Scoping review of biosimilar disease-modifying antirheumatic drugs in pregnancy: evidence gaps and proposed outcome reporting framework

Living with autoimmune or autoinflammatory diseases, such as rheumatic diseases (e.g., rheumatoid arthritis, ankylosing spondylitis) and inflammatory bowel disease (IBD), can profoundly impact every aspect of one’s daily life. Commonly characterized by chronic pain, physical limitations and fatigue [1], many of these conditions often strike at the childbearing age [1, 2]. Furthermore, conditions such as rheumatoid arthritis and systemic lupus erythematosus, affect more females than males [1, 3]. Given this confluence, there are particular implications with pregnancy, as individuals with active inflammatory disease are at higher risk of adverse maternal (i.e., miscarriage, gestational diabetes) and fetal outcomes (i.e., low birth weight, congenital anomalies) [4, 5]. As such, the maintenance of low disease activity throughout pregnancy is essential with the use of disease-modifying antirheumatic drugs (DMARDs) to effectively reduce disease burden, slow disease progression and alleviate flares [6].

Since their first regulatory approval in 1998, biologic (b) DMARDs have revolutionized the treatment armamentarium for a wide range of autoinflammatory and autoimmune diseases [7]. Biologics are powerful, intricate molecules derived from living organisms (e.g., animals, bacteria, yeast) or their components (e.g., proteins or cells) [8]. These agents are meticulously designed to target specific immune system pathways or mechanisms to suppress inflammation, alleviate painful symptoms and delay progression of autoimmune diseases [6]. However, due to the complex manufacturing process and high precision required for adequate quality control, these agents are very costly to access, with each drug costing on average $10,000 to $30,000 United States (US) dollars per person annually [9]. Similarly, in 2018, tumor necrosis factor inhibitors (TNFis), the most common bDMARD class, accounted for the highest proportion (8.3%) of publicly funded medication expenditures of all drug classes in Canada [10]. To combat the high cost of biologics and improve access to care, biosimilars were developed in the recent decade, defined by the US Food and Drug Administration (FDA) as highly similar drugs that have “no clinically meaningful differences in terms of safety, purity, and potency (i.e., safety and effectiveness) from…an existing FDA-approved biologic, called a reference product”[11] or bio-originator. The first FDA-approved biosimilar DMARD, Inflectra (infliximab-dyyb), a member of the TNFi class, was introduced in April of 2016 as a biosimilar to the reference product Remicade (infliximab) and is indicated for the treatment of autoimmune conditions such as rheumatoid arthritis and IBD [12].

Given the transformative role of biologics in managing autoimmune diseases [13], there has been considerable interest in the perinatal impacts of these drugs, especially as improved disease control plays an important role in allowing more patients to consider pregnancy [14]. Over the past two decades, a substantial body of research has evaluated perinatal exposure to bDMARDs and its impact on outcomes, leading to a number of syntheses [15‐18]. However, the majority of syntheses on bDMARD evidence has primarily focused on TNFis, particularly the bio-originators of TNFis, rather than biosimilars of TNFis or those of other bDMARD classes such as B-cell inhibitors (i.e., rituximab). In 2020, Tsao et al. published a systematic review and meta-analysis evaluating the maternal and neonatal outcomes of women exposed to biologics, primarily TNF inhibitors (TNFis), before and during pregnancy. However, their search did not include biosimilars or biosimilars of other non-TNF biologic DMARDs [15]. Similarly, systematic reviews and meta-analyses on the safety of biologics during pregnancy by Komaki et al. (2017) and Nielsen et al. (2021) focused exclusively on bio-originator TNFis and did not incorporate search terms for any biosimilar DMARDs [16, 17]. More recently, O’Byrne et al. (2022) conducted a systematic review and meta-analysis on fetal and maternal outcomes related to biologic exposure during conception and pregnancy. However, their review also lacked a comprehensive search for approved biosimilar DMARD brand names, and no terms related to “biosimilar” were included in the search strategy [18].

Although biosimilars are considered highly similar to their bio-originator counterpart, Health Canada declares that the approval of a biosimilar does not imply equivalence [19]. Yet, data on the safety and risks of biosimilar DMARDs during pregnancy remains limited, as shown by the absence of pregnancy compatibility recommendations in the recent American College of Rheumatology (ACR) and the European Alliance of Associations for Rheumatology (EULAR) guidelines in 2020 and 2024 [20, 21]. In daily rheumatologic practice, this gap forces providers to rely on extrapolating existing bio-originator data when making clinical decisions about biosimilar DMARD use around pregnancy. As a result, rheumatologists may face uncertainty when counseling patients, offering family planning recommendations, or making decisions around treatment continuation or switching. This need for a clearer map of current perinatal evidence is becoming increasingly urgent, particularly as many jurisdictions, including Canada and parts of Europe, have already implemented mandatory biosimilar switching initiatives since 2019 to reduce healthcare costs [22, 23]. To address this, we conducted a scoping review to achieve our broad objective of mapping and synthesizing current evidence on the reproductive health impact of biosimilar DMARDs (Supplementary Table 1) on fetal/neonatal, maternal and paternal outcomes (e.g., fertility[24], disease activity[25]), in individuals living with chronic autoimmune diseases (i.e., rheumatoid arthritis, IBD, psoriasis). Mapping the perinatal literature landscape on biosimilar DMARDs is crucial to inform clinical decision-making and reveal gaps in literature to shape future research directions.

To our knowledge, this scoping review is the first map of literature to date on biosimilar DMARDs and pregnancy outcomes. With only 6 included studies, 5 of which examined biosimilars of TNFis, there is a clear limited body of evidence on the perinatal impact of biosimilar DMARDs. Critically, the lack of analytic studies precludes knowledge of the associations between biosimilar DMARD exposure and pregnancy outcomes, hindering the ability to guide patients and providers in making informed decisions about pregnancy and family planning. There is also a complete absence of data examining paternal biosimilar DMARD exposure.

As all 6 included studies examined biosimilars of the TNFi bDMARD class, our findings revealed an overall lack of pregnancy data on biosimilar DMARDs, particularly biosimilars of non-TNFi bDMARD classes. As the first biosimilar DMARD class approved by the FDA in 2016 [12], it is expected that the majority of current evidence may be focused on biosimilars of TNFis. Specifically, we observed a predominance of infliximab biosimilars in the literature, which likely reflects regulatory and market factors—namely, that infliximab was the first biosimilar approved by the FDA in 2016 [12]—which led to earlier market entry, broader clinical uptake and greater research efforts, rather than a reflection of scientific breadth. However, there is also a need for more perinatal data on the safety and risks of biosimilars of non-TNFi bDMARDs, as it is recommended for patients whose autoimmune disease activity is suboptimally controlled by TNFis to be switched to bDMARDs of another class [20, 21]. Notably, non-TNFi bDMARD agents such as rituximab—which also has approved biosimilars—are commonly used in women of childbearing age and, as Immunoglobulin G-based monoclonal antibodies, can cross the placenta [38, 39]. The overall scarcity of perinatal data on biosimilar DMARDs is concerning, given the substantial and growing use of biologics in the pregnant population. In 2009, Zelinkova et al. reported that 19% of 61 IBD patients with active plans for reproduction were taking TNFi bDMARDs [14]. Similarly, a population-based cohort study by Tsao et al. found a statistically significant rise in biologic use among 6,218 pregnant women (8,431 pregnancies) with autoimmune diseases from 2002 to 2012 (p < 0.001) [40]. However, the high cost of reference biologics has driven the sharp increase in biosimilar DMARD utilization over the past decade, including among individuals of childbearing age. Multiple mandatory biosimilar switching initiatives underway in many jurisdictions such as those in Canada and Europe [23, 41] are rapidly shifting the trajectory of biologic DMARD use toward biosimilars. This emphasizes the rising need to investigate the perinatal impact of biosimilar DMARDs, especially as more biosimilar DMARDs continue to be introduced to the market.

In addition to the overall limited evidence on biosimilar DMARDs, it is crucial to note that all included observational studies were primarily descriptive in nature, which reveals a lack of analytical research on biosimilar DMARDs. While 5 out of 6 included studies were descriptive, there are inherent limitations in case reports, case series and cross-sectional study designs—mainly the fact that they do not allow us to make causal inferences. Instead, they provide descriptive observations of a single case or a series of cases, without the statistical analyses required to evaluate causal relationships and properly adjust for confounders. Therefore, in terms of applicability to clinical practice and treatment guidelines, the strength of evidence they provide is limited. Analytical studies, on the other hand, are crucial for causal inference, which provide guidance for clinical decision-making. The singular cohort study included in our review compared biosimilar infliximab to their biologic originator but only reported the rates of each pregnancy outcome and had small sample sizes. Therefore, the calculated odds ratios may be confounded by factors such as disease activity and concomitant medications. As a result, no causal conclusions can be drawn regarding differences in pregnancy outcomes between biologic originators and biosimilars. Therefore, our scoping review reveals that the current perinatal evidence base is insufficient for drawing causal inferences, and future studies are required to assess associations between biosimilar DMARD exposure and pregnancy outcomes or evaluate differences in pregnancy outcomes between biologic originator DMARDs and their biosimilar counterparts. Due to this gap in analytical literature, there is an absence of pregnancy compatibility recommendations for biosimilar DMARDs in the ACR and EULAR treatment guidelines [20, 21]. Several factors may contribute to this gap, such as the relatively recent introduction of biosimilars and the exclusion of pregnant women clinical trials due to ethical and logistical challenges [42].

While biosimilars must show “no clinically meaningful differences in terms of safety, purity, and potency (i.e., safety and effectiveness) from…an existing FDA-approved biologic”[11, 43], Health Canada declares that the approval of a biosimilar does not imply equivalence [19]. Their complex manufacturing processes are difficult to replicate precisely, raising theoretical concerns about potential deviations in efficacy and safety [44‐46]. Even subtle differences in production may result in structural variations or post-translational modifications—such as glycosylation—that could, in turn, alter biological function or immunogenicity [47, 48]. These considerations are further supported by the FDA’s observation that pre-licensing trials may be underpowered to detect rare or delayed immune responses [49]. This is critical in the context of pregnancy—where immune tolerance and placental drug transfer mechanisms are highly dynamic. Therefore, due to the lack of perinatal evidence on biosimilars, providers must rely on extrapolating existing data from bio-originator DMARDs to make clinical decisions on biosimilar DMARDs in pregnant patients. This may contribute to reluctance in prescribing or using biosimilar DMARDs during pregnancy, as shown by Zelinkova et al.’s study where 30% of IBD patients who consulted a physician prior to conception were advised to undergo medication changes due to pregnancy or an active desire to reproduce, largely owing to limited information on peri-conceptional safety [14]. This is concerning, as changes to a stable DMARD regimen, particularly around the perinatal period, may lead to loss of inflammatory disease control, which is associated with adverse outcomes in both the mother and the baby [50]. Therefore, analytical studies and the leveraging of observational data are crucial to addressing this gap, such as through robust head-to-head comparisons evaluating the perinatal impact of biologic originator DMARDs versus their biosimilar counterparts or comparing biosimilars to no exposure. To support these investigations, there is value in national pregnancy registries and large multicenter observational studies, which would allow for larger sample sizes and confounder adjustment—ultimately generating more robust and generalizable findings.

Our review also evaluated methodologies used for reporting data in the included studies, where we found clear reporting of sample size units. However, significant heterogeneity was identified in the reporting of outcomes, exposure measures, and maternal disease activity, highlighting inconsistencies in the literature. In contrast to the findings from our recent scoping reviews on targeted synthetic DMARDs [29] and non-TNFi bDMARDs [51], we did not identify inconsistencies in the reporting of sample size units in the present review until our updated search in June 2025. In the cross-sectional studies identified from our original search, sample size units (i.e., number of women, pregnancies and newborns) were clearly stated in the original text which enabled ease of interpretation and synthesis. However, in the cohort study identified from our updated search, proportions appeared to be reported in terms of number of “women”, but data on the number of pregnancies and corresponding newborns were unavailable to be extracted. As one mother may have multiple pregnancies that may result in multiple newborns, it is important in perinatal research to comprehensively report the number of mothers, pregnancies and newborns to optimize interpretation and comparability across studies. We also identified inconsistencies in the reporting of biosimilar drug exposure timing and duration, which are critical to report in perinatal research given the varying impact of insults to the fetus at each trimester of pregnancy [52]. There was also incomplete reporting of biosimilar drug names, where some studies did report the specific biosimilar drug names (e.g., Infliximab-dyyb, CT-P13) and others only reported the generic drug names. To address this, it is crucial to report details on the biosimilar drug name (e.g., brand name or FDA’s suffix-based naming system for biosimilars [53]), exposure measures of each drug in terms of timing (i.e., initiated before or during pregnancy), duration (i.e., number of gestational weeks exposed), and, if applicable, discontinuation (i.e., weeks’ gestation of last dose). We also identified substantial variability in outcome reporting which constitutes a substantial barrier to drawing meaningful conclusions and comparisons. For example, Scott et al. stated no congenital anomalies were identified, but mentioned the presence of ‘floppy larynx’ or laryngomalacia in one of the newborns, which is generally considered a congenital anomaly in literature [54]. This highlights the subjective nature of reporting pregnancy outcomes in the literature and the importance of reader discernment along with evidence-based reporting. In addition, we found incomplete reporting of maternal outcomes across the studies, particularly on disease activity. Comprehensive data on maternal autoimmune disease activity (e.g., active disease at conception, in clinical remission at various points of pregnancy) is important to report as active disease itself is associated with adverse pregnancy outcomes [4, 5]. Lastly, we identified an absence of relevant maternal comorbidity outcome reporting such as diabetes mellitus [55] and gestational hypertension [56, 57], which are important to examine as they have been linked to increased risk of adverse neonatal outcomes such as congenital anomaly. To mitigate variations in outcome reporting and enhance comparability of findings, we encourage future researchers to adopt the use of our flexible Reproductive Health Outcomes Reporting Framework (Fig. 2) as a guide for consistent and comprehensive reporting.

To our knowledge, our Reproductive Health Outcomes Reporting Framework (Fig. 2) is the first comprehensive map and visual synthesis of the reproductive health outcomes reported in current literature on biosimilar DMARDs. Outcomes listed in the colored boxes of this framework provide a clear illustration of all reported outcomes in the literature, whereas the grey boxes represent current gaps in literature that were not reported in our included studies. For example, we discovered no reports on intrauterine fetal outcomes such as fetal death, as well as neonatal outcomes such as stillbirth. Our framework highlights the lack of data on these critical outcomes, which continues to hinder clinicians’ ability to make evidence-based decisions. This framework serves as a guide to not only standardize outcome reporting but also inform future directions of perinatal biosimilar DMARD research to improve our understanding of these emerging agents.

From a clinical standpoint, the limited perinatal evidence on biosimilar DMARDs may pose challenges for clinicians, particularly during patient counseling and clinical decision-making. As more jurisdictions implement biosimilar switching mandates, clinicians must remain vigilant by staying informed on emerging evidence and ensuring patient safety concerns are addressed. In the absence of robust analytical evidence, shared decision-making becomes essential while ensuring patients understand the importance of maintaining disease control. Patients and providers are recommended to engage in early, individualized pre-pregnancy planning ideally 1–2 years prior to conception [58], as these consultations result in fewer flares and lower disease activity in the first trimester—all indicators of positive pregnancy outcomes[59]. Lastly, supporting pregnancy registries and post-market surveillance are vital in expanding the perinatal evidence base on biosimilar DMARDs and ultimately enabling patients and clinicians to make informed treatment decisions.

Overall, our scoping review offers a novel contribution to the literature as the first to: (1) comprehensively synthesize and map perinatal evidence gaps of biosimilar DMARDs, (2) propose a Reproductive Health Outcomes Reporting Framework to standardize reporting, (3) evaluate methodological reporting gaps across studies, and (4) discuss clinical implications of the lack of perinatal biosimilar DMARD evidence. To our knowledge, no previous reviews have evaluated perinatal exposure to biosimilar DMARDs; instead, existing reviews have focused on their bio-originators, namely TNFis [15‐18]. For example, systematic reviews and meta-analyses on the safety of biologics in pregnancy by Komaki et al. (2017), Tsao et al. (2020) and Nielsen et al. (2021) did not search for biosimilars and instead primarily examined bDMARDs, particularly TNFis [15‐17]. Similarly, O’Byrne et al.’s 2022 systematic review and meta-analysis on fetal and maternal outcomes related to biologic exposure around pregnancy also did not include search terms particularly related to “biosimilar” or approved biosimilar DMARD brand names [18]. Furthermore, based on our analysis of gaps in methodological reporting across literature, our review is the first to provide concrete, actionable recommendations for improving comparability in future perinatal studies, such as consistently reporting sample size units (i.e., mothers, pregnancies and babies) and adopting our Reproductive Health Outcomes Reporting Framework.

Strengths and limitations of our scoping review warrant discussion. Although only search terms for DMARDs approved in our jurisdiction (i.e., by Health Canada) were included in the search strategy, a research librarian was involved in the development and application of the search strategy, which consisted of a comprehensive list of relevant search terms to optimize identification of all relevant studies. The study screening process was strengthened by consistent communication amongst the team regarding ambiguities, which ensured standardized inclusion of studies. Although our search identified studies in French, German, and Korean, owing to available time and resources, only publications available in English full texts were included. We acknowledge this as a potential source of publication bias, specifically language bias, and urge a more inclusive approach in future investigations to enhance representation of the literature. Lastly, the lack of perinatal data on non-TNFi biosimilars further limits the generalizability of our findings.

As biosimilar DMARDs continue to be integrated into healthcare systems worldwide, our scoping review provides a map for visualizing current evidence gaps to guide future perinatal research as new evidence emerges. Given the clear cost-saving benefits of biosimilars, it is crucial to leverage observational studies to generate robust analytical evidence, providing patients with greater assurance and insight into their safety and risks during pregnancy. Therefore, until more data become available, careful monitoring and individualized treatment decisions should guide the use of biosimilar DMARDs in patients with autoimmune disease who are pregnant or planning pregnancy.