| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES: |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES: |
|---|
Bangkok is the epicenter of dengue in Thailand and the site where the original clinical descriptions of DHF and its association with immunologic priming by a previous dengue infection (secondary dengue infection) were made.3,8–11 The Queen Sirikit National Institute of Child Health (QSNICH), formerly Bangkok Children’s Hospital, has admitted the greatest number of dengue cases for 40 years because it is government-financed, accessible to all, and a widely recognized center of excellence. Since 1973, the QSNICH and the Armed Forces Research Institute of Medical Sciences have collaborated to conduct longitudinal surveillance of dengue in Bangkok. In this report, we present the results of that collaboration to illustrate the long-term circulation and associations with disease severity and large epidemics of the four dengue virus serotypes.
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES: |
|---|
From 1973, children younger than 18 admitted with suspected dengue had acute and convalescent sera obtained for serology and virus isolation. The attending physician assigned DHF grade at the time of discharge using World Health Organization (WHO) criteria.12 Statistical analysis was performed using SPSS for Windows version 10.0 (SPSS Inc.).
Moderate to severe annual dengue epidemics.
The national annual dengue incidence surveillance was used to classify each year’s epidemic as moderate, or severe6 (Table 1
). Dengue incidence per 100,000 was ranked into percentiles; the 50th was 71, the 75th was 134, and 90th was 175. Annual epidemics with an incidence ranging from the 75th to 90th percentile were classified as moderately severe, and those greater than the 90th percentile as severe. Accordingly, there were moderately severe epidemics in 1984, 1985, 1989, 1990, and 1997, and severe epidemics in 1987 and 1998.
|
).
|
Statistical analysis.
Statistical analysis was performed using SPSS for Windows version 10.0 (SPSS Inc.). Student’s t test, analysis of variance, and Pearson’s correlation or linear regression are used to determine differences or associations among continuous variables, and 
2 or binomial test for proportions. Backward logistic regression was used to assess the independent influence of a number of factors on dengue disease severity.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES: |
|---|
Specificity of clinical diagnosis. There was no correlation between year and the number of laboratory-confirmed acute dengue cases of all hospitalized suspected dengue, suggesting that changes over time were not due to changes in hospital admission criteria. There was a correlation between the reported national annual disease incidence and the specificity of diagnosis of children diagnosed as suspected dengue by clinical criteria in our series (r = 0.424, P = 0.028). This suggests that the clinical criteria for acute dengue infection were unchanged and that the likelihood of a child’s having acute dengue depended on the incidence of disease in the population.
Dengue virus isolation rates.
The dengue virus isolation rate was calculated as the number of isolates divided by the number of dengue cases. No more than one dengue serotype was recovered per patient, in contrast to other reports.16 The overall isolation rate in children was 33% and increased during the study period (ß = 0.84, P < 0.0001 by linear regression) from 21% in 1973 to a peak of 57% in 1998 (Figure 1).
|
) and the nation (r = 0.852, P < 0.0001).6
|
). Before 1979, few dengue cases occurred during the cool, dry winter months of December and January. Since 1979, dengue cases have occurred year-round. From 1973 to 1986, DEN-2 was the dominant dengue isolate. In 1987, for the first time, DEN-3 became the most frequent isolate. DEN-2 regained predominance in 1988–89, followed by DEN-1 in 1990–92, DEN-4 in 1993–94, and DEN-3 in 1995–9. Each dengue serotype appears to have caused several clearly distinguishable outbreaks at 7- to 9-year intervals.
|
SD from the overall mean, with a small increase in mean age over time (ß = 0.49, P = 0.01 by linear regression). The overall modal age was 6 years, which also increased over time from 5 years during 1973–79 to 8 years during 1990–99 (ß = 0.50, P = 0.008 by linear regression). There were proportionately more males (52%) than females (48%) (P < 0.001 by binomial test).
Antibody response patterns.
Of all dengue virus infections, 16% were primary dengue infection (Table 3). The proportion with primary dengue increased over time with a 3- to 4-year cyclical oscillation in the rate of primary dengue (Figure 4) (ß = 0.67 by linear regression, P < 0.0001). The mean age of children with primary dengue was 6.6 years (SD 3.5) compared with 7.8 years (SD 3.3) for secondary dengue (P < 0. 001, 2-tailed Student’s t test).
|
|
). There was a correlation between the yearly proportion of primary dengue cases and the isolation of DEN-1 (R = 0.5, P = 0.0007 by Pearson correlation) but no other type, suggesting that Bangkok DEN-1 strains were especially pathogenic among children without immune priming from prior flavivirus infection.
DEN-2 was the most frequent isolate in secondary dengue (37%) followed by DEN-3 (29%), DEN-1 (22%), and DEN-4 (12%). There was a correlation between the yearly proportion of secondary dengue cases and the isolation of DEN-2 (R = 0.5, P = 0.004 by Pearson correlation) but no other type, suggesting that Bangkok DEN-2 strains were more pathogenic among children with immune priming from prior flavivirus infection. Overall, DEN-4 was isolated more often from secondary dengue cases (12%) than primary (2%) (P < 0.0001, 2). DEN-3 was the most frequent isolate among combined primary and secondary cases, suggesting that Bangkok strains combined pathogenicity attributes of DEN-1 and DEN-2.
Factors associated with severe dengue outbreak years. DEN-1, DEN-2, or DEN-3 was the dominant isolate during moderately severe dengue outbreak years. In contrast, only DEN-3 was the dominant isolate in the two severe outbreaks. These data further suggest that Bangkok dengue strains have differing ability to elicit severe disease and that DEN-3 has been the most successful.
Dengue antibody pattern and dengue disease severity.
There were 2,396 children (16%) hospitalized with a diagnosis of DF and 11,731 (80%) with DHF (Table 4). Of those with DHF, 16% were diagnosed with DHF grade I, 32% with DHF grade II, 45% with DHF grade III and 7% with DHF grade IV.
|
2 for linear trend). Secondary dengue DHF was associated with more-severe DHF: grade I (15% of all secondary DHF), grade II (31%), grade III (47%), and grade IV (7%) (P < 0.0001 by 2 for linear trend). The proportion of DHF due to primary infection varied yearly from a low of 1% in 1978 to a high of 14% in 1997 and has increased over time (ß = 0.61, P = 0.001 by linear regression).
Dengue serotypes and dengue disease severity.
The frequency of isolated dengue serotypes was examined by clinical severity (Table 4). DEN-3 was the most frequent isolate in DF patients (39% of all DF viral isolates) followed by DEN-1 (35%), DEN-2 (17%), and DEN-4 (9%). The frequency of isolated dengue serotypes differed between primary and secondary DF. DEN-3 and DEN-1 were the most frequent isolated dengue serotypes in primary DF (46% each of all primary DF viral isolates) followed by DEN-2 (6%), and DEN-4 (2%). DEN-3 was the most frequent isolated dengue serotype in secondary DF (34% of all secondary DF viral isolates) followed by DEN-1 (29%), DEN-2 (24%), and DEN-4 (13%).
Among DHF cases in children, DEN-2 was the most frequent isolate (35% of all DHF viral isolates) followed by DEN-3 (31%), DEN-1 (24%), and DEN-4 (10%). The frequency of isolated dengue serotypes differed between primary and secondary DHF. DEN-3 was the most frequently isolated dengue serotype in primary DHF (50% of all primary DHF viral isolates) followed by DEN-1 (44%), DEN-2 (4%), and DEN-4 (2%). DEN-2 was the most frequently isolated dengue serotype in secondary DHF (40% of all secondary DHF viral isolates) followed by DEN-3 (28%), DEN-1 (20%), and DEN-4 (12%).
To eliminate the interactions among serologic status, severe dengue years, and dengue serotypes on the presence of DHF, backward logistic regression analysis was used to assess the independent influence of each variable. Secondary dengue (log odds ratio [LOR] 3.8, 95% confidence interval [C.I.] 2.8, 5.1) and dengue serotype (DEN-1 [LOR 1.7, C.I. 1.5, 1.9], DEN-3 [LOR 1.5, C.I. 1.3, 1.7], and DEN-4 [LOR 1.5, C.I. 1.2, 1.9]) each independently contributed to the presence of DHF when the other factors were controlled. Severe dengue year independently contributed to the absence of DHF (LOR 0.8, C.I 0.8, 0.9) and DEN-2 did not independently contribute to DHF (LOR 0.8, 0.7, 1.0).
The overall case-fatality for all years was 0.5%. The peak case-fatality rate was 1.6% during the 1987 severe outbreak. Overall, there was a gradual but not significant decline in case-fatality, with a rate of 0.2% during the last 5 years of the study (ß = -0.19, P = 0.33 by linear regression). Case-fatality for primary and secondary dengue did not differ (0.6% and 0.4% respectively, P = 0.27 by 2) and was similar also among serotypes.
Dengue in infants.
From 1973 to 1999, there were a total of 773 infants with acute dengue virus infection (Table 5). The mean age was 7 months (range 1–12 months, SD 2.4 months) with a modal age of 6 months and an equal representation of males and females (51% and 49%, respectively). The peak number of cases occurred from 6–8 months of age (Figure 5). The total case-fatality among infants with confirmed acute dengue was 11/773 (1.4%) and was greater than the 0.5% case-fatality for all years in children (odds ratio 3.1, 1.5<OR<6.1, P < 0.0001). No trend was noted between year and case-fatality, with only 8 years having deaths from acute dengue.
|
|
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES: |
|---|
Our findings extend the early observations of dengue transmission in Thailand and illustrate the constancy in dengue disease severity and serotypes over time despite an increase in overall disease incidence and virus transmission. In 1962, all four dengue serotypes were isolated in the Bangkok area and associated with the occurrence of severe dengue disease.17–19 Our observations extend these observations by demonstrating the dramatic periodic shifts in the occurrence of the predominant dengue serotype over time.
Multiple factors influence dengue transmission, the cycle of epidemics, and the emergence of new dominant serotypes. Vector density and infection rates reflect access to breeding sources and environmental temperature changes due to urbanization and the El Nino phase of the Southern Oscillation.20–23 Herd susceptibility/immunity and other intrinsic population characteristics are additional factors that presumably influence the emergence of new serotypes and drive high disease rates that create epidemics.24,25
The finding that the yearly proportion of primary dengue cases is correlated to the isolation of DEN-1 and no other dengue serotype suggests that Bangkok DEN-1 strains were more pathogenic among children without immune priming from prior flavivirus infection. The occurrence of more DEN-1 over time may in part explain why the proportion of hospitalized children with primary dengue has increased in our population. Our findings with DEN-1 illustrate the fact that each serotype has unique characteristics manifested in how the disease presents in the population.
Infants represent a unique susceptible population for severe dengue disease. The study of dengue in this group has led to many of our current theories on dengue pathogenesis, maternal antibody, and dengue antibody enhancement.14,15,26,27 Our series demonstrated that case-fatality among infants with confirmed acute dengue was three times greater than that of other children. This may reflect the susceptibility of this population toward complications that arise from plasma leakage and shock syndrome. This may be due to the small amount of volume reserve that exists in infants when plasma leakage occurs, under-recognition of potential DHF in febrile infants, or overtreatment with aggressive volume replacement when volume depletion occurs. The isolation of DEN-2 and DEN-3 as the most frequently isolated virus in infants, as it is for secondary dengue infection in other children, suggests that the characteristics of these viruses in producing severe dengue disease in the host relies more on the presence of heterotypic dengue antibody than immunologic priming of cellular or T-cell immunity.
This frequency of dengue serotype isolation is summarized for all years as DEN-3 > DEN-2 > DEN-1 > DEN-4; for primary dengue, as DEN-3 > DEN-1 > DEN-2 > DEN-4; for secondary dengue, as DEN-2 > DEN-3 > DEN-1 > DEN-4; for all DF patients, as DEN-3 > DEN-1 > DEN-2 > DEN-4; and for all DHF, as DEN-2 > DEN-3 > DEN-1. The high degree of correlation of DEN-1 and DEN-2 isolation rates, and its absence with DEN-3, with primary and secondary dengue suggests a propensity of these serotypes to associate with these serologic responses. Our results also indicate that DEN-4 is unusual in that it is the least frequently isolated serotype in any dengue disease state and is not associated with severe dengue outbreaks, and that 97% of isolates are from secondary dengue infections. These observations underscore the importance of viewing each dengue serotype as a separate virus that has evolved with its own unique viral characteristics and propensity to produce disease in the presence or absence of heterotypic dengue antibody.
Examination of severe dengue disease, DHF, in our series demonstrates that it is primarily a condition of secondary dengue infection, with primary DHF in children associated with milder disease. Dengue serotype isolation and dengue disease severity demonstrate that for children with DF, DEN-3 > DEN-1 > DEN-2 > DEN-4, and for children with DHF, DEN-2 > DEN-3 > DEN-1 > DEN-4.
Over time, there was a subtle change in the demographics of children with hospitalized dengue, with a slight increase in mean age from 7.4–7.8 years and an increase in the modal age from 5 years in 1973 to 8 years in 1999. There also was an increase in the percentage of hospitalized primary dengue over time. These findings suggest that children are getting dengue infection at a later age. These observed changes may be due to the overall improvement in the standard of living in Thailand. Improved housing standards, screens on windows, and reduction of mosquito breeding containers around homes may all contribute to decreasing the intensity of potential exposure to dengue-infected mosquitoes in young children. Dengue infection occurs at an older age, though, as intensity of exposure to dengue-infected mosquitoes increases with greater exposure to the outside environment at home or school. Symptomatic dengue virus infection may be age-dependant and produce more-severe disease in older children and thus might explain the observation that the proportion of hospitalized children with primary dengue is increasing.
Our results suggest that dengue serotypes and their isolation from a spectrum of serologic responses to dengue infection and disease severity is not a random event, and that specific dengue serotypes are more frequently isolated in specific dengue disease states. Our results demonstrate the complexity of the dengue virus transmission and its interaction with the host response, and have important implications in identifying factors to predict severe dengue years, developing interventions to control dengue outbreaks, and developing an effective dengue vaccine.
This longitudinal serotype-specific study of dengue virus over 27 years is unique for its long-term characterization of dengue serology, disease severity, and circulating dengue serotypes and serves as a surveillance model to be instituted by other countries faced with the growing threat of dengue virus infection.
Received January 17, 2002. Accepted for publication August 29, 2002.
Acknowledgments: We thank the doctors and nurses at the Queen Sirikit National Institute of Child Health for their outstanding patient care and participation in this study, and the staff members at the Department of Virology, Armed Forces Research Institute of Medical Sciences, over the past 27 years who have carefully performed diagnostic testing, data entry, and specimen archival. We also thank Dr. Daniel Libraty for his critical reading of this manuscript and Dr. Scott Halstead for sharing his thoughts and vast knowledge base, and giving his comments and suggestions on improving the manuscript.
Financial support: This work was funded by the U.S. Army Medical Research and Materiel Command, Fort Detrick, MD.
The opinions or assertions contained herein are the private views of the authors and are not to be construed as reflecting the official views of the United States Army or the Department of Defense.
Reprint requests: Timothy P. Endy, Division of Virology, U.S. Army Medical Research Institute for Infectious Diseases, 1425 Porter St., Fort Detrick, MD 21702-5011
|
REFERENCES: |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES: |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. Thomas, O. Verlaeten, A. Cabie, S. Kaidomar, V. Moravie, J. Martial, F. Najioullah, Y. Plumelle, C. Fonteau, P. Dussart, et al. Influence of the Dengue Serotype, Previous Dengue Infection, and Plasma Viral Load on Clinical Presentation and Outcome During a Dengue-2 and Dengue-4 Co-Epidemic Am J Trop Med Hyg, June 1, 2008; 78(6): 990 - 998. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Nagao and K. Koelle Decreases in dengue transmission may act to increase the incidence of dengue hemorrhagic fever PNAS, February 12, 2008; 105(6): 2238 - 2243. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.-C. Wang, C.-C. Wu, J.-W. Liu, A.-S. Lin, S.-F. Liu, Y.-H. Chung, M.-C. Su, I.-K. Lee, and M.-C. Lin Chest Radiographic Presentation in Patients with Dengue Hemorrhagic Fever Am J Trop Med Hyg, August 1, 2007; 77(2): 291 - 296. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.-C. Wang, S.-F. Liu, S.-C. Liao, I.-K. Lee, J.-W. Liu, A.-S. Lin, C.-C. Wu, Y.-H. Chung, and M.-C. Lin Acute Respiratory Failure in Adult Patients with Dengue Virus Infection Am J Trop Med Hyg, July 1, 2007; 77(1): 151 - 158. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Clyde, J. L. Kyle, and E. Harris Recent Advances in Deciphering Viral and Host Determinants of Dengue Virus Replication and Pathogenesis J. Virol., December 1, 2006; 80(23): 11418 - 11431. [Full Text] [PDF]
|
|
|
|
|
|
B. Adams, E. C. Holmes, C. Zhang, M. P. Mammen Jr, S. Nimmannitya, S. Kalayanarooj, and M. Boots Cross-protective immunity can account for the alternating epidemic pattern of dengue virus serotypes circulating in Bangkok PNAS, September 19, 2006; 103(38): 14234 - 14239. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Zhou, M. P. Mammen Jr, C. Klungthong, P. Chinnawirotpisan, D. W. Vaughn, S. Nimmannitya, S. Kalayanarooj, E. C. Holmes, and C. Zhang Comparative analysis reveals no consistent association between the secondary structure of the 3'-untranslated region of dengue viruses and disease syndrome J. Gen. Virol., September 1, 2006; 87(9): 2595 - 2603. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. J. Wearing and P. Rohani Ecological and immunological determinants of dengue epidemics PNAS, August 1, 2006; 103(31): 11802 - 11807. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Zhang, M. P. Mammen Jr, P. Chinnawirotpisan, C. Klungthong, P. Rodpradit, A. Nisalak, D. W. Vaughn, S. Nimmannitya, S. Kalayanarooj, and E. C. Holmes Structure and age of genetic diversity of dengue virus type 2 in Thailand. J. Gen. Virol., April 1, 2006; 87(Pt 4): 873 - 883. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. BALMASEDA, S. N. HAMMOND, L. PEREZ, Y. TELLEZ, S. I. SABORIO, J. C. MERCADO, R. CUADRA, J. ROCHA, M. A. PEREZ, S. SILVA, et al. SEROTYPE-SPECIFIC DIFFERENCES IN CLINICAL MANIFESTATIONS OF DENGUE. Am J Trop Med Hyg, March 1, 2006; 74(3): 449 - 456. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. PODDER, R. F. BREIMAN, T. AZIM, H. M. THU, N. VELATHANTHIRI, L. Q. MAI, K. LOWRY, and J. G. AASKOV ORIGIN OF DENGUE TYPE 3 VIRUSES ASSOCIATED WITH THE DENGUE OUTBREAK IN DHAKA, BANGLADESH, IN 2000 AND 2001 Am J Trop Med Hyg, February 1, 2006; 74(2): 263 - 265. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Aaskov, K. Buzacott, H. M. Thu, K. Lowry, and E. C. Holmes Long-Term Transmission of Defective RNA Viruses in Humans and Aedes Mosquitoes Science, January 13, 2006; 311(5758): 236 - 238. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Zhang, M. P. Mammen Jr., P. Chinnawirotpisan, C. Klungthong, P. Rodpradit, P. Monkongdee, S. Nimmannitya, S. Kalayanarooj, and E. C. Holmes Clade Replacements in Dengue Virus Serotypes 1 and 3 Are Associated with Changing Serotype Prevalence J. Virol., December 15, 2005; 79(24): 15123 - 15130. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. A. T. Cummings, I. B. Schwartz, L. Billings, L. B. Shaw, and D. S. Burke Dynamic effects of antibody-dependent enhancement on the fitness of viruses PNAS, October 18, 2005; 102(42): 15259 - 15264. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Modis, S. Ogata, D. Clements, and S. C. Harrison Variable Surface Epitopes in the Crystal Structure of Dengue Virus Type 3 Envelope Glycoprotein J. Virol., January 15, 2005; 79(2): 1223 - 1231. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G N Malavige, S Fernando, D J Fernando, and S L Seneviratne Dengue viral infections Postgrad. Med. J., October 1, 2004; 80(948): 588 - 601. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
