Acute Brain Complications of Coronavirus Disease 2019

Several neurologic complications have been highlighted in the initial phase of COVID-19, suggesting direct or indirect neurotropism of the virus. SARS-CoV-2 has been, for example, associated with an increased risk of stroke, particularly in severe cases. The virus can lead to clotting abnormalities, which lead to ischemic strokes. These strokes can occur in young patients without typical risk factors, indicating a direct or indirect impact of the virus on the cerebrovascular system. Another significant acute complication of COVID-19 is encephalitis. This condition can again result from the direct invasion of the central nervous system by the virus, or from an abnormal immune response to the infection. Additionally, COVID-19 can cause a range of other acute neurologic symptoms, including Guillain–Barré syndrome (GBS). GBS is a rare immune-mediated acute polyradiculoneuritis in which the immune system attacks the nerves, leading to muscle weakness and, in severe cases, paralysis. Overall, these acute neurologic effects illustrate the possible direct or indirect effects of COVID-19 on the nervous system.

Long-Term Neurologic and Psychiatric Impacts from Other Viral Infections

Long-lasting impacts on the brain of SARS-CoV-2 are suggested by previous studies on postinfectious effects caused by other viruses. Experiences with Severe Acute Respiratory Syndrome (SARS) and Middle East Respiratory Syndrome (MERS), in particular, indicate that coronaviruses can have enduring effects on brain health, with both neurologic and psychiatric symptoms. While the full extent of these impacts is still under investigation, it is likely that coronaviruses, like other neurotropic viruses, have the potential to cause persistent and significant pathophysiological changes to the brain in survivors. In developing brains, the long-term neurologic impact of viral infections can be particularly severe. For instance, the Zika virus, which caused widespread developmental delays among newborns during the 2015 outbreak, demonstrated how viral infections can lead to serious neurologic issues, including microcephaly and cognitive delays. These effects are thought to result from both direct viral invasion of the central nervous system and the inflammatory responses triggered by the infection. In the longer term, viral infections could also act as triggers for neurodegenerative diseases, such as Alzheimer’s and Parkinson’s diseases, possibly accelerating the onset of these diseases in predisposed individuals. ^{, , ,} For example, the presence of certain viruses has been linked to increased amyloid-beta deposition in Alzheimer’s disease.

In addition to neurologic outcomes, viruses can also have significant psychiatric consequences. Adults who have survived previous coronavirus infections could have reported chronic psychiatric symptoms, including anxiety, depression, and posttraumatic stress disorder (PTSD). These symptoms may be exacerbated by the stress and trauma associated with the infection and the broader societal impacts of the pandemic, such as isolation and financial strain. Research suggests that viral infections may disrupt the normal functioning of the brain’s neurotransmitter systems, leading to these psychiatric symptoms. Children can also suffer from long-term psychiatric effects following viral infections. ^, The experience of illness, hospitalization, and separation from family can be traumatic, leading to increased rates of anxiety, depression, and behavioral problems. Additionally, the potential for neuroinflammation and direct viral effects on the brain may predispose to psychiatric disorders later in life.

The Historical Example of the Spanish Flu

The most significant historic example of long-term neurologic and psychiatric effects of viral infections is certainly provided by the 1918 Spanish flu pandemic. ^, Caused by the H1N1 influenza virus, it had a profound and lasting impact on global health, with its repercussions extending far beyond the acute phase of the illness. This pandemic not only claimed millions of lives but also left many survivors with enduring neurologic and psychiatric sequelae, providing a case study in understanding the potential long-term effects of viral infections on the brain.

One of the most striking neurologic complications associated with the Spanish flu was the development of encephalitis lethargica, also known as “sleeping sickness.” This condition was characterized by a wide range of symptoms, including lethargy, movement disorders, and, in severe cases, coma. Encephalitis lethargica led to a significant number of deaths and left many survivors with permanent neurologic damage, including Parkinsonism. The connection between the Spanish flu and encephalitis lethargica remains a subject of debate, but the temporal association and the overlapping symptoms have led many researchers to suspect that the influenza virus may have played a role in triggering this condition, either through direct viral invasion of the central nervous system or by provoking an aberrant autoimmune response that affected the brain. In addition to encephalitis lethargica, many survivors of the Spanish flu experienced a range of other neurologic symptoms that persisted long after the acute infection had resolved. These included chronic headaches, dizziness, and a general sense of malaise, which were often described as part of a “post-influenza syndrome.”

The psychiatric sequelae of the Spanish flu were also notable. Survivors of the pandemic reported high rates of psychiatric disorders, including anxiety, depression, and psychosis. There were widespread accounts of individuals who, having recovered from the flu, exhibited significant changes in personality, mood, and behavior. Some patients developed what was then referred to as “neurasthenia,” a condition marked by fatigue, irritability, and emotional instability, which, today, might be classified as a mood disorder. The relationship between the Spanish flu and psychiatric disorders is complex and likely multifactorial. On one hand, the direct effects of the virus on the brain, through either infection or inflammation, may have contributed to the development of psychiatric symptoms. Alternatively, the immense psychological stress of living through a pandemic—widespread death, uncertainty of survival, social ramifications, and economic disruptions—undoubtedly played a significant role in the observed mental health outcomes. The pandemic also occurred during a time when mental health support was rudimentary, exacerbating the long-term psychiatric toll.

The Clinical Model of Chronic Fatigue Syndrome and Myalgic Encephalomyelitis

More recently, postinfectious syndromes have been considered in the framework of chronic fatigue syndrome (CFS) and myalgic encephalomyelitis (ME) ^, that could also be extended to long COVID. These conditions share common features, such as debilitating fatigue, cognitive impairment, and a range of other symptoms, that can severely impact quality of life such as post-exertional malaise.

The pathophysiological mechanisms underlying these conditions are still being explored, but they may involve immune dysregulation, neuroinflammation, and metabolic dysfunction. The immune response triggered by the inciting virus may be characterized by an excessive immune response, leading to chronic inflammation and altered neurochemical signaling. These processes may then contribute to persistent fatigue and cognitive impairments as is seen in other viral infections that precede CFS/ME, such as Epstein–Barr virus or influenza.

Neuroimaging studies have provided valuable insights into the neurologic and metabolic changes associated with CFS and ME (for review, see Ref ). Functional MR (fMR) imaging scans have shown altered brain activity patterns in patients with CFS/ME, such as abnormal activation in regions related to cognitive processing like attention and memory. Both conditions may involve metabolic dysregulation affecting brain energy production and utilization. This hypometabolic state may reflect underlying neuroinflammation or metabolic dysfunction. Fluorodeoxyglucose (FDG)-PET imaging has revealed hypometabolism in specific brain regions, such as the frontal and temporal lobes, which are linked to cognitive function and emotional regulation. PET studies using translocator protein (TSPO) targets tracers like [11C] PK11195 have detected elevated microglial activation, suggesting a role for neuroinflammation in CFS/ME. This supports the autoimmune hypothesis that chronic inflammation, activation of microglia, and release of pro-inflammatory cytokines contribute to the persistent fatigue and cognitive impairments experienced by patients.

This potential link among long COVID, CFS, and ME underscores the complexity of post-viral syndromes and their impact on patient health. The overlapping symptoms could share pathophysiological mechanisms and neuroimaging signatures. Continued research is essential to further elucidate these connections.

Fluorodeoxyglucose Brain PET Findings

One of the key tools in understanding the neurologic and psychiatric impacts of COVID-19 has been neuroimaging, particularly the use of PET with the glucose analog 18F-fluorodeoxyglucose (18F-FDG). ^, Its ability to visualize metabolic activity provides critical insights into brain function and pathology, making it a valuable tool for clinicians in diagnosing and managing various brain disorders, including neurodegenerative diseases, epilepsy, brain tumors, and inflammatory conditions. ^, In patients with long COVID, 18F-FDG PET scans have consistently revealed a characteristic pattern of decreased glucose metabolism in specific brain regions. ^{, , ,} This individual hypometabolic profile has been observed in both adults and children.

Numerous FDG-PET studies have identified significant cerebral anomalies in adults suffering from long COVID. ^, The first retrospective analysis of brain PET scans revealed hypometabolism in several key brain regions, including the olfactory gyrus and more broadly the limbic and paralimbic regions, the brainstem and cerebellum ( Fig. 1 , Table 1 ). These findings of brain hypometabolism were supported by multiple monocentric case-control studies and a larger multicenter study, reinforcing the across different populations and settings.

Brain 18F-FDG PET hypometabolism in patients with long COVID. In comparison to healthy subjects, patients exhibit hypometabolism in the bilateral rectal/orbital gyrus, including the olfactory gyrus; the right temporal lobe, including the amygdala and the hippocampus, extending to the right thalamus; the bilateral pons/medulla brainstem; and the bilateral cerebellum .

( Adapted from Guedj E, Campion JY, Dudouet P, et al. F-FDG brain PET hypometabolism in patients with long COVID. Eur J Nucl Med Mol Imaging . Published online January 2021. doi:10.1007/s00259-021-05215-4; p-voxel < 0.001 uncorrected, p-cluster < 0.05 FWE-corrected; xjView rendering: https://www.alivelearn.net/xjview/ .)

The cerebral anomalies observed in patients with long COVID have been correlated with various clinical symptoms. Neuropsychological assessments revealed cognitive deficits associated with cerebral glucose hypometabolism, particularly in regions critical for cognitive functions, further underscoring the impact of COVID-19 on brain health. For instance, brain hypometabolism, particularly in the olfactory gyrus and limbic/paralimbic regions, has been correlated with persistent neurocognitive symptoms such as memory impairment, attention deficits, and executive dysfunction. ^, Another study identified a diffuse pattern of hypometabolism in the frontal and temporal lobes, where the degree of hypometabolism was inversely correlated with the number and duration of symptoms, as also initially reported by the first French study, suggesting that the extent of metabolic dysfunction may directly reflect the severity and persistence of clinical manifestations. Additionally, olfactory dysfunction of long COVID has been correlated to metabolic signatures in brain regions crucial for olfactory processing, further linking these cerebral anomalies to specific sensory deficits. Interestingly, the metabolism of the frontal cluster that included the olfactory gyrus was decreased more starkly in patients treated with angiotensin-converting enzyme (ACE) drugs, and higher among patients who had used nasal decongestant spray, suggesting a possible role of ACE receptors as an olfactory gateway for this neurotropism.

Moreover, FDG-PET studies reveal a dynamic and evolving metabolic profile of COVID-19, reflecting the disease progression from the acute phase to more chronic stages. During the acute and subacute phases, patients may exhibit notable hypermetabolism in critical brain regions such as the limbic/paralimbic system, brainstem, and cerebellum, likely resulting from intense inflammatory responses triggered by the virus. ^, Alternatively, this hypermetabolism may occur as a compensatory response to localized hypometabolism or as part of the brain’s adaptation to ongoing inflammatory stress. Overtime, these metabolic patterns can shift, with some regions that initially displayed hypermetabolism potentially transitioning into hypometabolism as the disease progresses.

Encouragingly, there is evidence that some of these metabolic anomalies may be reversible. A longitudinal study tracking patients over 9 months revealed minor improvements in the pons and cerebellum, suggesting very partial reversibility of these metabolic abnormalities in patients with persistent symptoms. On the opposite, this recovery was often associated with improvements in cognitive function, highlighting the brain’s resilience and capacity for metabolic normalization overtime. As patients move from the acute to chronic stages, these metabolic anomalies may fluctuate, reflecting the brain’s ongoing response to the virus and the body’s inflammatory state. The observed partial recovery, particularly in regions that initially exhibited hypermetabolism, suggests that the long-term neurologic impacts of COVID-19 might be mitigated, leading to a gradual normalization of brain function.

Concerning pediatric patients, research has shown that the cerebral metabolic profile observed in long COVID is remarkably similar to those reported in adults. Specifically, regions such as the olfactory gyrus, and more broadly limbic/paralimbic areas, and the brainstem and cerebellum were consistently affected in both children and adults. These findings highlight that, despite differences in symptom presentation between age groups, the metabolic disruptions caused by COVID-19 affect similar brain regions in both children and adults.

Overall, FDG-PET studies have provided profound insights into the brain impact of long COVID in both adults and children as individual-level biomarker. The affected areas predominantly include a network of brain regions such as the olfactory bulbs, limbic and paralimbic structures (including the amygdala, hippocampus, insula, and cingulate cortex), brainstem, and cerebellum, which are critical for cognitive, sensory, autonomic functions, and emotional regulation. The metabolic anomalies observed, including both hypometabolism and occasionally hypermetabolism, are distinct from those seen in other neurologic or psychiatric disorders, suggesting a unique metabolic signature of COVID-19’s impact on the brain. The consistent findings across different populations and age groups highlight the pervasive nature of long COVID’s brain effects and support the hypothesis that the virus may use a “nose-to-brain” pathway. The dynamic nature of these metabolic changes, with evidence of potential reversibility, offers hope for recovery but also underscores the need for continued monitoring and intervention to mitigate long-term neurologic and psychiatric consequences.

Other Brain PET Targets

While 18F-FDG PET imaging has been widely utilized to investigate the brain impacts of long COVID, other radiotracers have been mostly used to explore brain inflammation and neurochemical changes associated with long COVID. One notable area of research involves the use of PET imaging to evaluate neuroinflammation. Radiotracers which binds to the TSPO associated with activated microglia, have been employed to assess inflammatory processes in the brain. Studies have indicated elevated TSPO levels in patients with long COVID, especially within the hippocampus, prefrontal and cingulate cortex, basal ganglia, and the pons, suggesting that neuroinflammation may play a significant role in the neurologic and psychiatric symptoms experienced by these individuals. ^{, , , ,} This finding aligns with the hypothesis that COVID-19 can trigger an inflammatory response in the central nervous system.

Alternatively, research has also focused on the potential impact on neurotransmitter systems. ^{, ,} PET studies examining dopamine and serotonin receptors have provided insights into the neurochemical reductions that may accompany long COVID, particularly in relation to mood disorders and cognitive impairment. Another promising avenue of research involves the use of radioligands targeting specific neuroreceptors or proteins. For instance, studies utilizing PET tracers are conducted to assess amyloid deposition in the brains of patients. These studies will aim to determine whether COVID-19 may accelerate neurodegenerative processes, particularly in individuals with preexisting vulnerabilities. Preliminary findings suggest that some patients exhibit increased amyloid burden, raising concerns about the potential long-term consequences of COVID-19.

Whole-Body PET Imaging

Recent studies utilizing whole-body imaging techniques, particularly total-body multiparametric PET, have also provided significant insights into the long-term effects of COVID-19. Authors revealed increased FDG uptake in various non-brain tissues from a case-control study, highlighting the systemic nature of long COVID. Specifically, they observed elevated FDG uptake in the lungs, bone marrow, large joints, and vessels, with a distinct pattern of mild-to-moderate uptake in these areas among patients with long COVID. These findings were confirmed in another study with increased FDG uptake in the lungs and bone marrow. Additionally, a retrospective study focusing on vascular hypermetabolism identified aortic FDG-PET hypermetabolism in 21% of the patients with long COVID. This vascular hypermetabolism, particularly in the thoracic aorta, was associated with symptoms such as persistent chest pain and dyspnea. The findings of these various studies suggest a persistent inflammatory response in multiple organs, further supporting the role of whole-body metabolic imaging in understanding long COVID.

Furthermore, using more innovative PET targets, a study revealed persistent tissue-based T-cell activation and viral RNA presence up to 2 years postinfection, highlighting the protracted nature of immune responses in patients with long COVID. This underscores the potential for ongoing viral activity and immune dysregulation as a basis for the chronic symptoms observed in long COVID. Finally, by utilizing dynamic PET with a CD8-targeted minibody to measure CD8+ T-cell distribution in vivo, significant immune activity was found months after infection, demonstrating the potential of dynamic PET for studying the total-body immune response in patients with long COVID.

Together, these findings reinforce the importance of advanced imaging modalities in elucidating the complex pathophysiology of long COVID, paving the way for more targeted therapeutic interventions.

Anatomic and Diffusion MR Imaging

One of the most impactful MR imaging studies provided compelling evidence of both structural and functional changes in the brain by comparing pre-COVID and post-COVID scans of 394 participants from the UK Biobank. They reported reduction in gray matter (GM) thickness and contrast in the orbitofrontal and parahippocampal area along with an overall reduction of brain size in patients with post-COVID pointing toward structural alteration of the limbic system.

Studies on diffusion imaging observed aberrant WM fractional anisotropy in the thalamus that correlated with fatigue complaints or decreased WM integrity in the brainstem (also associated with fatigue). Evidence of neuroinflammatory processes has been supported by increased free-water and mean diffusivity in the WM that could be seen as proxy for local immune response. Altogether, these findings suggest structural changes both in the GM and WM of limbic and subcortical area that may come from neuroinflammatory response to the disease. Hypotheses on the neuropathological aspect of the infection suggest systemic inflammatory processes or infiltration of the blood–brain barrier.

Functional MR Imaging

Many patients show clear cognitive deficits despite the absence of apparent structural brain lesions. fMR imaging, which captures blood oxygen levels reflecting neuronal activity, is crucial in analyzing brain functions in such cases. While fMR imaging has not been clinically validated, it has led to pivotal findings in the context of COVID-19. fMR imaging studies have reported altered patterns of functional connectivity 6 to 9 months after the initial infection, varying according to the severity of respiratory symptoms in the acute phase. These differences in connectivity partly explain cognitive deficits. Overall, reduced functional connectivity between cortical areas—particularly within the dorsal attention and somatosensory motor networks—was observed, while increased cortico-subcortical connectivity (eg, putamen, cerebellum) correlated with acute severity. Overall, the authors observed that functional connectivity was lower between cortical areas, particularly within the dorsal attention and somatosensory motor networks, while cortico-subcortical (putamen, cerebellum) connectivity increased, correlating with acute severity. Alterations in networks involved in consciousness and self-awareness have also been reported, independent of the severity of the acute infection. ^, For instance, increased functional connectivity in the default mode network (DMN) has been observed in COVID-19 patients with disorders of consciousness 3 to 6 months postinfection. Similarly, higher fractional rates of dynamic functional brain states were identified in COVID-19 patients with high rates of PTSD 6 months postinfection. A study on resting-state fMR imaging conducted 2 years after infection observed increased amplitude of low-frequency fluctuation (ALFF) in subcortical areas (left putamen, right pallidum) and decreased ALFF in temporal and parietal networks. Additionally, lower regional homogeneity in the precentral and postcentral gyri was correlated with increased mental fatigue and cognitive complaints.

Are There Distinct MR Imaging/Neuropsychological Profiles?

The heterogeneity in neuropsychological profiles among patients with COVID-19 has been highlighted, with at least 3 distinct clinical phenotypes of cognitive disorders identified. ^, The most discriminant factor among these phenotypes was self-awareness of cognitive impairment. Some patients exhibited a lack of awareness of their cognitive deficits despite objective memory impairments (anosognosia), while others reported significant complaints like fatigue but demonstrated only mild attentional and executive dysfunctions. ^, A third group had normal cognitive performance, but their complaints matched their subjective experiences. These findings suggest that alterations in functional connectivity, particularly in the DMN and networks related to consciousness and self-awareness, could underpin these diverse clinical phenotypes. A significant challenge for future studies is that patients diagnosed with post-COVID syndrome often correspond to only one clinical phenotype. Consequently, the phenotype characterized by memory impairments with a lack of awareness of their deficits may not be adequately represented in these cohorts due to the absence of self-reported complaints. This could lead to an incomplete understanding of the full spectrum of cognitive impairments associated with post-COVID syndrome. Future research must consider using relevant assessments and targeted recruitment strategies to ensure all phenotypes are adequately studied.

Future Directions for MR Imaging

In summary, both structural, anatomic, and fMR imaging alterations of the brain are coherent with neurologic outcome in the limbic, subcortical, and cerebellar area. Evidence of brain functional impairment in the sensory and DMNs has been put in relationship with cognitive disorders and points to a general alteration of the systems of memory, self-awareness, and emotion processing. Findings, modalities, and acquisition scheme have, however, been quite heterogeneous and have sometime failed to reach consensus. Nevertheless, further investigating multimodal neuroimaging markers after SARS-CoV-2 infection either with anatomic, diffusion, or functional imaging is key to uncover the wide range of long COVID trajectories and provide accurate policies to take care of patients.