Programmatic Congenital CMV Universal Screening Program

BACKGROUND: CMV is the most common congenital infection (cCMV). Traditional identification strategies including hearing screen and physical exam are insensitive and miss affected infants. To improve identification of infected newborns, we established a universal, institutional cCMV newborn screening program. METHODS: All newborns born or transferred to nurseries in a hospital system in Memphis, Tennessee between March 2016 and April 2017 were screened for cCMV. Infant saliva was collected on a Copan swab prior to discharge and within 2 weeks of birth. Specimens were centrally processed using a real-time CMV PCR assay (Simplexa™ CMV) (DiaSorin, Cypress CA) amplifying the UL83 gene, and the 3M Integrated Cycler. Parents received educational materials on cCMV testing and natural history prior to specimen collection. All patients with a positive screen had a full evaluation including physical exam, eye exam, hearing testing, CBC, chemistries and head ultrasound (HUS). RESULTS: There were 35/6,114 (0.6%) positive screens. Of 35, 16 (45.7%) were male and 5 (14%) were less than 37 weeks gestation. Thirty-one of 35 saliva specimens were collected on day 0 or 1 of life. All patients were evaluated by an infectious disease specialist at a median of 15 days of age. Confirmatory urine PCR was positive in 25/33 (76%) tested. Overall, 11/25 (44%) with confirmed congenital CMV were symptomatic. This included 28% with microcephaly and 20% with low birth weight. Six (24%) failed newborn hearing screening of one or both ears. Other abnormalities included thrombocytopenia (5%), elevated ALT (10%), elevated direct bilirubin (5%), and abnormal HUS (11/25, 44%), of which 7/11 had lenticulostriate vasculopathy and 2/11 had intracranial calcifications. Twelve infected infants had an eye examination and none had retinitis. Eleven infants were offered therapy and five were treated. Ten of 25 congenitally infected infants had audiology follow-up by 6 months with four abnormal. All infants were referred for early intervention. CONCLUSION: We have demonstrated the feasibility of implementing large-scale, saliva-based cCMV screening program within one hospital system. Universal screening detected twice as many infected infants than would have targeted screening based on newborn hearing screen and growth parameters. DISCLOSURES: J. P. Devincenzo, AstraZeneca/MedImmune: Investigator, Research support