OUP user menu

Implications of previous subclinical dengue infection but not virus load in dengue hemorrhagic fever

Wen-Ting Yeh, Rong-Fu Chen, Lin Wang, Jien-Wei Liu, Men-Fang Shaio, Kuender D. Yang
DOI: http://dx.doi.org/10.1111/j.1574-695X.2006.00127.x 84-90 First published online: 1 October 2006

Abstract

In a study comparing the virus load and immune reaction between patients with primary and secondary dengue-2 (DEN-2) infections in a hospital-based analysis, we found that 40.7% (55/135) of the 135 patients had secondary DEN-2 infection following a DEN-2 outbreak in southern Taiwan. Most of the secondary infections had subclinical primary dengue infections (78.2%; 43/55). Patients with secondary DEN-2 infections had lower platelet counts, and blood interferon-α and virus load, but significantly higher interleukin-10 (P=0.030) and anti-DEN-1 neutralization titers (P=0.013) than those with primary infection. Patients with secondary DEN-2 infection also had a higher rate of dengue hemorrhagic fever (DHF) (61.7% vs. 36.3%). A previous subclinical dengue infection is involved in the secondary DEN-2 infection associated with altered immune reaction and higher DHF rate, but lower blood virus load.

Keywords
  • secondary dengue-2 (DEN-2) infection
  • dengue hemorrhagic fever (DHF)
  • cytokines
  • virus load

Introduction

Dengue fever (DF) is a mosquito-borne disease caused by single-stranded RNA viruses called dengue viruses, serotypes 1–4 (DEN-1–4). Classical DF is characterized by benign febrile illness with fever, skin rashes, headache and bone pain. A severe complication of DF, dengue hemorrhagic fever (DHF), has increased both in Asia and in Western countries, threatening two-thirds of countries worldwide and showing a higher mortality as a result of massive internal bleeding or irreversible dengue shock syndrome (Pinheiro & Corber, 1997; Castleberry & Mahon, 2003; Guzman, 2005). Some studies have shown that DHF occurs more frequently in sequential outbreaks of DEN-1 followed by DEN-2 infections (Sangkawibha et al. 1984; Kouri et al. 1987; Clarke, 2000; Guzman, 2005), because presence of anti-DEN antibodies from primary DEN infection might enhance secondary DEN infection, a process known as antibody-dependent enhancement (Ferguson et al. 1999; Morens, 2003). The last DEN-1 outbreak in Kaohsiung, Taiwan, occurred between 1988 and 1991 (Harn et al. 1993). An epidemiological study, however, showed that subclinical transmission of the DEN-1 virus might have occurred in the past decade in Kaohsiung (Chen et al. 1996). Recently, a large DEN-2 outbreak with definite DHF cases in southern Taiwan occurred between 2001 and 2002 (Liu et al. 2003). The possibility of immune enhancement of DEN-2 infections has raised considerable concern in the country. Although no case of DHF had been reported in the last DEN-1 outbreak (Harn et al. 1993; Chen et al. 1996), many cases occurred in the recent DEN-2 outbreak (Liu et al. 2003). We therefore conducted a hospital-based study to investigate the implications of subclinical primary DEN-1 infections in the subsequent DEN-2 outbreak associated with altered cytokine profiles and higher DHF development rate in southern Taiwan.

Two mechanisms have been proposed to explain DHF and dengue shock syndrome (DSS) (Kouri et al. 1987; Rosen, 1996; Morens, 2003; Guzman, 2005): viral virulence and immune enhancement. Molecular studies revealed, however, that DEN-2 viruses belonging to two distinct genotypic groups were equally isolated from both DF and DHF patients (Rico-Hesse et al. 1998). This suggests that pathogenesis of DHF may be not related to particular mutant strains of DEN-2 viruses. The other mechanism may be related to antibody-dependent enhancement of dengue infections in which presence of subneutralized anti-DEN antibodies may cause cross-enhancement of secondary infections (Guzman et al. 2000; Guzman, 2005). Guzman (2005) reported that secondary dengue infection was a significant factor in more than 97% of the severe cases of DHF. The immune mechanism contributing to DHF in secondary DEN infection has recently been related to altered cytokine production, especially higher interleukin-10 (IL-10) production (Chaturvedi et al. 1999; Perez et al. 2004; Guzman, 2005). In an in vitro model of heterotypic dengue infection, we previously showed that suppression of the T helper 1 (Th1) reaction was involved in the antibody-dependent enhancement of heterotypic dengue infections (Yang et al. 2001). A recent study of patients with and without DHF further demonstrated that altered Th1/Th2 reaction associated with higher IL-10 expression, but lower interferon-γ (IFN-γ) production was found in DHF patients (Chen et al. 2005). To explore further whether heterotypic secondary DEN-2 infection is involved in pathogenesis of DHF, we defined primary and secondary DEN-2 infections, and compared the clinical severity, virus load and immune reactions between patients with primary and secondary DEN-2 infections associated with the recent DEN-2 outbreak in southern Taiwan.

Materials and methods

Subjects

A total of 182 hospitalized dengue-suspected patients were recruited for this study. The study was approved by the Institute Review Board of the hospital. The first peripheral blood samples were collected upon admission, 3–7 days after onset of symptoms. Some patients allowed blood sampling every other day for a kinetic assessment of virus load. Of the 182 dengue-suspected patients, 135 were confirmed to have DEN-2 infections based on a type-specific real-time reverse transcriptase PCR (RT-PCR) detection of DEN-2 virus as described below. Patients with DHF were defined based on the World Health Organization criteria for DF complicated with reduced platelets (<100000mm−3), petechia or hemorrhagic manifestations, and plasma leakage with hemoconcentration ≥20%, pleural effusion, ascites or hypoalbuminemia (Pinheiro & Corber, 1997; Clarke, 2000; Halstead, 2000).

Assessment of dengue primary and secondary infections

Serological methods to detect dengue antibodies have been the most commonly used diagnostic procedures for differentiation of primary and secondary DEN infections (Makino et al. 1994). We used a capture dengue immunoglobulin G (IgG) immunoassay (ELISA Kit; PANBIO) to differentiate primary DEN-2 from secondary DEN-2 infections in the blood samples collected on days 3–7 days of the disease. We defined a DEN-2 infection by RT-PCR detection of DEN-2 virus, and defined a secondary DEN-2 infection by both detectable dengue IgG and RT-PCR detection of DEN-2 virus in blood within 7 days of the start of the disease. The cut-off value for the presence of DEN IgG antibodies was given by the antibody index (AI): AI=10 × sample absorbance/mean absorbance of triplicate reference values. The cut-off AI value for a positive detection of capture DEN IgG was set at 22, as per the manufacturer's recommendations.

RT-PCR analysis of blood virus load

We subjected viral RNA extracted from the sera of patients with DEN-2 to fluorogenic quantitative RT-PCR detection of total DEN-2 RNA copies as previously described (Chen et al. 2001, 2005). In brief, 140µL of serum was individually added with 560µL of QIAmp viral RNA extraction solution (Qiagen Inc.) to cause inactivation and lysis of all potential infectious agents. The viral RNA was further purified by the QIAmp spin column and finally suspended to 40µL, as per the manufacturer's recommendations. A fluorescence quantitative RT-PCR was carried out in an ABI 7700 quantitative PCR machine (Applied Biosystems) for 40 cycles using TaqMan technology (Chen et al. 2001, 2005). The forward primer, reverse primer and nested fluorescent probe sequence for detecting DEN-2 were 5′-GGCTTAGCGCTCACATCCA-3′, 5′-GCTGGCCACCCTCTCTTCTT-3′ and FAM-5′-CGCCCACCACTATAGCTGCCGGA-3′-TAMRA, respectively (Chen et al. 2001, 2005). The detection limit of DEN-2 virus RNA copies in our study model was 1.4PFUmL−1, as in our previous used method (Chen et al. 2001).

Measurement of blood IFN-α, IL-10, IFN-γ and IL-13 levels

Blood cytokines, IFN-α, IL-10, IFN-γ and IL-13, were measured to reflect early and late Th1/Th2 reactions. IFN-α and IL-10 are early innate inflammatory mediators that can affect Th1 and Th2 reaction in viral infections (Payvandi et al. 1998; Cousens et al. 1999); IFN-γ and IL-13 are lymphokines that can regulate Th1 cell immunity directed against viral infections (Paludan et al. 1997; Durbin et al. 2002). IFN-α, IL-10, IFN-γ and IL-13 levels in sera were measured by an enzyme-linked immunoassay (ELISA) kit purchased from Bender MedSystems Inc. The results were calculated from interpolation in a standard curve constructed from a series of well-known concentrations of manufacturer standards (Yang et al. 1995, 2001; Chen et al. 2001).

Preparation of DEN viruses and determination of DEN-1 and DEN-2 neutralization titers

DEN-1 and DEN-2 viruses obtained from the Institute of Preventive Medicine at the National Defense Medical Center were used for titration of serum anti-DEN-1 and anti-DEN-2 neutralization titers. Virus titers were determined based on a standard PFU assay on BHK-21 cells as described previously (Yang et al. 1995, 2001; Chen et al. 2001), and adjusted to 3 × 107PFUmL−1 in RPMI 1640 (Gibco BRL) with 10% fetal calf serum in a large-scale preparation. The same batch of viruses was collected and stored at −70°C before use.

The DEN-1- and DEN-2-specific neutralizing antibody titers in the patients' sera were determined by a plaque-reduction neutralization test in the BHK-21 cells (Morens et al. 1985; Yang et al. 2001). The viruses preneutralized with each individual serum were incubated with the BHK-21 cells, which had achieved 80% confluence culture on 24-well plates, for determination of viral plaque formation. The viral plaques became visible after coculture in medium with 1.5% carboxymethyl cellulose and 2% fetal calf serum for 6 days. An effective virus neutralization titer is defined as at least 80% reduction of viral plaques.

Data presentation and statistics

Data from subjects with confirmed DEN-2 infection were classified according to those with primary and secondary DEN-2 infections. Results from the measurement of immune mediators are presented as mean±SE. Virus load in blood is presented as total DEN-2 virus RNA copiesmL−1. Differences in the proportion of patients with primary and secondary DEN-2 infections as well as patients with and without DHF were analysed using the χ2 test. Student's t-test was used to analyse differences in immune mediators and virus neutralization titers in patients with primary and secondary DEN-2 infections.

Results

Clinical manifestations in patients with primary and secondary DEN-2 infections

As determined by real-time RT-PCR detection of DEN-2 virus RNA, 135 of 182 hospitalized dengue-suspected patients were shown to have DEN-2 infections. Of the 135 DEN-2 patients, 80 (59.3%) and 55 (40.7%) had primary and secondary DEN-2 infections, based on detectable dengue IgG. Of the 55 patients with secondary DEN-2 infection, only 12 (21.8%) had previously had dengue fever. As shown in Table 1, patients with secondary DEN-2 infections tended to have lower platelet counts (P<0.001). Patients with secondary DEN-2 infections had higher aspartate aminotransferase (ASL) and alanine aminotransferase (ALT) levels, but these levels were not significantly different between patients with primary and secondary infections. There were no differences of gender or age between patients with primary and secondary DEN-2 infections. However, patients with secondary DEN-2 infections had a higher rate of DHF than those with primary infection (61.7% vs. 36.3%, P=0.003).

View this table:
Table 1

Clinical and laboratory manifestations in patients with primary and secondary DEN-2 infections

PrimarySecondaryP
Platelet counts (× 10 000 mm−1)8.1 ± 0.63.7 ± 0.6<0.001
ALT (U mL−1)88.6 ± 22.193.4 ± 19.80.440
AST (U mL−1)122 ± 19.5151.6 ± 26.60.059
Age (years)53.8 ± 2.752.7 ± 2.70.385
Gender (female frequency)46/8034/550.475
Frequency of DHF29/80 (36.3%)34/55 (61.7%)0.003
  • Data are presented as mean ± SE.

  • Significant difference between both groups as determined by Student's t-test.

  • Significant difference between both groups as determined by the chi-squared test.

Cytokine profiles in patients with primary and secondary DEN-2 infections

Certain cytokines, such as IFN-α and IFN-γ, are known to protect against viral infections (Cousens et al. 1999; Sainz & Halford, 2002), whereas others, such as IL-10 and IL-13, are shown to promote certain viral infections (Paludan et al. 1997; Zhao et al. 1998). We measured cytokine levels in the blood samples from patients with primary and secondary DEN-2 infections. The results showed that IFN-α concentrations in blood were higher in patients with primary DEN-2 infections (102.6±32.1 vs. 44.0±6.3pgmL−1, P=0.041). By contrast, patients with secondary DEN-2 infections had significantly higher IL-10 blood concentrations (39.1±10.8 vs. 14.7±4.4pgmL−1, P=0.030). The IFN-γ and IL-13 concentrations in blood were not significantly different between patients with primary and secondary DEN-2 infections (Table 2).

View this table:
Table 2

Plasma cytokines in primary and secondary DEN-2 infections

Cytokine (pg mL−1)PrimarySecondaryP
IFN-α102.6 ± 32.144.0 ± 6.30.041
IL-1014.7 ± 4.439.1 ± 10.80.030
IFN-γ127.7 ± 17.6106.5 ± 12.00.174
IL-1313.9 ± 6.18.2 ± 6.20.263
  • IFN-α, interferon-α IL-10, interleukin-10; IFN-γ, interferon-γ IL-13, interleukin-13.

  • P values determined using Student's t-test.

Virus load in patients with primary and secondary infections

The virus load in blood was detected by a real-time quantitative RT-PCR. Patients with primary DEN-2 infections appeared to have higher virus RNA copies in blood upon admission than those with secondary DEN-2 infections (Fig. 1a). The peak virus load occurred in the early febrile stage and declined to almost undetectable levels 6–7 days after admission (Fig. 1b).

Figure 1

A real-time RT-PCR analysis of DEN-2 virus load in primary and secondary DEN-2 infections. The DEN-2 virus load in blood samples was determined by a real-time fluorogenic RT-PCR analysis of DEN-2 RNA copies. (a) Initial DEN-2 virus RNA copies in blood upon admission were significantly higher in patients with primary than those with secondary DEN-2 infections (P=0.030). (b) Kinetic changes of the virus RNA copies in blood from patients with primary and secondary DEN-2 infections revealed a similar trend showing higher virus load in primary DEN-2 infections. Data are presented as mean±SE calculated from samples collected 2, 4 and 6 days after admission. In the primary infection group (•), 17 samples were collected at 2, 4 and 6 days. In the secondary infection group (○), 14 samples were collected at 2, 4 and 6 days.

DEN IgG and DEN-1 and DEN-2 neutralization titers in patients with primary and secondary infections

Cross-reactive antibodies among heterotypic DEN viruses have long been implicated in the pathogenesis of secondary DEN infections. Experiments were next performed to investigate whether the serum anti-DEN-1 neutralization antibody titers were different between patients with primary and secondary DEN-2 infections. The rationale underlying our investigation of DEN-1-specific neutralization titers was that there had been a DEN-1 outbreak with possible subclinical infections in the past decade in southern Taiwan. The results indicated that patients with secondary DEN-2 infections had higher anti-DEN-1 and anti-DEN-2 neutralization titers than those with primary DEN-2 infections (P<0.05; Table 3). This is consistent with AI values measured via the capture dengue IgG assay (P<0.001; Table 3). The higher AI titers in secondary DEN-2 patients than in primary DEN-2 patients might reflect augmentation of cross-antibody production in the secondary DEN-2 infection.

View this table:
Table 3

DEN antibodies in primary and secondary DEN-2 infections

PrimarySecondaryP
Capture DEN IgG titer15.1 ± 4.5866.6 ± 119.5<0.001
Anti-DEN-1 NT titer17.2 ± 63.7138.2 ± 74.90.013
Anti-DEN-2 NT titer41.0 ± 72.4186.9 ± 72.60.047
  • Capture DEN IgG titers are represented as antibody index (AI) calculated from comparison of the mean calibrator value of the triplicate results from a reference serum. P-values were determined by Student's t-test.

  • Anti-DEN-1 and anti-DEN-2 neutralization titers (NT) were determined as that resulting in 80% reduction of virus PFU, and are presented as geometric mean titers.

Discussion

The roles of cytokines in DEN infections have mainly been studied based on comparisons between patients with DHF and DF (Kurane et al. 1989, 1991; Chaturvedi et al. 2000; Pacsa et al. 2000; Mustafa et al. 2001; Gagnon et al. 2002). Earlier studies showed that Th1 mediators such as IL-2, IFN-γ or IL-12 might be implicated in the pathogenesis of severe dengue infection (Kurane et al. 1989, 1991; Pacsa et al. 2000); later studies favored a Th2-dominant reaction in DHF (Chaturvedi et al. 2000; Mustafa et al. 2001). In the present study, we have tried to focus on the comparisons of clinical severity, immune mediators and virus load between patients with primary and secondary DEN-2 infections. We found higher IL-10 but lower IFN-α blood concentrations in patients with secondary DEN-2 infections. IFN-γ and IL-13 concentrations in blood were not significantly different between patients with primary and secondary DEN-2 infections. IFN-α is an early proinflammatory cytokine that acts as an upstream regulator of IFN-γ production (Cousens et al. 1999), and IL-10 is also an important proinflammatory cytokine that can regulate IFN-α production in virus infections (Payvandi et al. 1998). Both IFN-γ and IL-13 are, by contrast, later mediators directing the reciprocal polarization of Th1 and Th2 reactions (Paludan et al. 1997; Durbin et al. 2002). Taken together, this indicates that the earlier proinflammatory reaction associated with IFN-α and IL-10 production may be more critical than the later Th1/Th2 reaction polarized by IFN-γ and IL-13 production in the pathogenesis of heterotypic secondary DEN-2 infections.

In a previous in vitro study, we have found that only in the presence of certain titers of subneutralizing DEN-1 immune sera, human mononuclear leukocytes in response to DEN-2 viruses released lower IFN-γ but higher IL-10 levels (Yang et al. 2001). This observation has been further supported by the present study showing that a lower IFN-α but higher IL-10 in blood was associated with secondary DEN-2 infections, together with a higher frequency of DHF. This result is similar to that reported from Cuba (Perez et al. 2004; Guzman, 2005), but differs from those of Kurane (1991) and Libraty (2002), who demonstrated that blood IFN-γ levels were significantly higher in patients with DHF than those with DF. The variable results might be related to certain outbreaks with different virus subtypes or related to different sequential DEN infections in different populations. In southern Taiwan, the disease has been suspected to have subclinical DEN-1 transmission (Chen et al. 1996). In this study, we confirmed the subclinical DEN infections and demonstrated that patients with secondary DEN-2 infections tend to have a higher rate of DHF. This kind of subclinical transmission of DEN infections has also been reported from India (Singh et al. 2000). This suggests that heterotypic secondary DEN infections associated with a higher rate of DHF development may be more frequent in Taiwan and possibly in other Asian countries. A better vector control policy and identification of subclinical DEN infections should be taken to prevent DHF outbreaks in these countries.

Epidemiological studies have repeatedly demonstrated that secondary DEN infections contribute to higher rates of DHF in Thailand and Cuba (Sangkawibha et al. 1984; Makino et al. 1994; Guzman et al. 2000; Anantapreecha et al. 2005; Guzman, 2005). A recent report from Cuba showed that secondary dengue infection contributes to more than 97% of the severe cases of DHF, with younger infants and older adults being more susceptible to DHF and fatality (Guzman, 2005). Almost all of the DHF cases caused by DEN-2 and DEN-4 were as a result of secondary infection, whereas 20% of the DHF cases caused by DEN-1 and DEN-3 in Thailand were the result of primary infections (Anantapreecha et al. 2005). In our DEN-2 outbreak, we found a similar trend showing that patients with secondary DEN-2 infections tended to have a higher rate of developing DHF than primary infections (61.7% vs. 36.3%) in a hospital-based analysis. The reason why patients with primary DEN-2 infection also showed a higher DHF rate in our study may be because we analysed those hospitalized but not the general population.

Several studies have explored the relationship between virus load and clinical severity of dengue infections. Some revealed that clinical severity of the DEN-2 (Vaughn et al. 2000) and DEN-3 (Libraty et al. 2002) infections was related to an increase in virus load. Another study of a DEN-3 outbreak in Tainan, Taiwan, in 2000 showed that a higher and longer DEN-3 viremia was found in patients with DHF than those with DF (Wang et al. 2003). Murgue (2001), however, showed that there was no difference of virus load between patients with primary and secondary DEN-2 infections. Kuberski (1997) even reported that the magnitude of viremia was higher in patients with primary than secondary DEN-1 infections. In a simultaneous determination of virus load, disease severity and immune reactions in patients with primary and secondary DEN-2 infections, our results indicated that the virus load was significantly lower in patients with secondary than in those with primary DEN-2 infections. The lower virus load in patients with secondary DEN-2 infections was associated with higher anti-DEN antibody titers and a higher frequency of DHF in this DEN-2 outbreak. This suggests that patients with secondary DEN-2 infections tend to have more rapidly increasing and higher anti-DEN IgG antibody titers than those with primary infection, but elicit an altered immune response showing higher IL-10 but lower IFN-α concentrations in blood associated with the development of DHF. Modulation of early IFN-α and IL-10 production in secondary DEN infections may be considered while developing pharmacotherapeutics for DHF treatment.

Authors contribution

W.-T.Y. and R.-F.C. contributed equally to this work.

Acknowledgements

This work was in part supported by grants NSC-92-2314-B-182A-110 and NSC-93-2314-B-182A-010 from the National Science Council, Taiwan.

Footnotes

  • Editor: Patrick Brennan

References

View Abstract