PURPOSE: To evaluate MR imaging and lumbar cerebrospinal fluid enzymes as potential sensitive indicators of cerebral injury after open-heart valve replacement surgery. METHODS: Thirty-four patients with cardiac valvular disease were prospectively entered into this study and then underwent valve replacement or repair under cardiopulmonary bypass using a membrane oxygenator. In 26 patients, MR head images were obtained 12 to 24 hours before surgery; repeat MR images were obtained between 1 and 2 weeks after surgery. In 18 patients, lumbar puncture cerebrospinal fluid was analyzed 24 to 48 hours after surgery; the analyses included measurement of lactic dehydrogenase, creatine phosphokinase, adenylate kinase, and neuron-specific enolase. RESULTS: After surgery, MR imaging showed new ischemic lesions in 15 (58%) of 26 patients: 7 with deep white matter hyperintense lesions; 5 with brain stem, caudate, cerebellar, or thalamic/basal ganglia infarcts; 1 with intraparenchymal hemorrhage; 1 with a subdural hematoma and cortical infarct; and 1 with a corpus callosum lesion consistent with calcium or air. These new ischemic lesions seen on MR images were associated with a focal neurologic deficit in only 4 (27%) of the 15 patients. Neuron-specific enolase and lactic dehydrogenase were abnormally elevated after surgery in 5 (28%) of 18 patients. Adenylate kinase and creatine phosphokinase (brain isozymes) were elevated in one (67%) of the patients. Two (40%) of the five patients with abnormally high neuron-specific enolase or lactic dehydrogenase after surgery also showed a new focal neurologic deficit. CONCLUSIONS: MR imaging is a sensitive measure of subclinical cerebral ischemia after cardiac valve replacement under cardiopulmonary bypass. Cerebrospinal fluid neuron-specific enolase and lactic dehydrogenase are less sensitive than MR imaging for detecting subclinical cerebral ischemia, but these values were elevated after surgery more frequently than was adenylate kinase in our patients.
PURPOSE: To evaluate MR imaging and lumbar cerebrospinal fluid enzymes as potential sensitive indicators of cerebral injury after open-heart valve replacement surgery. METHODS: Thirty-four patients with cardiac valvular disease were prospectively entered into this study and then underwent valve replacement or repair under cardiopulmonary bypass using a membrane oxygenator. In 26 patients, MR head images were obtained 12 to 24 hours before surgery; repeat MR images were obtained between 1 and 2 weeks after surgery. In 18 patients, lumbar puncture cerebrospinal fluid was analyzed 24 to 48 hours after surgery; the analyses included measurement of lactic dehydrogenase, creatine phosphokinase, adenylate kinase, and neuron-specific enolase. RESULTS: After surgery, MR imaging showed new ischemic lesions in 15 (58%) of 26 patients: 7 with deep white matter hyperintense lesions; 5 with brain stem, caudate, cerebellar, or thalamic/basal ganglia infarcts; 1 with intraparenchymal hemorrhage; 1 with a subdural hematoma and cortical infarct; and 1 with a corpus callosum lesion consistent with calcium or air. These new ischemic lesions seen on MR images were associated with a focal neurologic deficit in only 4 (27%) of the 15 patients. Neuron-specific enolase and lactic dehydrogenase were abnormally elevated after surgery in 5 (28%) of 18 patients. Adenylate kinase and creatine phosphokinase (brain isozymes) were elevated in one (67%) of the patients. Two (40%) of the five patients with abnormally high neuron-specific enolase or lactic dehydrogenase after surgery also showed a new focal neurologic deficit. CONCLUSIONS: MR imaging is a sensitive measure of subclinical cerebral ischemia after cardiac valve replacement under cardiopulmonary bypass. Cerebrospinal fluid neuron-specific enolase and lactic dehydrogenase are less sensitive than MR imaging for detecting subclinical cerebral ischemia, but these values were elevated after surgery more frequently than was adenylate kinase in our patients.