1. DENGUE: EPIDEMIOLOGY PART 1
SCOTT B HALSTEAD, MD
Director, Research
PEDIATRIC DENGUE VACCINE INITIATIVE

2. TRANSMISSION Aedes aegypti breeds in clean water in and around houses.
Daytime biting.
Transmission from human to human requires the same female mosquito to bite a viremic human and then bite a susceptible human at an interval of around days.
Mosquito vectors of dengue viruses are Aedes aegypti and Aedes albopictus; the former supports efficient transmission in urban areas and is the principal vector throughout the tropical world, the latter transmits inefficiently in more rural areas of the tropics. Both are daytime biting.

3. FOUR VIRUSES Life time immunity follows infection to one type.
Second, third and possibly four infections are possible.
CHILDREN – first infections are mild, largely inapparent.
ADULTS - first infections may produce DF, some viruses more overt than others.
The four dengue viruses, members of the flavivirus genus, evolved from a common ancestor in subhuman primates and separately were introduced into the urban transmission cycle.

4. PRIMARY INFECTIONS Clinical Features
In children –
DEN 1 & 3 – mild illness
DEN 2 & 4 – no illness
In adults
DEN 1 & 3 – Disease/Infection ~1; g.i. hemorrhages may accompany peptic ulcer disease.
DEN 2 & 4 - mild - moderate
During initial infections marked differences in pathogenicity of different dengue viruses are observed with disease generally being more severe in adults than in children.

5. DENGUE FEVER Incubation period = 5 days Fever = 5 days Leukopenia
Moderate thrombocytopenia
In 1929, U.S. Army volunteers in the Philippines were infected serially via bites of dengue 1-infected Aedes aegypti and then carefully studied clinically. See Halstead SB Am J Trop Med Hyg 23: , 1974.
Simmons et al
Phil J Sci 44:1-252, 1931

6. DENGUE 1 MACULO- PAPULAR RASH. Day 5 after onset of fever.
Concurrent with defervescence many patients with Dengue Fever develop a measles-like generalized maculo-papular rash.

7. DISEASE SPECTRUM MILD SEVERE DF DHF + Thrombocytopenia Thrombocytopenia Hidden Vasc. Perm1? Overt Vasc. Perm. 1. Wills BA et al J Infect Dis 190: , 2004
Infections with dengue viruses result in a spectrum of responses, from inapparent to mild to severe (vascular permeability syndrome). Principal determinants of clinical outcomes are age and immune status.

8. DENGUE HEMORRHAGIC FEVER/DENGUE SHOCK SYNDROME (DHF/DSS) Dengue vasculopathy
The term dengue hemorrhagic fever derives from early outbreaks in the Philippines, thought to be Korean hemorrhagic fever (HF with renal syndrome). GI hemorrhages were common in early outbreaks because the need for rapid fluid replacement therapy was not recognized. Over time, the incidence and severity of hemorrhaging has declined. The central pathophysiological defiicit is vascular permeability.

9. DHF/DSS occurs predominantly in children, particularly in well nourished children. Vascular permeability is the significant pathophysiological phenomenon and hemorrhagic signs may be absent (see petechium mid-forehead)

10. DSS GRADE IV

11. DSS GRADE III
Classical dengue shock syndrome (incorrectly labeled as grade II) illustrated by narrow pulse pressure at time of peak elevation of hematocrit (hypovolemia) thrombocytopenia, pleural effusion after initiating fluid repletion. Note short time course, complete reversal of vascular leak within hours. There is grave danger in overhydration.

12. WHY IS DENGUE SUCH A BIG PROBLEM TODAY?
Global population growth
Rural to urban migration
Growth of cities
Deterioration of cities
Jet travel
Health services poorly organized/ underfunded
Lack of vector control professionals
Dengue is a quintessential 20-21st century emerging disease. By 1960, Aedes aegypti had been largely eradicated from the American tropics but after abandoning vector control programs, the mosquito quickly returned. Viral infections were accelerated by the demographic forces described in this slide.

13. Global Spread of Dengue
It has been estimated that in recent years as many as 10 million cases of dengue fever and 500,000 cases of DHF/DSS occur annually. Case fatality rates in some countries reach 5%
million infections/year
Countries with active dengue +
Aedes aegypti

14. WHY TWO SYNDROMES, BENIGN and SEVERE?
Observed in two immunological settings.
Primary infections in infants.
2. Secondary infections in children
and adults.
Virtually all DHF/DSS cases are observed in two epidemiological settings, in individuals circulating dengue antibodies passively acquired (infants less than one year) and from previous dengue infection(s). Despite numerous reports, epidemic DHF/DSS has never been associated with a virgin-soil dengue epidemic (Barnes & Rosen L. AJTMH ). Most DHF/DSS cases associated with primary infections occur in areas where two or more dengue viruses are endemic or have been sequentially epidemic. Because dengue vascular permeability is difficult to detect, controversies largely stem from inaccurate case ascertainment.

15. Two-infections The epidemiological data
DHF documented in children (> 1 yr) who circulate infection-acquired dengue antibody. Four prospective cohort and 6 prospective population-based studies.
In most studies, DHF comprises 2-5% of secondary infections
There is ample evidence supporting the role of antibodies as risk factors for DHF/DSS. These data have been obtained from rigorous prospective cohort studies and prospective population-based studies in defined populations.

16. DHF IN CHILDREN: PROSPECTIVE COHORT STUDIES
In 3 of 4 cohort studies the incidence of disease recognized as severe varied between 3.5 and 6% of secondary dengue infections. These studies were conducted over a period of 30 years in different populations and with different viruses circulating.

17. DHF IN CHILDREN: PROSPECTIVE POPULATION-BASED STUDIES
The Halstead study was conducted in Bangkok in , that of Russel et al on the island of Kah Samui in the Gulf of Siam, Sangkawibha is a neutralization test monitored cohort living in Rayong, Thailand in 1980.

18. DHF IN CHILDREN: PROSPECTIVE POPULATION- BASED STUDIES
Guzman et al, 1990 describes a retrospective cohort chosen to study the 1981 Cuban dengue 2 epidemic; Thein et al describe DHF/DSS cases a selected area of Yangon, Myanmar and Guzman et al, 2000 is a study of the 1997 dengue 2 epidemic in Santiago de Cuba.

19. SEQUENTIAL DENGUE INFECTIONS
Two infections can occur in twelve
possible combinations.
Sequential infections with the four dengue viruses can occur in 12 possible combinations.

20. Established second infection sequences leading to DHF
2 – 1 Thailand; Indonesia
3 – 1 Thailand
1 – 2 Cuba, 1981; Cuba 1997; Thailand
3 – 2 Thailand
4 – 2 Thailand
1 – 3 Cuba, 2001; Thailand; Indonesia
2 – 3 Thailand, DF in Cuba
1 – 4 Thailand
2 – 4 Indonesia
3 – 4 Thailand
The most frequently observed causes of DHF/DSS are infections in the combination DENV1 then DENV 2 (Asian genotypes) and DENV1 then DENV 3.

21. No data 4 – 1 4 – 3 KALAYANROOJ S et al AJTMH 2008 in press.
Only these two sequences have not been documented as causes of severe disease accompanying a second dengue infection.
KALAYANROOJ S et al AJTMH 2008 in press.

22. Third infections: resulting in DHF
1 – 3 – 2 Thailand
MAMMAN MP personal communication
No DHF
In Cuba, sequential infections with dengue types 1, 2 and 3 occurred in , 1981 and , respectively. Severe cases accompanying a third dengue viral infection were not detected.
1 – 2 – 3 Cuba, 2001.
GUZMAN MG personal communication

23. DENGUE VIRUSES, BANGKOK 1973 - 2001
Distribution of confirmed hospitalized DHF cases at 2 month intervals from 1973 through There has been general upward trend with larger outbreaks at occuring at 2 – 5 year intervals. Peak hospitalization rates occur during the rainy season month of July – August. Colored graphs show isolations of different dengue viruses demonstrating off-set outbreaks of DENV 3 and 4 compared with DENV 1 – 2.

24. Lags at Which Correlation Between Bangkok and other Provinces
Is Maximized
p<1e-8
~148 km/month
(months)
A mathematical model of reported hospitalization for DHF throughout Thailand gives evidence that different viruses move out of Bangkok in “travelling waves” moving at 148 km per month (Cummings DT et al. Nature 427: 344-7, 2004)

25. DHF AT BANGKOK CHILDRENS HOSPITAL
1O INFECT.
2O INFECTIONS
Decadal average age distribution of hospitalized DHF in Bangkok during the 1970s, 80s and 90s. Two features are notable: 1) bimodal distribution, with one mode among infants; 2) modal age of hospitalization steadily moves up (from 5 to 9 years). DHF in infants accompanies primary infections while DHF in children 1 year and older accompanies secondary dengue infections.

26. The endemicity of the four dengue infections in Bangkok during the 1960s resulted in this age-specific prevalence of dengue antibodies in a large randomly selected population. While infection rates vary markedly in different years (measured at 42% in 1962) am average annual dengue infection rate of 15% produces this prevalence curve.

27. Fischer and Halstead Yale J Biol Med 42:329-349,1970
The age distribution of DHF cases (Bangkok) is uniquely related to second dengue infections. The dotted line above is the predicted age distribution of cases during a second dengue infection assuming all ages are at equal intrinsic risk to DHF.
Fischer and Halstead Yale J Biol Med 42: ,1970

28. Fischer and Halstead Yale J Biol Med 42:329-349,1970
Predicted age distribution of triple sequential dengue infections does not fit the observed age distribution of hospitalized DHF cases.
Fischer and Halstead Yale J Biol Med 42: ,1970

29. Fischer and Halstead Yale J Biol Med 42:329-349,1970
However, when the higher intrinsic risk to DHF in younger children is considered, the predicted secondary dengue infection age distribution (dotted line) fits with the observed (solid line).
Fischer and Halstead Yale J Biol Med 42: ,1970

30. DHF AT BANGKOK CHILDRENS HOSPITAL
Attention is directed to the shift in modal age of hospitalization from age 5 in the 1970s, 6 in the 1980s and 8 in the 1990s. Increasing numbers of adult cases of DHF have also been observed over this time period.

31. EFFECT OF AVERAGE FORCE OF INFECTION (Ro) ON
AGE SPECIFIC SECONDARY INFECTION INCIDENCE
Ro = 30%
Ro = 20%
Ro = 10%
This shift in age at occurrence of second dengue infections where four dengue viruses circulate is diirectly related to force of infection (infection rate). Low annual infection rates retard the probability of experiencing second dengue infections compared with higher infection rates.

32. Dengue hemorrhagic fever/dengue shock syndrome has occurred in some (but not all) dengue epidemics since the 1950s, Why?

33. DHF does not occur if antibodies from first infection neutralize the second infecting virus.
An important controlling mechanism during second dengue infections is the degree of neutralization of the heterotypic second infecting virus by antibodies raised following the first infection.

34. BANGKOK STUDY Kliks et al AJTMH 40:444, 1989.
40 Bangkok school children had documented secondary DEN 2 infections (pre-infection blood sample contained dengue antibodies).
7 were hospitalized; 33 silent.
Undiluted pre-infection sera tested for neutralization or enhancement in human PBL cultures.
This phenomenon was first demonstrated in a study of Bangkok school children, 40 of whom experienced a secondary dengue infection, probably with dengue 2 virus. Seven children were hospitalized during this infection; 33 secondary infections were silent. Pre-secondary infection serum samples had been collected from all children. These were tested for ability neutralize or enhance dengue infections in cultures of monocytes from dengue-susceptible human donors.

35. ADE AND DHF BLOCKED BY NEUTRALIZING ANTIBODIES
Twenty-nine of 33 sera from children experiencing a silent dengue infection neutralized dengue 2 in human monocyte cultures; 6 of 7 sera from children experiencing overt disease enhanced or failed to neutralize dengue 2 virus. Children with high levels of heterotypic neutralizing antibodies were protected from severe disease accompanying secondary dengue infections. Heterotypic antibodies provide partial, not complete protection.

36. ANTIGENIC STRUCTURE OF VIRUS: IQUITOS STUDY
School children cohorts followed from 1990 until now.
DEN 1 transmitted in
DEN 2 transmitted from 1995.
Prevalence of neutralizing antibodies measured in 1993, 1994 and 1995 cohorts.
In 1995, secondary DEN 2 infection rate estimated at 60.5%
A second example of the protective role of heterotypic dengue neutralizing antibodies was observed in 1995, when dengue 2 (American genotype) was introduced into Iquitos, Peru, a city of 300,000 on the Amazon River. Previously, dengue 1 had been endemic. Serial infections with dengue 1 then 2 in Cuba in 1981 had resulted in a large outbreak of DHF/DSS. In Iquitos, 60.5% of the population were infected in the sequence dengue 1 then dengue 2.

37. NO DHF with Secondary DEN 2 (American genotype) infections
Total population, yrs-old = 81,479.
Total 2ndary DEN 2 infections = 49,266.
Estimated hospitalized DHF =
Estimated deaths =
DHF cases observed = 0
Watts DM et al Lancet 354:1431-4, 1999
This resulted in an estimated 49,000 secondary dengue 2 infections in Based upon data from other dengue endemic countries, this number of secondary dengue 2 infections might have yielded up to 10,000 DHF cases and perhaps 200 deaths. However, no severe dengue cases were observed.

38. NEUTRALIZATION OF AMERICAN GENOTYPE DEN 2 VIRUSES by 34 DEN 1- IMMUNE HUMAN SERA
It was discovered that dengue 1 antibodies were able to neutralize virtually all American genotype dengue 2 viruses to high titer. Geometric mean titers observed in 34 dengue 1 sera when tested with and American genotype dengue 2 virus was 1:262 compared with 1:875 versus homologous dengue 1 virus. However, two strains of Asia genotype dengue 2 (one isolated in the Americas and the other in Asia) were poorly neutralized by antibodies to dengue 1.

39. ONE-WAY CROSS: 17 DENGUE 2-IMMUNE SERA DO NOT NEUTRALIZE DENGUE-1 VIRUSES
However, although dengue 2 viruses appear to possess an epitope that reacts with dengue 1 antibodies, infection of humans with the American genotype dengue 2 virus did not raise antibodies against dengue 1 (shown here with American and Asian isolates)

40. American genotype dengue 2 viruses are neutralized in vitro by human antibodies to dengue 1
BUT …
dengue 1 antibodies do not prevent but
may down regulate dengue 2 infections
The epidemiological observations from Iquitos in 1995 can be explained by the laboratory findings summarized in this slide. Other workers hypothesize that American genotype dengue 2 is replication defective producing low titered infections in mosquitoes and in humans. However, this virus has been circulating in the Americas for possibly hundreds of years and arrived in Peru earlier than did the allegedly more replication-competent Asian dengue 2 virus (Rico-Hesse R. Microevolution and virulence of dengue viruses. Adv Virus Res 59: , 2003)