according to whether or not they were exposed to neurotoxic d-drugs. It is not clear why these results differ from those previously reported . However, the number of subjects in that study was small, and a control group of subjects with primary HIV-associated neuropathy, unexposed to d-drugs, was not included. In contrast, the current study, although the largest controlled therapeutic trial in HIV neuropathy reported to date, did not include HIV-infected subjects without peripheral neuropathy. Notably, our study revealed that the proportion of the subjects with abnormally low levels of total and free carnitine is higher than normative controls. It is possible that an analysis of carnitine levels in HIV-infected subjects without neuropathy would reveal similar abnormalities. However, the absence of an association between carnitine levels and any measures of neuropathy in this study do not provide support for a role of carnitine deficiency in the pathogenesis of peripheral neuropathy associated with HIV or dideoxynucleoside therapy.
The authors would like to thank Sigma–Tau Pharmaceuticals who provided partial support for the carnitine assays, and Taiwo Baker who provided expert technical assistance.
David M. Simpsona
and the AIDS Clinical Trials Group Protocol 291/860 Study Team
1. Simpson D, Tagliati M. Nucleoside analogue-associated peripheral neuropathy in human immunodeficiency virus infection. J Acquir Immune Defic Syndr 1995, 9: 153–161.
2. Famularo G, Moretti S, Marcellini S. et al
. Acetyl-carnitine deficiency in AIDS patients with neurotoxicity on treatment with antiretroviral nucleoside analogues. AIDS 1997, 11: 185–190.
3. McArthur JC, Yiannoutsos C, Simpson DM. et al
. A phase II trial of nerve growth factor for sensory neuropathy associated with HIV infection. AIDS Clinical Trials Group Team 291.
Neurology 2000, 54: 1080–1088.
4. Stevens RD, Hillman SL, Worthy S, Sanders D, Millington DS. Assay for free and total carnitine in human plasma using tandem mass spectrometry. Clin Chem 2000, 46: 727–729.
5. McCarthy BG, Hsieh ST, Stocks A. et al
. Cutaneous innervation in sensory neuropathies: evaluation by skin biopsy. Neurology 1995, 45: 1848–1855.
6. Dyck PJ, Zimmerman I, Gillen DA, Johnson D, Karnes JL, O'Brien PC. Cool, warm, and heat-pain detection thresholds: testing methods and inferences about anatomic distribution of receptors. Neurology 1993, 43: 1500–1508.