Kaufmann P, Pascual JM, Anziska Y, Gooch CL, Engelstad K, Jhung S, DiMauro S, De Vivo DC. Nerve Conduction Abnormalities in Patients With MELAS and the A3243G Mutation. Arch Neurol. 2006;63(5):746-748. doi:10.1001/archneur.63.5.746
Mitochondrial DNA point mutations are especially deleterious to tissues with high energy demand, including the peripheral nervous system. Neuropathy has been associated with several mitochondrial diseases, including MELAS (mitochondrial encephalomyopathy, lactic acidosis, and strokelike episodes).
To evaluate nerve conduction in a genotypically and phenotypically homogeneous group of patients with MELAS and the A3243G mutation.
We studied 30 patients with MELAS and the A3243G mutation using neurophysiological techniques, medical history questionnaires, laboratory tests, and a standardized neurological examination.
Twenty-three subjects (77%) had abnormal nerve conduction measures. Symptoms suggestive of neuropathy were present in only half of the patients, but almost all had decreased reflexes or distal sensory findings on examination. Nerve conduction abnormalities were predominantly axonal and sensory and mainly present in the legs. Patients with nerve conduction abnormalities tended to be older and were more likely male.
Peripheral nerve impairment is common in those with MELAS and the A3243G mutation, and may be subclinical. Male sex and older age may add to the genetic disposition to develop neuropathy.
Mitochondrial DNA point mutations impair cellular energy metabolism. Patients usually present with symptoms of multiple organ involvement, especially of the nervous system, because of the high energetic demand of neurons. Central nervous system involvement was recognized early and is one of the defining features of the mitochondrial encephalomyopathies.
Peripheral nervous system involvement is well recognized in several mitochondrial syndromes: NARP (neurogenic muscle weakness, ataxia, and retinitis pigmentosa) is associated with a predominantly axonal neuropathy. Patients with MNGIE (mitochondrial neurogastrointestinal encephalomyopathy) and Leigh syndrome are affected by predominantly demyelinating neuropathies. Patients with MERRF (myoclonus epilepsy and ragged red fibers) have mixed demyelinating and axonal neuropathies.1- 3
Whereas peripheral neuropathy is a defining feature in these mitochondrial syndromes, it has been reported less frequently in those with MELAS (mitochondrial encephalomyopathy, lactic acidosis, and strokelike episodes), a well-recognized phenotype most commonly associated with an A→G transition at nucleotide 3243, in the transfer RNA leucine gene of mitochondrial DNA.1,4,5 Few studies have systematically evaluated MELAS patients electrophysiologically for peripheral neuropathy. Previous studies6- 8 based on clinical criteria have estimated the prevalence of neuropathy to be 15% to 22%.
The neuropathy in MELAS has been characterized as mostly axonal, but exceptional patients with a predominantly demyelinating presentation have been described.9 The neuropathy is usually chronic and progressive, but acute presentations have been described.10 Clinically, patients present with mostly mild sensory complaints and loss of proprioception and vibration in a “stocking-glove” distribution.3 The distal lower limbs are usually affected first. Motor neuropathies are less common than sensory neuropathies. There have also been reports11 of autonomic nervous system involvement, manifesting as ileus, diarrhea, or neurogenic bladder. To our knowledge, only 1 study8 has systematically evaluated A3243G mutation carriers using neurophysiological techniques, and found that 22% of patients fulfilled the diagnostic criteria for neuropathy.
To study the frequency of peripheral nerve impairment and potential contributing factors, we describe clinical and electrophysiological studies in a well-characterized cohort of patients with MELAS and the A3243G mutation.
We studied 30 individuals with the MELAS phenotype and the A3243G point mutation as part of their baseline evaluation for a clinical trial with dichloroacetate.12
We used a standardized medical questionnaire to obtain information on medical history, symptoms, and habits. The disease duration was determined from the onset of symptoms attributed to MELAS (strokelike episodes or focal seizures) to the time of evaluation. In addition, patients were examined by a neurologist (D.C.D. or P.K.) using a previously described semiquantitative tool, the Columbia Neurological Score, to evaluate the following: (1) height, weight, and head circumference; (2) general medical examination; (3) ophthalmoscopic examination; (4) cranial nerves; (5) stance and gait; (6) involuntary movements; (7) sensation; (8) cerebellar function; (12) muscle bulk, tone, and strength; (10) myotactic reflexes; (11) toe signs; and (9) other findings. These domains were scored as normal or abnormal, and summarized in the Columbia Neurological Score, ranging from 0 to 76 (with 76 being normal).13
We used the Karnofsky score, an established instrument, to evaluate performance in activities of daily living and independence of an individual on a semiquantitative scale. For example, someone able to work without complaints and without evidence of disease is given a score of 100. On the other hand, someone who is moribund or with a rapidly progressing fatal process is given a score of 10. Within the range of 0 to 100, the investigators assign a score in increments of 10. The instrument has a descriptive statement next to each numeric level (eg, a score of 50 describes someone who is moderately disabled, is dependent, and requires considerable assistance and frequent care).14
We measured fasting glucose and thyrotropin levels to screen for diabetes mellitus and thyroid disease, respectively. By using normal values established at the Columbia University Medical Center Laboratory as a reference, we considered a thyrotropin level of greater than 4.25 μU/mL and a fasting glucose level of greater than 125 mg/dL (6.9 mmol/L) as abnormal. Because dichloroacetate has been associated with peripheral nerve toxic effects, we performed nerve conduction studies at the dichloroacetate clinical trial entry visit as a baseline for safety monitoring during the treatment period. To minimize patient discomfort, we did not perform electromyography and we limited our investigation to the following nerve conduction studies: ulnar and peroneal motor studies and radial and sural sensory studies. Skin temperature was monitored during the testing. We used computerized electromyographic equipment and standard techniques. The results were considered abnormal when they were outside of 2.5 SDs from the mean of a healthy control group previously studied in our laboratory. We considered slowed conduction velocities or prolonged distal latencies to indicate demyelinating disease and reduced amplitudes to indicate axonal disease. When both types of abnormalities were present, we classified the pattern as “mixed.” The study was approved by the Columbia University Medical Center Institutional Review Board.
We compared patients with and without impaired nerve conduction measures using a t test for continuous and a χ2 test for dichotomous variables.
Of the 30 patients, 15 had symptoms such as imbalance, paresthesias, and numbness possibly related to peripheral neuropathy. Almost all (29 of 30) had abnormal findings suggestive of peripheral neuropathy on neurological examination, including abnormalities on reflex testing, sensory examination, distal muscle strength testing, and gait. None of the patients reported alcohol use or occupational exposure to potential neurotoxins.
Of the 30 patients, 23 (77%) had abnormal nerve conduction measures. Many patients (10 of 23) had sensory abnormalities only, some (8 of 23) had sensory and motor abnormalities, and a few (5 of 23) had pure motor abnormalities (Table). Most abnormalities were indicative of axonal (12 of 23) or mixed (7 of 23), rather than demyelinating (4 of 23), neuropathy. Lower extremity nerves (19 of 23 patients) were more commonly affected than upper extremity nerves.
Comparing patients with normal nerve conduction measures with those with neurophysiological abnormalities showed that disease duration, the Karnofsky score, and the Columbia Neurological Score were similar in both groups. However, patients with abnormal nerve conduction tended to be older and more were male (Table). On blood testing, 9 patients with abnormal nerve conduction study results had an abnormal fasting glucose level compared with 2 patients with normal nerve conduction study results. For thyrotropin level, 3 patients with abnormal nerve conduction had increased levels compared with 0 patients with normal nerve conduction. None of these differences was statistically significant (P>.05). A complete listing of the electrophysiological results can be found online (eTable).
Our data indicate that peripheral nervous system involvement is a frequent feature of those with MELAS and the A3243G mutation, and may be more common than previously reported.8 Our findings of predominantly axonal impairment concur with those of previous studies.8 Several patients with mixed axonal and demyelinating features had an elevated fasting glucose level, suggesting that impaired glucose homeostasis may have exacerbated the nerve damage due to impaired energy metabolism.
We did not find a statistically significant association between elevated glucose level and neuropathy, probably because both conditions were highly prevalent in our sample and the overall sample was small. However, thyroid dysfunction and impaired glucose homeostasis are common in those with MELAS; because they are possibly modifiable risk factors of neuropathy, they should be tested for in MELAS patients.
Our data suggest that males with MELAS may be more susceptible to developing neuropathy than females. This finding is in agreement with those of previous studies8 in MELAS and with the sex predilection reported for Leber hereditary optic neuropathy15 and for hearing impairment.16
Recognizing the high frequency of nerve conduction abnormalities in MELAS is important because peripheral neuropathy can be easily missed in this population experiencing cognitive deficits and coexisting central nervous system involvement. The propensity for peripheral nerve impairment in those with MELAS and the A3243G mutation may have contributed to the prohibitive peripheral nerve toxic effects associated with dichloroacetate, a lactate-lowering agent that we tested for the treatment of MELAS.12
Correspondence: Darryl C. De Vivo, MD, Department of Neurology, Columbia University, 710 W 168th St, New York, NY 10032 (firstname.lastname@example.org).
Accepted for Publication: December 22, 2005.
Author Contributions:Study concept and design: Kaufmann and De Vivo. Acquisition of data: Gooch, Engelstad, Jhung, and DiMauro. Analysis and interpretation of data: Kaufmann, Anziska, Jhung, and DiMauro. Drafting of the manuscript: Kaufmann, Anziska, Engelstad, Jhung, and De Vivo. Critical revision of the manuscript for important intellectual content: Anziska, Gooch, and DiMauro. Statistical analysis: Kaufmann, Anziska, Jhung, and DiMauro. Obtained funding: DiMauro and De Vivo. Administrative, technical, and material support: Engelstad. Study supervision: Kaufmann, Gooch, and De Vivo.
Funding/Support: This study was supported by grant K12 RR017648 (Dr Kaufmann) from the National Institutes of Health; grant PO1-HD32062 (Drs DiMauro and De Vivo) from the National Institute of Child Health and Human Development; and by the Colleen Giblin Foundation (Dr De Vivo). Dr Kaufmann is the recipient of an Irving Research Scholar Award.
Acknowledgment: We thank J. Carolina Regus for her clerical support, and Alina Melendez and Jose Garcia, MD, for their technical assistance.
Additional Resources: The online-only eTable is available.