Dengue

Summary

Dengue, caused by four closely related viruses (DENV-1–4), is a growing global public health concern, with outbreaks capable of overwhelming healthcare systems and disrupting economies. Climate change, along with other factors, bears substantial relevance to dengue by impacting the virus, mosquitoes, and human populations. Dengue is endemic in more than 100 countries across tropical and subtropical regions worldwide, and the expanding range of the mosquito vector increases risk in new areas. Notably, locally acquired cases are increasingly documented in areas like Spain, Portugal, France and the southern United States, while emerging evidence points to silent epidemics in Africa. Recent years have witnessed significant advancement in our understanding of the virus, immune responses, and disease progression, thanks to ongoing research efforts. Nevertheless, the quest for a reliable immune correlate of protection remains a crucial challenge in the assessment of dengue vaccines. Encouragingly, novel interventions have emerged including partially effective vaccines and innovative mosquito control strategies. These developments mark the beginning of a new era in dengue prevention and control, offering promise in addressing this pressing global health issue.

Introduction

Dengue is a systemic viral infection of increasing global importance. Epidemics of an illness compatible with dengue were first reported in 1779 (1) and the virus was first isolated in 1943 (2). Currently, dengue is endemic in more than 100 countries in tropical and subtropical regions of Southeast Asia, Africa, the West Pacific, and the Americas (3). Dengue is also seen in some regions of Europe, including France, Croatia, Portugal, and Germany, and some parts of the United States (US). Ongoing climate change, population growth, mobility, and urbanization are anticipated to exacerbate dengue burden, primarily by increasing risk in endemic areas, as well as secondarily by expanding range of the primary vector, Aedes aegypti mosquitoes, into new areas. (4) Due to the confluence of these factors, studies predict that the global population at risk will increase from 53% of the world’s population in 2015 to 63% in 2080, with high environmental suitability for dengue in tropical and subtropical areas worldwide (Figure 1). (5)

Figure 1.

Predicted global dengue risk. Means (A) and Standard deviations (SDs) (B) of Force of infection (FOI) estimates in dengue endemic countries across 200 geographically stratified bootstrap samples. Average FOI was estimated from age-stratified seroprevalence or case notification date using a set of environmental explanatory variables. Modified from Cattarino L and colleagues (3) Environmental suitability for dengue occurrence in 2080 (C). Adapted from Messina JP and colleagues (5)

Dengue has an important economic impact, resulting in estimated global healthcare costs of over $8·9 billion (95% uncertainty interval $3·7 billion–$19·7 billion) annually. (6, 7) High costs are associated with loss of productivity, as well as direct medical costs from hospitalization. (8, 9) Dengue outbreaks can overwhelm health care systems, disrupt economies, and reduce public confidence in government responses. Many commonly used vector control strategies, like insecticide spraying, have failed to curb disease incidence but continue to be employed in the absence of robust evidence for their effectiveness or optimal implementation (10). However, increased understanding of dengue epidemiology and immune mediators of symptomatic and severe disease, as well as the availability of effective clinical management, partially effective vaccines, candidate vaccines in the pipeline, and novel approaches to mosquito control, have the potential to inform and significantly improve the effectiveness of dengue control programs. This seminar reviews the latest research on dengue virus (DENV), current gaps in our knowledge, and areas for future study.

Dengue viruses

DENV-1, 2, 3 and 4 are single-stranded ribonucleic acid viruses in the genus Flavivirus, family Flaviviridae. Flavivirus include other viruses transmitted by mosquitoes and ticks such as Zika, West Nile, Japanese encephalitis, and tick-borne encephalitis viruses. The four dengue viruses are called serotypes because each has different interactions with the antibodies in human blood (2). They share approximately two thirds of their genomes (2), with different genotypes existing within each serotype, which can vary in disease severity. DENV is primarily transmitted through the bite of an infected mosquito vector, with Aedes aegypti as the most common vector, although other species including Aedes albopictus may also sustain transmission. Other rare transmission routes include perinatal transmission, blood transfusion, organ transplantation, and two cases of sexual transmission. (11–15) The incubation period from exposure to symptom development is typically four to ten days. (16)

Epidemiology

The global burden of dengue illness has continued to rise over the past decade, with large outbreaks in endemic areas and more cases of dengue in travelers. During 2007–2017, deaths from dengue increased by 65.5% to more than 40,500 (17,600–49,800) annually (17). Expansion of Aedes mosquito vectors and increasing dengue incidence in non-endemic areas is also a growing concern. New detections of local dengue transmission in areas without previous transmission and higher case numbers in areas with sporadic transmission have been documented in southern Europe and the United States, as well as unprecedented outbreaks at high altitudes as seen in Nepal. (18–20) Additionally, more DENV serotypes are co-circulating in endemic areas, producing increased case numbers and greater probability of severe disease from serotype re-introduction or replacement. (21)

DENV seroprevalence estimates vary widely across countries and regions, shaped by differences in underlying DENV transmission intensity as well as differences in methods and assays applied. High IgG antibody seroprevalence (>60%) has been reported in highly endemic areas in Southeast Asia and the Americas, mid-range (10%–60%) seroprevalence in areas with frequent or sporadic transmission such as many countries in Africa and the Middle East, and low seroprevalence in non-endemic areas such as the United States and Europe. (22) Vaccine trials have revealed high variability in DENV incidence across study sites in Latin America and Asia, ranging from 1·5 to 6·6 episodes of symptomatic dengue per 100 person-years among children 2–16 years old. (23) The estimated force of infection (FOI), which is the per capita rate at which susceptible individuals become infected, similarly varies by region, with highest FOI observed in areas near the tropics (Figure 1). However, FOI estimates are limited by surveillance data availability as well as temporal and interannual variability in dengue incidence (3).

Risk of infection is driven by susceptibility to the four DENV serotypes; therefore, DENV incidence in hyperendemic locales is concentrated in children and young adults. (24) The average age of infection in these areas has been increasing in recent decades, for reasons which may include a decrease in the FOI due to changes in the population age structure (25), effective vector control (26), or possibly increased awareness and diagnosis of dengue in adults. In areas hyperendemic for DENV transmission, the risk for enhanced disease has been suggested to be concentrated within two age-related peaks: the first in infants with possible contributions from waning maternal antibodies (27–29), and the second in individuals experiencing a second DENV infection. However, further studies of infants with severe dengue, are needed to unequivocally confirm that maternally derived anti-DENV antibodies are a critical risk factor for severe disease in infants (30). Delayed diagnosis or detection of shock can lead to severe disease and death. Risk of severe disease and death has also been shown to be higher for those with comorbidities such as diabetes or pulmonary, heart, or renal disease. A recent meta-analysis suggests that the relative risks of severe dengue associated with underlying chronic diseases and co-morbidities may be much higher than that of secondary infection alone. (31)

Disparities in dengue risk have been identified, with increased risk occurring in areas with higher population density and poor housing conditions. (32, 33) Increasing population mobility and tourism have also been linked to increased dengue transmission, while imported DENV cases have led to outbreaks in non-endemic areas. (34) DENV transmission occurs commonly within and around households, with a potential increased risk among people in endemic areas who stayed near the home compared to people with greater mobility. (35, 36) However, COVID-19-related disruptions and lockdowns in 2020 were found to result in a decrease in dengue incidence across endemic regions, with the strongest associations related to school closures and reduced time in non-residential areas, although changes in healthcare-seeking behavior could have also contributed to lower reported case numbers. (37)

Dengue is the most frequent arboviral disease encountered among travelers, with an increasing global trend during 1995 to 2020. (38) Among travelers to Southeast Asia during 1995–2020, incidence ranged from 50 to 159 dengue cases per 1000 ill travelers who sought care in the Geosentinel network of travel medicine providers during non-epidemic and epidemic years, respectively. (38) Although severe dengue often occurs during second infections in dengue-endemic regions, primary dengue infections can also be severe and result in fatal outcomes, which has been documented among travelers without previous dengue infection. (39) Additionally, asymptomatic infections have been documented to occur among travelers to dengue endemic areas in a ratio of approximately four to one, creating risks for introduction of dengue viruses or novel serotypes from asymptomatic persons into areas with competent mosquito vectors. (40)

Classification and clinical course

Dengue is a self-limiting acute febrile illness with non-specific manifestations. Among people infected with DENV, approximately 60–80% are asymptomatic or have subclinical infections, with increased risk of disease among those with older age and longer time between subsequent DENV infections. (41, 42) WHO guidelines previously classified symptomatic dengue virus infections as dengue fever, dengue hemorrhagic fever (DHF), and dengue shock syndrome (DSS). However, revised WHO guidelines in 2009 (Panel a) classified symptomatic dengue as dengue without warning signs, dengue with warning signs, and severe dengue (Figure 2). (16) (41, 42)

Figure 2.

Dengue clinical course, classification, laboratory abnormalities and management. From WHO Handbook for Clinical Management of Dengue (43).

Symptomatic dengue generally follows the following clinical course: febrile, critical, and recovery phases (Figure 3). (16, 43) During the febrile phase, which lasts from 2–7 days, there is typically an acute onset of high-grade fever (≥38·5°C) that may be accompanied by nausea, vomiting, a transient macular rash, aches, pains, and other constitutional symptoms. (44) The mucocutaneous manifestations of dengue are varied and may include transient facial erythema, petechial rash, conjunctival and scleral injection, and a maculopapular or morbilliform eruption three to six days after the onset of fever that may coalesce but with areas of sparing. (45) The tourniquet test (46) can be positive and minor bleeding, such as skin petechiae or bruises may occur. (47) Most commonly, fever resolves followed by recovery phase; if this is the case, this illness would be categorised as uncomplicated dengue.

Figure 3.

Dengue laboratory findings, virus detection and immune response

Data from WHO Dengue Guidelines for Diagnosis, Treatment, Prevention and Control (16), Hunsperger E. and colleagues (210) , Chaloemwong and colleagues (211) and Dussart and colleagues (212).

Some dengue patients experience the critical phase, which generally occurs around days 4–6 of illness and often coincides with defervescence. (16, 48) The hallmark of severe dengue is plasma leakage, when the blood’s protein-rich fluid component flows from blood vessels into surrounding tissue, which can lead to shock and is sometimes associated with haemorrhage. ^{(49, 50)} There is some evidence that less severe capillary leakage may be more common in clinically diagnosed uncomplicated dengue than previously recognized. ⁽⁵¹⁾ Plasma leakage becomes clinically apparent near the time of defervescence and spontaneously improves after about 48 to 72 hours (52). Dengue warning signs of possible clinical deterioration can precede the critical phase. Warning signs are intended to be used to detect disease progression and include abdominal pain or tenderness, persistent vomiting, clinically detectable extra-vasal fluid accumulation, mucosal bleed, lethargy or restlessness, liver enlargement, and an increase in haematocrit usually concurrent with rapid decrease in platelet count (Figure 3). (16, 53) If the patient improves and recovers, the illness is classified as dengue with warning signs. However, the disease may continue to advance towards severe dengue, which occurs in approximately 2%–5% of dengue patients (53, 54). Rates of progression to severe disease are highly variable by age, underlying comorbidities, clinical resources, expertise in managing dengue, and possibly infecting DENV serotype and genotype. (53)

The median case fatality rate for patients with dengue is 5% (range 0.01%–39%). (55) The criteria for severe dengue include: (1) severe plasma leakage leading to shock or to fluid accumulation with respiratory distress, (2) severe bleeding as evaluated by clinician, and (3) severe organ involvement including the central nervous system, heart, or liver (indicated by an aspartate aminotransferase [AST] or alanine aminotransferase [ALT] level of 1000 IU/l or greater). (16) Shock is signalled by rising haemoconcentration followed by an increase in diastolic pressure with narrowing pulse pressure, rapid pulse, restlessness, hypotension, signs of poor peripheral perfusion (cold extremities and slow capillary refill time) and decreased urinary output. Repeat shock episodes may occur during the critical phase (56). Circulatory compromise is generally worse in the extremes of age, likely due to increased vascular permeability and lower capacity to maintain cardiovascular homeostasis in young children, while factors in the elderly include comorbidities and vascular aging. (57) Epistaxis, gum bleeding, hypermenorrhea, hemoglobinuria, and other haemorrhagic manifestations are most seen during the critical phase. (47) The risk for severe bleeding (such as from the gastrointestinal or vaginal tract) increases in profound or prolonged shock, in association with coagulation abnormalities combined with tissue hypoxia and acidosis. (58)

Involvement of other organ systems can also occur during the various phases of dengue. Hepatitis and elevated liver enzymes are common among patients with symptomatic dengue, while acute liver failure, encephalitis, myocarditis, and acute kidney injury are infrequent. (59–62) Sight-damaging ophthalmic inflammation during dengue has also been described. (63) Dengue during pregnancy has implications for the mother and the fetus. A greater risk of DHF/DSS has been reported among pregnant women compared to non-pregnant women, (64) as well as an increased risk of maternal death. (65, 66) In addition to increased maternal morbidity and mortality, dengue also poses a risk to the fetus, with an increased risk of miscarriage (67), stillbirths, and neonatal deaths. (66)

During the recovery phase, extravasated fluids are resorbed and well-being improves. The patient may develop an erythematous, sometimes pruritic, “Herman’s rash” with white islands of normal skin. (68) In adults, post-viral fatigue and depression of several weeks to months have been described. (69)

Most dengue patients recover uneventfully, but it is crucial to promptly recognise those who will require medical intervention. A systematic review confirmed that warning signs were associated with progression to severe dengue but identified potential additional markers, including low serum albumin and elevated AST and ALT concentrations. (31) Thrombocytopenia is also commonly seen in dengue patients, and lower platelet counts have been associated with progression to severe disease. (31) Ultrasound can be used to detect plasma leakage in dengue. (70) Ascites, pleural effusion and gallbladder wall thickening are the most common findings, but standard protocols for sonographic procedures, as well as improved information about the positive predictive value of early and low-volume plasma leakage for development of severe dengue, are needed. (71, 72) Investigations are ongoing to determine whether inflammatory and vascular markers in the febrile phase of dengue may be useful to predict severe outcomes. (73, 74)

Dengue management

Currently there is no effective prophylactic or therapeutic agent against dengue. (75) Chloroquine, balapiravir, celgosivir, lovastatin, corticosteroids, ivermectin, plasma infusion, recombinant activated factor VII, anti-D globulin, immunoglobulin, and interleukin 11 have not been shown to be beneficial. (76–78) However, clinical trials are limited by small sample sizes, heterogeneous populations, and difficulties in assessing outcomes, and additional work is needed to thoroughly assess potential benefits. A randomized controlled trial of montelukast, a leukotriene receptor antagonist used to reduce asthma exacerbation, is ongoing among adult patients with dengue to evaluate its efficacy in preventing dengue with warning signs (ClinicalTrials.gov Identifier: NCT04673422). The small molecule, JNJ-A07, which blocks the intracellular replication of DENV, has shown promise in pre-clinical trials. (79) There are also ongoing pre-clinical trials of therapeutic monoclonal antibodies against dengue. (80)

In the absence of specific therapy, the management of dengue remains supportive. The 2012 WHO handbook, (43) since adapted to other guidelines, (81, 82) focuses on a stepwise approach of assessment and treatment according to Groups A, B, and C (Figure 2). Group A patients (no warning signs, comorbidities, or difficult social circumstances) are sent home with daily in-person monitoring. Groups B (comorbidities or warning signs) and C (severe dengue) require hospital management and intravenous fluids (Figure 2). (83) Fluid replacement is lifesaving in severe dengue but must be administered cautiously and discontinued when plasma leakage subsides to avoid iatrogenic fluid overload. (84) Among patients with thrombocytopenia, prophylactic platelet transfusion does not prevent bleeding and may contribute to fluid overload. (85, 86) Although these guidelines are based mainly on expert opinion and small randomized controlled trials, (87–90) case fatality rates have been considerably reduced with judicious fluid replacement. Some areas of uncertainty remain, such as in the choice of colloid solution and blood products, the use of fluid boluses, and the optimal treatment of recurrent shock episodes. Steroid use is not recommended as it has not shown clinical benefit. (57, 76) Training in clinical management, including early recognition of plasma leakage, remains an essential strategy to reduce morbidity and mortality.

Dengue diagnosis

Dengue differential diagnosis is broad. During the febrile phase, it includes other viral infections (measles, rubella, enterovirus, adenovirus, influenza and other arboviruses), as well as bacterial (leptospirosis and typhoid fever) and parasitic (malaria) illnesses that may be present in dengue-endemic areas. (91). Diagnosis of dengue infection during the acute phase can be achieved using whole blood, plasma, or serum collected up to seven days after symptom onset by detection of viral RNA through nucleic acid amplification tests (NAAT) (92), detection of viral antigens such as dengue non-structural protein 1 (NS1), by enzyme-linked immunosorbent assay (ELISA), or rapid diagnostic tests, and IgM antibodies from day 4 to approximately 12 weeks post onset through serologic testing (93) (Fig 3). NAAT assays are the preferred method of dengue diagnosis; (92) in addition to their diagnostic specificity, molecular methods can be used to identify the virus serotype.

NS1 can be detected in other body fluids such as urine, saliva, and cerebrospinal fluid. (94) NS1 tests can be as sensitive as molecular tests during the first seven days of symptoms in primary infections, although sensitivity is lower in secondary infections; after seven days, although sensitivity is lower, NS1 has been detected in serum up to 12 days after symptom onset (94, 95). Dengue IgM antibodies are detectable for a longer period, from day four to approximately 12 weeks post symptom onset (96). Dengue IgG is detectable around day seven in the first infection (primary infection); the antibody concentration increases slowly thereafter and is thought to persist for life. In patients with a previous dengue infection (secondary infection), anti-dengue IgG titres rise rapidly within the first week of illness (Figure 3). Although serologic assays provide less certainty than NAAT or NS1 due to cross-reactivity with other flaviviruses and longer antibody duration, a positive anti-DENV IgM suggests recent infection. Additionally, seroconversion or a four-fold rise in titres on anti-DENV IgM or IgG assays in paired samples is strongly suggestive of recent infection. (96) Many rapid tests are available in the market and are an important tool for the early diagnosis of dengue. Meta-analyses suggest that immunochromatographic tests that combine IgM, IgG, and NS1 detection have the best performance compared to tests detecting individual analytes, with pooled sensitivity of 90%–91% and specificities of 89%–96% (97, 98). Unfortunately, these tests are not widely available in dengue endemic areas. In patients from areas with ongoing transmission of another flavivirus, plaque reduction neutralization tests (PRNT) can help distinguish DENV from other flaviviruses. PRNTs, however, are rarely available in clinical laboratories and typically do not provide results within a meaningful timeframe for clinical management. PRNTs may be valuable in circumstances such as pregnancy where differentiating between Zika and dengue may have important clinical implications. (93) NS1 antibody ELISA tests for Zika virus (ZIKV) have high specificity due to the substantial amino acid differences between DENV and ZIKV NS1 and can be useful for differential diagnosis (94).

Dengue immunology

DENV infection is initiated in the skin when an infected mosquito takes a blood meal, injecting virus along with salivary proteins that increase recruitment of susceptible immune cells to the site of infection (99). Myeloid cells are a key target of DENV infection, including monocytes, macrophages, and dendritic cells. A first DENV infection results in an early innate response characterized by stimulation of interferon gamma (IFNγ). In contrast, in subsequent DENV infections, binding but not neutralizing antibodies induced by prior exposure to DENV (or a related flavivirus) facilitate infection of myeloid cells via the fragment crystallizable gamma receptor (FcγR), producing a larger population of viruses and further exacerbating disease severity in a process called extrinsic antibody-dependent enhancement (ADE) (100–104). Entry via the FcγR also mediates intrinsic ADE, which suppresses IFNγ stimulation and innate immunity. This results in a shift instead towards a T helper 2 (Th2) response dominated by secretion of Interleukin-10, minimizing induction of other pro-inflammatory cytokines and hindering the early cellular and humoral immune response (Figure 4) (100). ADE is thought to increase replication of the virus at this key early stage and elevate the risk of progressing to severe dengue and DHF/DSS. (105, 106) However, while pre-infection binding antibody levels are associated with increased viremia and risk of DHF/DSS, the causal link from ADE to higher viremia to DHF/DSS has not been demonstrated (103).

Figure 4.

Correlates of dengue pathogenesis and protection. Types of immunological responses associated with increased dengue disease (pathogenic, generally during a secondary infection: ADE of myeloid cells (100–102), strong plasmablast response (107, 109, 110), and weak T cell response (114)) or reduced disease (protective, including broadly neutralizing antibodies (135, 136) and effective, cytotoxic CD8+ and CD4+ T cells (115–118)).

Mediators of severe disease

CD14+CD16+ monocytes increase in early DENV infection (107, 108) and trigger a strong plasmablast response characterized by secretion of high levels of anti-DENV antibodies (107, 108). The role that excess antibody production plays in acute dengue is not clear, but may contribute to disease pathogenesis by furthering ADE, increasing autoantibodies, and potentially changing glycosylation of antibodies, which has been shown to be strongly associated with DHF/DSS (109, 110). DENV can also directly infect B cells, and while B cell infection does not contribute to viremia it does drive proliferation of B cells and stimulation of cytokines (111). Higher viral load may also mediate severe disease by increasing secretion of NS1. In in vitro and animal models, NS1 levels directly and indirectly triggered vascular leakage, disrupting the glycocalyx and tight junctures between endothelial cells lining blood vessels and further facilitating dissemination of virus into tissue as well as plasma leakage (112). However, the association between NS1 levels and severe disease has not been clearly demonstrated in clinical studies. (49) Severe dengue is also associated with liver and spleen pathology, with autopsies revealing DENV tropism for liver macrophages (Kupffer cells) and splenic macrophages, which secrete high levels of cytokines and mediate damage due to deposition of complement, resulting in necrosis in liver and splenic endothelial cells (113).

T cell responses

The role of DENV-specific T cells has been debated, as they have been implicated in both pathogenic and protective immunity (Figure 4). On the side of a pathogenic role, CD8+ T cells induced by prior DENV infection have been shown to have low affinity for the new infecting serotype, proliferating but with limited cytotoxic function, resulting in delayed viral clearance and stimulation of pro-inflammatory cytokines that mediate leakage and disease (114). However, other studies suggest a protective role, with higher magnitude and more multifunctional, cytokine-producing DENV-specific CD8+ T cell responses and specific HLA alleles associated with reduced viremia, and lower probability of progression to symptomatic disease and severe dengue (115–117).

DENV-specific CD4+ cells both promote CD8+ T cells and stimulate B cells. Individuals with multiple prior DENV exposures have populations of clonally expanded cytotoxic CD4+ cells that may be protective (118). In contrast, while T follicular helper (Tfh) cells are critical to facilitating the maturation of the B cell response, Tfh expansion in acute dengue has been associated with secondary and severe dengue as well as a strong plasmablast response, suggesting a possible role in immunopathology (119, 120).

DENV serotypes and post-secondary DENV antibodies

Although all DENV serotypes can result in symptomatic or severe disease, differences have been identified in risk by serotype and infection number (primary or secondary). DENV-2 and DENV-4-associated dengue illnesses are more frequently identified as secondary DENV infections, whereas DENV-1 and DENV-3 can cause significant primary disease (121–123) DENV-4 has been associated with a reduced risk of disease compared to the other serotypes, which is supported by observations of silent DENV-4 epidemics in Thailand. (124) All sequences of infecting serotypes can cause severe disease and there is not a defined order in the sequence of DENV infections for worse outcomes, although different patterns have been described (125, 126). Antigenic differences between serotypes may be important in explaining the magnitude of dengue epidemics (127–129), and viral sequence, genotype, and changes in non-structural proteins may also help determine epidemic severity (130, 131).

Sequential infection with distinct DENV serotypes induces antibodies capable of neutralizing DENV-1–4 without triggering ADE, likely by activating cross-reactive memory B cells to undergo further affinity maturation and target quaternary epitopes conserved across serotypes, thus providing broad protection. High levels of cross-reactive binding antibodies and multiple prior DENV infections are associated with reduced risk of symptomatic disease (105, 124, 132). Consistent with this observation, Dengvaxia and the Qdenga dengue vaccines have high efficacy against symptomatic and severe disease in DENV-immune individuals but not naive individuals, further suggesting that sequential exposure to distinct DENV strains is important to inducing broad protection (133, 134). A few post-secondary broadly neutralizing antibodies have been identified such as envelope-dimer-dependent epitope (EDE) antibodies. These antibodies target conserved quaternary epitopes and neutralize by disrupting the conformational changes required for viral entry (135, 136). EDE-like broadly neutralizing antibodies may be good correlates of protection for the evaluation of new dengue vaccines.

Dengue Vaccines

The need for a tetravalent formulation that induces simultaneous and balanced protection against all 4 serotypes has slowed the development of a dengue vaccine. Among people with a DENV infection before vaccination, even a vaccine dominated by one vaccine serotype is likely to induce cross protective immunity by activating memory. However, in DENV naïve individuals with no immune memory, the protective immune response will strongly depend on the immunogenicity of each serotype-specific vaccine component (137).

There are three leading dengue vaccines. Dengvaxia (CYD-TDV), developed by Sanofi-Pasteur, was the first licensed dengue vaccine (138). Qdenga (TAK-003), developed by Takeda, was approved by the European Commission in December 2020 and is licensed in several countries including Indonesia, Brazil, Argentina, the United Kingdom, and Germany (139–141). The third, TV003, was developed by the National Institutes of Health (NIH) and is in late-stage trials (142). All three are live vaccines and contain 4 different attenuated vaccine viruses (tetravalent) targeting each of the DENV serotypes. However, they differ in the number of doses required (1 to 3) and time to complete the series (up to 1 year), which could affect feasibility and preferences for use in different settings. Additionally, the need for a dengue test to determine eligibility (pre-vaccination screening) poses a logistical barrier for vaccines only recommended for use among people with a previous DENV infection. There are several other dengue vaccine candidates undergoing clinical trials or pre-clinical evaluation, including other live-attenuated vaccines, inactivated vaccines, recombinant vaccines, and DNA vaccines (143). The successful mRNA vaccine technology used for SARS-CoV-2 is also being evaluated for dengue and could provide dengue vaccine candidates in the future. (144)

Dengvaxia (CYD-TDV)

Dengvaxia uses a three-dose schedule, with doses administered six months apart. It was first recommended by WHO in 2016 for persons nine years of age and older living in highly endemic areas (145). Long term follow-up data over five years from the phase three trials and further analyses of the efficacy results (146) demonstrated that children with evidence of previous DENV infection (seropositive) were protected from severe dengue if they were vaccinated with Dengvaxia. However, risk of hospitalization for dengue and severe dengue was increased among children two to 16 years of age without previous DENV infection (seronegative) who were vaccinated with Dengvaxia and had a subsequent infection in the years after vaccination (hazard ratio for hospitalization, 1·75, 95% Confidence Intervals [CI]: 1·14–2·70, severe dengue 2·87, 95% CI: 1·09, 7·61). After these findings, WHO revised the recommendations for the vaccine to only be given to children with laboratory-confirmed evidence of a past DENV infection (147).

For children aged nine to 16 years with evidence of previous DENV infection, Dengvaxia had an efficacy of about 80% against the outcomes of symptomatic virologically confirmed dengue (VCD), hospitalization for dengue, and severe dengue. (146, 148) Among seropositive children the efficacy by serotype varied (149), with highest protection against DENV-4 (89%), followed by DENV-3 (80%), and lowest against DENV-1 (67%) and DENV-2 (67%) (Table 1). (146)

Table 1.

Comparison of vaccine efficacy for the target use population for live attenuated tetravalent dengue vaccines licensed or in phase three trials

Vaccine	Dengvaxia/CYD-TDV146	Qdenga/TAK-003152	TV003161
Manufacturer	Sanofi Pasteur	Takeda	NIH/Butantan/Merck
Status	Recommended by World Health Organization (WHO) with two options: 1) seropositive persons ages 9–45y; or 2) all persons regardless of serostatus in high seroprevalence areas (>80% seropositivity). Licensed in 20 countries. In U.S. recommended for seropositive children ages 9–16y living in endemic areas.138	Recommended by WHO to be considered for introduction among children ages 6–16 years in settings with high transmission intensity. Licensed in several countries including Indonesia, Brazil, Argentina, United Kingdom and Germany.	Ongoing phase three trial
Platform	4 chimeric viruses for each DENV serotype on a yellow fever virus (YFV) backbone	Attenuated DENV-2 and 3 chimeric viruses each one of the four DENV serotypes	Attenuated DENV-1, DENV-3 and DENV-4 and a chimeric virus for DENV-2 on a DENV-4 backbone
Ages of trial participants	9–16y	6–16y	2–59y
Doses	3 doses 6 months apart	2 doses 3 months apart	1 dose
Pre-vaccination antibody screening recommended	Yes	No	Unknown
Timeframe for efficacy endpoint	25 months for virologically confirmed dengue (VCD and 60 months for hospitalization)	54 months for VCD and hospitalization	24 months for VCD
Efficacy among seropositive persons	Efficacy and 95% Confidence Interval (CI)	Efficacy and 95% CI	Efficacy, 95% CI
Virologically Confirmed Dengue
Overall	76 (64, 84)	64 (58, 69)	89 (78, 96)
By serotype
DENV-1	67 (46, 80)	56 (45, 65)	97 (81, 100)
DENV-2	67 (47, 80)	80 (73, 86)	84 (63, 94)
DENV-3	80 (67, 88)	52 (37, 64)	NR
DENV-4	89 (80, 94)	71 (40, 86)	NR
Hospitalization
Overall	79 (46, 80)	86 (79, 91)	NR
By serotype
DENV-1	78 (55, 90)	67 (37, 82)	NR
DENV-2	82 (66, 90)	96 (90, 98)	NR
DENV-3	63 (18, 83)	74 (39, 89)	NR
DENV-4	93 (62, 99)	100 (NE, NE)	NR
Efficacy among seronegative persons
Virologically confirmed disease
Overall	39 (−1, 63)	54 (42, 63)	74 (58, 84)
By serotype
DENV-1	41 (−7, 67)	45 (26, 60)	86 (69, 94)
DENV-2	−21 (−136, 38)	88 (79, 93)	58 (21, 78)
DENV-3	52 (−6, 78)	−16 (−108, 36)	NR
DENV-4	65 (24, 84)	−106 (−629, 42)	NR
Hospitalization
Overall	−41 (−168, 93)	79 (64, 88)	NR
By serotype
DENV-1	−37 (−219, 41)	78 (44, 92)	NR
DENV-2	−141 (−795, 35)	100 (NE, NE)	NR
DENV-3	15 (−225, 78)	−88 (−573, 48)	NR
DENV-4	7 (−712, 89)	100 (NE, NE)	NR

Note: Dengvaxia trial included participants ages 2–16 years of age but only licensed for seropositive individuals ≥9 years. Estimates of efficacy against hospitalization by serotype in seronegative participants ages 2–8 years are: DENV-1 –42 (34, −205), DENV-2 –436 (−58, −1723), DENV-3 –141 (9, −540), DENV-4 –16 (70, −344). NE=Not possible to estimate due to zero cell in one of the groups, NR= not reported. Y=years

The requirement for a laboratory test prior to vaccine administration creates a unique challenge for Dengvaxia implementation. Qualifying laboratory tests include a positive NAAT or NS1 test performed during an episode of acute dengue, or a positive result on pre-vaccination screening tests for serologic evidence of previous infection (presence of IgG antibodies) that meet specific performance characteristics. To reduce the risk of vaccinating someone without previous DENV infection, high specificity in a pre-vaccination screening test is a priority. International working groups and the CDC recommend using tests with a minimum sensitivity of 75% to 85% and minimum specificity of 95% to 98% (138, 150). Few commercially available tests currently meet these requirements. (151)

Qdenga (TAK-003)

Qdenga, developed by Takeda, consists of two doses given three months apart. Among children aged four to 16 years, efficacy against VCD was 64% among seropositive and 54% among seronegative children at three years after vaccination. Efficacy against hospitalization for dengue was higher, at 86% among seropositive and 79% among seronegative children. (152) Differences in efficacy were observed by serotype. Among seronegatives, there was no efficacy against DENV-3 and DENV-4 (Table 1). Notably, estimates indicated a potential increased risk for hospitalization after infection with DENV-3, although numbers were small (three cases in the placebo group and eleven cases in the vaccine group), and mainly observed at one site (134). In December 2022, the European Commission approved the use of Qdenga regardless of serostatus following a positive opinion from the European Medicines Agency (139). The next step for its use in Europe is official recommendations from each European Union country (153). Qdenga has been approved in several countries and Germany has started vaccination among travelers (154). In September 2023, it received a recommendation from the Strategic Advisory Group of Experts (SAGE) on immunization that recommended that Qdenga be considered for introduction in settings with high transmission intensity to maximize the public health impact and minimize any potential risk in seronegative persons. SAGE recommended that the vaccine be introduced to children aged 6 to 16 years of age, and that post-authorization studies should be conducted to further study vaccine effectiveness and safety against serotypes 3 and 4. (155)

The company also plans to submit filings to other regulatory agencies (156).

TV003

TV003 was developed by NIH and was formulated by selecting serotype-specific components to provide a balanced safety and immunogenicity profile based on an evaluation of multiple monovalent and tetravalent candidates (Table 1). (157, 158) TV003 consists of a single dose and has been licensed to several manufacturers globally, including Merck & Co in the U.S. and the Instituto Butantan in Brazil. Phase three trials in Brazil are underway (159, 160). Preliminary results from two-year follow-up of the phase three trial were recently released. Through 2 years of follow-up, the efficacy against VCD was 89% among seropositives and 74% among seronegatives. Results by serotype were available for DENV-1 and DENV-2, with higher efficacy among seropositive participants compared to seronegative (DENV-1 97% and 86%, DENV-2 84% and 58%, respectively) (161). Efficacy for other serotypes is not currently available but is expected as part of the ongoing phase three trial. (162)

Vector control

People who live in or travel to dengue endemic areas can prevent mosquito bites by using EPA-approved insect repellents and wearing clothing that covers arms and legs. The use of screened windows and doors and air conditioning have been shown to be protective (163–165).Bednets can reduce mosquito populations and may have an impact on dengue transmission. (166) Chemical control of Aedes species mosquitoes is limited by widespread insecticide resistance in endemic areas. (167) Novel vector control methods have been developed, including the use of genetically modified mosquitoes (GMM) and Wolbachia-based methods (168). GMM carry a gene that is passed to their offspring and kills females in the larval stage. Male offspring, however, survive and pass this gene to future generations. As a result, mosquito populations decrease over time. (168, 169)

Strategies utilizing Wolbachia, an intracellular bacterium found in about 60% of all insects (170) but not commonly found in wild Aedes mosquitos, (171) have been used for vector control. Wolbachia-mediated suppression refers to a reduction in wild populations of Aedes mosquitoes and is achieved by releasing Wolbachia-infected males into the environment to mate with uninfected wild females, as the resulting eggs do not hatch. (172) In Wolbachia replacement, both Wolbachia-infected males and female mosquitoes are released, which pass the bacteria to their offspring and gradually replace the wild population. (173, 174) In mosquitoes, Wolbachia infection reduces transmission of arboviruses, including dengue, chikungunya, and Zika viruses, when infected female mosquitoes take a bloodmeal. This method has demonstrated reductions of nearly 80% in dengue cases and related hospitalizations in areas where it has been implemented (175) and is currently being deployed in Brazil and Indonesia (176).

Dengue controversies, gaps, and opportunities (panel b)

Many of the key questions in dengue research revolve around the role and behavior of antibodies in protective immunity and ADE, as well as how these change at different time points after infection. It was previously thought that a first DENV infection induced antibodies that wane over two years to titers that can subsequently enhance severe dengue disease, and that secondary DENV infection with a different serotype induced stable, cross-serotype protective antibodies. (41, 42, 177) However, instead of waning over two years, cross-reactive binding antibodies associated with protection or enhancement are stable by ~8 months after primary infection and are maintained at that ‘set point’ for many years after (124, 178). In contrast, anti-DENV antibodies induced after secondary DENV are less stable, wane rapidly for 8 months and then gradually decay over longer periods. One study showed vaccine efficacy waned much faster than geometric mean antibody titers to DENV-1–4, suggesting a component of immunity other than waning antibodies may explain loss of protection (179).

While serotype-specific immunity (i.e., ‘homotypic’ immunity) has classically been thought to impart lifelong protection, there is some evidence that reinfections with a given DENV serotype may occur (180). This bears further evaluation; if proven, the absence of lifelong homotypic immunity would have important implications for dengue vaccines and our understanding of DENV epidemiology.

Additionally, factors shaping the variability observed in severity of epidemics remain poorly understood. Investigators in Taiwan have reported increasing severity throughout the time course of a given epidemic, associated with increased viral diversity, which they hypothesize to be driven by cross-protective immunity (181, 182). It remains to be seen whether these same phenomena are replicated in other regions, with different levels of population immunity and transmission patterns.

Identification of a satisfactory immune correlate of protection — a biomarker measuring immune response to vaccination that is associated with vaccine efficacy — remains an important challenge for DENV epidemiologic studies and assessments of vaccine immunogenicity. This is mainly due to the dominance of immunity to cross-reactive epitopes which do not provide effective protection. In the Dengvaxia pediatric vaccine trial, the discordance between vaccine efficacy and neutralization response rates indicated that PRNT neutralization response is not a completely valid correlate of protection. However, higher PRNT titers after three doses of Dengvaxia were associated with a lower rate of VCD and hospitalization overall and for each infecting serotype (183). Following vaccination, geometric neutralizing antibodies by PRNT to DENV-1–4 of ≥1:100 were associated with ~50% protection against symptomatic dengue, while titers of ≥1:500 were associated with 80% vaccine efficacy; high titers were also associated with protection against hospitalized dengue (183, 184). In the Qdenga vaccine trials, neutralizing antibody titers were lower in participants who experienced VCD (i.e., cases) compared to control, largely driven by data from seropositive participants (134). Techniques such as antibody depletion of cross-reactive antibodies provide improved information about serotype-specific immunity and have shown associations with serotype-specific vaccine efficacy (134, 183–185). Depletion assays have shown that Dengvaxia is dominated by DENV-4 type-specific antibodies and the Takeda dengue vaccine by DENV-2 type-specific antibodies, with mostly cross-reactive antibodies against other serotypes (186, 187). Many of these serotype-specific antibodies bind quaternary epitopes (i.e., two adjacent E proteins simultaneously). The role of neutralizing antibodies binding quaternary epitopes is an ongoing area of research into correlates of protection.

The role of immunological ‘boosting’, defined here as qualitative or quantitative changes in immunity associated with re-exposure to DENV in an individual already exposed to that serotype, remains poorly understood. The effects of boosting are proposed to be evident in dynamic antibody patterns among individuals residing in hyperendemic locales over time. (188) If boosting were a significant contributor to maintenance of DENV immunity and durability of protection, interventions that decrease the force of infection (such as incompletely-protective vaccination programs or partially-effective mosquito control programs) could yield paradoxical effects by lowering levels of boosting and increasing the susceptible population. However, partially effective vaccines and mosquito control interventions complement each other, contributing effectiveness when the other is lacking. Mathematical modeling suggests that combining interventions may yield consistent high effectiveness. (189) Designing and implementing a dengue control program that uses a combination of available interventions is a public health priority.

The immunological and clinical interactions between DENV and non-DENV flaviviruses remain poorly understood, but existing data suggest that the relationship may be highly context-dependent and not bidirectional. ZIKV infection has been suggested to predispose to DHF with a subsequent infection with some, but not necessarily all, serotypes (104). In reverse, higher levels of DENV immunity are associated with protection against Zika in adults and children (190), while lower levels may increase risk of Zika microcephaly. (191, 192)Japanese encephalitis virus (JEV) immunity has been variably associated with risk of dengue illness (193) but also with possible protection (194). Given significant immunological cross-reactivity observed between DENV and non-DENV flaviviruses, it is plausible that these interactions may have consequences for the clinical outcomes of DENV infection as well as the immunological outcomes of DENV vaccination. Further research on these interactions is needed, across a range of flavivirus-endemic regions of the world, to inform DENV vaccine development and evaluation efforts, as well as diagnostics.

Dengue surveillance remains a key challenge in assessing global dengue burden and temporal and geospatial trends. The large proportion of asymptomatic and subclinical cases contribute substantially to transmission but complicate detection, and the non-specific presentation of acute febrile illness can easily be mistaken for other etiologies. This is particularly challenging in malaria-endemic regions where cases have similar presentations, in areas with limited diagnostic test availability, and regions with infrequent or sporadic dengue transmission. National dengue surveillance systems also vary widely in surveillance and laboratory capacity, as well as case definitions used; future efforts should work toward strengthening country-level surveillance and laboratory capacity, harmonizing case definitions, and encouraging public data sharing to better inform dengue preparedness and response efforts.

Search strategy and selection criteria

We searched Pubmed for the years September 30, 2012, to October 10, 2022. We used the search terms “dengue” or “dengue virus”. We also included references cited in these publications and relevant older references from the authors’ personal files. We consulted international dengue control, prevention and treatment guidelines and WHO policy documents.

Panel a: The 1997 versus the 2009 WHO dengue classification and case definitions.

• The 1997 WHO classification and case definitions of dengue, dengue haemorrhagic fever (DHF) and dengue shock syndrome (DSS) (195) originated from a clinical study in the 1960s of 123 children in Thailand. (196) A case definition of DHF is met when all four criteria of fever, haemorrhagic manifestations, thrombocytopenia and evidence of plasma leakage are present. DHF has four increasing grades of severity towards DSS (Grades III and IV). DHF definition is required to classify cases as dengue shock syndrome (DHF plus circulatory failure). The 1997 WHO classifications offer distinct advantages for research studies, as the clinical phenomena described relate to underlying mechanisms of immuno-pathogenesis (i.e., ADE), and may guide treatment pathways (e.g., timing and degree of fluid replacement with plasma leakage). Disadvantages of the 1997 criteria include failure to capture severe disease beyond DHF (e.g., cardiac or hepatic end-organ damage) and limited capacity to facilitate triage of DENV-infected patients. The WHO regional office for South-East Asia proposed the term Expanded Dengue Syndrome to describe cases with atypical yet serious manifestations. (197)

• The 2009 WHO classification system evolved to facilitate triage and management of dengue patients (53), as well as to capture a broader spectrum of dengue-related disease. Individuals are classified as having dengue, dengue with warning signs and severe dengue. (16)

• The 2009 WHO criteria have been widely implemented to guide dengue clinical management decisions, but they have also been criticized due to the broad recommendations for hospitalization of patients with dengue warning signs. Multiple studies indicate the general approach of hospitalizing for dengue with warning signs, as per the 2009 WHO criteria, may increase the identification of individuals who will progress to severe disease (i.e., improved sensitivity), at the expense of increasing hospitalization of individuals who will not progress to severe disease (198, 199). The positive predictive value of individual warning signs for severe dengue has been reported to range from 12–58%, with some warning signs (e.g., clinical fluid accumulation) performing better than others (200). The availability and increasing use of dengue rapid diagnostic tests may facilitate the use of the 2009 simpler dengue classification scheme. However, rapid diagnostic tests are not available in most endemic countries. More precise definitions of the 2009 warning signs and severe dengue are needed, as well as improved risk prediction tools for clinical and research use (201–204).

• Currently, the 2009 classification of dengue without warning signs, dengue with warning signs and severe dengue (16) has been adapted in several countries and international guidelines, (82, 83, 205, 206) while the 1997 dengue, DHF, DSS classification (195) continues to be used in others. (207–209)

Panel b: Priorities for future dengue virus (DENV) research.

DENV prevention and control:

• Effectiveness of combined vaccine/vector control programs in decreasing or eliminating DENV infection and illness.

• Long term data on effectiveness of interventions including vaccines and Wolbachia-based vector control.

Immune correlates of protection:

• Immune signatures durably associated with immunopathogenesis and immunoprotection for DENV.

• Clinical and immunological interactions between DENV and non-DENV flaviviruses.

• Durability of homotypic protection – is it lifelong?

• Importance of immune boosting in maintaining protective immunity for DENV.

• Gaps in vaccine efficacy for specific sub-groups that could lead to enhancement of severe disease.

DENV epidemiology:

• Homotypic DENV reinfections – confirmation, frequency, and risk factors.

• Increases in DENV severity within the course of an epidemic –confirmation and features of settings at risk.

• Harmonization of surveillance and laboratory methods across regions experiencing DENV transmission; increased data sharing, establish a coordinated genomic surveillance strategy.

Management and diagnosis

• Improved point of care diagnostics.

• Clinical evaluation of the impact of antivirals.

• Improved triage and risk assessment tools (biomarkers, ultrasound, other).

Online Summary.

Dengue Overview

• DENV-1, 2, 3 and 4 are single-stranded ribonucleic acid viruses in the genus Flavivirus, family Flaviviridae. DENV-1–4 are called serotypes because each has different interactions with the antibodies in human blood. They share approximately two thirds of their genomes, with different genotypes existing within each serotype, which vary in disease severity.

• DENV is primarily transmitted through the bite of an infected mosquito vector, with Aedes aegypti as the most common vector, although other species including Aedes albopictus may also sustain transmission.

• Other rare transmission routes include perinatal transmission, blood transfusion, organ transplantation, and two cases of sexual transmission. The incubation period from exposure to symptom development is typically 4–10 days.

• Dengue affects more than 100 countries in tropical and subtropical areas. Ongoing climate change, population growth, mobility and urbanization are anticipated to exacerbate dengue burden, primarily by increasing risk in endemic areas, as well as secondarily by expanding range of Aedes species mosquitoes, into new areas.

Presentation and diagnosis

• Dengue is a self-limiting acute febrile illness with non-specific manifestations. Warning signs (e.g. abdominal pain, persisting vomiting, fluid accumulation) are used to detect disease progression. The hallmark of severe dengue is plasma leakage that may lead to shock and less frequently, haemorrhage and there may be liver, central nervous system, heart and other organ involvement.

• Currently there is no effective prophylactic or therapeutic agent against dengue. The management of dengue remains supportive with an emphasis on maintaining hydration and, when indicated by the clinical manifestations, providing intravenous fluids.

• Diagnosis of dengue infection during the acute phase can be achieved using blood collected ≤7 days after symptom onset by detection of viral RNA through nucleic acid amplification tests (NAAT), detection of viral antigens such as dengue non-structural protein 1 (NS1), by enzyme-linked immunosorbent assay (ELISA), or rapid diagnostic tests, and IgM antibodies through serologic testing.

Prevention

• Persons who live in or travel to dengue endemic areas can prevent mosquito bites by using EPA-approved insect repellents and wearing clothing that covers arms and legs. The use of screened windows and doors, air conditioning, and bed nets have been shown to be protective. Chemical control of Aedes species mosquitoes is limited by widespread insecticide resistance in endemic areas.

• Novel vector control methods have been developed, including the use of genetically modified mosquitoes (GMM) and Wolbachia-based methods (168). There are three leading dengue vaccines. Dengvaxia (CYD-TDV), developed by Sanofi-Pasteur, was the first licensed dengue vaccine. Qdenga (TAK-003), developed by Takeda, was recommended by SAGE in September 2023 to be considered for introduction in settings with high transmission intensity. The third, TV003, was developed by the National Institutes of Health (NIH) and is in late-stage trials.