Is artificial intelligence key to dengue prevention?

Image via Sanofi Pasteur

Scientists hope new dengue warning system and big data can control outbreaks

Dengue fever outbreaks are increasing in both frequency and magnitude. Not only that, the number of countries that could potentially be affected by the disease is growing all the time.

This growth has led to renewed efforts to address the disease, and a pioneering Malaysian researcher was recently recognized for his efforts to harness the power of big data and artificial intelligence to accurately predict dengue outbreaks.

Dr. Dhesi Baha Raja received the Pistoia Alliance Life Science Award at King’s College London in April of this year, for developing a disease prediction platform that employs technology and data to give people prior warning of when disease outbreaks occur.

Image of Dr Dhesi, creator of a dengue prevention apllication using artificial intelligence.

Dr Dhesi, public health advocate and PhD student at Universiti Malaysia Sarawak

The medical doctor and epidemiologist has spent years working to develop AIME (Artificial Intelligence in Medical Epidemiology) together with his colleague Mr. Rainier Mallol, Dr. Peter Ho & Dr. Ting along with a team of six people. This is the third international award that the prediction platform has won, the first being the Global Impact Competition that received recognition from Singularity University in Silicon Valley as well as the Clinton Foundation and the second award was being the Best Health Startup in Latin America.
President of Malaysian Integrated Medical Professional Association (MIMPA), Dr. Dhesi said that winning the latest award, which was organized by the Pistoia Alliance of King’s College London, proves and validates the artificial intelligence technology used as a tool for dengue prevention.

Despite inherent challenges in gathering such large scale data, ultimately, the powerful platform will provide the authorities with critical information about where and when an outbreak is likely to occur. This priceless information could aid in rolling out intervention strategies, such as warning the public and clearing potential breeding sites.

The science bit

For those of us who think Java is our morning caffeine hit, exactly how does the platform work? Dr. Dhesi explains that it relies on a complex algorithm, which analyses a wide range of data collected by local government and also satellite image recognition systems. Over 20 variables such as weather, wind speed, wind direction, thunderstorm, solar radiation and rainfall schedule are included and analyzed. Population models and geographical terrain are also included. The ultimate result of this intersection between epidemiology, public health and technology is a map, which clearly illustrates the probability and location of the next dengue outbreak.

The ground-breaking platform can predict dengue fever outbreaks up to two or three months in advance, with an accuracy approaching 88.7 per cent and within a 400m radius. Dr. Dhesi has just returned from Rio de Janeiro, where the platform was employed in a bid to fight dengue in advance of this summer’s Olympics. In Brazil, its perceived accuracy was around 84 per cent, whereas in Malaysia in was over 88 per cent – giving it an average accuracy of 86.37 per cent.

The web-based application has been tested in two states within Malaysia, Kuala Lumpur, and Selangor, and the first ever mobile app is due to be deployed across Malaysia soon. Once its capability is adequately tested there, it will be rolled out globally. Dr. Dhesi’s team are working closely with mobile digital service provider Webe on this.

By making the app free to download, this will ensure the service becomes accessible to all, Dr Dhesi explains.
“With the web-based application, this could only be used by public health officials and agencies. We recognized the need for us to democratize this health service to the community, and the only way to do this is to provide the community with the mobile app.”
This will also enable the gathering of even greater knowledge on the possibility of dengue outbreaks in high-risk areas, as well as monitoring the changing risks as people move to different areas, he adds.

Giving everyone the ability to predict dengue fever outbreaks in their pockets is nothing short of revolutionary. Dr. Dhesi agrees; “We are trying to be the Elon Musk of public health,” he explains, referencing the renegade engineer, investor, and electric car pioneer who is fighting to reduce global warming.

The wrong focus

Dengue fever is not only a major public health problem, it is an expensive one. Dr. Dhesi explains to Break Dengue that, according to a 2010 World Health Organisation report, dengue-affected countries spend an average of $440 million per year combatting the disease. It is his opinion that these efforts are inefficient at best, and completely misdirected at worst. For example, some 43 per cent of this spend on dengue goes towards ‘fogging’ i.e., spraying to kill, or ‘knock-down’, any adult dengue mosquitoes that may be carrying the virus. Dr. Dhesi says that this narrow focus on vector control is challenging, adding that current reactionary methods in dealing with disease outbreaks are insufficient and costly.

“Dengue is so expensive because public health efforts have been focused on a reactive approach. This has been happening since the 1960s, that only when an outbreak actually occurs do we react to it.” As these numerous ways to deal with dengue have been without success, it is Dr. Dhesi’s belief that using data is the best way to shift to a more pro-active and thus effective approach.

Image of the team who developed the award winning application to combat dengue outbreaks using artificial intelligence, and big data.

Dr Dhesi’s team of six developed the award winning application to combat dengue outbreaks using artificial intelligence, and big data

“Instead of using the money we spend on nationwide health campaigns, we could direct financial resources towards tech, in other words, predicting exact outbreaks in advance in order to strategize a more focused prevention program which will allow us to spend less money & have a more effective outbreak management approach”

A neglected disease no more

Dengue was originally thought to be confined to Africa and some Asian countries, but in recent times, several other countries have reported cases. It has become clear that the total population now at risk is between 2.5 billion to 3 billion, and this new reality must alter the way the disease is perceived and addressed, according to Dr. Dhesi. “It is my view that dengue fever falls into the category of neglected tropical diseases. Now we are seeing the trend that dengue is moving towards the rest of the world. The Bay Area in California has recently reported a number of cases, we have also seen cases in Japan. It is moving into Latin America and Europe as well, so there is a clear need for us to pay more attention to dengue.” He also warns that prevention, as always, is better than cure.

“We must understand that it is not just dengue fever we must be concerned with. Dengue is carried by the Aedes mosquito, and this same mosquito also carries the Zika virus.”

Yet while the WHO has been reacting strongly to threats posed by outbreaks of Zika virus and Ebola, dengue fever has not received the same attention or prominence. “When it comes to dengue, people can be somewhat complacent. Now we have seen the same mosquito carrying Zika… I think it is time we shifted our attention.”

The next step

What is the next step for the award-winning AIME? Dr. Dhesi is now seeking venture capitalists as well as Government agencies to fund AIME’s endeavors as he continues to work on the roll-out of the mobile app and platform for government officials. Those who wish to pledge their support can also do so by voting for his project on the Webe website

The doctor, researcher, and entrepreneur is extremely active on social media, tweeting as @DhesiMD to almost 6,000 followers. His goal is to get people talking about dengue fever and raise awareness of his disease prevention efforts. “We realize that in order to kick this off globally we will need collaboration partners. By rolling it out globally, we will help to reduce the number of outbreaks, as well as significantly bringing down the costs of healthcare as related to dengue fever. There is a need for us to act as soon as possible.”

—

Find out more about how dengue impacts people’s lives in our short documentary below.

Bigger than Zika? The pregnancy risks of dengue

image of woman getting checked for dengue during pregnancy.

A pregnant woman is examined for dengue.

Zika is the topical tropical disease on everyone’s lips due to the devastating consequences of the latest outbreak; namely, the microcephaly suffered by offspring of women who have become infected during pregnancy.

Zika virus is a flavivirus that is primarily transmitted by infected Aedes mosquitoes. This vector also transmits the dengue virus and is commonly found in tropical and sub-tropical environments in Africa, the Americas, Asia and the Pacific.

The recent outbreak of Zika virus has been accompanied by an unprecedented rise in the number of children being born with microcephaly, or an unusually small head. In addition, several countries, including Brazil, reported a steep increase in the neurological condition Guillain-Barré syndrome, which can cause paralysis and is often fatal. The epidemic of birth defects caused by the Zika virus has led researchers to look more closely at the potential consequences of other viruses during pregnancy.

Dengue during pregnancy and associated fetal adverse outcomes

Although Zika and dengue viruses are closely related, there is limited data on the possible adverse effects of dengue infection during pregnancy on fetal outcomes. Yet, just like Zika, dengue virus is spreading fast across Brazil and other South American countries.A recently published systematic review and meta-analysis aimed to estimate the increase in the risk of four adverse fetal outcomes in women who had dengue infection during pregnancy; namely, stillbirth, miscarriage, preterm birth, and low birthweight.

A recently published systematic review and meta-analysis aimed to estimate the increase in the risk of four adverse fetal outcomes in women who had dengue infection during pregnancy; namely, stillbirth, miscarriage, preterm birth, and low birthweight.

Researchers at the London School of Hygiene & Tropical Medicine engaged in a systematic review and meta-analysis, searching the available literature for any relevant studies.

The review included any original studies that reported any fetal outcomes for pregnant women who had dengue infection during the gestational period. They excluded case reports, ecological studies, reviews, in-vitro studies, and studies without data for pregnancy outcomes, but case-control, cohort, and cross-sectional studies and unselected case series were eligible for inclusion.

They identified 16 studies that were eligible for inclusion in the systematic review and eight that were eligible for the meta-analyses; these included over 6,000 pregnant women, with almost 300 of these having been exposed to dengue during pregnancy.

The results were nothing sort of significant; evidence emerged that dengue infection alone, in the absence of clinical symptoms, does not affect the outcome of pregnancy, but suggested clinical dengue during pregnancy seems to increase the frequency of stillbirth, prematurity, and low birth weight.

For example, with miscarriage, the researchers determined women with dengue infection during pregnancy were three and a half times more likely to experience a miscarriage compared with those without.

In addition, the risk of stillbirth was estimated to be up to seven times as high for a woman with dengue infection versus those without; although a meta-analysis for stillbirth could not be carried out because that outcome was investigated in just one study, the researchers calculated the crude relative risk to be 6.7 in women with symptomatic dengue compared with women without dengue. Preterm birth and low birth weight were the most common adverse pregnancy outcomes.

The authors concluded that the evidence suggests symptomatic dengue during pregnancy might be associated with fetal adverse outcomes.

“If confirmed, it would be important to monitor pregnancies during which dengue is diagnosed and to consider pregnant women in dengue control policies,” they said.

Mrs. Enny Paixão, of the LSHTM, was the lead author on the study. Dengue during pregnancy is the topic of her PhD at the School, and she is also part of a research group on dengue and other arboviruses at the Federal University of Bahia in Brazil.

She explains that this group is one of the most important in Brazil in the field of epidemiology of infectious diseases and has been conducting investigations on dengue and latterly Zika and chikungunya, as well as other infectious diseases of interest to public health.

In terms of what instigated the recent meta-analysis, Mrs. Paixao says that while there a broad understanding that dengue fever is potentially dangerous in pregnancy, exact data was not known. The recent emergence of severe antenatal side effects of Zika virus was another factor.

Zika and dengue risks to expecting mothers

Zika is from the same family of diseases as dengue fever, yet has been receiving far more attention in the past few months – are people unaware of how common and potentially dangerous dengue infection is?

“Zika is an emerging disease, and there is a lack of information about the effects of the disease in the general population and in pregnant women. So it is important to investigate the potential effects of both diseases. There is evidence that other infections can be harmful during pregnancy so it was pertinent to produce information about the effects of dengue during pregnancy. Another systematic review published in 2010, had concluded that vertical transmission is possible and that the virus could potentially be dangerous in pregnancy,” she explains.

It is estimated that each year, 390 million people are infected with dengue and 96 million develop clinical symptoms. The results of the meta-analysis and review provide further evidence of a link between symptomatic dengue virus and adverse outcomes in pregnancy.
“Most cases of dengue are asymptomatic or self-limiting, however, some cases progress to severe illness. Infections without any clinical symptoms do not appear to be harmful to the fetus, but also that clinical dengue during pregnancy seems to increase the frequency of stillbirth, prematurity, and low birth weight,” says Mrs. Paixão.

The results are extremely interesting but not conclusive, she stresses; further insights must be gained into how, why, and when dengue virus impacts fetal development.

“The results of this study are key to attract attention to this important topic but there are many questions that remain without an answer. These should be better investigated, such as the relevance of clinical symptoms and the gestational age at which the disease occurs.”
The research team at LSHTM have recommended the development of further epidemiological studies on this theme, with larger sample sizes, adequate comparison groups, as well as control groups to avoid confounding.

Right now, the implications of these findings for clinicians treating people in areas where dengue fever is rampant are unclear but Paixao says that there are a plethora of strategies that could be employed if these findings are confirmed in the future.

“If this association is confirmed, recommendations should be made for the close monitoring of pregnancies during which dengue is diagnosed and also strategies for dengue control to include pregnant women as an at-risk population.”

The LSHTM team call for original research into this issue to be carried out; but while there are scientists interested in the topic, developing an appropriate methodology to carry out these studies is not an easy task, admits Paixao.

“Acquiring dengue during the nine months of pregnancy is relatively rare in a population where dengue has already been circulating for a long time. Yet we need a large number of individuals in order to be able to observe if there is a relationship between dengue and adverse fetal outcomes. Also, it is difficult to establish a direct relation with a fetal outcome that sometimes occurs seven to eight months after the disease onset,” she explains.

According to Paixao, people are already aware of the potential dangers of Zika can cause, and a similar level of awareness must be raised in relation to dengue fever.

“Dengue and Zika are both vector-borne disease so the way to be protected is the same, avoiding a bite from the mosquito. But it is also important to warn of the potential risk of dengue harmful outcomes when acquired during pregnancy because the risk of dengue infection will be continued in the areas where arboviruses are circulating until the development of a vaccine.”

—

Learn more about the risks of dengue and long-term effects this disease can have on people’s lives.

Watch ‘The Lingering Effect of Dengue Fever’

Report dengue fever activity near you. Join Dengue Track

Dengue Track

Researchers discover game-changing dengue treatment

Clinical trials for a new dengue fever treatment could start within a year, following a discovery by the research team of University of Queensland scientists.

A new dengue fever treatment for victims may not be as far off as you might think. Australian scientists took a massive leap forward following a recent, ground-breaking discovery. The finding that the body reacts similarly to dengue fever and bacterial infections could lead to existing therapies being repurposed as effective drugs, and researchers say clinical trials could begin within 12 months.

The research, conducted at University of Queensland’s School of Chemistry and Molecular Biosciences and the Australian Infectious Disease Research Centre, was published recently in Science Translational Medicine.

The research team was led by virologist and Head of School Professor Paul Young, who has been working on dengue virus for many years. His long-term interest in a viral protein secreted from infected cells, NS1, has now borne fruit as its pivotal role in dengue fever has finally been elucidated.

He says he became interested in dengue virus NS1 back in the late 1980s when he obtained WHO funding to work on projects at the London School of Hygiene and Tropical Medicine. Over the years, Professor Young and his team have characterized the protein, recombinantly expressed it, purified it, and used it as the basis of vaccine candidates, as well as the foundation for a diagnostic assay that was commercialized and is now the gold standard for early diagnosis used around the world.

20-year-old hypothesis unlocks potential dengue fever treatment

The hypothesis that arose nearly 20 years ago was that NS1 was a stimulant for the immune system and would directly activate macrophages (an immune cell type), to release factors promoting inflammation. Dengue typically causes a debilitating fever but this can progress to potentially fatal dengue hemorrhagic fever and dengue shock syndrome.

The team started working in earnest on the project about three years ago, and the original hypothesis has now proven to be correct; the NS1 protein does indeed activate macrophages, and also endothelial cells, the cells that line our blood vessels. This results in the production of proteins by macrophages that lead to inflammation and fever, explains Prof Young.

“Back in the mid-1990s my group identified a possible role for NS1 in exacerbating disease and now our latest findings have shown that this viral protein is, in fact, a viral toxin that can induce leaks in capillaries. The clinical symptoms associated with severe dengue disease, hemorrhage and shock may be directly caused by NS1.”

The UQ team have now carried out early laboratory experiments supporting that assumption as they have been able to block NS1 activity in cellular systems, explains UQ immunologist Professor Kate Stacey.

“We have shown in cell culture systems that NS1 has effects on the endothelial cells consistent with an increase in blood vessel ‘leakiness’. When we treat an intact cell layer with NS1 the junctions between cells are weakened,” she says, explaining that this would be consistent with the hemorrhage and shock that is seen in severe cases of dengue virus infection.

A similar pathway to ‘leakiness’

The pathway that eventually leads to this “leakiness” appears to be similar to the potentially fatal leak caused by bacteria in cases of septic shock but the problem was to identify a cellular receptor for NS1 that was eliciting these responses. The team then found a protein called TLR4, which is responsible for mediating the response to NS1. TLR4 is primarily known as a receptor for lipopolysaccharide (LPS) which is a major bacterial cell wall component.

“This instantly makes immunologists very suspicious of our findings, because LPS is everywhere, and can easily contaminate biological materials. So we had an extensive job to satisfy ourselves and critics that we were not just looking at a response to contaminating LPS,” Professor Stacey explains.

Determining that NS1 is mediating effects through the TLR4 receptor provides a particular advantage in the traditionally slow process of drug development, as this pathway has already intensively studied, particularly for treatment or prevention of bacterial septic shock, says Professor Young.

“As it is the same pathway stimulated by dengue NS1 we believe that we could successfully repurpose these sepsis drugs for treating dengue patients. These compounds have already gone through the many years of clinical trials necessary to establish safety and efficacy. Consequently, we could bypass the normal lengthy early clinical trials process saving many years to application.”

Dramatic effects as target identified

Indeed, Professor Stacey says she was shocked by the dramatic effects observed when the team treated mice infected with dengue virus with an inhibitor of the TLR4 pathway.

“We found this treatment prevented the leakage of fluid into the tissues from the blood. This suggests that TLR4 is a very good target for reducing the symptoms of dengue infection.”

An interesting aspect of this is that receptors like TLR4 have evolved to try and help the immune system identify the presence of invaders and mount defensive responses. But when these responses are too strong, they can become part of the pathology and it is often the body’s response to infection that causes much of the problem. This then leads to the possibility of having drugs that don’t directly target the virus, but instead modify the body’s response to the virus to reduce the disease symptoms.

Bigger than Zika? Learn more about the pregnancy risks of dengue

If you experience dengue symptoms, see a medical professional at once.

Clinical trials possible within a year

Drugs targeting the TLR4 pathway are in various stages of development; one such drug, eritoran, has already been through phase III clinical trials for bacterial sepsis and is also currently under investigation for influenza.

“So if TLR4 is a good target for treating dengue, then we have the advantage of all this prior development, rather than having to make drugs from scratch. We are not at the stage of clinical trials yet and we are just commencing trials of eritoran in mice. If this is successful like our earlier TLR4 inhibitor, then clinical trials could be organized within a year,” states Professor Stacey.

The significance of these findings is not confined to dengue; they have wide-ranging implications throughout the immunology field. Professor Stacey says the research draws interesting and surprising parallels between dengue viral disease and bacterial sepsis.

“The dangerous complications of both involve blood vessel leakiness and shock, leading to organ failure. We would propose that the activation of the TLR4 pathway is common to both these conditions. There are a growing number of viral infections where responses through TLR4 are implicated, including influenza, ebola virus and Japanese encephalitis virus. This is an interesting twist for something that looked for many years like a receptor specific for response to bacteria.”

A drug that lessens the severity of the disease

The hope from this work is for a drug that inhibits the response through TLR4, lessening the severity of the disease and decreasing the length of hospital stays. This is important for countries suffering severe outbreaks of dengue where the total number of hospitalizations can be a substantial strain on the medical system, as well as a significant economic burden. Up to 500,000 cases of dengue hemorrhagic fever are diagnosed each year, with as many as 25,000 deaths.

Ultimately the solution to dengue fever will require control of the transmission cycle, and this approach will not affect that, notes Prof Stacey.

“However, in the medium term, we certainly still need dengue treatments to reduce disease severity.”

A completely new viral player

The researchers will now move quickly to exploit this latest knowledge, says Professor Young.

“We have identified a completely new viral player in the dengue disease process, opening up the opportunity of developing new classes of inhibitors directed at this new target. The next steps are to complete pre-clinical studies to confirm inhibition of the NS1 activation pathway in ameliorating dengue virus disease and to identify clinical trial partners and sites for testing re-purposed compounds.”

For now, he remains extremely positive about the new opportunities his team’s research will open up in terms of dengue fever treatment.

“I’ve been working in the field of dengue research for more than 30 years and I’ve never been more excited about the prospects for successfully treating dengue disease. This is why I got involved – to make a difference.”

—

Get started improving public health using Dengue Track. Click below!

Dengue Track