Dengue is one of the most widespread mosquito-borne diseases, with estimates suggesting that between 284 and 528 million people are infected annually (Harapan et al., 2020). The World Health Organization (WHO) has classified dengue as one of the most significant viral mosquito-borne diseases, affecting over half of the global population, especially those residing in high-risk areas (Kularatne and Dalugama, 2022). The vectors responsible for dengue transmission are highly prevalent in urban environments, especially in tropical and subtropical regions where the disease is endemic (Liu et al., 2023). Dengue is also an increasing concern among international travelers and among migrants arriving from highly endemic countries (Riddell and Babiker, 2017; Masyeni et al., 2018; Halstead and Wilder-Smith, 2019).

Clinical manifestations range from mild febrile illness to severe complications such as dengue hemorrhagic fever (DHF) and dengue shock syndrome (DSS), conditions that may lead to fatal outcomes (Moraes et al., 2013). Disease severity is influenced by several factors, including viral serotype (DENV-1 to DENV-4), host immune status, and other determinants. The lack of specific antiviral therapies emphasizes the urgent need for effective preventive measures. Although vector control remains the primary approach to reducing transmission, it has proven insufficient to contain outbreaks, underscoring the need for an effective vaccine (Kant et al., 2025).

CYD-TDV, Sanofi Pasteur (Lyon, France), was the first dengue vaccine to be approved worldwide. It received initial authorization in 2015 in the Philippines and, in 2019, by the U.S. FDA, although it has not yet been approved by the European Medicines Agency (EMA) in Europe. CYD-TDV is a recombinant, live - attenuated, tetravalent dengue vaccine constructed using the yellow fever 17D vaccine strain (YFV-17D) backbone with chimeric prM/E RNA sequences derived from the four dengue virus serotypes (Tonto et al., 2025). The vaccine is administered in a three-dose schedule (0, 6, and 12 months) to individuals aged 9–45 years who are already seropositive for dengue (European Medicines Agency, 2025a). In seronegative individuals, vaccine-induced antibodies may promote antibody-dependent enhancement (ADE), increasing the risk of severe dengue (Yang et al., 2018). Its efficacy has been demonstrated in several endemic countries, including the phase 3 CYD15 trial (NCT01374516), which reported a vaccine efficacy (VE) of 60.8% against virologically confirmed symptomatic dengue, with higher protection observed against hospitalized and severe forms (Villar et al., 2015). By contrast, TAK-003, Takeda Vaccines (Cambridge, MA, United States), is a more recently developed live attenuated tetravalent vaccine designed using a DENV-2 backbone with recombinant components from the other serotypes, conferring immunity against all four dengue virus serotypes (Wilder-Smith, 2024). TAK-003 induces robust humoral and T-cell–mediated immune responses (Sáez-Llorens et al., 2025). A 5-year clinical trial involving over 20,000 children and adolescents demonstrated its effectiveness in preventing severe dengue and reducing hospitalizations (Tricou et al., 2024). The vaccine was first approved in Indonesia for individuals aged 6–45 years. It subsequently received marketing authorization in the European Union in 2023 for individuals aged 4 years and over, regardless of previous dengue infection.

Although both vaccines have demonstrated efficacy and favorable safety profiles in randomized clinical trials (Reynales et al., 2020; Tricou et al., 2020; 2023; Biswal et al., 2021; de Moraes et al., 2022; Sirivichayakul et al., 2022; Luhm et al., 2023), they differ substantially in their indications and target populations. CYD-TDV is restricted to seropositive individuals due to the risk of ADE in seronegative recipients, whereas TAK-003 is indicated for both seropositive and seronegative individuals and therefore has broader potential use, including among travelers from non-endemic regions. These differences make a direct comparison of their respective safety profiles particularly relevant.

Furthermore, despite robust evidence from clinical trials, there are still important pharmacovigilance considerations. Live attenuated vaccines require continuous monitoring to detect rare or unexpected adverse events, and real-world post-marketing experience for both products is still limited. The use of dengue vaccines in populations not represented in clinical trials may reveal different safety patterns. Currently, there is a lack of post-marketing safety data from real-world conditions for both vaccines, emphasizing the need for studies based on spontaneous reporting systems.

Therefore, the aim of this study was to describe adverse drug reactions (ADRs) associated with both dengue vaccines as reported in a European pharmacovigilance database and, secondarily, to compare their respective safety profiles.

A total of 2,288 ICSRs reporting TAK-003 or CYD-TDV as suspected medicinal products were retrieved from the EV platform. Of these, 77.3% (n = 1,768) involved TAK-003, while 20.7% (n = 520) involved CYD-TDV.

The characteristics of ICSRs associated with TAK-003 (N = 1,768) and CYD-TDV (N = 520) were reported in Table 1. A significant difference in sex distribution was observed between the two vaccine groups (p < 0.01). TAK-003-related ICSRs showed a higher proportion of female cases (57.9%), whereas CYD-TDV-related reports demonstrated a more balanced sex distribution (male 41.2% and female 42.7%). Significant differences were also identified in age group distribution (p < 0.01). TAK-003-related ICSRs predominantly involved adult populations (≥18 years; 49.8%), while CYD-TDV-related reports more frequently concerned pediatric and adolescent populations, particularly individuals aged 12–17 years (32.3%).

The majority of ICSRs in both groups were classified as serious according to regulatory criteria; however, a significantly higher proportion of serious reports was observed for CYD-TDV compared with TAK-003 (81.1% vs. 76.4%; p = 0.04).

Marked differences were observed in reported clinical outcomes. Notably, fatal outcomes were significantly more frequently reported in CYD-TDV-related ICSRs compared with TAK-003-related reports (50.8% vs. 0.6%; p < 0.01). Conversely, TAK-003-related reports showed significantly higher proportions of outcomes classified as not recovered/not resolved, recovering/resolving, and recovered/resolved (all p < 0.01). No significant differences were observed for the outcome recovered with sequelae (p = 0.18). A substantial proportion of ICSRs in both groups had outcomes classified as unknown (23.6% for TAK-003% and 18.9% for CYD-TDV). Finally, the primary source qualification differed significantly between the two groups (p < 0.01). TAK-003-related ICSRs were predominantly reported by healthcare professionals (71.8%), whereas most CYD-TDV-related reports originated from non-healthcare professionals (54.4%) (Table 1).

The SOC level disproportionality analysis revealed relevant differences in the reporting profiles of TAK-003 and CYD-TDV. For the majority of SOCs, the ROR was below 1.0, indicating lower disproportional reporting for TAK-003 compared with CYD-TDV. Specifically, significant RORs with 95% confidence intervals entirely below unity were observed for infections and infestations (ROR 0.07; 95% CI 0.056–0.089), gastrointestinal disorders (ROR 0.22; 95% CI 0.18–0.27), general disorders and administration site conditions (ROR 0.29; 95% CI 0.23–0.36), nervous system disorders (ROR 0.42; 95% CI 0.34–0.51), cardiac disorders (ROR 0.35; 95% CI 0.23–0.51), and eye disorders (ROR 0.48; 95% CI 0.36–0.64).

Similarly, reduced disproportionality for TAK-003 was observed across several clinically relevant SOCs, including hepatobiliary (ROR 0.09; 95% CI 0.04–0.19), metabolism and nutrition (ROR 0.09; 95% CI 0.06–0.14), renal and urinary (ROR 0.12; 95% CI 0.06–0.23), and vascular disorders (ROR 0.13; 95% CI 0.10–0.17).

Conversely, a signal of disproportionate reporting in favor of TAK-003 was identified for skin and subcutaneous tissue disorders, which showed a statistically significant ROR above unity (ROR 1.98; 95% CI 1.59–2.48), suggesting a higher relative frequency of reporting compared with CYD-TDV. For musculoskeletal and connective tissue disorders, the ROR was close to unity, and the confidence interval included 1 (ROR 1.04; 95% CI 0.83–1.30), indicating no meaningful difference in reporting frequency between the two vaccines for this SOC (Table 2).

A total of 10,901 suspected adverse reactions were observed in all ICSRs, corresponding to a mean of 4.8 reactions per case. Notably, 6,106 suspected reactions were observed in the TAK-003 group, averaging at 3.5 reactions per ICSR, compared to 4,795 suspected reactions in the CYD-TDV group, with a mean of 9.2 reactions per ICSR.

The PT disproportionality analysis revealed a consistent pattern of disproportionate reporting in favor of TAK-003 compared to CYD-TDV for cutaneous and hypersensitivity-related events. The strongest association was identified for injection site erythema (ROR 10.44; 95% CI 3.30–33.06), followed by signals for pruritus (ROR 4.81; 95% CI 2.43–9.50), angioedema (ROR 3.74; 95% CI 1.62–8.65), urticaria (ROR 3.30; 95% CI 1.65–6.57), and rash-related PTs, including cutaneous rash (ROR 2.36; 95% CI 1.13–4.17) and rash pruritic (ROR 3.06; 95% CI 1.09–8.59). Disproportionate reporting was also observed for common reactogenicity-related PTs, including fatigue (ROR 2.48; 95% CI 1.63–3.77), myalgia (ROR 1.77; 95% CI 1.26–2.48), and arthralgia (ROR 1.62; 95% CI 1.12–2.34). Several additional PTs showed RORs above unity (e.g., influenza-like illness and limb discomfort), although some estimates were based on small numbers and wide confidence intervals. Conversely, multiple PTs indicative of more severe clinical presentations were reported less frequently in association with TAK-003 than with CYD-TDV, including dengue fever (ROR 0.05; 95% CI 0.03–0.06), and other infection- and complication-related PTs (e.g., pneumonia, sepsis), with confidence intervals consistently below 1 (Table 3).

The analysis of ICSRs identified several PTs that were reported exclusively in association with CYD-TDV and were not observed in ICSR related to TAK-003. These PTs primarily comprised serious neurological, hemorrhagic, and multi-system conditions. Among the most frequently reported were cerebral and internal hemorrhage, noninfective encephalitis, brain oedema, and pulmonary or visceral hemorrhage. Other clinically relevant PTs included multiple organ dysfunction syndrome, coma, septic shock, disseminated intravascular coagulation, and myocardial hemorrhage. While the presence of these PTs does not imply a causal relationship, their exclusive reporting in CYD-TDV ICSR suggests that they warrant further monitoring and careful evaluation in post-marketing safety assessments.

PTs reported exclusively in TAK-003 ICSR were also identified and are presented in the Supplementary Material. However, this analysis focused on PTs uniquely associated with CYD-TDV due to their frequency and potential clinical relevance.

The complete list of PTs reported exclusively for CYD-TDV or for TAK-003 is provided in the Supplementary Material (Supplementary Table S1).

In this pharmacovigilance study based on EV data, we compared the post-marketing safety profiles of the two currently available dengue vaccines, TAK-003 and CYD-TDV. Overall, our findings highlight substantial differences in reporting patterns, seriousness, clinical outcomes, and adverse reaction profiles between the two vaccines, likely reflecting not only intrinsic differences in vaccine design but also distinct target populations, indications, and historical contexts of use.

TAK-003 accounted for more than three-quarters of all dengue vaccine–related ICSRs retrieved from the EV database. This predominance is plausibly explained by its more recent authorization in the European Union and its broader indication irrespective of baseline dengue serostatus, in contrast to CYD-TDV, whose use is restricted to seropositive individuals due to the risk of ADE in seronegative recipients (Sridhar et al., 2018; Author anonymous, 2019). Differences in age distribution, baseline dengue serostatus, and target populations are likely to play a central role in shaping the observed reporting patterns between CYD-TDV and TAK-003. Thus, CYD-TDV-related ICSRs predominantly concerned pediatric and adolescent populations, reflecting its historical use in dengue-endemic settings and its restriction to seropositive individual (Hadinegoro et al., 2015; Godói et al., 2017). In contrast, TAK-003-related reports mainly involved adults and individuals vaccinated irrespective of baseline serostatus, consistent with its broader indication and more recent use, including in non-endemic or travel-related contexts. These substantial differences in target populations and vaccination strategies should be taken into account when interpreting comparative pharmacovigilance findings.

A higher proportion of female cases was also observed among ICSRs related to TAK-003, indicating sex-related differences. Such findings are commonly reported in spontaneous reporting systems and may reflect sex-related differences in immune response, healthcare-seeking behavior, or reporting practices rather than a true difference in vaccine-related risk (Klein and Flanagan, 2016).

In terms of severity, most ICSRs for both vaccines were classified as serious according to regulatory criteria; notably, a significantly higher proportion of serious reports was observed for CYD-TDV.

Importantly, results derived from spontaneous reporting systems such as EudraVigilance should be interpreted with caution, as they primarily reflect reporting behaviors rather than causal associations. In this context, the higher proportion of serious and fatal outcomes reported in CYD-TDV-related ICSRs should not be interpreted as evidence of increased vaccine-related risk. Instead, these findings are more plausibly explained by confounding by indication, differences in epidemiological settings, baseline disease severity, and reporting bias, all of which are well-recognized limitations of pharmacovigilance databases.

Moreover, CYD-TDV has been available on the market for longer and has been subject to enhanced surveillance due to concerns regarding severe dengue and ADE in specific populations (Hadinegoro et al., 2015; Wilder-Smith et al., 2019). Moreover, CYD-TDV has primarily been deployed in dengue-endemic areas characterized by high background morbidity and mortality, which may have influenced the clinical context and reporting patterns of serious outcomes observed in spontaneous reports. TAK-003 accounted for more than three-quarters of all dengue vaccine–related ICSRsThe analysis of clinical outcomes revealed marked differences between the two vaccines. Fatal outcomes were reported substantially more frequently in CYD-TDV-related ICSRs, whereas TAK-003-related reports more commonly described non-fatal outcomes, including recovered/resolved or recovering/resolving cases. In addition, long-term follow-up of CYD-TDV clinical trials has shown an increased risk of severe dengue in seronegative individuals, which led to revised recommendations and restricted use (Hadinegoro et al., 2015; Sridhar et al., 2018).

Disproportionality analyses at both SOC and PT levels further highlighted a distinct reporting profile for TAK-003. Most SOCs showed significantly lower reporting odds for TAK-003 compared with CYD-TDV, particularly for infections, nervous system and cardiovascular disorders, and dengue-related conditions. Conversely, TAK-003 showed higher disproportionality for skin and subcutaneous tissue disorders, including injection site erythema, pruritus, urticaria, angioedema, and rash-related terms. These findings are consistent with the reactogenicity profile observed in TAK-003 clinical trials, where local and mild systemic reactions were the most frequently reported adverse events (Biswal et al., 2019; Tricou et al., 2024; European Medicines Agency, 2025b).

At the PT level, increased reporting of fatigue, myalgia, and arthralgia was also observed for TAK-003. These events are well described in the Summary of Product Characteristics and in phase 2 and phase 3 trials of TAK-003 and are generally considered transient and self-limiting (Biswal et al., 2019; Tricou et al., 2024). A recent post-marketing signal of anaphylaxis associated with TAK-003 vaccination reported by Brazilian pharmacovigilance authorities further supports the importance of continued monitoring of hypersensitivity reactions, although regulatory measures were promptly implemented and no vaccine-related deaths were identified (Agência Nacional de Vigilância Sanitária. Brazil, 2024).

In addition, the comparison of vaccine-specific PTs showed that several PTs were exclusively reported in CYD-TDV-related ICSRs and were not observed in TAK-003 ICSRs. These PTs mainly included serious neurological, hemorrhagic, and multi-organ conditions, such as cerebral and internal hemorrhage, noninfective encephalitis, brain oedema, and multiple organ dysfunction syndrome. The exclusive reporting of serious neurological and haemorrhagic PTs in CYD-TDV-related ICSRs represents an important observation that requires careful contextualization. Dengue virus infection is well recognized to be associated with neurological involvement (e.g., encephalitis/encephalopathy and immune-mediated neuroinflammatory disorders) and haemorrhagic complications driven by thrombocytopenia, platelet dysfunction, and vasculopathy/endothelial dysfunction (Naderian et al., 2025). In this context, the occurrence of such events in spontaneous reports is likely influenced by the clinical severity of dengue disease and the epidemiological settings in which CYD-TDV has historically been used, rather than indicating a direct vaccine-related adverse effect. Therefore, while these PTs warrant continued post-marketing monitoring, their presence should be interpreted within the broader clinical and epidemiological background of dengue infection. Importantly, this pattern is likely influenced by differences in target populations, age distribution, baseline dengue serostatus, and epidemiological settings in which CYD-TDV has historically been used. These settings are often characterized by a higher baseline risk of severe dengue and related complications, as reported in systematic reviews and long-term safety follow-up studies (Godói et al., 2017; Forrat et al., 2021). In contrast, available clinical trial and meta-analytic evidence indicate that TAK-003 is predominantly associated with non-serious and reactogenic adverse events, with no consistent signals of severe outcomes identified to date (Flacco et al., 2024). The complete list of vaccine-specific PTs is provided in the Supplementary Material.

Several limitations inherent to spontaneous reporting systems must be acknowledged. These include under-reporting, differential reporting stimulated by media attention or regulatory actions, lack of exposure denominators, and incomplete clinical information (Hazell and Shakir, 2006; Pariente et al., 2010). Additionally, differences in reporting source qualification between TAK-003 and CYD-TDV may have influenced the type and severity of reported events.

Beyond epidemiological and population-related factors, recent evidence suggests that dengue virus infection is associated with profound immune modulation and metabolic reprogramming, involving alterations in glucose and lipid metabolism, mitochondrial function, and inflammatory signaling pathways. These immunometabolic changes have been shown to influence innate and adaptive immune responses, endothelial activation, and systemic inflammatory processes, all of which contribute to disease severity and multisystem involvement (Chermahini et al., 2025). While these mechanisms are primarily related to dengue virus pathophysiology rather than vaccine-induced effects, they provide a broader biological framework for interpreting severe clinical manifestations reported in pharmacovigilance databases. In this context, immunometabolic dysregulation may shape the clinical background in which adverse events are reported following dengue vaccination, particularly in high-risk populations, without implying a direct causal relationship with vaccination.

In conclusion, this real-world pharmacovigilance analysis indicates that TAK-003 is predominantly associated with non-serious, reactogenic, and hypersensitivity-related adverse reactions, whereas CYD-TDV reports more frequently involve serious and fatal outcomes, likely influenced by its restricted indication and historical context of use.

At SOCs level, TAK-003 demonstrated a lower rate of disproportional reporting across most clinically relevant SOCs. Meanwhile, a consistent signal emerged for skin and subcutaneous tissue disorders.

At PTs level, this signal was mainly driven by local and hypersensitivity-related reactions, including injection site reactions, pruritus, urticaria, angioedema, and rash-related terms, which were largely non-serious and self-limiting. In contrast, the exclusive reporting of certain PTs in CYD-TDV-related ICSRs should be interpreted in light of the differences in target populations, baseline risk profiles, and epidemiological settings discussed above, rather than as an indication of vaccine-related causality.

Overall, these findings are consistent with the available clinical and post-marketing evidence, supporting the favourable safety profile of TAK-003, while underscoring the need for ongoing post-marketing surveillance of both dengue vaccines (Biswal et al., 2019; Tricou et al., 2024; European Medicines Agency, 2025a).

One of the key strengths of this study is its utilization of the EV database, which is among the largest spontaneous reporting systems worldwide. This allowed for the evaluation of real-world safety data for dengue vaccines across multiple countries and healthcare settings. The broad coverage of the database enhances the generalizability of the findings beyond the controlled conditions of clinical trials. The comparative design represents an additional strength. By directly comparing TAK-003 and CYD-TDV, the analysis was able to identify differential reporting patterns that may not emerge from single-product evaluations, providing a more informative context for signal interpretation. The combined use of descriptive statistics and disproportionality analyses at both SOC and PT levels further strengthened the assessment, enabling identification of coherent and biologically plausible safety patterns.

Finally, the consistency observed between post-marketing findings and ADRs reported in pre-authorization clinical trials reinforces the robustness and relevance of the results for regulatory pharmacovigilance activities.

This study is subject to the inherent limitations of spontaneous reporting systems, including under-reporting, reporting biases, and the absence of exposure denominators, which preclude reliable estimation of incidence rates and causal inference. Differences in reporting behavior related to time since market authorization, media attention, and regulatory actions may have influenced the observed reporting patterns. Moreover, EV is a regulatory pharmacovigilance database that includes reports originating from both European Economic Area and non-European countries and does not systematically capture information on the geographical location of dengue virus acquisition or the total number of vaccinated individuals. Consequently, regional epidemiological patterns and disease incidence within Europe cannot be reliably assessed using this data source.

In addition, the completeness and quality of ICSRs are variable, with frequent missing information on patient characteristics, clinical history, comorbidities, and outcomes, limiting the ability to adjust for potential confounding factors. Finally, differences indications, target populations, and vaccination contexts between TAK-003 and CYD-TDV should be taken into account when interpreting comparative safety findings.