The Neuroradiology Journal 22: 558-563, 2009
www.centauro.it
The Brain-Heart Connection in Mitochondrial Respiratory Chain Diseases M. CORDEIRO*, F. SCAGLIA**, S. LOPES DA SILVA*, P. GARCIA***, M. GRAZINA****, C. MOURA*, L. DIOGOL** * Neuroradiology, Medical Imaging Department, Faculty of Medicine, University Hospital Coimbra; Portugal ** Department of Molecular and Human Genetics, Baylor College of Medicine and Texas Children’s Hospital; Houston, Texas, USA *** Metabolic Diseases Unit, Center for Child Development “Luís Borges”, Pediatric Hospital of Coimbra - CHC EPE; Coimbra, Portugal **** Center for Neuroscience and Cell Biology, University of Coimbra; Coimbra, Portugal
Key words: mitochondrial respiratory chain diseases, metabolic diseases, magnetic resonance, heart, cerebellum, brainstem, thalamus
SUMMARY – Mitochondrial respiratory chain disorders (MRCD) are a heterogeneous group of diseases leading to an inadequate production of ATP. Brain and heart are among the most affected organs. Thus far, no specific relationship has been found between specific affected areas in the central nervous system and cardiac involvement. This study investigated the relationship between specific brain involvement and heart disease in mitochondrial disorders. We hypothesize that specific areas of brain lesions in children with MRCD are more frequently correlated to heart disease than others. A retrospective evaluation of the clinical records of 63 children with a definite MRCD, was performed searching for heart disease, namely, dilated and hypertrophic cardiomyopathy and arrhythmia. Brain MR images were evaluated and characterized regarding specific areas of atrophy and involvement. These findings were correlated using the Fischer exact test whose strength was determined with the Phi coefficient. During the period analyzed, 13 children (20.6%) developed cardiac disease, of whom nine (14.3%) exhibited isolated cardiomyopathy, one (1.6%) exhibited arrhythmia and three both. The main MRI abnormalities observed were brain atrophy (65.1%) and among this group 17.5% of subjects had cerebellar involvement. In addition, supratentorial, cerebellar and brainstem white and grey matter lesions were also found. There was a statistically significant relationship between progression to cardiac disease and cerebellar atrophy (Fisher’s Exact Test –0.005 and Phi 0.394) and lesions in the cerebral peduncles (0.035/0.358). Moreover, there was an additional correlation between thalamic lesions and progression to hypertrophic myocardiopathy (0.029/0.397). A statistical relationship between thalamic, mesencephalic and cerebellar involvement and cardiac disease in children with definite MRCD was observed. The true significance of this connection warrants further assessment.
Introduction Mitochondrial respiratory chain disorders (MRCD) are a genetically and clinically heterogeneous group of disorders caused by a primary derangement of oxidative phosphorylation 1 and impaired cellular ATP production. They can present at any age and may have any mode of inheritance. MRCD have a minimal established incidence of 1 in 10,000 births and are likely to be more frequent 2. The defini558
tion of mitochondrial disease is not consensual since some experts in the field consider this diagnosis in the case of any disorder resulting from any abnormality occurring inside the mitochondria (including abnormal cardiolipin deficiency as observed in Barth syndrome or abnormal iron deposition as observed in Friedreich ataxia), whereas others only consider this diagnosis in the context of diseases resulting from a malfunctioning respiratory chain. Mitochondria are more critically needed in those or-
M. Cordeiro
Figure 1 Axial fast spin-echo T2-weighted image (FSE. TR/TE 3500/96 ms) of a brain MRI of a three-year-old girl with a definite MRCD (MRC complex IV deficiency in muscle and liver) showing cerebellar atrophy. She developed cardiomyopathy as her disease progressed.
The Brain-Heart Connection in Mitochondrial Respiratory Chain Diseases
Figure 2 Axial fast spin-echo T2-weighted image (FSE. TR/TE 3500/96 ms) of a brain MRI of a 2-year-old boy with a definite MRCD showing cerebral peduncle lesions. He developed cardiac arrhythmia during disease progression.
gans with a higher metabolic rate such as the brain, heart and skeletal muscle and to a lesser extent kidney, liver or bowel which explains why these conditions exhibit a wide range of clinical symptoms. Organs with higher metabolic demands (e.g. the brain) are more likely to show clinical manifestations of mitochondrial dysfunction. Thus, neurological signs and symptoms are the most common presenting features in all age groups. The heterogeneity of the clinical presentation makes these diseases difficult to diagnose. Diagnostic criteria for pediatric MRCD have been therefore adapted from previously published adult criteria 3. These parameters involve clinical data, histological analysis (namely the presence of red ragged fibers or subsarcolemmal accumulation of mitochondria), the activity of the mitochondrial respiratory chain (MRC), quantification of fibroblast ATP synthesis rates, identification of a nuclear or mtDNA mutation of undisputed pathogenicity and metabolic indicators of impaired MRC function. Brain imaging is a non invasive test that may enhance the probability of the diagnosis during the diagnostic work-up. Patients with mitochondrial disease often present a combinaFigure 3 Axial fast spin-echo T2-weighted image (FSE. TR/TE 4000/100 ms) of a brain MRI of a 5-year-old boy with a definite MRCD showing symmetrical thalamic lesions. During the course of disease he developed hypertrophic carmyodiopathy.
559
The Brain-Heart Connection in Mitochondrial Respiratory Chain Diseases
tion of abnormalities from different patterns on their MRI scans. Brain volume loss, leukodystrophy, delayed myelination, brain stem and deep gray matter lesions have been frequently reported 4,5,6. In a previous, still unpublished study 7, we tried to find a predictive value to brain MRI lesions in children with MRCD. We correlated brain MRI findings with clinical and laboratory data both at disease presentation and after the full manifestations of disease were present and correlated them using Fisher’s exact test. We found that some brain lesions were significantly associated with specific patterns of clinical progression. Most could easily be explained by our knowledge of physiology and neurology. We found, for instance, a strong and significant correlation between brainstem lesions and later development of respiratory failure. However, one association was not easily explainable: children with a definite MRCD and posterior fossa involvement on brain MRI frequently developed heart disease. This paper describes the study where such an association was found and discusses a possible mechanistic explanation comparing this finding with similar data found in the literature. Methods Two hundred children investigated for MRCD in our institution from 1997 to 2006 were classified according to Bernier criteria for MRCD in children 3. Diagnosis was based on clinical features and metabolic, enzymatic, histological and genetic analyses results. Skeletal muscle, liver and/or myocardium samples were investigated in all patients 8. A definite diagnosis was considered if two major or one major and two minor criteria were fulfilled. The presence of both clinical and laboratory criteria was a prerequisite. The remaining patients were excluded from this study. It was a retrospective study approved by the local ethics committee. MRI was done in 149 of the 179 pediatric patients having brain image analysis performed (ultrasound, computed tomography and/or MRI). Sixteen of those 149 patients had a final diagnosis other than a MRC disorder. Brain imaging findings were not used as a single diagnostic criterion, in agreement with the Bernier diagnostic criteria 3. From the 133 children with a primary disorder suggestive of MRCD who underwent brain MRI, we selected 63 with a definite diagnosis. Median age at clinical onset was six months (mean±sd: 9.2±7.21 months; 560
M. Cordeiro
range: one day of life - six years old); 33 were female with a sex ratio of M/F is 0.9. Sixty-three per cent of patients had a MRC defect (enzymatic activity below 30% of the normal mean, expressed as a ratio using citrate synthase) in at least one tissue sample (skeletal muscle, liver and/or myocardium: isolated deficiencies of complexes IV and I were the most frequent (42% and 29%, respectively). Four patients had nDNA abnormalities (three in SURF1 gene and one in desoxiguanosine kinase gene) and four presented mtDNA mutations (8993T>G; 3302A>G and 11778G>A single base substitutions and a 4404 bp deletion). MRI scans were performed on a whole body MR system (1.5-Tesla scanner (Signa Infinity; General Electric Medical Systems, Milwaukee, WI, USA) using a standard circularly polarized head coil. Every child was anesthetized for the examination. Five sequences were acquired for each patient: sagittal spin-echo T1-weighted images (TR, 500-600 ms; TE, 12 ms), axial and coronal fast spin-echo T2-weighted images (TR, 3500-5000 ms; TE, 90-120 ms) and fast spin-echo fluid-attenuated inversion recovery (FLAIR) images in either the axial or sagittal plane of 5 mm thickness with 6 mm spacing. The field of view was 22-24 cm. The children had a thorough clinical evaluation already published elsewhere 8. For this study we focused on the presence of cardiomyopathy (diagnosed clinically and with echocardiogram) or arrhythmias (diagnosed clinically and with electrophysiological studies) both at presentation and during the course of disease (a definite timing was not defined for this interval, but it was in every case superior to six months and when a full clinical picture was present). Some patients had more than one MRI, in which case the most recent scan was chosen for analysis. Median age of the MRI brain study selected for evaluation was 3.7 years [mean±sd: 3.9 years ± 3.8 (five days of life - 18 years old)]. Brain images were independently reviewed by two neuroradiologists and one neuroradiology resident. Brain atrophy was evaluated qualitatively upon agreement between the two neuroradiologists. It was further analyzed according to the region involved. Other causes of volume loss in the developing brain, such as brain injury sequelae, hydrocephalus, ischemia or encephalitis were excluded. White or gray matter lesions, defined as high signal intensity at long TR sequences, were also evaluated according to the affected brain area. The neuroimaging features of the 63 children
www.centauro.it
The Neuroradiology Journal 22: 558-563, 2009
were correlated with the type of cardiac involvement, using Fisher´s exact test to assess the significance and Phi test to assess the strength of the correlation. SPSS software package version 169 was used for the analysis. Test results were considered significant when p= G mutation in the mitochondrial tRNALys gene has been observed in cerebellum, cerebrum and heart, establishing a link between these organs 14. Kearns-Sayre syndrome is characterized by a triad of features including 1) onset in subjects younger than 20 years; 2) chronic, progressive, external ophthalmoplegia; and 3) pigmentary degeneration of the retina. In addition. KearnsSayre syndrome may include cardiac conduction defects, cerebellar ataxia and raised cerebrospinal fluid (CSF) protein levels (>100 mg/dL) 15. Deleted mtDNA can be found in skeletal and cardiac muscle, brain cortex, cerebellum, liver and kidney in postmortem studies of patients with Kearn-Sayre Syndrome 16. Histologicaly, spongiosis in the frontal cortex, diffuse loss of Purkinje cells in the cerebellum, and heart fibrosis with mitochondrial abnormalities have been shown 17. Coenzyme Q(10) is a mobile lipophilic electron carrier located in the inner mitochondrial membrane. Defects of coenzyme Q(10) biosynthesis represent one of the few treatable mitochondrial diseases. These disorders can present with hypertrophic cardiomyopathy and cerebellar atrophy with ataxia in the context of CoQ9 mutations 18 providing a defined molecular link for mitochondrial dysfunction affecting the heart and cerebellum mithocondria. The role of the brainstem regulating heart rate is well known 19. Thus, lesions in the brainstem of MRCD patients would expectedly lead to cardiac rhythm disturbances. Such an association was not observed, but this potential as562
M. Cordeiro
sociation would deserve a more careful evaluation as part of a prospective study. Nonetheless, a more generic relationship was found between cerebral peduncular lesions and progression to cardiac disease in general (arrhythmia and cardiomyopathy). A relationship was found between thalamic lesions and progression to hypertrophic cardiomyopathy. A potential relationship between thalamic involvement and cardiac disease has been documented in the past. The thalamus has been involved in cardiac rhythm modulation 20. Some potential weaknesses of our study may stem from the following facts, i.e. It is a retrospective study that resulted from a new analysis of data previously collected for a study designed to show the specificity of brain imaging features in MRCD 6 and the predictive value of these lesions 7. We could have analyzed more general areas of brain damage (for instance. the brainstem as a whole) instead of specific regions. Ideally, a more comprehensive genetic and biochemical analysis should have been performed in cardiac tissue or other pertinent tissues of children with cardiomyopathy, and the observed specific defects should have been analyzed in conjunction with the brain MRI findings to establish a potential association with an underlying biochemical and molecular mitochondrial signature. The major strengths are the relatively large population of children analyzed (63 children with a definite diagnosis) and the rigorous statistical analysis. The observed associations are by no means conclusive. One way to strengthen a possible connection between brain and heart would be to look from a longitudinal prospective at a cohort of patients with a definite MRCD. A relationship between brain MRI findings, cardiac involvement, and specific MRC defects and molecular (mitochondrial or nuclear) defects observed from endomyocardial or other pertinent tissues should be performed to further elucidate the significance of these observations. Conclusion Neurocardiology is a recent and growing medical specialty. Mitochondria are essential for cardiac muscle and brain functioning; a potential relationship between both of them is understandable. The data suggest a greater prevalence of heart disease in children with MRCD and cerebellar atrophy, cerebral pedun-
www.centauro.it
cular lesions and thalamic lesions. Further investigations are needed to determine the clinical impact of this observation and to determine
The Neuroradiology Journal 22: 558-563, 2009
common biochemical and molecular mitochondrial abnormalities that could underlie this association.
References 1 Naviaux RK: Developing a systematic approach to the diagnosis and classification of mitochondrial disease. Mitochondrion 4 (5-6): 351-61, 2004. 2 Scaglia F: Genetics of White Matter Disorder – Cardiomyopathy and Mitochondrial Cytopathy, 2008 ASNR Proceedings, ASNR meeting, June 2-5, New Orleans, USA, 2008. 3 Bernier FP, Boneh A, Dennett X et Al: Diagnostic criteria for respiratory chain disorders in adults and children. Neurology 59: 1406-1411, 2002. 4 Kang PB, Hunter JV, Melvin JJ, Selak MA, Faerber EN, Kaye EM: Infantile leukoencephalopathy owing to mitochondrial enzyme dysfunction. J Child Neurol 17: 421-8, 2002. 5 Scaglia F, Wong LJ, Vladutiu GD et Al: Predominant cerebellar volume loss as a neuroradiologic feature of pediatric respiratory chain defects. Am J Neuroradiol 26: 1675-80, 2005. 6 Cordeiro M, Diogo L, Garcia P et Al: Brain MR Imaging in the Diagnosis of Mitochondrial Respiratory Chain Disorders in Children; New Orleans ASNR 2008 Proceedings 2008: 7. 7 Cordeiro M, Diogo L: Predictive Value of Brain MR Findings in Children with Mitochondrial Respiratory Chain Disorders; New Orleans ASNR 2008 Proceedings 2008: 7. 8 Diogo L, Grazina M, Garcia P et Al: Pediatric mitochondrial respiratory chain disorders in the Centro region of Portugal. Pediatr Neurol 40: 351-6, 2009. 9 Einstein G, Abernethy K: Statistical Package for the Social Sciences: SPSS 10.0. Applications guide: Chicago: SPSS, 1990. 10 Natelson BH: Neurocardiology. An interdisciplinary area for the 80s. Arch Neurol 42: 2, 1985. 11 Samuels MA: The brain-heart connection. Circulation 116 (1): 77-84, 2007. 12 Pandolfo M: Friedreich ataxia. Arch Neurol 65 (10): 1296-303, 2008. 13 Al-Mahdawi S, Pinto RM, Ismail O et Al: The Friedreich ataxia GAA repeat expansion mutation induces comparable epigenetic changes in human and transgenic mouse brain and heart tissues. Hum Mol Genet 1; 17 (5): 735-46, Epub Nov 27, 2007. 14 Lertrit P, Noer AS, Byrne E et Al: Tissue segregation of a heteroplasmic mtDNA mutation in MERRF (myoclonic epilepsy with ragged red fibers) encephalomyopathy. Hum Genet 90 (3): 251-4, 1992. 15 van der Knaap MS, Valked J: Kearns–Sayre Syndrome. In: van der Knaap MS, Valked J, editors. Magnetic Resonance of Myelination and Myelin Disorders. 3rd ed. Berlin Heidelberg: Springer-Verlag2005: 215-20. 16 Ponzetto C, Bresolin N, Bordoni A et Al: Kearns-Sayre syndrome: different amounts of deleted mitochondrial DNA are present in several autoptic tissues. J Neurol Sci 96 (2-3): 207-10, 1990. 17 Bresolin N, Moggio M, Bet L: Progressive cytochrome c oxidase deficiency in a case of Kearns-Sayre syndrome: morphological, immunological, and biochemical studies in muscle biopsies and autopsy tissues. Ann Neurol 21 (6): 564-72, 1987.
18 Duncan AJ, Bitner-Glindzicz M, Meunier B et Al: A nonsense mutation in COQ9 causes autosomal-recessive neonatal-onset primary coenzyme Q10 deficiency: a potentially treatable form of mitochondrial disease. Am J Hum Genet 84 (5): 558-66, 2009. 19 Novak V, Novak P, deMarchie M: The effect of severe brainstem injury on heart rate and blood pressure oscillations. Clin Auton Res 5 (1): 24-30, 1995. 20 McCabe PM, McEchron MD, Green EJ et Al: Destruction of neurons in the VPM thalamus prevents rabbit heart rate conditioning. Physiol Behav 57 (1): 159-63, 1995.
Miguel Cordeiro, MD S. Neurorradiologia Serviços de Radiologia Clinica Universitaria de Imagiologia Hospitais da Universidade de Coimbra Praceta Prof. Mota Pinto 3000-075 Coimbra Codex, Portugal Tel.: +351 914655121 E-mail: [email protected]
563
www.centauro.it
The Brain-Heart Connection in Mitochondrial Respiratory Chain Diseases M. CORDEIRO*, F. SCAGLIA**, S. LOPES DA SILVA*, P. GARCIA***, M. GRAZINA****, C. MOURA*, L. DIOGOL** * Neuroradiology, Medical Imaging Department, Faculty of Medicine, University Hospital Coimbra; Portugal ** Department of Molecular and Human Genetics, Baylor College of Medicine and Texas Children’s Hospital; Houston, Texas, USA *** Metabolic Diseases Unit, Center for Child Development “Luís Borges”, Pediatric Hospital of Coimbra - CHC EPE; Coimbra, Portugal **** Center for Neuroscience and Cell Biology, University of Coimbra; Coimbra, Portugal
Key words: mitochondrial respiratory chain diseases, metabolic diseases, magnetic resonance, heart, cerebellum, brainstem, thalamus
SUMMARY – Mitochondrial respiratory chain disorders (MRCD) are a heterogeneous group of diseases leading to an inadequate production of ATP. Brain and heart are among the most affected organs. Thus far, no specific relationship has been found between specific affected areas in the central nervous system and cardiac involvement. This study investigated the relationship between specific brain involvement and heart disease in mitochondrial disorders. We hypothesize that specific areas of brain lesions in children with MRCD are more frequently correlated to heart disease than others. A retrospective evaluation of the clinical records of 63 children with a definite MRCD, was performed searching for heart disease, namely, dilated and hypertrophic cardiomyopathy and arrhythmia. Brain MR images were evaluated and characterized regarding specific areas of atrophy and involvement. These findings were correlated using the Fischer exact test whose strength was determined with the Phi coefficient. During the period analyzed, 13 children (20.6%) developed cardiac disease, of whom nine (14.3%) exhibited isolated cardiomyopathy, one (1.6%) exhibited arrhythmia and three both. The main MRI abnormalities observed were brain atrophy (65.1%) and among this group 17.5% of subjects had cerebellar involvement. In addition, supratentorial, cerebellar and brainstem white and grey matter lesions were also found. There was a statistically significant relationship between progression to cardiac disease and cerebellar atrophy (Fisher’s Exact Test –0.005 and Phi 0.394) and lesions in the cerebral peduncles (0.035/0.358). Moreover, there was an additional correlation between thalamic lesions and progression to hypertrophic myocardiopathy (0.029/0.397). A statistical relationship between thalamic, mesencephalic and cerebellar involvement and cardiac disease in children with definite MRCD was observed. The true significance of this connection warrants further assessment.
Introduction Mitochondrial respiratory chain disorders (MRCD) are a genetically and clinically heterogeneous group of disorders caused by a primary derangement of oxidative phosphorylation 1 and impaired cellular ATP production. They can present at any age and may have any mode of inheritance. MRCD have a minimal established incidence of 1 in 10,000 births and are likely to be more frequent 2. The defini558
tion of mitochondrial disease is not consensual since some experts in the field consider this diagnosis in the case of any disorder resulting from any abnormality occurring inside the mitochondria (including abnormal cardiolipin deficiency as observed in Barth syndrome or abnormal iron deposition as observed in Friedreich ataxia), whereas others only consider this diagnosis in the context of diseases resulting from a malfunctioning respiratory chain. Mitochondria are more critically needed in those or-
M. Cordeiro
Figure 1 Axial fast spin-echo T2-weighted image (FSE. TR/TE 3500/96 ms) of a brain MRI of a three-year-old girl with a definite MRCD (MRC complex IV deficiency in muscle and liver) showing cerebellar atrophy. She developed cardiomyopathy as her disease progressed.
The Brain-Heart Connection in Mitochondrial Respiratory Chain Diseases
Figure 2 Axial fast spin-echo T2-weighted image (FSE. TR/TE 3500/96 ms) of a brain MRI of a 2-year-old boy with a definite MRCD showing cerebral peduncle lesions. He developed cardiac arrhythmia during disease progression.
gans with a higher metabolic rate such as the brain, heart and skeletal muscle and to a lesser extent kidney, liver or bowel which explains why these conditions exhibit a wide range of clinical symptoms. Organs with higher metabolic demands (e.g. the brain) are more likely to show clinical manifestations of mitochondrial dysfunction. Thus, neurological signs and symptoms are the most common presenting features in all age groups. The heterogeneity of the clinical presentation makes these diseases difficult to diagnose. Diagnostic criteria for pediatric MRCD have been therefore adapted from previously published adult criteria 3. These parameters involve clinical data, histological analysis (namely the presence of red ragged fibers or subsarcolemmal accumulation of mitochondria), the activity of the mitochondrial respiratory chain (MRC), quantification of fibroblast ATP synthesis rates, identification of a nuclear or mtDNA mutation of undisputed pathogenicity and metabolic indicators of impaired MRC function. Brain imaging is a non invasive test that may enhance the probability of the diagnosis during the diagnostic work-up. Patients with mitochondrial disease often present a combinaFigure 3 Axial fast spin-echo T2-weighted image (FSE. TR/TE 4000/100 ms) of a brain MRI of a 5-year-old boy with a definite MRCD showing symmetrical thalamic lesions. During the course of disease he developed hypertrophic carmyodiopathy.
559
The Brain-Heart Connection in Mitochondrial Respiratory Chain Diseases
tion of abnormalities from different patterns on their MRI scans. Brain volume loss, leukodystrophy, delayed myelination, brain stem and deep gray matter lesions have been frequently reported 4,5,6. In a previous, still unpublished study 7, we tried to find a predictive value to brain MRI lesions in children with MRCD. We correlated brain MRI findings with clinical and laboratory data both at disease presentation and after the full manifestations of disease were present and correlated them using Fisher’s exact test. We found that some brain lesions were significantly associated with specific patterns of clinical progression. Most could easily be explained by our knowledge of physiology and neurology. We found, for instance, a strong and significant correlation between brainstem lesions and later development of respiratory failure. However, one association was not easily explainable: children with a definite MRCD and posterior fossa involvement on brain MRI frequently developed heart disease. This paper describes the study where such an association was found and discusses a possible mechanistic explanation comparing this finding with similar data found in the literature. Methods Two hundred children investigated for MRCD in our institution from 1997 to 2006 were classified according to Bernier criteria for MRCD in children 3. Diagnosis was based on clinical features and metabolic, enzymatic, histological and genetic analyses results. Skeletal muscle, liver and/or myocardium samples were investigated in all patients 8. A definite diagnosis was considered if two major or one major and two minor criteria were fulfilled. The presence of both clinical and laboratory criteria was a prerequisite. The remaining patients were excluded from this study. It was a retrospective study approved by the local ethics committee. MRI was done in 149 of the 179 pediatric patients having brain image analysis performed (ultrasound, computed tomography and/or MRI). Sixteen of those 149 patients had a final diagnosis other than a MRC disorder. Brain imaging findings were not used as a single diagnostic criterion, in agreement with the Bernier diagnostic criteria 3. From the 133 children with a primary disorder suggestive of MRCD who underwent brain MRI, we selected 63 with a definite diagnosis. Median age at clinical onset was six months (mean±sd: 9.2±7.21 months; 560
M. Cordeiro
range: one day of life - six years old); 33 were female with a sex ratio of M/F is 0.9. Sixty-three per cent of patients had a MRC defect (enzymatic activity below 30% of the normal mean, expressed as a ratio using citrate synthase) in at least one tissue sample (skeletal muscle, liver and/or myocardium: isolated deficiencies of complexes IV and I were the most frequent (42% and 29%, respectively). Four patients had nDNA abnormalities (three in SURF1 gene and one in desoxiguanosine kinase gene) and four presented mtDNA mutations (8993T>G; 3302A>G and 11778G>A single base substitutions and a 4404 bp deletion). MRI scans were performed on a whole body MR system (1.5-Tesla scanner (Signa Infinity; General Electric Medical Systems, Milwaukee, WI, USA) using a standard circularly polarized head coil. Every child was anesthetized for the examination. Five sequences were acquired for each patient: sagittal spin-echo T1-weighted images (TR, 500-600 ms; TE, 12 ms), axial and coronal fast spin-echo T2-weighted images (TR, 3500-5000 ms; TE, 90-120 ms) and fast spin-echo fluid-attenuated inversion recovery (FLAIR) images in either the axial or sagittal plane of 5 mm thickness with 6 mm spacing. The field of view was 22-24 cm. The children had a thorough clinical evaluation already published elsewhere 8. For this study we focused on the presence of cardiomyopathy (diagnosed clinically and with echocardiogram) or arrhythmias (diagnosed clinically and with electrophysiological studies) both at presentation and during the course of disease (a definite timing was not defined for this interval, but it was in every case superior to six months and when a full clinical picture was present). Some patients had more than one MRI, in which case the most recent scan was chosen for analysis. Median age of the MRI brain study selected for evaluation was 3.7 years [mean±sd: 3.9 years ± 3.8 (five days of life - 18 years old)]. Brain images were independently reviewed by two neuroradiologists and one neuroradiology resident. Brain atrophy was evaluated qualitatively upon agreement between the two neuroradiologists. It was further analyzed according to the region involved. Other causes of volume loss in the developing brain, such as brain injury sequelae, hydrocephalus, ischemia or encephalitis were excluded. White or gray matter lesions, defined as high signal intensity at long TR sequences, were also evaluated according to the affected brain area. The neuroimaging features of the 63 children
www.centauro.it
The Neuroradiology Journal 22: 558-563, 2009
were correlated with the type of cardiac involvement, using Fisher´s exact test to assess the significance and Phi test to assess the strength of the correlation. SPSS software package version 169 was used for the analysis. Test results were considered significant when p= G mutation in the mitochondrial tRNALys gene has been observed in cerebellum, cerebrum and heart, establishing a link between these organs 14. Kearns-Sayre syndrome is characterized by a triad of features including 1) onset in subjects younger than 20 years; 2) chronic, progressive, external ophthalmoplegia; and 3) pigmentary degeneration of the retina. In addition. KearnsSayre syndrome may include cardiac conduction defects, cerebellar ataxia and raised cerebrospinal fluid (CSF) protein levels (>100 mg/dL) 15. Deleted mtDNA can be found in skeletal and cardiac muscle, brain cortex, cerebellum, liver and kidney in postmortem studies of patients with Kearn-Sayre Syndrome 16. Histologicaly, spongiosis in the frontal cortex, diffuse loss of Purkinje cells in the cerebellum, and heart fibrosis with mitochondrial abnormalities have been shown 17. Coenzyme Q(10) is a mobile lipophilic electron carrier located in the inner mitochondrial membrane. Defects of coenzyme Q(10) biosynthesis represent one of the few treatable mitochondrial diseases. These disorders can present with hypertrophic cardiomyopathy and cerebellar atrophy with ataxia in the context of CoQ9 mutations 18 providing a defined molecular link for mitochondrial dysfunction affecting the heart and cerebellum mithocondria. The role of the brainstem regulating heart rate is well known 19. Thus, lesions in the brainstem of MRCD patients would expectedly lead to cardiac rhythm disturbances. Such an association was not observed, but this potential as562
M. Cordeiro
sociation would deserve a more careful evaluation as part of a prospective study. Nonetheless, a more generic relationship was found between cerebral peduncular lesions and progression to cardiac disease in general (arrhythmia and cardiomyopathy). A relationship was found between thalamic lesions and progression to hypertrophic cardiomyopathy. A potential relationship between thalamic involvement and cardiac disease has been documented in the past. The thalamus has been involved in cardiac rhythm modulation 20. Some potential weaknesses of our study may stem from the following facts, i.e. It is a retrospective study that resulted from a new analysis of data previously collected for a study designed to show the specificity of brain imaging features in MRCD 6 and the predictive value of these lesions 7. We could have analyzed more general areas of brain damage (for instance. the brainstem as a whole) instead of specific regions. Ideally, a more comprehensive genetic and biochemical analysis should have been performed in cardiac tissue or other pertinent tissues of children with cardiomyopathy, and the observed specific defects should have been analyzed in conjunction with the brain MRI findings to establish a potential association with an underlying biochemical and molecular mitochondrial signature. The major strengths are the relatively large population of children analyzed (63 children with a definite diagnosis) and the rigorous statistical analysis. The observed associations are by no means conclusive. One way to strengthen a possible connection between brain and heart would be to look from a longitudinal prospective at a cohort of patients with a definite MRCD. A relationship between brain MRI findings, cardiac involvement, and specific MRC defects and molecular (mitochondrial or nuclear) defects observed from endomyocardial or other pertinent tissues should be performed to further elucidate the significance of these observations. Conclusion Neurocardiology is a recent and growing medical specialty. Mitochondria are essential for cardiac muscle and brain functioning; a potential relationship between both of them is understandable. The data suggest a greater prevalence of heart disease in children with MRCD and cerebellar atrophy, cerebral pedun-
www.centauro.it
cular lesions and thalamic lesions. Further investigations are needed to determine the clinical impact of this observation and to determine
The Neuroradiology Journal 22: 558-563, 2009
common biochemical and molecular mitochondrial abnormalities that could underlie this association.
References 1 Naviaux RK: Developing a systematic approach to the diagnosis and classification of mitochondrial disease. Mitochondrion 4 (5-6): 351-61, 2004. 2 Scaglia F: Genetics of White Matter Disorder – Cardiomyopathy and Mitochondrial Cytopathy, 2008 ASNR Proceedings, ASNR meeting, June 2-5, New Orleans, USA, 2008. 3 Bernier FP, Boneh A, Dennett X et Al: Diagnostic criteria for respiratory chain disorders in adults and children. Neurology 59: 1406-1411, 2002. 4 Kang PB, Hunter JV, Melvin JJ, Selak MA, Faerber EN, Kaye EM: Infantile leukoencephalopathy owing to mitochondrial enzyme dysfunction. J Child Neurol 17: 421-8, 2002. 5 Scaglia F, Wong LJ, Vladutiu GD et Al: Predominant cerebellar volume loss as a neuroradiologic feature of pediatric respiratory chain defects. Am J Neuroradiol 26: 1675-80, 2005. 6 Cordeiro M, Diogo L, Garcia P et Al: Brain MR Imaging in the Diagnosis of Mitochondrial Respiratory Chain Disorders in Children; New Orleans ASNR 2008 Proceedings 2008: 7. 7 Cordeiro M, Diogo L: Predictive Value of Brain MR Findings in Children with Mitochondrial Respiratory Chain Disorders; New Orleans ASNR 2008 Proceedings 2008: 7. 8 Diogo L, Grazina M, Garcia P et Al: Pediatric mitochondrial respiratory chain disorders in the Centro region of Portugal. Pediatr Neurol 40: 351-6, 2009. 9 Einstein G, Abernethy K: Statistical Package for the Social Sciences: SPSS 10.0. Applications guide: Chicago: SPSS, 1990. 10 Natelson BH: Neurocardiology. An interdisciplinary area for the 80s. Arch Neurol 42: 2, 1985. 11 Samuels MA: The brain-heart connection. Circulation 116 (1): 77-84, 2007. 12 Pandolfo M: Friedreich ataxia. Arch Neurol 65 (10): 1296-303, 2008. 13 Al-Mahdawi S, Pinto RM, Ismail O et Al: The Friedreich ataxia GAA repeat expansion mutation induces comparable epigenetic changes in human and transgenic mouse brain and heart tissues. Hum Mol Genet 1; 17 (5): 735-46, Epub Nov 27, 2007. 14 Lertrit P, Noer AS, Byrne E et Al: Tissue segregation of a heteroplasmic mtDNA mutation in MERRF (myoclonic epilepsy with ragged red fibers) encephalomyopathy. Hum Genet 90 (3): 251-4, 1992. 15 van der Knaap MS, Valked J: Kearns–Sayre Syndrome. In: van der Knaap MS, Valked J, editors. Magnetic Resonance of Myelination and Myelin Disorders. 3rd ed. Berlin Heidelberg: Springer-Verlag2005: 215-20. 16 Ponzetto C, Bresolin N, Bordoni A et Al: Kearns-Sayre syndrome: different amounts of deleted mitochondrial DNA are present in several autoptic tissues. J Neurol Sci 96 (2-3): 207-10, 1990. 17 Bresolin N, Moggio M, Bet L: Progressive cytochrome c oxidase deficiency in a case of Kearns-Sayre syndrome: morphological, immunological, and biochemical studies in muscle biopsies and autopsy tissues. Ann Neurol 21 (6): 564-72, 1987.
18 Duncan AJ, Bitner-Glindzicz M, Meunier B et Al: A nonsense mutation in COQ9 causes autosomal-recessive neonatal-onset primary coenzyme Q10 deficiency: a potentially treatable form of mitochondrial disease. Am J Hum Genet 84 (5): 558-66, 2009. 19 Novak V, Novak P, deMarchie M: The effect of severe brainstem injury on heart rate and blood pressure oscillations. Clin Auton Res 5 (1): 24-30, 1995. 20 McCabe PM, McEchron MD, Green EJ et Al: Destruction of neurons in the VPM thalamus prevents rabbit heart rate conditioning. Physiol Behav 57 (1): 159-63, 1995.
Miguel Cordeiro, MD S. Neurorradiologia Serviços de Radiologia Clinica Universitaria de Imagiologia Hospitais da Universidade de Coimbra Praceta Prof. Mota Pinto 3000-075 Coimbra Codex, Portugal Tel.: +351 914655121 E-mail: [email protected]
563
