Characteristics and temporal trends of EU teratological safety signals and information dissemination to healthcare professionals in the Netherlands—A contribution of the ConcePTION project

Abstract

Aim

Although regulatory responses to safety signals have been studied, pregnancy‐specific data and their dissemination to medical professionals remain unclear. This study examined the characteristics and temporal trends of teratological safety signals prompting the European Medicines Agency to implement risk minimization measures. A secondary aim was to investigate how information on these signals, independent of regulatory communication, was disseminated to medical professionals in the Netherlands.

Methods

Teratological safety signals from 2002 to 2022 were retrieved from the European Pharmacovigilance Issues Tracking Tool. For each signal, we recorded the dates of market authorization, first spontaneous report, scientific publication, regulatory decision and dissemination through the Teratology Information Service (TIS) and Dutch professional journals. The source of the signal (spontaneous reports, scientific literature or otherwise) was assessed. Time intervals between events were calculated and visualized.

Results

Twenty‐four safety signals pertaining to fetal teratogenesis were identified. At the time of regulatory action, medicines had been on the market for a median of 23.9 years (range 3.2–59.3). Cases or studies published in the scientific literature were the origin for most (n = 18, 75%) teratological safety signals. Amendments of the product information was a regulatory action for all signals; eight (33%) also triggered Direct Healthcare Professional Communications. Apart from these regulatory communications, information may have reached healthcare professionals through scientific literature or other channels, sometimes before and sometimes alongside regulatory decisions.

Conclusion

Teratological safety signals are predominantly scientific literature‐driven and often arise decades after marketing. Scientific publications and TIS may provide early guidance, frequently preceding regulatory decisions.

What is already known about this subject

• Drug safety information on pregnancy exposure is largely obtained post‐marketing.

• The median time for decisions on non‐teratological specific safety signals in the EU is 12 years after marketing authorization.

What this study adds

• In the EU, safety signals related to teratological risk are generated with a median time of 23.9 (range 3.2–59.3) years after the drug has received market authorization.

• Unlike other safety signals, which are frequently triggered by spontaneous reports, teratological safety signals are mainly literature driven.

• Beyond regulatory communications, information on teratological safety signals may have reached healthcare professionals through scientific literature or other channels, sometimes before and sometimes alongside information on regulatory decisions.

1. INTRODUCTION

About 200 million women become pregnant across the world every year. ¹ The use of medications during pregnancy is common, though the prevalence differs among various regions of the world. Estimates suggest that between 40% and 96% of pregnant women take at least one prescribed medication. ² , ³ , ⁴ , ⁵ , ⁶ Despite its common occurrence, obtaining reliable safety data on medication use in pregnancy remains challenging. Pregnant women are generally excluded from clinical trials that evaluate the efficacy and safety of new medicines, although in some cases their inclusion may be scientifically justified and ethically appropriate. ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ As a result, safety data is mainly obtained after market authorization during use in daily clinical practice. ¹² Well‐known examples include the association between thalidomide and phocomelia, as well as the link between mycophenolate mofetil and congenital malformations, both identified through observations in clinical practice prompting clinically led research. ¹³ , ¹⁴ Similarly, the connection between valproate and neurodevelopmental disorders has been established based on observational studies arising from clinical cases. ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹

The European Medicines Agency (EMA) defines a safety signal as ‘information on a new or known adverse event that is potentially caused by a medicine and that warrants further investigation’. ²² Teratological safety signals can be identified through various sources. These may include the assessment of Individual Case Safety Reports (ICSRs) submitted voluntarily by healthcare professionals or pregnant women to adverse event monitoring systems, dedicated pregnancy registries, enhanced pharmacovigilance programmes, Teratology Information Services (TIS) and observational studies. ²³ , ²⁴ Such information provides initial insights, although further research is often needed to confirm these signals. In the European Union (EU), ICSRs are captured in EMA's EudraVigilance database and transmitted to VigiBase. This global database is maintained by the Uppsala Monitoring Centre, the WHO Collaborating Centre for International Drug Monitoring, and contains information of ICSRs of more than 180 countries worldwide. ²⁵ , ²⁶ , ²⁷

Signal detection and management involve coordinated efforts by marketing authorization holders (MAHs), national regulatory authorities and the EMA. ²⁸ The latter two can validate signals, by evaluating evidence and register them in the European Pharmacovigilance Issues Tracking Tool (EPITT) for monitoring regulatory actions. Subsequently, the signal can be assessed by EMA's Pharmacovigilance Risk Assessment Committee (PRAC). ²² If confirmed with a likely causal relationship, regulatory recommendations may be issued to minimize risk, including Direct Healthcare Professional Communications (DHPC), updating the Summary of Product Characteristics (SmPC), revising the risk management plan (RMP) and/or modifying the patient information leaflets (PILs). ²⁹ In exceptional cases, a medicine may be suspended or withdrawn from the market. ²² Studies examining EMA regulatory decisions following safety signals have found a median interval of 12 years between marketing authorization and decision, with most signals originating from adverse event reports. ³⁰ , ³¹ However, these studies did not focus specifically on pregnancy‐related signals or examine dissemination of safety information to healthcare professionals. Beyond the regulatory pathway, healthcare professionals also receive safety information through additional sources including scientific literature, specialist networks, and TIS. This information is also regularly published in professional journals for medical practitioners and pharmacists. The purpose, nature and speed of information dissemination may differ across these communication channels. By examining the nature and context of these safety signals, as well as how the information is communicated, we can gain a clearer understanding of their significance.

This study aimed to examine the characteristics and temporal trends of teratological safety signals that prompted the EMA to implement risk minimization measures between 2002 and 2022. A secondary aim was to investigate how information on these signals, irrespective of communication by regulatory authorities, was communicated to healthcare professionals in the Netherlands.

2. METHODS

2.1. Study design and cohort

A retrospective, descriptive study was conducted to examine signals related to teratological safety which were assessed by EMA between 2002 and 2022. This could either concern the PRAC or its predecessor the Pharmacovigilance Working Party (PhVWP). Teratological safety signals had to involve pregnant women exposed to medicines potentially affecting the (unborn) child. Only signals that had been discussed by the PhVWP between 1 January 2002 and 19 July 2012 or the PRAC between 19 July 2012 and 31 December 2022 were taken into account.

Teratological safety signals were obtained from the EPITT database, which was established in 2012 but also contains data dating back to 2002. Suspected teratological adverse events were coded using the Medical Dictionary for Regulatory Affairs (MedDRA—V25.0), and drug names mapped to Anatomical Therapeutic Chemical (ATC) codes. ³² , ³³ All pregnancy‐related safety signals in the study period were collected by two staff members of the Dutch Medicines Evaluation Board (MEB) (JvW, HC). To focus on clinically relevant safety signals, only those leading to a regulatory risk minimization action were included.

2.2. Characteristics of the teratological safety signals

For each teratological safety signal, the following information was extracted from EPITT: Reference number, the date of the regulatory decision, description of the safety signal and the corresponding regulatory action, as well as the origin of the data underpinning the decision. If multiple decisions applied to the same association, this information was also noted.

Marketing authorization dates of the medicines were primarily obtained from the European Union Reference Dates (EURD) list, which is publicly available on the EMA website. ³⁴ This list includes only active substances subject to assessment at the European level. For medicines authorized nationally, data were retrieved from the Medicines Information Bank (Geneesmiddeleninformatie‐bank) on the website of the Dutch MEB. ³⁵ When a signal encompassed multiple medicines, such as those belonging to the same drug class, the marketing authorization date of the earliest approved active substance was used.

Information from PRAC‐assessed signals was also available in publicly accessible EMA meeting minutes, which were reviewed. ³⁶ Earlier signals were assessed by the PhVWP, whose meeting minutes and recommendations were publicly available from September 2009 to July 2012. ³⁷ For signals predating this period, non‐public minutes were available for review by the Dutch MEB (JvW, HC).

The origin of the initial safety signal was classified using the following (combination of) categories: spontaneous case reports; cases or studies published in the scientific literature; and other sources (e.g., other regulatory agencies or MAHs). The EMA regulatory action was classified as (a combination of) amendment of the SmPC; DHPC; amendment of the RMP; suspension or withdrawal from the market; and/or other measures to minimize the risk. For associations with multiple decisions at different moments, the date of the first decision was used.

To determine the first ICSR globally for each safety signal, a search was conducted in VigiBase. Although signals were also identified via the European database, this database allows capture of initial reports that may also have originated outside Europe. Searches used the active substance and relevant MedDRA terms, limited to spontaneous reports of medicinal product use during pregnancy with potential effects on the offspring. Results were filtered by date, and the earliest report was recorded on the timeline.

2.3. Dissemination of information to healthcare professionals

Beyond regulatory communications aimed at healthcare professionals, such as SmPC amendments or DHPCs, information on the associations underlying safety signals may also have been disseminated through scientific publications and, in the Netherlands, via the Dutch TIS and professional medical and pharmaceutical journals. In some cases, this information may precede regulatory decisions; in others, it triggers or coincides with them.

To identify the earliest published reference in scientific literature, a PubMed search was conducted for full‐text articles in English or Dutch. Additionally, the journal Reactions Weekly was searched for supplementary case reports, as it compiles content related to adverse drug reactions (ADRs) from scientific journals, media releases and regulatory websites. ³⁸ Relevant publications and their references were reviewed, and the earliest available article was selected in consensus with an epidemiologist (YvRW) and a medical doctor/clinical pharmacologist (EvP).

Information may also have been shared through the Dutch TIS, which is part of the national Pharmacovigilance Centre Lareb. This TIS disseminates information on the potential teratogenicity of medicines through the TIS Knowledge Bank, an online library or information pages which are updated as new scientific literature becomes available. ³⁹ The findings are also published in the online reference work for healthcare professionals ‘Commentaren Medicatiebewaking (CM)’ enabling use in pharmacist and physician information systems for clinical decision support, particularly maternal exposure to potentially teratogenic medication. ⁴⁰ , ⁴¹ , ⁴² To study the changes in the information provided by the Dutch TIS, editions of CM published between 2000 and 2022 were manually reviewed. Access to the PDF versions of CM was granted by its publisher Health Base Foundation. ⁴³ CM is published annually, and any update related to a safety signal was recorded with the date of June 1st of the corresponding year.

In respect to the Dutch professional journals, articles published between 1 January 2000 and 1 January 2023, in the journals Nederlands Tijdschrift voor Geneeskunde, Nederlands Tijdschrift voor Obstetrie en Gynaecologie and Pharmaceutisch Weekblad that described the medicinal product and ADR in the safety signal or reported EMA regulatory decisions, were included. The name of the medicinal product or drug class was used to search within the journals.

2.4. Timeline of events

To map the temporal trends of events related to the safety signals, the intervals between the dates of the events listed below were calculated for all signals, as well as separately for those with and without an issued DHPC.

1. Marketing authorization

2. Earliest mention in scientific literature

3. First ICSR reported globally

4. Information published by the Dutch TIS

5. Publication in Dutch professional journals

2.5. Data analysis

Median time intervals between the events outlined in paragraph 2.5 and the regulatory decision date (used as the reference point) were calculated using descriptive statistics. Boxplots were generated to visualize these time intervals. A Mann–Whitney U test was applied (SPSS Version 25.0) to compare the timing of events between signals that prompted a DHPC and those that resulted solely in an SmPC amendment.

2.6. Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://wwwguidetopharmacology.org and are permanently archived in the concise Guide to PHARMACOLOGY 2021/2022. ⁴⁴

3. RESULTS

During the study period, a total of 35 teratological safety signals were reviewed by the EMA. For two signals, only additional information was requested, while for nine signals, it was decided that routine pharmacovigilance should be continued. The remaining 24 safety signals, for which risk minimization measures were either introduced or modified, were included in this study. In two of these signals, existing measures were reduced; the recommendation for dual contraception was changed to single contraception for one product, and another product was subsequently considered safe for use during pregnancy.

3.1. Characteristics of the teratological safety signals

Most discussed signals concerned only one medicinal product (n = 17; 71%). The remaining signals were associated with at least two different medicines, but more often to an entire class.

The medicinal products involved belonged to the following ATC classes: antidepressants (n = 2), antivirals (n = 4), antiemetics and antinauseants (n = 1), antiepileptics (n = 3), thyroid therapy (n = 2), immunosuppressants (n = 2), analgesics (n = 1), antipsychotics (n = 1), antimycotics (n = 1), agents acting on the renin–angiotensin system (n = 1), antithrombotics (n = 1), antibacterials (n = 1), diuretics (n = 1), antihistamines for systemic use (n = 1) and urologicals (n = 1) and a combination preparation containing both analgesics and decongestants (n = 1). Adverse events were related to congenital disorders (n = 19), miscarriage and congenital disorders (n = 3), miscarriage and stillbirth (n = 1) and preterm birth (n = 1).

With regard to the origin of the safety signals, one signal was based on spontaneous case reports, 20 originated from cases or studies published in the scientific literature and three were identified from other sources (two from the Food and Drug Administration [FDA] decisions and one from a post‐authorization safety study).

All signals led to SmPC amendments, with eight accompanied by a DHPC; one of these also required an RMP update. No signal resulted in the suspension or withdrawal of a medicinal product.

3.2. Dissemination of information to healthcare professionals

The Dutch TIS updated information on 18 signals, including all eight DHPC‐related signals and 10 out of 16 signals that lead to an SmPC amendment.

First references in scientific literature were identified for 22 of the 24 signals, with 20 of these published before the corresponding regulatory decision. In the main Dutch medical and pharmaceutical professional journals 12 publications containing relevant information on eight safety signals were identified, for six of which a DHPC had been issued. Two signals were discussed in all three journals, while the Dutch Pharmacy journal was the only journal to publish relevant articles on the remaining six signals.

3.3. Timeline of events

An overview of the timing of the various events for all safety signals relative to the date of the regulatory decision, is presented in Table 1. The median interval between initial marketing authorization and regulatory decision was 23.9 years, with the longest‐marketed drug having been available for 59.3 years. Although considerable overlap between the events exist, they generally occurred in the following consecutive order: First ICSR; the first reference in scientific literature; TIS information; regulatory decision EMA; dissemination in Dutch professional journals (Figure 1). ICSRs associated with DHPC signals were reported sooner than those linked to SmPC signals (5.1 vs. 13.2 years), the difference being statistically significant.

TABLE 1.

Median timespan between key events for all signals and by outcome (DHPCs vs. only SmPC amendments). The timespan before the regulatory decision is shown, unless marked with *, which refers to the period after the decision. MA = marketing authorization date; DHPC = Direct Healthcare Professional Communication; SmPC = Summary of Product Characteristics; TIS = Teratology Information Service; CM = Commentaren Medicatiebewaking; ICSR = Individual Case Safety Report. Differences between timing of signals leading to DHPCs and SmPC were tested with a Student's t‐test.

Events	All signals			Signal resulting in DHPC			Signal resulting in SmPC only			Difference signals resulting in DHPC vs. SmPC only
	N	Median	Range	N	Median	Range		Median	Range	Mann–Whitney U
Time from MA to regulatory decision (years)	24	23.9	59.3/3.2	8	22.8	47.4/4.7	16	25.7	59.3/3.2	0.490
Time from first ICSR to regulatory decision (years)	20	10.8	32.4/0.2	6	5.1	10.6/0.2	14	13.2	32.4/0.7	0.012
Time from first reference in scientific literature to regulatory decision (years)	22	7.1	48.4/4.0*	7	5.3	36.9/0.0	15	8.6	48.4/4.0*	0.482
Time from TIS information update to regulatory decision (years)	18	2.4	19.0/13.6*	7	2.8	14.0/1.7*	11	1.5	19.0/13.6*	0.592
Time from publication in Dutch professional journals to regulatory decision (years)	8	0.1*	2.2/5.7*	6	0.1*	2.2/1.1*	2	2.1*	1.6/5.7*	1.000

FIGURE 1.

Box and whisker plots showing the timing (minimum, first quartile, median, third quartile, maximum and outliers) between the date of the regulatory decision and the date of marketing authorization, first ICSR in VigiBase, the first reference in scientific literature, advise by the Dutch TIS and article in one of the Dutch professional journals. Abbreviations: ICSR, Individual Case Safety Report; TIS, Teratology Information Service.

Regarding the dissemination of information, the intervals between the first scientific publication and the regulatory decision were generally shorter for DHPC‐related signals (median 5.3 vs. 8.6 years). TIS updates occurred earlier for DHPC‐related signals than for those linked to SmPC amendments (median 2.8 vs. 1.5 years), while the median interval between the regulatory decision and publication in Dutch professional journals was shorter for DHPC‐related signals (median 0.1 vs. 2.1 years), although none of these differences were statistically different.

4. DISCUSSION

This study investigated the characteristics and temporal patterns of key events related to teratological safety signals evaluated by the EMA between 2002 and 2022 and examined how these signals were translated into information for healthcare professionals in the Netherlands. Although not statistically significant, signals that triggered a DHPC tended to have shorter intervals between the regulatory decision and the occurrence of the first reference in scientific literature, the first ICSR, as well as time on the market, compared with signals leading only to SmPC amendments. This may reflect differences in regulatory urgency.

In addition to the dissemination of knowledge about regulatory measures, scientists and healthcare professionals play a clear role in spreading knowledge regarding safety of products used by pregnant women. The typical sequence of events, as presented in Figure 1, shows that information about safety signals is shared with healthcare professionals, through scientific publications or via the TIS or professional publications often before a formal regulatory decision is made. Women using medication during pregnancy benefit from being informed as is illustrated by the EMA's public hearing on valproate, where existing risk minimization measures showed to be insufficient. ⁴⁵

TIS actively monitors scientific literature, enabling rapid updates as relevant studies emerge. This allows TIS to inform healthcare professionals of potential concerns before regulatory action, supporting informed decision‐making. Our analysis showed Dutch TIS often updated its guidance before EMA safety signal issuance. However, TIS updates may differ in scope and nature from formal regulatory decisions. While both TIS and regulatory authorities may use overlapping literature sources, regulatory processes are more time‐intensive due to required coordination across member states and MAHs, requirements for supplementary evidence and implementation of risk minimization measures.

4.1. Previous research

Pacurariu et al. studied the characteristics of signals reviewed by the PRAC during its first 18 months from July 2012 till December 2013. ³⁰ Farcaş et al. analysed safety signals of the PRAC between January 2014 and December 2017. ³¹ Both studies reported a low proportion of pregnancy‐related signals, with 1.6% in utero exposure in the study of Pacurariu et al., and 2.9% in the study of Farcaş et al. The median time leading to a general safety signal in these studies, including non‐pregnancy‐related signals, was 10.5–12 years. ³⁰ , ³¹ The median time from approval to 2010 general FDA safety‐related label change showed a comparable time span of 11 years. ⁴⁶ Shorter intervals in non‐pregnancy studies likely reflect clinical trial monitoring and broader pre‐marketing use, whereas limited safety data and reduced medication exposure during pregnancy tend to delay signal detection. ⁷ , ⁸ The extended timeframe may reflect that pregnancy‐related signals typically originate from research studies rather than spontaneous reports.

In the study of Pacurariu et al. ³⁰ , three signals led to a DHPC and 27 to SmPC amendments, while our study found seven DHPCs and 22 SmPC amendments—a statistically significant difference (p = 0.012, Chi‐square), suggesting pregnancy‐related signals are more likely to trigger a DHPC. Similarly, Farcaş et al. reported 17 DHPCs and 86 label amendments, which is also significantly lower than in our study (p < 0.001, Chi‐square). ³¹ It should be noted that regulatory procedures during the early years of PRAC's operation may have differed from those throughout the entire duration of our study.

Both studies found that spontaneous reports were the main source of information for 62% and 64% of the safety signals in the studies by Pacurariu and Farcaş respectively. ³⁰ , ³¹ These figures exclude literature case reports, which were responsible for an additional 10% and 16% of the signals in both studies. In contrast, the present study identified spontaneous reports as the source for only 1 of the 24 signals, with the majority of teratological signals being based on scientific literature. Although some of these literature sources may have included ICSR‐based case reports or case series, the findings suggest that the sources of pregnancy‐related safety signals differ from those of non‐pregnancy‐related signals. Clinical observations and case reports rarely identify teratogenic medicines, except in cases of rare congenital disorders, because detecting an association requires both a strong teratogenic effect and a low background frequency. Formal epidemiological studies are considered more effective for investigating moderate effects and more common outcomes. ⁴⁷

4.2. Methodological considerations

While the analysis was primarily based on data from EPITT, supplementary information from PRAC meetings was obtained from publicly available EMA meeting minutes. Data on older safety signals assessed by the PhVWP were not publicly accessible and were instead provided by the Dutch MEB.

To evaluate the completeness of teratological signals prior to the establishment of EPITT, a separate analysis was performed in which a random sample of approximately 20% of all safety signals introduced between 2002 and 2012 was cross‐checked against PhVWP meeting minutes. This analysis confirmed that all included safety signals were indeed present in the EPITT database.

Although this study provided insights into the timing of key events, the wide error bars in the boxplots indicate low precision. This is likely due to the relatively small number of included safety signals and the variability inherent to each medicinal product. Since not all event dates could be retrieved, some boxplots were based on a limited subset of cases (see Table 1 and Figure 1).

The receipt of the first ICSR merely indicates when an association was initially observed and reported, rather than when a safety signal could have been identified. It has been recommended that at least three well‐documented cases are needed to identify an early signal. ⁴⁸ However, the broad definition of some conditions—such as ‘cardiac malformations’ or ‘impaired cognitive development’—made determining the date of the first and subsequent ICSRs in VigiBase difficult. Moreover clinical details of the ICSRs were not always available.

Although we assessed the origin of the teratological safety signals, regulatory decisions are not based solely on this information; a variety of other data are also considered to reach an informed decision. For example, alongside scientific studies, information from ICSRs in the EudraVigilance database is typically taken into account, even if it did not directly trigger the signal. ²²

Finally, it should be noted that the search for publications in Dutch medical and pharmaceutical journals and TIS information covered literature and CM from 2000 onwards. As a result, publications and TIS information related to some safety signals at the beginning of the study period may not have been captured, which could have introduced bias.

4.3. Strengths and limitations

While previous research has characterized safety signals discussed by the PRAC, these studies were not specific to pregnancy. ³⁰ , ³¹ By narrowing the focus to pregnancy‐related cases, this study provides information on the timing and characteristics of regulatory decisions on exposure to medicinal products during pregnancy. It also demonstrates how information was shared by the Dutch TIS and communicated in Dutch professional publications, regardless of the regulatory decision. Finally, the study covers a substantial timespan of two decades, allowing for a detailed analysis of the timing of regulatory actions related to safety signals.

A limitation of this study is that the actual impact on clinical practice could not be assessed. For confirmed teratogens, prescribing practices may be regularly reviewed where Risk minimization measures such as a Pregnancy Prevention Programmes are implemented and their impact assessed. ⁴⁹ A relevant example is the study by Abtahi et al., which demonstrated that risk minimization measures for valproate use in several European countries resulted in a reduction in exposure to the drug during pregnancy. ⁵⁰

Another limitation is that this study focuses on drug safety signals reviewed by the EMA and is therefore relevant primarily to European countries. Furthermore, implementation in clinical practice was examined only within the Netherlands, whereas practices in other countries may have differed.

Finally, identifying the first ICSR in VigiBase proved difficult for several signals due to involvement of multiple drugs (29%, n = 7) and nonspecific event reporting.

4.4. Implications for pharmacovigilance and research in pregnancy

Improving the quality and consistency of reporting pregnancy PV data could facilitate earlier detection of potential safety signals and support the timely initiation of studies to substantiate these findings. Several steps have already been taken in this direction. Notably, a reference framework of core data elements has been developed to standardize primary‐source pregnancy PV data collection, thereby enhancing harmonization and evidence synthesis and enabling more rapid generation of high‐quality evidence on medication safety in pregnancy. ⁵¹ Nevertheless, data quality may still vary depending on the reported event, and sufficiently detailed, high‐quality reporting remains essential for accurate assessment. ⁵² Increasing awareness of reporting requirements is therefore crucial. In this context, the recently published EMA Guideline on Good Pharmacovigilance Practices (GVP) for pregnant and breastfeeding women, including children exposed in utero or via breast milk, underscores the importance of targeted and comprehensive data collection and provides guidance for a more systematic and robust approach. ⁵³ Further progress may be achieved by exploring the inclusion of pregnant and breastfeeding women in clinical research. To this end, the World Health Organization has established a global task force aimed at promoting their timely and ethical inclusion in clinical trials by 2030. ⁵⁴ Beyond the regulatory solutions outlined above, additional efficiency can be achieved through the timely communication of safety signals and the implementation of appropriate risk minimization measures. International networks of TIS provide an important foundation for translating signals from research and clinical practice into actionable recommendations for healthcare professionals. ⁵⁵ , ⁵⁶ Integrating this knowledge into clinical decision support systems used in both primary and secondary care can support healthcare providers in pragmatically applying this information in daily practice. This is particularly relevant for ensuring safe prescribing in patients with specific characteristics, such as comorbidities and pregnancy. ⁴²

Our study focused on safety signals addressed by the EMA. Future research could also consider signals that did not lead to formal regulatory action but may still have influenced clinical decision‐making. Equally valuable would be the investigation of ‘signals of safety’ or favourable outcomes. In the absence of teratogenic concerns, accumulating evidence can ultimately support the safe use of specific medicines during pregnancy. This can at least prevent unnecessary concern among pregnant women about the medication used for their (chronic) condition, but it can also prevent unnecessary discontinuation of medication, and even termination of wanted pregnancies due to unfounded concerns.

5. CONCLUSION

The time between market authorization and regulatory decision regarding teratological safety signals is notably longer compared to non‐pregnancy‐related safety signals. On average, these signals emerge decades post‐authorization, with a median interval of 23.9 years (range 3.2–59.3). This extended timeline can be attributed to several factors, including the slower availability of safety data post‐market and the fact that these signals are predominantly identified through case reports and studies from scientific literature rather than spontaneous reporting systems. Consequently, the process of reaching a well‐supported regulatory decision is more protracted.

In addition to the regulatory authorities' dissemination of safety information, healthcare professionals often receive updates through scientific and professional publications next to the formal regulatory process. The ongoing role of the Dutch TIS in providing frequent updates further highlights the importance of timely and accurate information‐sharing in safeguarding maternal and fetal health.

These findings underline the need for continuous monitoring and data‐sharing strategies to accelerate the identification and evaluation and exchange of teratological risks and regulatory decisions, ultimately enhancing the safety of medicinal products for pregnant women.

AUTHOR CONTRIBUTIONS

Yrea R. J. van Rijt‐Weetink, Rosanne Huisman and Eugène P. van Puijenbroek conceptualized the study and drafted the manuscript. J‐P. van Wieringen and H.J.M.J Crijns collected and analysed the data in EPITT; Yrea R. J. van Rijt‐Weetink and Rosanne Huisman collected and analysed the data in literature and VigiBase and combined these data with those from EPITT; Rosanne Huisman, Sander D. Borgsteede and Eugène P. van Puijenbroek collected and analysed the data in Commentaren Medicatiebewaking. All authors reviewed and contributed to the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors report no conflict of interest.

van Rijt‐Weetink YRJ, Huisman RE, van Wieringen J‐P, et al. Characteristics and temporal trends of EU teratological safety signals and information dissemination to healthcare professionals in the Netherlands—A contribution of the ConcePTION project. Br J Clin Pharmacol. 2026;92(10):3365‐3374. doi: 10.1002/bcp.70615

DATA AVAILABILITY STATEMENT

The data that support the signals identified from European Pharmacovigilance Issues Tracking Tool (EPITT) and used in this study are only partially publicly available. For this reason, only aggregated data are presented. Access to the pseudonymised data is available from the authors on reasonable request.