A Clinical guide to pediatric infectious disease
Infections can be transmitted from mother to developing fetus transplacentally, at the time of delivery, and through breast milk.
The diagnosis of congenital and perinatal infections is complicated by several issues:
- Infected babies may be asymptomatic or minimally symptomatic when born, only to develop symptoms and disabilities in the following weeks, months, or even years.
- Babies born symptomatic may have findings such as anemia, thrombocytopenia, jaundice, and microcephaly, which can be the result of infections with several different pathogens (e.g., cytomegalovirus [CMV], syphilis, rubella).
- Laboratory evaluation for congenital and some perinatal infections has often included the use of TORCH titers (i.e., toxoplasmosis, rubella, CMV, and herpes). TORCH titers are immunoglobulin G (IgG) antibodies that readily cross the placenta and may persist for months in the neonatal circulation. As 50% of adults have serologic evidence of previous infection with CMV, herpes simplex virus, or toxoplasmosis, up to 50% of newborns tested will have positive TORCH titers for these organisms. TORCH titers are of no value in the diagnosis of congenital and perinatal infections. Rather, a disease-specific approach using precise clinical and laboratory methods is required.
CMV is the most common congenital infection, affecting 1% to 2% of all newborns. The rate of infection in a fetus of a mother with primary infection is between
40% and 50%; the rate of infection is less than 1% if the mother has reactivated infection. Most infected infants (90%) are asymptomatic at birth; only 10% of these children have long-term sequelae. The most common sequela is asymptomatic congenital CMV infection sensorineural hearing loss. This hearing loss can be progressive and may be unilateral or bilateral.
Ten percent of infants with congenital CMV infection are symptomatic at birth. These patients often present with multiple organ system involvement, including intrauterine growth retardation, petechiae, anemia, leukopenia, and thrombocytopenia. A classic finding on computed tomography of the head in congenital CMV is periventricular calcification (Fig. 3.1). The predilection for the calcification in this area is believed to be related to the tendency of CMV to infect the rapidly dividing
germinal matrix cells. Magnetic resonance imaging of neonates with congenital CMV infection may also reveal irregular areas in the junction of the gray and white matter, accompanied by the presence of ventricular pseudocysts. Patients with congenital CMV may have continual neurologic damage even after birth, which is thought to be secondary to persistent viral replication.
FIG. 3.1. Periventricular calcifications seen in congenital cytomegalovirus infection.
The gold standard of diagnosis is isolation of the virus, usually from the urine. Urine viral cultures are often positive within days because the kidney is a principal site of viral replication. Proof of congenital infection is based on obtaining appropriate specimens within 3 weeks of birth. After this time, distinguishing between intrauterine infection and perinatal infection may be difficult.
Patients with congenital CMV infection are at risk for a variety of disabilities, including developmental delay and hearing loss. Symptomatic children are considered to have the highest risk for long-term abnormalities, although recent longitudinal studies have found predicting disability difficult. It is known that the major disability in asymptomatic congenital CMV infection is sensorineural hearing loss. This hearing loss may be progressive in affected infants. After a diagnosis of congenital CMV is made, routine assessments should be instituted. It has been recommended that a careful ophthalmology exam be performed at 12 months, 3 years, and at entrance to preschool. Audiology examinations should be done every 3 months until 3 years of age and then annually.
Diagnosis of Cytomegalovirus
· Urine viral culture
· Hearing evaluation
· Ultrasound or computed tomograph of brain
· Eye examination
No definitive protocols exist for the treatment of congenital CMV infection. Clinical trials are in progress; a recent randomized clinical trial comparing outcomes in symptomatic infants given the antiviral agent ganciclovir with those in patients receiving no treatment suggested that there may be a benefit from treatment, the greatest benefit in treated children being a reduction in hearing loss. In early protocols of treatment of CMV, symptomatic infants were administered intravenous ganciclovir for 6 weeks; later studies extended treatment of affected infants with intravenous and then oral ganciclovir for up to 1 year. Currently, it is not recommended that asymptomatic
infants found to be congenitally infected receive ganciclovir. These children should be followed carefully for the development of sensorineural hearing loss. Neonates with life-threatening symptomatic disease, including intractable thrombocytopenia, pneumonia, or hepatic failure, are candidates for antiviral therapy. It is my experience and the experience of investigators nationally that therapy can be very beneficial in these cases.
Complications of ganciclovir include difficulties in maintaining intravenous access and neutropenia. There will likely be continued efforts to identify precisely and treat those infants likely to have long-term sequelae from congenital CMV infection and those most likely to benefit from therapy.
Herpes Simplex Virus
Depending on the population surveyed, as many as 10% to 15% of women report a history of genital herpes. This number may actually be higher because many infected women do not have a history of visible genital lesions. Previously, pregnant women who reported a history of genital herpes had weekly viral cultures obtained as they approached delivery. However, when results of the viral cultures become available, typically in 7 days, they were of little value. This practice of weekly viral cultures has been replaced by a more basic approach; women who do not have active lesions at the time of delivery are often allowed to deliver vaginally.
The current practice is such that exposure to herpesvirus can never be avoided, and pediatricians must be vigilant to the various manifestations of neonatal herpes. The manifestations of neonatal herpes usually present in the first month of life.
Herpetic vesicles can appear on the scalp, mouth, eyes, and skin. Although children with mucocutaneous disease may initially be well appearing, 75% progress to disseminated disease (Fig. 3.2).
Encephalitis may present with lethargy, seizures, and a cerebrospinal pleocytosis that shows a lymphocytic predominance. In infants with meningitis or encephalitis, positive temporal lobe spikes on electroencephalogram (EEG) are found in 80% of cases and represent a noninvasive means of diagnosis. This is often misdiagnosed as aseptic meningitis or as the more common enteroviral meningitis. Aseptic meningitis in the first month of life should be considered herpes until proved otherwise.
FIG. 3.2. Scalp vesicles in a 2-week-old infant representing mucocutaneous herpes infection (see color plate).
Rarely, herpes can present with progressive pneumonia in the neonatal period. This is usually not associated with skin vesicles but can be associated with dissemination, with increased liver function tests, coagulopathy, and organ failure.
Disseminated disease is often associated with fever, coagulopathy, and abnormalities of liver function tests. Vesicles are usually not present initially. Disseminated disease has a very high mortality rate, even when treated appropriately. A persistently febrile neonate, particularly if blood cultures remain negative and liver enzymes are elevated, should raise strong suspicion of herpes (Table 3.1).
The first issue in diagnosis is a clinical suspicion of herpes simplex disease. Herpetic vesicles in the neonatal period are an emergency. If there is any suspicion regarding whether a skin lesion is a vesicle, empiric acyclovir should be started while appropriate studies are obtained. This is very similar to the management of the febrile
neonate, in whom empiric antibiotics are started until blood cultures are final. Aseptic meningitis in the neonate should also be considered herpes until proved otherwise.
TABLE 3.1. Neonatal Herpes: Clinical Manifestations
There are several ways of making the diagnosis of neonatal herpes. Vesicles can be scraped to obtain viral culture and tested for direct fluorescent antibodies; these have relatively high yield. Polymerase chain reaction (PCR) of the serum and spinal fluid is now considered the gold standard for diagnosis. Even in neonates with only mucocutaneous disease, the rate of a positive serum PCR can be greater than 50%. There are increasing reports of reduced sensitivity of cerebrospinal fluid (CSF) PCR in diagnosing pediatric herpes simplex encephalitis, particularly if CSF has been obtained on the first day of illness. In patients with a compatible clinical picture of herpes simplex encephalitis, particularly if an alternative diagnosis such as enterovirus has not been confirmed, it may be prudent to continue treatment until a second sample of CSF has been obtained and tested.
Treatment of neonatal herpes is intravenous acyclovir. Herpes simplex disease in the neonate is never treated either empirically or definitively with topical or oral acyclovir. The dose is 60 mg/kg per day in three divided doses. The treatment duration is 14 days for mucocutaneous disease and 21 days for disseminated disease or in neonates in whom central nervous system (CNS) involvement is suspected. Supportive care in terms of controlling seizures and coagulopathy is also critical in the management of these patients.
Herpes simplex virus is a latent virus, and it is assumed that most neonates with infection will develop latent infection. There is ongoing concern that reactivation of
latent virus in infants with a history of neonatal infection may be associated with continued morbidity. Children with a history of mucocutaneous disease often have recurrences of their vesicles; the effect of this on long-term development is uncertain. Studies conducted at the National Institute of Allergy and Infectious Disease have found that infants with more than three recurrences of mucocutaneous vesicles have a higher incidence of neurologic sequelae. It is postulated that cutaneous reactivation may be associated with subclinical reactivation in the CNS.
There have been initial clinical trials to enroll children with recurrent cutaneous herpes on long-term oral acyclovir therapy given at 300 mg/m2 per dose two or three times daily for 6 months. Although this does seem to prevent herpes recurrences, nearly half of the children develop neutropenia, defined as white blood cell counts of less than 1,000/m3. Currently, there is no consensus regarding preventative therapy; this should be answered in the future by prospective clinical trials.
Diagnosis of Herpes Simplex Virus
· Direct immunofluorescent antibodies (DFA), viral culture of vesicles
· Serum, CSF PCR
· EEG (80% show temporal lobe spikes)
Congenital syphilis results from inadequate or late treatment of syphilis in a pregnant woman. Infection passed to the newborn can result in stillbirth, symptomatic neonatal disease, or infection that does not manifest until later in childhood.
Infected infants may appear normal at birth, only to manifest symptoms in the first year of life. These infants may present with mucocutaneous lesions, anemia, and organomegaly. Congenital syphilis nephrotic syndrome or an osteochondritis that causes decreased movement of the affected extremity (pseudoparalysis of Parrot) may be seen. Late manifestations of congential syphilis include interstitial keratitis and notching of the central incisors (Hutchinson's teeth).
There are two types of serologic tests used for the diagnosis of syphilis in adults. The nontreponemal tests include the VDRL and RPR. These can be quantitated by titers and, after successful therapy, usually become nonreactive. The treponemal
tests, which include MHA-TP and FTA-ABS, are not quantitative and are thought to remain positive for life. All syphilis serology crosses the placenta and, like other TORCH titers, cannot be used for diagnosis in neonates. Because an infected infant may be asymptomatic at birth (only to develop the disease in later life), we therefore have an unusual situation in which both laboratory tests and physical examination are not helpful in the diagnosis of congenital syphilis.
In the early 1990s, new criteria were established for the presumptive diagnosis of congenital syphilis. An infant born to a mother considered untreated or inadequately treated receives the presumptive diagnosis of congenital syphilis. The definition of inadequate maternal treatment includes the following:
- Syphilis during pregnancy treated with a nonpenicillin regimen, such as erythromycin
- Syphilis during pregnancy treated with an appropriate penicillin regimen without the expected posttherapy decrease in nontreponemal titers
- Syphilis treated less than 1 month before delivery
- Syphilis treated before pregnancy but with insufficient serologic follow-up to assess the response to treatment
Some investigators believe that the rate of fetal infection from a mother with secondary syphilis in pregnancy is so high that treatment should also be considered for all those infants as well.
Children with a presumptive diagnosis of congenital syphilis should have the following evaluation:
- Physical exam, although it should be stressed that infected infants appear be normal at birth
- Nontreponemal and treponemal tests done from infant blood rather than cord blood
- CSF for cells, protein, and VDRL
- Complete blood count and platelet count
- Long-bone radiographs looking for osteolytic lesions
A major part of the evaluation for congenital syphilis in the infant includes the evaluation for congenital neurosyphilis. The diagnosis of congenital neurosyphilis is important because several forms of penicillin, including penicillin G benzathine, do not reach treponicidal concentrations in the CSF. The gold standard for CNS disease is the rabbit infectivity test. This test involves the inoculation of clinical specimens into rabbits and is sensitive enough to detect as few as 10 organisms. Unfortunately, this test is not routinely available outside research laboratories.
Controversy continues about the best diagnostic approach to document CNS involvement in congenital syphilis. Recently, PCR of blood and CSF, in addition to IgM immunoblotting, has been developed. Compared with the rabbit infectivity test, these studies were found to identify 94% and 100% of infants with CNS disease, respectively. It is hoped that these tests will become widely available in the future. In infants with otherwise normal clinical, laboratory, and radiographic studies, CNS involvement is unusual.
Diagnosis of Neonatal Syphilis
· Maternal history
· Long bone films
· Serum, CSF VDRL
· Complete blood count
Asymptomatic infants who have normal CSF results, complete blood counts, and radiographic examination (and therefore low likelihood of CNS involvement) may be treated with a single dose of intramuscular benzathine penicillin at a dose of 50,000 units/kg. Neonates who have abnormalities on physical exam or laboratory evaluation should be treated with aqueous penicillin G, 100,000 to 150,000 units/kg per day administered at 50,000 units/kg per dose intravenously every 12 hours for the first 7 days of life and every 8 hours thereafter for a total of 10 days. An additional option for these patients with proven or highly probable disease includes procaine penicillin, 50,000 units/kg per dose given intramuscularly once a day for 10 days.
Follow-up is critical in infants treated for congenital syphilis. Serologic nontreponemal tests should be performed 3, 6, and 12 months after conclusion of treatment. Nontreponemal titers should be nonreactive by 6 months of age if treatment was adequate. Children with increasing titers or persistent stable titers should be considered for repeat treatment.
Congenital rubella infection is unusual because the readily available vaccine prevents infection in mothers of childbearing age. However, cases still occur in populations that are not optimally vaccinated. Most cases of congenital rubella are the result of primary maternal disease during pregnancy. The risk for fetal infection is highest during early gestation, although the most severe manifestations occur when infection is in the last trimester.
Patients born with congenital rubella usually suffer from severe intrauterine growth retardation as well as a variety of cardiac, ophthalmologic, and neurologic defects. The most common congenital heart defects include patent ductus arteriosis and pulmonary artery stenosis. Cataracts occur in more than 30% of cases. Severe cases often have “celery stalk” appearance of long bones visible on plain radiographs. Organomegaly, thrombocytopenia, and purpuric skin lesions may also be seen. There is considerable clinical overlap in patients with congenital CMV. As
with many congenital infections, many children may be asymptomatic at birth but develop manifestations later in life.
Diagnosis is made by detection of rubella-specific serum IgM antibodies. Congenital infection can also be diagnosed by documenting increasing serum concentrations of infant rubella IgG over several months. Virus is readily excreted from throat, urine, and CSF, but viral isolation of rubella is difficult and usually not achieved.
Diagnosis of Congenital Rubella
· Rubella IgM
· Persistence of IgG
· Viral culture of throat, urine, CSF
At the present time, there is no treatment for infants with congenital rubella. Defects of the eyes are managed as needed. Contact isolation is recommended for children with proven or presumed congenital rubella during the first year of life.
Congenital toxoplasmosis occurs when a woman acquires primary infection during pregnancy. Overall transmission rate is between 20% and 30%. The rate of transmission is low when women are infected in the first trimester, but when infection occurs during this time, infants are more severely affected. Conversely, if infections occur later in pregnancy, the risk for transmission is high, but infected infants are less likely to have the most severe clinical features of the disease.
Studies have found that most causes of toxoplasmosis in humans include eating rare to medium cooked beef, working in an outside environment, and cat exposure. In some cases, the actual source of transmission remains unclear despite repeated investigation. Factors that predict congenital toxoplasmosis have been found to include mother's birth outside the United States, multiparity, and the level of education of the mother (the least and most educated appear to be at highest risk).
The classical clinical triad of congenital toxoplasmosis includes retinitis, hydrocephalus, and intracranial calcification, although this is present in only a small
number of cases (Fig. 3.3). Congenital toxoplasmosis is usually asymptomatic at birth. As they grow older, a significant percentage of these children have learning disabilities, visual impairment, or mental retardation.
FIG. 3.3. Hydrocephalous, intracranial calcifications seen in congenital toxoplasmosis.
Traditional TORCH titers are again of little value in the diagnosis of congenital toxoplasmosis. Maternal IgG antibodies cross the placenta and are positive in a good percentage of newborns. Commercially available IgM assays are not recommended for routine use because they have very high false-positive and false-negative rates. The reference laboratory at the Palo Alto Medical Foundation, under the direction of Dr. Jack Remington, is the standard way of making the laboratory diagnosis of congenital toxoplasmosis (Palo Alto Medical Foundation; telephone number 650–853–4828.) This laboratory performs a variety of specialized assays, including the following:
Sabin Feldman dye test. This test measures primarily IgG antibodies. Any titer is considered positive.
Immunosorbent agglutination assay (IgM-ISAgA). This test is highly sensitive and recommended for infants in whom congenital toxoplasmosis is suspected.
IgA enzyme-linked immunosorbent assay (ELISA). This test is actually more sensitive for detection of infection in newborns than the IgM assay. Both tests are done if congenital toxoplasmosis is suspected.
Evaluation of infants with suspected congenital toxoplasmosis should also include ophthalmologic examination and computed tomography or ultrasound of the brain to determine whether hydrocephalus or calcifications are present.
Diagnosis of Neonatal Toxoplasmosis
· Ophthalmology exam
· Ultrasound, computed tomography of brain
· Toxoplasmosis IgA, IgM by “double-sandwich” ELISA
Management of Confirmed Congenital Toxoplasmosis
Infants who are diagnosed with congenital toxoplasmosis require treatment. The therapy for congenital toxoplasmosis is as follows:
- Sulfadiazine, 100 mg/mL; half of child's weight (in kilograms) equals number of mls, given twice a day
- Pyrimethamine, 2 mg/mL; half of child's weight (in kilograms) equals number of mls, given once a day
- Folinic acid (leucovorin) 5-mg tablets, 2 tablets crushed in formula Monday, Wednesday, and Friday
- For all infants with both asymptomatic and symptomatic congenital infection, therapy is recommended for 1 year.
Diagnosis of Acute Toxoplasmosis in Pregnant Women
Diagnosis of toxoplasmosis in pregnancy is difficult because most patients have minimal symptoms; a mononucleosis-like illness or posterior cervical adenitis is sometimes seen. In women at increased risk for primary toxoplasmosis, screening for IgG antibodies can be performed. It is important to confirm all screenings with follow-up testing at a reference laboratory. The panel of testing available can often give an estimate of the timing of maternal infection.
If the mother has had primary infection during pregnancy, fetal infection can be evaluated using PCR of amniotic fluid. Serial ultrasound examinations looking for hydrocephalus, organomegaly, and intracranial calcifications are used to determine whether fetal infection is causing specific organ system disease. Maternal antitoxoplasmosis
treatment is frequently given in this situation because it may reduce fetal damage.
Lymphocytic Choriomeningitis Virus
Lymphocytic choriomeningitis virus (LCMV) is a single-stranded RNA virus that causes chronic infection in house mice and pet hamsters, which then shed the virus in the urine and feces. Humans may be infected by direct contact with excreta or aerosolization of viral particles.
It has become increasingly appreciated that LCMV may cause a congenital infection characterized by periventricular calcification, hydrocephalus, and chorioretinitis. This may give a symptom complex similar to other congenital infections, particularly toxoplasmosis or CMV.
The true incidence of LCMV congenital infection is not known because it is likely that only the most extreme cases are diagnosed. The diagnosis should be considered in the neonate with intracranial calcifications or hydrocephalous, especially if the workup for toxoplasmosis and CMV are negative. Diagnosis is by immunofluorescent antibody or ELISA.
At the present time, there is no therapy for LCMV. Because of this newly described syndrome, expectant mothers should be cautioned regarding contact with mice and other rodents.
Human Immunodeficiency Virus
A pregnant woman with untreated human immunodeficiency virus (HIV) has an approximately 20% to 25% chance of giving birth to an infected infant. HIV can be transmitted transplacentally at any stage of the pregnancy. Most infants are infected at the time of delivery by exposure to maternal blood and body fluids. HIV can also be transmitted during the postnatal period through breast-feeding.
HIV antibody crosses the placenta; thus, 100% of infants born to HIV-positive women are HIV antibody positive. Maternal HIV antibody can persist in the child for up to 18 months. The only certain statement regarding an HIV antibody–positive infant is that the mother is infected.
There are numerous presentations of the child with HIV infection. Some children present in the first year of life with severe thrush, organomegaly, and failure to thrive. Pneumocystis jiroveci (formerly P. carinii), in a previously thriving infant is a classic presentation and should always be considered in a child who suddenly develops severe interstitial pneumonia and respiratory failure. A subset of children with HIV infection remain relatively well for years, with the only signs being isolated developmental delay or thrombocytopenia.
Diagnosis of HIV infection in the infant is made by serum HIV PCR. This test is obtained at 1 and 4 months of age. If these tests are negative, the child is assumed uninfected. Positive test at this time indicates that true infection is likely. Some clinicians follow infants for up to 18 months to document seroreversion to HIV-negative antibody status.
Ongoing efforts continue to reduce mother-to-child transmission. In 1994, AIDS Clinical Trial Group (ACTG) 076 was developed. In this protocol, zidovudine (AZT) was given to the mother starting in the second trimester of pregnancy. During delivery, intravenous AZT was given to the mother. After delivery, AZT suspension (10 mg/mL) was given at a dose of 2 mg/kg every 6 hours for 6 weeks. The transmission rate in this study dropped from 25% to 8%; the study was stopped early because the statistically and clinically significant findings could not ethically be kept from the control arm of the study.
During the past 10 years, the development of highly active antiretroviral therapy (HAART) has led to further reduction in mother-to-child transmission. Strong consideration for elective cesarean section is given, particularly if maternal viral load exceeds 5,000 copies/mL. Intravenous AZT continues to be given at time of delivery. After delivery, oral AZT is given to the child for 6 weeks. The 10 mg/mL suspension is given at a dose of 2 mg/kg orally every 6 hours. Currently, the incidence of transmission under these aggressive protocols is less than 2%. It is likely that further refinements to reduce maternal transmission (including additional drugs administered to the infant) will be developed. Maternal transmission of HIV to infants should be considered a highly preventable condition. All pregnant women should be counseled about HIV testing. This ensures the diagnosis of women unaware of their infection and protection of their unborn children.
If effective protocols for prevention of maternal transmissions are used, transmission of HIV from mother to child should be less than 2%. The diagnosis of true HIV infection is made by PCR, showing viral copies of HIV at 1 and 4 months of age. If this indeed is documented, HIV infection is considered to have occurred in the infant. Protocols for the management of infant HIV infection are continually revised. Most specialist agree that HAART therapy should be given once infant HIV infection has been documented. Although monotherapy is used in the protocols preventing maternal transmission, for actual treatment of confirmed disease, combination of at least three antiretroviral drugs should be considered.
Azimi PH, Janner D, Berne P, et al. Concentrations of procaine and aqueous penicillin in the cerebrospinal fluid of infants treated for congenital syphilis. J Pediatr 1994;124(4):649–653.
Bale JF. Congenital infections. Neurol Clin 2002:20(4):1039–1060.
Bechtel RT, Haught KA, Mets MB. Lymphocytic choriomeningitis virus: a new addition to the TORCH evaluation. Arch Ophthalmol1997;115(5):680–681.
Boyer KM. Diagnostic testing for congenital toxoplasmosis. Pediatr Infect Dis J 2001;20(1):59–60.
Litwin CM, Hill HR. Serologic and DNA-based testing for congenital and perinatal infections. Pediatr Infect Dis J 1997;16(12):1166–1175.
McAuley J, Boyer KM, Patel D, et al. Early and longitudinal evaluations of treated infants and children and untreated historical patients with congenital toxoplasmosis: the Chicago Collaborative Treatment Trial. Clin Infect Dis 1994;18(1)38–72.
Michaels MG, Greenberg DP, Sabo DL, et al. Treatment of children with congenital cytomegalovirus infection with ganciclovir. Pediatr Infect Dis J 2003;22(6):504–509.
Michelow IC, Wendel GD, Norgard MV, et al. Central nervous system infection in congenital syphilis. N Engl J Med 2002;346(23):1792–1798.