Surgery and Cognition: Delirium, Cognitive Decline, and Opportunities to Protect the Brain

Resources

I. Surgery and the Potential Intersections of Delirium, Cognitive Decline, and Dementia

Cavallari M, Dai W, Guttmann CR, et al. Neural substrates of vulnerability to postsurgical delirium as revealed by presurgical diffusion MRI. Brain. 2016;139(Pt4):1282-94.

Fong TG, Davis D, Growdon ME, et al. The interface between delirium and dementia in elderly adults. Lancet Neurol. 2015;14(8):823-32.

Inouye SK, Marcantonio ER, Kosar CM, et al. The short-term and long-term relationship between delirium and cognitive trajectory in older surgical patients. Alzheimers Dement. 2016;12(7):766-75.

Saczynski JS, Marcantonio ER, Quach L, et al. Cognitive trajectories after postoperative delirium. N Engl J Med. 2012 5;367(1):30-9.

Witlox J, Eurelings LS, de Jonghe JF, et al. WA. Delirium in elderly patients and the risk of postdischarge mortality, institutionalization, and dementia: a meta-analysis. JAMA. 2010;304(4):443-51.

II. Anesthesia and Sedation

Boveroux P, Vanhaudenhuyse A, Bruno MA, et al. Breakdown of within- and between-network resting state functional magnetic resonance imaging connectivity during propofol-induced loss of consciousness. Anesthesiology. 2010;113(5):1038-53.

ClinicalTrials.gov. Trajectory of Recovery in the Elderly (TORIE). June 23, 2016.

ClinicalTrials.gov. Electroencephalography Guidance of Anesthesia ENGAGES) Study. December 23, 2015.

Escallier KE, Nadelson MR, Zhou D, Avidan MS. Monitoring the brain: processed electroencephalogram and peri-operative outcomes. Anaesthesia. 2014;69(8):899-910.

Fritz BA, Kalarickal PL, Maybrier HR, et al. Intraoperative electroencephalogram suppression predicts postoperative delirium. Anesth Analg. 2016;122(1):234-42.

Mashour GA, Avidan MS. Postoperative delirium: disconnecting the network? Anesthesiology. 2014;121(2):214-6.

Murray C, Sanderson DJ, Barkus C, et al. Systemic inflammation induces acute working memory deficits in the primed brain: relevance for delirium. Neurobiol Aging. 2012 Mar;33(3):603-616.e3.

Peng M, Zhang C, Dong Y, et al. Battery of behavioral tests in mice to study postoperative delirium. Sci Rep. 2016;6:29874.

van Dellen E, van der Kooi AW, Numan T, et al. Decreased functional connectivity and disturbed directionality of information flow in the electroencephalography of intensive care unit patients with delirium after cardiac surgery. Anesthesiology. 2014;121(2):328-35.

van der Kooi AW, Zaal IJ, Klijn FA, et al. Delirium detection using EEG: what and how to measure. Chest. 2015;147(1):94-101.

Whitlock EL, Torres BA, Lin N, et al. Postoperative delirium in a substudy of cardiothoracic surgical patients in the BAG-RECALL clinical trial. Anesth Analg. 2014;118(4):809-17.

Xie Z, Dong Y, Maeda U, et al. The common inhalation anesthetic isoflurane induces apoptosis and increases amyloid beta protein levels. Anesthesiology. 2006 May;104(5):988-94.

Xie Z, Culley DJ, Dong Y, et al. The common inhalation anesthetic isoflurane induces caspase activation and increases amyloid beta-protein level in vivo. Ann Neurol. 2008 Dec;64(6):618-27.

Zhang Y, Dong Y, Wu X, et al. The mitochondrial pathway of anesthetic isoflurane-induced apoptosis. J Biol Chem. 2010 Feb 5;285(6):4025-37.

Zhang Y, Dong Y, Xu Z, Xie Z. Propofol and magnesium attenuate isoflurane-induced caspase-3 activation via inhibiting mitochondrial permeability transition pore. Med Gas Res. 2012 Aug 17;2(1):20.

Zhang B, Tian M, Zheng H, et al. Effects of anesthetic isoflurane and desflurane on human cerebrospinal fluid Aβ and τ level. Anesthesiology. 2013 Jul;119(1):52-60.

III. Clinical Biomarkers and Therapeutics

Berger M, Burke J, Eckenhoff R, Mathew J. Alzheimer's disease, anesthesia, and surgery: a clinically focused review. J Cardiothorac Vasc Anesth. 2014;28(6):1609-23.

Berger M, Nadler JW, Friedman A, et al. MAD-PIA trial team. The effect of propofol versus isoflurane anesthesia on human cerebrospinal fluid markers of Alzheimer's Disease: results of a randomized trial. J Alzheimers Dis. 2016;52(4):1299-310.

Biomarker Development for Postoperative Cognitive Impairment in the Elderly. BioCog project. August 29, 2016.

Cape E, Hall RJ, van Munster BC, et al. Cerebrospinal fluid markers of neuroinflammation in delirium: a role for interleukin-1β in delirium after hip fracture. J Psychosom Res. 2014;77(3):219-25.

ClinicalTrials.gov. Markers of Alzheimers Disease and Cognitive Outcomes After Perioperative Care (MADCO-PC) Study. January 7, 2016.

Clark CM, Pontecorvo MJ, Beach TG, et al. Cerebral PET with florbetapir compared with neuropathology at autopsy for detection of neuritic amyloid-β plaques: a prospective cohort study. Lancet Neurol. 2012;11(8):669-78.

Dillon ST, Vasunilashorn SM, Ngo L, et al. Higher c-reactive protein levels predict postoperative delirium in older patients undergoing major elective surgery: a longitudinal nested case-control study. Biol Psychiatry. 2016.

Faulkner S, Bainbridge A, Kato T, et al. Xenon augmented hypothermia reduces early lactate/N-acetylaspartate and cell death in perinatal asphyxia. Ann Neurol. 2011;70(1):133-50.

Franks NP, Dickinson R, de Sousa SLM, et al. How does xenon produce anaesthesia? Nature. 1998;396(6709):324.

Fries M, Brücken A, Çizen A, et al. Combining xenon and mild therapeutic hypothermia preserves neurological function after prolonged cardiac arrest in pigs. Crit Care Med. 2012;40(4):1297-303.

Hoogland ICM, Houbolt C, van Westerloo DJ, et al. Systemic inflammation and microglial activation: systematic review of animal experiments. J Neuroinflammation. 2015;12:114.

Ji MH, Yuan HM, Zhang GF, et al. Changes in plasma and cerebrospinal fluid biomarkers in aged patients with early postoperative cognitive dysfunction following total hip-replacement surgery. J Anesth. 2013;27(2):236-42.

Laitio R, Hynninen M, Arola O, et al. Effect of inhaled xenon on cerebral white matter damage in comatose survivors of out-of-hospital vardiac arrest: a randomized clinical trial. JAMA. 2016;315(11):1120-8.

Ma D, Hossain M, Chow A, et al. Xenon and hypothermia combine to provide neuroprotection from neonatal asphyxia. Ann Neurol. 2005;58(2):182-93.

Marcantonio ER. Postoperative delirium: a 76-year-old woman with delirium following surgery. JAMA. 2012; 308(1):73-81.

Munster BC, Aronica E, Zwinderman AH, et al. Neuroinflammation in delirium: a postmortem case-control study. Rejuvenation Res. 2011;14(6):615-22.

Sheng SP, Lei B, James ML, et al. Xenon neuroprotection in experimental stroke: interactions with hypothermia and intracerebral hemorrhage. Anesthesiology. 2012;117(6):1262-75.

Stoppe C, Coburn M, Fahlenkamp A, et al. Elevated serum concentrations of erythropoietin after xenon anaesthesia in cardiac surgery: secondary analysis of a randomized controlled trial. Br J Anaesth. 2015;114(4):701-3.

Tang JX, Baranov D, Hammond M, et al. Human Alzheimer and inflammation biomarkers after anesthesia and surgery. Anesthesiology. 2011;115(4):727-32.

Xie Z, Swain CA, Ward SA, et al. Preoperative cerebrospinal fluid β-Amyloid/Tau ratio and postoperative delirium. Ann Clin Transl Neurol. 2014;1(5):319-328.

Xie Z, McAuliffe S, Swain CA, et al. Cerebrospinal fluid aβ to tau ratio and postoperative cognitive change. Ann Surg. 2013;258(2):364-9.

Westhoff D, Witlox J, Koenderman L, et al. Preoperative cerebrospinal fluid cytokine levels and the risk of postoperative delirium in elderly hip fracture patients. J Neuroinflammation. 2013;10:122.

Sponsors

Presented by

The Brain Dysfunction Discussion Group is proudly supported by

How to cite this eBriefing

The New York Academy of Sciences. Surgery and Cognition: Delirium, Cognitive Decline, and Opportunities to Protect the Brain . Academy eBriefings. 2016. Available at: www.nyas.org/POCD-eB

Highlights

Postoperative delirium, postoperative cognitive decline, and dementia are interrelated.

Patients who experience postoperative delirium have an increased risk of cognitive impairment three years afterwards.

Researchers must develop a clear definition of postoperative cognitive decline and use it consistently in clinical studies to better understand its causes and effects.

Does delirium lead to dementia?

Adults over 65 years of age undergo 19.2 million surgical procedures in the United States each year, representing fully half of the country's surgical population, and those numbers are rising sharply, said Sharon K. Inouye of Harvard Medical School / Hebrew Senior Life.

Older adults also experience complications at a rate 4 or 5 times that of younger patients. Many studies have identified postoperative delirium as the most frequent and the most impactful surgical complication. In a significant number of cases, delirium can lead to cognitive decline, and according to one recent meta-analysis, raises the risk of dementia 12.5-fold. Many clinicians who treat the elderly can recall their patients' families telling them, "Mom was fine until she had that surgery, but after the surgery she changed." Both delirium and dementia are associated with inflammation and cholinergic mechanisms. Yet the interrelationship between these conditions is unclear.

The interrelationship between postoperative delirium, postoperative cognitive decline, and dementia remains poorly understood

In her plenary talk, Inouye stacked the evidence that can support a causal relationship between delirium and subsequent dementia. A recently-published study which she helped conduct in a cohort of 560 patients age 70 and older called SAGES — Successful Aging After Elective Surgery — found that three years after surgery, the patients who experienced postoperative delirium on average showed a rate of cognitive decline consistent with mild cognitive impairment, while non-delirium patients returned to their baseline cognitive scores.

Delirium is a risk factor for dementia, and conversely, dementia is a risk factor for delirium. However, some clinical pathological evidence suggests that dementia following delirium may develop through a different biological pathway than conventional Alzheimer's pathology. For example, the SAGES study found that biomarkers associated with Alzheimer's disease such as ApoE4-e4 status and volumetric measurements (e.g., hippocampal volume) determined by MRI are not predictive of postoperative delirium.

In the SAGES study, patients who experienced postoperative delirium showed long-term decline in cognitive performance when assessed three years later.

Biomarker evidence to date does not paint a clear picture of the link between delirium and dementia, Inouye said. Neuroinflammation appears to be an important contributor to the relationship between surgery and delirium — perhaps through pre-existing inflammation or by triggering an inflammatory response during or after surgery that causes cytokines to flood the brain through a leaky blood–brain barrier. Studies show that rats, mice, or other animals that receive experimentally induced inflammatory challenges show cognitive deficits that can be at least partially rescued by anti-inflammatory molecules.

The neuroinflammation hypothesis predicts that increased inflammation due to disease state and/or following surgery and anesthesia could contribute to postoperative delirium and cognitive decline.

Understanding the interrelationship between predisposing factors, postoperative delirium, and postoperative cognitive decline has been even more challenging because unlike postoperative delirium, there is no accepted definition for postoperative cognitive decline. In fact, it is defined in the literature in as many as 100 different ways. "That's really a problem for the field," Inouye said. One point of divergence is time of onset. "Everyone — even healthy normals, will decline postoperatively, and there's no accepted time standards" for when a reduction in cognitive abilities should be diagnosed as postoperative cognitive decline, she said. In the SAGES study, Inouye and her colleagues used two variations of a psychological test battery to assess the condition and got different results depending on which they were using. Many studies do not even define a preoperative cognitive baseline, Inouye noted, which makes it impossible to determine if a change has occurred.

Inouye ended her talk with several recommendations for the field. First, identifying baseline cognitive status is absolutely essential, she said. Furthermore, all studies should measure postoperative delirium, and the field should agree on what domains should be tracked for neuropsychological testing. Also, she said, the field must recognize that full recovery after postoperative cognitive decline even for healthy older individuals may take several months or even a year.

Highlights

Researchers are using fMRI imaging to explore how anesthesia affects brain activity.

A battery of cognitive tests in mice appears to recapitulate some aspects of postoperative delirium and cognitive decline.

Electroencephalography conducted during surgery may help guide clinicians on the use of anesthesia and reduce the risk of postoperative delirium.

Anesthesia, neural networks, and the trajectory of recovery

The role of anesthesia in cognitive function and brain activity after surgery in the elderly is currently unknown. Joshua Mincer from the Icahn School of Medicine at Mount Sinai in New York described his work using functional magnetic resonance imaging (fMRI) to assess the effects of anesthesia in healthy elderly people.

fMRI detects changes in blood oxygen levels that correspond to an increase in brain activity. When researchers first began using the technique in the 1990s, they recorded brain activity of subjects performing a specific sensory or motor task, or a task probing a cognitive domain such as memory or attention. They found that such tasks often activated brain areas that were disconnected anatomically but that formed functionally linked networks.

More recently, scientists have found that so-called resting state brain activity, which occurs spontaneously when subjects are not engaged in a specific task, can be mapped to essentially the same set of functional networks. The same pattern of resting state brain activation is also seen under anesthesia. Studying the changes within and between resting state and functional connectivity networks can reveal how anesthesia may be affecting the brain, Mincer said.

In healthy volunteers age 40–80 recruited to an ongoing study called Trajectory of Recovery in the Elderly (TORIE), Mincer and his colleagues first conduct extensive baseline cognitive testing and perform baseline fMRI scans. They then conduct fMRI at several time points: while the subjects undergo two hours of anesthesia at a depth appropriate for surgery, immediately after they emerge from anesthesia, and two and seven days later.

Dimensionality, measured with fMRI, varies widely across 11 subjects under anesthesia. Each colored line represents a single subject's dimensionality across time, normalized to zero at baseline.

Mincer is especially interested in a brain activity measure termed "dimensionality" — defined as the number of networks activated in the brain at any given point in time. A decrease in dimensionality reflects a network turning off, while an increase can occur when a network fragments — that is, its component parts cease to fire together. In the eleven subjects tested so far, dimensionality has varied widely under anesthesia, though it returned to baseline by the second day afterwards. In one person, for example, dimensionality increased by 150% under anesthesia, while in another it decreased by 50%. "We can begin to think about this dimensionality concept as a biomarker of a subject's response to anesthesia," Mincer said. Responses could broadly be broken down by whether dimensionality increased or decreased compared to baseline, he added.

Mincer and his colleagues are just beginning to piece together what changes in dimensionality might mean functionally for the brain. Recent studies on how anesthetics cause loss of consciousness are well aligned with the idea of network fragmentation and an increased dimensionality under anesthesia. But older research suggests that anesthesia quiets or shuts down the brain, fitting with a decrease in dimensionality under anesthesia.

The researchers hope to test a total of seventy to eighty subjects of different ages to explore how exactly networks change as dimensionality changes, how robust these changes are, and whether they vary with age. They also plan to correlate dimensionality with other biomarkers and with cognitive data.

Animal models for postoperative delirium and cognitive dysfunction

Despite the frequency of postoperative delirium and postoperative cognitive dysfunction in geriatric patients, the mechanisms underlying these conditions are poorly understood, in part due to a lack of good animal models. Zhongcong Xie from Massachusetts General Hospital and Harvard Medical School in Boston described his work developing animal tests for these conditions.

Isoflurane (but not desflurane) induces apoptosis and cytotoxicity in mice.

A previous study used a single maze test to assess postoperative delirium in mice. Ideally, though, animal tests should be conducted as a battery, mimicking the several tests administered to patients to assess changes in thinking and consciousness. Tests should also not require training the animal. To address these gaps, Xie and his colleagues developed a three-test battery for mice analogous to the Confusion Assessment Method used for patients. Six and nine hours after anesthesia and abdominal surgery, mice showed deficits on these tests, but returned to baseline after 24 hours. Mice and humans are very different, and demonstrating disorganized thinking in mice is difficult, but the results suggest that this combination of tests recapture some aspects of delirium in patients.

To study postoperative cognitive decline, Xie and his colleagues combined rodent tests for learning and memory with cell biology studies. Early work in cell studies and mice revealed that isoflurane, a commonly used anesthetic, induced the activation of capsase-3, a marker of apoptosis, and increased levels of amyloid beta, a hallmark of Alzheimer's disease. Further studies found that isoflurane increased the expression of the gene Bax, a promoter of apoptosis, and decreased expression of Bcl-2, an inhibitor of apoptosis — leading to the accumulation of reactive oxygen species, subsequent mitochondrial damage, and ultimately apoptosis and cytotoxicity. However, desflurane, another commonly used anesthetic, has not been associated with any of these effects. Isoflurane but not desflurane also provoked learning and memory deficits. "This shows us that potentially we have some good guys and bad guys here," Xie said. Preliminary clinical data is also hinting at a difference between the two anesthetic drugs.

A new series of studies hints at possible mechanisms underlying postoperative delirium and cognitive decline. Mice undergoing anesthesia and surgery show decreased levels of ATP following the procedure, and a drug called cyclosporine A — which inhibits mitochondrial dysfunction — can rescue both ATP levels and cognitive deficits, implicating an energy deficit in these conditions. Mice undergoing anesthesia and surgery also showed age-dependent blood brain barrier dysfunction and olfactory dysfunction, implicating these processes in these conditions.

Electroencephalography and post-operative delirium

One reason that postoperative delirium has been largely ignored until recently is that the condition is challenging to diagnose, said Michael S. Avidan of Washington University School of Medicine in St Louis. Clinicians must distinguish it from states caused by a reduced level of arousal, such as coma, but patients in the ICU frequently experience chemical coma induced by drugs.

Avidan proposed using electroencephalography (EEG) — "essentially...a voltmeter for the brain" — as a tool for diagnosing delirium. In 2014, he and his colleagues published a meta-analysis of four studies showing that when EEG was used to guide the use of anesthesia, patients experienced a decrease in postoperative delirium. One recent study in intensive care patients demonstrated that burst suppression — an EEG pattern characterized by short periods of high electrical activity alternating with no activity — was associated with increased post-coma delirium and mortality. A subsequent prospective cohort trial by Avidan and his colleagues found a similar pattern in surgery: patients with more burst suppression tended to experience more postoperative delirium. Some patients turned out to be more sensitive to anesthesia, showing burst suppression upon receiving lower doses of it than others. Of those anesthesia-sensitive patients, 36% experienced delirium, compared to 17% of non-sensitive patients.

The effects of anesthesia depend on whether the brain is healthy or compromised.

Avidan hypothesized that a healthy brain generally experiences a typical response to anesthesia, with no burst suppression or postoperative delirium, while a somehow-vulnerable brain — perhaps due to Alzheimer's disease pathology — is at risk of burst suppression even with low concentrations of anesthesia. "When people are vulnerable, maybe that's when you need to intervene and prevent the delirium," Avidan said.

He and his colleagues are now conducting a 1200-patient randomized controlled trial called the Electroencephalography Guidance of Anesthesia to Alleviate Geriatric Syndromes (ENGAGES) Study at multiple centers in the US and a parallel multicenter trial in Canada to determine whether EEG can be used to titrate the dose of anesthesia. The trials will also determine whether preventing EEG burst suppression during surgery reduces the incidence of postoperative delirium. Within these trials, the researchers are also exploring the use of EEG during delirium to better understand the condition. "EEG is more variable but less complex during delirium," he said.

EEG studies of brain networks in patients with delirium have suggested that brain connectivity during delirium is decreased. In postoperative patients presenting with delirium, clinicians must rule out that they are experiencing nonconvulsant status epilepticus, which is a prolonged seizure. Continuous EEG could also help clinical assessment of delirium by excluding this diagnosis, Avidan said.

Highlights

Levels of proinflammatory cytokines and the Alzheimer's disease marker tau in the cerebrospinal fluid increase dramatically immediately after surgery.

Amyloid beta status before surgery may be a predictive biomarker of postoperative decline in cognitively normal patients.

A European initiative called BioCog aims to identify biomarkers for postoperative cognitive disorders, especially ones that implicate the cholinergic system and inflammation.

Inflammatory molecules such as interleukin-2 and C-reactive protein may be risk markers for delirium.

Xenon is an anesthetic gas that has neuroprotective qualities and may protect patients against postoperative delirium and its subsequent effects.

Alzheimer's Disease biomarker changes after anesthesia and surgery

Studies performed in vitro and in animal systems suggest that anesthesia and surgical stress may accelerate many of the molecular and cellular pathways implicated in Alzheimer's disease, such as increased amyloid beta, blockade of microglial activation, increased intracellular calcium, neuroinflammation, and tau pathology. Researchers have struggled to demonstrate whether these changes that play important roles in Alzheimer's disease actually happen after anesthesia and surgery in our patients, said Miles Berger of Duke University Medical Center, but if they do, the public health implications would be considerable.

From Berger M et al, J Cardiothorac Vasc Anesth. 2014 Dec;28(6):1609-23

Early studies failed to find a statistically significant association between Alzheimer's disease biomarkers before and after surgery and postoperative cognitive decline, but a 2011 study has shown a link. However, the association may not be causative: perhaps people with early-stage Alzheimer's, many of whom have motor dysfunction, simply fall more often and therefore require surgery more often, he said.

To explore the causality of Alzheimer's disease biomarker changes after surgery, Berger and his colleagues conducted a study called Markers of Alzheimers Disease After Propofol or Isoflurane Anesthesia (MAD-PIA) that measured changes in levels of amyloid beta and tau in cerebrospinal fluid. They found a three-fold increase in tau levels 24 hours after surgery, but no changes in amyloid beta levels. Because the surgeries were neurological, however, it was unclear whether this change was caused by anesthesia, surgery, or both. Also unclear was how long such biomarker changes persist, and what their effects are on patients, and to what extent they are specific to the neurological or ear-nose-throat surgeries the study examined, which require cutting into the brain or the dura. "These tau changes could be very meaningful, or they could be some type of acute phase reaction in an operative setting," he explained.

To investigate the time course and functional effects of these biomarker changes, the researchers are currently conducting another study, Markers of Alzheimers Disease and Cognitive Outcomes After Perioperative Care (MADCO-PC), which will measure changes in amyloid beta and tau immediately after surgery as well as six weeks and one year later, and explore whether changes in these biomarkers correlate with changes in cognitive functioning and neuroimaging. Initial findings suggest a massive increase in proinflammatory cytokine levels in cerebrospinal fluid 24 hours after surgery; interleukin-6, for example, increased sixteen-fold. fMRI data also point to changes in a type of resting state activation called the default mode network that correlate tightly with cognitive changes after anesthesia and surgery, Berger said.

The MADCO-PC study is evaluating postoperative cognition but it lacks the statistical power to determine whether or not a subject has postoperative cognitive decline, Berger said. It also includes a heterogeneous group of patients, and doesn't randomize nonsurgical controls. Still, he said, studies have shown "there is significant neuroinflammation after non-cardiac and non-neurological surgery at 24 hours post-op that resolves by six weeks later."

Beta-amyloid as predictor of postoperative cognitive decline

Whether or not anesthesia can induce Alzheimer's disease pathology in the brains of cognitively normal patients is currently under debate. At the time of death, 30 to 40% of elderly people who are cognitively normal already have amyloid deposition in their brains — a hallmark of Alzheimer's disease pathology.

Marek Brzezinski of the University of California, San Francisco similarly reported that beta-amyloid load, a central feature of Alzheimer’s pathology, did not appear to correlate with the risk of incident delirium.

Biomarker development for postoperative cognitive impairment

Many elderly patients undergoing surgery take medicines that have anticholinergic effects, such as drugs to treat depression, insomnia, and urinary incontinence. These medicines are thought to be risk factors for mild cognitive impairment. These drugs may also raise the risk of postoperative cognitive decline, said Claudia Spies of Charity–University Medicine Berlin. Cholinergic deficiency can interfere with anti-inflammatory pathways, and this sequence has been proposed as a possible pathophysiological mechanism for postoperative cognitive decline, she explained.

However, studies aimed at identifying molecules involved in postoperative cognitive decline, and thus potential biomarkers for the condition, have been conflicting. Indeed, there are no clinical or biomarker-based algorithms for predicting the risk or outcome of this commonly occurring condition. Spies described a large-scale European initiative called BioCog to identify biomarkers and to establish a biobank, which will allow researchers to test the anticholinergic-inflammatory hypothesis, as well as other potential pathophysiological mechanisms for postoperative cognitive disorders.

BioCog, which stands for Biomarker Development for Postoperative Cognitive Impairment in the Elderly, aims to enroll 1200 patients 65 to 80 years of age who are scheduled to have major elective surgery lasting more than 60 minutes. (Patients with neurological or psychiatric disorders, long-term drug or alcohol dependence, and other conditions are excluded from the study.) Patients will receive neuropsychological testing before surgery, upon discharge, at 3 months, and at 1, 2, and 3 years following surgery. Daily delirium screening will be conducted during the week after surgery, and MRI scans will be done at baseline, as well as 3 months and 1 year after surgery.

The BioCog Study will measure cognition among elderly patients for 3 years after surgery.

As of mid-June, 640 patients were enrolled, and 485 had undergone their first MRI. Spies described the tests that will be used for measuring delirium and postoperative cognitive dysfunction; the type of neuroimaging data that will be collected and its possible associations with postoperative delirium and cognitive decline; the molecular biomarkers that will be investigated; and the plan for bioinformatic analysis of the data.

Is inflammation the connection between surgery, postoperative delirium, and dementia?

According to one prevalent hypothesis, neuroinflammation drives postoperative delirium and cognitive dysfunction. Surgery, infection, or another inflammatory insult in the body may cause the release of cytokines, which could in turn cross the blood brain barrier, activate microglia in the brain, and spur neuroinflammation, said Edward R. Marcantonio of the Beth Israel Deaconess Medical Center in Boston. Inflammation that resolves quickly might cause delirium of short duration, but a persistent episode may lead to longer delirium, and if neuronal cell death ensues, perhaps postoperative cognitive dysfunction or even dementia.

Data from the Successful Aging after Elective Surgery (SAGES) study, conducted by Marcantonio, Inouye, and others, have provided some evidence for this hypothesis. The study included 566 adults over age 70 who did not have dementia and were scheduled for major noncardiac surgery. The researchers measured inflammatory biomarkers in blood samples collected preoperatively, immediately after surgery, and two days and one month after surgery, and compared them with matched nonsurgical control subjects. Subjects who experienced delirium had significantly higher levels of the cytokine interleukin-6 on postoperative day two than matched controls, and levels returned to baseline a month after surgery. The cytokine interleukin-2, which is thought to regulate blood–brain barrier permeability, showed a different pattern, remaining elevated at all time points in patients who experienced delirium. The preoperative increase suggests interleukin-2 may be a risk marker for delirium.

IL-6 and IL-2 expression levels correlate with postoperative delirium at different timepoints after surgery.

The researchers also conducted proteomics analysis on blood samples from participants in the SAGES study and identified C-reactive protein, also an inflammatory marker, to be most closely associated with delirium, particularly preoperatively. "There may be some priming, where patients at risk of delirium may have higher inflammation even before the surgery," Marcantonio said.

The traditional view of inflammation holds that peripheral inflammation does not cross into the central nervous system, but more recent research contradicts this idea. A meta-analysis of 51 studies in which inflammation was induced in rodents found that peripheral inflammation reduced cognitive performance, and those studies that examined the central nervous system found molecular evidence for neuroinflammation. A postmortem human study of 15 patients found microglial activation and elevated interleukin-6 levels in the 9 patients who experienced postoperative delirium.

Marcantonio proposed several studies that would move the field forward. First, a database containing information on demographic and clinical factors, types of surgery and anesthesia, postoperative course, complications, cognitive and functional factors, and biomarkers would help develop predictive models for patients at risk for delirium and cognitive decline. Second, the field needs better markers of neuroinflammation — preferably blood-based ones. Conducting in vivo biomarker studies is challenging, partly because collecting cerebrospinal fluid during surgery or delirium is difficult. Finally, intervention studies that of potential inflammation-based targets could help identify treatments for these conditions.

Xenon for protection from acute neurological injury

Mervyn Maze from the University of California, San Francisco, described the neuroprotective and potentially therapeutic properties of the volatile anesthetic xenon. As a so-called noble gas, xenon is chemically inert, but biologically active; used as an anesthetic since 1946, it is now marketed in 14 countries for its anesthetic properties. Even though it is not as potent as other volatile anesthetics, it is relatively insoluble and therefore induces anesthesia very rapidly.

Xenon works by blocking the NMDA glutamate receptor, and in doing so, interferes with excitotoxicity, which damages neurons. The gas also interferes with apoptosis, both by inhibiting pro-apoptotic factors and upregulating anti-apoptotic factors, and increases the expression of mTOR, an enzyme involved in cell growth regulation. Additionally, xenon dramatically boosts erythropoietin levels, which could enhance red blood cell production. Multiple studies in cell culture and different animal models demonstrate xenon's protective effects. These effects are particularly prominent when combined with hypothermia, with which the gas has a synergistic effect.

Maze and his colleagues demonstrated xenon's neuroprotective potential in neonatal piglets experiencing ischemia and in adult pigs after cardiac arrest. Based on these results, the researchers tested xenon in 110 patients who experienced cardiac arrest, were successfully resuscitated, but remained comatose. They compared standard of care after cardiac arrest — at this medical center, hypothermia to 32–34 degrees Celsius — with hypothermia plus xenon, and found significantly less white matter injury and demyelination in the xenon group. The study included patients whose cardiac function had been restored within 45 minutes, and although xenon treatment did not show a survival advantage in the entire group, it significantly improved survival in the subset whose cardiac function had been restored within 30 minutes.

Xenon improved brain injury in patients experiencing coma after cardiac arrest.

Xenon has a good pharmacokinetic profile and has several features that suggest it may interrupt the progress of acute neurologic injury, Maze said. The downside, however, is that it is expensive, difficult to produce, and requires complex equipment for delivery. Maze and his colleagues have conducted a clinical trial of the substance for preventing postoperative cognitive decline. The study compares xenon to another general anesthetic in hip fracture surgery in 230 elderly patients, and will be published soon.

Highlights

Developing a set of standardized definitions, measures, and diagnostic tests for postoperative delirium and cognitive decline is key for moving the field forward.

More research is also needed on whether neuroinflammation could be involved in the development of postoperative cognitive problems.

Drugs approved for other indications, such as cholinergic compounds or NMDA receptor antagonists, should be tested for their ability to prevent delirium.

Identifying those most vulnerable

Howard Fillit of The Alzheimer's Drug Discovery Foundation opened the discussion by noting that there have been very few clinical trials of interventions for postoperative delirium, and asking speakers what such a trial would ideally look like — when it would be administered, what outcomes would be used, and when follow-up would occur. Inouye noted that to prevent postoperative cognitive decline she would start with an intervention that prevents delirium, preferably by interfering with neuroinflammation. One might begin by targeting high-risk patients — that is, people who are on the trajectory for underlying cognitive impairment in some way, perhaps identified by EEG or fMRI.

Maze agreed that the ability to identify an at-risk group is crucial for translational research. He believes neuroinflammation is the culprit, although it remains an open question how exactly neuroinflammation affects cognition. Maze noted that people with a dysregulated neuroinflammatory response are the ones to target for prevention and treatment strategies. It will be important to identify anyone who has inflammation going in to surgery, but probably more important will be to measure a prospective patient's innate immune system and understand pre-operatively how they will respond to a surgical insult.

Spies agreed that researchers should target patients who show a preoperative immune pathology in a trial that aims to prevent delirium and its sequelae. These patients are likely different from those that undergo elective surgery for something like hip or knee replacement, so the populations shouldn't be mixed in a study, or they should be stratified into separate groups.

Berger noted that measuring risk with fMRI, PET scanning for amyloid beta, or even EEG would almost involve conducting a study within a study, because these measures are not done within routine care. A practical challenge may be the ability to quickly analyze such data. Berger also noted that while neuroinflammation currently appears to be the most promising target, previous studies looking at anti-neuro-inflammatory compounds have not shown an association with postoperative delirium or cognitive dysfunction. It will be important to identify specifically which inflammatory pathways are involved and to selectively target them, he said.

Fillit asked whether a drug study might want to focus on a particular type of surgery. Marcantonio noted that homogeneity in surgery type would be ideal; to understand a population at risk, many patients will have to be screened, but for a small, proof-of-concept study, homogeneity could be achieved. Regardless, he said, such studies should make sure to use state-of-the-art highly sensitive delirium measures that can also determine delirium severity. He offered to consult with researchers launching studies to consult on existing measures.

Spies raised several other study design issues. First, how and how often to measure delirium? Second, how to account for pre-existing factors such as diastolic dysfunction or metabolic syndrome that can boost the risk profile of surgery, and how to account for the fact that patients are often anesthetized too deeply, also raising the risks? These factors must be very carefully controlled for, she said.

One realization he took from the meeting, Brzezinski said, was that "everyone is doing something different" in terms of diagnostic tests, measures, and definitions. Regardless of the study, researchers must begin collecting some basic, commonly agreed-on data, and make them freely available to the community. Having such a database that grows over time will allow researchers to devise hypothesis-generating studies.

Berger agreed, adding that he and a group of researchers are working on standardized nomenclature for postoperative cognitive decline that should come out later this year. Journals should also be encouraged to insist on the creation and the use of clinically defined definitions.

Fillit raised the question of the type of study design that could be used to determine which anesthetics are most neuroprotective. The field needs to conduct a comparative effectiveness study of the anesthetics used in surgery, Xie said. Fillit noted that this could be accomplished using already-known biomarkers, and its results would allow clinicians to provide patients with some concrete answers about factors that might affect their cognitive outcome after surgery. Spies said that the BioCog project will conduct risk profiling, but that depth rather than type of anesthetic drug is likely the key factor. She added that researchers must make sure to measure delirium in all patients in the recovery room. Later, too many factors can mask its occurrence or intensity.

Speakers noted that the problem of postoperative delirium and cognitive decline is multifactorial, with no single-drug solution. Rather, researchers will have to stratify risk, identify biomarkers, look at anesthesia strategies, and determine how to measure outcomes. Although the task will be formidable, it is very much achievable, Inouye said.

Fillit raised the possibility of repurposing drugs from other fields. There's increasing interest in the cholinergic hypothesis, and the study of cholinergic drugs in the prevention and treatment of postoperative cognitive problems; perhaps NMDA receptor antagonists such as xenon can also play a role. Marcantonio replied that he conducted a pilot trial some years ago of donepezil (Aricept), an Alzheimer's drug that increases cholinergic activity, for treatment of delirium, but it didn't work; he added that European trials with another substance also didn't work and in fact had to be stopped early due to safety concerns. Spies noted that there were complicated study design issues at play, and repeating such trials with drugs that more specifically target the brain rather than also targeting other tissues in the body would be worthwhile. Maze said that he and colleagues just conducted a study using dexmedetomidine (Precedex), an analgesic and sedative, and they found it prevented delirium in a surgical population.

Inouye suggested making a list of 10–15 promising drugs, so that researchers could determine their safety in a presurgical population as well as their impact on inflammatory biomarkers. Fillit proposed forming a study group to do so. There have been so few clinical trials that the first few will probably fail but, he said, "we'll learn a lot."

Do postoperative delirium and cognitive decline cause long-term dementia, or do they merely unmask a vulnerability already present in the brain?

Is there a pathway separate from Alzheimer's disease pathology that drives the onset of dementia after postoperative delirium?

How does anesthesia affect brain activity and consciousness?

How exactly does neuroinflammation alter cognition, and which specific inflammatory pathways are involved?

Do different anesthetics carry different risks of postoperative delirium?

Can continual EEG monitoring help diagnose delirium?

Is there a definite causal link between Alzheimer's disease pathology and perioperative care?

What makes a brain vulnerable to postoperative delirium and subsequent long-term cognitive decline?

Can xenon or other neuroprotective factors prevent postoperative delirium and cognitive decline?