Neurological syndromes driven by postinfectious processes or unrecognized persistent infections

Abstract

Purpose of review

The immune system serves a critical role in protecting the host against various pathogens. However, under circumstances, once triggered by the infectious process, it may be detrimental to the host. This may be as a result of nonspecific immune activation or due to a targeted immune response to a specific host antigen. In this opinion piece, we discuss the underlying mechanisms that lead to such an inflammatory or autoimmune syndrome affecting the nervous system. We examine these hypotheses in the context of recent emerging infections to provide mechanistic insight into the clinical manifestations and rationale for immunomodulatory therapy.

Recent findings

Some pathogens endure longer than previously thought. Persistent infections may continue to drive immune responses resulting in chronic inflammation or development of autoimmune processes, resulting in damage to the nervous system. Patients with genetic susceptibilities in immune regulation may be particularly vulnerable to pathogen driven autoimmune responses.

Summary

The presence of prolonged pathogens may result in chronic immune stimulations that drives immune-mediated neurologic complications. Understanding the burden and mechanisms of these processes is challenging but important.

INTRODUCTION

It is clear that a persistent infectious process will lead to immune activation which may be detrimental to the host but is necessary to control the infectious process. Thus, the possibility that autoimmune diseases may be driven by such an infectious process has been investigated extensively and the inability to consistently find such an infection has led to the idea that while such diseases may be triggered by infections, once the immune system is activated, the host antigens against which the immune response is mediated against will continue to drive the syndrome. This concept assumes that the infectious agent is no longer necessary. Three lines of evidence strongly support this notion. First, clinical experience has demonstrated for many postinfectious neurologic diseases, patients respond to immune-modulatory therapies. For example, acute disseminated encephalomyelitis (ADEM) is typically managed with high-dose corticosteroids [1^▪,2], although plasmapheresis [3,4] and intravenous immunoglobulin (IVIG) [5] also show benefit. However, there have not been controlled trials for the majority of immune-mediated neurologic complications of infections and the treatment regimens are often case report driven. Significantly, the loss of immune regulation via checkpoint inhibitor therapy also demonstrates the immune-mediated nature of neurological disorders such as Guillain–Barré syndrome (GBS) [6,7] and myasthenia gravis [8^▪, 9] suggesting that manipulation of the host immune responses might be sufficient to initiate the immune responses against host antigens. The second line of evidence of the immune-based pathology of these disorders stems from the rare neurologic complications associated with vaccinations (reviewed in [10]). For example, ADEM has been linked to polio vaccination [1^▪] and increase incidents of narcolepsy are associated with some forms of the influenza H1N1 vaccine [11] but not others [12]. The third line of evidence that implicates the immune system in some neurologic disorders is the induction of disease by passive transfer of immune components, not infectious agents, into animal models. For example, axonal neuropathy can be induced by implanting an antiganglioside antibody secreting hybridoma into mice [13]. Further studies have shown that passive transfer of human antibodies from patients with GBS prevent axon repair after peripheral nerve injury in mice [14]. Importantly, inoculation with lipo-oligosaccharide, the component of the infectious agent linked to GBS, in the absence of intact bacteria, induced a similar disease in rabbits along with antiganglioside antibodies [15]. Similarly, adoptive transfer of primed T cells to myelin antigen has been well studied in murine models of experimental allergic encephalomyelitis [16].

Despite these clear lines of evidence, our understanding of the mechanisms driving the observed immunopathologies is still limited. As postinfectious neurologic diseases are highly heterogeneous and a single manifestation can be associated with many different pathogens, it stands to reason that multiple mechanisms of immune dysfunction are involved in these disorders. Here, we explore the evidence for persistent infections driving chronic immune responses and evidence for self-sustained autoimmune processes after infections. Although all breaches of tolerance can result in immune-mediated pathologies, understanding the underlying disease process may ultimately be important. Identification of chronic infectious agents as the driving force for the immune response may allow for the development of antivirals or adjuvant therapies that reduce the burden of the foreign antigens. Similarly, identifying self-sustaining immune responses lays the foundation for the development of antigen-specific immune modulatory therapies. Furthermore, understanding the underlying driving forces of immune dysregulation may allow the synthesis of models to predict who may be at risk to develop these complications.

MECHANISMS DRIVING POSTINFECTIOUS IMMUNE PROCESSES

Several hypotheses exist on the pathophysiological mechanisms driving postinfectious immune-mediated disorders. These include autoimmunity driven by molecular mimicry [17], epitope spread [18], bystander activation [19], neoplasm [20], presentation of cryptic self-antigens [21], skewing of antigen presentation by inflammation [22^▪], loss of regulatory T cells [23], and pathogen or inflammation provoked damage resulting in a feed-forward loop of damage and repair that induces self-sustained autoimmune responses [24] (Fig. 1). Although there is evidence to support all of these mechanisms in systemic autoimmune diseases, there is still limited direct evidence for these processes in nervous system diseases such as ADEM, transverse myelitis, or multiple sclerosis. Elucidating the mechanisms behind the neurological complications of infectious agents is challenging for multiple reasons. First, although potentially devastating, these are fairly uncommon complications. Only during large outbreaks, such as the recent Zika virus epidemics, do patterns of neurologic diseases emerge. Still, it remains unclear why some individuals with a particular infection develop postinfectious neurologic disorders and others with the same infectious do not. Second, postinfectious neurologic complications may be recognized only months to years after illness from infection and in some cases, the neurologic manifestations may be the only symptom of infection [25,26^▪▪,27]. Access to brain tissue during the acute manifestations is another major limitation in elucidating the nature of the immune response or the infectious agent. As many infections are asymptomatic in the majority of patients, neurologic complications due to these infections maybe under-appreciated and unattributed to the infectious process. Despite these challenges, much progress has been made in our understanding of neurologic complications of infectious diseases.

FIGURE 1.

Pathophysiology of immune-mediated parainfectious CNS syndromes: The syndrome is either due to a persistent or transient infection. A persistent infection can result in a chronic inflammation which may in part be driven by the antigens on the pathogen or via epitope spreading, leading to immune response against host antigens. A transient infection can result in chronic immune responses against the host antigens due to the process of molecular mimicry between the antigens of the pathogen and the host. In some patients, a genetic predisposition can drive the autoimmune response triggered by an infection.

PERSISTENT INFECTIONS

Historically there have been several nervous system disorders that have been attributed to postinfectious immune-mediated disease processes. However, more recent insights into the persistence of infectious agents have raised doubts about a truly post-infectious nature of some of these complications. By definition, postinfectious nervous system disorders occur during the convalescent phase of a disease, in which the pathogen driving the disease process has already been eliminated or is below the threshold of detection from tissue or cerebrospinal fluid (CSF). Yet, data supporting this assumption are not often readily available. Indeed, it may be that chronic infections continue to drive immune responses to these infections, explaining relapsing symptoms observed after resolution of acute stage of infections. A classic example is immune reconstitution inflammatory syndrome seen in patients with HIV infection [28] or with progressive multifocal leukoencephalopathy due to John Cunningham virus infection [29]. In these situations, the inflammatory response far exceeds what may be necessary to control the infection. As the brain is encased in a tight bony cavity, it can only sustain small amounts of swelling before physical damage ensues. Hence, immunomodulatory therapy such as corticosteroids may be necessary even if there is a persistent viral infection [29,30]. Similarly, drawing on data from previous and recent epidemics, including West Nile virus (WNV), Zika virus, Nodding syndrome, and Ebola virus, the possibility that persistent infections may be drivers of neurologic immune-mediated disorders is discussed below which may have important implications for understanding disease mechanisms and treatments.

WNV is a Flavivirus that, like most Flaviviruses, typically causes an asymptomatic infection [31]. However, 20% of infected individuals develop West Nile fever and in approximately 1% of individuals, WNV can directly infect the central nervous system (CNS). Neuroinvasive disease can cause encephalitis, meningitis, or an acute poliomyelitis-like syndrome [32,33]. Some patients with WNV fever recover completely; however, persistent neurological complications including movement disorders [34], weakness [35], headache [34], cognitive impairments, and fatigue [34,36] have been reported. Indeed, up to one-third of patients develop a chronic fatigue lasting more than six months with the average duration of five years [37]. Importantly, the incidence of long-term neurologic complications in patients with history of WNV is not associated with neuroinvasive disease. Those without neuroinvasive disease are just as likely to develop chronic fatigue [37] and other neurologic complications [34], suggesting a peripheral mechanism for induction and maintenance of neurologic complications. When markers of inflammation were examined, patients with fatigue had increased levels of inflammatory cytokines including GM-CSF, INF-gamma, IL-2, IL-6, and IL-12p70 as compared with those patients without fatigue [37]. This may indicate that in patients who develop fatigue after WNV infection that an immunopathology is responsible for the fatigue after viral clearance. Alternatively, it may also suggest that in patients who develop fatigue the immune response remains active due to lack of complete viral clearance, resulting in sustained fatigue. Some reports have demonstrated the presence of WNV RNA as long as 6 years post infection in up to 20% of patients [38]. These findings were recently confirmed by immunogold labeling transmission electron microscopy detection of WNV from urine from patients infected up to 9 years prior [39^▪▪]. The persistence of WNV in kidney may explain both the long-term neurological complications secondary to chronic immune activation and the increased incidence of chronic kidney disease observed in patients with a history of WNV [40]. As WNV remains the most important arbovirus in the continental United States [41] increased awareness of the potential for chronic infection is important.

Zika virus is a re-emerging pathogen that has garnered much attention due to serious complications of infection on fetal development [42,43]. In addition to skewing neural development Zika virus has been reported to be associated with GBS [44–46,47^▪▪]. In a case series from seven countries the incidence of GBS was found to increase from 100–877% during periods of active Zika virus transmission as compared with pre-Zika incidence levels [47^▪▪]. GBS is classically described as a postinfectious process associated with multiple infectious agents, most commonly respiratory infections and Campylobacter jejuni [48,49^▪,50]. However, in a study following patients in Columbia, approximately 40% of the patients that develop GBS have symptom onset, while Zika virus is still detectable. The mean onset of neurologic symptoms occurred seven days after viral symptoms with a mean duration of viral symptoms lasting four days [26^▪▪]. Of particular interest, two patients had no previous symptoms of viral infection prior to onset of neurological complications [39^▪▪], suggesting Zika may also cause asymptomatic infections in people. Zika virus has also been associated with ADEM that occurred three weeks after the presentation of viral symptoms [51^▪,52]. Recent evidence suggests that maternal [53] and fetal immune responses [54] directed against the virus may contribute to the nervous system pathologies associated with Zika infection. This is important as recent studies have demonstrated that Zika can persist for extended periods in subsets of patients. Zika virus has been detected from multiple body fluids including serum, semen, urine, salvia, CSF, and vaginal secretions. In a prospective study, patients with Zika virus were followed over time and body fluids were repeatedly tested by PCR and, in a subset of samples, by viral isolation and propagation [55^▪▪]. These studies demonstrated that the median time for clearance of the virus depended upon the body fluid. For example, in urine, the median time until clearance was eight days were as in serum the median time until clearance was 14 days. However, some individuals had detectable virus in the serum for as long as 80 days postinfection. Viral clearance appears to occur more rapidly from vaginal secretions as compared with the semen where virus may persist up to 188 days [56,57]. In addition, intermittent detection of virus occurred in both serum and urine [55^▪▪], suggesting that Zika virus activity may fluctuate over time. Why some individuals have persistent viral infections and others appear to eliminate the virus quickly is unknown. In primate models of Zika infection, virus persisted the peripheral nervous system [58] and CSF [59^▪▪] three times as long as compared with the sera. Importantly, no antiviral immune response was detected in the CSF, suggesting the CNS may serve as a reservoir for the virus [59^▪▪]. The continuum of neurologic clinical presentations associated with Zika virus suggests that multiple mechanisms may contribute to the neuropathology coupled with this infection. The long-term neurological complications associated with extended Zika virus infections may not yet be recognized.

Nodding syndrome is one of many neurological complications thought to be associated with Onchocerca volvulus infection [60]. The majority of current evidence suggests this parasite cannot enter the CNS [61], and therefore, neurologic complications are hypothesized to be immune mediated [62]. Recent work has demonstrated cross-reactive immune responses against the parasite and host proteins expressed in regions of the brain impacted in children with Nodding syndrome [63]. Adequate anti-parasitic therapeutic regimes and vector control strategies have appeared to decrease the incidence of Nodding syndrome [64]. However, once a person is infected the current therapeutic strategies only kill microfilariae [65,66]. Once developed into a mature parasite current annual doses of ivermectin are unable to effectively kill the adult worm [66,67]. These major worms persist for 9–11 years [68] in a host and continue to propagate and produce further microfilariae albeit at a reduced level of productivity [69,70]. If indeed the parasite is driving the immune response responsible for neurologic complications of O. volvulus infection new therapeutics targeting the mature parasite or increased frequency and dosing of ivermectin [67] are desperately needed.

It has recently been recognized that Ebola can cause result in frank infection of the CNS [71] as cause well as long-term neurologic sequelae in survivors [72^▪]. Ebola survivors have cognitive impairments, depression, anxiety, vision and hearing disturbances, persistent headache, and extreme fatigue as compared with their non-Ebola infected counterparts [72^▪,73^▪,74,75]. Similar to the Flaviviruses, Ebola virus, a Filovirus, may also persist for extended periods in some body fluids. Infectious Ebola virus has been found in breast milk and semen for 15 and 40 days respectively after symptom resolution [76] and at least one case of sexual transmission has been reported [77]. The CNS may serve as a long-term reservoir for Ebola virus. Evidence from both animal models [78] and case reports [79^▪▪,80] strongly support this concept. However, in a small cohort, a systematic evaluation of CSF from Ebola survivors did not find any viral RNA 1 year after infection [81^▪]. The development and detection of Ebola antibodies in patients with no reported clinical signs or symptoms [82] suggest that Ebola infection may be asymptomatic in some infected individuals. This may in turn lead to an underestimate of the true burden of disease as well as make it difficult to attribute any long-term neurologic complications to an unrecognized previous infection.

SELF-SUSTAINED AUTOIMMUNE PROCESSES AFTER INFECTIONS

Infections commonly lead to upregulation of immune check point molecules such as programmed cell death protein 1 and cytotoxic T lymphocyte antigen 4 (CTLA-4) These pathways are important to prevent immune-related pathology. The molecules are critical for maintaining self-tolerance and for modulating the length and magnitude of effector immune responses to minimize collateral tissue damage [83]. Hence it stands to reason that patients with mutations in these pathways are more vulnerable to develop autoimmune disorders triggered by infections. This phenomenon has been validated in animal models where blockage of CTLA-4 results in paraneoplastic neurological disease [84]. Patients with mutations in CTLA-4 develop recurrent enhancing brain lesions with lymphocytic infiltrates in other organs as well [85]. Check point inhibitors are now being widely used for treatment of various cancers. It is being increasingly recognized that many of these patients develop a variety of autoimmune neurological complications such as myositis [86], vasculitis [87], or limbic encephalitis [88].

Some patients with herpes encephalitis may develop antibody-mediated autoimmune encephalitis after they recover from the viral infection [89]. Although the vast majority of the immune response is targeting the N-methyl D-aspartate (NMDA) receptor [90], other immune specificities including α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor, gamma-Aminobutyric acid_A receptor [91,92], and dopamine-2 receptor [93] have also been identified. No genetic abnormality or molecular mimicry has been identified as a cause of the autoimmune encephalitis. One possibility might be that as the virus buds off the neuronal cell membrane, the NMDA receptor might get incorporated on the viral envelope. The viral antigens then act as an adjuvant resulting in an immune response against the receptor.

CONCLUSION

The burden of neurological complications triggered by infections is grossly underestimated. In many cases the infection may appear benign or remote to the clinical syndrome. However, infections can cause immune-mediated syndromes that affect the nervous system by a variety of mechanisms. Although elimination of the infectious organism with appropriate antimicrobials is necessary, it may not be sufficient. Immunomodulatory therapies given either concurrently or sequentially may be needed to control the inflammation. The duration of such therapy requires close monitoring of the specific immune activation resulting in the syndrome. Viral infections are of particular concern; as for most viruses, there are no specific antiviral agents available; hence, timing of immunomodulatory therapy would be key. If the infection is persistent then such therapy could be harmful.

KEY POINTS.

• Immune-mediated complications are very common with any infectious process, although the severity maybe host-dependent.

• Persistent low-level infectious process can masquerade as a postinfectious syndrome, particularly in the CNS where detection of an infection may be difficult.

• Infections through the process of molecular mimicry and epitope spreading can trigger an autoimmune-mediated encephalitis that may perpetuate even after the infection has been cured.