Influence of oxygenator type on the prevalence and extent of microembolic retinal ischemia during cardiopulmonary bypass. Assessment by digital image analysis

Abstract

We have previously reported the occurrence of microembolic ischemia in the retina during cardiopulmonary bypass, as revealed by fluorescein angiography. This method has been extended by digital image analysis to include quantification of the extent of retinal ischemia and has been applied to a prospective comparative study of 64 patients undergoing elective coronary operations with either a bubble or a membrane oxygenator. Patients with diabetes or clinically evident cerebrovascular disease were excluded. Bypass procedures were standardized in all cases with pulsatile flow and a 40 microns arterial line filter (Pall EC Plus). Thirty patients had bypass with a bubble oxygenator (Harvey H1700) and 34 patients had bypass with a flat sheet membrane oxygenator (COBE CML). In each case retinal fluorescein angiograms were obtained preoperatively and 5 minutes before the end of bypass and were processed with a digital image analyzer (Context Vision GOP-302). Microembolic perfusion defects were identified by digital subtraction of preoperative and end-bypass angiograms and their total area was computed. Results. In the bubble oxygenator group retinal perfusion defects indicative of microembolism occurred in all 30 (100%) patients. In contrast, over half the patients in the membrane oxygenator group had normal retinal perfusion, and the prevalence of perfusion defects (44%; 70% confidence limits 34% to 54%) was significantly less than in the bubble group (p less than 0.001). In addition, those patients in the membrane group had significantly fewer lesions (median 0; 70% confidence limits 0 to 1) than patients in the bubble group (median 2; 70% confidence limits 2 to 2; p less than 0.001) and also had significantly smaller total areas of retinal ischemia (median 0 mm2; 70% confidence limits 0 to 0.16 mm2) than the bubble group (median 0.22 mm2; 70% confidence limits 0.21 to 0.27 mm2; p less than 0.001). There was no relationship between the extent of retinal ischemia and bypass time, arterial blood gas concentrations, volume of cardiotomy suction or donor blood returned to the pump, or recent medication with aspirin. Inferences. Digital image analysis of retinal fluorescein angiograms may provide a method of quantifying microembolic ischemia in the central nervous system during cardiopulmonary bypass. Flat sheet membrane oxygenation appears to provide significantly better protection against microembolic ischemia than bubble oxygenation.