I. INTRODUCTION. Vertically transmitted (mother to child) viral infections of the fetus and newborn can generally be divided into three distinct categories by transmission modes. The first is congenital infections, which can be transmitted to the fetus via the placenta in utero. The second category is peripartum infections, which are acquired intrapartum or during the delivery. The final category is postnatal infections, viruses transmitted in the postpartum period,
commonly via breast milk feeding. Classifying these infections into congenital and perinatal categories highlights aspects of their pathogenesis in the fetus and newborn infant. When these infections occur in older children or adults, they are typically benign. However, if the host is immunocompromised or if the immune system is not yet developed, such as in the neonate, clinical symptoms may be quite severe or even fatal. Congenital infections can have manifestations that can lead to spontaneous fetal loss, or become clinically apparent antenatally by ultrasonography or when the infant is born, whereas perinatal infections may not become clinically obvious until after the first few weeks of life.

Although classically the congenital infections have gone by the acronym TORCH (T = toxoplasmosis, O = other, R = rubella, C = cytomegalovirus, H = herpes simplex virus), the concept of obtaining “TORCH titers” for diagnostics in an infant is out of date with current viral diagnostic testing platforms. When congenital or perinatal infections are suspected, the diagnosis of each of the possible infectious agents should be considered separately and the appropriate most rapid diagnostic test is requested in order to implement therapy as quickly as possible. Useless information is often obtained when the diagnosis is attempted by drawing a single serum sample to be sent for measurement of “TORCH” titers. These immunoglobulin G (IgG) antibodies are acquired by passive transmission to the fetus and merely reflect the maternal serostatus. Pathogen-specific immunoglobulin M (IgM) antibodies do reflect fetal/infant infection status but with variable sensitivity and specificity. The following discussion is divided by pathogen as to the usual timing of acquisition of infection (congenital or peripartum or postnatal) and in approximate order of prevalence. A summary of the diagnostic evaluations for separate viral infections is shown in Table 48.1.

II. CYTOMEGALOVIRUS (CMV) (CONGENITAL, PERIPARTUM, AND POSTNATAL). CMV is a double-stranded enveloped DNA virus that results lifelong infection. It is a member of the herpesvirus family, is highly species-specific, and derives its name from the histopathologic appearance of infected cells, which have abundant cytoplasm and both intranuclear and cytoplasmic inclusions. The highly common rate of congenital infection following maternal infection and its resulting brain damage and birth defects has led the Institute of Medicine to name the development of a CMV vaccine as a top priority.

A. Epidemiology. CMV is present in saliva, urine, genital secretions, breast milk, and blood/blood products of infected persons and can be transmitted by exposure to any of these sources. Primary infection (acute infection) is usually asymptomatic in older infants, children, and adults but may manifest with mononucleosis-like symptoms, including a prolonged fever and a mild hepatitis. Latent infection is asymptomatic unless the host becomes immunocompromised. CMV infection is very common, with seroprevalence in the United States between 50% and 85% by age 40 years. Approximately 40% of pregnant women in the United States are infected, which is in contrast to over 90% seropositivity in underdeveloped nations. Yet, the U.S. seroprevalence rates are highly dependent on race distribution and geography, with Southeastern

states having higher seroprevalence rates compared to the rest of the country and Latin and African American women demonstrating the highest preconception seroimmunity. Primary CMV infection occurs in approximately 1% of pregnant women, likely via sexual transmission or exposure to mucosal fluid of CMV-infected toddlers who shed high amounts of the virus. Primary maternal infection is a high-risk setting for the infant, with a fetal transmission rate of 30% to 40%. This high fetal transmission rate in primary infection is in contrast to the 1% to 2% transmission rate in women infected with CMV prior to pregnancy. In this setting, virus is transmitted following maternal virus reactivation or reinfection. Approximately half of congenital infections are due to primary maternal CMV infection during pregnancy. Although transmission in the setting of nonprimary maternal infection can result in hearing loss, congenital CMV infection in the setting of no preexisting immunity disproportionately contributes to the symptomatic infections. The risk of transmission to the fetus as a function of gestational age is uncertain, but infection during early gestation likely carries a higher risk of severe fetal disease.

Congenital CMV occurs in approximately 1% of all live births in the United States and is the leading infectious cause of sensorineural hearing loss (SNHL), developmental delay, and occasional childhood death. In fact, CMV contributes to more cases of childhood deafness than that of Haemophilus influenzae bacterial meningitis in the prevaccine era. Annually, 30,000 to 40,000 CMV-infected infants are born in the United States (at least 1 in 150 live births), with 10% presenting with symptomatic disease at birth. Additionally, 10% to 15% of the asymptomatic neonates will develop significant sequelae in the first year of life, most commonly hearing loss. Therefore, over 5,000 infants are severely affected or die from CMV infection in the United States each year (1 in 750 live births). Congenital CMV infection is more common among HIV-exposed infants, and coinfected infants may have more rapid progression of HIV-1 disease. Therefore, screening for congenital CMV infection in HIV-exposed infants is advised.

Finally, only in unique settings can perinatal or postnatal transmission of CMV lead to neonatal disease, including postnatal infection of very low birth weight preterm infants and children with congenital immunodeficiencies, such as severe combined immunodeficiency (SCID). In the neonatal intensive care unit, many postnatal CMV infections were previously caused by transfusion of CMV seropositive blood products, which has mainly been eliminated through the use of seronegative and leukoreduced blood products. Currently, postnatal transmission via breast milk feeding is the most common mode of infection in preterm infants, which can lead to a sepsis-like illness, pneumonitis, and enteritis. The impact of this infection on long-term outcome and neurodevelopment is an area of ongoing investigation.

B. Clinical disease in congenital infection may present at birth or may manifest with symptoms later in infancy. Only very low birth weight preterm infants (<1,500 g) or immunosuppressed infants will have symptomatic disease from peripartum or postnatal CMV acquisition.

1. Congenital symptomatic CMV disease can present as an acute fulminant infection involving multiple organ systems with as high as 30% mortality. Signs include petechiae or purpura (79%), hepatosplenomegaly (HSM) (74%), jaundice (63%), pneumonitis, and/or “blueberry muffin spots” reflecting extramedullary hematopoiesis. Laboratory abnormalities include elevated hepatic transaminases and bilirubin levels (as much as half conjugated), anemia, and thrombocytopenia. Hyperbilirubinemia may be present at birth or develop over time and can persists beyond the period of physiologic jaundice. Approximately one-third of these infants are preterm, and one-third has intrauterine growth restriction (IUGR) and microcephaly.

A second early presentation includes infants who are symptomatic, most commonly with SNHL but without life-threatening complications. These babies may also have IUGR or disproportionate microcephaly (48%) with or without intracranial calcifications. These calcifications may occur anywhere in the brain but are classically found in the periventricular area. Other findings of central nervous system (CNS) disease can include ventricular dilatation, cortical atrophy, and migrational disorders such as lissencephaly, pachygyria, and demyelination as well as chorioretinitis in approximately 10% to 15% of infants. Babies with CNS manifestations almost always have developmental abnormalities and neurologic dysfunction. These range from mild learning and language disability or mild hearing loss to IQ scores below 50, motor abnormalities, deafness, and visual problems. Because SNHL is the most common sequela of CMV infection (60% in symptomatic and 5% in asymptomatic infants at birth), any infant failing the newborn hearing screen also should be screened for CMV infection. Conversely, infants with documented congenital CMV infection should be assessed for hearing loss as neonates and throughout the first 2 years of life.

2. Asymptomatic congenital infection at birth in 5% to 15% of neonates can manifest as late disease in infancy, throughout the first 2 years of life. Abnormalities include developmental abnormalities, hearing loss, seizures, mental retardation, motor spasticity, and acquired microcephaly.

3. Peripartum and postnatally acquired CMV infection may occur (i) from intrapartum exposure to the virus within the maternal genital tract, (ii) from postnatal exposure to infected breast milk, (iii) from exposure to infected blood or blood products, or (iv) nosocomially through urine or saliva. The time from infection to disease presentation varies from 4 to 12 weeks. Almost all term infants who are infected perinatally and postnatally remain asymptomatic, with the exception of severely immunocompromised infants. Although longterm developmental and neurologic abnormalities are rarely seen, an acute infection syndrome including neutropenia, anemia, thrombocytopenia, and hepatosplenomegaly can occur in preterm infants. Data suggest that all infants regardless of gestational age should have hearing testing over the first 2 years of life if documented to have acquired CMV.

4. CMV pneumonitis. CMV has been associated with pneumonitis occurring primarily in preterm infants <4 months old. Symptoms and radiographic findings in CMV pneumonitis are similar to those seen in afebrile pneumonia of other causes in neonates and young infants, including Chlamydia trachomatis, Ureaplasma urealyticum, and respiratory syncytial virus (RSV). Symptoms include tachypnea, cough, coryza, and nasal congestion. Intercostal retractions and hypoxemia may be present, and apnea may occur. Radiographically, there is hyperinflation, diffusely increased pulmonary markings, thickened bronchial walls, and focal atelectasis. A small number of infants may have symptoms that are severe enough to require mechanical ventilation or require an increased level of respiratory support. Long-term sequela include recurrent pulmonary problems, including wheezing, bronchopulmonary dysplasia (defined as prolonged oxygen dependence), and, in some cases, repeated hospitalizations for respiratory distress. Whether this presentation reflects congenital or perinatal CMV infection is unclear. Conversely, merely finding CMV in respiratory secretions of a preterm infant does not prove causality because CMV is present in saliva of infected infants.

5. Transfusion-acquired CMV infection. In the past, significant morbidity and mortality could occur in newborn infants receiving CMV-infected blood or blood products. Because both the cellular and humoral maternal immune systems are helpful in preventing infection or in ameliorating clinical disease, those most severely affected were preterm, low birth weight infants born to CMV-seronegative women. Mortality was estimated to be 20% in very low birth weight infants. Symptoms typically developed 4 to 12 weeks after transfusion; lasted for 2 to 3 weeks; and consisted of respiratory distress, pallor, and hepatosplenomegaly. Hematologic abnormalities were also seen, including hemolysis, thrombocytopenia, and atypical lymphocytosis. Transfusion-acquired CMV is now rare in the United States, prevented by using blood/blood products from CMV-seronegative donors or filtered, leukoreduced products (see Chapter 42).

C. Diagnosis. CMV infection should be suspected in any infant having typical symptoms of infection or if there is a maternal history of seroconversion or a mononucleosis-like, febrile illness in pregnancy or ultrasound findings consistent with CMV infection (i.e., echogenic bowel, intracranial calcifications). The diagnosis is made if CMV is identified in amniotic fluid or urine, saliva, blood, or respiratory secretions of the infant and defined as congenital infection if found in the infant within the first 3 weeks of life and as peripartum or postnatal infection if negative in the first 3 weeks and positive after 4 weeks of life. Depending on when the fetus or infant infection occurred, blood is the earliest specimen to become positive and is highly specific for congenital disease when CMV is detected in blood of a newborn; however, not all congenitally infected infants are viremic at birth. Thus, detection of CMV shedding in urine or saliva provides the highest sensitivity for diagnosis. A negative viral test from blood cannot rule out CMV infection, but a negative urine or saliva test in an untreated infant symptomatic for
4 weeks or more does rule out infection. There are three rapid diagnostic techniques:

1. CMV polymerase chain reaction (PCR). CMV may be detected by PCR in urine, saliva, or blood. The sensitivity and specificity of using this test for diagnosis is quite high for urine and saliva, but a negative PCR in blood does not rule out infection. Saliva is a preferred specimen in infants due to ease of collection. In fact, a CMV PCR testing platform based on dried saliva “spots” on filter paper has been validated as highly sensitive and specific and may be amenable to being added to the current newborn screening tests that utilize dried blood spots.

2. Spin-enhanced or “shell vial” culture. Virus can be isolated from saliva and in high titer from urine. Depending on local laboratory specifications, the specimen is collected as fluid or with a Dacron swab, inoculated into viral transport medium, then inoculated into viral tissue culture medium containing a coverslip on which tissue culture cells (MRC5) have been grown and incubated. Viable CMV infects the cells, which are then lysed and stained with antibody to CMV antigens. Virus can be detected with high sensitivity and specificity within 24 to 72 hours of inoculation. It is much more rapid than standard tissue culture, which may take from 2 to 6 weeks for replication and identification. A negative result generally rules out CMV infection except in infants who may have acquired infection within the prior 2 to 3 weeks.

3. CMV antigen. Peripheral blood can be centrifuged and the buffy coat spread on a slide. The neutrophils are then lysed and stained with an antibody to CMV pp65 antigen. Positive results confirm CMV infection and viremia; however, negative results do not rule out CMV infection. This test only used to follow efficacy of therapy and may be replaced by quantitative blood PCR tests.

4. CMV IgG and IgM. The determination of serum antibody titers to CMV has limited usefulness for the neonate, although negative IgG titers in both maternal and infant sera are sufficient to exclude congenital CMV infection. A positive IgM during pregnancy without the detection of CMV-specific IgG should be repeated to look for a new seroconversion, whereas a positive IgM in the presence of IgG should be further assessed with a CMV IgG avidity assay. Low maternal CMV IgG avidity would indicate recent infection and therefore the infant should be tested for CMV and followed closely after birth. The interpretation of a positive IgG titer in the newborn is complicated by the presence of transplacentally derived maternal IgG. Uninfected infants usually show a decline in IgG within 1 month and have no detectable titer by 4 to 12 months, whereas infected infants will continue to produce IgG. Tests for CMV-specific IgM have limited specificity but may help in the diagnosis of an infant infection.

If the diagnosis of congenital CMV infection is made, the newborn should have a thorough physical and neurologic examination, a head ultrasound of the brain, potentially followed by magnetic resonance imaging (MRI) scan of the brain, an ophthalmologic examination, and repeated hearing tests. Laboratory evaluation should
include a complete blood count, liver function tests, and, preferably, cerebrospinal fluid (CSF) examination. In CMV-infected infants with symptomatic disease, approximately 90% with abnormal brain imaging will have CNS sequelae. However, about 30% of infants with normal brain imaging will also have sequelae. Infants with evidence of neurologic involvement should be considered as candidates for antiviral treatment.

D. Treatment. Ganciclovir and the oral prodrug, valganciclovir, have been effective in the treatment of and prophylaxis against dissemination of CMV in immunocompromised patients and infants. The earliest studies of infants with symptomatic CMV disease showed a strong trend toward efficacy in the IV ganciclovir-treated infants as assessed by stabilization or improvement of SNHL. Further studies indicated that extended treatment of symptomatic infants with valganciclovir for 6 months showed improvements in hearing loss and developmental delay over 6 weeks of treatment. The primary reported toxicity of valganciclovir treatment is mild neutropenia. Yet, the occurrence of neutropenia was equally common between 6 weeks and 6 months of age in infants who received the prolonged or short course of valganciclovir—indicating that the viral infection, as opposed to the drug treatment, is the primary contributor to the observed neutropenia. Families should be advised that although evidence is increasing as to ganciclovir’s ability to improve long-term neurologic outcomes, there is a potential for future reproductive system effects because testicular atrophy and gonadal tumors were found in some animals treated with pharmacologic doses of ganciclovir. Moreover, the efficacy of initiating treatment at >1 month of age in symptomatic infection is not known, demonstrating the importance of early diagnosis. Finally, although the treatment of postnatally acquired CMV infection is recommended in highly immunosuppressed infants, the effectiveness of treatment of symptomatic postnatal CMV infection in preterm infants to ameliorate the disease course or improve long-term outcome is unknown. Thus, treatment should be recommended and supervised by a pediatric infectious disease specialist.

1. Screening. Because only about 1% of women acquire primary CMV infection during pregnancy and there are no currently available prevention strategies in pregnant women that have been shown to be effective in randomized trials, screening for women at risk for seroconversion is generally not recommended. Isolation of virus from the cervix or urine of pregnant women cannot be used to predict fetal infection. In cases of documented primary maternal infection or seroconversion, quantitative PCR testing of amniotic fluid can determine whether the fetus acquired infection. However, counseling about a positive finding of fetal infection is difficult because approximately 80% of infected fetuses will only have mild or asymptomatic disease. Some investigators have found that higher CMV viral loads from the amniotic fluid tended to correlate with abnormal neurodevelopmental outcome. One case-control study suggested a protective benefit against severe neonatal disease by administering hyperimmune CMV immunoglobulin
antenatally to women with low affinity antibody to CMV, yet a subsequent randomized controlled trial did not demonstrate benefit in preventing congenital infection. Yet, pregnant women, and particularly those that are exposed to toddlers, can be counseled to reduce their CMV acquisition risk. The Centers for Disease Control and Prevention (CDC) recommends that (i) pregnant women practice hand washing with soap and water after contact with diapers or oral secretions; do not share food, utensils, toothbrushes, and pacifiers with children; and avoid saliva when kissing a child; (ii) pregnant women who develop a mononucleosis-like illness during pregnancy should be evaluated for CMV infection and counseled about risks to the unborn child; (iii) antibody testing can confirm prior CMV infection; (iv) the benefits of breastfeeding outweigh the minimal risk of acquiring CMV; (v) there is no need to screen for CMV or exclude CMV-excreting children from schools or institutions.

2. Immunization. Passive immunization with hyperimmune anti-CMV immunoglobulin and active immunization with a live-attenuated CMV vaccine represent attractive therapies for prophylaxis against congenital CMV infections. However, data from clinical trials have not shown adequate efficacy of either of these approaches with current passive and active vaccine products. Two live-attenuated CMV vaccines have been developed, but their efficacy has not been clearly established. A subunit vaccine consisting of the major immunodominant glycoprotein present on the surface of the virus, glycoprotein B (gB), was studied for the prevention of maternal CMV acquisition following delivery, yet was only 50% efficacious in preventing CMV acquisition. Ongoing vaccine development has focused on distinct glycoprotein complexes and the elicitation of both humoral and cellular immunity, holding promising eventual development of a maternal CMV vaccine that will eliminate congenital CMV transmission, much like that of the rubella virus vaccine.

3. Breast milk feeding. Although breast milk is a common source for postnatal CMV infection in the newborn, symptomatic infection is rare in term infants. In this setting, protection against disseminated disease may be provided by transplacentally derived maternal IgG or antibody in breast milk. However, there may be insufficient transplacental IgG to provide adequate protection in preterm infants. For mothers of extremely premature and low birth weight infants known to be CMV seropositive, freezing breast milk will reduce the titer of CMV but will not eliminate active virus. At present, there is no recommended method of minimizing the risk of exposure to CMV in breast milk for preterm infants; maternal breast milk is the preferred enteral nutrition in preterm infants. Methods to reduce acquisition of CMV via breast milk feeding for preterm infants is needed to eliminate this risk for preterm infants.

4. Environmental restrictions. Day care centers and hospitals are potential high-risk environments for acquiring CMV infection. Not surprisingly, a number of studies confirmed an increased risk of infection in day care workers. However, there does not appear to be an increased
risk of infection in hospital personnel, indicating that hand hygiene and infection control measures practiced in hospital settings are sufficient to control the spread of CMV to workers. Unfortunately, such control may be difficult to achieve in day care centers. Good hand-washing technique should be suggested to pregnant women with children in day care settings and with children attending day care, especially if the women are known to be seronegative. The determination of CMV susceptibility of these women by serology may be useful for counseling.

5. Transfusion product restrictions. The risk of transfusion-acquired CMV infection in the neonate has been almost eliminated by the use of CMV antibody-negative donors, by freezing packed red blood cells (PRBCs) in glycerol, or by removing the white blood cells. It is particularly important to use blood from one of these sources in preterm, low birth weight infants (see Chapter 42).

III. HERPES SIMPLEX VIRUS (HSV: PERINATAL). HSV, a lifelong infection, is a double-stranded, enveloped DNA virus with two virologically distinct types: types 1 and 2. HSV-2 was previously the primary cause of genital lesions, yet HSV-1 has become the predominant virus type in genital lesions of young women. Both types produce clinically indistinguishable neonatal syndromes. The virus can cause localized disease of the infant’s skin, eye, or mouth (SEM) or may disseminate by cell-to-cell contiguous spread or viremia. After adsorption and penetration into host cells, viral replication proceeds, resulting in cellular swelling, hemorrhagic necrosis, formation of intranuclear inclusions, cytolysis, and cell death.

A. Epidemiology. Acquisition of HSV results in lifelong disease, with periodic virus reactivation and mucosal shedding. At least 80% of the U.S. population is infected with HSV type 1 by the fifth decade of life, the cause of recurrent orolabial disease and an increasing cause of genital disease. According to the 2005 to 2008 National Health and Nutrition Examination Survey, the overall seroprevalence of HSV-1 and -2 in the United States in 19- to 49-year-olds is 54% and 16%, respectively. Women without prior exposure to HSV have a 4% chance of primary infection during pregnancy and a 2% chance of a nonprimary acute infection with either HSV-1 or -2 (previously infected with the alternate HSV type). The majority of these new HSV acquisitions will be asymptomatic.

Infection in the newborn occurs as a result of direct exposure to the virus, most commonly in the perinatal period from maternal genital disease or asymptomatic virus shedding. In one study, the characteristic ulcerations of the genitalia were present only in two-thirds of the genital tracts from which HSV could be isolated. It is estimated that up to 0.4% of all women presenting for delivery are shedding virus, and more than 1% of all women with a history of recurrent HSV infection asymptomatically shed HSV at delivery. Yet, it is critical to recognize that most mothers of infants with neonatal HSV do not have a history of HSV outbreaks. Approximately 30% to 50% of infants will acquire HSV infection if maternal primary infection occurs near delivery, whereas <1% of infants are infected if born to a woman with preexisting immunity (recurrent disease). Additionally,
one-third of infants born to mothers with newly acquired HSV-2 or -1, although already infected with the other HSV type (nonprimary, first episode defined by detection of virus in the maternal genital tract at the time of delivery but no IgG response for the type-specific HSV identified), may acquire HSV infection. This may be due to protective maternal type-specific antibodies in the infant’s serum or the birth canal. The overall incidence of newborn infection with HSV is estimated to be 1 in 3,000 to 1 in 20,000 (or 200 to 1,333 infants per year) in the United States.

1. Intrapartum transmission is the most common cause of neonatal HSV infection. It is primarily associated with active shedding of virus from the cervix or vulva at the time of delivery. Up to 90% of newborn infections occur as a result of intrapartum transmission. Maternal immunity and the related amount and duration of maternal virus shedding are major determinates of peripartum transmission. Transmission risks are greatest with primary maternal infection during pregnancy, with nonprimary acute infection with HSV-1 or -2 being the next highest risk setting. In fact, when maternal antibody is present, the risk of acquisition of HSV, even for the newborn exposed to HSV in the birth canal, is much lower than that of primary maternal infection. The exact mechanism of action of maternal antibody in preventing perinatal infection is not known, but transplacentally acquired antibody is associated with reduced risk of severe newborn disease following perinatal HSV exposure. The risk of intrapartum infection increases with ruptured membranes, especially when ruptured longer than 4 hours. Finally, direct methods for fetal monitoring, such as with scalp electrodes, increase the risk of fetal transmission in the setting of active shedding. It is best to avoid these techniques if possible in women with a history of recurrent infection or suspected primary HSV disease.

2. Antenatal transmission. In utero infection with HSV has been documented but is uncommon. Spontaneous abortion has occurred with primary maternal infection before 20 weeks’ gestation, but the true risk to the fetus of early-trimester primary infection is not known. Fetal infections may occur by either transplacental or ascending routes and have been documented in the setting of both primary and, rarely, recurrent maternal disease. There may be a wide range of clinical manifestations, from localized skin or eye involvement to multiorgan disease and congenital malformations. Chorioretinitis, microcephaly, and hydranencephaly may be found in these small numbers of congenitally infected patients.

3. Postnatal transmission. A small percentage of neonatal HSV infections result from postnatal HSV exposure (˜10%). Potential sources include symptomatic and asymptomatic oropharyngeal shedding by either parent, hospital personnel, or other contacts, and maternal breast lesions. Measures to minimize exposure from these sources are discussed in the following text.

C. Clinical manifestations. The morbidity and mortality of neonatal HSV best correlates with three categories of disease. These are (i) infections localized to
the SEM, (ii) encephalitis with or without localized mucocutaneous disease, and (iii) disseminated infection with multiple organ involvement.

1. SEM infection. Approximately 50% of infants with HSV have disease localized to the skin, eye, or mucocutaneous membranes. Vesicles typically appear on the sixth to ninth day of neonatal life. A cluster of vesicles often develops on the presenting part of the body, where extended direct contact with virus may occur. Vesicles occur in 90% of infants with localized mucocutaneous infection, and recurrent disease is common. Significant morbidity can occur in these infants despite the absence of signs of disseminated disease at the time of diagnosis. Up to 10% of infants later show neurologic impairment, and infants with keratoconjunctivitis can develop chorioretinitis, cataracts, and retinopathy. Thus, ophthalmologic and neurologic follow-up is important in all infants with mucocutaneous HSV. Infants with three or more recurrences of vesicles, likely reflecting poor immunologic control of virus replication, have an increased risk of neurologic complications.

2. CNS infection. Approximately one-third of neonates with HSV present with encephalitis in the absence of disseminated disease, and as many as 60% of these infants do not have mucocutaneous vesicles. These infants usually become symptomatic at 10 to 14 days of life with lethargy, seizures, temperature instability, and hypotonia. In the setting of disseminated disease, HSV is thought to invade the CNS from hematogenous spread. However, CNS infection in the absence of disseminated disease can occur, most often in infants having transplacentally derived viral-neutralizing antibodies, which may protect against widespread dissemination but not influence intraneuronal viral replication. Mortality is high without treatment and is approximately 15% with treatment. Late treatment is associated with increased mortality, highlighting the need for early treatment when neonatal HSV infection is suspected. Approximately two-thirds of surviving infants have impaired neurodevelopment. Long-term sequelae from acute HSV encephalitis include microcephaly, hydranencephaly, porencephalic cysts, spasticity, blindness, deafness, chorioretinitis, and learning disabilities.

3. Disseminated infection. This is the most severe form of neonatal HSV infection. It accounts for approximately 22% of all infants with neonatal HSV infection and can result in mortality for over half. Pneumonitis and fulminant hepatitis are associated with greater mortality. Symptoms usually begin within the first week of neonatal life. The liver, adrenals, and other visceral organs are usually involved. Approximately two-thirds of infants also have encephalitis. Clinical findings include seizures, shock, respiratory distress, disseminated intravascular coagulation (DIC), and respiratory failure. A typical vesicular rash may be absent in as many as 20% of infants. Forty percent of the infants who survive have long-term morbidity.

D. Diagnosis. HSV infection should be considered in the differential diagnosis of ill neonates with a variety of clinical presentations. These include CNS abnormalities, fever, shock, DIC, and/or hepatitis. HSV also should be considered in infants with respiratory distress without an obvious bacterial cause or a clinical course and findings consistent with prematurity.
The possibility of concomitant HSV infection with other commonly encountered problems of the preterm infant should be considered. Viral isolation or PCR detection of viral DNA in the appropriate clinical setting remains critical to the diagnosis. For the infant with mucocutaneous lesions, tissue should be scraped from vesicles, placed in the appropriate viral transport medium, and promptly processed for culture and/or PCR by a diagnostic virology laboratory. Virus also can be isolated or detected from the oropharynx and nasopharynx, conjunctivae, stool, urine, and CSF. In the absence of a vesicular rash, viral isolation or detection from these sites may aid in the diagnosis of disseminated HSV or HSV encephalitis. With encephalitis, an elevated CSF protein level and pleocytosis are often seen, but initial values may be within normal limits. Therefore, serial CSF examinations may be very important. Electroencephalography and computed tomography (CT)/MRI are also useful in the diagnosis of HSV encephalitis. Viral isolation from CSF is reported to be successful in as many as 40% of cases, and rates of detection in CSF by PCR may reach close to 100%. Combined HSV-1 and -2 serology is of little value, because many women are infected with HSV-1 and because these tests usually have a relatively slow turnaround time; however, obtaining type-specific antibody (HSV-1 or -2) has an 80% to 98% sensitivity and >96% specificity for identifying previous maternal infection and, thus, will assist in assessing infant risk of acquiring HSV. Infant HSV-specific IgM detection is not useful. Laboratory abnormalities seen with disseminated disease include elevated hepatic transaminase levels, direct hyperbilirubinemia, neutropenia, thrombocytopenia, and coagulopathy. A diffuse interstitial pattern is usually observed on radiographs of infants with HSV pneumonitis.

E. Treatment. Antiviral therapy (acyclovir, a nucleoside analog that selectively inhibits HSV replication) is highly efficacious in this setting, but the timing of therapy is critical. Treatment is indicated for all forms of neonatal HSV disease. Initial antiviral studies were carried out with vidarabine, which reduced morbidity and mortality of HSV-infected neonates. Mortality with encephalitis was reduced from 50% to 15% and in disseminated disease from 90% to 70%. Later studies found that acyclovir is as efficacious as vidarabine for the treatment of neonatal HSV. Furthermore, acyclovir is a selective inhibitor of viral replication with minimal side effects on the host and can be administered in relatively small volumes over short infusion times. Recommendations include treating infants with disease limited to the SEM disease with 20 mg acyclovir per kilogram every 8 hours for 14 days, and those with CNS or disseminated disease for at least 21 days, or longer if the CSF PCR remains positive. Infants with ocular involvement should have an ophthalmologic evaluation and treatment with topical ophthalmic antiviral agents in addition to parenteral therapy. Oral therapy such as with valacyclovir is not recommended for initial treatment. Yet, oral acyclovir suppressive therapy following initial acute treatment at a dose of 300 mg/m²/dose three times a day for 6 months of life was beneficial in improving the developmental outcome for infants with neonatal HSV infection. Other reports have demonstrated good outcomes in perinatally infected infants treated with suppressive therapy with higher doses of oral acyclovir for up to 2 years of life.

1. Pregnancy strategies. Pregnant women known to be HSV-seronegative (or seronegative for HSV-1 or -2) should avoid genital sexual intercourse with a known HSV-seropositive partner in the third trimester. For women who do acquire primary HSV during pregnancy or have recurrent outbreaks, several trials have shown efficacy and safety of treating pregnant women with clinically symptomatic primary HSV infection with a 10-day course of acyclovir (oral therapy or IV if more severe disease) and its subsequent reduction in cesarean section. It is also recommended that women with HSV-2 be tested for HIV because HSV-2 seropositive persons have a twofold greater risk for acquisition of HIV than those who are seronegative for HSV-2.

2. Delivery strategies. Cesarean section is recommended for women with active genital lesions or prodromal symptoms at the time of delivery. The principal problem in developing antenatal strategies for the prevention of HSV transmission is the inability to identify maternal shedding of virus at the time of delivery. Viral identification requires isolation in tissue culture or PCR, so any attempt to identify women who may be shedding HSV at delivery would require antenatal cervical sampling and rapid turnaround with virus detection. Unfortunately, such screening cultures taken before labor fail to predict active excretion at delivery. Until more rapid HSV detection techniques are available, the only clear recommendation that can be made is to deliver infants by cesarean section if genital lesions are present at the start of labor. The efficacy of this approach may diminish when membranes are ruptured beyond 4 hours. Nevertheless, it is generally recommended that cesarean section be considered even with membrane rupture of longer durations. For women with a history of prior genital herpes, careful examination should be performed to determine whether lesions are present when labor commences. If lesions are observed, cesarean section should be offered. If no lesions are identified, vaginal delivery is appropriate, but a cervical swab should be obtained for culture and/or PCR and obtain maternal serology to determine if a new acquisition of a nonprimary infection with HSV-1 or -2 has occurred. Women with known clinical disease or serologic evidence of primary or nonprimary first-episode infection can be offered acyclovir near term until delivery, enabling a vaginal delivery if there are no visible lesions, but the impact of this strategy on prevention of neonatal disease is not established.

3. Management of the newborn at risk for HSV (Table 48.2). At this time, there are no data to support the prophylactic use of antiviral agents or immunoglobulin to prevent transmission to the newborn infant. Infants inadvertently delivered vaginally in the setting of cervical lesions should be isolated from other infants in the nursery, and swabs should be obtained from the oropharynx/nasopharynx, conjunctivae, and anus for viral detection at 12 to 24 hours of age. If the mother has no prior history of HSV, initiate acyclovir treatment while awaiting the laboratory results. If the mother can be identified as having recurrent infection, the risk of neonatal infection rate is low, and parents should be
instructed to consult their pediatrician if a rash or other clinical changes (lethargy, tachypnea, poor feeding) develop. Weekly pediatric followup during the first month is recommended. If the mother is found to have either recent primary or nonprimary, first-episode infection and a genital lesion, it is recommended by some experts to treat the infant for 10 days of acyclovir even without symptomatology or detection of virus in the infant. Infants with a positive culture or PCR from any site or the evolution of clinical symptomatology should immediately have cultures repeated and antiviral therapy started. Before starting acyclovir therapy, the infant should have conjunctival, nasopharyngeal, anal swabs for culture/PCR, plasma viral load, and a CSF evaluation for pleocytosis and HSV DNA PCR. Evidence of dissemination should be evaluated with hepatitic transaminases, blood counts and coagulation tests for hematologic and clotting disorders, and a chest radiograph if respiratory symptoms develop.