Progression of Cardiac Dysfunction in a Case of Mitochondrial Diabetes: A case report

Although there are some reports that have evaluated the correlation between mitochondrial tRNA(Leu) (UUR) mutation at position 3243 and cardiomyopathy, the degree of cardiac involvement in mitochondrial diabetes is still largely unknown. Here we report the case of a patient with mitochondrial diabetes who developed cardiac dysfunction, evaluated by echocardiography, along with worsening of glycemic control over a relatively short period, within a year.

In April 1997, a 22-year-old Japanese man presented to a medical practice because of thirst, polyuria, and body weight loss of ∼8 kg over a few months. Because blood testing revealed a high plasma glucose level (23.6 mmol/l), and because urine was positive for ketone bodies, he was immediately admitted to the hospital. After initial therapy with insulin for 5 months, he was mainly treated with oral hypoglycemic agents. In July 1998, his glycemic control worsened and he presented to our hospital; insulin therapy was restarted. Because his maternal grandmother and paternal uncle had diabetes and he complained of hearing impairment, he was thought to have “mitochondrial diabetes.” After obtaining informed consent from the patient, mutation of the mitochondrial gene was analyzed. In leukocytes, A→G transition at the nucleotide pair 3243 of the mitochondrial gene was detected. Although he had no cardiac symptoms, to rule out cardiac involvement, further investigations (i.e., electrocardiography, chest X-ray, and echocardiography) were performed. No abnormal findings were found in all of these tests. Subsequently, his glycemic control worsened again (HbA_1c 9.5%) because of poor compliance with insulin injections. Therefore, he was admitted to our hospital in May 1999. On admission, his BMI was 18 kg/m², his blood pressure was 103/64 mmHg, and his pulse rate was 70/min. Physical examination revealed no sign of neuropathy, skeletal myopathy, or heart failure. The postprandial plasma glucose level was 8.3 mmol/l, HbA_1c 8.6% (normal range 4.2–5.5%), urine ketone bodies negative, serum lactate level normal, serum pyruvate level slightly elevated, GAD65 antibody negative (cutoff <1.3 U/ml; 100% sensitivity and 100% specificity of the assay in GAD antibody proficiency test [Immunology of Diabetes Workshop]; lab ID no. 305), urine C-peptide level 25.8 μg/day (normal range 18.3–124.4 μg/day), fasting C-peptide level 0.13 nmol/l (0.17–0.76 nmol/l), and glucagon-stimulated serum C-peptide level 0.56 nmol/l (6 min after intravenous injection of 1 mg glucagon). He did not consume alcohol, and his thyroid function was normal. Although he had no cardiac symptoms and his electrocardiography and chest X-ray were normal, we performed follow-up echocardiography to evaluate “latent” mitochondrial cardiomyopathy. To our surprise, echocardiography revealed diffuse hypokinesis of the left ventricle, especially in the apex, which was not apparent in his former study performed the previous year. Left ventricular hypertrophy was not observed. Left ventricular ejection fraction calculated by automated quantification by modified Simpson method using an Agilent Sonos 5500 was decreased to 37% (normal >65%). Myocardial perfusion scintigraphy using 99mTc-tetrofosmin showed no perfusion defect. Radioisotope multigated ventriculography also showed diffuse hypokinesis of the left ventricle, with a low ejection fraction (41% [normal >65%]). Ambulatory Holter recording showed no abnormality. To rule out ischemic heart disease, coronary angiography was performed, and no significant stenosis was found. Transvenous endomyocardial biopsy was performed after obtaining informed consent from the patient. Histological examination of right ventricular myocardial specimens revealed mild hypertrophy and a disarray of myofibrils with marked vacuolar degeneration, which was considered to be the swelling of mitochondria. These findings are compatible with mitochondrial cardiomyopathy. Moreover, using a ventricular septal endomyocardial specimen, A→G transition at nucleotide pair 3243 of the mitochondrial gene was detected; heteroplasmic concentration was 82.4%, whereas that in leukocytes was 10.5%. Interestingly, the proportion of heteroplasmy was markedly higher in the heart, which may reflect the pathological heterogeneity of organ involvement.

In Japan, nearly 1% of patients with diabetes possess a mitochondrial tRNA(Leu)(UUR) mutation at position 3243 (1), which was originally found in MELAS (mitochondrial encephalomyopathy, lacticacidosis, and stroke-like episodes) (2). Some recent articles have reported that this mutation may be a cause of cardiomyopathy in diabetes (3,4,5,6). The degree of cardiac involvement in mitochondrial diabetes, however, is still largely unknown. In our case, even though the patient had no cardiac symptoms and electrocardiography and chest X-ray showed no obvious change within a year, echocardiography revealed diffuse hypokinesis of the left ventricle, which was not apparent a year previously. It has been reported that left ventricular hypertrophy is a clinical feature of patients with MELAS (7,8,9,10). Generally, in mitochondrial disease, both hypertrophic and dilated cardiomyopathy can be observed (5,6,7). Because this case cannot be diagnosed as MELAS because of the lack of abnormality in neurological findings, and given that the clinical features of cardiac involvement in mitochondrial disease vary among different subgroups (7), left ventricular systolic dysfunction without left ventricular hypertrophy, as seen in our patient’s case, is not a surprising feature of this disorder. Presently, this patient is clinically asymptomatic (NYHA functional class I). Left ventricular thickness might increase in the future as the disease progresses; therefore, strict follow-up with echocardiography is considered to be essential. Regarding the progression of cardiomyopathy in mitochondrial diabetes, Momiyama et al. (5) reported a patient with left ventricular hypertrophy who developed left ventricular systolic dysfunction and dilatation over 8 years of observation, as compared with progression of cardiac dysfunction within 1 year in our patient. Our search of the literature revealed no report of change of cardiac function using echocardiography over such a short period. We would like to emphasize that such a change in left ventriclular function could occur in a short period of time in mitochondrial diabetes. Yoshida et al. (3) reported that patients with mitochondrial diabetes should be examined by echocardiography periodically, even if their electrocardiogram and chest X-ray are normal. Our report confirmed the importance of serial echocardiography to evaluate the cardiac involvement in mitochondrial diabetes.

Then, what is the cause of left ventricular deterioration over a relatively short period? In this case, coronary artery disease; hypertensive heart disease; alcoholic cardiomyopathy; congenital heart disorders, including glycogen storage disease, thyroid disorders, and pericardial disease; and myocarditis were excluded. We speculate that poor glycemic control may have been responsible for the progression of cardiac dysfunction in this case, because no factor other than worsening of glycemic control seemed to have changed. Therefore, we suggest that it may be worth trying to evaluate cardiac function by echocardiography once a year in patients with mitochondrial diabetes, especially in those with poor glycemic control. Although more cases should be accumulated to reach a conclusion, we propose that recognition of cases like this is clinically important for the proper management of mitochondrial diabetes.