TARGET is registered in the Brazilian Clinical Trials database with registry code RBR-62zdkk and constitutes a multicenter, prospective cohort study of consecutive COVID-19 cases, consisting of participants recruited from secondary and tertiary health care facilities in Brazil, namely, the University Hospital of Brasília (University of Brasília [UnB]) and the public regional hospitals of Asa Norte, Gama, Santa Maria, and Taguatinga (Brasília, Distrito Federal). Clinical procedures will be conducted according to a standard protocol approved by local institutional review boards under accession number CAAE 30846920.7.0000.0008, with participants included after voluntary signing of an informed consent form. The study protocol follows the recommendations of the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) statement for observational studies [16].

Inclusion criteria will include the following: laboratory confirmation of COVID-19 by reverse transcription polymerase chain reaction (RT-PCR; KIT Molecular SARS-CoV2 [E/RP]; Bio-Manguinhos), a signed informed consent form, and older than 18 years. The exclusion criteria are as follows: patients with inconclusive, unavailable, or solely serological laboratory diagnosis of COVID-19; a dementia condition (Alzheimer, Parkinson, vascular frontotemporal, Lewy body, or others); schizophrenia or schizoaffective disorders with psychotic characteristics; and patients with previously known congenital hemorrhagic diseases or thrombophilia or use of anticoagulants (for coagulation tests; Figure 1).

The clinical evaluation aims to identify the signs and symptoms associated with COVID-19 and monitor the status (mild/moderate or severe/critical) of the disease. Clinical assessments include identification of comorbidities, COVID-19–associated symptoms and treatment, use of medications in general, clinical chemistry findings, and life support regimen care. Severe cases are defined for the purposes of the study by dyspnea (respiratory rate >30 breaths per minute) coupled with any of the following criteria: pulsed wave oxygen saturation <93% at rest, PaO₂/FiO₂ ≤300 mmHg, respiratory failure requiring mechanical ventilation, multiple organ failure, shock, or admission to the intensive care unit. Patients who do not meet the criteria for severe/critical forms will be considered mild/moderate cases. No asymptomatic patients will be included (Textbox 1 and Table 1).

The TARGET protocol comprises two arms, referred to as the timeline (T)-branch and the days of hospitalization (D)-branch. For the purposes of this study, the T-branch refers to patients with mild/moderate forms that did not require hospitalization. The D-branch refers to patients with severe/critical forms who required hospital admission. Patients on the T-branch will be analyzed on days 0, 9, and 18 (T0, T9, and T18), while patients on the D-branch will be analyzed on days 0, 3, 9, and 18 (D0, D3, D9, and D18; Figure 1).

The TARGET protocol encompasses the collection of blood samples in four different types of containers. For these purposes, whole peripheral blood samples will be collected in vacuum tubes without an anticoagulant (10 mL) to obtain serum samples for metabolomic and biochemical analysis, in sodium citrate (5 mL) for coagulation assays, and in heparin (10 mL) to obtain plasma samples and to isolate polymorphonuclear and peripheral blood mononuclear cells (PBMC). Samples will be collected at distinct time points as follows: day 0, at inclusion by confirmation of COVID-19; day 3, within 2-4 days after inclusion; day 9, within 7-10 days after inclusion; and day 18, within 15-20 days after inclusion (Figure 2).

The whole blood samples will be processed to obtain serum or plasma samples and PBMC batches. Tubes will be centrifuged at 1900 g for 10 minutes at 22 °C to obtain serum or heparinized plasma and processed immediately within-facility for routine PBMC isolation and cryopreservation [17].

Plasma samples obtained in ethylenediamine tetraacetic acid will be immediately analyzed at local clinical laboratory facilities. Biochemical analysis will be performed with a Rocas Cobas E411 system using compatible reagents to assess the levels of ferritin, troponin T, creatine kinase, creatinine, aspartate aminotransferase, and alanine aminotransferase. All procedures will be performed according to the manufacturer’s recommendations. Cortisol and prolactin in serum samples will be quantified at the Sabin Medicina Diagnóstica (Brasília, Brazil) using the chemiluminescent method (Siemens, Advia Centaur). Hematological analyses will be performed using whole blood obtained in sodium citrate tubes in an ABBOTT Cell Dyn 3700 analyzer or equivalent with samples from days 0, 9, and 18.

The coagulation profile will be assessed at the DASA Laboratory (Brasília, Brazil) by thromboelastometry tests carried out within a 3-hour interval after blood collection [13,25] as follows: extrinsic system assay (EXTEM), intrinsic system assay (INTEM), fibrinogen assay (FIBTEM), and nonactivated coagulation assay (NATEM; ROTEM, Werfen, Barcelona, Spain; Figure 3).

Serum biomarkers will be assessed using the Luminex Bio-Plex Pro platform set for 27 human immune mediators (Bio-Rad Laboratories, California) for the simultaneous assessment of the following analytes: chemokines (CXCL8, CCL11, CCL3, CCL4, CCL2, CCL5, and CXCL10), proinflammatory cytokines (IL-1β, IL-6, TNF-α, IL-12p70, IFN-γ, IL-17A, and IL-15) and regulatory cytokines (IL-1Ra, IL-4, IL- 5, IL-9, IL-10, and IL-13), and cell growth factors (IL-2, IL-7, FGF-basic, PDGF, VEG, G-CSF, and GM-CSF). All procedures will follow the manufacturer’s recommendations, with levels obtained at days 0, 3, and 9.

The expression levels of immune mediators will be assessed by TaqMan quantitative RT-PCR (qRT-PCR) assays (Applied Biosystems, Foster City, CA) for each transcript, with retro-transcription set up with 1 ng/μL of RNA from PBMCs and from neutrophils. The reactions will be carried out on a QuantiStudio 3 Thermal Cycler (Applied Biosystems, CA) in standard mode [26]. Alternatively, qRT-PCR experiments for selected genes will be based on the SYBR Green (family of dyes for molecular biology) detection methodology using the SYBR Green PCR Master Mix and universal cycling conditions on the ABI Prism 7000 Sequence Detection System. Expression levels will be assessed on days 3 and 9.

Immunophenotypic/functional profiles of circulating leukocytes will be assessed upon short-term culture in vitro. The cell suspension will be stained with distinct panels of monoclonal antibodies comprising CD3, CD4, CD8, HLA-DR, CD25, CD19, CD5, CD27, CD38, PD-1, CD28, CD14, CD16, and CD56 for cell surface analysis, and IL-1β, IL-6, TNF-α, IFN-α, IL-5-PE, and IL-10 for evaluation of intracytoplasmic functional status. Cells will be run on a FORTESSA flow cytometer (BD Bioscience, CA). An average of 100,000 cells will be analyzed per sample. The phenotypic/functional parameters will be assessed using FlowJo software (BD, New Jersey). Leukocyte phenotyping will be obtained at days 0, 3, and 9.

The immunoglobulin profile will be assessed by multiplexed PCR using cryopreserved PBMCs [19,27]. Sequencing libraries will be produced and sequenced using the Illumina MiSeq 2x300 bp platform. For each sample, 1 million reads corresponding to IgG and IgA will be produced. The reads will be processed using pRESTO [28], Ig heavy chain genes will be annotated using the MIXCR platform [29], and clonotypes will be determined as described in [30]. The immunoglobulin profile will be assessed in samples obtained on day 3.

Neutrophils will be obtained from healthy donors and from patients at two different times, T9/T18 and D9/D18. Neutrophils isolated from each donor will be divided into two aliquots and incubated in buffer (control group) or phorbol 12-myristate 13-acetate (PMA; activated group). Cells from each group will be submitted to the functional and molecular analyses described in the following sections.

Cell isolation will be performed from peripheral venous blood by centrifugation in Percoll gradients, as previously described [27]. The remaining erythrocytes will be removed by hypotonic lysis. Samples containing >95% neutrophils and >98% viable neutrophils will be prepared in 50% autologous plasma and divided into aliquots for control or activation in PMA [28].

The isolated neutrophils, after incubation in each condition, will be tested for reactive oxygen species production as described by Tahir et al [29].

At the final 20 minutes of incubation in each condition, an aliquot of neutrophils will be removed and incubated with SYTOX green (high-affinity nucleic acid stain; 10 µmol/L) and 4′,6-diamidino-2-phenylindole (blue-fluorescent DNA stain; 1 µg/mL). Fluorescence readings will be performed both by fluorescence microscopy and on a spectrophotometer [30].

Neutrophils from each condition will be applied to the upper chamber of a xCELLigence Real-Time Cell Analysis instrument and dynamically measured regarding the migration toward the chemoattractant N-Formylmethionyl-leucyl-phenylalanine (potent polymorphonuclear leukocyte chemotactic factor) [31].

Neutrophils obtained in all experimental conditions will be subjected to cell lysis by ultrasound cavitation and then trypsin digested. Eluted peptides will be subjected to capillary chromatography in a two-column online vented system. The fractions will be eluted directly into the ionization chamber of an Orbitrap Elite mass spectrometer to be analyzed in data-dependent acquisition mode. The 20 most intense ions will be selected for high-energy collision dissociation fragmentation, and the fragments will be analyzed in the ion-trap detector. Mass spectra will be analyzed for protein quantification and identification using the Progenesis and Peaks software packages, as well as in-house developed R scripts for statistical analysis [32].

We will aim to analyze each serum sample collected from mild (T-branch group) and severe (D-branch group) case patients and healthy volunteers at each time point in the same manner (Figures 1 and 2). Untargeted metabolomics will be performed using a Bruker UPLC-PDA-QqTOF (ultra-performance liquid chromatography photodiode array quadrupole-quadrupole-time-of-flight) compact (UPLC-MS/MS). The obtained data will be used to build molecular networking using the Global Natural Products Social Molecular Networking platform, which gives a better overview of the metabolic profile and group compound classes in clusters. For statistical validation of the metabolite differences between the groups, the data will be processed in MZmine and analyzed by means of chemometric tools, as provided by the MetaboAnalyst platform. Based on this broad chemical outline, specific metabolites will be quantitatively analyzed using targeted metabolomics using a CLAE Quadrupole/Quadrupole. This approach will be used to validate and quantify specific components of interest, such as metabolites involved in the glycolytic and hexosamine pathways, and tricarboxylic acid cycle, among other possible biomarkers.

The sample size calculation was based on the assumption that 20% of patients who are critically ill have elevated TNF-α levels compared to 5% of patients who are noncritically ill [8]. TNF levels will be normalized according to the study population. Based on a bilateral 95% significance level, a power of 80%, and a case-control ratio of 1, a sample size of 200 patients can be predicted.

Clinical and demographic data between groups of patients (severe vs nonsevere) as well as treatment profiles and differences in successive time points will be compared. To compare mean values of soluble biomarkers between groups, the t student test (or analysis of variance with post hoc test for multiple groups) or the nonparametric Mann-Whitney (or Kruskal-Wallis for multiple groups) test will be used as appropriate, according to the observed sample distribution features. The SAS 9.3 software (SAS Institute Inc) will be used with a significance level set at .05. The main binary outcomes will be analyzed by the estimation of relative risks using a log-binomial generalized model or by the area under the curve using trapezoidal methods. All personnel conducting laboratory procedures will be blinded regarding the patient status (severity of COVID-19). Receiver operating characteristic curves will determine cutoff points at which individual or combined scores of the molecular and cellular biomarkers become informative of the progression of COVID-19 cases, according to optimal sensitivity and specificity to identify the signature that best characterize each clinical status.