Clinical manifestations

Presentation and course

The clinical manifestations of lymphocytic choriomeningitis virus infection depend strongly on the developmental stage of the human at the time of infection (12). In particular, clinical signs and symptoms depend on whether the infection occurs during the prenatal period or during postnatal life. Congenital infection of a human with LCMV involves both the postnatal infection of a pregnant woman and the prenatal infection of a fetus. History of a maternal illness during pregnancy consistent with LCMV infection can aid in the diagnosis of congenital LCMV infection. Thus, although the focus of this chapter is congenital LCMV infection, the clinical manifestations of both postnatal and prenatal infection will be discussed.

Acquired (postnatal) LCMV infection. LCMV infection during postnatal life (during childhood or adulthood) typically manifests itself as a brief febrile illness from which the patient fully recovers (37). Classic LCMV infection is a biphasic disease in which the initial symptoms include myalgia, fever, malaise, headache, anorexia, nausea, and vomiting. Defervescence and abatement of constitutional symptoms ensues, often followed by a second phase, consisting of central nervous system disease. The symptoms of this CNS phase of the disease are usually those of aseptic meningitis, including headache, nuchal rigidity, fever, vomiting, and photophobia. The entire course of the disease is usually only 1 to 3 weeks (41).

Laboratory abnormalities frequently observed during the initial febrile phase include thrombocytopenia, leukopenia, and mild elevations of liver enzymes. Infiltrates may appear on chest radiographs. The hallmark laboratory abnormality during the second CNS phase of the illness is a CSF pleocytosis. The CSF may contain hundreds to thousands of white blood cells, most of which are lymphocytes. CSF eosinophilia has also been reported (24). Hypoglycorrhachia and mild elevations of CSF protein can also occur.

The clinical spectrum of acquired LCMV infection is broad. In as many as one-third of postnatal infections, the disease is asymptomatic. Others develop extraneural disease that extends beyond the usual symptoms and may include orchitis, pneumonitis, myocarditis, parotitis, dermatitis, and pharyngitis (51). In other patients, the CNS disease may be considerably more severe than usual and may include transverse myelitis, Guillain-Barré syndrome, hydrocephalus, and encephalitis. Although recovery is usually complete, fatalities from acquired LCMV infection occasionally occur (55).

Infections acquired by solid organ transplantation are always severe. Recipients of infected organs typically develop fever, leukopenia, and lethargy several weeks post-transplantation and then rapidly progress to multi-organ system failure and shock. Most of these cases are fatal (36; 22).

Congenital LCMV infection. Infection of the human fetus can induce spontaneous abortion and death of the fetus (08). Among those fetuses that survive, the two cardinal signs of congenital LCMV infection are (1) vision impairment and (2) brain dysfunction (07).

The vision impairment is due principally to chorioretinitis and the formation of chorioretinal scars. The most common location of the chorioretinal scarring is in the periphery of the fundus, whereas scarring of the macula is somewhat less common (02). The chorioretinopathy is usually bilateral. Other ocular abnormalities reported in congenital LCMV infection include optic atrophy, nystagmus, vitreitis, strabismus, microphthalmos, and cataract (06).

Although the vision impairment induced by congenital LCMV infection is often severe, it is the effect of LCMV on the developing fetal brain that causes the greatest disability. Prenatal infection with LCMV often leads to either macrocephaly or microcephaly. The macrocephaly observed in infants with congenital LCMV is almost invariably due to a noncommunicating hydrocephalus. The microcephaly accompanying congenital LCMV infection is due to a virus-induced failure of brain growth. Periventricular calcifications are also cardinal features of congenital LCMV infection (77; 17; 29).

Although hydrocephalus, microencephaly, and periventricular calcifications are most commonly observed, other forms of neuropathology, alone or in combination, can also occur.

These include periventricular cysts, porencephalic cysts, encephalomalacia, intraparenchymal calcifications, cerebellar hypoplasia, and neuronal migration disturbances (17).

Brain function in children with congenital LCMV infection is often severely and permanently impaired. Mental retardation, cerebral palsy, ataxia, seizure disorders, and decreased visual acuity are common neurologic sequelae (48). However, one study has demonstrated that the outcomes of children with congenital LCMV infection may be broad (17). All children with the combination of microencephaly and periventricular calcifications are profoundly neurologically impaired. Spastic quadriparesis, profound mental retardation, blindness, and medically refractory epilepsy are typical of this group. But other children with congenital LCMV infection, who do not have the combination of microencephaly and periventricular calcifications, typically have a more favorable outcome with less severe motor, mental, and vision impairments. Children with isolated cerebellar hypoplasia are often ataxic but have only mild or moderate mental retardation and vision loss (17). The differences in outcome among children with congenital LCMV infection are likely due to differences in the gestational timing of infection (14).

Unlike many other congenital infections, LCMV does not typically induce systemic manifestations. Birth weight is typically appropriate for gestational age. Skin rashes and thrombocytopenia, which are common in several other prominent congenital infections, are unusual in congenital LCMV infection. Hepatosplenomegaly is only rarely observed, and serum liver enzyme levels are usually normal. Auditory deficits are unusual.

Prognosis and complications

Most patients who acquire LCMV postnatally have a full recovery with no long-term sequelae. However, occasionally, LCMV infection induces persistent problems. The ventriculitis and meningitis induced by LCMV sometimes leads to the complication of noncommunicating hydrocephalus, thus, necessitating placement of a permanent ventriculo-peritoneal shunt (47). In addition, some adult patients with LCMV have lingering symptoms, including headaches, arthralgias, and neuropsychological problems that may persist for more than a year after the acute symptoms of infection have resolved (66).

The prognosis for children with congenital LCMV infection is generally poor. A meta-analysis of all reported cases of congenital LCMV infection revealed a mortality rate of 35% by 21 months of age (77). Of those that survive, most have severe neurodevelopmental disorders, including microcephaly, poor somatic growth, profound vision impairment, severe seizure disorders, spastic weakness, and substantial mental retardation (48). However, some of these children have only moderate neurologic and mental handicaps, and a few have been described as having a normal outcome (14). Hearing is relatively spared in children with congenital LCMV infection, and developmental regression is virtually absent.

Complications in children with congenital LCMV infection are nonspecific and consist of the medical problems that commonly arise in scenarios involving ventriculoperitoneal shunts, severe seizure disorders, and static encephalopathy. These complications include shunt failure or infection, aspiration pneumonia, injuries from falls, and joint contractures.

Clinical vignette

The patient was the 3486 g (appropriate for gestational age) product of a 41-week gestation born to a 25-year-old gravida 3 para two Caucasian married woman. The pregnancy was uncomplicated until the third trimester, when the mother experienced a “flu-like” illness with symptoms that included fever, headache, malaise, nausea, and vomiting. The illness was monophasic, self-limited, and ended after 3 to 4 days. Several weeks later, a routine ultrasound was performed that revealed fetal ventriculomegaly. Weekly ultrasounds were performed for the duration of the pregnancy and showed stable ventriculomegaly.

The male infant was delivered vaginally without complications. Apgar scores were 6 at 1 minute and 8 at 5 minutes. The physical examination at birth was remarkable for plagiocephaly, with a protuberance of the right parieto-occipital region. The anterior fontanelle was open, soft, and flat. The fronto-occipital head circumference was 34.0 cm (50th percentile for gestational age). The infant appeared healthy and was transported to the newborn intensive care unit because of the prenatally observed ventriculomegaly. In the NICU, the infant was irritable but alert and had moderately increased muscle tone in all four extremities but had no contractures. The remainder of the physical examination was unremarkable. In particular, no dysmorphic features were present (except for the plagiocephaly), nor was there any rash, heart murmur, or hepatosplenomegaly.

A head CT scan revealed bilateral periventricular mineralizations and an area of severe encephalomalacia adjacent to the posterior horn of the right lateral ventricle.

Cerebrospinal fluid examination revealed 0 red blood cells/µl, 4 white blood cells/µl (2 lymphocytes and 2 histiocytes), a protein content of 122 mg/dl, and a glucose content of 69 mg/dl. Blood glucose content was 85 mg/dl. Complete blood count and white blood cell differential were unremarkable.

Studies for infectious pathogens obtained during the first two weeks of life included a negative urine culture for cytomegalovirus and negative serologic studies for Toxoplasma gondii, herpes simplex virus, rubella, and Treponema pallidum (syphilis). Serologic studies for LCMV, assayed by the state hygienic laboratory using a complement fixation method and by the Centers for Disease Control and Prevention using an enzyme-linked immunosorbent assay, were consistent with intrauterine infection with LCMV. Table 1 shows the child's serologic titers for LCMV at various ages.

Table 1. Titers for Lymphocytic Choriomeningitis Virus Infection

At 3 months of age, the patient’s parents reported that he would not fixate or visually follow and that he was not yet rolling over. Examination confirmed the lack of visually directed behavior. His pupils reacted only minimally to light, and he had bilateral exotropia. Funduscopy revealed marked chorioretinal scarring bilaterally.

Further physical examination at this same age revealed that the child had poor head control and that the diffuse hypertonia, which had been evident at birth, had worsened. The child had “cortical” thumbs bilaterally and abnormally brisk muscle stretch reflexes in all four extremities. In addition, the head circumference had increased out of proportion to body growth. Repeat head CT scan revealed worsened ventriculomegaly, which necessitated placement of a ventriculoperitoneal shunt.

At 6 months of age, the child began having seizures of several varieties. One type of seizure consisted of behavioral arrest, staring, and eyelid fluttering, with duration of 30 to 90 seconds. The second type consisted of generalized tonic-clonic activity, with duration of several minutes. EEG was abnormal interictally, both during sleep and while awake, with multifocal sharp and spike-waves. MRI scan of the brain was markedly abnormal with bilateral colpocephaly, extensive regions of cerebrocortical encephalomalacia, and cerebellar hypoplasia.

The seizures were reduced in frequency with phenobarbital.

At 11 months of age, the child was seen for follow-up in the child neurology clinic. His mother reported that the child had multiple seizure types, each of which occurred several times per week. The seizures were improved, but not eliminated, by treatment with phenobarbital. His developmental progress was very slow. On examination, he continued to have plagiocephaly, was diffusely hypertonic and hyperreflexic, and he showed no visually directed behavior. Serologic studies were consistent with a past infection with LCMV.

At five years of age, the child had severe cognitive, motor, and vision impairment. He had no use of language. He was non-ambulatory but could maneuver around a room by crawling. He had severe vision deficits and would respond behaviorally to bright lights only. He continued to have occasional seizures, but these were relatively well controlled on carbamazepine monotherapy.

Biological basis

Etiology and pathogenesis

Congenital LCMV infection is due to the transplacental infection of the fetus with the arenavirus, lymphocytic choriomeningitis. In some cases, the infection may also be acquired by the fetus during the intrapartum period, presumably by exposure of the fetus to maternal blood or vaginal secretions during maternal viremia.

The mechanism by which LCMV damages the fetal human brain is unknown. LCMV is not a cytolytic virus in most cell types, including neurons. Thus, unlike herpes and several other pathogens that directly induce brain damage by killing host neurons, LCMV teratogenesis must have some other underlying pathogenesis.

The pathogenic mechanisms of LCMV have been studied extensively in the acutely infected adult mouse. Following intracranial inoculation of the adult mouse, LCMV replicates to high titers in the choroid plexus and meninges within several days. Viral antigen within these tissues becomes the target of an acute mononuclear cell infiltration driven by antigen-specific, CD8+ T-lymphocytes.

During the acute infection, LCMV replication is limited to the meningeal and ependymal membranes of the brain and to the choroid plexus, and very little virus penetrates into the brain parenchyma. Acutely infected adult mice typically die between 6 and 9 days post-inoculation as a result of this T-cell driven acute meningitis. Evidence that the disease is immune-mediated lies in the findings that generalized immunosuppression or specific deletion of T-cell subsets blocks onset of the acute disease. In these immunocompromised adult mice, viral clearance is delayed or blocked, and a persistent infection results, but the mice survive and remain free of acute disease (19).

As is true of virtually all infections, the genetic background of the host is an important factor in LCMV infectivity, immune response, and mortality (27). Among three common mice strains with different major histocompatibility complex (MHC) haplotypes, there were marked differences in CD8+ T-cell activation, viral load, and immunopathology 30 days after infection. Likewise, genetic differences in humans likely underlie the range in LCMV disease.

Although the adaptive immune response (and, in particular, T-cells) plays a critical role in LCMV-induced disease, the innate immune system plays a crucial role as well. In particular, type 1 interferons (IFN-1) can substantially worsen the effects of LCMV infection (76). IFN-1 is a cytokine that is produced and secreted by many cell types in response to pathogens. It facilitates communication among cells and can trigger protective defenses of the adaptive immune system. IFN-1 has long been thought to protect the host against viral infections due to its antiviral and immune-stimulating effects. However, studies have shown that IFN-1 can substantially worsen the vascular pathology and immune disruption of LCMV infection and can impair clearance of the virus. Conversely, blockade of IFN-1 using an IFN-1 receptor neutralizing antibody restored vascular health and the integrity of the immune system and facilitated clearance of the virus (71). Thus, IFN-1 can harm the host during LCMV infection. The results suggest that compounds or signaling pathways that interfere with IFN-1 signaling may have a therapeutic effect in LCMV infections.

The presence of other viral pathogens can also alter IFN-1 expression and can, thus, interact with LCMV. For example, ectromelia virus (ECTV) is usually lethal in mice, due to blockade of IFN-1. However, coinfection with LCMV rescues the mice via the LCMV-induced upregulation of IFN-1. Conversely, the ECTV infection partially suppresses the IFN-1 production induced by LCMV, thus, leading to diminished CD8 T-cell responses by LCMV. These findings demonstrate that coinfection of LCMV with other viruses, which is an event that likely occurs commonly in nature, can substantially alter the outcome of both infections (52). The immune-mediated mechanisms of disease observed in adult mice with LCMV probably also contribute to some of the pathology induced in the human fetus infected with the virus. In particular, the hydrocephalus commonly observed in children with congenital LCMV is likely due to ependymal inflammation within the ventricular system, particularly at the cerebral aqueduct, with a secondary blockage of CSF egress.

Mice may also serve as an excellent model for studying LCMV- induced chorioretinitis. In contrast to intracerebral inoculation, intraocular injection of adult mice with LCMV is not lethal, but instead leads to inflammation of the choroid and retina, resulting in disorganized retinal cytoarchitecture, abnormal electrophysiological responses, and vision loss, similar to that observed in children congenitally exposed to LCMV. This pathological process is primarily mediated by the host immune system and can be prevented by pharmacological, immunological, or genetic inhibition of the immune system. Successive adoptive transfer of primed CD8 T-cells into these immunocompromised animals recapitulates the disease, thus demonstrating that CD8 cells are solely responsible for induced retinopathy, with negligible contribution from CD4 cells (79).

Adult mice that were infected as neonates with LCMV have persistent infection with the virus, accompanied by long-term learning and behavior issues. Studies have shown that these behavioral effects are due, at least in part, to virus-induced interference with the production of neuroblasts in the normally mitotically active dentate gyrus and subventricular zone of adult mice. The chronic upregulation of chemokines, induced by the persistent infection, appears to underlie these defects in neurogenesis (69).

The adult mouse is probably not the best animal model system for the study of congenital LCMV infection. However, intracerebral inoculation of the neonatal rat with LCMV is a superb model system of human congenital LCMV infection. Inoculation of the neonatal rat with LCMV results in a distinct pattern of infection in which specific brain regions are consistently and heavily infected with the virus, whereas other brain regions are entirely spared (57; 56). Importantly, unlike infection of the adult mouse, LCMV infection of the neonatal rat results in the infection of multiple populations of neurons as well as infection of the meninges and choroid plexus (16).

Not all neuronal populations of the developing brain are vulnerable to LCMV infection. Indeed, the neonatal rat model has shown that the only brain regions vulnerable to neuronal infection are those in which neurogenesis is occurring (15). The converse is also true: all brain regions in which neurogenesis is occurring are vulnerable to LCMV infection.

This one-to-one correspondence between infectability and neurogenesis suggests that the metabolic machinery present within mitotically active neurons facilitates the propagation and survival of LCMV (13; 14). The periventricular calcifications commonly observed in humans with congenital LCMV infection probably reflect the infection and death of mitotically active neuronal precursors of the subependymal periventricular region.

Neuronal migration defects are often present in children congenitally infected with LCMV. Likewise, in the developing rat brain, LCMV infection disrupts the migration of recently generated neurons.

A study examining the effect of LCMV on cerebellar development in the neonatal rat has shed considerable light on the pathogenesis of congenital LCMV infection (43). In particular, when the neonatal rat is infected with LCMV, three separate forms of pathology occur in the developing cerebellum: (1) destructive lesions that lead to porencephalic cysts, (2) cerebellar hypoplasia that leads to regional or global reductions in cerebellar size, and (3) neuronal migration disturbances that lead to permanently ectopically located granule neurons. To determine which of these forms of pathology are immune-mediated and driven by T-lymphocytes, the effects of infection were compared in wild type rats and in congenitally athymic rats, which lack T-cells. If a form of pathology is immune-mediated and driven by T-lymphocytes, then that pathology will be absent in congenitally athymic rats. The study showed that the destructive lesions and the neuronal migration disturbances, which are very prominent in the wild type rats, fail to occur in the athymic rats. In contrast, the cerebellar hypoplasia occurs in both wild type and athymic rats. Thus, the destructive lesions and the neuronal migration disturbances are immune-mediated, whereas the cerebellar hypoplasia is not immune-mediated and is likely due to virus-induced interference with cellular mitosis.

The rat model of congenital LCMV infection has shown that the effect of LCMV on the developing brain depends critically on the developmental stage of the host at the time of infection (14). The cellular targets of infection, the peak viral titers in tissues, and the nature and severity of neuropathology all depend strongly on age of the developing host. For example, neurons of the cerebral cortex are highly infectible on postnatal day 1 but are not infectible just 3 days later, on postnatal day 4.

Within the developing brain, infectability and pathology can change dramatically over the course of only a few days. All of the various neuropathologic changes evident in humans with congenital LCMV infection can be recapitulated in the rat model by infecting developing rats at different developmental stages. These findings suggest that the variability in outcome among children with congenital LCMV infection is due to differences in the gestational timing of infection (15).

Further evidence that the gestational timing of infection is a critically important variable is derived from a study examining LCMV in placental tissues. Enninga and Theiler exposed first trimester and third trimester explants of human placentas to LCMV in vitro (34). They found that LCMV could infect and replicate in first trimester placentas, but not in third trimester placentas. In addition, they found that LCMV induced no innate immune response in first trimester placentas, but did induce a robust innate immune response with high expression levels of proinflammatory and antiviral cytokines in third trimester placentas. These findings suggest that, later in gestation, the placenta can exert protective effects against LCMV infection that are not present earlier in gestation. Thus, there is strong evidence that both the fetus and the placenta evolve over the course of gestation in ways that alter the outcome of LCMV infection (14; 34).

Glial cells play critical roles in the propagation and spread of LCMV within the developing brain. Astrocytes and Bergmann glial cells are the initial cells of the brain parenchyma infected with LCMV. Furthermore, these glial populations are the principal site of LCMV propagation and are the conduit by which the virus spreads throughout the brain and into vulnerable neuronal populations (13).

Glial cells may also play an important role in LCMV-induced inflammation. Astroglia are the dominant expressors of certain chemokines and their ligand/receptors (eg, CXCR 10 and IFN), and these same astroglia-expressed molecules play critical roles in controlling the pattern of T cell infiltration. Thus, astrocytes may be a major agent through which the virus spreads through the brain and a principal effector of the cytokine-chemokine cascade elicited by the infection (26).

LCMV is a highly neurotropic virus, both in the human and in the neonatal rat. The mechanism underlying this tropism for certain tissues and cell types remains unknown but may be due to the presence of a "receptor" for the virus. Certain cells may express a molecule on their external surfaces to which LCMV may become anchored in preparation for viral entry. A particular membrane glycoprotein, referred to as "alpha-dystroglycan," has been identified as a "receptor" for LCMV on cultured astrocytes (20). Cultured cells containing a null mutation for alpha-dystroglycan are not susceptible to LCMV infection, whereas cells in which the gene for alpha-dystroglycan is replaced are susceptible (65).

Alpha-dystroglycan is part of the dystroglycan complex, a set of proteins on the cell surface that spans the cell membrane and links molecules of the extracellular matrix, including laminin, agrin, and neurexin, with the internal actin-cytoskeletal machinery. The dystroglycan complex stabilizes the cell membrane and may also transmit signals from the extracellular environment to internal cellular machinery. When LCMV binds to alpha-dystroglycan, the virus displaces components of the extracellular space that normally bind to the dystroglycan complex. As a result, LCMV may produce some of its pathologic effects by disrupting dystroglycan function. For example, because the interactions of dystroglycan with extracellular laminin are important for membrane stability, disruption of these interactions by LCMV may increase the fragility of the cell membrane and contribute to the death or dysfunction of the affected cell (61).

Various strains of LCMV differ in their avidity for alpha-dystroglycan (67) and can be divided into high affinity and low affinity groups (65). In mice, the high affinity group depends on alpha-dystroglycan for cell entry, localizes to the spleen’s white pulp, dampens the cytotoxic T-lymphocyte response, and induces a persistent infection. In contrast, the low affinity group does not depend on alpha-dystroglycan for cell entry, localizes to the spleen’s red pulp, induces a strong cytotoxic T-lymphocyte response, and leads to viral clearance (65). Thus, the low and high affinity LCMV strains have different cellular targets and cause different diseases.

Interestingly, the presence of alpha-dystroglycan on a cell surface is neither necessary nor sufficient for infection, even among those LCMV strains that have a high affinity for alpha-dystroglycan. A study has shown that heparan sulfate, an abundant proteoglycan present on mammalian cells, can mediate the interaction of LCMV with cell surfaces (04). Thus, like alpha-dystroglycan, heparan sulfate may contribute to LCMV infection of certain cells. Following engagement of LCMV with alpha-dystroglycan or heparan sulfate at cellular surfaces, viral entry occurs via endocytosis and fusion with the lysosomal glycoprotein CD164 (04).

Early studies with LCMV revealed that most tissues in mice are vulnerable to infection. However, the virus has been notably absent from muscle cells. It was assumed that muscle cells must lack the receptor for the virus. However, when it was later discovered that alpha-dystroglycan is the receptor for LCMV, and that skeletal muscle contains large quantities of alpha-dystroglycan, it became evident that there must be some other explanation for the uninfectability of muscle cells. It has been shown that myotubes are refractory to LCMV infection due to impaired cell entry. In particular, the surface glycoprotein on LCMV impairs penetration through the muscle cell membrane (40).

LCMV surface glycoproteins are being studied extensively due to their key role in receptor binding, leading to internalization and fusion of the virion with the cell membrane. Studies in cell cultures using recombinant LCMV with various N glycosylation mutants have revealed that N glycosylation of these glycoproteins is important in determining viral fitness and cell tropism. Mutations leading to deletion of N glycans in the LCMV glycoprotein provide an advantage to the virus for infection of neurons but a disadvantage for infection of macrophages. Due to effects on protein folding, N glycosylation may mask antibody binding sites, which may have important implications for development of an LCMV vaccine (10).

When fetal mice acquire LCMV vertically from their mothers, the virus persists within their neurons, where the virus can alter gene expression and lead to long-term specific behavioral deficits. Mice with life-long high viral loads exhibit impaired learning and memory. Specifically deficient is the acquisition phase of learning on a novel avoidance task (46). These learning and memory problems appear to be due to viral-induced interference with adult neurogenesis. In particular, persistent LCMV infection in mice substantially impairs the generation of neuroblasts from neuroprogenitor cells within the dentate gyrus and subventricular regions. This interference with neurogenesis is likely related to chronic upregulation of proinflammatory cytokines, which can impair cellular production (69). Whether LCMV persists in humans following congenital infection and whether LCMV infection can alter human neuronal gene expression, as it does in mice, are unknown.

Epidemiology

The incidence and prevalence of congenital LCMV infection are unknown. Although the published case reports of LCMV infection during pregnancy make it clear that LCMV can be a severe neuroteratogen, it is not known whether the profoundly affected infants described in the case reports represent the typical outcome of gestational LCMV infection or whether they represent only the most severely affected cases. No prospective epidemiological or clinical studies of congenital LCMV infection have been conducted. Information regarding the incidence and spectrum of LCMV-induced teratogenicity is further limited by the fact that LCMV is not one of the infectious agents for which infants with a suspected congenital infection are routinely checked. Therefore, it is possible that congenital LCMV infection, like many other congenital infections, produces a spectrum of pathologic effects that range from minimal to profound.

Although the incidence of gestational LCMV infection is unknown, it is known that LCMV is prevalent in the environment, has a great geographic range, and infects large numbers of humans. LCMV is endemic in wild mice throughout temperate regions (01) and probably occurs wherever the species Mus musculus, its natural reservoir, exists. Until recent years, LCMV had not been detected in Africa. However, Mus musculus has advanced and colonized in many areas in Africa. As these domestic mice have spread, so has the virus. In 2015, LCMV RNA was detected for the first time in Mus musculus within a country of Western Africa. Indeed, 13% of the Mus musculus mice were found to be infected with LCMV. A later study, in 2021, found that the seroprevalence for LCMV in humans in Gabon, Central Africa is 21.5%, and the LCMV RNA could be detected in several novel rodent species, including porcupines and shrews (73). These findings demonstrate that LCMV is emerging as a new threat to the people and native rodent species of Africa (59).

An epidemiologic study several decades ago demonstrated that 9% of the house mice in urban Baltimore are infected with LCMV and that significant clustering occurs where the prevalence is higher (25). A study examining the seroprevalence of LCMV in Colombia found that 10% of wild mice are seropositive for LCMV (21). This similar prevalence to the Baltimore study suggests that the LCMV infection rate among mice is relatively stable over time and geography, at least across the Americas.

However, mouse infection rates with LCMV may not be equivalent everywhere. One factor that may greatly increase local mouse infection rates with LCMV is the presence of illegal waste sites. In illegal waste sites, because of abundant food supplies and ideal shelters, rodents have optimal conditions to breed and survive. As a result, the densities of rodents are high, thus, facilitating the transmission of rodent-borne viral infections. The prevalence of LCMV infection among mice in waste sites may also be due to a change in mouse behavior. When food resources are abundant, male mice become more aggressive and are more likely to inflict bite wounds, thus, facilitating the horizontal transmission of viral infections. One study demonstrated that, within illegal waste sites, seroprevalence for LCMV in Mus musculus was greater than 47%, a rate that far exceeds the seroprevalence rates in natural habitats (31).

Serological studies have found that 4.7% of adults in Baltimore and 5.1% of healthy black women in Birmingham possess antibodies to LCMV, indicating prior exposure and infection (25; 68). A study in 2021 found similar seroprevalence rates in Croatia, with 6.1% seroprevalence in the general population and 3.9% in pregnant women (75). Together, the data suggest that congenital LCMV infection is an underdiagnosed disease and that the virus is responsible for more cases of congenital neurologic and vision dysfunction than has previously been recognized (41; 54; 06).

Acquired LCMV infections can occur at any time of the year; however, most occur during the late autumn and winter months. This probably reflects seasonal variations in the cohabitation of humans with mice. During the cold months of autumn and winter, wild mice are driven indoors, carrying LCMV with them and increasing the likelihood of a human infection.

In the wake of the September 11, 2001 terrorist attacks on the United States, greater emphasis has been placed on the potential of bioterrorism. Along with the other arenaviruses, LCMV has been listed by the Centers for Disease Control as a Category A Priority Pathogen. LCMV ranks as a high-priority agent and as a risk to national security because it is easily disseminated, has the potential for major public health impact, and could induce public panic and social disruption (23).

Prevention

No vaccine exists to prevent LCMV infection; however, measures can be taken to reduce the risk of infection. Congenital LCMV infection will not occur unless a woman contracts a primary infection with LCMV during pregnancy. Rodents, especially house mice, are the principal reservoir of LCMV. Women can reduce their risk of contracting LCMV by minimizing their exposure to the secretions and excretions of mice. This can be accomplished most effectively by eliminating cohabitation with mice. Pregnant women should also avoid contact with pet rodents, especially mice and hamsters. Laboratory personnel who work with rodents have an increased risk of infection with LCMV. Pregnant women who work in animal care facilities or laboratories at research institutions should wear gloves, gowns, and face masks to avoid potential aerosolized or secreted LCMV.

Because congenitally infected mice can secrete large quantities of LCMV and remain essentially asymptomatic, LCMV may be rampant within animal colonies at research institutions, thus, putting laboratory workers, and especially pregnant women and their fetuses, at risk. Therefore, all animal colonies should be tested periodically for LCMV (32).

Acquisition of LCMV from solid organ transplantation represents a substantial risk to the life of the recipient. Donors with LCMV meningitis or encephalitis pose a clear risk for transmitting a fatal infection to recipients. Health care providers, transplant centers, and organ procurement organizations should be aware of the risks posed by LCMV and should consider LCMV in any potential donor with signs of aseptic meningitis but no identified infectious agent. The risks and benefits of offering and receiving organs from donors with possible LCMV infection should be carefully weighed (22).

Differential diagnosis

Confusing conditions

The principal items in the differential diagnosis of congenital LCMV infection are the other infectious pathogens that can cross the placenta and damage the developing fetus. These infectious agents are linked conceptually by the acronym "TORCHS" and include Toxoplasma gondii, rubella virus, cytomegalovirus, herpes simplex virus, and syphilis. Cytomegalovirus and toxoplasmosis are particularly difficult to differentiate from LCMV because infection with any of these three infectious agents can produce microcephaly, intracerebral mineralization, and chorioretinitis. Although clinical clues may aid in distinguishing one congenital infection from another, definitive identification of the causative infectious agent usually requires laboratory data, including cultures and serologic studies (18).

Symptomatic congenital cytomegalovirus infection is usually associated with several systemic signs that are not observed in congenital LCMV infection. These include hepatosplenomegaly, jaundice, anemia, intrauterine growth retardation, and a petechial or purpuric rash. Hearing deficits are common in infants infected with cytomegalovirus but are unusual in congenital LCMV infection. The diagnosis of congenital cytomegalovirus infection is best established by isolating the virus from urine or saliva within the first three weeks of life. Detection beyond this time point may reflect postnatal acquisition of the virus (18).

Of all of the congenital infections, toxoplasmosis most closely mimics congenital LCMV infection. Microencephaly, hydrocephalus, chorioretinitis, and intracranial calcifications are hallmarks of both congenital infections. Approximately 10% of newborns infected with Toxoplasma gondii will exhibit hepatosplenomegaly, jaundice, and rash (38). These systemic signs are usually absent in congenital LCMV infection. The two infections also tend to differ neuroradiographically. In cases of congenital toxoplasmosis, the intracranial mineralization tends to be diffuse within the brain parenchyma. In contrast, in congenital LCMV infection, the mineralizations are typically periventricular. Congenital infection with toxoplasmosis and LCMV are so clinically similar, however, that differentiating among them requires laboratory testing. The diagnosis of congenital toxoplasmosis can be established by serologic studies and confirmed by detecting the infectious organisms in tissues, blood, or cerebrospinal fluid.

Zika virus has emerged as a major pathogen worldwide, especially in the Americas. An arbovirus that is spread by the bite of mosquitoes, Zika can cause congenital infections that, like LCMV, can induce severe brain and retinal injuries. Zika is further similar to LCMV in that it is an intensely neurotropic virus that targets neuroprogenitor cells and tends to spare other fetal organs. Congenital Zika infection leads to microencephaly that is often even more severe than that observed with LCMV. The most severe form of microencephaly is referred to as fetal brain disruption sequence, in which the sutures overlap and there are scalp rugae. This condition is commonly induced by Zika, but rarely occurs with LCMV. Both Zika and LCMV can induce cerebral calcifications. However, in congenital Zika virus infection, the calcifications tend to occur at the gray-white matter junction, whereas in LCMV they are usually periventricular (50).

The incidence of congenital rubella syndrome has diminished since the institution of universal immunization, but the condition has not been eliminated and must be considered in any newborn with signs suggestive of a congenital infection. A woman contracting rubella during pregnancy typically has symptoms of sore throat, headache, fever, malaise, and myalgias. These symptoms of rubella closely resemble the usual symptoms of acquired LCMV infection. Congenital rubella syndrome, however, typically includes intrauterine growth retardation, cataracts, congenital heart disease, hepatosplenomegaly, jaundice, purpuric rash, and sensorineural hearing loss, which are systemic signs not typically present in congenital LCMV infection. Congenital rubella syndrome can be confirmed by isolating rubella virus from urine, cerebrospinal fluid, or blood. Detecting rubella-specific IgM or persistently high levels of rubella-specific IgG in an infant's sera also strongly supports the diagnosis of congenital rubella infection.

Herpes simplex viruses types 1 and 2 are prevalent viruses that commonly cause substantial damage to newborn infants. Like LCMV, infection with Herpes simplex virus can cause chorioretinitis and encephalitis during the newborn period. However, most cases of neonatal Herpes simplex virus infection are acquired during the intrapartum or early postnatal periods, unlike congenital LCMV infection, which occurs prenatally. Thus, unlike cases of congenital LCMV, most infants with Herpes simplex virus will not have microcephaly or hydrocephalus at birth. In addition, many neonates with the infection will exhibit multiple systemic signs, including vesicular skin lesions, conjunctivitis, hepatic dysfunction, pneumonitis, and disseminated intravascular coagulopathy. These systemic signs are typically absent in congenital LCMV infection. The diagnosis of Herpes simplex virus infection can be confirmed by polymerase chain reaction or by isolating the virus from the oropharynx, cerebrospinal fluid, skin vesicles, or conjunctiva.

Congenital syphilis, the disease caused by infection of the fetus with the spirochete Treponema pallidum, is an important cause of developmental brain injury worldwide, despite rigorous public health measures and effective antibiotic therapy. Like congenital LCMV, congenital syphilis can induce retinopathy, meningo-encephalitis, developmental delay, and seizure disorders. However, congenital syphilis can usually be recognized by the physical stigmata accompanying the infection. Features of congenital syphilis in the neonate reflect disseminated infection and include low birth weight, skin rash (particularly on the palms and soles), lymphadenopathy, hepatosplenomegaly, jaundice, persistent rhinitis, and periostitis or osteochondritis. The osseous abnormalities, which are the most consistent clinical features of congenital syphilis, may inhibit limb movement (pseudoparalysis) and are absent in congenital LCMV infection. Infants with suspected congenital syphilis require a thorough evaluation that should include radiography of long bones, cerebrospinal fluid examination, serologic assays, and histopathologic examination (with dark field microscopy) of the umbilical cord, placenta, and skin lesions.

The differential diagnosis of congenital LCMV infection also includes several noninfectious entities. Chromosomal abnormalities are prominent causes of microencephaly. However, abnormalities in the structure or number of chromosomes commonly induce dysmorphic features (especially of the hands, feet, and facies) or structural abnormalities (especially of the heart or genitourinary system) that are not observed in congenital LCMV infection.

Several genetic disorders can mimic congenital LCMV infection (63; 35). In particular, Aicardi-Goutières syndrome is an autosomal recessive disorder that often presents as neonatal encephalopathy and intracranial calcifications (62). However, its progressive course and identifiable mutations in the TREX1 and RNASEH2 genes distinguish it from congenital LCMV infection.

A second genetic disorder that mimics congenital LCMV is pseudo-TORCH syndrome (60; 44). In this autosomal recessive disorder, infants have many of the classic features of the common congenital infections that gave rise to the TORCH acronym (toxoplasmosis, other infections, rubella, cytomegalovirus, herpes viruses). However, in pseudo-TORCH syndrome, no serological or microbiological evidence of a congenital infection is ever identified. Congenital LCMV infection can be distinguished from pseudo-TORCH syndrome in several ways. First, most mothers of infants with congenital LCMV infection have a history of exposure to wild mice and experience a definite “flu-like” illness during the pregnancy; these historical factors are typically absent in pseudo-TORCH syndrome. Most importantly, chorioretinitis is present in all cases of congenital LCMV infection and absent in cases of pseudo-TORCH syndrome (11).

One other genetic disorder, Aicardi syndrome, can closely mimic congenital LCMV infection (77; 78; 42). Aicardi syndrome causes lacunar retinopathy and infantile seizures, and is usually accompanied by microencephaly and developmental delay. In addition, Aicardi syndrome can induce polymicrogyria or pachygyria and ventriculomegaly. All of these features of Aicardi syndrome are likewise commonly observed in congenital LCMV. However, agenesis of the corpus callosum and vertebral anomalies, which are hallmarks of Aicardi syndrome, are not clinical features of congenital LCMV infection. Nevertheless, Aicardi syndrome shares so many features in common with congenital LCMV infection that some authors have suggested that all children suspected of Aicardi syndrome should undergo LCMV serology testing (42).

Finally, infants with tuberous sclerosis, a genetic disorder with an autosomal dominant inheritance pattern and a high spontaneous mutation rate, may have intracranial calcifications and seizures (especially infantile spasms), which are also clinical features of congenital LCMV infection. However, unlike congenital LCMV infection, infants with tuberous sclerosis will often have affected family members and will not typically have microcephaly or hydrocephalus (though they may develop hydrocephalus later in life as a result of periventricular brain tumors).

Diagnostic workup

Acute human LCMV infections can be diagnosed by isolation of the virus from cerebrospinal fluid. However, by the time of birth, a baby prenatally infected with LCMV may no longer harbor the virus. Thus, congenital LCMV infection is usually diagnosed by means of serologic testing. The immunofluorescent antibody test detects both IgM and IgG and has greater sensitivity than the more widely available complement fixation method (49). The immunofluorescent antibody test is commercially available, and its specificity and sensitivity make it an acceptable diagnostic tool. An even more sensitive test for the detection of congenital LCMV infection is the ELISA, which measures titers of LCMV IgG and IgM and is performed at the Centers for Disease Control and Prevention.

Although serology may be utilized to confirm congenital LCMV infection, such serological testing may be unavailable, or healthcare providers may neglect to consider congenital LCMV infection. In these situations, knowledge of prenatal neuroimaging findings consistent with congenital LCMV infection may enhance its detection and diagnosis. A case of congenital LCMV was detected following visualization of several abnormal head ultrasound findings, including dysmorphic caudate nuclei with calcifications, enlarged subarachnoid fluid spaces, and cerebellar hypoplasia (72). Likewise, several cases of congenital LCMV have been detected postnatally by head ultrasound. In each of these cases, maternal history was consistent with rodent exposure, and serological data confirmed prenatal LCMV infection (64).

Polymerase chain reaction has been utilized as a means of detecting LCMV RNA in an infected infant (33). The use of polymerase chain reaction offers exciting possibilities for detection of LCMV infection. Polymerase chain reaction for the detection of viral genes is both highly specific and sensitive. Indeed, real-time polymerase chain reaction can detect genomic sequences reliably in CSF, even when the viral concentration is less than 10 live viruses per milliliter (28). However, LCMV does not induce persistent infections in humans, and the time course of viral clearance from an infected human fetus is unknown. A fetus may sustain substantial brain damage from LCMV but effectively clear the virus and have no LCMV RNA to be detected by polymerase chain reaction in the postnatal period. Thus, in many cases, it would be most ideal to test for LCMV during the prenatal period. In 2017, the first case was published describing the use of polymerase chain reaction to prenatally diagnose a fetus infected with LCMV (30). The mother, who had been exposed to mice, had suffered from a flu-like illness during the 16th week of pregnancy. Subsequent prenatal ultrasounds revealed fetal ventriculomegaly and hyperechogenicity of cerebral parenchyma. Amniocentesis at 24 weeks gestation produced amniotic fluid with negative polymerase chain reaction screening for the common “TORCH” infections but was positive for LCMV. Thus, polymerase chain reaction can be used to diagnose congenital LCMV infection prenatally.

Management

There is no specific treatment for LCMV infection. An effective antiviral therapy for LCMV infection has not yet been developed. Ribavirin has had some mixed success in the treatment of severe infections but is limited to off-label use and is associated with substantial toxicity. Ribavirin displays a dual mutagenic and inhibitory activity on LCMV that can be relevant to future treatment designs (58). Another promising antiviral compound is favipiravir, a pyrazine derivative with broad antiviral activity against RNA viruses. By disrupting early stages of viral replication, favipiravir has robust antiviral activity against arenaviruses (53). In addition, unlike ribavirin, favipiravir has little cellular toxicity. However, to date, favipiravir’s antiviral effects against LCMV have been tested only in cell culture. Whether favipiravir could protect animals or humans infected with LCMV is unknown.

Valproic acid, a medication commonly used to treat epileptic seizures, has also shown promise in controlling LCMV infection by inhibiting budding and release of viral particles. To date, the effect of valproic acid on LCMV infection has been studied only in culture, but the findings suggest that therapeutic doses used to treat epilepsy could produce antiviral effects (74).

One study, conducted in vitro, has shown that arbidol, an approved influenza antiviral, when combined with other approved drugs, including aripiprazole, amodiaquine, sertraline, and niclosamide, can inhibit the infection of cells by arenaviruses, including LCMV. This study provides proof of concept that combinations of approved drugs could act synergistically to inhibit LCMV infections (39).

Because most acquired LCMV infections have relatively mild symptoms that resolve spontaneously, no treatment is necessary. In cases in which LCMV produces more severe symptoms of meningitis or encephalitis, a course of steroids may substantially improve the symptoms and hasten recovery (66).

Children with hydrocephalus due to congenital LCMV infection often require placement of a ventriculoperitoneal shunt during infancy. Seizures often begin during early postnatal life, are difficult to control, and require administration of multiple antiepileptic medications. The mental retardation induced by congenital LCMV infection is often profound. In most cases, affected children should be referred for education intervention during early life. The spasticity accompanying congenital LCMV infection is often severe. Although physical therapy can help to maintain range of motion and minimize painful spasms and contractures, implantation of a baclofen pump is often necessary and helpful.

In many cases, identifying the source of the LCMV infection for a patient is important in order to take potential action against the offending rodent infestation and to minimize the risk of infection to others. This often involves trapping mice in homes and businesses and confirming the LCMV infection through necropsy, serology, and tissue testing of the trapped mice. However, trapping mice can sometimes be problematic and time-consuming. One report has demonstrated that the source of LCMV can sometimes be confirmed by testing fecal pellets of rodents by polymerase chain reaction (70).

Special considerations

Pregnancy

Pregnancy is a critically important issue with respect to LCMV infection. Most adults infected with LCMV have only a mild flulike illness and fully recover. However, when the infection occurs during pregnancy, the fetus can be severely and permanently affected. Because house mice and hamsters are reservoirs of the virus, pregnant women should avoid cohabitation with or exposure to these rodents.