Carbon 13–Labeled Magnetic Resonance Spectroscopy Observation of Cerebral Glucose Metabolism: Metabolism in MELAS: Case Report

Background  Carbon 13–labeled magnetic resonance spectroscopy (¹³C-MRS) with [^1-13C]-glucose administration, the ¹³C atom that behaves as a radio inactive tracer in the brain, can differentiate aerobic and anaerobic glucose metabolism by detecting [^4-13C]-glutamate (Glu C4) and [^3-13C]-lactate (Lac C3).

Objective  To investigate the cerebral metabolic derangement resulting from mitochondrial dysfunction in mitochondrial myopathy, encephalopathy, lactic acidosis, and strokelike episodes (MELAS).

Design  Application of a new ¹³C-MRS technique to a patient with MELAS compared with control subjects (n = 7).

Patient  A 19-year-old woman with an A3243G mitochondrial mutation who underwent ¹³C-MRS for 30 minutes after oral administration of [^1-13C]-glucose (0.75 g/kg).

Result  Decreased Glu C4–labeling (P<.001) and increased Lac C3 synthesis (>2 SDs) compared with controls were demonstrated in the patient with MELAS.

Conclusions  This first report on ¹³C-MRS observation of cerebral glucose metabolism in a patient with MELAS demonstrated the presence of low glutamate production via the tricarboxylic acid cycle compared with high lactate synthesis by glycolysis. The present findings suggest that the clinical use of ¹³C-MRS can be extended to diagnose mitochondrial dysfunction and monitor cerebral glucose metabolism in a variety of mitochondrial disorders.

Carbon 13 (¹³C) is a stable isotope with a very low natural abundance, at only 1.1%, but is detectable by magnetic resonancespectroscopy (MRS). Since MRS has the advantage not only of chemical specificity but also of specificity of particular atom positions within molecules, it permits the monitoring of the fate of ¹³C atoms appearing in metabolized products. When [^1-13C]-glucose is administered, for instance, the glucose is metabolized to [^3-13C]-pyruvate, which is metabolized aerobically through the tricarboxylic acid (TCA) cycle or/and anaerobically to [^3-13C]-lactate.¹ In nonactivated cortex at least 85% to 90% of the glucose is consumed by the TCA cycle, and the first nuclear magnetic resonance–visible product of this metabolism is [^4-13C]-glutamate since the concentrations of TCA cycle intermediates prior to glutamate are not nuclear magnetic resonance detectable in the brain. This neuronal TCA cycle and subsequent glutamate-glutamine cycle are recently looked on as a critical metabolic pathway for normal cerebral glutamatergic function.²

The syndrome of mitochondrial myopathy, encephalopathy, lactic acidosis, and strokelike episodes (MELAS) is known to be associated with point mutations of the mitochondrial DNA, in particular the A to G transition at nucleotide 3243 of the transfer RNA gene,³ which causes abnormalities in the respiratory electron transport cascade. Disruption of the respiratory chain and consequent depletion of NAD⁺ and NADP⁺ shift the predominant catabolic metabolism from the TCA cycle to anaerobic glycolysis.⁴ This in turn produces an accumulation of pyruvate and its reduced product lactate. Magnetic resonance spectroscopic measurement of brain phosphocreatine and lactate levels, therefore, has been shown to be helpful in the diagnosis and monitoring of mitochondrial diseases.^4-6 Little is known, however, regarding the metabolic changes occurring in the TCA cycle in MELAS seeing as no efficient diagnostic technique for in vivo monitoring of aerobic cerebral metabolism has been developed. In this study, we applied a ¹³C-MRS technique with oral administration of [^1-13C]-glucose^7-9 to investigate the metabolic derangement resulting from mitochondrial dysfunction in MELAS.

A 19-year-old woman diagnosed as having MELAS with an A3243G mitochondrial mutation was investigated by ¹³C-MRS under oral administration of [^1-13C]-glucose. Headaches and vomiting initially developed when the patient was 11 years old. An episode of cerebral infarction causing transient motor paralysis of the lower extremities followed 2 years later, when treatment with sodium dichloroacetate was started.

She was seen with mild cognitive decline and general muscle weakness at the time of the investigation, which had been mostly stable during the previous few years. She had no systemic conditions including diabetes mellitus. Her magnetic resonance imaging showed no detectable brain lesions in the area of investigation except for mild atrophy. Informed consent was obtained using forms and procedures approved by the Ethical Board Committee of the National Center of Neurology and Psychiatry, Tokyo, Japan.

The details of this procedure have been described elsewhere.^7-9 After [^1-13C]-glucose (ISOTEC Co; Sigma-Aldrich Co, St Louis, Mo; 0.75 g/kg of body weight) in a 30% wt/vol carbonated water solution was administered orally, the subject was placed in a 2.0-T whole-body magnetic resonance scanning tube (Toshiba, Tochigi, Japan) equipped with hydrogen 1–labeled-¹³C–heteronuclear single-quantum coherence correlation sequence wearing earplugs with eyes closed in a dark condition. Blood samples were obtained at predetermined intervals to observe the time course of the fractional enrichment (FE) of plasma [^1-13C]-glucose concentration through an antecubital intravenous catheter.

During the period from 30 to 60 minutes after glucose administration, ¹³C -MRS spectra were acquired simultaneously from 2 voxels of 64 mL (4×4×4 cm) both placed symmetrically on the patient’s bilateral parieto-occipital lobes using multislice scout magnetic resonance imagings. Single acquisition time was 5 minutes.

After correction of distortion due to eddy current magnetic field, 2-dimensional curve fitting using a nonlinear complex least squares method was applied for 2-dimensional spectra. The sum of ¹³C spectra acquired for 30 minutes was calculated, and the contents of [^4-13C]-glutamate (Glu C4) and [^3-13C]-lactate (Lac C3) were obtained using weighting functions decided by the water signal in the same voxel.¹⁰ Adjustments for the atrophy were made by image segmentation. The concentrations of Glu C4 and Lac C3 in the patient’s brain were compared with those in healthy controls aged from 22 to 28 years (n = 7) obtained by the same protocol.

The time course and level of FE of plasma [^1-13C]-glucose in the patient was mostly identical to that of the controls; that is, 50% to 60% of the FE level was obtained during ¹³C-MRS observation. Figure 1 shows ¹³C-MR spectra obtained from 30 to 60 minutes after oral administration of [^1-13C]-glucose. In the patient with MELAS, Glu C4 peaks were lower and Lac C3 peaks were higher compared with controls. The mean (SD) Glu C4 concentration was 1.356 (0.404)mM in the controls and 0.636 (0.007)mM in the patient with MELAS (P<.001) (Figure 2). However, the mean (SD) Lac C3 concentration was 0.1366 (0.0494)mM in the controls and 0.4996 (0.784)mM in the patient with MELAS, with the latter exceeding the former by more than 2 SDs (Figure 3).

To our knowledge, this is the first report describing ¹³C-MRS observation of cerebral glucose metabolism in a patient with MELAS and suggesting the presence of metabolic derangement characterized by low glutamate production and high lactate synthesis in the brains of patients with MELAS. In a normal human brain, Gruetter et al¹¹ observed that [^4-13C]-glutamate labeling rapidly reached close to the maximum value 60 minutes after intravenous continuous infusion of [^1-13C]-glucose. The concentration of [^4-13C]-glutamate quantified from the spectra obtained during 120 and 180 minutes was 2.4 mol/mL with a 63% mean plasma [^1-13C]-glucose FE concentration. This gave a calculated brain glutamate concentration of 9.1 mol/mL, almost identical to previous estimates of total brain glutamate concentration, suggesting that most brain glutamate is synthesized rapidly from plasma glucose.

Observation of the time course of [^4-13C]-glutamate spectra with its further metabolites, [^3-13C]-glutamate and [^4-13C]-glutamine, yields the rates of TCA cycle turnover and glutamate-glutamine cycling, thereby providing a powerful tool for the investigation of glutamate and glutamine metabolism.^1,2,7 Despite these advantages, however, a limitation of the previous ¹³C-MRS method was that it required continuous intravenous infusion of ¹³C-labeled glucose together with somatostatin and insulin to maintain a hyperglycemic glucose clamp for more than 2 hours,^7,11 which is impractical for clinical investigations.

The investigation protocol applied in this study has an advantage and a disadvantage.⁷ The advantage is simplicity; that is, the patient is examined in the MRS scanner for 30 minutes after the oral administration of a [^1-13C]-glucose solution of 0.75 g/kg body weight, which was less than 50 g for most subjects. The disadvantage is the uncertainty present in the level of FE of plasma [^1-13C]-glucose; that is, in the presence of diabetes mellitus, for instance, FE of plasma [^1-13C]-glucose does not increase sharply compared with healthy subjects because of a high nonlabeled plasma glucose level and limited insulin secretion, so that ¹³C incorporation into metabolites can be observed as being falsely decreased.

This was not the case, however, in our patient who showed the same time course and level of plasma [^1-13C]-glucose FE as controls before and during the MRS spectra measurement. To eliminate this uncertainty, in general, the ratio of Lac C3/Glu 4 will be useful for diagnosing the presence of an imbalance between anaerobic and aerobic metabolism, which may be associated with various mitochondrial disorders.

In this study, we observed decreased glutamate labeling down to 47% (P<.001) and increased lactate synthesis up to 366% (>2 SDs) in a patient with MELAS compared with controls. Since glutamate is a major transmitter in the brain, decreased glutamate production may relate to various neurological symptoms such as mental retardation associated with MELAS. If glutamate production is activated and accelerated, on the contrary, increased formation of toxic free radicals may result because of impaired electron transport chain function in mitochondria.

The clinical use of ¹³C-MRS can be extended to diagnose the presence of mitochondrial dysfunction and monitor the state of the cerebral metabolic derangements in a variety of mitochondrial disorders.

Correspondence: Taisuke Otsuki, MD, PhD, Department of Neurosurgery, National Center Hospital for Mental, Nervous, and Muscular Disorders, National Center of Neurology and Psychiatry, Ogawahigashi-4-1-1, Kodaira, Tokyo 187-8511 Japan (otsukit@ncnpmusashi.gr.jp).

Accepted for Publication: March 23, 2004.

Author Contributions:Study concept and design: Otsuki and Tsukada. Acquisition of data: Otsuki, Tsukada, Goto, Okamoto, and Watanabe. Analysis and interpretation of data: Otsuki, Kanamatsu, Tsukada, and Watanabe. Drafting of the manuscript: Otsuki, Goto, and Okamoto. Critical revision of the manuscript for important intellectual content: Kanamatsu, Tsukada, and Watanabe. Obtained funding: Otsuki. Administrative, technical, and material support: Otsuki, Kanamatsu, Tsukada, Goto, Okamoto, and Watanabe.

Funding/Support: This study was supported by Health Science research grants from the Ministry of Health, Labor, and Welfare of Japan, and grants from the Research on Brain Science Project and National Research and Development Program for Medical and Welfare Apparatus under entrustment by NEDO.