Development of standard definitions and grading for Maternal and Fetal Adverse Event Terminology

Adverse event (AE) monitoring is central to assessing therapeutic safety. The lack of a comprehensive framework to define and grade maternal and fetal AEs in pregnancy trials severely limits understanding risks in pregnant women. We created AE terminology to improve safety monitoring for developing pregnancy drugs, devices and interventions.

decisions around dose-escalation and the Maximum Tolerated Dose (MTD) are based on observation of Adverse Reactions (ARs) of given severity. The most widely used system, the Common Terminology Criteria for Adverse Events (CTCAE) (Version 5.0) comprises 837 potential AE, of which only 4 relate to 'pregnancy, the puerperium and perinatal conditions'. 9 Some condition-specific severity grading for pregnancy-specific events have been developed (e.g., in HIV-AIDS and surgery). [10][11][12] However, there remain no standard general severity grading criteria. This contrasts with recent Delphi consensus work to integrate neonatal terminology and definitions into wider dictionaries. 13,14 We convened an international multidisciplinary group to develop standard definitions and severity grading criteria to enable objective reporting of AEs in all clinical trials involving pregnant women.

| METHODS
We undertook a stepwise consensus process through a Delphi approach, between January 2015 and December 2019 ( Figure 1). 15 Following a comprehensive review of existing terminology, consensus statements and guidelines, draft maternal and fetal AE definitions and severity grading criteria were agreed by the Steering Group. New fetal AE terms were integrated with MedDRA (Medical Dictionary for Regulatory Activities) requirements and added to MedDRA terms list. The draft fetal and maternal AE definitions and severity criteria were reviewed by a Patient Public Advisory Group (PPAG) before undergoing a two-stage international modified Delphi consensus process. A final set of maternal and fetal AE definitions and severity criteria were agreed.

| Review of existing AE terminology
In January 2015, we consulted internationally with academic and industry pregnancy and fetal therapy clinical triallists to identify existing maternal and fetal AE definitions and severity grading criteria for clinical trials in pregnancy (Table S1). 10,11,16 Existing definitions and severity grading criteria were reviewed with MedDRA preferred terms for AEs.  (Table S2); key references were retrieved (March-April 2015) F I G U R E 1 An overview of the three-stage development process of the standard maternal and fetal AE severity grading criteria. Stakeholder involvement is indicated by C (clinicians), S (scientists), I (industry), M (midwifery representatives), P (patient and/or charity representatives) and R (regulatory authority employees the first four were not specifically pregnancy related AE systems. Two included no pregnancy-specific AEs 63,64 and the other three were considered insufficient to meet the needs of trials of maternal and fetal therapies, particularly as few fetal AEs were included. 10,11,65 MedDRA AE terms for 'f(o)etal complications' were limited and related generally to teratogenic effects, or fetal malposition and malpresentation in labour.

| Review of consensus statements and guidelines literature
Handsearching full lists of guidelines from National Colleges and Professional Societies (Table S2)  identified maternal AEs that were grouped into 10 categories.
Identified documents were examined for definitions of AE terms and criteria used to indicate their severity. A summary was prepared for the Steering Group before the first meeting (Table S3).

| Review by Patient Public Advisory Group
The PPAG review supported the general principles of assessing AE severity in pregnancy proposed by the Steering Group and the preliminary generic and specific AE severity grading criteria. The group also identified broader considerations on the AE assessment in pregnancy (Table 3). Among these were the potential psycho-

| Main findings
We describe the development of the most comprehensive set of maternal and fetal AE definitions and severity grading criteria available, which can guide investigators and clinicians in assessing the severity of AEs so as to increase the quality of safety information. By reviewing international guidelines and consensus statements, building on the existing CTCAE framework, liaising with MedDRA and seeking consensus from international experts through a modified Delphi consensus process we have made this terminology as robust as possible.
T A B L E 3 Patient and Public Advisory Group recommendations for best practice in clinical trials of maternal and fetal therapies � Define the timing of AE assessment in the clinical trial protocol. Depending on the intervention short-term, medium-term, and/or long-term AEs may be appropriate and the severity grading may change across these timepoints. Ideally report as much as possible, potentially as supplementary online information. � Record information on antenatal decisions to terminate the pregnancy or to have only palliative neonatal care after birth. State in the clinical trial protocol how these outcomes will be analysed and consider them part of the same group when analysing data and measuring mortality. � Mode of labour onset and mode of delivery are outcomes that should be reported, including whether the mode of delivery is likely to impact on decisions regarding delivery in future pregnancies (i.e., women undergoing classical Caesarean section or open fetal surgery are advised to delay another pregnancy for a year and to avoid labour). � Assess the psychological impact of the intervention on the pregnant woman and her partner including the psychological impact of any fetal AEs.
Evaluate using validated measures in comparison with an 'untreated' group with the same condition. � Where possible include costs in clinical trial funding applications for psychological support for pregnant women and their partners and, especially in phase I trials, consider qualitative assessment of the women's experience. � Include in the clinical trial protocol assessment methods to capture data about the fetal response to an intervention, including indications of fetal pain or stress. Appropriate measures are likely to vary depending on whether the intervention involves the fetus directly (fetal surgery) or indirectly (maternal medication). � Capture data on subsequent fertility and pregnancies over a time period proportionate and relevant to the intervention under investigation. This should include whether women were trying to conceive, and their pregnancy outcomes and complications if they were successful.  66 We hope that this terminology will now be adopted by triallists, industry and regulatory authorities to address this deficit in treatments for pregnant women and their fetuses.

| Interpretation
The MFAET version 1.0 system can also be used to standardise recording and severity grading of AEs in untreated populations, to provide comparison and valuable reproducible contextual 'control' data with which to interpret safety and AEs in interventional trials.
The EVERREST prospective study, for example, is defining characteristics of pregnancies affected by severe early onset fetal growth restriction, generating control data for a future first-inhuman intervention trial of maternal gene therapy. 67 The terminology is being used to derive the dose escalation plan of the

| Strengths and limitations
The strength of our process is that we carefully examined existing terminology and literature to identify gaps in AE assessment for the mother and the fetus. Our Steering Group and Delphi consensus members included multiple key stakeholders involved in developing maternal and fetal drugs, devices and surgical interventions. We ensured international participation in the process with academic and non-academic clinicians and researchers, industry representatives, scientists, midwifery and parent representatives.
This terminology should however not be considered final or exhaustive. Future refinement and expansion, such as that undertaken by other AE severity grading systems, will continue to improve these criteria with revised versions to be released in the future. More criteria are likely to be added as the process matures, fetal anaemia being one such criterion that has not yet been included. Until the next version is available, triallists can apply the generic criteria to estimate the severity for AEs that may not be currently included in MFAET version 1.0. We did not specifically develop an AE definition for analysis of Cardiotocography (CTG, non-stress test 'NST') as this will be accommodated in the fetal heart rate AEs (Fetal Bradycardia and Fetal Tachyarrhythmia).
Future versions may need to accommodate new methods of fetal assessment including objective fetal movement monitoring, computerised CTG analysis of short-term variability (STV) and machine learning algorithms.
Laboratory-based AEs for the pregnant woman such as abnormal liver or thyroid function are not included in any existing grading criteria such as CTCAE, DAIDS, etc., and will require future development. Deriving laboratory-based AEs for the fetus depends on data from invasive sampling of amniotic fluid, placental villi, fetal blood or urine which carries a risk of miscarriage or preterm labour. Historical data is available and has been used to develop non-invasive ultrasound Doppler methods to diagnose fetal anaemia. 71 As analysis of fetal circulating DNA, RNA and proteins in the maternal blood advances, it is possible that fetal laboratory-based AEs will be developed to further assess fetal well-being. 72,73 The goal of a standardised AE severity scale is to reduce subjectivity in severity assessments and thus reduce interobserver