Charles J. Pfau
Most infections never go beyond "flu-like" illness, but sometimes these symptoms herald the onset of neurologic diseases or hemorrhagic fevers of varying severity. Only lymphocytic choriomeningitis, Junin, Machupo, Guanarito, and Lassa viruses have demonstrated natural disease potential. About 70 percent of human lymphocytic choriomeningitis virus infections are asymptomatic or so mild that they cannot be distinguished from common respiratory or gastrointestinal illnesses. The more severe LCM infections present such a wide spectrum of manifestations that a typical case is difficult to describe. Nevertheless, headache, photophobia, listlessness, apathy, memory defects, confusion, and subtle mental difficulties are among the most common symptoms. Even though this infection can be temporarily debilitating, it is rarely fatal and even when neurologic involvement occurs, complete recovery is usually seen.
The clinical presentation of Argentine hemorrhagic fever (Junin virus), Bolivian hemorrhagic fever (Machupo virus), Venezuelan hemorrhagic fever (Guanarito virus), and Lassa fever (Lassa virus) is similar in several ways, yet sufficiently different to warrant brief mention. Disease onset usually begins with insidious progression of general malaise and fever. Progression is the norm for Argentine, Bolivian, and Venezuelan hemorrhagic fevers, in contrast to Lassa fever. Hepatitis is unusual or mild in the three Latin American hemorrhagic fevers, whereas it is frequent and moderately severe in Lassa fever. Hemorrhaging, neurologic signs and symptoms, and leukopenia and thrombocytopenia are much more common in the three Latin American hemorrhagic fevers than in Lassa fever. Both increased hemoconcentration and urinary protein are associated with a high mortality in the three hemorrhagic fevers.
The virus is round, oval, or pleomorphic, 110 to 130 nm in diameter, and enveloped. The genome consists of two distinct single-stranded viral RNA species, called L and S. The arenaviruses have ambisense genomes: the 3' half is antisense, whereas the 5' half is positive-sense. Cellular ribosomes appear to be incorporated into the virion.
Of the fifteen arenaviruses known to infect animals, five cause disease in humans: Lassa virus, Junin virus, Machupo virus, Guanarito virus, and lymphocytic choriomeningitis virus. All arenaviruses contain a set of internal cross-reacting antigens as well as species-specific envelope antigens.
Arenaviruses are thought to multiply like typical antisense RNA viruses (e.g. bunyaviruses).
Infection is suspected to be caused by invasion through broken skin or the aerosol/respiratory route. The gross pathology caused by these diseases is unimpressive and is of little help in explicating the pathogenetic mechanism.
Interferon is induced by arenavirus infection, but is of questionable benefit. The humoral response is exceptionally slow. Cell-mediated immunity is probably of prime importance.
The arenaviruses that affect humans exist in nature as benign infections in restricted rodent hosts; human disease is usually due to contact with rodent excreta.
Differential clinical diagnosis is complex; the diagnosis is confirmed only by detecting a rise in antibody titers or by isolating the virus.
Elimination of rodents is effective but often not practical; effective vaccines and antiviral agents (e.g., ribavirin) are becoming available.
The hallmark of arenaviruses is their tendency to cause persistent silent infections in their natural hosts (rodents) and severe, often lethal, disseminated disease in humans. Arenaviruses are pleomorphic enveloped particles that contain two RNA segments of virus origin and ribosome-like components. Suitable conditions for transmission of virus to humans occur in areas where humans come in contact with rodent urine that contains virus. Persistent viremia and viruria in rodents result from a slow or insufficient immune response when immunologically immature fetuses or neonates are infected. In humans, the disease is acute.
There are five pathogens of humans, who are only accidental hosts. Four cause severe hemorrhagic fever with a mortality of about 15 percent among hospitalized patients in circumscribed areas (Lassa virus in West Africa, Junin virus in the Argentine pampas, Machupo virus in Bolivia, and Guanarito virus in Venezuela). The fifth, lymphocytic choriomeningitis (LCM) virus, is much more widely distributed, but causes milder infections, often neurologic.
Only lymphocytic choriomeningitis, Junin, Machupo, Guanarito, and Lassa viruses have demonstrated natural disease potential. About 70 percent of human lymphocytic choriomeningitis virus infections are asymptomatic or so mild that they cannot be distinguished from common respiratory or gastrointestinal illnesses. The more severe LCM infections present such a wide spectrum of manifestations that a typical case is difficult to describe. Nevertheless, headache, photophobia, listlessness, apathy, memory defects, confusion, and subtle mental difficulties are among the most common symptoms. Even though this infection can be temporarily debilitating, it is rarely fatal, and even when neurologic involvement occurs, complete recovery is usually seen.
The clinical presentation of Argentine hemorrhagic fever (Junin virus), Bolivian hemorrhagic fever (Machupo virus), Venezuelan hemorrhagic fever (Guanarito virus), and Lassa fever (Lassa virus) is similar in several ways, yet sufficiently different to warrant brief mention. The incubation period is probably around 10 to 14 days (as is also true for lymphocytic choriomeningitis virus infections). Disease onset usually begins with insidious progression of general malaise and fever over a 2- to 4-day period. Progression beyond this stage is the norm for Argentine, Bolivian, and Venezuelan hemorrhagic fevers, in contrast to Lassa fever. Hepatitis is unusual or mild in the three Latin American hemorrhagic fevers, whereas it is frequent and moderately severe in Lassa fever. With Lassa fever, marked elevations of serum glutamic oxaloacetic transaminase (SGOT) is associated with a poor prognosis. Hemorrhaging, neurologic signs and symptoms, and leukopenia and thrombocytopenia are much more common in the three Latin American hemorrhagic fevers than in Lassa fever. Both increased hemoconcentration and urinary protein are associated with a high mortality in the three hemorrhagic fevers.
Arenaviruses appear in ultrathin sections as round, oval, or pleomorphic enveloped particles with a mean diameter of 110 to 130 nm. The viral envelope, which is acquired by budding through the host cell plasma membrane, carries club-shaped surface projections about 10 nm long. During morphogenesis, sandy-appearing granules resembling ribosomes are found within the unstructured interior of nascent viruses. These particles give arenaviruses their name: arena is Latin for "sand." Highly purified arenaviruses appear to contain 18S and 28S host ribosomal RNAs. These RNAs do not seem to have a required role in virus replication.
The viral RNA comes in two distinct segments or species, designated L and S. The RNAs of LCM virus and three "New World" arenaviruses (Pichinde, Junin and Tacaribe) are ambisense, as is the case for members of one genus of the Bunyaviridae (phleboviruses). In these RNAs, the 3' half is of negative polarity and the 5' half is positive. Consequently, some viral proteins are encoded in subgenomic, virus-complementary mRNA species, whereas other proteins are encoded in subgenomic, virus-sense mRNA sequences. The ambisense strategy of replication is assumed to be shared by all members of the family.
All arenaviruses contain a distinctive set of internal cross-reacting antigens. Other antigens distinct for each species are primarily structures of the envelope. In the past, intraspecific differences were recognized by pathogenicity, but now they can also be determined by oligonucleotide fingerprints. Little is known about the molecular biology of arenavirus-cell interactions.
Knowledge of the multiplication of arenaviruses is fragmentary. Most of what is known comes from studies with LCM virus. Lymphocytic choriomeningitis virus replicates in a wide variety of cell types. Although the virus receptor has not been identified, it must be highly conserved and widely distributed. Transcription of the genome and replication is confined to the cytoplasm. The small RNA in the virion encodes in the negative sense a nucleoprotein (NP), and in the positive or message sense a precursor glycoprotein (GPC), which is cleaved into two virion glycoproteins (GP1 and GP2). The large RNA in the virion encodes in the negative sense an RNA-dependent RNA polymerase (L), and in the positive sense a zinc-binding protein (Z) which binds to the ribonucleoprotein complex. The virus buds from the plasma membrane, incorporating host lipids into the virus membrane.
The arenaviruses are not ordinarily contagious among humans, and they are nonpathogenic in their rodent hosts, but rodent-to-human infections can cause severe, sometimes fatal, disease (Fig. 57-1). This type of situation is not uncommon after interspecies transmission of viruses to humans, as with Marburg and Ebola viruses (see Ch. 41). However, it remains one of the central unresolved questions in arenavirus research. Perhaps infrequent exposure has prevented the weeding out of unsuitable immune response genes in populations that for 5,000 years have been subject to major epidemics caused by viruses. The explosive nature of some of the arenavirus diseases in clinical settings may indicate rapid attenuation of the viruses as human-to-human transmission occurs.
FIGURE 57-1 Pathogenesis of lymphocytic choriomeningitis (LCM) and hemorrhagic fevers.
Although aerosol and respiratory spread, as well as cuts and abrasions in the skin, are suspected, portal of entry of arenaviruses and time course of their systemic distribution are uncertain. The onset of the hemorrhagic fevers caused by Lassa, Junin, Machupo, and Guanarito viruses may be insidious, with the disease presentation within 7 to 14 days after infection simply as pyrexia, headache, sore throat, and myalgia. Virus can be recovered from the blood and serum for up to 3 weeks after onset of the infection, and Lassa virus can be recovered from the urine for up to 5 weeks. Hemorrhagic phenomena, heralded by unremitting high fever, can begin after day 5 of illness and are followed by dehydration and hemoconcentration, shock syndrome, hemorrhagic manifestations, and cardiovascular collapse. The pantropic nature of these viruses is revealed by their presence in various dysfunctional organs.
Compared with the dramatic clinical course and mortality, the gross pathology is unimpressive and of little help in constructing a pathogenetic scheme. Complete autopsies have not been performed on patients with Lassa and Bolivian hemorrhagic fevers; however, autopsies performed on patients with Argentine hemorrhagic fever show a lack of deposited immunoglobulin and complement component C3 in the kidneys and small blood vessels. Mediators released from infected cells have a potential role in the pathogenesis of dysfunction of some target organs. Although LCM virus can produce severe human disease, characterized by prominent neurologic manifestations, pathologic lesions have not been studied extensively. However, in the mouse model the immune response against LCM virus (specifically in the T-cell compartment) is central to the development of fatal neurologic disease. Furthermore, mice infected with a lethal dose of this virus can invariably be saved by treatment with antibody to alpha/beta interferon raising the possibility that endogenous interferon alpha/beta enhances the immunopathology.
Antibodies develop following overt human infection with arenaviruses and are detectable by enzyme-linked immunosorbent assay (ELISA), complement fixation, neutralization, and fluorescent antibody techniques. The humoral response is exceptionally slow, but ultimately a long-lasting and vigorous production of antibodies occurs. Usually, antibodies demonstrable by immunofluorescence are the first to appear, followed by complement-fixing antibodies. The complement-fixing antibodies are short-lived, with titers diminishing rapidly 5 to 12 months after onset. In contrast, neutralizing antibodies remain detectable for many years. Cell-mediated immunity is important in arenavirus infections of experimental animals; it is sometimes harmful, but is probably beneficial in human infections, at least for Lassa fever. In Lassa fever passive transfer of early-convalescent-phase human antibodies does not protect monkeys or guinea pigs, whereas late antibodies neutralize virus and are protective. Induction of alpha interferon has been shown in patients with Argentine hemorrhagic fever. In general, arenaviruses are relatively resistant to the antiviral action of alpha/beta interferon. Interferon titers are significantly higher in fatal cases than in survivors (perhaps owing to higher levels of virus in the former). All evidence suggests that viral clearance in humans is complete and that chronic infection is not established. Reinfection with Lassa virus is possible, but appears to be uncommon.
The arenaviruses exist in nature as benign infections in restricted rodent hosts (Fig. 57-2). The only exception is Tacaribe virus, which was isolated from Artibeus bats. In every case in which a human arenavirus disease has been studied, an interface between humans and rodents has been described; the one common characteristic of these zoonotic infection patterns is human contact with rodent excreta. For example, Argentinean agricultural workers are exposed to Junin virus in maize fields when large numbers of reservoir rodents are disturbed by crop harvesting. In recent years, significant numbers of lymphocytic choriomeningitis infections have been attributed to silently infected pet hamsters and hamsters in biomedical laboratory colonies.
FIGURE 57-2 Transmission of arenaviruses from rodent reservoirs to humans.
Frequent and explosive hospital-acquired infections in West Africa brought Lassa virus to the attention of the medical world 25 years ago. It is now clear that the virus is transmitted mostly at the village level and that most infections are mild or asymptomatic. For those sick enough to be admitted to the hospital, mortality is about 15 percent. Outside the
hospital setting there can be an astonishingly high seroconversion rate (nearly 30 percent per year) and an even more perplexing reversion to seronegativity (about 6 percent). The case fatality rates associated with Junin, Machupo, and Guanarito viruses range from 5 to 35 percent. LCM virus, which is found worldwide in Mus musculus (the common field mouse), is considered to be the agent in about 5 percent of central nervous system infections of virus origin; these infections may be debilitating but are rarely fatal. The other arenavirusesAmpari, Flexal, Ippy, Mobala, Mopeia, Latino, Parana, Pichinde, Tacaribe, and Tamiamican cause infections in laboratory personnel, especially when high concentrations of virus are being processed.
Differential clinical diagnosis of the arenavirus hemorrhagic fevers is complex. The arenaviruses must be suspected if they are prevalent in geographic areas where infections have occurred and in regions known to harbor reservoir rodent species. Various diseases leading to sepsis, with disseminated intravascular coagulation and shock, can be confused with diseases caused by arenaviruses. Other viruses also must be considered along with LCM virus in differential diagnosis of aseptic meningitis. There is an obvious need for early clinical diagnosis of the arenavirus hemorrhagic fevers in order to begin therapeutic (chemotherapy) intervention. Initial reports, using Lassa Fever as the model, indicate that this may be achievable through use of the reverse transcription-polymerase chain reaction assay.
Junin and Machupo viruses are isolated primarily by intracerebral inoculation of newborn hamsters. Lassa virus is regularly isolated by inoculation of Vero cells. The most sensitive method for isolating LCM virus is intracerebral inoculation of weanling mice. If identification of specific viral antigens is the goal, antigen capture enzyme-linked immunosorbent assays (ELISA) are available. All arenaviruses appear to share antigenic determinants in the ribonucleoproteins, as well as antigenically distinct determinants in their outer glycoproteins. Positive immunofluorescent staining of acetone-fixed infected cells is definitive for more than just family identification, since with limiting dilutions of antibody, old World viruses (Lassa and lymphocytic choriomeningitis viruses) can be readily distinguished from New World viruses (Junin, Machupo, and Guanarito viruses). Arenavirus species may be identified by their unique surface glycoproteins and infectivity neutralization.
A successful rodent control program in areas affected by Bolivian hemorrhagic fever has been described. Although elimination of rodents that shed virus has protected humans, it is not a reasonable long-term approach for other arenaviruses, because of the rodent ecology. A live attenuated Junin virus vaccine has now been tested in about 100 volunteers, with a resulting humoral and cell-mediated response frequency of more than 95 percent. A Lassa virus glycoprotein gene has been cloned and expressed in vaccinia virus. This vaccine has offered a high degree of protection against disease and death in monkeys challenged with the intact Lassa virus. Plasma from convalescent patients has become the single specific therapeutic adjunct for patients severely ill with Bolivian and Argentine hemorrhagic fevers. Use of plasma is not indicated yet for patients with Lassa fever.
Physicians attending patients are convinced that such plasma is valuable if given during the first 8 days of disease, but more controlled trials are needed. At least seven serologically distinct strains of Lassa virus have been isolated; animal studies suggest that effective therapy should involve geographic matching of immune plasma and virus strain. Early admission to the hospital, bed rest, oral hydration, sedation, and analgesia are important. In view of the frequency of Lassa virus transmission from person to person in a hospital setting, strict measures must be taken to isolate patients who have or are suspected to have the disease. Isolation of patients with the other pathogenic arenaviruses is also probably desirable.
Although several classes of antiviral compounds have been found with specific in vitro activity against arenaviruses, only ribavirin has been proven to be effective against Lassa fever in humans. It may be used at any point in the illness, as well as for postexposure prophylaxis. Junin and Guanarito viruses have been found to be highly sensitive to ribavirin in vitro, but their efficacy in humans has not been established yet.
Demby AH, Chamberlain J, Brown DWG, Clegg CS: Early diagnosis of Lassa Fever by reverse transcription-PCR. J Clin Invest, 32: 2898, 1994
McCormick JB: Lassa, Junin, Machupo and Guanarito Viruses. p. 776. In: Webster, RG, Granoff, A (eds.): Encyclopedia of Virology. Academic Press, San Diego, 1994
Peters CJ: Arenaviruses. In Belshe R (ed): Textbook of Human Virology. 2nd Ed. PSG Publishing, Littleton, Mass, 1991
Salvato M.S. (ed): The Arenaviridae. Plenum, New York, 1993
Southern PJ, Meyer, BJ: Lymphocytic Choriomeningitis VirusMolecular Biology. p. 806. In: Webster, RG, Granoff, A (eds.): Encyclopedia of Virology. Academic Press, San Diego, 1994
Vainrub B, Salas R: Latin American Hemorrhagic Fever. Infect Dis Clinics N Amer, 8: 47, 1994
Welsh RM: Lymphocytic Choriomeningitis VirusGeneral Features. p. 801. In: Webster, RG, Granoff, A (eds.): Encyclopedia of Virology. Academic Press, San Diego, 1994