Disulfiram neuropathy: Two cases of distal axonopathy

Abstract

Background. Disulfiram may cause a peripheral neuropathy that is considered dose- and duration-of-exposure-related. Axonal degeneration has been described as a pathological hallmark of disulfiram toxicity, but experiments have reported a primary toxic effect of the molecule on Schwann cells and myelin. Case Reports. Case 1: At the end of two months of treatment with disulfiram 250 mg/day, a 31-year-old woman complained of weakness in distal segments of the lower limbs associated with burning dysesthesias, numbness and pain in the soles of the feet and the legs below the knees; bilateral walking steppage, reduction in foot strength, absence of ankle jerk and knee tendon reflexes, and tactile stocking pin-pick and vibratory sensory impairment in the lower limbs below the knee. Disulfiram was discontinued and she recovered partially over three months. Case 2: After one month of treatment with disulfiram 1600 mg/day, a 27-year-old man reported walking impairment, distal lower limb weakness and paresthesias. He had unsteady gait with bilateral steppage and foot drop, absence of ankle jerks and overall sensation impairment below the knee. Disulfiram was discontinued and nine months later there was almost complete recovery of motor deficits, only minor motor weakness in distal leg muscles, and no dysesthesia, sensation deficits or areflexia. In both of them clinical and neurophysiological patterns were indicative of a distal axonopathy. Discussion. The mechanisms by which disulfiram cause injury in human nerves are unclear, though may involve carbon disulfide. The discrepancy between experimental and clinical observations is still unexplained. Conclusion. We report two cases of disulfiram axonal toxicity and the partial response following discontinuation of the drug.

Keywords :

• Disulfiram
• Toxic neuropathy

Introduction

Drugs in current clinical use that may cause a peripheral neuropathy include antimicrobials, antineoplastic agents, cardiovascular drugs, hypnotics, and psychotropics. Most of these drugs cause axonal changes and a few may cause demyelinating (e.g.,. chloroquine, FK506-tacrolimus, perhexiline, procainamide, zimeldine) or mixed demyelinating/axonal neuropathies (e.g., amiodarone, suramin). Electrodiagnostic and pathological studies help in distinguish axonal from demyelinating alterations. In axonal neuropathies, electroneurography (ENG) typically shows low amplitudes of sensory/motor compound action potentials and normal or borderline slow conduction velocities (1). By contrast, in the demyelinating neuropathies, slow nerve conduction velocities, prolonged distal latencies and only slight reduced amplitudes of compound action potentials were observed (1).

Axonal degeneration with neurofilaments and organelles accumulation and “dying back” processes are the pathological hallmarks of distal axonopathy while segmental absence of myelination or thin-regenerating-myelin and “onion bulbs” (concentric layers of Schwann cell processes around an axon caused by repetitive segmental demyelination and regeneration of myelin) are typical features of demyelinating neuropathies (2).

Disulfiram (tetraethylthiuram disulphide) is a dithiocarbamoyl drug used in the treatment of alcoholism that irreversibly inhibits acetaldehyde oxidation by competing with nicotinamide adenine dinucleotide (NAD) for binding sites on liver aldehyde dehydrogenases (3,4). Increased acetaldehyde levels are thought to produce the unpleasant side effects associated with acetaldehyde syndrome (3).

Disulfiram may induce a peripheral neuropathy characterized by distal sensory impairment with loss of coordination, painful paresthesias and distal weakness with foot drop (3–18). The severity of the neuropathy is directly related to dose and duration of exposure (3,5–7,11–18). If disulfiram is not discontinued, sensory and motor impairment can progress proximally (19).

Although isolated descriptions of both axonal and myelin involvement were reported (20), wallerian-type axonal degeneration, intermediate filament accumulation, marked loss of larger myelinated fibers with no evidence of primary demyelination have been longer considered the pathological hallmarks of disulfiram toxicity (5–7,10,13–18,21,22). However, some experiments in animals showed that disulfiram causes Schwann cell damage and demyelination of peripheral nerves (4,23,24).

In order to clarify the characteristics of nerve involvement, we studied clinical and neurophysiological findings in two patients suffering from disulfiram-related neuropathy.

Case report

Case 1

A 31-year-old woman drank heavily for over five years, consuming one to two litres of wine daily. Disulfiram therapy was administered at the dosage of 250 mg daily for two months, while the patient abstained from alcohol. Two months after starting disulfiram therapy, she complained of weakness in distal segments of the lower limbs associated with burning dysesthesias, numbness and pain in the soles of the feet and the legs below the knees. About two weeks later, she noted gait difficulty with foot drop and disequilibrium. Neurological examination at this stage revealed bilateral walking steppage, bilateral reduction in foot strength (Medical Research Council [MRC] strength score: 2), bilateral absence of ankle jerk and knee tendon reflexes, and tactile stocking pin-pick and vibratory sensory impairment in the lower limbs below the knee. General examination results were normal.

Routine laboratory tests, serum B12 and folate assays, thyroid function tests, immunological and virological tests, antibody IgG and IgM to the gangliosides GM1, GM2, GD1a, GD1b, IgM to myelin-associated glycoprotein (MAG) and IgG to the anti neuronal Hu, Yo, Ri, cerebrospinal fluid examination and spinal cord magnetic resonance imaging were normal.

Motor nerve conduction velocity of the right peroneal nerve was normal while the amplitude of the compound action potential was very reduced (0.30 mV). Left peroneal nerve and both tibialis nerves were not excitable. Suralis conduction velocity was normal while amplitude of potential was markedly reduced (3 μV). Medianus and motor ulnaris nerves were normal. Sensitive conduction velocity of ulnaris was slightly decreased bilaterally (41 m/s on the right and 40 m/s on the left). Needle electromyography (EMG) showed fibrillation and polyphasic potentials with increased amplitude and duration in distal legs muscles.

During hospitalization, disulfiram treatment was discontinued. One month after stopping disulfiram, neurological examination showed some improvement in painful paresthesias, while strength and sensation deficits were unchanged. Three months later, improvement in strength (MRC 4-) was observed. Patient still had sensation deficits and complained of painful paresthesias. ENG study of peroneal and tibial nerves were unchanged while the amplitude of suralis potential was improved (5 μV). Needle EMG still showed a denervation pattern in distal leg muscles. Three further months later, neurological examination and EMG-ENG results were unchanged.

Case 2

A 27-year-old man was admitted to our Unit for a convulsion. He had a history of seizures (first episode one year earlier), a personality disorder (DSM-IV) and alcohol abuse since the age of 24 years. One and a half months before admission, he had stopped drinking and started disulfiram treatment at the unusually high dosage of 1600 mg daily. The patient reported walking impairment, distal lower limb weakness and paresthesias that had started fifteen days before admission.

Examination revealed an unsteady gait with bilateral steppage and foot drop (MRC 3), bilateral absence of ankle jerks and overall sensation impairment below the knee. General examination results were normal.

Routine laboratory tests, serum B12 and folate assays, thyroid function tests, immunological and virological tests, antibody IgG and IgM to the gangliosides GM1, GM2, GD1a, GD1b, IgM to MAG and IgG to the anti neuronal Hu, Yo, Ri, cerebrospinal fluid examination, and spinal cord magnetic resonance imaging produced normal values. Motor nerve conduction velocity of the peroneal and tibial nerves was normal, while the amplitudes of compound action potential were reduced (respectively, 0.10 mV and 2.90 mV on the right side and 0.10 mV and 3 mV on the left one). Suralis was normal. Right medianus was not excitable.

Compound action potential of the left medianus was very low (1.7 mV) while motor conduction velocity and sensitive component were normal. Motor ulnaris was normal and sensitive conduction velocity was slightly decreased bilaterally (41 m/s on the right, 42 m/s on the left). Needle EMG did not show fibrillation; isolated polyphasic potentials with increased amplitude and duration in distal leg muscles were found.

During hospitalization disulfiram treatment was discontinued. Patient was examined again nine months after the disulfiram treatment was stopped. Neurological examination showed an almost complete recovery of motor deficits, only minor motor weakness in distal leg muscles (MRC 4+), no dysesthesia, sensation deficits or areflexia. Follow-up EMG-ENG was not available.

Discussion

We have described two cases of disulfiram axonopathy. Although our patients did not report any neurological symptoms before starting disulfiram, we can not exclude a sub-clinical pre-existing alcohol-related nerve injury which may have amplified disulfiram toxicity because they were not submitted to baseline electrodiagnostic studies. Anyway, a causal role of disulfiram in the clinical picture can be inferred from the close temporal correlation between exposure and appearance of the symptoms as well as discontinuation and clinical improvement.

Neuropathy is a relatively frequent side-effect of disulfiram treatment for chronic alcoholism (19). It can be mild to severe, completely or partially reversible in a few months or, sometimes, very disabling (8,10,13,14,22). Isolated cases of acute neuropathy associated with encephalopathy following drug intoxication were described (10,17,25) but, usually, a time interval from 10 days to 18 months between starting treatment and manifesting symptoms was reported (5,8,10,14). Its course and severity are considered to be directly proportional to the dosage and duration of exposure (3,5,8,9,11–14,16,18,19). The mechanisms by which disulfiram cause injury in human nerves are unclear. Disulfiram is reduced by glutathione reductase to N,N-diethyldithiocarbamate (DEDC) which, in turn, may be converted to diethylamine and carbon disulfide (CS₂) or methyldiethyldithiocarbamate (ME-DEDC) (4).

Since prominent axonal damage has been repeatedly reported and deemed to be the hallmark of disulfiram toxicity (5–7,10,13–16,18,21,22), CS₂, which is known to induce axonal degeneration and neurofilament accumulation in experimental systems (26–28), could be a suggestive intermediary of disulfiram nerve toxicity. However, some studies in animals have not demonstrated neurotoxic CS₂ levels after direct oral intake of disulfiram which instead resulted in ME-DEDC pathway activation, Schwann cell damage and primary demyelination of peripheral nerves (4,23,24). Therefore, a DEDC-mediated demyelinating pathogenic mechanism was also supposed. The discrepancy between experimental and clinical observations is still unexplained. Our data show that, at least in humans, disulfiram neuropathy is both clinically and neurophysiologically a distal axonopathy. This appears to be in line with previous clinical reports of no evidence of primary demyelination (5–7,10,13–16,18,21,22).

Taken together, these findings indirectly seem to suggest a role for CS₂ in determining nerve toxicity; nevertheless the reason why in humans the likely CS₂-mediated axonal damage should be predominant over the ME-DEDC-induced myelin damage remains unclear.

We suggest that disulfiram effects on peripheral nerves are complex and heterogeneous. Some unknown mechanisms (intermediate or “collective” intermediate effects, environmental, individual and genetic factors, pharmacodynamic or pharmacokinetic variables, different starting and maintenance doses) might influence the metabolic pathways and therefore determine course and final characteristics of the clinical picture. It would be useful to investigate these mechanisms to better understand the toxic effect of the drug, as well as the nerve response to toxic damage.