Tocilizumab Failed to Reduce Mortality in Severe COVID-19 Patients: Results From a Randomized Controlled Clinical Trial

is associated with an increased level of blood IL-6. It is suggested that adjusting the IL-6 level or its effects may diminish the inammation of the airway and alter the course of the disease. We performed a controlled, randomized, double-blind clinical trial including severe hospitalized COVID-19 patients in Iran. The patients were randomly distributed by block randomization to take usual care plus IL-6 inhibitor (tocilizumab 8mg/kg, 1 or 2 dosages) or usual care alone (ratio 1:1). The endpoint was dened by the clinical improvement and discharge or death. Main and safety analysis was performed in the overall population. We enrolled 40 eligible patients (20 patients in each group) from July 10, till October 10, 2020, in this study. After randomization, one patient in the usual care arm and three patients in the tocilizumab arm refused to participate in the study and were eliminated from the study and overall analysis. The mean age of participants was 59.62 ± 15.80 and 63.52 ± 12.83 years old in Tocilizumab (8 women and 9 men) and standard care (9 women and 10 men) groups respectively (P-value= 0.426). The number of recovered patients did not signicantly differ between tocilizumab and usual care groups (70.6% (n=12) vs 78.9% (n=15) respectively) (P-value=0.563). Also, the survival time was not signicantly varied between participants in the tocilizumab intervention and usual care group [Log rank test: P=0.973; Hazard ratio: 1.25; 95% C I: 0. 249-4.209]. The results not time. Improvement and discharge were determined by the patients’ clinical, general, and laboratory conditions and were determined by consciousness, ameliorated dyspnea, stopped fever for 3 days, SO2 upper than 93%, normal range of urinary output, tolerated oral regimen (PO), blood pressure more than 10 millimeters of mercury (mm Hg), a respiratory rate less than 20 breaths per minute, heart rate less than 90 beats per minutes, and reduced CRP amount and no any other side effects. All data were assessed during the study were registered on case report forms (CRFs) and the Excel database. of on of level the reduction of neutrophil percent in patients, we did not nd any effects of tocilizumab on improvement of clinical and laboratory characteristics of patients. Same as our results, previously the effect of tocilizumab on the improvement of anemia and reduction of the neutrophil count has been shown in rheumatoid arthritis (RA) patients 21). Our results in this randomized, controlled, double-blind clinical trial showed no signicant effect for tocilizumab intervention on the treatment of non-intubated, severe COVID-19 patients who had elevated IL-6 levels. Tocilizumab treatment did not signicantly change recovery rate, survival time, and time to event in affected individuals. Regarding the positive effects of corticosteroids administration on COVID-19 patients(29, 30), it seems that using systemic immunosuppression or a combination of various therapeutic agents might be more ecient than blocking a specic cytokine pathway.


Abstract Background
Severe COVID-19 is associated with an increased level of blood IL-6. It is suggested that adjusting the IL-6 level or its effects may diminish the in ammation of the airway and alter the course of the disease.

Methods
We performed a controlled, randomized, double-blind clinical trial including severe hospitalized COVID-19 patients in Iran. The patients were randomly distributed by block randomization to take usual care plus IL-6 inhibitor (tocilizumab 8mg/kg, 1 or 2 dosages) or usual care alone (ratio 1:1). The endpoint was de ned by the clinical improvement and discharge or death. Main and safety analysis was performed in the overall population.

Results
We enrolled 40 eligible patients (20 patients in each group) from July 10, till October 10, 2020, in this study. After randomization, one patient in the usual care arm and three patients in the tocilizumab arm refused to participate in the study and were eliminated from the study and overall analysis. The mean age of participants was 59.62 ± 15.80 and 63.52 ± 12.83 years old in Tocilizumab (8 women and 9 men) and standard care (9 women and 10 men) groups respectively (P-value= 0.426). The number of recovered patients did not signi cantly differ between tocilizumab and usual care groups (70.6% (n=12) vs 78.9% (n=15) respectively) (P-value=0.563). Also, the survival time was not signi cantly varied between participants in the tocilizumab intervention and usual care group [Log rank test: P=0.973; Hazard ratio: 1.25; 95% C I: 0. 249-4.209].

Conclusion
The study results recommended that tocilizumab could not be a bene cial agent for treating severe cases of COVID-19 patients and have not signi cantly affected patients' clinical improvement.

Background
Severe acute respiratory syndrome coronavirus (SARS-CoV)-2 that generates a pandemic of coronavirus disease 2019 (COVID-19) viral infection, has threatened human health and become a global concern since its emergence in late 2019 (1). COVID-19 presentations may range from light upper respiratory involvements to severe pneumonia that is characterized by immunological abnormalities, hyper-in ammation, and cytokine storm (2)(3)(4). Cytokine storm or excessive release of in ammatory cytokines leads to lethal organ damages, lung brosis, acute respiratory distress syndrome (ARDS), and even patients' death (5).
It has been demonstrated that the level of serum interleukin (IL)-10, Il-8, IL-6, and tumor necrosis factor (TNF) was remarkably increased in severe or deceased patients compared to moderate or recovered cases (1,6). IL-6 is a pro-in ammatory and pro-brotic cytokine that produce by immune cells including monocytes and lymphocytes as well as broblasts and bronchial epithelial cells and participates in the pathogenesis of lung brosis (7)(8)(9). It is found that IL-6 serum level in critically ill patients is almost 10 times more than in moderate cases (10). The elevated level of serum IL-6 is positively correlated with the SARS-CoV-2 RNAaemia (10), need for intubation (11), and disease severity stages of patients (12) and can be used as prognostic and diagnostic markers for COVID-19 severity. Therefore, it is suggested that administration of IL-6 signaling inhibitors might have e cient effects on the resolution of cytokine storm and treatment of individuals affected with COVID-19 who are in the pulmonary stage. Tocilizumab is a monoclonal antibody to the IL-6 receptor and is mainly used for the treatment of rheumatic disease (13). There are contradictory results regarding tocilizumab e cacy on COVID-19 patients (14)(15)(16)(17). Therefore, in this randomized controlled trial (RCT), we investigated the effects of tocilizumab administration on the recovery, oxygen therapy, clinical and laboratory features of COVID-19 inpatients.
Patients And Methods

Trail design
This research is a double-blind, randomized, two-arm parallel, phase-2, controlled trial that was performed from July 10 till October 10, 2020. 40 cases from Imam Khomeini and Shariati Hospital, Tehran University of Medical Sciences (TUMS) were included in this research. The trial protocol has been designed and mediated in adherence to the Consolidated Standards of Reporting Trials (CONSORT) guideline (18) and was approved by the Iranian Registry of Clinical Trials (ID: IRCT20081027001411N4). on 9 July 2020. According to the contradictory results about the effect of tocilizumab on COVID-19 patients, the lowest sample size with a 10% drop-out rate was estimated 40 patients according to the effect size of tocilizumab on the pulmonary disease (19). The trial was conducted under the Declaration of Helsinki guidelines and was agreed by the Ethics Committee of the Tehran University of Medical Science (Approval Number: IR.TUMS.VCR.REC1399.290).

Participants
The con rmation of COVID-19 infection in patients was con rmed by the following speci cations: 1. Con rmation of SARS-CoV-2 in nasopharyngeal swab via polymerase chain reaction (PCR) and 2. Atypical computed tomography (CT) features (subpleural, bilateral, peripheral ground-glass opacity). 18 years or older patients were eligible for enrollment in the trial if they complied with the following conditions: 1.
Con rmed COVID-19 infection; 2. Elevated C-reactive protein (CRP higher than 10mg/L) or IL-6 (higher than 18 pg/ml) or lymphopenia (lymphocyte count under 1100/ MCL); 3. At the pulmonary stage of the disease with blood oxygen saturation <93% or respiratory rate (RR) higher than 24; 4. Not connecting to the mechanical ventilator; 5. Not responding to standard COVID-19 treatment. All participants provided informed consent before enrollment. Patients were not eligible for enrollment in the trial if they complied with the following conditions: 1. Were allergic or intolerant to any therapeutic factors used in this study; 2. With positive pro-calcitonin (PCT) and had an active infection (including latent or active tuberculosis (TB) infection); 3. Had a history of receiving immunosuppressive drugs and corticosteroids; 4. With a history of active malignancies.

Randomization and treatment
Eligible patients (one or two days after hospitalization) were randomly distributed to the control and intervention groups by block randomization (1:1 ratio). One patient in the standard care arm and three patients in the tocilizumab arm refused to participate and were excluded from the study and overall analysis before baseline measurements. Patients allocated to receive usual care alone or usual care plus 8mg/kg tocilizumab, (Actemra, Roche) (if the patient's conditions were not stable, 2 doses by 12 hours were administrated, maximum dose: 800 mg). All patients received usual care for the disease based on the Iranian protocol for diagnosis and treatment of COVID-19. The major standard medical cares that were received by COVID-19 patients are summarized in Table 1. In this study, patients, investigators, and outcome assessors did not inform which group received an intervention. Besides, a placebo was not used in the control group.

Outcome
The clinical, demographic, and laboratory data of the patients were recorded before their enrollment. Participants were followed through day 0 to 5, improvement, discharge, or death. The primary outcome was improvement and discharge or death after administration of intervention whichever came rst. The 28 days' mortality rate, time to improvement, and time to the event were also evaluated in each group.
Clinical characteristics of the patients such as RR, heart rate (HR), blood pressure, fever, and general symptoms including gastrointestinal involvement (GI), thorax pain, dyspnea, cough, weakness, and myalgia were assessed at enrollment and after 3-5 days of starting the treatment.
Besides, the degree of oxygen saturation (SO2) and the necessity and type of oxygen supports including 1. Nasal cannula, 2. Simple oxygen mask, 3. Reservoir mask, 4. Noninvasive ventilation (NIV) and 5. Invasive ventilation were assessed at enrollment and after 3-5 days of starting the intervention. CT scan ndings including pulmonary involvement percent and radiological properties were also evaluated at baseline and 6 weeks after treatment (only in participants who consented). Laboratory examinations to assess the level of complete blood count (CBC), alanine aminotransferase (ALT) and aspartate aminotransferase (AST) liver enzymes, CRP, erythrocyte sedimentation rate (ESR), fasting blood sugar (FBS), vein blood gas (VBG) (containing PH, the partial pressure of carbon dioxide (PCO2), and bicarbonate (HCO3)), IL-6, D-dimer, troponin, creatine phosphokinase (CPK) and ferritin were done at baseline and by 3-5 days of starting the treatment and also at hospital discharge time.
Improvement and discharge were determined by the patients' clinical, general, and laboratory conditions and were determined by consciousness, ameliorated dyspnea, stopped fever for 3 days, SO2 upper than 93%, normal range of urinary output, tolerated oral regimen (PO), blood pressure more than 10 millimeters of mercury (mm Hg), a respiratory rate less than 20 breaths per minute, heart rate less than 90 beats per minutes, and reduced CRP amount and no any other side effects. All data were assessed during the study were registered on case report forms (CRFs) and the Excel database.

Side events
All unfavorable adverse effects experienced by participants during this trial, whether or not related to tocilizumab intervention, were assessed and stored.

Statistical analysis
All continuous variables were reported as the mean ± standard deviation (SD) and categorical variables are expressed as N (%). The Kolmogorov-Smirnov (K-S) normality test was used for continuous variables. Repeated measures ANOVA was performed for comparison between both groups over time. Besides, Mann Whitney or Student's t-tests was applied to assess the statistical contrasts between two arms in each time point. Moreover, Wilcoxon signed-rank or paired t-tests was performed to test for statistical comparison between two-time points in each of the studied intervention arms. Two-sided Chi-square/Fisher's exact tests were applied to test the associations between two intervention arms and the categorical variables. To analyze and draw the time-to-death curve between two intervention arms, Kaplan-Meier curve and the logrank test were applied. Using the Benjamini-Hochberg correction method for multiple comparisons, the false discovery rate was corrected. Statistical signi cance was de ned at p<005 and all statistical analysis was done by STATA software (Versions 11.2).

Patients
Between July 10 to October 10, 2020, 40 eligible patients were randomly assigned to receive usual care plus tocilizumab (n=20), or usual care alone (n=20). One patient in the standard care arm and 3 patients in the tocilizumab arm refused to participate in the trial and were excluded from the study and overall analysis. The results for the outcomes were not signi cantly distinct from the results of intention to treat (ITT), according to the analysis per protocol. The patients' enrollment, randomization, and assignment are summarized in Figure 1. The mean age of participants was 61.74 ± 14.19 years old (19 (52.8%) men and 17 (47.2%) women). There are no notable differences between the two arms regarding their sex and age (P=0.99 and P=0.426 respectively). In the tocilizumab group, 12 patients (70.6%) received one dose and 5 patients (29.4%) received two doses of tocilizumab. The coexisting conditions and clinical characteristics of patients at enrolment were shown in Table 1. Nine patients (25%) had RR upper than 24 breaths per minute and 14 patients (38.9%) had an HR of more than 100 beats per minute. Dyspnea, weakness, and cough were the most prevalent symptoms in patients. Except for RR which was higher in patients of the tocilizumab arm, there were no substantial between-group differences in clinical features, comorbidity conditions, and administrated standard medical care at enrollment. Besides, except for ALT which was signi cantly elevated in the tocilizumab group, and ESR which was notably elevated in the usual care group, we did not observe signi cant between-group differences regarding laboratory features at enrolment.
The average blood SO2 level was 84.6 ± 8.92 at baseline. All of the patients had peripheral ground-glass opacity at enrollment and the majority of them had 10-30% (38.8%) and 30-50% (27.7%) respiratory involvements respectively. All participants were receiving oxygen supports at baseline and the majority of them (80%) were received nasal cannula. Pulmonary involvement level and CT scan ndings of patients in each group at baseline were summarized in Table 2.

Primary outcome
Mortality at 28 days was not considerably different between patients in the tocilizumab arm and standard care, with death reported in 5 (29.4%) and 4 (21.1%) patients respectively (Table 3) We also did not observe a major difference in the time to event (improvement, or death) between participants assigned to the tocilizumab and standard care arms (median, 12.88 ± 9.05 days vs 12.05 ± 8.92 days; P=0.784). Also, the time to improvement was not considerably different between the two arms (median, 9.42 ± 3.55 days in the tocilizumab group vs 9.73 ± 6.05 days in the usual care group; P=0.874) ( Table 3).
The number of patients in each oxygen support group at baseline and after 3-5 days of starting the treatment and the primary outcomes of them in both standard care and tocilizumab arms were shown in Supplementary Figure 1. We did not observe notable differences in the incidence of death according to the type of oxygen therapy in patients. Of the dead patients, 55.5% (2 patients in the control arm and 3 patients in the tocilizumab arm) had pulmonary involvement percent more than 70%, and 22% had pulmonary involvement percent between 50 to 70% (one patient in each group).

Secondary outcome
The level of participants' blood SO2 was substantially ameliorated by 3-5 days of treatment in both groups (P=0.001 and P=0.002 in the tocilizumab and standard care respectively). Although the body temperature was reduced in both groups after 3-5 days of treatment, the reduction was only signi cant in patients in the usual care arm. We did not observe signi cant differences in the blood pressure HR and RR before and after 3-5 days of treatment in patients of both arms. After 3-5 days of starting the treatment, the number of patients suffered from dyspnea was signi cantly reduced in both the tocilizumab and standard care treatment group (P=0.012 and P=0.031 respectively). However, we did not observe signi cant bene cial effects of tocilizumab on the reduction of GI symptoms, myalgia, and cough in patients, and the symptoms only were signi cantly improved in the usual care arm. Clinical features of participants before and after treatment were summarized in Table 4.  Figure S1).

Laboratory ndings
The hemoglobin level was notably reduced in the control group of patients by 3-5 days of starting the treatment and also at discharge time (P=0.002 and P=0.025 respectively). Besides, the hematocrit percentage was diminished in patients who received standard care treatment at discharge time. While the level of hemoglobin and hematocrit percentage did not change in patients of the tocilizumab group. Tocilizumab also reduced neutrophil percentage in patients, nevertheless, we did not detect a signi cant reduction in the percentage of neutrophil in patients who received standard care. Besides, although tocilizumab reduced IL-6 level, the reduction was not signi cant. Except for hemoglobin and hematocrit levels, we did not detect any signi cant differences in the level of laboratory characteristics of patients during treatment in either group (Table 5).

Safety
Although we did not see serious adverse events in the usual care arm, a total of three patients (17.6%) in the tocilizumab arm showed serious side events between initiation of intervention and the end of the trial. Shock and infection developed in three and one patient in the tocilizumab arm respectively. Of these four adverse events, 3 were considered to be related or possibly related and one was considered to be unrelated to intervention by the site investigators. All serious adverse events are shown in Supplementary Table 1.

Discussion
It has been found that the level of serum IL-6 is elevated in critically ill COVID-19 patients and it is correlated with the severity of illness in patients (1,6,12). It is suggested that administration of IL-6 receptor inhibitor might prevent the progression of cytokine storm in COVID-19 patients who are in the pulmonary stage. Therefore, we aimed to investigate the effects of tocilizumab administration as an IL-6 receptor blockage, on the treatment and clinical and laboratory characteristics of hospitalized COVID-19 patients. We did not nd bene cial effects for the administration of tocilizumab in the treatment of severe COVID-19 patients in our RCT. Although tocilizumab administration improved the oxygen support status of more patients in the intervention arm compared to the control arm by 3-5 days of treatment, it could not improve the survival time, reduce the mortality rate and decrease the time to discharge in the intervention group. Except for the signi cant effects of tocilizumab treatment on the maintenance of hemoglobin level and the reduction of neutrophil percent in patients, we did not nd any therapeutic effects of tocilizumab on the improvement of clinical and laboratory characteristics of patients. Same as our results, previously the effect of tocilizumab on the improvement of anemia and reduction of the neutrophil count has been shown in rheumatoid arthritis (RA) patients (20,21).

Some observational and cohort studies investigated the effect of tocilizumab on COVID-19 patients and reported contradictory results regarding
its treatment e cacy (14)(15)(16)(17). In a case-control study by Marte et al., on severe to critical COVID-19 patients who received tocilizumab, no difference in mortality rate has been found in all studied patients. However, they found that the treatment can reduce the mortality rate in nonintubated patients (14). PharmD et al., in an observational study, reported limited improvement effects for tocilizumab. In their study, from patients who received tocilizumab, 42% were expired and 49% were improved (17).
In contrast to our results, Rossotti et al. found tocilizumab as potentially effective care for COVID-19 patients which was associated with improved overall survival in patients compared to controls.(15) Moreover, results from an observational cohort research on patients with severe COVID-19 pneumonia showed that intravenous or subcutaneous tocilizumab may reduce the risk of mortality and the risk of intubation in patients (22). Xu et al. also prospectively observed tocilizumab effects on treating COVID-19 patients. They found that tocilizumab improved clinical outcomes including fever, oxygen intake, lymphocytes percentage, and CRP level immediately in patients (23). Besides, in a Cohort study by Somers et al., tocilizumab administration was associated with reduced mortality rate and increased reinfection occurrence in intubated COVID-19 patients (24). Moreover, according to the observational study by Jordan et al., tocilizumab might offer bene cial effects in reducing COVID-19 associated mortality, in ammation, and oxygen supports in patients (25). Our study also has some limitations and weaknesses such as the possible existed bias, lack of placebo, and low sample size. Therefore, to determine the exact therapeutic effect of tocilizumab on COVID-19 patients, more clinical trials on COVID-19 patients on different disease stages are needed.

Conclusions
Our results in this randomized, controlled, double-blind clinical trial showed no signi cant effect for tocilizumab intervention on the treatment of non-intubated, severe COVID-19 patients who had elevated IL-6 levels. Tocilizumab treatment did not signi cantly change recovery rate, survival time, and time to event in affected individuals. Regarding the positive effects of corticosteroids administration on COVID-19 patients(29, 30), it seems that using systemic immunosuppression or a combination of various therapeutic agents might be more e cient than blocking a speci c cytokine pathway.

Consent to participate
The written informed consent was signed by all patients before enrolling in the study.

Availability of data and materials
Data are available upon reasonable request.

Competing interests
The authors declare that they have no con icts of interest.

Funding
This study was supported by a grant from the Deputy of Research, Tehran University of Medical Sciences (Grant No. 99-1-96-47101).

Role of the funding source
The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had nal responsibility for the decision to submit it for publication.   Table 4. Clinical characteristics of patients before and after treatment measurements in each of groups.
Microliter; mg/L Milligrams per liter; meq/L: Milliequivalent per liter; mm Hg: Millimeter of mercury; mm/hr: Millimeters per hour; ng/ml: Nanogram per milliliter; PCO2: Partial pressure of carbon dioxide; pg/ml: Pictogram per milliliter; RBC: Red blood cell; SD, Standard deviation; U/L: Unit per liter; VBG: Venous blood gas; WBC: White blood cell; NA: Not available. Figure 1 Randomization, enrollment, and treatment assignment. Kaplan-Meier curves for statistical comparison of time to death between both groups (P-value=0.973).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.