Nomenclature

The ISPOCD group coined the term postoperative cognitive dysfunction (POCD), which reflected an objectively measured decline in cognitive function that typically persists beyond the period expected for normal recovery from the physiological and pharmacological effects of anesthesia and surgery. When patients presenting for anesthesia and surgery are identified with cognitive impairment at baseline, this has been referred to as preexisting cognitive impairment (PreCI).

Postoperative delirium (POD) refers to a form of acute cognitive disruption characterized by inattention, a fluctuating course, and cognitive disturbance, and is diagnosed according to the Diagnostic and Statistical Manual for Mental Disorders, version 5 (DSM-5). Delirium in the community conforms to the same definitions and criteria. In contrast to POD, which conforms to the criteria for delirium diagnosed in any situation, POCD and PreCI have been confined to the field of perioperative medicine research where they have been historically defined by objective criteria with no attention given to subjective or functional criteria. In contrast, cognitive impairment and decline diagnosed in the general community conforms to either DSM-5 definitions and criteria, and/or the National Institute of Aging-Alzheimer’s Association (NIA-AA) definitions, all of which require a subjective component and an assessment of activities of daily living (ADL). The NIA-AA terms Mild Cognitive Impairment (MCI and dementia ) are more familiar than the DSM-5 terms (mild and major NCD), but the definitions and criteria roughly map onto each other.

The NIA-AA nomenclature is more granular, making provisions for the inclusion of biomarkers (biochemical and imaging). This is currently useful for research purposes, but is likely to apply to clinical scenarios in the future. In addition to the subjective complaint and functional criteria, another important difference between DSM-5/NIA-AA and POCD/PreCI has been the objective criteria applied. Although variable, many POCD/PreCI studies require a decline of 1.96SD below controls on two or more tests from a battery of 8-10 neuropsychological tests. Mild NCD and major NCD require 1-2SD and ≥ 2SD, respectively, below controls/norms in only one cognitive domain.

These differences, combined with the variability of POCD criteria (and timing) prompted an international multidisciplinary group to consider a new nomenclature for POCD. Not only would diagnostic standardization facilitate further research in the area, but it would allow effective communication between clinicians at a clinical level. This new nomenclature recommends “perioperative neurocognitive disorders” (PND) as an overarching term for cognitive impairment or change, including delirium, identified in the perioperative period. Constructs previously used to define cognitive change associated with the perioperative period are discussed below with the recommended new nomenclature ( Table 82.1 ).

Measurement and Diagnosis of Perioperative Neurocognitive Disorders

Historically, a diagnosis of POCD has relied only on an objective assessment of decline using a battery of neuropsychological tests that cover many cognitive domains (e.g., executive function, memory, attention, visuospatial, psychomotor, and language). In most studies, very conservative cut-points were applied (e.g., 1.96 standard deviations below controls on two or more tests out of a battery of 8-10). These research applications contrast sharply with the simple objective criteria required for mild NCD (postoperative) or major NCD (postoperative) as described above, and are described in detail elsewhere.

Cognitive Decline Outcome Criteria

There are several criteria that studies have used to define POCD in the past. These include the 1SD rule, the 20% rule, and the reliable change index (RCI). Each has advantages and limitations, but the latter has the large advantage of incorporating the changes observed in a control group over a similar time period, largely to account for practice and time effects. Because the RCI can be related to a control group or normative data, it satisfies the DSM-5 criteria for NCD (postoperative).

Sensitivity

The number of tests directly impacts the sensitivity of the measure; for example, if the definition of POCD is a decline of 2SD in 2 tests from a battery of 8 tests, the probability of identifying a decline of 2SD in 2 tests from 10 tests is doubled to 0.10. For the assessment of NCD, the DSM-5 only requires decline in one or more cognitive domains (complex attention, executive function, learning and memory, language, perceptual-motor, or social cognition), but does not recommend specific neuropsychological tests or a specific number of tests. Sensitivity can be improved by considering small declines across a number of tests, or a single very large decline in what is termed, “combined z-scores.”

While seemingly esoteric, the various methods of analysis can markedly impact the diagnosis of POCD/PND, highlighted by Keizer et al., who reported a POCD incidence of 10.5%, 31%, and 7.7% using the same test results with different methods of analysis (1SD definition, 20% definition and RCI, respectively). Clearly, standardization will be required to interpret results across many studies.

More recently, computerized cognitive assessment batteries have been developed that avoid practice effects, offer easier administration, are quicker and standardized, and have the potential to overcome cultural and language difficulties. To date, however, computerized test batteries have received limited attention in POCD/PND research mainly due to a lack of relevant validation studies.

group change versus individual change

Studies investigating POCD have variously used either individual or group change. Individual change refers to a dichotomous outcome of “decline” versus “no decline,” whereas group analysis considers differences between groups on a continuous scale. Clinical studies typically then use statistics to compare an a priori primary outcome between two groups to test whether any observed difference is due to chance. But this approach misses important individual information within each group. In particular, the individuals who decline the most may be a small fraction of the group, but arguably are the most critical to consider. Some individuals may even demonstrate cognitive improvement, due to correction of underlying pathology and improved function (e.g., reduction in pain, improved mobility, improved ADL). Statistics that simply average these groups together will miss these important individual changes ( Fig. 82.1 ), which might be due to identifiable and perhaps correctable factors.

Perioperative cognitive trajectories.

Perioperative patients come into surgery having either a stable (black) or declining trajectory of cognition (red) . After surgery, the majority are unchanged (black) , with a small fraction showing postoperative cognitive improvement ( POCI , green ). Some postoperative patients have an acute decline into POD or dNCR, from which most recover (green) . A fraction of these will decline again later, perhaps matching their preoperative trajectory. The bottom red trajectory is meant to indicate that a very small fraction of perioperative patients never fully recover, and assume a steeper downward trajectory than if they had not had surgery. Thickness of line is intended as a rough reflection of probability of following the indicated trajectory. Other trajectories are possible. dNCR , Delayed neurocognitive recovery; NCD , neurocognitive disorder; POD , postoperative delirium.

Risk Factors

Multiple studies have investigated risk factors for the development of postoperative NCDs, with the most commonly cited being advanced age, a history of PreCI, and type of surgical procedure. Other factors include a prior history of delirium, frailty, psychotropic medications, ASA physical status, number of medications, impairments in IADLs or ADL, and smoking. PND is the likely combination of patient vulnerability along with the risks of the surgical procedure and its related complications that enhance the risk for the development of POD and potentially POCD. While many of these risk factors such as age and a prior history of delirium are not modifiable, there is growing interest in identifying modifiable risk factors and the use of multidisciplinary teams combined with prehabilitation to decrease the risk of developing PND. Moreover, there has been growing interest in whether genetic risk factors, specifically the apolipoprotein-Ee4 genotype, are associated with the development of either POD or PND. Studies performed to date suggest that the presence of apolipoprotein-Ee4 genotype is not associated with the development of POD, but the data are less clear for other forms of PND, with some studies suggesting that it is, or is not, a predictor for the development of both dNCR and NCD (postoperatively).

Informed Consent

It is currently uncommon that older adults are informed of their risk of PND in the consent process, despite clear evidence that PND is more common than most complications of which patients are routinely informed. However, informed consent is difficult to obtain in older operative patients because of a high prevalence of unrecognized MCI in this population. Among community-dwelling elders, as many as 70% may have some degree of cognitive impairment, although the prevalence appears to be lower for those presenting for elective surgical procedures. Preexisting MCI confounds their ability to understand the complex anesthesia and surgical procedures and risks, but also their risk for the development of subsequent NCDs. One must also remain cognizant that there may be substantial risks of forfeiting a surgical procedure that may be of benefit. While it is beyond the scope of this discussion to describe the entirety of the informed consent process, at a minimum, all older patients consenting for anesthesia care should clearly understand and actively participate in discussions about their risks and be capable of understanding the information provided about the planned management and potential adverse outcomes, including the risks of postoperative NCDs.

Preoperative Management

Older patients often have geriatric syndromes that are not considered to be a part of the routine preoperative evaluation, yet are associated with an increased risk for the development of PND. For example, the risk of developing delirium is significantly increased in patients with preoperative cognitive impairment, frailty, functional impairment, depression, and among those taking certain psychotropic medications. Accordingly, the American College of Surgeons together with the American Geriatrics Society have developed guidelines for the perioperative evaluation of older surgical patients that are relevant to the development of PND, including an assessment of preoperative cognitive performance, depression, functional status, frailty, and a review of prescription and over-the-counter medications, in order to identify those associated with a risk of PND. In the presence of these risk factors, the perioperative physician should consider referral to either a primary care physician, geriatrician, or mental health specialist for optimization or prehabilitation prior to the surgical procedure. Identification of patients at highest risk for the development of PND may enhance patient-centered outcomes by providing patients and their families with a better understanding of their perioperative course and allow practitioners an opportunity to deploy resources to those at highest risk.

Type of Anesthesia

Researchers and clinicians have widely debated whether regional anesthesia is preferable to general anesthesia in suitable surgical cases for the prevention of PND. For all forms of PND, including POD, it is intuitive that regional anesthesia will be preferable, as general anesthesia targets the CNS, and data now suggest that deep anesthesia is associated with a higher incidence of NCDs. However, most studies have been unable to show differences in the risk of either POCD or POD following surgery under either general or regional anesthesia. There are a number of potential reasons for this lack of difference. One is that regional anesthesia is often accompanied by sedation at levels comparable with general anesthesia based on processed electroencephalogram monitoring. However, even in the few studies that have limited or randomized sedation during regional anesthesia, little difference was found. This is the basis for the current hypothesis that most forms of PND, including POD, are the result of the surgery itself, combined with preexisting vulnerabilities.

Dexmedetomidine

Studies have demonstrated that dexmedetomidine, as compared to benzodiazepines and/or propofol for sedation, decreased the incidence of delirium in the intensive care unit. This observation has led to investigations of dexmedetomidine as an intraoperative adjunct to either regional or general anesthesia in older patients. The majority of these studies have found that perioperative dexmedetomidine was associated with a lower prevalence of PND as compared to propofol in both cardiac and noncardiac surgery patients.

However, it remains unclear whether the reduction in PND is due to the pharmacologic actions of dexmedetomidine itself or simply less brain depression. There is a growing body of evidence that deep sedation or general anesthesia based upon EEG monitoring during a regional anesthetic is associated with a higher prevalence of PND as compared to lighter levels of sedation or general anesthesia. In other words, it may be the state of the brain (and duration) that is responsible for a lower risk of PND, rather than any specific effect of the drugs themselves.

Ketamine

Ketamine is rarely used as a sole anesthetic, but it is frequently administered intraoperatively to reduce postoperative pain. Most studies performed to date have not shown that intraoperative administration of ketamine, despite its ability to decrease postoperative opioid requirements, decreases the incidence of POD and some studies suggest that it may increase it. In addition, at least one study noted that intraoperative administration of ketamine may be associated with postoperative hallucinations and nightmares.

Intraoperative Brain Monitoring

There has been a growing interest in whether intraoperative management based on processed EEG or cerebral oxygen monitoring decreases the risk of either PND or POD. There is evidence to suggest that the use of the processed EEG to guide anesthetic management may reduce the risk of both, although the evidence is greater for POD. Questions remain as to the mechanism, whether PND is a direct result of the anesthetic or its dosage, or the development of EEG burst suppression patterns. Evidence indirectly supports the latter, as the dosage of anesthetic being administered to the patients has not been rigorously correlated with the development of PND. As there is little risk of EEG monitoring in most surgical procedures, some have recommended that EEG-based management be used to reduce the risk of PND in older surgical patients.

The evidence is less clear for the use of intraoperative cerebral oximetry to reduce the risk of PND. While some studies have suggested that higher perioperative cerebral oxygen saturation is associated with a lower risk of POD, therapeutic restoration of regional cerebral oxygen desaturation has not been shown to lower POD risk. Overall, the evidence for the use of cerebral oximetry for the prevention of POCD is greater than that for POD, although the majority of studies performed to date have been in the setting of cardiac surgery. Taken together, the evidence makes it difficult to make a recommendation regarding the measurement of cerebral oxygenation during routine surgical procedures.

Blood Pressure Control

There is currently significant interest in whether hypotension, hypertension, or blood pressure variability are associated with the development of PND, but the data are often difficult to interpret. Early studies that randomized patients to either high or low blood pressure management failed to find a difference in cognitive outcomes. However, subsequent observational studies have suggested that intraoperative hypotension, hypertension, and blood pressure variability were each associated with the development of PND, as were vasopressor administration and postoperative hypertension. While there is no clear evidence for what constitutes optimal blood pressure management in older surgical patients, maintenance of normal blood pressures without significant variability may decrease the risk of developing PND.

Pain Management

The national opioid crisis has focused anesthesiologists’ attention to the use of opioid-sparing techniques in the perioperative period. This is particularly relevant for older surgical patients, in that both postoperative pain and the administration of opioids have been associated with the development of POD. However, the outcomes from opioid-sparing pain management studies are often mixed. For example, while regional analgesic techniques to manage postoperative pain are associated with decreased opioid consumption, their role for preventing the development of POD is less clear.

As described above, perioperative administration of dexmedetomidine has been shown to decrease the development of POD in the majority of studies. Contributing to this action may be that the α-2 adrenergic receptor agonist analgesic effects of dexmedetomidine decrease opioid requirements. Interestingly, there exists little evidence on whether clonidine, another α-2 agonist, is associated with a reduction in POD risk in older surgical patients. Similarly, early studies suggested that the perioperative use of gabapentin reduces the development of POD, but subsequent investigations have demonstrated that, despite decreasing opioid consumption, gabapentin administration did not reduce the development of POD, and may increase respiratory side effects. Acetaminophen and COX2 inhibitors have also been suggested as a part of multimodal pain management strategy to reduce POD. While acetaminophen administration has been associated with a reduction in postoperative opioid utilization, it has not been associated with a reduction in POD. In contrast, COX2 inhibitors have been shown to decrease pain severity, opioid utilization, and POD in patients greater than 60 years of age undergoing lower extremity joint replacement surgery. This provides hope that multimodal management of postoperative pain in elders will be associated with a decreased risk of POD, yet may be associated with increases in the costs of pain management.

At this point, the effective strategy to reduce PND in older patients is nonpharmacologic—for example, assuring good sleep and nutritional hygiene, rapid mobilization, and early orientation to familiar aspects of their environment, such as family members. Older patients may need their glasses and hearing aids to facilitate orientation. It is also important to remove devices such as Foley catheters to avoid urinary tract infections, to avoid Beers’ criteria medications, and to promote normal bowel function. One cannot know the success of these interventions without testing for POD using validated instruments like the CAM. Without formal screening, many cases of delirium, especially the hypoactive type, go undetected.

If a patient is found to be delirious, the cause may be identifiable. For example, delirium might be due to unrecognized hypoxia, pneumonia, urinary tract infection, electrolyte abnormalities, urinary retention, fecal impaction, acute renal failure, hypoglycemia, and dysrhythmias. The medication record may reveal drugs associated with the development of delirium, which should be discontinued, if possible. Nonpharmacologic interventions might include a calm environment and reorientation tools such as the presence of a clock, calendar, family members and familiar items, and the removal of restraint devices. Pharmacologic interventions should be reserved for those who pose a threat to themselves or others, as these drugs may mask, rather than treat, delirium. Effective drugs include dexmedetomidine and haloperidol, but the adage of “start low and go slow” to avoid exacerbating the duration or severity of delirium should be adopted.

For patients who develop PND of any form, it is important to refer them to their primary care provider, geriatrician, or mental health expert, as evidence is beginning to suggest that POD and PND are predictors of subsequent cognitive decline.

Amyloidopathy

The involvement of a small proteolytic fragment of the membrane protein amyloid precursor protein (APP) in neurodegeneration was strongly implicated from both pathological as well as genetic studies. The pathognomonic feature of Alzheimer disease (AD), the senile plaque, contains considerable quantities of this peptide in a characteristic aggregated form, but data now suggest that plaques are a less toxic, sequestered form of the peptide; the neurotoxic forms are thought to be the more labile small (∼dodecamers) oligomers of the amyloid β peptide. These oligomers are produced at various rates over a period of decades prior to the onset of symptoms ( Fig. 82.2 ), and are variously eliminated or sequestered into plaque, making a causal association with symptoms exceedingly difficult. In fact, the mechanism by which such small aggregates produce cytotoxicity is not at all clear, but may involve a detergent-like action whereby an amphiphilic peptide inserts into, and disrupts, cell membranes. Despite being a pathognomonic feature of AD, imaging and cerebrospinal fluid (CSF) studies (discussed later) have demonstrated a poor relationship between amyloid β accumulation and cognitive loss.

A model of the progression of neurodegeneration and cognition.

Cerebrospinal fluid (CSF) Aβ42 (purple) and deposition of amyloid on PET imaging (red) occur early and are then followed by elevations in CSF tau (blue) and probably tau aggregates on PET imaging (data not yet available). Neurodegeneration as measured by FDG PET and structural magnetic resonance imaging (MRI, yellow) , respectively, follow and precede functional changes such as Mild Cognitive Impairment (MCI) or dementia. By definition, all curves converge at the top right-hand corner, the point of maximum abnormality (not levels). The horizontal axis is time, which although highly variable, is on the order of decades. A vertical line drawn between normal and MCI approximates a large fraction of older patients presenting for surgery, which implies that many of our patients will have significant neuropathology at the time of their operation.

From Jack C, Knopman, DS, Jagust WJ, et al. Tracking pathophysiological processes in Alzheimer’s disease: an updated hypothetical model of dynamic biomarkers. Lancet Neurol. 2013;12:207–216.

Cell and Molecular Studies

Initial studies found that inhalational anesthetics like halothane enhanced the aggregation of amyloid β in the test tube, and that when combined with cells in culture, also enhanced the cytotoxicity of exogenously added amyloid β peptide. Further studies found that the activity of the enzyme responsible for cleaving and releasing the amyloid β peptide, the β-acting cleavage enzyme, was enhanced by anesthetics like isoflurane, so that amyloid β production was also increased. So even without exogenous amyloid β, this pathway was capable of causing apoptosis in cell culture systems. Moreover, isoflurane enhances the cytotoxicity of presenilin-1 mutations in cultured neurons. Presenilin-1 is an important regulator of amyloid β production through effects on the γ-secretase complex, and the mutation in question upregulates this complex to amplify the production of amyloid β. Thus, it is clear that certain anesthetics intersect the amyloidopathy pathways at several points that result in exaggerated cytotoxicity.

There appeared to be variation in the degree that different anesthetics caused these effects. Halothane and isoflurane both caused substantial amyloid β production, aggregation, and cytotoxicity in cell culture. On the other hand, desflurane caused less amyloid β production and apoptosis than isoflurane. Perhaps most importantly, propofol at clinical concentrations (∼1 uM) actually inhibited amyloid aggregation in vitro, and was without cytotoxicity in cultured cells. Such rank-order effects are extremely important means of translating hypotheses to intact animals.

Animal Studies

In early PND/POCD research, it was found that wild-type rats had impaired learning and memory after isoflurane anesthesia, and this was most apparent in older animals. In addition to age, other vulnerabilities were introduced, such as human transgenes for AD, so that the amyloid mechanism could be explored in vivo. The Tg2576 animal contains the APP Swedish mutation that overproduces amyloid β and acquires cognitive deficits by 10 to 12 months of age. Several 2-hour exposures of aged animals to either halothane or isoflurane caused no further deficits in the already impaired learning or memory, but significantly increased the number and density of amyloid plaques in their brains. Other rodent work, even in wild-type animals, has confirmed that anesthetic exposure (largely isoflurane) increases amyloid β levels in the brain. Rank-order effects appear to be mostly retained, as halothane exposure increases amyloid plaque more than isoflurane, and propofol had little effect, although this was in a different transgenic (3xTgAD) animal.

Human Studies

Amyloid β can be assayed in humans either postmortem or in the living by ELISA analysis of CSF, or PET imaging of amyloid β plaque. Low amyloid β in the CSF is consistent with AD pathology because the peptide is thought to be sequestered into plaque and less available for distribution into CSF. In addition, it might also reflect fewer active synapses, as the APP tends to be highly expressed in synaptic regions. Consistent with this, PET imaging that shows high density of amyloid-ligand binding is consistent with AD. Studies have examined amyloid β preoperatively as a risk biomarker, or postoperatively as a marker of PND occurrence and progression. In the former case, it is relatively straightforward to get a sample of CSF when a spinal anesthetic is initiated. Analysis, however, is nontrivial, and should be conducted by standardized laboratories specifically established for this purpose. In one study, low CSF amyloid β predicted dNCR at 3 months postoperatively. PET imaging with any of several amyloid β ligands can also be performed preoperatively for risk assessment. In a recent study of cardiac surgical patients, amyloid imaging was performed at 6 weeks and 1 year postsurgery. At neither time point was amyloid load associated with PND, but it was noted that amyloid load increased more between the two time points than in matched nonsurgical controls. This result is directly analogous to the above animal study in that the increased amyloidopathy was not reflected in worse cognitive outcomes. The reasons are suggested by the mixed role that amyloidopathy represents, both reflecting the presence of, and the sequestering, of toxic oligomers, as well as the delay between this particular proteinopathy and neurodegeneration.

In addition to its use in risk prediction, the use of amyloid β as a marker of disease progression has been studied by several investigators, but in general for very limited periods (days). For example, in patients with lumbar drains in place for the operation and up to 2 days after surgery, repetitive sampling of CSF showed no significant change in the amyloid β. An uncontrolled study in cardiac patients showed a significant decrease in CSF amyloid β 6 months after surgery, consistent with progressive amyloidopathy. Finally, one study has shown that CSF amyloid β increases acutely after anesthesia and surgery, which may be consistent with prior data showing anesthetic-induced increases in amyloid β production.

Amyloidopathy Summary

It is now clear that anesthetics interact directly with the amyloid β pathways, and in a differential manner. The haloalkanes, exemplified by halothane, are the most provocative, predicting acceleration, whereas propofol has little effect. The critical in vivo data confirms these in vitro studies, but they also support the growing recognition that amyloidopathy is itself not temporally associated with cognitive ability, suggesting that amyloidopathy might not underlie the most prevalent form of PND, dNCR. It may, however, contribute to later forms of PND, such as mild or major NCD (postoperative), but the separation in time makes this association difficult to establish. Amyloid β will continue to be useful as a risk biomarker, especially if amyloid imaging or CSF is available preoperatively. Assays of amyloid β in blood have not reached the stage where they are considered sensitive and specific, but this is obviously a highly desired goal.

Tauopathy

The other major pathognomonic feature of AD (as well as various tauopathies) is the intracellular neurofibrillary tangle (NFT), now known to consist largely of tau protein. Tau is a microtubule-associated protein (MAP) that regulates microtubule stability and dynamics, both of which are critical for neuronal structure and intracellular transport. Phosphorylation of tau at several sites causes it to detach from the microtubule, which if extensive, ultimately results in loss of microtubular function. Like amyloid, high concentrations of phosphorylated tau will begin to self-assemble into ordered fibrillary structures, which either directly cause further cell stress, or serve to isolate tau from the microtubule. In either case, both neurodegeneration and cognitive loss relate better to NFT prevalence (both by imaging and CSF levels) than does amyloid β, as it is thought to be a later event reflecting cell demise (see Fig. 82.2 ). Release of tau from damaged cells is thought to enhance tauopathy in neighboring cells in a prion-like manner. Release also deposits tau into the CSF and ultimately bloodstream, making it a useful biomarker of CNS stress/injury.

Cell and Molecular Studies

The effect of isoflurane has been evaluated in a few cell culture studies, and in each case, isoflurane (∼2%) caused tau hyperphosphorylation, perhaps by upregulation of the associated kinase, GSK-3b. Similarly, both propofol and dexmedetomidine caused tau phosphorylation in a neuronal cell line. The mechanism appears to center around inhibition of the associated phosphatase, PP2A, either directly by hypothermia or direct anesthetic binding, or through inhibition of the PP2A-tau complex.

Animal Studies

Early work showed that wild-type mice exposed to isoflurane had a dramatic increase in tau phosphorylation in their brains. However, it was found that this was largely due to the extreme temperature sensitivity of the PP2A phosphatase, and the fact that the animals had been allowed to cool during the anesthetic. Subsequent work has determined that a degree of excessive tau phosphorylation exists for several days after the anesthetic, even when animal temperature is carefully maintained. A transgenic mouse that incorporates human genes associated with both amyloid and tau pathology also found that abdominal surgery under volatile anesthesia (desflurane) increased tau deposits, and that this was associated with learning and behavioral defects. Interestingly, neither enhanced tau nor cognitive deficits were observed if propofol was used for the brief surgical procedure.

Human Studies

Clinical studies to date have been limited to assays of tau in CSF before and after surgery of various types. As stated previously for amyloidopathy, these studies often take advantage of the need for a subarachnoid catheter for the surgical procedure itself, but in some cases, repeated single CSF draws have been used. Results have been remarkably consistent. Total tau, but not phosphotau, is elevated after surgery, and it continues to rise for at least a day or two, when most of the sampling is terminated. Presumably, this elevation of tau reflects CNS cellular damage, but the implications in terms of future cognition are not yet clear. It is important to note that PET ligands with affinity for NFT tau are now available, yet transsurgical studies have yet to appear in the literature. Sensitive and specific blood assays for tau and other injury biomarkers (neurofilament light, S100β) are now available, and a recent study shows elevations after surgery. While far more convenient than CSF, blood is likely to be contaminated by peripheral tissue sources that are almost invariably damaged during surgery.

Tauopathy Summary

The MAP tau is gaining much attention as a more proximal biomarker of CNS injury and neurodegeneration than amyloid β. The effect of anesthesia and surgery on tau is remarkably consistent across species and models, and suggests a form of CNS stress and injury in the immediate postoperative period. The upstream mechanism may reside in surgery-induced neuroinflammation (discussed later), but animal studies suggest that direct effects on tau phosphorylation may also contribute. Postoperative tau imaging studies hold substantial promise for defining the magnitude and time course of any surgery-induced CNS injury.

Calcium Dysregulation

Intracellular calcium is a critical cell function regulator, which, if excessively elevated, also leads to cell stress, apoptosis, autophagy, and necrosis. Calcium dysregulation has therefore been considered an important link between amyloidopathy/tauopathy and cell death, and has recently been proposed as a target for mitigation of neurodegeneration in disorders like AD, although no trials are underway. The primary source of elevated calcium in neurons is thought to be the endoplasmic reticulum, released via the two major channels, the ryanodine receptor (RyR) and the inositol 1,4,5-trisphosphate receptor (InsP3R). An antagonist of RyR, dantrolene, is available and well known to anesthesiologists, but has two principle drawbacks. First, it is distributed only poorly to the brain with the usual parenteral administration, and second, it has the significant side effect of causing weakness due to effects on skeletal muscle. Dantrolene has been proposed as a treatment for AD.

Cell and Molecular Studies

Electrophysiology and cell culture studies have shown that the volatile anesthetics enhance the opening of both the RyR and InsP3R, resulting in concentrations of intracellular calcium that activate apoptosis pathways. These effects were not apparent in cell lines devoid of the InsP ₃ R, and were completely mitigated in normal cells treated with either dantrolene or an InsP ₃ R antagonist, xestospongin-c. It is important to note that fairly high concentrations and long durations of volatile anesthetic exposure were necessary to produce these isolated cell effects, arguing against the in vivo relevance. Also, even higher concentrations of propofol were required to induce cytotoxic calcium levels, suggesting that potentially useful rank-order effects exist among the general anesthetics.

Animal and Human Studies

There are only a few rodent studies that have demonstrated modest efficacy of chronic dantrolene in slowing cognitive decline and neurodegeneration, but none to date on the efficacy of acute dantrolene in preventing POD or losses in memory or cognition. Perioperative human studies have not yet been reported.

Calcium Summary

Calcium dysregulation is probably an important upstream event in anesthesia and surgery-induced cognitive decline, and yet has remained largely unstudied. While chronic dantrolene administration to slow the progression of AD might be considered implausible, the potential for acute, intraoperative use is attractive, especially given that the various formulations of dantrolene are already clinically approved, and that muscle relaxation during surgery is typically desirable.

Neuroinflammation

The consistent observation in animals that anesthetics alone induce only modest effects on pathology and cognition, and that more robust and consistent effects are produced in both when a surgical procedure is added, have led to the hypothesis that surgery-induced inflammation due to activation of the innate immune system is the principle contributor to PND. It is well established that surgery induces a classical systemic inflammatory response, triggered by damage-associated molecular pattern molecules, and propagated systemically by a variety of cytokines and chemokines. This inflammatory response is essential for wound healing and microbial defense, but can also exaggerate ongoing inflammatory processes in some individuals. In general, an intact blood-brain barrier (BBB) isolates the brain from significant inflammatory mediators and cells. But the BBB can become “leaky” in the elderly, and in those with preexisting neuroinflammation, so that the peripheral innate immune response can be more effectively transduced into the brain. This is the probable reason that age, preexisting cognitive decline, cardiovascular disease, diabetes, and frailty are important risk factors for PND.

Cell and Molecular Studies

It is not possible to simulate surgery in cell culture systems, but attempts at pro-inflammatory influences have been combined with anesthetic exposure to understand the response of a variety of cell types. For example, a cultured microglia cell line had little cytokine response to sevoflurane and isoflurane (1 and 2 MAC) or to 1 or 2 ED ₅₀ of propofol. However, in the presence of low-dose lipopolysaccharide (LPS), robust cytokine responses occurred, which were amplified by exposure to the volatile drugs, sevoflurane more than isoflurane. On the other hand, clinical concentrations of propofol completely ablated the cytokine response to LPS. While translation of cell culture results continues to be vexing, such studies will continue to be important to define molecular pathways contributing to cytotoxicity in specific cell types.

Animal Studies

The literature contains numerous animal studies demonstrating the contribution of the inflammatory response to PND/POCD. For example, while subtle changes in behavior can be detected after anesthesia alone (typically isoflurane), the effects are considerably larger and longer lasting when combined with even a brief surgical procedure. The effects of surgery-induced inflammation are further magnified in aged animals, metabolic syndrome animals, and in those with human disease transgenes, consistent with predictions above based on BBB disruption. The cognitive effects are usually accompanied by more pronounced elevations in both peripheral and brain cytokine levels, and activation of astrocytes and microglia, disrupting the precise glia-neuron signaling axis. Removal of pro-inflammatory components, either via cytokine antibodies (to IL-6 or TNFα) or antiinflammatory drugs (e.g., dexamethasone, statins), at least partially mitigates the effect of anesthesia and surgery on both behavior and pathology. Because the pro-inflammatory effect is important for surgical recovery, some of these studies have noted that blocking these cytokines impairs wound healing. Thus, attention is shifting to modulation of inflammation via cholinergic pathways or to the inflammatory resolution pathways, a complex mixture of specific cytokines (IL-4, 10) as well as small lipid mediators (resolvins). Two studies suggest efficacy of either exogenous cholinergic agonists or resolvin therapy to mitigate PND in rodents. Finally, the bidirectional involvement of pain and analgesics in neuroinflammation is well appreciated, but the role in PND/POCD has not yet been systematically studied.

Although the evidence to date suggests that the anesthetic plays only a minor role in producing PND/POCD, the choice of drug may modulate the magnitude and duration of the neuroinflammatory response. For example, the cell culture studies suggested that propofol mitigated the effect of LPS on cultured microglia; accordingly, the use of propofol as compared to desflurane significantly reduced both the pathology and the behavioral effects of surgery in a transgenic animal. In fact, one study suggested that chronic (subanesthetic) doses of propofol reduced the trajectory of decline in a model of murine AD.

Human Studies

As is becoming more the rule than the exception, the above rodent studies have only modestly translated to humans. This is undoubtedly a result of extreme variability in how patients are treated, and in the patients themselves. For example, most patients receive both propofol and a volatile agent; few receive a volatile agent alone. There is clear evidence of a peripheral pro-inflammatory effect of surgery, but of widely varying magnitude and duration. Very consistent evidence for this peripheral immune response being mirrored in the CNS has also been reported. A series of studies sampled CSF before and out to 24 or 48 hours after surgery and measured several cytokines. In each case, significant elevations in pro- and antiinflammatory cytokines were detected, which appeared to still be rising even at 48 hours. In addition, one study confirmed preclinical evidence of early changes in BBB integrity as indicated by an abrupt and transient increase in CSF/plasma albumin ratio after orthopedic surgery. Interestingly, the nonrandomized study found that total IV anesthesia was associated with lower CSF pro-inflammatory cytokines than inhalational anesthesia, although the larger randomized study could detect no difference between isoflurane- and propofol-based anesthetics. Again, it may be important to note that the isoflurane group also received propofol for induction in both studies, and that the dose-response relationship for antiinflammatory effects of this drug have not been established. A more recent PET imaging study noted that a marker of activated astrocyte microglia activation was significantly suppressed in patients 4 days after abdominal surgery and followed by an elevation at 3 months postsurgery that was associated with a reduction in cognitive capacity ( Fig. 82.3 ). These findings are consistent with earlier studies with the same PET probe in primates and humans, where an initial and rapid (4 hours) pro-inflammatory signal due to LPS was followed by depression at 22 hours. Although difficult and expensive, more time points will be necessary in these biomarker studies to define the contribution of innate immunity to the cognitive outcomes.

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

Local Anesthetics Opioids Prehospital Care for Medical Emergencies and Trauma Anesthesia for Orthopedic Surgery Anesthesia for Otolaryngologic and Head-Neck Surgery Anesthesia for Pediatric Cardiac Surgery

Stay updated, free articles. Join our Telegram channel

Join