Outcome of pregnancies with recent primary cytomegalovirus infection in first trimester treated with hyperimmunoglobulin: observational study
ABSTRACT
Objective
To examine the efficacy of hyperimmunoglobulin (HIG) treatment in women with a recent primary cytomegalovirus (CMV) infection up to 14 weeks' gestation.
Methods
This is an ongoing observational study conducted at the prenatal medicine departments of the University Hospitals of Tübingen, Bonn, Cologne and Erlangen, Germany, as well as at the Laboratory Prof. Gisela Enders and Colleagues in Stuttgart, Germany and the Institute for Medical Virology at the University of Tübingen, Tübingen, Germany. Enrolment criteria were the presence of confirmed recent primary CMV infection in the first trimester and a gestational age at first HIG administration of ≤ 14 weeks. The following inclusion criteria indicated a recent primary infection: low anti-immunoglobulin (Ig)-G levels, low anti-CMV-IgG avidity in the presence of a positive CMV-IgM test and no positive reactivity or just seroconversion anti-gB2-IgG-reactivity. HIG administration was started as soon as possible within a few days after the first visit. HIG was administered intravenously at a dose of 200 IU/kg maternal body weight and repeated every 2 weeks until about 18 weeks' gestation. The primary outcome was maternal–fetal transmission at the time of amniocentesis. Multivariate logistic regression analysis was used to determine significant covariates that could predict maternal–fetal transmission.
Results
We included 149 pregnancies (153 fetuses) that completed the treatment. Median maternal age and weight were 32.0 years and 65.0 kg, respectively. Median gestational age at the time of first referral to one of the four centers was 9.4 weeks. Median anti-CMV-IgG level, anti-CMV-IgM index and CMV-IgG avidity were 5.7 U/mL, 2.5 and 22.3%, respectively. HIG treatment was started at a median gestational age of 10.6 weeks and ended at a median of 17.9 weeks. Within this time frame, HIG was administered on average four times in each patient. Amniocentesis was carried out at a median gestational age of 20.4 weeks. In 143 (93.5%) of the 153 cases, the fetus was not infected. Maternal–fetal transmission occurred in 10 cases (6.5% (95% CI, 3.2–11.7%)). On uni- and multivariate logistic regression analysis, the level of anti-IgM index was the only factor associated significantly with maternal–fetal transmission at amniocentesis. However, only four (40.0%) of the 10 cases with maternal–fetal transmission had an anti-IgM index above 11.4, which corresponds to the 95th centile of pregnancies without transmission.
Conclusions
HIG is a treatment option to prevent maternal–fetal transmission in pregnancy with a primary CMV infection. However, HIG treatment seems to be beneficial primarily in women with a recent primary infection in the first trimester or during the periconceptional period, and when it is administered at a biweekly dose of 200 IU/kg. © 2021 The Authors. Ultrasound in Obstetrics & Gynecology published by John Wiley & Sons Ltd on behalf of International Society of Ultrasound in Obstetrics and Gynecology.
CONTRIBUTION
What are the novel findings of this work?
Primary cytomegalovirus (CMV) infection in the first trimester carries a substantial risk of developmental disorder after birth. We found that hyperimmunoglobulin (HIG) treatment can prevent maternal–fetal CMV transmission if patients are well selected.
What are the clinical implications of this work?
Administration of HIG should be discussed in cases of a primary CMV infection in the first trimester. HIG treatment is mostly successful in women with a recent primary infection in the first trimester or during the periconceptional period, and when HIG is administered at a biweekly dose of 200 IU/kg.
INTRODUCTION
Fetal cytomegalovirus (CMV) infection is the most frequent and relevant viral infection during pregnancy, affecting about 0.2–2.2% of all live births. It is the leading cause of non-genetic hearing loss and belongs to the most relevant reasons of neurological disability1-6. However, the risk of sequelae depends highly on the gestational age at the time of maternal and fetal infection7, 8. A recent meta-analysis by Chatzakis et al.8, evaluating the risk of maternal–fetal transmission after primary CMV infection, included 10 studies with about 2900 fetuses and concluded that during the pre- and periconception period and the first, second and third trimesters of pregnancy, the risk of transmission was 5.5%, 21.0%, 36.8%, 40.3% and 66.2%, respectively. The authors also reviewed the risk of fetal insult in case of transmission, defined as prenatal central nervous system findings leading to termination of pregnancy or presence of neurological symptoms at birth, and found a risk of 19.3% for a primary infection in the first trimester which decreased to 0.9% and 0.4% for an infection in the second and third trimester, respectively8.
It is an ongoing discussion whether there is an effective treatment option after a primary maternal CMV infection9, either to help prevent maternal–fetal transmission or, in the case of fetal infection, to avoid developmental impairment. Several studies have focused on the efficacy of hyperimmunoglobulin (HIG) to prevent maternal–fetal transmission9, 10. HIG treatment failed to reduce the maternal–fetal transmission rate in two randomized controlled trials (RCTs)11, 12. Revello et al.11 treated 123 women with either HIG or placebo and did not find a significant difference in the congenital infection rate between the two groups (30% vs 44%). Another large RCT by Hughes et al.12 (NCT01376778) enrolled 399 women with a primary CMV infection, however, the study was stopped because the transmission rate was similar between the HIG and placebo groups (22.7% and 19.4%, respectively)13 and the complete details and findings of the study have not been published yet. Both of these RCTs included pregnant women up to 24–26 weeks' gestation and used an HIG dose of 100 IU/kg maternal body weight administered on a monthly basis. In contrast, our group observed a maternal–fetal transmission rate of 2.5% at the time of amniocentesis and a congenital infection rate of 7.5% at birth in a series of 40 women with a primary CMV infection who were treated with HIG14. In that study, we included only women with a recent primary infection who had the first HIG administration at ≤ 14 weeks' gestation and the treatment dose of HIG was 200 IU/kg body weight administered every 2 weeks up to 20 weeks' gestation.
In this extension of our previous study, we set out to examine the efficacy of the HIG treatment in about 150 women with a recent primary CMV infection up to 14 weeks' gestation.
METHODS
This is an ongoing observational study conducted at the prenatal medicine departments of the University Hospitals of Tübingen, Bonn, Cologne and Erlangen, as well as the Laboratory Prof. Gisela Enders and Colleagues in Stuttgart and the Institute for Medical Virology and Epidemiology of Viral Diseases at the University Hospital of Tübingen, Tübingen, Germany. The study started in 2013 at the University Hospital of Tübingen, and the other centers joined in 2017. The study presented here is an extension of the one published in 2019 and involves about four times the initial number of patients including the 40 cases reported previously14. The last woman included in the study started HIG treatment in June 2020. Ethical approval was obtained from the ethics committees of the University Hospitals of Tübingen (749/2020BO2), Cologne (20-1525), Bonn (285/17) and Erlangen (85_19 Bc). All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
In Germany, there is no antenatal screening program for CMV and, therefore, CMV testing in the first trimester is offered upon patient request or after recommendation as part of an individualized healthcare service. Pregnant women with a primary maternal CMV infection were referred to one of the four prenatal medicine departments involved in this study.
The following commercial tests were used for the diagnosis of CMV infection: Elecsys® (ECLIA) anti-immunoglobulin (Ig)-G, anti-IgM and IgG avidity (Roche Diagnostics, Rotkreutz, Switzerland) using fully automated Cobas 6000/Cobas e601 analyzer (Roche Diagnostics) and a recomLine CMV IgG, IgM and IgG avidity assay (Mikrogen, Neuried, Germany). The assays were used as described in previous virological monitoring studies14, 15. For patients managed at the prenatal medicine departments of the University Hospitals of Tübingen, Bonn and Cologne, the virological studies were carried out at the Institute of Medical Virology, University Hospital Tübingen. For women who were treated at the University Hospital of Erlangen, the blood samples were analyzed at the Laboratory Prof. Gisela Enders and Colleagues. At this center, the anti-IgG ECLIA system, the anti-IgM ELA Test PKS (Medac GmbH, Wedel, Germany), the VIDAS® (ELFA; bioMérieux, Marcy l'Etoile, France) CMV IgG Avidity I/II (bioMérieux) and a recomLine CMV IgG, IgM and IgG avidity assay (Mikrogen) were used16.
Enrolment criteria were the presence of confirmed, recent primary CMV infection in the first trimester and a gestational age at first HIG administration of ≤ 14 weeks. The following inclusion criteria indicated a recent primary infection: anti-CMV-IgG level (ECLIA) < 60 U/mL, no positive reactivity or just seroconversion reactivity against anti-rec-gB2-IgG and a borderline or reactive anti-IgM index. In the few cases in which the IgM index was < 1.0 or the IgG level was 60–100 U/mL, an early CMV primary infection was still assumed because there was anti-p150-IgM reactivity in the immunoblot assay and low IgG avidity. The CMV-IgG avidity level was evaluated using both the ECLIA or ELFA system and a recomLine immunoblot assay1. The following recombinant antigens were included to categorize CMV-IgG avidity by immunoblot analysis: IE1, CM2, p150 and gB2. Cases were included if the CMV-IgG avidity level was low in both assays or if one test indicated an intermediate avidity and the other low avidity4, 10.
All cases with a recent primary CMV infection were classified according to the potential timing of the maternal infection. This decision had to be made based on the one blood sample that was obtained up to 14 weeks' gestation. Therefore, we assumed that, if the blood sample was taken prior to 8 weeks' gestation, the onset of maternal infection was during the periconceptional period. In women for whom the blood sample was taken between 8 and 14 weeks, the decision was based on the IgG avidity. If CMV-IgG avidity was ≤ 20% or not measurable due to low anti-IgG levels, the onset of infection was assumed to be in the first trimester, whereas, if CMV-IgG avidity was > 20%, the onset of the infection was assumed to be in the periconceptional period.
Before HIG (Cytotect CP®, Biotest AG, Dreieich, Germany) treatment was started, each woman received detailed information about the off-label use of HIG and the potential side effects. The HIG administration was started as soon as possible within a few days after the first visit in one of the four units involved in the study and after the healthcare insurance providers agreed to cover the medical expenses. HIG was administered intravenously at a dose of 200 IU/kg of maternal body weight and the treatment was repeated every 2 weeks until about 18 weeks' gestation.
The primary outcome measure was maternal–fetal transmission at the time of amniocentesis.
In all cases that are included in this study, amniocentesis was carried out at least 6 weeks after first presentation, generally at about 20 weeks' gestation. In the virological department of the University Hospital of Tübingen, the amniotic fluid was tested by two polymerase chain reactions (PCRs; nested PCR (nPCR) and quantitative real-time PCR), short-term microculture and long-term viral culture until generation of a cytopathic effect, ranging from 18 h to 5 days. Short-term 18-h fibroblast microculture, followed by CMV-IE1 immunoperoxidase staining and long-term (10 days) virus isolation from amniotic fluid, was performed by a virus-concentration step using high-speed centrifugation of 50 000 g for 1 h at 4°C prior to virus inoculation17. DNA was purified by spin columns using QIAmp DNA Blood Mini Kit (Qiagen, Düsseldorf, Germany) and used for qualitative nPCR of the IE1Ex4-target region. The limit of detection for nPCR was 200 copies/mL. Quantitative real-time PCR from plasma, serum or ethylenediaminetetraacetic acid (EDTA) whole blood as well as from amniotic fluid was performed using CMV-R-GENE® real-time PCR kits (Argene®; bioMérieux) with a limit of detection of 600 copies of CMV DNA/mL using target gene CMV UL83. At the Laboratory Prof. Gisela Enders and Colleagues, a RealStar® CMV PCR Kit 1.0 (altona Diagnostics GmbH, Hamburg, Germany) was used. These samples were tested additionally by rapid cell culture using embryonic lung fibroblasts and CMV-IE/EA immunoperoxidase staining.
If no maternal–fetal transmission occurred, no further treatment was carried out. In case of maternal–fetal transmission, further treatment with valacyclovir 8 g/day was offered18. In these cases, the medication was given until delivery. Furthermore, ultrasound monitoring was scheduled on a biweekly basis and a magnetic resonance imaging scan was offered at about 30 weeks' gestation. Termination of pregnancy was discussed, but none of the women chose this option. Furthermore, none of the women opted for discontinuation of the treatment.
Within the first few days after delivery, the umbilical cord, urine and/or saliva were tested for CMV-DNA and for viral isolation. In case of a positive virus detection, viral load in EDTA whole blood and urine or saliva was performed. Newborns with congenital CMV infection were included in a neurodevelopmental follow-up program as part of the local clinical service. This protocol includes valganciclovir for symptomatically congenital CMV-infected newborns19.
Information about the fetal development, transmission status at the time of amniocentesis and outcome data after birth were recorded in the perinatal databases (Viewpoint, Solingen, Germany). From 2017 onwards, we also recorded all pregnancy complications during and after HIG treatment and during the subsequent course of pregnancy, as well as the head circumference of the newborns.
Statistical analysis
Pregnancy and treatment characteristics were compared according to transmission status at the time of amniocentesis. Data are presented as median (25th to 75th interquartile range (IQR)) or percentage as appropriate. Comparisons between the two groups were performed using Student's t-test or Mann–Whitney U-test followed by Kolmogorov–Smirnov test for normal distribution. For proportions, 95% CI were calculated using the method of Clopper and Pearson.
Uni- and multivariate logistic regression analysis was carried out to identify significant covariates for the prediction of maternal–fetal transmission. Numerical data were included as continuous variables and dichotomous data as categorical variables. Birth-weight and head circumference centiles were computed according to the reference ranges of Voigt et al.20. The significance level was set at 0.05.
RESULTS
Study population
Since 2013, HIG treatment has been initiated in 165 women. Of these, 15 women were excluded due to missed miscarriage (n = 5) or a chromosomal defect detected during the first-trimester risk assessment (n = 2) or because they declined amniocentesis (n = 8). In one further case, the therapy was stopped after the second HIG administration due to an allergic reaction directly after the treatment; this case was also excluded from analysis. Thus, 149 pregnant women completed the treatment, including four with a twin pregnancy, resulting in a study group of 153 fetuses (Figure 1).
In 52 women (53 fetuses), CMV infection occurred in the first trimester and, in 97 women (100 fetuses), during the periconceptional period. Median maternal age was 32.0 years and median maternal weight was 65.0 kg. One-hundred and twenty-five (83.9%) women had at least one previous child and in 110 (73.8%) cases, the youngest child in the family was 3 years of age or younger. Median gestational age at the time of first referral to one of the four centers was 9.4 weeks. Median anti-IgG level, anti-IgM index and CMV-IgG avidity were 5.7 U/mL, 2.5, and 22.3%, respectively. The complete dataset is given in Table S1.
HIG treatment
Treatment with HIG was started at a median gestational age of 10.6 weeks and stopped at a median of 17.9 weeks. Within this time frame, HIG was administered on average four times in each patient. Amniocentesis was carried out at a median gestational age of 20.4 weeks.
In 143 (93.5%) of the 153 cases, the fetus was not infected. Thus, maternal–fetal transmission occurred in 10 cases (6.5% (95% CI, 3.2–11.7%)) (Figure 1). In a subgroup analysis, we examined the maternal–fetal transmission rate according to the time of maternal infection. There were five (5.0% (95% CI, 1.6–11.3%)) infected fetuses in the group with periconceptional CMV infection and five (9.4% (95% CI, 3.1–20.6%)) fetal infections in the group with first-trimester infection.
The pregnancy and virological study characteristics of the cases with and without transmission are shown in Table 1. In the group with maternal–fetal transmission, compared to the group without, the anti-IgM index was higher, the anti-IgG levels were lower and there was a higher proportion of cases with a positive PCR result at the time of first presentation. There were no further significant differences regarding the pregnancy and treatment characteristics between the two groups.
| Parameter | No maternal–fetal transmission (n = 139) | Maternal–fetal transmission (n = 10) | P |
|---|---|---|---|
| Maternal age (years) | 32.0 (29.6–34.8) | 34.7 (30.3–38.7) | 0.974 |
| Maternal weight (kg) | 65.0 (58.0–72.0) | 66.4 (52.0–75.5) | 0.635 |
| GA at presentation (weeks) | 9.6 (8.1–11.4) | 8.8 (8.0–10.7) | 0.365 |
| Had child ≤ 3 years of age at presentation | 103 (74.1) | 7 (70.0) | 0.890 |
| GA at first HIG administration (weeks) | 10.6 (9.1–12.3) | 9.2 (8.3–11.1) | 0.123 |
| GA at last HIG administration (weeks) | 17.9 (16.7–18.7) | 18.0 (16.9–19.1) | 0.465 |
| GA at amniocentesis (weeks) | 20.4 (20.1–20.9) | 20.2 (20.1–20.3) | 0.591 |
| Positive CMV-DNA PCR at first presentation* | 27 (19.4) | 5 (50.0) | 0.019 |
| Anti-CMV-IgG level (ECLIA) (U/mL) | 5.9 (2.4–16.5) | 2.7 (1.3–4.9) | < 0.001 |
| Anti-IgM index (ECLIA) | 2.3 (1.3–4.3) | 6.4 (4.4–20.1) | 0.002† |
| CMV-IgG avidity (ECLIA) (%) | 22.8 (9.9–32.0) | 15.1 (4.8–23.1) | 0.537 |
- Data are given as median (interquartile range) or n (%).
- * Maternal EDTA blood or serum.
- † Mann–Whitney U-test, otherwise Student's t-test or chi-square test. ECLIA, commercial test for CMV antibody status; GA, gestational age; IgG, immunoglobulin-G; IgM, immunoglobulin-M; PCR, polymerase chain reaction.
On uni- and multivariate logistic regression analysis, only the level of the anti-IgM index was associated significantly with maternal–fetal transmission at the time of amniocentesis (Table 2). However, only four (40.0%) of 10 cases with maternal–fetal transmission had an anti-IgM level above 11.4, which corresponds to the 95th centile of pregnancies without transmission.
| Covariate | Univariate analysis | Multivariate analysis | ||
|---|---|---|---|---|
| OR (95% CI) | P | OR (95% CI) | P | |
| Maternal age (in years) | 1.000 (0.973–1.027) | 0.974 | — | — |
| Maternal weight (in kg) | 1.018 (0.971–1.068) | 0.453 | — | — |
| GA at presentation (in weeks) | 0.868 (0.640–1.178) | 0.364 | — | — |
| Had child ≤ 3 years of age at presentation | 0.906 (0.223–3.678) | 0.890 | — | — |
| GA at first HIG administration (in weeks) | 0.776 (0.560–1.1077) | 0.129 | — | — |
| GA at last HIG administration (in weeks) | 1.189 (0.750–1.886) | 0.462 | — | — |
| GA at amniocentesis (in weeks) | 0.821 (0.404–1.670) | 0.587 | — | — |
| Positive CMV-DNA PCR at first presentation | 4.296 (1.161–15.898) | 0.029 | 2.988 (0.735–12.157) | 0.126 |
| Anti-CMV-IgG level (in U/mL) | 0.836 (0.687–1.018) | 0.075 | — | — |
| Anti-IgM index | 1.203 (1.089–1.330) | < 0.001 | 1.193 (1.076–1.323) | 0.001 |
| CMV-IgG avidity (in %) | 0.985 (0.938–1.034) | 0.535 | — | — |
- GA, gestational age; HIG, hyperimmunoglobulin; IgG, immunoglobulin-G; IgM, immunoglobulin-M; OR, odds ratio; PCR, polymerase chain reaction.
Side effects during or shortly after HIG administration were examined in a subgroup of 133 women. Of these, 109 (82.0% (95% CI, 74.4–88.1%)) were asymptomatic and the other 24 women noticed flu-like symptoms, tiredness, headache or a rash, predominantly after the first administration. The symptoms disappeared in all patients the following day.
Outcome of pregnancies with maternal–fetal transmission
The median viral load in the amniotic fluid of the 10 cases with maternal–fetal transmission was 1 245 000 (IQR, 49 000–8 230 000) copies/mL. CMV short-term culture indicated a median of 5390 (IQR, 1550–16 000) infected fibroblast nuclei per mL. The details of these 10 cases are listed in Table 3. One fetus died at 21.1 weeks' gestation. In the remaining nine cases, median gestational age at delivery was 37.6 (IQR, 36.7–38.9) weeks and median birth weight was 3040 (IQR, 2760–3100) g, corresponding to the 37th birth-weight centile (Figure 2). Four newborns were asymptomatic, three had some degree of hearing loss, one had elevated liver enzymes and two had heterotopic brain lesions, one with and the other without cystic brain lesions. The median head circumference was 34.0 (IQR, 33.0–34.0) cm (Figure 3).
| Case | Virological data at presentation: anti-IgM index, anti-IgG level (U/mL), IgG avidity (%), PCR status | Clinical data: MA (years), child ≤ 3 years old, weight (kg) | HIG administration: GA at first HIG (weeks), GA at last HIG (weeks), GA at amnio (weeks) | Examination of AF: viral load (copies/mL), short-term microculture (infected fibroblasts/mL) | Delivery data: GA (weeks), BW (g)/BW centile, HC (cm)/HC centile | Current status: age of child (years)*, outcome |
|---|---|---|---|---|---|---|
| 1 | 5.3, 3.2, 9.4, neg | 34.8, yes, 72.0 | 8.3, 16.4, 20.3 | 3 740 000, 5390 | 37.6, 2620/11, 33/22 | 2.8, asymptomatic |
| 2 | 11.6, 3.8, 51.1, pos | 38.8, no, 48.0 | 8.4, 18.4, 20.3 | 875 000, 1550 | 38.3, 3100/37, 35/62 | 1.5, asymptomatic |
| 3 | 7.4, 1.3, 23.1, pos | 42.2, yes, 98.0 | 8.9, 16.9, 20.1 | 82 000 000, 4400 | 36,7, 2760/46, 33/37 | 2.2, cystic brain lesions, heterotopia and hearing loss |
| 4 | 20.1, 8.0, 3.8, neg | 34.7, yes, 50.0 | 9.6, 17.6, 20.1 | 1 420 000, 8000 | 35.6, 2400/49, 33/54 | 2.4, hearing loss† |
| 5 | 4.4, 4.9, 40.3, neg | 23.8, yes, 52.0 | 10.1, 18.1, 20.3 | 27 300, 26 | 21.1, 311/1, ND | Intrauterine death |
| 6 | 20.6, 6.3, 4.8, neg | 30.3, yes, 60.7 | 11.1, 19.1, 20.1 | 9810, 8 | 38.9, 3040/38, 34/33 | 1.0, increased ALT/AST† |
| 7 | 0.6, 1.2, NE, pos | 38.7, yes, 59.1 | 11.9, 17.9, 20.9 | 8 230 000, 40 000 | 37.3, 3100/50, 34/46 | 0.7, asymptomatic |
| 8 | 21.0, 1.2, 22.9, pos | 28.5, yes, 102.0 | 6.6, 16.7, 20.7 | 1 070 000, 16 000 | 40.0, 3170/15, 34/18 | 0.5, hearing loss† |
| 9 | 4.2, 2.2, 14.2, pos | 30.5, no, 75.5 | 13.1, 19.1, 20.1 | 31 700 000, 16 000 | 36.4, 2780/44, 32/16 | 0.2, heterotopia† |
| 10 | 5.0, 1.4, 16.0, neg | 35.0, no, 74.0 | 8.0, 20.0, 20.0 | 49 000, ND | 40.6, 3230/28, 34/18 | 1.2, asymptomatic† |
- * At last examination.
- † Treatment with valganciclovir after delivery. AF, amniotic fluid; ALT, alanine transaminase; amnio, amniocentesis; AST, aspartate transaminase; BW, birth weight; GA, gestational age; HC, head circumference; HIG, hyperimmunoglobulin; IgG, immunoglobulin-G; IgM, immunoglobulin-M; MA, maternal age; ND, not done; NE, not estimable due to too low IgG; neg, negative; PCR, polymerase chain reaction; pos, positive.
) and without (
) transmission of cytomegalovirus (CMV) at time of amniocentesis. Birth weight of CMV-positive fetus that died at 21 weeks is not shown. (b) Box-and-whiskers plot showing birth-weight centile of fetuses with and without CMV transmission at time of amniocentesis. Boxes are median and interquartile range and whiskers are range.
) and without (
) transmission of cytomegalovirus (CMV) at time of amniocentesis. Head circumference of CMV-positive fetus that died at 21 weeks was not measured. (b) Box-and-whiskers plot showing head circumference centile of fetuses with and without CMV transmission at time of amniocentesis. Boxes are median and interquartile range and whiskers are range.At birth, we observed two additional cases with maternal–fetal transmission. The amniocentesis in these two cases was carried out at 19.4 and 20.3 weeks' gestation and was CMV-negative in both. The course of pregnancy was uneventful, and the two newborns were asymptomatically CMV infected.
Outcome of pregnancies without maternal–fetal transmission
In the pregnancies without transmission, delivery occurred at a median gestational age of 39.6 (IQR, 38.7–40.3) weeks. Within this group, there was one termination of pregnancy due to a chromosomal defect and one late miscarriage due to rupture of membranes. Ten fetuses were born preterm (one at 34 weeks and the remaining nine at 36 weeks).
The median birth weight of the 141 newborns that were liveborn was 3400 (IQR, 3100–3730) g, which corresponds to the 44th birth-weight centile (Figure 2). Four (2.8%) and 11 (7.8%) newborns had a birth weight below the 5th and 10th centile, respectively (Figure 2). Median head circumference was 35.0 (IQR, 34.0–35.5) cm (Figure 3). There were three pregnancies that were complicated by pre-eclampsia (delivery at 36.5, 38.0 and 38.9 weeks' gestation) and five women were treated due to preterm contractions. Four women had gestational diabetes, one woman suffered from depression and three pregnancies were complicated by an abnormally invasive placenta or placenta previa. In one case, there were recurrent episodes of vaginal bleeding with unknown origin leading to delivery at 34 weeks' gestation.
DISCUSSION
In this study of 149 pregnancies with a recent primary CMV infection up to 14 weeks' gestation, HIG treatment was used to prevent maternal–fetal transmission. The transmission rate at the time of amniocentesis was only 6.5%, HIG administration was predominantly well tolerated and the proportion of women with complications during the subsequent course of pregnancy was similar to that in the general population. The maternal–fetal transmission rate was similar between the group of pregnancies in which primary infection occurred in the first trimester and in the group in which primary infection was during the periconceptional period.
Multivariate logistic regression analysis indicated that the only significant predictor of maternal–fetal transmission at the time of amniocentesis was a higher anti-IgM index at the time of first presentation. Given that all pregnancies had a recent CMV infection, as established by several test systems, we believe that the increased anti-IgM level in cases with maternal–fetal transmission may indicate a more pronounced immune response compared to the non-transmission group.
The findings of previous studies with respect to the effectiveness of HIG treatment in pregnancies with CMV infection are inconsistent9. In the majority of the studies, HIG was given on a monthly basis and the treatment was continued up to the third trimester. However, we have demonstrated that the half-life of HIG is much shorter than previously thought, which necessitates more frequent administration21.
Two RCTs have examined the effectiveness of HIG to prevent fetal infection in pregnancy with primary CMV infection11, 12 and both failed to demonstrate a lower transmission rate in pregnancies that received HIG. In the study of Revello et al.11, 30% of the newborns were infected and in the trial of Hughes et al.12, the transmission rate was 22.7%. In the placebo arms of the two studies, the transmission rates were 44% and 19.4%, respectively. The study protocols of the two RCTs were similar to each other and differed in many aspects from our study. First, in both RCTs, the HIG dosage was 100 IU/kg maternal body weight and the treatment was repeated on a monthly basis up to almost the end of the pregnancy. In this study, we used a HIG dose of 200 IU/kg maternal body weight and administered the medication on a biweekly basis up to 18 weeks' gestation. The rationale for this protocol was based on the findings of our previous study14 and on our pharmacological studies indicating that the half-life of HIG is only about 10 days21. Second, we included only pregnant women with a recent primary infection22. We used several test systems and stricter inclusion criteria than those proposed recently by the ISUOG guidelines1 to make an unequivocal diagnosis of a recent infection. In our experience, there has to be a lack of IgG antibodies with high avidity in order for the HIG therapy to be useful. Third, we included only women with a primary infection up to 14 weeks' gestation whereas the RCTs recruited patients with a primary infection up to 26 and 24 weeks, respectively. As the risk of sequelae for second- and third-trimester infections is only 0.9% and 0.4%, respectively, we did not see an indication for HIG in these cases8. Fourth, in contrast to the two RCTs, we included only cases that underwent amniocentesis 6 weeks after the diagnosis of the maternal infection at the earliest. This gave us the chance to assess the infection status of the fetus during the most vulnerable time until 20 weeks' gestation23.
In contrast to other studies, we did not observe a higher proportion of obstetric complications during the subsequent course of the pregnancy after HIG treatment11, 12, 24. Revello et al.11 observed a preterm delivery rate of 7.6%, fetal growth restriction in 3.8% and eclampsia in 1.9% of the women treated with HIG. In contrast, in this study the preterm birth and pre-eclampsia rates, as well as the proportion of fetuses with a birth weight below the 10th centile, were consistent with the respective complication rates in the general population25-28. This difference may be explained by the fact that the HIG treatment was stopped before 20 weeks' gestation, while in the two RCTs, the treatment was continued up to the third trimester.
We acknowledge that this study has some weaknesses. We report on a large series of women, all of whom fulfilled the inclusion criteria and who were treated in four different centers. However, this is not a RCT. Such a study comparing HIG to placebo would no longer be feasible, certainly in Germany. In view of the results of the present and our previous experience14, it would be almost impossible to get ethical approval for such a study and recruitment would be extremely difficult.
We also acknowledge that, in this study, we do not report on the transmission rate of a control group that did not receive HIG. In our previous study, we reported on the outcome of a historical cohort of 108 women who had amniocentesis between 19 and 22 weeks' gestation due to a primary CMV infection in the first trimester. In this group, the transmission rate was 35.2% (95% CI, 26.2–45.0%)14.
In a recent RCT by Shahar-Nissan et al.29, the authors used valacyclovir (8 g/day) to prevent maternal–fetal transmission of CMV. The transmission rate was 30% and 11% in the placebo and in the treatment groups, respectively. Further studies are needed to compare the efficacy of valacyclovir vs HIG and to assess if a combination of the two therapies could be beneficial.
In conclusion, compared with the CMV transmission rate in a historical placebo group, HIG treatment is associated with lower maternal–fetal transmission rate in pregnancies with a primary CMV infection. However, HIG treatment seems to be beneficial primarily in women with a recent primary infection in the first trimester or during the periconceptional period, and when administration is biweekly at a dose of 200 IU/kg.
Disclosures
K.O.K. and K.H. have given paid presentations for Biotest, a company that produces immunoglobulin preparations. K.H. is a member of the Scientific Advisory Board of the Initiative for the Prevention of Congenital Cytomegaly Disorders (ICON). All related honoraria are paid into a UKT Institute for Medical Virology grant account to support the Tübingen Congenital CMV Study.
Acknowledgment
Open Access funding enabled and organized by ProjektDEAL.
Open Research
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are openly available.
