Dengue is a leading public health problem with an expanding global burden. Dengue virus is also a significant cause of illness in international travelers with an increasing number of cases of dengue fever identified in travelers returning from dengue-endemic countries.
This review focuses on the clinical illness of dengue infection in international travelers and provides a summary of the risk of infection for travelers, clinical features of infection, and an overview of dengue vaccines and their potential applicability to travelers.
Four prospective studies of travelers to dengue-endemic destinations have shown that the dengue infection incidence ranges from 10.2 to 30 per 1,000 person-months. This varies according to travel destination and duration and season of travel. Dengue is also a common cause of fever in returned travelers, accounting for up to 16% of all febrile illnesses in returned travelers. Although the majority of infections are asymptomatic, a small proportion of travelers develop dengue hemorrhagic fever. The diagnosis of dengue in travelers requires a combination of serological testing for IgG and IgM together with either nucleic acid or NS1 antigen testing. Several vaccine candidates have now entered into clinical trials including ChimeriVax Dengue, which is currently in phase 3 trials, live-attenuated chimeric vaccines (DENV-DENV Chimera, Inviragen), live-attenuated viral vaccines, recombinant protein subunit vaccines, and DNA vaccines.
Dengue infection in international travelers is not infrequent and may be associated with substantial morbidity. Furthermore, an accurate diagnosis of dengue in travelers requires the use of a combination of diagnostic tests. Although a vaccine is not yet available a number of promising candidates are under clinical evaluation. For now travelers should be provided with accurate advice regarding preventive measures when visiting dengue-endemic areas.
Dengue is the most common arthropod-borne viral infection in the world and is endemic in over 100 countries throughout Africa, the Americas, Asia, Eastern Mediterranean, and Western Pacific. Dengue fever (DF) is transmitted by mosquitoes that remain in close proximity to humans and can be caused by any of the four dengue virus serotypes (DENV-1, DENV-2, DENV-3, and DENV-4) belonging to the genus Flavivirus, family Flaviridae. The main vectors for dengue viruses are Aedes aegypti and Aedes albopictus, both of which are widely spread throughout Asia, the Pacific, and the Americas.
There has been a dramatic increase in the number of dengue cases worldwide with an estimated 30-fold increase of documented dengue infections in the last 50 years. Regional surveillance data of dengue infections by the Pan American Health Organization (PAHO) reflects this upward trend in global dengue disease burden (Figure 1). The World Health Organization (WHO) estimates that each year over 100 million cases of DF and 500,000 cases of dengue hemorrhagic fever (DHF) are reported resulting in 25,000 deaths, although it is likely that significant underreporting occurs.[1, 2] However, Bhatt and colleagues have recently shown that each year there are approximately 390 million dengue infections, of which 96 million are clinically apparent. The increased mobility of humans across international borders has contributed to sharp increases in dengue spread. Additionally, the ease by which viremic travelers can introduce dengue into different regions poses significant challenges for public health interventions aimed at reducing the global burden of dengue.
Infection with dengue virus is often asymptomatic or presents as an acute self-limiting febrile illness. Acute dengue may be associated with additional symptoms including rash, headache, retro-orbital pain, myalgia, arthralgia, nausea or vomiting, diarrhea, and hemorrhagic manifestations including petechiae or ecchymoses. Most clinical dengue infections are mild to moderate in severity although a small proportion (1%–3%) of patients develop DHF or dengue shock syndrome (DSS) which are associated with case-fatality rates ranging from <1% to 5%. The greatest burden of severe illness occurs in children and infants in endemic countries.
In 2009, the WHO revised the case definitions for DHF and DSS because severe dengue cases that did not fulfill the criteria were poorly categorized and also because of reported difficulties with the applicability of the existing criteria. The new model distinguishes between dengue and severe dengue and proposes warning signs for disease progression. Its performance is currently being tested in over 18 countries (Table 1).
Table 1. WHO criteria for the classification of dengue fever and severe dengue infection
DSS = dengue shock syndrome; WHO = World Health Organization; ALT = alanine transaminase; AST = aspartate transaminase; HCT = hematocrit; CNS = central nervous system.
Severe plasma leakage
Live in/travel to dengue-endemic area
•Abdominal pain or tenderness
Fever and two of the following criteria:
•Clinical fluid accumulation
•Fluid accumulation with respiratory distress
•Aches and pains
As evaluated by clinician
•Tourniquet test positive
•Liver enlargement >2 cm
Severe organ involvement
•Laboratory: increase in HCT concurrent with rapid decrease in platelet count
•Liver: AST or ALT ≥ 1,000
•Any warning sign
•CNS: Impaired consciousness
•Heart and other organs
Incidence of Dengue in Travelers
Four published prospective studies of travelers to dengue-endemic destinations have been performed over the last decade (Table 2).[9-12] Among 447 short-term (mean travel duration of 1 month) Dutch travelers visiting tropical and sub-tropical destinations, the incidence rate (number of dengue infections per person-months of exposure) of DF was 30 per 1,000 person-months [95% confidence interval (CI) 17.4–51.6]. This translates to an attack rate (number of dengue infections divided by the number of travelers that were exposed) of 2.8%. Season of travel, travel duration, and specific destination were not identified as predictors of infection, possibly because of an insufficient sample size. Only 3 of 13 travelers reported symptoms suggestive of DF, translating into a clinical-to-subclinical rate of 1 : 3.3.
Table 2. Available prospective studies quantifying incidence estimates of dengue in travelers
A second prospective study examined dengue incidence among Israelis who had traveled to Asia, South America, and Africa for 3 to 6 months. Of 104 travelers, 7 (6.7%) had serological evidence of dengue infection, 4 of whom were symptomatic. The dengue incidence in long-term travelers was estimated at 11 infections per 1,000 person-months. The attack rates by region of travel were 1.1% (5/451) for Southeast Asia, 0.6% for South America (1/159), and 4% (1/25) for Africa. Many travelers had been vaccinated for Japanese encephalitis or yellow fever virus, but cross-reactivity with other flaviviruses was not explored, so the actual incidence may have been overestimated.
A third recent incidence-estimate study was performed among a cohort of 387 Australian tourists to Asia. Over two-thirds of participants were short-term travelers (<30 days of travel). The dengue incidence was found to be 10.2 infections per 1,000 person-months (95% CI 2.7–26.1). The four travelers with serological evidence of dengue infection were asymptomatic. The baseline dengue seroprevalence was 4.4%, possibly related to previous travel to Asia by over 70% of participants.
The final study was performed prospectively among short-term Dutch travelers to dengue-endemic areas in Asia, Africa, South and Central America, and the Caribbean. The incidence density of recent dengue infection was 14.6 per 1,000 person-months (95% CI 8.3–23.9), with most infections occurring in travelers to Southeast and Southern Asia, especially during the rainy season. The majority (64%) of infections were asymptomatic. Previous dengue infection was identified in 6.5% and positively correlated with increasing age, previous trips to dengue-endemic countries, being born in a dengue-endemic country, and traveling for the purpose of visiting friends and relatives. The incidence estimates of dengue in travelers by region of travel as determined by the two most recent studies are illustrated in Figure 2.
These four studies demonstrate that a significant risk of dengue infection exists in international travelers to dengue-endemic regions and that the incidence ranges from 10.2 to 30 per 1,000 person-months, varying according to travel destination and duration. This risk estimate is notably higher than the incidence of other travel-related infections such as hepatitis A and typhoid.
Dengue Infection as a Cause of Fever in Returned Travelers
Most cases of dengue infections in febrile travelers are diagnosed in travelers returning from Asia, followed by the Americas then Africa. In a recent large retrospective study of infections in returned travelers, dengue was either as frequent as or more frequent than malaria as a cause of febrile illness in travelers returning from all parts of the world except for sub-Saharan Africa.
Retrospective studies have identified dengue infection as the cause of fever in 2% to 16.5% of febrile returned travelers.[15-19] An Australian study of fever in returned travelers found that 5% of febrile travelers who visited dengue-endemic regions between 1998 and 2004 had dengue infection, with the odds of travel to Asia being five times greater among those with dengue. A study of French travelers hospitalized after returning from tropical destinations (1999–2003) identified dengue as the cause of illness in 2% of cases. Febrile travelers who had visited Asia were again significantly more likely to have dengue infection compared with travelers from other regions. In a study of febrile Swedish travelers who visited malaria-endemic countries between 2005 and 2008, 4% were diagnosed with dengue infection. A more recent study from Italy (conducted in 2009–2010) identified dengue as the cause of fever in 16% of febrile returned travelers, the higher proportion perhaps indicative of the cyclical increase in dengue activity, development of better diagnostic assays, and/or an increasing burden of dengue as a cause of illness in travelers. Increasing case numbers of travel-associated dengue were reported between 2005 and 2010.[21, 22] However, as the overall number of international travelers visiting dengue-endemic countries has also increased, this increase in dengue cases cannot definitely be ascribed to increasing absolute risk of dengue infection.
Clinical Severity in Travelers
The determinants of clinical versus subclinical infection are not clearly defined, but infecting viral strains, cross-reactive immunity from infection with different viral serotypes, and preexisting host immunity may all contribute. The ratio of clinical to subclinical infection identified in studies of local endemic populations has ranged from 1 : 0.2 to 1 : 7.[23-25] Seroprevalence studies of travelers who have spent variable time periods in dengue-endemic countries have demonstrated dengue antibody presence in 8.7% to 19.5% of individuals, many of whom have reported no consistent clinical illness.[26, 27] This finding was also observed in the prospective studies of dengue infections in travelers where 43% to 100% of dengue infections were asymptomatic.[9-12] Traveler surveillance data likely underestimates dengue infections as only clinical infections are reported.[9, 11, 12] This has major implications for dengue naive countries with vectors able to transmit the disease.
Symptomatic Dengue Infection in Travelers
Of the travelers who experience a clinical illness, fever (86%–100%) is the most commonly reported symptom. Other common symptoms are myalgia (42.2%–79%), headache (59.2%–68%), nausea (34%–37%), vomiting (8%–19%), diarrhea (20.4%–37%), abdominal pain (15%), and the development of a rash (29.2%–74%) The laboratory features that are typically seen are thrombocytopenia (<100,000 μL, 52.6%–72%), leukopenia (80%–89.5%), and elevated liver transaminases (ALT, 47%–70%).[28-30]
Hospitalization of Dengue Infected Travelers
The number of dengue-infected travelers requiring hospitalization is increasing, with a recent US study reporting a tripling of hospitalized cases between 2000 and 2007. The increase in both the number of cases in dengue-endemic regions and the number of US travelers to dengue-endemic regions are likely to be contributing factors. Several single center studies have reported that dengue infection is among the four commonest reasons for hospitalization of ill travelers returning from Asia, following malaria and undifferentiated febrile illness, accounting for 13% of all post-travel hospitalizations.[18, 32] In contrast, travelers from Africa were most commonly hospitalized with malaria. Overall, up to 25% of returned travelers require hospitalization and of these average length of the hospital stay was up to 5 days, with a range of 1 to 11 days.[30, 33]
Dengue Hemorrhagic Fever in Travelers
DHF has been described infrequently in travelers and no prospective studies have reported its travel-associated incidence.[28, 34, 35] Heterogeneous risk factors for DHF have been reported but overall a lack of a consensus persists about what places travelers at risk of DHF or DSS.
In endemic countries, DHF previously occurred mainly in children. However, an increasing average age of DHF among children and more frequent diagnosis of DHF among adults are now being reported. Also a relationship between subsequent infection with a different serotype and the development of antibody-mediated immune enhancement has been suggested to be associated with an increased risk of DHF and DSS.[36, 37] This may contribute to more severe dengue infection in preexposed travelers. Several cases of DHF occurring with both primary and secondary dengue infection have been reported in travelers.[33, 38] However, the development of capillary leakage which has been considered to be the hallmark of DHF/DSS has not been shown to be associated with secondary infection or with dengue severity in travelers with dengue.[39, 40]
Retrospective and surveillance studies of febrile returned travelers have described a 0.9% to 3% prevalence of DHF,[30, 41-44] which mirrors the DHF prevalence in endemic populations. However, the proportion of DHF in travelers may be overestimated as individuals experiencing more severe symptoms are more likely to seek medical attention.
Mortality in Travelers
Whilst dengue infection rarely leads to a fatal outcome in travelers, several have been reported in the literature. Primary dengue infections have been identified in some of these fatal cases and causes of death have been attributed to intra-cranial hemorrhages, other hemorrhagic events, and fulminant hepatic failure.[45-48]
Autochthonous Transmission of Dengue
Recent reports of autochthonous transmission of dengue infection in Madeira (Portugal), metropolitan and coastal France, Florida, and Croatia highlight the propensity for its establishment in disease-naive countries that house A albopictus.[49-52] Immediate steps that were taken to reduce the spread of the virus and the risk of an epidemic in metropolitan France included targeted vector control measures, active case finding, and enhanced dengue virus education of health professionals. In response to these cases and the likelihood that concomitant increases in vector density favors secondary transmission of imported dengue viruses, a surveillance program for dengue infections in travelers from dengue-endemic areas was initiated in 2010 in France.[50, 51]
Other measures such as airport screening of febrile travelers in Taiwan did not demonstrate any preventive benefit. Avoidance of mosquito bites whilst in the febrile period, which occurs inadvertently as cases are often hospitalized in the febrile period, is another measure to reduce transmission in areas where the vectors are present. In Northern Queensland, Australia, dengue is not endemic but this region experiences annual epidemics originating from imported dengue by viremic travelers. In 2009, four separate outbreaks of DF occurred, with dengue importations from Southeast Asia (57% of cases) from the South Pacific islands of Samoa, Vanuatu, Fiji, Cook Islands, and Tonga (32%).
Diagnosis of Dengue
Dengue diagnosis relies on serological assays, molecular diagnostics, and antigen detection. Serology using hemagglutination-inhibition, immunofluorescence antibody assays, and enzyme-linked immunosorbent assay (ELISA) tests is widely used but is limited by variable sensitivity and specificity due to cross-reactivity to other flavivirus infections and flavivirus vaccines. In acute dengue infection, serum IgM becomes positive within 4 to 5 days and is followed by development of IgG after day 7 of the illness. However, the sensitivity of commercially available ELISAs for diagnoses of acute dengue varies from 60% to 90% and specificity from 80% to 99%.
Dengue specific tests include RNA detection methods and nonstructural (NS1) antigen assays. Polymerase chain reaction (PCR) assays enable the early identification of dengue RNA during the viremic phase (usually in the first 5 days after onset of illness).[57-60] PCRs have many advantages such as rapidity, serotype specificity (including detection of concurrent infections by different serotypes), ability for quantitative measurements, and high sensitivity (92%–98.5%) and specificity (92.4%–100%).[57-60] More recently, NS1 antigen detection has become widely available. NS1 antigen detection is highly specific for acute dengue (95%–100%) but has variable sensitivity (65%–85%).[61-65] The high specificity of the NS1 antigen means that a negative result does not exclude the diagnosis of dengue; however, a positive test means a very high likelihood of confirming the diagnosis. NS1 antigen testing is more sensitive when used in the first 3 days after fever onset, in patients with primary infection, high-level viremia, DENV-1 infection, and in patients with DF compared with DHF and DSS.[62, 64, 66] The kinetics of the clearance of viremia and NS1 antigenemia from serum is different in primary versus secondary infection, across dengue serotypes, and with disease severity. Both in endemic populations and in travelers, combining NS1 antigen or PCR with IgM/IgG testing can increase the diagnostic yield in acute dengue infection (85%–99%).[61, 66, 68] In one study looking at the performance of NS1 antigen in travelers, the sensitivity of NS1 antigen detection was highest on days 6–7 after the onset of the illness which is different to NS1 antigen detection studies of endemic populations where the sensitivity is reported as highest around day 3 after fever onset.[68, 69]
Distinguishing between primary and secondary dengue infection via serology is difficult, but may be possible by determining the ratio of IgM to IgG. An IgM : IgG ratio of ≥1.78 on day 6 of illness is considered to be consistent with primary infection while a ratio of <1.2 indicates secondary infection.[57, 70] IgG avidity tests in nonvaccinated (flavivirus) individuals can also be used to determine primary versus secondary dengue infections.[71-73]
Development of a dengue vaccine has been hampered by the complex virology of dengue viruses and the essential requirement of producing a balanced immunological response against four serotypes. With live-attenuated viral (LAVs) vaccines, interference between dengue viruses has presented significant problems in developing an effective quadrivalent vaccine. Over-attenuation of viruses is also problematic and results in poor immunogenicity. Balancing adequate viral attenuation to avoid pathogenicity while ensuring retention of immunogenicity and production of long-term immune responses has proven difficult.
Several vaccine candidates have now entered into clinical trials,[74-80] including ChimeriVax Dengue (Sanofi Pasteur, Lyon, France),[74, 81] live-attenuated chimeric vaccines (DENV-DENV Chimera, Inviragen, Fort Collins, CO, USA), LAV vaccines (US National Institutes of Health, NIH, Bethesda, MD, USA), recombinant protein subunit vaccines (Hawaii Biotech, Inc., Aiea, HI, USA), and DNA vaccines (US Naval Medical Research Center, Silver Spring, MD, USA).
The ChimeriVax TDV vaccine is based on four recombinant viruses containing the prM and E genes from each dengue serotype, which have been introduced into the yellow fever virus 17D vaccine backbone. This vaccine is most advanced in clinical development and is highly immunogenic,[83, 84] able to induce broad neutralizing antibodies against dengue 1 to 4, and has been found safe and immunogenic in phase 1 and 2 clinical trials. Administration of three doses of vaccine at 0, 6, and 12 months induces production of high titer neutralizing antibodies against all four dengue serotypes in 77% to 100% of vaccine recipients.[74, 81, 83, 86]
The results of a phase 2b efficacy study have recently been reported. The trial, conducted in northern Thailand and enrolling 4,002 individuals, demonstrated an overall efficacy of 30.2%. The efficacy against DENV1, 3, and 4, was 61.2, 81.9, and 90.0%, respectively, while for DENV2 the efficacy was only 3.5%, in spite of vaccine recipients developing high titer neutralizing antibodies against all serotypes. Reasons for the failure of the DENV2 component are unclear, but the hospitalization rate was nevertheless reduced by 45% in vaccine recipients compared with unvaccinated controls. The vaccine is currently in phase III trials in Asia and Brazil to further evaluate the efficacy of ChimeriVax TDV. These trials have recently completed recruitments and the results will be available in the near future.
A number of LAV vaccines have also entered clinical trials.[75, 76] The National Institute of Allergy and Infectious Disease (NIAID)/Laboratory of Infectious Diseases (LID) has developed a LAV tetravalent vaccine comprised of recombinant dengue viruses containing deletions in the 3′ noncoding region of the virus that reduces the ability of the virus to replicate.[75, 87, 88] This vaccine has entered into phase 1 clinical trials and has been shown to be safe and able to induce production of high titers of neutralizing antibodies. In addition, the LAV vaccine may result in balanced immune responses against all four serotypes.[88, 89] The NIAID/LID vaccine has been licensed to a number of developers including Instituto Butantan in Brazil, Biological E and Panacea Biotec in India, and VaBiotech in Vietnam for further development and for progression to clinical trials to determine vaccine efficacy. Inviragen has also developed a chimeric tetravalent LAV vaccine based on DENV2 and consisting of viruses containing envelope proteins of all four serotypes. This vaccine appears immunogenic and has entered into phase II trials.
An alternative approach to chimeric and LAV vaccines are subunit protein vaccines[76, 77, 90] which avoid the problems of viral interference and use of replicating virus associated with LAV vaccines. This approach may be safer and result in induction of balanced immune responses against all four serotypes. Merck & Co in conjunction with Hawaii Biotech, Inc. have developed a recombinant subunit protein vaccine containing recombinant C-terminally truncated recombinant envelope glycoproteins (DEN-80E) of the four dengue viruses. This vaccine appeared safe and immunogenic in a phase 1 clinical trial and is undergoing further clinical development.
A DNA vaccine (D1ME-VR-P) has also been developed by the US Naval Medical Research Center[76, 78, 90] and has been tested in a phase 1 clinical study. It appears safe and moderately immunogenic with 40% of vaccine recipients developing neutralizing antibody and 83% developing T-cell responses.
Vaccine Prospects for International Travelers
The introduction of dengue vaccines into national immunization programs of dengue-endemic countries with high disease burden takes precedence over a travel vaccine.However, the high incidence of dengue in travelers visiting endemic countries coupled with the role that travelers play in the global spread of dengue argues a place for vaccine availability for travelers in the future. When this becomes available, it will be important to determine which traveler groups are at greatest risk of dengue infection in order to prioritize vaccine recipients. From the studies presented in this review it is reasonable to suggest that travelers visiting Southeast Asia during the rainy season or those who travel frequently to dengue-endemic regions could be most strongly considered for vaccination. Travelers visiting destinations with epidemic dengue activity may also be appropriate for targeted vaccination. Individuals with a prior dengue diagnosis may also be vaccine candidates as this may help reduce the likelihood of a secondary and possibly more severe dengue infection. Dengue vaccination of travelers poses some significant challenges. The vaccine most likely to first become commercially available is ChimeriVax DEN (Sanofi Pasteur). The three-dose schedule over a 12-month period required for ChimeriVax DEN vaccine means that many travelers will be unable to achieve full vaccination prior to departure. The long-term safety of travelers who receive only one or two pre-travel doses is unknown. Although the possibility of severe dengue infection exists, the recently reported efficacy trial in Thailand failed to demonstrate more severe infection in partially immunized vaccine recipients. Vaccines requiring fewer doses or with a shorter dosing schedules would be more suitable for travelers. However, other vaccine candidates have not yet entered phase 3 safety, immunogenicity, or efficacy studies and consequently will not be commercially available for several years. Sustained immunity in travelers and evidence of long-term efficacy of the vaccine with only one or two doses would be ideal features of a dengue vaccine. As with all vaccines, the risk of infection versus the potential negative consequences of vaccination will need to be considered for each individual traveler.
In conclusion, dengue infection in international travelers occurs frequently and may be associated with substantial morbidity. Although there have been relatively few prospective studies these have confirmed that the incidence rate for dengue infection in short-term travelers to dengue-endemic countries is substantial. An accurate diagnosis of dengue in travelers is best done using more than one diagnostic test, and like any infection in endemic countries requires a combination of serological testing for IgG and IgM together with either nucleic acid testing or NS1 antigen testing. A safe and effective vaccine providing effective protection against all four serotypes of DENV is not yet available, although a number of promising candidates are under clinical evaluation. Until a vaccine becomes available, travelers should be provided with accurate advice regarding preventive measures when visiting areas where dengue is endemic, particularly during the rainy season.
Declaration of Interests
The authors state that they have no conflicts of interest to declare.