European Journal of Obstetrics & Gynecology

GYNECl)Lg
ELSEVIER and Reproductive Biology 63 (1995) 7 - 1 6

Review article
Congenital cytomegalovirus infection
Yair Daniel, Ilan Gull, M. Reuben Peyser, Joseph B. Lessing*
Department ~/' Obstetrics and Gynecology "A ', Tel Aviv Sourasky Medical ('enter and Sackler Facul O' o f Medicine, Tel Avic Unicersity,
Tel A viv, Israel

Received 13 December 1994: revision received 16 June 1995; accepted 22 June 1995

Abstract

Ob/ective: Congenital cytomegalovirus is the most common viral infection affecting approximately 1% of newborns. The virus
can be transmitted to the fetus during both primary and recurrent infection. Although most of the infants are asymptomatic at
birth, up to 15% develop late complications. The annual cost of treating cytomegalovirus infection complications in the USA is
two billion US dollars. Many issues regarding cytomegalovirus infection such as routine screening, antenatal diagnosis and
vaccination during pregnancy are unsettled and disputed. The aim of this article is to review the current literature on the subject
and to draw some conclusions. Design: Review of the current literature. Conclusions: At present, it appears that there is no
indication for routine prenatal screening, while other issues, such as the most accurate method for antenatal diagnosis and the
indications for pregnancy termination are, as yet, unsettled.

Keywor&': CMV; Pregnancy; Prenatal screening; Antenatal diagnosis; Vaccine

1. Introduction tomatic infants develop late complications up to the age

of 2 years, the most important of which is sensorineural
Cytomegalovirus (CMV) is an ubiquitous agent that hearing loss that can be progressive. Cytomegalovirus is
commonly infects many animals and humans [1]. These considered as the most common infectious cause of
viruses belong to the herpes virus group that contain mental retardation and deafness [5].
deoxyribonucleic acid (DNA). The viruses are high-spe- In the last 20 years, more than 800 000 infants in the
cies specific, and similar to other herpes virus infections, USA had congenital CMV: 50 000 were symptomatic.
primary CMV infection is followed by persistent and/or The annual cost of treating CMV infectious complica-
recurrent infections, most often due to reactivation of a tions in the USA is two billion US dollars. Late neuro-
latent infection [2]. logic complications developed in 120000 of the
Reinfection is less prevalent and is due to the presence asymptomatic infants [6].
of several different strains of CMV [3]. In the vast
majority of cases, including perinatal infections, CMV 1.1. Maternal infection
infections are subclinical [3]. In the host with a compro- The prevalence of both primary and recurrent CMV
mised immune system CMV causes serious infections, infection in pregnant women varies in different types of
some leading to death [1]. populations and ranges from 0.7-4% for primary CMV
The most c o m m o n viral congenital infection is that of infection [7-12] and up to 13.5% for recurrent infection,
CMV affecting about 1% (0.2%- > 3'7,,) [3] of infants, as manifested by genital shedding near term [13]. The
90% of whom are asymptomatic at birth and about 10'7, rates are higher in the lower socioeconomic sectors [14].
who develop cytomegalic inclusion disease with a 2 0 - In the USA, 50 60% of pregnant women from the high
30% mortality rate [4]. About 10-15% of the asymp- socioeconomic sector are seropositive, compared to 7 0 -
80% of the low socioeconomic group [12]. In Europe,
45% of pregnant women are seropositive at the beginning
* Corresponding author. of pregnancy [15].

0301-2115/95/$09.50 ~C 1995 - Elsevier Science Ireland Ltd. All rights reserved

S S D I 0301-2115(95)02198-2
8 Y. Daniel et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 63 (1995) 7 16

The differences seem to be related to crowded living tion. There are three mechanisms for recurrent infection
conditions and poor personal hygiene in the lower socio- [3]: (1) Low-grade chronic infection following primary
economic groups. Risk factors for CMV infection are: infection; (2) Reinfection with a different strain of CMV
black, low education, other genital tract infections, < 30 that differs genetically and antigenetically [33]; (3) Reac-
years and close contact with children < 2 years [14,16]. tivation of a latent virus following primary infection in
Spread of CMV is mainly by way of oral and venereal response to different stimuli (which is the most common).
routes most frequently during childhood and early adult- Primary infection and recurrent virus shedding are
hood [17]. more common among young adults. After the age of 30,
As a result, the rates of infection, virus shedding and virus excretion is markedly reduced [34].
the incidence of reinfection and coinfection with multiple Pregnancy itself is not a risk factor for infection [12].
strains of CMV, are far higher in a promiscuous popu- Between 0.7% and 4% of seronegative pregnant women
lation [18-20]. are infected during pregnancy [12].
The major source of infection are young infants and During pregnancy, genital shedding increases from
children with subclinical infection [21,22]. Seroconver- 1.5% in the first trimester to 13.5% near term [13].
sion rate as high as 50% has been described in parents Postpartum, the breast is the most common site of
of children < 3 years of age excreting CMV [23,24]. reactivation with rates ranging from 14 to 27% of the
Day-care centers are a high-risk setting for primary CMV total population [35].
infection with increased risk for congenital infection [25]. Cytomegalovirus can be transmitted in utero during
The prevalence of CMV shedding among children in both primary and recurrent infection, although it is far
day care centers is very high (23%-72%) [26], and more infrequent in the latter (40% vs. 0.15-1%) [36-38].
workers at child day care centers are at high risk for CMV Since the prevalence of CMV immunity (seropositivity)
infection. The annual seroconversion rate in this group is high among women (50%-80%) [12], transmission of
was 11%, as compared to 2.2% in the general population reactivated virus is the major cause of congenital infec-
[26]. tions [28,36,37]. The mechanism that transmits the virus
has not been determined, but infected maternal leuko-
The common use of day care centers and the increase
cytes are most probably transmitted through the placenta
in breast feeding, particularly among families of middle
to the fetal circulation, resulting in viremia with infection
and upper income groups, i.e. those more likely to be
of multiple organs [17]. Endometrial, cervical or even
seronegative, has changed the epidemiology of CMV,
maternal or paternal germ cells, might also be the source
and may cause an increase of congenital CMV in certain
[17].
groups [27].
Maternal immunity, although not protective, modifies
The second major source of infection is venereal
fetal infection [37-40]. Most neonatal cytomegalic inclu-
transmission which occurs during both heterosexual and
sion diseases are the result of primary maternal infection.
homosexual activity [28].
Serious fetal damage is rare following recurrent infection.
Higher rates of cervical infection were found in women
Intrauterine transmission occurs in about 40% of
who were highly-sexually active and who were examined primary infections and clinically apparent disease occurs
for other venereal diseases [19]. The rate of CMV infec- in 10-15%. Only 0.15-1% of recurrent infections are
tion was correlated with a large number of sexual part- transmitted to the fetus, with a higher prevalence among
ners, early age of intercourse and a non-white race [29]. low socioeconomic groups [41], and the disease is clini-
More direct evidence of venereal transmission was cally apparent in only 0 - 1 % of the infants [41].
provided by analysis of CMV strains isolated from sex It is, as yet, unknown why some women, even with
partners by DNA-restriction fingerprinting [30]. recurrent infection and high levels of antibodies, transmit
The uterine cervix and the semen are important the virus to the fetus, while others without antibodies
reservoirs of CMV and play a role in disease transmis- during primary infection do not. Some studies suggest
sion. Both are common sites of active and chronic that viral load as reflected by a greater and more
infection [28]. The frequency of cervical shedding in- persistent antibody response may be an important factor
creases with multiple sexual partners, youth and ad- [42]. Some studies found a depressed maternal-cell-medi-
vanced stages of pregnancy [31]. ated immunity in women who transmitted the virus, but
Prevention of infection should focus on good hygiene, these findings were not consistent [43-46]. Stern et al. [47]
limited contact with infected children and responsible and Fernando et al. [48] found that women with positive
sexual practices, although the efficacy of this is uncertain. blast lymphocyte reactions to the CMV antigen did not
Clinically, most primary infections (up to 90%) are transmit the virus to the fetus.
subclinical. In symptomatic patients, the disease usually As with other intrauterine infections, placental infec-
manifests as a mononucleosis-like disease with a negative tion was detected with or without fetal infection [49].
hetrophil antibody test [32]. Fetal infection can occur throughout gestation, but
Asymptomatic, intermittent excretion of the virus infection may be more virulent when acquired early in
from single or multiple sites characterizes recurrent infec- gestation [11].
 Y. Daniel et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 63 (I995) 7 16 9

1.2. Congenital CMV infection in the internal ear and the endothelial cells [53]. Dam-
Throughout the world, CMV is the most common age to the endothelial cells in the brain vessels results in
cause of congenital and perinatal viral infection [1,50]. ischemia and necrosis with calcifications that develop
Congenital infection affects approximately 1% (0.2- later, most of which are periventricular and appear in
> 3%) [3,50] of live births with an even higher percent- 20% of cases [2]. The calcifications that are demon-
age in developing nations [50]. strated by ultrasound, computerized tomography (CT)
There is a direct relationship between prevalence of and magnetic resonance imaging (MRI) [41] are not
congenital CMV infection and the rate of maternal pathognomomic but support the diagnosis of CMV
immunity, resulting from the inability of maternal im- infection.
munity to prevent reactivation and transmission of the Ninety percent of congenital CMV infections are
latent virus to the fetus [36,37,51]. As a result from the asymptomatic with a better outcome: 5-15% are at risk
high rate of seropositivity, congenital CMV infection is of having developmental abnormalities such as sen-
more common among lower socioeconomic sectors and sorineural hearing loss, microcephaly, motor defects,
in developing countries. mental retardation, chorioretinitis and dental defects.
Intrauterine transmission occurs in 40% of primary Usually, these defects become apparent within the first
infections, 10% of whom have a clinically apparent 2 years, but hearing loss can be progressive or occur
disease. Of the 10% symptomatic infants, half have later [39,54-57]. Hearing loss is the most important
typical cytomegalic inclusion disease and half have late-occurring sequela of congenital CMV. The impair-
atypical involvement [3]. ment can be bilateral in up to 50% of cases, and is
Typical cytomegalic inclusion disease involves multi- severe enough to produce serious difficulties with verbal
organs in particular, the reticuloendothelial system and communication and learning [58]. In 25% of these
the central nervous system. infants, hearing loss develops or becomes more severe
Weller and Hanshaw [52] described the most fre- after the first year of life. Serial audiometric examina-
quently found abnormalities in cytomegalic inclusion tions should follow for every infant with proven con-
disease as hepatomegaly, splenomegaly, microcephaly,
genital CMV.
jaundice and petechia (10%). Microcephaly is fre-
Microcephaly, with various degrees of mental retar-
quently combined with cerebral periventricular calcifi-
dation and neuromuscular defects, occurs in 2-7% of
cation, intrauterine growth retardation and
infants with asymptomatic infection. The risk of chori-
prematurity. Less common symptoms include inguinal
oretinitis is approximately l%. Dental defects are less
hernia in boys, and chorioretinitis with or without optic
severe than in the symptomatic group and are charac-
atrophy. Pneumonitis is rare (<1%). Congenital
terized by generalized yellowish discoloration of the
anomalies associated with CMV include those of the
teeth and opaque, hypocalcified, and in many cases
third brachial arch and defects of tooth enamel.
absent, enamel [3].
Laboratory findings in infants with symptomatic in-
fection include: increased levels of cord immunogiobu- Fowler et al. [38] compared the outcome of CMV
lin (Ig)-M ( > 20 mg%); atypical lymphocytosis (5%); infection in infants born to mothers who had primary
elevated levels of liver enzymes; thrombocytopenia, and recurrent infection during pregnancy for a mean of
conjugated hyperbilirubinemia and increased levels of 4.7 years. Only infants born to mothers with primary
protein in CSF [40]. infection had symptomatic infections at birth (18%).
The mortality rate of cytomegalic inclusion disease is In the primary infection group (125 patients), 25%
between 20-30% [40]. Of the surviving infants, 90% had one or more sequela that included 13% with mental
develop severe mental, developmental and hearing retardation (I.Q < 70), 15% with sensorineural hearing
problems. In the original group of 17 patients described loss, 8% bilaterally, 6% with chorioretinitis, 5% with
by Weller and Hanshaw [52], only two were normal at microcephaly and 1% with motor defects.
14 and 20 months of age. These problems include Five percent of the infants in this group had seizures
mental retardation (70%), seizures, blindness, motor and 2% died.
handicaps, hearing loss (53%), ocular and dental defects In the recurrent infection group (64 patients), only
and language and learning difficulties [40]. The hearing 8% had one or more sequela that included: 5% with
loss can be progressive. Bad prognostic factors include: sensorineural hearing loss (none bilaterally), 2% with
primary infection; cytomegalic inclusion disease; neuro- chorioretinitis and 2% with microcephaly.
logical complications; microcephaly; brain calcifica- Mental retardation, seizures, motor defects or death
tions; and high immunoglobulin (IgM) titers [53]. did not occur in this group. These results provide more
In contrast to the rubella virus, CMV does not affect evidence that maternal antibodies before conception,
organogenesis and only damages organs that have al- although they cannot prevent infection, provide sub-
ready developed. Cytomegalovirus has an affinity with stantial protection against damaging congenital CMV
the ependimal cells, the eighth cranial nerve, corti organ infection in the newborn.
10 Y. Daniel et ,l. European Jourmd o! Obstetrics & Gynecology and Repro~h,'tive Biology 63 (1995) 7 16

It is unknown why some infants develop severe dis- This defect, that is CMV specific, persists for years
ease while others are asymptomatic. Usually the hu- whether the infection is symptomatic or not [43,71].
moral and cell-mediated immunity of the infants are The blastogenic response to mitogens and the numbers
not affected and the defect might be CMV specific. of T4 and T8 cells are normal.
Britt and Vugler [59] found higher IgG titers in
mothers to asymptomatic infants, which may reflect 2. Diagnosis
late acquired infection with less severe damage to the
fetus. Diagnosis of primary CMV infection during preg-
nancy is difficult because 90% are asymptomatic and
10% are present as a mononucleosis-like disease with
1.3. P e r i n a t a l #Tfection
negative heterophil antibody test [12]. Diagnosis of
Perinatal infection refers to those infections acquired
CMV infection can be made by serologic assays or by
during delivery due to exposure to infected maternal
detection of the virus in different clinical specimens.
genital secretion [60], acquired during the postnatal
Recently developed techniques have greatly increased
period from breast feeding [35,61,62], or acquired as a
the sensitivity and detection speed of CMV and of host
result of a blood transfusion [63,64].
antibody responses [72]. Cytomegalovirus antibodies
Cytomegalovirus infection occurs in 40 60% of in-
are generally detectable within a few weeks of onset of
fants who are breast-fed by seropositive mothers for
viral shedding in primary infection and persist indefin-
> 1 month [35,61,62] and in 25-50% of infants exposed
itely. Thus, the serologic tests for CMV are usually
to the virus in the birth canal [60]. As a result, 1-15%
used for determining susceptibility to primary infection
of infants become infected by 6 months of age. The
and for diagnosing primary infection through serocon-
incubation period ranges from 4 to 12 weeks and the
version (seronegative to seropositive). Seroconversion is
infection is chronic with shedding of the virus for years
a reliable means of diagnosing primary CMV infection.
[39]. The vast majority remain asymptomatic but the
A negative first serum sample is the only observation
infection may be associated with interstitial pneumoni-
that, in combination with any later proof of CMV
tis [65].
infection, reliably demonstrates primary infection. In
In premature infants who require intensive care,
addition, the rise in triers may be delayed for up to 4
blood transfusions are an important iatrogenic cause of
weeks after primary infection, thus a negative result
CMV infection [63]. This infection is associated with
may also require follow-up. Paired samples that demon-
rapid deterioration, septic appearance, hepatospleno-
strate an increase in titers (usually a four-fold increase)
megaly, pneumonitis and death [63,64].
are also important for diagnosing active infection, irre-
spective of being primary or recurrent. It is imperative
1.4. h m m m o l o g i c responses that the clinician be informed of the sensitivity and
Intrauterine, perinatal or postnatal CMV infection is specificity of the test performed. There are many sero-
characterized by chronic shedding of the virus in high logical tests for CMV antibodies. The traditional com-
quantities, due to the immaturity of the infants' im- plement fixation (CF) technology has been replaced by
mune system [43,66,67]. immunofluorescence, enzyme-linked immunoabsorbent
During pregnancy, both transplacental transfer of assay (EL1SA), latex agglutination, radioimmunoassays
maternal IgG and fetal production of lgM occur and, (RIA) and indirect hemagglutination assays [72]. These
at delivery, the amount of IgG antibodies in the cord assays are accurate and results (presence of antibody)
blood and maternal sera are equivalent [66,68,69]. As are available after several hours.
the levels of maternal antibodies decline, the infant Most assays for CMV antibodies measure the total
begins to produce his own IgG antibodies and contin- antibody titre (IgG + IgM). However, there are assays
ues to produce these for weeks to months after birth. which measure CMV-specific lgM such as ELISA and
Production of lgG continues for years, except for anti- RIA with a sensitivity of 73% and 78%, respectively
bodies to early CMV antigen and complement fixation [73,74], and can be used for diagnosing primary infec-
antibodies [66]. Some studies suggest that some infants tion. However, IgM antibodies may be found not only
born with symptomatic infection have a defect in the in primary infection but can also be detected in preg-
lgG antibody response to CMV-specific antigens, com- nant women with secondary CMV infection, in up to
pared to infants with asymptomatic infection [70]. 10% of cases, probably because IgM may persist for 18
In infants with chronic congenital CMV infection, months following primary infection [74]. Due to these
there is only one consistent defect in the cell mediated problems, IgM assays, as a rapid means of diagnostic
immunity - - an inability of their lymphocytes to elicit active infection, were replaced by the new rapid meth-
a blastogenic response and induce production of inter- ods of detecting CMV antigens [72,75].
feron in vitro, when challenged with CMV antigens The detection of infant viruria is preferable to IgM
[43,46,71]. detection for diagnosis of congenital CMV infection
 Y. Daniel et al. ,' Eur~q~ean Journal g/" Obstetrics & GynecoloL~v and Reproductive Biology 63 (1995) 7 16 11

[72]. Isolation of C M V still remains the definitive diag- the pregnancy; (6) as 90% of the infants are asymp-
nostic procedure to which all newer assays are com- tomatic, there is no absolute indication for termination
pared. M a n y clinical specimens, such as urine, blood, of the pregnancy; (7) it is extremely difficult to avoid
throat smear and amniotic fluid, can be cultured, al- infection in seronegative women during pregnancy be-
though a single sample will reveal only 50% of serocon- cause the virus is endemic internationally, although
verted subjects [75]. preventive measures such as good hygiene, limited con-
The virus is cultured on human fibroblast monolayers tact with infected children and responsible sexual prac-
and isolation is confirmed by detection of visible cyto- tices are advisable. There is also no need for routine
pathic effects [75,76] which, on average, takes 1 - 2 weeks vaginal or cervical culture during pregnancy because the
and sometimes up to > 6 weeks when the amount of prediction is minimal, and virus isolation will identify
virus in the specimen is low [72]. This prolonged interval only 50% of seroconverted women in one sample, and
required for the culture was the reason for developing will not discriminate primary from recurrent infection
methods of rapid detection of C M V antigens. The [75]. Adler [89] does not recommend routine screening of
monoclonal antibodies to the 72-KDA major immedi- all pregnant women. However, he suggested limiting
ate-early protein of C M V can detect a nuclear antigen screening before conception to women at high risk (such
in fibroblasts within hours of infection [77-80], thus as day care center workers and mothers of children < 2
permitting evaluation of cultures long before the appear- years of age), and to advise them about preventive
ance of identifiable cytopathology. The cultures for measures during pregnancy.
rapid antigen detection are performed on flat monolay- Yow and Demmler [6] suggested that all women be
ers on cover slips in shell vials, or in 24-well cluster plates tested before conception.
[801. However, these preconception screening programs
The sensitivity reported was comparable, or even should not be introduced until the protective efficacy of
superior in cases of negative culture, to that of conven- the preventive measures is proven.
tional culture [78,80]. There are also monoclonal anti-
bodies to other early and late C M V antigens which are 2.2. Antenatal diagnosis
useful for rapid antigen detection. Antenatal diagnosis is possible by: (1) ultrasound, CT,
Hybridization techniques for C M V detection in urine and M R I [41], (2) amniotic fluid culture or C M V - D N A
were also developed with a very high sensitivity, even detection by PCR; [3] cordocentesis for IgM antibodies
more than that of conventional culture [81 83]. and blood indices.
In situ hybridization for detection of C M V inclusions Ultrasound, CT, and M R I can demonstrate cerebral
is increasingly used in cultured cells or directly in calcification or necrosis, which are usually periventricu-
infected tissue [84]. However, these methods of virus lar in C M V infections [90]. Other ultrasound findings
identification does not discriminate primary from recur- that can suggest C M V infection are hyperechoic areas in
rent infection and correlation with the clinical and the fetal abdomen [91], and echogenic vessels in the fetal
serological status is necessary. thalami and basal ganglia [92].
The polymerase chain reaction (PCR), which amplifies In 1971, Davis et al. [93] were the first to isolate the
a specific short segment of D N A , was recently intro- virus by amniocentesis. L a m y et al. [94] performed a
duced for the diagnosis of C M V infection and has been prospective study to evaluate the risk of fetal C M V
used successfully for antenatal diagnosis [75]. transmission: 1771 women were tested for CMV; 861
were seronegative, 20 women seroconverted during preg-
2.1. Maternal screening nancy (2.3%), and seven cases with primary C M V
Most authors do not recommend maternal screening infection were investigated by ultrasonography, amnio-
for C M V during pregnancy for several reasons [41,85- centesis and cordocentesis. In five women, an amniotic
89]: (1) as discussed previously, the reliability of the fluid culture was positive: in three cases, IgM was
current serological tests available for diagnosis of pri- identified in cord blood and in two cases brain ultra-
mary C M V infection are limited. Serologic tests for sonography results were positive. Biopsy of fetal tissue
detection of specific C M V IgM are negative in almost after abortion confirmed C M V infection in the five
25% of patients with primary infection. In addition, IgM fetuses with positive amniotic fluid culture. In the two
antibodies may persist for months and, in some cases, cases with negative amniotic fluid culture and no IgM
may represent infection prior to conception: (2) even antibodies in cord blood, the neonates were free of
with proven primary C M V infection there is no way the infection (no IgM, negative culture).
damage, if any, to the fetus is predictable; (3) currently, Hohlfeld et al. [95] investigated 16 fetuses with pri-
there is no approved treatment: (4) there is no approved mary maternal C M V infection during pregnancy by
vaccine for pregnant women, as yet; (5) maternal anti- ultrasonography, amniocentesis and cordocentesis. Posi-
bodies do not protect the fetus from transmission of the tive prenatal diagnosis was established in eight cases: of
virus during reactivation that can occur at all stages of these, positive amniotic fluid culture was found in eight,
12 Y. Daniel et al. ,; European Journal o! Obstetrics & Gvnecolog.v and Rt,produclire Biology 63 (1995) 7 16

IgM antibodies in six and abnormal ultrasonography in interferon, transfer factor and nucleoside drugs, none
two cases. Among infected fetuses, abnormal labora- of which were beneficial clinically.
tory findings in cord blood included: anemia, thrombo- Recently, gancyclovir proved to have potent activity
cytopenia and elevated liver function test. Three against CMV in vitro. The drug was clinically effective
pregnancies were terminated because of abnormal ul- in the treatment of immunocompromised patients
trasonography or abnormal laboratory indices. Four [108 111], but caused myelosuppression in a significant
infants were born; three with subclinical infection, and number of patients, thus limiting its use [111],
one with bilateral hearing loss. In this study, the shell The drug may also cause sterility in males. There are
vial assay [96] was used to detect early viral antigen in no adequate studies on the use of gancyclovir for
the amniotic fluid; the test was positive in all eight cases treatment of congenital CMV infection. Consequently,
with positive amniotic fluid culture. the drug should not yet be used for this purpose.
Grose et al. [97] reviewed 20 cases where congenital Foscarnet is another experimental drug that is con-
CMV infection was diagnosed by amniotic fluid cul- traindicated in infants (nephrotoxic) [3].
ture.
Previously, detection of fetal IgM antibodies against 3. t. Vaccines
CMV was considered the best marker for prenatal Two live attenuated vaccines have been developed
diagnosis of CMV infection [61,66,69,98,99]. These an- and are undergoing trials. The AD-169 laboratory
tibodies are detected only after 20 weeks of pregnancy adapted strain of CMV was developed in England
[100]. An immature immune system, or insufficient test [l 12]. The Towne vaccine was derived by passing wild-
sensitivity could lead to false negative results of IgM type CMV isolate, obtained from congenitally-infected
[101]. The above mentioned studies, and several others women, 125 times in human fibroblast until it became
[102-104], have shown that amniotic fluid culture is the attenuated [113,114]. The vaccines have been given to
most sensitive and specific indicator of CMV transmis- both susceptible and immune-healthy individuals, and
sion. Further studies are needed to demonstrate to renal transplant recipients [115]. The vaccines were
whether abnormal cord blood indices can be used as an highly immunogenic. Both humoral and cell-mediated
indicator for the severity of the disease. immunities were provoked [115]. However, the immu-
Recently, CMV infection was detected in a chorionic nity wanes more rapidly than following natural infec-
villi sample using PCR of the major immediate early tion [116]. The vaccine's virus replication was restricted
antigen gene l104,105]. compared to natural infection and excretion, and trans-
Based on these studies, it appears that when CMV is mission to daily contacts was limited [114].
isolated from the amniotic fluid, congenital infection is Reactivation of the vaccine virus has not been de-
present. However, a negative result, whether by culture tected in either normal or transplant recipients
or PCR, is not predictive, as transmission of the virus [117,118]. In renal transplant recipients, the vaccine did
can occur later in the pregnancy, resulting in congenital not protect against primary CMV infection but reduced
infection. In addition, a positive amniotic fluid culture its virulence [117.118].
cannot predict the damage, and statistically most of the As is known to us today, maternal antibodies, whilst
infected fetuses will be healthy [106]. In a recent study not preventing transmission and fetal infection, protect
by Donner et al. [107], it was demonstrated that five of the fetus from severe damage [38]. The best form of
eight amniotic fluid samples tested before 21 weeks prevention would be to vaccinate all seronegative
were negative in congenitally infected fetuses. Most women of childbearing age [6] and in light of the
authors agree that in the presence of fetal ultrasonogra- positive results, this approach sounds logical. However,
phy abnormality and primary CMV infection, the preg- the use of the vaccine remains controversial for a
nancy should be terminated [106]. It seems that in the number of reasons: (1) we do not yet know how lasting
presence o f abnormal cord blood indices and primary and protective the immunity is; (2) will latency prove to
CMV infection the pregnancy should also be termi- be problematic; (3) whom to vaccinate all women at
nated [94]. In primary CMV infection with positive childbearing age, or only seronegative women - - both
amniotic fluid culture and normal ultrasonography, attitudes were found cost effective [5]; (4) we do not
there is no absolute indication for pregnancy termina- know if the vaccine protects against reinfection with
tion as most of the fetuses will be healthy, and the wild strains: (5) we do not know whether the vaccine
decision should be left to the parents after appropriate protects the fetus from intrauterine infection or may
consultation [106]. actually infect the fetus; (6) does the vaccine have
oncogenic potential, as it is known that CMV can cause
3. Treatment and prophylaxis transformation of both human and nonhuman cell cul-
tures. To date, there has been no increase in the inci-
Various antiviral agents have been used in attempts dence of cancer among the recipients of Towne vaccine
to treat symptomatic CMV infection [106], including [6,1191.
 Y. Daniel et al. /European Journal ~[ Obstetrics & Gynecology and Reproductive Biology 63 (1995) 7 16 13

The solution to these problems would be a subunit [8] Grant S, Edmond E. Syme J. A prospective study of cy-
vaccine [6]. The candidate for this vaccine is the gB tomegalovirus infection in pregnancy. 1. Laboratory evidence of
congenital infection following maternal primary and reactivated
protein complex that is the most abundant protein
infection. J Infect Dis 1981: 3 : 2 4 31.
complex within the virion envelope. This protein is [9] Griffiths PD, CampbelI-Benzie A, Heath RB. A prospective
expressed on the surface of infected cells and generates study of primary cytomegalovirus infection ill pregnant women.
neutralizing antibodies soon after infection, which are Br J Obstet Gynecol 1980: 87:308 314.
constantly found in the serum during the convalescent [10] Ahlfors K. Epidemiological studies of congenital cy-
tomegalovirus infection. PhD Diss Malmo Sweden. University
phase [120]. Recently, the gene corresponding to the gB
of Lund 1982.
protein was inserted into vaccinia virus and adeno virus [11] Stagno S. Pass RF, Cloude G e t al. Primary cytomegalovirus
recombinant vaccine vector system, which will allow infection in pregnancy: incidence, transmission to fetus and
large quantities of protein to produce the vaccine [121]. clinical outcome. J A M A 1986: 256:1904 1908.
[12] Stagno S, Whithly RJ. Herpes virus infections of pregnancy,
part four. N Engl J Med 1985: 313:1270 1274.
4. Conclusions [13] Stagno S. Reynolds DW, Tsiantos A et al. Cervical cy-
tomegalovirus excretion in pregnant and nonpregnant women:
suppression in early gestation. J Infect Dis 1975: 131:522 527.
Cytomegalovirus is the most common congenital in- [14] Walmus BF, Yow MD, Leter JW et al. Factors predictive of
fection affecting about 1% of infants. The virus can cytomegalovirus immune status in pregnant women. J Infect
infect the fetus during primary and recurrent infection Dis 1988: 157:172 177.
and at any stage of pregnancy. Ninety percent of the [15] Griffiths PD, Baboonian C, Rutter D, Peckham C. Congenital
and maternal cytomegalovirus infections in the London popula-
infected infants are asymptomatic at birth. Although tion. Br J Obstet Gynaecol 1991: 98:135 140.
maternal antibodies cannot prevent infection of the [16] Fowler KB. Pass RF. Sexually transmitted diseases in mothers
fetus, they offer a substantial protection against damag- of neonates with congenital cytomegalovirus infection. J Infect
ing CMV infection. At present there is no indication for Dis 1991: 164:259 264.
prenatal serologic screening for CMV. The most accu- [17] Gold E, Nankervis GA. Cytomegalovirus. In: Evans AS ed.
Viral infection of humans. Epidemiology and Control. New
rate way to make an antenatal diagnosis is through York: Plenum Press, 1976:143 161.
amniotic fluid culture. No test is available that can [18] Stagno S. Pass RF, Dworsky ME, Jr. et al. Maternal cy-
predict the damage to the infected fetus. Two live tomegalovirus infection and perinatal transmission. Clin Obstet
attenuated CMV vaccines exist that have not yet been Gynecol 1982: 25:563 576.
approved for the general population. A subunit vaccine [19] Jordan MC, Rousseau WE, Noble GR et al. Association of
Cervical cytomegalovirus with venereal disease. N Engl J Med
is under development. To date there is no approved 1973: 288:932 934.
specific antiviral treatment against congenital CMV [20] Drew WL, Mintz L, Miner RC et al. Prevalence of" cy-
infection. Preventive efforts at this time must focus on tomegalovirus infection in homosexual men. J lnf Dis 1981:
good hygiene, limited intimate contact with infected 143:188 192.
children and responsible sexual practices, although [2H Stagno S, Cloud J, Pass RF et al. Factors associated with
primary cytomegalovirus infection during pregnancy. J Med
more studies are needed to prove their efficacy.
Virol 1984: 13:347 353.
[22] Pass RF, Little EA, Stagno S e t al. Young children as a
probable source of maternal and congenital cytomegalovirus
References infection. N Engl J Med 1987: 316:1366 1370.
[23] Murph JR, Bale JF Jr. The natural history of aquired cy-
[1] Weller TH. The cytomegalovirus, ubiquitous agents with protein tomegalovirus infection a m o n g children in group day care. Am
clinical manifestation. Second of two parts. N Engl J Med 1971: J Dis Child 1988: 142:843 846.
285:203 214, 267 274. [24] Klapper PE. Morris DJ. Screening for viral and protozoan
[2] Overall JC Jr. Viral infections of the fetus and neonate. In: infections in pregnancy: a review. Br J Obstet Gynaecol 1990:
Feigin RD, Cherry JD, eds. Textbook of pediatrics infectious 97:974 983.
diseases (2nd ed). Philadelphia: W.B. Saunders Company. 1987: [251 Pass RF, Kinney JS. Child care workers and children with
1:966 1007. congenital cytomegalovirus infection. Pediatrics 1985: 75:971
[3] Alford AA. Stagno S, Pass R F et al. Congenital and perinatal 973.
cytomegalovirus infection. Rev lnf Dis 1990:12 (Suppl 7): [26] Adler SP. Cytomegalovirus transmission and child day care.
745 751. Adv Pediatr Infect Dis 1992: 7:109 122.
[4] Grose C, ltani O, Weiner CP. Prenatal diagnosis of fetal infec- [27] Stagno S, Cloud GA. Working parents: the impact of day care
tion. advances from amniocentesis to cordocentesis congeni- and breast-feeding on cytomegalovirus infections in offspring.
tal toxoplasmosis, rubella, cytomegalovirus varicella virus, Proc Natl Acad Sci USA 1994; 91:2384 2389.
parvovirus and HIV. Ped Inf Dis 1989: 8 : 4 5 9 468. [28] Ho M. Epidemiology of cytomegalovirus infections. Rev Infect
[5] Porath A, McNutt RA, Smiely LM et al. Eflectiveness and cost Dis 1990:12 (Suppl 7): 701 710.
benefit of proposed live cytomegalovirus vaccine in the preven- [291 Chandler SH, Holmes K K , Wentworth BB et al. The epidemi-
tion of congenital disease. Rev lnf Dis 1990: 12:31 40. ology of cytomegaloviral infection in women attending a sexu-
[6] Yow MD, Demmler GJ. Congenital cytomegalo,,irus infection ally transmitted disease clinic. J Infect Dis 1985: 152:597 605.
20 years is long enough. N Engl J Med 1992: 326:702 703. [301 Handsfield HH, Chandler SH, Caine VA et al. Cy-
[7] Stern H, Tucker SM. Prospective study of cytomegalovirus in tomegalovirus infection in sex partners: evidence for sexual
pregnancy. BMJ 1973: 2 : 2 6 8 270. transmission. J Infect Dis 1985: 151:344 348.
14 Y. Daniel et al. ,' Eut+ol)ean Journal o~ Ohsletries & Qt'#wcoloAu" am/Re7~rmh.'tire Biolo~zy 63 (1995) 7 16

[31] Chandler SH, Alexander ER, Holmes KK. Epidemiology of [52] Weller TIt. Hanshaw JB. Virological and clinical observations
cytomegaloviral infection in a heterogeneous population of on cytomegalic inclusion disease. N Engl J Med 1964; 266:
pregnant women. J Infect Dis 1985; 152:249 -256. 1233 1244.
[32] Pass RF. Epidemiology and transmission of cytomegalovirus. J [53] Fattal AV. Harel S. Congenital cytomegalic virus infection.
Inf Dis 1985: 152:243 248. Harefuah 1992: 123:341 345.
[33] Huang ES, Kilpatrick BA, Huang YT et al. Detection of [54] Kamar ML, Nankervis GA. Gold E. lnapparent congenital
human cytomegalovirus and analysis of strain variation. Yale ,I cytomegalovirus infection a l\)llow-up study. N Engl J Med
Biol Med 1976: 49:29 43. 1973; 288:1370 1372.
[34] Knox GE, Pass RF, Reynolds DW el al. Comparative preva- [55] Melish ME, Hanshaw JB. Congenital cytomegaloviras infec-
lence of subclinical cytomegalovirus in genital and urinary tion: developmental progress of infants detected by routine
tracts of low-income urban women. J Infect Dis 1979: 140: screening. Am J Dis Child 1973; 126:190 194.
419--422. [56] Saigal S, Lunyk O. Larke RPB et al. The outcome in children
[35] Dworsky ME. Yow M, Stagno S e t al. Cytomegalovirus infec- with congenital cytomegalovirus infection: a longitudinal fol-
tion of breast milk and transmission in infanc). Pediatrics 1983: low-up study. Am J Dis Child 1982; 136:896 901.
72:295 299. [57] Reynokts DW, Stagno S, Strubbs KG et al. lnapparent congen-
[36] Stagno S. Reynolds DW. Huang ES et al. Congenital cy- ital cytomegalovirus infection with elevated cord IGM levels:
tomegalovirus infection: occurrence in an immune population. causal relationship with auditory and mental deficiency. N Engl
N Engl J Med 1977; 296:1254 1258. J Med 1974; 290:291 296.
[37] Stagno S, Pass RF, Dworsky ME et al. Congenital cy-
[58] Stagno S, Reynolds DW. Amos CS et al. Auditory and visual
tomegalovirus infection: the relative importance of primary and
defects restlhing from symptomatic and subclinical congenital
recurl+ent maternal infection. N Engl J Med 1982; 306:945
cytomegaloviral and toxoplasma infections. Pediatrics 1977; 59:
949.
669--678.
[38] Fowler BK, Stagno S+ Pass RF et al, The outcome of congeni-
[59] Britt W J, Vugler LG. Antiviral antibody responses in mothers
tal cytomegalovirus infection in rehltion to maternal antibody
and their newborns with clinical and subclinical congenital
status. N Engl J Med 1992: 326: 663-667.
cytomegalovirus infections. J lnf Dis 1990; 161:214 219.
[39] Stagno S, Pass RF, Dworsky ME et al. Congenital and perina-
[60] Reynolds DW, Stagno S, Hosty TS et al. Maternal cy-
tal cytomegalovirus infection. Semin Perinatol 1983: 7:31 42.
tomegalovirus excretion and perinatal infection. N Engl J Med
[40] Pass RF, Stagno S. Myers GJ et al. Outcome of symptomatic
1973: 289: I.. 5.
congenital cytomegalovirus infection: resuhs of long term longi-
[61] Stagno S, Reynolds DW, Pass RF el al. Breast milk and the
tudinal follow-up. Pediatrics 1980: 66:758 762.
risk of cytomegalovirus infection. N Engl J Med 198+,}: 302:
'[41] Cunningham FG, MacDonald PC, Gant NF et al. eds. Cy-
1073 1076.
tomegalovirus. In: Williams Obstetrics. London: Prentice Hall,
1993, 1287 1289. [62] Hayes K. Dankes DM, Gibas H el al. Cytomegalovirus in
[42] All\~rd CA, Hayes K, Britt WJ. Primary cytomegalovirus infec- hulnan milk. N Engl J Med 1972; 287:177 178.
tion in pregnancy: comparisml of antibody responses to virus- [63] Yeager AS, Grumet FC, Hafleigh EB et al. Prevention of
encoded proteins between women with and without intrauterine transfusion acquired cytomegalovirus infection in newborn in-
infection. J Infect Dis 1988; 158:917 924. [+ants. J Pediatr 1981: 98: 281-287.
[43] Pass RF, Stagno S, Britt WJ ct al. Specific cell-mediated [64] Ballard RA, Drew WL, Huflmgle KG et al. Acquired cy-
immunit+v and the natural history of" congenital infection with tomegalovirus infection in preterm infants. Am J Dis Child
cytomegalovirus. J Infect Dis 1983: 148:953 961. 1979: 133:482 485.
[44] Starr SE. Tolpm MD. Friedman ItM et al. hnpaired cellular [65] Brasfield DM, Stagno S, Whithley RJ et al. Infant pneumonitis
imnmnity to cytyomegalovirus in congenital infected chiklren associated with cytomegalovirus, chlamydia pneumocystis and
and their mothers. J Infect Dis 1979: 140:500 505. urcaplasma: follow-up. Pediatrics 1987: 79:76 83.
[45] Reynolds DW, Dean PH, Pass RF el al. Specific cell-mediated [66] Stagno S. Reynolds DW. Tsiantos A e t al. Comparative serial
inm+tanity in children with congenital at+u.l neonatal cy- virologic and serologic studies of symptomatic and subclinical
tomegalovifus infection and their mothers. J Infect Dis 1979; congenital and natally acquired cytomegalo,Arus infection. J
140:493 499. Infect Dis 1975: 132:568 577.
[46] Gehrz RC, Marker SC, Knorr SO, Balfour HH Jf et at. Specific [67] Slagno S, Pass RF+ Dworsky ME et al. Congenital and perina-
cell mediated immune defect in active cytomegalovirus infection tal cytomegalovirus infection, clinical characteristics and patho-
of young children and their mothers. Lancet 1977: 2:844 847. genic factors. Birth defects. CMV. Pathos Prey Hum Infect
[47] Stern H, Hannington J. Booth J et al. An early lnarker of fetal 1984" 20:60 85.
infection after primary cytomegalovirus infection in pregnancy. [68] Hanshaw JB, Steinfeld H J, White CJ. Fluorescent-antibody test
Br Med J Clin Res 1986: 292:718 720. for eytomegalovirus macroglobulin. N Engl J Med 1968; 279:
[48] Fernando S, Pearce ,IM, Booth JC. Lymphocyte responscs and 566 -. 570.
virus excretion as risk factors for intrauterine infection with [69] Stagno S. Pass RF. Reynolds DW et al. Comparative study of
cytomegalovirus. J Med Virol 1993: 41:108 113. diagnostic procedures lor congenital cytomegalovirus infection.
[49] Sinzger C, Muntefering tt. Lonmg T et al. Cell types infected in Pediatrics 1980; 65:251 257.
human cytomegalovirus placentitis identified by immunohisto- [70] Pereira L, Stagno S. Hoffman M e t al. Cytomegalovirus-in-
chemical doable staining. Virchows Arch A Pathol Anat Histo- fected cell polypeptides immune-precipitated by sera from chil-
pathoi 1993: 423:249 256. dren with congcnital and perinatal infections, lnt'ect lmmun
[50] AIford CA, Stagno S. Pass RF et al. Epidemiology of cy- 1983: 39:100 108.
tOlnegalovirus. In: Nahmias A J, Dowdle WR, Schinazi RF, eds. [71] Pass RF. Dworsky ME, Whitley RJ et al. Specific lymphocyte
The Human Herpes Virus: An Interdisciplinary t:'erspective. blastogenic responses in children with cytomegalovirus and
New York: Elsevier, 1981:159 171. herpes simplex virus infections acquired early in infancy. Infect
[51] Schopfer K. Lauber E, Krech U. Congenital cytomegalovirus hnmun 1981; 34:166 170:
infection in newborn infimts ol +mothers infected before preg- [72] Chou S. Newel- methods for diagnosis of cytomegalovirus
nancy. Arch Dis Child 1978: 53: 536--539. infection. Rev Infect Dis 1990:12 (Suppl 7): 727 736.
 Y. Daniel et al. / European Journal o f Obstetrics & Gynecology and Reproductive Biology 63 (1995) 7 16 15

[73] Griffiths PD, Stagno S, Pass RF et al. Infection with cy- [93] Davis LE, Tweed GV, Stewards JA et al. Cytomegalovirus
tomegalovirus during pregnancy: specific lgM antibodies as a mononucleosis in a first trimester pregnant female with trans-
marker of recent primary infection. J Infect Dis 1982; 145: mission to the fetus. Pediatrics 197l; 48:200 206.
647-653. [94] Lamy EM, Mulongo NK, Gadisseux FJ et al. Prenatal diagno-
[74] Stagno S, Tinker MK, Elrod C et al. IgM antibodies detected sis of fetal cytomegalovirus infection. Am J Obstet Gynecol
by ELISA and RIA in the diagnosis of cytomegalovirus infec- 1992; 66: 9l 94.
tions in pregnant women and newborn infants. J Clin Microbiol [95] Hohlfeld P, Vial Y, Brignon MC et al. Cytomegalovirus fetal
1985; 21:930 935. infection. Prenatal diagnosis. Obstet Gynecol 1991; 78:615
[75] Daiminger A, Schalasta G, Betzl D, Enders G. Detection of 618.
human cytomegalovirus in urine samples by cell culture, early [96] Stirk PR, Griffiths PD. Use of monoclonal antibodies for the
antigen assay and polymerase chain reaction. Infection 1994: diagnosis of cytomegalovirus infection by detection of early
22:24 28. antigen fluorescent foci in cell culture. J Med Virol 1987: 21:
[76] Stagno S, Pass RF, Alford CA. Perinatal infections and malde- 329 337.
velopment. In: Bloom AD, James LS, eds. The Fetus and the [97] Grose C. Meehan T, Weiner CP. Prenatal diagnosis of congen-
Newborn. New York: Alan R. Liss, 1981; 31-50. ital cytomegalovirus infection by virus isolation following am-
[77] Gleaves CA, Smith TF. Shuster EA, Pearson GR. Rapid niocentesis. Pediatr Infect Dis J 1992; 11:605 607.
detection of cytomegalovirus in MRC-5 cells inoculated with [98] Stagno S, Pass RF, Thomas JP et al. Defects of tooth structure
urine specimens by using low-speed centrifugation and mono- in congenital cytomegalovirus infection. Pediatrics 1982; 69:
clonal antibody to an early antigen. J Clin Microbiol 1984; 19: 646 648.
917 919. [99] Stagno S, Reynolds DW, Amos CS et al. Auditory and visual
[78] Gleaves CA, Smith TF, Shuster EA, Pearson GR. Comparison defects resulting from symptomatic and subclinical congenital
of standard tube and shell vial cell culture techniques for the cytomegalovirus and toxoplasma infections. Pediatrics 1977; 59:
detection of cytomegalovirus in clinical specimens. J Clin Mi- 669 678.
crobiol 1985: 21:217 221. [100] Voision GA. Edelman P, Genetet N e t al. Immunologie de la
[79] Swenson PD, Kaplan MH. Rapid detection of cytomegalovirus reproduction. Paris: Flammarion 1990; 435-439.
in cell culture by indirect immunoperoxidase staining with [101] Ahlfors K, Forsgren M, Griffiths P e t al. Comparison of four
monoclonal antibody to an early nuclear antigen. J C[in Micro- serological tests for the detection of specific immunoglobulin M
biol 1985; 21:669 673. in the cord sera of infants congenitally infected with cy-
[80] Paya CV, Wold AD, Smith TF. Detection of cytomegalovirus tomegalovirus. Scand J lnfect Dis 1987; 19:303 308.
infections in specimens other than urine by the shell vial assay [102] Meisel R. Fetal cytomegalovirus infection. A case report. Am J
and conventional tube cell cultures. J Clin Microbiol 1987; 25: Obstet Gynecol 1990; 162:663 664.
755 -757. [103] Weiner CP, Grose C. Prenatal diagnosis of congenital cy-
[81] Chou S, Merigan TC. Rapid detection and quantitation of tomegalovirus infection. Two decades later. Am J Obstet Gy-
human cytomegalovirus in urine through DNA hybridization. necol 1990; 163:447 450.
N Engl J Meal 1983: 308:921-925. [104] Dong ZW, Yan C, Yi W e t al. Detection of congenital cy-
[82] Schuster V. Matz B, Wiegand H, Traub B, Kampa D Neu- tomegalovirus infection by using chorionic villi of the early
mann-Haefelin D. Detection of human cytomegalovirus in pregnancy and polymerase chain reaction, lnt J Gynaecol Ob-
urine by DNA-DNA and RNA-DNA hybridization. J Infect stet 1994: 44:229 231.
Dis 1986; 154:309 314. [105] Hogge WA, Buffone, Hogge JS. Prenatal diagnosis of cy-
[83] Spector SA, Rua JA, Spector DH, McMillan R. Detection of tomegalovirus (CMV) infection: a preliminary report. Prenat
human cytomegalovirus in clinical specimens by DNA-DNA Diagn 1993; 13:131 136.
hybridization. J Infect Dis 1984; 150:121 126. [106] Pass RF. Commentary. Is there a role for prenatal diagnosis of
[84] Unger ER, Budgeon LR, Myerson D, Brigati DJ. Viral diagno- congenital cytomegalovirus infection'? Pediatr Infect Dis J 1992:
sis by in situ hybridization. Description of a rapid simplified 11: 608---609.
colorimetric method. Am J Surg Pathol 1986: 10:1 8. [107] Donner C, Liesnard C, Content J, Busine A, Aderca J Rodesch
[85] Kevin AA, Faro S. Virus, bacteria and protozoans in preg- F. Prenatal diagnosis of 52 pregnancies at risk for congenital
nancy: a sample of each. Clin Obstet Gynecol 1993: 36:878 cytomegalovirus infection. Obstet Gynecol 1993; 82:481 486.
885. [108] Alford CA. Chronic intrauterine and perinatal infections. In:
[86] Editorial. Screening for congenital cytomegalovirus. Lancet Galasso G J, Merigan TC, Buchanan RA, eds. Antiviral agents
1989; If: 599 600. and viral disease of man. New York: Raven Press, 1984;
[87] Jeffries JD. Cytomegalovirus infections in pregnancy. Br J 433 486.
Obstet Gynecot 1984: 91:305 306. [109] Collaborative DHPG treatment study group. Treatment of
[88] Peckham CS, Coleman JC, Hurley R et al. Cytomegalovirus serious cytomegalovirus infection with 9-(1,3-Dihydroxy-2-Pro-
infection in pregnancy: preliminary findings from a prospective poxymethyl) guanine in patients with AIDS and other im-
study. Lancet 1983; 1:1352 1355. munodeficiencies. N Engl J Med 1986; 314:801 805.
[89] Adler SP. Cytomegalovirus and pregnancy. Curr Opin Obstet [110] Kotler DP, Culpepper-Morgan JA, Tierney AR et al. Treat-
Gynecol 1992; 4:670 675. ment of disseminated cytomegalovirus infection with 9-(1,3-Di-
[90] Tassin GB, Maklad NF, Stewart RR et al. Cytomegalic inclu- hydroxy-2-Propoxymethyl) guanine: Evidence of prolonged
sion disease: intrauterine sonographic diagnosis using findings survival in patients with AIDS. AIDS Res 1986: 2:229 308.
involving the brain. AJNR Am J Neuroradiol 1991; 12:117 [111] Holland GN, Sakamoto M J, Hardy D et al. UCLA CMV
122. retinopathy study group treatment of cytomegalovirus retinopa-
[91] Forouzan 1. Fetalabdominal echogenic mass: an early sign of thy in patients with AIDS: use of the experimental drug 9-(l,3-
intrauterine cytomegalovirus infection. Obstet Gynecol 1992: Dihydroxy-2-Propoxymethyl) guanine. Arch Opthalmol 1986;
80:535 537. 104:1794-1800.
[92] Estroff JA, Parad RB, Teele RL et al. Echogenic vessels in the [112] Elek SD, Stern H. Development of vaccine against mental
fetal thalami and basal ganglia associated with cytomegalovirus retardation caused by cytomegalovirus infection in utero.
infection. J Ultrasound Med 1992; 11:686 688. Lancet 1974: 1:1 5.
16 Y. Daniel et al. ~ European Jourmd o! Obstetrics & Gvnecolo,~,~l' and R~7~rorh.'lit'e Biology 63 (1995) I 16

[113] Plotkin SA, Furukawa T. Zygraich N et al. Candidate cy- [118] Balfour HH Jr, Sachs GW, Welo P e t al. Cytomegalovirus
tomegalovirus strain for h u m a n vaccination. Infect l m m u n vaccine in renal transplant candidates: progress report of a
1975: 12:521 527. randomized, placebo-controlled, double-blind trial. Birth de-
[114] Plotkin SA, Friedman HM, Starr SE et al. Prevention and fects. CMV. Pathog Prey Hum Infect 1984; 2(1:289 304.
treatment of cytomegalovirus infection. In: Nahmias AJ
[119] Gibson W, lrmiere A. Selection of particles and proteins for use
Dowdle WR. Schinazi. RF, eds. The H u m a n Herpes Virus. An
as h u m a n cytomegalovirus subunit vaccines. Birth defects.
Interdisciplinary Perspective. New York: Elsevier, 1981:403
413. CMV. Pathog Prey Hum Infect 1984: 20:305 324.
[115] Plotkin SA. Farquhar J, Horenberger E. Clinical trials o1" [120] Britt WJ, Vugler L, Butfiloski EJ el al. Cell surface expression
immunization with the Towne 125 strain of h u m a n cy- ot" h u m a n cytomegalovirus gP55-116 (gB): Use of HCMV-re-
tomegalovirus. J Infect Dis 1976; 134:470 475. combinant vaccinia virus-infected cells in analysis of h u m a n
[116] Stern H. Live cylomegalovirus vaccination of healthy volun- neutralizing antibody response. J Virol 1990: 64:1079 1085.
teers. Eight years follow-up studies. Birth defects. CMV.
[121] Marshall GS, Ricciardi RP. Rando RF et al. An adenovirus
Pathog Prey H u m Infect 1984: 20:263 269.
recombinant that expresses the h u m a n cytomegalovirus major
[117] Plotkin SA. Smiley ML, Friedman HM et al. Prevention of
cytomegalovirus disease by Towne strain live attenuated vac- envelope glycoprotein and induces neutralizing antibodies. J
cine. Birth defects. CMV. Pathog Prey H u m Infect 1984: 20: Infect Dis 1990: 162:1177 II81.
271 287.