35. STRUCTURAL MRI AND COGNITIVE AND NEUROPSYCHIATRIC SYMPTOMS IN POST-ACUTE SEQUELAE SARS-COV-2 INFECTION (PASC)

Abstract

Introduction

Approximately 30% of COVID-19 patients exhibit symptoms of post-acute sequelae of SARS-CoV-2 infection (PASC), or long COVID, and 90 % of those present with neuropsychiatric symptoms. As of January 2023, there have been 660M confirmed cases of COVID-19 worldwide as per the World Health Organization (WHO). Here, we used structural magnetic resonance imaging (MRI) to examine differences in gray matter thickness and volume for the limbic (cingulate gyrus) and the dorsolateral prefrontal cortex ROIs between PASC (COVID+) and healthy controls (COVID-) as well as their correlates to cardiovascular risk and resilience measures. Given the high prevalence of neuropsychiatric symptoms in PASC patients, understanding structural brain changes is essential for anticipating long-term cognitive and cardiovascular outcomes. These findings have significant implications for geriatric health, as persistent brain alterations may contribute to accelerated cognitive aging and increased vulnerability to neurodegenerative diseases.

Methods

Participants and Clinical Assessments: Participants for this study were recruited from the UCLA hospital and the broader Los Angeles community. They included 36 individuals (14 males and 22 females) ranging from ages 20 to 67 years. 28 of these participants received a COVID-19 diagnosis. COVID-19 tests were not conducted at the time of the study, participants self-reported their test results along with their test date in the case of COVID + groups. Additional demographic information collected included years of education, handedness, race, and native language. Inclusion criteria for PASC included self-report symptoms of brain fog, depression, fatigue, etc. and other symptoms of PASC 6 months after the onset of symptoms and receiving a COVID-positive diagnosis.

Neuropsychiatric, Behavioral, and Neurocognitive assessments: Measures of comorbid neuropsychiatric symptoms included the 24-item the Hamilton Depression Scale (HAMD) to quantify mood symptoms, the Hamilton Anxiety Scale (HAMA), a widely used measure of anxiety symptoms, and the Apathy Evaluation Scale (AES), a measure of the severity of apathy. Measures of medical comorbidity included the Stroke Risk Factor Prediction Chart (CVRF) of the American Heart Association for rating cerebrovascular risk factors and the Cumulative Illness Rating Scale-Geriatric (CIRS-G) used for rating the severity of chronic medical illness in several organ-systems. Resilience was determined using the Connor-Davidson Resilience scale (CDRISC), as a measure of stress coping ability. All study procedures were conducted under an approval by the UCLA IRB.

Images were acquired using a Siemens 3T Prisma MRI system at UCLA's Brain Mapping Center with a 32-channel phased array head coil. Acquisition protocol was identical to the Huma Connectome Project Lifespan studies for Aging and Development[5]. Structural MRIs included T1-weighted (T1w) multi-echo MPRAGE (voxel size=0.8mm isotropic; repetition time (TR)=2500ms; echo time (TE)=1.81:1.79:7.18ms; inversion time (TI)=1000ms; flip angle=8.0o; acquisition time (TA)=8:22min) and T2-weighted (T2w; voxel size=0.8mm isotropic; TR=3200ms; TE=564ms; TA=6:35min) acquisitions with real-time motion correction.

Multimodal imaging data were visually inspected and preprocessed with the HCP minimal pipelines [5] using the BIDS-App. Utilizing T1w and T2w structural MRI data, PreFreeSurfer, FreeSurfer, and PostFreeSurfer preprocessing streams were used to obtain accurate cortical surface reconstructions for the estimation of cortical gray matter thickness and segmented using the Desikan-Killiany Atlas [5]. Subcortical volumes and intracranial volumes were also estimated following these optimized Freesurfer preprocessing streams.

Statistical analysis of gray matter thickness for group differences between COVID+ and COVID- populations was conducted using ANCOVA with age, sex, and intracranial volume as covariates on regions of interest (Figure 2, 3). Pearson correlations were also computed to determine relationships between gray matter thickness and clinical variables (Figure 2, 3).

Results

(Please see attachment for figures)

Qualitative visualization of T2-weighted MRI scans (Figure 1) shows increased white-matter hyperintensities in the cingulum, thalamic radiation, and corpus callosum of a 60-year-old COVID+ subject with severe neuropsychiatric symptoms, compared to a healthy control and a younger COVID+ subject. Cortical thickness analyses (Figure 2) revealed significantly greater thickness in the COVID+ group for both the left (p=0.03) and right (p=0.01) caudal anterior cingulate cortex (ACC), as well as the left (p=0.029) and right (p=0.02) posterior cingulate (PC). Additionally, the COVID+ group demonstrated increased gray matter volume and cortical thickness in the rostral middle frontal (RMF) and lateral orbitofrontal (LOBF) cortices (Figure 3). Correlational analyses indicated that greater right RMF and LOBF gray matter volume were associated with lower cardiovascular risk and higher resilience scores in the COVID+ group.

Conclusions

Our results report an increase in gray matter volume and thickness in PASC patients compared to the COVID- group, which agrees with the findings from Tu et al. [2], Lu et al. [3], and Besteher et al. [4]. These observed volume enlargements are attributed to a persistent neuroinflammation hypothesis as suggested by Golderg et al. [5]. Finally, in the COVID+ group higher cortical thickness was associated with greater resilience and at the same time lower cardiovascular risk. These findings suggest that persistent neuroinflammation may play a role in long-term brain health, particularly in aging populations. As neuroinflammatory processes have been implicated in neurodegenerative diseases, understanding these structural brain changes could have significant implications for geriatric health, including cognitive resilience and cardiovascular risk management over time.