Bringing to Light the Risk Factors And Incidence of Neuropsychological Dysfunction in ICU Survivors, 2nd Study

Delirium affects 50-70% of patients in Intensive Care Units (ICUs) with respiratory failure or shock.1-8 ICU delirium predicts death,2,8 length of stay,3 cost,9 and acquisition of Alzheimer's Disease and Related Dementia (ADRD).10-14 Alzheimer's disease and related dementias" or "ADRD" refers to a diverse range of cognitive impairments that may specifically reflect Alzheimer's disease or, alternatively, may reflect any number of conditions that fall broadly under the rubric of dementia. The relationship between delirium and ADRD must be studied in ICU survivors.10-13,15-23 The risk factors, diagnostic tools, mechanisms, phenotypes, and histopathology of this ADRD are ill-defined.

Our NIA-funded, NEJM published, PAR-18-029 cited original BRAIN-ICU-1 Study10,14 of 821 subjects showed over one-third of ICU survivors (without preexisting dementia) emerged with new ADRD by 1 year.10-12,17,19,24-27 Our pilot data found that many patients have a progressive ADRD phenotype. Some demonstrated transient decline followed by cognitive recovery. Our limited work indicates that cognitive rehabilitation may induce recovery,28 but we must discover which ADRD phenotypes are responsive to intervention.29,30 We must understand neurodegenerative and pathoanatomic risk factors for ADRD in survivors. Our pilot MRI data show acute ICU delirium is associated with atrophy of the whole brain, frontal lobes, and hippocampus.21,22 Specifically, a higher ventricle-to-brain ratio (VBR) appears linked with ADRD.20 We lack information on neuroinflammation, microvascular injury, white matter disease, and amyloid/tau-related neurodegeneration pathways.

In ICU survivors with/without delirium, the BRAIN-ICU-2 Study will define phenotypes, risk factors, diagnostics, mechanistic pathways, and histopathology of ADRD (in direct response to PAR-18-029). We will characterize which patients show cognitive reserve, recovery, and ADRD progression, and in so doing, address knowledge gaps between delirium and ADRD and facilitate future interventions.

Our overarching hypothesis:

Delirium results in neurodegenerative and neuropathological changes that can be measured with imaging biomarkers, plasma/cerebrospinal fluid (CSF) biomarkers, and autopsy measures of neuropathology, all of which ultimately drive post-ICU ADRD phenotypes, which we will characterize in BRAIN-ICU-2.

1.2. Specific Aims (Figure 1) Aim 1: To test the hypothesis that ICU survivorship and mortality over a decade is associated with unique clinical ADRD phenotypes. With our extended BRAIN-ICU-1 and new cohort (with/without delirium), we will amass survivor follow-up data at 3mo \& 12mo (original + new cohort, below), 4y \& 6y (completed), and now 12-14y data. Clinical ADRD Consensus Diagnosis will include Repeatable Battery for Assessment of Neuropsychological Status (RBANS) 31-33, and tests of memory, executive function, functional outcomes, \& quality of life. We will conduct trajectory analyses to define phenotypes (reserve, recovery, progression).

Aim 2: To test the hypothesis that global brain atrophy as assessed by structural MRI will mediate the association between delirium duration and clinical ADRD phenotypes at 12-month follow-up. We will enroll a new ICU cohort of 567 patients from Vanderbilt \& Rush and engage 3mo \& 12mo ADRD follow-up. In-hospital, we will assess delirium (Confusion Assessment Method-ICU), baseline risk (e.g., education, frailty, comorbidity, ApoE), neuroinflammatory plasma biomarkers (e.g., HMGB1, S100B, PAI-1), and clinical risks (e.g., sepsis, drug exposures, immobility). In addition to enrollment MRIs, we will also obtain 3T MRI, plasma, and CSF biomarkers at discharge and 12-month. Our primary analysis will assess if the change in ventricle-to-brain ratio (VBR) mediates the association between delirium and clinical ADRD. Secondary analyses will evaluate MRI measures of brain structure (white matter integrity) \& function (perfusion), and plasma/CSF biomarkers for AD neuropathology, inflammation, \& synaptic/axonal integrity as additional mediating paths.

Aim 3: To test the hypothesis that microvascular pathology underlies post-ICU ADRD. We will establish a repository at Rush and analyze decedents' brains from both the extended BRAIN-ICU-1 and new ICU cohort. We will assess the association between delirium and autopsy measures of neuropathology focusing on cerebrovascular disease (e.g., microinfarcts due to endothelial activation in sepsis). We will also fully characterize the presence of other common neuropathologies (e.g., plaques, tangles, TDP-43).