To be included in the pregCHIKV study, cases had to have the laboratory-confirmed symptomatic Chikungunya virus infection during pregnancy. Thus, included women had to have fever and detection of the viral RNA genome by reverse transcriptase-polymerase chain reaction (RT-PCR) or seroconversionIgM/IgG antibodies in the collected samples.
To be included in the control group, women had to have no fever, no dengue infection, and no CHIKVinfection at any time during pregnancy. Pregnant women were unexposed to CHIKV if they had given birth before the CHIK epidemic in the Caribbean area (before December 2013) or if specific IgM and IgG anti-CHIKV antibodies by “in-house”enzyme-linked immunosorbent assay testson blood sample at delivery were negative.
2.3. Biological AnalysisDuring the epidemic, biological diagnosis was performed by the French National Reference Center for arboviruses in French Guiana. For serological diagnosis, both IgM and IgG anti-CHIKV-specific antibodies were screened in sera using an “in-house” enzyme-linked immunosorbent assay (IgM antibody capture ELISA and ELISA, respectively) as described by Talarmin et al. [17]. For the molecular investigation, viral RNA was extracted from each biological sample with the QIAamp® Viral RNA kit (Qiagen, Hilden, Germany). Finally, extracted RNA samples were analyzed by real-time RT-PCR targeting the NSP1 gene of CHIKV, as described by Panning et al. [18]. Classification of patients as having laboratory-confirmed cases required either a positive RT-PCR, IgM/IgG seroconversion, or an isolated IgG detection with, at least, a subsequent 4-fold increase in titer between the first collected sample and the one obtained at delivery. 2.4. Data Collection and OutcomeData were collected from medical records and at the patient’s bedside (data about the standard of care). The quality of pregnancy follow-up was defined by having a first ultrasound before 14 weeks of gestation (WG) and at least 7 prenatal visits.
The pregnancy outcomes were collected are defined in Table S1. Given the risks of insufficient power for single pathological events, different pregnancy adverse outcomes were grouped into a single composite end point. Specifically, the criterion consisted of: preterm labor, preterm delivery, intra-uterine growth retardation, post-partum hemorrhage, caesarean section for abnormal fetal heart rate, or stillbirth.The neonatal outcomes were collected (Table S1) from birth to discharge (immediate postpartum). Similarly to what was done for pregnancy outcomes, different neonatal adverse outcomes were grouped into a single composite end-point. Specifically, the criterion consisted of: five-minutes Apgar score 2.5. Statistical AnalysisSTATA 16.0 (STATA Corporation, College Station, TX, USA) was used.
First, the analysis plan consisted of checking the comparability of the 2 groups (CHIKV group versus Control group) with respect to socio-demographic characteristics, quality of pregnancy follow-up, and history using Pearson’s chi2 or Fisher’s exact tests.
Then, pregnancy and neonatal outcomes were compared between the control group and the CHIKV group using Pearson’s chi2 or Fisher’s exact tests. The adjustment for potential confounders was performed using unconditional logistic regression. Adjustment factors were selected based on the scientific literature. The adjustment variables for the model studying the association between adverse events during pregnancy and maternal CHIKV infection were obstetrical history, primigravida, and quality of pregnancy follow-up. For the study of the association between neonatal outcomes and CHIKV infection, the adjustment variables selected were gestational age at delivery, mode of delivery, and primigravida.
Finally, for exploratory purposes, we studied pregnancy and neonatal outcomes only in the CHIKV group according to the time between fever and delivery. Thus, the CHIKV group was divided into 2 subgroups: one group of women who had CHIKV fever within 7 days before delivery and another group where women had CHIKV fever before 7 days before delivery. The 7-day threshold was chosen based on the duration of CHIKV viremia [2].The initial significance level was 5%; p values were corrected for multiple testing using the Bonferroni method.
2.6. Ethics StatementWritten informed consent was provided and signed by all subjects before enrolment. For minors, written consent was signed by both the participant and her legal representative. The study was approved by the French regulatory authorities CCTIRS (Advisory Committee on Information Processing in Material Research in the Field of Health n° 12323), the CNIL (National Commission on Informatics and Liberty; authorization DR-2012-585), and the CEEI (Inserm’s ethics committee/Institutional Review Board, approval n° 00003888).
3. Results 3.1. Participant CharacteristicsIn this case-control study, 246 pregnant women were enrolled: 73 cases in the CHIKV group and 173 in the control group. Among the CHIKV group, biological confirmation was done by detection of CHIKV viral RNA genome for 42 women (56.5%) and by IgM/IgG seroconversion for 31 women (42.5%). Virus sequencing was not performed in patients with a positive CHIKV RT-PCR.
Socio-demographic characteristics, obstetrical and medical history, and quality of pregnancy follow-up did not differ significantly between the two groups (Table 1). 3.2. Description of Maternal CHIKV Maternal CHIKV symptoms appeared in the third trimester in 63% of cases, in the second trimester in 33% of cases, and in the first trimester in 4% of cases. The median term of CHIKV infection was 30.7 weeks (IQR = 8.2). The main symptoms reported were: fever (100%), joint pain (63.0%), headache/retro-orbital pain (35.6%) (Table 2). Joint pain was more frequently reported for maternal infections occurring in the second trimester than for infections occurring in other trimesters of pregnancy (p-value = 0.023). The hospitalization rate for maternal CHIKV was 65.7%.Hospitalization occurred within 24 h of symptom onset in 60% of cases and between 24 and 48 h in 32% of cases. Length of hospital stay for CHIKV fever was between 1 and 3 days for 75% of women. No women were transferred to the intensive care unit (ICU). 3.3. Pregnancy OutcomesThe mean gestational age was 38.5weeks (±2.1) in the CHIKV group versus 38.9 weeks (±1.8) in the control group (p = 0.1232). Most deliveries were vaginal (79.4 % in CHIKV group versus 75.7% in control group; p = 0.527). Most fetal presentations were cephalic (97.3 % in CHIKV group versus 97.1% in control group). The median duration of maternal hospitalization for vaginal delivery was 3.5 days (IQR = 1) in the CHIKV group versus 3 days (IQR = 1) in the control group (p = 0.2367). The median duration of maternal hospitalization for cesarean delivery was 5 days (IQR = 2) in both groups (p = 0.630).
The occurrence of pregnancy adverse outcomes did not differ significantly between the two groups (Table 3). After adjustment for obstetrical history, being a primigravida, and quality of pregnancy follow-up, the association between pregnancy adverse outcomes (composite variable) and CHIKV during pregnancy (group variable) was still not significant (Wald test p = 0.361, AOR = 1.33 [0.72–2.43]). 3.4. Neonatal OutcomesOf 246 pregnancies, 114 newborns were male (46.3%). The mean birth weight of newborns was not different between the 2 groups: 3176 g (+/− 556 g) in the control group versus 3122 g (+/− 526) in the CHIKV group (p = 0.480). After excluding scheduled cesarean section (25/246), labor was spontaneous in 71.4% of cases for the CHIKV group and in 79.1% of cases for the control group (p = 0.221).
Congenital anomalies detected were: polymalformative syndrome (1), renal anomalies (1), undescended testis (3), Down syndrome (1), omphalocele (1), clubfoot (5), and cutaneous lesions (4).
The frequency of neonatal outcomes was not significantly different between the two groups (Table 4). After adjustment for preterm delivery, being a primigravida, and mode of delivery, the association between neonatal adverse outcomes (composite variable) and CHIKV during pregnancy (group variable) was still not significant (Wald test p = 0.327, AOR = 1.45 [0.69–3.08])The status of the newborn at birth in the CHIKV group according to the time from infection to delivery (>7 days and ≤7 days) is presented in Table 5. It shows that neonatal ICU admission seemed more frequent among women having had CHIKV infection within 1 week before delivery than in those with earlier infection. 4. DiscussionIn this carefully designed nested case-control study, the two groups were similar with respect to socioeconomic characteristics, obstetrical history, and medical follow-up. Considering pregnancy, delivery, or the neonatal period, we show that, overall, CHIKV infection during pregnancy was not associated with any significant difference in outcomes compared with controls. In addition, we observed that admission to the neonatal ICU appeared to be greater in mothers infected within 7 days of delivery than in mothers infected earlier in pregnancy.
As other studies, the present study may lack statistical power because of the small number of positive CHIKV cases. It, however, contributes additional information on a specific subject which is still underexplored. Although limited by its retrospective design, the collection of data and samples was similar in the two groups. In addition, the risk of underreporting is mainly related to minor or non-severe symptoms and not to major or severe events. Despite the limitations mentioned above, two methodological features should be highlighted in this study. First, because the study was nested in a cohort, inclusion of participants was done during pregnancy, which allowed for more reliable data collection. Second, strict biological and clinical criteria were used to define the comparison groups. Indeed, the group of women infected during pregnancy were required to have had at least one fever and biological confirmation of CHIKV infection; the presence of only positive IgM for CHIKV was not sufficient to prove infection during pregnancy [19] and the diagnosis had to be confirmed either by seroconversion on two successive samples during pregnancy or by a positive RT-PCR result. Women in the control group had to either have given birth before the arrival of CHIKV in the Caribbean area, or to have negative IgM/IgG serology at delivery. The epidemic dynamic in French Guiana and the arrival of CHIKV were well documented by the National Reference Center for Arboviroses, Institut Pasteur de la Guyane) who began close surveillance of when CHIKV emergence began in January 2013 in the general population [6].During the 2014–2015 epidemic, the majority of patients hospitalized for CHIKV infection in French Guiana presented a typical clinical picture. A few atypical or severe cases were reported, including neurological forms [7]. In our study, no atypical or severe cases were reported during pregnancy. The clinical presentation did not differ from that observed in non-pregnant patients [9,10,20]. Indeed, the three most frequently reported symptoms, regardless of the trimester of infection, were fever, joint pain, and headache/retro-orbital pain [4,6,13]. It appears that joint pain was more frequently reported when the infection occurred in the second trimester. Given the small sample size in this subgroup analysis, larger studies are needed to confirm these observations. Few severe cases of maternal CHIKV infection requiring intensive care admission have been reported in the literature [6]. Perhaps these publications are rare exceptions to the rule or result from the effect of different sociodemographic and maternal contexts (history, immunity, pregnancy follow-up, etc.).The study in French Guiana confirmed the results observed in other comparative studies of Chikungunya infection during pregnancy [8,10,21]. Maternal CHIKV did not increase the risk of pathological pregnancy. The comparability of the groups in terms of sociodemographic characteristics, obstetrical history, and pregnancy follow-up allowed us to manage potential confounders. The study compared clinical and non-biological endpoints, particularly in neonates. No biological information regarding vertical transmission of CHIKV was available, which is a limitation of the present study. However, events and neonatal management were not different between cases and controls (Table 4). It is notable, however, that neonatal ICU admission appeared to be more frequent when maternal infection occurred shortly before delivery, in the period of maternal viremia (Table 5—significant difference).Although, this result should be interpreted with caution because of the lack of statistical power, it echoes similar observations reported in the literature. The study of vertical transmission of CHIKV during the 2005–2006 epidemic in the Reunion Island also suggested an increased risk of vertical transmission of CHIKV from mother to child during the period of maternal intrapartum viremia with a risk of severe neonatal infection (encephalopathy, hemorrhagic fever) [13,14]. Indeed, the transmission rate of CHIKV from mother to child reaches 50% when delivery occurs during the viremic period [14,22,23]. Neonatal meningoencephalitis in the context of maternal infection with CHIKV close to delivery has also been reported in French Polynesia [13] and in Latin America (El Salvador, Colombia, Dominican Republic) [24,25]. In our study, as in the West Indies, no cases of encephalopathy were observed [5,25]. One of the hypotheses put forward to explain these differences in neonatal infection is the viral genotype: the Asian isolate at the origin of the epidemics in the West Indies and French Guiana could be less virulent than the one from the Indian Ocean [26,27]. 5. ConclusionsWhile the impact of CHIKV during pregnancy does not appear to increase the risk of maternal and fetal complications, vigilance is warranted when delivery occurs during the maternal viremia period. Indeed, the results of this study suggest an increased risk of admission to intensive care for the newborn when maternal CHIKV infection occurs within 7 days before delivery, i.e., during the period of viremia.
Further studies and meta-analyses of published results could focus on women infected in the week prior to delivery to estimate vertical transmission of CHIKVand newborn health status with greater statistical power.
留言 (0)