Background: Rapid assessment of COVID-19 vaccine safety during pregnancy is urgently needed. Methods: We conducted a rapid systematic review, to evaluate the safety of COVID-19 vaccines selected by the COVID-19 Vaccines Global Access-Maternal Immunization Working Group in August 2020, including their components and their technological platforms used in other vaccines for pregnant persons. We searched literature databases, COVID-19 vaccine pregnancy registries, and explored reference lists from the inception date to February 2021 without language restriction. Pairs of reviewers independently selected studies through COVIDENCE, and performed the data extraction and the risk of bias assessment. Discrepancies were resolved by consensus. Registered on PROSPERO (CRD42021234185). Results: We retrieved 6757 records and 12 COVID-19 pregnancy registries from the search strategy; 38 clinical and non-clinical studies (involving 2,398,855 pregnant persons and 56 pregnant animals) were included. Most studies (89%) were conducted in high-income countries and were cohort studies (57%). Most studies (76%) compared vaccine exposures with no exposure during the three trimesters of pregnancy. The most frequent exposure was to AS03 adjuvant, in the context of A/H1N1 pandemic influenza vaccines, (n = 24) and aluminum-based adjuvants (n = 11). Only one study reported exposure to messenger RNA in lipid nanoparticles COVID-19 vaccines. Except for one preliminary report about A/H1N1 influenza vaccination (adjuvant AS03), corrected by the authors in a more thorough analysis, all studies concluded that there were no safety concerns. Conclusion: This rapid review found no evidence of pregnancy-associated safety concerns of COVID-19 vaccines or of their components or platforms when used in other vaccines. However, the need for further data on several vaccine platforms and components is warranted, given their novelty. Our findings support current WHO guidelines recommending that pregnant persons may consider receiving COVID-19 vaccines, particularly if they are at high risk of exposure or have comorbidities that enhance the risk of severe disease.
For this rapid review, we followed the Cochrane methods [10], [11] and the 2020 Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [12] for reporting results. This review was registered in PROSPERO (CRD42021234185). We included studies that used comparative or non-comparative study designs. Case series were only included if they reported more than 50 exposed pregnant persons. We also included experimental studies of any sample size with exposed pregnant animals. We excluded systematic reviews (SRs) but explored their reference lists as an additional primary study source. The exposures or interventions of interest are the COVID-19 candidate vaccines that the COVAX-MIWG selected for review in August 2020; or the vaccine platforms (protein/subunit, vectored, nucleic acid/mRNA-LNP); or the components (antigen, vehicle, construct, adjuvants, lipid nanoparticles or other components) used by the selected COVID-19 vaccines (Table 1 ). At least one of these exposures was explicitly described in the report. Main characteristics of the vaccines that were selected for review by the COVAX-MIWG# in August 2020. # COVAX-MIWG: COVID-19 Vaccines Global Access – Maternal Immunization Working Group. * Merck discontinued the development of this vaccine on January 25, 2021. LNP: lipid nanoparticle; AS: Adjuvant System; CpG: Cytosine phosphoGuanosine; MRC: Human Fetal Lung Fibroblast Cells; CHO: Chinese hamster ovary; TM: transmembrane domain; S: Spike; FI: formalin-inactivated; rS: recombinant Spike. We considered outcomes concerning exposure to the vaccines based on the reported gestational age at vaccination (based on validated methods including ultrasound or last menstrual period [LMP] for human studies). We used the 21 standardized case definitions developed by the Global Alignment of Immunization Safety Assessment in Pregnancy (GAIA) of prioritized obstetric and neonatal outcomes based on the Brighton Collaboration process [13]. The ten GAIA obstetric outcomes include hypertensive disorders of pregnancy, maternal death, non-reassuring fetal status, pathways to preterm birth, postpartum hemorrhage, abortion/miscarriage, antenatal bleeding, gestational diabetes, dysfunctional labor, and fetal growth retardation. The 11 neonatal outcomes include congenital anomalies, neonatal death, neonatal infections, preterm birth, stillbirth, low birth weight, small for gestational age, neonatal encephalopathy, respiratory distress, failure to thrive, and microcephaly. For this rapid review, we considered the integrative outcome “safety concerns” as any statistically significant adverse outcome reported in the comparative studies, or unexpected frequencies with respect to the published incidences in the peer-reviewed literature reported in uncontrolled studies. We described all the adverse events as they were reported by the authors of the original studies. For the full review, safety outcomes will be analyzed according to the US Food and Drug Administration (FDA) Toxicity Grading Scale for Healthy Adult and Adolescent Volunteers Enrolled in Preventive Vaccine Clinical Trials [14]. An adverse event (AE) is defined as any untoward medical occurrence in a patient or clinical investigation subject administered a pharmaceutical product regardless of its causal relationship to the study treatment [15]. An AE can therefore be any unfavorable and unintended sign (including an abnormal laboratory finding), symptom, or disease temporally associated with the use of a medicinal (investigational) product. These include local reactions at the injection site (pain, tenderness, erythema, edema, pruritus, other) and systemic reactions (fever ≥ 38 °C or 100.4°F, headache, malaise, myalgia, fatigue, etc.). We will also consider other post-vaccination medical events (unsolicited in the studies, reported by organ system as per Medical Dictionary for Regulatory Activities – MedDRA) [16]. We will use the classification in a four grade for the severity of AEs. We also will consider other classifications of AEs commonly reported in safety studies, including: The operative definition of each specific AE was reported elsewhere (PROSPERO- CRD42021234185). We searched published and unpublished studies, without restrictions on language or publication status, from inception date to February 2021 (See the full search strategies and search terms in Appendix 1) in the Cochrane Library databases, MEDLINE, EMBASE, Latin American and Caribbean Health Sciences Literature (LILACS), Science Citation Index Expanded (SCI-EXPANDED), China Network Knowledge Information (CNKI), WHO Database of publications on SARS CoV2, TOXLine, preprint servers (ArXiv, BiorXiv, medRxiv, search.bioPreprint), and COVID-19 research websites (PregCOV-19LSR, Maternal and Child Health, Nutrition: John Hopkins Centre for Humanitarian health, the LOVE database). We also searched reference lists of relevant primary studies and systematic reviews retrieved by the search strategy and the adverse events/safety reported in active COVID-19 pregnancy registries. The Food and Drug Administration (FDA), the European Medicines Agency (EMA), and clinical trials websites will be searched for the full review. We will then contact original authors and experts in the field for clarification or to obtain extra information. For the full review, we will re-run the search strategy, between March 2021 and the current date and time, to capture any new evidence in databases. Pairs of authors independently screened each identified record by title and abstract and retrieved all the full texts of the potentially eligible studies. Pairs of review authors independently examined the full‐text articles for compliance with the inclusion criteria and selected the eligible studies. We resolved any disagreements by discussion. We documented the selection process with a PRISMA flow chart [12], conducted through COVIDENCE [17], a software for systematic reviews. Pairs of review authors independently extracted data from eligible studies using a data extraction form designed and pilot‐tested by the authors. Any disagreements were resolved by discussion. Extracted data included study characteristics and outcome data. Where studies have multiple publications, we collated multiple reports of the same study under a single study ID with multiple references. In Appendix 2, we describe the risk of bias assessment tools used for each study design. Briefly, we independently assessed the risk of bias of the included clinical trials using the Cochrane risk of bias assessment tool [18]. We used the Cochrane EPOC group tools [19] to assess controlled before‐after studies (CBAs), nationwide uncontrolled before‐after studies (UBAs), interrupted time series (ITSs), and controlled-ITSs (CITSs). We rated the risk of bias in each domain as “low”, “high”, or “unclear”. For observational cohort, case-control, cross-sectional, and case-series studies we used the NIH Quality Assessment Tool [20]. After answering the different signaling questions “Yes”, “No”, “Cannot determine”, “Not applicable”, or “Not reported”, the raters classified the study quality as “good”, “fair”, or “poor”. For consistency with the other designs, we use the classifications low, high, or unclear risk of bias, respectively. The primary analysis was the comparison of participants exposed and unexposed to the vaccines or their components. For this rapid review, we tabulated the study exposure characteristics and compared them against the unexposed. We analyzed the results of each study to determine any safety concerns as “Yes”, “No”, or “Unclear”. Data from non-comparative studies, including registries, were collected and analyzed in the context of background rates of neonatal and obstetric outcomes. For specific indicators, we take into consideration group-specific definitions such as low-to-middle-income countries (LMICs). We described the effect estimates as reported by the authors of the included studies. For dichotomous data, we used the numbers of events in the control and intervention groups of each study to calculate Risk Ratios (RRs), Hazard Ratios (HRs), or Mantel‐Haenszel Odds Ratios (ORs). We planned to conduct meta-analysis and subgroup analyses by the trimester of exposure and sensitivity analysis restricted to studies with a low risk of bias. However, these were not pursued for this rapid review, given the lack of safety concerns identified. We plan to perform a meta-analysis and present GRADE ‘Summary of findings’ tables [10], [21] for the full review as was previously stated (PROSPERO- CRD42021234185).