Enterovirus, Parechovirus, and Saffold Virus Infections

Taxonomy

Enteroviruses, parechoviruses and Saffold viruses are RNA viruses belonging to the family Picornaviridae (pico = small). They are grouped together because they share certain physical, biochemical, and molecular properties. In electron micrographs, the viruses are seen as 27- to 30-nm particles that consist of naked (nonenveloped) protein capsids, constituting approximately 70% to 75% of the mass of particles, and dense central cores containing the single-stranded message-sense genomic RNA.

The original classification of human enteroviruses is shown in Table 25-1 . The enteroviruses were originally distributed into four groups based on their different effects in tissue culture and pattern of disease in experimentally infected animals: polioviruses (causal agents of poliomyelitis in humans and nonhuman primates); coxsackie A viruses (coxsackievirus A), associated with herpangina, human central nervous system (CNS) disease, and flaccid paralysis in suckling mice); coxsackie B viruses (coxsackievirus B) (human CNS and cardiac disease, spastic paralysis in mice); and the echoviruses (nonpathogenic in mice and not initially linked to human disease).

Although this scheme was initially useful, many strains were subsequently isolated that do not conform to such rigid specificities. For example, several coxsackievirus A strains replicate and have a cytopathic effect in monkey kidney tissue cultures, and some echovirus strains cause paralysis in mice. For this reason, and to simplify the nomenclature, subsequent enteroviruses were assigned sequential numbers. After this convention, the prototype enterovirus strains Fermon, Toluca-1, J 670/71, and BrCr (identified from 1959-1973) were numerically designated enterovirus (EV) 68 through 71, respectively. Additional enteroviruses continued to be identified that could not be identified using antisera specific for the classic serotypes. More than 50 additional such EV types have been assigned, although not all have been linked to human disease.

Complicating matters somewhat, studies of echoviruses 22 and 23 found that they exhibited genomic and proteomic differences from other enteroviruses, and hence they were reclassified in the new genus Parechovirus as parechoviruses types 1 and 2. Similarly, hepatitis A virus was initially assigned the designation of EV72 but was reclassified as the sole member of the Hepatovirus genus within the picornavirus family because of marked genetic and biologic distinctions from the enteroviruses.

In recent years, genetic, biologic and molecular properties have been used to revise picornavirus taxonomy, leading to a reorganization of the human enteroviruses into four alphabetically designated human enterovirus species (HEV-A, -B, -C, and -D) ( Table 25-2 ). Determining the nucleotide sequence encoding the viral VP1capsid protein now plays a major role in the approach to taxonomy and predictably identifies viruses originally classified by serologic means, leading to the term “molecular serotyping.” This approach will, no doubt, dominate future phylogenetic studies. Of interest, the application of molecular phylogenetic approaches has also revealed that recombination between circulating enteroviruses is a frequent event and is likely to increase their genetic diversity. This propensity for recombination has played a role in recent outbreaks of paralytic diseases involving vaccine-derived stains and in well-documented cases of HEV-B infections.

Similarly, molecular analysis has led to the identification of 16 distinct types of human parechoviruses. Of these, types 1 and 3 have most often been associated with human illness. Substantial recombination among the parechoviruses has also been noted, allowing rapid genetic diversification.

Morphology and Replication

The genome of picornaviruses is a single-stranded, positive-sense RNA molecule approximately 7.4 to 8 kb in length. It consists of a 5′ noncoding region, followed by a single long open-reading frame, a short 3′ noncoding region, and a polyA tail. The 5′ noncoding region folds into highly conserved structures that are thought to play a role in the initiation of the replication of the viral genome and contain an internal ribosome entry site, which is essential for the initiation of translation. Similarly, the 3′ noncoding region is well conserved within each picornavirus genus and is thought to be involved in replication of the viral genome. The viral genome is packaged into naked capsids that exhibit icosahedral symmetry with 20 triangular faces and 12 vertices.

Enterovirus replication begins with the adsorption of virions to cell surface receptors, which are typically integrins or immunoglobulin-like proteins ( Table 25-3 ). The virions penetrate the surface of the cell, uncoat, and the viral genome functions as messenger RNA for the viral polyprotein. This polypeptide contains three domains, P1 to P3, which are cleaved into three to four proteins each. The P1 region is liberated from the polyprotein by the viral 2A protein, a chymotrypsin-like protease. P1 is initially split into three proteins—VP0, VP1, and VP3—by the viral 3C protease. VP0 is then further processed into two smaller proteins, VP4 and VP2. Portions of VP1, VP2, and VP3 are exposed at the surface of the virion, whereas VP4 is entirely internal. VP1, VP2, and VP3 have no sequence homology but share the same topology. Specifically, they form an eight-stranded antiparallel β-barrel that is wedge shaped and composed of two antiparallel β-sheets. The amino acid sequences in the loops that connect the β-strands and the N- and C-terminal sequences that extend from the β-barrel domain of VP1, VP2, and VP3 give each enterovirus its distinct antigenicity.

The replication of enteroviruses occurs in the cytoplasm in membrane-associated replication complexes, and is completed rapidly (5-10 hours). Studies of polioviruses and coxsackieviruses have shown that enteroviral replication is associated with disruption of cellular protein secretion, and host cell protein synthesis is suppressed because of cleavage of eIF4G by the enteroviral 2A protein. The coxsackievirus 2A protein also cleaves dystrophin, a cytoskeletal protein; this activity has been hypothesized to play a role in damage to the myocardium.

The parechoviruses replicate in a similar fashion. Integrins (α _v β ₃ and perhaps α _v β ₁ ) have been identified as key receptors. In contrast to the enteroviruses, the parechovirus 2A protein does not function as a protease but may function in genome replication. Moreover, parechoviral capsids are composed of three proteins: VP1, VP3, and an uncleaved VP0 protein. In addition, parechovirus replication occurs in small, discrete foci in the cytoplasm, rather than in large accumulations of membranous vesicles like the enteroviruses. Moreover, transcription and translation do not appear to be disrupted by the parechoviruses, perhaps explaining their relatively mild and delayed cytopathic effect when grown in tissue culture.

The replication of Saffold viruses has received little specific study thus far, but they are likely to exhibit features of other members of the Cardiovirus genus.

Replication Characteristics and Host Systems

Enteroviruses and parechoviruses are relatively stable viruses in that they retain infectivity for several days at room temperature and can be stored indefinitely at ordinary freezer temperatures (−20° C). They are inactivated quickly by heat (>56° C), formaldehyde, chlorination, and ultraviolet light but are refractory to ether, ethanol, and isopropanol.

Enterovirus strains grow rapidly when adapted to susceptible host systems and cause cytopathology in tissue culture in 2 to 7 days. The typical tissue culture cytopathic effect is shown in Figure 25-1 ; characteristic pathologic findings in mice are shown in Figures 25-2 and 25-3 . Final titers of virus recovered in the laboratory vary markedly among different viral strains and the host system used; typically, concentrations of 10 ³ to 10 ⁷ infectious doses per 0.1 mL of tissue culture fluid or tissue homogenate are obtained. Unadapted viral strains frequently require long periods of incubation. In both tissue culture and suckling mice, evidence of growth usually is visible. Blind passage occasionally is necessary for the cytopathology to become apparent.

Although many different primary and secondary tissue culture systems support the growth of various enteroviruses, primary rhesus monkey kidney cultures generally are accepted to have the most inclusive spectrum. Other simian kidney tissue cultures, however, also have the same broad spectrum. Tissue cultures of human origin have a more limited spectrum, but several echovirus types have shown more consistent primary isolation in human than in monkey kidney. A satisfactory system for the primary recovery of enteroviruses from clinical specimens would include primary rhesus, cynomolgus, or African green monkey kidney tissue cultures; a diploid, human embryonic lung fibroblast cell strain; rhabdomyosarcoma cell line tissue cultures; and intraperitoneal and intracerebral inoculation of suckling mice younger than 24 hours.

Human parechovirus types 1 to 6 can be isolated in several commonly used tissue culture systems. Human parechovirus types 1, 4, and 6 can be recovered in HT29 (primary colorectal adenocarcinoma) cells. Types 1 and 4 can be isolated in Vero (African green monkey kidney), A549 (human lung adenocarcinoma), and RD (rhabdomyosarcoma) cells. Human parechovirus 3 grows in Vero and A549 cells.

Antigenic Characteristics

Although some minor cross-reactions exist between several coxsackievirus and echovirus types, common group antigens of diagnostic importance have not been defined well. Heat treatment of virions and the use of synthetic peptides have produced antigens with broad enteroviral reactivity. These antigens have been used in enzyme-linked immunosorbent assay (ELISA) and complement-fixation tests to determine IgG and IgM enteroviral antibodies and for antigen detection. In one study, Terletskaia-Ladwig and colleagues reported the identification of patients infected with enteroviruses by the use of an immunoglobulin M (IgM) enzyme immunoassay (EIA). This test used heat-treated coxsackievirus B5 and echovirus 9 as antigens, and it identified patients infected with echoviruses 4, 11, and 30. The sensitivity of the test was 35%. In another study involving heat-treated virus or synthetic peptides, the respective sensitivities were 67% and 62%. However, both tests lacked specificity. Intratypic strain differences are common findings, and some strains (prime strains) are neutralized poorly by antisera to prototype viruses. However, in animals, these prime strains induce antibodies that neutralize the specific prototype viruses.

The identification of polioviral, coxsackieviral, and echoviral types by neutralization in suckling mice or tissue culture with antiserum pools is relatively well defined. Neutralization is induced by the epitopes on structural proteins VP1, VP2, and VP3; in particular, several epitopes are clustered on VP1. Prime strains do cause diagnostic difficulty because frequently they are not neutralized by the reference antisera, which is a particular problem with echoviruses 4, 9, and 11 and enterovirus 71. If these types are suspected, in some instances, this problem can be overcome by using antisera in less diluted concentrations or antisera prepared against several different strains of problem viruses. Recently, Kubo and associates have been able to type enteroviral isolates not identified through neutralization by nucleotide sequence analysis of the VP4 gene. They specifically identified prime strains of echovirus 18 and enterovirus 71. Sequence analysis of the VP1 gene also is useful for typing enteroviral prime strains not identified by neutralization.

Host Range

It was long believed that humans were the only natural hosts of enteroviruses. Enteroviruses have also been recovered in nature from sewage, flies, swine, dogs, a calf, a budgerigar (i.e., small Australian parakeet), a fox, mussels, clams, and oysters. Serologic evidence of infection with enteroviruses similar to human strains has been found in chimpanzees and other nonhuman primates, cattle, rabbits, a fox, a chipmunk, and a marmot. It is possible that infection in some of these reports was the result of their direct contact with infected humans or infected human excreta. However, genetically distinct enteroviruses have been identified in cattle, possums, domesticated pigs, sheep, and nonhuman primates.

Although enteroviruses do not multiply in flies, they appear to be a possible significant vector in situations of poor sanitation and heavy human infection. The contamination of shellfish is also intriguing because, in addition to their possible role in human infection, they offer a source of enteroviral storage during cold weather. Contaminated foods are another possible source of human infection.

General Considerations

Enteroviruses and parechoviruses are spread from person to person by fecal-oral and possibly by oral-oral (respiratory) routes. Swimming and wading pools may serve as a means of spread of enteroviruses during the summer. Children are the main susceptible cohort; they are immunologically susceptible, and their unhygienic habits facilitate spread. Spread is from child to child (by feces to skin to mouth) and then within family groups. Recovery of enteroviruses is inversely related to age; the prevalence of antibodies to common enteroviruses and parechoviruses increases with age. The incidence of infections and the prevalence of antibodies do not differ between boys and girls. Oral-oral transmission by way of the contaminated hands of health care personnel and transmission by fomites have been documented on a long-term care pediatric ward. Echovirus 18 was isolated from human breast milk, and it was possible that enterovirus transmission to the baby occurred through the breast milk. Chang and colleagues detected coxsackievirus B3 in breast milk of two symptomatic mothers, and their babies both suffered severe illnesses with hepatic necrosis and meningitis caused by coxsackievirus B3.

Geographic Distribution and Seasonality

Enteroviruses and parechoviruses have a worldwide distribution. ^∗

∗ References .

Neutralizing antibodies for specific viral types have been found in serologic surveys throughout the world, and most strains have been recovered in worldwide isolation studies. In any one area, there are frequent fluctuations in predominant types. Epidemics probably depend on new susceptible persons in the population rather than on reinfections; they may be localized and sporadic and may vary in cause from place to place in the same year. Pandemic waves of infection also occur.

In temperate climates, enteroviral infections occur primarily in the summer and fall, but in the tropics, they are prevalent all year. A basic concept in understanding their epidemiology is the far greater frequency of unrecognized infection than that of clinical disease. This is illustrated by poliomyelitis, which remained an epidemiologic mystery until it was appreciated that unrecognized infections were the main source of contagion. Serologic surveys were instrumental in elucidating the problem. In populations living in conditions of poor sanitation and hygiene, epidemics do not occur, but wide dissemination of polioviruses has been confirmed by demonstrating the presence of specific antibodies to all three types in nearly 100% of children by the age of 5 years.

Epidemics of poliomyelitis first began to appear in Europe and the United States during the latter part of the 19th century; they continued with increasing frequency in the economically advanced countries until the introduction of effective vaccines in the 1950s and 1960s. The evolution from endemic to epidemic follows a characteristic pattern, beginning with collections of a few cases, then endemic rates that are higher than usual, followed by severe epidemics with high attack rates.

The age group attacked in endemic areas and in early epidemics is the youngest one; more than 90% of paralytic cases begin in children younger than 5 years. After a pattern of epidemicity begins, it is irreversible unless preventive vaccination is carried out. Because epidemics recur over a period of years, there is a shift in age incidence such that relatively fewer cases are in the youngest children; the peak often occurs in the 5- to 14-year-old group, and an increasing proportion is in young adults. These changes are correlated with socioeconomic factors and improved standards of hygiene; when children are protected from immunizing infections in the first few years of life, the pool of susceptible persons builds up, and introduction of a virulent strain often is followed by an epidemic. Extensive use of vaccines in the past 5 decades has resulted in elimination of paralytic poliomyelitis from large geographic areas, but the disease remains endemic in various parts of the world. Although seasonal periodicity is distinct in temperate climates, some viral activity does take place during the winter. Infection and acquisition of postinfection immunity occur with greater intensity and at earlier ages among crowded, economically deprived populations with less efficient sanitation facilities.

Molecular techniques have allowed the study of genotypes of specific viral types in populations over time. For example, Mulders and colleagues studied the molecular epidemiology of wild poliovirus type 1 in Europe, the Middle East, and the Indian subcontinent. They found four major genotypes circulating. Two genotypes were found predominantly in Eastern Europe, a third genotype was circulating mainly in Egypt, and the fourth genotype was widely dispersed. All four genotypes were found in Pakistan.

The epidemiologic behavior of nonpolio enteroviruses and parechoviruses parallels that of polioviruses; unrecognized infections far outnumber those with distinctive symptoms. The agents are disseminated widely throughout the world, and outbreaks related to one or another type of virus occur regularly. These outbreaks tend to be localized, with different agents being prevalent in different years. In the late 1950s, however, echovirus 9 had a far wider circulation, sweeping through a large part of the world and infecting children and young adults. This behavior has been repeated occasionally with other enteroviruses; after a long absence, a particular agent returns and circulates among the susceptible persons of different ages who have been born since the previous epidemic occurred. Other agents remain endemic in a given area, surfacing as sporadic cases and occasionally as small outbreaks. Multiple types are frequently active at the same time, although one agent commonly predominates in a given locality.

There are no available data on the incidence of symptomatic congenital and neonatal enteroviral infections. From the frequency of reports in the literature, it appears that severe neonatal disease caused by enteroviruses decreased slightly during the late 1960s and early 1970s and then became more common again. In 2007, there was an increase in the detection of severe neonatal disease caused by coxsackievirus B1 infection, which has subsequently persisted as the most common enterovirus type reported to the National Enterovirus Surveillance System.

Although more than 100 nonpolio enterovirus types and 16 parechovirus types have been identified, only 24 different virus types have been noted in the 48 years from 1961 to 2005. The five most prevalent nonpolio enterovirus isolations per year in the United States are shown in Table 25-4 . Most patients from whom viruses were isolated had neurologic illnesses. It is possible that other enteroviruses were also prevalent but did not produce clinical disease severe enough to cause physicians to submit specimens for study. Many coxsackievirus A infections, even in epidemics, have probably gone undiagnosed because suckling mouse inoculation was not performed. The use of molecular detection methods may better reveal these in the future.

An analysis of the Centers for Disease Control and Prevention (CDC) nonpolio enterovirus data for 14 years found that early isolates in a particular year were predictive of isolates for the remainder of that year. The six most common isolates during March, April, and May were predictive of 59% of the total isolates during July through December of the same year. Khetsuriani and associates at the CDC presented an extensive report on enterovirus surveillance in the United States for the period 1970 to 2005. During this period, the five most common enterovirus isolates, in order, have been echovirus 9, echovirus 11, echovirus 30, coxsackievirus B5, and echovirus 6. During the most recent period (2006-2008), the most common isolates, in order, have been coxsackievirus B1, echovirus 6, echovirus 9, echovirus 18, and coxsackievirus A9. Similar data are available for the most common enteroviral isolates in Spain from 1988 to 1997 and in Belgium from 1980 to 1994. The most common enterovirus isolated in both countries was echovirus 30. In 1997 and 1998, major epidemic disease caused by enterovirus 71 occurred in Taiwan, Malaysia, Australia, and Japan Similar outbreaks have since continued in other countries such as Australia and in Southeast Asia.

Live-attenuated trivalent polioviral vaccine was used until 2000 in the United States and has eliminated epidemic poliomyelitis in the Western Hemisphere. It is unclear if circulation of the polio vaccine strains had any effects on enteroviral ecology. In 1970, polioviruses accounted for only 6% of the total enteroviral isolations from patients with neurologic illnesses. Although the figures are not directly comparable, more than one third of the enteroviral isolations in 1962 from similar patients were polioviruses. However, Horstmann and associates studied specimens from sewage and asymptomatic children during the vaccine era and found that the number of yearly polioviral isolations (presumably vaccine strains) was greater than the number of nonpolio enteroviruses. However, the prevalence of oral polio vaccine viruses did not seem to affect the seasonal epidemiology of other enteroviruses.

Transplacental Transmission

Ascending Infection and Contact Infection During Birth

Definitive evidence is lacking for ascending infection or contact infection with enteroviruses during birth. In prospective studies of genital herpes simplex and cytomegaloviral infections, there have been no enteroviral isolations. These results suggest that ascending infections with enteroviruses, if they occur at all, are rare. However, Reyes and associates recovered coxsackievirus B5 from the cervix of four third-trimester pregnant women. Three of the four positive cultures were obtained 3 weeks or more before delivery. In the fourth case, the cervical culture was obtained the day before delivery, and the child was delivered by cesarean section. All of the infants were healthy, but unfortunately, culture for virus was possible only from the infant delivered by cesarean section; the result was negative. In an earlier study, Reyes and colleagues reported a child who died of a disseminated echovirus 11 infection. The illness had its onset on the third day of life, and the virus was recovered from the mother’s cervix at that time.

Enteroviral infection during the birth process seems probable. The fecal carriage rate of enteroviruses in asymptomatic adult patients varies between 0% and 6% or higher in different population groups. Cherry and associates found that in 2 (4%) of 55 mothers, enteroviruses were present in the feces shortly after delivery. Katz, in a discussion of a child with neonatal coxsackievirus B4 infection, suggested that the infant might have inhaled maternally excreted organisms during birth. The fact that this child had pneumonia tends to support the contention. Infections occurring 2 to 7 days after birth could have been acquired during passage through the birth canal.

Neonatal Infection

Neonatal infections and illnesses from enteroviruses are relatively common. Transmission of enteroviruses to newborns is similar to that for populations of older people. The main factor in the spread of virus is human-to-human contact. During the summer and fall of 1981 in Rochester, New York, 666 neonates were cultured for enteroviruses within 24 hours of birth and then weekly for 1 month. The incidence of acquisition of nonpolio enteroviral infections during this period was 12.8%. Two risk factors were identified: lower socioeconomic status and lack of breastfeeding.

Events During Pathogenesis

Congenital infections with enteroviruses result from transplacental passage of virus to fetus. The method of transport from mother to fetus is poorly understood. Maternal viremia during enteroviral infections is common, and because virus has been recovered from the placenta on several occasions, it is probable that active infection of the placenta also occurs. Benirschke found no histologic evidence of placental disease in three cases of established transplacentally acquired coxsackievirus B infections. Batcup and associates found diffuse perivillous fibrin deposition with villous necrosis and inflammatory cell infiltration of the placenta in a woman who 2 weeks earlier, at 33 weeks of gestation, had coxsackievirus A9 meningitis. The woman was delivered of a macerated, stillborn infant. At birth, virus was recovered from the placenta but not from the stillborn infant.

It is assumed that infection in the fetus results from hematogenous dissemination initiated in the involved placenta. It is also possible that some in utero infection results from the ingestion of virus contained in amniotic fluid; in this situation, primary fetal infection involves the pharynx and lower alimentary tract. The portal of entry of infection during the birth process and the neonatal period is similar to that for older children and adults.

Figure 25-4 shows a schematic diagram of the events of pathogenesis. After initial acquisition of virus by the oral or respiratory route, implantation occurs in the pharynx and the lower alimentary tract. Within 1 day, the infection extends to the regional lymph nodes. On about the third day, minor viremia occurs, resulting in involvement of many secondary infection sites. In congenital infections, infection is initiated during the minor viremia phase. Multiplication of virus in secondary sites coincides with the onset of clinical symptoms. Illness can vary from minor to fatal infections. Major viremia occurs during the period of multiplication of virus in the secondary infection sites; this period usually lasts from the third to the seventh days of infection. In many echovirus and coxsackievirus infections, CNS involvement apparently occurs at the same time as other secondary organ involvement. This occasionally appears to happen with polioviral infections; however, more commonly, the CNS symptoms of poliomyelitis are delayed, suggesting that seeding occurred later in association with the major viremia.

Cessation of viremia correlates with the appearance of serum antibody. The viral concentration in secondary infection sites begins to diminish on about the seventh day. However, infection continues in the lower intestinal tract for prolonged periods.

Factors that Affect Pathogenesis

The pathogenesis and pathology of enterovirus and parechovirus infections depend on the virulence, tropism, and inoculum concentration of virus, as well as on many specific host factors. Enteroviruses have marked differences in tropism and virulence. Although some generalizations can be made in regard to tropism, there are marked differences even among strains of specific viral types. Differences in virulence of specific enteroviral types may be the result of recombination among enteroviruses or point mutations.

Enterovirus infections of the fetus and neonate are thought to be more severe than similar infections in older individuals. This is undoubtedly true for coxsackievirus B infections and probably also true for coxsackievirus A, echovirus, and poliovirus infections. Although the reasons for this increased severity are largely unknown, several aspects of neonatal immune mechanisms offer clues. The similarity of coxsackievirus B infections in suckling mice to those in human neonates has provided a useful animal model. Heineberg and coworkers compared coxsackievirus B1 infections in 24-hour-old suckling mice with similar infections in older mice. They observed that adult mice produced interferon (IFN) in all infected tissues, whereas in suckling mice, only small amounts of IFN were identified in the liver. They thought that the difference in outcome of coxsackievirus B1 infections in suckling and older mice could be explained by the inability of the cells of the immature animal to elaborate IFN. Additional studies of abnormalities of innate immunity in neonates may enhance our understanding of the severity of enterovirus infections in newborns.

Others thought that the increased susceptibility of suckling mice to severe coxsackievirus infections was related to the transplacentally acquired, increased concentrations of adrenocortical hormones. Kunin suggested that the difference in age-specific susceptibility might be explained at the cellular level. He showed that a variety of tissues of newborn mice bound coxsackievirus B3, whereas tissues of adult mice were virtually inactive in this regard. It has been suggested that the progressive loss of receptor-containing cells or of receptor sites on persisting cells with increasing age might be the mechanism that accounts for infections of lesser severity in older animals. Supporting this suggestion, Ito and colleagues showed that expression of the coxsackie and adenovirus receptor (CAR) (see Table 25-3 ) decreases as rats age. Teisner and Haahr suggested that the increased susceptibility of suckling mice to severe and fatal coxsackievirus infections might be from physiologic hypothermia and poikilothermia during the first week of life.

In the past, it was assumed that specific pathology in various organs and tissues in enteroviral infections was caused by the direct cytopathic effect and tropism of a particular virus. However, a large number of studies using murine myocarditis model systems have suggested that host immune responses contribute to the pathology. These studies suggest that T-cell–mediated processes and virus-induced autoimmunity cause acute and chronic tissue damage. Other studies suggest that the primary viral cytopathic effect is responsible for tissue damage and that various T-cell responses are a response to the damage, not the cause. A review of various murine myocarditis model systems suggests that the genetics of the hosts and of the viral strains determine the likelihood of autoimmune, cell-mediated cellular damage. ^†

† References .

However, none of the model systems is appropriate for the evaluation of the pathogenesis of neonatal myocarditis. Although available studies suggest that enterovirus-induced myocarditis in older children and adults occasionally may have a delayed cell-mediated component, the short incubation period and fulminant nature of neonatal disease, as well as the similar infection in suckling mice, suggest that autoimmune factors are not major in the pathogenesis of acute enteroviral myocarditis in neonates.

During the last 45 years, the clinical manifestations caused by several enteroviral serotypes have changed. For example, echovirus 11 infection initially was noted in association with an outbreak of upper respiratory infection in a day nursery more than 50 years ago. Then in the 1960s, it was found to be related to exanthem and aseptic meningitis. After this, and occurring presently, is the association of echovirus infection and severe sepsis-like illnesses with hepatitis in neonates. ^‡

‡ References .

Another example relates to enterovirus 71 infections. Initially, this virus was noted in association with aseptic meningitis, with only a small number of cases also having exanthem. During the last several decades, severe epidemic disease with enterovirus 71 has occurred in Taiwan, Singapore, Australia, Malaysia, Japan, and other countries in Southeast Asia. In these epidemics, hand-foot-and-mouth syndrome is a major finding, and the neurologic disease is more severe than in the past.

These phenotypic changes could be the result of point mutations or the result of recombination among enteroviruses. ^§

§ Reference .

Chan and AbuBakar have presented evidence indicating that a recombination event occurred between enterovirus 71 and coxsackievirus A16.

General Considerations

Great variations in the clinical signs of congenital and neonatal enterovirus infections are paralleled by wide variations in pathology. Because pathologic material usually is available only from patients with fatal illnesses, the discussion in this section considers only the more severe enteroviral manifestations. It is worth emphasizing, however, that these fatal infections account for only a small portion of all congenital and neonatal enterovirus infections. The pathologic findings in infants with milder infections, such as nonspecific febrile illness, have not been described.

Polioviruses

The pathologic findings in fatal neonatal poliomyelitis are similar to those seen in disease of older children and adults. The major findings have involved the CNS, specifically the anterior horns of the spinal cord and the motor nuclei of the cranial nerves. Involvement is usually irregularly distributed and asymmetric. Microscopically, the anterior horn cells show neuronal destruction; gliosis; and perivascular, small, round cell infiltration. Myocarditis has also been observed, characterized by focal necrosis of muscle fibers and various degrees of cellular infiltration.

Coxsackieviruses A

Records of neonatal illnesses associated with coxsackieviruses A are rare. Gold and coworkers, in a study of sudden unexpected death in infants, recovered coxsackievirus A4 from the brains of three children. Histologic abnormalities were not identified in the brains or spinal cords of these patients. Baker and Phillips reported the death of twins in association with coxsackievirus A3 intrauterine infections; the first twin was stillborn, and the second twin died when 2 days old of viral pneumonia.

Eisenhut and associates described a full-term neonate with coxsackievirus A9 infection with meningitis, myocarditis, and disseminated intravascular coagulation who died on the seventh day of life.

Coxsackieviruses B

Of the enteroviruses, coxsackieviruses B have been most frequently associated with severe and catastrophic neonatal disease. The most common findings in these cases have been myocarditis or meningoencephalitis, or both. Involvement of the adrenals, pancreas, liver, and lungs has occurred.

Brain and Spinal Cord

The meninges are congested, edematous, and occasionally mildly infiltrated with inflammatory cells. ^щ

? References .

Lesions in the brain and spinal cord are focal rather than diffuse but frequently involve many different areas. The lesions consist of areas of eosinophilic degeneration of cortical cells, clusters of mononuclear and glial cells ( Fig. 25-5 ), and perivascular cuffing. On occasion, areas of liquefaction necrosis unassociated with inflammation are seen.

Heart

Grossly, the heart is usually enlarged, with dilation of the chambers and flabby musculature. ^¶

¶ References .

Microscopically, the pericardium frequently contains some inflammatory cells; and thickening, edema, and focal infiltrations of inflammatory cells may be found in the endocardium. The myocardium ( Fig. 25-6 ) is congested and contains infiltrations of inflammatory cells (i.e., lymphocytes, mononuclear cells, reticulum cells, histiocytes, plasma cells, and polymorphonuclear and eosinophil leukocytes). Involvement of the myocardium is often patchy and focal but occasionally is diffuse. The muscle shows loss of striation, edema, and eosinophilic degeneration. Muscle necrosis without extensive cellular infiltration is common.

Echoviruses

In an earlier period, although frequently responsible for neonatal illnesses, echoviruses were rarely associated with fatal infections. During the past 40 years, however, there have been many reports of fatal illnesses in newborns from echovirus type 11. ^#

# References .

In virtually all cases, the major pathologic finding was massive hepatic necrosis; other findings included hemorrhagic necrosis of the adrenal glands, hemorrhage in other organs, myocardial necrosis, and acute tubular necrosis of the kidneys. Wang and colleagues studied four neonates (three with echovirus 11 and one with echovirus 5 infections) with fulminant hepatic failure and observed two histopathologic patterns associated with minimal inflammation but extensive hemorrhagic necrosis. One pattern indicated ongoing endothelial injury with endotheliitis and fibrinoid necrosis. The second pattern, which was seen in the two neonates who initially survived, was that of venoocclusive disease. Virus has not been identified in hepatocytes. Extensive myositis of the strap muscles of the neck occurred in one case. Massive hepatic necrosis has also occurred in infections with echoviruses 3, 5, 6, 7, 9, 14, 19, 20, and 21. ^∗a

∗a References .

Wreghitt and associates described a neonate with a fatal echovirus 7 infection. This infant was found to have massive disseminated intravascular coagulation, with bleeding in the adrenal glands, renal medulla, liver, and cerebellum.

At autopsy, one infant with echovirus 6 infection was found to have cloudy and thickened leptomeninges, liver necrosis, adrenal and renal hemorrhage, and mild interstitial pneumonitis. One infant with echovirus 9 infection had an enlarged and congested liver with marked central necrosis, and another with this virus had interstitial pneumonitis without liver involvement. Three infants with echovirus 11 infections had renal and adrenal hemorrhage and small-vessel thrombi in the renal medulla and in the medulla and the inner cortex of the adrenal glands. In these patients, the livers were normal. Two infants, one with echovirus 6 and the other with echovirus 31 infection, had only extensive pneumonias. Willems and colleagues described an infant with echovirus 11 infection who had pneumonia, persistent pulmonary hypertension, and purpura fulminans.

Parechoviruses

Neonatal parechovirus type 3 infections often involve the CNS, and extensive white matter injury has been repeatedly noted.

Abortion

Prematurity and Stillbirths

Congenital Malformations