DOI: 10.1002/pd.4362
ORIGINAL ARTICLE
Congenital cytomegalovirus infection and small for gestational age infants Giuliana Simonazzi1*, Alessandra Curti1, Paola Murano1, Francesca Cervi1, Margherita Contoli1, Tiziana Lazzarotto2, Maria Grazia Capretti3, Nicola Rizzo1 and Brunella Guerra1 1
Department of Medical Surgical Sciences, Division of Obstetrics and Prenatal Medicine, St. Orsola Malpighi Hospital, University of Bologna, Bologna, Italy Department of Specialised, Experimental, and Diagnostic Medicine, Division of Microbiology and Virology, St. Orsola-Malpighi Hospital, University of Bologna, Bologna, Italy 3 Department of Medical Surgical Sciences, Division of Neonatology, St. Orsola-Malpighi Hospital, University of Bologna, Bologna, Italy *Correspondence to: Giuliana Simonazzi. E-mail: [email protected] 2
ABSTRACT Objective To evaluate the incidence of infants with birth weight less than the 10th centile for small for gestational age (SGA) in primary maternal cytomegalovirus (CMV) infection and to determine whether SGA predicts poor neurodevelopmental outcome. Methods A retrospective cohort study, which included singleton live-born infants from pregnancies complicated by primary maternal CMV infection. Infants were classified as uninfected or infected based on viral virus isolation and real-time PCR in urine at birth. SGA neonates rate and long-term sequelae were evaluated. Results Between 2000 and 2012, 848 women with primary CMV infection were referred to our center with 588 infants assessed at birth. Congenital CMV infection was diagnosed in 119 cases (20%), of which 8 were SGA (6.7%) compared with 27 out of the 469 uninfected infants (5.7%) (p-value = 0.69). Among the 119 infected babies, 14 infants were symptomatic at birth or at postnatal follow-up (12%), of whom two were SGA compared with six out of the 105 asymptomatic neonates (5.7% versus 15%, respectively, p-value = 0.22).
Conclusion Congenital CMV infection does not seem to be associated with a higher incidence of SGA, and long-term outcomes do not seem to be affected by isolated impaired fetal growth. © 2014 John Wiley & Sons, Ltd.
Funding sources: None Conflicts of interest: None declared
INTRODUCTION Cytomegalovirus (CMV) is the most common cause of intrauterine infection, occurring in 0.3–2% of live-born infants.1 CMV can be transmitted to the fetus in both primary and non-primary maternal infection. Primary maternal infection seems to represent a higher risk for transmission and severity of fetal infection, but congenital infection and postnatal disability can also occur following recurrent or exogenous reinfection.2–5 Although only from 10 to 15% of infected fetuses show symptoms at birth, the clinical manifestations may be so severe as to lead to a high perinatal mortality rate and major neurologic sequelae in most of the surviving babies. In addition, from 10 to 15% of asymptomatic neonates will develop long-term sequelae, namely progressive hearing loss and mental retardation.6 Intrauterine infection may be the primary etiology of inadequate fetal growth in approximately from 5 to 10% of cases.7 Viral infection, including CMV,8–10 is usually implicated Prenatal Diagnosis 2014, 34, 765–769
as a fetal factor affecting growth. Infection might also be involved in placental insufficiency, both in terms of initiation (by increasing the risk of acute uteroplacental atherosis) and/ or potentiation (by amplifying the maternal systemic inflammatory response).11,12 The aim of our study was to evaluate the incidence of infants with birth weight less than the 10th centile for gestational age, defined as small for gestational age (SGA), in a cohort of pregnancies complicated by primary CMV infection and to determine whether SGA predicts poor neurodevelopmental outcome.
MATERIAL AND METHODS This retrospective cohort study included pregnancies complicated by maternal primary CMV infection referred for prenatal counseling to Sant’Orsola-Malpighi Hospital, University of Bologna, Bologna, (Italy) from other cities and regions of Italy. The study was carried out following the ethical rules of St. Orsola-Malpighi General Hospital, Bologna, Italy (reference number: 28/2012/1/CMV; date of approval: 12 April 2012). © 2014 John Wiley & Sons, Ltd.
G. Simonazzi et al.
766
Multiple pregnancies and patients who received CMV-specific hyperimmune globulin (i.e. Efficacy Study of Human Cytomegalovirus Hyperimmune Globulin to Prevent Congenital Human CMV Infection and Trial ClinicalTrials.gov Identifier: NCT00881517) were excluded, as were therapeutic abortions and miscarriages. Maternal primary CMV infection was identified as seroconversion in previously CMV-seronegative women or by detection of CMV-specific IgM combined with low avidity anti-CMV IgG.13 As is usual at our institution, women with primary CMV infection acquired early in gestation were offered an anatomy scan at 20 to 22 weeks and amniocentesis at 20 to 21 weeks for diagnosis of fetal infection, in all cases, at least 6 weeks after seroconversion. The diagnostic performances of viral isolation and PCR (real time) on amniotic fluid with respect to the presence or absence of fetal CMV infection have been described elsewhere.13,14 According to our protocol, serial ultrasound scans are recommended at 28 and 33 weeks’ gestation throughout pregnancy.15 Because most pregnant women referred to our institution are from other cities and regions, patients can decide whether to undergo ultrasound scans at our division or not. Cytomegalovirus congenital infection was identified by rapid virus isolation16 and real-time PCR in urine within the second week of life. CMV-DNA was identified and quantified with a real-time PCR commercial assay (CMV ELITe MGB kit, ELITechGroup, Turin, Italy) performed with the ABI PRISM 7300 platform (PE Applied Biosystems, Foster City, CA, USA). The detection limit was 11 copies/reaction. Information concerning pediatric outcomes was obtained directly when the mothers had given birth at our hospital. If the mothers had received care elsewhere, this information was elicited by postal questionnaire, which was included in a letter sent to patients’ doctor to be returned to our institution for data collection. Doctors were asked to report on the following: (1) the results of CMV isolation from urine in the newborn during the first week of life, if performed, and (2) their child’s health status. Outcomes of interest were the incidence of SGA infants and long-term sequelae in case of symptomatic CMV infection. Infants were defined as SGA when their birth weight was less than the 10th centile, and appropriate for gestational age (AGA) when their birth weight was more than the 10th centile, according to Italian charts used in our hospital.17 The following were considered long-term sequelae: hearing impairment, including deafness (complete unilateral or bilateral hearing loss) and hearing loss (any clinically reduced hearing not defined as deafness), and neurodevelopmental delay, including motor, speech, and learning disabilities. Descriptive statistical analysis was performed by routine tests. Results are presented as mean (± standard deviation) for continuous variables or percentage for categorical variables (n,%). Continuous and categorical variables were compared using sample t-test and Fisher’s exact test, respectively. Statistical significance was defined as a p-value < 0.05. A power analysis was performed. Given an incidence of SGA neonates of about 10% in the general population, 115 case patients (infected babies) and 460 controls (uninfected babies) were Prenatal Diagnosis 2014, 34, 765–769
needed to have 80% power to detect a 50% increase in the rate of SGA at birth.
RESULTS Between January 2000 and June 2012, 848 women with primary maternal CMV infection were referred to our division for prenatal counseling: 157 patients (18 multiple pregnancies, 48 patients treated with CMV-specific hyperimmune globulin, 37 miscarriages and 54 therapeutic abortions in the first or second trimester) were excluded, whereas 103 patients were lost to follow-up. Therefore, the study population comprised 588 live-born infants. Data on the study population are summarized in Table 1. Congenital CMV infection was diagnosed in 119 cases (20%). No difference was found between uninfected and infected babies, except for estimated time of onset of maternal infection (10.8 ± 7.5 weeks versus 15.1 ± 8.82 weeks, respectively, pvalue = 0.0001). Thirty-five out of 588 neonates were SGA (6%), 27 (5.7%) and 8 (6.7%) among uninfected and infected babies, respectively (p-value = 0.69). Gestational age at the estimated time of maternal infection was similar in CMVinfected SGA and AGA infants (11 ± 6.9 weeks versus 15.4 ± 8.89 weeks, respectively, p-value = 0.16). Three hundred and fifty-five patients underwent amniocentesis to diagnose fetal infection. A high viral load in the amniotic fluid was found in 28 cases; eight of these infected infants were symptomatic. Table 1 lists details on birth weight centiles according to neonatal infection status. Abnormal findings were detected on mid-trimester ultrasound in 29 babies (Table 2). Two infants were SGA: one uninfected baby presented cerebral ventriculomegaly, whereas bowel hyperechogenicity was diagnosed in the infected neonates. As regards maternal complications, 11 out of 588 pregnancies were complicated by hypertensive disorders (1.8%), gestational hypertension in six cases, preeclampsia in four and HELLP syndrome in one case. Only two of these patients had a CMV-infected baby: neither was SGA at birth or had symptoms of disease at follow-up.
Table 1 Study population characteristics Uninfected infants (n = 469)
Infected infants (n = 119)
p-value
31 ± 5
31 ± 4
1.00a
219 (46.7%)
53 (44.5%)
0.67b
Week of infection
10.8 ± 7.5
15.1 ± 8.82
0.0001a
Gestational age at delivery (weeks)
39.15 ± 2.28
38.94 ± 1.88
0.35a
3311 ± 473
3247 ± 529
0.19b
27 (5.7%)
8 (6.7%)
0.69a
rd
7
2
0.95a
rd
20
6
0.95a
Maternal age (years) Nulliparous
Birth weight (g) SGA 3 centile th and
ORIGINAL ARTICLE
Congenital cytomegalovirus infection and small for gestational age infants Giuliana Simonazzi1*, Alessandra Curti1, Paola Murano1, Francesca Cervi1, Margherita Contoli1, Tiziana Lazzarotto2, Maria Grazia Capretti3, Nicola Rizzo1 and Brunella Guerra1 1
Department of Medical Surgical Sciences, Division of Obstetrics and Prenatal Medicine, St. Orsola Malpighi Hospital, University of Bologna, Bologna, Italy Department of Specialised, Experimental, and Diagnostic Medicine, Division of Microbiology and Virology, St. Orsola-Malpighi Hospital, University of Bologna, Bologna, Italy 3 Department of Medical Surgical Sciences, Division of Neonatology, St. Orsola-Malpighi Hospital, University of Bologna, Bologna, Italy *Correspondence to: Giuliana Simonazzi. E-mail: [email protected] 2
ABSTRACT Objective To evaluate the incidence of infants with birth weight less than the 10th centile for small for gestational age (SGA) in primary maternal cytomegalovirus (CMV) infection and to determine whether SGA predicts poor neurodevelopmental outcome. Methods A retrospective cohort study, which included singleton live-born infants from pregnancies complicated by primary maternal CMV infection. Infants were classified as uninfected or infected based on viral virus isolation and real-time PCR in urine at birth. SGA neonates rate and long-term sequelae were evaluated. Results Between 2000 and 2012, 848 women with primary CMV infection were referred to our center with 588 infants assessed at birth. Congenital CMV infection was diagnosed in 119 cases (20%), of which 8 were SGA (6.7%) compared with 27 out of the 469 uninfected infants (5.7%) (p-value = 0.69). Among the 119 infected babies, 14 infants were symptomatic at birth or at postnatal follow-up (12%), of whom two were SGA compared with six out of the 105 asymptomatic neonates (5.7% versus 15%, respectively, p-value = 0.22).
Conclusion Congenital CMV infection does not seem to be associated with a higher incidence of SGA, and long-term outcomes do not seem to be affected by isolated impaired fetal growth. © 2014 John Wiley & Sons, Ltd.
Funding sources: None Conflicts of interest: None declared
INTRODUCTION Cytomegalovirus (CMV) is the most common cause of intrauterine infection, occurring in 0.3–2% of live-born infants.1 CMV can be transmitted to the fetus in both primary and non-primary maternal infection. Primary maternal infection seems to represent a higher risk for transmission and severity of fetal infection, but congenital infection and postnatal disability can also occur following recurrent or exogenous reinfection.2–5 Although only from 10 to 15% of infected fetuses show symptoms at birth, the clinical manifestations may be so severe as to lead to a high perinatal mortality rate and major neurologic sequelae in most of the surviving babies. In addition, from 10 to 15% of asymptomatic neonates will develop long-term sequelae, namely progressive hearing loss and mental retardation.6 Intrauterine infection may be the primary etiology of inadequate fetal growth in approximately from 5 to 10% of cases.7 Viral infection, including CMV,8–10 is usually implicated Prenatal Diagnosis 2014, 34, 765–769
as a fetal factor affecting growth. Infection might also be involved in placental insufficiency, both in terms of initiation (by increasing the risk of acute uteroplacental atherosis) and/ or potentiation (by amplifying the maternal systemic inflammatory response).11,12 The aim of our study was to evaluate the incidence of infants with birth weight less than the 10th centile for gestational age, defined as small for gestational age (SGA), in a cohort of pregnancies complicated by primary CMV infection and to determine whether SGA predicts poor neurodevelopmental outcome.
MATERIAL AND METHODS This retrospective cohort study included pregnancies complicated by maternal primary CMV infection referred for prenatal counseling to Sant’Orsola-Malpighi Hospital, University of Bologna, Bologna, (Italy) from other cities and regions of Italy. The study was carried out following the ethical rules of St. Orsola-Malpighi General Hospital, Bologna, Italy (reference number: 28/2012/1/CMV; date of approval: 12 April 2012). © 2014 John Wiley & Sons, Ltd.
G. Simonazzi et al.
766
Multiple pregnancies and patients who received CMV-specific hyperimmune globulin (i.e. Efficacy Study of Human Cytomegalovirus Hyperimmune Globulin to Prevent Congenital Human CMV Infection and Trial ClinicalTrials.gov Identifier: NCT00881517) were excluded, as were therapeutic abortions and miscarriages. Maternal primary CMV infection was identified as seroconversion in previously CMV-seronegative women or by detection of CMV-specific IgM combined with low avidity anti-CMV IgG.13 As is usual at our institution, women with primary CMV infection acquired early in gestation were offered an anatomy scan at 20 to 22 weeks and amniocentesis at 20 to 21 weeks for diagnosis of fetal infection, in all cases, at least 6 weeks after seroconversion. The diagnostic performances of viral isolation and PCR (real time) on amniotic fluid with respect to the presence or absence of fetal CMV infection have been described elsewhere.13,14 According to our protocol, serial ultrasound scans are recommended at 28 and 33 weeks’ gestation throughout pregnancy.15 Because most pregnant women referred to our institution are from other cities and regions, patients can decide whether to undergo ultrasound scans at our division or not. Cytomegalovirus congenital infection was identified by rapid virus isolation16 and real-time PCR in urine within the second week of life. CMV-DNA was identified and quantified with a real-time PCR commercial assay (CMV ELITe MGB kit, ELITechGroup, Turin, Italy) performed with the ABI PRISM 7300 platform (PE Applied Biosystems, Foster City, CA, USA). The detection limit was 11 copies/reaction. Information concerning pediatric outcomes was obtained directly when the mothers had given birth at our hospital. If the mothers had received care elsewhere, this information was elicited by postal questionnaire, which was included in a letter sent to patients’ doctor to be returned to our institution for data collection. Doctors were asked to report on the following: (1) the results of CMV isolation from urine in the newborn during the first week of life, if performed, and (2) their child’s health status. Outcomes of interest were the incidence of SGA infants and long-term sequelae in case of symptomatic CMV infection. Infants were defined as SGA when their birth weight was less than the 10th centile, and appropriate for gestational age (AGA) when their birth weight was more than the 10th centile, according to Italian charts used in our hospital.17 The following were considered long-term sequelae: hearing impairment, including deafness (complete unilateral or bilateral hearing loss) and hearing loss (any clinically reduced hearing not defined as deafness), and neurodevelopmental delay, including motor, speech, and learning disabilities. Descriptive statistical analysis was performed by routine tests. Results are presented as mean (± standard deviation) for continuous variables or percentage for categorical variables (n,%). Continuous and categorical variables were compared using sample t-test and Fisher’s exact test, respectively. Statistical significance was defined as a p-value < 0.05. A power analysis was performed. Given an incidence of SGA neonates of about 10% in the general population, 115 case patients (infected babies) and 460 controls (uninfected babies) were Prenatal Diagnosis 2014, 34, 765–769
needed to have 80% power to detect a 50% increase in the rate of SGA at birth.
RESULTS Between January 2000 and June 2012, 848 women with primary maternal CMV infection were referred to our division for prenatal counseling: 157 patients (18 multiple pregnancies, 48 patients treated with CMV-specific hyperimmune globulin, 37 miscarriages and 54 therapeutic abortions in the first or second trimester) were excluded, whereas 103 patients were lost to follow-up. Therefore, the study population comprised 588 live-born infants. Data on the study population are summarized in Table 1. Congenital CMV infection was diagnosed in 119 cases (20%). No difference was found between uninfected and infected babies, except for estimated time of onset of maternal infection (10.8 ± 7.5 weeks versus 15.1 ± 8.82 weeks, respectively, pvalue = 0.0001). Thirty-five out of 588 neonates were SGA (6%), 27 (5.7%) and 8 (6.7%) among uninfected and infected babies, respectively (p-value = 0.69). Gestational age at the estimated time of maternal infection was similar in CMVinfected SGA and AGA infants (11 ± 6.9 weeks versus 15.4 ± 8.89 weeks, respectively, p-value = 0.16). Three hundred and fifty-five patients underwent amniocentesis to diagnose fetal infection. A high viral load in the amniotic fluid was found in 28 cases; eight of these infected infants were symptomatic. Table 1 lists details on birth weight centiles according to neonatal infection status. Abnormal findings were detected on mid-trimester ultrasound in 29 babies (Table 2). Two infants were SGA: one uninfected baby presented cerebral ventriculomegaly, whereas bowel hyperechogenicity was diagnosed in the infected neonates. As regards maternal complications, 11 out of 588 pregnancies were complicated by hypertensive disorders (1.8%), gestational hypertension in six cases, preeclampsia in four and HELLP syndrome in one case. Only two of these patients had a CMV-infected baby: neither was SGA at birth or had symptoms of disease at follow-up.
Table 1 Study population characteristics Uninfected infants (n = 469)
Infected infants (n = 119)
p-value
31 ± 5
31 ± 4
1.00a
219 (46.7%)
53 (44.5%)
0.67b
Week of infection
10.8 ± 7.5
15.1 ± 8.82
0.0001a
Gestational age at delivery (weeks)
39.15 ± 2.28
38.94 ± 1.88
0.35a
3311 ± 473
3247 ± 529
0.19b
27 (5.7%)
8 (6.7%)
0.69a
rd
7
2
0.95a
rd
20
6
0.95a
Maternal age (years) Nulliparous
Birth weight (g) SGA 3 centile th and
