The clinical spectrum and long-term outcome of 73 children diagnosed with mitochondrial diseases between 1985 and 2005 were investigated at the Universities of Montreal and Toronto, Canada. Phenotypic categories included neonatal-onset lactic acidosis (10%), Leigh syndrome (18%), nonspecific encephalopathy (32%), mitochondrial encephalomyopathy (19%), visceral (11%), and Leber hereditary optic neuropathy (5%). Age at onset was a median of 7 months (range, prenatal to 16 years). Presenting symptoms were neurologic in 22%, including seizures, ataxia, extrapyramidal movement disorders, muscle weakness, ptosis, and headache. Neurologic presentation was acute in 35%, with stroke-like episodes, intermittent ataxia, episodic peripheral weakness, and recurrent muscle cramps. MRI showed basal ganglia hyperintensities in 46%, cerebral atrophy in 47%, brainstem lesions in 34%, and cortical infarcts in 10%. One third of the cohort developed acute acidotic crises, usually associated with benign infectious disease and usually in infancy. At follow-up, 66 patients (90%) showed clinical signs of cerebral involvement, and 29% had visceral involvement. Molecular diagnoses were established in 81%, and a mitochondrial DNA mutation was found in 20%. Mortality was 46% at a median age of 13 months, 80% <3 years of age. Patients with first symptoms before age 6 months had a tenfold increased risk of mortality, and age at first symptoms was an independent predictor of mortality. Cardiac or visceral involvement and neurologic crises were not independent outcome factors. Of 32 patients with disease onset >5 years, 62% had a favorable outcome, with only mild impairment or normal functional independence. 
COMMENT. A high level of suspicion for mitochondrial disease (MD) is recommended in the evaluation of patients with unexplained organ dysfunction. “Any age, any symptom, any organ” is an appropriate description of MD (Munnich A et al. 1996; cited by Debray et al). In diagnosis of MD, the above authors propose initial least-invasive techniques such as fibroblast culture, and blood DNA testing for specific etiologies. Muscle and/or liver biopsy should be deferred, pending the results of fibroblast culture and blood DNA. In a large series of MD patients published in 1995 , the most useful confirmatory diagnostic test was histochemical analysis of muscle. Elevated plasma and CSF lactate are good indicators of MD (an especially high plasma lactate is a predictor of poor outcome), but specific etiologies require molecular diagnosis.