Congenital cytomegalovirus infection following primary maternal infection in the third trimester

Authors

  • L Gindes,

    Corresponding author
    1. a Department of Obstetrics and Gynecology and b Central Virology Laboratory, Chaim Sheba Medical Center, Tel Hashomer, Israelc Department of Obstetrics and Gynecology, Assaf-Harofe Medical Center, Zerifin, Israeld Department of Obstetrics and Gynecology, Rabin Medical Center, Petach Tiqwa, Israele Department of Clinical Microbiology and Infectious Diseases, Hadassah Medical Center and School of Medicine, The Hebrew University, Ein-Kerem, Jerusalem, Israel
    Search for more papers by this author
  • a M Teperberg-Oikawa,

    1. a Department of Obstetrics and Gynecology and b Central Virology Laboratory, Chaim Sheba Medical Center, Tel Hashomer, Israelc Department of Obstetrics and Gynecology, Assaf-Harofe Medical Center, Zerifin, Israeld Department of Obstetrics and Gynecology, Rabin Medical Center, Petach Tiqwa, Israele Department of Clinical Microbiology and Infectious Diseases, Hadassah Medical Center and School of Medicine, The Hebrew University, Ein-Kerem, Jerusalem, Israel
    Search for more papers by this author
  • b D Sherman,

    1. a Department of Obstetrics and Gynecology and b Central Virology Laboratory, Chaim Sheba Medical Center, Tel Hashomer, Israelc Department of Obstetrics and Gynecology, Assaf-Harofe Medical Center, Zerifin, Israeld Department of Obstetrics and Gynecology, Rabin Medical Center, Petach Tiqwa, Israele Department of Clinical Microbiology and Infectious Diseases, Hadassah Medical Center and School of Medicine, The Hebrew University, Ein-Kerem, Jerusalem, Israel
    Search for more papers by this author
  • c J Pardo,

    1. a Department of Obstetrics and Gynecology and b Central Virology Laboratory, Chaim Sheba Medical Center, Tel Hashomer, Israelc Department of Obstetrics and Gynecology, Assaf-Harofe Medical Center, Zerifin, Israeld Department of Obstetrics and Gynecology, Rabin Medical Center, Petach Tiqwa, Israele Department of Clinical Microbiology and Infectious Diseases, Hadassah Medical Center and School of Medicine, The Hebrew University, Ein-Kerem, Jerusalem, Israel
    Search for more papers by this author
  • and d G Rahav e

    1. a Department of Obstetrics and Gynecology and b Central Virology Laboratory, Chaim Sheba Medical Center, Tel Hashomer, Israelc Department of Obstetrics and Gynecology, Assaf-Harofe Medical Center, Zerifin, Israeld Department of Obstetrics and Gynecology, Rabin Medical Center, Petach Tiqwa, Israele Department of Clinical Microbiology and Infectious Diseases, Hadassah Medical Center and School of Medicine, The Hebrew University, Ein-Kerem, Jerusalem, Israel
    Search for more papers by this author

  • The first five affiliations are affiliated with the Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel.

Dr L Gindes, Department of Obstetrics and Gynecology, The Chaim Sheba Medical Center, 52621, Tel Hashomer, Israel. Email gindesl@zahav.net.il

Abstract

Objective  To determine the effect of primary cytomegalovirus (CMV) infection in the third trimester on fetal outcome.

Design  Observational study.

Setting  Four perinatal departments in tertiary hospitals in Israel.

Population  Twenty-eight women with primary CMV infection acquired after 25 weeks of gestation.

Methods  Prenatal evaluation included amniocentesis and ultrasonographic examinations. Maternal infection was determined from seroconversion and presence of low avidity anti-CMV immunoglobulin G after 25 weeks of gestation. Fetal CMV infection was diagnosed from CMV isolated or CMV DNA amplified from the amniotic fluid. Neonatal infection was established from CMV presence in their urine or anti-CMV IgM was in their peripheral blood immediately after birth. All liveborn neonates underwent cerebral ultrasonography, hearing assessment, and psychomotor development evaluation. Infected neonates were followed up for a median of 36 months (range 6–36 months).

Main outcome measures  Intrauterine CMV infection and neonatal CMV disease throughout follow up.

Results  Vertical transmission of CMV was documented in 21 (75%) of the 28 pregnancies. None of the 20 live infected newborn had symptomatic congenital infection. One pregnancy was terminated at 34 weeks following evidence of prenatal infection. Most of the patients (75%) had CMV serology test due to clinical signs of CMV disease.

Conclusions  Although CMV infection during the third trimester of pregnancy is highly transmissible, sequelae were not found among infected offspring.

Introduction

Cytomegalovirus (CMV) is the most common cause of congenital viral infection, occurring in approximately 1% of all newborns.1 Primary CMV infection occurs during pregnancy in 2% of women, with intrauterine transmission in approximately 40% of the cases. Approximately 10% of the infected liveborn neonates have symptomatic disease at birth. In addition, between 10 and 15% of the asymptomatic newborns develop late sequelae, particularly sensorineural hearing loss and neurodevelopmental disorders.2–4 CMV infections acquired early and late in gestation have similar rates of in utero transmission,5 but little is known about the relationship between the gestational age at the primary maternal infection and the outcome of the fetal infection. It is presumed that infections acquired in late pregnancy present with less prominent signs in the neonates because neuronal growth and migration are completed at about 26 weeks of gestation, and therefore, they are not affected by late pregnancy infections.6

Primary CMV infections during pregnancy are generally asymptomatic, and therefore, it is difficult to define the time of onset of maternal infection.2 In recent years, supplementary serologic assays, such as the immunoglobulin G (IgG) avidity test, have been demonstrated as helpful in determining the onset of infection.7

Congenital CMV infection is relatively a common cause of hearing loss and mental retardation, and its antenatal diagnosis and prevention is a major challenge in perinatology. Because the relationship between timing of maternal CMV infection and fetal outcome has not been clearly delineated, the prenatal counselling of women with evidence of primary CMV infection is complicated. Therefore, the objective of our study was to determine how primary CMV infection in the third trimester of pregnancy affects fetal outcome.

Methods

All cases of primary CMV infection acquired during the third trimester of pregnancy, which were referred to the prenatal or infectious diseases clinics in four tertiary hospitals in Israel from 1998 to 2005, were enrolled in this cohort observational study. Approval of the Institutional Ethics Committee was obtained.

Criterion for inclusion was maternal primary infection during the third trimester, evidenced by seroconversion later than 25 weeks of gestation.

Demographic, clinical and pregnancy-related information was obtained. Clinical CMV was diagnosed when woman developed prolonged fever >38°C, liver function tests abnormalities, and lymphocytosis with atypical lymphocytes.

Serum CMV-specific IgG and IgM were measured by enzyme immunoassay (ImX; Abbott Laboratories, North Chicago, IL, USA or Eti-Cytok IgM; Sorin Biomedica, Vercelli, Italy). An IgG avidity assay (CMV IgG avidity EIA; Radim, Rome, Italy) was used, and avidities of less than 25% indicated recent infection that has developed in the past 12 weeks. In all cases where first positive serological test demonstrated positive IgM and negative IgG, subsequent tests demonstrated positive IgG.

Fetal diagnosis was made by amniocentesis 6 weeks or more after the maternal seroconversion. Amniotic fluid samples were inoculated into shell vials containing MRC5 monolayers for isolation of CMV.8 Polymerase chain reaction amplification of CMV DNA in the amniotic fluid samples was also performed.9 A detailed ultrasonographic examination was performed on each woman after seroconversion and every 2 weeks thereafter. Elective terminations of pregnancy required evaluation and approval by an external committee. Each neonate was evaluated by a neonatologist. Neonatal evaluations included physical examination, testing of urine for CMV by 2 weeks of age, blood cell and platelet count, measurement of liver enzymes, bilirubin, and anti-CMV IgM. The newborns underwent ophthalmoscopy, cerebral and abdominal ultrasonography, and brain-stem auditory evoked responses. Among fetuses with suspected neurological involvement, cerebrospinal fluid was obtained and electroencephalography, cerebral computed tomography, and magnetic resonance imaging were performed. Symptomatic congenital CMV disease was defined by fetal or infant death or by neurological involvement, including microcephaly, periventricular calcifications, cerebral dysplasias, seizures in an infant with CMV DNA in cerebrospinal fluid, ventricular and subependymal abnormalities, chorioretinitis, or auditory impairment. Intrauterine growth restriction was defined by head and abdominal circumferences that were below the 10th percentile for fetuses of similar age. Infected neonates were followed up for a mean of 32 months (range 6–36 months). CMV disease at 2–3 years of age was defined by mental retardation (IQ below 70) or motor delay, auditory or visual impairment. The Bayley Scale for Infant Development, 2nd edn (BSID-II), and the Wechsler Preschool and Primary Scale of Intelligence, 3rd edn (WPPSI-III) were used for evaluation. During the first 2 years of life, the children’s hearing was assessed by brain-stem auditory evoked responses. The threshold for normal hearing was defined as 0–20 dB. Abnormal responses were defined as mild (threshold, 21–45 dB), moderate (threshold, 46–70 dB), or severe (threshold, at least 71 dB).

Odds ratios and their 95% CIs were calculated and were compared by the Fisher’s exact test

Results

Twenty-eight pregnant women were eligible and enrolled in the study (Table 1). Twenty-one (75%) underwent serologic tests following clinical signs of CMV disease, which included fever, lymphadenopathy, and disturbed liver functions. Symptoms began between 25 and 37 weeks of pregnancy (mean 28.9 weeks). Five seronegative women underwent serial CMV serology testing that had been performed on a monthly basis, and two seronegative women were tested due to exposure to a sick household. The median gestational age upon seroconversion was 30 weeks (26–37 weeks). None of our cases was diagnosed due to ultrasonographic findings suggestive of congenital CMV infection.

Table 1.  Detailed information of each case participates in the study
No.Clinical CMV*Week of symptoms during pregnancyWeek of last negative serologyWeek of seroconversion**Week of first positive serologyResults of first positive serology***AmniocentesisCMV in neonatal urineNeonatal CMV IgMComments
  • ND, not done; TOP, termination of pregnancy.

  • *

    Clinical CMV: prolonged fever >38°C, hepatitis, lymphocytosis with atypical lymphocytes.

  • **

    Seroconversion is the appearance of IgM followed by that of IgG. The avidity assist in the calculation of the timing of seroconversion.

  • ***

    In all cases where first positive serological test demonstrated positive IgM and negative IgG, subsequent tests demonstrated positive IgG.

  • ****

    Pregestational serology was performed before current pregnancy.

1Yes37273737IgG positiveNDNegativeNegative 
IgM positive
Avidity 13%
2Yes32323333IgG negativeNDPositiveNegative 
IgM positive
3Yes30273131IgG positivePositiveNDND 
IgM positive
4Yes29253030IgG positivePositivePositiveND 
IgM positive
Avidity 9%
5Yes282429–3032IgG positiveNDNegativeND 
IgM positive
Avidity 13%
6Yes27272828IgG negativePositivePositiveND 
IgM positive
7No262526–2730IgG positivePositivePositiveNDRoutine follow-up serology for seronegative, fatigue at 26 weeks
IgM positive
Avidity 21%
8No 252627IgG positiveNDPositivePositiveRoutine follow-up serology for seronegative
IgM positive
 
9No 2627–2931IgG positivePositivePositiveNDRoutine follow-up serology for seronegative
IgM positive
Avidity 18%
10No 183035IgG positiveNDPositivePositiveExposure to a child with CMV at 30 weeks
IgM positive
Avidity 23%
11No 283434IgG positivePositiveTOPTOPRoutine follow-up serology for seronegative
IgM positive
 
12Yes27123131IgG positiveNDNegativePositive 
IgM positive
13Yes28112828IgG positiveNDNegativeNegative 
IgM positive
 
14Yes26112626IgG negativePositivePositiveNegative 
IgM positive
15Yes30103133IgG positivePositivePositiveNegative 
IgM positive
Avidity 7%
16Yes3083131IgG positiveNDPositiveNegative 
IgM positive
17Yes2983030IgG positiveNegativeNDNegative 
IgM positive
Avidity 5%
18Yes2882830IgG negativeNDPositiveNegative 
IgM positive
19Yes2782832IgG positivePositivePositiveND 
IgM positive
20Yes25–2652626IgG negativePositivePositiveND 
IgM positive
21Yes29Pregestation****3333IgG negativeNDNegativeNegative 
IgM positive
22Yes30Pregestation3232IgG negativeNDNegativeNegative 
IgM positive
23Yes30Pregestation3131IgG positiveNDNegativePositive 
IgM positive
Avidity 9%
24Yes33Pregestation3435IgG positiveNDPositiveNegative 
IgM positive
25Yes26Pregestation2728IgG positivePositivePositivePositive 
IgM positive
26Yes25Pregestation2727IgG positiveNegativeNegativeNegative 
IgM positive
27No Pregestation2728IgG positivePositivePositivePositiveExposure to husband with CMV at 26 weeks
IgM positive
28No 2829–3234IgG positiveNDNegativePositiveRoutine follow-up serology for seronegative
IgM positive

Fourteen women underwent amniocentesis, while 14 elected to deliver without further testing. Vertical transmission was detected in 21 of the 28 pregnancies (75%): in 12 of the 14 women who underwent amniocentesis (85.7%) and in 9 of the 14 who elected to deliver without amniocentesis (64%) (OR 4.5, P = 0.065, 95% CI 0.93–42.8). Twenty-seven women elected to continue their pregnancies and delivered at term or near term (36–42 weeks). One woman elected to terminate her pregnancy in week 34 following detection of CMV DNA in the amniotic fluid. All women delivered healthy infants without CMV disease during the follow-up period (average 33 months, median 36 months). All had normal hearing and developmental evaluations.

Two women with evidence of vertical transmission had abnormal ultrasound findings. One had oligohydramnios at 36 weeks of gestation and delivered a healthy infected child 2 weeks later. Another woman, who had CMV disease at week 26, had false-positive ultrasonographic features compatible with microcephaly that disappeared on follow-up evaluation and gave birth to a healthy newborn that developed uneventfully during the follow-up period.

Discussion

In the present study, we found a transmission rate of 75% following third-trimester CMV infection without any sequelae in the infected offspring. In previous studies, a definite association was noted between the gestational age of maternal infection and the outcome of the fetus—almost all infants with symptomatic congenital infection were exposed in the first half of pregnancy.5,10–12 However, few cases of congenital disease were reported following third-trimester maternal infection, presenting mainly with hearing loss.11,13 Severe neurological damage following maternal infection in the third trimester was described in seven infants.14 The authors assumed that these children were infected in the third trimester of gestation, but all children were referred for investigation of neurological and developmental problems when they were between 10 months and 11 years old. In only one child, serological evidence of intrauterine CMV infection at gestational week 27 existed, but information on the exact timing of seroconversion during pregnancy was missing.

Contrary to our findings, Stagno et al.5 found no difference in the transmission rate when they compared pregnancies where maternal CMV infection occurred at between weeks 4 and 22, 16 and 27, and 23 and 40. Others reported transmission rates between 58 and 77% in late gestational periods.12,15,16

It is difficult to study the effect of gestational age on the outcome of congenital CMV because it is often impossible to time the onset of the maternal infection. When the first test is performed at week 28, IgM antibodies against CMV could indicate first-, second- or third-trimester infection. To evaluate the effect of genuine third-trimester CMV infection, only women with documented clinical disease or a precise date of IgG CMV seroconversion were included in our study; this accounts for the small number of women enrolled.

Seventy-five percent of our cohort had clinical disease, much higher than the reported rate of 5%.10,17,18 Due to insistence of pregnant women and the fear of litigation, most obstetricians test pregnant women for CMV during the first trimester. Testing for CMV during the third trimester, however, is usually performed only after suspected clinical signs of disease.

Half of the women underwent amniocentesis, but although infection was demonstrated in 85.7% of them, all but one elected to continue their pregnancy. Israeli law does not entirely prevent abortion at any stage of pregnancy, but the degree of permissiveness differs between trimesters of pregnancy (Ministry of Justice, 1977). Until week 23, elective termination of pregnancy is permissible if one or more of the following conditions is present: the pregnant woman is a teenager, over 40 years, or not married; the woman has severe emotional problems; the woman’s health is at risk; or there is evidence of fetal abnormalities. The decision to terminate pregnancy is made by hospital committees composed of a social worker and a gynaecologist. However, after week 23 of pregnancy, the regulation of pregnancy termination becomes more restrictive, and termination is limited to medical reasons when approaching the third trimester. The Israeli Ministry of Health has authorised six hospitals to create high committees for the elective termination of late pregnancy. These committees are larger and multidisciplinary and are made up of the highest level administrators: the director of the Medical Centre, the director of the Obstetrics and Gynaecology Department, the director of the Neonatal Department, the director of the Genetic Counselling Department, and the director of the Social Work Department. Nevertheless, the rate of late termination of pregnancy in Israel is 5–10 times higher than that reported for countries such as Denmark, the UK, Canada, and the USA, where policies are relatively more restrictive.19

Ultrasonography has limited sensitivity in detection of fetal infection.20 Furthermore, since ultrasonographic abnormalities develop dynamically several weeks after the viral transmission,21 the yield of such tests in third-trimester infections is low. Interestingly, in our study, the CMV intrauterine transmission rate seemed to increased with advancing stage of pregnancy, which is in accordance with the data by Revello and Gerna22 who reported transmission rates of 45.4% (49/108), 45.6% (21/46), and 78.6% (11/14) following primary infections in the first, second and third trimesters, respectively. Higher rates of CMV transmission during late gestation were also reported by Daiminger et al.12 and by Bodeus et al.16 An increase in congenital infection rates with advancing pregnancy is known for acute maternal infection with rubella virus, varicella-zoster virus, and Toxoplasma gondii when increased placental blood flow facilitates inoculation of the placenta. In contrast, the severity of congenital disease is related inversely to gestational age and is much greater when infections occur in the first trimester. Sequelae of fetal infection are worse during organogenesis. As the duration of pregnancy increases, fetal infections are less likely to cause congenital malformations.22 Although passive immunisation may alter the natural course of CMV during pregnancy,23 the possible efficacy of hyperimmune globulin therapy for maternal infection in the last trimester is uncertain.

The major limitation of this study is the small number, so it may not be adequately powered to detect low rates of neonatal consequences of CMV infection. However, the exact timing of seroconversion during the third trimester was possible only in those with symptomatic CMV disease during the third trimester or among those who performed serology every 2–3 weeks, causing diminution of the group.

Conclusion

Although third-trimester CMV infection carries a high transmission rate, it has no deleterious effect on the fetal outcome. Although the study is limited by its small size and its observational design, we think that performing amniocentesis for the diagnosis of third-trimester intrauterine CMV infection is an unnecessary invasive procedure.

Contribution to authorship

L.G. collected all the information of the patients from medical files, interviewed the patients, analysed statistically, and wrote the paper. M.T.O. carried out the CMV laboratory tests, identified the patients suspected to have primary CMV infection during the third trimester. D.S. added information of his patients. J.P. added information of his patients. G.R. collected all the information of the patients from medical files, interviewed the patients, and assisted in writing the paper.

Ancillary