Tweet
This week on Infection Landscapes we will discuss yellow fever, one of the first diseases identified as being vectored by mosquitoes as well as one of the first recognized hemorrhagic fevers. This post will conclude the extended special series on arthropod-borne infections that began over six months ago.
The Pathogen. Yellow fever is caused by an enveloped, single-stranded RNA virus. It is a Flavivirus, in fact the original Flavivirus, in the Flaviviridae family, which also includes dengue virus, hepatitis C virus, West Nile virus, Japanese encephalitis virus, and the other encephalitide viruses. Here is a picture of the yellow fever virus from the Hardin Library at the University of Iowa:
Flaviruses take their name specifically from the yellow fever virus, which is derived from the Latin word for yellow, flavus. Yellow fever virus invades primarily the macrophages and dendritic cells. Here is a nice graphic that depicts common life cycle elements that are generic to Flaviviruses, including yellow fever virus:
So let's examine this mosquito vector more closely. Here is a map produced by the World Health Organization (WHO) showing the current global distribution of A. aegypti:
This week on Infection Landscapes we will discuss yellow fever, one of the first diseases identified as being vectored by mosquitoes as well as one of the first recognized hemorrhagic fevers. This post will conclude the extended special series on arthropod-borne infections that began over six months ago.
The Pathogen. Yellow fever is caused by an enveloped, single-stranded RNA virus. It is a Flavivirus, in fact the original Flavivirus, in the Flaviviridae family, which also includes dengue virus, hepatitis C virus, West Nile virus, Japanese encephalitis virus, and the other encephalitide viruses. Here is a picture of the yellow fever virus from the Hardin Library at the University of Iowa:
Electron micrograph of yellow fever virions
Flaviruses take their name specifically from the yellow fever virus, which is derived from the Latin word for yellow, flavus. Yellow fever virus invades primarily the macrophages and dendritic cells. Here is a nice graphic that depicts common life cycle elements that are generic to Flaviviruses, including yellow fever virus:
The Vector and the Landscape: Aedes aegypti mosquitoes are the most important vectors for the human transmission of yellow fever, which, you will recall, was also the case for dengue fever. Several other Aedes and Haemagogus mosquito species are also relevant for the transmission of yellow fever, but Aedes aegypti is particularly important because of its adapted ecology to the human domestic environment. Even though they share a primary vector, the transmission of yellow fever is different from dengue fever in that there are three quite distinct cycles of disease transmission. Because several vector species engage these different cycles, yellow fever reflects different disease ecologies and thus a unique landscape epidemiology for each of the transmission pathways. Dengue fever remains (for the most part) bound to an urban transmission cycle, which narrows the ecology of transmission considerably.
Historically, A. aegypti became famous because of its transmission of yellow fever. The realization that A. aegypti was a major problem with respect to yellow fever came from Cuba. In fact, since yellow fever was the first infectious disease identified as being transmitted by way of mosquito, the mosquito hypothesis of disease transmission, and thus vector-borne disease transmission, was established through work done in Cuba. Often, Walter Reed is credited with the work in Cuba that established the vector relationship and led to the elimination of much of the yellow fever endemic there in the late 1800s. However, it was the Cuban doctor, Carlos Finlay, who identified and proposed the mosquito as the vector for yellow fever in 1881, and which Walter Reed was able to mobilize his significant United States military resources to test extensively. Walter Reed, nevertheless, did assign the credit for the discovery properly to Dr. Finlay. But Western history remembers it differently, probably because Western history was not written by the Cubans.
Carlos Finlay
So let's examine this mosquito vector more closely. Here is a map produced by the World Health Organization (WHO) showing the current global distribution of A. aegypti:
Red+Blue: A. aegypti is present
And here is a close-up picture of this vector:
To begin, we must emphasize that this mosquito has a very particular preference for the water environment it selects for laying its eggs. It likes SMALL containers that collect rainwater. And the mechanics work as follows. This mosquito does not lay its eggs either in the water or on the surface of the water, as most other species do. Instead, A. aegypti lays its eggs above the water on the interior wall of the vessel containing the water so that when the water vessel is refilled, from the water line at which the mosquito laid its eggs to the lip of the vessel, the eggs will have enough time to complete their developmental cycle to adulthood before evaporation depletes the water source. A truly incredible evolutionary adaptation.
Here are the stages of the mosquito life cycle (picture by Hopp MJ and Foley J. Global-scale Relationships Between Climate and the Dengue Fever Vector Aedes Aegypti. Climate Change. 2001; 48: 441-463):
Aedes aegypti
To begin, we must emphasize that this mosquito has a very particular preference for the water environment it selects for laying its eggs. It likes SMALL containers that collect rainwater. And the mechanics work as follows. This mosquito does not lay its eggs either in the water or on the surface of the water, as most other species do. Instead, A. aegypti lays its eggs above the water on the interior wall of the vessel containing the water so that when the water vessel is refilled, from the water line at which the mosquito laid its eggs to the lip of the vessel, the eggs will have enough time to complete their developmental cycle to adulthood before evaporation depletes the water source. A truly incredible evolutionary adaptation.
Here are the stages of the mosquito life cycle (picture by Hopp MJ and Foley J. Global-scale Relationships Between Climate and the Dengue Fever Vector Aedes Aegypti. Climate Change. 2001; 48: 441-463):
In this picture below taken by the Centers for Disease Control and Prevention (CDC), notice the ovipositioning that leaves a ring of eggs just above what was, at one point, the water line:
This mosquito is originally adapted to a forest habitat wherein it would seek out holes in trees that would regularly collect rainwater. Tree holes are much more ubiquitous than you might think in a forest (think woodpeckers), and so this was quite an effective niche for this mosquito. As humans encroached more and more on forest habitat establishing agriculture, and building increasingly dense communities and living conditions, A. aegypti readily adapted to the new circumstances. The mosquitoes found an abundance of new and highly effective small containers strewn in and around households that can easily collect, or are intended to store, water. The mass production of plastics has been a major factor in the proliferation of potential water containers. Today A. aegypti is just as much an urban mosquito as it is a forest mosquito and probably more so. As such, A. aegypti is now uniquely adapted to the human environment. Unlike other mosquito species, they will often live in the household with humans, and can complete their whole life cycle here. They also bite during the day, so they have unlimited access to humans for taking blood meals. And finally, this mosquito's preferred host, as you may have already guessed, is humans.Because this mosquito is so very effective at exploiting the human environment, it is also very effective at transmitting any viruses that it is capable of carrying and which are infective to humans.
So, what of the landscape of yellow fever? The reality is that the ecology of yellow fever is more complex than that of the other virus transmitted by A. aegypti that we covered previously (i.e. dengue fever).
Yellow fever has, in fact, three distinct cycles of transmission. All are mosquito-borne, but the ecology and landscape epidemiology of each is unique. Importantly, humans, themselves, often serve as the bridges between these different transmission cycles.
The urban cycle is most clearly delineated by the human environment and, thus, primarily involves A. aegypti mosquitoes since these are some of the most anthropophilic mosquitoes known. Humans are the reservoir for this cycle, but since mosquitoes can transmit the virus transovarially they may also serve as a reservoir.
The sylvan cycle (i.e. jungle cycle) involves two genera of mosquitoes: Aedes and Haemagogus. These mosquitoes transmit the virus primarily to non-human primates, which serve as the reservoir. Other mammals may be infected in this cycle but are not likely to be important for the maintenance of the viral ecology. These two cycles, the urban and the sylvan, occur in both South America and Africa and so are relevant transmission cycles in both macro-regions. The sylvan cycle in Africa requires A. africanus to maintain the viral ecology. However, the sylvan cycle in South America is the only transmission cycle that does not involve aedine mosquitoes.
Haemagogus mosquitoes are the critical vectors in the South American sylvan cycle of yellow fever. They are distinctly tropical forest mosquitoes, typically living out their lives in the forest canopy. Having adapted to a somewhat narrower ecologic niche, these mosquitoes oviposition in tree holes, the crevices of tree bark, or within exposed bamboo stalks. The mosquitoes deposit their eggs directly on the surface where the eggs, similar to the aedine mosquitoes, will develop once they are immersed in water following the next rain. As mentioned above these mosquitoes transmit the virus primarily to non-human primates, especially those that occupy the canopy. Humans are at risk of infection in this transmission cycle when they come into contact with this forested habitat, particularly when that contact actually disrupts the habitat. This scenario arises with deforestation and the development of natural habitat for agricultural or resource extraction purposes. Transmission to humans via this cycle is typically limited to workers in industries that encounter this forested habitat, where sporadic cases, or occasionally, small-scale outbreaks are the norm.
In the African sylvan cycle, the analogue to Haemagogus mosquitoes is Aedes africanus. This mosquito has a very similar ecology to the South American Haemagogus species. It lives in the forest canopy and takes blood meals primarily from non-human primates, which serve as the reservoir for the yellow fever virus. The ovipositioning of A. africanus also mimics that of the Haemagogus mosquitoes. Transmission to humans in the African sylvan cycle also affects workers involved in deforestation, but is not limited to this population. Unfortunately, conflict in certain regions of sub-Saharan Africa where sylvan yellow fever is endemic exposes refugees and displaced communities to this cycle, and has ultimately resulted in more severe outbreaks in this setting than has been experienced in South America.
The intermediate cycle (i.e. the savannah cycle, or the rural cycle) occurs only in Africa and involves several different species of Aedes mosquitoes in the transmission of the virus. As the name suggests, this cycle occurs in the landscape between the strictly forest, or sylvan, cycle and the strictly domestic, or urban, cycle. As such, the Aedes spp. involved typically obtain blood meals from both humans and monkeys. It is essentially an ecotonal cycle, wherein the geography of transmission is determined by landscapes of transition from one habitat to another.
The graph below depicts the relevant mosquito vectors for the different transmission cycles occurring in Africa and South America (published in Annu Rev Entomol. 2007. 52:209-29)
We can see from this presentation that the occurrence of yellow fever is not geographically uniform. Rather, it results from transmission processes that occur in response to distinct, yet sometimes overlapping, vector ecologies, which themselves are geographically demarcated by unique features of the landscape. The ways in which humans encounter and navigate such landscapes will often determine the nature of disease endemicity.
Yellow fever has, in fact, three distinct cycles of transmission. All are mosquito-borne, but the ecology and landscape epidemiology of each is unique. Importantly, humans, themselves, often serve as the bridges between these different transmission cycles.
The urban cycle is most clearly delineated by the human environment and, thus, primarily involves A. aegypti mosquitoes since these are some of the most anthropophilic mosquitoes known. Humans are the reservoir for this cycle, but since mosquitoes can transmit the virus transovarially they may also serve as a reservoir.
The sylvan cycle (i.e. jungle cycle) involves two genera of mosquitoes: Aedes and Haemagogus. These mosquitoes transmit the virus primarily to non-human primates, which serve as the reservoir. Other mammals may be infected in this cycle but are not likely to be important for the maintenance of the viral ecology. These two cycles, the urban and the sylvan, occur in both South America and Africa and so are relevant transmission cycles in both macro-regions. The sylvan cycle in Africa requires A. africanus to maintain the viral ecology. However, the sylvan cycle in South America is the only transmission cycle that does not involve aedine mosquitoes.
Haemagogus mosquitoes are the critical vectors in the South American sylvan cycle of yellow fever. They are distinctly tropical forest mosquitoes, typically living out their lives in the forest canopy. Having adapted to a somewhat narrower ecologic niche, these mosquitoes oviposition in tree holes, the crevices of tree bark, or within exposed bamboo stalks. The mosquitoes deposit their eggs directly on the surface where the eggs, similar to the aedine mosquitoes, will develop once they are immersed in water following the next rain. As mentioned above these mosquitoes transmit the virus primarily to non-human primates, especially those that occupy the canopy. Humans are at risk of infection in this transmission cycle when they come into contact with this forested habitat, particularly when that contact actually disrupts the habitat. This scenario arises with deforestation and the development of natural habitat for agricultural or resource extraction purposes. Transmission to humans via this cycle is typically limited to workers in industries that encounter this forested habitat, where sporadic cases, or occasionally, small-scale outbreaks are the norm.
Bridge between sylvan and urban transmission cycles in South American yellow fever
In the African sylvan cycle, the analogue to Haemagogus mosquitoes is Aedes africanus. This mosquito has a very similar ecology to the South American Haemagogus species. It lives in the forest canopy and takes blood meals primarily from non-human primates, which serve as the reservoir for the yellow fever virus. The ovipositioning of A. africanus also mimics that of the Haemagogus mosquitoes. Transmission to humans in the African sylvan cycle also affects workers involved in deforestation, but is not limited to this population. Unfortunately, conflict in certain regions of sub-Saharan Africa where sylvan yellow fever is endemic exposes refugees and displaced communities to this cycle, and has ultimately resulted in more severe outbreaks in this setting than has been experienced in South America.
Bridges between sylvan, intermediate, and urban transmission cycles in African yellow fever
The graph below depicts the relevant mosquito vectors for the different transmission cycles occurring in Africa and South America (published in Annu Rev Entomol. 2007. 52:209-29)
We can see from this presentation that the occurrence of yellow fever is not geographically uniform. Rather, it results from transmission processes that occur in response to distinct, yet sometimes overlapping, vector ecologies, which themselves are geographically demarcated by unique features of the landscape. The ways in which humans encounter and navigate such landscapes will often determine the nature of disease endemicity.
The Disease: As with many other arboviruses and/or hemorrhagic fevers, yellow fever manifests as a range of clinical disease rather than one distinct condition. In areas where the virus is endemic, some infections may be entirely asymptomatic; the opposite end of the spectrum constitutes a severe, life-threatening illness. Overall, however, the case-fatality is high: approximately 20% among populations where the disease is endemic, and possibly as high as 90% among persons from non-endemic areas (typically travelers). Under-reporting of subclinical or mild disease undoubtedly inflates these numbers somewhat, but it is, nevertheless, an extremely serious infection.
There are typically three distinct phases of the clinical presentation of yellow fever:
Phase 1 corresponds to the host's initial viremia and typically presents with nausea and vomiting, fever, headache, dizziness, and myalgia. Leukopenia may be identifiable at the beginning of the acute symptoms, while liver enzymes typically increase a few days later. The viral load will typically reach its maximum 2 to 3 days after the initial infection. Importantly, those cases that end in fatality will often have a higher and extended duration of viremia relative to non-fatal cases.
Phase 2 is known as the remission, which, as the name suggests, can be identified by a break in the fever and relief from the clinical symptoms. The remission can last up to 2 days, but is not a necessary part of the natural history of disease. Some persons do not experience any remission in their clinical course. On the other hand, some infected persons will recover entirely during this remission and are referred to as abortive infections. These persons also do not demonstrate the jaundice associated with the severe third phase, from which the disease and the virus derive their names.
Phase 3 is referred to as the intoxication phase, and is the period of greatest clinical severity. Approximately 15% of infected individuals enter the intoxication period, when more advanced liver dysfunction begins to manifest. Jaundice is typically present, along with the return of nausea, vomiting and fever. Due to the associated coagulopathy, bleeding diathesis is also common at this stage. Multiple organ failure is common, with kidney, heart and neurologic involvement, depending on the extent of liver damage. As mentioned above, the case-fatality is 20% among endemic populations.
Epidemiology Summary and Control: Roughly 200,000 cases of yellow fever occur each year, with about 30,000 associated deaths. Here is a map produced by the World Health Organization (WHO) showing those geographic areas at risk of yellow fever virus transmission and those areas with reported outbreaks:
The continent of Africa experiences 90% of these cases, with the rest occurring in South America and Panama. Dramatic epidemics involving up to 100,000 cases each have been documented in the tropical countries of Africa. Outbreaks have also been identified in the Americas, but these are much less common in the western hemisphere in modern times. Historically, however, large-scale epidemics were experienced throughout all of the Americas on a regular basis. Here are two more WHO maps showing where yellow fever has occurred during the last half of the twentieth century. Also shown are the lines demarcating current endemicity in Africa and the Americas:
There was a point in the history of yellow fever when great gains were achieved toward its control, at least in the western hemisphere. Following the recognition of A. aegypti in transmitting yellow fever, a massive campaign was undertaken to eliminate the mosquito from the surrounding tropical forests during the building of the Panama Canal, with modest effect. However, in 1947 the Pan-American Health Organization (PAHO) began a massive program to eliminate this mosquito from the whole of the Americas. And this time the campaign was extraordinarily successful. By 1972, A. aegypti had been eliminated from 73% of the habitat that it occupied prior to 1947,primarily without the use of insecticides. However, when it was discovered that a jungle cycle of yellow fever existed that could not be eliminated by the campaign, tragically the intensive program was abandoned. By the close of the twentieth century, A. aegypti had newly colonized more geographic area than it had prior to 1947, and occupied a much greater global distribution as well. This has important implications not only for yellow fever, but also for dengue fever. This map published by Emerging Infectious Diseases tells the story best:
The large disparity in yellow fever incidence and mortality that currently exists between Africa and the Americas is largely due to the massive vaccinations campaigns that have been undertaken in many countries in South America, which have thus eliminated the urban cycle of disease in the western hemisphere. In many endemic African countries the same resources have not been available to mobilize widespread vaccination. This lack of vaccination, coupled with the added intermediate transmission cycle, have been important contributors to the much higher disease burden we see in Africa today.
This concludes the extended series on arthropod-borne infections. The series was by no means exhaustive and I will, of course, return to these and new arthropod-borne infections in future on an individual basis. Nevertheless, I think this series has served as a good general survey of many of the important arthropod-borne infections.
Next time at Infection Landscapes I will cover measles. Stay tuned.
There are typically three distinct phases of the clinical presentation of yellow fever:
Phase 1 corresponds to the host's initial viremia and typically presents with nausea and vomiting, fever, headache, dizziness, and myalgia. Leukopenia may be identifiable at the beginning of the acute symptoms, while liver enzymes typically increase a few days later. The viral load will typically reach its maximum 2 to 3 days after the initial infection. Importantly, those cases that end in fatality will often have a higher and extended duration of viremia relative to non-fatal cases.
Phase 2 is known as the remission, which, as the name suggests, can be identified by a break in the fever and relief from the clinical symptoms. The remission can last up to 2 days, but is not a necessary part of the natural history of disease. Some persons do not experience any remission in their clinical course. On the other hand, some infected persons will recover entirely during this remission and are referred to as abortive infections. These persons also do not demonstrate the jaundice associated with the severe third phase, from which the disease and the virus derive their names.
Phase 3 is referred to as the intoxication phase, and is the period of greatest clinical severity. Approximately 15% of infected individuals enter the intoxication period, when more advanced liver dysfunction begins to manifest. Jaundice is typically present, along with the return of nausea, vomiting and fever. Due to the associated coagulopathy, bleeding diathesis is also common at this stage. Multiple organ failure is common, with kidney, heart and neurologic involvement, depending on the extent of liver damage. As mentioned above, the case-fatality is 20% among endemic populations.
Epidemiology Summary and Control: Roughly 200,000 cases of yellow fever occur each year, with about 30,000 associated deaths. Here is a map produced by the World Health Organization (WHO) showing those geographic areas at risk of yellow fever virus transmission and those areas with reported outbreaks:
Countries at risk of yellow fever and countries that have reported at least one outbreak of yellow fever, 1985-1999
The continent of Africa experiences 90% of these cases, with the rest occurring in South America and Panama. Dramatic epidemics involving up to 100,000 cases each have been documented in the tropical countries of Africa. Outbreaks have also been identified in the Americas, but these are much less common in the western hemisphere in modern times. Historically, however, large-scale epidemics were experienced throughout all of the Americas on a regular basis. Here are two more WHO maps showing where yellow fever has occurred during the last half of the twentieth century. Also shown are the lines demarcating current endemicity in Africa and the Americas:
There was a point in the history of yellow fever when great gains were achieved toward its control, at least in the western hemisphere. Following the recognition of A. aegypti in transmitting yellow fever, a massive campaign was undertaken to eliminate the mosquito from the surrounding tropical forests during the building of the Panama Canal, with modest effect. However, in 1947 the Pan-American Health Organization (PAHO) began a massive program to eliminate this mosquito from the whole of the Americas. And this time the campaign was extraordinarily successful. By 1972, A. aegypti had been eliminated from 73% of the habitat that it occupied prior to 1947,primarily without the use of insecticides. However, when it was discovered that a jungle cycle of yellow fever existed that could not be eliminated by the campaign, tragically the intensive program was abandoned. By the close of the twentieth century, A. aegypti had newly colonized more geographic area than it had prior to 1947, and occupied a much greater global distribution as well. This has important implications not only for yellow fever, but also for dengue fever. This map published by Emerging Infectious Diseases tells the story best:
Shaded areas indicate presence of A. aegypti
And here again is the map demonstrating the current global distribution of this mosquito:
Red+Blue: A. aegypti is present
The PAHO program was so effective in eliminating A. aegypti because it utilized a campaign to eliminate all open water containers in and around the household in every community.While this was an extremely labor intensive public health intervention, it was simple to employ and it only needed to target places of human habitation. However, the identification of a jungle cycle that could still make yellow fever viable resulted in the abandonment of the intensive campaign, and, thus, the resurgence of A. aegypti followed. Because of massive and effective yellow fever vaccination campaigns in the Americas, the consequences of this resurgence in the western hemisphere are far more relevant to the emergence of epidemic dengue fever than they are for epidemics of yellow fever.
This concludes the extended series on arthropod-borne infections. The series was by no means exhaustive and I will, of course, return to these and new arthropod-borne infections in future on an individual basis. Nevertheless, I think this series has served as a good general survey of many of the important arthropod-borne infections.
Next time at Infection Landscapes I will cover measles. Stay tuned.