Use of glucocorticoids megadoses in SARS-CoV-2 infection in a spanish registry: SEMI-COVID-19

Objective To describe the impact of different doses of corticosteroids on the evolution of patients with COVID-19 pneumonia, based on the potential benefit of the non-genomic mechanism of these drugs at higher doses. Methods Observational study using data collected from the SEMI-COVID-19 Registry. We evaluated the epidemiological, radiological and analytical scenario between patients treated with megadoses therapy of corticosteroids vs low-dose of corticosteroids and the development of complications. The primary endpoint was all-cause in-hospital mortality according to use of corticosteroids megadoses. Results Of a total of 14,921 patients, corticosteroids were used in 5,262 (35.3%). Of them, 2,216 (46%) specifically received megadoses. Age was a factor that differed between those who received megadoses therapy versus those who did not in a significant manner (69 years [IQR 59–79] vs 73 years [IQR 61–83]; p < .001). Radiological and analytical findings showed a higher use of megadoses therapy among patients with an interstitial infiltrate and elevated inflammatory markers associated with COVID-19. In the univariate study it appears that steroid use is associated with increased mortality (OR 2.07 95% CI 1.91–2.24 p < .001) and megadose use with increased survival (OR 0.84 95% CI 0.75–0.96, p 0.011), but when adjusting for possible confounding factors, it is observed that the use of megadoses is also associated with higher mortality (OR 1.54, 95% CI 1.32–1.80; p < .001). There is no difference between megadoses and low-dose (p .298). Although, there are differences in the use of megadoses versus low-dose in terms of complications, mainly infectious, with fewer pneumonias and sepsis in the megadoses group (OR 0.82 95% CI 0.71–0.95; p < .001 and OR 0.80 95% CI 0.65–0.97; p < .001) respectively. Conclusion There is no difference in mortality with megadoses versus low-dose, but there is a lower incidence of infectious complications with glucocorticoid megadoses.

It has been postulated that in the development of the disease it is possible to distinguish a hyper-inflammatory phase, in which using glucocorticoids might play an essential role in preventing acute pulmonary distress syndrome (ARDS) [3][4][5].
Corticosteroids (CS) have been widely adopted, there are still questions on dosing, timing and duration of CS that have not been systematically studied at large.
The aim of this study is to describe the impact of different doses of corticosteroids on the evolution of patients with COVID-19 pneumonia.

Literature search
A literature search was conducted using the MEDLINE database with the following search terms: "corticosteroids and COVID-19," "megadoses and SARS-CoV-2," and "immunomodulatory and COVID-19." The most up-to-date evidence and all information regarding use of corticosteroids in COVID-19 reported in English or Spanish were selected.

Material and methods
This work is a multicenter, nationwide, observational study based on patient data obtained from the SEMI-COVID- 19 Registry, an enterprise of the Spanish Society of Internal Medicine (SEMI, for its initials in Spanish) to advance knowledge of the patients infected with SARS-CoV-2. The SEMI-COVID- 19 Registry was approved by the Provincial Research Ethics Committee of Málaga (Spain).

Study design and population
The registry is an anonymized online database of retrospective data on consecutive adult patients with COVID-19 hospitalized in internal medicine departments from 131 Spanish hospitals. The diagnosis was confirmed microbiologically by reverse transcription polymerase chain reaction (RT-PCR) testing of a nasopharyngeal or bronchoalveolar lavage sample. Exclusion criteria were subsequent admissions of the same patient and denial or withdrawal of informed consent. Patients were cared for at their attending physician's discretion, according to local protocols and their clinical judgment.
The registry includes data on more than 300 variables in categories such as: • Sociodemographic and epidemiological data More in-depth information on the registry and preliminary results are available in a previously published work [6].

Study endpoints
The primary endpoint was all-cause in-hospital mortality according to use of corticosteroids megadoses, defined as > 150 mg of prednisone in 24h. The follow-up period was from admission to discharge or death, including early readmissions.
We analyzed the criteria for the use of megadoses, any relationship to epidemiological, clinical, laboratory, and radiologic parameters, and the development of complications depending on the use of megadoses of corticosteroids.

Data analysis
We initially selected patients who received corticosteroids (5,262 out of a sample of 14,921). We further subdivided this population into two groups according to the amount of corticosteroids received: low-dose and megadoses. We defined megadoses therapy as the use of > 150 mg prednisone in 24h.
Continuous quantitative variables were tested for normal distribution using rates of skewness and kurtosis, Levene's test, or the Kolmogorov-Smirnov test, as appropriate. These variables were expressed as medians and interquartile range (IQR). Comparisons between groups were made using the Student's T-test, Mann-Whitney U test, Wilcoxon test, analysis of variance (ANOVA), or the Kruskal-Wallis test. Categorical variables were expressed as absolute values and percentages. Differences in proportions were analyzed using the chi-square test, McNemar's test, or Fisher's exact test, as appropriate.
Measures of association were expressed as odds ratio (OR) with 95% confidence intervals (95% CI). Statistical analysis was carried out using STATA software (v14.2). Statistical significance was established as p < 0.005.
We also used logistic regression to evaluate the relationship between use of megadoses and mortality. A multivariate analysis was carried out to adjust for confounding variables using clinically relevant, statistically significant variables (p < 0.001) identified in the previous analysis.

Demographics, and clinical features
Demographics and comorbidities in patients with corticosteroids or megadose therapy are shown in Table 1. Age differed between those who received megadose therapy versus those who did not in a significant manner (69 years [IQR 59-79] vs 73 years [IQR 61-83]; p < .001). There was a lower rate of megadose therapy among patients with dyslipidemia, arterial hypertension, heart and respiratory diseases. Regarding patients' previous treatment, a lower percentage of patients who were taking systemic corticosteroids therapy or other immunosuppressive received megadoses therapy.

Laboratory and radiologic findings
Radiologic and laboratory findings showed a higher use of megadose therapy among patients with an interstitial infiltrate and elevated inflammatory markers associated with COVID-19, such as elevated lactate dehydrogenase and C-reactive protein, on admission. Full data are presented in Table 2.

Other treatments
In Table 3 we studied the use of other treatments concomitantly for SARS-CoV-2 infection. There was a trend towards greater use of other immunomodulatory medications in those patients who also received corticosteroid megadoses.

Complications and mortality
There are differences in the use of megadoses versus low-dose in terms of complications. This is reflected in The analysis of outcome and mortality is shown in Table 5. We found no difference between ICU admission (14.4% low dose; 15% megadoses p .54) and average in hospital stay per days in both groups (12 days, IQR 7-18 low-dose; 12 days, IQR 8-19 megadoses. p .88). Tables 6 and 7 show the relationship between steroid use and mortality. Patients were initially divided into two groups according to whether or not they received steroid therapy, and specifically the use of megadoses or low-dose of corticosteroids. In the univariate study it appears that steroid use is associated with increased mortality (OR 2.07 95% CI 1.91-2.24 p <0.001) and megadose use with increased survival (OR 0.84 95% CI 0.75-0.96, p 0.011), but when adjusting for possible confounding factors, it is observed that the use of megadoses is also associated with higher mortality (OR 1.54, 95% CI 1.32-1.80; p < .001). The low-dose

Discussion
Throughout these long months of the covid19 pandemic, there has been much controversy about the role of corticosteroids in covid19 pneumonia, as there was no evidence of benefit in previous similar viral infections [5]. In March 2020, the World Health Organization (WHO) advised against its use [7], however the results of the RECOVERY clinical trial showed a reduction in mortality in the patients treated with 10 days of dexamethasone 6 mg compared to the placebo group [8].
To begin with, it is worth highlighting in this study the high number of patients who received corticosteroids. Specifically, 2216 out of 4794 patients received megadoses, considering megadose an amount of prednisone greater than 150 mg per day [9], representing a 46% in comparison with the 40% of people receiving pulse therapy in the Irastorza et al. observational

PLOS ONE
study of 242 patients [10], and the 20% described in the one published by López Zuñiga with 318 participants [11]. The particular interest in evaluating the difference between megadoses in contrast to lower doses was based on the hypothesis that they could have different impact on the evolution of the disease, as well as different side effects, since pulse therapy uses the non-genomic mechanisms of glucocorticoids to enhance the anti-inflammatory power and reduce the metabolic side effects and incidence of infections [12], as it has been demonstrated in other systemic autoimmune diseases [13].
Even though the advanced age, hypertension and dyslipidemia were postulated as risk factors for ADRS [14], in our work the patients receiving megadoses were younger and suffered less comorbidities than the patients of the other group, in contrast to the Irastorza series [10] in which no significant differences were observed. However, in his study, patients who received pulse therapy out of the second week of the disease and patients who did not receive corticosteroids were included in the same group. Returning to our study, a significant greater number of men received megadoses compared to women, 67% vs 33% respectively, perhaps because of the serious incidence of ADRS in men and their worse prognosis compared to women [15].
It should also be noticed that patients receiving previous immunosuppressive therapy received lower proportion of megadoses; we could not know if this was due to the fear of viral persistence already described in immunocompromised hosts [16], or because of a milder course in these patients, as current evidence only shows a probably increased risk of severe COVID-19 and death in patients with malignancy or solid organ transplant recipients, but this is less clear in other immunocompromised patients [17].
As for the situation at admission, the patients receiving megadoses had significant higher levels of lactate-dehydrogenase, c-reactive-protein and D-Dimer in contrast to the low-dose group, considering the higher these inflammatory markers are, the more they have been associated with lung damage, ADRS and worse prognosis. We found similar results about the radiological scenario, since the people treated with megadoses had a significant interstitial infiltrate in the X ray at admission in comparison with the low-dose group, and the radiological extension has also been associated to ADRS [18].
Regarding concomitant use with other immunomodulatory treatments or adjunctive treatments such us lopinavir-ritonavir, the group of patients of megadoses treatment received also more lopinavir-ritonavir, tocilizumab and bariticinib. A sub-analysis in our study of their effect on mortality showed that patients on lopinavir-ritonavir survived longer, as did those on tocilizumab and bariticinib. In this sense, it was hypothesized whether the concomitant effect of both, megadoses of corticosteroids and the specific immunomodulator, may influence the survival of COVID-19 patients. On tocilizumab, the EMPACTA clinical trial [19] included 249 patients in the tocilizumab group and 128 patients in the placebo group and the results suggested that patients who were most likely to benefit from tocilizumab had moderate or severe disease and that tocilizumab may add a potential benefit to antiviral treatment and glucocorticoids. Concerning bariticinib, there are few studies reflecting its use and impact on covid-19 and they include few patients. [20,21] In our study there were 86 cases registered who received corticosteroids at the same time, mostly megadoses, with a protective effect on mortality which is an interesting finding that requires further study. As for lopinavir-ritonavir, a randomized trial found that this treatment added to standard supportive care was not associated with clinical improvement or mortality in seriously ill patients with COVID-19 compared to standard care alone [22].
As for the development of complications during admission, significantly more complications of heart failure, arrhythmias and renal failure were observed in the non-megadose group, probably influenced by the higher proportion of comorbidities observed in this group compared to the megadose group. On the other hand, the incidence of venous thromboembolic disease and stroke in patients who used megadoses was higher, maybe explained by the greater inflammation in these patients, as it has been demonstrated in other studies [23,24]. In other series, there have been reported a 7% of bacterial coinfections in hospitalized COVID-19 patients, increasing to 14% in studies that only included ICU patients [25]. We would like to highlight that no higher proportion of bacterial pneumonia or sepsis were observed in the megadose group, supporting the initial hypothesis on the use of the non-genomic pathway of megadoses, as explained earlier [12].
Regarding the evolution of the patients who received corticosteroids, no increase in the average hospital stay was described among those who used megadoses. The patients in the megadoses group required more high-flow devices and non-invasive mechanical ventilation, but there were no differences between groups in terms of transfers to ICU or invasive ventilation.
In the observational study of Fernandez Cruz et al, a significant reduction in mortality was demonstrated among glucocorticoids users in the group classified as moderate-severe disease, but there were not significant differences between the patients receiving 1 mg/kg/d of methylprednisolone or pulse therapy (up to 500mg/d) [26]. In this study, the preliminary bivariate analysis showed an increased mortality among the patients receiving corticosteroids, however, in the group treated with megadoses (OR 0.85 CI 0.75-0.96) the survival rate was higher compared to the no megadoses group. The statistical significance disappeared in the multivariate analysis due to the introduction of confounding factors. Increased mortality is observed in the megadose group, as opposed to this type of regimen in other systemic autoimmune diseases [13]. Several studies have been recently published showing the effectiveness of high dose glucocorticoid pulse therapy in the prognosis of patients with COVID19 pneumonia in the inflammatory stage of the disease [10,11,27] in contrast to a Brazilian double-blind, randomized, placebo-controlled trial which reported no benefit from the use of methylprednisolone [28].
Thus, there is still no clear answer to which dose should be used, how long it should last or even if there are significant differences between dexamethasone and methylprednisolone. A new randomized controlled trial (CORTIVID) is coming soon with the intention to evaluate the role of pulse therapy [29].

Conclusion
This study includes a huge number of patients treated with corticosteroids and specifically with megadoses. There is no difference in mortality with megadoses versus low-dose of corticosteroids, but there is a lower incidence of infectious complications in megadoses group.

Limitations
It is a retrospective study. We could not evaluate the impact of megadoses in the respiratory situation, radiological evolution, nor in the inflammatory parameters, as we only had the data at the moment of hospital admission and a week later, so we could not establish a direct relationship with the glucocorticoid treatment, since it is difficult to establish the temporal relationship between the evolution of the evolution of the clinical parameters and the treatment. Besides, although the treatment regimens were divided into megadoses vs low-dose of corticosteroids based on > 150mg of prednisone/day or <150mg/day respectively, we could not establish the exact treatment regimens, the type of glucocorticoid used, nor its duration. Moreover, there are other glucocorticoid-related infections and side effects that have not been evaluated in this registry, so further studies and specific clinical trials to evaluate the differences between regimens are needed.