Review articles Yellow fever and dengue: a threat to Europe?

The introduction and rapidly expanding range of Aedes albopictus in Europe is an iconic example of the growing risk of the globalization of vectors and vector-borne diseases. The history of yellow fever and dengue in temperate regions confirms that transmission of both diseases could recur, particularly if Ae. aegypti, a more effective vector, were to be re-introduced. The article is a broad overview of the natural history and epidemiology of both diseases in the context of these risks.


Background
There is logic in dealing with yellow fever and dengue together, for they have much in common: • Both are caused by viruses of the family Flaviviridae, genus Flavivirus.• Both viruses are strictly primatophilic -they only infect primates, including man.• In their original habitat, both are zoonotic infections transmitted by forest-dwelling mosquitoes.• Both can cause haemorrhagic illness in humans, often with fatal consequences.• Both owe their importance as human pathogens to two forest mosquitoes that have become closely associated with the peridomestic environment.• The viruses and their urban vectors owe their worldwide distribution to transportation of goods and people.• Both diseases have a history of transmission in temperate regions, including Europe.
According to the World Health Organization, there are currently 200,000 worldwide cases and 30,000 deaths from yellow fever per year, 90% of them in Africa [1], and as many as 50 million cases of dengue [2].
Epidemics of yellow fever, sometimes catastrophic, were once common in North America as far north as New York and Boston (Table ), and in European ports as far north as Cardiff and Dublin [3].Large epidemics of dengue occurred in the same regions from the 18th century onwards.A massive epidemic, estimated at one million cases, with at least 1,000 deaths, occurred in Greece in 1927-28 [4,5] Aedes aegypti, the primary urban vector for both viruses, was once established as far north in Europe as Brest and Odessa (Figure 1).It disappeared from the entire Mediterranean region in the mid-20th century, for reasons that are not clear.Ae. albopictus, generally regarded as a less important vector of dengue [7], is also capable of transmitting yellow fever.It was introduced to Europe in the 1970s, is well established in at least twelve countries (Figure 2) [8], and is likely to spread northwards, perhaps as far as Scandinavia.
The number of persons who visit countries endemic for dengue and yellow fever is continually rising [11,12].It is therefore cogent to consider whether introduction of these viruses is likely to lead to autochthonous and even endemic transmission in Europe.

Transmission
Five factors are key to the epidemiology of vector-borne diseases: the ecology and behaviour of the host, the ecology and behaviour of the vectors, and the degree of immunity in the population.A holistic view of this complexity is key to assessing the likelihood of transmission in Europe [13].

Origin of the viruses
There is little doubt that the yellow fever virus (YFV) originated in Africa, and that viruses circulating in the New World are of African origin.Curiously, yellow fever has never been recorded in Asia, although Ae. aegypti is widespread there.
There are four antigenically distinct DENV serotypes that cause very similar disease in humans.It is widely accepted that all four are of Asian origin [14], although DENV-2 is enzootic in Africa [15].

Zoonotic vectors and hosts
In the Old World, the sylvatic vectors of yellow fever and dengue are canopy-dwelling mosquitoes of the genus Aedes and three subgenera, Stegomyia, Finlaya, and Diceromyia, that feed exclusively on monkeys.In the Americas, the principal zoonotic vectors of yellow fever are Sabethes and Haemogogus species; both are also strictly primatophilic [3].

Sylvatic transmission to humans
Sylvatic infections are acquired when humans enter woodland where there is zoonotic transmission.In recent years, a number of unvaccinated tourists have died of yellow fever after visiting enzootic areas [16,17].

Vector-host specificity
Host specificity is a characteristic of many vectors; it is conceivable that it improves the chances of locating hosts.This may be particularly useful in the sylvatic environment, where bands of monkeys roam between established sleeping sites.
The specificity of DENV and YFV to primatophilic vectors may have evolved to exploit this relationship, and/or to surmount barriers to infection in the insect.
Whatever the reason, given the absence of wild primates, it is unlikely that any vector species native to Europe is able to transmit these viruses.

Peridomestic transmission
Neither Dengue is endemic in many urban and rural populations throughout the tropics.'Virgin soil' epidemics in large cities are often explosive.In 1988, for example, there were an estimated 420,000 cases in four months in the coastal city of Guayaquil, Ecuador [18] The large urban outbreaks of yellow fever that were common until the early 20th century remain a real and constant danger in enzootic countries that do not enforce routine vaccination.Moreover, it is reasonable to assume that areas that are prone to dengue transmission are equally prone to yellow fever, so areas without history of the latter, including those in southeast Asia, may well be at risk.

Vectors The yellow fever mosquito, Aedes aegypti
Ae. aegypti is the quintessential urban vector of yellow fever and dengue.It is a remarkable species because Source: [10].

Figure 1 Historical distribution of Aedes aegypti
Dark grey areas: maximum range distribution of Ae. aegypti, black lines: January 10°C isotherm in the northern hemisphere; mid grey lines: the July 10°C isotherm in the southern hemisphere.
The distribution limit broadly fits the 10°C isotherm in the southern hemisphere, but far less so in the northern hemisphere.Source: adapted from a map published by Christophers [9].
the 'domesticated' form is rarely found more than 100 m from human habitation and feeds almost exclusively on human blood.Nevertheless, like its forest ancestor, it remains day-active with a preference for heavy shade.It freely enters homes and other buildings and spends much of its time hidden in dark places, often among clothing, a stable microclimate with few predators.Its human host is abundant and lives under the same roof, an arrangement that minimises the hazards of questing for a blood meal.It lays eggs in man-made objects that contain water, from discarded tires and buckets to the saucers under flowerpots and waterstorage barrels.In short, humans are the perfect host: they provide safe shelter, plentiful food and abundant sites for procreation.Indeed, in most cities of the tropics, homes are so close together and breeding sites so abundant that they can be regarded as a single factory for mosquitoes in an urban jungle.In the past three decades, attempts to reduce populations of the species have rarely been successful and never sustained [19,20].

The Asian Tiger mosquito, Aedes albopictus
Ae. albopictus is often abundant in the peridomestic environment, particularly in areas with plentiful vegetation.However, in addition to humans, it feeds freely on animals and birds, and so can exist far from human habitation.Since non-primates are not susceptible to the viruses, such blood meals do not contribute to the transmission cycle, and for this reason, Ae. albopictus has generally been regarded as a secondary vector [7].Nevertheless, dengue epidemics have been recorded in places where Ae. albopictus is the only vector [21], and in recent years, the species has proved highly effective in urban transmission of another African sylvatic virus, chikungunya virus [22,23].

Aedes aegypti
Ae. aegypti and yellow fever arrived in the New World together, as passengers in the slave trade.Slave ships generally made the passage from Africa to the Americas in four to six weeks.The virus was enzootic in regions where the slave caravans captured local inhabitants, and urban transmission was rife in the ports of dispatch.The casks used for shipboard storage of water must have been prolific breeding sites for the mosquito, and the slaves were an abundant source of blood.With the slaves and the mosquito came the virus, and it was not uncommon for ships to arrive in port with large numbers of dying persons aboard, hence the yellow flag of quarantine.
In the United States, the species has been recorded from 21 states (Alabama, Arkansas, Florida, District of Colombia, Georgia, Illinois, Indiana, Kansas, Kentucky, Louisiana, Maryland, Missouri, Mississippi, New York, North Carolina, Ohio, Oklahoma, South Carolina, Tennessee, Texas, and Virginia) [24].In many of these, winter temperatures below -20°C are not unusual.
Presumably the mosquitoes survive in sheltered sites, for they are not resistant to freezing.Thus there is no obvious climatic reason why the species, were it to be re-introduced, could not survive in most areas in Europe.

Aedes albopictus
In its original range, Ae. albopictus was present from Beijing and northern Japan to tropical Asia [25].In 1983, however, the mosquito was found in Memphis, Tennessee [26], and, two years later, a survey revealed that it was widely distributed, often common, in the southern United States.Investigation revealed a global trade in used tyres that were frequently infested with eggs and larvae of the species [27].Japan was the principal exporter, and a study of winter diapause at various latitudes in Asia confirmed that the day-length that triggered diapause was identical in the southern United States and in southern Japan [28].The mosquito is now widespread in the United States, and is a major nuisance species as far north as Nebraska and Illinois, where winter snowfall can be well above 200 cm, average January night-time temperatures are -10ºC, and temperatures as low as -33ºC have been recorded.It is also established in Mexico and all the countries of Central and South America except Chile.In Africa it is well established in Nigeria, Gabon, Equatorial Guinea and Cameroon [29,30], and in Europe it has been reported from 16 countries [8].Recent infestations in the Netherlands have been traced to imports of 'lucky bamboo' from sub-tropical China [31], but these mosquitoes do not appear to have survived the winter, perhaps because they have no winter diapause.

Yellow fever
As with most viral diseases, yellow fever can present with a wide spectrum of symptoms, from mild to fatal.In clinical cases, there is generally a sudden onset of fever with severe headache, arthralgias, and myalgia.The striking yellowing of the eyes and skin, caused by hepatic dysfunction, may appear on the third day and indicates a poor prognosis.The fever often follows a 'saddleback' curve, with a brief drop in temperature and symptoms after the third day, followed by a return with increased severity that can lead to spontaneous haemorrhage ('coffee ground' vomit), delirium, renal failure, coma and death.Fatality rates of clinical cases can be as high as 80% [3], on a par with Ebola, Marburg and other haemorrhagic viral infections.

Dengue
As many as 80% of all dengue infections are asymptomatic.Among clinical cases, early stages are similar to those of yellow fever, although with excruciating arthralgia and myalgia, hence the term 'break-bone fever'.Fever and other symptoms rarely last more than seven days, but convalescence can be prolonged and debilitating.The later stages of the illness often include a widespread rash [32].
A portion of dengue cases, usually less than 5%, can be severe and a fraction of these may be fatal [33].Severe dengue, commonly referred to as dengue haemorrhagic fever/dengue shock syndrome (DHF/DSS) to distinguish it from 'classic' dengue, is associated with spontaneous haemorrhage and an increase of vascular permeability that can lead to life-threatening hypovolemic shock.The causes of this condition have been debated for decades, but remain unresolved [34][35][36].A widely held but hotly contested hypothesis is that after infection with one serotype, secondary infections by one or more of the others can precipitate the syndrome by a process referred to as antibody-dependent enhancement, but the occurrence of severe dengue in epidemics of primary infection, such as the Greek epidemic and a recent epidemic in Cape Verde [37], contradicts this hypothesis.An associated controversy is the validity of graded sets of criteria to categorise severity that are recommended by the World Health Organization, and these have been revised several times in recent years [38].Both issues are of prime importance for the management and treatment of patients.
It is a common misconception that DHF/DSS first appeared in the 1950s in south-east Asia.It is certainly true that the syndrome became a serious public health problem in that period, but it was not a new phenomenon: significant mortality associated with haemorrhagic symptoms had been described in the earliest epidemic of dengue-like disease on record, in Philadelphia in 1789, as well as in later epidemics in East Africa and in Australia [14,39].Moreover, as already mentioned, at least 1,000 people died in the Greek epidemic in 1927-28.In the years after the Second World War, however, rapid expansion of densely populated urban areas, coupled with enormous infestations of Ae. aegypti, led to a massive increase in the prevalence and incidence of the disease in south-east Asia, so a plausible explanation for the emergence of this 'new' syndrome is that escalating numbers of classic infections simply led to an increased awareness of the relatively rare manifestations -the 'iceberg effect'.

Treatment
There is no specific treatment for yellow fever or dengue virus infections; supportive therapy is the only option, although there is active research into antiviral drugs against these diseases [40].For dengue fevers, intravenous fluids are used to counter haemoconcentration, and platelet transfusions in the event of severe thrombocytopaenia [41].Strict avoidance of anticoagulants, including aspirin, is important.

Prevention Vaccination
Yellow fever A safe, effective yellow fever vaccine, based on a live attenuated strain, has been available for more than half a century, and mass vaccination is a highly effective approach to prevent urban transmission, but the incidence of the disease, particularly in Africa, confirms that coverage is inadequate, and there is a real and present danger of a major urban epidemic.Moreover, there is good reason to believe that the 2.5 billion people who live in regions at risk of dengue infection are also at risk of yellow fever; if so, then, given the lax attitude towards vaccination of travellers in most countries, the danger of a catastrophic epidemic beyond regions generally associated with transmission is also real, and this could include parts of Europe infested with Ae. albopictus.If such an event were to occur, current stocks of vaccine would probably be inadequate to respond to worldwide demand.

Dengue
No vaccine against dengue is available, but attenuated virus vaccines and second-generation recombinant vaccines are in active development [42].A large-scale trial (phase IIb) of a chimeric tetravalent vaccine [43] has been under way since February 2009 [44].If successful, then a vaccine might be licenced within five years.

Vector control
At the beginning of the 20th century, urban yellow fever was eliminated from many countries by energetic campaigns to eliminate Ae. aegypti breeding sites.After the Second World War, focal application of the synthetic pesticide dichlorodiphenyltrichloroethane (DDT) to infested containers and their surroundings was an outstanding success; according to the Pan American Health Organization, the species was eradicated from 22 countries in the Americas [45].The reason for the efficacy of this method has only recently become apparent: 'skip-oviposition' (the deposition of small numbers of eggs in many different sites) made it highly probable that they would encounter treated sites [19].
No substitute for DDT is currently available, so many authorities resort to spraying insecticidal aerosols (ultra-low-volume) of organophosphates or pyrethroids from hand-held machines, road vehicles or aircraft.Unfortunately, the method is expensive and generally ineffective, at least against Ae.aegypti, because the species spends much of its time indoors at sites that are inaccessible to the aerosol [20,46].Moreover, even if a large number of mosquitoes were to be eliminated by this treatment, the impact on adult mosquito populations would probably be too short for an effective impact on transmission [47].Although the World Health Organization recommends that health authorities evaluate the technique under local circumstances [6], their principal strategy is community-based source reduction, the elimination of breeding sites by the community.Unfortunately, there is no evidence that this approach has been successful in any part of the world.
Control of Ae. albopictus is probably even more difficult than for Ae.aegypti, given its ability to breed away from human habitation, but insecticidal aerosols may be more effective for Ae.albopictus because the mosquito tends to rest outdoors.

The future in Europe
Dengue is essentially an urban disease because of the urban ecology of its vectors and the behaviour of its hosts.Rapid urbanisation has made it an increasingly serious public health problem in the tropics [48].Millions of people travel from the tropics to Europe and North America each year (for example, 1.2 million people who live in the UK visit the Indian subcontinent, with average stays of 29 days) and, after malaria, dengue infection is the second most frequent reason for hospitalisation after their return [11,12].
The history of dengue and yellow fever in Europe is evidence that conditions are already suitable for transmission.The establishment of Ae. albopictus has made this possible, and the possibility will increase as the species expands northwards, or if Ae. aegypti is re-established.The epidemic of chikungunya in northern Italy in 2007 [8,49] confirms that Ae. albopictus is capable of supporting epidemic transmission, although laboratory studies indicate that the strain of virus involved was particularly adapted to this species [50,51].Nevertheless, it is not unreasonable to assume that climatic conditions that permit malaria transmission will also support transmission of yellow fever and dengue, in which case transmission could extend into northern Europe [52].
Lastly, it is widely stated that the incidence of vectorborne diseases will increase if global temperatures increase.While there is no doubt that temperature and rainfall play a role in their transmission, it is clear that many other factors are involved [6].A more urgent emerging problem is the quantum leap in the mobility of vectors and pathogens that has taken place in the past four decades, a direct result of the revolution of transport technologies and global travel [53].The potential impact of this globalisation of vector-borne diseases is a challenge for the future.

Figure 2
Figure 2 Current (2009) distribution of Aedes albopictus in Europe by administrative unit