Skeletal Muscle Mitochondrial Dysfunction in Alternating Hemiplegia of Childhood G. J. Kemp, DM,' D. J. Taylor, DPhil," P. R. J. Barnes, MRCP,* J. Wilson, FRCP,t and G. K. Radda, FRS" Alternating hemiplegia of childhood is an uncommon disease characterized by repeated, transient attacks zyxwvut of hemiplegia. Its pathophysiology is uncertain, but atten- tion recently has focused on possible mitochondrial ab- normalities. Using 3'P magnetic resonance spectroscopy, we studied gastrocnemius muscle in zyxwvuts 5 patients with al- ternating hemiplegia, aged 8 to 30 (mean, 18) years, at rest and during incremental aerobic exercise and recov- ery. There were no significant differences in resting muscle between patients and a control group aged 7 to zyxwvu 42 (mean, zyxwvutsrqp 19) years. Exercise performance was grossly impaired in the patients, the mean duration being 30% of normal. The total change in pH during exercise was somewhat less than in control subjects, while the changes in phosphocreatine concentration and intracel- Mar ADP were similar. Thus the average overall rate of fall of phosphocreatine concentration during exercise was three-fold greater than in control subjects. How- ever, the initial rate of ATP turnover at the start of exercise (a measure of muscle mass and efficiency) was not abnormal. During recovery, both the initial rate of phosphocreatine resynthe9is and the calculated mito- chondrial capacity were reduced by about 35%. This mitochondrial defect probably explains most of the ab- normalities seen during exercise. Kemp GJ, Taylor DT, Barnes PRJ, Wilson J, Radda GK. Skeletal muscle mitochondrial dysfunction in alternating hemiplegia of childhood. Ann Neurol 1995;38:681-684 Alternating hemiplegia of childhood (AH) is an un- common disease characterized by repeated, transient attacks of hemiplegia, affecting one side or the other or both zyxwvutsrqpo El-41. Recently, a study of 4 patients with AH using 31P magnetic resonance spectroscopy (31PMRS) showed abnormalities in the muscle at rest consistent with mitochondrial dysfunction [5]. Analysis of exer- cise and (especially) recovery from exercise is a sensi- From the "MRC Biochemical and Clinical Magnetic Resonance Unit, Oxford Radcliffe Hospital NHS Trust, Oxford and tGreat Ormond Street Hospital for Children NHS Trust and the Institute of Child Health, London, United kngdom. Received Mar 21, 1995, and in revised form Jun 23. Accepted for publication Jun 26, 1995. Address correspondence to Dr Kemp, MRC Biochemical and Clini- cal Magnetic Resonance Unit, Oxford Radcliffe Hospital Trust, Headington, Oxford OX3 9DU, United Kingdom. tive way to detect abnormalities of mitochondrial func- tion under conditions of high ATP turnover 16, z 71. Here we report a study of 5 patients with AH in which gastrocnemius muscle was studied at rest and during exercise and recovery. Subjects and Methods S2c bjects These were 5 females, aged 8 to 30 (mean, 18) years, diag- nosed as having AH. Results were compared with those of 10 female and 2 male control subjects, aged 7 to 42 (mean, 19) years. All patients were active and mobile, except that 2 usually used a wheelchair when out of doors "as a precau- tion." None had any fixed neurological deficit. Informed con- sent was obtained from each patient or their parent, and studies were carried out with approval of the local hospital ethical committee. "P Magnetic Resonance Spectroscopy and Exercise Protocol Subjects were placed in a 90-cm bore 2-T magnet (Oxford Instruments, Eynsham, Oxford, United Kindom) interfaced to a spectrometer (Bruker, Coventry, United Kmgdom) with the right calf overlying a 6-cm-diameter surface coil. Data were collected with an 80-psec pulse width and a 2-second interpulse delay. Two 64-scan spectra were acquired from the muscle at rest. Patients exercised by performing plantar flexion, lifting a weight of 10% lean body mass a distance of 7 cm at 30 liftdmin. After 5 minutes of exercise, the weight was increased by 29% lean body mass every 1.25 minutes until fatigue. 32-scan spectra were collected throughout exercise. Postexercise recovery was monitored by collecting four 8- scan, four 16-scan, three 32-scan, and two 64-scan spectra (1 3 minutes in total). Analysis zyxwv of Magnetic Resonance Spectroscopic Data Signals were processed by exponential multiplication and Fourier transformation, and signal intensities were obtained using a time domain fitting program (VARPRO, R. de Beer, Utrecht, Holland). Cytosolic concentrations of inorganic phosphate (Pi) and phosphocreatine (PCr) (mM, i.e., mmol/ liter of intracellular water) were calculated from the relative signal intensities of Pi, PCr, and ATP corrected for differen- tial magnetic saturation and assuming a normal intracellular ATP ([ATP)) of 8.2 mM: cytosolic pH was determined from the chemical shift of Pi from PCr; and free {ADP] (FM) was calculated from pH and [PCr) and the equilibrium constant of the creatine kinase equilibrium (6 x lo-'" M), assuming a normal [total creatine) of 42.5 mM 18). Phosphorylation potential is calculated in reciprocal form zy as [Pi]{ADP]/ {ATP]. During exercise [PCr] is more conveniently ex- pressed as phosphocreatine concentration (PCr/[PCr + Pi)), which corrects for any signal loss due to movement with respect to the coil. Abnormalities of exercise duration and of metabolic changes during the whole of exercise are taken into account by calculating average mean rates of fall of pH and PCr/(PCr + Pi). The initial rate of ATP turnover, which is inversely proportional to "effective" muscle mass, is calculated from the initial rates of change of pH and {PCr] (i.e., from rest to first exercise spectra) using the known stoichiometry of pro- ton consumption by net PCr hydrolysis and ATP production Copyright 0 1995 by the American Neurological Association 681