Prenatal diagnosis of fetal cytomegalovirus infection Lauren Lynch, MD,. Fernand Daffos, MD: David Emanuel, MD,c Yves Giovangrandi, MD: Richard Meisel, MD,. Francois Forestier, PhD: Gieri Cathomas, MD, and Richard L. Berkowitz, MD' C
New York, New York, and Paris, France Twelve fetuses were evaluated with a combination of ultrasonography, amniocentesis, and blood sampling for possible cytomegalovirus infection. In seven the mother had a documented primary cytomegalovirus infection. All seven women had normal ultrasonographic findings and one fetus was found to be infected. In the other five cases fetal cytomegalovirus infection was diagnosed in association with abnormal utlrasonographic findings. There was no history of maternal infection in the latter group. All positive and negative diagnoses were confirmed and none of the six infected fetuses survived. In this series, the most reliable parameters of infection were the isolation of the virus from amniotic fluid and elevations of total immunoglobulin M and -y-glutamyl transpeptidase in fetal blood. In the majority of infected fetuses cytomegalovirus-specific immunoglobulin M was not detected in blood. Prenatal diagnosis of fetal cytomegalovirus infection is possible with a combination of amniocentesis and fetal blood sampling. (AM J OSSTET GVNECOL 1991 ;165:714-8.)
Key words: Cytomegalovirus, prenatal diagnosis, fetus
Cytomegalovirus is the most common cause of intrauterine infection, affecting an estimated 1% of all live births in the United States. 1 Fetal infection may follow either primary or reactivated maternal infection, although symptomatic congenital disease is more likely to occur in association with primary maternal infection. 2 A total of 30% to 40% of pregnant women with primary cytomegalovirus infection transmit the virus to their fetuses. 3 Although only 10% of those infected are symptomatic at birth, 20% to 30% die and late complications almost invariably develop among the survivors. In addition, long-term neurologic sequelae develop later in childhood in 5% to 15% of congenitally infected infants without symptoms! The natural history of intrauterine cytomegalovirus infection is not well understood, but it is clear that some fetuses are irreversibly damaged by the virus before delivery. Those infants would not derive much benefit from postnatal therapy, but if infected fetuses could be detected before this irreversible stage has been reached, treatment in utero might have a significant effect on the course of the disease. The recent development of effective antiviral agents with in vitro and in vivo activity against human cytomegalovirus may make such therFrom the Department of Obstetrics, Gynecology and Reproductive Science, Mount Sinai Medical Center," the Department of Fetal Medicine, Institut de Puericulture de Paris,' and the Department of Pediatrics, Memorial Sloan-Kettering Cancer Institute.' Received for publication November 16, 1990; accepted February 27, 1991. Reprint requests: Lauren Lynch, MD, Department of Obstetrics, Gynecology and Reproductive Science, Mount Sinai Medical Center, One Gustave L. Levy Place, Box 1171, New York, NY 10029. 611129143
714
apy possible.' Daffos et al. 6 have shown that in utero treatment of fetuses infected with toxoplasmosis significantly improves the outcome in those cases. However, before consideration can be given to treatment of cytomegalovirus in utero, it is necessary to be able to accurately diagnose fetal infection with this organism. In addition to traditional fibroblast cultures, new methods including the use of monoclonal antibodies, deoxyribonucleic acid (DNA)-ribonucleic acid (RNA) hybridization and DNA amplification by the polymerase chain reaction have been applied to the rapid diagnosis of cytomegalovirus infection. The sensitivity and specificity of these methods in the detection of fetal infection remains to be determined. This article describes our experience with the prenatal diagnosis of cytomegalovirus infection in 12 fetuses with either exposure to documented primary maternal infection during pregnancy or abnormal ultrasonographic results suggestive of fetal cytomegalovirus infection.
Material and methods Patients at risk of fetal cytomegalovirus infection were evaluated at the Mount Sinai Medical Center or the Institut de Puericulture de Paris between 1982 and 1989. Seven women were seen because of documented primary maternal cytomegalovirus infection, and five additional patients were investigated because of abnormal ultrasonographic findings suggestive of fetal infection. All patients underwent detailed ultrasonographic examinations and 11 of the 12 also had percutaneous fetal blood sampling and simultaneous amniocentesis. Amniotic fluid was cultured for cytomegalovirus. Complete blood cell count, 'Y-glutamyl
Volume 165 Number 3
transpeptidase (GGTP), total immunoglobulin M (IgM), cytomegalovirus-specific IgM, and cytomegalovirus cultures were performed on the fetal blood samples. In addition, polymerase chain reaction with cytomegalovirus-specific oligonucleotides was performed on two amniotic fluid and fetal blood samples obtained at the Mount Sinai Medical Center. All cytomegalovirus cultures, specific IgM assays, and polymerase chain reaction studies of those fetuses evaluated at the Mount Sinai Medical Center were performed in the laboratory of one of the authors (D.E.) at Memorial Sloan-Kettering Cancer Center. Cytomegalovirus-specific IgM was determined by enzyme-linked immunosorbent assay (ELISA). Cytomegalovirus isolation was performed in fibroblast culture by indirect immunofluorescence staining with cytomegalovirusspecific monoclonal antibodies. Results were generally available within 48 hours after innoculation. The polymerase chain reaction studies on amniotic fluid and fetal blood were performed by previously described techniques. 7 We used oligonucleotide primers flanking a 421 base pair sequence of the cytomegalovirus immediate-early gene and an oligonucleotide probe to an internal sequence of the amplified sequence. Positive and negative controls were analyzed simultaneously with each study sample. Specific details regarding the polymerase chain reaction primers and probe and the cytomegalovirus-specific monoclonal antibodies will be furnished on request. At the Institute of Puericulture de Paris, cytomegalovirus-specific IgM studies were also performed by ELISA. Cytomegalovirus cultures were done by fibroblast innoculation without immunofluorescence, and results were available within 3 to 4 days after innoculation. Complete blood counts were done by standard automatized techniques, and all abnormal platelet counts were confirmed manually. Total IgM and GGTP were assayed by standard techniques at both centers. Hematocrit values, platelet counts, and total IgM values were compared with normal fetal values for the corresponding gestational age. s GGTP levels do not change with gestational age, and the values in this series were compared with known fetal norms (24.4 ± 9.6 UI/L)." Fetal blood sampling was done percutaneously under ultrasonic guidance, and amniotic fluid was obtained immediately after the blood had been aspirated. The fetal origin and purity of the blood was confirmed in all cases by red blood cell size analysis.
Results Primary infections. Seven pregnancies were studied because of documented primary infections in the mothers (Table I, patients 1 to 7). All these women had
Prenatal diagnosis of cytomegalovirus infection
715
clinical manifestations of cytomegalovirus infection, and cytomegalovirus-specific IgM present in their serum. The infections all occurred during the first half of pregnancy (range, 11 to 18 weeks), and the patients were referred for fetal assessment between 21 and 32 weeks' gestation. One of these fetuses was found to be infected with cytomegalovirus. This case (Table I, No. 1) has been previously reported. lO Briefly, the mother had fever, neurologic symptoms, and liver dysfunction at 11 weeks' gestation and she was shown to have cytomegalovirus-specific IgM. At 21 weeks' gestation ultrasonographic examination revealed an appropriately grown fetus and no structural abnormalities or evidence of hydrops. Fetal blood studies revealed a normal hematocrit value and platelet count, and cytomegalovirus-specific IgM was absent. The GGTP value was elevated, however, and cultures of the amniotic fluid and fetal blood were positive for cytomegalovirus. This pregnancy was terminated. Autopsy revealed a grossly normal fetus with histologic evidence of nonspecific acute infection (white cell infiltrates in the lung and liver). There were rare inclusion cells suggestive of cytomegalic inclusion disease. Immunofluorescence tissue studies were positive for cytomegalovirus in the fetal kidney, lung, adrenal, and placenta. The six remaining fetuses of mothers with documented infection were sampled between 24 and 32 weeks' gestation. All were structurally normal on ultrasonographic examination. The fetal hematocrit, platelet count and total IgM levels were normal. Specific cytomegalovirus IgM and viral cultures of amniotic fluid and fetal blood were negative in all cases. These fetuses were delivered at term, and all except one were appropriately grown. None had manifestations of cytomegalovirus infection. Specific IgM and viral cultures were negative in all neonates. Abnormal ultrasonographic findings. Five additional fetuses were studied because of abnormal ultraso no graphic findings and all were infected. None of these five fetuses survived; three pregnancies were terminated and the two live-born infants died within a few days (Table I, patients 8 to 12). Patient No.9 did not undergo amniocentesis or fetal blood sampling. In this instance the pregnancy was terminated because of the ultrasonographic findings and the diagnosis was made retrospectively by culturing tissues from the abortus. In all five cases cytomegalovirus infection was confirmed by isolation of the virus from fetal tissues after termination of pregnancy or in the urine of live-born infants during the first week after birth. All the mothers were cytomegalovirus IgG seropositive but none had cytomegalovirus-specific IgM. Therefore it is likely that these cases were a result of recurrent infection, although a primary infection earlier during the pregnancy cannot be ruled out. One case was confirmed to
716
Lynch et al.
September 1991 Am J Obstet Gynecol
Table I. Patient data
Patient No.
1*
Reason for investigation
Type of infection
Maternal IgM
Gestational age at PUBS (wk)
+
21
+
24-32
Primary CMV infection at 13 wk Maternal infection during pregnancy
Primary
Abnormal ultrasonographic findings at 21 wk Abnormal ultrasonographic findings at 16 wk
Recurrent
22
Unknown
ND
lOt
Fetal growth retardation at 30 wk
Unknown
33
llt
Abnormal ultrasonographic findings at 31 wk Abnormal ultrasonographic findings at 30 wk
Unknown
31
Unknown
30
2-7t
8*
9*
12*
Primary
(all cases)
Fetal blood Ultrasonogmphic findings
ITotallgM ICMV IgM
Platelet count
GGTP
Normal
N
t
N/A
Normal anatomy, one mild growth retardation Ascites, dilatation of posterior horns of lateral ventricles Oligohydramnios, dilatation of posterior horns of lateral ventricles, bowel echodensity Oligohydramnios, fetal growth retardation
N
N
N
t
t
NA
NA
NA
N
NA
t
N
t
t
N
t
t
Oligohydramnios, cerebral calcification, ventricular dilatation Severe hydrops
NA
+
PUBS, Percutaneous umbilical blood sampling; CMV, cytomegalovirus; NA, not available; N, normal; AF, amniotic fluid; TOP, termination of pregnancy; +, positive; -, negative; PCR, polymerase chain reaction; ND, not done. *Evaluated at Mount Sinai Medical Center. tEvaluated at the Institut de Puericulture de Paris.
be a recurrence because at 10 weeks' gestation the mother was found to have cytomegalovirus-specific IgG and no IgM. The ultrasonographic findings in this group were varied. The most common abnormalities were ventricular dilatation (three cases) and oligohydramnios (three cases). One fetus (patient No. 10) was thought to be growth retarded, and this was the only one with a birth weight below the 10th percentile. Not only were the ultrasonographic findings diverse among the patients, but in a given fetus they sometimes changed. For example, in the fetus of patient No.8 moderate ascites at 22 weeks' gestation resolved almost completely within 1 week.
Comment Approximately 10% of infants congenitally infected with cytomegalovirus are symptomatic at birth" The most common findings are petechiae, hepatosplenomegaly, jaundice, microcephaly, small size for gestational age, inguinal hernias in males, and chorioretinitis. The mortality in this group may be as high as 30%, and 90% of the survivors have permanent disabilities such as mental retardation, neuromuscular dis-
orders, and hearing loss. Permanent sequelae, which is usually hearing loss, develop in 5% to 15% of the infected infants without symptoms, but mental retardation and severe neuromuscular disorders may also occur,,· l1 Between 30% and 40% ofinfants exposed to primary maternal cytomegalovirus infection will be congenitally infected.' It is not known why most exposed fetuses do not acquire the infection. Gestational age at the time of maternal infection does not affect the transmission rate, although it does seem to influence the severity of the infection. In the report by Stagno et al.' of 37 infected infants, all those who were symptomatic at birth and those in whom clinically significant handicaps subsequently developed were born to women with primary infection before 27 weeks' gestation. If 1% of all babies delivered in the United States are congenitally infected, and approximately 20% of these have permanent sequelae, it is obvious that cytomegalovirus is a significant public health problem. Surprisingly, there are only a few isolated reports regarding the prenatal diagnosis of this infection. 12 The most obvious explanation for this is that the great majority of primary cytomegalovirus infections are without
Prenatal diagnosis of cytomegalovirus infection
Volume 165 Number 3
PCR
CMV cultures AF
+
I
Blood
AF
+
I
Blood
Outcome
Autopsy findings
NA
NA
TOP
NA
NA
Normal fetus; histologic evidence of infection
Normal infants, one small for gestational age
+
+
TOP
Ventricular dilatation with massive necrosis of right hemisphere; placental villi tis
NA
NA
TOP
Disseminated CMV infection; ventricular dilatation with brain necrosis and calcifications; placental villi tis
+
NA
NA
Disseminated CMV infection; normal brain
+
NA
NA
Delivery at 36 wk; hepatic failure; thrombocytopenia; small for gestational age; neonatal death TOP, microcephaly, petechia, hepatosplenomegaly
+
+
Delivery at term; pulmonary hypoplasia; neonatal death
Disseminated CMV infection; normal brain
NA
NA
717
symptoms, and pregnant women are not routinely screened for cytomegalovirus antibodies. Another problem is that there is no intrauterine treatment for this disease at this time. In this series we evaluated seven fetuses exposed to documented primary maternal infection and found only one (14%) to be infected in utero. All these fetuses, including the infected one, had normal ultrasonographic findings. By combining viral cultures of amniotic fluid and fetal blood with assessment of specific cytomegalovirus IgM, total IgM, platelet count, and GGTP in fetal blood we were able to make a correct diagnosis in each of these cases. All negative cases were confirmed after birth and cytomegalovirus was isolated from the infected fetus after termination of that pregnancy. It must be noted, however, that given the normal morphologic findings in the infected fetus it is impossible to predict what manifestations if any may have occurred after birth in that case. Five infected fetuses were evaluated because of abnormal ultrasonographic findings. The most common finding in this group was cerebral ventricular dilatation and oligohydramnios. However, no single abnormality was consistently seen in all fetuses. Clearly, any fetus with ventricular dilatation, microcephaly, or cerebral calcifications should be suspected of being infected. Also, fetuses with any evidence of unexplained hydrops (ascites, pleural effusions, or skin edema) should be evaluated for cytomegalovirus infection. None of the
ND
fetuses with abnormal ultrasonographic findings and cytomegalovirus infection survived; three pregnancies were terminated and the two live-born infants died within a few days. All these fetuses were found to have severe manifestations of cytomegalovirus infection including three with central nervous system abnormalities. The other two had normal central nervous system findings but died of pulmonary hypoplasia and liver failure. Five of the six infected fetuses underwent fetal blood sampling and amniocentesis. A positive viral culture of amniotic fluid and elevated total IgM and GGTP values in fetal blood were found in all five fetuses tested. Cytomegalovirus culture of fetal blood was positive in only one of five fetuses tested. Cytomegalovirus-specific IgM was present in fetal blood in only one of the five fetuses tested, compared with rates of 50% to 90% found in cord blood of infected neonates.'" 14 The fetal immune system has been found to be capable of consistently responding to certain antigenic stimuli by 22 weeks' gestation. '5 However, production of specific IgM appears to be dependent on the organism involved. For example, it has been found that the vast majority of fetuses with congenital rubella infection produce specific IgM after 22 weeks," whereas only 15% of fetuses infected with toxoplasma tested between 24 to 29 weeks produce specific IgM." The low rate of positivity in our patients may be because two of the patients were sampled at 21 and 23 weeks' gestation, or it may be related
718 Lynch et al.
to the IgM assays used. Ahlfors et al. l3 sent serum from culture proved cytomegalovirus-infected neonates to four different laboratories for analysis. The three that used an ELISA detected cytomegalovirus-specific IgM in 63% to 80% of the cases, whereas the one with a radioimmunoassay detected positivity in only 50% of cases. Other authors 14 have reported that the radioimmunoassay is 90% sensitive. It is clear from all studies that viral cultures remain the "gold standard." In this report only ELISAs were used, but they were performed in two different laboratories. Because the number of patients is small, the true sensitivity of specific IgM analysis cannot be determined from this series, but it is evident that this parameter alone cannot be used to exclude congenital cytomegalovirus infection. In conclusion, the prenatal diagnosis of cytomegalovirus infection is possible by use of a combination of ultrasonography, amniocentesis, and fetal blood sampling. Although amniotic fluid culture for cytomegalovirus seems to be very sensitive, the number of cases in this series is not large enough to advocate use of amniocentesis as the only diagnostic approach. Furthermore, in a review Stagno et al. 16 mentioned that they were aware of three women with primary gestational cytomegalovirus infection in whom the virus was isolated from fetal or placental tissues but not from amniotic fluid. In our opinion, until more data become available, fetal blood should also be studied when evaluating a fetus for cytomegalovirus infection. Furthermore, it is possible that fetuses with hematologic evidence of systemic involvement, such as thrombocytopenia and liver damage, may have a worse prognosis than those infected without these findings. Once the diagnosis has been made, if the fetus has abnormal ultrasonographic findings this series suggests that the prognosis for intact survival appears to be very poor. However, if the ultrasonographic findings are normal the prognosis is far from clear. Until follow-up data of such cytomegalovirus-infected fetuses are available, we will have to rely on information obtained from neonates about the natural history of these infants.
September 1991 Am J Obstet Gyneco1
REFERENCES 1. Stagno S. Cytomegalovirus. In: Infectious diseases of the fetus and newborn infant. Remington JS, Klein JO, eds. Philadelphia: WB Saunders, 1990:241-81. 2. Stagno S, Pass RF, Dworsky ME, et al. Congenital cytomegalovirus infection: the relative importance of primary and recurrent maternal infection. N Engl J Med 1982;306:945-9. 3. Stagno S, Pass RF, Cloud G, et al. Primary cytomegalovirus infection during pregnancy. JAMA 1986;256:1904-8. 4. Stagno S, Whitley RJ. Herpes virus infection of pregnancy. I: cytomegalovirus and Epstein-Barr virus infections. N EnglJ Med 1985;313:1270-4. 5. Fan-Havara P, Nanata MC, Brady MT. Ganciclovir-a review of pharmacology, therapeutic efficacy and potential use for treatment of congenital cytomegalovirus infection. J Clin Pharm Ther 1989; 14:329-40. 6. Daffos F, Forestier F, Capella-Pavlovsky M, et al. Prenatal management of 746 pregnancies at risk for congenital toxoplasmosis. N EnglJ Med 1988;318:271-5. 7. Saiki RK, Gelfond DH, Stoffel S, et al. Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 1988;239:487-91. 8. Forestier F, Daffos F, Galacteros F, et al. Hematological values of 163 normal fetuses between 18 and 30 weeks of gestation. Pediatr Res 1986;4:342-7. 9. Daffos F, Forestier F. Biologie du sang foetal. In: Medicine et biologie du foetus humain. Daffos F, Forestier F, eds. Paris: Maloine, 1988. 10. Meisel RL, Alvarez M, Lynch L, et al. Fetal cytomegalovirus infection: a case report. AM J OBSTET GYNECOL 1990;162:663-4. 11. MacDonald H, Tobin J O'H. Congenital cytomegalovirus infection: a collaborative study in epidemiological, clinical and laboratory findings. Dev Med Child Neurol 1978;20:471-82. 12. Grose C, Itani 0, Weiner CPo Prenatal diagnosis of fetal infection: advances from amniocentesis to cordocentesis-congenital toxoplasmosis, rubella, cytomegalovirus, varicella virus, parvovirus and human immunodeficiency virus. Pediatr Infect Dis J 1989;8:459-68. 13. Ahlfors K, Forsgren M, Griffiths P, et al. Comparison of four serological tests for the detection of specific immunoglobulin M in cord sera of infants congenitally infected with cytomegalovirus. ScandJ Infect Dis 1987;19:303-8. 14. Griffiths PD, Kangro HO. A user's guide to the indirect solid phase radioimmunoassay for the detection of cytomegalovirus specific IgM antibodies. J Viral Methods 1984;8:271-82. 15. Daffos F, Forestier F, Grangeot-Keros L, et al. Prenatal diagnosis of congenital rubella. Lancet 1984;2:1-3. 16. Stagno S, Pass RF, Dworsky ME, Alford CA. Maternal cytomegalovirus infection and perinatal transmission. Clin Obstet Gynecol 1982;25:563-76.
New York, New York, and Paris, France Twelve fetuses were evaluated with a combination of ultrasonography, amniocentesis, and blood sampling for possible cytomegalovirus infection. In seven the mother had a documented primary cytomegalovirus infection. All seven women had normal ultrasonographic findings and one fetus was found to be infected. In the other five cases fetal cytomegalovirus infection was diagnosed in association with abnormal utlrasonographic findings. There was no history of maternal infection in the latter group. All positive and negative diagnoses were confirmed and none of the six infected fetuses survived. In this series, the most reliable parameters of infection were the isolation of the virus from amniotic fluid and elevations of total immunoglobulin M and -y-glutamyl transpeptidase in fetal blood. In the majority of infected fetuses cytomegalovirus-specific immunoglobulin M was not detected in blood. Prenatal diagnosis of fetal cytomegalovirus infection is possible with a combination of amniocentesis and fetal blood sampling. (AM J OSSTET GVNECOL 1991 ;165:714-8.)
Key words: Cytomegalovirus, prenatal diagnosis, fetus
Cytomegalovirus is the most common cause of intrauterine infection, affecting an estimated 1% of all live births in the United States. 1 Fetal infection may follow either primary or reactivated maternal infection, although symptomatic congenital disease is more likely to occur in association with primary maternal infection. 2 A total of 30% to 40% of pregnant women with primary cytomegalovirus infection transmit the virus to their fetuses. 3 Although only 10% of those infected are symptomatic at birth, 20% to 30% die and late complications almost invariably develop among the survivors. In addition, long-term neurologic sequelae develop later in childhood in 5% to 15% of congenitally infected infants without symptoms! The natural history of intrauterine cytomegalovirus infection is not well understood, but it is clear that some fetuses are irreversibly damaged by the virus before delivery. Those infants would not derive much benefit from postnatal therapy, but if infected fetuses could be detected before this irreversible stage has been reached, treatment in utero might have a significant effect on the course of the disease. The recent development of effective antiviral agents with in vitro and in vivo activity against human cytomegalovirus may make such therFrom the Department of Obstetrics, Gynecology and Reproductive Science, Mount Sinai Medical Center," the Department of Fetal Medicine, Institut de Puericulture de Paris,' and the Department of Pediatrics, Memorial Sloan-Kettering Cancer Institute.' Received for publication November 16, 1990; accepted February 27, 1991. Reprint requests: Lauren Lynch, MD, Department of Obstetrics, Gynecology and Reproductive Science, Mount Sinai Medical Center, One Gustave L. Levy Place, Box 1171, New York, NY 10029. 611129143
714
apy possible.' Daffos et al. 6 have shown that in utero treatment of fetuses infected with toxoplasmosis significantly improves the outcome in those cases. However, before consideration can be given to treatment of cytomegalovirus in utero, it is necessary to be able to accurately diagnose fetal infection with this organism. In addition to traditional fibroblast cultures, new methods including the use of monoclonal antibodies, deoxyribonucleic acid (DNA)-ribonucleic acid (RNA) hybridization and DNA amplification by the polymerase chain reaction have been applied to the rapid diagnosis of cytomegalovirus infection. The sensitivity and specificity of these methods in the detection of fetal infection remains to be determined. This article describes our experience with the prenatal diagnosis of cytomegalovirus infection in 12 fetuses with either exposure to documented primary maternal infection during pregnancy or abnormal ultrasonographic results suggestive of fetal cytomegalovirus infection.
Material and methods Patients at risk of fetal cytomegalovirus infection were evaluated at the Mount Sinai Medical Center or the Institut de Puericulture de Paris between 1982 and 1989. Seven women were seen because of documented primary maternal cytomegalovirus infection, and five additional patients were investigated because of abnormal ultrasonographic findings suggestive of fetal infection. All patients underwent detailed ultrasonographic examinations and 11 of the 12 also had percutaneous fetal blood sampling and simultaneous amniocentesis. Amniotic fluid was cultured for cytomegalovirus. Complete blood cell count, 'Y-glutamyl
Volume 165 Number 3
transpeptidase (GGTP), total immunoglobulin M (IgM), cytomegalovirus-specific IgM, and cytomegalovirus cultures were performed on the fetal blood samples. In addition, polymerase chain reaction with cytomegalovirus-specific oligonucleotides was performed on two amniotic fluid and fetal blood samples obtained at the Mount Sinai Medical Center. All cytomegalovirus cultures, specific IgM assays, and polymerase chain reaction studies of those fetuses evaluated at the Mount Sinai Medical Center were performed in the laboratory of one of the authors (D.E.) at Memorial Sloan-Kettering Cancer Center. Cytomegalovirus-specific IgM was determined by enzyme-linked immunosorbent assay (ELISA). Cytomegalovirus isolation was performed in fibroblast culture by indirect immunofluorescence staining with cytomegalovirusspecific monoclonal antibodies. Results were generally available within 48 hours after innoculation. The polymerase chain reaction studies on amniotic fluid and fetal blood were performed by previously described techniques. 7 We used oligonucleotide primers flanking a 421 base pair sequence of the cytomegalovirus immediate-early gene and an oligonucleotide probe to an internal sequence of the amplified sequence. Positive and negative controls were analyzed simultaneously with each study sample. Specific details regarding the polymerase chain reaction primers and probe and the cytomegalovirus-specific monoclonal antibodies will be furnished on request. At the Institute of Puericulture de Paris, cytomegalovirus-specific IgM studies were also performed by ELISA. Cytomegalovirus cultures were done by fibroblast innoculation without immunofluorescence, and results were available within 3 to 4 days after innoculation. Complete blood counts were done by standard automatized techniques, and all abnormal platelet counts were confirmed manually. Total IgM and GGTP were assayed by standard techniques at both centers. Hematocrit values, platelet counts, and total IgM values were compared with normal fetal values for the corresponding gestational age. s GGTP levels do not change with gestational age, and the values in this series were compared with known fetal norms (24.4 ± 9.6 UI/L)." Fetal blood sampling was done percutaneously under ultrasonic guidance, and amniotic fluid was obtained immediately after the blood had been aspirated. The fetal origin and purity of the blood was confirmed in all cases by red blood cell size analysis.
Results Primary infections. Seven pregnancies were studied because of documented primary infections in the mothers (Table I, patients 1 to 7). All these women had
Prenatal diagnosis of cytomegalovirus infection
715
clinical manifestations of cytomegalovirus infection, and cytomegalovirus-specific IgM present in their serum. The infections all occurred during the first half of pregnancy (range, 11 to 18 weeks), and the patients were referred for fetal assessment between 21 and 32 weeks' gestation. One of these fetuses was found to be infected with cytomegalovirus. This case (Table I, No. 1) has been previously reported. lO Briefly, the mother had fever, neurologic symptoms, and liver dysfunction at 11 weeks' gestation and she was shown to have cytomegalovirus-specific IgM. At 21 weeks' gestation ultrasonographic examination revealed an appropriately grown fetus and no structural abnormalities or evidence of hydrops. Fetal blood studies revealed a normal hematocrit value and platelet count, and cytomegalovirus-specific IgM was absent. The GGTP value was elevated, however, and cultures of the amniotic fluid and fetal blood were positive for cytomegalovirus. This pregnancy was terminated. Autopsy revealed a grossly normal fetus with histologic evidence of nonspecific acute infection (white cell infiltrates in the lung and liver). There were rare inclusion cells suggestive of cytomegalic inclusion disease. Immunofluorescence tissue studies were positive for cytomegalovirus in the fetal kidney, lung, adrenal, and placenta. The six remaining fetuses of mothers with documented infection were sampled between 24 and 32 weeks' gestation. All were structurally normal on ultrasonographic examination. The fetal hematocrit, platelet count and total IgM levels were normal. Specific cytomegalovirus IgM and viral cultures of amniotic fluid and fetal blood were negative in all cases. These fetuses were delivered at term, and all except one were appropriately grown. None had manifestations of cytomegalovirus infection. Specific IgM and viral cultures were negative in all neonates. Abnormal ultrasonographic findings. Five additional fetuses were studied because of abnormal ultraso no graphic findings and all were infected. None of these five fetuses survived; three pregnancies were terminated and the two live-born infants died within a few days (Table I, patients 8 to 12). Patient No.9 did not undergo amniocentesis or fetal blood sampling. In this instance the pregnancy was terminated because of the ultrasonographic findings and the diagnosis was made retrospectively by culturing tissues from the abortus. In all five cases cytomegalovirus infection was confirmed by isolation of the virus from fetal tissues after termination of pregnancy or in the urine of live-born infants during the first week after birth. All the mothers were cytomegalovirus IgG seropositive but none had cytomegalovirus-specific IgM. Therefore it is likely that these cases were a result of recurrent infection, although a primary infection earlier during the pregnancy cannot be ruled out. One case was confirmed to
716
Lynch et al.
September 1991 Am J Obstet Gynecol
Table I. Patient data
Patient No.
1*
Reason for investigation
Type of infection
Maternal IgM
Gestational age at PUBS (wk)
+
21
+
24-32
Primary CMV infection at 13 wk Maternal infection during pregnancy
Primary
Abnormal ultrasonographic findings at 21 wk Abnormal ultrasonographic findings at 16 wk
Recurrent
22
Unknown
ND
lOt
Fetal growth retardation at 30 wk
Unknown
33
llt
Abnormal ultrasonographic findings at 31 wk Abnormal ultrasonographic findings at 30 wk
Unknown
31
Unknown
30
2-7t
8*
9*
12*
Primary
(all cases)
Fetal blood Ultrasonogmphic findings
ITotallgM ICMV IgM
Platelet count
GGTP
Normal
N
t
N/A
Normal anatomy, one mild growth retardation Ascites, dilatation of posterior horns of lateral ventricles Oligohydramnios, dilatation of posterior horns of lateral ventricles, bowel echodensity Oligohydramnios, fetal growth retardation
N
N
N
t
t
NA
NA
NA
N
NA
t
N
t
t
N
t
t
Oligohydramnios, cerebral calcification, ventricular dilatation Severe hydrops
NA
+
PUBS, Percutaneous umbilical blood sampling; CMV, cytomegalovirus; NA, not available; N, normal; AF, amniotic fluid; TOP, termination of pregnancy; +, positive; -, negative; PCR, polymerase chain reaction; ND, not done. *Evaluated at Mount Sinai Medical Center. tEvaluated at the Institut de Puericulture de Paris.
be a recurrence because at 10 weeks' gestation the mother was found to have cytomegalovirus-specific IgG and no IgM. The ultrasonographic findings in this group were varied. The most common abnormalities were ventricular dilatation (three cases) and oligohydramnios (three cases). One fetus (patient No. 10) was thought to be growth retarded, and this was the only one with a birth weight below the 10th percentile. Not only were the ultrasonographic findings diverse among the patients, but in a given fetus they sometimes changed. For example, in the fetus of patient No.8 moderate ascites at 22 weeks' gestation resolved almost completely within 1 week.
Comment Approximately 10% of infants congenitally infected with cytomegalovirus are symptomatic at birth" The most common findings are petechiae, hepatosplenomegaly, jaundice, microcephaly, small size for gestational age, inguinal hernias in males, and chorioretinitis. The mortality in this group may be as high as 30%, and 90% of the survivors have permanent disabilities such as mental retardation, neuromuscular dis-
orders, and hearing loss. Permanent sequelae, which is usually hearing loss, develop in 5% to 15% of the infected infants without symptoms, but mental retardation and severe neuromuscular disorders may also occur,,· l1 Between 30% and 40% ofinfants exposed to primary maternal cytomegalovirus infection will be congenitally infected.' It is not known why most exposed fetuses do not acquire the infection. Gestational age at the time of maternal infection does not affect the transmission rate, although it does seem to influence the severity of the infection. In the report by Stagno et al.' of 37 infected infants, all those who were symptomatic at birth and those in whom clinically significant handicaps subsequently developed were born to women with primary infection before 27 weeks' gestation. If 1% of all babies delivered in the United States are congenitally infected, and approximately 20% of these have permanent sequelae, it is obvious that cytomegalovirus is a significant public health problem. Surprisingly, there are only a few isolated reports regarding the prenatal diagnosis of this infection. 12 The most obvious explanation for this is that the great majority of primary cytomegalovirus infections are without
Prenatal diagnosis of cytomegalovirus infection
Volume 165 Number 3
PCR
CMV cultures AF
+
I
Blood
AF
+
I
Blood
Outcome
Autopsy findings
NA
NA
TOP
NA
NA
Normal fetus; histologic evidence of infection
Normal infants, one small for gestational age
+
+
TOP
Ventricular dilatation with massive necrosis of right hemisphere; placental villi tis
NA
NA
TOP
Disseminated CMV infection; ventricular dilatation with brain necrosis and calcifications; placental villi tis
+
NA
NA
Disseminated CMV infection; normal brain
+
NA
NA
Delivery at 36 wk; hepatic failure; thrombocytopenia; small for gestational age; neonatal death TOP, microcephaly, petechia, hepatosplenomegaly
+
+
Delivery at term; pulmonary hypoplasia; neonatal death
Disseminated CMV infection; normal brain
NA
NA
717
symptoms, and pregnant women are not routinely screened for cytomegalovirus antibodies. Another problem is that there is no intrauterine treatment for this disease at this time. In this series we evaluated seven fetuses exposed to documented primary maternal infection and found only one (14%) to be infected in utero. All these fetuses, including the infected one, had normal ultrasonographic findings. By combining viral cultures of amniotic fluid and fetal blood with assessment of specific cytomegalovirus IgM, total IgM, platelet count, and GGTP in fetal blood we were able to make a correct diagnosis in each of these cases. All negative cases were confirmed after birth and cytomegalovirus was isolated from the infected fetus after termination of that pregnancy. It must be noted, however, that given the normal morphologic findings in the infected fetus it is impossible to predict what manifestations if any may have occurred after birth in that case. Five infected fetuses were evaluated because of abnormal ultrasonographic findings. The most common finding in this group was cerebral ventricular dilatation and oligohydramnios. However, no single abnormality was consistently seen in all fetuses. Clearly, any fetus with ventricular dilatation, microcephaly, or cerebral calcifications should be suspected of being infected. Also, fetuses with any evidence of unexplained hydrops (ascites, pleural effusions, or skin edema) should be evaluated for cytomegalovirus infection. None of the
ND
fetuses with abnormal ultrasonographic findings and cytomegalovirus infection survived; three pregnancies were terminated and the two live-born infants died within a few days. All these fetuses were found to have severe manifestations of cytomegalovirus infection including three with central nervous system abnormalities. The other two had normal central nervous system findings but died of pulmonary hypoplasia and liver failure. Five of the six infected fetuses underwent fetal blood sampling and amniocentesis. A positive viral culture of amniotic fluid and elevated total IgM and GGTP values in fetal blood were found in all five fetuses tested. Cytomegalovirus culture of fetal blood was positive in only one of five fetuses tested. Cytomegalovirus-specific IgM was present in fetal blood in only one of the five fetuses tested, compared with rates of 50% to 90% found in cord blood of infected neonates.'" 14 The fetal immune system has been found to be capable of consistently responding to certain antigenic stimuli by 22 weeks' gestation. '5 However, production of specific IgM appears to be dependent on the organism involved. For example, it has been found that the vast majority of fetuses with congenital rubella infection produce specific IgM after 22 weeks," whereas only 15% of fetuses infected with toxoplasma tested between 24 to 29 weeks produce specific IgM." The low rate of positivity in our patients may be because two of the patients were sampled at 21 and 23 weeks' gestation, or it may be related
718 Lynch et al.
to the IgM assays used. Ahlfors et al. l3 sent serum from culture proved cytomegalovirus-infected neonates to four different laboratories for analysis. The three that used an ELISA detected cytomegalovirus-specific IgM in 63% to 80% of the cases, whereas the one with a radioimmunoassay detected positivity in only 50% of cases. Other authors 14 have reported that the radioimmunoassay is 90% sensitive. It is clear from all studies that viral cultures remain the "gold standard." In this report only ELISAs were used, but they were performed in two different laboratories. Because the number of patients is small, the true sensitivity of specific IgM analysis cannot be determined from this series, but it is evident that this parameter alone cannot be used to exclude congenital cytomegalovirus infection. In conclusion, the prenatal diagnosis of cytomegalovirus infection is possible by use of a combination of ultrasonography, amniocentesis, and fetal blood sampling. Although amniotic fluid culture for cytomegalovirus seems to be very sensitive, the number of cases in this series is not large enough to advocate use of amniocentesis as the only diagnostic approach. Furthermore, in a review Stagno et al. 16 mentioned that they were aware of three women with primary gestational cytomegalovirus infection in whom the virus was isolated from fetal or placental tissues but not from amniotic fluid. In our opinion, until more data become available, fetal blood should also be studied when evaluating a fetus for cytomegalovirus infection. Furthermore, it is possible that fetuses with hematologic evidence of systemic involvement, such as thrombocytopenia and liver damage, may have a worse prognosis than those infected without these findings. Once the diagnosis has been made, if the fetus has abnormal ultrasonographic findings this series suggests that the prognosis for intact survival appears to be very poor. However, if the ultrasonographic findings are normal the prognosis is far from clear. Until follow-up data of such cytomegalovirus-infected fetuses are available, we will have to rely on information obtained from neonates about the natural history of these infants.
September 1991 Am J Obstet Gyneco1
REFERENCES 1. Stagno S. Cytomegalovirus. In: Infectious diseases of the fetus and newborn infant. Remington JS, Klein JO, eds. Philadelphia: WB Saunders, 1990:241-81. 2. Stagno S, Pass RF, Dworsky ME, et al. Congenital cytomegalovirus infection: the relative importance of primary and recurrent maternal infection. N Engl J Med 1982;306:945-9. 3. Stagno S, Pass RF, Cloud G, et al. Primary cytomegalovirus infection during pregnancy. JAMA 1986;256:1904-8. 4. Stagno S, Whitley RJ. Herpes virus infection of pregnancy. I: cytomegalovirus and Epstein-Barr virus infections. N EnglJ Med 1985;313:1270-4. 5. Fan-Havara P, Nanata MC, Brady MT. Ganciclovir-a review of pharmacology, therapeutic efficacy and potential use for treatment of congenital cytomegalovirus infection. J Clin Pharm Ther 1989; 14:329-40. 6. Daffos F, Forestier F, Capella-Pavlovsky M, et al. Prenatal management of 746 pregnancies at risk for congenital toxoplasmosis. N EnglJ Med 1988;318:271-5. 7. Saiki RK, Gelfond DH, Stoffel S, et al. Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 1988;239:487-91. 8. Forestier F, Daffos F, Galacteros F, et al. Hematological values of 163 normal fetuses between 18 and 30 weeks of gestation. Pediatr Res 1986;4:342-7. 9. Daffos F, Forestier F. Biologie du sang foetal. In: Medicine et biologie du foetus humain. Daffos F, Forestier F, eds. Paris: Maloine, 1988. 10. Meisel RL, Alvarez M, Lynch L, et al. Fetal cytomegalovirus infection: a case report. AM J OBSTET GYNECOL 1990;162:663-4. 11. MacDonald H, Tobin J O'H. Congenital cytomegalovirus infection: a collaborative study in epidemiological, clinical and laboratory findings. Dev Med Child Neurol 1978;20:471-82. 12. Grose C, Itani 0, Weiner CPo Prenatal diagnosis of fetal infection: advances from amniocentesis to cordocentesis-congenital toxoplasmosis, rubella, cytomegalovirus, varicella virus, parvovirus and human immunodeficiency virus. Pediatr Infect Dis J 1989;8:459-68. 13. Ahlfors K, Forsgren M, Griffiths P, et al. Comparison of four serological tests for the detection of specific immunoglobulin M in cord sera of infants congenitally infected with cytomegalovirus. ScandJ Infect Dis 1987;19:303-8. 14. Griffiths PD, Kangro HO. A user's guide to the indirect solid phase radioimmunoassay for the detection of cytomegalovirus specific IgM antibodies. J Viral Methods 1984;8:271-82. 15. Daffos F, Forestier F, Grangeot-Keros L, et al. Prenatal diagnosis of congenital rubella. Lancet 1984;2:1-3. 16. Stagno S, Pass RF, Dworsky ME, Alford CA. Maternal cytomegalovirus infection and perinatal transmission. Clin Obstet Gynecol 1982;25:563-76.
