Magnetic resonance spectroscopy for assessing myocardial rejection in the transplanted rat heart

Traditionally, detection of heart rejection after heart transplantation is based on histologic grading of endomyocardial biopsy specimens. The value of magnetic resonance spectroscopy for determining heart rejection was assessed in rejecting and nonrejecting isografts and allografts using energy-rich phosphate spectroscopy. In 46 rats a heterotopic abdominal heart transplantation was performed, and animals were divided into the following groups: six isografts (no rejection), five untreated allografts (severe rejection), and 35 immunosuppressed allografts (mild to moderate rejection). One week after transplantation magnetic resonance spectroscopy was performed, and data were correlated to histologic findings (rejection grades according to Stanford and the New International Working Formulation classifications and relative volume of viable myocardium). Magnetic resonance spectroscopy allows detection of moderate to severe rejection with significant alterations in the energy-rich phosphates such as a decrease in the ratio of phosphocreatine/inorganic phosphate, phosphomonoester/inorganic phosphate, and beta-adenosine triphosphate/inorganic phosphate. A significant correlation was found between spectroscopic changes (phosphocreatine/inorganic phosphate) and histologic rejection (correlation coefficient r = 0.47, p < 0.005) and/or the amount of relative volume of viable myocardium and phosphocreatine/inorganic phosphate (r = 0.58) or beta-adenosine triphosphate/inorganic phosphate (r = 0.63), respectively. In conclusion magnetic resonance spectroscopy permits detection of moderate to severe degrees of heart rejection with a sensitivity of 85% and a specificity of 61%. Changes in the energy-rich phosphates correlate with the histologic grading of heart rejection and the relative volume of viable myocardium. Magnetic resonance spectroscopy appeared to be a valid technique for detecting myocardial rejection after heart transplantation in the reported experimental model.