Abstract

A prospective study of clinical and cytokine profiles of 107 infants with dengue hemorrhagic fever (DHF)/dengue shock syndrome (DSS) was conducted. Fever, petechiae on the skin, and hepatomegaly were the most common clinical findings associated with DHF/DSS in infants. DSS occurred in 20.5% of the patients. Hemoconcentration and thrombocytopenia were observed in 91.5% and 92.5% of the patients, respectively. Serologic testing revealed that almost all of the patients (95.3%) had primary dengue virus infections. These data demonstrate that clinical and laboratory findings of DHF/DSS in infants are compatible with the World Health Organization’s clinical diagnostic criteria for pediatric DHF. The present study is the first to report evidence of production of cytokines in infants with DHF/DSS and to describe the difference between the cytokine profile of infants with primary dengue virus infections and children with secondary infections. Overproduction of both proinflammatory cytokines (interferon-γ and tumor necrosis factor–α) and anti-inflammatory cytokines (interleukin-10 and -6) may play a role in the pathogenesis of DHF/DSS in infants

Dengue viruses are members of the family Flaviviridae and are of 4 serotypes (DEN-1, DEN-2, DEN-3, and DEN-4). Dengue virus infections may be asymptomatic or may lead to undifferentiated fever; dengue fever (DF) or dengue hemorrhagic fever (DHF) with capillary leakage that may progress to hypovolemic shock; or dengue shock syndrome (DSS), which may be fatal if it is not treated correctly [1, 2]. Dengue virus infections are serious public health problems in many regions of the world. More than 2.5 billion people in >100 countries worldwide are now at risk for dengue virus infection. An estimated 50 million infections occur annually, including 500,000 cases of DHF/DSS, and at least 2.5% of individuals with DHF/DSS die. The vast majority of cases, nearly 95%, occur in children <15 years of age [3], and ⩾5% of all DHF/DSS cases occur in infants. However, since DHF/DSS was reported to occur in infants during their first dengue infections by Halstead et al. [4] in 1969, only a few research studies, with rather small numbers of patients, have been published on this subject [5–9]. Technical difficulties in obtaining the rather large blood samples required by research protocols from small subjects and possible constraints imposed by human experimentation protocols may partially explain why DHF/DSS in infants has not been more comprehensively studied, as has DHF/DSS in older children [10]

A frequently cited model of the immunopathogenesis of DHF/DSS focuses on secondary dengue virus infections in children >1 year of age [11–13]. In this model, aberrant or memory immune activation directed against dengue virus–infected cells initiates cytokine release, increased vascular permeability, and activation of the blood clotting system, the hallmarks of DHF/DSS [13]. Comparable immunologic studies of the features of severe dengue in infants during their first dengue virus infection have not been published. The research challenge presented by infant DHF/DSS has been described recently by Halstead et al. [10]. Studies involving infants with DHF/DSS accompanying primary dengue virus infections should provide powerful new evidence with regard to the immune mechanisms underlying all cases of DHF [14]. In response to the challenge of studying this unique pathogenetic group, we conducted a prospective study of clinical and cytokine profiles in infants. In the present study, we demonstrate that clinical and laboratory findings associated with DHF/DSS in infants are compatible with the clinical diagnostic criteria for DHF in children of the World Health Organization (WHO). We also report evidence of cytokine overproduction in infants with DHF/DSS

Patients, Materials, and Methods

Patients

One hundred eighteen infants <12 months of age who were admitted to the Department of Dengue Hemorrhagic Fever at Children’s Hospital No. 1–Ho Chi Minh City (Ho Chi Minh City, Vietnam) between January 1998 and March 2002 with a clinical diagnosis of suspected DHF according to the criteria of the WHO [1] were enrolled in the study, provided that a parent or guardian gave informed consent. According to WHO criteria, which are based on information derived from children, patients were classified as having nonshock DHF (grades I and II) or DSS (grades III and IV) [1]

Patients received routine care from ⩾1 of the authors. For each patient, basic demographic data, medical history, physical examination findings, and subsequent progress were recorded on a standard data form. Informed consent was obtained from all parents or guardians. The study was approved by the Scientific and Ethical Committee of Children’s Hospital No. 1–Ho Chi Minh City

Sample Collection

Paired blood samples were obtained from each patient in the study, 1 during the acute phase and 1 during the convalescent phase. Acute-phase blood samples (2–3 mL) were obtained at admission to the hospital (day 3 to day 7 after the onset of fever). Convalescent-phase blood samples (2–3 mL) were obtained during the convalescent phase of the disease (day 8 to day 19 after the onset of fever). A blood sample from each of 99 mothers of these patients was obtained at the time of the infant’s hospital admission. Serum was separated as quickly as possible, and serum samples were stored at −70°C until they were used

Laboratory Methods

Serologic testingDengue virus infections in infants were confirmed by viral envelope and membrane (E/M)–specific capture IgM and IgG ELISAs, using dengue virus and Japanese encephalitis virus (JEV) as antigens, at the Center for Disease Control, Department of Health, Taipei, Taiwan. Dual analyses of E/M-specific capture IgM and IgG ELISAs and NS1 serotype–specific IgG ELISAs were used to differentiate between primary and secondary dengue virus infections, as described elsewhere [15]. Serum samples from the mothers were tested to determine their immune status to dengue virus by use of NS1 serotype–specific IgG ELISAs

E/M-specific capture IgM and IgG ELISAs in which diluted pooled viral antigens from culture supernatants of DEN-1–, DEN-2–, DEN-3–, DEN-4–, and JEV-infected Vero cells were used as antigens were performed to measure levels of IgM and IgG antibodies in paired serum samples from the infants. The enzyme activity was finally developed, and the optical density (OD) was determined 30 min later at the dual wavelengths of 405 and 630 nm with a Dynatech MR700 microplate reader [15]

NS1 serotype–specific IgG ELISAs in which diluted NS1-containing culture supernatants of DEN-1–, DEN-2–, DEN-3–, DEN-4–, and JEV-infected Vero cells were used as antigens were performed to measure levels of NS1-specific IgG antibody in serum samples from infants and mothers. The enzyme activity was developed, and ODs were determined [15]

The ODs of culture supernatants of Vero cells with and without dengue virus infection were designated the test absorbance and negative control value, respectively, for each sample in the ELISA. Positivity was determined by comparison to individual negative controls. A positive sample was a sample with a ratio of test absorbance to negative control of ⩾2.0, and a negative sample was a sample with a ratio of <2.0. For serum samples with positive results of capture IgM and IgG ELISAs, a ratio of IgM to IgG of ⩾1.2 was used to identify a primary dengue virus infection, and a ratio of <1.2 was used to identify a secondary dengue virus infection. For serum samples with positive NS1-specific IgG antibody responses, NS1 serotype was determined by the ratio of the highest OD value to the second highest OD value read from the 4 dengue virus serotypes. Positive serotype specificity was defined by an OD ratio ⩾1.2, and negative serotype specificity was defined by an OD ratio <1.2. NS1 serotype–specific IgG ELISA was used to identify primary dengue virus infection; primary infection was defined by (1) a negative NS1-specific IgG antibody response for serum samples obtained between days 1 and 14 of illness or (2) a positive serotype specificity for serum samples obtained after ⩾9 days of illness. Secondary dengue virus infection was defined by (1) a positive NS1-specific IgG antibody response for serum samples obtained between days 1 and 8 of illness or (2) a positive NS1-specific IgG antibody response and negative serotype specificity at any time after the onset of symptoms [15]

Cytokine assaysThe plasma levels of 6 cytokines (interferon [IFN]–γ, tumor necrosis factor [TNF]–α, interleukin [IL]–10, IL-6, IL-4, and IL-2) of the patients were measured simultaneously by the BD Human Th1/Th2 Cytokine Cytometric Bead Array Kit-II (BD Biosciences Pharmigen) in a 50-μL sample, according to the manufacturer’s instructions. The detection limits of the kit for IFN-γ, TNF-α, IL-10, IL-6, IL-4, and IL-2 were 7.1, 2.8, 2.8, ⩽3.0, 2.6, and 2.6 pg/mL, respectively

Protein C and S assaysPlasma levels of proteins C and S were measured by use of commercial ELISA kits (Helena Laboratories). The laboratory methods used and the interpretation of results for all kits followed the manufacturer’s instructions. Determinations of cytokine and protein C and S levels in the patients’ samples were performed at the Departments of Microbiology and Immunology and of Medical Technology, Medical College, National Cheng Kung University, Tainan, Taiwan

Complete blood counts, coagulation tests, liver and renal function tests, and ionogramsComplete blood counts (Careside H-2000 counter), coagulation tests (Diagnostica Stago), liver and renal function tests, and ionograms (R5A 800A kit; Hycel Groupe Lisabio) were performed at the hospital laboratory

Data and Statistical Analysis

The statistical significance of differences in normally distributed data was compared using analysis of variance; the Kruskal-Wallis test for 2 groups was used when the variances in the samples differed. Statistical analyses were performed with EpiInfo 2000, version 1.1 (Centers for Disease Control and Prevention). P<.05 was considered to be statistically significant

Results

Clinical findingsOn the basis of E/M-specific capture IgM and IgG ELISAs using dengue virus and JEV as antigens, dengue virus infection was confirmed to be the cause of illness for 107 of 118 infants hospitalized with DHF. Dengue virus–specific IgM and/or IgG antibodies with limited cross-reactivity to JEV were found in serum samples from these 107 infants (data not shown). The limited cross-reactivity was most likely the result of the naive status of immune memory to other flavivirus of these infants. Eighty-five of these 107 infants were categorized as having nonshock DHF (grade I, 1 infant; grade II, 84 infants), and 22 were categorized as having DSS (grade III, 17 infants; grade IV, 5 infants). The age distribution is shown in figure 1. Some cases were observed in infants <3 months of age, but the largest numbers of cases were in infants who were 6–9 months of age (mean, 6.7 months; range, 1–11 months). The mean age of infants with DSS was significantly higher than that of infants with nonshock DHF (8.2 vs. 6.4 months; P=.002). Twenty (90%) of 22 DSS patients were infants ⩾6 months. The male-to-female ratio was 57:50 (1.1:1)

Figure 1

Age distribution, by month, of 107 infants with dengue hemorrhagic fever (DHF)/dengue shock syndrome (DSS). Nos. of infants are shown above columns. Few cases were observed in infants <3 months of age; most cases occurred in infants between the ages of 6 and 9 months. White columns All infants with DHF or DSS; black columns infants with DSS

Figure 1

Age distribution, by month, of 107 infants with dengue hemorrhagic fever (DHF)/dengue shock syndrome (DSS). Nos. of infants are shown above columns. Few cases were observed in infants <3 months of age; most cases occurred in infants between the ages of 6 and 9 months. White columns All infants with DHF or DSS; black columns infants with DSS

Table 1 shows the clinical characteristics, laboratory findings, and treatment for the patients. All patients had high, continuous fever that lasted 2–13 days (mean, 5.2 days). Almost all patients had petechiae on the skin (106 [99.0%] of 107). Eight patients (7.4%) had gastrointestinal (GI) bleeding (hematemesis or melena). Surprisingly, epistaxis and gum bleeding were not recorded for any patient. Hepatomegaly was detected in 104 patients (97.1%). Liver size ranged from 1 to 6 cm below the right costal margin. Interestingly, infants with DSS had a much greater degree of liver enlargement than did those with nonshock DHF (mean, 3.8 vs. 2.8 cm; P=.000). Splenomegaly was detected in 7 patients (6.5%). DSS was observed in 22 patients (20.5%). Shock occurred on day 3 to day 6 after the onset of fever (mean, 4.6 days). Thirteen (59.0%) of these 22 patients still had fever at the onset of shock; fever lasted from 1 to 4 days after onset of shock (mean, 1.8 days). Ten (9.3%) of the 107 patients had neurologic signs (dengue encephalopathy) manifested by convulsion (7 patients), lethargy (4 patients), and coma (4 patients). In addition, nonspecific signs and symptoms (cough, runny nose, and diarrhea) were noted in 46 patients (42.9%), 29 patients (27.1%), and 20 patients (18.6%), respectively. Coinfections seen in infants with DHF/DSS in the present study were pneumonia (4 patients), bronchitis/bronchiolitis (4 patients), and shigellosis (2 patients)

Table 1

Clinical characteristics, laboratory findings, and treatments for 85 infants with nonshock dengue hemorrhagic fever (DHF) and 22 infants with dengue shock syndrome (DSS)

Table 1

Clinical characteristics, laboratory findings, and treatments for 85 infants with nonshock dengue hemorrhagic fever (DHF) and 22 infants with dengue shock syndrome (DSS)

Laboratory findingsHemoconcentration, manifested by an increase in hematocrit (Hct) of ⩾20% above the convalescent-phase value, was observed in 98 patients (91.5%). In the other 9 patients (8.4%), Hct increased by ⩾10%–19%. The peak Hct, as well as increase in Hct, for patients with DSS was significantly higher than that for patients with nonshock DHF (mean, 42.6% vs. 39.1% [P=.000] and 41.9% vs. 32.5% [P=.003], respectively). Thrombocytopenia was found in 99 patients (92.5%). The lowest platelet count for infants with DSS was significantly lower than that for infants with nonshock DHF (P=.02; table 1). The lowest platelet count was significantly lower for infants with GI bleeding than for infants without (mean, 39.6×103 vs. 64.1×103 cells/mm3; P=.02; table 2)

Table 2

Comparison between infants with nonshock dengue hemorrhagic fever/dengue shock syndrome who had gastrointestinal (GI) bleeding and those who did not

Table 2

Comparison between infants with nonshock dengue hemorrhagic fever/dengue shock syndrome who had gastrointestinal (GI) bleeding and those who did not

Liver function tests and ionograms were performed for only 28 and 42 patients, respectively, because of the difficulty of obtaining enough blood for many tests from infants. Elevated serum levels of aspartate aminotransaminase (AST) and alanine aminotransaminase (ALT) were seen in 26 patients (92.8%) and 23 patients (82.1%), respectively. There was a significant difference in serum levels of AST and ALT in infants with and without GI bleeding (mean, 3739.8 vs. 356 U/L [P=.03] and 1119.4 vs. 139.1 U/L [P=.002], respectively) (table 2). However, serum levels of AST and ALT did not differ significantly between infants with nonshock DHF and those with DSS (P = .1 and P=.3, respectively). Hyponatremia was detected in 23 patients (54.7%), and hypocalcemia and hypokalemia were detected in 6 patients (14.2%) and 2 patients (4.7%), respectively

Dual analyses of E/M-specific capture IgM and IgG ELISA and NS1 serotype–specific IgG ELISA results showed that 102 patients (95.3%) had primary dengue virus infections and only 5 patients (4.6%) had secondary dengue virus infections. Ninety-eight (98.9%) of 99 mothers whose serum samples were available were found to be seropositive for dengue virus by means of NS1 serotype–specific IgG ELISA. Based on NS1 serotype-specific IgG ELISA results, most of those 98 mothers (86 mothers [87.7%]) had experienced ⩾2 dengue virus infections during their lifetime, whereas only 12 mothers (12.2%) had previously experienced a single dengue virus infection. Representative results of E/M-specific capture IgM and IgG ELISAs and/or NS1 serotype–specific IgG ELISAs for paired infants and mothers are presented in table 3. Patients 1–4 had primary dengue virus infections (all caused by DEN-3); patient 5 had secondary dengue virus infection. Positive NS1-specific IgG antibody responses and negative serotype specificity (NS1 serotyping OD ratio <1.2) in serum samples from the mothers of patients 1–4 (M6, M26, M52, and M65, respectively) showed that these mothers had experienced ⩾2 previous dengue virus infections, and the other mother (M58) had experienced a single dengue virus infection (caused by DEN-2), demonstrated by positive NS1-specific IgG antibody responses and an NS1 serotyping OD ratio ⩾1.2 (table 3)

Table 3

Representative results of capture IgM and IgG ELISAs and NS1 serotype–specific IgG ELISAs for samples from paired infants with nonshock dengue hemorrhagic fever/dengue shock syndrome and samples from mothers

Table 3

Representative results of capture IgM and IgG ELISAs and NS1 serotype–specific IgG ELISAs for samples from paired infants with nonshock dengue hemorrhagic fever/dengue shock syndrome and samples from mothers

Chest radiography was performed for 31 patients. Right-side pleural effusion was demonstrated in 8 patients, bilateral pleural effusion in 8 patients, bronchitis/bronchiolitis in 4 patients, and pneumonia in 4 patients

Treatment and resultOne hundred patients (93.4%; 78 patients with nonshock DHF and 22 patients with DSS) received intravenous fluid replacement with Ringer’s lactate (RL) or 5% dextrose in RL solution (5% dextrose in RL solution is used to treat patients with nonshock DHF, not patients with DSS). The mean amount of intravenous fluid was 104.3 mL/kg (range, 27.5–211.7 mL/kg), administered over a mean period of 24.9 h (range, 6–53 h). In the present study, there was no significant difference in the mean amount or in the duration of intravenous fluid replacement between patients with nonshock DHF and those with DSS (P=.1 and P=.9, respectively). Administration of colloid solution (dextran 40) was indicated in 20 patients (18.6%). The mean amount administered was 53.6 mL/kg (range, 28–100 mL/kg); this was correlated with severity of the disease (P=.01; table 1). Eight patients (7.4%), 5 of whom had severe GI bleeding, received a transfusion of fresh whole blood (FWB; mean amount, 24.7 mL/kg; range, 13–71 mL/kg)

With proper case management and good nursing care, almost all of the patients completely recovered, but 4 patients died; thus, the case-fatality rate was 3.7% among infants with DHF/DSS in this study. Of the patients who died, 2 had prolonged shock, 3 had encephalopathy, 2 had respiratory failure, and 3 had massive GI bleeding. Clinical characteristics and laboratory findings for these 4 patients are shown in table 4

Table 4

Clinical and laboratory data for 4 infants with fatal dengue shock syndrome

Table 4

Clinical and laboratory data for 4 infants with fatal dengue shock syndrome

Cytokine profile in infants with DHF/DSSThe levels of IFN-γ in the acute-phase samples from infants with DHF/DSS were significantly higher than those in samples from healthy control infants (mean, 56.2 vs. 4.1 pg/mL; P=.01; table 5). When the data are pooled and the kinetic changes in cytokine blood levels are plotted by day after fever onset, levels of IFN-γ in infants with DHF/DSS are found to have increased on day 4 to day 6 after the onset of fever and to have rapidly decreased on day 7 and during the convalescent phase. The concentration of IFN-γ in convalescent-phase samples from the patients was somewhat higher than that for control subjects, but was not significantly different (mean, 18.5 vs. 4.1 pg/mL; P=.3; figure 2). Similar results were observed when the TNF-α levels in the acute- and convalescent-phase samples were compared with those in samples from control subjects. Infants with DHF/DSS had significantly higher TNF-α levels in acute-phase samples than did control subjects (mean, 9 vs. 0.8 pg/mL; P=.01; table 5). TNF-α levels were elevated on day 4 to day 7 after the onset of fever and decreased on day 8 to day 19; thus, the concentration of TNF-α in convalescent-phase samples from infants with DHF/DSS was not significantly higher than that in samples from control subjects (mean, 4.4 vs. 0.8 pg/mL; P=.1). Significantly elevated IL-10 and IL-6 levels were detected in the acute-phase samples from infants with DHF/DSS, compared with samples from control subjects (mean, 73.8 vs. 0.3 pg/mL [P=.000] and 28.2 vs. 1.4 pg/mL [P=.000], respectively). In contrast, IL-10 and IL-6 levels in convalescent-phase samples from infants with DHF/DSS decreased, compared with levels in acute-phase samples, but were still significantly higher than in samples from control subjects (mean, 8.3 vs. 0.3 pg/mL [P=.002] and 22.7 vs. 1.4 pg/mL [P=.009], respectively). However, further analysis showed that the mean serum levels of IFN-γ, TNF-α, IL-10, and IL-6 did not differ significantly between patients with nonshock DHF and those with DSS (table 5) or between patients with primary and secondary infections (data not shown). The mean serum levels of IFN-γ, TNF-α, and IL-10 in patients with fatal outcomes were not significantly higher than those in survivors, but a significantly higher elevation of IL-6 was observed in patients who died than in patients who survived the infection (mean, 131.2 vs. 23.0 pg/mL; P=.007). On the other hand, compared with levels in samples from control subjects, levels of IL-2 and IL-4 in acute- and convalescent-phase samples in infants with DHF/DSS were not elevated

Table 5

Cytokine profile in acute-phase serum samples from 43 infants with nonshock dengue hemorrhagic fever (DHF), 19 infants with dengue shock syndrome (DSS), and 6 control infants

Table 5

Cytokine profile in acute-phase serum samples from 43 infants with nonshock dengue hemorrhagic fever (DHF), 19 infants with dengue shock syndrome (DSS), and 6 control infants

Figure 2

Kinetic change of the levels of cytokines, by day after fever onset, in infants with dengue hemorrhagic fever (DHF)/dengue shock syndrome (DSS). A Levels of interferon (IFN)–γ in infants with DHF/DSS increased on day 4 to day 6 after fever onset and rapidly decreased on day 7 and during the convalescent phase. B Similarly, serum levels of tumor necrosis factor (TNF)–α increased on day 4 to day 7 after fever onset in infants with DHF/DSS and then decreased on day 8 to day 19. Interleukin (IL)–10 (C) and IL-6 (D) levels in acute-phase samples were elevated; levels of these cytokines in convalescent-phase samples decreased but were still significantly higher than those in controls. There was no significant elevation in the levels of IL-2 (E) or IL-4 (F) in acute-phase and convalescent-phase samples, compared with controls

Figure 2

Kinetic change of the levels of cytokines, by day after fever onset, in infants with dengue hemorrhagic fever (DHF)/dengue shock syndrome (DSS). A Levels of interferon (IFN)–γ in infants with DHF/DSS increased on day 4 to day 6 after fever onset and rapidly decreased on day 7 and during the convalescent phase. B Similarly, serum levels of tumor necrosis factor (TNF)–α increased on day 4 to day 7 after fever onset in infants with DHF/DSS and then decreased on day 8 to day 19. Interleukin (IL)–10 (C) and IL-6 (D) levels in acute-phase samples were elevated; levels of these cytokines in convalescent-phase samples decreased but were still significantly higher than those in controls. There was no significant elevation in the levels of IL-2 (E) or IL-4 (F) in acute-phase and convalescent-phase samples, compared with controls

Coagulation tests and plasma protein C and S levels in infants with DHF/DSSCoagulation tests were done for 13 patients. Prolonged prothrombin time (PT) and activated partial thromboplastin time (APTT), decreased fibrinogen levels, and positive results of testing for d-dimer were seen in 5 patients (38.4%), 11 patients (84.6%), 10 patients (76.9%), and 2 patients (15.3%), respectively. PT and APTT were significantly prolonged and fibrinogen levels were significantly lower in patients with DSS, compared with patients with nonshock DHF (P=.008, P=.01, and P=.002, respectively; table 1)

Plasma levels of anticoagulant proteins C and S in the acute-phase samples from infants with DHF/DSS were significantly lower than levels in samples from healthy control subjects (mean, 54.5% vs. 105% [P=.000] and 55.4% vs. 110.9% [P = .000], respectively). The levels of proteins C and S in the convalescent-phase samples from infants with DHF/DSS were significantly lower than levels in samples from control subjects (mean, 70.1% vs. 105% [P=.000] and 82.5% vs. 110.9% [P=.000], respectively). However, decreases in levels of proteins C and S were not correlated with the severity of DHF (nonshock DHF vs. DSS) or with GI bleeding (tables 1 and 2). Some correlation between plasma levels of protein C (not of protein S) and serum levels of AST and ALT (r=0.34 [P=.02] and r=0.37 [P = .005], respectively) was indicated by linear regression analysis. Linear regression analysis also showed significant negative correlations between protein C and S levels and PT (r = −0.7 [P=.004] and r=-0.8 [P=.02], respectively). However, no association between levels of proteins C and S, APTT, and fibrinogen levels was noted. Similarly, there was no correlation between levels of proteins C and S and an increase in Hct, which is a sign of capillary leakage (P=.2 and P=.3, respectively)

Correlation of plasma concentrations of cytokines with serum levels of transaminases and activation of coagulation In this study, serum IL-6 levels strongly correlated with PT (r=0.87; P=.001) but were not correlated with APTT or fibrinogen levels (r=0.53 [P=.1] and r=-0.46 [P=.1], respectively) or with serum levels of transaminases. Further analysis shows that serum IL-6 levels had some positive association with the duration of fever in infants with DHF/DSS (r=0.31; P=.01). In contrast, serum IL-10 levels strongly correlated with serum levels of AST and ALT (r=0.91 [P=.000] and r=0.91 [P=.000], respectively) but were not related to PT, APTT, and fibrinogen levels. Moreover, serum TNF-α levels were significantly associated with an increase in Hct (r=0.34; P=.005; table 6)

Table 6

Results of linear regression analysis showing significant associations of cytokines with plasma levels of transaminases and results of coagulation tests

Table 6

Results of linear regression analysis showing significant associations of cytokines with plasma levels of transaminases and results of coagulation tests

Discussion

High fever (100% of patients), petechiae on the skin (99% of patients), and hepatomegaly (97.1% of patients) were the most common clinical findings associated with DHF/DSS in infants in the present study. DSS (hypovolemic shock secondary to capillary leakage) occurred in 22 (20.5%) of our 107 patients. In contrast to children with DHF/DSS, in whom shock nearly always accompanies or follows defervescence [16, 17], more than one-half of the infants in the present study still had fever at the time that shock occurred. As a result, physicians needed to differentiate between DSS and septic shock. GI bleeding was recorded in 8 infants (7.4%), 5 of whom received an FWB transfusion. The finding that patients with GI bleeding had significantly higher serum levels of AST and ALT than did those without GI bleeding suggests that hepatocellular damage may underlie severe hemorrhaging. This has been observed in older children and adults [18, 19]. Activation of the coagulation system was indicated by prolonged PT (38.4% of patients) and APTT (84.6% of patients), decreased fibrinogen levels (76.9% of patients), and positive results of testing for d-dimer (15.3% of patients). There was no significant difference in PT, APTT, and fibrinogen levels in patients with and patients without GI bleeding in the present study. As evidence that platelet abnormalities underlie severe hemorrhage, patients with GI bleeding had significantly lower platelet counts than did those without GI bleeding. Massive GI bleeding was often associated with a grave prognosis; 3 of 4 infants with fatal outcomes had massive GI bleeding

Hemoconcentration, manifested by an increase in Hct of ⩾20%, as defined by the WHO criteria [1], was observed in 98 patients (91.5%). Hct increased in the remaining 9 patients (8.4%) by ⩾10%–19%; all but 1 of these patients had evidence of capillary leakage, demonstrated by right-side or bilateral pleural effusion on chest radiography. Thrombocytopenia was observed in 99 patients (92.5%). In a study of 31 Thai infants with DHF, Witayathawornwong [8] reported that all patients had thrombocytopenia and pleural effusion on chest radiography; however, an increase in Hct of >20% was recorded in only 20 patients (64.5%), and another 11 patients (35.4%) had an increase in Hct of 11%–20%. The peak Hct, the increase in Hct, and the lowest platelet count in infants with DHF/DSS in the present study correlated with severity of the disease (nonshock DHF vs. DSS; P=.007, P=.000, and P=.02, respectively). Our data in the present study, as mentioned above, demonstrate that clinical characteristics and laboratory findings for infants with DHF/DSS are compatible with the WHO case definition, which is based primarily on findings in older children

Serologic responses, measured by IgM and IgG ELISAs, demonstrated that almost all of our patients (95.3%) suffered from primary dengue virus infections and that only 5 patients (4.7%) had secondary infections. In our previous study, which included 47 infants with DHF/DSS, all patients had a primary infection pattern detected by hemagglutination inhibition test [6]. Ninety-eight (98.8%) of the 99 mothers of the infants in the study had dengue virus antibodies, as shown by NS1 serotype–specific IgG ELISA. These results are consistent with the published observation that DHF/DSS in infants occurs during primary dengue virus infections [4, 5, 8, 9] and with the hypothesis that severe disease is linked to transplacentally transmitted maternal dengue virus antibodies [4, 5]. In a large hospitalized series in Bangkok, Thailand, performed in 1962–1964, the vast majority of infants hospitalized with DHF had primary antibody responses measured in paired serum samples by the hemagglutination inhibition test. A serologic survey of adults in Bangkok showed that almost all women of child-bearing age had dengue virus antibodies [4]. Kliks et al. [5] studied 13 infants from whom DEN-2 viruses were isolated, all of whom had been hospitalized for DHF and had well-documented primary dengue virus infections. All of the mothers of these infants were immune to multiple dengue viruses; none had recently had dengue virus infections. At birth, maternal dengue virus antibody protects infants from dengue virus infection, but later, the risk of development of DHF/DSS increases when IgG antibodies are catabolized to lower titers. Continued catabolism results in the loss of enhancing antibodies and a decrease in risk for DHF/DSS at ∼1 year of age. This dual effect of passively acquired antibodies also explains the age distribution of DHF/DSS among infants in the present study. The largest numbers of cases were in the 6–9-month age group. The age distribution of DHF/DSS in infants in the present study is similar to that in Thailand, Indonesia, and Myanmar [10]

Early diagnosis, proper treatment, and good nursing care are the principles of management of DHF/DSS in infants, as well as in children. Although the incidence of DHF/DSS in infants was low (∼5%), the number of severe cases and the mortality rate were high among infants, compared with among older children. In the present study, almost all patients completely recovered, but 4 patients died, resulting in a rather high case-fatality rate of 3.7%. In comparison, the case-fatality rate was <1% among all patients with DHF/DSS in the same time period [17, 20]

There are many reports that cytokine blood levels in patients with DF differ from those in patients with DHF/DSS. All such data are from studies of secondary dengue virus infections in older children and adults [21–28]. The present study is the first to provide data on the serum levels of both proinflammatory cytokines (IFN-γ and TNF-α) and anti-inflammatory cytokines (IL-10 and IL-6) in infants with DHF/DSS. Unusually high levels of cytokines may play a role in the pathogenesis of DHF in infants. To investigate whether there is any difference in the cytokine profiles of infants with primary dengue virus infections and older children with secondary infections, we measured the concentrations of cytokines in 39 children with DHF/DSS aged 4–15 years who had secondary dengue virus infections, using the same techniques described in the present study. We found significant elevations of serum IFN-γ and IL-10 levels in children with DHF/DSS (mean±SD, 109.5±225.7 and 78.9±57.6 pg/mL, respectively), compared with levels in control subjects (mean±SD, 6.2±6.2 and 0.5±1.1 pg/mL, respectively). Serum levels of TNF-α, IL-6, IL-4, and IL-2 were not elevated in children with DHF/DSS who had secondary dengue virus infections (mean±SD, 2.5±2.1, 8.8±6.7, 1.0±1.2, and 0.3 ± 0.9 pg/mL, respectively), compared with levels in control subjects (mean±SD, 1.2±1.4, 3.3±2.3, 0.3±0.7, and 1.3 ± 1.6 pg/mL, respectively; unpublished data). The levels of IFN-γ and IL-10 in children with secondary dengue virus infections did not differ from those in infants with primary dengue virus infections. Instead, the cytokine profile differed. In contrast to children with DHF/DSS, who produce elevated levels of IFN-γ and IL-10, infants produced elevated levels of TNF-α and IL-6, in addition to IFN-γ and IL-10. It is evident that, in opposition to the view of Rothman and Ennis [12], cross-reactive T cells that are activated in response to secondary infection with a different serotype are not required to produce the high levels of cytokines that accompany severe DHF. A unifying hypothesis is that enhancing antibodies acquired passively from mothers (primary infection in infants) or actively from previous infection (secondary infection in children) promotes dengue virus infection of Fc-bearing cells, resulting in a large, infected cell mass. T cell and cytokine responses are simply proportional to the infected cell mass. Levels of the inflammatory cytokine IFN-γ and the anti-inflammatory cytokine IL-10 are elevated simultaneously in patients with DHF/DSS. Infants experiencing DHF during primary dengue virus infection also have elevated blood levels of TNF-α and IL-6

Linear regression analysis revealed that levels of cytokines such as TNF-α, IL-6, and IL-10 are correlated with some biological responses. An elevated serum level of TNF-α has a significant correlation with the increase in Hct that is a sign of capillary leakage in infants with DHF/DSS. This may be the result of TNF-α toxicity to vascular endothelial cells and may increase vascular permeability. There were significantly higher levels of IL-6 in infants with fatal outcome, compared with levels in those who survived the infection (P=.007). But no significant difference exists between patients with and patients without shock. The serum level of IL-6 in infants with DHF/DSS was also strongly correlated with activation of the extrinsic pathway of coagulation (PT; P=.001) but not with the intrinsic pathway (APTT) or with fibrinogen levels. This is consistent with reports that IL-6 levels are positively associated with the severity of dengue virus infection in both children and adults [29]; of increased serum levels of IL-6 and IL-8 in patients with DHF, but not in patients with DF [30]; that the highest levels of IL-6 are found in patients with DSS [31]; and that IL-6 is significantly associated with the activation markers of coagulation and fibrinolysis [32]. Furthermore, a significant and strong correlation between IL-10 and serum levels of AST and ALT was found in infants with DHF/DSS. Green et al. [33] also reported that elevated levels of IL-10 in children with dengue virus infection were associated with disease severity (DF vs. DHF) and the degree of capillary leakage, quantified by the size of pleural effusions. The elevation of IL-6 and IL-10 levels in infants with DHF/DSS suggests that, when an infant responds to dengue virus infection via generation of inflammatory cytokines (IFN-γ and TNF-α), there is simultaneous generation of inhibitory cytokines (IL-6 and IL-10) to counteract the inflammation. The more stimulation of the pathogen, the higher the levels of stress mediators, such as IL-6 or IL-10, induced in the host to counteract the response. Cytokines can cause cell activation synergistically or antagonistically; the net outcome will depend on the balance between various cytokine actions [13]

The present study emphasizes that DHF/DSS in infants is not uncommon and that clinical characteristics and laboratory findings for infants with DHF/DSS are compatible with the WHO case definition for DHF/DSS in children. This study is the first to describe cytokine blood levels during the acute stage of DHF/DSS in infants. The cytokine profile in infants with primary dengue virus infections and children with secondary infections were studied in the same laboratory, using the same assay methods. Overproduction of both proinflammatory cytokines (IFN-γ and TNF-α) and anti-inflammatory cytokines (IL-10 and IL-6) may play a role in the pathogenesis of DHF/DSS in infants

Acknowledgments

We thank Tran Tan Tram, the Director of the Children’s Hospital No. 1–Ho Chi Minh City (Ho Chi Minh City, Vietnam), and Chung-Ming Chang, National Health Research Institutes (Taiwan), for help and support during the course of this study. Thanks also to Hoang Le Phuc, for help with data analyses, and the doctors and nurses of the Department of Dengue Hemorrhagic Fever, Children’s Hospital No. 1–Ho Chi Minh City, for providing excellent patient care

References

1
World Health Organization (WHO)
Dengue haemorrhagic fever: diagnosis, treatment, prevention and control. 2nd edition
1997
Geneva
WHO
2
Gubler
DJ
Dengue and dengue hemorrhagic fever
Clin Microbiol Rev
 , 
1998
, vol. 
11
 (pg. 
480
-
96
)
3
World Health Organization (WHO)
Dengue and dengue haemorrhagic fever
Fact sheet N.117
 , 
2002
Geneva
WHO
4
Halstead
SB
Scanlon
J
Umpaivit
P
Udomsakdi
S
Dengue and chikungunya virus infection in man in Thailand, 1962–1964. IV. Epidemiologic studies in the Bangkok metropolitan area
Am J Trop Med Hyg
 , 
1969
, vol. 
18
 (pg. 
997
-
1021
)
5
Kliks
SC
Nimmannitya
S
Nisalak
A
Burke
DS
Evidence that maternal antibodies are important in the development of dengue hemorrhagic fever in infants
Am J Trop Med Hyg
 , 
1988
, vol. 
38
 (pg. 
411
-
19
)
6
Lan
NT
Hung
NT
Lien
LB
Dengue hemorrhagic fever in infants [abstract 43]
Program and abstracts of the 4th International Symposium on Dengue Fever
 , 
1997
Geneva
World Health Organization
pg. 
43
 
7
Hidayah
N
Darmowandowo
W
Ismoedijanto
Soegijanto
S
Dengue hemorrhagic fever in infant (clinical aspects, immunological responses related to the clinical degrees) [abstract]
Program and abstracts of the 1st International Conference on Dengue/Dengue Hemorrhagic Fever
 , 
2000
Geneva
World Health Organization
pg. 
11
 
8
Witayathawornwong
P
Dengue hemorrhagic fever in infancy at Petchabun Hospital, Thailand
Southeast Asian J Trop Med Public Health
 , 
2001
, vol. 
32
 (pg. 
481
-
7
)
9
Pancharoen
C
Thisyakorn
U
Dengue virus infection during infancy
Trans R Soc Trop Med Hyg
 , 
2001
, vol. 
95
 (pg. 
307
-
8
)
10
Halstead
SB
Lan
NT
Myint
TT
, et al.  . 
Dengue hemorrhagic fever in infants: research opportunities ignored
Emerg Infect Dis
 , 
2002
, vol. 
8
 (pg. 
1474
-
9
)
11
Kurane
I
Ennis
FA
Gubler
DJ
Kuno
G
Immunopathogenesis of dengue virus infections
Dengue and dengue haemorrhagic fever
 , 
1997
Wallingford, Oxon, UK
CAB International
(pg. 
273
-
90
)
12
Rothman
AL
Ennis
FA
Immunopathogenesis of dengue hemorrhagic fever
Virology
 , 
1999
, vol. 
257
 (pg. 
1
-
6
)
13
Lei
HY
Yeh
TM
Liu
HS
Lin
YS
Chen
SH
Immunopathogenesis of dengue virus infection
J Biomed Sci
 , 
2001
, vol. 
8
 (pg. 
377
-
88
)
14
Halstead
SB
Enhanced dengue disease—epidemiological and experimental evidence
Presented at: Symposium on Accelerating the Development and Introduction of a Dengue Vaccine for Poor Children, Ho Chi Minh City, Vietnam, 5–8 December 2001
 
15
Shu
PY
Chen
LK
Chang
SF
, et al.  . 
Comparison of capture immunoglobulin M (IgM) and IgG enzyme-linked immunosorbent assay (ELISA) and nonstructural protein NS1 serotype–specific IgG ELISA for differentiation of primary and secondary dengue virus infections
Clin Diagn Lab Immunol
 , 
2003
, vol. 
10
 (pg. 
622
-
30
)
16
Lan
NT
Experience in treatment of dengue shock syndrome [PhD thesis]
Ho Chi Minh City, Vietnam, The University of Medicine and Pharmacy–Ho Chi Minh City
 , 
1994
17
Hung
NT
Lan NT Treatment of dengue hemorrhagic fever at Children’s Hospital No
1–Ho Chi Minh City, 1991–2000
Med J Ho Chi Minh City
 , 
2001
, vol. 
3
 (pg. 
149
-
52
)
18
Kuo
CH
Tai
DI
Chang-Chien
CS
Lan
CK
Chiou
SS
Liver biochemical tests and dengue fever
Am J Trop Med Hyg
 , 
1992
, vol. 
47
 (pg. 
265
-
70
)
19
Nguyen
TL
Nguyen
TH
Tieu
NT
The impact of dengue hemorrhagic fever on liver function
Res Virol
 , 
1997
, vol. 
148
 (pg. 
273
-
7
)
20
Lan
NT
Hung
NT
Phuong
BTM
, et al.  . 
Treatment of dengue hemorrhagic fever at Children’s Hospital No. 1–Ho Chi Minh City, 1991–1996
Dengue Bull
 , 
1998
, vol. 
22
 (pg. 
150
-
61
)
21
Yadav
M
Kamath
KR
Iyngkaran
K
Sinniah
M
Dengue hemorrhagic fever and dengue shock syndrome—are they tumor necrosis factor–mediated disorders
FEMS Microbiol Immunol
 , 
1991
, vol. 
4
 (pg. 
45
-
9
)
22
Vitarana
T
de Silva
H
Withana
N
Gunasekera
C
Elevated tumor necrosis factor in dengue fever and dengue hemorrhagic fever
Ceylon Med J
 , 
1991
, vol. 
36
 (pg. 
63
-
5
)
23
Hober
D
Poli
L
Roblin
B
, et al.  . 
Serum levels of tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6) and interleukin-1β (IL-1β) in dengue-infected patients
Am J Trop Med Hyg
 , 
1993
, vol. 
48
 (pg. 
324
-
31
)
24
Hober
D
Lan
NT
Shen
L
, et al.  . 
Tumor necrosis factor alpha levels in plasma and whole-blood culture in dengue infected patients: relationship between virus detection and pre-existing specific antibodies
J Med Virol
 , 
1998
, vol. 
54
 (pg. 
210
-
8
)
25
Kurane
I
Innis
BL
Nimmannitya
S
Nisalak
A
Meager
A
High levels of interferon alpha in the sera of children with dengue virus infection
Am J Trop Med Hyg
 , 
1993
, vol. 
48
 (pg. 
222
-
9
)
26
Bethell
DB
Flobbe
K
Cao
XT
, et al.  . 
Pathophysiologic and prognostic role of cytokines in dengue hemorrhagic fever
J Infect Dis
 , 
1998
, vol. 
177
 (pg. 
778
-
82
)
27
Braga
ELA
Moura
P
Pinto
LMO
, et al.  . 
Detection of circulant tumor necrosis factor-α, soluble tumor necrosis factor p75 and interferon-γ in Brazilian patients with dengue fever and dengue hemorrhagic fever
Mem Inst Oswaldo Cruz
 , 
2001
, vol. 
96
 (pg. 
229
-
32
)
28
Gagnon
SJ
Mori
M
Kurane
I
, et al.  . 
Cytokine gene expression and protein production in peripheral blood mononuclear cells of children with acute dengue virus infections
J Med Virol
 , 
2002
, vol. 
67
 (pg. 
41
-
6
)
29
Kuno
G
Bailey
RE
Cytokine responses to dengue infection among Puerto Rican patients
Mem Inst Oswaldo Cruz
 , 
1994
, vol. 
89
 (pg. 
179
-
82
)
30
Huang
YH
Lei
HY
Liu
HS
Lin
YS
Liu
CC
Dengue virus infects human endothelial cells and induces IL-6 and IL-8 production
Am J Trop Med Hyg
 , 
2000
, vol. 
63
 (pg. 
71
-
5
)
31
Juffrie
M
Meer
GM
Hack
CE
, et al.  . 
Inflammatory mediators in dengue virus infection in children: interleukin-6 and its relation to C-reactive protein and secretory phospholipase A2
Am J Trop Med Hyg
 , 
2001
, vol. 
65
 (pg. 
70
-
5
)
32
Suharti
C
van Gorp
ECM
Setiati
TE
, et al.  . 
The role of cytokine in activation and fibrinolysis in dengue shock syndrome
Thromb Haemost
 , 
2002
, vol. 
87
 (pg. 
42
-
6
)
33
Green
S
Pichyangkul
S
Vaughn
DW
, et al.  . 
Early CD69 expression on peripheral blood lymphocytes from children with dengue hemorrhagic fever
J Infect Dis
 , 
1999
, vol. 
180
 (pg. 
1429
-
35
)
Financial support: partial support from National Health Research Institutes, Taiwan (grant NHRI-CN-CL-8901P)