Cohort Study to Investigate the Association Between Changes in Brain Volume and Postoperative Cognitive Dysfunction

Background:

Anesthesia care for elective surgery of patients that are in their seventies or eighties is an ever-growing entity in clinical practice, given the demographic development in many countries of the Western world. Cognitive decline of variable degree is frequently observed after major surgery in elderly patients. A potential association between anesthesia and the development of dementia and brain atrophy in elderly patients is controversial. However, it is not uncommon that during preoperative visits by the anesthesiologist, patients express fear to suffer some sort of cognitive decline after the operation, and raise concerns about psychological or neurological long-term adverse effects (e.g. memory loss) related to anesthesia. Clinicians have identified this problem decades ago, and are adding postoperative cognitive dysfunction (POCD) to the list of frequent complications of anesthesia in elderly patients; yet, there is a lack of prospective clinical data and poor understanding of the pathophysiology of POCD, and no preventive strategy or treatment has been described so far.

One study reported an incidence of postoperative cognitive dysfunction (POCD) of 41% in surgical patients 60 years of age, or older. In a previous study conducted by our group, as many as 47% of patients over 65 years presented with POCD one week after surgery. In many patients, this is a transient problem, but some patients retain permanent deficits, with increased mortality and important consequences on the socioeconomic situation. An association between surgery and long-term cognitive decline has not been established yet; however, patients who develop postoperative delirium (a transient, but particularly severe form of POCD) have an increased risk of developing dementia in the years following surgery. However, in a cohort follow-up study, the diagnosis of POCD itself did not represent a significant risk factor for dementia. The pathophysiology of postoperative cognitive decline remains unclear. However, studies in animals have shown that the systemic inflammatory response elicited by the surgical procedure is associated with microglial activation and postoperative cognitive impairment. Furthermore, volatile anesthetics have been shown to be neurotoxic in cell cultures, and in some animal experiments exposure to clinical concentrations of volatile anesthetics was associated with postoperative cognitive decline. A recent retrospective analysis of cohort data suggests that surgery may be associated with a decrease in brain volume in patients. However, due to methodological limitations in this study, no analysis of a potential link between atrophy and cognitive functions was made.

Study Design:

Prospective cohort study

Study Groups:

Two groups of patients will be investigated. Seventy patients undergoing major surgical procedures will be recruited for this study, and 30 healthy study participants (no surgical intervention) will be recruited to serve as controls. In all patients, baseline cognitive functions and structural MRI imaging including determination of gray matter volumes relevant for cognitive functions (hippocampus) will be performed.

• Group 1: Surgery group Sevoflurane (n = 70)
• Group 2: Control group (n = 30)

Recruitment:

Study participants of the Surgery group Sevoflurane (Group 1) will be recruited as patients of the Basel University Hospital, a tertiary medical care center in Basel, Switzerland, associated with the University of Basel, Switzerland. Study participants of the Control group (Group 2) will be recruited randomly.

Endpoints:

Targeted primary endpoint of the study is the difference in change of hippocampal volume over time between patients with POCD and patients without POCD.

Secondary endpoint is the correlation between change in cerebral volume and change in cognitive function.

Anesthetic Management:

Anesthesia will be standardized (general anesthesia using sevoflurane and fentanyl for maintenance, thiopental for induction and atracurium or rocuronium for neuromuscular blockade). Intraoperatively, the dose of anesthetics will be controlled using an EEG derived index (depth of anesthesia monitoring (Bispectral Index (BIS), Aspect Medical Systems), and end-tidal sevoflurane concentration monitoring). Furthermore, intraoperative bilateral cerebral near-infrared spectroscopy (NIRS) monitoring (NIRO-200, Hamamatsu Photonics, Japan) will be applied. Data will be downloaded directly from the monitors and the anesthesia machine (Dräger Perseus) using ICM+ software.

Neuropsychological Assessment:

The cognitive assessment test battery CERAD-Plus consistent of the Consortium to Establish a Registry for Alzheimer's Disease - Neuropsychological Assessment Battery (CERAD-NAB), Trail Making Tests A+B, and phonetic fluency (s-words), as well as the Test for Attentional Performance (TAP) developed to analyze different aspects of attention, will be performed preoperatively, at 7 days, 3 month and 12 months postoperatively. Training of all study personnel and supervision of cognitive testing will be carried out by the Memory Clinic at Basel University Hospital. Cognitive functions will be quantified using the CERAD total score, in its demographically corrected form. A correction for short-term practice effects will be carried out based on previous work.

Blood Samples:

Blood samples will be taken from study participants preoperatively, 12h postoperatively, and on day 2 and 7, to determine levels of C-reactive protein (CRP), interleukin-6 (IL-6), tumor necrosis factor-alpha (TNF-alpha). Blood samples will be stored for 1 year to allow additional determinations.

Demographic, Procedural and Subjective Patient Data:

• Age, sex, and years of education.
• Cognitive testing as outlined: CERAD-Plus (CERAD-NAB, Trail Making Tests A+B, phonetic fluency (s-words)), Test for Attentional Performance (TAP), preoperatively, at 7 days, 3 months and 12 months postoperatively.
• Charlson Comorbidity Score preoperatively, at 3 months and 1 year postoperatively
• Instrumental Activities of Daily Living (IADL) preoperatively, at 3 months and 1 year postoperatively.
• Geriatric Depression Scale (GDS) preoperatively, at 3 months and 1 year postoperatively.
• Short Form Health Survey (SF-36) preoperatively, at 3 months and 1 year postoperatively.
• ASA physical status, type and duration of anesthesia, type and duration of surgery.
• Postoperative delirium (Confusion Assessment Method; CAM), repeat surgery and postoperative infections.
• Subjective grading of cognitive functions on a VAS scale (0-10) preoperatively, at 7 days, 3 months and 1 year postoperatively.
• Subjective grading of change in cognitive functions on a five-point Likert scale (much worse, worse, equal, better, much better) at 7 days, 3 months and 1 year postoperatively.
• Subjective grading of successful surgery (Yes/No) at 7 days, 3 months and 1 year postoperatively.
• Subjective grading of pain on a VAS scale (0-10) preoperatively, 7 days, 3 months and 1 year postoperatively.

Magnetic Resonance Imaging (MRI):

MRI will be performed preoperatively, at 3 months, and 1 year postoperatively on the same day as cognitive function testing. No intravenous contrast dye will be administered. Training of all study personnel and MRI evaluation, and/or supervision of MRI evaluation, will be carried out by the Division of Neuroradiology at the University Hospital Basel. High resolution anatomic and diffusion MRI will be performed using the hospital's 3T MAGNETOM Prisma™ MRI scanner (Siemens, Zurich, Switzerland). For our initial MRI analysis, we assume that the same regions as those described in mild cognitive impairment (MCI) and Alzheimer's disease (AD) are relevant to POCD. We will perform a region of interest (ROI) analysis during this part of the analysis. In a second step, we will perform a whole brain analysis.

The whole brain MRI protocol includes high resolution three-dimensional (3D)-T1 (MPRAGE sequence, repetition time (TR) = 1620 ms, echo time (TE) = 2.48 ms, 192 slices), clinical standard at the University Hospital Basel) and 3D-T2 (SPACE sequence, TR = 3200 ms, TE = 408 ms, 192 slices) imaging for segmentation. Furthermore, a diffusion sequence (DTI sequence, TR = 5800 ms, TE = 77 ms, 50 slices, 64 directions, 8 B0) for quantitative imaging will be used. We added a two-dimensional (2D)-PD/T2 weighted sequence (TSE sequence, TR = 4480 ms, TE1 = 9 ms, TE2 = 106 ms, 46 slices) for diagnostic purposes. The total acquisition time for the MRI is approximately 24 minutes.

Statistical Analysis:

Cortical reconstruction and volumetric segmentation will be performed with the FreeSurfer image analysis suite, which is documented and freely available for download online (http://surfer.nmr.mgh.harvard.edu/). The technical details of these procedures are described in prior publications.

Comparisons (preoperatively to 3 months postoperatively) of hippocampal volume and the results of voxel-based volumetry of the medio-temporal lobe and lateral association areas of the temporal and parietal lobes will be performed using region of interest (ROI) analysis. The quantitative imaging results will be correlated to the CERAD total score performance obtained on the same day as the MRI scans. The relationship of changes in volume (if present) with intraoperative depth of anesthesia (BIS) and the administered dose of sevoflurane expressed as age-corrected minimum alveolar concentration (MAC) equivalent multiplied by time of administration will be calculated. We will also use multivariate approaches (voxel-wise multivariate analysis (MANOVA) and classification/prediction procedure) on multi-parameter MRI data in order to build optimal composite predictors of patients' neuropsychological outcomes. In whole brain analysis, we will use a multiple regression and repeated measures ANOVA whole brain voxel-wise analyses in FreeSurfer to correlate the estimated modulated gray matter volume with the neuropsychological scores.

Sample Size Calculation:

Assuming that 41% of operated patients develop POCD, and a standard deviation of the hippocampal volume change of 0.45 in both patients with POCD and without POCD, a total sample size of 56 subjects may detect a difference of 0.4 mm3 in hippocampal volume with a power of 90% at a two-sided significance level of 5%. To compensate for the high loss to follow-up which is unfortunately typical for studies on POCD, and estimated at 20%, we will recruit 70 patients in the surgery group.