Exposure to quadrivalent influenza vaccine during pregnancy: Results from a global pregnancy registry

Abstract Background The Fluzone® Quadrivalent (IIV4, Sanofi Pasteur) Pregnancy Registry was created to monitor vaccine safety during pregnancy (clinicaltrials.gov, NCT01945424). Here, we describe maternal, pregnancy, obstetrical and neonatal outcomes after vaccine exposure in pregnant women between August 2013 and September 2019. Methods All women exposed to IIV4 during their pregnancy were eligible for inclusion. Outcomes were prospective (reported following vaccine exposure but before knowledge of pregnancy outcome ascertained through prenatal tests) or retrospective (prenatal tests were undertaken before the exposure was reported). Results Among 239 IIV4 vaccine exposure reports received, there were 105 prospective and 10 retrospective reports of maternal adverse events (AEs). The most frequent prospectively reported maternal AEs were medication errors (expired product [n = 8, 3.8%]; extra dose [n = 7, 3.3%]) and injection site pain (n = 7, 3.3%). Among 62 prospectively reported pregnancy and obstetrical events with available follow‐up information, seven AEs were reported, four (6.4%) of which were spontaneous abortions. A further seven AEs were reported among the 29 retrospective pregnancy and obstetrical events with available follow‐up information. Among neonatal outcomes (15 prospective; 28 retrospective), >85% were reported as full‐term births. One premature birth was reported prospectively. Four other neonatal AEs were reported, all retrospectively: two cases of talipes (club foot), one central nervous system anomaly and one atrial septal defect. All infants with available information had normal APGAR scores at 5 minutes. Conclusions The frequency of AEs following exposure to IIV4 during pregnancy did not indicate new safety concerns.


| INTRODUCTION
Seasonal influenza causes a substantial disease burden globally, with approximately 3-5 million cases of severe illness and up to 650,000 deaths annually. 1 Influenza infection during pregnancy, particularly during the second and third trimesters, is associated with an increased risk of developing severe complications such as pneumonia, leading to hospitalization and/or death. [2][3][4] Women who develop influenza during pregnancy also have an increased risk for adverse obstetric outcomes, including preterm labour and delivery. 2 Additionally, infants younger than six months old are at an increased risk of influenzarelated mortality and morbidity. 3,5 Maternal immunization against influenza is thus considered to be an essential component of prenatal care. 6 Immunization directly protects the mother and indirectly protects their unborn infant through the transfer of transplacental antibodies and secreted IgA antibodies from breastfeeding. 2,7,8 Infants often respond poorly to vaccines as a result of an immature immune system and are not included in the indication for influenza vaccination, but maternal antibodies may circulate in the infant until six months of age. 7,8 It is recommended that all pregnant women receive an influenza vaccine irrespective of their trimester of pregnancy. 1,[9][10][11] Nevertheless, vaccine coverage among pregnant women remains low in many countries. 12,13 In the United States, influenza vaccine coverage for pregnant women increased from 8.8% during the 2002-2003 influenza season to 50.9% for 2011-2012 14 but continues to remain below the 2020 Healthy People target to vaccinate 80% of pregnant women. 15 Coverage estimates below 10% have been reported among pregnant women in a number of Southeast Asian countries 16,17 ; in a cross-sectional study in Singapore, fewer than half (46%) of pregnant women knew that influenza vaccination was recommended during pregnancy. 16 Similarly, during the 2019 flu season in Canada, 45% of pregnant women received the recommended influenza vaccine. 18 Concerns over vaccine safety have been found to be a major factor contributing to reduced vaccine uptake among pregnant women. 12,19,20 Safety data collected after vaccine exposure during pregnancy in randomized clinical trials are generally limited to women who are inadvertently vaccinated during pregnancy, as pregnancy tends to be listed as an exclusion criterion at enrolment. 21 However, numerous observational studies have shown that inactivated influenza vaccination during pregnancy is not associated with an increased risk of pregnancy-related complications such as preterm birth and spontaneous abortion. 22

| Study population
The study population eligible for inclusion in the IIV4 pregnancy registry was all women of reproductive age who were exposed to IIV4 during their pregnancy or within 30 days of their last menstrual period (LMP). The 30-day window following their LMP ensured that women who may have been exposed immediately preceding conception were monitored; this also helps to account for inaccuracies in participant recall of LMP. For cases where the LMP was not reported, the estimated date of delivery (EDD) was used to approximate the LMP by counting back 280 days.
The exposure status for each report was verified. If not provided in the reporting information, where possible, the gestational week at the time of exposure was calculated by comparing the vaccination date to the LMP.

| Pharmacovigilance
All voluntary, spontaneous reports of exposure submitted to the registry by health care practitioners (HCPs) and pregnant women were captured in the Sanofi Global pharmacovigilance database. A recommendation to contact Sanofi Pasteur regarding all exposure occurring during pregnancy for inclusion in the registry is included in the prescribing information for a number of countries including the United States and Canada 28 and on the Sanofi Pasteur pregnancy registry website. 29 All individuals reporting pregnancy exposure to the vaccine were sent a Pregnancy Report Form to collect any missing or additional follow-up information on relevant maternal, pregnancy, obstetrical and neonatal outcomes, including any adverse events (AEs). For pregnancies ending in a live birth, a structured Infant Data Collection Form was sent 6 months after the EDD to collect follow-up information on the infant's condition and the diagnosis of any congenital anomalies. If either of these forms were not returned, three follow-up reminders were sent by Sanofi Pasteur. If the initial reporter did not have access to information on the infant, an attempt to obtain the contact information of the infant's paediatrician or doctor was made.
The initial pregnancy exposure reports were separated for analysis depending on whether the case was prospective or retrospective in order to reduce reporting bias. Pregnancy exposure reports were classified as prospective if the report was made following vaccine exposure but before knowledge of the pregnancy outcome was ascertained through prenatal testing. A report was considered to be retrospective if prenatal tests had already been undertaken before the exposure was reported. Reports with no information available on the timing of prenatal tests were considered prospective.
All reported AEs were coded according to the Medical Dictionary of Regulatory Activities preferred terms (MedDRA PTs), including diagnoses and symptoms. 30 Each event was treated independently, such that the total number of events was assessed separately from the number of women for whom AEs were reported.

| Outcomes
Exposure reports were categorised according to three predetermined outcome categories: (i) maternal outcomes, (ii) pregnancy and obstetrical outcomes and (iii) neonatal outcomes. Maternal events were defined as those impacting maternal health but independent of the pregnancy (e.g., injection site reactions); medication errors, including vaccine storage and administration issues, were included in line with good pharmacovigilance practices. 31,32 Pregnancy and obstetrical outcomes were those directly affecting pregnancy, labour or delivery (e.g., spontaneous abortion); neonatal outcomes were those that were directly related to the infant, evaluated at birth or within the first 28 days of life (including congenital abnormalities). For infants reported to have been born full-term and for whom the information was available, the reported birth weight and APGAR scores at the 5-minute mark were also reported. All AEs, including those reported through follow-up information and regardless of outcome category, were evaluated for medical seriousness at the time of receipt. Serious AEs (SAEs) included congenital anomalies, persistent or significant disability, life-threatening, hospitalizations, death and other medically serious events.

| Statistical methods
The IIV4 pregnancy registry is descriptive, and no predetermined target sample size was established. The baseline characteristics of the exposure reports were presented using descriptive statistics. The frequency of AEs within each outcome category and corresponding 95% confidence intervals (CIs) were calculated, stratified by prospective or retrospective reporting. The Agresti-Coull binomial proportions CI was utilized, supplemented by the Jeffreys interval when indicated. 33 Losses to follow-up were accounted for by adjusting the denominator used to calculate frequencies within each outcome category. When assessing maternal outcomes, all exposure reports were included in the denominator, regardless of the availability of follow-up information. For pregnancy and obstetrical outcomes, all cases where at least one successful follow-up was completed were included. For neonatal outcomes, all cases with follow-up information on the infant were included. For birth weight and APGAR score, the denominator included infants for whom this information was available. All analyses were performed using SAS Enterprise Guide 7.1 (SAS Institute, Cary NC).

| RESULTS
Between August 2013 and September 2019, 239 reports of pregnancy exposure to IIV4 were captured in the registry (210 prospective and 29 retrospective reports; Figure 1). The characteristics of these spontaneous reports are summarized in Table 1. Reports were received from nine different countries, with over 85% originating from the United States, Australia and Canada; 75% of the reports were submitted by HCPs.
There were 62 prospective reports of pregnancy and obstetrical outcomes with at least one successful follow-up, including four spontaneous abortions (6.5%), one case of foetal hypokinesia (1.6%) and two cases of morning sickness (3.2%) ( Table 2) Among the 29 retrospective reports of pregnancy and obstetrical outcomes with follow-up information, seven adverse pregnancy and obstetrical AEs were described (Table 2). These included two cases (6.9%) of gestational diabetes, one case of abnormal foetal heart rate (3.4%), one case of severe premature separation of the placenta  Table 3). The first case of talipes was in a male infant born at 39 weeks of gestation and was reported after prenatal testing had occurred. The mother, aged 37 years, received IIV4 at 29 weeks of gestation. No family history of talipes and no other AEs during pregnancy were reported. The second case of talipes was reported in a male infant born to a 36-year-old mother who was vaccinated at six weeks of gestation. During the pregnancy, the mother had a severe respiratory infection, which was treated with amoxicillin on an unspecified date for five days. The mother's prenatal laboratory tests, prenatal vitamins and ultrasound results were not reported. The report was filed almost six months after the infant was born.
The infant with a central nervous system anomaly was born to a F I G U R E 1 Flow diagram for reports received to the IIV4 pregnancy registry The APGAR score at 5 minutes was normal (≥7) for all 32 neonates with available information, and birth weight was normal (2,500-4,000 g) for 28/30 neonates. Low birth weight was reported for two infants prospectively.
Overall, 30 AEs reported for 19 pregnant women were considered an SAE (Appendix B). One SAE concerned a 31-year-old mother who experienced hypovolemic shock following H1N1 influenza virus infection 36 days after vaccination. The mother was vaccinated at 24 weeks of pregnancy and was considered to be a case of vaccination failure. She had a medical history of hypothyroidism treated with levothyroxine sodium. She was recovering at the time of the report, and the outcome of the pregnancy is unknown. A 29-year-old woman vaccinated during the third trimester of pregnancy developed Guillain-Barré syndrome (GBS) with onset five days after vaccination.
She gave birth to a baby girl who was reported to be healthy and was  The overall frequency of occurrence of congenital abnormalities among total exposure reports described here is comparable with the estimated background of clinically recognized major birth defects in the United States general population (2%-4%). 39 Notably, most neonatal AEs were reported retrospectively, meaning the occurrence of the event may have triggered the initial exposure report to be submitted to the registry. Among these were two cases of talipes in newborn boys. Although talipes is twice as likely to occur in male than female infants, 40,41 it occurred at a higher frequency in the current report (7.1%) than would be expected in the general population (estimated prevalence, 1 per 1,000 live births). 40 The use of prenatal vitamins and folic acid is known to help prevent the occurrence of congenital heart defects and neural-tube defects, including talipes. 42 However, the use of, or absence of, prenatal vitamins was not reported for these two cases. Because one case was reported following prenatal testing and the other six months after the birth, the potential for reporting bias through retrospective reporting should be considered.
Over 65% of the pregnancy exposure reports captured here were from the United States. Therefore, the outcome frequencies can be qualitatively compared to the AEs captured in the Vaccine Adverse T A B L E 2 Pregnancy and obstetrical adverse events following vaccination with IIV4  35 Notably, the rate of spontaneous abortion among pregnant women described in the current registry is no higher than background rates of miscarriage in clinically recognized pregnancy in the United States general population (15%-20%). 39 The IIV4 pregnancy registry is a passive pharmacovigilance surveillance system; as such, our data are subject to the limitations of passive surveillance that have been previously well-described. 43,44 These are likely to have resulted in an underreporting of pregnancy Jefferies method used to ensure attained intervals fell between 0,1 in accordance with the binomial distribution.