Pausing a vaccine is not a failure, but a call for continuous dengue surveillance

• Maíra Aguiar, group leader at BCAM and Ikerbasque Research Associate Professor, publishes an analysis in the prestigious scientific journal Science, calling for active and long-term pharmacovigilance following the recent preventive suspension of the Butantan-DV vaccine in Brazil

Maíra Aguiar, group leader at BCAM and Ikerbasque Associate Professor, publishes a commentary in Science calling for active, transparent, and long-term surveillance following the preventive suspension of the Butantan-DV dengue vaccine in Brazil

The detection of a potential safety signal that leads health authorities to temporarily pause a vaccine should not automatically be interpreted as a failure of vaccination policy. Rather, it demonstrates why robust pharmacovigilance and epidemiological surveillance are essential once vaccines move from clinical trials into large-scale population use.

This is the central message of a new publication in Science by Maíra Aguiar, Ikerbasque Associate Professor and leader of the Mathematical and Theoretical Biology group at the Basque Center for Applied Mathematics (BCAM). The article examines the recent decision by the Brazilian Ministry of Health to temporarily suspend administration of the single-dose Butantan-DV dengue vaccine following reports of 42 severe adverse events among approximately 500,000 vaccinated individuals.

Aguiar argues that the evaluation of dengue vaccines cannot end with regulatory approval or the completion of clinical trials. Once vaccination is introduced at population scale, safety and effectiveness must continue to be assessed under the epidemiological conditions in which the vaccine is actually being used.

The article highlights three priorities:

• Active pharmacovigilance: surveillance systems should actively search for potential safety signals rather than rely exclusively on spontaneous reporting.
• Transparent epidemiological data: timely access to information on vaccination, dengue infections, hospitalization, baseline serostatus where feasible, and circulating serotypes is essential for evaluating safety signals in context.
• Post-marketing studies capable of distinguishing causes: analyses should be designed to separate events potentially associated with vaccination from severe outcomes caused by subsequent natural dengue infection.

A long-standing challenge in dengue vaccination

Dengue vaccination is particularly complex because four antigenically distinct dengue virus serotypes circulate. Immune history matters: protection following a first dengue exposure can differ according to the serotype encountered subsequently, and under some immunological conditions pre-existing antibodies may contribute to increased disease severity through antibody-dependent enhancement.

“Safety, real-world effectiveness, and protection against each of the four serotypes need to be monitored throughout population rollout, not only during clinical development,” Aguiar emphasizes.

This question has been a longstanding focus of her research. In 2016, Aguiar and collaborators used mathematical modelling to investigate the population-level impact of Dengvaxia, showing that vaccine outcomes could differ substantially according to previous dengue exposure. That work, together with subsequent analyses of the risks of vaccinating dengue-seronegative individuals, contributed to the scientific discussion surrounding the reassessment of Dengvaxia vaccination strategies and WHO recommendations.

Aguiar also co-authored the 2016 commentary “The risks behind Dengvaxia recommendation” and continued to examine the implications of serostatus-dependent vaccine performance as longer-term safety evidence became available. These earlier studies anticipated a principle that remains central to the new Science publication: average vaccine efficacy can conceal important differences between individuals with different immune histories and between infections caused by different dengue serotypes.

The Dengvaxia experience is particularly relevant. Long-term follow-up showed an increased risk of hospitalization and severe dengue among some vaccinated individuals who had not previously been infected with dengue. This evidence ultimately led to major changes in how the vaccine was recommended and implemented.

For Butantan-DV, Aguiar stresses that the current safety signal should therefore be investigated rapidly and transparently, but without assuming causality before the epidemiological and clinical evidence has been fully evaluated. In parallel, Brazil continues to use other dengue-control tools, including vaccination with QDENGA.

From endemic dengue to the growing European risk

The relevance of this work extends beyond countries where dengue has traditionally been endemic.

Aguiar and the Mathematical and Theoretical Biology group at BCAM have also been studying how dengue and other Aedes[MA1] -borne diseases can be introduced into regions where transmission has historically been uncommon, including Europe. Their recent work combines epidemiological data, mosquito surveillance, deterministic and stochastic modelling, and environmental information to quantify the probability that imported infections can generate local transmission.

In northern Spain, the group has analysed the expanding presence and abundance of Aedes albopictus and its implications for the risk of local transmission of mosquito-borne viruses. Their research indicates that, in non-endemic settings, mosquito abundance is a key determinant of whether an imported infection remains an isolated event or has the potential to initiate local transmission.

The team has also analysed recent European dengue outbreaks, including the 2024 outbreak in Fano, Italy, using stochastic transmission models. Such studies show why outbreak risk cannot always be summarized by a single deterministic epidemic threshold: when transmission is close to the critical regime, chance events can determine whether an introduction disappears or develops into a measurable outbreak.

These results reinforce the broader message of the Science article. As competent mosquito vectors become established across larger parts of Europe and imported dengue infections continue to occur, surveillance systems will increasingly need to integrate human cases, mosquito abundance, serotype information, vaccination history, and probabilistic outbreak-risk assessment.

“Europe is not in the same epidemiological situation as dengue-endemic countries, but that is precisely why preparedness matters,” Aguiar explains. “The objective is to identify the conditions under which local transmission becomes possible before sustained transmission is established.”

Surveillance should be built into vaccine policy

For Aguiar, the temporary suspension of a vaccine should therefore be understood as part of a functioning safety system rather than as evidence that surveillance has failed.

The broader challenge is to ensure that monitoring is not assembled only after a safety concern emerges.

Dengue vaccine programmes should incorporate active pharmacovigilance, serotype-specific epidemiological surveillance, monitoring of hospitalization and severe disease, and evaluation of baseline immune status where feasible from the beginning of vaccine rollout. Safety signals can then be assessed rapidly while also being interpreted in relation to the expected population-level benefits of vaccination under local epidemiological conditions.

“The key question is not simply whether a vaccine works on average,” Aguiar says. “We need to understand for whom it works, against which serotypes, under which epidemiological conditions, and how those answers change over time.”