Reorganization of brain connectivity in post-COVID condition A 18F-FDG PET study

doi:10.21203/rs.3.rs-4999400/v1

Background. A hypometabolic brain pattern has been reported in patients with post-COVID condition. The aim of this study was to investigate reorganization in metabolic connectivity in patients with post-COVID condition (PCC).

Results. One hundred eighty-eight patients who underwent brain ¹⁸F-FDG PET for PCC were retrospectively included from two university hospital centres. These patients were age- and sex-matched to 120 healthy controls who underwent brain ¹⁸F-FDG PET before the COVID-19 outbreak. A voxel-based group comparison between patients and controls was performed (p-voxel at 0.005 uncorrected, p-cluster at 0.05 FWE corrected). Interregional correlation analyses (IRCA) of the identified clusters as well as sparse inverse covariance estimations (SICEs) at whole-brain scaling were also conducted. Both analyses were performed at the group level for all patients and then secondarily according to the postinfection delay; 88 and 100 patients, respectively, had a delay of less than or greater than 9 months (± 9 M). Three hypometabolic clusters, namely, the right frontotemporal, right and left cerebellar, were identified from the voxel-based group comparisons of PCC patients. Within this hypometabolic PCC pattern, a modification in metabolic connectivity was observed in patients compared with controls, which was more marked in the + 9 M group than in the − 9 M group. On the other hand, the graph analysis revealed a decrease in connectivity efficiency metrics in the PCC.

Conclusions. Metabolic connectivity is modified in patients with PCC within the hypometabolic post-COVID-19 network, with lasting reorganization evolving over time, suggesting functional adaptation.

Post-COVID condition (PCC), commonly known as “long COVID”, affects approximately 10–20% of people infected with SARS-CoV-2, with more than 17 million people across Europe potentially had suffering from PCC during the first two years of the pandemic [1]. Approximately 90% of patients presented a slow recovery 2 years after the onset of symptoms, on a participative study certainly including most severe presentations, with a reduction in the number of symptoms of only 25% [2]. Among them, the cognitive complaints are often the most disabling and lasting symptoms [3]. This recovery may be supported by functional brain adaptations.

Brain ¹⁸F-FDG PET is a useful diagnostic tool for patients with PCC reporting cognitive or dysautonomia symptoms for differential diagnosis but also for assessing cerebral impairment [4]. A hypometabolic pattern involving the orbitofrontal cortex, the temporal cortex (especially the internal temporal areas), the brainstem, and the cerebellum has been reported in patients with PCC [5].

To further understand the pathophysiology of cerebral impairment, different affected brain regions can be considered as a connectome rather than an independent sum of brain regions [6]. Such connectivity analyses are usually conducted via data-driven methods such as interregional correlation analysis (IRCA) at the voxel level or with no a priori hypotheses methods such as sparse inverse covariance estimation (SICE) [7].

We hypothesize that metabolic PET connectivity analysis might reflect the functional reorganization of patients with PCC over time. In this context, this study aims to investigate reorganizations in metabolic connectivity in patients with PCC according to the postinfection delay.

We retrospectively identified patients who presented with PCC between May 2020 and December 2022, satisfied the WHO definition [1] and had a brain ¹⁸F-FDG PET scan to investigate suspected brain impairment (fatigue, anosmia/ageusia, cognitive symptoms and/or dysautonomia) at two University Hospitals [4]. This study was approved by the CEMEN (Nuclear Medicine Research Ethics Committee) with the record number CEMEN 2024-04.

These patients were age- and sex-matched with healthy controls who underwent brain ¹⁸F-FDG PET before the COVID-19 outbreak with no neuropsychiatric antecedents and normal neuropsychological tests from the two University Hospital local databases (NCT00484523 and NCT03345290).

Brain ¹⁸F-FDG PET was performed according to recent guidelines [8] and as previously described [5].

After normalization of images to the Montreal National Institute (MNI) space and smoothing with an 8 mm Gaussian filter, voxel-based group comparisons were performed with SPM 12 (https://www.fil.ion.ucl.ac.uk/spm/software/spm12/) using age, sex and University Hospital centre as covariates, proportional scaling for intensity normalization and an inclusive grey-matter mask.

First, a voxelwise group comparison was performed with SPM between the whole group of patients and the controls (p-voxel value of 0.005, uncorrected, with cluster volume corrected from the familywise error (FWE)). This voxelwise group comparison was then performed between the subgroups of patients equally dichotomized by the median time from COVID-19 infection: less than or more than 9 months (respectively, -9 M and + 9 M). The identified clusters were reported according to the anatomical automatic labelling (AAL) atlas (https://www.gin.cnrs.fr/en/tools/aal/). Second, identified cluster values from the whole group comparison were extracted via Marsbar (https://marsbar-toolbox.github.io/) and used as seeds for IRCA [9] (p-voxel value of 0.005, uncorrected, with cluster volume correction on the basis of Monte Carlo simulations [10]). Third, SICE [11] analyses were also performed. A series of nodes representing brain regions of interest (ROIs) for connectivity analysis were selected (116 ROIs from automated anatomical labelling (AAL) [12] and 3 ROIs from the brainstem) and then extracted via the Marsbar. Using the skggm package [13], the SICE method was applied to the subject-by-node/ROI matrix to estimate the metabolic connectivity matrices for each group (whole population, -9 M, + 9 M and control groups), as detailed elsewhere [11]. The Stability Approach to Regularization Selection (StARS) was applied for optimal density selection [14]. Metabolic connectivity matrices were represented in a brain template via the Nilearn package (https://nilearn.github.io/stable/). A bootstrapping procedure (N = 100) was implemented to rigorously test differences in the number of connections between two groups via multiple independent t tests. An absolute threshold of the T score of 15 was used to highlight significant differences. Finally, a graph theory analysis was carried out with BRAPH (BRain Analysis using graPH theory (http://www.braph.org/)), a software package used to perform graph theory analysis of the brain connectome, to compare the connectivity of the patient and control brains. The SICE matrices for patients and controls previously defined were used as input matrices. The tested global parameters were degree, strength, radius, diameter, eccentricity, path length, global and local efficiency, clustering nodes, transitivity, modularity, assortativity and small-worldness. Nonparametric permutation tests with 100 permutations were carried out to assess differences between groups.

Continuous and categorical variables were compared with the Mann‒Whitney and chi‒square tests, respectively. A p value ≤ 0.05 was considered significant. Statistical analyses were performed in MATLAB version 9.13.0 (R2022b).

This study included 188 patients with PCC (48.6 ±11.8 years old, 123 women, 77 and 111 from the first and second centres, respectively). Details of the -9 M and +9 M groups of patients and healthy controls are available in Table 1. Patients in the +9 M group presented more sleeping and cognitive disorders (p≤0.03).

As depicted in Figure 1, the voxelwise comparison between patients and controls revealed three significant hypometabolic clusters, one right frontotemporal, one right cerebellar, and one left cerebellar (detailed volumes in Supplemental Table 1). When dichotomized into two subgroups, more involvement of the pericentral regions was found in the +9 M group, as was the left cerebellum (Figure 1 and Supplemental Table 1). A direct comparison between the two groups of patients is available in Supplemental Figure 1.

At the voxel level represented in Figure 2, the IRCA of identified clusters showed the greatest ~~increase~~ in molecular connectivity from the right frontotemporal cluster in PCC patients compared with controls (combining increases and decreases in connectivity), which was even more marked in the more than 9 M group than in the -9 M group (detailed clusters identified with the IRCA in Supplemental Tables 2 to 4).

SICE analyses are presented in Figure 3. The upper part confirms that metabolic connectivity was modified in PCC patients compared with healthy controls, with modifications mainly affecting the hypometabolic COVID-19 pattern. The middle and lower parts show that these modifications were predominant in the +9 M group of patients, mainly between the frontal lobe and other lobes and within the frontal, temporal and cerebellar lobes. A direct comparison of the less and +9 M groups is available in Supplemental Figure 2.

The graph theory analysis identified four significant parameters when patients were compared with controls: patients with PCC presented lower average strength, local and global efficiency, and clustering than controls did (p≤0.02). No other parameter was significant. These differences were also observed when the lower and +9 M groups were compared with the control group but not when the lower and +9 M groups were compared.

This bicentric study confirms that the hypometabolic pattern in PCC patients can persist at least 9 months after infection, with modifications of metabolic connectivity within this network.

This study is the first to highlight metabolic connectome modifications in PCC patients. Our results are in line with those of previous fMRI studies [15–22] showing BOLD connectivity modifications within the right frontal, temporal and cerebellar cortices in large series of PCC patients [15, 16, 18]. Compared with fMRI, studies of connectivity with PET imaging have the advantage of providing more reproducible results with low variance, such as those obtained with BOLD signals in fMRI [23].

To study metabolic connectivity in PCC patients, methods with a priori and no a priori information, IRCA and SICE, respectively, were applied. Both methods revealed modifications in metabolic connectivity within the frontal regions and the PCC hypometabolic network (as depicted in Figs. 2 and 3), which was more marked in the group of patients with longer-duration disease (Supplemental Tables and Supplemental Fig. 2).

These modifications in connectivity within the brain regions affected by long COVID-19, combining increases and decreases in connectivity, suggest a functional adaptation of the brain [24]. Notably, the + 9 M group presented more cognitive symptoms and more hypometabolism than did the − 9 M group did, which might reflect a more severe group of patients and hence greater brain reorganization.

On the other hand, graph theory analyses at the global level revealed a decrease in global efficiency connectivity metrics in PCC patients, which was observed both in the lower and + 9 M groups of patients.

Our results confirm that the hypometabolic network in PCC patients can persist over time, with metabolic modifications observed throughout the entire course of the disease. This finding is in line with the clinical results of patients with PCC experiencing symptoms at least 1 to 2 years after infection [1, 2], corroborating the persistence of brain metabolic impairment in brain ¹⁸F-FDG PET [25]. Notably, the pericentral hypometabolic cluster seems to appear later in the course of the disease (Fig. 1). The implicated brain regions are those of the motor cortex, which could be linked to the reduction in physical activity accompanying the decrease in social interactions among PCC patients, similar to what was observed in patients during the COVID-19 lockdown [26]. The modifications of connectivity observed within the frontal lobe in patients with longer durations of symptoms (Figs. 2 and 3) could be related to the sleeping and cognitive disorders most commonly observed in the + 9 M group of patients (Table 1).

Our study has several limitations. No longitudinal design was applied in the current study, limiting the direct comparison of metabolic connectivity. In addition, the retrospective design and bicentric nature of the current study do not allow us to correlate our metabolic connectivity findings with clinical characteristics.

This bicentric study in confirmed PCC patients revealed that modifications in metabolic connectivity within the PCC hypometabolic network can be observed early in the course of the disease and evolve over time. These metabolic connectivity reorganizations could suggest an attempt at functional adaptation of the brain to metabolic impairment, even if connectivity might be less efficient. These findings are opening a portal are essential for a better understanding of the chronic aspects of PCC physiopathology as well as for guiding future strategies of rehabilitation for PCC patient care.

Conflicts of interest/competing interests

The authors disclose no potential conflicts of interest related to the present work.

Ethics approval and consent to participate

This study was approved by the CEMEN (Nuclear Medicine Research Ethics Committee) with the record number CEMEN 2024-04.

Consent for publication

Informed consent was obtained from all individual participants included in the study.

Funding

The local PET database of healthy controls was funded by APHM (regional PHRC; NCT00484523).

Authors' contributions

All authors contributed significantly to the collection, analysis and interpretation of the data (AV, MD, SH, FG, AB, EK, MC, AM, TH, EG), to the writing of the manuscript (AV, MD, EG) and to the revision of the manuscript (SH, FG, AB, EK, MC, AM, TH).

Acknowledgement

Not applicable

Availability of data and material

Data that support the findings of this study may be requested from the corresponding author (AV).

Code availability

Not applicable.

WHO COVID-19: Case Definitions. World Health Organisation website. [Internet]. 2020. https://www.who.int/publications/i/item/WHO-2019-nCoV-Surveillance_Case_Definition-2020.2
Servier C, Porcher R, Pane I, Ravaud P, Tran V-T. Trajectories of the evolution of post-COVID-19 condition, up to two years after symptoms onset. Int J Infect Dis. 2023;133:67–74.
Hampshire Adam A, Adriana A, Christina T, William, Hellyer Peter J, Valentina G, et al. Cognition and Memory after Covid-19 in a Large Community Sample. N Engl J Med. 2024;390:806–18.
Verger A, Barthel H, Tolboom N, Fraioli F, Cecchin D, Albert NL et al. 2-[18F]-FDG PET for imaging brain involvement in patients with long COVID: perspective of the EANM Neuroimaging Committee. Eur J Nucl Med Mol Imaging. 2022;s00259-022-05913–7.
Verger A, Kas A, Dudouet P, Goehringer F, Salmon-Ceron D, Guedj E. Visual interpretation of brain hypometabolism related to neurological long COVID: a French multicentric experience. Eur J Nucl Med Mol Imaging. 2022;49:3197–202.
Yakushev I, Drzezga A, Habeck C. Metabolic connectivity: methods and applications. Curr Opin Neurol. 2017;30:677–85.
Sala A, Lizarraga A, Caminiti SP, Calhoun VD, Eickhoff SB, Habeck C, et al. Brain connectomics: time for a molecular imaging perspective? Trends Cogn Sci. 2023;27:353–66.
Guedj E, Varrone A, Boellaard R, Albert NL, Barthel H, van Berckel B, et al. EANM procedure guidelines for brain PET imaging using [18F]FDG, version 3. Eur J Nucl Med Mol Imaging. 2022;49:632–51.
Lee DS, Kang H, Kim H, Park H, Oh JS, Lee JS, et al. Metabolic connectivity by interregional correlation analysis using statistical parametric mapping (SPM) and FDG brain PET; methodological development and patterns of metabolic connectivity in adults. Eur J Nucl Med Mol Imaging. 2008;35:1681–91.
Lieberman MD, Cunningham WA. Type I and Type II error concerns in fMRI research: re-balancing the scale. Soc Cogn Affect Neurosci. 2009;4:423–8.
Huang S, Li J, Sun L, Ye J, Fleisher A, Wu T, et al. Learning brain connectivity of Alzheimer’s disease by sparse inverse covariance estimation. NeuroImage. 2010;50:935–49.
Tzourio-Mazoyer N, Landeau B, Papathanassiou D, Crivello F, Etard O, Delcroix N, et al. Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain. NeuroImage. 2002;15:273–89.
Laska J, Narayan M. skggm 0.2.7: A scikit-learn compatible package for Gaussian and related Graphical Models [Internet]. Zenodo; 2017. https://doi.org/10.5281/zenodo.830033
Liu H, Roeder K, Wasserman L. Stability Approach to Regularization Selection (StARS) for High Dimensional Graphical Models. Adv Neural Inf Process Syst. 2010;24:1432–40.
Li R, Liu G, Zhang X, Zhang M, Lu J, Li H. Altered intrinsic brain activity and functional connectivity in COVID-19 hospitalized patients at 6-month follow-up. BMC Infect Dis. 2023;23:521.
Voruz P, Cionca A, Jacot De Alcântara I, Nuber-Champier A, Allali G, Benzakour L, et al. Brain functional connectivity alterations associated with neuropsychological performance 6–9 months following SARS‐CoV ‐2 infection. Hum Brain Mapp. 2023;44:1629–46.
Muccioli L, Sighinolfi G, Mitolo M, Ferri L, Jane Rochat M, Pensato U, et al. Cognitive and functional connectivity impairment in post-COVID-19 olfactory dysfunction. NeuroImage: Clin. 2023;38:103410.
Churchill NW, Roudaia E, Chen JJ, Gilboa A, Sekuler A, Ji X, et al. Effects of post-acute COVID-19 syndrome on the functional brain networks of non-hospitalized individuals. Front Neurol. 2023;14:1136408.
Barnden L, Thapaliya K, Eaton-Fitch N, Barth M, Marshall-Gradisnik S. Altered brain connectivity in Long Covid during cognitive exertion: a pilot study. Front Neurosci. 2023;17:1182607.
Arrigoni A, Previtali M, Bosticardo S, Pezzetti G, Poloni S, Capelli S, et al. Brain microstructure and connectivity in COVID-19 patients with olfactory or cognitive impairment. NeuroImage: Clin. 2024;43:103631.
Han M, He C, Li T, Li Q, Chu T, Li J, et al. Altered dynamic and static brain activity and functional connectivity in COVID-19 patients: a preliminary study. NeuroReport. 2024;35:306–15.
Del Serrano VM, Serrano-Heras G, Romero Sánchez CM, Piqueras Landete P, Rojas-Bartolome L, Feria I et al. Brain and cognitive changes in patients with long COVID compared with infection-recovered control subjects. Brain. 2024;awae101.
Verger A, Guedj E. The renaissance of functional 18F-FDG PET brain activation imaging. Eur J Nucl Med Mol Imaging. 2018;45:2338–41.
Yakushev I, Drzezga A, Habeck C. Metabolic connectivity: methods and applications. Curr Opin Neurol. 2017;30:677–85.
Horowitz T, Dudouet P, Campion J-Y, Kaphan E, Radulesco T, Gonzalez S et al. Persistent brain metabolic impairment in long COVID patients with persistent clinical symptoms: a nine-month follow-up [18F]FDG-PET study. Eur J Nucl Med Mol Imaging [Internet]. 2024 [cited 2024 Jun 24]; https://link.springer.com/10.1007/s00259-024-06775-x
Guedj E, Campion J, Horowitz T, Barthelemy F, Cammilleri S, Ceccaldi M. The impact of COVID-19 lockdown on brain metabolism. Hum Brain Mapp. 2022;43:593–7.

Table 1. Population characteristics

	Less than 9 months from the infection group n = 88	More than 9 months from the infection group n = 100	Controls N=120	P values
Age (years old)	48.7 (±13.1)	48.5 (±10.7)	49.6 (±16.1)	0.77
Sex (female)	59 (67%)	64 (64%)	73 (61%)	0.66
Time from COVID infection (days)	153.4 (±66.4)	447.0 (±168.8)	X	<0.001*
COVID diagnosis confirmed by Antigenic or PCR test (%)	85 (96.6%)	100 (100%)	X	0.06
Number of patients with:
Asthenia (%)	50 (56.8%)	63 (63%)	X	0.44
Dyspnoea, Shortness of breath (%)	51 (57.9%)	65 (65%)	X	0.37
Chest pain (%)	24 (27.3%)	18 (18%)	X	0.11
Headaches (%)	24 (27.3%)	35 (35%)	X	0.30
Sleep disorders (%)	36 (40.9%)	57 (57%)	X	0.03*
Cognitive disorders, Memory loss (%)	58 (65.9%)	90 (90%)	X	<0.001*
Anosmia, Ageusia (%)	31 (35.2%)	34 (34%)	X	0.82
Arthro-muscular symptoms (aches, arthralgia, myalgia) (%)	28 (32%)	39 (39%)	X	0.58
Total average number of symptoms/person from above symptoms : Mean, SD	3.43 (1.72)	4.01 (1.73)

*p value significant for the comparisons between the two groups of patients

Reorganization of brain connectivity in post-COVID condition A 18F-FDG PET study

Status:

Version 1

Abstract

Figures

Introduction

Methods

Results

Discussion

Conclusions

Declarations