An estimated 3.9 billion people in 128 countries are at risk of dengue virus infection. Of the estimated 390 million dengue infections that occur each year, 96 million will manifest clinically with flu-like symptoms including fever, headache, nausea and muscle and joint pain. Unlike the flu virus, dengue virus cannot be transmitted directly from person to person. It instead relies on an insect vector, the mosquito Aedes aegypti. Female mosquitoes contract the virus when they bite and feed on an infected human. After a period of four to ten days, the virus disseminates to various tissues in the mosquito, where it remains for the rest of the mosquito’s life. At this point, the mosquito is infectious and can transmit the virus through its saliva and bite.

Earlier studies showed that the time during which dengue virus-infected humans can transmit the virus to mosquitoes coincides with the onset of clinical symptoms and an increase in viral load in their blood. These observations led to the assumption that infected, asymptomatic humans are so-called “dead-end hosts” for the virus because their viral levels are so low as to make them noninfectious to mosquitoes, essentially breaking the transmission chain.

In a new paper published last week in the Proceedings of the National Academy of Sciences, an international group of researchers challenged a long-held assumption that asymptomatic patients infected with the dengue virus are not infectious. The team sought to experimentally test the assumption that asymptomatic people are noninfectious and to determine how human-to-mosquito transmission varied with timing of symptom onset.

To identify potentially asymptomatic, infected people, the researchers devised a strategy wherein hospitalized patients confirmed as having a dengue virus infection were treated as index cases. The researchers then sent out teams to recruit “cluster participants”: family members from the index patient’s household or neighbours living in houses within a 200 m radius or in the 20 closest houses. The researchers used blood tests to determine their infection status and conducted follow up visits to monitor symptom development. They also took blood samples to measure the amount of viral RNA in their blood.

So how does one measure the transmission of dengue virus from humans to mosquitoes? By letting the mosquitoes feed on human blood and then measuring the amount of the virus’ genetic material in the mosquito. In the direct feeding technique, two groups of 25 hungry female mosquitoes were trapped underneath a plastic cup on a participant’s leg for five minutes and allowed to feast. In the indirect method, blood samples were taken from the participant and placed in feeders with hungry female mosquitoes.

The study recruited a total of 181 participants: 126 participants had symptoms of dengue fever when the mosquito feedings took place; 42 were classified as presymptomatic because they did not develop symptoms until after the mosquito feedings; and 13 asymptomatic participants remained symptom-free from recruitment through to the 10 day follow-up period.

For symptomatic and presymptomatic participants, dengue virus levels in the blood peaked between day zero, when clinical symptoms were first observed, and two days thereafter, although viral RNA would still be detected in some patients as far eight days after symptom onset. Using both the direct and indirect feeding methods, the researchers showed that the human-to-mosquito transmission occurred primarily from two days before symptoms appear to six days after the onset of illness. Not surprisingly, asymptomatic participants had, on average, lower levels of dengue virus in their blood than presymptomatic or symptomatic individuals. The average viral load in asymptomatic individuals was similar to that in symptomatic patients two to three days before or five to eight days after symptom onset.

The researchers unexpectedly found that asymptomatic dengue virus-infected participants were significantly more infectious to mosquitoes than symptomatic participants at all times during the course of the illness, except between days zero and two. Both asymptomatic and presymptomatic participants were more likely to transmit the virus to feeding mosquitoes than symptomatic patients, regardless of the viral load in the individual’s blood. The 50% mosquito infectious dose, which is the amount of virus needed in the blood to infect 50% of mosquitoes, was roughly 100 times lower for asymptomatic and presymptomatic infections than for symptomatic cases. What’s more, mosquitoes infected by asymptomatic or presymptomatic participants had higher viral loads than those infected by symptomatic participants.

So even though the asymptomatic participants had lower viral loads in their blood, they had a higher transmission potential with a greater proportion of mosquitoes being infected and those mosquitoes themselves having a higher viral load.

As paradigm-shifting as these results are, there were a number of limitations to this study. First, the numbers of presymptomatic and asymptomatic participants were actually quite small. It will be interesting to see if these effects hold up in a larger population. Second, in this study, the data on how infectiousness changed over the course of the illness were generated from a composite of the data from all the participants, who were all tested at different points in their illness. This is because each participant only participated in mosquito feeding once. Ideally, you would want to determine the infectiousness of the same individual every day before, during and after symptoms appear. That would mean either subjecting someone to mosquito bites every day to allow direct feeding or taking daily blood samples for the indirect feeding technique, neither of which sound particularly enjoyable. Presumably, if you compile single data points from enough participants to make a time course composite, the individual differences will average out and you should be left with a fairly reliable infectiousness profile. But again, more participants equal more power.

These findings challenge the hypothesis that asymptomatically infected people are dead-end hosts for dengue virus and add a new dimension to the transmission dynamics and epidemiology of this disease. Within a public health context, these results highlight the need to be able to detect all infections quickly, both symptomatic and asymptomatic, and to broaden surveillance efforts to include those who are likely to be exposed to the virus but do not have any clinical signs of disease. A more complete understanding of transmission dynamics will inform better prevention strategies and more effective control measures, both of which are desperately needed in the fight against dengue.

Reference: Duong V, Lambrechts L, Paul RE, Ly S, Lay RS, Long KC, Huy R, Tarantola A, Scott TW, Sakuntabhai A, & Buchy P (2015). Asymptomatic humans transmit dengue virus to mosquitoes. Proceedings of the National Academy of Sciences of the United States of America PMID: 26553981