Pulsatile compared with nonpulsatile perfusion using a centrifugal pump for cardiopulmonary bypass during coronary artery bypass grafting. Effects on systemic haemodynamics, oxygenation, and inflammatory response parameters.

Abstract

The present study investigated the influence of pulsatile or nonpulsatile flow delivery with a centrifugal pump for cardiopulmonary bypass (CPB) during coronary artery bypass grafting (CABG) in two randomized groups of 19 patients each. All patients received a standard anaesthetic and surgical protocol. Pulsatile perfusion during CPB was created by accelerating the baseline pump speed of the Sarns centrifugal pump at a rate of 50 cycles per minute. Measurements included perioperative systemic haemodynamics and oxygen exchange, total haemolytic complement (CH50), polymorphonuclear (neutrophil) granulocyte (PMN) count and plasma granulocyte elastase bound to alpha 1-proteinase inhibitor (E-alpha 1-PI). Laboratory measurements were corrected for haemodilution. During and after CPB there were only a few significant differences between the groups in systemic haemodynamics and oxygenation, i.e. a lower mean arterial blood pressure after the end of CPB in the nonpulsatile group (65 mmHg, SD = 11 vs 76 mmHg, SD = 11) and a lower SvO2 during rewarming on CPB in the nonpulsatile group (62%, SD = 8 vs 67%, SD = 8). The decrease in percentage of PMNs in the total white blood cell count during CPB was greater in the nonpulsatile group than in the pulsatile group (from 61 to 46% vs 63 to 53% of prebypass value). The steep increase of PMN count at the end of CPB and postoperatively was comparable in both groups. The maximal decrease of CH50 levels, occurring after surgery, was significantly higher in the nonpulsatile group (70% SD = 15 vs 79%, SD = 16, of baseline value), suggesting a greater complement activation. E-alpha 1-PI levels increased significantly in both groups during and after CPB with higher peak levels, obtained at one hour after admission to an intensive care unit, in the nonpulsatile group (316 micrograms/l, SD = 102) than in the pulsatile group (247 micrograms/l, SD = 106). There was a partly inverse correlation between the peak postoperative elastase levels and the PaO2/FiO2 ratios at the first postoperative morning. This ratio was significantly lower in the nonpulsatile group (211, SD = 56) than in the pulsatile group (247, SD = 62). Postoperative respiratory tract infection was more frequent in the nonpulsatile group (n = 9) than in the pulsatile group (n = 2). Adding a pulsatile component to centrifugal blood pumping during CPB may have benefits with regard to the possibly detrimental whole body inflammatory response to CPB. Further studies are warranted to investigate whether these differences will affect clinical outcome.