| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Antibody-dependent enhancement of dengue virus infection leading to increased viral load is a leading theory for severe diseases caused by dengue virus. Increased levels of circulating dengue virus have been documented in DHF cases compared with DF cases.4–6 However, the immunologic events that occur in the early stages of infection that lead to severe disease have not been characterized in patients. In the absence of an animal model for DHF, researchers are forced to try to characterize early immunologic events in patients with recently acquired dengue virus infections. A major limiting factor in such an investigation is identifying individuals prior to the onset of their clinical illness so that early events can be characterized. The current study was designed to overcome that problem by identifying individuals very early in the infection. Family members and nearest neighbors of index cases hospitalized with acute dengue virus infections were enrolled and followed longitudinally to determine the incidence rate of dengue virus infection. This proof-of-concept study was conducted to evaluate the hypothesis that monitoring persons clustered around an index case would be an efficient and practical way to obtain blood samples before, during, and after dengue infection.
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Dengue virus laboratory tests. Serum samples were tested for the presence of IgM antibodies to dengue virus using commercial enzyme-linked immunosorbent assay (ELISA) kits (Focus Technologies, Cypress, CA). Hemagglutination inhibition (HI) assays and plaque reduction neutralization tests (PRNTs) were also performed to confirm dengue virus infection and to classify the infection as primary or secondary based on the antibody response.9 Dengue virus RNA in blood was detected using a reverse transcriptase–polymerase chain reaction (RT-PCR) assay.10 Blood samples collected from symptomatic volunteers during the acute stage of febrile illness and from suspected asymptomatically infected volunteers were tested for virus by culture in C6/36 cells, and virus was identified using serotype-specific monoclonal antibodies for dengue virus.11
Study design. Each cluster of volunteers was identified by an index case from the pediatric ward at Sumber Waras Hospital in West Jakarta. Physicians at this local community hospital have a well-established relationship with the surrounding community and are experienced in clinically diagnosing dengue virus infections. For the index cases, children 4–14 years of age were recruited. Ward clinicians were asked to select 1–2 cases per week based on clinical manifestations, together with the detection of IgM antibody to dengue virus by ELISA and/or dengue viral RNA by RT-PCR. In most cases, these diagnostic assays were performed within 48 hours of index case identification. Other considerations in selecting index cases included automobile accessibility to their community and the willingness of the family to participate.
Usually within 48 hours of index case identification, family members and nearest neighbors greater than four years old who lived within a 10-meter radius of the index case’s home were invited to participate. Volunteers who were afflicted with or who had a history of severe anemia, bleeding disorder, or any known immunologic disorder were excluded from the study. On enrollment, we measured the otic temperature, and collected blood samples and demographic data from each study volunteer. Volunteers were followed for 14 consecutive days for fever and other clinical manifestations suggestive of acute dengue virus infection. During the monitoring period, we collected additional blood samples from each volunteer every 2–3 days. We also collected blood specimens from each volunteer who developed a fever or dengue-like signs or symptoms and processed them for a diagnosis of dengue. To confirm that dengue virus was the cause of fever, we tested samples within 24–48 hours of collection for dengue viral RNA by RT-PCR and for IgM antibody to dengue virus by ELISA. If a sample collected from a febrile volunteer was positive for IgM antibodies to dengue virus, the enrollment blood sample for that volunteer was tested for IgM antibodies to dengue virus to determine if seroconversion had occurred. These laboratory tests were performed at the Virology Laboratory of the U.S. Naval Medical Research Unit No. 2. Blood samples treated with EDTA were also obtained at the onset of fever and sent to Sumber Waras Hospital for hematocrit and platelet count determinations.
Febrile volunteers with a positive IgM ELISA or RT-PCR result were encouraged to be hospitalized for close monitoring and serial laboratory tests. Volunteers refusing hospitalization were monitored closely as outpatients until they were afebrile for two consecutive days. Convalescent blood samples were obtained two weeks later and again at six months.
Hospitalized volunteers were bled daily until two days after defervescence, two weeks later, and approximately six months after discharge from the hospital. Blood samples were routinely tested for hemoglobin, hematocrit, platelets, white blood cells, protein, and albumin. A tourniquet test was done on admission and daily ultrasound examinations for ascites and pleural effusions were performed until two days after defervescence to detect evidence of plasma leakage. We classified hospitalized volunteers as DF or DHF (World Health Organization, 1997)12 following evaluation of the clinical data.
We categorized all volunteers clustered around the index cases into one of three groups: non-dengue (ND) infection, dengue infection at enrollment (ED), or post-enrollment dengue (PED) infection. A PED case was defined as a volunteer who developed fever after enrollment and demonstrated seroconversion for IgM antibodies to dengue virus and/or dengue viremia by RT-PCR or virus isolation. Volunteers who developed fever but were negative for IgM antibodies to dengue virus or viremia were later classified as PED infections if there was a four-fold increase in the HI antibody titer to dengue virus between enrollment and the two-week post-enrollment blood samples. For equivocal HI results (less than a four-fold increase in titer), pre-enrollment and two-week post-enrollment PRNT50 titers were compared for a 
four-fold increase against one or more dengue virus serotypes. For volunteers who never developed fever, laboratory evaluation of serum samples (dengue serology and virus detection) was still conducted throughout the 14-day monitoring period. Asymptomatic volunteers who were positive for viremia and/or showed a four-fold increase in HI titer between the enrollment and two-week post-enrollment blood samples were also classified as PED cases.
Volunteers that demonstrated evidence of infection (IgM, RT-PCR, or virus isolation), with or without symptoms at the time of enrollment were classified as ED cases. All other volunteers were classified as ND cases.
The U.S. Naval Medical Research Unit No. 2 Institutional Review Board and the Indonesian Ministry of Health Ethical Review Committee reviewed and approved the human use protocol (DoD# 30861) used in the study, in compliance with all U.S. federal regulations governing the protection of human subjects. Informed written consent was obtained from all participants, and if < 15 years of age, from a parent or legal guardian.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and Figure 1). Infections by all dengue virus serotypes were identified among the index cases. The predominant serotype identified was DEN-1 (10), followed by DEN-2 (8), DEN-3 (5), and DEN-4 (2). The serotype was not determined in 28 index cases. None of the index dengue infections was fatal.
|
|
shows the demographics of the volunteers in the study.
Among the 785 cluster volunteers, 175 (22.3%) were classified as ED cases. Table 1 also shows the demographics of ED cases. Eleven ED cases were found to be viremic for dengue at enrollment by the RT-PCR. The DEN-1 serotype was identified in three cases and DEN-2 in eight cases. Confirmatory virus isolation was accomplished in 5 of the 11 cases. The remaining 164 ED cases had evidence of recent dengue virus infection, as indicated by IgM antibodies to dengue virus at enrollment and were classified as convalescent infections (Figure 1).
Nine of the 11 viremic ED cases had fever at enrollment. The nine febrile volunteers were hospitalized for close monitoring within 24–48 hours of enrollment after dengue serologic and/or RT-PCR results were known. Four of the febrile hospitalized ED cases were subsequently diagnosed with DHF (three grade I and one grade II). All four DHF cases were caused by a DEN-2 virus. Analysis of HI and PRNT titers suggested that two of the DHF cases were secondary infections and two were primary infections. The volunteers with primary infections were 10 and 8 years old, whereas those with the secondary infections were 20 and 21 years old.
Seventeen (2.2%) volunteers were diagnosed as PED cases during the 14-day monitoring period (Table 2), and most (15 of 17, 88%) were among nearest neighbors (Table 1). Of the 17 PED cases, 10 were positive by the RT-PCR (1 DEN-1, 7 DEN-2, 1 DEN-3, and 1 DEN-4), and the serotype was confirmed by virus isolation in 5 cases. The other seven PED cases became positive for IgM antibodies to dengue virus during monitoring, and dengue infections were also evident by serocoversion ( 4-fold increase in antibody titer) by HI or PRNT. Four of the 17 PED cases were primary infections based on HI and PRNT50 results, and all others were secondary infections.
|
). The one PED DHF case was hospitalized five days after enrollment with a fever of 38.8°C and a positive tourniquet test result, along with a history of epistaxis prior to admission that continued for two days while hospitalized. Small pleural effusions and ascites developed on day three, but resolved by day five. The platelet count was normal on admission but decreased to 40,000/mm3 by hospital day six. From admission, total protein and albumin levels decreased from 8.1 g/dL to 5.2 g/dL and from 3.5 g/dL to 2.9 g/dL, respectively (normal levels: protein = 6.4–8.7 g/dL and albumin = 3.5–5.2 g/dL) on day six. The patient remained hemodynamically stable throughout hospitalization and was discharged on day 10.
Asymptomatic dengue virus infections occurred in eight of the PED cases (Table 2). Dengue viremia was detected in two of the eight cases, with DEN-1 identified in one case by RT-PCR and DEN-2 identified in the second case by both RT-PCR and virus isolation. In the case with DEN-1, virus was detected in the blood sample obtained on day 10. For the case with DEN-2, virus was detected in the blood sample obtained on day 4. To our knowledge, these cases, together with the two asymptomatic ED viremic cases mentioned earlier, represent the first documentation of asymptomatic dengue viremia in human volunteers from naturally acquired infections.
Table 3 shows the serotype-specific relationship of dengue viruses identified from index and cluster cases. There were seven instances in which the virus was identified from both the cluster case and the corresponding index case (four ED cases and three PED cases). The infecting serotype of the cluster case corresponded with the infecting serotype of the index case in four instances (all DEN-2) and in three instances the infecting serotypes were different. We did not detect a specific pattern of the infecting serotype between the ED and PED cases.
|
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Among 785 volunteers in this investigation, we observed 17 PED (new) dengue infections that were both symptomatic and asymptomatic, including one case of DHF. The calculated incidence rate of dengue infection was 567 cases per 1,000 person-years of follow-up. Based on the one observed DHF case, the calculated DHF incidence rate was 33 DHF cases per 1,000 person-years of follow-up, resulting in a DHF to DF incidence rate ratio of 1:18. This ratio closely approximates other epidemiologic observations, showing that DHF occurs in roughly 5% of dengue virus infections in areas where all four serotypes of dengue virus are endemic.13–15
We detected 175 individuals who had either a recent dengue virus infection or who were acutely infected at the time of enrollment and diagnosed with DHF. Eleven of the 175 volunteers were found to be viremic. Combining these 11 cases with the PED cases, a total of 28 acutely infected dengue cases were observed among the volunteers (Figure 1), resulting in a calculated incidence of 933 acute dengue cases per 1,000 person-years of follow-up. Of the 28 acute dengue cases, five were classified as DHF resulting in an overall DHF incidence of 166 cases per 1,000 person-years.
Virologic analysis of dengue cases confirmed that all four serotypes circulated in West Jakarta. The DEN-1 serotype was the predominant virus identified among the index cases, but DEN-2 caused most of the cluster cases, including all five DHF cases. We plan to genetically characterize these viruses and compare them to other circulating Asian strains. Using banked sera, we will also examine the kinetics of infection in both symptomatic and asymptomatic cases by a quantitative RT-PCR. Although an association between the level of dengue viremia and severity of symptomatic disease was previously established, these analyses may provide some insight into whether there is a correlation between the level of viremia and symptomatic or asymptomatic dengue virus infection.
An attractive feature of the cluster investigation method is that it does not rely on outbreak events to accumulate a sizeable number of cases. When longitudinally following a randomly selected cohort, the study population may or may not experience outbreak-related dengue virus transmission, so that to arrive at a sufficient number of DHF and DF cases for statistical analysis, one may have to follow a particular cohort for many months to years. With the cluster investigation method that selects index cases as a reference point, only days to weeks of follow-up are required to define past and present virus activity in an area. A shorter follow-up period would significantly reduce cost and personnel needs.
Cohort studies to date have been unsuccessful in documenting asymptomatic viremia in humans from naturally acquired dengue infection.16 Given the natural history of dengue virus infection, such silent viremia would seem plausible, but with the design of typical longitudinal cohort studies, infected volunteers are only identified upon the development of symptoms. Asymptomatic infections are only detected by seroconversion of blood samples collected at pre-determined time points. Using a cluster design approach, we were able to document eight asymptomatic dengue infections, two of which demonstrated viremia. To our knowledge, this is the first documentation of asymptomatic viremia in naturally acquired human dengue infections.
The one obvious disadvantage of the cluster method is that volunteers are required to participate in an intensive period of surveillance where, despite being healthy at the time, they are asked to donate blood samples every 2–3 days for a period of two weeks. We had little difficulty recruiting the requisite 10–15 people per cluster and often had to refuse additional enrollment of willing volunteers. With augmented resources and experience, we plan to expand our study design to new areas (East Jakarta and another major city in West Java, Indonesia) and to increase the number of index cases (6–8 per week) and persons recruited per cluster (up to 20). The major challenge to our cluster study design is maintaining adequate support (finances and trained personnel) for the intensive community monitoring. We believe that the yield of the cluster investigation method is equal to or substantially greater than that of other prospective study methods used to examine the epidemiology of naturally acquired dengue virus infection. As this method is applied to other countries with varying cultural sensitivities and expectations, differing levels of success may be encountered.
Received September 20, 2004. Accepted for publication December 23, 2004.
Acknowledgments: We thank the nurses and physicians in the Internal Medicine Department at Sumber Waras Hospital (Soesilowati Soerachmad, Hansa Wulur, and Fajar Kurniawan Frans J. V. Pangalila); and the Viral Diseases laboratory staff (Sri Hartati and Dasep Purwaganda) for conducting the serologic testing; Yurike S. Tobing and Akterono Dwi for conducting tissue culture and virus isolation; Gustiani and Ungke Anton Jaya for molecular testing; Agus Rachmat for assistance with geographic information system mapping; and Andy Whitehurst for editorial assistance with the manuscript.
Financial support: This work was supported by the U.S. Naval Medical Research Center (Silver Spring, MD) work unit 63002A810SD0012 and the Military Infectious Diseases Research Program.
Disclaimer: The opinions and assertions herein are not considered as official or views of the U.S. Navy or the Naval Service at large.
Disclosure: None of the authors have a financial or personal conflict of interest related to this study. The corresponding author had full access to all data in the study and final responsibility for the decision to submit this publication.
Authors’ addresses: Charmagne G. Beckett, Curtis G. Hayes, and Kevin R. Porter, Naval Medical Research Center, Silver Spring, MD 20910. E-mails: beckettc{at}nmrc.navy.mil and porterk{at}nmrc.navy.mil. Herman Kosasih, Indra Faisal, Nurhayati, Ratna Tan, Susana Wid-jaja, Erlin Listiyaningsih, Chairin Ma’roef, and Suharyono Wuryadi, Viral Diseases Program, American Embassy Jakarta Unit 8132, U.S. Naval Medical Research Unit No. 2, FPO, AP 96520-8132, Telephone: 62-21-421-4457, Fax: 62-21-424-4507, E-mail: viro{at}namru2.org. Michael J. Bangs, Navy Disease Vector Ecology and Control Center, Silverdale, WA 98315. Tatang K. Samsi, Pediatrics Department, Sumber Waras Hospital, Jakarta, Indonesia. Djoko Yuwono, National Institute of Health Research and Development, Jakarta, Indonesia.
Reprint requests: Robiyati, Kompleks, Publications Department, U.S. Naval Medical Research Unit No. 2, P2M-PLP/LITBANGKES, Jl. Percetakan Negara No. 29, Jakarta Pusat, Indonesia.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C.-C. Kan, P.-F. Lee, T.-H. Wen, D.-Y. Chao, M.-H. Wu, N. H. Lin, S. Y.-J. Huang, C.-S. Shang, I-C. Fan, P.-Y. Shu, et al. Two Clustering Diffusion Patterns Identified from the 2001-2003 Dengue Epidemic, Kaohsiung, Taiwan Am J Trop Med Hyg, September 1, 2008; 79(3): 344 - 352. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
