Resting-state functional connectivity and cognition following major cardiac surgery in older adults without preoperative cognitive impairment: Preliminary findings

Abstract

Background/Objectives

Resting-state functional magnetic resonance imaging was used to examine for changes in intrinsic functional brain connectivity associated with postoperative changes in cognition; a common complication in seniors undergoing major surgery.

Design

Objective cognitive testing and functional brain imaging were prospectively performed at preoperative baseline and 6-weeks after surgery, and at the same time intervals in non-surgical controls.

Setting

Academic medical center.

Participants

12 senior patients undergoing cardiac surgery and 12 non-surgical senior controls with a history of coronary artery disease; both groups without cognitive impairment at preoperative baseline (MMSE > 27).

Measurements

Differences in resting-state functional connectivity (RSFC) and global cognitive change relationships were assessed using a voxel-wise intrinsic connectivity method, controlled for demographic factors and pre- and perioperative cerebral white matter disease volume. Analyses were corrected for multiple comparisons, p-FDR <0.01.

Results

Global cognitive change after cardiac surgery was significantly associated with intrinsic RSFC changes in regions of the posterior cingulate cortex (PCC) and right superior frontal gyrus (rSFG); anatomical and functional locations of the brain's default mode network (DMN). No statistically significant relationships were found between global cognitive change and RSFC change in the non-surgical controls.

Conclusion

Clinicians have long known that some older patients develop postoperative cognitive dysfunction (POCD) after anesthesia and surgery, yet the neurobiological correlates of POCD are not well defined. Our results suggest that altered resting state functional connectivity within specific DMN regions is positively correlated with global cognitive change 6-weeks after cardiac surgery, suggesting that DMN activity and connectivity could be important diagnostic markers of POCD and/or intervention targets for potential POCD treatment efforts.

Introduction

Postoperative cognitive dysfunction (POCD) is a syndrome that has been described in both cardiac and non-cardiac surgery patients, and which occurs more often in patients over the age of 60.¹ However, there is no current consensus on exactly how to measure POCD, when after surgery to measure it, or on how much a patient's cognitive performance must decline to be considered clinically relevant. Human clinical and animal studies have suggested that POCD may result from unresolved neuroinflammation,² Alzheimer's disease-associated pathology,³ specific anesthetics,⁴ perioperative cerebrovascular damage,⁵ and/or an effect of pre-existing health conditions such as metabolic syndrome.⁶ However, the relative contribution of these possible factors to human POCD is unclear. Further, we do not yet know whether structural or functional deficits in specific brain regions or global dysfunction underlie human POCD.

Resting-state functional connectivity (RSFC) is a reliable neurophysiological phenomenon characterized by shared coherent, spontaneous low frequency blood oxygen level dependent (BOLD) signal fluctuations among functionally related brain regions in the absence of task-related activation.⁷ RSFC networks have been identified for the motor system, task-negative/introspective thought and memory, stimulus salience, sensorimotor integration, auditory processing, and higher order executive functioning. The most studied RSFC network is the default mode network (DMN), a group of functionally connected posterior and anterior cortical regions that are active when individuals are at rest or internally focused, and which become less active when individuals shift their focus and cognitive effort to external tasks.⁸ Regions critical to the DMN, such as the posterior cingulate cortex (PCC), have been consistently identified as being densely anatomically interconnected with the rest of the brain and play a central role in global cerebral communication processes.⁹ RSFC alterations in DMN regions have been found in several conditions associated with cognitive decline or dysfunction including advanced aging,¹⁰ Alzheimer's disease¹¹ and delirium.¹² We hypothesized that perioperative changes in DMN RSFC would be correlated with postoperative cognitive changes in cardiac surgery patients, but not in non-surgical controls.

Methods

This study was approved by the DUMC Institutional Review Board (IRB), and all enrolled study participants provided written informed consent.

Participants

Patients over age 60 scheduled to undergo coronary artery bypass grafting (CABG) and/or valve replacement (VR) surgery with cardiopulmonary bypass were prospectively enrolled. Non-surgical controls with coronary artery disease (CAD), as evidenced by a prior myocardial infarction or by evidence of CAD on cardiac catheterization, were recruited from a cardiology clinic. Non-surgical controls could not be under consideration for surgical revascularization within 6 weeks of the baseline study visit. Patients and controls with a history of cortical stroke, alcoholism, psychiatric illness, renal failure, <7^th grade education, non-native English speaking, baseline Mini Mental Status Exam (MMSE) total ≤26, or unsafe for 3 Tesla MRI were excluded. See Table 1 and Supplemental Appendix 1.

Table 1. Demographic, General Health Factors and Surgical Characteristics.

	Variable	Group		Statisticb	p-value
		Surgical (n=12)	Controls (n=12)
Demographics	Mean age, yrs. (SD)	69.7 (7.3)	70.4 (7.9)	-0.24	0.82
	Sex (M/F)	8/4	11/1	2.27	0.32
	Mean Weight, kg (SD)	76.9 (17.4)	86.4 (10.7)	-1.61	0.12
	Education, yrs. (SD)	15.1 (2.1)	15.0 (2.9)	0.08	0.92
	Race (Afr.Amer/Cauc/Asian/Other)	0/11/1/0	0/12/0/0	1.04	1.00
Health Factors	Diabetes (Type I or II)	3	5	0.75	0.67
	History of Hypertension	6	11	3.23	0.07
	Previous Myocardial Infarction	1	1	0.00	1.00
	Baseline Ischemic White Matter Damage Burden a	0.005 (0.007)	0.003 (0.004)	0.86	0.40
	White Matter Lesion Volume Change from Baseline (mL)	2.28 (5.46)	1.97 (2.70)	0.18	0.86
Surgical	# Coronary Grafts	0(7), 1(0), 2(0), 3(1), 4(1)		---	---
Characteristics	Surgical Procedure	Valve (10), CABG (1), CABG+VR (1)		---	---
	LVEF (SD)	54.2 (5.1)	---	---	---
	Mean Pump CPB Time, min. (SD)	145.9 (65.6)	---	---	---
	Mean Cross Clamp Time, min (SD)	83.7 (35.3)	---	---	---

^aRatio of white matter hyperintensity volume to total intracranial volume (in mL).

^bIndependent, two-sample t-test or chi-square comparisons (two-tailed).

LVEF: left ventricular ejection fraction; CPB: cardiopulmonary bypass

Perioperative Management

In surgical patients, anesthesia was induced and maintained with midazolam, fentanyl, and isoflurane or sevoflurane. All surgical patients underwent nonpulsatile hypothermic (30°-32°C) CPB with a membrane oxygenator and an arterial line filter. The pump was primed with crystalloid, and serial hematocrit levels were kept at ≥0.21. Before initiation of CPB, all surgical patients received heparin anticoagulation (300–400U/kg) to achieve a target activated coagulation time of >480s. Perfusion was maintained at pump flow rates of 2-2.4 L·min−1·m−2 throughout CPB to maintain mean arterial pressure at 50-80mm Hg. See Table 1 for surgical variable data.

Neurocognitive Procedures

Participants were administered standardized neuropsychological assessment measures designed to assess auditory-verbal learning, immediate and delayed memory recall, visual immediate and delayed memory recall, complex attention, visuomotor performance and processing speed, manual dexterity, and complex executive functioning skills (see Table 2 & Supplemental Appendix 1). All measures were administered at preoperative baseline (0.70+/-0.38 weeks prior to surgery) and ∼6-weeks postoperatively (6.56+/-1.96 weeks). There were no participant losses and cognitive testing was completed for all participants at baseline and follow-up.

Table 2. Cognition at Presurgical Baseline and Postoperative Change at 6-week Follow-up.

Cognitive Variables a	Baseline Performance			Postoperative Cognitive Change b
	Surgical (n=12)	Controls (n=12)	p	Surgical (n=12)	Controls (n=12)	Statisticc	p	d
WRAT-4 Reading Subtest	50.3 (5.8)	48.4 (6.5)	0.47	---	---	---	---	---
MMSE Total	28.6 (1.0)	28.8 (0.6)	0.58	---	---	---	---	---
RAVLT – Initial Learning	5.6 (2.1)	5.9 (1.7)	0.72	-0.35 (0.6)	-0.03 (0.9)	-1.02	0.32	-0.42
RAVLT – Total Learning	46.5 (10.8)	45.2 (11.2)	0.29	-0.87 (1.0)	0.24 (1.0)	-2.72	0.01f	-1.11
RAVLT – Delayed Recall	10.3 (3.5)	8.4 (4.4)	0.27	-0.60 (1.1)	0.04 (1.1)	-1.42	0.17	-0.58
WVR – Immediate Recall	6.1 (2.8)	8.8 (5.3)	0.14	0.34 (0.5)	0.0 (1.3)	0.40	0.41	0.34
WVR – Delayed Recall	5.3 (2.8)	5.3 (2.7)	0.99	0.1 (0.6)	0.2 (1.0)	-0.29	0.77	-0.12
WDSF – Total Correct	7.8 (2.1)	7.0 (1.5)	0.28	0.1 (0.6)	0.0 (0.7)	0.38	0.71	0.15
WDSB – Total Correct	7.0 (2.6)	6.5 (2.2)	0.61	-0.4 (0.9)	0.0 (0.9)	-1.09	0.28	-0.44
SCWT –Interference	-5.0 (9.9)	-3.6 (5.3)	0.67	-1.0 (1.3)	0.0 (1.0)	-2.11	0.04f	-0.86
TMT – Trails A (Inverse)	-30.8 (11.9)	-30.9 (9.5)	0.97	-1.1 (2.1)	0.1 (1.1)	-1.75	0.09	-0.72
TMT – Trails B (Inverse)	-115.3 (101.8)	-89.75 (41.1)	0.50	-0.7 (1.6)	0.0 (1.1)	-1.25	0.22	-0.51
DSST – Total Correct	43.33 (13.2)	51.33 (13.9)	0.16	-0.5 (1.4)	0.0 (1.0)	-1.01	0.32	-0.41
GPT – Mean Bilateral Speed (Inverse)	-100.5 (35.2)	-94.9 (40.2)	0.72	-1.0 (2.5)	0.0 (1.0)	-1.28	0.21	-0.53
Global Cognitive Changed	---	---	---	-0.5 (0.5)	0.0 (0.3)	-2.97	0.007g	-1.21

^aDuke Neurologic Outcomes Research Group (NORG) Neurocognitive Battery [Mini-Mental Status Examination (MMSE); Wide Range Achievement Test (4^th Revision) Reading Subtest (WRAT-4); Rey Auditory Verbal Learning Test (RAVLT); Wechsler Adult Intelligence Scale – 3^rd Revision & Wechsler Memory Scale - Revised Digit Span Forward (WDSF), Digit Span Backward (WDSB) & Visual Reproduction (WVR) Subtests; Stroop Color and Word Test (SCWT); Trail Making Test (TMT); Wechsler Adult Intelligence Scale – 3^rd Revision Digit Symbol Substitution Test (DSST), Lafayette Grooved Pegboard Test (GPT)]. See Supplemental Appendix 1.

^bExpressed as reliable change index (RCI)¹⁶ scores relative to non-surgical control group test-retest data. See Supplemental Appendix 1 for RCI formula and methods.

^cIndependent, two-sample t-test comparisons (two-tailed).

^dCalculated as the mean RCI value from all cognitive variables.

^eCohen's d effect-size values.

^fp<0.05

^gp<0.01

Surgical patients were screened for agitation/sedation and postoperative delirium at post-op days 1-3. While sedation and partial delirium symptoms were a factor in four of our twelve surgical patients during the postoperative period, none of these surgical patients met full Confusion Assessment Method (CAM)¹³ delirium diagnostic criteria.

Neuroimaging Procedures & Data Acquisition

Standard anatomical images and functional data were acquired on a 3T GE MR750 with an 8-channel head coil. Anatomical data consisted of high-resolution T1-weighted fast spoiled gradient-echo oblique axial acquisition [256×256 matrix, 256mm field-of-view (FOV), 11° flip angle, 136 1mm thick slices, echo time (TE) 3.0ms, repetition time (TR) 6.93ms] and T2 fluid attenuation inversion recovery (FLAIR) oblique axial acquisition [128×128 matrix, 256mm FOV, 90° flip angle, 68 2mm thick slices, TE 145.6ms, TR 11000ms, inversion time (TI) 2250ms] scans. rsfMRI data were acquired with a sensitivity encoding (SENSE) spiral-in oblique axial, slice interleaved acquisition (64×64 matrix, 256mm FOV, 60° flip angle, 34 4mm thick slices, TE 30ms, TR 3000ms, SENSE factor 2). 18sec. were discarded from the beginning of the rsfMRI sequence to correct for initial MR signal fluctuation, after which 124 time points (6.2min.) of data were retained for RSFC analysis. During the rsfMRI data collection, all subjects were instructed to look at a black crosshair centered on a white background. Physiological (i.e., respiration and heart rate) and movement data were collected during rsfMRI scanning for later RSFC data signal correction methods. Physiological data were detected using a respiration stretch transducer and finger pulse-oximeter connected to Biopac amplifers sampling at 100Hz and were synchronized to the trigger pulse associated with the first MR image acquisition using CIGAL software.¹⁴ See Supplemental Appendix 1.

Neuroimaging Data Preprocessing

Neuroimaging data were spatially preprocessed and analyzed with Statistical Parametric Mapping MATLAB software (SPM ver.8, Wellcome Institute, London). SPM default gray and white matter and CSF anatomical segmentation parameters were applied to both the T1 anatomical segmented and coregistered FLAIR and functional data for nonlinear alignment with MNI atlas space (2mm³ isotropic voxels). Functional data was spatially smoothed with a Gaussian 8mm FWHM filter. Coregistered and normalized FLAIR data were interrogated for ischemic white matter lesion volumes at each time point using the MATLAB LST toolbox for SPM8.

First-level covariate correction for each participants' rsfMRI data included composite motion greater than 0.5mm, physiological signal (i.e., respiration and heart rate), and BOLD scan-to-scan signal artifact greater than z=3.0. Linear regression of confounding effects was conducted using Artifact Detection Tools (NITRC; https://www.nitrc.org/projects/artifact_detect/) and CompCor¹⁵ to maintain temporal resolution of processed data and to avoid the induction of anti-correlations or potential DMN signal loss associated with global signal regression. Functional data were band-pass frequency filtered (0.008–0.09Hz) and session-specific temporal linear detrending occurred after confound removal regression.

Analyses

Demographic Variables & Cognitive Outcomes

Independent, two-sample t-tests or Chi-square test comparisons (where appropriate) were conducted for all demographic variable comparisons and to assess differences between groups in cognitive outcomes (two-tailed, p<0.05; see Table 1). Planned cognitive baseline and outcome comparisons involved raw scores at baseline and comparison of computed reliable change index (RCI) for each variable. RCI values, expressed as z-scores, reflect the magnitude of perioperative cognitive change controlling for expected mean performance change, practice effects and normal test-retest reliability observed in the non-surgical controls (see Supplemental Appendix 1).¹⁶ Mean total RCI, calculated as the mean of all the RCI scores for each of the 12 cognitive variables (see Table 2), was used as an aggregate measure of global perioperative cognitive change. Five of the twelve surgical patients had RCI values <-1.64 on 2 or more test variables, which has been previously used by some groups as a criterion for dichotomous determination of POCD.¹⁷

White Matter Lesion Volumes

FLAIR variables of interest were total lesion volume (mL) and the ratio of lesion volume to total intracranial volume (TIV). Independent, two-sample tests were conducted between groups for baseline lesion burden and pre-/postoperative changes in total lesion volume and corrected white matter lesion burden (p<0.05, two-tailed). Baseline TIV-corrected total white matter hyperintensity volume and TIV-corrected perioperative white matter legion volume change variables were retained as covariates for entry into our RSFC analyses (see section below). See Supplemental Appendix 1.

Intrinsic Functional Connectivity

Intrinsic Connectivity Contrast (ICC) characterizes the strength of the global connectivity pattern between each gray matter voxel and gray matter voxels in the rest of the brain. In an ICC analysis, values reflect the root sum of squares of the functional connectivity between each gray matter voxel and the remainder of gray matter voxels in the brain.^{18, 19} See Supplemental Appendix 1.

Random-effects analysis of ICC was conducted between surgical and control groups to examine any pre-surgical differences in RSFC, as well as any potential differences between groups in the relationship among baseline global cognitive performance and baseline ICC, adjusting for age, education and baseline white matter hyperintensity volume. To examine RSFC differences in relation to postoperative cognitive change, a random-effects, mixed-model analysis was conducted to interrogate for any differences in the relationship between ICC baseline and 6-week follow-up change and the subject-level global cognitive outcome variable, adjusting for participant age, education, and baseline and perioperative white matter hyperintensity lesion volume variables. Age, education and white matter hyperintensity volume covariates were mean centered and orthogonalized prior to model insertion. For all analyses, change in RSFC was expressed as 6-week follow-up minus the presurgical baseline. All neuroimaging analyses were conducted with statistical thresholds set to a false discovery rate (FDR) multiple comparison correction²⁰ with a spatial extent (k_E) of p-FDR<0.01 and peak voxel p<0.001.

Results

Demographic Variables & Cognitive Outcomes

Presurgical demographics and health variables for cardiac surgical and non-surgical ambulatory controls are noted in Table 1. No statistically significant differences were found between groups for age, education, sex or race. Baseline health factors such as weight, history of diabetes and previous history of myocardial infarction also did not differ between groups.

Baseline preoperative intellectual levels were similar between groups (see Table 2). No statistically significant baseline differences were found between groups for any of the neurocognitive battery variables (Table 2). However, the surgical group showed worsened auditory-verbal list learning abilities (t=-2.72, p=0.01) and response inhibition skills (t=-2.011, p=0.04) after surgery. The mean global RCI score was significantly lower in surgical patients than controls (t=-2.97, p=0.007).

White Matter Lesion Volume

There were no significant differences in baseline ischemic white matter burden between surgical patients and controls, or between groups for baseline to 6-week changes in white matter lesion volume (see Table 1). Among the surgical group and adjusting for age and education, baseline white matter lesion volume was not significantly associated with overall postoperative cognitive change (r=-0.48, p=0.17) or change in any of the other neurocognitive battery variables with the exception of Trail Making Test – Form A (r=-0.80, p=0.006).

Voxel-wise Intrinsic Functional Connectivity

There were no significant differences between groups in ICC at baseline or in the relationship between baseline global cognition and baseline ICC, adjusting for age, education and baseline white matter hyperintensity volume. Two regions of statistically significant positive association between perioperative ICC change and cognitive change exceeded statistical thresholds in our surgical group; neither was detected in our control group. These regions were located in the posterior division of the left posterior cingulate cortex (PCC) with two local maxima [cluster 154 k_E; 13.05T (-10x, -40y, 38z), 7.05T (-4x, -40y, 32z)] and the right superior frontal gyrus (rSFG) with a single cluster maxima [cluster 64 k_E; 12.81T (24x, 30y, 58z). (See Figure 1). Fisher-transformed r-values for these regions of significant ICC difference were extracted from the PCC and rSFG ROIs, averaged within each ROI, and then individually plotted with the global cognitive change values for participants in the surgical and non-surgical control groups to allow for visualization of the relationship between FC change and global RCI change. Our group-wise differences were observed to reflect a significant positive relationship between baseline-to-6 week postoperative functional connectivity change and global cognition change, adjusting for age, education and cerebral white matter hyperintensity volumes covariates, in the PCC (see Figure 1, A.2) and rSFG (see Figure 1, B.2) for surgical patients that was not shared by the non-surgical controls.

Figure 1. Regional Changes in Intrinsic Functional Connectivity Associated with Postoperative Cognitive Change in Cardiac Surgical Patients Relative to Nonsurgical Controls.

Note: Significant positive relationship between perioperative changes in left posterior cingulate cortex (PCC) intrinsic connectivity contrast (ICC) values and mean global postoperative cognitive change (i.e., white region in A.1 and scatterplot A.2), and perioperative changes in right superior frontal gyrus (rSFG) ICC values and mean global postoperative cognitive change (i.e., white region in B.1 and scatterplot B.2). Statistical parametric mapping (SPM) random-effects comparison of rs-fMRI voxel-wise intrinsic functional connectivity signal change (i.e., ICC) and mean global cognitive change, adjusting for age, education, baseline white matter hyperintensity volume, and perioperative white matter hyperintensity volume change [surgical patients (n = 12), controls (n =12); peak voxel threshold p<0.001, cluster spatial extent (k_E) threshold p-FDR < 0.01]. Scatterplot x-axis = perioperative change in ICC (i.e., Fisher-transformed mean r-values) in the PCC (A.2) and rSFG (B.2); y-axis = postoperative change in mean global reliable change index (RCI) values from baseline. See Supplemental Appendix 1 for greater methodological detail on ICC and RCI values and their calculation.

Discussion

We report that postoperative change in global cognitive function is associated with perioperative changes in intrinsic RSFC in DMN-associated posterior cingulate cortex (PCC)/precuneus and right superior frontal gyrus (rSFG) cortical regions (see Figure 1). Perioperative RSFC changes in these DMN regions were associated with changes in global cognitive performance, after adjusting for age, education, baseline cerebral white matter hyperintensity volume and peri-operative white matter hyperintensity volume change.

In our samples, global perioperative volumetric changes in leukoaraiosis did not significantly alter the positive relationship between RSFC change in these DMN regions and global cognitive outcome. Studies have found that perioperative leukoariaosis volume does not fully account for POCD.^{21, 22} However, it remains unclear whether strategically located perioperative leukoaraiosis, independent of lesion volume, may significantly impact resting-state or task-based functional connectivity in critical global cortical control networks, like the DMN.

Our finding of a continuous relationship between postoperative global cognitive change and RSFC changes in regions of the DMN raises the possibility that POCD and its functional neurological correlates represent a spectrum rather than a dichotomous trait. DMN activity is inversely correlated with activity in task-related brain networks such as the dorsal attention network (reviewed in ²³). Cortical association areas, like the DMN, are known to have greater dendritic spine density than primary unimodal cortices²⁴ and higher white matter organization.²⁵ The PCC/precuneus region, a central network “hub” of the DMN,²⁶ has 40% higher metabolic activity that most other brain regions,²⁷ and has been implicated in autobiographical memory to emotional salience. However, the exact role of the DMN in cognition remains unclear.²³ The pathophysiological basis for the decrease in DMN “hub” functional connectivity in association with increased postoperative cognitive decline seen in our sample is unknown. The decrease in intrinsic RSFC in these regions could correlate with or even cause difficulty in shifting between internal processes to external tasks requiring concerted cognitive effort and attention. Altered coordination between latent and active functional brain networks may explain some of the postoperative difficulties in “cognitive efficiency” described by older surgical patients.²⁸ Future studies using both resting state and task-based fMRI studies will be required to address this hypothesis.

In addition to our study small sample size, an important limitation of this study is the exclusion of participants with pre-operative cognitive impairment (i.e., MMSE >27). Thus, it is unclear to what extent these findings may generalize to patients with baseline cognitive impairment. Understanding the association between postoperative delirium, presurgical cognitive dysfunction and functional brain connectivity changes that might mediate such associations, are important questions for future studies. Furthermore, our results do not address the root cause of POCD. They do, however, indicate that in cardiac surgery patients without cognitive impairment at presurgical baseline poorer postoperative global cognitive outcomes share a significant association with perioperative decreases in intrinsic functional connectivity of regions important to the DMN. Our preliminary findings support the need for larger future studies to determine the extent to which altered postoperative resting and/or task-based functional brain connectivity may be expressed by varying POCD cognitive and behavioral phenotypes and how pre- and perioperative neurological and psychiatric complications may mediate those relationships.

Supplemental Appendix 1: Detailed study participant information, perioperative management, neurocognitive assessment procedures, characterization and measurement of cognitive change, postoperative delirium assessment procedures, and neuroimaging procedures, data acquisition and analyses.