Changes in cerebral oxygen saturation during transcatheter aortic valve replacement

Cerebral oxygen saturation (rSO(2)) is a non-invasive monitor used to monitor cerebral oxygen balance and perfusion. Decreases in rSO(2) > 20 % from baseline have been associated with cerebral ischemia and increased perioperative morbidity. During transcatheter aortic valve replacement (TAVR), hemodynamic manipulation with ventricular pacing up to 180 beats per minute is necessary for valve deployment. The magnitude and duration of rSO(2) change during this manipulation is unclear. In this small case series, changes in rSO(2) in patients undergoing TAVR are investigated. Ten ASA IV patients undergoing TAVR with general anesthesia at a university hospital were prospectively observed. Cerebral oximetry values were analyzed at four points: pre-procedure (baseline), after tracheal intubation, during valve deployment, and at procedure end. Baseline rSO(2) values were 54.5 +/- 6.9 %. After induction of general anesthesia, rSO(2) increased to a mean of 66.0 +/- 6.7 %. During valve deployment, the mean rSO(2) decreased < 20 % below baseline to 48.5 +/- 13.4 %. In two patients, rSO(2) decreased > 20 % of baseline. Cerebral oxygenation returned to post-induction values in all patients 13 +/- 10 min after valve deployment. At procedure end, the mean rSO(2) was 67.6 +/- 8.1 %. As expected, rapid ventricular pacing resulting in the desired decrease in cardiac output during valve deployment was associated with a significant decrease in rSO(2) compared to post-induction values. However, despite increased post-induction values in all patients, whether related to increased inspired oxygen fraction or reduced cerebral oxygen consumption under anesthesia, two patients experienced a significant decrease in rSO(2) compared to baseline. Recovery to baseline was not immediate, and took up to 20 min in three patients. Furthermore, baseline rSO(2) in this population was at the lower limit of the published normal range. Significant cerebral desaturation during valve deployment may potentially be limited by maximizing rSO(2) after anesthetic induction. Future studies should attempt to correlate recovery in rSO(2) with recovery of hemodynamics and cardiac function, provide detailed neurological assessments pre and post procedure, determine the most effective method of maximizing rSO(2) prior to hemodynamic manipulation, and provide the most rapid method of recovery of rSO(2) following valve deployment.