DGHI Affiliate Contributes to Development of WHO Dengue Guidelines

January 31, 2017

In summer 2016, the World Health Organization (WHO) issued guidelines on the implementation of Dengvaxia, the first vaccine against all four dengue serotypes, or strains. Katia Koelle, associate professor of biology and Duke Global Health Institute (DGHI) affiliate, was among a consortium of researchers selected to generate model-based predictions of the long-term safety, health and economic impact of the vaccine. 

These models—and the researchers’ subsequent conclusions—helped inform the WHO’s position on the use of the vaccine. The model comparison study was recently published in PLOS Medicine.

About Dengue

According to the WHO, the global annual incidence of dengue is estimated at 50 million to 100 million symptomatic cases per year, primarily in Asia, followed by Latin America and Africa. As with Zika, dengue’s primary vector is the Aedes aegypti mosquito. The vaccine, developed by Sanofi Pasteur, has been licensed in several Asian, Central American and South American countries. 

Dengue Infection Pattern Complicates Vaccination Strategy

Several factors make formulating a dengue vaccination strategy particularly challenging.

An individual’s first dengue infection is usually very mild, often asymptomatic, but provides long-term immunity to that serotype of the virus and short-term (about one to two years) immunity to the other serotypes. After that brief period of immunity, if a person contracts a different serotype of the virus, the infection tends to be far more severe. After the second infection, subsequent infections are, like the first infection, typically mild or asymptomatic. 

Given this pattern, and trial data that suggest that the Dengvaxia vaccine acts as a “silent,” or asymptomatic natural infection, vaccination would ideally occur after the first dengue infection, with the vaccination acting as the second infection, but without the possibility of developing into severe disease. However, since the first infection is often undetectable, it’s nearly impossible to identify individuals who have been infected with dengue one time without administering a costly serological test. And vaccinating a person before he or she has contracted the dengue virus puts the individual at higher risk for a severe second infection. 

“We have to be extremely careful about how we vaccinate people,” said Koelle. “Some vaccines—like this one—can turn out to be detrimental in certain epidemiological contexts.” 

A further vaccination complexity is that in some countries where dengue is endemic, the transmission intensity varies by region. In one area of the country, it may be reasonable to assume the majority of people of a certain age have already had an initial dengue infection, while in another region, the chance of infection may be significantly lower. Many countries have nationwide vaccination policies that fail to take into account these variations in transmission intensity. 

Eight Epidemiological Models Were Developed

The consortium of researchers brought to the table eight different models to predict the impact and cost-effectiveness of the vaccine in a variety of situations. Each model reflects some different assumptions about dengue epidemiology and is based on different parameters, but a common element is the assumption—based on trial data—that the Dengvaxia vaccine acts as a “silent,” or asymptomatic infection and provides temporary cross-protection against all serotypes. The models also assume that vaccination is implemented in settings with existing dengue control interventions and treatment.

“The models take into consideration different unknowns about, for example, how the different dengue strains interact, how much complexity there is in terms of seasonality and how people mix with one another,” said Koelle.

What the Models Tell Us

All of the models found that in areas of moderate to high dengue transmission intensity, and particularly in high transmission settings, the vaccine would result in a reduction of dengue cases and hospitalizations. 

In contrast, the researchers concluded from the models that implementing the vaccine in low transmission intensity areas would probably increase dengue hospitalizations. In these settings, the vaccination is more likely to act as a person’s first exposure to dengue, thus priming him or her to have a secondary-like infection if naturally exposed to dengue again.

WHO Recommendations Reflect Modeling Conclusions

Last spring, the research consortium presented its findings to the WHO Strategic Advisory Group of Experts on Immunization. The WHO drew heavily on these findings when formulating its position on the Dengvaxia vaccine.

The WHO recommended that the vaccine should be considered for use only in areas where the dengue disease burden is high—with prior infection of dengue virus at 50 percent or higher in the age group targeted for vaccination. Implementation of the vaccine in populations with seroprevalence (prior infection) of 70 percent or higher is likely to yield the most significant public health impact and cost-effectiveness, while vaccinating populations with seroprevalence between 50 and 70 percent may have a lower impact.

The WHO recommended against implementing the vaccine in populations with low seroprevalence (under 50 percent) in the age group considered for vaccination, due to low cost-effectiveness and the potential long-term risks of severe dengue in people who receive the vaccination with no prior dengue infections.

The WHO also specifies that the dengue vaccine should be a part of a comprehensive dengue control strategy that should also include vector control, evidence-based clinical care for dengue patients, and strong dengue surveillance.

Want stories like this in your inbox? Sign up to receive DGHI emails!


We have to be extremely careful about how we vaccinate people. Some vaccines—like this one—can turn out to be detrimental in certain epidemiological contexts.

Katia Koelle, associate professor of biology and DGHI affiliate

Share This