NEW YORK (GenomeWeb) – By combining epidemiological and genomic approaches, researchers have traced the spread of the Brazilian yellow fever outbreak.
Brazil has recently experienced its largest-ever recorded yellow fever outbreak. Between December 2016 and March 2018, more than 2,000 people had confirmed yellow fever cases and 676 people died. A University of Oxford-led team used a suite of tools to determine how the outbreak occurred.
Yellow fever, a mosquito-borne disease, is a caused by the yellow fever virus (YFV), a member of the Flaviviridae family. It can be transmitted through either a sylvatic or urban cycle. Because of the scale of the outbreak, there was concern that the virus had begun to be spread through an urban cycle.
As they reported today in Science, the Oxford-led team found, though, that the viral transmission characteristics suggested a sylvan cycle transmission and they traced the virus' origins to northern or central-western Brazil.
"Ultimately, the combination of epidemiological and genomic data tells us that the YFV outbreak in Brazil was driven by sylvatic transmission," first author and Oxford researcher Nuno Faria said in a statement. "Although the conditions for urban transmission seem to be there, that fortunately didn't happen."
In a sylvan cycle — which is the typical yellow fever transmission means in the Americas — mosquitoes that live in trees, like Haemagogus and Sabethes, infect non-human primates which leads to human infections in forests. But in an urban cycle — which has occurred in Angola, the Democratic Republic of the Congo, and in 1942 in Brazil — Aedes mosquitoes that feed mostly on humans transmit the virus.
If the virus was being transmitted through an urban cycle, the researchers noted there could be serious consequences, especially if it spread to São Paulo or Rio de Janeiro. Though there is an effective yellow fever vaccine, only people living near forested areas in Brazil were vaccinated until recently.
Faria and his colleagues examined the characteristics of 683 PCR-confirmed yellow fever cases in humans and 313 non-human primates in Brazil, mostly Minas Gerais. In a time series, they found cases in the two hosts to peak around the same time, though the human cases lagged behind by four days.
In addition, human cases were more likely to live closer to a possible sylvatic infection source or to a recent non-human primate case than a randomly chosen resident. They also found yellow fever infection patterns to differ in timing from chikungunya infections — which is spread by Ae. aegypti and Ae. albopictus — indicating they might be transmitted by different vectors.
When they compared the age and sex distribution of observed YFV cases with what would be expected under a sylvan or urban transmission cycle, the researchers found the data to fit with a sylvan cycle: 85 percent of cases were in men, who are more likely to travel to remote regions of the forest.
Using YFV genome data from this outbreak, two prior Brazilian outbreaks, and a publicly available dataset, the researchers constructed a YFV phylogenetic tree that placed the recent epidemic as a single clade within the South American I genotype. The outgroup of the outbreak clade was from a human case from 2002 in north Brazil.
As samples from previous outbreaks were distantly related, this suggested to the researchers that the epidemic was due to a new strain, rather than the re-emergence of a previous strain.
By combining genomic and epidemiology data, the researchers pieced together when and where the virus emerged. They estimated the time of the most recent common ancestor of the outbreak clade to be late July 2016 in northeastern Brazil, which is around the time the first case was confirmed in a non-human primate in Minas Gerais. The virus then spread south toward more populated area, moving about 4.25 kilometers a day.
That, they noted, is fairly quick and may have been expedited by infected mosquitoes being moved in cars or by the illegal trade of non-human primates. By the time yellow fever was declared a public health emergency in Minas Gerais in January 2017, it had spread about 600 kilometers.
Faria and his colleagues said the results show that combining epidemiological and genomic data can help monitor disease exposure risk. In a related commentary in Science, the University of Texas Medical Branch's Alan Barrett said that such mapping could contribute to reaching the World Health Organization's goal to eliminate yellow fever epidemics by 2026.