Meningitis and encephalitis are defined as inflammatory conditions of the meninges or brain, and often occur together as meningoencephalitis, affecting both meninges and cerebrum. Meningitis may present with various symptoms including headache, fever, painful eye movements, neck stiffness, and cranial or limb radiculopathies. However, it is specifically defined by evidence of cellular infiltration of the meninges. As such, it is generally evident through laboratory testing, finding abnormal numbers of white blood cells in the cerebrospinal fluid (CSF), and sometimes also evident radiologically through abnormal densities or enhancements of the meninges. Encephalitis is more difficult to define clinically, as the symptoms may include varying degrees and regional involvements of brain dysfunction. Indeed, many disorders are described as encephalitis even when they may involve a subacute injurious process to the brain that does not specifically include inflammation per se. (These include toxic, metabolic, mitochondrial, prion, or neoplastic disorders discussed in other chapters that best would be termed encephalopathies, rather than encephalitides.) The gold standard for defining encephalitis is brain biopsy, but this is required infrequently. Encephalitis can generally be inferred by the presence of abnormal clinical findings referable to the brain, with accompanying evidence of inflammation of the brain, often by neuroimaging. Supportive features can include electroencephalographic findings suggesting acute brain injury, and accompanying meningitis. Classically, meningitis and encephalitis were perceived as attributable to infectious etiologies. Causative organisms included more or less relatively identifiable typical or atypical bacteria, fungi, or parasites or often less identifiable viruses. However, recent decades have brought with them the advent of increasing general good health, decreasing infectious causes of these disorders, and increasingly potent molecular tools allowing better diagnosis. Thus, it has become clear that a significant fraction of meningitis and encephalitis, particularly after the period of infancy, and particularly in the developed world, has no apparent infectious cause. In recent years, a number of inflammatory syndromes without known external organismal causative underpinnings have been defined as clinicopathologic entities. The acute multiple sclerosis-like syndrome, involving mostly white matter, called acute disseminated encephalomyelitis (ADEM), still has no clear pathoetiology. However, improvements of molecular assay techniques have led to the categorization of a number of other clinical encephalitis syndromes, mostly involving cerebral gray matter, through identification of the presence in serum and/or CSF of characteristic antibodies to nervous system proteins. The antigens to which these antibodies are directed include intraneuronal proteins (Hu, Yo, Ma), synaptic proteins, and cell surface proteins. The cell surface antigens are actually the most common and most notably include the voltagegated potassium channel complex (VGKCC) receptor and glutamate N-methyl-D-aspartate receptor (NMDAR). Despite incomplete understanding of their pathogenic basis, these various encephalitic disorders characterized by presence of antineuronal antibodies are best termed the immune-mediated encephalitides. They may be further categorized as being of autoimmune idiopathic nature or autoimmune and associated with a neoplasm. These latter are termed paraneoplastic syndromes. However, because so many of the same syndromes occur in the presence or absence of a neoplasm, it is most useful to consider these disorders as immune-mediated, with or without an identified underlying associated factor such as a malignancy.

The incidence and prevalence of encephalitis is not well established due to differences in populations, diagnosis, and reporting. Recent studies suggest an overall incidence rate across all age ranges of about 50 cases of all-cause encephalitis per million persons. Of these, about 40% are infectious, 40% unknown cause, 10% ADEM, and 5% each are NMDAR and VGKCC antibody encephalitis. A number of studies have suggested that, excluding infants and excluding epidemics, NMDAR encephalitis is more common than any other single sporadic viral encephalitic etiology. Thus in adults, the incidence of VGKCC and NMDAR antibody disorders probably is each somewhat similar to that of Creutzfeldt-Jakob disease with an incidence rate of about two per million persons per year. VGKCC antibody encephalitis occurs more often in persons older than the age of 50 years and affects men more than women (about 65% men). NMDAR encephalitis is increasingly recognized in the pediatric population, and more than half of cases occur in children. A series of about 500 cases has shown a median age of onset of 21 years with a range of 8 months to 85 years. There is a strong female predominance (81%), but among patients younger than 12 years, or older than 45 years, as many as 40% of cases are male. Only about 40% of patients have tumors, but of those with tumors, more than 95% are women. Tumors are more common among patients of Asian or African ancestry. Individuals with tumors that are resected improve more quickly, require less treatment, and are much less likely to relapse. With or without tumor, 2 years after treatment, there is a good outcome in 80% of cases, with independent living, without disability, or only mild symptoms or disability. However, there is also a fatality rate of about 5%.

The general defining feature of the immune-mediated meningoencephalitides are the presence of specific autoantibodies directed toward nervous system antigens. It is not entirely clear that the antibodies themselves are pathogenic, and some have suggested that they might be bystanders in disease pathogenesis. For the antibodies directed toward intraneuronal antigens, evidence is indeed weak that the antigen relates to the syndromes and that the antibodies themselves are pathogenic; perhaps cytotoxic T cells are more responsible for disease pathogenesis and neural destruction. For the cell surface antigens, a variety of observations are suggestive
that the antibodies themselves may play an important part in the pathogenesis of the disorders. Antibody levels may correlate with disease activity; procedures to reduce immune response may result in clinical improvement; and in some cases, animal models have been developed mimicking the effects of antigen depletion. The genesis of these disorders is generally conceptualized as beginning with the development of an immune response to an autoantigen in the nervous system, either spontaneously due to exposure of the immune system to antigens displayed by a tumor, virus, or microbe or due to unmasking of otherwise privileged nervous system antigens due to brain injury or infection. The antigen provoking the response may be the actual neural antigen or a cross-reacting closely related molecule due to “molecular mimicry.” Once an immune response has occurred, there is immune-mediated attack on normal nervous system tissue. Such attack may occur after any inciting infection or injury has already resolved.

Immune-mediated encephalitides may be associated with antibodies to cell surface proteins, such as the VGKCC, NMDAR, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor, or to γ-aminobutyric acid-B (GABA_B) or glycine (GLY) receptors or may be associated with antibodies to synaptic proteins such as glutamic acid decarboxylase (GAD) or amphiphysin or may be associated with intraneuronal proteins such as Hu, Yo, Ma, or Ri (Table 71.1). Any of these syndromes may occur in the context of paraneoplasia, in which presumably some onconeural antigen on the tumor prompted the immune response, or “spontaneously” without the presence of a tumor. In general, the presumed deleterious effects of an antibody may relate simply to interference with neural transmission or to the development of immune-mediated attack on the patients neural tissue. NMDAR encephalitis is associated with the presence of anti-NR1 subunit antibodies, which apparently bind the NMDAR causing internalization of these receptors important for neurotransmission, particularly in the hippocampus. Pathologic studies of biopsies performed on patients with NMDAR antibody encephalitis often additionally show mild inflammation with perivascular infiltrates, which likely brings to the region injurious cytokines, but neurons and their connections are often intact. To the extent that there is only loss of neurotransmitter function and/or presence of noxious inflammatory molecules, elimination of the injurious antibody and the inflammatory response can allow for cellular production of new receptor proteins providing for complete recovery of function. To the extent that there is an immune attack on neurons, recovery is unlikely. In paraneoplastic cerebellar degeneration due to anti-Yo, pathologic studies show complete destructive depletion of all Purkinje cells, severely, and permanently impairing cerebellar function. Likewise in anti-Hu syndrome, there appears to be hippocampal atrophy with loss of neurons.

The clinical hallmarks of the immune-mediated meningoencephalitides are subacute deterioration of central nervous system function, accompanied typically by some evidence of an inflammatory process in the central nervous system on CSF testing and/or neuroimaging. It is often the speed, rather than the particular clinical characteristics that indicates an immune-mediated meningoencephalitis. Rapidly progressive central neurologic dysfunction, in the face of no clear evidence of toxic, vascular, neoplastic, or infectious nervous system involvement, should promptly raise the suspicion of an immunemediated or prion-mediated (see Chapter 68) disorder.

Diagnostic testing, reviewed in the following text, is essential to interpret the clinical findings. Magnetic resonance imaging (MRI) of the brain in these immune-mediated disorders may show MRI signal abnormalities in gray or white matter on T2-weighted or fluid-attenuated inversion recovery (FLAIR) imaging sequences. Signal abnormalities may also be seen on diffusion-weighted imaging (DWI) of the cortical ribbon and deep nuclei, similar to changes seen in prion diseases or ischemic disorders. However,
the distinguishing feature of the immune-mediated disorders is inflammation. Even if DWI signal abnormalities are seen, they are accompanied by more extensive FLAIR or T2-weighted imaging abnormalities not common in prion disease. And contrast-enhanced imaging may show areas of breakdown of the blood-brain barrier, definitely inconsistent with prion disease but common in inflammatory disease. But it is CSF and blood studies that best help define the immune-mediated disorders. CSF may show evidence of central nervous system inflammation such as the presence of pleocytosis and elevated protein. Cellular fluid is not characteristically seen in prion disease. Blood and CSF studies may identify marker antibodies that characterize distinct clinicopathologic entities with characteristic diagnostic, prognostic, and therapeutic features. Some of these immune syndromes are sufficiently recognizable disorders through their clinical phenotype, even though it may take some days to weeks to confirm an immune-mediated disorder with an antibody test. Other syndromes may not be as easily recognized but nonetheless may show a somewhat consistent epidemiology and clinical presentation.