Cytomegalovirus DNA Detection by Polymerase Chain Reaction in Cerebrospinal Fluid of Infants With Congenital Infection: Associations With Clinical Evaluation at Birth and Implications for Follow-up.

BACKGROUND: DNA detection of human cytomegalovirus (hCMV) in cerebrospinal fluid (CSF) by polymerase chain reaction (PCR) is a marker of central nervous system (CNS) involvement in congenital hCMV infection (cCMV), but its prognostic value is unknown. METHODS: A multicenter, retrospective study was performed using the Spanish Congenital Cytomegalovirus Infection Database (REDICCMV; http://www.cmvcongenito.es). Newborns with cCMV and a lumbar puncture performed were included and classified according to their hCMV-PCR in CSF result (positive/negative). Clinical characteristics, neuroimaging abnormalities, plasma viral load, and audiological and neurological outcomes of both groups were compared. RESULTS: A total of 136 neonates were included in the study: 21 (15.4%) with positive CSF hCMV-PCR and 115 (84.6%) with negative results. Seventeen patients (81%) in the positive group were symptomatic at birth compared with 52.2% of infants in the negative group (odds ratio [OR], 3.86; 95% confidence interval [CI], 1.28-14.1; P = .01). Only 4 asymptomatic newborns (6.8%) had a positive CSF hCMV-PCR. There were no differences between groups regarding the rate of microcephaly, neuroimaging abnormalities, neurological sequelae at 6 months of age, or plasma viral load. Sensorineural hearing loss (SNHL) at birth was associated with a positive CSF hCMV-PCR result (OR, 3.49; 95% CI, 1.08-11.27; P = .04), although no association was found at 6 months of age. CONCLUSIONS: A positive hCMV-PCR result in CSF is associated with symptomatic cCMV and SNHL at birth. However, no differences in neuroimaging studies, plasma viral load, or outcomes at 6 months were found. These results suggest that hCMV-PCR in CSF may not be a useful prognostic marker in cCMV.

[Can we rule out a congenital cytomegalovirus infection when the result of polymerase chain reaction in dried blood spots is negative?]. / ¿Podemos descartar infección congénita por citomegalovirus cuando la reacción en cadena de la polimerasa viral es negativa en la prueba del talón?

INTRODUCTION: The detection of cytomegalovirus (CMV) DNA by real time polymerase chain reaction (rt-PCR) in dried blood spots collected routinely for metabolic screening has been assessed for the retrospective diagnosis of congenital CMV (cCMV) infection in many studies, but not in Spain. The aim of this study is to analyze the diagnostic accuracy of this technique in our hospital. METHODS: A cross-sectional retrospective observational study was conducted including all patients born between January, 2007 and September, 2012 with confirmed cCMV infection. The assessment of CMV DNA was made by using rt-PCR in dried blood spots of these patients. RESULTS: Fourteen patients were included: 4/14 were symptomatic and 4/14 had sequelae. The detection of CMV DNA by rt-PCR was positive in only 7 patients. A statistically significant relationship between low viral load at birth and negative rt-PCR in dried blood spots was demonstrated. CONCLUSIONS: Despite the low number of patients included, our data highlight an important amount of false negative results in the DNA CMV detection by rt-PCR in these samples for the retrospective diagnosis of cCMV infection, especially in cases with low viral load at birth.

Is Polymerase Chain Reaction in Neonatal Dried Blood Spots Reliable for the Diagnosis of Congenital Cytomegalovirus Infection?

BACKGROUND: Detection of cytomegalovirus (CMV) DNA by real-time polymerase chain reaction (rt-PCR) in dried blood spots (DBSs) collected for newborn screening has been assessed for retrospective diagnosis of congenital CMV (cCMV) infection, with variable results (sensitivities ranging from 34% to 100%). We aimed to assess the accuracy of this technique in Spain in a large patient series. METHODS: Ambispective, multicenter study including patients with confirmed cCMV from the Spanish Registry of cCMV patients. cCMV was established on the presence of CMV DNA in any body fluid, by positive culture findings or by molecular techniques during the first 2 weeks of life. Children in whom cCMV had been excluded were used as negative controls. Neonatal DBS samples were collected from both groups. The presence of CMV DNA was assessed by rt-PCR (RealStar CMV, Altona, Germany) in a central laboratory. RESULTS: One-hundred three patients and 81 controls from 10 hospitals were included. The performance of CMV DNA determination in DBS for the diagnosis of cCMV was as follows (95% confidence interval): sensitivity 0.56 (0.47-0.65), specificity 0.98 (0.91-0.99), positive likelihood ratio 22.81 (5.74-90.58) and negative likelihood ratio 0.45 (0.36-0.56). Sensitivity increased with the birth viral load (bVL) log category. In cCMV patients, lower bVL was the single variable associated with a negative DBS rt-PCR result (P = 0.017). CONCLUSIONS: The sensitivity of CMV rt-PCR in DBS in our series was low and correlated with the bVL. Thus, a negative DBS result would not rule out cCMV infection, especially in patients with a low viremia level at birth.

Growth Patterns in Children With Congenital Cytomegalovirus Infection.

BACKGROUND: Congenital cytomegalovirus infection (CMVc) affects 0.7%-6% of recent births. Among its clinical manifestations are low weight and length at birth. OBJECTIVE: Describe the growth patterns of children with CMVc in their early years. METHODS: Observational, multicenter study of patients with CMVc. Anthropometric data were collected during the first 2 years of life and compared with World Health Organization standards. RESULTS: Anthropometric characteristics of 383 children with CMVc were studied, of which 198 (51%) were symptomatic at birth. At birth, 9% were small for gestational age (SGA) in terms of their weight and length and 17% had microcephaly. At 24 ± 3 months, 10% had a weight and length ≤2 SD, and 13% a head circumference ≤2 SD. Of those who were SGA at birth, at 24 ± 3 months >20% remained at ≤2 SD of their weight and length. Conversely, 75% of children with low weight or length at 24 ± 3 had not been SGA at birth. 20% of infants with microcephaly at birth remained with microcephaly, and 10% of those without microcephaly developed it at 24 ± 3 months. The average growth rate in length and weight was normal. Patients who were symptomatic at birth, premature and with motor and neurocognitive impairment had a significantly higher risk of low weight and length at 24 ± 3 months. CONCLUSION: Around 10% of children with CMVc are at ≤2 SD in weight, length and head circumference at 24 ± 3 months. The lack of adequate growth is associated with symptoms at birth, prematurity and motor and neurocognitive impairment. Growth impairment could be incorporated into the symptomatic spectrum of CMVc.