Cytomegalovirus (CMV) is a member of the Herpesviridae family, which belongs to DNA viruses, and is common in the general population(1). The rates of seropositivity in adult women are ranging from 40% (in most European countries) to 90% (in African and Asian countries)(2,3).
Congenital cytomegalovirus infection (cCMV) is the most common non-genetic cause of sensorineural hearing loss (SNHL), accounting for 25-30% of cases of childhood sensorineural deafness(4), and one of the congenital infections that cause significant neurological impairment in children from USA and Northern Europe(5).
Several factors contribute decisively to the underdiagnosis of CMV in childhood, which leads to an increased rate of morbidity and mortality in these infants. These include limited awareness of the CMV infectious risk in both parents and doctors(6,7), lack of recognition of the infection (due to nonspecific symptoms in both pregnant women and newborns), lack of screening programs in some countries, and low effective vaccines and treatment regimens(8).
Primary CMV infection occurs as a result of close personal contact and is transmitted through fluids, body secretions or vertically (transplacental), leading to congenital infection in the fetus(3).
CMV infection is the most common intrauterine infection, affecting 0.3-2% of live births(9), being defined as active CMV infection when detected in the first three weeks of life(10).
The overall prevalence of cCMV has been estimated around 0.7%, but it varies widely worldwide, being estimated between 0.48% and 1.3% in the United States(10), 0.54% in The Netherlands(11), and 1.08% in Brazil(12).
In the past, symptomatic cCMV infection was considered to occur only after the mother’s primary infection during pregnancy, and preexisting maternal immunity prevented the fetus from being at risk for infection in recurrent maternal infection(3). These assumptions led to the conclusion that populations with high rates of seroprevalence may have a lower risk of primary maternal CMV infection and therefore lower risks of cCMV, which is not entirely true(3). This assumption was questioned because it was observed that populations with low socioeconomic status and high seropositivity rates in women of reproductive age usually have higher overall rates of cCMV infection (1-2%), compared to the global average (0.4-0.7%)(3).
These observations led to the hypothesis of secondary CMV infections in pregnant women immunized for CMV at conception (either by reactivation of the latent virus or reinfection with a new CMV strain). These infections can also lead to cCMV and fetal infection(13), but with a lower risk rate, around 1.4%(14).
Furthermore, some authors hypothesized that reinfection with a new strain of CMV can lead to a higher risk of cCMV compared with the viral reactivation, but this has not been clearly established(15).
The incidence of primary CMV infection in pregnant women varies between 0.5% and 4%(1), but only about one third (30-39%) of these infections will result in transmitting the virus to the fetus(3). The risk of fetal CMV infection is higher with primary maternal infection and less likely with recurrent infection(3).
Also, the severe forms of disease occur more commonly among fetuses whose mothers suffer from a primary infection in the first half of pregnancy, although the viral transmission rate is 8.3%(16), up to 8.8%(17) in the first 18 weeks of pregnancy, compared to the third trimester, when CMV transmission varies between 40% and 70%, but when newborns are more commonly asymptomatic(17,18).
For these reasons, the Society of Maternal-Fetal Medicine recommends explaining to the pregnant women with primary CMV infection that the risk of congenital infection varies, on average, between 30% and 50%, but the severity of the disease cannot be anticipated(19).
What is certain is that determining the timing of CMV infection is important in order to determine the risk of fetal infection(19), the primary infection having a 20-fold higher risk of vertical transmission compared to CMV reactivation or reinfection (30% versus 1.4%)(3).
Maternal diagnosis of CMV infection
In pregnant women, CMV infection is asymptomatic or usually appears with mild clinical symptoms, and therefore this condition is difficult to recognize if the serology testing on CMV is not done as a routine screening in the first trimester of pregnancy(1). For these reasons, the CMV infection can be easily overlooked, leading to fetal infection and to congenital CMV(1).
Most studies have shown that less than 5% of pregnant women with primary CMV infection have flu-like syndrome, defined by simultaneous occurrence of fever, pharyngitis, cervical lymphadenopathy, fatigue, malaise, myalgia, headache, hepatosplenomegaly or rash(20,21).
Changes in laboratory tests, such as lymphocytosis and increased transaminases, may indicate the presence of CMV infection in 12% to 36% of pregnant women affected by this condition, but these are nonspecific and may be caused by other viral infections(22).
Infectious serological screening is essential for the early detection of CMV infection, but in some countries, such as Australia, New Zealand or USA(3), it is no longer performed, while in other countries (Austria, Belgium, France, Germany, Israel, Italy, Portugal, Spain and The Netherlands) it is carried out only for certain risk groups(23).
However, in Romania, the screening is mandatory in the first trimester of pregnancy, being added to the TORCH profile, along with hepatitis B, C, VDRL and syphilis.
Usually, the presence of IgM antibody indicates an acute infection, but this antibody can be produced during secondary infections, or may be false positive in response to other viral infections, such as Epstein-Barr virus, herpes simplex virus (HSV) or varicella-zoster virus (VZV)(19,24). IgM can also persist for several months following the primary infection, potentially predating pregnancy by a significant period of time(19). Therefore, the presence of IgM alone should not be used for diagnosis(19).
The diagnosis of primary CMV infection is made by documenting seroconversion and the appearance of virus-specific IgG antibodies in the pregnant woman serum, previously known as seronegative. The presence of IgG antibodies indicates a past infection, but the time interval from this infection, especially if IgM are also positive, varies from two weeks to one year(25). Approximately 1-4% of the seronegative women will have a primary infection during pregnancy, and the majority of these women will be asymptomatic(19).
If the IgM CMV antibodies are detected without specific IgG CMV antibodies, the primary infections diagnosis is clear(26). Unfortunately, this clinical situations is rare, because of the rapid increase of IgG antibodies after the acute infection and due to a decreased of CMV screening in many countries(26).
For this reasons, in countries without CMV screening programs or when we have both the IgM CMV and IgG CMV present, the IgG CMV avidity test is considered essential to establish the infection time or to differentiate a primary infection from the secondary one(3,27,28).
The IgG antibodies produced as a response for the primary CMV infections usually have low avidity, evolving in time for a better bond with the CMV antigens(27,28).
The avidity levels are mentioned as an avidity index, described as the IgG percent bonded to the antigens as a result of the agents distortion treatment, and can be a sensitive marker for a primary CMV infection in the last 4 months(19).
For these reasons, the Society of Perinatal Medicine recommends that for women suspected of primary CMV infection during pregnancy, the diagnose should be made using IgG seroconversion, or by IgM CMV +, IgG CMV + and low IgG avidity(19).
4. Polymerase chain reaction (PCR) testing is a fast and sensitive method in detecting CMV, which has become widely available in recent years and has as its mechanism the amplification of nucleic acids. This technique examines either the well preserved genetic antigens fragments, or some specific DNA fragments – showing a high sensibility in pathogen agent detection(29).
Molecular methods can be used for CMV nucleic acids detections, as well as for viral replication diagnosis from specific cells culture(24).
The CMV may be isolated from several biological samples, including blood, urine, saliva, semen, vaginal discharge and amniotic fluid(24). Viremia may persist for up to one month after the primary infection, but detectable viremia has also been reported in seropositive women with recurrent infections usually at lower values(30,31).
PCR identification of CMV DNA can be qualitative or quantitative. Quantitative PCR (real-time PCR) allows both the diagnosis and the treatment surveillance in the severe forms of the disease, but a threshold above which the risk of vertical transmission is higher could not be established(32,33).
The prenatal diagnosis of CMV infection
Only 10-20% of the fetuses exposed to CMV in utero will show signs of infection at birth, such as: intrauterine growth restriction, microcephaly, hepatosplenomegaly, thrombocytopenia, parenchymal calcifications of the brain, ventriculomegaly, cerebellar hypoplasia(34). Furthermore, neurodevelopmental sequelae such as mental retardation and motor/auditory or visual impairment may occur later in life, even in asymptomatic newborns(34).
In case of a primary maternal CMV infection, but without a confirmed fetal infection, the risk of severe fetal sequelae is approximately 3% and the overall risk of fetal impairment is around 8%(19).
The methods used for assessing primary infection are invasive (amniocentesis, cordocentesis) and noninvasive (imaging techniques), the disadvantage of the latter being that some changes may not be visible in the early stages of the infection and also that it cannot assess the severity of the lesions, especially the neurological lesions(35).
1. Invasive prenatal diagnostic techniques: amniocentesis or cordocentesis
The most common method used for prenatal diagnosis of CMV fetal infection is amniocentesis(36,37). The cordocentesis may be performed, having a sensitivity and specificity similar to those of amniotic fluid CMV testing, but with a higher complications rate compared to amniocentesis(36,37), which is why the usefulness in cCMV prenatal diagnosis is disputed(24).
The prenatal diagnosis of cCMV can be made either by viral cell culture or by detecting CMV DNA in an amniotic fluid sample using PCT or real time PCR(24). The molecular diagnosis by PCR is currently preferred to CMV culture, due to a higher sensitivity (90% to 100%)(5).
Although the amniotic fluid CMV-PCR is a test with increased sensitivity and specificity, it depends on when the amniocentesis is performed(22,24). The amniotic fluid taken before 21-22 weeks of pregnancy or less than 6 weeks after the primary maternal infection may have an undetectable CMV load due to the time required for CMV to be excreted by the fetal kidney, which is the primary site of viral shedding(22,24).
As a result, the Maternal-Fetal Medicine Association recommends carrying out amniocentesis for CMV detection after 21 weeks in pregnancy or more than 6 weeks since the primary maternal infection(19).
The use of real-time PCR allows a quantitative determination of the viral load, but currently there are differences of opinion regarding its value due to an increased risk of fetal complications. Some authors claim that a CMV viral load of more than 103 copies/ml in amniotic fluid correlates with fetal infection, and that viral loads higher than 105 copies/ml are associated with more severe forms of the disease(38). Other groups of investigators claim that a viral load in the amniotic fluid of more than 105 copies/ml was associated with symptomatic infection in the newborn or fetus(38,39), while others failed to establish a clear link between the viral load and the degree of fetal/neonatal impairment(40,41).
For this reasons, the professional forums recommend caution in assessing the degree of fetal impairment only on the basis of the CMV viral load from the amniotic fluid(5).
2. The imagistic/imaging investigations
The imagistic studies in cCMV have two main objectives: detecting the fetal structural abnormalities for confirming the fetal impairment, and providing information on fetal prognosis(42).
However, certain sequelae of cCMV, such as chorioretinitis, petechiae and neurodevelopmental disorders, are not detectable by prenatal imaging methods, so the absence of imaging abnormalities does not rule out fetal impairment(43).
At a cellular level, CMV is able to affect multiple cell lines, nonetheless his action is targeted on the fetal brain and kidney(2,44,45). In neurons, astrocytes, glial cells and endothelial cells, CMV causes a number of changes in the neuronal proliferation and migration, as well as in the organization of cortical cells(44,45).
While ultrasound is the method of choice for fetal imaging, being easy and more accessible, MRI can bring additional information especially in the detection of fetal brain abnormalities, both structural and functional (metabolic)(46).
Fetal structural changes identifiable through prenatal ultrasounds are: cerebral abnormalities, such as ventriculomegaly, brain calcifications, microcephaly, occipital horn variations, and non-cerebral abnormalities, such as echogenic bowel, intrauterine growth restriction, hepatosplenomegaly, ascites and cardiomegaly(35,47), growth in placental width (enlarged placenta – due to placental inflammation)(48), oligohydramnios and, seldom, polyhydramnios(2). Rare lesions, such as subependymal cysts or intrahepatic calcifications, have also been reported, but these changes need to be confirmed by other studies(47).
The neurological ultrasound examination in the cCMV-confirmed cases is essential, because the cranial abnormalities like ventriculomegaly and microcephaly are associated with a poor neurocognitive prognosis(49).
The intracranial calcifications are a classic sign found in cCMV, usually described by ultrasound, being observed in the basal ganglia, as a manifestation of lenticular vasculopathy, but also in the cerebral parenchyma(50). Furthermore, according to Doneda et al., the suspicion of cCMV should be raised when these intracranial calcifications are detected in the anterior temporal lobe, also called “polar temporal lesion”(51).
The hyperechoic intestine is a classic ultrasound sign of cCMV, although it is sometimes overestimated(52), being caused either by direct injury of the virus to the intestine, or by intraamniotic hemorrhage(53).
cCMV is associated with an increased risk of fetal death in utero, even in the absence of suspicious ultrasound changes(43).
Magnetic Resonance Imaging (MRI)
Fetal MRI can be used especially in the detection of neurological abnormalities, particularly if the ultrasound has detected abnormalities at this level(54).
Fetal MRI is complementary and often superior to ultrasound in detecting abnormal gyration and myelination(51,55).
The timing to perform fetal MRI in cCMV is not well established, being preferred at the beginning of the third trimester (27-33 weeks of gestation), because some lesions, such as neuronal migration, are more difficult to detect in the first two trimesters(50,54).
The lesions identified by MRI in cCMV are often nonspecific, the most common abnormalities observed being ventriculomegaly and white matter (WM) signal abnormalities or cerebellar hypoplasia(42,56). Other brain lesions that can be seen on MRI are cerebral cysts, dilation of the temporal horns, ventriculitis and intracranial calcifications(50,51).
Congenital CMV infection affects the germinal matrix, leading to the loss of neurons and their disruption of migration when the infection occurs before 16-18 weeks of gestation(2,50). Lissencephaly and polymicrogyria can be seen in infections that occur between 18 and 24 weeks, while fetuses exposed to CMV in the third trimester usually have a normal gyral pattern(2,50).
Polimicrogyria occurs mainly in the frontal and perisilvic regions, due to a neural migratory disorder(57).
Hoffman et al.(58) found, using MRI examination, that the volume of the brain is smaller in fetuses with cCMV, an observation also confirmed by Grinberg et al.(9) One possible explanation of these MRI modifications could be the loss of neuronal stem cells, the disruption of migration and the differentiation of brain stem cells, as well as accentuated local inflammation and hypoxia(44).
Furthermore, Grinberg et al. found on MRI examination that there is a correlation between low cerebellar volume and neurobehavioral disorders in childhood, especially for daily activities and communication skills(9). There is an association between cerebellar lesions and the presence of some cognitive disorders, such as speech, behavioral, social and motor deficiencies(9,59).
Cerebellar abnormalities are a common feature in postnatal MRI for cCMV(60), but are less common in fetal MRI, either due to the fact that they occur later in pregnancy, or to the fact that children with symptomatic cCMV undergo more detailed imaging examinations(51,61).
Cerebellar hypoplasia and dysplasia can affect the vermis, cerebellar hemispheres, or both, and are associated with a high risk of childhood sequelae(2,51).
Other abnormalities of cCMV identifiable on MRI are: hepatomegaly, splenomegaly, intrauterine growth restriction (FGR), hyperechoic bowel, cutaneous edema, ascites, pleural and pericardial effusion, hydrops, oligo-/polyhydramnios and placentomegaly(48,62).
The isolated cases of hyperechoic bowel detected on ultrasound without other organic lesions are associated with normal MRI and a good prognosis(43).
In situations where both MRI and ultrasound are normal, the fetal prognosis is good and the rate of major complications in childhood is low(63).
However, normal brain imaging does not completely rule out the presence of neurodevelopmental disorders in childhood, especially since hearing loss is frequently progressive in cCMV.
The postnatal diagnosis of CMV infection
Congenital cytomegalovirus infection is the most common non-genetic cause of sensorineural hearing loss, accounting for 25-30% of cases of childhood sensorineural deafness(4).
Between 5% and 10% of the newborns with congenital cytomegalovirus infection are symptomatic, mainly the central nervous system being affected(64). On the other hand, the majority of the asymptomatic infants remain unidentified due to the lack of clinical manifestations(5).
The most common symptoms in the cCMV infection are jaundice, petechiae, hepatosplenomegaly and neurological anomalies, such as microcephaly and intracranial calcifications, which are found in up to 75% of cases(3,65). Less frequent signs that can also be observed are cataract, microphthalmia, myocarditis and cardiac defects(3,65).
Hepatobiliary abnormalities, confirmed by increased transaminases and conjugated hyperbilirubinemia, can be seen in 23-80% of symptomatic newborns, but they are only transient, with normalization in a few weeks(3,65).
The congenital CMV infection diagnostic can be confirmed by identifying the virus or the viral antigens in the newborns’ urine or saliva in the first two weeks of life(5). The presence of the virus or viral antigens after three weeks alone cannot confirm cCMV, because after this interval the CMV infection can be one acquired at birth or postnatal(5).
The serological methods are not very accurate for the postnatal confirmation of the diagnosis of cCMV, because the detection of CMV-IgG antibodies in fetal blood may be due to transplacental transfer of maternal antibodies, and the detection of CMV-IgM antibodies do not have a high level of sensitivity and specificity(66).
The effectiveness of antiviral treatment during pregnancy is still a hot topic, due to the questionable efficacy and potential teratogenic risk of drugs. The clinical benefits of CMV-specific hyperimmune globulin treatment are also disputed, due to the discordant results obtained from various studies(16,67).
The treatment options for cCMV infections are still limited, because most drugs act by inhibiting CMV viral replication, but cannot remove the virus from the human body. It has also been observed that, after stopping antiviral therapy, the viral load in the blood tends to increase(68).
Studies have shown that a six-week course of ganciclovir, started in the neonatal period, is effective in reducing the severity of neurological sequelae and hearing loss in both symptomatic and asymptomatic infants(69,70), but unfortunately there is no benefit of a long-term use(68).
Regarding the antenatal use of CMV hyperimmune globulin (CMV HIG) to prevent infection in newborns, in a study published in 2014, the authors found no benefit of treatment, but there were a number of side effects in those who received CMV HIG(67).
The American College of Obstetricians and Gynecologists (ACOG) does not currently recommend the prenatal treatment with ganciclovir or valaciclovir because it has not been shown to be effective(71). These observations are also reinforced by the Society of Maternal-Fetal Medicine, which does not recommend the prenatal treatment with ganciclovir or valaciclovir or the antenatal therapy, either with antivirals or with CMV HIG(19).
CMV remains the leading infectious cause of birth defects in both the fetus and the newborn.
An increasing number of evidence indicates that the CMV reinfection during pregnancy contributes to a much larger proportion of symptomatic cCMV than was previously supposed.
For this reason, avoiding exposure of pregnant women to CMV or serological screening should be recommended for both seronegative and seropositive pregnant women.
In cases of confirmed or suspected primary maternal CMV infection, amniotic fluid viremia obtained by amniocentesis, as well as prenatal ultrasound and, possibly, MRI evaluation are required to determine the risk of cCMV and to assess the possible fetal damage.
MRI can bring additional information especially in the case of brain abnormalities and can achieve a better assessment of the neonatal prognosis.