Volume 57, Issue 4 p. 551-559
Systematic Review
Free Access

Outcome of fetuses with congenital cytomegalovirus infection and normal ultrasound at diagnosis: systematic review and meta-analysis

D. Buca

D. Buca

Centre for High Risk Pregnancy and Fetal Care, Department of Obstetrics and Gynecology, University of Chieti, Italy

D.B. and D.D.M. are joint first authors.Search for more papers by this author
D. Di Mascio

D. Di Mascio

Department of Maternal and Child Health and Urological Sciences, “Sapienza” University of Rome, Rome, Italy

D.B. and D.D.M. are joint first authors.Search for more papers by this author
G. Rizzo

G. Rizzo

Division of Maternal and Fetal Medicine, Ospedale Cristo Re, University of Rome Tor Vergata, Rome, Italy

Department of Obstetrics and Gynecology, The First I.M. Sechenov Moscow State Medical University, Moscow, Russia

Search for more papers by this author
A. Giancotti

A. Giancotti

Department of Maternal and Child Health and Urological Sciences, “Sapienza” University of Rome, Rome, Italy

Search for more papers by this author
A. D'Amico

A. D'Amico

Centre for High Risk Pregnancy and Fetal Care, Department of Obstetrics and Gynecology, University of Chieti, Italy

Search for more papers by this author
M. Leombroni

M. Leombroni

Centre for High Risk Pregnancy and Fetal Care, Department of Obstetrics and Gynecology, University of Chieti, Italy

Search for more papers by this author
A. Makatsarya

A. Makatsarya

Department of Obstetrics and Gynecology, The First I.M. Sechenov Moscow State Medical University, Moscow, Russia

Search for more papers by this author
A. Familiari

A. Familiari

Department of Clinical and Community Sciences, University of Milan, and Department of Woman Child and Neonate, Fondazione IRCCS Cà Granda Ospedale Maggiore Policlinico, Milan, Italy

Search for more papers by this author
M. Liberati

M. Liberati

Centre for High Risk Pregnancy and Fetal Care, Department of Obstetrics and Gynecology, University of Chieti, Italy

Search for more papers by this author
L. Nappi

L. Nappi

Department of Obstetrics and Gynecology, Department of Medical and Surgical Sciences, University of Foggia, Foggia, Italy

Search for more papers by this author
M. E. Flacco

M. E. Flacco

Department of Medical Sciences, University of Ferrara, Ferrara, Italy

Search for more papers by this author
L. Manzoli

L. Manzoli

Department of Medical Sciences, University of Ferrara, Ferrara, Italy

Search for more papers by this author
L. J. Salomon

L. J. Salomon

Hôpital Necker-Enfants Malades, Assistance Publique-Hôpitaux de Paris, Université Paris Descartes and Fetus & LUMIERE team, Paris, France

Search for more papers by this author
G. Scambia

G. Scambia

Department of Obstetrics and Gynecology, Catholic University of the Sacred Heart, Fondazione Policlinico A. Gemelli, Rome, Italy

Search for more papers by this author
F. D'Antonio

Corresponding Author

F. D'Antonio

Department of Obstetrics and Gynecology, Department of Medical and Surgical Sciences, University of Foggia, Foggia, Italy

Correspondence to: Dr F. D'Antonio, Fetal Medicine Unit, Department of Medical and Surgical Sciences, Department of Obstetrics and Gynecology, University of Foggia, Viale L. Pinto, 71100 Foggia, Italy (e-mail: [email protected])Search for more papers by this author
First published: 08 October 2020
Citations: 42

ABSTRACT

en

Objective

To report the outcome of fetuses with congenital cytomegalovirus (CMV) infection and normal ultrasound at the time of diagnosis, and to evaluate the rate of an additional anomaly detected only on magnetic resonance imaging (MRI).

Methods

Medline, EMBASE, CINAHL and Cochrane databases were searched for studies reporting on the outcome of fetuses with congenital CMV infection. Inclusion criteria were fetuses with confirmed CMV infection and normal ultrasound assessment at the time of the initial evaluation. The outcomes observed were an anomaly detected on a follow-up ultrasound scan, an anomaly detected on prenatal MRI but missed on ultrasound, an anomaly detected on postnatal assessment but missed prenatally, perinatal mortality, symptomatic infection at birth, neurodevelopmental outcome and hearing and visual deficits. Neurodevelopmental outcome was assessed only in cases of isolated CMV infection confirmed at birth. Subgroup analysis was performed according to the trimester in which maternal infection occurred. Random-effects meta-analysis of proportions was used to analyze the data.

Results

Twenty-six studies were included, comprising 2603 fetuses with congenital CMV infection, of which 1178 (45.3%) had normal ultrasound at the time of diagnosis and were included in the analysis. The overall rate of an associated central nervous system (CNS) anomaly detected on a follow-up ultrasound scan was 4.4% (95% CI, 1.4–8.8%) (32/523; 15 studies), while the rates of those detected exclusively on prenatal MRI or on postnatal imaging were 5.8% (95% CI, 1.9–11.5%) (19/357; 11 studies) and 3.2% (95% CI, 0.3–9.0%) (50/660; 17 studies), respectively. The rate of an associated extra-CNS anomaly detected on a follow-up ultrasound scan was 2.9% (95% CI, 0.8–6.3%) (19/523; 15 studies), while the rates of those detected exclusively on MRI or on postnatal imaging were 0% (95% CI, 0.0–1.7%) (0/357; 11 studies) and 0.9% (95% CI, 0.3–1.8%) (4/660; 17 studies), respectively. Intrauterine death and perinatal death each occurred in 0.7% (95% CI, 0.3–1.4%) (2/824; 23 studies) of cases. In cases without an associated anomaly detected pre- or postnatally, symptomatic infection was found in 1.5% (95% CI, 0.7–2.7%) (6/548; 19 studies) of infants, the overall rate of a neurodevelopmental anomaly was 3.1% (95% CI, 1.6–5.1%) (16/550; 19 studies), and hearing problems affected 6.5% (95% CI, 3.8–10.0%) (36/550; 19 studies) of children. Subanalyses according to the trimester in which maternal infection occurred were affected by the very small number of included cases and lack of comparison of the observed outcomes in the original studies. Compared with fetuses infected in the second or third trimester, those infected in the first trimester had a relatively higher risk of having an additional anomaly detected on follow-up ultrasound or MRI, abnormal neurodevelopmental outcome and hearing problems.

Conclusions

In fetuses with congenital CMV infection in which no anomalies are detected on prenatal ultrasound or MRI, the risk of adverse postnatal outcome is lower than that reported previously in the published literature when not considering the role of antenatal imaging assessment. The results from this review also highlight the potential role of MRI, even in fetuses with no anomalies detected on ultrasound, as an anomaly can be detected exclusively on MRI in about 6% of cases. The findings from this study could enhance prenatal counseling of pregnancies with congenital CMV infection with normal prenatal imaging. © 2020 International Society of Ultrasound in Obstetrics and Gynecology.

RESUMEN

es

Desenlace de los fetos con infección congénita por citomegalovirus y ecografía normal en el momento del diagnóstico: revisión sistemática y metaanálisis

Objetivo

Informar del desenlace de los fetos con infección congénita por citomegalovirus (CMV) y ecografía normal en el momento del diagnóstico, y evaluar la tasa de una anomalía adicional detectada sólo mediante imagen por resonancia magnética (IRM).

Métodos

Se realizaron búsquedas en las bases de datos Medline, EMBASE, CINAHL y Cochrane con respecto a estudios que informaran sobre el desenlace de los fetos con infección congénita por CMV. Los criterios de inclusión fueron fetos con infección por CMV confirmada y evaluación ecográfica normal en el momento de la evaluación inicial. Los desenlaces observados fueron una anomalía detectada en una ecografía de seguimiento, una anomalía detectada en la IRM prenatal pero que no se detectó en la ecografía, una anomalía detectada en la evaluación postnatal pero que no se detectó prenatalmente, mortalidad perinatal, infección sintomática al nacer, desenlaces relacionados con el desarrollo neurológico y deficiencias auditivas y visuales. El desenlace relacionado con el desarrollo neurológico se evaluó sólo en los casos de infección aislada por CMV confirmada en el momento del nacimiento. Se realizó un análisis de subgrupos según el trimestre en el que se produjo la infección materna. Para el análisis de los datos se utilizó un metaanálisis de efectos aleatorios.

Resultados

Se incluyeron 26 estudios, con un total de 2.603 fetos con infección congénita por CMV, de los cuales 1.178 (45,3%) mostraron una ecografía normal en el momento del diagnóstico y por tanto fueron incluidos en el análisis. La tasa global de una anomalía asociada con el sistema nervioso central (SNC) detectada en una ecografía de seguimiento fue del 4,4% (IC 95%, 1,4–8,8%) (32/523; 15 estudios), mientras que las tasas de aquellas detectadas exclusivamente en la IRM prenatal o en la imagen postnatal fueron del 5,8% (IC 95%, 1,9–11,5%) (19/357; 11 estudios) y del 3,2% (IC 95%, 0,3–9,0%) (50/660; 17 estudios), respectivamente. La tasa de una anomalía asociada del extra-sistema nervioso central (extra-SNC) detectada en una ecografía de seguimiento fue del 2,9% (IC 95%, 0,8–6,3%) (19/523; 15 estudios), mientras que las tasas de aquellas detectadas exclusivamente en la IRM o en la imagen postnatal fueron del 0% (IC 95%, 0,0–1,7%) (0/357; 11 estudios) y del 0,9% (IC 95%, 0,3–1,8%) (4/660; 17 estudios), respectivamente. La muerte intrauterina y la muerte perinatal ocurrieron cada una en el 0,7% (IC 95%, 0,3–1,4%) (2/824; 23 estudios) de los casos. En los casos sin una anomalía asociada detectada pre o post-natalmente, se encontró una infección sintomática en el 1,5% (IC 95%, 0,7–2,7%) (6/548; 19 estudios) de los bebés, la tasa total de una anomalía del desarrollo neurológico fue del 3,1% (IC 95%, 1,6–5,1%) (16/550; 19 estudios), y los problemas de audición afectaron al 6,5% (IC 95%, 3,8–10,0%) (36/550; 19 estudios) de los bebés. Los subanálisis según el trimestre en el que se produjo la infección materna se vieron afectados por el bajo número de casos incluidos y la falta de comparación de los resultados observados en los estudios originales. En comparación con los fetos infectados en el segundo o tercer trimestre, los infectados en el primer trimestre tenían un riesgo relativamente mayor de presentar una anomalía adicional detectada en la ecografía o en la resonancia magnética de seguimiento, un desenlace anómalo del desarrollo neurológico y problemas de audición.

Conclusiones

En aquellos fetos con infección congénita por CMV en los que no se detectan anomalías en la ecografía prenatal o en la IRM, el riesgo de un desenlace postnatal adverso es menor que el publicado previamente en la literatura cuando no se tiene en cuenta el papel de la evaluación prenatal mediante imágenes. Los resultados de esta revisión destacan también el papel potencial de la IRM, incluso en fetos sin anomalías detectadas en la ecografía, ya que una anomalía puede detectarse exclusivamente en la IRM en aproximadamente el 6% de los casos. Los resultados de este estudio podrían mejorar el asesoramiento prenatal de los embarazos con infección congénita por CMV mediante imágenes prenatales normales. © 2020 International Society of Ultrasound in Obstetrics and Gynecology.

摘要

zh

先天性巨细胞病毒感染和普通超声检查对胎儿的诊断:系统回顾和荟萃分析

目的

报告先天性巨细胞病毒感染的诊断(CMV)以及常规超声波在诊断时的应用

并评估仅在磁共振成像(MRI)上发现的其他异常发生率。

原理

用Medline,EMBASE,CINAHL和Cochrane数据库搜索,以获取有关:

先天性巨细胞病毒感染胎儿的预后。 纳入搜索的标准包括确诊为先天性巨细胞病毒感染(CMV)的胎儿,以及初次评估时常规超声检查 。 观察到的结果是在后续超声扫描中发现异常,在产前做核磁共振(MRI)时发现异常,但超声波并未查出,在产后评估中发现异常但在产前并未查出,围产期死亡率,出生时症状感染,神经发育预后和听力和视力缺陷。神经发育预后评估仅针对出生时确诊有孤立的CMV感染病例。 根据每三个月母体发生的感染的进行亚组分析。 用单组率的荟萃分析来分析数据。

结果

对 26项研究进行分析,包括2603例先天性巨细胞病毒感染的胎儿,其中1178例(45.3%)在诊断时超声检查正常,也被纳入分析。后续超声扫描发现的相关中枢神经系统(CNS)异常的查出率为4.4%(95%置信区间,1.4–8.8%)(32/523; 15项研究),而产前专项MRI 检查或产后影像学检查的查处率分别为5.8%(95%置信区间,1.9–11.5%)(19/357; 11项研究)和3.2%(95%置信区间,0.3-9.0%)(50/660; 17项研究) 。后续超声扫描发现与中枢神经系统相关的异常检出率为2.9%(95%zhi xin qu jian,0.8–6.3%)(19/523; 15项研究),而专项MRI或出生后影像学检查检出率为0%(95%置信区间,0.0-1.7%)(0/357; 11项研究)和0.9%(95%置信区间,0.3-1.8%)(4/660; 17项研究)。宫内死亡和围产期死亡率分别发生0.7%(95%置信区间,0.3–1.4%)(2/824; 23项研究)的病例中。在产前或产后均未发现异常的相关病例中,在1.5%(95%置信区间,0.7–2.7%)(6/548; 19项研究)的婴儿中发现有症状的感染,儿童中总计神经发育异常率为3.1% (95%置信区间,1.6–5.1%)(16/550; 19项研究),后遗症中听力障碍占6.5%(95%置信区间,3.8–10.0%)(36/550; 19项研究) 。对妊娠早孕产妇感染进行的亚组分析由于纳入病例数量极少,并且对原始研究中观察到的结果缺乏比较而受到影响。与中期或晚期妊娠感染的胎儿相比,妊娠早期感染的胎儿在后续的超声或MRI检查中有相对较高的其他神经发育异常的预后和听力问题。

结论

对于先天性CMV感染的胎儿,在产前超声或MRI上未发现异常,其不良的产后预后风险比以前发表的文献中所报道的未考虑产前影像学检查要低。 这篇综述的结果也凸显了MRI的潜在作用,即使在超声检查中未查出异常的胎儿中也是如此,因为大约6%的病例仅可在MRI上发现异常。 这项研究发现通过常规的产前影像学检查可提升先天性巨细胞病毒感染的孕妇的产前咨询效果。 ©2020国际妇产科超声学会。

CONTRIBUTION

What are the novel findings of this work?

In fetuses with congenital cytomegalovirus (CMV) infection in which no anomalies are detected on prenatal imaging, the risk of adverse postnatal outcome is lower compared with that reported previously in the published literature when not considering the role of antenatal imaging assessment.

What are the clinical implications of this work?

Women with CMV infection and normal prenatal imaging should be counseled that the risk of symptomatic infection or neurodevelopmental anomaly of the newborn is relatively low, particularly when maternal infection occurs after the first trimester. Prenatal assessment could benefit from fetal magnetic resonance imaging, which finds an anomaly in 6% of sonographically normal cases.

INTRODUCTION

Cytomegalovirus (CMV) is the most common congenital viral infection occurring during pregnancy, with a reported prevalence at birth of 0.48 to 1.3%1-4. Maternal infection is asymptomatic in the large majority of cases, although it can present sporadically as a flu-like syndrome. The chance of seroconversion depends on social, behavioral and environmental factors1-3, and has been estimated to be around 2% in pregnant women5. Maternal infection can lead to congenital CMV infection, which is caused by in-utero mother-to-fetus transmission. The likelihood of congenital infection is higher in primary than in non-primary maternal CMV infection, although, once acquired, the risk of symptomatic infection and long-term sequelae remains high even in non-primary maternal infection6. Furthermore, the risk of transmission is strongly related to gestational age, and is higher when seroconversion occurs during the third trimester of pregnancy7.

Parental counseling of pregnancies with congenital CMV infection is challenging and still based on old pediatric series which report that about 15–25% of CMV-infected infants develop short- and long-term sequelae, such as abnormal neurodevelopmental outcome and hearing loss8, 9.

However, in the last two decades, advances in prenatal imaging have led to a significant improvement in the detection rate of congenital anomalies, which has allowed more accurate prediction of the short- and long-term outcomes of these children10. The presence of a structural anomaly on ultrasound in CMV cases has been shown to be associated with a high risk of symptomatic infection and abnormal neurodevelopmental outcome11, 12. However, the burden of abnormal outcome in those fetuses presenting with no apparent anomalies on prenatal ultrasound has yet to be quantified.

The aim of this systematic review was to explore the outcome of fetuses with CMV infection and normal prenatal ultrasound at the time of diagnosis, and to evaluate the rate of an additional anomaly detected only on magnetic resonance imaging (MRI).

METHODS

Protocol, eligibility criteria, information sources and search

This review was performed according to an a-priori designed protocol recommended for systematic reviews and meta-analyses13-15. Medline, EMBASE, CINAHL and ClinicalTrials.gov databases were searched electronically in July 2020, utilizing combinations of the relevant medical subject heading (MeSH) terms, keywords and word variants for ‘cytomegalovirus’ or ‘infection’ and ‘outcome’. The search and selection criteria were restricted to the English language. Reference lists of relevant articles and reviews were searched manually for additional reports, and PRISMA and MOOSE guidelines were followed16-18.

Study selection, data collection and data items

Inclusion criteria were a fetus with confirmed CMV infection, defined as a positive polymerase chain reaction test for CMV in amniotic fluid or fetal blood samples, or a neonatal urine sample positive for CMV during the first few days after delivery, and with no associated central nervous system (CNS) or extra-CNS anomalies detected on ultrasound at the time of diagnosis.

The outcomes observed were: an associated anomaly detected on a follow-up ultrasound scan; an associated anomaly detected exclusively on fetal MRI and missed on ultrasound; an associated anomaly detected exclusively on postnatal imaging and missed on prenatal assessment (either ultrasound or MRI); an associated anomaly detected exclusively on clinical examination at birth, including microcephaly, cutaneous rash, elevated liver enzymes, thrombocytopenia and chorioretinitis; perinatal mortality, including intrauterine death, neonatal death and perinatal death; symptomatic infection at birth; abnormal neurodevelopmental outcome (overall, mild, moderate and severe), defined as the presence of a neurological or neuropsychological anomaly; hearing loss; visual problems; abnormal motor function; and abnormal cognitive function. Furthermore, we aimed to perform a subgroup analysis according to gestational age at maternal infection (first, second or third trimester). Data from studies reporting the incidence of these outcomes in fetuses with isolated CMV infection at diagnosis were considered eligible for inclusion in the analysis. Studies reporting cases of CMV infection presenting with an associated anomaly at the first ultrasound assessment and those from which information on prenatal imaging could not be extrapolated were excluded. Autopsy-based studies were excluded on the basis that fetuses undergoing termination of pregnancy are more likely to show an associated major structural anomaly. Likewise, series reporting information on CMV infection in the setting of a specific CNS or extra-CNS malformation, and studies including only cases diagnosed postnatally, were also excluded because they report higher rates of adverse outcome and may not reflect the natural history of the anomaly. Studies published before 2000 were also excluded because we felt that advances in prenatal imaging techniques and improvements in the diagnosis of CMV infection make them less relevant. Only full-text articles were considered eligible for inclusion; case reports, conference abstracts and case series with fewer than five cases were also excluded to avoid publication bias.

Two authors (F.D.A., D.B.) independently reviewed all abstracts. Agreement regarding potential relevance or inconsistencies was reached by consensus between the two reviewers or by discussion with a third reviewer (M.Li.). Full-text copies of those papers considered potentially eligible were obtained, and the same reviewers independently extracted relevant data regarding study characteristics and pregnancy outcomes. If more than one study had been published on the same cohort with identical endpoints, the report containing the most comprehensive information on the population was included to avoid overlapping populations.

For the purpose of the analysis, associated MRI anomalies were classified as a structural malformation or abnormal signal intensity without a clear morphological anomaly. Neurodevelopmental outcome was ascertained exclusively in infants with isolated CMV infection confirmed at birth.

Quality assessment of the included studies was performed using the Newcastle–Ottawa Scale (NOS) for cohort studies. According to the NOS, each study is judged on three broad perspectives: selection of the study groups, comparability of the groups and ascertainment of the outcome of interest19. Assessment of the selection of a study includes evaluation of the representativeness of the exposed cohort, selection of the non-exposed cohort, ascertainment of the exposure and demonstration that the outcome of interest was not present at the start of the study. Assessment of the comparability of the study includes evaluation of the comparability of cohorts based on the design or analysis. Ascertainment of the outcome of interest includes evaluation of the type of assessment of the outcome of interest and the length and adequacy of follow-up. According to the NOS, a study can be awarded a maximum of one star for each numbered item within the selection and outcome categories, and a maximum of two stars can be given for comparability19.

Statistical analysis

Meta-analysis of proportions was used to combine data, and pooled proportions are reported. Funnel plots displaying the outcome rate from individual studies vs their precision (one per standard error) were carried out with an exploratory aim. Tests for funnel-plot asymmetry were not used when the total number of publications included for each outcome was less than 10. In such cases, the power of the tests is too low to distinguish chance from real asymmetry.

Between-study heterogeneity was explored using the I2 statistic, which represents the percentage of between-study variation that is due to heterogeneity rather than chance. A value of 0% indicates no observed heterogeneity, whereas I2 values ≥ 50% indicate a substantial level of heterogeneity. A random-effects model was used to compute the pooled data analysis. All proportion meta-analysis was carried out using StatsDirect version 2.7.9 (StatsDirect, Ltd, Altrincham, Cheshire, UK).

RESULTS

General characteristics

In total, 434 articles were identified, of which 102 were assessed with respect to their eligibility for inclusion and 2610–12,20–42 were included in the systematic review (Figure 1, Table 1, Table S1). These 26 studies included 2603 fetuses affected by CMV infection; of these, 1178 (45.3% (95% CI, 43.4–47.2%)) had no associated CNS or extra-CNS anomalies detected on ultrasound at the time of diagnosis. The majority of the included studies reported data on primary CMV infection, while six22, 26, 36, 37, 41, 42 reported the outcome of both primary and secondary infections, although these studies included very few cases. The timing of follow-up ultrasound ranged from 2 to 4 weeks. In four studies22, 28, 33, 40, some of the included cases underwent prenatal therapy, mainly with valacyclovir.

Details are in the caption following the image
Flowchart summarizing inclusion in systematic review of studies reporting on outcome of fetuses with congenital cytomegalovirus infection.
Table 1. General characteristics of studies included in systematic review on outcome of fetuses with congenital cytomegalovirus infection
Study Country Study design Study period Type of infection Prenatal imaging GA at infection Ultrasound follow-up Age at postnatal follow-up* Prenatal therapy Fetuses (n)
Lipitz (2020)20 Israel Retro 2011–2018 Primary US+MRI 1T 3–4 weeks 2 years (6 months to 10 years) No 123
Birnbaum (2017)21 Israel Retro 2011–2014 Primary US+MRI NS 3–4 weeks 48 (27 to 64) months No 81
Leruez-Ville (2016)22 France Retro 2008–2013 Primary and secondary US 1T, 2T, 3T NS (‘serial’) 24 (1 to 53) months Yes 82
Leyder (2016)12 Belgium Prosp 1996–2012 Primary US 1T, 2T 4 weeks NS No 69
Cannie (2016)23 Belgium Retro 2004–2014 Primary US+MRI 1T, 2T NS 4.4 (0.6 to 9.7) years No 121
Amir (2016)24 Israel Retro 2007–2013 Primary US 1T, 2T NS 32 (12 to 83) months No 98
Zavattoni (2016)25 Italy Prosp 1995–2010 Primary US+MRI 1T NS 6 years No 46
Picone (2014)26 France Retro 2004–2013 Primary and secondary US+MRI NS 2 weeks NS NS 69
Garcia-Flores (2013)27 Spain Retro 2008–2012 Primary US+MRI NS NS 1 year NS 7
Farkas (2011)28 Israel Retro 2001–2007 Primary US+MRI NS 3–4 weeks 11 to 81 months Yes 69
Feldman (2011)29 Israel Retro 2000–2006 Primary US+MRI NS 3–4 weeks 18 months No 118
Lipitz (2010)10 Israel Prosp 2004–2009 Primary US+MRI 1T, 2T, 3T 3–4 weeks 6 to 55 months No 38
Doneda (2010)30 Italy Retro 2000–2008 Primary US+MRI 1T, 2T NS 1 month to 8 years NS 38
Simonazzi (2010)31 Italy Prosp 2007–2008 Primary US NS NS NS No 218
Picone (2008)32 France Retro 1991–2002 NS US+MRI NS NS 23 (0 to 72) months No 38
Benoist (2008)33 France Retro 2000–2007 Primary US+MRI 3T 2 weeks NS Yes§ 49
Benoist (2008)34 France Retro 2000–2006 Primary US+MRI 3T NS NS No 73
Gindes (2008)35 Israel Prosp 1998–2005 Primary US 3T 2 weeks 36 (6 to 36) months No 21
Romanelli (2008)36 France Retro 2005–2006 Primary and secondary US+MRI 1T, 2T 2 weeks NS NS 13
Guerra (2008)11 Italy NS 1996–2006 Primary US NS NS NS No 650
Yinon (2006)37 Israel Retro 1997–2004 Primary and secondary US 1T, 2T NS 36 (4 to 72) months No 40
Lipitz (2002)38 Israel Prosp 1999–2000 Primary US+MRI 1T, 2T, 3T 3–4 weeks 32 (6 to 96) months No 51
Gouarin (2002)39 France Retro 1992–1999 Primary US 2T NS NS NS 30
Enders (2001)40 Germany Retro 1988–2000 Primary US 1T, 2T, 3T NS 22.5 months (3 months to 8 years) Yes 189
Azam (2001)41 Switzerland Prosp 1989–1998 Primary and secondary US 1T, 2T, 3T 4 weeks 39 (14 to 76) months No 26
Lazzarotto (2000)42 Italy Prosp 1994–1998 Primary and secondary US NS NS NS No 246
  • Only first author's name given.
  • * Median (range), or as reported in study.
  • Valacyclovir in 37 cases.
  • Therapy in four cases.
  • § Valacyclovir in 17 cases.
  • Therapy in two cases. 1T, first trimester; 2T, second trimester; 3T, third trimester; GA, gestational age; MRI, magnetic resonance imaging; NS, not specified; Prosp, prospective; Retro, retrospective; US, ultrasound.

The results of the quality assessment of the included studies using the NOS are presented in Table 2. All the included studies showed an overall good score regarding the selection and comparability of the study groups and for ascertainment of the outcome of interest. The main weaknesses of these studies were their retrospective design, small sample size, heterogeneity of the outcomes observed and lack of stratification of the analysis according to gestational age at infection. The results of funnel-plot analysis for each of the study outcomes can be found in Appendix S1.

Table 2. Quality assessment of included studies according to Newcastle–Ottawa Scale (NOS) for case–control studies
Study Selection Comparability Outcome
Lipitz (2020)20 ★★★ ★★
Birnbaum (2017)21 ★★★ ★★
Leruez-Ville (2016)22 ★★★ ★★
Leyder (2016)12 ★★★ ★★
Cannie (2016)23 ★★★ ★★
Amir (2016)24 ★★★ ★★
Zavattoni (2016)25 ★★★ ★★
Picone (2014)26 ★★★ ★★
Garcia-Flores (2013)27 ★★★ ★★
Farkas (2011)28 ★★★ ★★
Feldman (2011)29 ★★★ ★★
Lipitz (2010)10 ★★★ ★★
Doneda (2010)30 ★★★ ★★
Simonazzi (2010)31 ★★★ ★★
Picone (2008)32 ★★★ ★★
Benoist (2008)33 ★★★ ★★
Benoist (2008)34 ★★★ ★★
Gindes (2008)35 ★★★ ★★
Romanelli (2008)36 ★★★ ★★
Guerra (2008)11 ★★★ ★★
Yinon (2006)37 ★★★ ★★
Lipitz (2002)38 ★★★ ★★
Gouarin (2002)39 ★★★ ★★
Enders (2001)40 ★★★ ★★
Azam (2001)41 ★★★ ★★
Lazzarotto (2000)42 ★★★ ★★
  • Only first author's name given. According to NOS, a study can be awarded maximum of one star for each numbered item within selection and outcome categories. Maximum of two stars can be given for comparability.

Synthesis of results

Table 3 shows the rate of an additional anomaly found on follow-up prenatal imaging or at birth in fetuses with congenital CMV infection and normal ultrasound assessment at the time of diagnosis.

Table 3. Incidence of anomaly detected on prenatal imaging or at birth in fetuses with congenital cytomegalovirus infection and normal ultrasound assessment at time of diagnosis
Finding Studies (n) Fetuses (n/N) Pooled proportion (%) (95% CI) I2 (%)
Follow-up US
CNS anomaly 15 32/523 4.35 (1.4–8.8) 74.4
Extra-CNS anomaly 15 19/523 2.93 (0.8–6.3) 68.2
Prenatal MRI only (missed on US)
CNS anomaly 11 19/357 5.77 (1.9–11.5) 68
Extra-CNS anomaly 11 0/357 0 (0–1.7) 0
Abnormal signal intensity 11 21/357 5.24 (1.5–11.0) 70.8
Postnatal imaging only (missed prenatally)
CNS anomaly 17 50/660 3.23 (0.3–9.0) 89.3
Extra-CNS anomaly 17 4/660 0.91 (0.3–1.8) 0
Postnatal clinical examination only
Microcephaly 16 8/442 2.01 (0.1–3.5) 0
Rash 10 2/308 1.03 (0.2–2.5) 0
Elevated liver enzymes 11 19/343 4.05 (0.5–10.1) 77.8
Thrombocytopenia 11 7/343 2.25 (1.0–4.1) 2
Chorioretinitis 10 3/326 1.39 (0.4–2.9) 0
  • CNS, central nervous system; MRI, magnetic resonance imaging; US, ultrasound.

The overall rate of an associated CNS anomaly detected on follow-up ultrasound was 4.4% (95% CI, 1.4–8.8%) (32/523; 15 studies), while the rates of those detected exclusively by prenatal MRI or postnatal imaging were 5.8% (95% CI, 1.9–11.5%) (19/357; 11 studies) and 3.2% (95% CI, 0.3–9.0%) (50/660; 17 studies), respectively. The rate of an associated extra-CNS anomaly detected exclusively on follow-up ultrasound was 2.9% (95% CI, 0.8–6.3%) (19/523; 15 studies), while the rates of those detected exclusively by prenatal MRI or postnatal imaging were 0% (95% CI, 0.0–1.7%) (0/357; 11 studies) and 0.9% (95% CI, 0.3–1.8%) (4/660; 17 studies), respectively.

Abnormal signal intensity was detected in 5.2% (21/357) of cases, although two studies10, 21 reporting the postnatal outcome of fetuses with hyperintensity on MRI found no neurological impairment in this subset of children.

Among abnormalities found exclusively on clinical examination, elevated liver enzymes (4.1% (19/343)), thrombocytopenia (2.3% (7/343)) and microcephaly (2.0% (8/442)) were the three most common conditions.

Intrauterine death and perinatal death each occurred in 0.7% (95% CI, 0.3–1.4%) (2/824) of cases, while no cases of neonatal death were reported in the included studies (Table 4).

Table 4. Incidence of intrauterine (IUD), neonatal (NND) and perinatal (PND) death in fetuses with congenital cytomegalovirus infection and normal ultrasound assessment at time of diagnosis
Outcome Studies (n) Fetuses (n/N) Pooled proportion (%) (95% CI) I2 (%)
IUD 23 2/824 0.74 (0.3–1.4) 0
NND 23 0/824 0 (0–1.4) 0
PND 23 2/824 0.74 (0.3–1.4) 0

When considering the postnatal outcome of cases without an associated anomaly detected pre- or postnatally, symptomatic infection was seen in 1.5% (95% CI, 0.7–2.7%) (6/548) of cases, and the overall rate of a neurodevelopmental anomaly was 3.1% (95% CI, 1.6–5.1%) (16/550), ranging from mild (1.2% (4/471)) to severe (1.9% (8/471)). Hearing problems affected 6.5% (95% CI, 3.8–10.0%) of children, while motor and cognitive delays and visual problems were reported, respectively, in 2.3% (9/487), 1.1% (3/487) and 1.0% (1/471) of cases (Table 5).

Table 5. Incidence of abnormal postnatal outcome in fetuses with isolated congenital cytomegalovirus infection confirmed at birth
Outcome Studies (n) Fetuses (n/N) Pooled proportion (%) (95% CI) I2 (%)
Symptomatic infection 19 6/548 1.49 (0.7–2.7) 22.1
Abnormal neurodevelopmental outcome 19 16/550 3.10 (1.6–5.1) 22.1
Mild 17 4/471 1.19 (0.4–2.3) 0
Moderate 17 0/471 0 (0–1.7) 0
Severe 17 8/471 1.90 (0.9–3.3) 0
Visual problems 18 1/471 0.95 (0.3–2.0) 0
Hearing problems 19 36/550 6.54 (3.8–10.0) 48.8
Cognitive delay 18 3/487 1.14 (0.4–2.3) 0
Motor delay 18 9/487 2.28 (1.2–3.8) 0
Other neurological anomaly 18 2/487 1.10 (0.4–2.2) 0

Subgroup analysis according to trimester of infection

Subgroup analysis according to the trimester in which infection occurred was affected by the small number of included cases and even smaller number of events, which may have affected the robustness of the results.

An anomaly was detected on a follow-up ultrasound assessment in 4.4% of fetuses infected in the first trimester and in none of those infected in the second or third trimester. Likewise, the rate of an associated structural anomaly detected exclusively on prenatal MRI was 6.3% in fetuses infected in the first trimester, while no additional anomaly was found on MRI in those infected in the second or third trimester (Table S2).

In cases without an associated anomaly detected pre- or postnatally, the incidence of abnormal neurodevelopmental outcome was 5.4% in children infected in the first trimester and 0% in those infected in the second or third trimester. Hearing problems occurred in 11.4% of children infected in the first trimester, 7.0% of those infected in the second trimester and 0% of those infected in the third trimester (Table S3).

DISCUSSION

Main findings

The findings of this systematic review show that, in fetuses with congenital CMV infection and no anomalies detected on prenatal imaging, the risk of adverse postnatal outcome is lower than that reported previously in the published literature when not considering the role of antenatal imaging assessment. In fetuses with normal ultrasound at the time of diagnosis, a CNS anomaly was detected on follow-up ultrasound in about 4% of cases, thus highlighting the need for longitudinal evaluation of affected fetuses throughout pregnancy. The results of this study also highlight the potential role of MRI even in fetuses with no anomalies on ultrasound, as an anomaly can be detected exclusively on MRI in about 6% of cases. The risk of perinatal mortality in congenitally infected fetuses was negligible. More importantly, symptomatic infection and abnormal neurodevelopmental outcome occurred, respectively, in 1.5% and 3.1% of fetuses with normal prenatal imaging confirmed at birth, while hearing problems occurred in 6.5%. Subanalysis according to the trimester in which infection occurred was affected by the very small number of included cases and lack of comparison of the observed outcomes in the original studies. Compared with fetuses infected in the second or third trimester, those infected in the first trimester had relatively higher risks of having an additional anomaly detected on follow-up ultrasound or MRI, abnormal neurodevelopmental outcome and hearing problems.

Strengths and limitations

To the best of our knowledge, this is the first systematic review to explore the role of prenatal imaging in fetuses with congenital CMV infection. The extensive literature search, the multitude of outcomes explored and the strict inclusion criteria are the major strengths of this review. The small number of cases in some of the included studies, their retrospective non-randomized design, the lack of standardized criteria for surveillance and follow-up examinations and the heterogeneity of the outcomes observed are the major limitations of the review. Furthermore, we acknowledge that it is probable that not all these cases underwent multiplanar neurosonography to evaluate the fetal brain, and it is well known that the chance of finding an additional anomaly on MRI significantly depends on the quality of the sonogram, and is higher when only standard ultrasonography has been performed43. Another major limitation of the present review is the very small number of included cases in the subgroup analysis according to the timing of maternal infection, which may affect the robustness of the reported results. Furthermore, we could not stratify the analysis according to the viral load of the amniotic fluid (although it has been reported that a higher viral load is more likely to be associated with a symptomatic fetus or newborn42) or the type of maternal infection (primary vs non-primary) owing to the lack of reporting of these data in the included studies. Despite these limitations, to our knowledge, this is the first and most comprehensive review specifically focusing on the outcomes of fetuses with CMV infection and normal prenatal ultrasound at the time of diagnosis.

Clinical and research implications

CMV infection is the most common fetal infection and is the leading cause of non-hereditary sensorineural hearing loss (SNHL) in high-income countries44-46. However, despite the clinical relevance of the infection, although maternal serological screening has been improved, it is currently offered in only a few countries, such as Germany and Italy47.

Currently, parental counseling in pregnancies affected by congenital CMV infection is still based on observational studies that date back many years. Parents are commonly counseled that, even if the infant is asymptomatic, congenital CMV infection is one of the main determinants of SNHL, delayed-onset SNHL and fluctuating SNHL48, with an overall incidence of SNHL ranging from 15–20% up to 44% in symptomatic children8, 44. However, current counseling of women with a pregnancy complicated by congenital CMV infection does not fully take into account the role of prenatal imaging in predicting perinatal and neurodevelopmental outcome.

The findings from the current systematic review show that the rate of abnormal neurocognitive delay (both mild and severe) and the incidence of hearing loss and visual defects in fetuses with congenital CMV infection and no anomalies seen on prenatal imaging are lower than those reported previously, with the highest incidence occurring in cases of first-trimester infection. In this scenario, parents whose pregnancy is complicated by congenital CMV infection can be counseled on the low risk of symptomatic infection and severe neurodevelopmental outcome when prenatal imaging is normal, although the risk of hearing problems cannot be ruled out completely by ultrasound.

Fetuses with isolated subtle CNS anomalies, such as ventriculomegaly, are commonly referred for prenatal MRI assessment in order to rule out additional anomalies that can significantly impact on the prediction of postnatal outcome43, 49. Despite this, there is still insufficient evidence on whether prenatal MRI should be offered for fetuses with congenital infection. In the present review, the overall rate of an additional finding detected exclusively on MRI was about 6%, and abnormal hyperintensity was found in 5%. It is noteworthy that no anomalies at birth and no significant neurological sequelae were found in studies reporting the postnatal outcome of fetuses with abnormal signal intensity10, 21. Moreover, the clinical significance and prognostic value of a hyperintense signal on MRI are still uncertain, particularly when it affects the frontal and parieto-occipital white matter, while hyperintensity in the temporal lobe is more likely to be associated with adverse neurological outcome50. In such situations, prenatal MRI should always be considered in the diagnostic work-up of fetuses with congenital CMV infection, even when no anomalies are detected on ultrasound. Nevertheless, parents might be reassured that the risk of detecting an associated anomaly on MRI is low when a thorough sonographic assessment has been performed.

The timing of maternal infection is another issue during the management of congenital CMV infection. A recent systematic review showed that severe fetal impairment is rare when maternal infection occurs after the first trimester of pregnancy51. These results are concordant with the findings from our review, as we found that fetuses infected in the first trimester, compared with those infected in the second or third trimester, had a relatively higher risk of an additional anomaly detected on follow-up ultrasound or MRI, abnormal neurodevelopmental outcome and hearing problems, although the strength of the associations was affected by the small number of included cases.

Conclusion

In fetuses affected by congenital CMV infection with no anomalies detected on prenatal imaging, the risk of adverse postnatal outcome, particularly in terms of hearing loss and neurodevelopmental impairment, is lower than that reported previously when not considering the role of antenatal imaging assessment. Moreover, the results from this review highlight the potential role of MRI even in fetuses with no anomalies on ultrasound, as an anomaly can be detected exclusively on MRI in about 6% of cases. The findings from this study could enhance the prenatal counseling of pregnancies with congenital CMV infection and normal prenatal imaging.

ACKNOWLEDGMENT

We thank Professor S. Lipitz, Dr T. E. Miller and Dr M. Leruez-Ville for providing further information for this meta-analysis.

    DATA AVAILABILITY STATEMENT

    The data that support the findings of this study are available from the corresponding author upon reasonable request.