Objective CMV urine load has not been studied previously in infants with postnatally acquired CMV infection. We compared CMV urine loads of postnatally infected infants and congenitally infected infants.

Methods Infants treated in our NICU (2003–2007) were included. Congenitally infected infants were diagnosed at birth, postnatally infected preterm infants at term. In the latter, congenital infection was excluded by examination of urine collected shortly after birth or of Guthrie cards. CMV infection was diagnosed by real-time PCR in urine. Differences between CMV urine loads were analyzed by Mann Whitney test. Statistical significance was assumed for p<0.05.

Results Ten infants with congenital CMV (group I) and 25 infants with postnatal CMV infection (group II) were included. Nine of 10 group I infants were symptomatic including cerebral abnormalities. Twenty-three of postnatally infected infants were asymptomatic; 2 developed pneumonitis. In 14/25 postnatally infected infants calcifications and/or germinolytic cysts were identified at term (subgroup IIA) on cerebral ultrasonography (CUS) and 11/25 had normal CUS (subgroup IIB). The median CMV urine load in group I was 2.49×10̂6 copies/ml, group II 1.66×10̂5 copies/ml (subgroup IIA: 6.52×10̂5 copies/ml, subgroup IIB: 7.2×10̂4 copies/ml). CMV urine load was significantly higher in group I than in group II (p = 0.001), subgroup IIA (p = 0.026) and subgroup IIB (p = 0.0001). Subgroups IIA and IIB did not differ significantly (p = 0.22).

Conclusions CMV urine load is significantly lower in infants with postnatally acquired CMV infection (irrespective of cerebral abnormalities) than in infants with congenital CMV infection and seems to correlate with disease severity.