Summary:Purpose: To determine whether changes in the high-energy phosphates occur with use of the ketogenic diet in patients with intractable epilepsy.Methods: 3 '~ magnetic resonance spectroscopic imaging studies were performed at 4.1 T in seven patients with intractable epilepsy (four Lennox-Gastaut syndrome, one absence, one primary generalized tonic-clonic, and one partial complex) before and after institution of the ketogenic diet. Coronal 'H anatomic imaging also was performed to provide correlation to the "P data.Results: Taking the patients as a group, the ratio of phosphocreatine (PC)ly-adenosine triphosphate (ATP) measured at baseline (regular diet) compared with that measured after the ketogenic diet showed a small but significant increase from 0.61 k 0.08 to 0.69 k 0.08 (p < 0.05). Comparing the ratio of PCr inorganic phosphorus (P,) measured at baseline with the postketogenic diet, there was a significant increase from 2.45 -C 0.27 to 2.99 5 0.44 (p < 0.05).
Conclusions:As a group, improvement of energy metabolism occurs with use of the ketogenic diet. This is in agreement with the chronic ketosis studies performed earlier in rodents.Key Words: Ketosis-Epilepsy-"P spectroscopy-Metabolism-Bioenergetics.The ketogenic diet has long been available as a treatment for intractable epilepsy, particularly in the pediatric population (1-3). How it achieves seizure control is unknown. Earlier studies performed in patients on the diet have failed to show any specific relation between success of diet with systemic changes induced by the diet, for example, pH, electrolytes, or cholesterol (4-6). In particular, the level of serum ketones does not specifically relate to degree of seizure control, although it has been shown that with resumption of a normal diet or by a glucose infusion, there is rapid loss of seizure control (6,7). Thus although ketosis is a requirement for seizure control, it is not the defining factor. Other studies focused on the cerebral metabolic changes seen in animal models of ketosis. Biochemical-extract studies performed on the whole brain of rats fed a high-fat diet suggested that there are increases in brain energy reserve [i.e., a sum of brain adenosine triphosphate (ATP), phosphocreatine (Per), glucose, and glycogen (8,9>]. No changes in glutamate, y-aminobutyric acid (GABA) and other amino acids were seen. However, significant difAccepted November 6, 1998. ferences potentially exist between human and rat brain with respect to adaptation to starvation and ketosis. For example, analyses of cerebral substrate utilization in 2-day starved ketotic conscious or lightly anesthetized rats suggested that the glucose contribution to total energy expenditure decreases only slightly to none, with ketone bodies contributing 3%, depending on brain region (1 0). This is in contrast to human studies as reported by Hasselbalch et al. (11). By using ["F]2-fluor0-2-deoxyglucose positron emission tomography (PET) and arteriovenous (AV) difference measurements, they reported that after 3 days of starvation...