Objective
The objective of the study was to evaluate the vertical transmission rate and fetal risk following primary maternal cytomegalovirus infection before and around conception.
Study Design
Data of patients referred to fetal medicine clinic in 1 tertiary center in Israel were evaluated. Each included subject had primary maternal cytomegalovirus infection determined by serology, precise gestational dating, and testing of fetal infection. Subjects were assigned to five subgroups: pregestational, periconception, and first, second, or third trimester of pregnancy.
Results
Five hundred eight pregnancies were included. None of the 97 pregnancies in the preconception group and 6 of the 130 periconception subjects (4.6%) were congenitally infected. Transmission rates were 34.8%, 42.0%, and 58.6% for the first, second, and third trimesters, respectively ( P = .049). Prenatal and postnatal follow-up indicated that third-trimester infection has no clinical effect on the fetus.
Conclusion
Pre- and periconception maternal infection carries small risk for fetal infection, whereas it is positively correlated to time of maternal infection during pregnancy.
Cytomegalovirus (CMV) is the leading cause of fetal infection, and the consequential congenital disease is a major medical problem. Each year in the United States alone, hundreds of children die and thousands develop permanent disabilities as a result of congenital CMV infection. More children may be affected by congenital CMV than by any other better known congenital conditions such as Down syndrome or neural tube defects. Therefore, prenatal CMV diagnosis and prevention of its associated fetal infection and disease is a major challenge in perinatology.
Although recurrent maternal CMV infection can occur, the risk of vertical transmission is much higher following primary maternal CMV infection. Many published studies have documented the epidemiology of congenital CMV following primary maternal infection. Nevertheless, most large studies are focused on maternal infection during the first half of pregnancy. Available data on fetal outcome following prepregnancy, periconception, and late pregnancy maternal infection is based on very small study groups.
We summarized the data and evaluated the risk of fetal infection and congenital disease of 508 pregnancies complicated by primary maternal CMV infection before conception or during the different stages of pregnancy. The aim of the present study was to compare the vertical transmission rate and fetal outcome following pre- and periconception primary maternal CMV infection with those following infection at different gestational trimesters.
Materials and Methods
Our study group included pregnant patients who were referred to a specialized fetal medicine clinic at the Sheba Medical Center in Israel for counseling and prenatal diagnostic workup for primary maternal CMV infection from January 2000 to December 2006.
Although there is no formal CMV screening program in Israel, it is a common practice of most obstetricians to advise patients to undergo CMV screening before planned pregnancy or during the first trimester of an ongoing pregnancy. CMV serological screening was performed with different commercially available kits for specific anti-CMV immunoglobulin G (IgG) and immunoglobulin M (IgM) as well as an IgG avidity test in IgG-positive patients. IgG seronegative women were reassessed during pregnancy. Patients considered screen positive by their primary care obstetricians were referred for further prenatal investigation and follow-up. Twelve patients had also minor, mostly unrelated, ultrasound findings or unspecific maternal symptoms described in the referral letter. However, each study subject had CMV infection documented by serologic assays.
All referred patients were primarily interviewed and counseled by fetal medicine specialist. Medical and obstetrical history as well as the detailed course of the ongoing pregnancy was documented. Precise gestational dating was determined by the first day of the last menstrual period (LMP) and ultrasonography at the first trimester.
Diagnosis of primary maternal CMV infection was exclusively determined by IgG seroconversion (the appearance of de novo-specific IgG antibodies in a previously seronegative patient) or positive specific anti-CMV IgG and IgM antibodies associated with low IgG avidity.
Timing of maternal infection was very carefully related to gestational age. Each patient was assigned to 1 of the 5 following subgroups: pregestational (12 months to 8 weeks before conception), periconception (between 8 weeks before and 6 weeks after conception), first trimester (up to full 13 weeks), second trimester (up to full 26 weeks), and third trimester (gestational age beyond 26 weeks).
Clear-cut differentiation between gestational and pregestational maternal infection is desired, however impossible in many around-conception cases. Therefore, we chose to define a group of periconception. Patients were assigned to first-trimester maternal infection when the serological data clearly supported it (IgG seroconversion during the first trimester or positive IgG and IgM associated with low IgG avidity beyond 8 weeks after conception). Patients were assigned to the periconception group when the serology was highly indicative of infection during the time period of 8 weeks before and 6 weeks after conception (IgG seroconversion) or when the data were not clearly indicative of first-trimester infection (low IgG avidity before 8 weeks after conception).
Following the initial evaluation and counseling, all patients were offered to participate in our prenatal CMV diagnosis program, which includes amniocentesis and repeated detailed ultrasonographic evaluation. Amniocentesis was performed after 21 weeks from the LMP, allowing an interval of at least 7 weeks between the estimated date of maternal infection and the date of the invasive procedure. Transabdominal ultrasound-guided amniocentesis was performed using a 21-gauge needle to collect 30 mL of amniotic fluid for CMV assays and fetal karyotyping. Infectious virus was detected in amniotic fluid samples by rapid virus isolation in cell cultures (shell vial procedure), whereas the presence of viral genome was determined by polymerase chain reaction.
Patients with proven fetal infection had a postamniocentesis counseling session. The risks of fetal CMV infection were discussed, and patients were offered to participate in a pre- and postnatal follow-up program. Serial detailed ultrasound examinations were performed every 3-4 weeks until delivery. Magnetic resonance imaging examination directed to identify fetal central nervous system pathology associated with CMV infection was also performed in these cases since August 2004.
Neonatal urine was tested after birth for CMV during the first 7 days of life to determine whether congenital CMV infection is present. All neonates with positive urine culture or prenatally documented CMV infection had a fundus examination, hearing evaluation, and brain ultrasound scan during the first few days of life. Results of these tests were obtained from hospital charts. The parents were referred for long term follow-up including a physical examination and hearing test in a pediatric infectious diseases unit at a near-home tertiary medical center. Information on the long-term follow-up was obtained from telephone interviews of the parents.
Data were prospectively collected and retrospectively evaluated for each case. Transmission rates were compared using the χ 2 test. The study was approved by the institutional review board.
Results
Five hundred eight women were diagnosed with primary maternal CMV infection, matched the inclusion criteria, and were included in the study group. All patients were referred to the fetal medicine clinic at the Sheba Medical Center between January 2000 and December 2006.
These 508 patients included 16 twin pregnancies, which summarize the number of cases included in the present study group to be 524 fetuses.
All 508 pregnancies were assigned to 1 of the 5 study groups based on gestational age at the time of maternal infection. Ninety-seven pregnancies were assigned to the pregestation group, 130 to the periconception group, 152 to the first-trimester group, 100 to the second-trimester group, and 29 to the third-trimester primary maternal CMV infection group.
Amniocentesis was performed in 485 of the 524 patients (92.6%), whereas diagnosis of congenital CMV infection was based on a postnatal CMV urine test alone in the remaining 39 neonates (7.4%). Amniocentesis and urine test results were both available in 379 of the 446 live-born neonates (85.0%).
The vertical transmission rate is presented in Table 1 . Each twin pregnancy was considered as 1 data point for the calculation of transmission rates. Both fetuses of 14 sets of twins were negative (11 pairs) or positive (3 pairs) for CMV infection. Two twin pregnancies were considered as positive transmission, although 1 twin was positive and the other was negative for CMV.
| Study group | Pregnancies, n a | CMV positive pregnancies, n | Transmission rate, % |
|---|---|---|---|
| Pregestation | 97 | 0 | 0 |
| Periconception | 130 | 6 | 4.6 |
| First trimester | 152 | 53 | 34.8 |
| Second trimester | 100 | 42 | 42.0 |
| Third trimester | 29 | 17 | 58.6 |
| All cases | 508 | 118 b | 23.2 |
| Gestational cases c | 281 | 112 | 39.9 |
a Each of the 16 twin pregnancies is considered as 1 data point;
b Each of the 2 twin pregnancies with 1 positive twin and 1 negative twin was calculated as 1 positive data point;
The overall transmission rate in the present study group was 23.2% (118 of 508). The vertical CMV transmission rate of primary maternal infection during pregnancy was 39.9% (112 of 281). None of the 97 fetuses in the pregestation group was infected, whereas the virus was vertically transmitted in 6 of the 130 pregnancies included in the periconception study group (4.6%). The transmission rate was significantly correlated to gestational age at the time of maternal infection (34.8%, 42.0%, and 58.6% for first second and third trimesters, respectively, P = .049 by χ 2 ).
Of 121 diagnosed positive fetal infection cases, 105 underwent amniocentesis, whereas the other 16 were diagnosed by postnatal urine culture alone. Two of the 105 amniocentesis subjects have shown negative results on amniotic fluid samples but had positive urine cultures for CMV infection immediately after birth. These 2 false-negative amniocenteses were performed strictly according to our protocol, at a gestational age beyond 21 weeks and also allowing at least 7 weeks from the time of the maternal infection.
We also found that rapid virus isolation in cell cultures (shell vial) and polymerase chain reaction for the detection of the viral genome share very high sensitivity and specificity because both were concurrently positive or negative in all studied cases.
Abnormal findings of detailed ultrasound imaging in 17 infected fetuses are described in Table 2 . Our results indicate that severe disease may affect fetuses infected in the first as well as the second trimester, whereas no findings were identified in fetuses affected later in pregnancy.
