• Arterial conduits
• cardiac surgery
• cardiopulmonary bypass
• surgery–coronary bypass
• surgery–valve

Adverse neurological outcomes have plagued cardiac surgical procedures since the origins of the speciality. The spectrum of adverse neurological events range from clinically overt postoperative stroke to minor cognitive impairment. Despite an abundance of clinical investigations that have sought to identify and modify risk factors for these events, practice patterns have not changed drastically with regards to extracorporeal circulation and aortic manipulation, with most centres establishing protocols within the confines of their institutions and comfort level.

In their paper published in Heart, Anastasiadis and colleagues1 report results comparing patients undergoing coronary artery bypass surgery (CABG) using a minimal extracorporeal circuit (MECC) compared with a conventional circuit (CECC). The authors are to be congratulated on a well-designed randomised prospective study that adds to the growing body of literature supporting the use of MECC during cardiac operations requiring cardiopulmonary bypass (CPB). In this study, 32 patients undergoing CABG were randomly allocated to each group. Regional cerebral oxygen saturation was monitored using near-infrared spectroscopy (NIRS), and neurocognitive function was assessed preoperatively, at discharge and at 3 months postoperatively using a battery of standardised neurocognitive tests. Patient demographics and clinical characteristics were similar between the groups, with no significant difference in mortality and no major postoperative neurological events in either group. From their NIRS data, there was a significantly higher incidence of cerebral oxygen desaturation episodes (CODE) in the CECC group compared with the MECC group (55% vs 38%, p=0.04). A regional oxygen desaturation score greater than 3000, a risk marker of greater early postoperative neurocognitive decline, was more common in the CECC group compared with the MECC group (35% vs 21%, p=0.01).2 Three-month neurocognitive results showed better neurocognitive performance on most tests for the MECC group compared with the CECC group and even improved results from baseline in the MECC group for most of these tests. Importantly, although there were significant differences between MECC and CECC at 3 months, neurocognitive function in the CECC group was relatively stable compared with baseline values, the differences being due to improved neurocognitive scores in the MECC group. Although differences within each group were not statistically compared between 3-month and baseline values in either group, group differences were detected at the various time points. In the CECC group, 61% of patients compared with 21% of patients in the MECC group (p<0.01) had evidence of cognitive decline at 3 months compared with baseline. Finally, cognitive decline was significantly correlated with at least one CODE in the CECC group compared with the MECC group (63% vs 33%, p=0.04). This last result suggests that at least part of the neurocognitive differences observed were due to cerebral hypoperfusion during CPB.

What can we conclude based on these results as well as others about strategies to minimise postoperative neurocognitive dysfunction? In cardiac surgical operations, several interventions and precautions are likely to have a cumulative effect in improving not only neurocognitive impairment but also other neurological outcomes. In patients requiring CPB the following precautions should be taken: maintaining adequate systemic perfusion pressures (60–80 mm Hg), moderate hypothermia, haematocrit greater than 20%, CPB flow rates between 2.0 and 2.4 l/min per m², α stat pH management; the avoidance of direct reinfusion of shed blood from pericardial space into the CPB circuit by cardiotomy sequestration or using cell saver, and minimising aortic manipulation using a single cross-clamp technique when feasible. The MECC system probably represents an additional level of precaution because it takes advantage of improvements in perfusion technology.3–5 The MECC system utilises a centrifugal pump and small-prime heparin bonded circuits with no cardiotomy suction. The priming volume is reduced by using retrograde autologous priming, and arterial and venous line filters are not included. At least in this study, the incidence of cerebral hypoperfusion was lower with the MECC system compared with CECC.

Other methods have been investigated that may reduce neurocognitive impairment that were not assessed in this study. These include avoiding CPB altogether using off-pump revascularisation techniques and minimising intraoperative cerebral embolic events by avoiding aortic cannulation and clamping. Off-pump coronary artery bypass provides the opportunity for a ‘no touch’ aortic technique that eliminates the risk of cerebral embolisation associated with aortic manipulation and the inflammatory response associated with CPB.6 Intuitively, even the most ideal perfusion circuit would be associated with a greater degree of aortic manipulation and CPB-induced inflammation compared with a no aortic touch off-pump operation, although this has been challenged in recent trials.3 In patients undergoing CABG with CPB support, a single cross-clamp strategy may be more effective than two clamps (cross-clamp and partial occluding clamp). Hammon and colleagues7 showed reduced neuropsychological deficits in patients undergoing on-pump CABG with a single cross-clamping strategy compared with a two-clamp strategy (the strategy utilised in this study). Assessing ascending aortic atherosclerosis with epiaortic ultrasonography may provide valuable information by quantifying the burden of disease. This non-invasive test is quick and effective and provides the surgeon with valuable information about the safety of aortic cannulation and aortic clamping.8 Although the literature supporting any one of these strategies is not definitive or overwhelmingly supportive, combining several of these strategies is likely to be beneficial in reducing the neurological sequelae of cardiac operations.

Neurocognitive dysfunction represents an elusive endpoint that is assessed with a complex but well-validated battery of tests.9 Unlike MRI, NIRS or transcranial Doppler, which identify hard endpoints that are tangible and used to document stroke, cerebral oxygenation and cerebral embolic events, respectively, neurocognitive testing yields scores that can be difficult to interpret. Each of these modalities has its strengths and weaknesses and can often provide valuable information when corroborated with another modality. In this study, the short-term improvement in neurocognitive function in the MECC group has been corroborated by others,10 11 although long-term neurocognitive function tends to decline in patients with coronary disease regardless of whether they undergo CABG or not.11 Selnes and colleagues11 showed that patients with coronary artery disease had lower baseline cognitive performance and greater degrees of decline over 72 months when compared with a group of heart-healthy subjects, although there were actual improvements in all groups from baseline to 12 months. Those results imply that patients undergoing CABG have some degree of cerebrovascular disease that inevitably affects long-term cognitive performance, regardless of whether they undergo on-pump or off-pump CABG or medical therapy.12 However, early neurocognitive deficits differ from late neurocognitive decline, with early deficits (within 3 months) probably related to a host of surgical variables including cerebral hypoperfusion, intraoperative and postoperative emboli, the overall inflammatory response and the use of general anaesthesia. Although the CECC group did not show resolution of neurocognitive impairment at 3 months compared with the MECC group in this trial, others have shown that patients undergoing conventional CPB return to baseline after CABG.11

The improvement in neurocognitive function in the MECC group suggests some component of practice effect. In other words, patients performed better because they remembered the test. Whether this occurred because of less memory impairment with MECC is unknown, but the lack of improvement in the CECC group compared with MECC suggests less early cognitive impairment with the MECC system. Given the higher incidence of CODE and worse desaturation scores in the CECC group, it is plausible that these differences could account for the different neurocognitive scores between groups. This suggests that perfusion is improved with the MECC system, which should be more thoroughly investigated.

With neurocognitive testing, do statistically significant differences in scores translate into clinically significant differences? In other words, does a difference in the digit span-backward test 4.7±1.4 vs 3.5±1.1 translate into a meaningful clinical difference despite a p value of 0.008? Although the authors compared neurocognitive scores between MECC and CECC at three different time points, it would have been helpful to see if there were differences within each group over time. Nonetheless, this is a common limitation of neurocognitive testing and does not significantly detract from the value of the paper.

The MECC system probably provides some protective benefit to neurocognitive function after CABG, which has been demonstrated in other studies. In addition to a reduced inflammatory response by a smaller heparin-bonded circuit and centrifugal pump, the use of cell-saver technology in lieu of cardiotomy suction probably reduces lipid cerebral microembolisation by minimising the autotransfusion of mediastinal blood.13 14 Combining these refinements in perfusion technology with strategies to reduce aortic manipulation (single cross-clamp technique) will probably decrease the incidence and impact of adverse neurological events. In patients revascularised without CPB (off-pump), minimising aortic manipulation using clampless facilitating devices or in-situ arterial grafts represents a valuable alternative strategy, especially in patients with more advanced ascending aortic atherosclerosis.

Neurocognitive impairment and postoperative stroke continue to plague the results of CABG and cardiac surgical operations in general. Improvements in perfusion technology with miniaturised extracorporeal circulation represents a valuable strategy that can be combined with other techniques to attenuate early postoperative neurocognitive dysfunction in patients undergoing CABG, despite an ageing population of patients with multiple comorbidities being referred for these procedures.