Progressive and fluctuating sensorineural hearing loss in children with asymptomatic congenital cytomegalovirus infection Karen B. Fowler, DrPH, Faye P. McCollister, EdD, Arthur J. Dahle, PhD, Suresh B o p p a n a , MD, William J. Britt, MD, a n d Robert F. Pass, MD From the Department of Pediatrics, University of Alabama at Birmingham School of Med- icine, the Department of Special Education, University of Alabama College of Education, Tuscaloosa, and the Department of Biocommunication, University of Alabama at Birming- ham Objective: To determine the prevalence and temporal changes of sensorineural hearing loss (SNHL) among children with clinically inapparent (asymptomatic) congenital cytomegalovirus (CMV) infection identified from a cohort of newborn infants screened for congenital CMV infection. Methods: The study population consisted of 307 children with documented asymptomatic congenital CMV infection, 76 uninfected siblings of children with asymptomatic congenital CMV infection, and 201 children whose neonatal screen for congenital CMV infection showed negative results. Audiologic evalu- ations were completed for all children to determine their hearing status. Results: SNHL occurred only in children with congenital CMV infection. Of the chil- dren with asymptomatic congenital CMV infection, 22 (7.2% 95% confidence in- terval, 4.5% to 10.6%) had SNHL. Among the children with hearing loss, further de- terioration of hearing occurred in 50.0%, with the median age at first progression at 18 months (range, 2 to 70 months). Delayed-onset SNHL was observed in 18.2% of the children, with the median age of detection at 27 months (range, 25 to 62 months). Fluctuating SNHL was documented in 22.7% of the children with hearing loss. Conclusions: Asymptomatic congenital CMV infection is likely a leading cause of SNHL in young children. The continued deterioration of hearing and delayed on- set of SNHL in these children emphasizes the need for continued monitoring of their hearing status. (J Pediatr 1997 130:624-30) Supported in part by a research grant (5 P01 HD 10699) from the National Institute of Child Health and Human Development, a re- search grant (5 R01 DC 02139) from the National Institute on Deafness and Other Communication Disorders, and a research grant (5 M01 RR 00032) from the General Clinical Research Center, Na- tional Institutes of Health, the Deafness Research Foundation, and the Civitan International Research Center. Submitted for publication June 20, 1996 accepted Sept. 27, 1996. Reprint requests: Karen B. Fowler, DrPH, Department of Pediatrics, University of Alabama at Birmingham, 1600 7th Ave. South, Suite 752, Birmingham, AL 35233. Copyright �� 1997 by Mosby-Year Book, Inc. 0022-3476/97/$5.00 + 0 9/21/78377 Cytomegalovirus is the leading cause of congenital infection in the United States. An estimated 1% of newborn infants are infected with CMV prenatally, a rate that equates to around 40,000 new cases each year. 1 The vast majority, 90% to 95% ABR Auditory brain-stem response CI Confidence interval CMV Cytomegalovims SNHL Sensorineural hearing loss of these infected infants, will have no evidence of disease in the newborn period. However, infants with clinically inap- parent (asymptomatic) congenital C M V infection may later have central nervous system and sensory impairments, with 6 2 4

The Journal of Pediatrics Fowler et al. 6 2 5 Volume 130, Number 4 T a b l e I. Characteristics of children with congenital CMV infection, their uninfected siblings, and control children without congenital CMV infection Children with congenital CMV infection (n = 307) Siblings of children with congenital CMV infection (n = 76) Children without congenital CMV infection (n = 20 I) Characteristics No. (%) No. (%) No. (%) Race* Black 233 White 74 Other 0 Sex Female 149 Male 158 Insurance status* Private insurance 67 Medicaid or no insurance 239 Unknown 1 Prenatal care* Private physicians 61 Health department 242 None/unknown 4 Weeks' gestation at delivery 37 Completed wk 51 -37 Completed wk 256 Sensorineural hearing loss* 22 (75.9) 56 (73.7) 177 (88.1) (24.1) 20 (26.3) 23 (ll.4) 0 1 (0.5) (48.5) 43 (56.6) 111 (55.2) (51.5) 33 (43.4) 90 (44.8) (21.8) 16 (21.1) 8 (4.0) (77.9) 60 (78.9) 193 (96.0) (0.3) 0 0 (19.9) 18 (23.7) 5 (2.5) (78.8) 55 (72.4) 196 (97.5) (1.3) 3 (3.9) 0 (16.6) 7 (9.2) 25 (12.4) (83.4) 69 (90.8) 176 (87.6) (7.2) 0 0 *p 0.05. sensorineural heating loss being the most commonly ob- served deficit. 2-11 Congenital CMV infection is an important cause of audi- tory system damage however, information regarding audi- ologic findings in children with asymptomatic congenital CMV infection is fimited. To study this hearing loss, we must identify children with infection by virologic tests in the newborn period and continue to monitor them with audio- logic evaluations. Previous studies identified children with asymptomatic congenital CMV infection and reported rates and severity of hearing loss in these children. 3-5' 10, 12, 13 However, the rates and severity of hearing loss in these chil- dren have varied considerably, reflecting differences in methods of patient selection, types of audiologic evaluation, and small study populations with SNHL. 3-5' 10, 12, 13 Ill addi- tion, many of these studies did not evaluate hearing loss in young infants because improved audiologic methods to al- low more precise documentation of auditory function dur- ing early infancy were unavailable. 3-5, 12, 13 To define the prevalence of hearing impairment and temporal changes in audiologic function in children with asymptomatic con- genital CMV infection, we performed serial audiologic examinations o f 307 infected children identified from a cohort screened as neonates for congenital CMV infec- tion. Their audiologic outcomes were compared with those of 201 randomly selected children who had cultures nega- five for CMV in the newborn period and 76 uninfected sib- lings of children with asymptomatic congenital CMV infec- tion. METHODS Study population. The study population consisted of children born between 1980 and December 1995, who were identified by newborn screening for congenital CMV infec- tion at two hospitals in Birmingham, AlaJ 4, 15 Congenital CMV infection was identified by isolation of virus in urine or saliva in the first or second week of life. 16-18 These chil- dren were assessed and their medical records were system- aticaUy reviewed by study personnel to determine whether any of the following symptoms were observed in the new- born period: microcephaly, thrombocytopenia, petechiae, hepatosplenomegaly, or jaundice with conjugated hyperbil- irubinemia. Four hundred twenty-four children met the cri- teria of no clinically apparent disease (asymptomatic) in the newborn period. No child received antiviral or antiinfiam- matory therapy. The children were followed up in a special clinic that provides serial andiologic, psychometric, vision, and medical evaluations. To document progression or fluc- tuation or both of hearing loss, we included in this study only 307 children with two or more audiologic evaluations. These children did not differ with respect to race, gender, gesta- tional age, birth weight, insurance status, prenatal care loca-