Seven out of the 12 patients investigated in this study received cycles of antiretroviral therapy before starting the regimens described above: subjects 1, 6, 7 and 11 were treated with ZDV (200 mg every 8 h) plus ddI (250 mg every 12 h) (mean duration of treatment 13.2 ±3 weeks, range 10–19 weeks); subjects 1, 6 and 7 then received ZDV (200 mg every 8 h) plus ddC (0.75 mg every 8 h; mean duration of treatment 6.3 weeks, range 5–8 weeks); subject 8 received a course of ZDV (200 mg every 8 h) plus ddC (0.75 mg every 8 h) for 7 weeks; subjects 2 and 12 were treated with ddC (0.75 mg every 8 h) alone and ZDV (100 mg every 4 h) alone for 9 and 33 weeks, respectively. Remarkably, during these courses of antiretroviral therapy none of them had evidence of peripheral neuropathy.
All patients reported the early onset of aching feet (mean 9.2 weeks) followed 3–4 weeks later by a burning dysesthesia of the feet but not of the hands. Clinically, at this timepoint all patients had diminished pin-prick, light touch, vibration, and temperature sensation in the feet up to midcalf level. Deep tendon reflexes of the ankles were decreased (n = 7) or absent (n = 5) with the rest of reflexes preserved. Four patients also had a mild weakness of the ankle and toe extensors. Although a coexisting, subclinical neuropathy related to HIV could not be ruled out, the abrupt onset of the neuropathy seen in these subjects was in striking contrast to the slowly progressive development of HIV-related painful, peripheral neuropathy . Furthermore, even the close temporal relationship linking the onset of the neuropathy and the administration of neurotoxic nucleoside analogues pointed to the causal role of these compounds. All patients had a normal neurological examination before starting antiretroviral treatment.
Other causes of peripheral neuropathy, such as alcohol use, diabetes, cancer, and autoimmune disorders, were ruled out on the basis of clinical and laboratory evaluation.
The withdrawal of the treatment with ddI, ddC and d4T was followed within 8 weeks by the disappearance of clinical symptoms in two subjects or an improvement of variable degree in the remaining 10.
Patients with AIDS who did not experience peripheral neuropathy while taking ZDV (n = 11; 100 mg every 4 h, mean duration of treatment 24 ± 5 weeks, range 16–26 weeks) or ddI (n = 10; 250 mg every 12 h, mean duration of treatment 21 ± 7 weeks, range 13–29 weeks) were also examined. These two groups were matched for sex and age with the study group (10 men and one woman in the ZDV group, and 10 men in the ddI group; mean age 38 ±6.4 years), range 33–51 years in the ZDV group; mean age 36 ± 9.1 years, range 31–47 years in the ddI group). These patients were comparable with subjects experiencing peripheral neuropathy also with regard to the degree of immunodeficiency, as shown by CD4 counts (85 ± 161 x 106/l, range 0–498 in ZDV group; 93 ± 175 × 106/l, range 0–521 in the ddI group).
The nutritional status of all subjects with AIDS, whether they had neuropathy or not and irrespective of the antiretroviral treatment, was preserved with respect to their clinical condition, as assessed by clinical examination, measurements of triceps skin-fold thickness and mid-arm circumference, and laboratory data. Remarkably, all patients had stable body weight or maintained their weight within a 10% variation range during the previous 4 months. Furthermore, they had albumin levels above 4 g/dl and total carnitine levels comparable with healthy individuals. None had persistent or severe diarrhoea in the previous 6 months and nutritional support was not requested. None had clinical or laboratory evidence of kidney dysfunction.
Ten HIV-negative subjects with non-HIV-related neuropathies (five with acute idiopathic demyelinating polyradiculoneuropathy, four with idiopathic axonal neuropathy, one with idiopathic motor neuropathy), and 13 healthy individuals served as additional control groups.
Study subjects and controls were accrued over a period of 26 months from various centres.
Measurement of carnitines in serum. Case and control sera were stored at −80°C until assayed in a blinded and randomized manner for acetyl-carnitine levels. The radioenzyme method described by McGarry and Foster, with minor modifications, was used for measuring carnitines, as previously described [24,25].
Statistical comparisons between the study subjects and the control groups were made by the two-tailed t test.
The serum levels of total carnitine were comparable between the study subjects and the control groups (43.4 ± 11.3, 46.6 ± 9.7, 45.7 ± 8.7, 44.2 ± 10.4, and 41.6 ±7.8 µmol/l in subjects with nucleoside analogue-related neuropathy, ZDV-treated subjects without evidence of neuropathy, ddI-treated subjects without evidence of neuropathy, HIV-negative subjects with axonal or autoimmune neuropathies, and healthy individuals, respectively). The finding of normal levels of total carnitine in the study subjects was considered to further indicate their preserved nutritional status. The serum levels of acetyl-carnitine in the study subjects and the control groups are shown in Fig. 1.
Subjects who developed axonal peripheral neuropathy on treatment with nucleoside analogues had significantly lower levels of acetyl-carnitine (3.2 ± 1.4, range 1.2–6.0 µmol/l) than subjects who did not experience peripheral neuropathy while taking ZDV or ddI (6.7 ± 2.2 µmol/l, range, 3.9–11 µmol/l, P<0.001, in ZDV group 5.7 ± 1.0 µmol/l, range 4.1–7.3 µmol/l, P<0.001, in the ddI group) and healthy individuals (6.5 ±1.0 µmol/l, range 5.1–9.0 µmol/l, P < 0.001).
Remarkably, nine (75%) out of 12 subjects in the study group had acetyl-carnitine levels below the lowest limit (3.9 µmol/l) we measured in ZDV- or ddI-treated subjects without evidence of peripheral neuropathy.
The patient in the study group who had just started treatment with ZDV, having changed from ddC administered over a 17-week period, had serum levels of acetyl-carnitine (5.3 µmol/l) approximately around the lower limit of the normal range.
The mean acetyl-carnitine levels (6.3 ± 3.07 µmol/l, range 2.01–12.95 µmol/l) in sera from HIV-negative subjects with axonal or demyelinating autoimmune neuropathies were as comparable as in healthy individuals. Only two subjects in this group had levels of acetyl-carnitine below the lowest limit measured in ZDV- or ddI-treated subjects who did not develop the neuropathy, but the remaining had levels in the normal or even the high range.
The use of the nucleoside analogues ddI, ddC and d4T may be burdened by the onset of a severe dose-limiting peripheral neuropathy [1–4]. Their mitochondrial toxicity, as shown by the inhibition of mitochondrial DNA synthesis, is considered central to the pathogenesis of this disorder [5,6] and explains why, despite interrupting the treatment, patients may continue to worsen for a number of weeks.
We report here that patients requiring evaluation for clinically manifest axonal, painful peripheral neuropathy developed during treatment with neurotoxic nucleoside analogues had reduced levels of acetyl-carnitine in their sera. By contrast, subjects who did not experience peripheral neuropathy while taking ZDV or ddI had normal levels of acetyl-carnitine. Thus, the finding of low levels of acetyl-carnitine is associated with the development of peripheral neuropathy during treatment with nucleoside analogues.
These subjects, however, had normal levels of total carnitine in their sera. This appears to suggest that certain pathways of carnitine metabolism, such as the acetylation of carnitine to acetyl-carnitine, are impaired in these subjects, probably as a result of the toxicity of ddI, ddC and d4T. Other causes of acetyl-carnitine deficiency  were ruled out on the basis of clinical and laboratory assessment. Furthermore, the normal levels of total carnitine and albumin point to the preserved nutritional status of these subjects and support the view that the acetyl-carnitine deficiency observed is independent of nutritional factors.
The comparable CD4 counts observed in the study group and in ZDV- and ddI-treated subjects who did not experience neuropathy excluded the hypothesis that the lower levels of acetyl-carnitine found in AIDS patients with neuropathy might be related to a greater degree of immunodeficiency. We also investigated the possibility that acetyl-carnitine deficiency was a function of the duration of antiretroviral therapy. However, the duration of drug exposure in subjects with neuropathy was comparable or even shorter than in subjects taking ZDV or ddI who did not develop peripheral neuropathy.
Our study demonstrates that in ZDV- or ddI-treated patients who did not develop peripheral neuropathy acetyl-carnitine levels may be slightly reduced, although marked acetyl-carnitine deficiency is unusual. In the individuals in this group other factors, including HIV infection itself, could be responsible for the slight reduction of acetyl-carnitine. In fact, previous reports have shown that acetyl-carnitine deficiency may be occasionally detected in sera from patients with AIDS [24,27] and a deficiency of carnitine is present in the skeletal muscles of subjects taking ZDV [28,29]. Whether a prolonged duration of treatment with ddI could result in marked acetyl-carnitine deficiency, irrespectively of the development of peripheral neuropathy, remains to be established.
Differences in the degree of mitochondrial DNA polymerase inhibition by nucleoside analogues or their tissue specificity have been shown [5,6]. The metabolism of nucleoside analogues in active nucleotides can vary from one tissue to another and, as a consequence, the tissues affected by nucleoside analogues can be quite different although these compounds have the same ultimate mechanism of action. The hypothesis that acetylcarnitine deficiency may be involved in the mechanism of neurotoxicity of nucleoside analogues is reasonable given the injury to mitochrondria mediated by these agents and the effects of acetyl-carnitine on the peripheral nerves [9–19], as well as on mitochondrial metabolism [20–22].
Two out of 10 patients with non HIV-related autoimmune neuropathies had levels of acetyl-carnitine below the lower limit measured in ZDV- or ddI-treated subjects without evidence of peripheral neuropathy. This appears to reinforce the view that acetyl-carnitine deficiency, in combination with other toxic or immune mechanisms, may be one determinant of injury to the peripheral nerves.
The significance of acetyl-carnitine deficiency in neuropathic patients needs to be further investigated. Nevertheless, our study suggests that, among AIDS patients, only those experiencing peripheral neuropathy during treatment with ddI, ddC and d4T have acetylcarnitine deficiency. Since acetyl-carnitine is currently available for the treatment of neuropathies and has no toxic effects, the exogenous supplementation of acetylcarnitine in order to prevent or treat the neurotoxicity of nucleoside analogues appears a reasonable hypothesis.
We thank M.C. Dalakas (National Institutes of Health, Bethesda, Maryland, USA) for providing us sera assayed in this study and critically discussing the paper.
1. Simpson DM, Tagliati M: Neurologic manifestations of HIV infection
. Ann Intern Med
2. Faulds D, Brogden RN: Didanosine. A review of its antiviral activity, pharmacokinetic properties and therapeutic potential in human immunodeficiency virus infection
3. Whittington R, Brogden RN: Zalcitabine. A review of its pharmacology and clinical potential in acquired immunodeficiency syndrome (AIDS)
4. Browne MJ, Mayer KH, Chafee SB, et al.
: 2′, 3′-didehydro-3′-deoxythymidine (d4T) in patients with AIDS or AIDS related complex: a phase I trial
. J Infect Dis
5. Chen CH, Vasquez-Padua M, Cheng YC: Effect of anti-human immunodeficiency virus nucleoside analogs on mitochrondrial DNA and its implication for delayed toxicity
. Mol Pharmacol
6. Lewis W, Dalakas MC: Mitochondrial toxicity of antiviral drugs
. Nature Med
7. Bremer J: The role of carnitine in intracellular metabolism
. J Clin Chem Clin Biochem
8. Colucci WJ, Gandour RD: Carnitine acetyltransferase: a review of its biology, enzymology, and bioorganic chemistry
. Bioorg Chem
9. Pettorossi VE, Brunetti V, Carrobi C, Della Torre G, Grassi S: L-acetylcarnitine enhances functional muscle reinnervation
. Drugs Exp Clin Res
10. Rampello L, Giammona G, Aleppo G, Favit A, Fiore L: Trophic action of acetyl-L-carnitine in neuronal cultures
. Acta Neurol
11. Forloni G, Angeretti N, Smiroldo S: Neuroprotective activity of acetyl-L-carnitine: studies in vitro
. J Neurosci Res
12. Rosenthal RE, Williams R, Bogaert YE, Getson PR, Fiskum G: Prevention of postischemic canine neurological injury through potentiation of brain energy metabolism by acetylcarnitine
13. De Angelis C, Scarfo C, Falcinelli M, Reda E, Ramacci MT, Angelucci L: Levocarnitine acetyl stimulates peripheral nerve regeneration and neuromuscular junction remodelling following sciatic nerve injury
. Int J Clin Pharmacol Res
14. Tenconi B, Donadoni L, Germani E, et al.
: Intraspinal degenerative atrophy caused by sciatic nerve lesions prevented by acetyl-L-carnitine
. Int J Clin Pharmacol Res
15. Piovesan P, Pacifici L, Taglialatela G, Ramacci MT, Angelucci L: Acetylcarnitine treatment increases choline acetyltransferase activity and NGF levels in the CNS of adult rats following total fimbria fornix transection
. Brain Res
16. De Simone R, Ramacci MT, Aloe L: Effect of acetyl-L-carnitine on forebrain cholinergic neurons of developing rats
. Int J Dev Neurosci
17. Taglialatela G, Angelucci L, Ramacci MT, Werrbach Perez K, Jackson GR, Perez Polo JR: Stimulation of nerve growth factor receptors in PC12 cells by acetyl-L-carnitine
. Biochem Pharmacol
18. Taglialatela G, Navarra D, Cruciani R, Ramacci MT, Alema GS, Angelucci L: Acetyl-L-carnitine increases nerve growth factor levels and choline acetyltransferase activity in the central nervous system of aged rats
. Exp Gerontol
19. Castorina M, Ferraris L: Acetyl-L-carnitine affects aged brain receptorial system in rodents
. Life Sci
20. Gadaleta MN, Petruzzella V, Renis M, et al.
: Reduced transcription of mitochondrial DNA in the senescent rat. Tissue dependence and effect of L-carnitine
. Eur J Biochem
21. Villa RF, Turpenoja L, Benzi G, et al.
: Action of L-acetylcarnitine on age-dependent modifications of mitochondrial membrane proteins from rat cerebellum
. Neurochem Res
22. Villa RF, Gorini A: Action of L-acetylcarnitine on different cerebral mitochondrial populations from hippocampus and striatum during aging
. Neurochem Res
23. Cornblath DR, McArthur JC: Predominantly sensory neuropathy in patients with AIDS and AIDS-related complex
24. De Simone C, Tzantzoglou S, Jirillo E, Marzo A, Vullo V, Arrigoni Martelli B: L-carnitine deficiency in AIDS patients
25. De Simone C, Famularo G, Tzantzoglou S, Trinchieri V, Moretti S, Sorice F: Carnitine depletion in peripheral blood mononuclear cells from patients with AIDS: effect of oral L-carnitine
26. Editorial: Carnitine deficiency
27. Famularo G, De Simone C: A new era for carnitine? Immunol Today
28. Dalakas MC, Leon-Monzon ME, Bernardini I, Gahl WA: The AZT-induced myopathy is associated with muscle carnitine deficiency and lipid storage
. Ann Neurol
29. Semino-Mora MC, Leon-Monzon ME, Dalakas MC: Effect of L-carnitine on the zidovudine-induced destruction of human myotubes. Part I: L-carnitine prevents the myotoxicity of AZT in vitro
. Lab Invest
Keywords:© Lippincott-Raven Publishers.
Acetyl-carnitine; didanosine; zalcitabine; stavudine; zidovudine; axonal peripheral neuropathy