Abstract
The ageing population in Western countries combined with advances in anaesthetic management and surgical techniques results in a growing number of frail patients presenting for surgery in the last 30 years. Increased age comes with reduced physiological reserves of major organs, and only worsens in the presence of concurrent disease. This fact undoubtedly contributes to an increase in perioperative complications.
Introduction
The ageing population in Western countries combined with advances in anaesthetic management and surgical techniques results in a growing number of frail patients presenting for surgery in the last 30 years. Increased age comes with reduced physiological reserves of major organs, and only worsens in the presence of concurrent disease. This fact undoubtedly contributes to an increase in perioperative complications.
Of all perioperative complications, cognitive impairment is a dreaded one. The association between perioperative changes and impaired cognitive status of elderly patients has been repeatedly confirmed since the first study of Bedford et al in 1955 [1–3]. However, contrary to common belief it seems that postoperative cognitive impairment is not caused by ‘anaesthesia’ and more specifically anaesthetic agents. Experimental and epidemiological evidence increasingly suggests that surgical stress is a key element in its genesis. Also, all cognitive dysfunction is not the same. The postoperative period is often followed by a short period of disrupted psychomotor performance and attention span. These phenomena are probably related to ‘sickness behaviour’ which is a normal behaviour in the aftermath of biological stress. However, cognitive dysfunction after the second postoperative day (on day three) should be considered pathological. Even though postoperative cognitive disorders remain poorly categorized, it is possible to discern two chronologically distinct pathologies: postoperative delirium (days three to seven), and postoperative cognitive dysfunction (POCD) (begins after the first week and can consolidate). More and more data suggest these two entities may have a common aetiological substratum and, for a subgroup of patients, there may be a continuum between the two pathologies.
Perioperative Cognitive Function and Dysfunction
Fragility in Brain Ageing
It is now well established that the brain, like any organ, goes through a normal ageing process.
Common sense dictates there should be a clinicopathological correlate to cognitive dysfunction, but it has proven difficult to find evidence for it. The normal ageing process is characterized by anatomical changes that have been visualized through standard imaging (atrophy or decreased fractional anisotropy in MRI-DTI), but it has no clinically detectable associated changes. Chronic diseases like hypertension and diabetes may accelerate these anatomical changes. Contrary to the normal ageing process, neurodegenerative disorders (e.g. Alzheimer’s or Parkinson’s diseases as well as vascular dementia) will directly affect cognitive skills. Acute brain diseases will inflict major brain damage and quickly degrade cognitive function.
The normal ageing process can be expressed as a loss of reserves relative to an initial cognitive potential evaluated through the Mini-Mental State Examination (MMSE) – see Fig. 13.1. Loss of reserves varies significantly between persons, and depends on socio-economic and environmental factors [4]. It is difficult to measure this loss without a proper neurocognitive measurement tool, and even in-depth neurocognitive assessment may not detect this loss. Nevertheless it seems crucial to detect preexisting loss of cognitive reserve (i.e. the frailty of the patient) early on during the perioperative phase [5]. It is challenging to identify and quantify loss of cognitive reserves, assess its different clinical presentations, and to track postoperative cognitive decay or recovery of postoperative cognitive function.
Fig. 13.1 Brain function over time. The normal ageing process can be expressed as a loss of reserves relative to an initial cognitive potential. When it decreases below a certain threshold, acute confusion/delirium results. Whether or not surgical stress results depends on (1) the frailty, which is the baseline position on the Y-axis at the time of stress and (2) the degree of stress (small versus big).
Early and Late Postoperative Cognitive Dysfunction
Early or Acute Postoperative Cognitive Dysfunction: Delirium
The American Psychiatry Association has recently defined delirium in the DSM-V Diagnostic and Statistical Manual of Mental Disorders 5th Edition. Postoperative delirium is a sudden event, usually occurring within 24 to 48 hours postoperatively [6]. It is characterized by disturbance of consciousness that fluctuates over time, reversal of sleep–wake rhythm, and cognitive impairment. The latter is global, with varied symptoms ranging from memory and behavioural disorders to temporo-spatial disorientation and anxiety. Three subtypes of delirium are discerned: hyperactive (25%), hypoactive (50%) and mixed (25%). This cerebral dysfunction is organic in origin and is in theory reversible [7].
Persistent Postoperative Cognitive Dysfunctions
The persistent form of POCD is less easily characterized than the acute one. While there is no formal definition in the DSM-V, it is conventionally accepted that a patient has ‘persistent POCD’ when one or more cognitive domains remain dysfunctional one week after a surgical procedure. The first reports on POCD in the 60s described how anterograde episodic memory and attention were affected after cardiac surgery [8], a finding later also described after non-cardiac surgery [9, 10]. Some authors have challenged this entity to be pathological, and argue these cognitive problems may represent normal age related decline of the cognitive function of a subgroup of patients. During the peri-operative period, the decline would only temporarily be accelerated; later, it would resume its preoperative rate of decline in such a manner that there would be no net change in decline relative to that to be expected had the patient not undergone surgery [11].
Epidemiology and Risk Factors
Delirium is one of the most common postoperative complications in patients over 60 years old, with an estimated incidence of 20% and 40% [12]. Risk factors associated with delirium in patients admitted for non-surgical (medical) reasons may also be involved in those admitted for surgery: age, chronic brain disease, depression, addiction, neurosensory deficit, chronic organ failure, endocrine disorders, iatrogenic factors and psychological factors. Certain types of surgery (orthopaedic, vascular or cardiac surgery) and emergency surgery, and a number of other perioperative factors have also been associated with an increased risk of postoperative confusion (Table 13.1). Delirium is associated with a deterioration of the patient’s functional prognosis, reduced life expectancy [13, 14] and persistent POCD after cardiac surgery [15].
The incidence of persistent POCD in patients older than 60 undergoing non-cardiac surgery is between 10% and 54% [16]. After cardiac surgery, some studies show a higher incidence of POCD, regardless of whether or not extracorporeal circulation was used [17]. Because the diagnostic criteria are not clear yet, interpretation of the results of studies on POCD remains controversial [9]. Known risk factors include age, preoperative cognitive impairment, depression, perioperative sepsis and the metabolic syndrome. Individual and social consequences of POCD include an increase in mortality, and loss of ability to live autonomously and return to work [3].
Role of Neuroinflammation
Some pathophysiological mechanisms of delirium and persistent POCD seem to be similar. They appear to be related to the combination of preexisting vulnerability factors (cognitive impairment, addiction, age), the degree of operative stress (itself related to the type of surgery), and additional acute organic factors such as hydro-electrolytic disorders, sepsis and repeated episodes of low arterial blood pressure. In the past, these cognitive problems were often hypothesized to result (directly or indirectly) from general anaesthesia, i.e. from the pharmacological effect of the anaesthetic agents used, from haemodynamic instability, or from ventilation or oxygenation problems. However, epidemiological studies have found no causal link between low blood pressure, hypoxaemia and the occurrence of POCD. The type of anaesthesia (general versus locoregional) could not be correlated with persistent POCD [18].
The most likely cause is now hypothesized to be perioperative inflammatory stress, both in its initial phase and in its resolution. The immune response protects the body against pathogens by recognizing elements of ‘non-self’ pathogen associated molecular patterns (PAMPs). The same immune system is also involved in recognizing elements of ‘self’ named damage associated molecular patterns (DAMPs) in certain situations such as cancer and cell injury. Perioperative tissue damage releases DAMPs (both locally and into the bloodstream) that are recognized by PPRs (pattern recognition receptors), molecules expressed on the surface of leucocytes. Activation of these PPRs causes nuclear translocation of transcription factors that serve to rapidly increase the release of proinflammatory cytokines such as IL-1β, TNF-α and IL-6. This response to PAMPs and DAMPS characterizes the initial phase of innate immunity (Fig. 13.2). It is hypothesized this is the manner by which surgery activates immune cells that can massively release proinflammatory agents.
Fig. 13.2 Neuro-inflammation. The immune response protects the body against pathogens by recognizing elements of ‘non-self’ PAMPs (pathogen associated molecular patterns). The same immune system is also involved in recognizing elements of ‘self’ named DAMPs (damage associated molecular patterns) in certain situations such as cancer and cell injury. Perioperative tissue damage releases DAMPs (both locally and into the bloodstream) that are recognized by PPRs (pattern recognition receptors), molecules expressed on the surface of leucocytes. Activation of these PPRs causes nuclear translocation of transcription factors that serve to rapidly increase the release of proinflammatory cytokines such as IL-1β, TNF-α and IL-6. This response to PAMPs and DAMPS characterizes the initial phase of innate immunity. It is hypothesized this is the manner by which surgery activates immune cells that can massively release pro-inflammatory agents.
The local and systemic effects of this immunological reaction are controlled by a mechanism called ‘the resolution phase’. One of its local mechanisms is a vagal reflex [19] that inhibits macrophage activity via activation of the α-7 nicotinic acetylcholine receptor [20]. The brain and particularly the hippocampal region appear to be a particular target for these activated immunocytes. The cerebral response to surgical stress is characterized by local secretion of pro-inflammatory cytokines, recruitment of immune cells and activation of microglia (inflammatory brain cells), which combined generate a provisional neuro-inflammatory profile. This can alter synaptic plasticity (disrupting hippocampal long-term potentiation, a neurobiological correlate of learning and memory) and has been associated with cognitive disorders in animals that clinically resemble perioperative cognitive impairment in humans [21] (Fig. 13.3). Age, systemic inflammation, infectious diseases and metabolic syndrome can all modulate the inflammatory response by increasing the initial response and preventing the resolution phase.
While surgical stress appears to be the trigger of this inflammatory response, other perioperative events could also modulate the inflammatory response and thus play a role in the aetiology of cognitive disorders. These factors include (1) postoperative sleep disorders (especially in patients with sleep apnoea); (2) drugs used to provide anaesthesia and perioperative analgesia; (3) postoperative pain; and (4) perioperative infection. Finally, by reducing surgical stress, minimally invasive approaches (intraoperative techniques, avoiding cardiopulmonary bypass) are considered as an attractive therapeutic target to limit the incidence of POCD.
Early Postoperative Delirium: Causal Factor of Persistent Postoperative Cognitive Dysfunction?
Delirium after cardiac surgery is a risk factor for persistent pathology [16, 22]. Actually, because the risk factors for both diseases are so similar, more and more authors believe there to be a POCD continuum. Even though the issue is not yet completely resolved, two hypotheses exist that link delirium to POCD. A first hypothesis claims a causal link between delirium and POCD exists, and predicts that preventing delirium will have an impact on the incidence or severity of POCD. A second hypothesis claims delirium to be a marker associated with POCD, and therefore identifying delirium could enable us to determine which patients are at high risk of POCD and focus on this group to try to come up with both aetiological and symptomatic treatment.
With both hypotheses, it seems important to diagnose delirium early on, to establish when it started, and whether it is hyper- or hypo-active, and adopt the necessary measures to treat it. The cornerstone of diagnosing POCD is a reliable, easy-to-use diagnostic tool to repeatedly assess the patient’s cognitive state.
Is There a Marker of Cognitive Dysfunction?
One of the major challenges in the management of POCD is diagnosing it. There are currently no specific criteria to define persistent POCD. POCD screening tools reported in the literature are very heterogeneous, yet for optimal management and to facilitate comparisons between studies measurement of POCD ought to be standardized and able to be performed at the bedside. Better still, a specific brain marker should be identified.
Biomarkers
In post-traumatic or post-stroke patients, several brain injury biomarkers have been validated. Based upon these data, and assuming that POCD also involves some degree of brain injury, several authors have studied the link between the dynamic changes in the levels of these biomarkers and the changes in cognitive function. Besides allowing POCD to be diagnosed per se, a specific biomarker would allow us to diagnose POCD before the appearance of the first clinical symptoms, opening the possibility for individualized prophylactic measures. However, a blood test to diagnose POCD remains elusive at this time. In one study that followed 150 patients after non-cardiac surgery [23], increased plasma GFAP (an astrocyte protein) measured immediately after surgery could be correlated with POCD one month later, yet two other biomarkers, PS100 and NSE, could not. In another study [24], the same biomarkers and the APOE4 genotype (involved in the aetiology of Alzheimer’s disease) again could not predict POCD.
Brain Imaging
MRI has been studied as a tool to predict or diagnose POCD. Preoperative temporal lobe abnormalities on MRI correlated with the risk of POCD. In a prospective study in elderly subjects undergoing gastrointestinal surgery, a small hippocampal volume predicted the occurrence of early POCD well [25]. This was expected since it is known that one of the main risk factors for POCD is preexisting cognitive impairment, which has been associated with changes in hippocampal volume [26, 27]. Because temporal MRI abnormalities may occur several years before the onset of clinical signs, MRI could detect patients at risk of POCD even at a very early stage.
But some investigators also reported negative results when exploring the use of imaging techniques such as brain MRI in diagnosing POCD. Acute MRI lesions after cardiac (CABG) [28] and non-cardiac surgery did not correlate with delirium or persistent POCD. Summarized, brain imaging is currently not a completely effective tool to detect or predict early persistent POCD at every point during the surgical process, but it can help to preoperatively define patients at risk for POCD. Therapeutic implications remain, however, limited.
Neuropsychological Tests
The ‘Troponin like biomarker’ that diagnoses POCD will probably result from neuropsychological tests. POCD is widely underdiagnosed, especially in the elderly population. Indeed, a postoperative delirium in an elderly patient it is often trivialized and put on the account of preexisting dementia. Delirium can present in several forms clinically, as previously described. In more than 50% of patients symptoms are not recognized because of the high percentage of patients that have a ‘hypoactive’ presentation without any associated agitation. That is why it is essential to come up with a reliable tool to screen for POCD: standardized, repeatable, fast, simple and easily performed at the bedside by a nurse or physician.
Early POCD, delirium, can be diagnosed with the CAM (Confusion Assessment Method), which is the only validated and reliable clinical tool that can detect the symptoms of delirium. The CAM can be performed by health professionals without neuropsychiatric training [29], takes less than 5 minutes, can be done at the bedside, and has a high sensitivity and specificity [30]. CAM assesses the acute nature of POCD, and whether any of the following are present: inattention or lack of focus, disorganized thinking, psychomotor agitation or apathy, and fluctuations and changes in the sleep–wake cycle. The CAM has been adapted for use in the ICU patient (CAM-ICU).
Diagnosing persistent POCD is more complex. In the absence of well-defined criteria, it is generally considered that a patient presents with late POCD if there is a deterioration of cognitive tests compared to the preoperative or immediate postoperative period. Different clinical studies have used a battery of neuropsychological tests with variable diagnostic thresholds [5, 31]. But a standardized, simple and reliable screening test that would allow us to routinely diagnose and grade cognitive function throughout the perioperative course has not yet been developed.
Related posts:
Full access? Get Clinical Tree
