Central Nervous System involvement in mitochondrial encephalomyopathies

Abstract

The term “Mitochondrial Encephalomyopathies” (MEs) was coined by Shapira in 1977 to describe the simultaneous involvement of central nervous system (CNS) and skeletal muscle. The CNS is metabolically very demanding and therefore vulnerable to defects of the mitochondrial respiratory chain. Cerebellar involvement, leukoencephalopathy, bilateral striatal necrosis are all manifestations of specific mitochondrial disorders. Subclinical CNS involvement is also present in the apparently pure myopathic Progressive External Ophthalmoplegia (PEO). Epilepsy and migraine are common events, particularly in MERRF and MELAS syndromes, and epilepsy strongly influences course and prognosis of mitochondrial diseases, often triggering potentially life-threatening conditions such as metabolic crisis or SLEs. In our cohort of 93 mitochondrial patients (age 16- 78), migraine was reported in 35.5 % with a much higher prevalence compared to general population independently from gender, genotype or phenotype. Migraine without aura was the most common headache (84.8 %). Mitochondrial patients with migraine showed (vs non migraineurs) younger age, significantly increased prevalence of epilepsy (p=0.0103), myoclonus (p=0.0309), stroke-like episodes (p=0.0290), EEG focal slow abnormalities (p=0.0359) and EEG epileptic focal abnormalities (p=0.0425). In the same series, 40.5 % of the 93 patients showed epileptiform abnormalities consisting of focal or multi-focal slow wave activities and spikes and 21.5 % presented clinically manifested epilepsy (partial seizures, with or without secondary generalization), mostly, but not only, MELAS and MERRF. Thus, migraine and epilepsy are not merely phenotypic aspects of specific MEs but rather the expression of vulnerability of CNS probably directly related with defects of respiratory chain. Our data support the hypothesis that mitochondrial dysfunction plays a key role in the pathophysiology of migraine and epilepsy. Stroke-Like episodes (SLEs) are the hallmark of MELAS, and can be observed even in other MEs, as POLG1-related disorders. MELAS is characterized by sudden neurological deficits similar to vascular strokes, but the pathophysiology of SLEs is still debated. Out of 41 SLEs in our patients with A3243G MELAS mutation, 64 % followed migraine and 74 % were associated with seizures. Attacks involved occipital (62 %), parietal (59 %) and temporal (51 %) lobes. Acute lesions were bilateral in 66 % of SLEs and symmetric at onset in 51 % with a very high incidence, compared with vascular strokes, of bilateral clinical syndromes as cortical visual loss or cortical deafness. Brain MRI demonstrated cortical and subcortical lesions with non-vascular distribution, migrational behavior and a mixed pattern of restricted and increased diffusion on ADC maps. Follow-up MRIs demonstrated partial reversibility of the lesions, with subsequent development of pseudolaminar necrosis and gliosis. However cerebral and/or cerebellar atrophy by far exceeded the extension of SLEs demonstrating two mechanisms playing in MELAS patophysiology: abrupt loss of function due to cell injury followed by partial recovery and an independent slowly progressive degenerative process. Beside classical CNS manifestations, dementia and extrapyramidal signs may be also present in mitochondrial disorders, sometime dominating the phenotype. In our patients, cognitive impairment was frequent revealing a consistent pattern independent from phenotype. Mental control, short term memory and visual selective attention were in fact selectively impaired in 75 % of the patients, but global intellectual decline or dementia, according to DSM-5 criteria, was also present in a minority of cases, including Alzheimer’s like syndromes. We did not found any strict correlation between neuropsychological profile, and age, clinical phenotypes and genetics, again supporting the idea that even cognitive deficits are linked to mitochondrial dysfunction and not to the effects of specific genetic changes. As far brain imaging, in our cohort, abnormal MRI findings, excluding SLEs, consisted in cerebral cortical atrophy (35 patients), cerebellar atrophy (24 patients), brain stem atrophy (11 patients), white matter abnormalities (45 patients) and signal changes in deep grey matter (25 patients). Brain atrophy was global, severe and progressive in MELAS, was milder and quite stable in MERRF and mostly absent in PEO. The study by Proton Magnetic Resonance Spectroscopy (H1-MRS) revealed normal spectra in all patients with single or multiple deletions of mitochondrial DNA at any stage of the disease, independently from severity of the phenotype and CNS involvement. Instead, in patients carrying A8344G or the A3243G non-MELAS patients we found increased peak of choline-containing compounds (Cho), variable reduction of N-acetyl-L-aspartate (NAA) and no Lactate (LA) peak in brain, while in a minority of these patients LA was present in CSF. In MELAS, LA was instead found in both affected and non-affected brain areas and in CSF. Neuroimaging, MRI spectroscopy and functional studies, as in vivo biomarkers, may help to better characterized MEs. MRI can identify common pattern in patients with MD, but only the integration with MRS, can provide useful information to guide genetic studies and follow-up in this heterogeneous group of disorders. Moreover, mitochondrial abnormalities may also occur in common neurodegenerative diseases, implying a mechanistic link between mitochondrial dysfunction and neurodegeneration. Thus, there is a strong need of biomarkers. The appropriate use of reliable biomarkers may in fact be helpful a) to identify primary mitochondrial disorders among common diseases b) to characterize mitochondrial dysfunction in neurodegenerative diseases c) for diagnosis and disease monitoring d) in defining natural history e) to characterize and better understanding mechanisms of mitochondrial dysfunction that lead to heterogeneous phenotypes f) to orient therapeutic strategies.