LOGBOOK · 2026-09-17 · SOURCE: ANVISA GGMON 06/2026, PORTARIA 715

Butantan-DV paused: 42 in 501,000,
and a system doing its job.

On 8 June 2026 Brazil's Ministry of Health temporarily discontinued the country's vaccination strategy with Butantan-DV, the world's first single-dose dengue vaccine. The reason was not a trial failure and not a recall. It was a pharmacovigilance number: 42 reports of rare, severe reactions among 501,000 people vaccinated, 0.008 percent, including three serious cases, two of them deaths. None of the 42 occurred in the municipalities where vaccination had expanded beyond health workers.

Three months later, the pause is still in force. Anvisa convened a working group and an expert panel (Portaria 715, 15 June). Minister Padilha, asked in July whether the vaccine would return, answered in the conditional tense: "Os especialistas estão investigando detalhadamente nessas 42 pessoas se tem alguma condição clínica, algum motivo... Pode ser que essa investigação sinalize que a vacina deva ficar restrita a um determinado público ou região." The experts are investigating, in those 42 people, whether there is some clinical condition, some reason. The investigation may signal that the vaccine should stay restricted to a particular group or region.

The Ministry's own FAQ puts the system's self-image in one line: "O sistema de monitoramento da vigilância epidemiológica funcionou e detectou um sinal de alarme que precisa ser melhor investigado." The monitoring system worked, and detected a warning signal that needs better investigation. Not a crisis voice. A maintenance voice.

The arithmetic that started it

Dengue kills. In the 2024 record season the Americas logged more than 13 million cases and over 8,000 deaths. Any vaccine that reaches a public system in Brazil arrives into hundreds of thousands of arms within months, which means even one-in-a-million events surface fast. Forty-two events in 501,000 doses is roughly one report per 12,000 recipients, and the two deaths make the signal impossible to wave through, whatever the final causality finding turns out to be.

Set against that: the phase 3 trial followed 16,235 participants for five years and found 65.0% efficacy against symptomatic dengue and 80.5% against severe dengue or dengue with warning signs, with zero hospitalizations in the vaccinated arm against eight under placebo. Both facts are true at once. That is exactly what a safety signal looks like before adjudication: a number too rare for trials, too serious to ignore, and unproven as cause.

Butantan-DV rollout, 2026	Date	Stage
Anvisa registration, ages 12–59	2025-11/12	World's first single-dose dengue vaccine approved
Pilot in 3 cities + 1 region	2026-01-17	Botucatu, Nova Lima, Maranguape, Araguaína
National: primary-care workers	2026-02-09	~216,000 workers targeted, 3.9M doses bought
Widened: all health workers, age 59	2026-05-04	São Paulo and other municipalities
Strategy discontinued	2026-06-08	42 severe reports in 501,000 doses (0.008%)
Anvisa working group + panel	2026-06-15	Portaria 715; benefit-risk assessment open

Rows from the Anvisa GGMON alert 06/2026, Ministry of Health FAQ, and São Paulo municipal bulletins. Checked 2026-09-17.

Vaccine cold-chain staging room: open insulated carrier with single-dose vials and a data logger on a worn stainless table, clinic corridor behind — The cold chain that moved 501,000 single doses between January and May. A pause notice on a clinic door is what precaution looks like when a signal is real but unadjudicated. Illustrative staging, not an actual Butanutan facility.

What suspension does and does not mean

Three confusions travel with every headline like this one. First: the vaccine is not banned, and its efficacy data are not invalidated. The Ministry's discontinuation text says plainly that the decision does not invalidate the clinical findings; it stops new exposure while causality is investigated in the 42 individuals, one by one, for underlying conditions.

Second: Qdenga continues. Brazil's other dengue vaccine, the two-dose Takeda product, remains in the public system for children and adolescents aged 10 to 14, roughly eight million doses since 2024. A signal on one product is a signal on one product.

Third: this is the system working, not failing. The surveillance net caught a 0.008 percent anomaly within months of rollout, and the response was to stop, investigate, and publish the working group's mandate in the federal gazette. The alternative timeline, where 42 scattered reports never aggregate into a visible number, is the one that ends badly.

Public-health surveillance room at night: wide desk with stacked report folders, a highlighted spreadsheet printout, CRT monitor with a bar chart, one desk lamp — Forty-two reports did not arrive here as a number. They arrived as forty-two separate faxes, calls and forms, from forty-two separate clinics, over four months. The room's whole job is to notice when the pile becomes a pattern. Illustrative staging.

The honest open question

One caveat belongs beside every efficacy figure on this page, and it predates the suspension. During five years of follow-up, DENV-3 and DENV-4 simply did not circulate enough to be measured. The trial's protection claims rest on DENV-1 and DENV-2. Whether single-dose protection generalizes to the other two serotypes is unknown, and unknown in the same way for the competitor vaccine. Brazil's investigators now carry two open questions at once: what happened in those 42 people, and what happens when serotype 3 or 4 returns in force.

For a household, nothing here changes the basics

If you were vaccinated, the Ministry's standing guidance is to watch the 21 days after any dose and seek care for symptoms, not to act on rumors. If you were not, the suspension removes a tool but not the work: the ten-minute yard survey and the warning-signs page are the same instruments they were in May. Mosquito control did not get paused in June.

We will log the panel's conclusion when it is published.