key: cord-0269872-pv7tgwcg
authors: Satchidanandam, V.; Pradeep, S. P.; Venkatesh, P. H.; Manchala, N. R.; Veedu, A. V.; Basavaraju, R. K.; Selvasundari, L.; Ramakrishna, M.; Chandrakiran, Y.; Krishnamurthy, V.; Holigi, S.; Thomas, T.; Ross, C. R.; Dias, M.
title: Innate immune cytokine profiling and biomarker identification for outcome in dengue patients
date: 2020-10-15
journal: nan
DOI: 10.1101/2020.10.14.20212001
sha: 3bd5bc01b7737a265fa128681c6a30f1ba55bdfa
doc_id: 269872
cord_uid: pv7tgwcg

Biomarkers of progression to severe dengue are urgently required for effective patient management. Innate immune cells have been implicated in the enhancement of infection and cytokine storm associated with dengue severity. Using intracellular cytokine staining and flow cytometry, we observed significantly higher proportions of innate immune cells secreting inflammatory cytokines dominated by IFN-{gamma} and TNF- at admission associated with good prognosis. Secondary dengue predisposed to severe outcomes. In patients with severe dengue and those with liver impairment, early activation as well as efficient down-regulation of innate responses were compromised. IFN-{gamma}+CD56+CD3+ NKT cells and IL-6+ granulocytes served as novel biomarkers of progression to severity (composite AUC=0.85-0.9). Strong correlations among multiple cytokine-secreting innate cell subsets pointed to coordinated activation of the entire innate immune system by DENV.

TNF-α + CD16 + monocytes, (E) TNF-α + granulocytes and (F) TNF-α + CD19 + B cells from dengue patients compared to febrile controls (FC) and healthy controls (HC). Each dot represents one 5 patient; P value displayed with median and IQR. 5 When assessed as a function of disease severity, a significantly greater proportion of TNFα + granulocytes were evident in DF relative to DFWS/SD ( Fig. 2A) , a trend also evident for dual secretion of TNF-α and IL-6 by this cell subset and IL-10 secretion by CD56 + CD3 + NKT cells ( Fig. 2B, C) . When we used bleed-scores or liver enzyme levels as a surrogate of severity, those with no bleeding or normal liver enzyme levels carried a significantly greater proportion of innate 5 cells secreting inflammatory cytokines compared to those with varying degrees of hemorrhage or abnormal liver enzyme levels ( 6 S10E), suggesting a requirement for high viral antigen levels to achieve efficient innate cell activation.

associated with better prognosis. Frequency of (A) total TNF-α + granulocytes, (B) TNF-α + IL-6 + 5 granulocytes, and (C) total IL-10 + CD56 + CD3 + NKT cells compared between DF and DFWS/SD.

(D) Frequency of total TNF-α + CD14 + CD16 + monocytes compared between bleed-scores (BS).

Frequency of total (E) TNF-α + CD14 + monocytes and (F) TNF-α + CD14 + CD16 + monocytes compared between normal and elevated ALT (>55 IU/L; E) and AST (>48 IU/L; F). P values with median and IQR reported. 7 In order to query the link if any, between kinetics of innate immune activation by DENV and disease severity, we compared innate cell cytokine secretion between different measures of severity at early (days 1-3), intermediate (days [4] [5] [6] and late (days 7-15) times of hospital presentation. Patients admitted 1-3 days post symptom onset (dpso) had impressively higher TNFα (Fig. 3A) and IFN-γ (Fig. 3B , S11A) -secreting innate cells in those with normal compared to 5 above-normal liver enzyme levels. Those admitted 4-6 dpso had a significantly greater proportion of IFN-γ + CD56 + CD3 + NKT cells as well as TNF-α-secreting granulocytes, CD19 + B cells and CD56 + CD3 + NKT cells in DF relative to DFWS/SD (Fig. 3 , C-F). In contrast, patients with high liver enzymes failed to down-regulate secretion of TNF-α, IL-6 and IP-10 from different innate cell subsets during the late stage (7-15 dpso) that was prominent in patients with normal levels 

Frequency of total (A) TNF-α + CD14 + monocytes and (B) IFN-γ + CD56 + CD3 + NKT cells during 1-3 dpso compared between normal and high levels of ALT (A; IU/L) and AST (B; IU/L). Frequency 5 of total (C) IFN-γ + CD56 + CD3 + NKT cells, (D) TNF-α + granulocytes, (E) TNF-α + CD19 + B cells and (F) TNF-α + CD56 + CD3 + NKT subsets during 4-6 dpso compared between DF and DFWS/SD. Frequency of total (G) TNF-α + CD56 + CD3 + NKT cells and (H) IL-6 + granulocytes during 7-15 dpso compared between normal and high AST levels. To identify potential biomarkers of progression to severity, we compared those who worsened after recruitment (as evidenced by a shift from DF/DFWS to DFWS/SD or death), with patients who readily recovered from DF and DFWS/SD. Both recovered DF and DFWS/SD patients carried a significantly greater proportion of IFN-γ + TNF-α + CD56 + CD3 + -, IFNγ + CD56 + CD3 + -NKT cells and monofunctional IL-6 + granulocytes (Fig. 4 , A-C) relative to 5 worsened patients. To assess the biomarker performance of these cells, receiver operating characteristic (ROC) curve analysis was performed. IFN-γ + TNF-α + CD56 + CD3 + -, IFNγ + CD56 + CD3 + NKT and monofunctional IL-6 + granulocytes provided AUC of 0.77, 0.76 and 0.75 respectively with 90% sensitivity and 60 to 66% specificity when DF was compared with the worsened group (fig. S12, A-C). Combining IL-6 + granulocytes and IFN-γ + CD56 + CD3 + NKT cells 10 using binary logistic regression resulted in composite AUC of 0.85 ( Fig. 4D ; table S5) and revealed that every one percentage rise in IFN-γ + CD56 + CD3 + NKT resulted in 3.12 fold lower odds of worsening (95% CI=1.1-9.2, P=0.035). In patients with elevated AST, this composite biomarker predicted the progression to severity with higher accuracy (AUC=0.9) and displayed 100% sensitivity with 81.9% specificity (table S5). 15 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint IFN-γ + TNF-α + CD56 + CD3 + , (B) total IFN-γ + CD56 + CD3 + NKT cells and (C) IL-6 + IL-10 -IP-10 -TNF-αgranulocytes compared between DF, DFWS/SD and worsened (W) patients. P value with median and IQR reported. (D) Composite ROC curve for total IFN-γ + CD56 + CD3 + NKT cells and 5 IL-6 + IL-10 -IP-10 -TNF-αgranulocytes comparing DF (purple) or DFWS/SD (green) with worsened patients. (*AUC), composite AUC.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 15, 2020. . This study, the first to query the cellular source of innate inflammatory cytokines in a large, blinded dengue cohort, conclusively demonstrated the beneficial role of early innate cell activation triggered by high viral antigen levels, in ensuring recovery from dengue. DENV activated all innate immune subsets resulting in mono and polyfunctional cytokine production that correlated with good outcome. The innate immune cytokine signature for each pathogen may be unique and 5 rewarding to investigate.

The observed persistence of innate cell-derived cytokines during late phase of disease pointed to dysregulated innate responses in SD (20). Severe COVID-19 patients also displayed sustained plasma levels of IP-10 and IL-6 (21); increased plasma IP-10 also correlated with liver impairment in HIV/HBV patients (22). In light of the reported requirement of IP-10 for B cell activation (23), 10 our finding of abnormal high IP-10 levels during late stages of disease is a likely contributor to the B cell mediated pathology attributed to severe dengue (24). A potential role for persistently elevated innate responses in provoking the reported inappropriate TCR signaling and T cell apoptosis in SD (8, 9) warrants further investigation. Thus, failure to achieve both robust activation and prompt attenuation of innate immune cells was a hallmark of severe dengue. 15 Our composite biomarker performed well despite limited number of patients with transitions to greater severity and inclusion of patients with varying disease duration. Potential enhancement of its performance by including additional hitherto unidentified cytokine secreting cells will enhance its utility in a clinical setting. Ease of processing and ready availability of flow cytometers in diagnostic laboratories assures feasibility of host blood-based biomarker deployment whereas 20 biomarkers reliant on expensive instruments and high-end technical skills, may have limited utility in resource constrained geographies (25, 26).

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 15, 2020. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 15, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint 16 supplementary materials. De-identified flow cytometry files (.fcs) that support the results reported in this article, and analyzed flow cytometry data (.xlsx) will be made available on request.

Supplementary Text 5 Figures S1-S13

Tables S1-S11

References (29, 30) STARD Checklist 10 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. 

Figs. S1 to S13 Tables S1 to S11

Other Supplementary Materials for this manuscript include the following:

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint

This study was carried out in accordance with the Declaration of Helsinki. Institutional ethics committee approval to conduct the study was obtained from the four participating hospitals; which represents the annual dengue season following onset of monsoon rains at the four hospitals listed above. Written informed consent was obtained from all participants before sample collection and analysis. Blood samples were coded and labeled as follows: DEN/hospital abbreviation/19/### before being sent to IISc for flow cytometry analysis. Consecutive suspected adult dengue patients (≥ 18 years) who tested positive using a dengue specific NS1/IgM rapid dengue day 1 test kit (J Mitra and Co., India) were recruited. Those who tested negative were recruited as febrile controls (FC); volunteers with no illness for the past 3 months were enrolled as healthy controls (HC; fig.   S1 ).

Sample size analysis for a desired power of 80%, type I error tolerance of 0.05, and a hypothesized effect size of 0.75, required at least 29 dengue patients who would transition post admission, to a worse condition as defined by the World Health Organization (WHO) categorization of dengue severity as follows: dengue fever (DF) was defined by headache, body ache, rash, nausea, or mild bleeding; dengue fever with warning signs (DFWS) included symptoms like persistent vomiting, . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. .

3 mucosal bleeding, pleural effusion, ascites, and hepatomegaly. Severe dengue (SD) included symptoms such as plasma leakage, ≥1000 IU/L of alanine aminotransferase (ALT) / aspartate aminotransferase (AST), severe bleeding which leads to shock, and/or organ impairment (3).

However, when the dengue cases ceased to appear in the hospitals by end November 2019, we had obtained only 10 patients who transitioned to a worse category, primarily owing to the clinical interventions following admission. The study was also designed to collect longitudinal samples at days 3 and 7 post-admission; however, only 154 patients provided a second sample and a single patient donated three consecutive samples. Demographic characteristics (i.e., gender and age), clinical features (i.e., days post symptom onset, nausea, head ache, body ache, abdominal pain, rashes, splenomegaly, hepatomegaly and bleeding manifestations) and routine hematological laboratory findings (i.e., complete blood cell count, serum albumin, liver enzymes, platelet count and hematocrit) were recorded. Patients were assigned bleed-scores (BS) as follows: no bleeding, 0; petechiae, 1; epistaxis/gingival bleeding/menorrhagia, 2; gastrointestinal bleeding, 3;

intracranial/intrapulmonary bleeding, 4. Plasma leakage in pleural and/or peritoneal cavities was confirmed using X-ray/ultrasound scans. HIV patients were not recruited. Data from samples of patients with co-infections (typhoid, sepsis, malaria, urinary tract infection, Hepatitis B) or those who were discharged against medical advice (DAMA) and samples with experimental errors (clotted blood samples, QC failure of flow cytometer, sample processing errors) were excluded from analysis.

Dengue virus (DENV) infection was confirmed using a commercial IgM, IgG and NS1 enzymelinked immunosorbent assay (ELISA; Panbio, Australia) and results were interpreted according to manufacturer's instructions. The kits were used to distinguish primary (IgM to IgG ratio >1.2)

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint 4 from secondary (IgM to IgG ratio <1.2) infection. Primary dengue status was also assigned to those who tested positive for DENV specific NS1 (index value >1.1) but were negative for IgM and IgG.

Blood samples collected in sodium citrate vacutainer tubes (BD Biosciences) were immediately processed, no later than 4 hours from collection. RBCs from 500µl blood were lysed using 4ml of 1X ammonium chloride buffer (166mM ammonium chloride, 9.9mM potassium bicarbonate and 0.126mM EDTA). The centrifuged cells were washed with 1X phosphate buffered saline (PBS) and stained with Fixable Viability Stain 450 [BD, Cat#562247] for 10 minutes at room temperature, to exclude dead cells. This was followed by staining for appropriate surface markers (table S6) for 30 minutes at 4°C. The surface marker TLR2 was superior to HLA-DR owing to its stable expression during infection in contrast to the latter which is reported to be down-regulated in all manner of inflammatory conditions, and was therefore used to identify monocytes (30). Cells is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint 5 intracellular cytokine-specific antibodies. FMF controls were used to set the positive gates for each cytokine ( fig. S3 ). Positive cytokine production was based on the criteria given for each cytokine from each cell subset in tables S7 and S8. Data were analyzed using FlowJo software (version 10.6.1). Polyfunctional cytokine secretion was assessed by Boolean gating. The analyzed data from FlowJo was submitted to the clinical statistician for unblinding of patient characteristics prior to statistical analysis.

Optimized t-Distributed Stochastic Neighbor Embedding (t-SNE; (31)) analysis to visualize the clusters within the NK/NKT cells was performed. All 32 patient samples with SD were included along with 32 each from DF and DFWS (WHO 2009 categorization) which were selected at random using the RAND function in Excel 2016. A subset of 10,000 events were selected from CD19cells for each sample using DownSample plugin, followed by concatenation of all events.

A total of 960,000 events and 7 markers (CD56, CD16, CD3, IP-10, IL-10, TNF-, and IFN-) were used to generate the t-SNE map. We used the KNN algorithm (random projection forest -ANNOY) and Barnes-Hut gradient algorithm implementation with the recommended parameters (iterations -1000; learning rate (eta) -67200) at perplexity = 50 in t-SNE plugin built within FlowJo.

All analyses were done using IBM SPSS statistics 23.0 and GraphPad prism version 8.

Significance between two or multiple groups was tested using Mann-Whitney U test (two-tailed) and non-parametric Kruskal-Wallis test with a Bonferroni correction for multiple comparisons, respectively. In patient cohort characteristics, normally distributed data were tested using one-way ANOVA. Chi square test of independence and Fisher's exact test were used to evaluate the association of clinical parameters with WHO categorization of patients based on severity.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint 6 Differences between proportions of primary and secondary infection across WHO categories were assessed using the Z test for proportions. Confidence intervals for odds ratio were determined using Baptista-Pike method. Spearman's correlation (two-tailed) analysis was performed to assess the positive or negative correlation between various cytokine secreting cell subsets. Receiver operating characteristic (ROC) curve analysis was performed to assess accuracy of proposed biomarker and 95% confidence intervals were calculated using Wilson/Brown method.

Multivariate binary logistic regression was performed to compare DF or DFWS/SD with worsened groups as the dependent variables. Independent variables for multivariate analysis were selected if they were significantly different in univariate analysis (two-tailed Mann-Whitney U test for nonparametric continuous data and Chi square test for categorical variables). Required assumptions such as dichotomous mutually exclusive dependent variable, two or more independent variables, linear relationship between each independent variable and odds ratio, absence of multicollinearity were all met. Even though three independent variables were significantly different two of them directly correlated with each other (IFN-γ + TNF-α + CD56 + CD3 + NKT cells and total IFNγ + CD56 + CD3 + NKT cells). Hence we used a combination of monofunctional IL-6 + granulocytes with either total IFN-γ + CD56 + CD3 + NKT cells or IFN-γ + TNF-α + CD56 + CD3 + NKT cells to generate logistic regression models. The latter was not significant and was not used.

Monofunctional IL-6 + granulocytes with total IFN-γ + CD56 + CD3 + NKT cells regression model was a good fit confirmed by the Hosmer and Lemeshow goodness of fit test. The estimated probabilities obtained from logistic regression model were used to plot composite ROC curves.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint

Of the 596 subjects with laboratory confirmed dengue who were included in the final data analyses, the mean age was 30.4±10.79 (mean±SD, range 17-69). 72% of enrolled patients were male and 28% were female (table S9) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint 8 S12B). CD19 + B cells also showed the presence of IL-10 + TNF-α + dual functional cells followed by IP-10 + TNF-α + and IP-10 + IFN-γ + . The triple positive cells from this subset included IFN-γ and TNF-α in combination with IL-10 or IP-10 ( fig. S12C ). Granulocytes and monocyte subsets had abundant IL-10 + TNF-α + cells ( fig. S12, E-H) . Granulocytes were the predominant secretor of multiple cytokines (IL-6, IL-10, IP-10 and TNF-α; table S1) in addition to multiple combinations of polyfunctional cells (fig S12H) . IL-6 and IL-10 were the least abundant cytokines secreted by all queried innate cell types against DENV.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint 11 CD14 + CD16classical monocytes (CM), CD14 + CD16 + intermediate monocytes (IM) and CD14 -CD16 + non-classical monocytes (NCM) based on CD16 vs. CD14. Live granulocytes were distinguished based on SSC-A scatter vs TLR-2 followed by those negative for live/dead dye.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint 13

of (A) IL-10 + TNF-α + granulocytes and (B) IFN-γ + TNF-α + CD56 + CD3 + NKT cells compared between dengue patients, febrile controls (FC) and healthy controls (HC). Each dot represents a patient sample. P value determined using Kruskal-Wallis test, followed by Bonferroni correction for multiple comparisons with median and IQR reported.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint 14

Two-dimensional representation identified five major CD56 + CD3 + NKT cell clusters (pink) visualized by t-SNE map. (B) Differential expression of CD56, CD3, IFN-γ, TNF-α, IP-10 and IL-10 in NKT cell clusters visualized by two-dimensional multicolored t-SNE maps.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint Levels of dengue NS1 compared between DF, DFWS and SD patients. P values were determined using Kruskal-Wallis test, followed by Bonferroni correction for multiple comparison between groups with median and IQR reported.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. .

of total (A) IFN-γ + CD56 + CD3 + NKT cells, (B) TNF-α + CD56 + CD3 + NKT cells, (C) IL-6 + granulocytes in total cohort and (D) IP-10 + CD56 + CD16 + NK cells in primary dengue cohort compared between normal (≤48 IU/L) and abnormal (>48 IU/L) levels of AST. Frequency of (E) monofunctional IP-10 + IL-10 -IFN-γ -TNF-α -CD19 + B cells in total cohort, (F) monofunctional IP-10 + IL-10 -IL-6 -TNF-αgranulocytes in total cohort, (G) total IFN-γ + CD56 + CD3 + NKT cells in secondary dengue and (H) total TNF-α + CD19 + B cells in total cohort compared between DF and . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint 21 DFWS/SD severity groups as well as between time window of hospital presentation during 1-3, 4-6 or 7-15 days post symptom onset (dpso). P values were determined using Mann-Whitney U test between normal and abnormal levels of AST or DF and DFWS/SD groups for any single time interval and Kruskal-Wallis test, followed by Bonferroni correction for multiple comparison of normal or abnormal AST / DF or DFWS/SD patients between the three time intervals. Medians with IQR are reported.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. Area under the curves (AUC) were determined by ROC curve analyses.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint Table S1 . Percentage producers and MFI of cytokines from innate cell subsets. % producers describe the percentage of patients in our cohort with detectable cytokine events from each cell subset; median cytokine secreting cells as a percent of parent and IQR are represented within Where the full study protocol can be accessed n/a 30

Sources of funding and other support; role of funders Main page 15 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint STARD 2015 AIM STARD stands for "Standards for Reporting Diagnostic accuracy studies". This list of items was developed to contribute to the completeness and transparency of reporting of diagnostic accuracy studies. Authors can use the list to write informative study reports. Editors and peer-reviewers can use it to evaluate whether the information has been included in manuscripts submitted for publication.

A diagnostic accuracy study evaluates the ability of one or more medical tests to correctly classify study participants as having a target condition. This can be a disease, a disease stage, response or benefit from therapy, or an event or condition in the future. A medical test can be an imaging procedure, a laboratory test, elements from history and physical examination, a combination of these, or any other method for collecting information about the current health status of a patient.

The test whose accuracy is evaluated is called index test. A study can evaluate the accuracy of one or more index tests. Evaluating the ability of a medical test to correctly classify patients is typically done by comparing the distribution of the index test results with those of the reference standard. The reference standard is the best available method for establishing the presence or absence of the target condition. An accuracy study can rely on one or more reference standards.

If test results are categorized as either positive or negative, the cross tabulation of the index test results against those of the reference standard can be used to estimate the sensitivity of the index test (the proportion of participants with the target condition who have a positive index test), and its specificity (the proportion without the target condition who have a negative index test). From this cross tabulation (sometimes referred to as the contingency or "2x2" table), several other accuracy statistics can be estimated, such as the positive and negative predictive values of the test. Confidence intervals around estimates of accuracy can then be calculated to quantify the statistical precision of the measurements.

If the index test results can take more than two values, categorization of test results as positive or negative requires a test positivity cut-off. When multiple such cut-offs can be defined, authors can report a receiver operating characteristic (ROC) curve which graphically represents the combination of sensitivity and specificity for each possible test positivity cut-off. The area under the ROC curve informs in a single numerical value about the overall diagnostic accuracy of the index test.

The intended use of a medical test can be diagnosis, screening, staging, monitoring, surveillance, prediction or prognosis. The clinical role of a test explains its position relative to existing tests in the clinical pathway. A replacement test, for example, replaces an existing test. A triage test is used before an existing test; an add-on test is used after an existing test.

Besides diagnostic accuracy, several other outcomes and statistics may be relevant in the evaluation of medical tests. Medical tests can also be used to classify patients for purposes other than diagnosis, such as staging or prognosis. The STARD list was not explicitly developed for these other outcomes, statistics, and study types, although most STARD items would still apply.

This STARD list was released in 2015. The 30 items were identified by an international expert group of methodologists, researchers, and editors. The guiding principle in the development of STARD was to select items that, when reported, would help readers to judge the potential for bias in the study, to appraise the applicability of the study findings and the validity of conclusions and recommendations. The list represents an update of the first version, which was published in 2003.

More information can be found on http://www.equator-network.org/reporting-guidelines/stard. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted October 15, 2020. . https://doi.org/10.1101/2020.10.14.20212001 doi: medRxiv preprint

and Notes

The global distribution and burden of dengue

A comparison of WHO guidelines issued in 1997 and 2009 for 15 dengue fever -single centre experience

Dengue: Guidelines for diagnosis, treatment, prevention and control

Dengue Antibody-Dependent Enhancement: Knowns and Unknowns

Antibody-Dependent Enhancement of Dengue Virus Growth in Human-Monocytes as a Risk Factor for Dengue Hemorrhagic-Fever

Inside-Out Control of Fc-Receptors