Page 33 - ANZCP Gazette May 2023

6
Perfusion
exposure in cardiac surgery should be a priority. Attempts to correlate emboli exposure to clinical out- comes such as cognitive function deficits following car- diac surgery have generated approximate equipoise in the relevant literature.25 In the absence of very large studies, such correlations may be too simplistic because the pathophysiology of deficits may involve multiple co- dependent risk factors. For example, emboli might impair autoregulation, but this might not become meas- urably harmful unless there is concurrent hypotension for a threshold period. Nevertheless, identifying emboli as a cause of autoregulation impairment would support an argument that exposure should be minimized. The latter argument would, in turn, be supported by the fact that measurement of autoregulation using a similar principle to that employed in the present study was pre- viously shown to be prognostic for stroke in patients undergoing CPB.7
We acknowledge that our approach to evaluating the effect of emboli exposure on cerebral autoregulation in humans constituted an imperfect experimental model. It is hard to conceive a design for a better study that would be ethical. The primary strength of our study lay in its being a human study performed ethically by utilizing the naturally occurring differences in emboli exposure in OCS and CCS. However, because the emboli exposure could not be controlled or manipulated, the groups were not as clearly separated as could be achieved in an animal study. On the other hand, it is not always clear that data from animal studies can be extrapolated to humans, whereas our data clearly do reflect real-world clinical practice. All patients in both groups were exposed to some emboli, which may well have affected cerebral vas- oreactivity in all subjects. This would have contributed to the difficulty of demonstrating differences between the groups, but it is the reality of CPB. Similarly, anesthetic drug administration, perfusion strategies and gas man- agement also had the potential to affect cerebral vasore- activity. Evaluation of anesthetic drug choices revealed an identically balanced use of propofol and vapors as the primary anesthetics prior to and during CPB across the two groups. Slightly more patients received propofol alone for maintenance of anesthesia after the aortic cross-clamp removal in closed chamber surgery (three patients) than open chamber surgery (no patients). However, if this were a significant disadvantage contrib- uting to the increase in Mx in the open chamber patients, it would be expected to manifest symmetrically in both cerebral hemispheres; not just the right side where the most plausible explanation for the observed asymmetric change is that it was exposed to many more emboli.
Finally, two other potential confounding issues need to be acknowledged. First, the study cohort was heavily skewed to male patients (Table 1). In a study investigat- ing a pathophysiological process and attempting to
minimize variables, this might be considered advanta- geous; animal studies are often restricted to a single- gender. However, in our case, this was entirely unintentional and simply reflected a chance outcome of patient and researcher availability. Second, the patients undergoing open and closed-chamber surgery started from different Mx baselines prior to the release of the aortic cross-clamp (Table 2), a finding we cannot explain. There was no order effect; patients undergoing both forms of surgery were recruited steadily and approximately evenly throughout the study. Nor were patients in either group cared for by particular surgeons, anesthesiologists or perfusionists. Perfusion and anes- thesia protocols, including drug and temperature man- agement, did not differ between the groups, and emboli exposure before the release of the aortic cross-clamp was not greater in the CCS group. Based on the distribu- tion of readings, there was no evidence for a ceiling effect at or near the mean baseline Mx for the CCS group. We, therefore, contend that the intergroup base- line difference does not invalidate the prospectively finalized analysis plan, which was to measure the change in Mx between the pre and post aortic cross-clamping phases and to compare the two groups on that basis.
Conclusions
Patients undergoing OCS were exposed to more emboli after the release of the aortic cross-clamp than patients undergoing CCS. There was no overall difference between groups in the change in autoregulation from before to after removal of the aortic cross-clamp. However, the majority of the emboli were directed to the right cerebral hemisphere, and the difference between groups in the increase in the correlation between MAP and CBFv in the right cerebral hemisphere was signifi- cant, implying a greater impairment in autoregulation associated with a larger embolic load. This secondary outcome supports a larger study designed and powered on the basis of our results to definitively resolve the ques- tion of whether cerebral emboli influence autoregulation during cardiac surgery. Such a study could include measures of mean arterial pressure and cognitive change after surgery with the aim of attempting to associate any adverse cognitive change with emboli, imparted auto- regulation and corresponding hypotension.
Acknowledgements
We thank Professors Charles Hogue and Alan Barber for their advice on study methods.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
MAY 2023 | www.anzcp.org 30