Different Profiles of Cytokines, Chemokines and Coagulation Mediators Associated with Severity in Brazilian Patients Infected with Dengue Virus

The incidence of dengue in Latin America has increased dramatically during the last decade. Understanding the pathogenic mechanisms in dengue is crucial for the identification of biomarkers for the triage of patients. We aimed to characterize the profile of cytokines (IFN-γ, TNF-α, IL-1β, IL-6, IL-18 and IL-10), chemokines (CXCL8/IL-8, CCL2/MCP-1 and CXCL10/IP-10) and coagulation mediators (Fibrinogen, D-dimer, Tissue factor-TF, Tissue factor pathway inhibitor-TFPI and Thrombomodulin) during the dengue-4 epidemic in Brazil. Laboratory-confirmed dengue cases had higher levels of TNF-α (p < 0.001), IL-6 (p = 0.005), IL-10 (p < 0.001), IL-18 (p = 0.001), CXCL8/IL-8 (p < 0.001), CCL2/MCP-1 (p < 0.001), CXCL10/IP-10 (p = 0.001), fibrinogen (p = 0.037), D-dimer (p = 0.01) and TFPI (p = 0.042) and lower levels of TF (p = 0.042) compared to healthy controls. A principal component analysis (PCA) distinguished between two profiles of mediators of inflammation and coagulation: protective (TNF-α, IL-1β and CXCL8/IL-8) and pathological (IL-6, TF and TFPI). Lastly, multivariate logistic regression analysis identified high aspartate aminotransferase-to-platelet ratio index (APRI) as independent risk factors associated with severity (adjusted OR: 1.33; 95% CI 1.03–1.71; p = 0.027), the area under the receiver operating characteristics curve (AUC) was 0.775 (95% CI 0.681–0.869) and an optimal cutoff value was 1.4 (sensitivity: 76%; specificity: 79%), so it could be a useful marker for the triage of patients attending primary care centers.


Introduction
Dengue is a mosquito-borne viral disease caused by the dengue virus (DENV), which belongs to the family Flaviviridae, genus Flavivirus. There are four different but genetically related serotypes (DENV-1, DENV-2, DENV-3 and DENV-4). Dengue is endemic in more than 100 countries, and it has been estimated that 96 million symptomatic infections

Study Population
In the period from January-March 2013, adults who met criteria for a suspected dengue case according to the guidelines of the Brazilian Ministry of Health [28] were enrolled at healthcare centers in the states of Rio de Janeiro (RJ) and Mato Grosso do Sul (MS): Rio Laranjeiras Hospital, RJ; Plantadores de Cana Hospital, RJ; Reference Center for Infectious and Parasitic Diseases (CEDIP), MS; and Guanandy Health Center, MS.

Sample Collection
After the participants gave their informed consent, approximately 20 mL of peripheral blood was obtained by means of venipuncture using BD Vacutainer™ tubes containing acid-citrate-dextrose (catalogue # BD 364606), and the plasma was centrifuged, aliquoted and stored at −70 • C until analysis. In addition, 15 adults who had not had febrile episode in the last three months or any history of other diseases were included as healthy controls for analysis of cytokines, chemokines and coagulation factors.

Dengue Diagnosis
All the dengue patients were positive by molecular and/or serological assays.

Serological Assays
Detection of nonstructural protein 1 (NS1) antigen was performed using the Platelia TM Dengue NS1 Ag ELISA kit (Bio-Rad Laboratories, Marnes La Coquette, France), in accordance with the manufacturer's instructions. Quantification of circulating DENV NS1 antigen was also performed, as described previously [29]. Detection of immunoglobulin M (IgM) anti-dengue was done using the Panbio ® Dengue IgM capture ELISA kit (Panbio Inc., Queensland, Australia), in accordance with the manufacturer's instructions.
IgG anti-dengue detection by means of the Dengue Virus IgG DxSelect TM kit (Focus Diagnostics, Cypress, CA, USA) was used to differentiate between primary and secondary infection. The infection was established as primary when the IgM/IgG optical density ratio was ≥1.2, as previously described [30].

Molecular Assays
For molecular tests, viral RNA was extracted using the QIAamp Viral RNA Mini kit (Qiagen, Hilden, Germany), in accordance with the manufacturer's instructions. DENV detection and typing were performed as described previously [31] and by using the Simplexa TM Dengue real-time RT-PCR assay (Focus Diagnostics, Cypress, CA, USA), in accordance with the manufacturer's instructions.

Classification of Cases
The participants were classified as presenting dengue without warning signs (DwoWS), dengue with warning signs (DwWS) or severe dengue (SD), in accordance with the WHO's 2009 guidelines [6]. Febrile patients with negative tests for dengue were classified as presenting acute undifferentiated febrile disease (AUFI).

Hematological and Biochemical Parameters
The results from laboratory tests such as complete blood count, aspartate aminotransferase (AST), alanine aminotransferase (ALT), fibrinogen, prothrombin time (PT) and activated partial thromboplastin time (aPTT) performed on admission were collected for analysis in this study. Furthermore, the aspartate aminotransferase-to-platelet ratio index (APRI) was calculated using the following formula: AST (IU/L)/upper limit normal/platelet count (×10 9 /L) × 100. The neutrophil-to-lymphocyte ratio (NLR), monocyteto-lymphocyte ratio (MLR) and platelet-to-lymphocyte ratio (PLR) were also calculated.

Statistical Analysis
The Shapiro-Wilk test was used to assess the normality of the measured data. Continuous variables were presented as means with standard deviations (SDs) or as medians with interquartile ranges (IQRs), as appropriate. The Kruskal-Wallis test (followed by Dunn's multiple comparisons test), one-way ANOVA, chi-square test and Fisher's exact test were used to compare differences between groups, as appropriate. Spearman's rank correlation between the variables studied was represented in a correlogram.
Principal component analysis (PCA) was used to reduce dimensionality and identify cytokines/chemokines and coagulation mediator patterns associated with disease severity. The Kaiser-Meyer-Olkin (KMO) test and Bartlett's sphericity test were used to assess whether the data were adequate for this analysis. PCA with subsequent varimax rotation determined the number of principal components and the cumulative percentage of variance. According to the Kaiser criterion, we use only principal components with associated eigenvalues greater than one. The individual scores for each component were transformed to a scale from 0 to 1, and the difference between the groups of patients was evaluated.
Bivariate analyses and stepwise multivariate logistic regression with backward elimination (p < 0.05) were used to identify laboratory markers associated with severity. The interactions of independent variables were assessed through a multicollinearity test using the variance inflation factor (VIF). The area under the receiver operating characteristic (ROC) curve (AUC) was calculated and the optimal cutoff point was determined using the Youden index.
All statistical analyses were conducted in the R Statistical Language version R-4.04 for Windows [R Core Team (2021). R: A language and environment for statistical computing and R Foundation for Statistical Computing, Vienna, Austria. URL: https://www.r-project. org/ (accessed on 20 April 2021)]; p values < 0.05 were considered statistically significant.

Baseline Clinical Characteristics of the Confirmed Dengue Cases
During the study period, 316 patients were recruited. The diagnosis of dengue was confirmed in 225 patients (71.2%), and 91 patients (28.8%) with negative serological and molecular tests were classified as AUFI. The median age of the dengue patients was 36 years (IQR: 26-50 years), 58% were women (male-to-female ratio = 0.7) and the most frequent comorbidity was hypertension (19%).

Type of Infection and Serotype
The dengue patients were investigated for prior dengue infection. Secondary infection was determined in 69.2% of the dengue patients. There was no significant difference in this proportion between the DwoWS (68.2%) and DwWS/SD (71.4%) patients.
In addition, DwWS/SD patients had significantly higher APRI values (p < 0.001) compared to DwoWS patients, and significantly higher APRI values were also found in patients with abdominal pain (p = 0.001) and mucosal bleeding (p = 0.001). Likewise, DwWS/SD patients had significantly lower PLR values (p = 0.018) compared to DwoWS patients, but no significant differences were found in relation to NLR and MLR (Table 1).

Cytokines and Chemokines
The circulating levels of cytokines and chemokines in Brazilian patients infected with DENV-1, DENV-2 and DENV-3 serotypes were determined previously by our group [19][20][21][22][23]. In the same way as previously, we now characterized the cytokine and chemokine profiles from patients during the dengue-4 epidemic using a multiplex-microbead and ELISA immunoassays.
In addition, we next investigated any association between laboratory parameters and the circulating levels of cytokines/chemokines in the dengue patients.

Coagulation Mediators
Our group has shown coagulopathies during DENV infection in previous epidemics [26,32,33]. In this context, we described some aspects of coagulation and fibrinolysis in patients during the dengue-4 epidemic. In addition to thrombocytopenia, the dengue patients also presented coagulopathy, as demonstrated by increased D-dimer levels (p = 0.01) and TFPI plasma levels (p = 0.04) and decreased fibrinogen levels (p = 0.04) and TF plasma levels (p = 0.005) ( Table 2). Coagulopathy also contributed to the presence of bleeding: prolonged aPTT was seen in a high proportion of DwWS/SD patients (78%; p = 0.004) ( Table 1).
The relationship between laboratory parameters and circulating levels of coagulation mediators was analyzed. A positive correlation was found between APRI values and plasma levels of TFPI (r s = 0.530; p < 0.001; n = 42) and between plasma levels of IL-6 and D-dimer (r s = 0.943; p = 0.005; n = 6) ( Figure 1). Additionally, we found a positive association between plasma levels of TFPI and TF (r s = 0.688; p = 0.003; n = 16) in DwWS/SD patients.

Principal Component Analysis (PCA) for Cytokines, Chemokines and Coagulation Mediators in Dengue Patients
PCA was performed for cytokines, chemokines and coagulation mediators from 54 dengue patients (39 DwoWS and 15 DwWS/SD). Two different combinations of inflammatory mediators were identified (eigenvalue-one criterion) with a cumulative percentage of variance of 80.8% after varimax rotation. The KMO test result was 0.59, and Bartlett's test was significant (p < 0.001). The pro-inflammatory cytokines TNF-α and IL-1β and the chemokine CXCL8/IL-8 contributed mainly to the first principal component (PC1), while the coagulation mediators TF and TFPI and the pro-inflammatory cytokine IL-6 had an orthogonal contribution to the second principal component (PC2) (Figure 2). For PC1, significantly higher individual score values were observed among DwoWS patients than among DwWS/SD patients (p = 0.015). However, for PC2, there was no significant difference between these two groups of dengue patients.

Coagulation Mediators
Our group has shown coagulopathies during DENV infection in previous epidemics [26,32,33]. In this context, we described some aspects of coagulation and fibrinolysis in patients during the dengue-4 epidemic. In addition to thrombocytopenia, the dengue patients also presented coagulopathy, as demonstrated by increased D-dimer levels (p = 0.01) and TFPI plasma levels (p = 0.04) and decreased fibrinogen levels (p = 0.04) and TF plasma levels (p = 0.005) ( Table 2). Coagulopathy also contributed to the presence of bleeding: prolonged aPTT was seen in a high proportion of DwWS/SD patients (78%; p = 0.004) (Table 1).
The relationship between laboratory parameters and circulating levels of coagulation mediators was analyzed. A positive correlation was found between APRI values and plasma levels of TFPI (rs = 0.530; p < 0.001; n = 42) and between plasma levels of IL-6 and D-dimer (rs = 0.943; p = 0.005; n = 6) ( Figure 1). Additionally, we found a positive association between plasma levels of TFPI and TF (rs = 0.688; p = 0.003; n = 16) in DwWS/SD patients.
In the ROC curve analysis, APRI performed better for distinguishing DwWS/SD patients compared to the other factors that were also associated with severity (Table 3).
lett's test was significant (p < 0.001). The pro-inflammatory cytokines TNF-α and IL-1β and the chemokine CXCL8/IL-8 contributed mainly to the first principal component (PC1), while the coagulation mediators TF and TFPI and the pro-inflammatory cytokine IL-6 had an orthogonal contribution to the second principal component (PC2) (Figure 2). For PC1, significantly higher individual score values were observed among DwoWS patients than among DwWS/SD patients (p = 0.015). However, for PC2, there was no significant difference between these two groups of dengue patients.

Discussion
The increased inflammatory response is considered to be the main factor responsible for the pathogenesis of dengue. Therefore, the components of the immune response, including cytokines and other cellular mediators, have been studied as biomarkers of severe disease [34].
This study was conducted in two cities in the regions most affected by the dengue epidemic of 2013 in Brazil (southeastern and central-western regions). DENV-4 and secondary infections were more frequently identified, as had previously been seen in another study conducted during this epidemic [35].

Discussion
The increased inflammatory response is considered to be the main factor responsible for the pathogenesis of dengue. Therefore, the components of the immune response, including cytokines and other cellular mediators, have been studied as biomarkers of severe disease [34].
This study was conducted in two cities in the regions most affected by the dengue epidemic of 2013 in Brazil (southeastern and central-western regions). DENV-4 and secondary infections were more frequently identified, as had previously been seen in another study conducted during this epidemic [35].
Although it has been established that the circulating levels of NS1 antigen are positively correlated with DENV viremia [36] and are associated with severity [37,38], there is still debate about its usefulness as a biomarker [39]. In our study, no significant difference was found in the circulating levels of the NS1 antigen between the dengue patient groups.
The hematological and biochemical parameters that we determined among dengue patients were consistent with what has been reported in studies of dengue epidemics with the other serotypes [40,41]. The dengue patients, especially DwWS/SD patients, had greater leukopenia, greater thrombocytopenia and higher elevated transaminase levels than the AUFI patients. Although it has been established that liver aminotransferases are not a good biomarker to discriminate severity [42], a recent meta-analysis suggested that the monitoring of platelet counts and transaminase levels during the acute phase of the disease could improve the early prediction of severe dengue [43].
The mechanisms involved in thrombocytopenia during DENV infection are not yet fully understood. Various mechanisms have been hypothesized, including the suppres-sion of bone marrow and the peripheral destruction of platelets, but it has already been established that platelets participate in the inflammatory and immune response [44]. In the present study, the dengue patients with thrombocytopenia had significantly higher levels of IL-10. This finding was in line with the hypothesis of formation of monocyte-platelet aggregates as a cause of thrombocytopenia and cytokine release such as IL-10 [45].
Liver involvement is common in DENV infection and is shown by increased levels of transaminases [46]. Many factors are thought to contribute to liver damage, including hepatocyte apoptosis directly due DENV, hypoxia due to impaired perfusion and immunemediated injury via upregulated cytokines/chemokines [47]. In our study, it was shown that the transaminase levels of the dengue patients had positive correlations with the plasma levels of IL-18.
The inflammatory response is essential for resistance to DENV infection, and it has been established that this response differs between mild and severe cases [48]. Thus, the cytokine/chemokine profile can serve as a laboratory tool for predicting severe disease [49,50]. In our study, higher plasma levels of the proinflammatory cytokines TNF-α, IL-6 and IL-18, and the anti-inflammatory cytokine IL-10 and the chemokines CXCL8/IL-8, CCL2/MCP-1 and CXCL10/IP-10 were found in the dengue patients compared to the healthy controls. Furthermore, PCA was performed to reduce the dimensionality of the data on the inflammatory and coagulation mediators studied and allowed us to identify two patterns in dengue patients. The first pattern was hyper-inflammatory and was characterized by increases in the levels of the cytokines TNF-α, IL-1β and CXCL8/IL-8, such that they became significantly higher in DwoWS patients. The second pattern was characterized by increases in the levels of the inflammatory cytokine IL-6 and the coagulation mediators TF and TFPI but which did not differ between the two groups of dengue patients, probably due to the fact that the majority were infected with DENV-4 and there were few cases of severe dengue. These findings are consistent with a Thai study that evaluated the cytokine profile in children with subclinical and symptomatic DENV infection and the identification of protective and pathological immunological profiles [24].
Thus, it has been hypothesized that TNF-α, which was found at significantly higher levels in DwoWS patients, belongs to the protective immunological profile, probably due to its ability to exert a protective effect on virus-induced, monocyte-derived dendritic cell apoptosis [51], while the findings on IL-1β and CXCL8/IL-8 could be associated with the predominant DENV-4 serotype and further studies are required. In addition, IL-6, which belongs to the pathological immunological profile and was found at significantly higher levels in DwWS/SD patients, has been associated with severe dengue in several studies [18].
In agreement with our previous studies [19], another pro-inflammatory cytokine that increased markedly in the dengue patients in our study was IL-18. DENV infection activates the NLRP3 inflammasome in macrophages and causes a release of IL-1β and IL-18 by caspase-1 [52]. Studies have shown that high plasma levels of IL-18 coincide with high levels of IL-18 binding protein (IL-18BP), a natural antagonist of IL-18, and that blocking IL-18 activity results in lower IFN-γ levels and higher TNF-α and IL-6 levels [53]. In our study, plasma IFN-γ levels were higher in the dengue patients than in healthy controls, although not significantly. On the other hand, in another study, hepatocytes and lymphocytes from the inflammatory infiltrate in the portal tract showed high expression of IL-18 [54]. In our study, a positive association between IL-18 and transaminase levels was observed, which could reflect the acute inflammatory response that occurs in the liver and contributes to liver damage.
Among the chemokines analyzed, elevated levels of CXCL10/IP-10 in the dengue patients were associated with high levels of NS1 antigen. CXCL10/IP-10 plays an important role in the recruitment of activated NK cells and may compete with DENV for binding to heparan sulfate on hepatocytes. Thus, high levels of CXCL10/IP-10 may limit the spread of infection [55]. In addition, another recent study showed that there was significant upregulation of CXCL10/IP-10 in secondary infection due to the dengue virus compared to the other cytokines analyzed [56].
In DENV infection, both the coagulation and the fibrinolysis processes are activated, thereby causing alterations to their laboratory parameters [44]. In our study, the dengue patients showed significantly elevated levels of fibrinogen, D-dimer and TFPI and decreased levels of TF compared to healthy controls, but no significant difference in the TM levels was seen. Moreover, more than half of the dengue patients presented prolonged aPTT and PT. Thus, the prolongation of aPTT and PT, together with decreased fibrinogen levels and increased D-dimer levels, suggest that increased procoagulant activity is associated with the activation of fibrinolysis as a response mechanism.
The decrease in fibrinogen levels can be explained by higher consumption due to the procoagulant state but also due to extravasation and decreased hepatic fibrinogen synthesis. However, it has been suggested that fibrinolysis in secondary dengue virus infections is not as prominent as in primary infections [57]. In our study, significantly low levels of fibrinogen, although still within the normal range, were found in DwWS/SD patients compared to DwoWS patients.
Studies have suggested monitoring aPTT, PT and D-dimer as early markers of severity and death [57][58][59]. In our study, prolonged aPTT values were found in DwWS/SD patients compared to DwoWS patients, but no significant differences in PT values or D-dimer levels were found between these two groups. Likewise, patients with mucosal bleeding presented elevated levels of IL-6 and IL-10 compared to those without mucosal bleeding, which is explained by the relationship between coagulopathy and the inflammatory response [60,61].
Among the coagulation mediators analyzed, TF and TFPI together with IL-6 formed the second pattern identified in the dengue patients through PCA analysis. Normally, exposure to membrane-bound TF initiates coagulation after vascular injury. Monocytes and endothelial cells do not express TF but do express it under pathological conditions and after exposure to inflammatory cytokines, thus activating the coagulation cascade through the TF pathway. In addition, TF exerts pro-inflammatory activity through activating PARs (protease-activated receptors), which are expressed in endothelial cells, mononuclear leukocytes, platelets, fibroblasts, smooth muscle cells and other cells. This produces upregulation of pro-inflammatory cytokines and adhesion molecules [44].
Overall, regulation of TF and TFPI in dengue infection has been analyzed by few studies to date. In our study, higher levels of TFPI and lower levels of TF were found in dengue patients than in healthy controls, which could be linked to the higher frequency of DENV-4 infected patients. A previous study published by our group had already found that patients with DENV1 and DENV2 infection had higher TF levels than DENV-4 infected patients [32]. TFPI levels increased in response to elevation of TF levels, and thus a positive correlation was shown between these two mediators in DwWS/SD patients in our study. TFPI is an anticoagulant that acts by inhibiting the TF-VIIa complex and factor X, but also has anti-inflammatory properties, which is why it is being studied for the treatment of infectious and inflammatory diseases [62,63].
In addition, it has been established that D-dimer correlates with the activation of the proinflammatory cytokine cascade in critically ill patients [64]. In our study, it was shown that the plasma levels of IL-6 of dengue patients had a positive correlation with the D-dimer levels. IL-6 promotes coagulation without affecting fibrinolysis, and elevated levels of Il-6 increase the surface expression of TF in monocytes and indirectly through C-reactive protein, an acute-phase protein that is released by hepatocytes after stimulation with IL-6, which also increases tissue factor activity in monocytes [65]. IL-6 and D-dimer have recently been shown to be closely related to severe cases of COVID-19 [66].
Lastly, in the multivariate adjusted logistic regression analysis, only APRI was independently associated with severity (adjusted OR: 1.33; 95% CI 1.03-1.71; p = 0.027). In addition, our study found a positive association between APRI values and the levels of pro-inflammatory cytokines, chemokines and coagulation mediators (IL-6, IL-18, CXCL10/IP-10 and TFPI) in dengue patients. Hence, this marker could be useful in the triage of patients who attend primary care facilities. Previous works had already proposed APRI as a marker of severity, however, there is still limited information in this regard [67,68].
However, standardized prospective studies with a larger sample size will be necessary to validate our results.
Our study has some limitations. The sample size and non-random selection of participants may not represent the heterogeneity of the general population. It was not possible to analyze the same number of samples for all laboratory markers, and PCA analysis was performed in only 54 patients, which could cause bias in our analysis. Likewise, due to the cross-sectional design of the study, it was not possible to obtain information on dynamic changes to the laboratory markers through paired samples.

Conclusions
Brazilian patients with dengue during the DENV-4 epidemic presented activation of inflammatory, anti-inflammatory and coagulation mediators, which were associated with alterations to hematological and biochemical parameters and the presence of warning signs. Two cytokine, chemokine and coagulation mediator profiles were identified: one that was protective, with high levels of TNF-α, IL-1β and CXCL8/IL-8, and another that was pathological, with high levels of IL-6, TF and TFPI. Lastly, APRI could be a useful marker in triage of patients attending primary care centers.  Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.