Initial CMV infection triggers the production of specific IgM antibodies, followed by lifelong IgG antibodies. However, the mechanisms by which CMV establishes latency and persists within the host remain poorly understood. Latency plays a critical role during pregnancy, as, in seropositive hosts, which constitute a reservoir of viruses, episodes of reactivation and viral replication may happen, increasing the risk of fetal exposure and viral transmission. This risk becomes more significant as the seroprevalence of CMV rises in the population, potentially increasing the likelihood of fetal exposure, even in pregnant women who are already IgG positive [7]. CMV seropositivity during pregnancy is common, although its prevalence varies substantially across countries, being approximately 60% in France [10], 72% in Finland [11], and 98% in Brazil [12].

Consistent with these differences, the proportion of congenital CMV cases attributable to maternal non-primary infection is considerably higher in Brazil (90%), where maternal seroprevalence is very high, than in Finland (53%) or France (48%). This pattern differs from that observed for many other viral infections, in which maternal infection before pregnancy usually induces an immune response that provides substantial fetal protection [13].

In CMV infection, however, pre-existing maternal immunity does not reliably prevent viral reactivation, reinfection with a different strain, or transmission to the fetus. This distinctive pathophysiology, together with the incomplete protection conferred by the immune response, helps explain the difficulty of developing a truly effective vaccine. More recently, mRNA-based strategies have emerged as promising approaches for CMV immunization. In a rodent model, transamniotically delivered mRNA reached the fetal circulation through the placenta and induced a CMV-specific humoral immune response that persisted into the neonatal period, without apparent adverse effects on maternal, fetal, or neonatal morbidity and mortality. Although still experimental, this minimally invasive approach may represent a potential preventive option in the future [14].

The probability of transplacental viral transmission to the fetus is a key clinical issue. Foetal infection in cases of early primary maternal infection occurs with an average incidence of 30% to 40% and increases in later pregnancy [15]. Immunological differences between mothers who transmit the virus and those who do not appear to play a role, although the exact mechanisms remain unclear. Maternal–fetal viral transmission is less common in immune women, occurring at a rate of 1% to 2%, indicating that some seropositive mothers can still transmit the virus. However, the literature supports that pre-existing adaptive immunity to CMV does not always prevent infection with a new CMV strain [16,17].

Early primary infection during pregnancy results in spontaneous abortion in approximately 15% of cases, typically with placental infection preceding fetal infection. Later in pregnancy, maternal CMV infection can cause preterm birth and intrauterine growth restriction in approximately 25% of affected newborns. These outcomes appear to be associated with placental pathology, suggesting that placental infection precedes viral transmission to the fetus [18].

The placenta protects the embryo and fetus through a complex structural and immunological network that blocks pathogens and contains antimicrobial factors [19]. In particular, the chorionic villi represent the first fully functional interface between the fetal and maternal circulations. During the second and third trimesters, the cytotrophoblast layer gradually regresses, leaving the syncytiotrophoblast as the primary barrier at the maternal–fetal interface. This layer facilitates the exchange of gases, nutrients, and waste products while also serving as a selective barrier against microbial invasion. Another key component of the placenta is its population of resident immune cells, known as Hofbauer cells, which are fetal-derived macrophages located within the chorionic villi. These cells perform essential immunological functions, including antigen presentation, phagocytosis, cytokine production, and coordination of both innate and adaptive immune responses [20]. Hofbauer cells are frequently targeted by various pathogens, including viruses, because of the expression of surface proteins such as CD4, which may facilitate viral entry. However, CD4 expression is believed to play a regulatory rather than pro-inflammatory role. Some studies also suggest that Hofbauer cells may contribute to the maternal-to-fetal transfer of antibodies via surface Fc receptors, potentially supporting neonatal immunity [21].

In the context of primary CMV infection during pregnancy, studies comparing women who transmit the infection to their fetus with those who do not have highlighted several immunological factors that may reduce the risk of transmission. These include the presence of neutralizing antibodies targeting CMV glycoprotein B, high-avidity IgG, and antibody-dependent, Fc-mediated cellular cytotoxicity, all of which appear to contribute to protection against fetal infection [22,23,24,25].

The group of Maidji and Pereira found that in women with a weak antibody response to CMV, IgG/virion immune complexes undergo transcytosis via neonatal FcR expressed by placental syncytium-trophoblasts. This process helps the virus avoid destruction by chorionic villi macrophages and spread to fetal vessels [26,27,28].

Thus, passive treatment with CMV hyperimmune IgG may be biologically plausible, although its efficacy remains controversial because it has not been confirmed in large randomized controlled trials [29]. Moreover, peripheral blood T-cell responses (CD4+, CD8+, and CD45RA+ memory cells) have been associated with lower rates of viral transmission to the fetus. Alfi et al. used an ex vivo model of CMV infection at the maternal–fetal interface to identify immune correlates of protection against transplacental transmission. Their findings indicate that CMV-specific, decidual tissue-resident memory CD8+ T cells are key drivers of protection. This protective response is absent in primary-infected tissues from CMV-seronegative women but is associated with resistance to non-primary infection in decidual tissues from HCMV-seropositive women [30].

However, these peripheral immune events may not fully reflect those occurring at the maternal–fetal interface, which could play a more significant role in determining the likelihood of fetal infection. The initial stage of placental infection probably occurs on the maternal side of the placenta, that is, in the maternal decidua, where invasive fetal cytotrophoblasts coexist with maternal immune cells. A study comparing placentas from women with primary or recurrent CMV infection found the virus mainly in fibroblasts and villous endothelial cells, as well as in macrophages, lymphocytes, and trophoblasts. Mothers of symptomatic infants showed more chronic villitis and greater villous alterations than mothers of asymptomatic infants. Additionally, mothers with non-primary infections had more damaged decidua than those with primary infections [31]. During pregnancy, chorionic villi are bathed in maternal blood, which comes into direct contact with the syncytiotrophoblast, the outer layer of the villi, and with cytotrophoblasts that invade the uterine vessels [31].

The hypothetical pathway of CMV fetal infection was demonstrated in a sophisticated experimental study conducted by Fisher et al. in the early 2000s. First-trimester chorionic villi and isolated cytotrophoblasts were exposed in vitro to CMV: syncytium-trophoblasts were not infected, while underlying cytotrophoblast clusters expressed viral proteins [32].

Cytotrophoblasts supported viral replication in vitro, impairing their differentiation and invasion of the uterine wall. Infection appeared to affect the overall cell population, as villi containing infected cytotrophoblasts showed altered architecture and reduced invasive capacity compared with uninfected villi. These findings suggest that CMV may reach the embryo–fetus through infection of underlying and invasive cytotrophoblasts. In addition, CMV impairs cytotrophoblast function, which may help explain the adverse pregnancy outcomes associated with maternal infection [33].

It is also relevant in the context of placental infection to consider the broad repertoire of CMV-encoded immune-modulatory genes that shape host antiviral responses at the maternal–fetal interface. CMV has evolved multiple gene products that interfere with antigen presentation and innate immune recognition, including viral homologues or functional mimics that modulate leukocyte activation, cytokine signaling, and natural killer (NK) cell surveillance. This is particularly important in the placenta, where decidual NK cells contribute to immune homeostasis, trophoblast invasion, vascular remodeling, and containment of infection. Several CMV immune-evasion mechanisms converge on NK-cell pathways, for example, by altering the expression or function of ligands involved in NK-cell activation and inhibition, thereby reducing the capacity of infected cells to be recognized and eliminated. In this setting, viral interference with NK-cell responses may facilitate local viral persistence, perturb placental immune balance, and potentially contribute to transplacental transmission and adverse fetal outcomes. Therefore, CMV genes targeting NK-cell immunity are of particular interest when discussing the impact of infection on placental health [34].

Infection in pregnancy and damage to the maternal–placental interface can lead to a hypoxic environment, resulting in compensatory placental remodeling. The balance between vascular endothelial growth factor (VEGF), which promotes angiogenesis, and placental growth factor (PlGF), which modulates trophoblast growth and differentiation, regulates normal placental development from early pregnancy onward. Under experimental hypoxic conditions, cultured cytotrophoblasts significantly upregulate VEGF expression and develop markedly dilated blood vessels around the villi to facilitate oxygen transfer to the fetus. This response suggests an inflammatory reaction aimed at increasing villous surface area to compensate for uteroplacental hypoxia. A similar phenomenon, also observed in women who smoke during pregnancy, may help explain the low birth weight of infants affected by congenital infection [35].

All women are susceptible to CMV infection, and diagnosis during pregnancy can be challenging in the absence of prenatal serological screening, regardless of whether the infection is primary or recurrent. Clinical manifestations in pregnant women with CMV infection are often vague, minimal, or absent. Symptoms occur in approximately one-third of cases and typically resemble a mild flu-like syndrome, including fever or low-grade fever, malaise and myalgia. Consequently, diagnosis of CMV infection rarely relies on clinical presentation alone [36]. Laboratory findings are generally non-specific and may include mild lymphocytosis and elevated transaminases. Therefore, diagnosis relies primarily on virological and serological testing [37]. Detection of CMV DNA by polymerase chain reaction (PCR) in biological fluids such as saliva, urine, or blood represents a major diagnostic approach, with high sensitivity and specificity, particularly when performed within the first weeks after maternal infection or in neonatal screening contexts. However, diagnostic accuracy may vary depending on whether the maternal infection is primary or non-primary, as well as on the timing between infection and testing [18]. Furthermore, Tanimura et al. reported that the presence of fetal ultrasound abnormalities and a positive PCR result in uterine cervical secretions were independent predictive factors for congenital CMV infection in CMV-IgM-positive women [38]. Table 1 summarizes the serological profiles observed in women undergoing evaluation of their CMV infectious status.

Primary infection is characterized by the absence of IgG, the presence of IgM, and variable PCR results. The highest risk of maternal–fetal CMV transmission and neonatal complications occurs during early primary infection, particularly in the periconceptional period and the first trimester. Chatzakis and Ville, in a large meta-analysis, found that first-trimester infections carry a 23% risk of late sequelae in neonates, compared with 0.1% and 0% when infection occurs in the second or third trimester, respectively [39]. Similarly, Faure-Bardon et al. studied 255 women with primary infection and their 260 infected children, analyzing follow-up data when the children were at least 1 year old. The timing of primary infection was determined by repeated measurements of maternal IgG and IgM, together with IgG avidity in serum collected during each trimester. Infection dating was prospective in 86% of cases and retrospective in 14%. Children were followed from birth to 48 months. At 24 months, rates of sensorineural hearing loss and/or neurological sequelae were 32.4% after first-trimester infection, 0% after second-trimester infection, and 0% after third-trimester infection (p < 0.0001). CMV infection appears to be more severe during the embryonic period, leading the authors to recommend auditory and neurological follow-up only for neonates with congenital infection after first-trimester maternal infection [40]. Therefore, CMV screening should ideally be performed before or at the beginning of pregnancy to assess maternal serological status and identify women at risk of primary infection during gestation. In women with negative CMV serology, retesting every 4 weeks until 14–16 weeks of gestation is advised, as primary maternal CMV infection is diagnosed by the presence of specific CMV IgM [2].

The latest-generation IgM assays have high sensitivity but lower specificity because IgM may persist after primary infection. If IgM is positive, IgG avidity testing is recommended to help confirm recent primary infection [41].

Conversely, current IgG avidity assays may help differentiate between recent (<3 months) and older (3–6 months) primary maternal infections. Consequently, high IgG avidity in the first trimester strongly argues against a primary infection acquired during that period, including the periconceptional period (±2 weeks from conception) and the preconceptional period (2–8 weeks before conception).

There are three key limitations of CMV IgG avidity testing:

Threshold variability: low- and high-avidity cutoffs vary across commercial assays;

Whenever preconception testing is performed, CMV serology should be recommended, and IgG avidity testing may be useful when recent primary infection is suspected.

In cases of recent primary infection, IgG avidity is often low (80–90% of cases) or intermediate. Notably, intermediate avidity can persist for up to six months after infection. When detected during suspected primary infection, retesting serum in specialized laboratories is recommended. Repeat testing can help clarify whether the infection is truly recent or has been present for a longer period, with important implications for clinical management [42,43].

Detection of CMV DNA in maternal blood may support the diagnosis of recent infection, particularly in the early phase, but its sensitivity is variable and DNAemia is transient. Viral DNA levels decline over time and may become undetectable, limiting their diagnostic value. Moreover, the presence of CMV DNA in blood or urine does not allow distinction between primary and non-primary infection nor accurate determination of the timing of infection, as viral shedding may persist for months with significant interindividual variability [44].

Diagnosing non-primary CMV infection without prior serological history is challenging [2]. No single laboratory test can reliably identify women with pre-existing immunity who are at risk of transmitting infection to the fetus, particularly in non-primary infections [45]. Serological markers are of limited value in non-primary infection, as CMV IgM is detected only in a minority of cases and may also persist or reappear during reactivation or reinfection, while increases in IgG titers are inconsistent and non-specific. Molecular testing is also limited, as CMV DNA may be detected intermittently in blood or other biological fluids, reflecting transient viremia and variable viral shedding. Thus, the diagnosis of non-primary CMV infection is often presumptive, and confirmation of fetal infection relies on invasive testing, particularly amniocentesis with CMV DNA detection, while serial serological and virological monitoring may provide only supportive information [8,46].

Approximately 1 in 200 neonates in high-income settings is affected by congenital CMV, making it the most common congenital infection [59]. Approximately 11–13% of infected infants are symptomatic at birth, and an estimated 40–58% of these infants will develop permanent sequelae. Even in the absence of neonatal symptoms, some infected children may later develop sequelae such as sensorineural hearing loss (SNHL), intellectual disability, and neurodevelopmental delay, which occur in approximately 8–15% of cases [60].

Congenital CMV infection should be suspected in a child born to a mother with documented infection (e.g., positive CMV PCR in amniotic fluid and seroconversion), suspected infection (e.g., flu-like syndrome during pregnancy or positive IgM results not further investigated), or prenatal ultrasound abnormalities suggestive of infection, such as ventriculomegaly, cerebral hyper echogenicity, subependymal cysts, hyperechogenic bowel, or hepatic calcifications [2].

In addition, any newborn presenting with symptoms, laboratory abnormalities, or ultrasound findings consistent with congenital CMV infection should be evaluated (Table 2) [2].

The gold standard for the diagnosis of congenital CMV infection is a positive real-time PCR result for CMV DNA in urine or saliva collected within the first 3 weeks of life. Comparison of PCR testing on saliva versus urine shows high sensitivity (93–100%) and negative predictive value (98–99%), together with moderate specificity (91–99.7%) and a relatively low positive predictive value (49–73%) [2]. False-positive saliva results may result from contamination by genital secretions or recent breastfeeding and are generally associated with low viral load. Therefore, a positive PCR result on saliva should be confirmed by PCR testing on urine [2]. After the first 3 weeks of life, the presence of CMV DNA in urine or saliva cannot definitively establish congenital infection. In children older than 3 weeks, routinely collected dried blood spots (DBS) from the first week of life can be examined retrospectively to detect CMV DNA in neonatal blood by PCR, thereby enabling diagnosis. However, the sensitivity of this method remains debated. According to a meta-analysis, PCR testing on dried blood spots (DBS) showed high specificity but variable sensitivity, with pooled estimates ranging from 84.4% to 99.9% [61]. Therefore, although a positive DBS PCR result is highly reliable for confirming congenital CMV infection, a negative result does not exclude the diagnosis. This limitation is mainly due to the lower viral load and smaller blood volume available in DBS samples compared with urine or saliva, which reduces the likelihood of detecting CMV DNA and makes DBS PCR less sensitive than testing performed on fresh neonatal urine or saliva collected within the first 3 weeks of life.

About 30% of infected neonates do not exhibit IgM at diagnosis, and IgM testing in neonates has a low sensitivity, making it unsuitable for diagnosing congenital CMV in neonates [62].

One of the major risk factors for long-term sequelae is MPI during the first trimester. A clinical virologist should assess serological findings to determine the timing of MPI as accurately as possible. The effects of MPI and maternal non-primary infection on long-term infant outcomes appear to be broadly comparable [45]. The clinical examination, laboratory testing, and specialized assessments required to determine whether congenital infection is symptomatic or asymptomatic are summarized in Table 3. Some of these findings at birth have variable prognostic value and should therefore be carefully considered when decisions about treatment and follow-up are made. Current management is increasingly focused on risk stratification and targeted follow-up. Children with CMV and confirmed transmission during the first trimester, or with unknown timing of transmission, should be followed for at least 6 years to ensure appropriate specialized management. By contrast, such prolonged follow-up may not be necessary in children born after documented MPI during the second or third trimester of pregnancy [2].

Microcephaly has high specificity for poor neurological outcome when compared with birth weight (i.e., HC z score − weight z score < −2). By contrast, in newborns without other symptoms, symmetric IUGR may not necessarily predict an adverse outcome [44].

When a newborn shows clinical signs of CMV infection at birth, sensorineural hearing loss, chorioretinitis, or cranial ultrasound abnormalities, brain MRI is recommended. MRI is also indicated in cases of maternal primary infection during the first trimester or when the timing of maternal infection is unknown. Normal neuroimaging predicts normal or near-normal neurodevelopmental outcomes, whereas significant abnormalities are associated with a poorer prognosis [63,64,65].

According to the recent European consensus (2024), antiviral therapy is recommended for symptomatic congenital CMV infection, particularly in cases with central nervous system involvement, microcephaly, or chorioretinitis (for 6 months), and in selected cases of isolated hepatitis or thrombocytopenia (for 6 weeks). By contrast, antiviral therapy is not recommended for asymptomatic infants or those with isolated intrauterine growth restriction [2].

Oral valganciclovir (16 mg/kg twice daily) is the treatment of choice, whereas intravenous ganciclovir is reserved for selected clinical situations. Therapy should be started as early as possible, ideally within the first month of life, and may be considered up to 12 weeks of age in selected cases [66].

Evidence regarding hearing outcomes is mixed: although early treatment may reduce hearing deterioration in some infants, initiation beyond the neonatal period does not appear to improve hearing, and prolonged therapy offers only modest long-term benefit. Careful monitoring is required because of potential haematological and hepatic adverse effects [66,67].

Although vaccine candidates aimed at reducing the risk of congenital CMV are advancing through clinical trials, it is likely to be several years before a licensed product becomes available. Until then, hygiene-based behavioral measures remain the only strategy currently available for primary prevention. The complex life cycle of CMV, its ability to evade host immunity, and the lack of clearly defined immune correlates of protection make vaccine development challenging [68].

Several vaccine platforms have been investigated, including live-attenuated, protein subunit, viral vector, DNA, and, more recently, mRNA vaccines, targeting key antigens such as glycoprotein B (gB) and the pentameric complex, as well as T-cell targets like pp65 [69]. Despite extensive research, several candidates have failed to demonstrate clinical efficacy in phase III trials [70]. Moderna® (Moderna, Inc., Cambridge, MA, USA) has developed an mRNA-1647 vaccine against CMV, but in October 2025, Moderna® announced that the Phase 3 mRNA-1647 vaccine candidate failed to meet its primary efficacy endpoint and its development was stopped [71].

Protective immunity is thought to involve both neutralizing antibodies against glycoprotein B and the pentameric complex, as well as cellular immune responses [69]. Vaccination of young children may provide indirect protection by reducing transmission to pregnant women [72].

However, the immunological deficits that allow reinfection in seropositive individuals, including pregnant women, remain unclear, and studies are urgently needed to define these deficits [73,74].