A multicenter comparison of mitochondrial respiratory chain and complex V enzyme activity tests was performed. The average reproducibility of the enzyme assays is 16% in human muscle samples. In a blinded diagnostic accuracy test in patient fibroblasts and SURF1 knock-out mouse muscle, each lab made the correct diagnosis except for two complex I results. We recommend that enzyme activities be evaluated based on ratios, e.g. with complex IV or citrate synthase activity. In spite of large variations in observed enzyme activities, we show that inter-laboratory comparison of patient sample test results is possible by using normalization against a control sample.
2012 Elsevier B.V. All rights reserved