Birth Prevalence of Congenital Cytomegalovirus Infection in HIV-Exposed Uninfected Children in the Era of Combination Antiretroviral Therapy

Abstract

Objectives: To estimate birth prevalence of congenital cytomegalovirus (cCMV) in HIV-exposed uninfected children born in the current era of combination antiretroviral therapy and describe cCMV-related neurodevelopmental and hearing outcomes.

Study design: The Surveillance Monitoring for ART Toxicities cohort study follows HIV-exposed uninfected children at 22 sites in the US and Puerto Rico. Birth cCMV prevalence was estimated in a subset of participants who had blood pellets collected within three weeks of birth and underwent ≥1 of 6 assessments evaluating cognitive and language development including an audiologic examination between 1 and 5 years of age. Detection of CMV DNA by polymerase chain reaction testing of peripheral blood mononuclear cells was used to diagnose cCMV. Proportions of suboptimal assessment scores were compared by cCMV status using Fisher exact test.

Results: Mothers of 895 eligible HIV-exposed uninfected children delivered between 2007 and 2015. Most (90%) were on combination antiretroviral therapy, 88% had an HIV viral load of ≤400 copies/mL, and 93% had CD4 cell counts of ≥200 cells/μL. Eight infants were diagnosed with cCMV, yielding an estimated prevalence of 0.89% (95% CI, 0.39%-1.75%). After adjusting for a sensitivity of 70%-75% for the testing method, projected prevalence was 1.2%-1.3%. No differences were observed in cognitive, language and hearing assessments by cCMV status.

Conclusions: Although birth cCMV prevalence in HIV-exposed uninfected children born to women with well-controlled HIV is trending down compared with earlier combination antiretroviral therapy-era estimates, it is above the 0.4% reported for the general US population. HIV-exposed uninfected children remain at increased risk for cCMV.

Keywords: HIV-exposed uninfected; cART; cCMV; congenital; cytomegalovirus; prevalence; women living with HIV.

Figure 1.

Flow chart describing the identification of cCMV infection in HIV-exposed uninfected children followed in the Surveillance Monitoring of Antiretroviral Therapy and Toxicity Study. ^aSurveillance Monitoring for Antiretroviral Therapy Toxicities cohort.

Figure 2.

Comparison of percent with lower and normal language, cognitive and audiologic scores by cCMV status for the cohort of 895 HIV-exposed uninfected children. The cutoff for a classification of “lower” on each assessment was: MacArthur-Bates Communicative Development Inventory: age-adjusted percentile score ≤10th percentile in any of the 4 domains (phrases understood, vocabulary comprehension, word production, or total gestures); Bayley-III: composite score ≤85 in any of the 5 domains (cognitive, language, motor, social-emotional, or adaptive behavior); Ages and Stages: total score ≥1 SD below age specific norms; Wechsler Preschool and Primary Scale of Intelligence: composite score ≤85 in any of the 3 quotients (verbal IQ, performance IQ, full scale IQ); Test of Language Development: Spoken Language Quotient £85; and audiologic evaluation: sensorineural hearing loss with worse ear pure tone average >25 dB hearing level at 0.5, 1.0, 2.0, and 4.0 kHz. ^aComparison by cCMV status was not significantly different for any of the neurodevelopment, language, and audiology assessments. The highest proportion classified as lower in these assessments was for the Bayley-III where 50% of cCMV negative children met cutoff criteria in ≥1 of 5 domains and are believed to be generally representative of children who are HIV exposed, and uninfected. ^bWithin normal limits. ^cFisher exact test. ^dSensorineural hearing loss.