Differences in regional cerebral oximetry during cardiac surgery for patients with or without postoperative cerebral ischaemic lesions evaluated by magnetic resonance imaging

Background: Near infrared spectroscopy (NIRS) is widely used to monitor regional cerebral tissue oxygenation (rScO(2)). We compared rScO(2) values during cardiac surgery in patients with or without new cerebral ischaemic lesions on diffusion weighted magnetic resonance imaging (DWI). We hypothesised patients with new cerebral lesions would have impaired tissue oxygenation reflected in their rScO(2) values. Methods: NIRS and DWI data were collected in 152 elective cardiac surgery patients. Absolute rScO(2) values, duration of desaturation below thresholds (baseline, 10%, and 20%), and accumulated cerebral desaturation load were compared between patients with or without new cerebral lesions on DWI. Primary outcome was time below 10% from rScO(2) baseline. Results: The time below 10% from rScO(2) baseline was significantly longer for patients with new cerebral lesions than for patients without [median (inter-quartile range): 11.0 (0.4; 37.5) min vs 1.8 inter-quartile range: (0.05; 20.9) min, P=0.02]. Furthermore, they had a higher accumulated desaturation load below baseline (P=0.02) and 10% below baseline (P=0.02). Finally, their absolute minimum rScO(2) value was significantly lower (P=0.01). However, the frequency of patients with desaturation below 10% and 20% was comparable between patients with and without new cerebral lesions. Receiver-operating characteristic curve analysis did not identify a clear-cut critical threshold among the investigated rScO(2) variables. Conclusions: Use of NIRS identified significant group differences in rScO(2) values between patients with or without new ischaemic lesions. However, a critical threshold could not be identified because of a high variation in NIRS values across both groups.