EXPERIMENTAL-STUDY OF CEREBRAL AUTOREGULATION DURING CARDIOPULMONARY BYPASS WITH OR WITHOUT PULSATILE PERFUSION

Twenty-four adult mongrel dogs were divided into four equal groups acc ording to the following method of cardiopulmonary bypass: normothermic continuous (so-called nonpulsatile)' perfusion, normothermic pulsatil e perfusion, hypothermic continuous perfusion, and hypothermic pulsati le perfusion. Cerebral blood flow was determined by measuring the volu me of sagittal sinus venous blood outflow with a transit-time ultrason ic flowmeter. Cardiopulmonary bypass was initiated at a Bow rate of 80 ml/kg per minute. Cerebral temperature was maintained at 37 degrees C in the normothermic groups and at 25 degrees C in the hypothermic gro ups. Arterial pH and carbon dioxide, were maintained within the physio logic range by alpha-stat acid-base regulation. Mean cerebral perfusio n pressure and blood flow were not affected during 90 minutes of the b ypass. The respective values were 67.1 mm Hg and 37.1 ml/100 gm brain per minute with normothermic continuous perfusion, 72.8 mm Hg and 39.0 ml/100 gm per minute with nonpulsatile perfusion, 98.0 mm Hg and 23.0 ml/gm per minute with hypothermic continuous perfusion, and 86.8 mm H g and 22.3 ml/100 gm per minute with hypothermic pulsatile perfusion. Pump flow rates were altered from 10 to 120 ml/kg per minute in a step wise fashion to obtain graded changes of perfusion pressure. Cerebral blood flow, however, was not changed significantly by cerebral perfusi on pressure so long as perfusion pressure was greater than 50 mm Hg. C onversely, cerebral blood flow changed proportionally with cerebral pe rfusion pressure at a pressure less than 50 mm Hg. The correlation bet ween cerebral blood flow and perfusion pressure was described as two s eparate lines determined by linear regression. The slope of the regres sion line relating cerebral blood flow to perfusion pressure was 0.16 +/- 0.08 for a cerebral perfusion pressure above 50 mm Hg and 0.68 +/- 0.11 below 50 mm Hg in the normothermic continuous perfusion group; 0 .14 +/- 0.09 and 0.32 +/- 0.09 with normothermic pulsatile perfusion; 0.10 +/- 0.04 and 0.62 +/- 0.18 with hypothermic continuous perfusion; 0.09 +/- 0.08 and 0.39 +/- 0.04 in the hypothermic pulsatile perfusio n group. The slope above 50 mm Hg was significantly smaller and closer to zero in all groups than it was at a perfusion pressure below 50 mm Hg (p < 0.05). The slope and cerebral blood Bow for a perfusion press ure above 50 mm Hg between pulsatile and continuous perfusion groups w as not significantly different, whereas the value of the slope and cer ebral blood Bow in pulsatile groups when cerebral perfusion pressure w as less than 50 mm Hg was significantly (p < 0.05) smaller and higher, respectively. These data suggest that cerebral autoregulation is inta ct at a cerebral perfusion pressure greater than 50 mm Hg during eithe r normothermic or hypothermic cardiopulmonary bypass. In addition, com pared to continuous (nonpulsatile) perfusion, pulsatile bypass generat ed a higher cerebral blood flow at a cerebral perfusion pressure less than 50 mm Hg.