Cerebral blood flow decreases with time whereas cerebral oxygen consumption remains stable during hypothermic cardiopulmonary bypass in humans

Recent investigations demonstrate that cerebral blood flow (CBF) progressively declines during hypothermic, nonpulsatile cardiopulmonary bypass (CPB). If CBF declines because of brain cooling, The cerebral metabolic rate for oxygen (CMRO₂) should decline in parallel with the reduction in CBF. Therefore we studied the response of CBF, the cerebral arteriovenous oxygen content difference (A - VD(cere)O₂), and CMRO₂ as a function of the duration of CPB in humans. To do this, we compared the cerebrovascular response to changes in the PaCO₂. Because sequential CBF measurements using xenon 133 (¹³³Xe) clearance must be separated by 15-25 min, we hypothesized that a time-dependent decline in CBF would accentuate the CBF reduction caused by a decrease in PaCO₂, but would blunt the CBF increase associated with a rise in PaCO₂. We measured CBF in 25 patients and calculated the cerebral arteriovenous oxygen content difference using radial arterial and jugular venous bulb blood samples. Patients were randomly assigned to management within either a lower (32-48 mm Hg) or higher (50-71 mm Hg) range of PaCO₂ uncorrected for temperature. Each patient underwent two randomly ordered sets of measurements, one at a lower PaCO₂ and the other at a higher PaCO₂ within the respective ranges. Cerebrovascular responsiveness to changes in PaCO₂ was calculated as specific reactivity (SR), the change in CBF divided by the change in PaCO₂, expressed in mL·100 g^-1·min^-1·mm Hg^-1. In the entire group of 25 subjects, SR was 0.69 ± 0.33 mL·100 g^-1·min^-1·mm Hg^-1 (SD) if PaCO₂ was reduced and 0.10 ± 0.30 mL·100 g^-1·min^-1·mm Hg^-1 if PaCO₂ was increased (P < 0.001). In patients managed within the lower range of PaCO₂, SR was 0.63 ± 0.31 and 0.21 ± 0.17, respectively, when PaCO₂ was reduced or increased (P < 0.05). In patients managed within the higher range of PaCO₂, SR was 0.76 ± 0.38 and -0.01 ± 0.36, respectively, when PaCO₂ was reduced or increased (P < 0.01). Estimated CMRO₂ remained constant within groups from the initial to the repeat measurements. These results confirm a significant time-dependent decline of CBF during CPB. Moreover, by demonstrating that CMRO₂ did not change significantly as PaCO₂ was altered in either direction, they suggest that the CBF reduction cannot be attributedd to progressive brain cooling during stable, hypothermic, nonpulsatile CPB, but must result from an alteration in cerebrovascular resistance.