Recent studies in Africa have shown an increased prevalence of congenital CMV infection among HIV-exposed infants compared with HIV-unexposed infants, despite near universal maternal use of highly active antiretroviral therapy. [1,2] However, data on the burden of congenital CMV-related sequelae among HIV-exposed and -unexposed children in Africa are limited. In this letter, we report updated estimates of prevalence by HIV-exposure status and clinical outcomes of CMV-positive children identified by screening 1078 newborns in western Kenya during 2015–2017. [1]

Among a total 39 CMV-positive children, 14 (35.9%) were HIV-exposed, 2 more than previously reported, because the mother’s infection–though diagnosed at delivery–was only documented during follow-up visits. The updated prevalence of congenital CMV infection among HIV-exposed vs. HIV-unexposed children was 6.5% (14/216) vs. 2.9% (25/862), with a prevalence ratio of 2.7 (95% confidence interval = 1.3–5.5; p<0.01), after adjusting for maternal age and parity, and excluding second-born twins to limit the effect of shared maternal risk factors by twins.

Per protocol, CMV-positive children would have two follow-up visits, at 4 and 24 months of age. A paediatric neurologist interviewed the mother and observed the child to assess gross motor, fine motor, speech, language comprehension, and social/cognitive development using the Malawi Developmental Assessment Tool. [3] Additionally, the neurologist performed age-appropriate physical and neurological examinations. Hearing was evaluated by observing whether the child turned the head to the sound of crumpling paper while being distracted. Vision was evaluated by observing whether the child was following objects at ~1-meter distance and, if older than 10 months, whether the child could localize and grasp small objects. Data on infant HIV status were collected from medical records.

Among 36 (92.3%) CMV-positive children followed, the first visit was at median age 11 (range: 5–30) months; 24 (66.7%) children had a second visit at median age 23 (21–29) months. Thirty-four (94.4%) children had normal neurodevelopmental, hearing and vision evaluations. Two (5.6%) children were diagnosed with a speech delay at approximately 2 years of age, despite normal development at the first follow-up visit. Child A was HIV-unexposed and had a history of possible seizures with a normal EEG and Child B was HIV-exposed with no other abnormalities. Among 14 HIV-exposed children, 9 (64.3%) were HIV-negative by PCR and/or antibody test at age ≥18 months, thus HIV infection was ruled out. Among 5 (35.7%) children for whom HIV infection could not be ruled out, two (14.3%) were HIV-negative by PCR at 6 weeks and one at 11 months, and 2 children (including Child B) were not tested.

In conclusion, most CMV-positive children followed had normal neurodevelopment when assessed at roughly 1 or 2 years of age. However, we did not follow CMV-negative infants for comparison. Additionally, we may have missed some children with neurodevelopmental delays because the follow-up was short or missed, and we did not perform formal vision or hearing assessment. Further studies assessing the burden of congenital CMV-related sequelae in low-middle income countries, particularly in regions with high maternal HIV prevalence, are warranted. [4]