Can corticosteroids improve the outcomes of patients with Covid-19? A retrospective cohort study of patients within and outside the epicentre

The novel coronavirus disease 2019 (Covid-19) has been a worldwide pandemic with more than 300,000 deaths. Corticosteroids have been used in some patients with severe Covid-19 in order to control the systemic inammation or cytokine storm, however, their effects and safety have not yet been elucidated. Patients with conrmed Covid-19 were retrospectively included from both the epicentre and out of the epicentre. Patients were classied into two groups according to the use of systemic corticosteroids, and the mortality and the rate of virus clearance were compared between the two groups. In addition, independent factors associated with death after corticosteroids treatment were also identied. This study was designed as a retrospective cohort study to compare the outcomes between patients with and without corticosteroids treatment. Corticosteroids treatment was classied as systematic administration of corticosteroids including oral prednisolone and intravenous injection of dexamethasone or methylprednisolone. Patients were further stratied based on different severities and other characteristics to assess the effectiveness of corticosteroids treatment. We also endeavoured to identify any subgroups who might potentially benet from corticosteroids treatment. percentage lymphocytes. Virus clearance dened as negative results real time-polymerase (RT-PCR) an at least 24 and virus clearance time was calculated from the diagnosis to the date of the second negative RT-PCR test.


Introduction
The coronavirus disease 2019 (Covid-19) is a severe acute respiratory syndrome (SARS) caused by the infection of a novel SARS coronavirus (SARS-CoV-2), which initially emerged in Wuhan, China in December, 2019. 1 Since the declaration of a public health emergency of international concern and subsequently a pandemic by the World Health Organization, the rapid outbreak of Covid-19 has led to a global anxiety with about 5 million infected cases and more than 300,000 deaths. Although numerous trials of different medications have been made such as antiviral drugs including Favipiravir, Remdesivir, Lopinavir and Ritonavir, and chloroquine and hydroxycholoroquine, no convincing results have been reported due to the lack of proper study designs as well as data for peer review. 2,3 So far, there are no effective treatments for Covid-19 and the potential vaccines are still under clinical trials.
Cytokine storm syndrome has been identi ed as a key character in patients with Covid-19 and the levels of some speci c cytokines are associated with disease severity and mortality such as IP-10, MCP-3, and IL-1 receptor antagonist. 4 Therefore, it is recommended that treatments of these hyperin ammations should be identi ed and addressed to reduce the mortality. 5,6 Corticosteroids are potent anti-in ammatory agents with the ability to modulate the in ammatory response, and they were reported to reduce the proin ammatory cytokines 7 and mortality 8 during 2002 SARS-CoV outbreak. However, there is a lack of study of corticosteroids in this novel coronavirus pandemic. In this study, we aimed to evaluate the effectiveness of systematic corticosteroids in patients with Covid-19, with the particular interests in any differences of treatment response within and out of the epicentre.

Methods
This study was approved by the Ethics Committee on Biomedical Research, West China Hospital of Sichuan University. All patients were consented for publications.

Patients
We retrospectively collected the data of adult patients with con rmed Covid-19 from 2 centres in Wuhan and 46 centres in Sichuan (out of Wuhan) from January to March, 2020. In order to study the e cacy of corticosteroids treatment in patients with acute respiratory distress syndrome (ARDS) caused by Covid-19, we collected ARDS patients from another 2 centres in Wuhan.
Covid-19 and different disease severities were diagnosed in accordance with the published guidelines. 9 Covid-19 was con rmed if suspects met one of the etiological and serological evidences: 1) positive for SARS-CoV-2 by real-time uorescent RT-PCR, 2) highly homologous to SARS-CoV-2 in viral gene sequence, or 3) serum SARS-CoV-2 speci c IgM and IgG detectable or reaching a titration of at least 4-fold increase during convalescence. Four severity categories were classi ed: 1) Mild, mild symptoms without pneumonia; 2) Moderate, fever and respiratory symptoms with pneumonia; 3) Severe, respiratory rate (RR) ≥ 30 breaths/min, oxygen saturation ≤ 93% at rest in room air, or the ratio of arterial partial pressure of oxygen to fraction of inspired oxygen (PaO 2 /FiO 2 ) ≤ 300mmHg; and 4) Critical, respiratory failure with the need for mechanical ventilation, shock, or organ failure requiring admission to intensive care unit (ICU).
Patients with ARDS was de ned according to the following criteria: 1) acute onset of respiratory symptoms within 7 days, 2) bilateral in ltrates on chest imaging, 3) without evidence of left atrial hypertension and cardiogenic pulmonary oedema, and 4) severe hypoxemia at PaO2/FiO2 ≤ 300mmHg. 10,11 Study design and Outcomes This study was designed as a retrospective cohort study to compare the outcomes between patients with and without corticosteroids treatment. Corticosteroids treatment was classi ed as systematic administration of corticosteroids including oral prednisolone and intravenous injection of dexamethasone or methylprednisolone. Patients were further strati ed based on different severities and other characteristics to assess the effectiveness of corticosteroids treatment. We also endeavoured to identify any subgroups who might potentially bene t from corticosteroids treatment.
The primary outcomes of this study were patients' prognosis including mortality and rate of virus clearance. The secondary outcomes included hospital length of stay (LOS), virus clearance time, and change of ratio of pulse oximetric saturation to fraction of inspired oxygen (SpO 2 /FiO 2 ) and percentage of blood lymphocytes. Virus clearance was de ned as at least two consecutive negative results of real time-polymerase chain reaction (RT-PCR) in an interval of at least 24 hours, and virus clearance time was calculated from the date of con rmed diagnosis to the date of the second negative RT-PCR test.
Statistical analysis SPSS Statistics V26 (© Copyright IBM Corporation and other(s) 1989-2019) was used for data analysis, and GraphPad Prism V8.4.2 was used for gure depiction. Continuous variables were shown as mean ± standard deviation, while dichotomous variables were reported as frequency and proportion. Independent Student's t-test and Spearman Chi-square test or Fisher's exact test were performed to compare the variables and outcomes between the patients with and without corticosteroids. Analysis of variance (ANOVA) was conducted to compare the SpO 2 /FiO 2 and blood lymphocytes among different time points in patients with corticosteroids treatment. Bivariate Pearson Correlation Matrix was used to analyse the correlation among various cofounding factors with potential in uence on corticosteroids treatment related mortality. Univariate and multivariate logistic regressions were conducted to identify potential independent factors associated with corticosteroids treatment related deaths. We further performed receiver operating characteristics (ROC) curves of each independent risk factors for death to analyse the area under the curve (AUC), and calculated the cut-off values by Youden index (=sensitivity + speci city -1). Statistical signi cance was de ned as a two-sided P value of < 0.05.

Results
A total of 775 patients were included in our nal analysis, of which 490 patients were from Wuhan and 285 patients were from Sichuan. Eventually, 238 (30.7%) patients received systemic corticosteroids treatment including 201 (41.0%) patients from Wuhan and 37 (13.0%) from Sichuan. Excluding the patients with ARDS separately collected from the 2 centres in Wuhan, the overall mortality rate was 5.4%, while the mortality rate in Wuhan and Sichuan was 7.7% and 1.3%, respectively. Although signi cant differences could also be seen in RR, total bilirubin (TB), alanine aminotransferase (ALT), aspartate transaminase (AST), albumin (ALB), and blood urea nitrogen (BUN), however, they all fell into the normal ranges.
Due to the inconsistent baseline characteristics between patients with and without corticosteroids treatment, we tried to investigate whether these signi cant differences were derived from the different disease severities in each group. We found that almost all of these characteristics were signi cantly different between the two groups (Table S1). Therefore, we further compared the outcomes after strati cation of the patients into different disease severities.

Primary outcomes
In total patients, compared with patients without corticosteroids treatment, patients with corticosteroids treatment had signi cantly higher mortality ( (Table 3) When compared the treatment outcomes of corticosteroids in patients from different regions, we found a similar pattern in patients from Wuhan, while in patients from Sichuan, no signi cant differences were shown in either disease severity groups. (Table 3)

and 2)
Risk factors for death associated with corticosteroids treatment Table 4 showed the potential factors associated with death in patients with corticosteroids treatment. We found that the factors with signi cant differences were similar to those demographics and characteristics identi ed as signi cantly different between the patients with and without corticosteroids treatment. Compared to survivals, patients who died with corticosteroids treatment were older (72±13 vs. 54±15, P<0.001) and there were more patients with symptoms (e.g. dyspnoea: 66.7% vs. 27 After the correlation matrix analysis of the above potential risk factors (Table S3), we nally chose the variates with less internal correlations in our multivariate logistic regression model, which included age, blood lymphocytes, D-dimer, ALT, CK, and the percentage of comorbidities and use of corticosteroids within 48 hours of admission. The multivariate logistic regression showed that blood lymphocytes (odds ratio (OR) 0.792, 95% con dence interval (CI) 0.672-0.932, P=0.005) and CK (OR 1.006, 95%CI 1.000-1.012, P=0.038) were independent risk factors associated with death in patients with corticosteroids treatment. (Table 5) The ROC curves identi ed a cut-off value of 10.45% for blood lymphocytes (AUC 0.863, 95%CI 0.778-0.947, P<0.001) with a sensitivity and speci city of 90.91% and 70.75%, and a cut-off value of 239.50 U/L for CK (AUC 0.672, 95%CI 0.504-0.840, P=0.023) with a sensitivity and speci city of 44.44% and 94.05%. (Figure 3 and Table 5)

Discussion
In this study, we found that corticosteroids inclined to be used in patients with severe and critical Covid-19, however, corticosteroids treatment was associated with higher mortality, lower rate of virus clearance, and increased days of hospital stay, although it could improve oxygenation and elevate the level of blood lymphocytes in severe patients. Low blood lymphocytes (<10.45%) and high blood CK level (>239.50U/L) were independent risk factors associated with mortality in patients receiving corticosteroids treatment.
Since the outbreak of SARS-CoV-2, debates have increased on the use of corticosteroids in patients with Covid-19. 12,13 Based on the experiences during the 2002 SARS and 2009 H1N1 pandemic, corticosteroids are expected to improve the outcomes if utilized in an early acute phase of infection. 14,15 However, controversial ndings have been reported in the treatment of patients with Covid-19. Case reports and studies with small number of patients found that corticosteroids treatment did not shorten duration of symptoms or hospital length of stays, but associated with greater risk of ICU admission. 16,17,18 Furthermore, high-dose corticosteroids use has been reported to be associated with death in patients with severe Covid-19. 19 On contrary, a retrospective cohort study of 201 patients with Covid-19 found that treatment with methylprednisolone could signi cantly decrease the risk of death by 62% among patients with ARDS. 20 In our study, we found that corticosteroids treatment was associated with increased mortality, especially in patients with severe Covid-19 or ARDS. While, in patients with mild-to-moderate Covid-19, corticosteroids treatment could not bring any bene ts. Although our results showed that corticosteroids could improve oxygenation in patients with severe Covid-19 and ARDS, but this increase of oxygenation still could not be able to change the nal outcomes, which we think might be due to the limited extent of increase in oxygenation. Therefore, our results support the recommendations by the World Health Organization against the administration of systemic corticosteroids to patients with Covid- 19. 21 There is another concern of the commencement of corticosteroids in patients with Covid-19, which is the delay of virus clearance. Studies in patients with SARS-CoV, in uenza A (H7N9), and Middle East Respiratory Syndrome (MERS) have already shown delayed viral clearance after systemic corticosteroids treatment. [22][23][24] In this study, we found that corticosteroids treatment was not associated with longer virus clearance time in patients with severe Cvoid-19 and ARDS, which is in consistent with the previous reports. 17,25 However, in patients with mild-to-moderate Covid-19, we noticed that the use of corticosteroids signi cantly delayed the virus clearance time. In addition, we also found that the rate of virus clearance was signi cantly lower in patients with corticosteroids regardless of the disease severities. Hence, the use of systemic corticosteroids in patients with Covid-19 should be cautious.
In terms of blood lymphocytes, previous studies have reported in consensus that lymphopenia is commonly developed in patients with Covid-19, 16,18 which is caused by upregulated cell apoptosis, autophagy, and migration to the lungs. 26,27 The level of lymphocytes has been found to be negatively associated with disease severity and a gradual increase of lymphocytes has been seen in recovered patients. 16,18,28 In addition, lymphopenia has also been reported to be an indicator for mortality in patients with 30 In our study, we found that the level of blood lymphocytes could also be served as a good predictor of death related to corticosteroids treatment. We also noticed an increase of lymphocytes after corticosteroids treatment especially in patients with severe Covid-19 and ARDS, however, it did not result in reduced mortality related to corticosteroids treatment even though the levels the lymphocytes were increased to by the corticosteroids were similar to those in patients without corticosteroids treatment. Thus, it seems the outcomes of the patients are determined by the severities of Covid-19 rather than the increase of blood lymphocytes, which is also true for the increase of oxygenation by corticosteroids. On the other hand, it is not clear how corticosteroids increase the blood lymphocytes. One possibility is the amelioration of in ammation by corticosteroids, which results in less cell migration to the in ammatory site. However, further studies are needed to address this phenomenon.
There are several limitations in our study. Firstly, albeit this is a study with relatively large numbers of patients, prospective controlled trials are needed for further validation, especially with comparable patients' baseline characteristics. Despite the strati cation of different disease severities in our study, we could still notice the signi cant differences in oxygenation and blood lymphocytes between the patients with and without corticosteroids treatment, which might potentially underestimate the effects of corticosteroids. Secondly, due to the miscellaneous types, doses, and durations of corticosteroids used, it is not feasible for us to draw a clear conclusion whether any speci c regime may potentially bring bene ts or is associated with worse outcomes. Thirdly, some other cofounding factors such as the access to ICU have not been investigated or considered because of the shortage of medical sources. Finally, due to the small numbers of dead patients with corticosteroids treatment, limited variates were included in the multivariate logistic regression model, which might omit some clinically important causes associated with mortality such as complications and multiple organ dysfunction syndrome, as SARS-CoV-2 could infect other tissues besides the lungs due to the wide expression of the virus receptors in different tissues 31 .

Conclusions
For patients with Covid-19, especially those with severe categories, corticosteroids treatment is associated with increased mortality and reduced rate of virus clearance. Therefore, commencement of corticosteroids in patients with Covid-19 should be cautious before more evidence. In addition, blood lymphocytes and CK levels could be used as predictors for corticosteroid treatment failure. Abbreviations ALB, albumin; ALT, alanine aminotransferase; ANOVA, analysis of variance; ARDS, acute respiratory distress syndrome; AST, aspartate transaminase; AUC, area under the curve; BUN, blood urea nitrogen; CI, con dence interval; CK, creatine kinase; Covid-19, coronavirus disease 2019; ICU, intensive care unit; LOS, length of stay; MERS, Middle East Respiratory Syndrome; OR, odds ratio; PaO 2 /FiO 2 , ratio of arterial partial pressure of oxygen to fraction of inspired oxygen; ROC, receiver operating characteristics; RR, respiratory rate; RT-PCR, real time-polymerase chain reaction; SARS, severe acute respiratory syndrome; SpO 2 /FiO 2 , ratio of pulse oximetric saturation to fraction of inspired oxygen; TB, total bilirubin. Tables   Table 1 Demographics between patients with and without corticosteroids treatment.    arterial partial pressure of carbon dioxide; PaO2, arterial partial pressure of oxygen; PLT, platelet; RR, respiratory rate; SBP, systolic blood pressure; SpO 2 , pulse oximetric saturation; T, body temperature; TB, total bilirubin; WBC, white blood cell.   Change of Lymphocytes in patients with and without steroids.