JOURNAL OF CLINICAL MICROBIOLOGY, Mar. 2009, p. 660–665 0095-1137/09/$08.00⫹0 doi:10.1128/JCM.01576-08 Copyright © 2009, American Society for Microbiology. All Rights Reserved.
Vol. 47, No. 3
Clinical Predictive Value of Real-Time PCR Quantification of Human Cytomegalovirus DNA in Amniotic Fluid Samples䌤 T. Goegebuer,1 B. Van Meensel,1 K. Beuselinck,1 V. Cossey,2 M. Van Ranst,1 M. Hanssens,3 and K. Lagrou1* Department of Medical Diagnostic Sciences,1 Department of Neonatology,2 and Department of Gynaecology,3 University Hospitals Leuven, Leuven, Belgium Received 14 August 2008/Returned for modification 24 September 2008/Accepted 15 December 2008
The aim of this study was to evaluate the diagnostic reliability and prognostic significance of the quantification of cytomegalovirus (CMV) DNA in amniotic fluid (AF). We retrospectively reviewed the results for 282 amniotic fluid samples that had been tested for CMV by a quantitative real-time PCR. We observed three cases in which no CMV genomes were detected in the AF but in which the children were nevertheless congenitally infected. Hence, we conclude that a negative result by PCR for CMV in AF cannot rule out the possibility of congenital infection. No false-positive PCR results were observed. A correlation between the CMV viral load in AF and the fetal and neonatal outcomes could not be demonstrated in our study. Instead, a correlation was found between the CMV viral load and the gestational age at the time of amniocentesis. allow any conclusions to be drawn (28). Despite the drawbacks of the diagnosis and treatment of a congenital CMV infection, gynecologists do screen their patients for CMV (18). Supporters of routine prenatal screening argue that the use of precautionary hygienic measures can be suggested to CMV-seronegative pregnant women. Otherwise, the knowledge of a CMV seroconversion during pregnancy can lead to an intensified follow-up by additional prenatal diagnostic procedures as well as ultrasound and/or nuclear magnetic resonance imaging, providing the mother to be with information and arguments that she can use to choose whether or not to terminate the pregnancy (15, 32). Opponents of routine prenatal screening fear that a positive serology for CMV will too often lead to needless abortions. Thus, great responsibility lies on health care professionals to provide sufficient and relevant information regarding test results, further options, and risks and to assist the pregnant women with making a decision consistent with their values (32). Two questions are to be considered with regard to the antenatal diagnosis of congenital CMV infection. The first question is whether the fetus is infected, and the second is whether the fetus is symptomatic and, if so, to what extent. The detection of CMV in amniotic fluid (AF) by viral culture or PCR is said to be effective in differentiating uninfected from infected fetuses (17, 22, 32). Because of a delay in the transplacental passage of the virus and because renal immaturity in the fetus before 21 weeks of gestation prevents its elimination into the AF, it is recommended that amniocentesis be performed after 21 weeks of gestation and following a time interval of at least 6 weeks after the maternal infection has been diagnosed in order to obtain more reliable results (8, 9, 23, 25). The value of the results of quantitative PCR for CMV in AF as a prognostic indicator of symptomatic congenital infection is even more controversial. Some authors have found that a high CMV viral load in AF is associated with a high risk of symptomatic infection in the fetus (13, 15, 19, 20). Others, however, failed to demonstrate such an association (27, 29, 36), whereas still others proposed a possible association between the viral load
Human cytomegalovirus (CMV) is the leading cause of congenital viral infection in developed countries, with the reported incidence varying between 0.2 and 2.2% of all live births (15, 35). The transmission rate following primary infection of the mother is about 40%. Only 10 to 15% of the CMV-infected children are symptomatic at birth, and the symptoms range from mild to life-threatening disease. The remaining 85 to 90% of the children are asymptomatic at birth, but 10% of them will develop complications later on, mainly neurodevelopmental defects and sensorineural hearing loss. Among pregnant women with recurrent infection, the rate of transmission to the infant is about 1%. Despite a preexisting immunity in the mother, epidemiological data suggest that the frequency and the severity of symptoms might be in the same range as those for a primary CMV infection (11, 12). The issue of whether pregnant women should be screened for CMV during pregnancy has been debated for many years, but no consensus has been agreed upon (6). None of the current international guidelines recommend routine serologic screening of pregnant women (1, 7, 16, 23, 26). Indeed, there is no prognostic marker in the mother to predict whether the virus will be transmitted to the fetus (32). To obtain more information, invasive prenatal diagnostic techniques, such as amniocentesis or cordocentesis, have been used. Moreover, CMV infection of the fetus can lead to a great variety of clinical and biological conditions, but there is no reliable marker that can be used to predict which infected fetuses will have serious sequelae (32, 33). Finally, no vaccine or prophylactic therapy is available at present (24, 32). Nigro et al. examined whether CMV-specific hyperimmune globulin therapy could be useful for the prevention and treatment of congenital CMV infection, yet the results of the study did not
* Corresponding author. Mailing address: Department of Medical Diagnostic Sciences, University Hospitals Leuven, Herestraat 49, Leuven 3000, Belgium. Phone: 32-16-33-70-19. Fax: 32-16-33-79-31. Email:
[email protected]. 䌤 Published ahead of print on 24 December 2008. 660
VOL. 47, 2009
QUANTIFICATION OF HUMAN CMV DNA IN AMNIOTIC FLUID
in AF and the gestational age at the time of amniocentesis (13, 29). In an attempt to clarify these discordant findings, we conducted a retrospective study and assessed the clinical predictive value of the quantification of human CMV DNA in AF samples by real-time PCR. The following two questions were to be answered: (i) is PCR for the detection of CMV in AF useful for the diagnosis of congenital CMV infection, and (ii) can quantification of the virus in AF aid with establishment of a prognosis for the infected fetus?
MATERIALS AND METHODS Patients and samples. All AF samples tested for CMV by PCR between November 2002 and October 2006 were traced back. Of the 282 samples tested, 241 had a negative result for CMV and 41 samples had a positive result for CMV. The timing of amniocentesis varied from 15.5 to 32.1 weeks of gestation, with the mean duration of pregnancy being 21.6 weeks. After receiving written permission from the patients concerned to obtain more information on the course of the pregnancy and the outcome of the child, we contacted the gynecologists who cared for the mothers and the pediatricians who cared for the children. A structured questionnaire was used to get clinical data, laboratory test results, pathology test results, and imaging study reports. Newborns were classified as uninfected or infected on the basis of isolation of the virus from urine sampled within the first week of life and aborted fetuses on the basis of histological tissue examination. Infected newborns were further classified as symptomatic or asymptomatic on the basis of the presence or the absence (at birth) of one or more of the following findings: preterm birth (⬍37 weeks of gestation); small size for gestational age (⬍3rd percentile); and the presence of petechiae or purpura, hepatosplenomegaly, central nervous system (CNS) abnormalities, elevated liver enzyme levels (alanine aminotransferase level, ⬎80 U/liter), thrombocytopenia (⬍100 ⫻103/mm3), or conjugated hyperbilirubinemia (⬎2 mg/dl). Imaging study results and the results of hearing tests, if they were performed, were collected as well. Finally, the pediatricians were asked to provide information on the child’s current health status, in particular, on the presence of neurologic disturbances, delays in psychomotor and/or mental developmental status, and CMV-related audiological or visual problems. At the end of the study, the children were between 2.5 months and 3 years of age. The study was approved by the University Hospitals Leuven Ethics Committee. Shell vial assay for CMV isolation. Flat-bottom tubes were seeded with embryonic skin and muscle cells suspended in minimal essential medium with 10% fetal calf serum to grow a monolayer. After a minimum of 24 h, the medium was removed and 500 l filtered urine was added. The flat-bottom tubes were centrifuged at 700 ⫻ g for 1 h at room temperature and were then incubated at 37°C for 15 to 24 h. The samples were evaluated for the presence of the immediateearly antigen by an immunofluorescence assay by subsequently adding monoclonal antibodies directed against the immediate-early antigen (clone E13; Argene SA, France), biotinylated polyclonal rabbit anti-mouse antibodies (Dako Diagnostics, Switzerland), fluorescein-streptavidin (Amersham Bioscience), and Evans blue. Extraction. The extraction of CMV DNA from AF was carried out on an easyMAG instument (bioMe´rieux, Marcy l’Etoile, France) by using generic protocol 1.0.6 and a sample volume of 220 l. An internal control (6 ⫻109 copies) and proteinase K were added to the sample before initiation of the protocol. The elution volume was 110 l. Real-time PCR for the CMV MCP gene. An in-house real-time PCR assay for the detection of DNA (4) was performed on an ABI 7900 real-time thermocycler (Applied Biosystems). Real-time PCR was carried out in a reaction volume of 40 l containing 10 l of the DNA extract, 20 l 2⫻ Universal Mastermix (Applied Biosystems), and primers and a probe targeting the CMV major capsid protein (MCP): 0.25 M primer CMCP11 (5⬘-CGTAACGTGGACCTGACGTTT-3⬘), 0.25 M primer CMCP12 (5⬘-CACGGTCCCGGTTTAGCA-3⬘), and 0.20 M probe CMCP3TM (6-carboxyfluorescein–5⬘-TATCTGCCCGAGGATCGCGG TTACA-3⬘–6-carboxytetramethylrhodamine). The limit of detection was 10 copies per reaction, which correlates to 500 copies of CMV per ml AF. Real-time PCR of the CMV glycoprotein H (gH) gene. The real-time PCR designed by Fukushima et al. (10) was carried out in a reaction volume of 40 l containing 10 l of the DNA extract, 20 l 2⫻ Universal Mastermix (Applied Biosystems), 0.25 M each primer, and 0.20 M probe. The reaction was performed on an ABI 7900 real-time thermocycler (Applied Biosystems). One negative control (negative plasma) and four positive controls (CMVAD169 quan-
661
TABLE 1. Timing of the amniocentesis and outcome of the babies in patients with false-negative results for CMV by PCR of AF Timing of amniocentesis Patient
Wk after maternal infection
Outcome for baby
Wk of pregnancy
1
16, 18
10, 12
2
21
10
3 4
21 23
7 6
MCDA twinsa; one baby was asymptomatic, and one baby had motor delay MCDA twinsa; both were asymptomatic Asymptomatic General developmental retardation
a
Both babies had a positive viral urinary culture during the first week of life.
titated viral DNA [Advanced Biotechnologies Inc., Columbia, MD] at 104, 103, 100, and 10 copies per reaction mixture) were included in the run. Nested PCR targeting the CMV glycoprotein B (gB) gene. Nested PCR was carried out with the primers and the cycling conditions described by Ziyaeyan et al. (37). The PCR master mixture for the first and second reactions contained 10⫻ PCR buffer II, 0.25 mM MgCl2, 0.2 mM each deoxynucleoside triphosphate, 25 pmol of either the outer or the inner primer, and 1.25 U AmpliTaq DNA polymerase (Applied Biosystems). In the first reaction, 10 l DNA extract was added to 40 l of the PCR master mixture; in the second reaction, 2 l of the outer PCR product was added to 48 l of the PCR master mixture. Both reactions were carried out on a GeneAmp 9700 thermocycler (Applied Biosystems). One negative control (negative plasma) and four positive controls (CMVAD169 quantitated viral DNA [Advanced Biotechnologies Inc.] at 104, 103, 100, and 10 copies per reaction mixture) were included in the run. Nine microliters of the inner PCR product (with 1 l loading buffer) was subjected to gel electrophoresis in a 2% agarose gel, and the gel was stained with ethidium bromide and photographed under UV light.
RESULTS AF samples negative for CMV. No CMV genome was detected in 241 AF samples. A urinary viral culture result was available for 38 children belonging to 36 mothers. Viral culture was positive for six children belonging to four mothers (Table 1). Both twins with positive urine cultures were monochorial diamniotic (MCDA). For patient 1, the mother of a male MCDA twin, two samples of AF were obtained before 21 weeks of pregnancy (16 and 18 weeks). The first AF sample was collected after laser therapy for early twin-twin transfusion syndrome, and the second was collected before intrauterine transfusion for anemia in the first recipient child. In both procedures, the amniotic sac of the recipient was sampled. CMV was detected in urine and blood samples from both children, but only the recipient child suffered from motor delay. For the other three women, prenatal diagnosis was performed at or after 21 weeks of gestation and at least 6 weeks after maternal infection had occurred. Upon retesting of the samples, the five AF samples belonging to four mothers remained negative. CMVAD169 quantitated viral DNA was included in the rerun of the test, and the test confirmed that the assay detects CMV at levels as low as 10 copies per reaction mixture (data not shown). Subsequently, an alternative realtime PCR targeting the gH gene was carried out, but it did not detect CMV nucleic acid in any of the selected AF samples (data not shown). Finally, a sensitive nested PCR targeting the gB gene also could not demonstrate the presence of the CMV genome in any of the five AF samples (data not shown).
662
GOEGEBUER ET AL.
J. CLIN. MICROBIOL.
TABLE 2. Preterm terminated pregnancies with positive results for CMV by PCR of AFa
Patient
Timing of amniocentesis (wk of pregnancy)
Wk of pregnancy after maternal infection
Viral load in AF (no. of copies/ml)
1
21.5
⬎12
1.29 ⫻ 107
35
2
21.4
⬎13
8.51 ⫻ 106
25.1
3
20.3
Unknown
3.21 ⫻ 105
Unknown
4
22
Unknown
6.76 ⫻ 107
23.3
5
21
⬎20
4.22 ⫻ 105
34.2
6
21.6
Unknown
5.22 ⫻ 105
22.5
7 8 9
14.6 21.3 21.0
Unknown ⫾11–13 ⫾16
2.80 ⫻ 103 8.60 ⫻ 105 4.80 ⫻ 106
Unknown 23.4 22.4
10
25.4
⫾8
1.72 ⫻ 107
27.3
11
15.5, 19.5
Unknown
Unknown
12 13
Unknown 22.4
Unknown ⫾12
4.7 ⫻ 104, 5.62 ⫻ 106 6.4 ⫻ 106 5.23 ⫻ 105
a b
Time of pregnancy termination (wk of pregnancy)
Unknown 24.5
Ultrasound result
Autopsy/pathology findings
Cerebral abnormalities, echogenic Focal encephalitis, splenomegaly; bowel (mild), OH CMV inclusions in kidneys, lungs, thyroid gland, and pancreas; chorioamnionitis and villitis Cerebral abnormalities, echogenic CMV inclusions in kidneys, liver, bowel, free intra-abdominal lungs, and pancreas; CMV and pericardial fluid, OH, placentitis hydrops placenta Abnormalities, not further CMV inclusions in CNS and in definedb other visceral organs,b villitis IUGR, echogenic bowel, free NA intra-abdominal and pericardial fluid Cerebral abnormalities, IUGR, CMV inclusions in kidneys, echogenic bowel (mild), OH, lungs, and CNS; villitis hydrops placenta Cerebral abnormalities, IUGR, HSM; CMV inclusions in cord decompensation, HSM, kidneys, liver, lungs, and OH, hydrops placenta (possibly) CNS; villitis No abnormalities Unknown No abnormalities CMV inclusions in lungs No abnormalities Encephalitis; myocarditis; CMV inclusions in kidneys, liver, lungs, pancreas, CNS, and thyroid gland; placentitis Cerebral abnormalities, IUGR Encephalitis; CMV inclusions in kidneys, liver, lungs, pancreas, and heart CNS calcifications and infection Unknown of the fetal heartb Unknown Unknown No abnormalities NA
HSM, hepatosplenomegaly; IUGR, intrauterine growth retardation; NA, no autopsy performed; OH, oligohydramnios. Information obtained from the mother only.
AF samples positive for CMV. The CMV genome was detected in the AF samples of 41 pregnant women. The CMV viral loads ranged from ⬍500 to 6.8 ⫻107 copies/ml. Clinical data were available for 35 of the 41 women with a positive result for CMV in AF (age range, 22 to 37 years; median age, 29 years). Twenty-two women went to term, while 13 pregnancies were terminated. (i) Terminated pregnancies. Amniocentesis was carried out at gestational ages ranging from 14.6 to 25.4 weeks (median, 21.4 weeks). The reasons for termination are shown in Table 2. Note that in four cases the decision to terminate the pregnancy was based only on the positive PCR result for CMV in AF, whereas the reason for termination was unclear in one more case. In one patient (patient 7), amniocentesis was performed before the 15th week of pregnancy. The CMV genome was detected in the AF, but the viral load was 2 log units lower than the mean viral load in the AF of the other women. Patient 11 underwent two amniocenteses with an interval of 4 weeks (at 15.5 and 19.5 weeks of gestation), and a marked increase in the viral load between the two samples was noticed. (ii) Nonterminated pregnancies. Twenty-two pregnant women with a positive result for CMV in AF continued their
pregnancies to term. Amniocenteses in this group were performed between 18.4 and 32.1 weeks of pregnancy (median, 22.6 weeks). The characteristics of the amniocentesis and the outcomes for the babies are summarized in Table 3. One sample of AF (patient 4) yielded a weak result (⬍500 copies/ml). For this patient, amniocentesis was carried out, as recommended, at 21 weeks of gestation and between 10 and 12 weeks after maternal infection. Despite the low viral load in the AF, the child was congenitally infected, as documented by the isolation of CMV from a urine sample collected from the newborn. At birth, the child was asymptomatic, and follow-up at 6 months of age revealed no problems. In Fig. 1, the CMV viral loads in AF samples are plotted against the presence or the absence of symptoms in the newborn. Data for terminated pregnancies in which no fetal damage upon ultrasound was detected and no result from histological tissue examination was available were excluded. A broad range of CMV viral loads was observed for both the symptomatic and the asymptomatic groups of newborns, and the viral loads in both groups overlapped to a great extent. It is clear that there was no significant difference between the mean and the median viral loads for the two groups. In Fig. 2, the CMV viral load is plotted against the gesta-
VOL. 47, 2009
QUANTIFICATION OF HUMAN CMV DNA IN AMNIOTIC FLUID
663
TABLE 3. Nonterminated pregnancies with positive results for CMV by PCR of AF Patient
Timing of amniocentesis (wk of pregnancy)
Wk of pregnancy after maternal infection
Viral load in AF (no. of copies/ml)
1 2 3 4 5 6 7 8 9 10 11 12 13
23 32.1 20.6 21.1 20.4 30.2 21 24.2 20 22 27.5 22.6 26
⫾8 ⫾8 ⬎11 ⫾10–12 ⫾11 ⫾10 ⫾11 ⫾8 ⫾8 Unknown ⫾8 ⫾12–18 ⫾7
6.30 ⫻ 107 1.03 ⫻ 106 1.98 ⫻ 105 ⬍500 2.28 ⫻ 105 1.12 ⫻ 107 8.37 ⫻ 105 2.09 ⫻ 107 7.10 ⫻ 104 1.05 ⫻ 107 2.18 ⫻ 105 2.20 ⫻ 106 1.19 ⫻ 102
14 15 16 17 18
20 22.1 25.3 18.4 19.4
⬎12 ⫾10–12 ⫾8 ⫾16 ⫾7
3.10 ⫻ 104 3.11 ⫻ 105 3.16 ⫻ 102 1.30 ⫻ 105 1.6 ⫻ 106
19 20 21 22
25.2 Unknown Unknown 24.6
Unknown Unknown Unknown Unknown
2.18 ⫻ 102 3.64 ⫻ 105 1.64 ⫻ 105 3.39 ⫻ 106
Outcome
Age at end of study
No problems No problems No problems No problems No problems No problems Petechiae, purpura, thrombopenia No problems Deaf on left side No problems No problems Cystic germinolysis of the brain Hypotrophic corpus callosum, enlarged ventricles No problems No problems No problems Bilateral deafness Cerebral cysts; moderately elevated liver enzyme levels, treated for 3 wk Possible hearing loss on right side No problems No problems Twins, no problems
2.5 mo 5 mo 3 mo 6 mo 6 mo 8.5 mo 6.5 mo 8 mo 8 mo 11 mo 1 yr 3 mo 1 yr 7 mo 1 yr 7 mo 1 1 1 1 1
yr yr yr yr yr
7 mo 8 mo 7 mo 8 mo 10 mo
2 2 2 3
yr 7 mo y, 5 mo yr 11 mo yr
tional age when amniocentesis was performed. In most cases (Fig. 2, circle 1), a correlation was found between the gestational age at the time of amniocentesis and the viral load in the AF (Spearman correlation, P ⫽ 0.0001). Two groups of outliers were observed (circle 2 and circle 3, Fig. 2). In the three women belonging to circle 2, amniocentesis was performed relatively late in gestation (at 27.5, 30.2, and 32.1 weeks of gestation; median, 30.2 weeks) because CMV infection occurred after 20 weeks of pregnancy. Finally, we found four cases with remarkably low CMV viral loads in AF, despite an appropriate timing of amniocentesis (Fig. 2, circle 3). Of the four infected children, two have remained asymptomatic, one displayed neural deformations at birth, and the fourth child is suspected of having hearing loss.
FIG. 1. CMV viral load in AF (log copies/ml) versus outcomes of the children (asymptomatic, symptomatic) from 31 pregnancies with congenital CMV infection.
FIG. 2. CMV viral load in AF (log copies/ml) versus gestational ages (weeks) of the children from 32 pregnancies with congenital CMV infection.
664
GOEGEBUER ET AL.
FIG. 3. Schematic overview of the study.
DISCUSSION The aim of this study was, first, to evaluate the usefulness of PCR for CMV in AF for the diagnosis of congenital CMV infection and, second, determine if quantification of the virus in AF could aid with determination of a prognosis for the infected fetus. Figure 3 provides a schematic overview of the study. In this small series of experiments, the quantitative PCR for CMV in AF had a high specificity, but negative results could not rule out congenital infection in the fetus, even when the hitherto accepted sampling conditions were met. Indeed, in four mothers we observed false-negative results by the realtime PCR for CMV in AF. Several authors have noted a suboptimal sensitivity of PCR-based methods for the detection of CMV in AF (3, 5, 14, 15, 21). The inadequate timing of amniocentesis, improper sample transport or processing, inhibition of the PCR by compounds in the AF, suboptimal primer or probe specificity, and an unsatisfactory intrinsic PCR sensitivity are mentioned as possible causes (2, 3, 21, 34). For one patient, two AF samples were obtained before the recommended 21 weeks of pregnancy, which may readily explain the negative PCR results. For the other three women, however, the amniocenteses for prenatal diagnosis were executed at the correct times. Since the three amniocentesis procedures for these three women were carried out at the University Hospitals Leuven, where the laboratory is located, long-distance transport was avoided, rendering improper sample handling less probable. Inhibition of the PCR by compounds in the AF was ruled out by the inclusion of an internal control for each sample. A revision of the target sequence of the in-house real-time PCR, the MCP gene, in BLAST (http://blast.ncbi.nlm .nih.gov) did not reveal any polymorphism at the primer or probe annealing sites. A real-time PCR targeting an alternative gene, the gH gene, did not detect CMV in any of the three AF samples. Therefore, the hypothesis that the specificity of the primer or the probe was suboptimal can also be abandoned. Finally, a sensitive nested PCR targeting the gB gene was performed, but it also could not demonstrate the presence of the CMV genome in any of the samples. In conclusion, through additional workup of the three negative AF samples, we ruled out the possibility that the false-negative results were due to a detection problem with the real-time PCR used. In the search for an explanation for the false-negative results for the detection of the CMV genome in AF, the hypothesis
J. CLIN. MICROBIOL.
that the fetus could be infected through the amniocentesis procedure itself has been brought up (14, 21, 32). The risk of CMV transmission during amniocentesis is considered minor; however, it is theoretically not impossible (21, 32). In a study by Liesnard et al., the detection by PCR of CMV DNA in maternal blood at the time of amniocentesis was infrequent. Moreover, the CMV transmission rates were not different between women with one amniocentesis and women with multiple amniocenteses and were comparable to the transmission rates determined from retrospective studies without antenatal intervention (21). However, since determination of the CMV viral load in blood was not carried out at the time of the prenatal diagnosis for any of the three mothers from our study, the possibility of iatrogenic infection cannot be excluded (21, 32). For future studies, it would be useful to include maternal viral load testing in the protocol to gain additional data for risk assessment and to be able to rule out iatrogenic transmission. In our opinion, however, the more probable explanation is the theory suggested by Revello and colleagues that the intrauterine transmission of CMV is characterized by an unpredictable delay, which prevents the possibility that a sensitivity of 100% for the detection of CMV in AF will ever be achieved (31, 32). In contrast to the findings of Lazzarotto et al. (19, 20), Gouarin et al. (13), and Guerra et al. (15), our study could not demonstrate a correlation between the CMV viral load in AF and the outcome for the fetus. Instead, in most of the nonterminated pregnancies, the CMV viral load in AF seemed to be related to the time during the pregnancy when the amniocentesis was performed. In addition, one patient underwent two consecutive amniocenteses within 4 weeks, and in this case, an increase in the viral load between the two tests was also seen. A similar correlation was observed by Gouarin et al. (13) and Picone et al. (29). An increase in the CMV viral load in AF during pregnancy could be explained by the accumulation of CMV in the AF, on the one hand, and an enhanced urinary flow of the fetus through pregnancy, on the other hand (30). Among the nonterminated pregnancies, we noticed two groups of outliers. Since CMV infection in the three women in circle 2 in Fig. 2 occurred after 20 weeks of pregnancy, the amniocentesis was performed at a later gestational age. We therefore hypothesize that fetuses infected later in pregnancy display better resistance to the virus, resulting in a lower CMV viral load in the AF. Four cases of relatively low CMV viral loads according to gestational age were observed (Fig. 2, circle 3). It is difficult to say whether in these cases CMV was cleared more effectively by the mother and the fetus or whether the women were in an early transmission state. Among our study population we noted a large variation in the time between maternal infection and prenatal diagnosis. In the majority of cases, the maternal infection was asymptomatic. The time of maternal infection was therefore deduced from serology results and must be considered an educated guess. In addition, the results of the baseline serology before pregnancy were not available for most of the patients, so it was difficult to distinguish a primary CMV infection from a reactivation or a reinfection. Even though the use of a minimum gestational age of 21 weeks for the timing of amniocentesis should be respected in order to obtain reliable results, a large variation in the time of amniocentesis was observed among our study patients, with the times ranging from 14.6 to 32.1 weeks
VOL. 47, 2009
QUANTIFICATION OF HUMAN CMV DNA IN AMNIOTIC FLUID
of gestation. Finally, as mentioned above, follow-up of the children was done by use of a questionnaire and not by use of a standardized neurological examination or an objective developmental scale. Also, since sensorineural hearing loss is progressive until 6 years of age, long-term follow-up is indicated to identify all symptomatic children. In conclusion, PCR for the detection of CMV in AF had a high specificity for the detection of congenital infection in the fetus. However, even when optimal sampling conditions are met, a negative result for CMV in AF cannot rule out the possibility of intrauterine infection. No correlation between the CMV viral load in AF and fetal outcome could be demonstrated in our study. Quantification of the CMV viral load in AF should therefore not be considered a reliable prognostic marker of the severity of fetal disease. Instead, a relation between the viral load and the time in pregnancy when the amniocentesis was performed was observed. Thus, whether a congenitally infected child develops symptoms or not is most probably correlated to factors other than the CMV viral load in AF (such as the CMV genotype or genetic diversity), but further investigation is warranted. REFERENCES 1. American College of Obstetricians and Gynecologists. 2000. Perinatal viral and parasitic infections. ACOG practice bulletin no. 20. American College of Obstetricians and Gynecologists, Washington, DC. http://www.acog.org. 2. Avidor, B., G. Efrat, M. Weinberg, Z. Kra-oz, J. Satinger, S. Mitrani-Rosembaum, Y. Yaron, L. Shulman, M. Tepperberg-Oikawa, D. Wolf, S. A. Berger, S. Lipitz, E. Mendelson, and M. Giladi. 2004. Insight into the intrinsic sensitivity of the PCR assay used to detect CMV infection in amniotic fluid specimens. J. Clin. Virol. 29:260–270. 3. Azam, A. Z., Y. Vial, C. L. Fawler, J. Zufferey, and P. Hohlfeld. 2001. Prenatal diagnosis of congenital cytomegalovirus infection. Obstet. Gynecol. 97:443–448. 4. Beuselinck, K., M. Van Ranst, and J. Van Eldere. 2005. Automated extraction of viral-pathogen RNA and DNA for high-throughput quantitative real-time PCR. J. Clin. Microbiol. 43:5541–5546. 5. Bodeus, M., C. Hubinont, P. Bernard, A. Bouckaert, K. Thomas, and P. Goubeau. 1999. Prenatal diagnosis of human cytomegalovirus by culture and polymerase chain reaction: 98 pregnancies leading to congenital infection. Prenat. Diagn. 19:314–317. 6. Collinet, P., D. Subtil, V. Houfflun-Debarge, N. Kacet, A. Dewilde, and F. Puech. 2004. Routine CMV screening during pregnancy. Eur. J. Obstet. Gynecol. Reprod. Biol. 114:3–11. 7. Devos, T., B. Spitz, and W. Peetermans. 2001. Het risico op congenitale cytomegalovirusinfectie bij zwangere gezondheidswerkers. Tijdschr. Geneeskunde 57:936–942. 8. Donner, C., C. Liesnard, F. Brancart, and F. Rodesch. 1994. Accuracy of amniotic fluid testing before 21 weeks’ gestation in prenatal diagnosis of congenital cytomegalovirus infection. Prenat. Diagn. 14:1055–1059. 9. Enders, G., U. Bader, L. Lindemann, G. Schalasta, and A. Daiminger. 2001. Prenatal diagnosis of congenital cytomegalovirus infection in 189 pregnancies with known outcome. Prenat. Diagn. 21:362–377. 10. Fukushima, E., K. Ishibashi, H. Kaneko, H. Nishimura, N. Inoue, T. Tokumoto, K. Tanabe, K. Ishioka, H. Ogawa, and T. Suzutani. 2008. Identification of a highly conserved region in the human cytomegalovirus glycoprotein H gene and design of molecular diagnostic methods targeting the region. J. Virol. Methods 151:55–60. 11. Gaytant, M. A., E. A. Steegers, B. A. Semmekrot, H. M. Merkus, and J. M. Galama. 2002. Congenital cytomegalovirus infection: review of the epidemiology and outcome. Obstet. Gynecol. Surv. 57:245–256. 12. Gaytant, M. A., G. I. Rours, E. A. Steegers, J. M. Galama, and B. A. Semmekrot. 2003. Congenital cytomegalovirus infection after recurrent infection: case reports and review of the literature. Eur. J. Pediatr. 162:248– 253. 13. Gouarin, S., E. Gault, A. Vabret, D. Cointe, F. Rozenberg, L. GrangeotKeros, P. Bariot, A. Garbarg-Chenon, P. Lebon, and F. Freymuth. 2002. Real-time PCR quantification of human cytomegalovirus DNA in amniotic fluid samples from mothers with primary infection. J. Clin. Microbiol. 40: 1767–1772. 14. Gouarin, S., P. Palmer, D. Cointe, S. Rogez, A. Vabret, F. Rozenberg, F. Denis, F. Freymuth, P. Lebon, and L. Gangreot-Keros. 2001. Congenital
15.
16.
17.
18.
19.
20.
21.
22.
23.
24. 25.
26.
27.
28.
29.
30.
31.
32.
33. 34.
35.
36.
37.
665
HCMV infection: a collaborative and comparative study of virus detection in amniotic fluid by culture and by PCR. J. Clin. Virol. 21:47–55. Guerra, B., T. Lazzarotto, S. Quarta, M. Lanari, L. Bovicelli, A. Nicolosi, and M. P. Landini. 2000. Prenatal diagnosis of symptomatic congenital cytomegalovirus infection. Am. J. Obstet. Gynecol. 183:476–482. Institute for Clinical Systems Improvement. 2007. Health care guideline: routine prenatal care, 11th ed. Institute for Clinical Systems Improvement, Bloomington, IN. http://www.icsi.org. Lazzarotto, T., B. Guerra, P. Spezzacatena, S. Varani, L. Gabrielli, P. Pradelli, F. Rumpianesi, C. Banzi, L. Bovicelli, and M. P. Landini. 1998. Prenatal diagnosis of congenital cytomegalovirus infection. J. Clin. Microbiol. 36:3540–3544. Lazzarotto, T., L. Gabrielli, M. Lanari, B. Guerra, T. Bellucci, M. Sassi, and M. P. Landini. 2004. Congenital cytomegalovirus infection: recent advances in the diagnosis of maternal infection. Hum. Immunol. 65:410–415. Lazzarotto, T., L. Gabrielli, M. P. Foschini, M. Lanari, B. Guerra, V. Eusebi, and M. P. Landini. 2003. Congenital cytomegalovirus infection in twin pregnancies: viral load in the amniotic fluid and pregnancy outcome. Pediatrics 112:e153–e157. Lazzarotto, T., S. Varani, B. Guerra, A. Nicolosi, M. Lanari, and M. P. Landini. 2000. Prenatal indicators of congenital cytomegalovirus infection. J. Pediatr. 137:90–95. Liesnard, C., C. Donner, F. Brancart, F. Gosselin, M. L. Delforge, and F. Rodesch. 2000. Prenatal diagnosis of congenital cytomegalovirus infection: prospective study of 237 pregnancies at risk. Obstet. Gynecol. 95(6 Pt 1): 881–888. Lipitz, S., S. Yagel, E. Shalev, R. Achiron, S. Mashiach, and E. Schiff. 1997. Prenatal diagnosis of fetal primary cytomegalovirus infection. Obstet. Gynecol. 89(5 Pt 1):763–767. Lodewyckx, K., G. Peeters, B. Spitz, S. Blot, M. Temmerman, W. Zhang, S. Alexander, F. Mambourg, and D. Ramaekers. 2004. Nationale richtlijn prenatale zorg: een basis voor een klinisch pad voor de opvolging van zwangerschappen. KCE reports 6A. Federaal Kenniscentrum voor de Gezondheidszorg, Brussels , Belgium. http://kce.fgov.be. Mosca, F., and L. Pugni. 2007. Cytomegalovirus infection: the state of the art. J. Chemother. 19(Suppl. 2):46–48. Mulongo, K. N., M. E. Lamy, and M. Van Lierde. 1995. Requirements for diagnosis of prenatal cytomegalovirus infection by amniotic fluid culture. Clin. Diagn. Virol. 4:231–238. National Collaborating Centre for Women’s and Children’s Health. 2003. Antenatal care: routine care for the healthy pregnant woman. Clinical guideline 6. National Institute for Health and Clinical Excellence, London, United Kingdom. http://www.nice.org.uk. Nedelec, O., N. Bellagra, L. Devisme, D. Hober, P. Wattre, and A. Dewilde. 2002. Congenital human cytomegalovirus infection: value of human cytomegalovirus DNA quantification in amniotic fluid. Ann. Biol. Clin. (Paris) 60:201–207. Nigro, G., S. P. Adler, R. La Torre, and A. M. Best. 2005. Passive immunization during pregnancy for congenital cytomegalovirus infection. N. Engl. J. Med. 353:1350–1362. Picone, O., J. M. Costa, M. Leruez-Ville, P. Ernault, M. Olivi, and Y. Ville. 2004. Cytomegalovirus (CMV) glycoprotein B genotype and CMV DNA load in the amniotic fluid of infected fetuses. Prenat. Diagn. 24:1001–1006. Rabinowitz, R., M. T. Peters, S. Vyas, S. Campbell, and K. H. Nicolaides. 1989. Measurement of fetal urine production in normal pregnancy by realtime ultrasonography. Am. J. Obstet. Gynecol. 161:1264–1266. Revello, M. G., A. Sarasini, M. Zavattoni, F. Baldanti, and G. Gerna. 1998. Improved prenatal diagnosis of congenital human cytomegalovirus infection by a modified nested polymerase chain reaction. J. Med. Virol. 56:99–103. Revello, M. G., and G. Gerna. 2002. Diagnosis and management of human cytomegalovirus infection in the mother, fetus, and newborn infant. Clin. Microbiol. Rev. 15:680–715. Revello, M. G., and G. Gerna. 2004. Pathogenesis and prenatal diagnosis of human cytomegalovirus infection. J. Clin. Virol. 29:71–83. Revello, M. G., D. Lilleri, M. Zavattoni, M. Furione, J. Middeldorp, and G. Gerna. 2003. Prenatal diagnosis of congenital human cytomegalovirus infection in amniotic fluid by nucleic acid sequence-based amplification assay. J. Clin. Microbiol. 41:1772–1774. Revello, M. G., M. Zavattoni, A. Sarasini, F. Baldanti, C. De Julio, L. De-Giuli, U. Nicolini, and G. Gerna. 1999. Prenatal diagnostic and prognostic value of human cytomegalovirus load and IgM antibody response in blood of congenitally infected fetuses. J. Infect. Dis. 180:1320–1323. Revello, M. G., M. Zavattoni, M. Furione, F. Baldanti, and G. Gerna. 1999. Quantification of human cytomegalovirus DNA in amniotic fluid of mothers of congenitally infected fetuses. J. Clin. Microbiol. 37:3350–3352. Ziyaeyan, M., A. Alborzi, A. Abbasian, M. Kalani, A. Moravej, J. Nasiri, A. Amiri, N. Hashemi, and F. Sefiddashti. 2007. Detection of HCMV DNA in placenta, amniotic fluid and fetuses of seropositive women by nested PCR. Eur. J. Pediatr. 166:723–726.
