Patients on long-term zidovudine (AZT) therapy experience muscle fatigue and weakness attributed to AZT-induced mitochondrial toxicity in skeletal muscle. To determine if the clinico-pathological abnormalities in these patients correspond to abnormal muscle energy metabolism, we used 31P in vivo magnetic resonance spectroscopy to follow phosphorylated metabolites during exercise. We studied 19 normal volunteers, 6 HIV-positive patients never treated with AZT, and 9 HIV-positive patients who had been treated with AZT for a mean period of 33 mo (range 12-48 mo) and had muscle biopsy-proven AZT-myopathy with abnormal mitochondria. Changes in phosphocreatine, ATP, and intracellular pH in the gastrocnemius muscle were followed during a graded steady state exercise protocol, and the recovery of phosphocreatine was followed on cessation of exercise. We found that graded steady state exercise produced a greater depletion of muscle phosphocreatine levels in the AZT-treated patients, compared to either HIV-positive patients who were not treated with AZT or normal controls. No differences in the effects of steady state exercise on muscle phosphocreatine levels were observed between the control group and the HIV-positive patients who had not been treated with AZT. The results suggest that the effect of AZT on muscle energy metabolism is significant, and similar to the effect observed in patients with known mitochondrial myopathies. Using a well-known model for control of mitochondrial metabolism, the observed differences in steady state phosphocreatine levels during exercise suggest that AZT treatment decreases the maximal work output and the maximal rate of muscle ATP synthesis.
T M Sinnwell, K Sivakumar, S Soueidan, C Jay, J A Frank, A C McLaughlin, M C Dalakas