Cytomegalovirus (CMV) is one of the most frequently transmitted intrauterine infections, detectable in an estimated 0.64%-0.70% of live births worldwide (Kenneson & Cannon, 2007). Most women in the United States know little about CMV infections and the impact the virus may have on their unborn infants (Jiyeon et al., 2006; D. S. Ross et al., in press). However, congenital CMV infection is as common a cause of serious developmental disability as are Down syndrome and neural tube defects (Cannon & Davis, 2005).

Sensorineural hearing loss is the most common sequela following congenital CMV infection (Dahle, et al., 2000; Fowler & Boppana, 2006; S. Ross et al., 2006). CMV is estimated to be the leading environmental cause of childhood hearing loss, accounting for approximately 15%-21% of all hearing loss at birth in the United States (Morton & Nance, 2006; Grosse et al., 2008). In addition, CMV-related hearing losses may also be progressive or late-onset, requiring more frequent audiological monitoring of infants and young children who have been diagnosed with congenital CMV infection.

What is CMV?

CMV belongs to the herpes virus family, and although not easily spread from person to person, CMV infection can occur following contact with the virus through saliva, blood, urine, breast milk, cervical secretions, or semen. CMV exposures increase with age; in the U.S. population, approximately 36% of children 6-11 years of age are CMV seropositive, with the seroprevalence increasing to 91% in adults who are 80 years of age and older (Staras et al., 2006). Most healthy children and adults infected with CMV will have no symptoms or might develop only a mild mononucleosis-like illness (Britt & Alford, 1996; Stagno et al., 2001).

Once a person is infected, the virus enters into a latent state (Stagno et al., 2001). Although the virus can reactivate, it is usually kept under immune control, so adults and children with normal immune function rarely exhibit CMV-related disease (Stagno et al., 2001). However, when CMV is transmitted in utero from the mother to the fetus, damage may occur and the infant may have acute symptoms at birth, permanent disabilities, or both (Stagno et al., 1986; Fowler et al., 1992; Boppana et al., 2001; Fowler et al., 2003).

In the United States, where newborns are not screened for CMV, it is possible to identify children with hearing loss caused by CMV only by obtaining their newborn blood spot for analysis. However, not all states store blood spot cards long enough for this kind of retrospective analysis. Knowledge that the mother was exposed during pregnancy, along with subsequent testing of the newborn, is another way in which children with hearing loss caused by CMV can be identified.

Hearing Loss and Congenital CMV

Hearing loss is the most common sequela of congenital CMV infection, occurring in 10%-15% of all infected children (Dahle et al., 2000; S. Ross et al., 2006; Fowler & Boppana, 2006; Dollard et al., 2007). Hearing loss reportedly occurs in 30%-40% of
children who have clinically apparent disease (symptomatic) at birth and in 5%-10% of children with clinically silent (asymptomatic) infections.

CMV-related hearing losses do not have a unique audiometric configuration and the loss may be present at birth or occur in the first years of life. Hearing loss from congenital CMV infection can be either unilateral or bilateral and varies from mild to profound. In addition, CMV-related hearing loss may be fluctuating and/or progressive. In fact, approximately half of the cases of hearing loss due to congenital CMV infection are late-onset and/or progressive and, therefore, will not be detected at birth through newborn hearing screening (Fowler et al., 1999).

Dollard et al. (2007) report a weighted average of sensorineural hearing loss (SNHL) of 14.1% due to congenital CMV infection. Fowler and Boppana (2006) reported hearing loss in 8.7% of children with asymptomatic infections and in 41% with symptomatic infections. The most complete data on hearing loss among children with congenital CMV infection come from studies conducted at the University of Alabama at Birmingham (UAB). Fowler et al. (1999) reported that 8% of all CMV-infected children identified with SNHL had a threshold of 30 dB or greater at 6 years of age, including 11% of children who were asymptomatic at birth and 36% of children who were symptomatic at birth (Fowler et al., 1999). Given that 59% of the children diagnosed with SNHL at 6 years of age had not been identified by 12 months of age, frequent and routine audiological follow-up is needed to identify children with CMV-related hearing loss (Fowler et al., 1999).

At least two studies have used retrospective newborn assessments among children with diagnosed SNHL to determine the fraction of cases of hearing loss attributable to congenital CMV infection. First, Barbi et al. (2003) reported congenital CMV infection in 17% of children with SNHL greater than 40 dB. This group included 10% of children with hearing loss detected soon after birth and 30% of children with hearing loss of unknown etiology who were diagnosed more than three months after birth.

The most direct and up-to-date estimate of the fraction of SNHL attributable to congenital CMV comes from a recent study from Japan. Ogawa et al. (2007) analyzed stored dried umbilical cord samples from 67 children diagnosed with severe SNHL; 15% were found to have congenital CMV infection. Among 55 children who had moderately severe to profound bilateral SNHL (>55 dB), 16% had congenital CMV infection. Among the subset of 36 children with profound bilateral SNHL (>90 dB), 22% were identified with congenital CMV infection.

Pathogenesis

The pathogenesis of hearing loss caused by congenital CMV infection is not well understood; however, animal models and studies of temporal bone structures from infants with severe congenital infection suggest some hypotheses (Schleiss & Choo, 2006). Studies of human temporal bones along with studies of animal models indicate that CMV (or a CMV antigen) may be present in the epithelium and the neural cells in the inner ear. CMV may cause damage to the inner ear by virus-mediated damage to the neural cells and/or secondary to host-derived inflammatory responses to CMV in the ear resulting in injury to the auditory apparatus and subsequent hearing loss. However, these explanations do not fully explain the progressive and late-onset nature of CMV-related hearing losses.

Treatment

Infants and children with congenital CMV infection need early and more frequent audiological evaluations throughout childhood [Joint Committee on Infant Hearing (JCIH), 2007]. The JCIH position statement recommends that the timing and number of hearing evaluations be customized or individualized for infants with congenital CMV infection. Previous recommendations have called for audiological evaluations every six months; however, at times when hearing loss appears to be changing, audiological evaluations may be needed every three months to assess and document the changes in the hearing status of the child infected with CMV.

Because this hearing loss may be progressive in nature, concerns of parents or primary health care providers that the child’s hearing status has changed should be followed up quickly through appropriate audiological evaluations for the child’s developmental age. Immediate evaluation and treatment of the child’s hearing loss is important for maximum opportunity for improved hearing. Because CMV-related hearing loss may be fluctuating and/or progressive, audiologists should consider providing training in communication methods that accommodate changing hearing loss and other options such as hearing aids with power- and frequency-response flexibility that accommodate changes in hearing.

Prevention

Because no CMV vaccine exists (Plotkin, 2002; Arvin et al., 2004; Schleiss & Heineman, 2005), alternative measures are needed to prevent congenital CMV infection. One way to prevent primary infection before and during pregnancy is through simple hygienic practices such as hand-washing (Cannon & Davis, 2005). Humans are the only reservoirs for the human herpes viruses and can transmit these agents through infected blood, tissues, bodily fluids, feces, and fomites (Stagno et al., 2001).

Women are infected with CMV most frequently through sexual activity or close contact with young children, who tend to secrete the virus in their saliva and urine for many months following their first—usually asymptomatic—infection (Dworsky et al., 1983; Yeager, 1983; Taber et al., 1985; Yow et al., 1987; Pass et al., 1990; Fowler & Pass, 2006). Day care providers and mothers of toddlers have an especially high risk of infection (Pass et al., 1984; Pass & Kinney, 1985; Pass & Hutto, 1986; Pass et al., 1986; Pass et al., 1987; Pass, 1990; Adler, 1991). Thus, exposure of the fetus to CMV is frequently the result of transmission from an infected toddler to a pregnant or soon-to-be pregnant mother. It is impossible, however, to separate CMV transmission by the oral route from the venereal route; unprotected sexual contact should be avoided and careful attention given to hygiene.

It is important to provide women of child-bearing age with information about preventing CMV infection before and after conception. Most women in the United States know very little, if anything, about congenital CMV (Jiyeon et al., 2006; D. S. Ross, et al., in press) and are therefore unlikely to take preventive measures during pregnancy. Female audiologists and speech-language pathologists (especially those who are pregnant or planning a pregnancy and who work closely with children) should practice universal precautions to reduce the risk of congenital CMV infection.

The most effective means for preventing infection is hand-washing (Adler et al., 1996; American College of Obstetricians and Gynecologists, 2002; Cannon & Davis, 2005; Finney et al., 1993; Onorato, et al., 1985; Stagno et al., 2001). More than 100 years of evidence demonstrates conclusively that hand-washing reduces risk of infection for a wide range of pathogens (Larson, 1988). An educational approach to motivate and inform women about careful hygiene and frequent hand-washing, especially after contact with the saliva or urine of young children, and carefully disposing of diapers, tissues, and other contaminated items, could substantially reduce the transmission of CMV. Avoiding the sharing of drinking glasses and eating utensils with young children as well as avoiding kissing young children on the mouth may also prevent transmission (Onorato et al., 1985; Finney et al., 1993; Stagno et al., 2001; American College of Obstetricians and Gynecologists, 2002; Adler, et al., 2004; Cannon & Davis, 2005).

Early and Frequent Assessments

Congenital CMV infection continues to be a cause of childhood hearing loss, and is the leading environmental cause of hearing loss among children in the United States. Congenital CMV infection is often underestimated because infants are not routinely screened for CMV infection at birth. To identify infants with congenital CMV infection conclusively, infants’ saliva or urine should be tested for CMV by standard cell-culture methods during the first three weeks of life. Retrospective identification of CMV by DNA polymerase chain reaction methods, using dried-blood spots collected from all newborns at birth, is also a method for identifying congenital CMV infection later, although the sensitivity and specificity for this method has not yet been established.

Because CMV-related hearing loss may be late-onset and/or progressive, infants with congenital CMV infection should have early and more frequent audiological evaluations to assess possible changes in their hearing status so that appropriate interventions may be implemented to minimize the impact of their hearing loss.

The findings and conclusions in this report are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.