Pro: during cardiopulmonary bypass for elective coronary artery bypass grafting, perfusion pressure should routinely be greater than 70 mmHg.

W ITHIN THE SPAN OF 50 years, cardiopulmonary bypass (CPB) has been transformed from an uncommon hazardous technique to a sophisticated, everyday occurrence. However, questions regarding the fundamental management of bypass remain unanswered: What is the optimum temperature management strategy during CPB? What is the optimum pump flow? Is pulsatile blood flow more physiologic during CPB? What is the optimum perfusion pressure during CPB ? Answering these questions is critical to reducing the substantial incidence of postoperative neurologic dysfunction that occurs after CPB. This incidence may be as high as 6% 1 for gross neurologic injury or 57% 2 for mild cognitive changes. The cause of postoperative neurologic dysfunction is not fully understood, but is believed to be a result of both embolic phenomena and hypoperfusion during CPB. The ideal perfusion pressure therefore represents a careful balance of risks. Reduced perfusion pressure and blood flow may decrease cerebral embolic load but increase the risk of hypoperfusion, whereas elevated perfusion pressures and flow may decrease the likelihood of hypoperfusion but increase cerebral embolic load. Although the ideal perfusion pressure has not yet been defined, using the lowest perfusion pressure that does not risk hypoperfusion would seem logical. This ideal or routine perfusion pressure is the subject of considerable debate, and the focus of this article. Cerebral autoregulation maintains a relatively constant blood flow for mean arterial blood pressure between 50 and 150 mmHg, assuming other physiologic variables remain constant (Table 1). During hypothermic CPB, the lower limit of cerebral autoregulation reaches 20 to 30 mmFlg. 3,4 Because cerebral autoregulation is left-shifted during hypothermic CPB, hypoperfusion would not be expected to occur unless perfusion pressure was significantly less than 50 mmHg, and routine CPB perfusion pressure near 50 mmHg would appear to be safe. The routine use of higher perfusion pressure (-->70 mmHg) may lead to worse nenrologic outcome. The evidence is as follows. During alpha-stat acid-base management, the pressureflow antoregulatory plateau has a slightly positive slope, 5 so that increasing perfusion pressure increases cerebral blood flow 6,7 and thus increases embolic load. These phenomena lead to an increased incidence of adverse neurologic outcomes. 8 Increased cerebral blood flow associated with increased adverse neurologic events is also seen during pH-stat acid-base management.3, 9 In addition to neurologic complications, higher perfusion pressures (->70 mmHg) may be associated with other adverse events. Increased pressure and blood flow may cause greater trauma to blood elements, exacerbate the inflammatory process associated with CPB, and reduce the effectiveness of circulating blood elements. Finally, excessively high perfusion pressure during CPB greatly compromises one of the primary imperatives of bypass, the opportunity for the surgeon to operate on a still heart in a bloodless field. Although it would be difficult, if not impossible, to quantify the effect of a blood-filled versus bloodless field on the long-term patency of coronary anastomoses, clinicians should carefully consider the opinion offered by Dr Norman Shumway: "The most important predictor of long-term outcome is the surgeon's ability to complete the coronary anastomoses." (Personal communication, 1992.) Of all the adverse events potentially associated with perfusion pressure management during CPB, neurologic outcome has received the most attention. Early studies 1°-~3 found severe hypotension during CPB was associated with postoperative neurologic deficits, highlighting the potential role of hypoperfusion in postoperative neurologic dysfunction. However, the magnitude and duration of hypotension in these studies were not well quantified. Stockard et a114 helped to correct this deficit by establishing the concept of ml 5° (the integral of perfusion pressure-< 50 mmHg over time). The subsequent study by Stockard et aP 5 of 75 cardiac surgical patients reported a 53% incidence of postoperative central nervous system dysfunction in patients who experienced an integrated hypotensive time of 100 mmHg. rain or greater versus a 10% incidence of postoperative central nervous system dysfunction in those who did not suffer this magnitude of hypotension. However, tm 5° or other measures of hypotension (equivalent to perfusion pressure-< 50 mmHg) do not consistently predict postoperative neurologic deficits (Table 2). For example, Slogoff et all6 in a prospective study of 204 patients, found perfusion pressures of 50 mmHg or less during CPB were not associated with an increase in the likelihood of postoperative cerebral dysfunction. In a subsequent study, ~7 neither the value of the lowest measured blood pressure nor integrated hypotension was predictive of postoperative neurologic dysfunction. Additional studies have both supported ls-2° and refuted 4,21-3° the relationship between intraoperative hypotension and postoperative neurologic dysfunction (Table 2). Direct comparison of these studies is difficult because outcomes are reported differently (neurologic v neuropsychologic), CPB management techniques have evolved over the last 30 years (eg, pH v alpha-stat acid-base management, use of arterial filters, etc), and study designs vary (retrospective v prospective). Most of the studies that suggested an association between hypotension and worsening neurologic outcome have used 50 mmHg as the limit for defining hypotension. Because of this, 50 mmHg has become a common, if not routine, lower limit for perfusion pressure during CPB, despite a substantial amount of conflicting data.