Hostname: page-component-cd9895bd7-dk4vv Total loading time: 0 Render date: 2024-12-27T22:24:31.232Z Has data issue: false hasContentIssue false

Pediatric CNS Inflammation and Infection: A Review of Immunopathology and Radiology

Published online by Cambridge University Press:  23 May 2024

Vivek Pai*
Affiliation:
Division of Neuroradiology, Department of Diagnostic and Interventional Radiology, The Hospital for Sick Children, Toronto, ON, Canada Department of Medical Imaging, University of Toronto, Toronto, ON, Canada
Shivaprakash Basavanthaiah Hiremath
Affiliation:
Department of Neuroradiology, London Health Sciences Centre - Victoria Hospital, London, ON, Canada
Manohar Shroff
Affiliation:
Division of Neuroradiology, Department of Diagnostic and Interventional Radiology, The Hospital for Sick Children, Toronto, ON, Canada Department of Medical Imaging, University of Toronto, Toronto, ON, Canada
*
Corresponding author: V. Pai; Email: vivek.pai@sickkids.ca
Rights & Permissions [Opens in a new window]

Abstract:

The innate and adaptive immune systems are critical in defense against pathogens and ensuring homeostasis. The central nervous system (CNS) was initially considered to be impermeable to immune cells due to the blood–brain barrier. However, this has now been debunked, with modern research delineating immune cell trafficking within the CNS, ensuring constant immune surveillance. However, these defenses may be breached in infections, which trigger an inflammatory cascade causing tissue damage. In addition, autoimmune conditions and genetic mutations may also lead to sustained proinflammatory molecule release causing significant CNS damage. Ensuing brain injury from most immune triggers is varied but may be associated with common patterns by virtue of a shared immune driver. MRI plays an important role in identifying these conditions and further enables understanding of their pathophysiology as well as their spatial predilection in the brain. In this review, we discuss basic immunology, the major CNS barriers to infections as well as the current understanding of selected pediatric infections and inflammatory processes.

Résumé :

RÉSUMÉ :

Les infections et l’inflammation du système nerveux central chez les enfants : revue de l’état des connaissances en immunopathologie et en radiologie.

Les systèmes immunitaires naturel et acquis jouent un rôle crucial dans la défense de l’organisme contre les agents pathogènes et par le fait même dans l’équilibre homéostatique. On croyait autrefois que le système nerveux central était impénétrable aux cellules immunitaires en raison de la barrière hématoencéphalique (BHE), mais cette hypothèse a été infirmée grâce aux techniques modernes de recherche sur la circulation des cellules immunitaires dans le SNC, qui assurent une surveillance immunitaire continue. Toutefois, ces moyens de défense peuvent subir des brèches en cas d’infection, ce qui déclenche une série de réactions inflammatoires qui entraînent elles-mêmes des lésions tissulaires. En outre, certaines affections auto-immunes ou mutations génétiques peuvent également provoquer la libération continue de molécules pro-inflammatoires, cause d’une atteinte importante au SNC. Les lésions cérébrales découlant de la plupart des déclencheurs immunitaires sont diverses, mais elles peuvent être associées à des manifestations communes en raison d’un même facteur immunitaire. L’imagerie par résonance magnétique joue un rôle important dans la reconnaissance de ces affections, ce qui permet de mieux comprendre leur physiopathologie ainsi que leur répartition spatiale particulière dans le cerveau. Il sera donc question, dans l’article, de notions fondamentales en immunologie, des principales barrières du SNC aux infections, de même que des connaissances actuelles sur certaines infections et certains processus inflammatoires chez les enfants.

Type
Review Article
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
© The Author(s), 2024. Published by Cambridge University Press on behalf of Canadian Neurological Sciences Federation

Introduction

Inflammation is the cardinal host defense response to invading microbes and injury. This host response has a clear benefit when under siege, particularly when regulated. However, sustained or inappropriate inflammation leads to multiple deleterious effects on the native tissue in the central nervous system (CNS). Neuroinflammatory disorders and neuroinfections are a broad category of disease processes, often with common mechanistic end points as a result of similar immunopathology. In this review, we briefly discuss key concepts of the immune system, with respect to the CNS, and the current understanding of important pediatric infections and inflammatory conditions.

Innate and adaptive immunity – revisiting basic concepts

Innate immunity (comprised of phagocytic cells, epithelial and endothelial cells, natural killer [NK] cells, innate lymphoid cells and platelets), the first line of defense, is responsible for the prompt detection of pathogens and homeostatic disruptions. Reference DeMaio, Mehrotra and Sambamurti1Reference Aristizábal, González, Anaya, Shoenfeld, Rojas-Villarraga and tal3 This is achieved with pattern recognition receptors (PRRs). Reference Kigerl, de Rivero Vaccari, Dietrich, Popovich and Keane2 Multiple subfamilies of PRRs have been identified including Toll-like receptors (TLRs), Nod-like receptors (NLRs), C-type lectin receptors and RIG-like receptors (RLRs). Reference Kigerl, de Rivero Vaccari, Dietrich, Popovich and Keane2 PRRs detect nonspecific “patterns” or structures located on pathogens, collectively called pathogen-associated molecular patterns (PAMPs). Reference Kigerl, de Rivero Vaccari, Dietrich, Popovich and Keane2 Upon PAMP recognition, PRRs initiate signaling pathways that induce antimicrobial effects and incite inflammation. Reference Medzhitov and Janeway4 PRRs also respond to “danger” signals or damage associated molecular patterns (DAMPs/alarmins) which originate from endogenous molecules released into the cytoplasm after CNS injury (e.g., heat shock proteins, hyaluronan, uric acid, galectins and thioredoxin) Reference Kigerl, de Rivero Vaccari, Dietrich, Popovich and Keane2 (Fig. 1).

Figure 1. Mechanism of action of innate immunity. Identification of PAMPs by PRRs, in this case toll-like receptors/TLRs, leads to activation of signalling pathways which cause production of antimicrobial molecules and incite inflammation. DAMPs on the other hand are released from dead and/or dying host cells causing a sterile inflammatory response. (Created with BioRender.com). DAMPs = damage associated molecular patterns; PAMPs = pathogen-associated molecular patterns; PRRs = pattern recognition receptors; TLRs = toll-like receptors.

Adaptive immunity provides delayed, but specific and long-lasting immunity. Its major components include B lymphocytes (humoral response) and T lymphocytes (cell-mediated responses). Reference DeMaio, Mehrotra and Sambamurti1 B cells react to specific antigen-stimulating proteins and synthesize immunoglobulins to neutralize pathogens. Reference DeMaio, Mehrotra and Sambamurti1 T-cells are subclassified into CD8 + and CD4 + cells, based on their functions and ability to recognize major histocompatibility complexes (MHC). Reference DeMaio, Mehrotra and Sambamurti1 CD8 + cells provide host defense against intracellular pathogens (e.g., viruses) as they detect surface antigens on infected cells. Reference DeMaio, Mehrotra and Sambamurti1 Upon target binding, CD8+ cells release lytic granules (perforins and granzymes) which perforate cell membranes, thus promoting apoptosis. Reference DeMaio, Mehrotra and Sambamurti1 CD8 + T cells also release cytokines (IFN-γ, TNF-α, TNF-β) which inhibit viral replication, activate macrophages and upregulate MHC class I expression (increasing T-cell detection). Reference DeMaio, Mehrotra and Sambamurti1 CD4 + T cells were grouped into Th1 cells (secreting pro-inflammatory molecules e.g., IFN-γ, TNF-α, TNF-β, IL-1β) and Th2 cells (secreting anti-inflammatory molecules e.g., IL4, IL-5, IL-6, IL-10, IL-13). Reference DeMaio, Mehrotra and Sambamurti1 Recently, more CD4+ cells have been identified (e.g., Th9, Th17, Th22, T-follicular cells, Treg cells). Reference DeMaio, Mehrotra and Sambamurti1 Most of these secrete proinflammatory cytokines leading to autoimmune and inflammatory diseases. Reference DeMaio, Mehrotra and Sambamurti1 A diagrammatic summary of the various components of innate and adaptive immunity is provided in Figure 2.

Figure 2. Diagrammatic representation of the components of innate and adaptive immunity. (Created with BioRender.com).

Innate immunity of the CNS: role of the microglia

The innate immune response within the CNS is mediated by resident tissue macrophages, known as microglia. Reference Goldman and Prinz5 These are derived from primitive yolk sac myeloid progenitors and constitute about 10%–15% of the cell population in the brain. Reference Goldman and Prinz5,Reference Subramaniam and Federoff6 Microglia play a vital role in mediating multiple regulatory processes within the brain which are critical for tissue development and maintenance of the neural homeostasis while playing a key role in response to brain injury and tissue repair. Reference Subramaniam and Federoff6 In physiological conditions, microglia acquire a relatively inactive phenotype (resting microglia). Reference Goldman and Prinz5 In this quiescent state, resting microglia demonstrate long cytoplasmic protrusions, a stable cell body with minimal to no movement. This state is regulated by intrinsic factors (e.g. Runx1 [Runt-related transcription factor 1], Irf8 [Interferon regulatory factor 8]) and extrinsic factors (e.g., TREM2 [triggering receptor also expressed on myeloid cells-2], chemokines [CX3CR1 and CD200R]). Reference Subramaniam and Federoff6 Additional factors that maintain the resting state of microglia include signals from healthy neurons (CX3CL1, neurotransmitters, neurotrophins and CD22) as well as microRNA-124 (expressed by microglial themselves) which reduces CD46, MHC-II and CD11b which enables maintenance of the quiescent state. Reference Subramaniam and Federoff6 Upon exposure to endogenous stimuli, microglia are ‘activated’. There are two polar states of activation, i.e., M1 and M2 phenotypes, which are associated phenomenologically with tissue injury and homeostasis, respectively. Reference Goldman and Prinz5,Reference Subramaniam and Federoff6 Typically activated microglia acquire the M1 phenotype and secrete proinflammatory and pro-killing cytokines (IL-1β, IL-6, IL-12, IL-17, IL-18, IL-23, TNF-α, IFN-γ, nitric oxide and CCL2), hence serving as the first line of CNS defense effectively purging out pathogens and prompting T-cells to elicit an adaptive immunological response. Reference Subramaniam and Federoff6 Alternatively, polarization into the M2 phenotypical state may, though poorly understood, is linked with immunoregulation, suppression of inflammation, tissue repair and injury resolution. Reference Subramaniam and Federoff6 Microglia in the M2 state lead to production of IL-4 and IL-10 via Arg1 (arginase 1), Ym1 (chitinase-like protein) and PPAR (peroxisome proliferator-activated receptor). Reference Subramaniam and Federoff6 The process of microglial activation is depicted in Figure 3.

Figure 3. Schematic depicting microglial polarization states and function. Under physiological conditions microglia perform active surveillance of the CNS (blue pathway). Multiple factors (e.g., colony stimulating factor 1 receptor [CSF1R], signal regulatory protein CD172 [SIRP1A], chemokine [CX3CL1 and CD200R]), help maintain this state. When triggered (red pathway) (e.g., exposure to lipopolysaccharide [LPS], IFN-γ, or GM-CSF), microglia acquire M1 phenotype (pro-inflammatory) which leads to neurotoxicity. Alternate pathway activation by IL-1, igG, or IL-10 causes microglia to attain an anti-inflammatory (M2) state (green pathway) ensuring neuroprotection. (Created with BioRender.com). CNS = central nervous system; GM-CSF = Granulocyte-macrophage colony-stimulating factor.

Demystifying “immune privilege” theory of the CNS

Historically, the CNS was erroneously considered to be “immune privileged”. Reference Louveau, Harris and Kipnis7 This concept originated from studies that demonstrated the successful survival of implanted neoplasms in neuroparenchyma but not in other tissues. Reference Galea, Bechmann and Perry8 Subsequently, microorganisms and vectors were thought to successfully evade immune recognition in the brain. Reference Galea, Bechmann and Perry8 The contemporaneous understanding of the blood–brain barrier (BBB) eventually deemed it to be responsible for this unique privilege. Reference Galea, Bechmann and Perry8 The BBB was thought to limit immune surveillance offered by peripheral lymphocytes in normal physiological conditions based on the fact that the CNS lacked antigen-presenting cells (APCs) and barely expressed MHC molecules. Reference Waisman, Liblau and Becher9 With an improved understanding of CNS immunity, this dogma has now been modified. Reference Waisman, Liblau and Becher9,Reference Clatworthy10 While there are no APCs in the brain, they exist along CNS vasculature. Reference Waisman, Liblau and Becher9 Neural-derived antigens are identified and released from the CNS and are found in deep cervical lymph nodes draining the CNS. Reference Waisman, Liblau and Becher9 Thus, immune responses can be mounted in these lymphoid structures against antigens within the CNS. Reference Waisman, Liblau and Becher9

Barriers to CNS infection

The CNS is protected by an elegant infrastructure of barriers that impede entry and spread of pathogens. In the following discussion, we elaborate on the role of microglia and the BBB in CNS immunity. We also briefly discuss the physiology of other barriers to CNS infection.

The blood–brain barrier (BBB)

The BBB forms the physical and immunological barrier in the brain parenchyma. The BBB is composed of endothelial cells expressing tight junctions (TJs), basement membrane and pericytes Reference Ampie and McGavern11 (Fig. 4). It prevents the entry of micro-organisms as well as neurotoxic components in plasma. Reference Sweeney, Zhao, Montagne, Nelson and Zlokovic12 A variety of enzymes, transporters and efflux pumps are also present, thus ensuring optimal transport of nutrients into the CNS and toxic metabolites out of it. Reference Mapunda, Tibar, Regragui and Engelhardt13 The TJs of the BBB are composed of transmembrane proteins such as claudin-5, occludin and junctional adhesion molecules (JAM). Reference Mapunda, Tibar, Regragui and Engelhardt13 Claudin-5 forms a diffusion barrier for small molecules while occludin regulates the movement of calcium across the BBB in addition to ensuring TJ stability. Reference Mapunda, Tibar, Regragui and Engelhardt13 Some authors suggest that JAM-A, JAM-B and JAM-C play a role in regulating the BBB stability and the migration of immune cells across it. Reference Mapunda, Tibar, Regragui and Engelhardt13 Other transmembrane proteins include platelet endothelial cell adhesion molecule 1 (PECAM-1) and CD99 which contribute to vascular integrity and mediate leukocyte trafficking, respectively. Reference Wimmer, Tietz and Nishihara14,Reference Bixel, Kloep, Butz, Petri, Engelhardt and Vestweber15

Figure 4. Anatomy of the blood–brain barrier. (Created with BioRender.com).

Immune trafficking across the BBB

Immune surveillance of the CNS is carried out by peripherally activated T-cells which traverse the BBB to gain access to the perivascular and subarachnoid spaces, even in normal physiological states. Reference Meyer, Martin-Blondel and Liblau16 This is due to the special properties of the BBB, which, unlike peripheral endothelium, lacks P-selectin protein and expresses an atypical chemokine receptor 1 (ACKR1). Reference Mapunda, Tibar, Regragui and Engelhardt13 The physiological immune entry is restricted to just T-cells which are independent of these trafficking cues. Reference Mapunda, Tibar, Regragui and Engelhardt13 T-cell entry is a complex, multistep process involving initial tether (capture) onto the endothelium, followed by selectin-facilitated rolling, effectively reducing the speed of the T-cells. Reference Nourshargh and Alon17 Chemotactic cues are later recognized allowing integrin-mediated cell arrest, crawling and finally diapedesis across the BBB. Reference Nourshargh and Alon17

During neuroinflammation, endothelial cells upregulate vascular cell adhesion molecule-1 (VCAM-1) and intercellular adhesion molecule 1 (ICAM1). These interact with αLβ2 and α4β1 on CD4+ T-cells, resulting in T cell arrest. Reference Kadry, Noorani and Cucullo18 Multiple studies have proposed that activated leukocyte cell adhesion molecule (ALCAM), melanoma cell adhesion molecule (MCAM) and nerve injury-induced protein (ninjurin-1) also play a role in the migration of T-cells across the BBB in conditions such as autoimmune encephalomyelitis and multiple sclerosis. Reference Mapunda, Tibar, Regragui and Engelhardt13,Reference Ifergan, Kebir and Alvarez19 It must be mentioned that the precise migration mechanisms of other immune cells (B cells, neutrophils, monocytes and dendritic cells) across the BBB are not fully understood. Reference Mapunda, Tibar, Regragui and Engelhardt13 B cell migration is thought to be mediated by endothelial ICAM-1 with α4β1integrin, chemokines (CCL2 and CXCL8) and ALCAM. Reference Marchetti and Engelhardt20

Exit pathway of T-cells

Interstitial fluid flows into the CSF across a paravascular route, termed glymphatics. CSF is eventually resorbed by arachnoid granulations into the dural venous blood. Reference Meyer, Martin-Blondel and Liblau16 Lymphatic vessels have been identified draining the CNS in close association with dural venous sinuses. Reference Louveau, Smirnov and Keyes21,Reference Aspelund, Antila and Proulx22 These lymphatics drain the interstitial fluid and CSF into deep cervical lymph nodes and contain immune cells, which suggests T-cells and MHC class II-positive myeloid cells are drained into peripheral lymph nodes. Reference Meyer, Martin-Blondel and Liblau16

Dural lymphatics express CCL21 chemokines which contribute to the migration of CCR7-expressing cells (dendritic cells and some T cells). Reference Louveau, Smirnov and Keyes21,Reference Aspelund, Antila and Proulx22 Hypothesizing from this, the lymphatic network is thought to permit the migration of antigen-presenting cells (APCs) into cervical lymph nodes, thereby initiating a Tcell response against antigens, including autoantigens. Reference Meyer, Martin-Blondel and Liblau16 This is seconded by the reduced severity of experimental autoimmune encephalomyelitis, following the removal of cervical lymph nodes. Reference Meyer, Martin-Blondel and Liblau16,Reference van Zwam, Huizinga and Heijmans23

Other barriers to CNS infection

Skull and meninges

The skull vault forms an obvious physical barrier, protecting the brain from exogenous pathogens. Reference Clatworthy10 Defects in the skull vault, including remnant embryonic connections, can be a source of entry for infections. Reference Malik and Shroff24 Within diploic spaces of the skull, pockets of marrow are present, which connect to the dural vasculature via diploic veins. Owing to a chemokine gradient between marrow and dura, egress of marrow-derived B cells takes place. Reference Ampie and McGavern11 B cell population expresses surface receptors which on engagement are activated and differentiate into antibody-secreting plasma cells. Reference Clatworthy10 The dural blood vessels are devoid of TJs and are lined by meningeal macrophages. Reference Ampie and McGavern11 Additional dural immune cells (T cells, B cells, dendritic cells, plasma cells, innate lymphoid cells and mast cells) are also encountered in peri-sinus spaces. Reference Ampie and McGavern11 Circulating microorganisms are most likely to exit across the fenestrated dural vessels hence explaining the need for a plethora of immune cells within the dura. Reference Ampie and McGavern11

Unlike the dural vasculature, blood vessels in the arachnoid mater demonstrate TJs. Reference Ampie and McGavern11 Nevertheless, microorganisms can still infect the CSF, particularly when the walls of blood vessels are infected, initiating a chain of inflammatory responses (meningitis). Reference Ampie and McGavern11 Fibroblasts within the arachnoid mater are also connected via TJs which limit the free diffusion of solutes and other water-soluble substances. Reference Mapunda, Tibar, Regragui and Engelhardt13 They express efflux pumps which enable adequate transport of toxins out of the CSF. Reference Mapunda, Tibar, Regragui and Engelhardt13 Pial blood vessels that penetrate the brain also demonstrate TJs and are defended by perivascular macrophages Reference Ampie and McGavern11 (Fig. 5)

Figure 5. Pott’s puffy tumor is a prototypical example of infection breaching multiple CNS barriers to gain access to the brain parenchyma. This example demonstrates Pott’s puffy tumor secondary to complicated frontal sinusitis in a 14-year-old year old presenting with fever, facial swelling and altered sensorium. Coronal fat-saturated T2-weighted sequence (A) through the face reveals heterogeneous contents within the left frontal sinus (arrow), obstructing the left fronto-ethmoid drainage pathway. Hyperintense mucosal thickening is also seen in the left maxillary sinus. Axial fat-saturated T2-weighted (B), diffusion-weighted (C), and contrast-enhanced axial 3D-T1 (D) sequences reveal multicompartment extra-calvarial and intracranial infective collections. A large, loculated heterogeneous abscess, with restricted diffusion, is seen in the frontal scalp tissues demonstrating peripheral enhancement and multiple septations within (arrows in B-D). Note the ill-defined T2-weighted hyperintensity in the frontal bone, representing osteomyelitis, with a focal bony breech (thin arrow in B). Intracranially, similar collections are seen in the right frontal epidural space (stepped arrows in B-D) and the right frontoparietal subdural spaces (dotted arrow in B-D) representing empyemas, as well as within the right frontal lobe parenchyma (arrowheads in B-D) representing abscesses. Contrast-enhanced axial FLAIR sequence (E) also reveals enhancement within the right parietal sulcal spaces, suggestive of meningitis (curved arrow). Phase contrast MR venography (F) reveals loss of flow signals along the rostral segment of the superior sagittal sinus due to infective thrombophlebitis. Overall, these findings are a prototypical representation of microorganisms breaching multiple physical and physiological CNS barriers. CNS = central nervous system; FLAIR = Fluid Attenuated Inversion Recovery.

Blood-cerebrospinal fluid barrier (BCSFB)

The choroid plexus projects into the ventricles of the brain and produces CSF. Blood vessels within the CP are devoid of TJs, thus permitting extravasation of blood-borne molecules into the CP stroma. Reference Ampie and McGavern11,Reference Mapunda, Tibar, Regragui and Engelhardt13 However, the CP stroma is patrolled by numerous cells of innate and adaptive immunity even under normal physiological states. Reference Ampie and McGavern11,Reference Mapunda, Tibar, Regragui and Engelhardt13 Additionally, the layer of epithelial cells lining the CP has TJs (occludin, claudin-1, -2, -3, -11, JAM-A and zonula occludens-1, -2, -3) which form a barrier between blood and CSF Reference Ampie and McGavern11,Reference Mapunda, Tibar, Regragui and Engelhardt13 (Fig. 6). In a pathological state, the integrity of this epithelial layer may be disrupted allowing pathogens to access the ventricular CSF, thereby inciting an inflammatory response known as ventriculitis. Reference Ampie and McGavern11

Figure 6. Diagrammatic representation of the choroid plexus demonstrating the blood-CSF barrier. Blood vessels within the choroid plexus are devoid of tight junctions. However, the outer epithelial layer demonstrates tight junctions which limits and monitors egress of molecules into the CSF. The subependymal space is patrolled by cells of innate and adaptive immunity even under normal physiological states. (Created with BioRender.com). CSF = cerebrospinal fluid.

Nasal epithelium and olfactory bulbs

Olfactory sensory neurons are located in the upper nasal airway and contain nonmotile sensory cilia (which bind to odor molecules) and an axon (responsible for signal transmission to the brain). Reference Patel and Pinto25 The axons traverse the cribriform plate and synapse with neurons in the olfactory bulb. Reference Patel and Pinto25 Given its location, the nasal mucosa is constantly exposed to air-borne microbes. However, the nasal mucosa is protected by macrophages, T- and B-cells. Nevertheless, certain neurotropic viruses may evade these defenses and travel along the olfactory sensory neurons across the cribriform plate to gain access to the olfactory bulbs, hence bypassing other CNS defenses. Reference Ampie and McGavern11 In response, the immune system attempts to purge these microbes while limiting cellular damage by releasing non-cytolytic cytokines and interferons (IFN-γ by CD4+ and 8+ as well as natural killer/NK cells). Reference Ampie and McGavern11 Once cleared, CD8+ cells establish permanent residency in the olfactory system, preventing re-infection. Reference Ampie and McGavern11

Selected radio-pathological considerations

In this section we discuss some of the current understanding of rather uncommon infections and neuroinflammatory disorders we have encountered in our clinical practice.

Infections

Parechovirus encephalitis

The Parechovirus A (PeV-A) or the Human Parechovirus (HPeV), is a single-stranded RNA virus belonging to the Picornaviridae family. Reference Sridhar, Karelehto, Brouwer, Pajkrt and Wolthers26,Reference Sarma, Hanzlik, Krishnasarma, Pagano and Pruthi27 Of the 19 known genotypes, PeV-A3 commonly affects young infants. Reference Sarma, Hanzlik, Krishnasarma, Pagano and Pruthi27 It is contracted via respiratory or gastrointestinal routes; CNS infection is secondary to hematogenous dissemination. Reference Sarma, Hanzlik, Krishnasarma, Pagano and Pruthi27 Maternofetal transmission is not uncommon (30%-50%). Reference Sarma, Hanzlik, Krishnasarma, Pagano and Pruthi27

Symptomatology may vary from minor gastroenteritis or respiratory symptoms to fever, rash, irritability and sepsis with occasional CSF pleocytosis. Reference Sarma, Hanzlik, Krishnasarma, Pagano and Pruthi27 Children < 2 years of age are at the highest risk of infection, with illness being particularly severe in children < 6 months of age (27). Reference Sarma, Hanzlik, Krishnasarma, Pagano and Pruthi27

Following infection, innate and adaptive immune responses are triggered. PRRs (TLRs, retinoic acid inducible gene I, and melanoma differentiation-associated gene 5) form the main defense mechanisms. Reference Sridhar, Karelehto, Brouwer, Pajkrt and Wolthers26 TLR7 and TLR8 identify the viral RNA and activate the NF-κβ signalling pathway leading to the secretion of pro-inflammatory cytokines (IFN-β, TNF-α, IL-6). Reference Sridhar, Karelehto, Brouwer, Pajkrt and Wolthers26 The role of T-cell immunity in parechoviral infection is not understood, though an increased T-cell proliferation has been noted in response to PeV-A1 capsids. Reference Sridhar, Karelehto, Brouwer, Pajkrt and Wolthers26 Owing to the immature immune system in neonates, the elevated levels of pro-inflammatory cytokines and systemic inflammatory response have been linked to the severe pathogenesis of parechoviral infection. Reference Sridhar, Karelehto, Brouwer, Pajkrt and Wolthers26 Neonatal sepsis (systemic inflammatory response syndrome) is a severe, but common, manifestation of PeV-A3 and is associated with increased vascular permeability and reduced BBB function, allowing leakage of the virus into the CNS. Reference Sridhar, Karelehto, Brouwer, Pajkrt and Wolthers26 Preferential neuroaxonal tropism may be one cause of the distinct imaging pattern of the white matter involvement. Reference Westerhuis, Koen and Wildenbeest28 Given the radiating pattern of involvement, the hypothesis of perivenular invasion or venous ischemia has also been proposed. This is supported by the postmortem examinations of two premature neonates, in which only periventricular inflammatory changes with marked macrophage infiltration and moderate astrocytosis were observed. Reference Sarma, Hanzlik, Krishnasarma, Pagano and Pruthi27,Reference Bissel, Auer and Chiang29

On cranial ultrasound, hyperechogenicity in the periventricular and deep white matter may be seen. On MRI, diffuse, multifocal signal abnormality and restricted diffusion in the subcortical white matter, with a frontal predominance, is a common finding Reference Sarma, Hanzlik, Krishnasarma, Pagano and Pruthi27 (Fig. 7). Branching T1- and T2 shortening along the deep medullary veins is also a common finding. Reference Sarma, Hanzlik, Krishnasarma, Pagano and Pruthi27 Thalamic involvement may be seen while the basal ganglia, hippocampal formations and posterior fossa structures may be spared. Reference Sarma, Hanzlik, Krishnasarma, Pagano and Pruthi27 Additionally, a lactate peak may be encountered on MR spectroscopy; however, contrast-enhanced MRI, MR angiography and MR venography are typically normal. Reference Sarma, Hanzlik, Krishnasarma, Pagano and Pruthi27

Figure 7. Parechoviral infection in a neonate presenting with fever and seizures. Diffusion-weighted imaging (DWI; A-F) reveals near symmetric restricted diffusion (ADC not shown) extensively involving the cerebral white matter bilaterally. Note the linear, radiating pattern of the signal abnormality (arrows) towards the subcortical and juxta-cortical white matter. Restricted diffusion is also noted in the corpus callosum (dotted arrows in C) and the dorsolateral thalami (dotted arrows in D), suggestive of preWallerian degeneration ADC = apparent diffusion coefficient.

Cerebral malaria

Malaria is a mosquito-borne disease, transmitted through the bite of a Plasmodium-infected female Anopheles mosquito. While multiple species are known to cause malaria, P. falciparum and P. vivax are the main cause of complications in humans, including cerebral malaria (CM). Reference Schiess, Villabona-Rueda, Cottier, Huether, Chipeta and Stins30 CM commonly presents with impairment in consciousness and encephalopathy, with or without seizures, eventually progressing to coma. Reference Schiess, Villabona-Rueda, Cottier, Huether, Chipeta and Stins30 Despite treatment, focal neurological signs and symptoms may persist, with signs of brainstem dysfunction being common in pediatric survivors. Reference Schiess, Villabona-Rueda, Cottier, Huether, Chipeta and Stins30

Understanding the pathogenesis of CM has been confounded by the fact that plasmodium-infected red blood cells (PRBCs) remain within the vascular lumen and do not physically damage the parenchyma. Reference Schiess, Villabona-Rueda, Cottier, Huether, Chipeta and Stins30 It is now known that cerebral microcirculatory disturbances cause the clinical manifestations of CM Reference Shikani, Freeman, Lisanti, Weiss, Tanowitz and Desruisseaux31 PRBCs undergo a process known as sequestration, by virtue of which they adhere to the endothelium with the help of a specific cell-surface ligand (P. Falciparum erythrocyte membrane protein 1). Reference Shikani, Freeman, Lisanti, Weiss, Tanowitz and Desruisseaux31 It is encoded by a variable gene (var-gene) and interacts with host adhesion receptors (ICAM-1, EPCR, CD36). Reference Schiess, Villabona-Rueda, Cottier, Huether, Chipeta and Stins30 Sequestration subsequently leads to multiple downstream cerebrovascular effects such as vasoconstriction, cerebral blood flow impediment, luminal obstruction and BBB disruption. Reference Shikani, Freeman, Lisanti, Weiss, Tanowitz and Desruisseaux31 Apart from these mechanical effects, an alternate hypothesis related to cytokine storm has also been proposed based on the observation of increased neutrophil activation and increased production of cytokines (TNF, IFN-γ, IL-2, IL-6, IL-8 and IL-10) Reference Schiess, Villabona-Rueda, Cottier, Huether, Chipeta and Stins30 (Fig. 8).

Figure 8. Pathophysiology of cerebral malaria. The rupture of infected red blood cells leads to the release of merozoites which activate macrophages and dendritic cells which in turn causes the release of TNF and IFN-γ. These increase the expression of VCAM1 and ICAM1 on the endothelial cells in the BBB which further recruits more infected RBS, hence promoting inflammation and stimulating pro-coagulation pathways. PRRs detect malarial PAMPs and trigger the secretion of cytokines including IL-6, IL-12, and TNF. This also leads to increased recruitment of basophils, neutrophils, and NK cells. Upon activation these cells release chemokines (MIP-1 α and MIP-1β) which damage the integrity of the BBB. In addition, the endothelial cells of the BBB present the malarial antigen to CD8+ T cells in the brain which incites CD8+ T cell–mediated endothelial cell death (mediated by granzyme B). Following BBB breakdown, parasite-derived proteins and peripheral immune cells gain access to the brain. This triggers astrocyte and microglial activation which increases proinflammatory cytokines and chemokines production (e.g., MAP kinase-mediated cytotoxic pathway) which causes axonal loss. (Created with BioRender.com). BBB = blood–brain barrier; NK = natural killer; PRRs = pattern recognition receptors.

The neuroimaging features of CM are variable, especially due to isolated case reports and small case series. Typical imaging findings include diffuse neuroparenchymal swelling with increased T2 signal intensity, affecting the cerebral hemispheres more than the posterior fossa structures. Reference Kampondeni, Potchen, Hommel and Kremsner33 In a cohort of 120 Malawian patients with CM, Potchen et al found moderate-to-severe cerebral edema, resulting in narrowing of CSF spaces and uncal herniation, was the most prevalent finding (47.5%). Reference Potchen, Kampondeni and Seydel32 Some variability in the distribution was also noted, with 24.6% and 7.0% demonstrating supratentorial and infratentorial swelling respectively. Reference Potchen, Kampondeni and Seydel32 In a study of pediatric patients admitted with CM by Seydel et al, 21 (81%) out of the 25 patients who died demonstrated MRI evidence of severe brain swelling at the time of admission (odds ratio for brain swelling among patients who died vs. survivors was 14.0; 95% CI, 4.5–43.4). These patients died as a result of respiratory arrest, consistent with the effects of elevated intracranial pressure. However, in this study, it was notes that brain swelling in the setting of CM does not always translate into fatality as the authors found that approximately 65% of the patients with severely increased brain volume survived with long-term outcomes similar to survivors with normal brain volumes. They concluded that interventions to decrease brain swelling and to sustain respiration temporarily could reduce mortality without an increased in morbidity. Reference Seydel, Kampondeni and Valim34

Signal alterations in the deep gray nuclei have also been described. Hyperintense T2 signals in the basal ganglia was common in the study conducted by Potchen et al (84.2%); predominant involvement of the globus pallidus was seen in a few cases. Reference Potchen, Kampondeni and Seydel32 Restricted diffusion in the basal ganglia was also seen (42%). Reference Potchen, Kampondeni and Seydel32 Thalamic involvement (64.2%) was observed, but associated restricted diffusion was uncommon. Reference Potchen, Kampondeni and Seydel32 Petechial cerebral hemorrhagic foci may also be noted on susceptibility weighted images (SWI). Reference Kampondeni, Potchen, Hommel and Kremsner33

Potchen et al also described T2-hyperintense cortical swelling (61.7%), typically not following a vascular distribution, however, cortical restricted diffusion was noted in 20.8% of cases. Reference Potchen, Kampondeni and Seydel32 These areas of restricted diffusion resolve in survivors, with no sequalae after a month, but persist or worsen in patients who succumb to the illness. The etiology of this phenomenon is still not completely understood. Reference Kampondeni, Potchen, Hommel and Kremsner33

Increased T2 signal in the white matter (71.7%) was commonly accompanied by DWI abnormalities (45.0%) in the study by Potchen et al. Reference Potchen, Kampondeni and Seydel32 Callosal T2 hyperintensity was also seen in (49.2%) commonly in the absence of other white matter changes. Reference Potchen, Kampondeni and Seydel32 In their cohort, cerebellar involvement was not uncommon (49.2%), ranging from diffuse involvement to multifocal areas or unifocal lesions, while restricted diffusion in the cerebellum was uncommon. Reference Potchen, Kampondeni and Seydel32 In our experience, we have encountered a case of CM showing restricted diffusion and T2-prolongation involving the juxta- and subcortical white matter, with a near symmetric distribution (Fig. 9).

Figure 9. Cerebral malaria in patient presenting to the emergency department in deep coma. Axial DWI (A, C) and corresponding ADC (B, D) images show restricted diffusion involving the juxta- and subcortical white matter bilaterally. A diagnosis of CM was made based on peripheral smears and the patient survived after intensive treatment with anti-malarial regimens and supportive management. ADC = apparent diffusion coefficient; CM = cerebral malaria; DWI = diffusion-weighted imaging.

Approximately 30% of those who survive CM develop structural brain abnormalities and are often associated with seizures, developmental delay, behavioral problems and multiple neurological deficits. Reference Kampondeni, Potchen, Hommel and Kremsner33 Volume loss of the brain parenchyma is the most common finding, seen as early as within 2 weeks after coma resolution. Periventricular and subcortical T2 hyperintense signal abnormality in almost 47 % and 18 % of all cases, respectively. Focal areas of cerebral cortical defects / encephalomalacia and isolated vermian atrophy may also be encountered. Reference Kampondeni, Potchen, Hommel and Kremsner33

Inflammatory basis of epilepsy syndromes

Febrile infection-related epilepsy syndrome (FIRES)

Status epilepticus (SE) is a common neurological condition, occasionally preceded by febrile illness in some patients. Reference Gaspard, Hirsch and Sculier35 This entity has various names including new-onset refractory SE (NORSE), febrile illness-related epilepsy syndrome (FIRES), acute encephalitis with refractory repetitive partial seizures (AERRPS) and devastating epilepsy in school-age children (DESC). Reference Gaspard, Hirsch and Sculier35 NORSE is defined as a clinical presentation, not a specific diagnosis, in patients with new onset of refractory SE without a clear acute /active structural or toxic-metabolic etiology. Reference Gaspard, Hirsch and Sculier35 FIRES, a subcategory of NORSE, present following a febrile illness with fever occurring 2 weeks to 24 hours prior to onset of refractory SE. Reference Gaspard, Hirsch and Sculier35 The exact etiology of FIRES is unknown. While few cases have tested positive for mycoplasma and adenovirus, however, no pathogen has been isolated in a vast majority of cases. Reference Reppucci and Datta36 Autoimmunity has also been implicated with reports of anti-glutamate receptor antibodies or anti-glutamic acid decarboxylase antibodies and IL6/IL8 obtained from CSF. Reference Reppucci and Datta36 Genes including SCN2A and PCDH19 have been reported. Reference Kessi, Liu and Zhan37 Febrile illness causes the release of pro-inflammatory cytokines and chemokines which lower the seizure threshold secondary to ion channel function modification while promoting neuronal excitability. Reference Reppucci and Datta36 These may also prime innate immunity activation in seizure-prone areas of the brain and alter cellular components in the BBB, thereby triggering a neuroinflammatory cascade. Reference Reppucci and Datta36 Activation of microglial NLRP3 inflammasome/IL1 axis has been proposed, leading to a proinflammatory-proconvulsive milieu. Reference Reppucci and Datta36 FIRES also shows elevation of pro-inflammatory profile including interleukin IL-6, C-X-X motif chemokine ligand 8 (CXCL8) and C-C motif ligand L3 and 4 (CCL4 and CCL3) as well as IL-1β. Reference Reppucci and Datta36,Reference Kessi, Liu and Zhan37 IL-1β in particular lowers the seizure threshold and is endogenously antagonized by IL-1Ra, encoded by the IL1RN gene. Reference Saitoh, Kobayashi and Ohmori38 Patients with the variable number of tandem repeat (VNTR) allele of the IL1RN, RN2, carry a higher risk of FIRES as the presence of RN2 is associated with reduced IL1RN-mRNA expression and enhanced IL-1β production. Reference Saitoh, Kobayashi and Ohmori38 FIRES typically affects children at a median age of 6–8 years. FIRES has 3 stages of disease evolution. Reference Reppucci and Datta36 The initial prodromal stage (approximately 13 days) is characterized by febrile illness (respiratory or gastrointestinal infection). Reference Reppucci and Datta36,Reference Kessi, Liu and Zhan37 This is followed by a new onset of seizures, which increase in frequency, eventually evolving into refractory SE. Reference Reppucci and Datta36 At this stage, most patients have recovered from fever. Reference Reppucci and Datta36 The final chronic stage ensues with patients requiring multiple anti-seizure medications with varying grades of residual disability. Seizures are typically multifocal or generalized, evolving into super-refractory SE reaching up to several seizures each day. Reference Reppucci and Datta36,Reference Vezzani, Häusler, Kluger and van Baalen39

The major role of MRI, in patients with FIRES, is to exclude other clinical mimics. The NORSE Institute recommends a contrast-enhanced MRI of the brain with MR angiography and MR venography within the first 24 hours of symptom onset. Reference Culleton, Talenti, Kaliakatsos, Pujar and D’Arco40 Most patients have an unremarkable initial MRI examination, though unilateral or bilateral temporal lobe signal abnormalities have been reported. Reference Culleton, Talenti, Kaliakatsos, Pujar and D’Arco40 Description of the spatial involvement of the temporal lobe is heterogeneous in literature, with authors reporting hippocampal, mesial temporal or just temporal lobe involvement. Reference Culleton, Talenti, Kaliakatsos, Pujar and D’Arco40 Cortical edema with T2/FLAIR hyperintensity and/or restricted diffusion may also be noted reflecting impairment of the Na/ATPase pump function Reference Culleton, Talenti, Kaliakatsos, Pujar and D’Arco40 (Fig. 10). Bilateral involvement of the basal ganglia is another frequent abnormality reported in the acute FIRES. Reference Culleton, Talenti, Kaliakatsos, Pujar and D’Arco40 Basal ganglia involvement may affect the striatal inhibitory function and reduce the seizure threshold, however, the exact mechanism leading to signal alterations in the basal ganglia is unclear. Involvement of the insular/peri-insular regions and the cerebral cortices (focal or diffuse) may also be noted. Reference Culleton, Talenti, Kaliakatsos, Pujar and D’Arco40 Symmetric T2/FLAIR hyperintensity involving the claustrum, referred to as the ‘claustral sign’ has been reported, however, this is neither diagnostic nor pathognomonic for NORSE or FIRES Reference Di Dier, Dekesel and Dekeyzer41 (Fig. 11). Involvement of thalami, brainstem and cerebellum is rare. Following contrast administration, parenchymal or leptomeningeal enhancement may be seen. Reference Culleton, Talenti, Kaliakatsos, Pujar and D’Arco40

Figure 10. FIRES in a 10-year-old presenting with acute failure of verbalization, reduced ambulation and poor intake along with abnormal repetitive movements of hands and pursing of lips. Axial FLAIR sequence (A-C) shows extensive, symmetric, gyriform hyperintense cortical swelling in the temporal lobes (A), the insula (B), and the frontal lobes (with a mesial predominance; C). Similar hyperintense swelling is seen involving the lentiform nuclei bilaterally. Axial DWI (D) and ADC (E) sequences show corresponding gyriform restricted diffusion in the temporal lobes. There was no restricted diffusion seen in the rest of the affected brain parenchyma. Coronal contrast-enhanced 3D-T1sequence (F) reveals no evidence of contrast enhancement. A differential diagnosis of infection, including viral etiologies such as herpes, as well as anti-NMDAR encephalitis were considered at the time of initial presentation. A biopsy of the temporal lobe revealed focal spongiotic changes with apoptotic neurons and neuronal loss but no immunohistochemistry was negative. Based on the clinical presentation and the pathological results, FIRES was considered and the patient was treated with methylprednisolone, IV immunoglobulin and plasma exchange (PLEX). ADC = apparent diffusion coefficient; DWI = diffusion-weighted imaging; FIRES = febrile infection-related epilepsy syndrome.

Figure 11. Claustral sign in a 3-year-old with suspected FIRES presenting with multiple seizures, following a bout of febrile illness. Axial FLAIR (A) and coronal T2-weighted (B) sequences reveal hyperintensities in the subinsular regions/claustrum bilaterally (arrows). Axial DWI (C) and axial contrast-enhanced 3D-T1 sequence (D) demonstrate no corresponding restricted diffusion or contrast enhancement, respectively. DWI = diffusion-weighted imaging; FIRES = febrile infection-related epilepsy syndrome.

Imaging in the chronic phase may reveal normal MRI appearances or global parenchymal atrophy with ex-vacuo dilation of the ventricles. Reference Culleton, Talenti, Kaliakatsos, Pujar and D’Arco40 Atrophy and/or T2 hyperintensity within the mesial temporal lobes is not uncommon. Reference Culleton, Talenti, Kaliakatsos, Pujar and D’Arco40 Abnormal white matter T2/FLAIR hyperintensities may be seen in the deep and/or subcortical white matter. Reference Culleton, Talenti, Kaliakatsos, Pujar and D’Arco40 Similar signal changes may be seen in the insular and peri-insular as well as the deep grey nuclei and cortices. Reference Culleton, Talenti, Kaliakatsos, Pujar and D’Arco40 Ependymal enhancement and signal abnormalities in the substantia nigra and corticospinal tracts have been described in single case reports. Reference Culleton, Talenti, Kaliakatsos, Pujar and D’Arco40

Rasmussen’s encephalitis (RE)

RE is a rare chronic neurological disorder of unknown origin, classically causing drug-resistant focal epilepsy with progressive unihemispheric atrophy. Reference Tang, Luan and Li42,Reference Varadkar, Bien and Kruse43 RE may begin anytime between infancy and adulthood, with a median age of 6 years. Reference Varadkar, Bien and Kruse43 Some may present with a prodromal period of hemiparesis or episodic seizures progressing to the acute phase characterized by frequent seizures arising from one cerebral hemisphere. Reference Varadkar, Bien and Kruse43 Nearly half the patients present with epilepsia partialis continua. Reference Varadkar, Bien and Kruse43 Untreated patients develop hemiparesis, hemianopia, cognitive decline and dysphasia (when the language dominant hemisphere is affected). Reference Varadkar, Bien and Kruse43

At present there is no obvious cause known to trigger RE. Viral infections were considered to be a trigger, however, to date no study has proved viral replication in the brain parenchyma of patients with RE. Reference Tang, Luan and Li42,Reference Bien44 Autoantibodies against GluR3 were considered, however, these were found only in a few patients with proven RE. Reference Varadkar, Bien and Kruse43 More recently, cytotoxic T cells have been evaluated in the pathogenesis of RE; pathological specimens have revealed that cytotoxic CD8+ T cell lymphocytes form the majority of infiltrated T cells in RE with 10% of which are granzyme-B positive CD8+ T cells. Reference Tang, Luan and Li42 This is viewed as compelling evidence of neuronal destruction and astrocytic degeneration caused by cytotoxic T cells, which gives rise to seizures. Reference Tang, Luan and Li42 Research has also shown certain genes (encoding interferon-γ, CCL5, CCL22, CCL23, CXCL9, CXCL10 and Fas ligand) were expressed in RE which are related to the activation and recruitment of T cells to sites of inflammation. Reference Varadkar, Bien and Kruse43 Activation of microglia has also been evaluated; their activation closely follows the T-cell infiltration patterns and progression of cortical damage. Reference Tang, Luan and Li42,Reference Varadkar, Bien and Kruse43 Activated microglia release interleukin 1 and other pro-inflammatory cytokines which participate in the seizure induction. Activated microglia may also take part in complement-mediated synaptic dysfunction which increases neuronal excitability. Reference Tang, Luan and Li42,Reference Varadkar, Bien and Kruse43

MRI forms the mainstay in the evaluation of RE. Typical MRI features include preferential atrophy of the frontal lobe and insula with volume loss of the caudate head, together leading to ex-vacuo enlargement of the CSF spaces Reference Tang, Luan and Li42 (Fig. 12). Volume loss typically ensues 4 months following symptom onset. Intra- and/or subcortical T2/FLAIR may also be observed Reference Tang, Luan and Li42,Reference Bien44 (Fig. 13). Areas of increased signals have been shown to correspond to sites of increased T cells, microglial nodules and GFAP + astrocytes. Most tissue loss is noted within the initial 12 months of disease onset; with disease progression, cortical atrophy develops in the rest of the hemisphere. Reference Tang, Luan and Li42,Reference Bien44 Fluorodeoxyglucose-positron emission tomography shows hypometabolism in the affected hemisphere while single photon emission computed tomography may show interictal hypoperfusion and ictal hyperperfusion of focal parenchymal lesions in patients with RE. Reference Tang, Luan and Li42

Figure 12. Rasmussen encephalitis in a 6-year-old patient presenting with sudden onset absence-like complex partial seizures originating from the right hemisphere. Axial FLAIR sequence (A, B) shows areas ill-defined areas of hyperintense signal involving the cortex and the juxtacortical white matter of the right parietal (A) and frontal (B) lobes. DWI (not shown) revealed no restricted diffusion in these locations. Corresponding contrast-enhanced 3D-T1 sequence (C, D) reveals no evidence of enhancement in these regions. DWI = diffusion-weighted imaging.

Figure 13. Evolution of Rasmussen encephalitis. Axial FLAIR (A) and axial T2-weighted (B) sequences reveal a small nonspecific focus of hyperintensity in the right frontal lobe (dotted arrows). Axial 3D-T1 sequence (C) shows normal volume of the right cerebral hemisphere. Follow-up MRI (D-F) obtained after 3 years due to medically refractory focal epilepsy, occurring multiple times daily, and emerging left hemiparesis. Axial FLAIR (D), axial T2-weighted (E), and axial 3D-T1 (F) sequences reveal marked volume loss of the right cerebral hemisphere, seen in the form of ex-vacuo dilation of the right lateral ventricle and prominence of the sulcal spaces. Note, apart from non-specific FLAIR/T2 hyperintensity along the right frontal horn, there is no other signal abnormality seen in the rest of the brain parenchyma.

Autoimmune encephalitis

Anti-NMDAR encephalitis

The N-methyl-D-aspartate receptor (NMDAR) is an ionotropic glutamate receptor, centered in the postsynaptic compartment. Reference Skrenkova, Hemelikova and Kolcheva45 It is comprised of two GluN1 and two GluN2 or GluN3 subunits; GluN1 and GluN3 bind glycine while GluN2 binds glutamate. Reference Skrenkova, Hemelikova and Kolcheva45 Autoantibodies against the GluN1 subunit of the NMDAR are the most common form of autoimmune encephalitis, especially in children. Reference Ding, Jian, Stary, Yi and Xiong46,Reference de Bruijn, Bruijstens and Bastiaansen47

Tumors and post-viral states are common immunologic triggers linked to anti-NMDAR encephalitis. Reference Huang, Xie, Hu and Tang48 However, in nearly half of the cases, the immunologic triggers remain unknown. Reference Huang, Xie, Hu and Tang48 Tumors containing nervous tissues expressing NMDAR (e.g., ovarian teratomas) incite expansion of T and B cells while virus-induced neuronal damage leads to direct transport of the receptors to draining lymph nodes or APC uptake stimulating memory B expansion. Reference Ding, Jian, Stary, Yi and Xiong46 Upon exposure to NMDAR, naïve B cells are activated and enter the CNS via the choroid plexus. Reference Huang, Xie, Hu and Tang48 Tumors also present potential ligands for the innate PRRs, including, but not limited to TLRs. Reference Ding, Jian, Stary, Yi and Xiong46 The activation of innate immunity-mediated cytokines and TLR ligands leads to BBB disruption, allowing B cell infiltration. Reference Ding, Jian, Stary, Yi and Xiong46 Interaction with the autoantibodies causes internalization of the NMDARs leading to decreased receptor density on cell surfaces and eventual imbalance between excitatory and inhibitory neurotransmitters Reference Skrenkova, Hemelikova and Kolcheva45 (Fig. 14).

Figure 14. Pathophysiology of anti-NMDAR encephalitis. (A) Simplified diagrammatic representation of neurotransmission at a glutamatergic synapse and (B) simplified diagrammatic representation of antibodymediated internalization. Binding of antibodies to the GluN1 subunit causes crosslinking of NMDAR, hence forming clusters (not shown). These are removed from the post-synaptic surface by endocytosis (receptor internalization). This eventually leads to a decreased density of NMDAR at the postsynaptic membrane. The net result is slowing down of glutamate evoked postsynaptic impulses. (Created with BioRender.com). NMDAR, N-methyl-D-aspartate receptor.

NMDAR encephalitis often shows a progressive clinical worsening. An initial viral-like prodrome may be encountered (fever, malaise, headache, etc). Reference Kelley, Patel, Marin, Corrigan, Mitsias and Griffith49 This is subsequently followed by psychiatric symptoms (anxiety, depression, schizophrenia, psychosis) with progression to temporal lobe dysfunction (amnesia and seizures). Reference Kelley, Patel, Marin, Corrigan, Mitsias and Griffith49 Ultimately, severe residual neurologic deficits may be seen (impairments of memory and psychiatric symptoms). Reference Kelley, Patel, Marin, Corrigan, Mitsias and Griffith49 Early diagnosis and treatment may be associated with a good prognosis and neurological outcomes, with patients returning to baseline. Reference Kelley, Patel, Marin, Corrigan, Mitsias and Griffith49

Most patients with anti-NMDAR encephalitis have normal brain MRI examinations at presentation (89%) and on follow-up (79%) Reference Kelley, Patel, Marin, Corrigan, Mitsias and Griffith49 Non-specific T2/FLAIR hyperintensity involving grey and/or white matter with variable supra- and infratentorial distribution has been described. Reference Kelley, Patel, Marin, Corrigan, Mitsias and Griffith49 In a study of 53 patients with anti-NMDAR encephalitis, including pediatric and adult patients, Zhang et al divided brain lesions into 4 groups: type 1 (normal MRI; 53%), type 2 (isolated hippocampal involvement; 13%), type 3 (lesions sparing the hippocampus; 13%) and type 4 (combined hippocampus and other focal brain involvement’ 21%). Reference Goldman and Prinz5 . Apart from the hippocampus, lesions were found in the frontal lobes, cingulate gyri, middle cerebellar peduncles, corpus callosum, insula, basal ganglia, thalami and brainstem. Reference Zhang, Duan and Ye50 They found no statistical difference in the imaging manifestations between pediatric and adult groups (p = 0.982). Reference Zhang, Duan and Ye50 Hippocampal involvement was associated with a poor neurological outcome. Reference Zhang, Duan and Ye50 The presence of restricted diffusion is uncommon and should prompt further evaluation of the etiologies including seizures, infectious encephalitis and vasculitis. Reference Van Mater51,Reference Budhram, Britton and Liebo52 (Fig. 15). Rare manifestations include isolated contrast enhancement of the leptomeninges and cranial nerve enhancement Reference Van Mater51,Reference Budhram, Britton and Liebo52 (Fig. 16).

Figure 15. Imaging findings in a 12-year-old patient with anti-NMDAR encephalitis. Axial FLAIR (A) and T2 (B) images reveal abnormal hyperintense swelling of the left frontal cortices (arrow), with resultant effacement of the left frontal sulci. Abnormal signal is also seen involving the subjacent white matter. Axial ADC (C) obtained at the corresponding level shows patchy areas low signals representing restricted diffusion (arrow) on the background of hyperintense signals representing vasogenic edema. Coronal FLAIR (D), axial T2 (E), axial ADC (F) images obtained through the temporal lobes reveals similar hyperintense swelling of the left anterolateral temporal cortices (arrows). Contrast was not administered as the patient was suffering from concomitant chronic renal failure. ADC = apparent diffusion coefficient; NMDAR = N-methyl-D-aspartate receptor.

Figure 16. Atypical imaging presentation of anti-NMDAR encephalitis in a 2-year-old presenting with new-onset seizures, behavioral changes, aphasia, and fever. Axial (A) and coronal oblique FLAIR (B) sequences reveal hyperintense signal abnormality within the hippocampal formations bilaterally (arrows) as well as gyriform, FLAIR hyperintense swelling of the left occipitotemporal gyrus (dotted arrow in A) and the left parahippocampal gyrus (dotted arrow in B). Contrast-enhanced axial FLAIR images (C, D) reveal diffuse leptomeningeal enhancement (stepped arrows) along both cerebral hemispheres. The patient received high-dose intravenous steroids with clinical status reaching back to baseline within a few weeks. NMDAR = N-methyl-D-aspartate receptor.

Neurological manifestations in monogenic autoimmune disorders

The monogenic autoimmune disorders affecting the CNS are further classified into Interleukin-1b mediated disorders, type 1 interferonopathies, hemophagocytic syndromes and immunodeficiency disorders. In this section, we discuss the current understanding of acute necrotizing encephalopathy and familial hemophagocytic lymphohistiocytosis.

Acute necrotizing encephalopathy (ANE) vs acute necrotizing encephalopathy 1 (ANE1)

ANE is a distinct parainfectious encephalopathy, triggered by a viral infection (influenza A, influenza B, novel influenza A/H1N1, parainfluenza, human herpes virus 6, herpes 7, etc.). While the pathomechanisms of ANE are still unclear, it is thought to be an exaggerated immune response to the viral infection (cytokine storm), resulting in multisystem involvement. The CNS manifestation of this is seen in the form of increased BBB permeability. Reference Wu, Wu, Pan, Wu, Liu and Zhang53 Research shows, that a CD56+ natural killer (NK) may be involved. Reference Wu, Wu, Pan, Wu, Liu and Zhang53 Elevation of other cytokines has also been described, with IL-6 and TNF-α playing a major role in neurotoxicity and BBB endothelial damage, respectively. Reference Wu, Wu, Pan, Wu, Liu and Zhang53 The outcome of ANE varies depending on the causative organism. In a study by Khamis et al, the rate of pediatric intensive care unit admission was 100% and 77% for influenza and non-influenza cases, respectively. The authors also found a higher mortality rate in patients with ANE secondary to influenza infection (42% vs 11% among those due to non-influenza infection) Reference Khamis, Brown, Kaliakatsos, Eyre and Lim54

ANE was initially thought to have a geographic distribution but is now considered a global phenomenon. Multiple cases of recurrent and/or familial ANE were reported suggesting an inherited disposition. Missense mutations in the Ran Binding Protein 2 (RANBP2), the gene encoding the nuclear pore protein, were found to be the susceptibility allele for familial/recurrent cases, with an autosomal dominant inheritance. Reference Wu, Peng, Jiang, He, Jiang and Hu55 This form of ANE has since been classified as “ANE1”. Reference Wu, Wu, Pan, Wu, Liu and Zhang53 Mutations in SCNIA R1575C and carnitine palmitoyl transferase II genes have also been linked to ANE1. Reference Neilson, Adams and Orr56Reference Kumakura, Iida, Saito, Mizuguchi and Hata61

RANBP2 is located on the cytoplasmic surface of the nuclear pore. It performs several functions such as protein transport facilitation, intracellular trafficking, neuronal energy maintenance and regulation of specific subsets of mRNAs. Reference Wu, Wu, Pan, Wu, Liu and Zhang53,Reference Mahadevan, Zhang and Akef62 Specifically, RANBP2 is a small ubiquitin-related modifier (SUMO) pathway protein that attaches and stabilizes target proteins (SUMOylation), thus enabling them to perform vital cellular functions. Reference Malik and Shroff24 In this context, RANBP2-mediated SUMOylation of Argonaute or AGO proteins ensure the silencing of ANE-related cytokines (IL-6, TNFα), STAT1 activity and type 1 interferon. Reference Malik and Shroff24 Failure of these silencing mechanisms leads to persistent cytokine production increased glutaminergic CNS excitotoxicity and eventual CNS injury Reference Levine, Ahsan, Ho and Santoro63,Reference Ichiyama, Endo, Kaneko, Isumi, Matsubara and Furukawa64 (Fig. 17).

Figure 17. Schematic of the ANE1 pathophysiology. The nuclear pore complex is a complex macromolecule located in the nuclear envelope. It serves as an important conduit for transport across the nuclear membrane. Argonaute (AGO protein) is involved in messenger RNA (mRNA) silencing, by facilitating attachment of micro-RNA (miRNA-induced silencing complex; a silencing RNA) to the target IL6 mRNA. The RANBP2-mediated SUMOylation ensures a stable attachment of during AGO to IL-6 mRNA passage during its transport across the nuclear pore protein. The SUMOylated AGO bound IL-6 mRNA later attaches to miRNA-induced silencing complex, thereby ceasing the production of IL-6. While the exact role of mutated RANBP2 interaction in this sequence of events is not clearly delineated, but failure of miRNA-induced silencing complex attachment to an otherwise unstable AG0 bound IL-6 mRNA complex may lead to sustained production of cytokines and hypercytokinemia. Adapted from reference 24. (Created with BioRender.com) ANE1 = acute necrotizing encephalopathy.

ANE1 is usually preceded by a nonspecific infection which triggers downstream hyperinflammatory mechanisms. Clinically, the disease is divided into 3 phases prodrome (fever, diarrhea, vomiting, upper respiratory tract infection), acute encephalopathy (seizure, altered consciousness, focal neurologic deficits) and recovery. Reference Park, Hwang, Kim, Lee and Ko65 The clinical course is diverse ranging from mild encephalopathy to a fulminant course leading to death. Reference Park, Hwang, Kim, Lee and Ko65

Neuroimaging manifestations include symmetric involvement of bilateral thalami with a characteristic trilaminar pattern This pattern refers to a central core of T2 hyperintensity representing perivascular hemorrhage and neuronal/glial necrosis, surrounded by a zone of cytotoxic edema, seen as restricted diffusion, due to vascular congestion and oligodendrocyte swelling. Reference Wu, Peng, Jiang, He, Jiang and Hu55 The outermost zone represents extravasation/vasogenic edema and demonstrates T2-shine through Reference Wu, Peng, Jiang, He, Jiang and Hu55 (Fig. 18). ANE1 has an increased predisposition to involve amygdala, hippocampus and medial temporal lobes compared to sporadic ANE. Reference Neilson, Adams and Orr56 Involvement of the cerebral hemispherical white matter, brainstem and cerebellum may also be noted. Reference Mizuguchi, Yamanouchi, Ichiyama and Shiomi66Reference Vanjare, Selvi and Karuppusami68 Uncommon manifestations include isolated involvement of the cerebral parenchyma, internal capsule and the corpus callosum. Reference Handryastuti, Silvana, Yunus, Taufiqqurrachman and Rafli69 The presence of cavitation and hemorrhage is linked to adverse clinical outcomes in pediatric ANE1. Reference Wong, Simon, Zimmerman, Wang, Toh and Ng70

Figure 18. ANE1 due to RANBP2 mutation in a 4-year-old child presenting with acute encephalopathy. Axial T2-weighted sequence (A) shows symmetric involvement of bilateral thalami (solid arrows). An evolving trilaminar pattern can be identified with a central area of T2 hyperintensity surrounded by a rim of intermediate signal and a peripheral rind of hyperintensity. A small focus of T2-weighted hyperintensity is also noted along the posterior aspect of the right lentiform nucleus (dotted arrow in A). Axial DWI sequence (B) shows patchy areas of restricted diffusion in the core of the thalamic lesions. Axial multiplanar gradient recalled (MPGR) sequence (C) shows areas of susceptibility within the left thalamic lesion suggesting foci of hemorrhage. Axial FLAIR sequence (D) shows homogeneous hyperintense swelling of the thalami as well as the right lentiform nucleus posteriorly. Axial (E) and coronal (F) contrastenhanced 3D-T1 sequence shows subtle enhancement within the thalamic lesions. DWI = diffusion-weighted imaging.

Familial hemophagocytic lymphohistiocytosis (FHLH)

Familial hemophagocytic lymphohistiocytosis (FHLH) is an autosomal recessive disorder with various subtypes (deficient genes): FHL1 (unknown), FHL2 (PRF1), FHL3 (Munc13-4 protein on UNC13D), FHL4 (Syntaxin 11 protein on STX11) and FHL5 (Munc18-2 protein on STXBP2) (71). PRF1 encodes perforin, contained in cytotoxic granules of lymphocytes. Reference Canna and Marsh72 Upon exposure to infected cells, CD8+ cytotoxic T cells and NK cells release these perforin-containing cytotoxic granules and granzymes resulting in target cell membrane destabilization and apoptosis. Reference Canna and Marsh72 Perforin deficiency leads to the inability of granule content to enter into target cells. Reference Canna and Marsh72 A prolonged synapse time ensues between perforin-deficient cytotoxic lymphocytes and target cells causing overproduction of pro-inflammatory cytokines. Reference Canna and Marsh72 APCs continue to accumulate and stimulate T cell activation and proliferation leading to excessive lymphohistiocytic proliferation and hypercytokinemia ultimately causing widespread tissue damage and hyperinflammation. Reference Canna and Marsh72 UNC13D, STX11 and STXBP2 gene products are important for normal cytotoxic granule exocytosis. Reference Canna and Marsh72 Defects in these genes limits the granule contents from being released into the immunological synapse, hence target cells are not destroyed. Reference Canna and Marsh72 HLH is also a manifestation of other genetic diseases, including pigmentary disorders (RAB27A, LYST and AP3B1 mutations), X-linked lymphoproliferative diseases (SH2D1A and XIAP mutations), CDC42 mutations, gain-of-function mutations in NLRC4 and Epstein-Barr virus (EBV) susceptibility disorders (MAGT1, ITK, CD27, CD70, CTPS1, RASGRP1 gene deficiency). Reference Canna and Marsh72 Discussion of their pathomechanisms is beyond the scope of this article, but summarized in Figure 19.

Figure 19. Schematic representation of HLH pathogenesis. (Circuit #1) in patients with familial HLH and some other forms of HLH related to primary EBV infection, genetic defects of virucidal property (e.g., Perforin 1 (PERF1) deficiency) causes sustained presentation of antigens and activation CD8+T cells, amplified by IFN-γ release and MHC-I upregulation. Prolonged IFN-γ release is also though to cause tissue macrophage/monocyte activation consequently leading to release of inflammatory mediators (IL-1b, IL-6, IL-18, IL-12). (Circuit #2) macrophages/monocytes are activated by activated by chronic inflammation and TLR stimulation (infection). In HLH associated with malignancies, malignant cells may drive HLH through autonomous cytokine release or presentation of EBV antigen. Drug (CAR-T cell therapy) induced occurs due to activation of therapeutic CD8+ CAR-T cells. (Created with BioRender.com). EBV = Epstein–Barr virus; CAR-T = Chimeric antigen receptor T cell; HLH = hemophagocytic lymphohistiocytosis.

CNS involvement is common in all forms of HLH. The reported incidence of CNS involvement in FHLH is ranges between 18-75% in FHLH. Reference Cooray, Sabanathan and Hacohen73 In a case series (n = 38), ataxia/gait disturbances (74%), seizures (50%), headache (47%), visual alterations (45%) and motor impairment (42%) were reported (74). Cranial nerve palsies, papilledema, dysmetria, dysarthria, sensory deficits and behavioral alterations were also reported. Reference Blincoe, Heeg and Campbell74 Histopathologically, pediatric CNS HLH shows meningeal lymphocytic and macrophage infiltration (stage 1), advancing to perivascular inflammatory infiltrate (stage 2), leading to diffuse parenchymal infiltration, multifocal necrosis and astrogliosis (stage 3). Reference Henter and Nennesmo75

MRI is the neuroimaging modality of choice for patients with CNS-HLH. Reference Cooray, Sabanathan and Hacohen73 In a study of 179 pediatric patients, CNS imaging findings were seen in up to 50% of cases with systemic HLH, most of which belonged to the FHLH group. Reference Zhao, Zhang and Li76 At neuroimaging, pediatric CNS-HLH presents with differing imaging patterns affecting various brain regions including multifocal cerebral and cerebellar white matter lesions, CLIPPERS (chronic lymphocytic inflammation with pontine perivascular enhancement responsive to steroids) like brainstem lesions and diffuse cerebellar involvement similar to cerebellitis. Reference Malik, Antonini and Mannam77 Atypical findings include diffuse hemorrhagic transformation with bilateral basal ganglia, thalamus and brainstem involvement. Reference Pak, Selehnia and Hunfeld78 Other findings reported in CNS-HLH are nodular or ring-enhancing lesions with restricted diffusion, leptomeningeal enhancement, diffuse cerebral edema, subdural collections as well as calcification and cerebral atrophy in late stages Reference Fitzgerald and MacClain79Reference Rego, Severino and Micalizzi81 (Fig. 20).

Figure 20. HLH in a 12-year-old with long-standing immune dysregulation (fever, cytopenia, splenomegaly). Axial FLAIR (A, B) images reveal bilateral, near-symmetric, gyriform FLAIR hyperintensity with cortical swelling with similar signal abnormality in the juxtacortical white matter in the frontal and parietal lobes (arrows in A, B). Axial FLAIR (C) reveals patchy hyperintensities in the cerebellar hemispheres as well (C). Contrast-enhanced axial 3D-T1 sequences (D-F) reveal multiple, punctate foci of cortical enhancement, most accentuated in the frontal lobes, more on the left (arrows in D), with few foci of enhancement in the right parietal lobe (arrow in E), and left cerebellum (arrow in F).

SARS-CoV-2 and COVID-19: the novel mimicker

The recent crippling pandemic, termed as Coronavirus disease (COVID-19), is caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) belonging to the positive-sense single-stranded RNA viruses beta coronavirus (β-CoVs) family Reference Yu, Sun and Feng82 The potential mechanisms to explain the SARS-CoV2 neurotropism include angiotensin-converting enzyme 2 receptor (ACE2; neurons, astroglia cells, microglia cells and endothelial cells) mediated direct entry into the neuronal cells secondary to hematogenous spread. Reference Desai, Manchanda, Kumar, Tiwari and Kumar83Reference Zubair, McAlpine, Gardin, Farhadian, Kuruvilla and Spudich85 Of note, the virus is approximately 60–140 nm in diameter allowing it to easily bypass the BBB. Reference Yu, Sun and Feng82 Transneuronal dissemination (antegrade or retrograde) is another route for CNS infection. ACE2, transmembrane serine protease 2, and neuropilin-1 (NRP1) expressed in the olfactory epithelium are thought to facilitate such transneuronal migration of the virus. Reference Yu, Sun and Feng82

The etiology of CNS manifestations in COVID-19 are multifactorial. Acute respiratory distress syndrome (ARDS), a common complication of pulmonary SARS-CoV-2 infection, causes severe hypoxia leading to accumulation of anaerobic metabolites in the mitochondria and failure of the Na+ K+-ATPase pump leading to cytotoxic edema and cerebral blood flow dysregulation. Reference Yu, Sun and Feng82 Neurological manifestations in COVID-19 are may also occur from a systemic inflammatory response (cytokine storm) characterized by an exaggeration of pro-inflammatory cytokine and chemokine release causing severe end-organ damage. Reference Yu, Sun and Feng82 Studies indicate that SARS-CoV-2 induces proinflammatory cytokine signaling from astrocytes and microglia cells causing a massive release of inflammatory factors (IL-6, IL-12, IL-15, IL-1β and TNF-α). Reference Yu, Sun and Feng82 Immune dysregulation is another underlying mechanism of CNS injury; the virus may induce an increased concentration of proinflammatory CCR6+ Th17 in CD4+ T cells and cytotoxic granules in CD8+ T cells, indicating T cell overactivation and associated immune-mediated CNS damage. Reference Yu, Sun and Feng82

Acute and subacute neurological complications

Acute encephalopathy or encephalitis typically presents with ill-defined areas of T2/FLAIR hyperintensity of the cortex and hemispherical white matter with gyral swelling; rare deep grey nuclei involvement; microbleeds; and variable contrast enhancement. Reference Orman, Desai and Kralik86Reference Lindan, Mankad and Ram88 Occasional reports of cerebellitis have also been published Reference O’ Neill and Polavarapu89 (Fig. 21). Parainfectious immune-mediated Guillain-Barre syndrome and Miller-Fisher variants are characterized by cauda equina nerve root thickening/enhancement and lower cranial nerve enhancement Reference Malik and Shroff24 (Fig. 22). On the other hand, endothelial injury and thrombosis lead to vasculitis, acute arterial stroke, venous thromboembolism, microbleeds and parenchymal hemorrhage. Moreover, cytokine and inflammatory mediators result in cerebral endothelial damage and blood-brain barrier disruption leading to posterior reversible encephalopathy syndrome. Reference Korkmazer, Ozogul and Hikmat90

Figure 21. COVID-related cerebellitis in a 6-year-old patient presenting with fever, headache, ataxia and decreased mental status. Axial DWI (A) and ADC (B) sequences reveal extensive restricted diffusion involving both the cerebellar hemispheres (arrows) with a focus of profound restricted diffusion in the left middle cerebellar peduncle (dotted arrows). Axial FLAIR (C) and T2-weighted (D) sequences reveal hyperintense swelling of the cerebellum with effacement of the interfolial spaces and the cerebellopontine angle cisterns. Contrast-enhanced axial 3D-T1 sequence (E) demonstrates no abnormal parenchymal enhancement. Delayed coronal contrast enhanced T1-weighted sequence (F) reveals leptomeningeal enhancement along the cerebellar hemispheres bilaterally. The patient was treated with intravenous immunoglobulins for 10 days along with 10 cycles of plasma exchange (PLEX). ADC = apparent diffusion coefficient; DWI, diffusion-weighted imaging.

Figure 22. Guillain-Barré syndrome (GBS) in a 2-year-old presenting with COVID infection detected on nasopharyngeal swab analysis, presenting with generalized weakness and gait disturbance. Sequential craniocaudal contrast-enhanced axial (A, B) and coronal (C) T1 sequences through the lumbar spine demonstrate enhancing thickening of the ventral cauda equina nerve roots. The patient was treated with a 3-day course of intravenous immunoglobulins.

Delayed neurological complications

The cytotoxic lesions of the corpus callosum splenium are associated with hyperinflammatory illness i.e., pediatric multisystem inflammatory syndrome in children (MIS-C); and the myositis of the neck and face typically occurs weeks after COVID-19 infection. On imaging, splenial lesions may be seen in up to 60% of children and demonstrate T2 FLAIR hyperintensity with restricted diffusion and postulated to represent intramyelinic edema due to cytokine-mediated glutamate toxicity. Reference Abdel-Mannan, Eyre and Löbel91

Cytokine storm-induced neuronal and autoimmune injury encompassing acute disseminated encephalomyelitis (ADEM) like phenotype, acute necrotizing encephalopathy (ANE) and transverse myelitis are reported in all phases (Fig. 23). Rarely, ADEM-like imaging phenotype in SARS-CoV-2 occurs due to immune-mediated encephalitis associated with anti-myelin oligodendrocyte (MOG) and Anti-NMDAR antibodies. Pediatric COVID-19-associated ADEM-like phenotype shows better treatment response and outcome on treatment with steroids compared to increased association with severe disease, risk of intracranial hemorrhage and poor outcome in adults. Reference Monti, Giovannini and Marudi92,Reference Manzano, McEntire, Martinez-Lage, Mateen and Hutto93

Figure 23. MRI in an 8-year-old patient presenting with fever and decreased level of consciousness. The patient tested positive for COVID infection. Axial DWI (A) and apparent diffusion coefficient (ADC; B) show areas of restricted diffusion involving both the thalami (arrowheads) with peripheral areas of facilitated diffusion (dotted arrows in B). Axial FLAIR (C) and coronal T2-weighted (D) sequences demonstrate marked swelling of the thalami (arrows). In addition, confluent T2-weighted hyperintensity is seen in the pons (stepped arrow in D), extending into the medulla. Corresponding FLAIR hyperintensity was noted in the affected brainstem (not shown). Contrast-enhanced axial (E) and coronal (F) 3D-T1 sequences show faint irregular enhancement in the thalami (arrows), with a nonenhancing core, corresponding to the areas of restricted diffusion. Subtle, irregular, ring-like enhancement is also seen in the right hemipons (stepped arrow in F). These imaging manifestations suggest a parainfectious CNS complication (ANE) following COVID-19 infection occurring due to a hyperimmune response. ANE = acute necrotizing encephalopathy; CNS = central nervous system; DWI = diffusion-weighted imaging.

Conclusion

In this review, we discuss some of the common infective and inflammatory conditions specific to the pediatric population. With an improved understanding of the immune system, the role of innate and adaptive immune mechanisms in the neurological and imaging manifestations of pediatric infections and inflammatory syndromes has now been elucidated. These conditions may utilize similar mechanisms, yet produce clinically and radiographically distinct manifestations with some overlap. Understanding these pathomechanisms is not only critical in accurate diagnosis but also paves the path for further research in terms of specific targeted treatments.

Author contributions

VP: Contributed to the study concept and design, manuscript preparation and literature search.

SBH: Contributed to the study concept and design and literature search.

MS: Contributed to the study concept and design and literature search.

Funding statement

None.

Competing interests

None.

References

DeMaio, A, Mehrotra, S, Sambamurti, K, et al. The role of the adaptive immune system and T cell dysfunction in neurodegenerative diseases. J Neuroinflammation. 2022;19:251.Google Scholar
Kigerl, KA, de Rivero Vaccari, JP, Dietrich, WD, Popovich, PG, Keane, RW. Pattern recognition receptors and central nervous system repair. Exp Neurol. 2014;258:516.Google Scholar
Aristizábal, B, González, Á. Innate immune system, autoimmunity. In: Anaya, JM, Shoenfeld, Y, Rojas-Villarraga, A, tal, e, editors. From Bench to Bedside. Bogota (Columbia): El Rosario University Press; 2013, pp. 3146.Google Scholar
Medzhitov, R, Janeway, CA. Decoding the patterns of self and nonself by the innate immune system. Science. 2002;296:298300.Google Scholar
Goldman, T, Prinz, M. Role of microglia in CNS autoimmunity. Clin Dev Immunol. 2013;2013:18.Google Scholar
Subramaniam, SR, Federoff, HJ. Targeting microglial activation states as a therapeutic avenue in Parkinson’s disease. Front Aging Neurosci. 2017;9:176.Google Scholar
Louveau, A, Harris, TH, Kipnis, J. Revisiting the mechanisms of CNS immune privilege. Trends Immunol. 2015;36:569–77.Google Scholar
Galea, I, Bechmann, I, Perry, VH. What is immune privilege (not)? Trends Immunol. 2007;28:12–8.Google Scholar
Waisman, A, Liblau, RS, Becher, B. Innate and adaptive immune responses in the CNS. Lancet Neurol. 2015;14:945–55.Google Scholar
Clatworthy, MR. The meninges—a cradle and school for nurturing and educating developing B cells. Immunity. 2021;54:2688–90.Google Scholar
Ampie, L, McGavern, DB. Immunological defense of CNS barriers against infections. Immunity. 2022;55:781–99.Google Scholar
Sweeney, MD, Zhao, Z, Montagne, A, Nelson, AR, Zlokovic, BV. Blood-brain barrier: from physiology to disease and back. Physiol Rev. 2019;99:2178.Google Scholar
Mapunda, JA, Tibar, H, Regragui, W, Engelhardt, B. How does the immune system enter the brain? Front Immunol. 2022;13:805657.Google Scholar
Wimmer, I, Tietz, S, Nishihara, H, et al. Pecam-1 stabilizes blood-brain barrier integrity and favors paracellular T-cell diapedesis across the blood-brain barrier during neuroinflammation. Front Immunol. 2019;10:711.Google Scholar
Bixel, G, Kloep, S, Butz, S, Petri, B, Engelhardt, B, Vestweber, D. Mouse CD99 participates in T-cell recruitment into inflamed skin. Blood. 2004;104:3205–13.Google Scholar
Meyer, C, Martin-Blondel, G, Liblau, RS. Endothelial cells and lymphatics at the interface between the immune and central nervous systems: implications for multiple sclerosis. Curr Opin Neurol. 2017;30:222–30.Google Scholar
Nourshargh, S, Alon, R. Leukocyte migration into inflamed tissues. Immunity. 2014;41:694707.Google Scholar
Kadry, H, Noorani, B, Cucullo, L. A blood-brain barrier overview on structure, function, impairment, and biomarkers of integrity. Fluids Barriers CNS. 2020;17:69.Google Scholar
Ifergan, I, Kebir, H, Alvarez, JI, et al. Central nervous system recruitment of effector memory CD8+ T lymphocytes during neuroinflammation is dependent on α4 integrin. Brain. 2011;134:3560–77.Google Scholar
Marchetti, L, Engelhardt, B. Immune cell trafficking across the blood-brain barrier in the absence and presence of neuroinflammation. Vasc Biol. 2020;2:H1H18.Google Scholar
Louveau, A, Smirnov, I, Keyes, TJ, et al. Structural and functional features of central nervous system lymphatic vessels. Nature. 2015;523:337–41.Google Scholar
Aspelund, A, Antila, S, Proulx, ST, et al. A dural lymphatic vascular system that drains brain interstitial fluid and macromolecules. J Exp Med. 2015;212:991999.Google Scholar
van Zwam, M, Huizinga, R, Heijmans, N, et al. Surgical excision of CNS-draining lymph nodes reduces relapse severity in chronic-relapsing experimental autoimmune encephalomyelitis. J Pathol. 2009;217:543–51.Google Scholar
Malik, P, Shroff, M. Infection and inflammation: radiological insights into patterns of pediatric immunemediated CNS injury. Neuroradiology. 2022;65:425–39.Google Scholar
Patel, RM, Pinto, JM. Olfaction: anatomy, physiology, and disease. Clin Anat. 2013;27:5460.Google Scholar
Sridhar, A, Karelehto, E, Brouwer, L, Pajkrt, D, Wolthers, K. Parechovirus a pathogenesis and the enigma of genotype A-3. Viruses. 2019;11:1062.Google Scholar
Sarma, A, Hanzlik, E, Krishnasarma, R, Pagano, L, Pruthi, S. Human parechovirus meningoencephalitis: neuroimaging in the era of polymerase chain reaction-based testing. AJNR Am J Neuroradiol. 2019;40:1418–21.Google Scholar
Westerhuis, BM, Koen, G, Wildenbeest, JG, et al. Specific cell tropism and neutralization of human parechovirus types 1 and 3: implications for pathogenesis and therapy development. J Gen Virol. 2012;93:2363–70.Google Scholar
Bissel, SJ, Auer, RN, Chiang, C-H, et al. Human parechovirus 3 meningitis and fatal leukoencephalopathy. J Neuropathol Exp Neurol. 2015;74:767–77.Google Scholar
Schiess, N, Villabona-Rueda, A, Cottier, KE, Huether, K, Chipeta, J, Stins, MF. Pathophysiology and neurologic sequelae of cerebral malaria. Malaria J. 2020;19:266.Google Scholar
Shikani, HJ, Freeman, BD, Lisanti, MP, Weiss, LM, Tanowitz, HB, Desruisseaux, MS. Cerebral malaria: we have come a long way. Am J Pathol. 2012;181:1484–92.Google Scholar
Potchen, MJ, Kampondeni, SD, Seydel, KB, et al. Acute brain MRI findings in 120 malawian children with cerebral malaria: new insights into an ancient disease. AJNR Am J Neuroradiol. 2012;33:17401746.Google Scholar
Kampondeni, SD, Potchen, MJ. MRI neuroimaging in pediatric cerebral malaria. In: Hommel, M, Kremsner, P, editors. Encyclopedia of malaria. New York: Springer; 2015, pp. 114.Google Scholar
Seydel, KB, Kampondeni, SD, Valim, C, et al. Brain swelling and death in children with cerebral malaria. N Engl J Med. 2015;372:1126–37.Google Scholar
Gaspard, N, Hirsch, LJ, Sculier, C, et al. New-onset refractory status epilepticus (norse) and febrile infection-related epilepsy syndrome (fires): state of the art and perspectives. Epilepsia. 2018;59:745752.Google Scholar
Reppucci, D, Datta, AN. Fires—pathophysiology, therapeutical approach, and outcome. Zeitschrift für Epileptologie. 2022;35:322–31.Google Scholar
Kessi, M, Liu, F, Zhan, Y, et al. Efficacy of different treatment modalities for acute and chronic phases of the febrile infection-related epilepsy syndrome: a systematic review. Seizure. 2020;79:61–8.Google Scholar
Saitoh, M, Kobayashi, K, Ohmori, I, et al. Cytokine-related and sodium channel polymorphism as candidate predisposing factors for childhood encephalopathy fires/AERRPS. J Neurol Sci. 2016;368:272276.Google Scholar
Vezzani, A, Häusler, M, Kluger, G, van Baalen, A. Febrile infection-related epilepsy syndrome: clinical review and hypotheses of epileptogenesis. Neuropediatrics. 2016;48:005–18.Google Scholar
Culleton, S, Talenti, G, Kaliakatsos, M, Pujar, S, D’Arco, F. The spectrum of neuroimaging findings in febrile infection-related epilepsy syndrome (FIRES): a literature review. Epilepsia. 2019;60:585–92.Google Scholar
Di Dier, K, Dekesel, L, Dekeyzer, S. The claustrum sign in febrile infection-related epilepsy syndrome (FIRES). J Belg Soc Radiol. 2023;107:45.Google Scholar
Tang, C, Luan, G, Li, T. Rasmussen’s encephalitis: mechanisms update and potential therapy target. Ther Adv Chronic Dis. 2020;11:204062232097141.Google Scholar
Varadkar, S, Bien, CG, Kruse, CA, et al. Rasmussen’s encephalitis: clinical features, pathobiology, and treatment advances. Lancet Neurol. 2014;13:195205.Google Scholar
Bien, CG. Pathogenesis, diagnosis and treatment of Rasmussen encephalitis: a european consensus statement. Brain. 2005;128:454–71.Google Scholar
Skrenkova, K, Hemelikova, K, Kolcheva, M, et al. Structural features in the glycine-binding sites of the GluN1 and glun3a subunits regulate the surface delivery of NMDA receptors. Sci Rep. 2019;9:12303.Google Scholar
Ding, H, Jian, Z, Stary, CM, Yi, W, Xiong, X. Molecular pathogenesis of Anti-NMDAR encephalitis. Biomed Res Int. 2015;2015:16.Google Scholar
de Bruijn, MAAM, Bruijstens, AL, Bastiaansen, AEM, et al. Pediatric autoimmune encephalitis. Neurol Neuroimmunol Neuroinflamm. 2020;7:e682.Google Scholar
Huang, Q, Xie, Y, Hu, Z, Tang, X. Anti-N-methyl-D-aspartate receptor encephalitis: a review of pathogenic mechanisms, treatment, prognosis. Brain Res. 2020;1727:146549.Google Scholar
Kelley, BP, Patel, SC, Marin, HL, Corrigan, JJ, Mitsias, PD, Griffith, B. Autoimmune encephalitis: pathophysiology and imaging review of an overlooked diagnosis. Am J Neuroradiol. 2017;38:1070–8.Google Scholar
Zhang, T, Duan, Y, Ye, J, et al. Brain MRI characteristics of patients with anti-n-methyl-D-aspartate receptor encephalitis and their associations with 2-year clinical outcome. AJNR Am J Neuroradiol. 2018;39:824–9.Google Scholar
Van Mater, H. Pediatric inflammatory brain diseases. Curr Opin Rheumatol. 2014;26:553–61.Google Scholar
Budhram, A, Britton, JW, Liebo, GB, et al. Use of diffusion-weighted imaging to distinguish seizurerelated change from limbic encephalitis. J Neurol. 2020;267:3337–42.Google Scholar
Wu, X, Wu, W, Pan, W, Wu, L, Liu, K, Zhang, H-L. Acute necrotizing encephalopathy: an underrecognized clinicoradiologic disorder. Mediators Inflamm. 2015;2015:110.Google Scholar
Khamis, S, Brown, A, Kaliakatsos, M, Eyre, M, Lim, M. Acute necrotizing encephalopathy of childhood: prevention is better than cure especially if the cure remains elusive. Dev Med Child Neurol. 2023;65:1139–40.Google Scholar
Wu, L, Peng, H, Jiang, Y, He, L, Jiang, L, Hu, Y. Clinical features and imaging manifestations of acute necrotizing encephalopathy in children. Int J Dev Neurosci. 2022;82:447–57.Google Scholar
Neilson, DE, Adams, MD, Orr, CMD, et al. Infection-triggered familial or recurrent cases of acute necrotizing encephalopathy caused by mutations in a component of the nuclear pore, RANBP2. Am J Hum Genet. 2009;84:4451.Google Scholar
Alawadhi, A, Saint-Martin, C, Bhanji, F, Srour, M, Atkinson, J, Sébire, G. Acute hemorrhagic encephalitis responding to combined decompressive craniectomy, intravenous immunoglobulin, and corticosteroid therapies: association with novel RANBP2 variant. Front Neurol. 2018;9:130.Google Scholar
Sell, K, Storch, K, Hahn, G, et al. Variable clinical course in acute necrotizing encephalopathy and identification of a novel ranbp2 mutation. Brain Dev. 2016;38:777–80.Google Scholar
Singh, TD, Fugate, JE, Rabinstein, AA. The spectrum of acute encephalitis: causes, management, and predictors of outcome. Neurology. 2014;84:359–66.Google Scholar
Saitoh, M, Shinohara, M, Hoshino, H, et al. Mutations of the SCN1A gene in acute encephalopathy. Epilepsia. 2012;53:558–64.Google Scholar
Kumakura, A, Iida, C, Saito, M, Mizuguchi, M, Hata, D. Pandemic influenza a-associated acute necrotizing encephalopathy without neurologic sequelae. Pediatr Neurol. 2011;45:344–6.Google Scholar
Mahadevan, K, Zhang, H, Akef, A, et al. RANBP2/NUP358 potentiates the translation of a subset of mrnas encoding secretory proteins. PLoS Biol. 2013;11:e1001545.Google Scholar
Levine, JM, Ahsan, N, Ho, E, Santoro, JD. Genetic acute necrotizing encephalopathy associated with RANBP2: clinical and therapeutic implications in pediatrics. Mult Scler Relat Disord. 2020;43:102194.Google Scholar
Ichiyama, T, Endo, S, Kaneko, M, Isumi, H, Matsubara, T, Furukawa, S. Serum cytokine concentrations of influenza-associated acute necrotizing encephalopathy. Pediatr Int. 2003;45:734–6.Google Scholar
Park, Y-J, Hwang, J-Y, Kim, Y-W, Lee, Y-J, Ko, A. Radiological manifestation of familial acute necrotizing encephalopathy with RANBP2 mutation in a far-East Asian family. Medicine. 2021;100:e25171.Google Scholar
Mizuguchi, M, Yamanouchi, H, Ichiyama, T, Shiomi, M. Acute encephalopathy associated with influenza and other viral infections. Acta Neurol Scand. 2007;115:4556.Google Scholar
Britton, PN, Dale, RC, Blyth, CC, et al. Influenza-associated encephalitis/encephalopathy identified by the Australian childhood encephalitis study 2013-2015. Pediatr Infect Dis J. 2017;36:1021–6.Google Scholar
Vanjare, HA, Selvi, BT, Karuppusami, R, et al. Clinical and radiologic findings of acute necrotizing encephalopathy in young adults. AJNR Am J Neuroradiol. 2020;41:2250–4.Google Scholar
Handryastuti, S, Silvana, S, Yunus, RE, Taufiqqurrachman, I, Rafli, A. Difficulties in diagnosis acute necrotizing encephalopathy of childhood: a case report. Child Neurol Open. 2022;9:2329048X2210956.Google Scholar
Wong, AM, Simon, EM, Zimmerman, RA, Wang, HS, Toh, CH, Ng, SH. Acute necrotizing encephalopathy of childhood: correlation of MR findings and clinical outcome. AJNR Am J Neuroradiol. 2006;27:1919–23.Google Scholar
Al-Samkari, H, Berliner, N. Hemophagocytic lymphohistiocytosis. Annu Rev Pathol. 2018;13:2749.Google Scholar
Canna, SW, Marsh, RA. Pediatric hemophagocytic lymphohistiocytosis. Blood. 2020;135:1332–43.Google Scholar
Cooray, S, Sabanathan, S, Hacohen, Y, et al. Treatment strategies for central nervous system effects in primary and secondary haemophagocytic lymphohistiocytosis in children. Curr Treat Options Neurol. 2022;24:5576.Google Scholar
Blincoe, A, Heeg, M, Campbell, PK, et al. Neuroinflammatory disease as an isolated manifestation of hemophagocytic lymphohistiocytosis. J Clin Immunol. 2020;40:901–1.Google Scholar
Henter, J-I, Nennesmo, I. Neuropathologic findings and neurologic symptoms in twenty-three children with hemophagocytic lymphohistiocytosis. J Pediatr. 1997;130:358–65.Google Scholar
Zhao, Y, Zhang, Q, Li, Z, et al. Central nervous system involvement in 179 Chinese children with hemophagocytic lymphohistiocytosis. Chin Med J. 2018;131:1786–92.Google Scholar
Malik, P, Antonini, L, Mannam, P, et al. MRI patterns in pediatric CNS hemophagocytic lymphohistiocytosis. AJNR Am J Neuroradiol. 2021;42:2077–85.Google Scholar
Pak, N, Selehnia, A, Hunfeld, MA, et al. Atypical neuroimaging characteristics of hemophagocytic lymphohistiocytosis in infants: a case series of hemorrhagic brain lesions in the deep grey matter. Neuroradiology. 2020;63:285–8.Google Scholar
Fitzgerald, NE, MacClain, KL. Imaging characteristics of hemophagocytic lymphohistiocytosis. Pediatr Radiol. 2003;33:392401.Google Scholar
Munoz Ruano, MM, Castillo, M. Brain CT and MR imaging in familial hemophagocytic lymphohistiocytosis. AJR Am J Roentgenol. 1998;170:802802.Google Scholar
Rego, I, Severino, M, Micalizzi, C, et al. Neuroradiologic findings and follow-up with magnetic resonance imaging of the genetic forms of haemophagocytic lymphohistiocytosis with CNS involvement. Pediatr Blood Cancer. 2011;58:810–4.Google Scholar
Yu, H, Sun, T, Feng, J. Complications and pathophysiology of COVID-19 in the nervous system. Front Neurol. 2020;11:573421.Google Scholar
Desai, I, Manchanda, R, Kumar, N, Tiwari, A, Kumar, M. Neurological manifestations of coronavirus disease 2019: exploring past to understand present. Neurol Sci. 2021;42:773–85.Google Scholar
Lin, JE, Asfour, A, Sewell, TB, et al. Neurological issues in children with COVID-19. Neurosci Lett. 2021;743:135567.Google Scholar
Zubair, AS, McAlpine, LS, Gardin, T, Farhadian, S, Kuruvilla, DE, Spudich, S. Neuropathogenesis and neurologic manifestations of the coronaviruses in the age of coronavirus disease 2019. JAMA Neurol. 2020;77:1018.Google Scholar
Orman, G, Desai, NK, Kralik, SF, et al. Neuroimaging offers low yield in children positive for SARSCOV-2. AJNR Am J Neuroradiol. 2021;42:951–4.Google Scholar
Kihira, S, Delman, BN, Belani, P, et al. Imaging features of acute encephalopathy in patients with COVID-19: a case series. Am J Neuroradiol. 2020;41:1804–8.Google Scholar
Lindan, CE, Mankad, K, Ram, D, et al. Neuroimaging manifestations in children with SARS-COV-2 infection: a multinational, multicentre collaborative study. Lancet Child Adolesc Health. 2021;5:167–77.Google Scholar
O’ Neill, KA, Polavarapu, A. Acute cerebellar ataxia associated with COVID-19 infection in a 5-yearOld boy. Child Neurol Open. 2021;8:2329048X211066755.Google Scholar
Korkmazer, B, Ozogul, M, Hikmat, E, et al. Posterior reversible encephalopathy syndrome in a pediatric COVID-19 patient. Pediatr Infect Dis J. 2021;40:e240e242.Google Scholar
Abdel-Mannan, O, Eyre, M, Löbel, U, et al. Neurologic and radiographic findings associated with COVID-19 infection in children. JAMA Neurol. 2020;77:1440.Google Scholar
Monti, G, Giovannini, G, Marudi, A, et al. Anti-NMDA receptor encephalitis presenting as new onset refractory status epilepticus in COVID-19. Seizure. 2020;81:18–20.Google Scholar
Manzano, GS, McEntire, CR, Martinez-Lage, M, Mateen, FJ, Hutto, SK. Acute disseminated encephalomyelitis and acute hemorrhagic leukoencephalitis following COVID-19. Neurol Neuroimmunol Neuroinflamm. 2021;8:e1080.Google Scholar
Figure 0

Figure 1. Mechanism of action of innate immunity. Identification of PAMPs by PRRs, in this case toll-like receptors/TLRs, leads to activation of signalling pathways which cause production of antimicrobial molecules and incite inflammation. DAMPs on the other hand are released from dead and/or dying host cells causing a sterile inflammatory response. (Created with BioRender.com). DAMPs = damage associated molecular patterns; PAMPs = pathogen-associated molecular patterns; PRRs = pattern recognition receptors; TLRs = toll-like receptors.

Figure 1

Figure 2. Diagrammatic representation of the components of innate and adaptive immunity. (Created with BioRender.com).

Figure 2

Figure 3. Schematic depicting microglial polarization states and function. Under physiological conditions microglia perform active surveillance of the CNS (blue pathway). Multiple factors (e.g., colony stimulating factor 1 receptor [CSF1R], signal regulatory protein CD172 [SIRP1A], chemokine [CX3CL1 and CD200R]), help maintain this state. When triggered (red pathway) (e.g., exposure to lipopolysaccharide [LPS], IFN-γ, or GM-CSF), microglia acquire M1 phenotype (pro-inflammatory) which leads to neurotoxicity. Alternate pathway activation by IL-1, igG, or IL-10 causes microglia to attain an anti-inflammatory (M2) state (green pathway) ensuring neuroprotection. (Created with BioRender.com). CNS = central nervous system; GM-CSF = Granulocyte-macrophage colony-stimulating factor.

Figure 3

Figure 4. Anatomy of the blood–brain barrier. (Created with BioRender.com).

Figure 4

Figure 5. Pott’s puffy tumor is a prototypical example of infection breaching multiple CNS barriers to gain access to the brain parenchyma. This example demonstrates Pott’s puffy tumor secondary to complicated frontal sinusitis in a 14-year-old year old presenting with fever, facial swelling and altered sensorium. Coronal fat-saturated T2-weighted sequence (A) through the face reveals heterogeneous contents within the left frontal sinus (arrow), obstructing the left fronto-ethmoid drainage pathway. Hyperintense mucosal thickening is also seen in the left maxillary sinus. Axial fat-saturated T2-weighted (B), diffusion-weighted (C), and contrast-enhanced axial 3D-T1 (D) sequences reveal multicompartment extra-calvarial and intracranial infective collections. A large, loculated heterogeneous abscess, with restricted diffusion, is seen in the frontal scalp tissues demonstrating peripheral enhancement and multiple septations within (arrows in B-D). Note the ill-defined T2-weighted hyperintensity in the frontal bone, representing osteomyelitis, with a focal bony breech (thin arrow in B). Intracranially, similar collections are seen in the right frontal epidural space (stepped arrows in B-D) and the right frontoparietal subdural spaces (dotted arrow in B-D) representing empyemas, as well as within the right frontal lobe parenchyma (arrowheads in B-D) representing abscesses. Contrast-enhanced axial FLAIR sequence (E) also reveals enhancement within the right parietal sulcal spaces, suggestive of meningitis (curved arrow). Phase contrast MR venography (F) reveals loss of flow signals along the rostral segment of the superior sagittal sinus due to infective thrombophlebitis. Overall, these findings are a prototypical representation of microorganisms breaching multiple physical and physiological CNS barriers. CNS = central nervous system; FLAIR = Fluid Attenuated Inversion Recovery.

Figure 5

Figure 6. Diagrammatic representation of the choroid plexus demonstrating the blood-CSF barrier. Blood vessels within the choroid plexus are devoid of tight junctions. However, the outer epithelial layer demonstrates tight junctions which limits and monitors egress of molecules into the CSF. The subependymal space is patrolled by cells of innate and adaptive immunity even under normal physiological states. (Created with BioRender.com). CSF = cerebrospinal fluid.

Figure 6

Figure 7. Parechoviral infection in a neonate presenting with fever and seizures. Diffusion-weighted imaging (DWI; A-F) reveals near symmetric restricted diffusion (ADC not shown) extensively involving the cerebral white matter bilaterally. Note the linear, radiating pattern of the signal abnormality (arrows) towards the subcortical and juxta-cortical white matter. Restricted diffusion is also noted in the corpus callosum (dotted arrows in C) and the dorsolateral thalami (dotted arrows in D), suggestive of preWallerian degeneration ADC = apparent diffusion coefficient.

Figure 7

Figure 8. Pathophysiology of cerebral malaria. The rupture of infected red blood cells leads to the release of merozoites which activate macrophages and dendritic cells which in turn causes the release of TNF and IFN-γ. These increase the expression of VCAM1 and ICAM1 on the endothelial cells in the BBB which further recruits more infected RBS, hence promoting inflammation and stimulating pro-coagulation pathways. PRRs detect malarial PAMPs and trigger the secretion of cytokines including IL-6, IL-12, and TNF. This also leads to increased recruitment of basophils, neutrophils, and NK cells. Upon activation these cells release chemokines (MIP-1 α and MIP-1β) which damage the integrity of the BBB. In addition, the endothelial cells of the BBB present the malarial antigen to CD8+ T cells in the brain which incites CD8+ T cell–mediated endothelial cell death (mediated by granzyme B). Following BBB breakdown, parasite-derived proteins and peripheral immune cells gain access to the brain. This triggers astrocyte and microglial activation which increases proinflammatory cytokines and chemokines production (e.g., MAP kinase-mediated cytotoxic pathway) which causes axonal loss. (Created with BioRender.com). BBB = blood–brain barrier; NK = natural killer; PRRs = pattern recognition receptors.

Figure 8

Figure 9. Cerebral malaria in patient presenting to the emergency department in deep coma. Axial DWI (A, C) and corresponding ADC (B, D) images show restricted diffusion involving the juxta- and subcortical white matter bilaterally. A diagnosis of CM was made based on peripheral smears and the patient survived after intensive treatment with anti-malarial regimens and supportive management. ADC = apparent diffusion coefficient; CM = cerebral malaria; DWI = diffusion-weighted imaging.

Figure 9

Figure 10. FIRES in a 10-year-old presenting with acute failure of verbalization, reduced ambulation and poor intake along with abnormal repetitive movements of hands and pursing of lips. Axial FLAIR sequence (A-C) shows extensive, symmetric, gyriform hyperintense cortical swelling in the temporal lobes (A), the insula (B), and the frontal lobes (with a mesial predominance; C). Similar hyperintense swelling is seen involving the lentiform nuclei bilaterally. Axial DWI (D) and ADC (E) sequences show corresponding gyriform restricted diffusion in the temporal lobes. There was no restricted diffusion seen in the rest of the affected brain parenchyma. Coronal contrast-enhanced 3D-T1sequence (F) reveals no evidence of contrast enhancement. A differential diagnosis of infection, including viral etiologies such as herpes, as well as anti-NMDAR encephalitis were considered at the time of initial presentation. A biopsy of the temporal lobe revealed focal spongiotic changes with apoptotic neurons and neuronal loss but no immunohistochemistry was negative. Based on the clinical presentation and the pathological results, FIRES was considered and the patient was treated with methylprednisolone, IV immunoglobulin and plasma exchange (PLEX). ADC = apparent diffusion coefficient; DWI = diffusion-weighted imaging; FIRES = febrile infection-related epilepsy syndrome.

Figure 10

Figure 11. Claustral sign in a 3-year-old with suspected FIRES presenting with multiple seizures, following a bout of febrile illness. Axial FLAIR (A) and coronal T2-weighted (B) sequences reveal hyperintensities in the subinsular regions/claustrum bilaterally (arrows). Axial DWI (C) and axial contrast-enhanced 3D-T1 sequence (D) demonstrate no corresponding restricted diffusion or contrast enhancement, respectively. DWI = diffusion-weighted imaging; FIRES = febrile infection-related epilepsy syndrome.

Figure 11

Figure 12. Rasmussen encephalitis in a 6-year-old patient presenting with sudden onset absence-like complex partial seizures originating from the right hemisphere. Axial FLAIR sequence (A, B) shows areas ill-defined areas of hyperintense signal involving the cortex and the juxtacortical white matter of the right parietal (A) and frontal (B) lobes. DWI (not shown) revealed no restricted diffusion in these locations. Corresponding contrast-enhanced 3D-T1 sequence (C, D) reveals no evidence of enhancement in these regions. DWI = diffusion-weighted imaging.

Figure 12

Figure 13. Evolution of Rasmussen encephalitis. Axial FLAIR (A) and axial T2-weighted (B) sequences reveal a small nonspecific focus of hyperintensity in the right frontal lobe (dotted arrows). Axial 3D-T1 sequence (C) shows normal volume of the right cerebral hemisphere. Follow-up MRI (D-F) obtained after 3 years due to medically refractory focal epilepsy, occurring multiple times daily, and emerging left hemiparesis. Axial FLAIR (D), axial T2-weighted (E), and axial 3D-T1 (F) sequences reveal marked volume loss of the right cerebral hemisphere, seen in the form of ex-vacuo dilation of the right lateral ventricle and prominence of the sulcal spaces. Note, apart from non-specific FLAIR/T2 hyperintensity along the right frontal horn, there is no other signal abnormality seen in the rest of the brain parenchyma.

Figure 13

Figure 14. Pathophysiology of anti-NMDAR encephalitis. (A) Simplified diagrammatic representation of neurotransmission at a glutamatergic synapse and (B) simplified diagrammatic representation of antibodymediated internalization. Binding of antibodies to the GluN1 subunit causes crosslinking of NMDAR, hence forming clusters (not shown). These are removed from the post-synaptic surface by endocytosis (receptor internalization). This eventually leads to a decreased density of NMDAR at the postsynaptic membrane. The net result is slowing down of glutamate evoked postsynaptic impulses. (Created with BioRender.com). NMDAR, N-methyl-D-aspartate receptor.

Figure 14

Figure 15. Imaging findings in a 12-year-old patient with anti-NMDAR encephalitis. Axial FLAIR (A) and T2 (B) images reveal abnormal hyperintense swelling of the left frontal cortices (arrow), with resultant effacement of the left frontal sulci. Abnormal signal is also seen involving the subjacent white matter. Axial ADC (C) obtained at the corresponding level shows patchy areas low signals representing restricted diffusion (arrow) on the background of hyperintense signals representing vasogenic edema. Coronal FLAIR (D), axial T2 (E), axial ADC (F) images obtained through the temporal lobes reveals similar hyperintense swelling of the left anterolateral temporal cortices (arrows). Contrast was not administered as the patient was suffering from concomitant chronic renal failure. ADC = apparent diffusion coefficient; NMDAR = N-methyl-D-aspartate receptor.

Figure 15

Figure 16. Atypical imaging presentation of anti-NMDAR encephalitis in a 2-year-old presenting with new-onset seizures, behavioral changes, aphasia, and fever. Axial (A) and coronal oblique FLAIR (B) sequences reveal hyperintense signal abnormality within the hippocampal formations bilaterally (arrows) as well as gyriform, FLAIR hyperintense swelling of the left occipitotemporal gyrus (dotted arrow in A) and the left parahippocampal gyrus (dotted arrow in B). Contrast-enhanced axial FLAIR images (C, D) reveal diffuse leptomeningeal enhancement (stepped arrows) along both cerebral hemispheres. The patient received high-dose intravenous steroids with clinical status reaching back to baseline within a few weeks. NMDAR = N-methyl-D-aspartate receptor.

Figure 16

Figure 17. Schematic of the ANE1 pathophysiology. The nuclear pore complex is a complex macromolecule located in the nuclear envelope. It serves as an important conduit for transport across the nuclear membrane. Argonaute (AGO protein) is involved in messenger RNA (mRNA) silencing, by facilitating attachment of micro-RNA (miRNA-induced silencing complex; a silencing RNA) to the target IL6 mRNA. The RANBP2-mediated SUMOylation ensures a stable attachment of during AGO to IL-6 mRNA passage during its transport across the nuclear pore protein. The SUMOylated AGO bound IL-6 mRNA later attaches to miRNA-induced silencing complex, thereby ceasing the production of IL-6. While the exact role of mutated RANBP2 interaction in this sequence of events is not clearly delineated, but failure of miRNA-induced silencing complex attachment to an otherwise unstable AG0 bound IL-6 mRNA complex may lead to sustained production of cytokines and hypercytokinemia. Adapted from reference 24. (Created with BioRender.com) ANE1 = acute necrotizing encephalopathy.

Figure 17

Figure 18. ANE1 due to RANBP2 mutation in a 4-year-old child presenting with acute encephalopathy. Axial T2-weighted sequence (A) shows symmetric involvement of bilateral thalami (solid arrows). An evolving trilaminar pattern can be identified with a central area of T2 hyperintensity surrounded by a rim of intermediate signal and a peripheral rind of hyperintensity. A small focus of T2-weighted hyperintensity is also noted along the posterior aspect of the right lentiform nucleus (dotted arrow in A). Axial DWI sequence (B) shows patchy areas of restricted diffusion in the core of the thalamic lesions. Axial multiplanar gradient recalled (MPGR) sequence (C) shows areas of susceptibility within the left thalamic lesion suggesting foci of hemorrhage. Axial FLAIR sequence (D) shows homogeneous hyperintense swelling of the thalami as well as the right lentiform nucleus posteriorly. Axial (E) and coronal (F) contrastenhanced 3D-T1 sequence shows subtle enhancement within the thalamic lesions. DWI = diffusion-weighted imaging.

Figure 18

Figure 19. Schematic representation of HLH pathogenesis. (Circuit #1) in patients with familial HLH and some other forms of HLH related to primary EBV infection, genetic defects of virucidal property (e.g., Perforin 1 (PERF1) deficiency) causes sustained presentation of antigens and activation CD8+T cells, amplified by IFN-γ release and MHC-I upregulation. Prolonged IFN-γ release is also though to cause tissue macrophage/monocyte activation consequently leading to release of inflammatory mediators (IL-1b, IL-6, IL-18, IL-12). (Circuit #2) macrophages/monocytes are activated by activated by chronic inflammation and TLR stimulation (infection). In HLH associated with malignancies, malignant cells may drive HLH through autonomous cytokine release or presentation of EBV antigen. Drug (CAR-T cell therapy) induced occurs due to activation of therapeutic CD8+ CAR-T cells. (Created with BioRender.com). EBV = Epstein–Barr virus; CAR-T = Chimeric antigen receptor T cell; HLH = hemophagocytic lymphohistiocytosis.

Figure 19

Figure 20. HLH in a 12-year-old with long-standing immune dysregulation (fever, cytopenia, splenomegaly). Axial FLAIR (A, B) images reveal bilateral, near-symmetric, gyriform FLAIR hyperintensity with cortical swelling with similar signal abnormality in the juxtacortical white matter in the frontal and parietal lobes (arrows in A, B). Axial FLAIR (C) reveals patchy hyperintensities in the cerebellar hemispheres as well (C). Contrast-enhanced axial 3D-T1 sequences (D-F) reveal multiple, punctate foci of cortical enhancement, most accentuated in the frontal lobes, more on the left (arrows in D), with few foci of enhancement in the right parietal lobe (arrow in E), and left cerebellum (arrow in F).

Figure 20

Figure 21. COVID-related cerebellitis in a 6-year-old patient presenting with fever, headache, ataxia and decreased mental status. Axial DWI (A) and ADC (B) sequences reveal extensive restricted diffusion involving both the cerebellar hemispheres (arrows) with a focus of profound restricted diffusion in the left middle cerebellar peduncle (dotted arrows). Axial FLAIR (C) and T2-weighted (D) sequences reveal hyperintense swelling of the cerebellum with effacement of the interfolial spaces and the cerebellopontine angle cisterns. Contrast-enhanced axial 3D-T1 sequence (E) demonstrates no abnormal parenchymal enhancement. Delayed coronal contrast enhanced T1-weighted sequence (F) reveals leptomeningeal enhancement along the cerebellar hemispheres bilaterally. The patient was treated with intravenous immunoglobulins for 10 days along with 10 cycles of plasma exchange (PLEX). ADC = apparent diffusion coefficient; DWI, diffusion-weighted imaging.

Figure 21

Figure 22. Guillain-Barré syndrome (GBS) in a 2-year-old presenting with COVID infection detected on nasopharyngeal swab analysis, presenting with generalized weakness and gait disturbance. Sequential craniocaudal contrast-enhanced axial (A, B) and coronal (C) T1 sequences through the lumbar spine demonstrate enhancing thickening of the ventral cauda equina nerve roots. The patient was treated with a 3-day course of intravenous immunoglobulins.

Figure 22

Figure 23. MRI in an 8-year-old patient presenting with fever and decreased level of consciousness. The patient tested positive for COVID infection. Axial DWI (A) and apparent diffusion coefficient (ADC; B) show areas of restricted diffusion involving both the thalami (arrowheads) with peripheral areas of facilitated diffusion (dotted arrows in B). Axial FLAIR (C) and coronal T2-weighted (D) sequences demonstrate marked swelling of the thalami (arrows). In addition, confluent T2-weighted hyperintensity is seen in the pons (stepped arrow in D), extending into the medulla. Corresponding FLAIR hyperintensity was noted in the affected brainstem (not shown). Contrast-enhanced axial (E) and coronal (F) 3D-T1 sequences show faint irregular enhancement in the thalami (arrows), with a nonenhancing core, corresponding to the areas of restricted diffusion. Subtle, irregular, ring-like enhancement is also seen in the right hemipons (stepped arrow in F). These imaging manifestations suggest a parainfectious CNS complication (ANE) following COVID-19 infection occurring due to a hyperimmune response. ANE = acute necrotizing encephalopathy; CNS = central nervous system; DWI = diffusion-weighted imaging.