Neuroimaging Risk of Postoperative Delirium

Delirium is a fluctuating disturbance of consciousness with accompanying impairments in attention, cognition, perception, emotional regulation, psychomotor abilities as well as sleep-wake cycle regulation. Delirium frequently occurs following surgery with general anaesthesia and presents a common postoperative complication in elderly patients (>=65 years). 'Postoperative delirium' (POD) refers to the occurrence of delirium from as early as the timepoint immediately following surgery, i.e., at emergence from general anaesthesia, and up to 30 days following surgery. The reported occurrence of POD is highly variable, ranging from 10% to 70%, depending on the age of patients, diagnostic criteria, and the type of surgery. Those especially vulnerable are the oldest patients with the most medically complex cases undergoing cardiac, vascular, or hip fracture surgery. POD is associated with functional decline; a linear association between mortality and number of delirious days has been reported. POD has also been shown to increase risk of dementia up to 10-fold alongside a 2-fold increase in risk of institutionalisation and death. POD increases the length of general hospital stay and is associated with more than doubling one-year medical costs. Hence, there is a drive to elucidate the mechanisms of POD and identify risk factors that can facilitate adequate patient care management to reduce POD occurrence, both from a health and a costs perspective.

Several risk factors for POD have been identified, which can be categorised into two groups based on when they occur relative to surgery. Preoperative 'baseline' risk factors for POD include older age, pre-existing cognitive impairment, sensory and functional impairment, use of psychotropic drugs, history of major depressive disorder (MDD), institutional residence and greater comorbidity, especially with regards to cardiovascular disease. Peri-/intraoperative 'precipitating' risk factors include perioperative pain, environmental unfamiliarity, type of surgery, disrupted sleep, anaesthetic dosage, use of a bladder catheter and the type of medications used. An individual's risk of POD has been modelled as a sum of these baseline and precipitating risk factors, where a heightened burden of baseline preoperative factors necessitates fewer peri/intraoperative precipitating factors to induce POD. This interaction between baseline and precipitating factors follows the hypothesis of neural functional reserve, which suggests that an individual's ability to tolerate situations that encompass stress to the brain or brain 'injury' is dependent upon their baseline functional reserve - representing a patient's 'resilience'. Baseline preoperative risk factors of POD can be seen as factors that reduce a patient's reserve or resilience, and peri-/intraoperative 'precipitating' risk factors can be seen as prime examples of situations encompassing stress on the brain. A low functional reserve has been shown to be an independent risk factor for POD. Therefore, investigations of the underlying baseline preoperative neurological impairments that have the effect of reducing a patient's resilience to peri-/intraoperative stressors - and increase their risk of POD - are highly warranted. The investigators will assess whether two potential baseline preoperative neural impairments contribute to a patient's risk of POD.

The glymphatic system is a recently discovered waste clearance system of the brain, utilising the perivascular channels to sweep cerebrospinal fluid (CSF) through and away from the brain's parenchyma, taking waste products and extracellular metabolites with it, essentially cleaning the brain. Aquaporin-4 water channels located at astrocytic end-feed form an essential component of this system, facilitating CSF movement from perivascular spaces into the interstitial space. There are a small number of non-invasive measures available to assess aspects of glymphatic system function in humans. The first measure was proposed by whereby fast fMRI acquisition (TR<400ms) - with the boundary edge of the volume (i.e., slice 0) placed at the level of the fourth ventricle - allows for the detection of fluid inflow arriving at the edge of the imaging volume with high signal intensity. The acquired CSF signal with slow oscillations at around 0.05Hz were found to be of significantly higher amplitude during sleep vs. wake, driven by slow neural oscillations in the delta range (~1-4Hz), and was anticorrelated with fluctuations in the global grey matter BOLD signal. A decrease in this anticorrelation was reported in patients with Parkinson's disease and was significantly related to lower cognitive performance on the Montreal Cognitive Assessment tool. A relationship between reduced global BOLD-CSF flow coupling and severity of small vessel disease has been reported, related to increased dilation of the perivascular space. Hence, the global BOLD-CSF coupling could provide an indication as to impaired glymphatic transport.

A second recent approach was developed using diffusion tensor imaging (DTI) of diffusion-weighted images to evaluate the motion of water molecules along deep medullary veins and their accompanying perivascular spaces running perpendicular to the ventricular walls on a transverse slice at the level of the lateral ventricle bodies. At this location, the distribution of different fibre types allows for the discernment of water diffusivity along the perivascular space independently from that within corticofugal corona radiata projection fibres (running in the craniocaudal direction, i.e., z-axis) and superior longitudinal fascicle association fibres (running in the anterior-posterior direction, i.e., y-axis). This approach, called the 'DTI-ALPS' method, i.e., 'diffusion tensor imaging along the perivascular space', can non-invasively measure glymphatic transport (i.e., CSF flow through the perivascular space) at this specific location, as the perivascular space runs perpendicular to both the projection and association fibres. Regions of interest are first manually labelled from a T1-weighted scan to measure diffusivities within these fibres along three directions (x, y, and z). Perivascular diffusion can be evaluated independently from diffusion along projection and association areas by identifying the ratio of the mean x-axis diffusivity in both areas to the mean diffusivity of these areas at the axes perpendicular to the x-axis and fibre tracts, i.e., the y-axis for projection fibres and z-axis for association fibres.

Perioperative factors such as the type of GA, in combination with various other perioperative factors, may alter postoperative brain function by affecting the glymphatic system. Dysfunction of this system that removes harmful metabolites from the brain can lead to neurological disorders and neuroinflammation. Reduced glymphatic system transport leads to an increased accumulation of amyloid-beta and tau, inducing neuroinflammation leading to a chain of processes that end in neuronal damage. Postoperative increases in amyloid-beta and tau have been reported and shown to be a risk factor for POD. As outlined above, different anaesthetics have differential effects on the glymphatic system, and anaesthetic type has been shown to be a risk factor for POD in patients. Volatile anaesthetics such as isoflurane or sevoflurane have a higher chance of causing POD, whereas intravenous anaesthetics such as dexmedetomidine and propofol seemingly rather have a protective effect against POD.

Elderly patients but not young people are highly susceptible to POD, triggering the question of why this group is selectively susceptible. As outlined previously, preoperative risk factors for POD include preclinical Alzheimer's disease, cerebral microemboli, vascular dementia, hyperglycaemia and other neurovascular diseases or neurodegeneration, all of which show glymphatic transport impairment. Ageing is associated with degenerative changes such as depolarised or downregulated aquaporin-4, sleep deprivation, and reduced cerebrovascular pulsatility, all of which have the potential to impair glymphatic system transport. In line with the functional reserve hypothesis, it has been suggested that susceptible patients at risk of POD have pre-existing glymphatic system impairment that progresses slowly without significant cognitive impairment, and that the perioperative risk factors exacerbate the pre-existing glymphatic damage beyond the compensatory limit, resulting in significant cognitive changes. This theory suggests that baseline preoperative measures of glymphatic system transport can be informative on the risk of POD. Our primary aim is to investigate whether baseline preoperative measures of glymphatic system transport can be useful in identifying patients at risk of POD, utilising the two non-invasive MR-based measures outlined previously.

The overall aim of the current study is to investigate potential underlying pathobiological mechanisms that contribute to POD, to further understand the mechanisms of this frequent disorder in the elderly and enhance perioperative care and patient management. Whether an impairment in glymphatic system transport at preoperative baseline can be informative on risk of POD will be assessed. All target values will be assessed at preoperative baseline. Assessment of whether a patient is classified as a 'POD patient' or a 'noPOD patient' is based on a diagnosis of delirium on any of the five days following surgery (including the day of surgery).