Solving the Mystery: Insights into the Patients who Develop Cognitive Dysfunction after Surgery

Cameron Bosinski, MD, MS

Persistent neurocognitive disorder is prevalent in patients, especially the elderly and those with preoperative cognitive dysfunction. In this session, “Updating the Confusion: Emerging Physiologic Markers of Postoperative Brain Dysfunction,” held Sunday, March 20 at the IARS 2022 Annual Meeting, the speakers Christopher Hughes, MD, MS, FCCM, Odmara Barreto Chang, MD, PhD, Marcos Lopez, MD, MS, and Eric Vu, MD, MSCI, discussed how biomarkers of brain activity, inflammation and cerebral autoregulation may be important in helping to identify patients that may likely develop cognitive dysfunction after surgery.

Christopher Hughes, MD, MS, FCCM, Professor of Anesthesiology and Chief of Anesthesiology Critical Care Medicine at Vanderbilt University Medical Center, moderated the session. Perioperative brain function is a combination of multiple factors, including preoperative cognitive trajectory, the surgery, anesthetic delivered and postoperative cognitive trajectory. Postoperative delirium occurs for up to 1 week postprocedure or until discharge and results in disturbances in attention and awareness, as well as fluctuating and disordered cognition. Persistent neurocognitive disorder (NCD) occurs after 1 week and must be reported by the patient and be present on objective testing. Impairment of activities of daily living meets criteria for major NCD. NCD is associated with increased length of stay, reduced patient satisfaction, functional decline and increased mortality.

Although there are many ways to determine the mechanisms of delirium, each has its drawback. Animal models do not always translate to humans. Cerebrospinal fluid biomarkers are difficult to obtain. Blood biomarkers are vast, but cannot directly measure specific pathophysiology. Imaging biomarkers are costly, lengthy and may expose patients to radiation. Physiologic markers are clinically relevant markers of cognitive function and were discussed more thoroughly by each speaker.

Odmara Barreto Chang, MD, PhD, an Assistant Professor in Residence at the University of California, San Francisco (UCSF) presented her talk, “EEG Changes as an Indicator of Neurocognitive Resilience.” Ketamine is used as an adjunct for general anesthesia and as an analgesic. Its mechanism is through NMDA receptor antagonism, AMPA receptor potentiation and reduction of inflammation via microglia inactivation and reduction of TNF-alpha and IL-6 signaling.

Dr. Barreto Chang (Barreto Chang 2022) performed a randomized controlled trial in which elderly patients were given ketamine intraoperatively. Their cognitive function was assessed preoperatively and postoperatively. Intraoperatively, cognitively normal patients had higher power in the 10-23 Hz frequency range. Cognitively impaired patients had higher power in the beta range (13-30 Hz) and less power in the alpha range (8-13 Hz) than cognitively normal patients. Cognitively normal patients that received ketamine had significantly increased electroencephalogram (EEG) power in the 10-25 Hz frequency band immediately after administration. Administration of ketamine did not change the EEG power spectrum of cognitively impaired patients from baseline.

Whether patients subsequently developed delirium was correlated with EEG tracing performed intraoperatively. There was no difference in EEG power spectra between patients that later developed delirium and those that did not. However, administration of ketamine did have an effect. Patients that received ketamine, but did not develop delirium, had significantly higher power around 10-25 Hz. Ketamine was not associated with the development of delirium in patients with normal cognitive function at baseline. Ketamine administration did lead to a higher likelihood of developing delirium if the patient was cognitively impaired at baseline 52% vs 20%.

Marcos Lopez, MD, MS, Assistant Professor, Anesthesiology Critical Care Medicine and Associate Program Director, Anesthesiology Critical Care Medicine Fellowship at Vanderbilt University Medical Center, presented on “Vascular and Endothelial Contributions to Postoperative Brain Dysfunction.” Surgery causes an acute stress response. Direct trauma to the vasculature can lead to hypoxia, oxidative stress, hemodynamic shear stress, the release of inflammatory mediators, vasospasm or coagulation (Riedel 2013). Acute brain dysfunction may occur as a result of toxic exposure, direct neuronal damage, endothelial dysfunction, impaired perfusion or increased permeability (Anesthesiology 2013, 118:631-9). Detection of three biomarkers has been used as a surrogate of an acute stress response. These include, S100B, a marker of blood brain barrier disruption, E-selection, which facilitates leukocyte chemotaxis on activated endothelium and plasminogen activator inhibitor-1, and a procoagulant that is released by damaged endothelium.

Previously, it has been shown that these three markers of vascular and blood brain barrier injury are associated with the development of postoperative delirium (Hughes 2016). Uncontrolled vascular inflammatory patterns in a mouse model of Alzheimer’s disease can lead to adverse cognitive effects in the short and long term (Alzheimer’s Dement 2020; 16:734-749). Older mice have been found to have greater neuroinflammation (CD68 and GFAP) and blood brain barrier disruption after surgery compared to younger mice.

Dr. Lopez investigated whether patients who developed delirium after undergoing cardiac surgery would have increased expression of markers of vascular injury intraoperatively. S100B was found to be similar preoperatively and 1 day postoperatively in all patients. However, patients that subsequently developed delirium had higher plasma concentrations of S100B intraoperatively compared to those patients that did not develop delirium. In order to quantify oxidative damage, breakdown products of arachidonic acid, such as isofurans and F₂-isoprostanes were measured. F₂-isoprostanes were elevated after induction of anesthesia, but declined after cardiopulmonary bypass. Isofurans did not increase until cardiopulmonary bypass was initiated, but remained elevated until admission to the ICU. Patients that developed delirium had higher levels of F₂-isoprostanes and isofurans measured intraoperatively. Intraoperative oxidative damage was independently associated with increased odds of delirium. Blood brain barrier damage, as measured by S100B, did not modify the effect of oxidative damage on delirium. Oxidative damage was associated with elevated levels of UCHL1, a marker of CNS neuronal injury. This suggests that systemic oxidative damage can cause CNS neuronal injury only when there is concomitant blood brain barrier damage. Altered control of perfusion and impaired vascular reactivity, may be contributors to postoperative brain dysfunction.

Reactive oxygen species can reduce vascular reactivity and thus perfusion by inhibiting endothelial nitric oxide synthase, by directly inactivating nitric oxide and by oxidizing the heme moiety on soluble guanylyl cyclase (Anesthesiology 2013 118:631-9). Previous studies have used pulse amplitude variability to show that that impaired vascular reactivity is associated with the development of delirium. Dr. Lopez examined arterioles from mediastinal fat isolated during cardiac surgery. Norepinephrine was applied to cause maximum vasoconstriction. Acetylcholine was then applied in increasing doses to measure the relative vascular reactivity. Patients that later developed delirium using CAM-ICU criteria had similar vascular reactivity to patients that did not develop delirium. Because acetylcholine indirectly causes vascular relaxation via the endothelium, the endothelium-independent vasodilator sodium nitroprusside was applied to these blood vessels. Patients that developed delirium had reduced vascular reactivity compared to patients without delirium when directly stimulated by sodium nitroprusside. Future studies will focus on examining neurovascular mechanisms of delirium, the effectiveness of prehabilitation programs as well as trials of agents that decrease vascular permeability or enhance vascular reactivity.

Eric Vu, MD, MSCI, Assistant Professor of Anesthesiology, Northwestern University Feinberg School of Medicine, and Medical Director, Cardiac Anesthesia at Anne & Robert Lurie Children’s Hospital of Chicago, presented on, “Cerebral Autoregulation Monitoring to Identify Patients at Risk for Delirium.” Postoperative delirium increased morbidity and mortality. Preoperative risk factors for delirium include age, ASA physical status, preoperative fasting and anticholinergic drugs. Intraoperative risk factors include abdominal or cardiac surgery, blood loss and surgery duration. A recent review paper (Aldecoa Eur J Anaesthesiol, 2017) provides an overview of the preventative measures, as well as other monitoring and interventions that can be performed to reduce the development of delirium. Currently, there is a lack of intraoperative monitoring and therapy available to prevent delirium.

Previous studies have associated the triple low (MAP < 75 mmHg, BIS < 45, MAC < 0.8) with mortality. When 2/3 are present, the risk of mortality is doubled. When 3/3 are present, the mortality risk is quadrupled. Monitors have been developed to alert anesthesiologists about the presence of a triple low. However, these alerts do not increase interventions by the anesthesiologist and thus do not reduce mortality. Alternatively, monitoring of cerebral autoregulation may be useful to reduce mortality and prevent cognitive dysfunction.

Cerebral autoregulation is able to maintain a constant cerebral blood flow across a range of cerebral perfusion pressures. Mechanisms of autoregulation include arteriolar contraction, neurovascular coupling, and pressure autoregulation. When patients are outside of the limit of autoregulation, they can develop cerebral ischemia or edema. For most people the autoregulation mean is a MAP of 66, but this has been found to vary widely from 43-90 mmHg. Because cerebral autoregulation is highly individual and highly variable, patients may benefit from real-time intraoperative monitoring. A randomized clinical trial implemented cerebral autoregulation monitoring during cardiopulmonary bypass (Brown JAMA Surg 2019) and compared outcomes to control patients that were not monitored. In response to the monitor, patients in the monitoring group received more doses of phenylephrine. The amount of time that patients spend below the lower limit of regulation was reduced by 50% in the monitored group. The primary outcome of this study was the incidence of postoperative delirium. The monitored group had a 30% reduction in delirium compared to the control group. Another study correlated cerebral perfusion with the BIS score during cardiopulmonary bypass (Liu Journal of Clinical Anesthesia 2021). As cerebral perfusion pressure decreased, so did the BIS score. The BIS score decreased significantly when the cerebral perfusion pressure was below the limit of autoregulation. Patients that developed delirium had a low mean BIS score and spent more time below a BIS score of 45.

Dr. Vu (Br J Anaesth 2022) has designed a device that is able to calculate the upper and lower limits of cerebral autoregulation when using an arterial line and near infrared spectroscopy and provide the anesthesiologist with real time feedback. In the future, implementation of cerebral perfusion monitoring may reduce the incidence of postoperative cognitive dysfunction.