Could IL-6 predict the clinical severity of COVID-19?

Study design

In this retrospective, single-center study, data were collected from 115 patients who were admitted to the Antalya Education and Research Hospital between 11 March 2020 and 4 April 2020.

The study population herein comprised 60 patients who were confirmed to be positive for COVID-19 infection as the result of SARS-CoV-2 nucleic acid real-time polymerase chain reaction (RT-PCR) analysis using specimens that had been derived from nasal and oropharyngeal swabs. Additionally, 55 patients were confirmed to be positive for COVID-19 infection as the result of thorax tomography.

The diagnosis, as well as the treatment, of the COVID-19 patients were performed according to the COVID-19 (SARS-CoV-2 infection) Patient Treatment Guidelines issued by the Turkish Ministry of Health, as well as the interim guidance of the World Health Organization (WHO).

Based on the guidelines, the patients herein were categorized as given below:

1. Mild cases:

   1. Symptoms such as fever, myalgia or arthralgia, cough, or sore throat, without respiratory distress (resting respiratory rates <24 breaths/min, and oxygen saturation >95%).

   2. Chest X-ray and/or chest computed tomography (CT) scan are normal.

2. Moderate cases:

   1. Symptoms such as fever, myalgia or arthralgia, cough, or sore throat, with resting a respiratory rate that was <24 breaths/min and oxygen saturation >93%.

   2. Patients with signs of mild to moderate pneumonia on chest X-ray or CT scan (presence of radiological findings in less than 50% of the lung parenchyma; patchy shadowing, focal ground glass opacity, and interstitial abnormalities).

3. Severe cases:

   1. Symptoms such as fever, myalgia or arthralgia, cough, or sore throat, with a resting a respiratory rate that was ≥30 breaths/min as well as oxygen saturation that was ≤90%.

   2. Patients with bilateral diffuse pneumonia findings on chest X-ray or CT scan (presence of radiological findings in more than 50% of the lung parenchyma; patchy shadowing, focal ground glass opacity, and interstitial abnormalities).

Patient data, comprising their demographic characteristics, any existing underlying comorbidities, symptoms exhibited, as well as laboratory and physical test results, were taken from the hospital electronic medical records and recorded.

The laboratory tests conducted comprised C-reactive protein (CRP), baseline IL-6, neutrophils (NEU), white blood cell (WBC), platelet (PLT), and lymphocyte (LYM) counts, in addition to the erythrocyte sedimentation rate (ESR), neutrophil lymphocyte ratio (N/L), and D-dimer, lactate dehydrogenase (LDH), high-sensitivity troponin T (hs-cTnT), ferritin, and procalcitonin (PCT) levels on admission to the hospital. IL-6 levels during the period of hospitalization in moderate and severe cases were included in the study.

Measurement of the body temperatures of the patients was performed using an infrared thermometer, and the diagnosis of fever was determined as a temperature that was above 37 °C.

This study was granted approval by the Institutional Research Ethics Committee (Process No. 2020-7/17).

Laboratory diagnosis

Specimens derived from both pharyngeal and nasal swabs were collected from each patient and placed individually into collection tubes that contained virus preservation solution (Bio-Speedy NAT transfer tube, specimen collection flocked swap. Bioeksen R&D Technologies Ltd., İstanbul, Turkey). Nucleic acid extraction were obtained with Bio-Speedy-nucleic acid isolation kits according to the to the manufacturer’s instructions. Nucleic acid isolates were studied with real‐time RT‐PCR technique via using Qiagen Rotorgene real time PCR cycler (QIAGEN, Hil-den, Germany). Bio-Speedy SARS CoV-2 RT-qPCR kit was used targeting the SARS-CoV-2 specific N and Orf1ab gene region. Cycle threshold (CT) value less than 38 is interpreted as positive for SARS‐CoV‐2 RNA.

Blood samples obtained in the morning during the first 24 h of hospitalization of all COVID-19 patients were included in the study. The IL-6 test results performed during the hospitalization period were evaluated together with the baseline test results. From each patient, 2.0 mL of total venous blood was collected from the antecubital vein into evacuated tubes (Greiner Bio-One, Austria) that contained 0.109 M, 3.2% trisodium citrate to determine the D-dimer levels; 2.0 mL of total venous blood was collected from the antecubital vein into evacuated tubes (Isotherm Co. Ltd., Turkey) that contained K3EDTA to determine the WBC, NEU, LYM, and PLT counts, N/L, and ESR; and 5.0 mL venous blood was collected into tubes that contained a spray-dried clot activator and gel separator (Isotherm Co. Ltd.Turkey) to measure the IL-6, CRP, hs-cTnT, LDH, ferritin, and PCT levels.

Separation of the citrated plasma and serum samples from the cells was conducted via centrifugation for 10 min at 1653 × g Relative Centrifugal Force (RCF), and all of the parameters were measured recurrently.

IL-6 was measured using commercially available kits on a Cobas e411 Immunoassay Analyzer (Roche Diagnostics, Germany). Analytical performance data of IL-6 measured by the electrochemiluminescence immunoassay method were as follows: Measuring interval was 1.5–5000 pg/mL (defined by the lower detection limit and the maximum of the master curve). The reference range was up to 7 pg/mL IL-6 (95th percentile). Precision: The repeatability values for human serums at different levels were as follows; 17.3 pg/mL CV% 6.0, 117 pg/mL CV% 2.5 and 891 pg/mL CV% 2.6 Analytical specificity: IL-6 assay does not show any significant cross-reactivity with the IL-1α, IL- 1β, IL-2, IL-3, IL-4, IL-8 IFN-γ, TNF-α, tested with IL-6 concentrations of approximately 3 pg/mL and 4000 pg/mL (maximum tested concentration)

Measurement of the CRP and LDH levels was conducted using kits that were commercially available with a fully-automated AU 5800 analyzer (Beckman CoulterInc.,USA).

The WBC, NEU, LYM, and PLT counts, and N/L were determined using kits that were commercially available with an XN 1000 autoanalyzer (Sysmex Corp. Japan).

The ESR was measured using an undiluted blood sample with ethylenediaminetetraacetic acid evacuated tubes with a fully-automatic Vision-C analyzer (YHLO Biotech Co., China).

hs-cTnT, levels were conducted kits that were commercially available with a fully-automated Cobas e411 analyzer (Roche Diagnostics, Germany)

The D-dimer levels were measured using commercially available kits with a fully-automated ACL TOP 700 analyzer (Instrumentation Laboratory Co., USA).

Ferritin was measured using commercially available kits with an a fully-automated DXI 800 analyzer (Beckman CoulterInc., USA).

The PCT levels were determined using commercially available kits with a fully-automated Cobas 8000 analyzer (Roche Diagnostics, Germany)

Statistical analysis

Statistical analyses of the data obtained herein were conducted using IBM SPSS Statistics for Windows 20.0 (IBM Corp., Armonk, NY, USA). Analytical methods were used to determine whether or not the variables were normally distributed. Descriptive analyses were presented using medians and interquartile (IQR) for the non-normally distributed and ordinal variables. Since the IL-6 measurement was not normally distributed, the Mann-Whitney U and Kruskal-Wallis tests were used to compare IL-6 and laboratory results between the groups. Multivariate logistic regression analysis was used to determine independent risk factors associated with severe COVID-19. The variables with p<0.2 in the univariate analyses were further tested in the multivariate model. Odds ratio (OR) with corresponding 95% confidence intervals (95% CIs) was reported.

Demographic and clinical characteristics

The study population herein comprised total of 115 patients, including 46 females (40%) and 69 (60%) males, who had received a diagnosis of COVID-19 infection. Of these 115 patients, 24 were classified as mild, 52 as moderate, and 39 as severe. There were no significant differences in the male to female ratio between the three groups (p=0.566). The relationship between mild, moderate and severe clinical conditions and age group cases was analyzed by dividing them into three age groups (20–39, 40–59 and up to 60 years old). It has been determined that the clinical condition was more severe with increasing age (p=0.008).

It was determined that there was a significant difference in the hospitalized clinic to intensive care unit ratio among mild, moderate, and severe groups (p<0.001).

More than half of the patients (66.9%) had, at minimum, one comorbidity. It was determined that hypertension (31.3%) was the most common comorbidity, after which was diabetes (29.5%).

Among the comorbidities, the rate of asthma was determined to be significantly different in all of the clinical groups (p=0.010).

Cough (58.3%), shortness of breath (43.5%), and fever (79.72%) were the most common symptoms, while fatigue (16.5%), gastrointestinal symptoms (diarrhea, vomiting, etc., 10.4%) and sore throat (7.8%) were relatively less common. Significant differences were determined between the three study groups with regards to shortness of breath and myalgia (p=0.042 and p=0.003).

Table 1 presents the demographic and clinical characteristics of the patients in all of the groups.

Clinical laboratory data

60 patients included in the study were confirmed to be positive for COVID-19 infection as the result of SARS-CoV-2 nucleic acid real-time polymerase chain reaction (RT-PCR) analysis. 55 patients were cofirmed to be negative for COVID-19 as the result of SARS-CoV-2 RT-PCR. However, thoracic tomography findings of these patients were consistent with COVID-19.

Table 2 presents the comparisons obtained by performing Kruskal Wallis for different parameters among the different clinical situations of the patients on admission. In this study, which covered the first months of the pandemic, there was no standardization regarding the test request yet. Test requests were made according to the symptoms and clinical findings of the patients, and the number of patients in each parameter was different. Age (p<0.001) and IL-6 test results (p=0.011) differed significantly according to the clinical classification of the disease. The mean age of the patients was 56.4 years. The median age for the mild group was 46.04 years (range 21–72 years), while it was 56.42 years for the moderate group (range 21–85 years), and 62.92 years (range 24–93) for the severe group (p<0.001). Post-hoc test was performed among age groups (mild<moderate, severe). A significant difference was determined between the mild and moderate (p=0.035) as well as the mild and severe (p<0.001) groups; however, no such difference was seen between the moderate and severe (p=0.162) groups. Post-hoc test was also performed on IL-6 results for all three groups (mild<severe). It was determined that a statistically significant difference existed between the mild and severe (p=0.013) groups; however, no such difference was seen between the mild and moderate (p=0.738) or the moderate and severe (p=0.093) groups.

Baseline IL-6 levels obtained by comparing the Mann-Whitney U test of the paired groups are shown in Figure 1. The IL-6 levels were significantly higher in the patients in the severe group than in those in the mild group (p=0.04). Furthermore, the IL-6 levels were increased significantly on admission in the patients in the severe group when compared with those in the moderate group (p=0.043, Figure 1). However, there was no difference in the baseline IL-6 levels between the patients in the mild group and those in the moderate group (p=0.223, Figure 1).

Figure 1:

The comparisons obtained by performing Mann-Whitney U test for IL-6 results of the paired groups.

Table 3 presents the result of multivariate logistic regression analysis performed to identify the independent factors associated with severe COVID-19. Increasing age (OR: 1.031; 95% CI: 1.004–1.060; p=0.026) and increasing IL-6 (OR: 2.258; 95% CI: 1.222–4.175; p=0.009) were identified as significant risk factors for severe COVID-19.

Analysis by receiver operating characteristic

In order to conduct an evaluation of the prognostic value that IL-6 exhibits in patients diagnosed with COVID-19, a receiver operating characteristic (ROC) curve was drawn for the patients in the mild and severe groups. The area under the ROC curve (AUC) of IL‐6, which was used for the prediction of the COVID‐19 severity, was 0.864 (95% CI 0.765–0.963 p<0.001), and was found to more accurately predict if COVID‐19 was complicated with severe pneumonia. Moreover, calculation of the specificity, sensitivity, and predictive positive and negative IL-6 values was performed to determine the optimal threshold value, which was found to be 28 pg/mL (Figure 2).

Figure 2:

Receiver operating characteristic (ROC) curve showing the predictive power of IL-6 for prediciting severe COVID-19.

The available longitudinal IL-6 measurement data of the patients in the subgroups were analyzed to conduct an investigation of their serial changes with regards to the severity of their COVID-19 infection. As a result, it was determined that only measurement of the baseline IL-6 levels was performed in patients who had mild COVID-19 infection. The levels of IL-6 were taken at five time intervals and presented. Day 0 was the first time interval, which comprised the baseline concentrations of the patients, taken within the first 24 h after admission. Measurements were conducted every other day for patients who were diagnosed with moderate and severe COVID-19 infection. When a comparison was conducted between the moderate and severe groups, it was determined that the patients who were in the severe group exhibited significantly increased serum concentrations of IL-6 during the course of hospitalization (Figure 3) The data presented in Figure 3 were determined by using the mean values of longitudinal IL-6 results.The IL-6 levels of five patients who did not survive COVID-19 were examined at most three times. The baseline IL-6 levels of these five expired patients were 59.48, 58.98, 115.2, 1408, and 20.95 pg/mL, respectively. All suspected or confirmed cases with COVID-19 were received hydroxychloroquine (or plus azithromycin/oseltamivir) on admission, according to the national protocol of Ministry of Health. As from 01 April 2020, favipiravir was introduced, favipiravir was started to be used for severe cases. Tocilizumab was not used during study period.

Figure 3:

Longitudinal changes of serum IL-6 levels in moderate and severe COVID-19 patients.