Immunological evidence of in utero infection
This may be by one of two ways: persistence of VZV-specific IgG at one year of age or the detection of VZV-specific immunoglobulin (Ig) M in the neonate or infant. As VZV IgM assays generally lack sensitivity, the former is used. Alternatively, if the child presents within the first two years of life with zoster but without a history of clinical chickenpox, in utero infection must have occurred.
Prevention of Infection
In order to prevent varicella in pregnancy any woman in contact with chickenpox or uncovered zoster who does not have a history of previous infection should have blood taken and tested within 48 hours for VZV-specific IgG. If shown to be antibody negative and the last contact is within the previous 10 days, varicella zoster immune globulin (VZIG) should be administered. Whilst this may not prevent maternal infection, especially if given late (>96 hours after contact), it should at least ameliorate the woman’s illness. This is especially important as life-threatening varicella pneumonia is more common in pregnancy, in particular in the third trimester. It is important to remember, however, that if a woman has a very good history of previous chickenpox, or indeed zoster, antibody testing is not necessary and may be misleading, as the tests lack the required sensitivity to detect every individual who has been infected with VZV. This could lead to the overuse of VZIG.
One of the most significant outcomes from the UK–German study was to demonstrate that, even if given as late as 10 days after a virus exposure, VZIG may still be effective in preventing fetal infection even if the woman goes on to develop chickenpox.2 This is because it is the maternal rash which signals the viraemia that infects the placenta. The presence of intravascular VZV IgG at this time should therefore mitigate against this and so reduce the proportion of fetuses which are infected. This was indeed borne out by the Enders study, which demonstrated that no cases of varicella embryopathy occurred in 97 fetuses in which maternal varicella developed despite post-exposure prophylaxis with VZIG.2 Indeed, with the caveats of the sensitivity of VZV-specific IgM assays set aside, it was also found that the proportion of these fetuses with VZV IgM was significantly lower compared with asymptomatic infants whose mothers did not have prophylaxis. Finally, there were no cases of varicella embryopathy in the 366 mothers who developed zoster in pregnancy as would be expected, because there is rarely any viraemia with zoster and, even if it were to occur, maternal VZV IgG would quickly neutralise it.
Prenatal Diagnosis
Prenatal diagnosis is difficult because the virus is restricted largely to nervous tissue and is not often shed in urine or does not always reactivate to produce skin lesions. This makes detection of viral DNA by PCR in amniotic fluid (AF) less sensitive than it is for other herpes viruses, e.g. cytomegalovirus (CMV). Given a transmission rate of VZV of 20–50% depending on the trimester during which maternal infection takes place, the sensitivity of AF by PCR is relatively low. Virus transmission to the fetus was seen in only 9/107 (8.4%) in cases of chickenpox before the 24th week.4,5
Varicella developing at the end of pregnancy potentially exposes the fetus to infection without the benefit of passive maternal antibody. Thus, if maternal infection occurs either seven days before or after delivery the newborn will not have sufficient antibody to prevent a severe infection, so is recommended to receive VZIG; consideration should also be given to giving prophylactic, post-exposure aciclovir, especially if the maternal rash developed up to four days before birth. Reports of mortality of c. 30% are exaggerated, but the mortality is not insignificant and probably lies at c. 1–5%.
Several unresolved problems with the management of pregnancy-related chickenpox remain, however. As the defects may be severe and potentially disabling due to viral reactivation with variable timing and extent, it is important that the value of prenatal diagnosis is explored more extensively and whether or not treating pregnant women with aciclovir has any effect on the risk of transplacental infection.
It has been suggested that antenatal screening for VZV IgG should be considered in order to vaccinate susceptible women postpartum, though because of the low risk of adverse outcomes and low incidence of varicella in pregnancy (c. 0.3–0.7/1000/year) this may not be necessary if varicella vaccine is added to the childhood immunisation schedule. What should not be forgotten, however, is to re-test women for VZV IgG who do not develop chickenpox at least eight weeks after receiving VZIG to exclude asymptomatic infection. This will allow those women who escaped infection, and who would, therefore, benefit from postpartum vaccination, to be identified.
Cytomegalovirus
The general characteristics of the pathogen are:
Herpesvirus; multiple strains; partial immunity – re-infection common
Global distribution
Enveloped dsDNA
Spread by person-to-person contact through contact with infected bodily fluids
Incubation period c. 3 – 4 weeks
Vaccination in clinical trials
Treatable with antivirals (ganciclovir [GCV], valganciclovir, foscarnet).
Cytomegalovirus (CMV) is the commonest congenital infection and complicates c. 0.5% of all pregnancies.6 Both primary infection and recurrent infection (a mixture of reactivation of the woman’s endogenous strain or re-infection with a different one) are capable of leading to congenital infection.7
Epidemiology, Transmission and Virus Survival
The women at highest risk of primary infection are those who already have their first child attending a day-care nursery. By the same token, nursery-care workers who do not adopt good infection control are at substantial risk, as are teenagers. The reasons behind these high-risk groups are principally to do with how the virus is acquired and where it is shed. Approximately 30–40% of nursery-aged children shed virus in urine and transmission depends on good personal hygiene and attention to environmental decontamination. This is because CMV remains viable on metal/wooden surfaces or glass/plastic for one or three hours respectively, though drying of such surfaces reduces the infectivity considerably. This explains why the annual sero-conversion rate can be as high as 10–20% in day-care nursery workers. Teenage pregnancies are at increased risk because of their young age, making it more likely that they are susceptible. Teenagers can become infected as a result of sexual exposure. This group mostly experiences primary infection. However, African and Asian women are mostly at risk of re-infection as the prevalence of past infection at childbearing age is greater than 95% among these ethnic groups compared with a prevalence of approximately 40–50% in Caucasian women. This difference may be due to differences rates of breastfeeding, since the most effective mode of transmission is via breastfeeding. This mode of transmission simply results in an asymptomatic infection of the infant takes place to be followed shortly afterwards by urinary excretion.
It is important to remember that there is no such thing as CMV immunity, and in fact that the level of protection that a previous infection affords against infection with a different strain is only c. 60–70%. Finally, there is an increased risk of placental transmission is if conception occurs within 6–12 months of a primary infection.
Table 21.2 shows the sources and routes of transmission of CMV infection.
Sources of virus | Route of transmission |
---|---|
Breast milk | Breast feeding |
Urine | Changing nappies and poor hand-washing |
Saliva | Socially |
Genital secretions | Sex Peri-/intra-partum |
Blood/organs | Iatrogenic |
Clinical Features of Maternal Infection
Often a primary or recurrent CMV infection is asymptomatic. If symptoms do occur, however, the classical syndrome of a glandular-fever illness akin to Epstein–Barr virus infection is actually much less common than an illness comprising vague lethargy and malaise associated with a moderate fever of c. 1–3 weeks duration. Often, if such patients are investigated a biochemical hepatitis with an alanine transaminase (ALT) level of up to 5–8 times normal and a lymphocytosis or lymphopaenia are seen. The monospot test is very often negative.
Clinical Features of Congenital Infection
The clinical features of congenital CMV infection are almost always not apparent at birth and reveal themselves later in infancy as failure to thrive, or because of a failure to reach milestones at the appropriate times. If the baby is symptomatic at birth, seen only in 7–10% of cases, it is often small for its gestational age and has microcephaly, hepatitis, splenomegaly and thrombocytopaenia. A chronic respiratory illness may also be a presenting feature.
Table 21.3 shows the range of clinical findings in babies symptomatic at birth with congenital CMV.
Clinical feature | Prevalence at birth (%) |
---|---|
Petechiae | 76 |
Jaundice | 67 |
Hepato-splenomegaly | 60 |
Sensorineural hearing loss | 56 |
Microcephaly | 53 |
Small for gestational age | 50 |
Hypotonia | 27 |
Retinitis | 10 |
In 7–10% of babies clinically affected at birth, half will have cytomegalic inclusion disease (CID), with multi-organ involvement of the reticulo-endothelial system and the central nervous system (CNS) – namely hepatosplenomegaly, jaundice, petechiae, microcephaly, seizures, hypotonia and intra-cerebral calcification. Less common manifestations include pneumonitis, dental defects and ocular problems (retinitis, cataracts, strabismus). The overall mortality in this group is 15–30%, mainly in the first weeks of life. Of the survivors, 90% will have long-term problems caused by the CNS involvement – sensori-neural hearing loss (SNHL), language delays, microcephaly, visual impairment and psychomotor retardation.
However, the majority (90–95%) are asymptomatic at birth and in this group the prognosis is much better, but still 15% will go on to develop sequelae, principally SNHL. Globally, congenital CMV is the cause of 10% of SNHL. There is a wide range of hearing loss from bilateral to unilateral high-frequency loss only. Overall, between 2% and 10% develop mental retardation with 1–2% developing retinitis, and less than 1% show signs of cerebral palsy in later life. Therefore, given the risk of placental transmission of 30–60% depending on gestational age at which maternal infection occurs, the overall risk of having an affected baby to a greater or lesser degree is c. 20–30% in a woman with a primary infection.
Gestational Age and Risk to the Fetus
In women experiencing a primary infection, the principal determinant of the degree of damage to the developing fetus is the trimester at which infection takes place. For a preconceptual or periconceptual infection, the risk of a bad outcome is not always low, however. Daiminger et al. found no infections in 3 women with infections 2–8 weeks before the last menstrual period (LMP), but among 20 women with infections closer to the LMP (1–5 weeks before), 9/20 (45%) suffered fetal infections with 5 fetal losses and of the 4 infected, 2 were severe.8 Revello et al. also demonstrated a rate transmission rate of 10% (2/25) for preconceptual transmission – both cases being subclinical, but of the 13 women with periconceptual infection transmission was seen in 4/13 (30%).9 Other authors have noted an increasing rate of transmission as the pregnancy advances, with approximate transmission rates of 30–40% first trimester, 40–50% second trimester and 70% third trimester.
It is well established that the risk of a poorer outcome is increased if infection occurs before 20 weeks, with approximately 25–30% having CNS abnormalities compared with just 6% where infection occurred after 20 weeks.
Whilst recurrent CMV infection is thought to pose a lower risk of transmission, in communities and ethnic groups where childhood infection is common and recurrent infection in pregnancy is responsible for 90% of maternal infections, the rates of CMV-induced SNHL are no different from countries where primary infection is the principal type of infection.
Laboratory Diagnosis of Maternal Infection
If infection is suspected in the woman because of a compatible clinical illness and/or ultrasound features in the fetus, close liaison between the obstetrician and microbiologist is essential as the diagnosis of CMV infection is often complicated and time critical as the gestation advances. One of the complications is because the available tests for CMV IgM are unreliable as they lack both sensitivity and specificity; low-level IgM may be significant and high-level IgM may be seen to persist for many months and even more than a year. If abnormal findings are present at the anomaly scan at 20 weeks, demonstrating a CMV IgG sero-conversion, whilst definitive, is an unusual event even with recourse to a booking blood sample as most often both IgG and IgM are present at booking. Thus, as CMV IgM may persist it is essential that maternal serum samples be tested by a CMV IgG avidity assay.10 This is a much more specific and sensitive test for primary infection, and because it can distinguish primary from nonprimary infections, it very often helps to resolve low-level IgM, which often proves to be of no significance.
CMV IgG Avidity Assay
This assay looks at the maturation of virus-specific IgG. This is always a time dependent process and takes a set time in any individual assay, but is often on average three months, and is ideally no longer than four months, to show the development of high-avidity IgG. The exception to this is in immune-compromised patients in whom the maturation of the IgG response is significantly prolonged. The assay compares the reactivities of a single serum sample in an enzyme immunoassay (EIA) in two individual wells after incubation and the binding of the patient’s CMV-specific IgG to virus antigen, which in one well is followed by a normal wash and in the other by the addition of a denaturating agent. The agent most often used is 6–8 M urea and it breaks the weak bonds between antigen and immature or low-avidity antibody compared with a control. An avidity index is then obtained by comparing the reactions in the two wells where, always, a low avidity indicates a current primary infection whilst high avidity indicates a past infection (at least 3–4 months old) or, at worst, a recurrent infection.
Assessing maternal viraemia is often performed, though diagnostically is of dubious value. Whilst a viraemia in a pregnant women with suspected CMV infection may be present, it requires a highly sensitive nested PCR to be used. With these assays the virus may be detectable, but often only at the limit of sensitivity of the assay and up to three months after diagnosis or symptom onset. Thus, serology is very often able to establish the diagnosis on its own. Furthermore, when used retrospectively and/or in samples with low IgM, the diagnostic utility is limited as no use as a negative PCR result does cannot exclude maternal infection.
Laboratory Diagnosis of Congenital/Fetal Infection
The diagnosis of congenital CMV infection after birth is very straightforward if the syndrome is suspected within the first three weeks of life, but this is an unusual event in ordinary clinical practice. In this example, a urine or blood specimen may be tested for virus, either by cell culture or, more likely, PCR. As the maximum incubation period of CMV is three weeks, a positive result must indicate that in utero infection has occurred. However, as most babies are not suspected clinically at birth, the often used approach to confirm or refute a congenital infection is to test the newborn blood spot taken within seven days of life for CMV DNA by PCR and to test current and booking maternal blood samples for CMV IgM, IgG and IgG avidity. The sensitivity of PCR in the newborn-dried blood spot is c. 70–90%. It may also be possible to make a diagnosis by detecting a maternal primary CMV infection by either a sero-conversion or by finding low-avidity CMV IgG at booking, or perhaps a re-infection by detecting a rise in IgG titre with or without CMV IgM. Congenitally infected newborns, whilst often shedding large quantities of virus in urine, should be treated in exactly the same manner as other babies on the ward or neonatal unit. Suffice to say that adherence to normal hand washing is extremely effective in preventing transmission to healthcare workers. Indeed, healthcare workers are at no greater risk of acquiring CMV compared with the general public if they adhere to good infection‐control practices.
Prenatal Diagnosis
The types of ultrasound findings in fetuses where virus transmission has occurred include: ventriculomegaly; increased periventricular echogenicity or pseudocysts; microcephaly; intra-cranial calcification; abnormalities of cortical development/cerebellum; echogenic bowel; intrauterine growth restriction (IUGR); splenomegaly; cardiomegaly; pericardial/pleural effusion; and hydrops.
Prenatal diagnosis of CMV is now relatively straightforward, providing that there is clear communication between the obstetrician and the microbiologist. Once maternal infection has been diagnosed and the ultrasound scan (USS) is normal, sufficient time must be allowed to elapse to allow transplacental transmission, fetal infection and replication of virus to take place, such that the virus is detectable in AF. It may be possible to attempt an amniocentesis early at c. 16–18 weeks of gestation if more than 6 weeks has elapsed after maternal diagnosis even though urinary production will not be guaranteed. This would allow an earlier termination if recommended or desired. Nevertheless, at this stage, CMV DNA PCR on AF at 16 weeks is only approximately 50% sensitive.11
Despite waiting for urine production to occur (>21st week), occasional false negative results occur, which may necessitate a repeat sample; of these a quarter may be due to early sampling whilst the rest are due to late transmission of the virus. Approximately 4% of babies with negative PCR results excrete virus at birth so the sensitivity of AF PCR is c. 95%.
The magnitude of the CMV viral load in AF is most closely related to the timing of the maternal infection and amniocentesis, it has no significance for the severity of the fetal damage. Even when the fetus is infected as shown by AF CMV PCR, it is possible that fetal blood sample may still assist in decision-making; thrombocytopenia is an independent predictor of a poorer outcome with regard to CNS development.
It is nevertheless important to remember that if suspicious fetal ultrasound findings (echogenic bowel, small for gestational age, microcephaly, etc.) are seen in a pregnant woman with a confirmed CMV infection, the two are almost always linked.