Viral load may not predict hearing loss in children with congenital CMV

Cytomegalovirus DNA levels were no different among children with hearing deficits and those without, suggesting that viral burden had poor predictive value as a marker for hearing loss.

“We could not confirm the association between systemic virus burden during early infancy and sensorineural hearing loss [SNHL] in children with asymptomatic congenital CMV infection that was observed in our previous study,” the researchers wrote.

The researchers performed DNA analysis on 202 blood samples from 135 children who were born with CMV at three Birmingham, Ala. hospitals between January 1994 and February 2005, including 75 children from the previous report.

Viral load was measured in three different age groups: younger than 2 months, 2 to 12 months and 12 to 36 months. The presence of SNHL was determined following standard protocol with age-appropriate audiologic evaluations, and the researchers performed independent analyses for children with and without hearing loss.

Data indicated no association between viral load measurements and hearing loss in any of the three age groups:

A viral load measurement less than 3,500 genomic equivalents per mL (ge/mL) in the younger than 2 months group had a negative predictive value of 94.4%, with two of 36 asymptomatic children with this measurement experiencing hearing loss.

A positive predictive value of 7.9% was calculated for viral load measurements greater than 3,500 ge/mL among children in the younger than 2 month group.

None of the children in the other two age groups experienced SNHL when peripheral blood virus burden was measured at 3,500 ge/mL or less, resulting in a 100% negative predictive value.

Lower viral loads in asymptomatic children with CMV may be associated with a lower risk for hearing deficit, according to the researchers.

“Our current study includes the largest cohort of children with asymptomatic congenital CMV infection that has been examined for the association between virus burden and SNHL,” the researchers wrote.

Ross SA. Pediatr Infect Dis J. 2009; doi:10.1097/INF.0b013e3181979a27.

This is an important study that adds to our current understanding of viral load as a predictor of long-term outcomes in congenital CMV disease. These negative data demonstrate that clear-cut associations between viral DNA concentration and outcomes are not readily apparent. Although viral load assessments certainly should be incorporated into ongoing therapeutic and natural history studies of congenital CMV, they should not be used as a basis for clinical decision-making at this time.

– David W. Kimberlin, MD

Infectious Diseases in Children Editorial Board