Protecting the Brain in Coronary Artery Bypass Graft Surgery

For much of the 35-year history of coronary artery bypass graft (CABG) surgery, the prevailing belief among cardiovascular physicians was that neurological complications of the procedure were infrequent, consisting primarily of a 1% to 5% incidence of stroke.¹ Some patients reported to their physicians that they were "not quite right" intellectually after the procedure, but such problems were not systematically examined and seemed to be uncommon. Using preoperative and postoperative neuropsychological testing, researchers from many centers have now convincingly demonstrated that measurable cognitive dysfunction is actually a common complication of CABG surgery, with an incidence of up to 80% to 90% at hospital discharge.² This apparent epidemic of operative brain injury has created a major dilemma for the physicians who refer more than 800 000 patients for CABG surgery worldwide each year.³ Because the cognitive dysfunction is often not evident on routine clinical or neurological examination, some physicians have accepted the problem as a small but necessary trade off in the quest to improve survival and functional status. However, some researchers have argued that this problem is a major reason the CABG procedure needs to be completely reengineered.

While there is no shortage of theories about the pathogenesis of cognitive dysfunction after CABG surgery, most can be classified into 2 broad categories: patient-related and procedure-related. Patient-related theories focus on the increasing age of the typical CABG patient and the underlying predisposition of elderly patients to develop cognitive dysfunction after any major operation. For example, the International Study of Postoperative Cognitive Dysfunction found that 26% of patients older than 60 years who underwent major abdominal or orthopedic surgery had cognitive dysfunction 1 week after surgery.⁴ Major risk factors included older age, increased duration of anesthesia, and postoperative respiratory and infectious complications. Only age was a risk factor for persistent cognitive dysfunction: at 3 months, 14% of patients aged 70 years or older had cognitive impairment.⁴ A study of 262 elderly patients undergoing elective total knee replacement found that 5% had residual cognitive dysfunction at 6 months, but that this adverse outcome did not depend on whether general or epidural anesthesia had been used.⁵

Thus, a certain proportion of patients subjected to a major pathophysiologic stress such as surgery will demonstrate abnormal postoperative cognitive function and the risk is most pronounced in elderly patients. These findings are particularly relevant in interpreting the cognitive dysfunction observed in CABG patients, because the age distribution has been increasing over the last decade with a current median age of 66 years.⁶ Furthermore, CABG patients typically have extensive atherosclerosis in multiple vascular systems and even "benign" invasive procedures in this population routinely produce detectable microemboli in the cerebral circulation.⁷ To the extent that these patient-related mechanisms account for the cognitive dysfunction observed after CABG surgery, the problem may not indicate any specific deficiency with how CABG surgery is currently performed and the optimal therapeutic approach may focus on finding effective adjunctive neuroprotective agents.⁸

There is now substantial evidence, however, to implicate procedural aspects of CABG surgery in the pathogenesis of cognitive dysfunction independent of any patient predispositions. Procedure-related theories focus on the use of cardiopulmonary bypass (CPB), manipulation of the aorta (especially by cross-clamping), or some other aspect of management such as temperature control in the postoperative period. Of these, most attention has been paid to CPB-related causes. Intraoperative transcranial Doppler monitoring has helped greatly in studying the pathogenesis of operative brain injury in CABG surgery. Most Doppler-detected microemboli occur during the CPB phase of the operation and the load of emboli detected correlates with postoperative neuropsychological deficits.⁹ The distribution of the Doppler emboli into the cerebral circulation seems to determine the extent of regional structural abnormalities of the brain, as well as the amount of detectable neuropsychological dysfunction.¹⁰

While all phases of the operation that involve aortic manipulation significantly increase the prevalence of cerebral microemboli, the period during which aortic cannulation is performed in preparation for the initiation of CPB had the highest prevalence of microemboli and the strongest correlation with brain tissue injury, as reflected by serum markers.¹¹ Other investigators have found that more frequent manipulation of the CPB circuit during the operation by the perfusionist correlated with significantly worse cognitive function 3 months postoperatively.¹² The presumption is that tiny amounts of air are introduced into the circuit with each manipulation resulting in gas microemboli to the brain. However, some of the Doppler-detected microemboli are clearly composed of particulate matter. Pathologic studies in patients who died within 3 weeks of CPB have revealed large numbers of fat-containing microemboli in the cerebral microvasculature.¹³ This embolic load correlated with the duration of CPB; for every extra hour on CPB, the embolic load increased by 90%. Whether these emboli originate in the CPB circuit and its tubing or in the patient's body remains unclear. Hence, the brain injury associated with CPB may result from a combined assault of microscopic amounts of gas and particulate matter on the cerebral circulation.

Despite residual uncertainty about the exact mechanism of CPB-induced brain injury in CABG surgery, the strong association of this problem with CPB has been a major impetus in the development of the off-pump, or beating-heart, CABG (OPCAB).¹⁴ Virtually, no CABG surgeries were performed without CPB before 1995. By 1999, 10% of CABG surgeries in the United States were off-pump.¹⁵ Current estimates are that 20% of CABG surgeries are now performed as off-pump procedures and expectations are that this proportion will reach 50% by 2005.¹⁴ Early enthusiasm for this version of CABG surgery was fueled largely by reports from single centers comparing OPCAB with historical or concurrent controls of CABG surgery with CPB. In these studies, OPCAB appeared to be associated with less myocardial damage, lower rates of other complications, a shorter hospital stay, and lower costs.^{16 - 18} In addition, the procedure was associated with a substantial reduction in Doppler-detected cerebral microemboli.¹⁹ However, 2 initial randomized control trials, each involving less than 100 patients, came to conflicting conclusions about whether OPCAB also reduced postoperative cognitive dysfunction.^{2 ,20}

In this issue of THE JOURNAL, Van Dijk et al²¹ report the neurocognitive and quality of life results of their randomized control trial comparing OPCAB and conventional CABG surgery with CPB in 281 patients. Given the recent marked growth in the use of OPCAB, such data are a welcome addition to the current knowledge base. Initial results from this trial showed that OPCAB reduced the intraoperative need for blood products and the postoperative extent of creatine kinase MB release. Length of stay was reduced by 1 day in OPCAB and there was no difference in event-free survival at 1 month.²² Cognitive dysfunction was determined by a standard battery of 10 neuropsychological tests administered before and after surgery by psychologists blinded to the patients' treatment assignment. At 3 months, the incidence of cognitive decline was 21% in the OPCAB group and 29% in the conventional CABG surgery group (P = .15). At 1 year, cognitive decline was present in 31% of the OPCAB group and 33% of the conventional CABG surgery group (P = .69). Both treatment groups showed a marked improvement in quality of life in the first 3 months after surgery, with some modest further improvements at 12 months, but no differences between treatment groups.

At face value, these results would appear to contradict the theory that avoiding CPB will substantially preserve the cognitive function of CABG surgery patients. However, for several reasons, these results cannot be regarded as definitive. First, the study sample size was selected assuming that the proportion of patients with cognitive decline at 3 months would be decreased by two thirds with OPCAB. The observed reduction was 35%, indicating that the study was substantially underpowered and that a more modest reduction could not be ruled out. Second, patients in the trial were at substantially lower risk for complications in general and for cognitive dysfunction in particular than the patients routinely undergoing OPCAB in the United States. For example, comparing the study cohort with the Society for Thoracic Surgery (STS) Registry cohort of OPCAB reveals that the trial patients averaged 5 years younger than the STS patients, had less advanced coronary artery disease, and had less extracoronary vascular disease.¹⁵ If a more representative group had been studied, the incidence of cognitive dysfunction would likely have been significantly higher than the 29% observed at 3 months and the power of the study to detect a benefit of OPCAB would have been enhanced.

Avoiding CPB may not be enough to have a brain-safe CABG surgery. Combining OPCAB with the new generation of anastomotic devices that permit proximal aortic graft attachment rapidly and without the need for aortic clamping may provide an even more powerful test of the hypothesis that making CABG surgery less invasive will provide better patient outcomes.

While the cognitive dysfunction seen after CABG surgery does improve in the months following surgery, long-term data demonstrate a disturbing trend for cognitive deterioration, particularly among patients who had abnormalities in the early postoperative period. In a prospectively studied cohort study, the 5-year incidence of cognitive dysfunction was almost as high (42%) as the incidence seen at discharge after CABG surgery (53%).²³ While the connection between early and late dysfunction remains speculative, the possibility exists that the early brain injury from CABG surgery increases the late risk of dementia, especially in elderly patients.

The technology of the CABG surgery procedure is evolving rapidly. There is often a very narrow window between when a medical technology is too immature to test in randomized trials and when it is so widely accepted that physicians are unwilling to randomize their patients. The trial by Van Dijk et al,²¹ performed in one of these strategic windows, provides a more balanced view of the current state of neuroprotection with OPCAB. We hope other innovative cardiac surgical groups will follow by demonstrating in even larger randomized trials that their changes to CABG surgery procedures are, in fact, a true advance.