Diagnosis and Management of CMV Infection in
Patrick Duff, M.D.
Perinatology 2010; 1:1-6
A twenty-seven-year-old primigravid woman had an ultrasound
examination at twenty-four weeks gestation because her uterus was smaller than
expected for dates. Of note, the placenta was moderately thickened, measuring
4.8 cm in diameter.
The amniotic fluid volume was in the low-normal range. The
estimated fetal weight was in the 15th percentile for gestational age, and the fetal bowel was highly echogenic.
abnormalities were noted. Upon detailed questioning, the patient indicated that
she had had a prolonged “flu-like” illness at approximately fourteen weeks’
gestation. No treatment had been administered for this illness. The patient
presently works as a pre-school teacher.
What is the most likely explanation for the ultrasound findings noted
above? What are the most appropriate steps in the evaluation of this patient?
Dr. Duff is Professor of Clinical Obstetrics and Gynecology, Division of
Maternal-Fetal Medicine at the University of Florida College of Medicine.
Cytomegalovirus is a DNA virus, and humans are its only known host.
Like other members of the herpes virus family, CMV can cause latent infection.
CMV may be transmitted horizontally as a result of organ donation,
blood transfusion, sexual contact, and contact with infected saliva and urine.
At least 50% of women of reproductive age have evidence of prior CMV infection.
The prevalence of prior infection is increased in women of lower socioeconomic
status. Most pregnant women acquire infection as a result of contact with their
own younger children or children in a daycare or pre-school setting.
Infants may acquire CMV infection by contact with contaminated blood
and genital secretions during delivery and via breast milk after delivery.
Fortunately, these two mechanisms of transmission are extremely unlikely to
cause any injury to the baby. However, when a pregnant woman develops primary
CMV infection in the first half of pregnancy, the risk of transplacental
infection is approximately 40%. Of these babies, 5 to 15% are acutely
symptomatic at birth. The principal clinical manifestations of severe
congenital CMV infection include hepatosplenomegaly, thrombocytopenia with
resultant petechiae, intracranial calcifications, intrauterine growth
restriction, hepatitis and jaundice, microcephaly, chorioretinitis, hearing
loss, mental retardation, and seizures. Of these severely infected infants, up
to 30% die, and approximately 80% of the survivors have serious sequelae.
Eighty-five to 95% of these infected infants are asymptomatic at birth. Ten to
15% of the initially asymptomatic children subsequently develop neurologic,
auditory, visual, and/or dental defects, which become evident as they enter
Interestingly, when primary maternal infection occurs in the third trimester of
pregnancy, the risk of transplacental transmission is much higher – 75 to 80%.
However, the risk of serious fetal injury is very low.
In contrast to primary maternal infection, recurrent or reactivated
maternal infection does not pose nearly so serious a risk to the fetus. When
recurrent or reactivated CMV infection develops during pregnancy, only 5 to 10%
of infants become infected. None of these children are symptomatic at birth.
Late sequela of infection include minor visual and auditory deficits and
developmental delays which become apparent as the child enters elementary
Diagnosis of CMV in the Mother
Some women who contract CMV infection will be asymptomatic, and the
diagnosis will only be suspected because abnormal fetal ultrasound findings have
been identified. When clinical manifestations are present, they include:
malaise, fever, generalized lymphadenopathy, and hepatosplenomegaly. Patients
who are immunocompromised may develop extremely serious sequelae of infection,
including chorioretinitis and pneumonitis.
A patient with suspected CMV infection should have a serologic assay
for IgM and IgG antibody. Typically, in the acute phase of the infection, the
IgM antibody is positive and the IgG antibody is negative. As the illness
evolves, IgG antibody becomes detectable. Patients who are acutely infected
also will usually test positive by culture or PCR for virus in the blood and
Clinicians must be aware that the initial serology for CMV can be
confusing because the IgM antibody may remain positive for 9 to 12 months after
an acute infection. Moreover, there can be anamnestic increases in IgM in the
face of a recurrent or reactivated infection. Therefore, IgG avidity testing
can be extremely helpful in differentiating between acute and chronic
infection. In the setting of acute infection, the avidity of IgG antibody for
the virus is low. In a recurrent or reactivated infection, IgG avidity is
high.2 Table 1 summarizes the key tests for diagnosing CMV infection in the
Table 1. Key Tests for the Diagnosis of CMV Infection in the Mother
Type of Infection
Absent on low
May be positive
but secondary response can occur
Positive for high
Diagnosis of Congenital Infection
The single best test for the diagnosis of congenital infection is
detection of virus in the amniotic fluid by culture or PCR. In an important
early study, Donner et al3 noted that amniocentesis correctly identified 12 of
13 (92%) infants with congenital CMV infection. In a subsequent
investigation, Lipitz and associates4 noted that amniocentesis was 100%
sensitive in diagnosing congenital CMV infection. In a more recent
investigation, Azam and associates5 showed that amniocentesis was 77% sensitive
in detecting congenital CMV infection; the specificity was 100%.
Cordocentesis has been suggested as a possible diagnostic test because
it permits the assessment of total IgM antibody and IgM-specific antibody in
umbilical cord blood. However, from a technical perspective, this test cannot
be performed at a gestational age much less than 20 weeks. Moreover, the fetus
does not develop an immunologic response to viruses such as CMV until
approximately 23 to 24 weeks’ gestation. Therefore, this test is not of
practical value in evaluating patients in the critical first half of pregnancy.
Detection of CMV in amniotic fluid by culture or PCR accurately
identifies an infected fetus. However, a comprehensive ultrasound examination
is then invaluable in determining the extent of fetal injury. The principal
ultrasound findings associated with congenital CMV infection are placentomegaly,
intrauterine growth restriction, microcephaly, ventriculomegaly, periventricular
calcifications, isolated serous effusions, and echogenic bowel.
Management of the Patient with Congenital CMV Infection
Once the diagnosis of congenital CMV infection is confirmed,
one option is pregnancy termination. A second proposed option has been
treatment of the mother with antiviral agents such as ganciclovir, foscarnet,
and cidofovir. These drugs are of moderate effectiveness in treating CMV
infection in the adult, particularly the immunocompromised patient. However,
they are not of proven value in preventing or treating congenital CMV
The most promising therapy for congenital CMV infection appears to be
hyperimmune globulin.6 In one of the earliest reports of this therapy, Nigro et
al7 described the use of hyperimmune globulin for treatment of a mother who had
a twin pregnancy, discordant for congenital CMV infection (Level III evidence).
Treatment was administered at 22 weeks’ gestation. The patient received hyperimmune globulin, 200 units/kg/day intravenously for three days. A separate
dose of 400 units of hyperimmune globulin was injected intra-amniotically into
the sac of the affected twin. The authors noted that treatment resulted in
decreased placental edema, improvement in fetal growth, an increase in IgG
avidity, and an increase in maternal cell-mediated immunity. At nine months of
age, both infants were negative for CMV.
In a subsequent widely publicized investigation8, this same group of
authors reported a prospective cohort study (Level II evidence) in which
patients received either treatment
with hyperimmune globulin,
with hyperimmune globulin, or no therapy. The study was performed at eight Italian
medical centers and included 157 pregnant women with serologically-confirmed
primary CMV infection. Of these women, 148 were asymptomatic and were
identified by routine serologic screening. Eight women had symptomatic viral
infection consistent with CMV, and one was identified because her fetus had
abnormal findings on ultrasonography. Forty-five women had a primary infection
more than six weeks before enrollment, underwent amniocentesis, and had CMV
detected in amniotic fluid by PCR or culture. Thirty-one of these women elected
to receive intravenous treatment with CMV-specific hyperimmune globulin (200
units/kg of body weight). Nine of the 31 received one or two additional
infusions of hyperimmune globulin into either the amniotic fluid or umbilical
cord because of persistent fetal abnormalities on ultrasonography. Fourteen
women with positive amniotic tests (culture or PCR ) for CMV declined
treatment. Seven of these individuals had infants who were symptomatic at
delivery. In contrast, only one of the 31 treated women had an infant with
clinical CMV disease at birth (adjusted odds ratio, 0.02; p < 0.001). This
treatment outcome was particularly impressive given that 15 of the women had
fetuses with unmistakable ultrasonographic evidence of CMV disease.
Eighty-four additional women did not undergo amniocentesis because
their infection occurred within six weeks before enrollment, their gestational
age was less than 20 weeks, or they declined the procedure. Thirty-seven of
these women received 100 units of hyperimmune globulin/kg every month until
delivery; the range of infusions was four to six. Forty-seven women declined
treatment. Among the treated women, 6 delivered infected infants as compared
with 19 of the untreated women (adjusted odds ratio, 0.32; p = 0.04). No
adverse effects of hyperimmune globulin were noted in either group receiving
These results are clinically impressive and biological plausible.
However, this investigation had several weaknesses.9 The design of the study
was not optimal. Although prospective, the study was not randomized and
controlled. The lack of a strict prospective randomized protocol resulted in a
curious blend of two cohorts—a treatment group and a prevention group. The
regimens of hyperimmune globulin were quite different in the two groups. In
actual fact, women in the prevention group received more intensive therapy (that
is, higher cumulative doses of hyperimmune globulin) than those in the treatment
The authors also did not address the financial and logistic issues associated
with screening large populations for CMV infection, triaging patients with
inevitable false-positive results, offering amniocentesis and comprehensive
ultrasound examinations to women who seroconvert, and then treating at-risk
women with hyperimmune globulin. The wholesale acquisition cost of hyperimmune
globulin at our facility is $850 for a 2.5 g vial. At a dose of 200 units
(mg)/kg, a 70 kg patient will require 14 grams of hyperimmune globulin for a
single treatment dose ($4760) and 7 grams for a prophylactic dose ($2380).
In a follow-up investigation, members of the same laboratory reported
a retrospective cohort study10
(Level II evidence) of 32 women in the previous
investigation who had received hyperimmune globulin. For women both with, and
without, an obviously injured baby, treatment with hyperimmune globulin resulted
in statistically significant reductions in placental thickness. This response
to treatment is of considerable importance, given the authors’ postulate that
placental thickening, leading to uteroplacental insufficiency, is the major
mechanism for fetal injury with congenital CMV infection.
Prevention of CMV Infection
Pregnant women who are susceptible to CMV infection should be advised of
the importance of careful handwashing and cleansing of environmental surfaces
when interacting with young children. Women who work in a day care or preschool
setting should particularly be urged to wear gloves when changing the diapers of
children and to wash their hands carefully after each diaper change. They also
should wash their hands after handling any toys or other objects that the
children might have placed in their mouths. Pregnant women who require blood
transfusions should always receive CMV-negative blood, and this same blood type
should be used for any intrauterine transfusion. In addition, because of the
possibility of CMV transmission through sexual intercourse, pregnant women
should be urged to adopt safe sexual practices if they are not engaged in a
mutually faithful monogamous relationship.
Clearly, development of an effective vaccine for prevention of CMV
infection would be of great importance. In this light, the recent publication
by Pass et al11 is extremely encouraging. These authors conducted a Phase II
placebo-controlled, randomized, double blind trial (Level I evidence) of a new CMV vaccine. This vaccine was prepared by recombinant technology and contained
envelope glycoprotein B along with the MF 59 adjuvant. Women were eligible for
the study if they were seronegative for CMV antibody, in good health, fourteen
to forty years old, and not pregnant or lactating. Participants received three
doses of vaccine or placebo at 0, 1, and 6 months. The women then underwent
serologic testing for CMV infection every three months for up to 42 months. The
serologic assay tested for IgG antibodies directed against viral proteins other
than glycoprotein B. Infection was confirmed by viral culture or immunoblotting.
The primary endpoint of the study was time until detection of CMV infection.
The vaccine was administered to 234 patients, and 230 women received
placebo. Eighteen infections occurred in the vaccine group, compared with 31 in
the placebo group. Vaccinated women were more likely to remain uninfected
during the follow-up period (p = .02). The observed efficacy of the vaccine was
50%; the 95% confidence interval was 7 to 73%. One of 81 infants (1%)
subsequently born to mothers in the vaccinated group had congenital CMV
infection compared with three of 97 (3%) infants born to infants in the placebo
group (p = .41). One infant in the placebo group had severe infection that was
evident at birth. The other three infants were asymptomatic at birth and free
of sequelae three to five years later.
The goal of a large scale vaccination program is to ensure that women
enter reproductive age with pre-existing immunity to CMV infection. In that
light, the study by Pass and colleagues11 is only partially encouraging.
Despite vaccination, 18 infections occurred, and the follow-up period was
relatively short. We do not yet know whether the protective effect of the
vaccine will be of extended duration. Moreover, it was discouraging to note
that one vaccinated mother delivered an infant who actually had congenital CMV
A detailed algorithm for diagnosis and management of CMV infection in
pregnancy is presented in Figure 1.
Algorithm for the Diagnosis and
Management of Congenital CMV Infection
The patient presented in the clinical case at the beginning of this
article should have a serologic assay for IgM and IgG antibody. IgG avidity
testing also should be performed. In addition, she should have a culture or PCR
test for virus in serum and urine. She should also have an amniocentesis, and
the amniotic fluid should be tested for CMV by either culture or PCR. Given
that there already are ultrasound findings suggestive of CMV infection, I
believe that treatment with hyperimmune globulin, at a dose of 200 units (mg)/kg
should be administered. A follow-up ultrasound examination should be performed
in approximately two weeks. If the abnormal ultrasound findings have resolved,
no additional treatment with hyperimmune globulin is indicated. However, if the
ultrasound findings persist, an additional dose of hyperimmune globulin should
1. Fowler KB, Stagno S, Pass RF, Britt WJ, Boll TJ, Alford CA. The outcome of
congenital cytomegalovirus infection in relation to maternal antibody status. N Engl J Med 1992;326:663-7.
2. Duff P. A thoughtful algorithm for the accurate diagnosis of primary CMV
infection in pregnancy. Am J Obstet Gynecol March 2007, pp. 196-7.
3. Donner C, Liesnard C, Content J, Busine A, Aderca J, Rodesch F. Prenatal
diagnosis of 52 pregnancies at risk for congenital cytomegalovirus infection.
Obstet Gynecol 1993;82:481-6.
4. Lipitz S, Yagel S, Shalev E, Achiron R, Mashiach S, Schiff E. Prenatal
diagnosis of fetal primary cytomegalovirus infection. Obstet Gynecol
5. Azam AZ, Vial Y, Fawer CL, Zuffrey J, Hohlfeld P. Prenatal diagnosis of
congenital cytomegalovirus infection. Obstet Gynecol 2001;97:443-8.
6. Adler SP, Nigro G, Pereira L. Recent advances in the prevention and treatment
of congenital cytomegalovirus infections. Semin Perinatol 2007;31:10-18.
7. Nigro G, LaTorre, Anceschi MM, Mazzocco M, Cosmi EV. Hyperimmunoglobulin
therapy for a twin fetus with cytomegalovirus infection and growth restriction.
Am J Obstet Gynecol 1999;180:1222-6.
8. Nigro G, Adler SP, LaTorre R, Best AM. Passive immunization during pregnancy
for congenital cytomegalovirus infection. N Engl J Med 2005;353:1350-62.
9. Duff P. Immunotherapy for congenital cytomegalovirus infection. N Engl J Med
10. LaTorre R, Nigro G, Mazzocco M, Best AM, Adler SP. Placental enlargement in
women with primary maternal cytomegalovirus infection is associated with fetal
and neonatal disease. Clin Infect Dis 2006;43:994-1000.
11. Pass RF, Zhang C, Evans A, Simpson T, Andrews W, Huang ML, et al. Vaccine
prevention of maternal cytomegalovirus infection. N Engl J Med 2009;360:1191-9.