Abstract

Background: We sought to detect microstructural brain injury after coronary artery bypass grafting (CABG) using cardiopulmonary bypass (CPB) and to define its anatomical distribution. Analyses were performed to determine possible correlation with postoperative neurocognitive deficit.
Methods: Seven patients undergoing CPB-CABG were assigned for serial cerebral designated diffusion tensor (DTI)-magnetic resonance imaging (MRI) examinations, preoperatively, on postoperative day (POD) 1 and 5. Levels of mean diffusivity (MD) and fractional anisotropy (FA) were analysed to assess gray and white matter injury. BBBdisruption and microembolic load were concomitantly evaluated by dynamic contrast enhancement (DCE)-MRI and diffusion-weighted imaging (DWI)-MRI methods, respectively. Neuropsychological tests were performed one day preoperatively and on POD 5.
Results: Increase in MD and decrease in FA were evident on POD 1 in gray matter structures including the hippocampus (MD, 0.000768 mm²/s ± 0.0000347 versus 0.000804 mm²/s ± 0.0000267, p<0.01; FA, 0.180 ± 0.03 mm²/s versus 0.163 mm²/s ± 0.032, p<0.01), middle frontal gyrus (MD, 0.000989 mm²/s ± 0.0000487 versus 0.001068 mm²/s ± 0.0000553, p<0.01), orbital gyrus (MD, 0.000867 mm²/s ± 0.0000374 versus 0.000908 mm²/s ± 0.0000356, p<0.01) and in the following white matter structures: Superior longitudinal fasciculus (MD, 0.000836 mm²/s ± 0.0000384 versus 0.000874 mm²/s ± 0.0000359, p<0.01; FA, 0.324 mm²/s ± 0.018 versus 0.305 mm²/s ± 0.022, p<0.01), frontal white matter (FA, 0.277 mm²/s ± 0.031 versus 0.26 mm²/s ± 0.033, p<0.01) internal capsule (MD, 0.000772 mm²/s ± 0.0000648 versus 0.000803 mm²/s ± 0.0000601, p<0.01; FA,0.364 mm²/s ± 0.032 versus 0.343 mm²/s ± 0.029, p<0.01) and corticospinal tract (MD, 0.000759 mm²/s ± 0.0000557 versus 0.00078 mm²/s ± 0.0000591, p<0.01; FA, 0.440 mm²/s ± 0.023 versus 0.411 mm²/s ± 0.023, p<0.01). Corticospinal tract fibre tracing showed significant decrease in postoperative fibre number. DTI findings persisted through POD 5 without spontaneous regression. BBB-disruption reflected by increased K^trans was detected in five patients (71%) on POD 1, mainly in frontal lobes and spontaneously restored by POD 5. Postoperative global cognitive score was reduced in all patients (98.2 ± 12 vs. 95.1 ± 11, p=0.032), predominantly in executive function and attention (91.8 ± 13 vs. 86.9±12, p=0.042). Postoperative FA decline significantly correlated with delta K^trans (R=0.6) and with postoperative decline in attention and Go-no-Go response inhibition tests (R=0.66 and 0.63, respectively).
Conclusion: CABG using CPB elicits gray and white matter microstructure injury in the above-mentioned anatomical distribution. These alterations are associated with transient BBB-disruption and may account for postoperative neurocognitive dysfunction.

Keywords: Magnetic resonance imaging; Diffusion tensor imaging; Cardiopulmonary bypass; Neuropsychological test; Cognition