News TVs 24

FAPESP: Change in vaccination strategy prevented urban yellow fever cycle in SP

Publicado em 10 novembro 2021

Yellow fever is a non-contagious viral disease that has two transmission cycles: wild and urban. In the first, the virus circulates among monkeys, being transmitted in Brazil by mosquitoes of the genera Haemagogus e Sabethes. In the second, it circulates among humans, being transmitted by the Aedes aegypti. Brazil has not had cases of urban yellow fever since 1942, but the wild version can, by accident, affect people who live near forests and forests, or who frequent them.

This is what happened in the last wild outbreaks in São Paulo, between 2016 and 2019. According to an article published in Scientific Reports, which describes the process of diffusion in the state, the disease killed people in places very close to the capital and in cities that had no vaccine recommendation, as they had never been affected by the disease, such as Campinas. Had the vaccination strategy not been adapted to the circumstances, the effects could have been much worse, resulting in more deaths, scientists now reveal.

“We are at great risk of having urban yellow fever reintroduced in cities adjacent to São Paulo. An urban epidemic would have very serious effects and would be a setback”, says Francisco Chiaravalloti Neto, professor at the Department of Epidemiology at the Faculty of Public Health at the University of São Paulo (FSP-USP). He is a co-author of the work, carried out with the support of FAPESP (São Paulo State Research Support Foundation).

Two waves were identified between 2016 and 2019: one coming from west to east (2016 and 2017) and another spreading from the Campinas region to the neighboring municipalities of Rio de Janeiro, Minas Gerais and Paraná, to the coast of São Paulo (2017 to 2019). The first wave left São José do Rio Preto in April 2016 and reached Campinas in August 2017, at a speed of 1 kilometer (km) per day. “At that time, due to the low vaccine coverage, cases in humans increased and the disease spread towards the capital of São Paulo, coast, Vale do Paraíba and Sorocaba at the same speed, and also to Vale do Ribeira, where it moved more slowly ”, recalls the FSP-USP professor.

The vaccination strategy of the Ministry of Health and the state secretariat in these cases is to immunize the entire municipality that is at risk and also those adjacent to it. “But when the disease hit Campinas, the State Immunization Division gave up this strategy, which included vaccinating the entire city plus neighboring municipalities, such as Indaiatuba and Paulínia, and decided to vaccinate cities like Jundiaí, Jarinu and Itatiba, as it was detected that that’s where the virus was migrating. At the time, Campinas and neighboring areas had more than 1 million inhabitants, but less than 1 million doses were available in the state”, summarizes Adriano Pinter, scientific researcher at the Superintendence for the Control of Endemic Diseases (Sucen), an agency of the Health Department of the State of São Paulo, and co-author of the article.

He claims that in Mairiporã, a city in which 181 human cases occurred, coverage was close to 80% when the disease arrived (unfortunately, around Christmas, when the number of visitors increases). “If Mairiporã had more than one hundred deaths even with vaccination coverage above 80%, imagine if they had not been vaccinated?” population. “People only get vaccinated when they are going to travel.”

Scientists warn that the situation experienced between 2016 and 2019 could be repeated. “We have to be careful, as we know the path that the disease can take. Studies show that these waves of yellow fever start in the Amazon and, through forest corridors, pass through Tocantins, Goiás and Triângulo Mineiro until reaching the State of São Paulo, spreading along the coast and migrating to the south. Today, the wave has arrived in Rio Grande do Sul. It is believed that they are waves that happen every five years. Campinas had never been affected by the wild cycle, it was the first time. And we describe this process well at work,” says Pinter.

Until 1999, the wild version of yellow fever was restricted to the North and Central regions of the country, with sporadic cases in the Southeast. Since the 2000s, São Paulo has been one of the centers of expansion and circulation of the disease. Between 2016 and 2019, 648 human cases were confirmed in the state, with 230 deaths, and 850 cases in monkeys or groups of monkeys. Yellow fever can present mild symptoms, which happens in most cases, but the fatality rate in severe cases is 40%.

propagation pattern

Despite not being a contagious disease, the diffusion pattern described by scientists is called “contagion” propagation, an expression that refers to transmission in the territory (as if forest fragments were “contagious” to each other).

According to the first author of the article, Alec Brian Lacerda, the spread of the disease can occur through expansion, relocation or in a hybrid way. “Diffusion by expansion is divided into two: by contagion, when the spread is observed due to the proximity of territory, or hierarchical, when large municipalities are generally affected. In this case, the propagation does not follow a pattern of continuity of territory, it takes leaps, which could be an indication of an urban outbreak.”

In diffusion by relocation, the migratory movement is observed: the disease leaves its point of origin, where it stops growing, and moves to a more favorable one, creating a new point of origin. “This happens, for example, when people leave areas without a vaccine recommendation and enter, without vaccine, into areas with a vaccine recommendation. In hybrid propagation, relocation also takes place, but with the old point of origin still active. We characterized the processes in this way and created maps showing the municipalities and the cases that occurred: whether they were cases in monkeys, in humans, or both”, explains Lacerda, who began developing the work while still on a FAPESP scientific initiation scholarship.

According to Pinter, at the time the vaccination proposal was made, it was taken into account that the virus was only spreading among wild mosquitoes and, as they are not able to fly very much, it was suggested that the spread would be due to the continuity of the territory. . “But there was still a question about whether human beings participated in the transmission or not. In the article, we confirm that this did not happen, that transmission actually occurred only between mosquitoes and non-human primates, and that man was eventually stricken, but he did not, and transmission did not occur in the city. If the human being had participated in the transmission, it would have been hierarchical, that is, people would appear infected in large cities, the transmission would take ‘leaps’ from one place to another. And what was happening was that the virus was spreading in small towns, due to territorial continuity.”

direction and speed

Lacerda clarifies that the group used data from the Professor Alexandre Vranjac Epidemiological Surveillance Center (CVE), from the São Paulo Health Department (cases involving monkeys and human cases, by municipality) and also vaccination coverage data from 2015 to 2018 from the Program National Immunization (PNI).

“We used vaccination coverage data for the population between 0 and 5 years of age and based on them we approximated the total coverage for the population, as the data available for this age group is a more accurate record of how the health system accesses the population. And the Ministry of Health recommends its use in research like ours”, he says.

With the data in hand, the scientists applied a statistical technique called kriging (kriging) to map the spread of the virus. “For each municipality, we establish a date, always the date of the first case, whether human or epizootic, then we create a numerical sequence corresponding to the months, linking the municipalities to the months in which the dissemination process started and we use the technique of kriging to map the phenomenon in space and time, creating curves that show the direction and speed of diffusion progression. Once the maps were ready, we were able to compare them with the vaccination coverage map and conclude, for example, that the epidemic reached Campinas when there was very low vaccination coverage or there was no vaccination recommendation”, recalls Chiaravalloti Neto.

hot winters

The arrival of the disease in places never affected before raises several hypotheses that need to be tested. “The one that makes the most sense to me has to do with warmer winters. There should be no mosquitoes, winged, in winter. There should only be eggs, as the larva does not survive the cold and dies. But what we saw is that transmission also happened during the winter in 2017. In other words, the mosquito was flying during the winter”, says Pinter.

He believes it is possible that colder winters were responsible for blocking the transmission of the virus in the past, protecting cities like Campinas and Mairiporã. “But now we have less cold winters. I’m not talking about a huge temperature difference, but 2°C, 3°C. There are articles about dengue showing that temperatures around 20 °C are good for the mosquito, while 16 °C are already detrimental. Much the same thing happens with yellow fever. So what makes sense to me amidst so many hypotheses is that we’ve had less cold winters, and the warmer they are, the faster these virus spreads.”

The main symptoms of the disease are feeling unwell, headache, high fever, muscle pain, chills, tiredness, vomiting, nausea and diarrhea, but it can also affect the kidneys and liver. Fortunately, there is a vaccine for yellow fever, produced in Brazil since 1937. Offered by the Unified Health System (SUS), it offers lifetime protection. “Our advice to the population is to get the vaccine”, say, in unison, the three authors.

The article Diffusion of sylvatic yellow fever in the state of São Paulo, Brazil, also signed by scientists Leila del Castillo Saad, from CVE, and Priscilla Venâncio Ikefuti, from FSP-USP, can be accessed at