New Insights on Effects of Glucocorticoids in Patients With SARS-CoV-2 Infection.

OBJECTIVE: Since January 2020, the highly contagious novel coronavirus SARS-CoV-2 has caused a global pandemic. Severe COVID-19 leads to a massive release of proinflammatory mediators, leading to diffuse damage to the lung parenchyma, and the development of acute respiratory distress syndrome. Treatment with the highly potent glucocorticoid (GC) dexamethasone was found to be effective in reducing mortality in severely affected patients.
METHODS: To review the effects of glucocorticoids in the context of COVID-19 we performed a literature search in the PubMed database using the terms COVID-19 and glucocorticoid treatment. We identified 1429 article publications related to COVID-19 and glucocorticoid published from 1.1.2020 to the present including 238 review articles and 36 Randomized Controlled Trials. From these studies, we retrieved 13 Randomized Controlled Trials and 86 review articles that were relevant to our review topics. We focused on the recent literature dealing with glucocorticoid metabolism in critically ill patients and investigating the effects of glucocorticoid therapy on the immune system in COVID-19 patients with severe lung injury.
RESULTS: In our review, we have discussed the regulation of the hypothalamic-pituitary-adrenal axis in patients with critical illness, selection of a specific GC for critical illness-related GC insufficiency, and recent studies that investigated hypothalamic-pituitary-adrenal dysfunction in patients with COVID-19. We have also addressed the specific activation of the immune system with chronic endogenous glucocorticoid excess, as seen in patients with Cushing syndrome, and, finally, we have discussed immune activation due to coronavirus infection and the possible mechanisms leading to improved outcomes in patients with COVID-19 treated with GCs.
CONCLUSION: For clinical endocrinologists prescribing GCs for their patients, a precise understanding of both the molecular- and cellular-level mechanisms of endogenous and exogenous GCs is imperative, including timing of administration, dosage, duration of treatment, and specific formulations of GCs.

Background

Synthetic glucocorticoids (GCs) are a highly effective and non-expensive class of drugs that have been extensively used for more than 70 years to treat a large range of immune-related disorders and diseases. Although common practice was to avoid prescription of GCs during acute infection, due to their immunosuppressive effects, in recent decades adjunctive treatment with GCs has been shown. The outbreak of the novel coronavirus disease has led to a revival of GC treatment in the context of viral infection.

For clinical endocrinologists prescribing GCs for their patients, a precise understanding of both the molecular and cellular action mechanisms of endogenous and exogenous GCs is imperative, including timing of administration, dosage, duration of treatment and specific formulations of GCs.

Recent studies raised the concept that severe inflammatory response (hyperinflammatory syndrome), responsible for the rapidly deteriorating COVID-19 infection, closely resembles macrophage activation syndrome (MAS) and other cytokine release syndromes, associated with autoimmune disorders(3). The treatment of MAS is based on high dose of intravenous GCs, to prevent the vicious cycle of a perpetuating self-injuring inflammatory state(4).

Indeed, recent studies demonstrated beneficial effects of GCs in COVID-19 disease, suggesting SARS-CoV-2 infections exhibit unique features that enable the therapeutic efficacy of GCs(5, 6, 7). SARS-CoV-2 infections induce significant changes in infected cells, many of which influence intracellular GC actions (8, 9, 10). These changes are generated mainly to promote virus replication, but inadvertently modulate the glucocorticoid receptor (GR) activity and contribute to the therapeutic effects of GCs in severe COVID-19 patients(10).

More than 50% of patients with COVID-19 pneumonia, admitted to ICU will develop acute respiratory distress syndrome (ARDS)(11). Remarkably, the hyperinflammatory phenotype of ARDS, characterized among other, by elevated IL-8 and CRP levels, is associated with elevated mortality rate(12). According to recent studies, alveolar macrophages play an important role in pathogenesis of ARDS.

Accumulation of macrophages with acquired profibrotic phenotype was found in lung tissue of patients with COVID-19 complicated by ARDS (14). In a recent study in patients with symptomatic post-COVID-19 diffuse parenchymal lung abnormalities it was shown that treatment with high dose of prednisolone was not superior to a low dose regimen(15).

In this review, we are focus on the current literature which relate to effects of GCs in the context of a novel coronavirus infection.

Hypothalamic–pituitary–adrenal (HPA) axis in critical illness

Under stressful conditions, the HPA axis is activated, leading to increased cortisol production(16). Levels of cortisol are found to be significantly higher in patients with sepsis, as well as in patients undergoing major surgery, to combat stress and major trauma(17). Critical care patients have higher cortisol and lower ACTH levels. Reduced cortisol clearance, in these patients is associated with reduced expression and activity of cortisol-metabolizing enzymes in the liver and kidney(18). Enhanced corticosteroid-binding globulin levels and the multidrug resistance P-glycoprotein transporter expression, as well as changes in 11β hydroxysteroid dehydrogenase enzymes activity (decrease in 11βHSD1 and increase in 11βHSD2) decrease the levels of bioactive GC, leading to a state of GC resistance(19). Critical illness-related corticosteroid insufficiency (CIRCI) is a term indicating an inadequate HPA axis response during critical illness (20,21). Critically ill patients, with reduced delta cortisol that is, a change in baseline cortisol at 60 min of <9 μg/dl after 250mcg-ACTH stimulation, presumably have relative adrenal dysfunction and were found to have poor outcomes(20). According to the updated guidelines adult patients with septic shock should be treated with IV hydrocortisone <400 mg/day (or methylprednisolone equivalent in persistent ARDS for ≥3 days(20). Methylprednisolone is the preferred GC for the treatment of ARDS, due to its greater affinity to GRs, high penetration in lung tissue and high potency for both genomic and non-genomic activity(21). Moreover, high-certainty evidence indicates that GCs result in significant reductions in ICU and hospital lengths of stay (21,22). In a systematic review presented by Annane et al.(22) that included 33 eligible studies, treatment with a long course of low‐dose corticosteroids significantly reduced 28‐day mortality as well as mortality rates in ICU. A recent small study has shown that 40% of patients admitted with COVID -19 disease had low basal level cortisol (<300 nmol/l), while the corresponding ACTH levels, measured in the same samples were also in the lower end of the normal range, thus pointing to COVID-19 related HPA dysregulation(23). On the other hand, Tan at al. found that COVID-19 patients whose baseline cortisol concentration was equal to or less than 744 nmol/L had a median survival of 36 days whereas, patients whose cortisol value was more than 744 nmol/L had a median survival of 15 days (p=0.0001)(24). Unfortunately, studies aiming to assess HPA function by performing low dose ACTH test in moderately to severely ill COVID-19 patients are lacking, albeit primary and secondary adrenal insufficiency were occasionally reported(25,26).

Mechanisms of anti-inflammatory and immunosuppressive effects of glucocorticoids

The anti-inflammatory effect of GCs includes the repression of genes related to inflammation, via the activation of monomeric GR which negatively regulates transcription factors, such as the NF-κB and AP-1, or through crosstalk with other transcription factors(27).

Annexin-1 is a ubiquitous phospholipid-binding protein, activated by GR, implicated in signal transduction processes, which plays an important role in regulation of inflammatory response. Annexin-1 mediates the adhesion and migration of leukocytes; upregulates anti-inflammatory cytokines, such as IL-10; modulates response to apoptosis and thereby prevents tissue necrosis(28,29). High annexin-1 expression was found in lung tissue of patients with inflammatory lung disease treated with GCs(28).

In the context of acute inflammation, GCs influence several steps downregulating activation of innate and adaptive immune systems.

Components of innate immune system influenced by glucocorticoids

Granulocytes

Endogenous and exogenous GCs promote neutrophil maturation and mobilization from bone marrow to blood, while reducing neutrophil apoptosis. GCs inhibit expression of adhesion molecules, and thus inhibit adhesion and extravasation of neutrophils to the site of inflammation(30). In asthma and rheumatoid arthritis, exogenous GCs significantly accelerated basophils and eosinophil apoptosis and/or reversal of cytokine-induced eosinophil survival(31,32).

Monocytes /macrophage

GCs inhibit activation of monocytes by microbial products like lipopolysaccharides (LPS) and thus cause resistance to LPS-induced apoptosis and also induce differentiation of an anti-inflammatory monocyte phenotype(33,34). In activated macrophages, high doses of GCs inhibit the expression of inflammatory cytokines, including IL-1β, IL-6, IL-12 (TNF-α), as well as nitric oxide production(35.)

Dendritic cells

Exogenous glucocorticoids enhance endocytic activity of dendritic cells, causing impairment of differentiation, maturation, and T cells activation(36).

Natural killer (NK) cell

GCs dampen expression of several genes influencing the capacity of the NK cells to bind to targets cells and reduce the production of granule constituents (perforin and granzyme B) thereby decreasing the cytotoxic activity of NK cells(37,38). Also, glucocorticoids inhibit the production of IFN γ, a key NK cell cytokine(39).

Components of adaptive immune system influenced by glucocorticoids

T-cell activity

Glucocorticoids modulate T-cell activity directly by interfering with T-cell receptor signaling, and indirectly by attenuating dendritic cell functions: antigen presentation, co-stimulation, and cytokine production(36). Also, GCs exert potent regulatory effects on T-cell activities; they influence the polarization of T helper (TH) cells, favoring the differentiation of TH2 cells and regulatory T (Treg) cells over that of TH1 cells and TH17 cells(39).

B-cell activity

Glucocorticoids impair B cell receptor and Toll-like receptor signaling, thus preventing activation of B cells by endogenous immune complexes. GCs also promote significant up-regulation of the genes, associated with the expression of the anti-inflammatory cytokine IL-10(40).

GCs also have an important role in the resolution phase of inflammation. They mediate survival of an anti-inflammatory monocytic phenotype, as well as the programming of alternatively activated M2c-like macrophages that are characterized by high expression of scavenger receptors, which clear apoptotic cells and secrete anti-inflammatory cytokines(34).

Genomic and non-genomic mechanisms of glucocorticoids activity

GCs act through binding to cytosolic glucocorticoid receptor (GR), predominantly to GR alpha (GRα) which is the most active isoform of GR(41,42). This classic genomic mechanism includes conformation changes of cytoplasmic GR, phosphorylation at serine 211, dissociation from hetero complex, and translocation into the nucleus, where it regulates gene expression(42).

Exogenous and endogenous GCs bind unequally to GRs, providing uneven effects on genomic and non-genomic immune response, as well as on HPA axis suppression(43), as illustrated in Table 1 .

Table 1

Genomic and non-genomic effects of common GCs. The table shows a schematic comparison between genomic and non-genomic potencies of various glucocorticoids. The data of relative non-genomic effects taken from ref 47, 49, 93. The data of relative genomics effects taken from ref 47.

Among synthetic GCs, dexamethasone had a higher affinity for the GR, minor mineralocorticoid activity, greater bioavailability, and a much longer half-life than endogenous GCs(44). Remarkably, anti-inflammatory effects (potency) of GCs positively correlated with effects on glucose metabolism, hepatic deposition of glycogen and glycogenesis(45). The degree of cytosolic GR saturation and the level of plasma concentration of GCs correlates with the extent of the therapeutic effect(44). Dose equivalents of 100 mg prednisone or more cause virtually full cytosolic receptor saturation; therefore, a higher dosage could change drug pharmacodynamics and lead to the appearance of non-classical non-genomic effects(44).

Non-genomic effects of GCs manifest within minutes of GC exposure and do not involve activation of classical genomic pathways(46). These effects can be classified as involving physicochemical interactions with cellular membranes (non-specific effects), interference with cytoplasmic signaling complexes that are mediated by cytosolic GR, or specific interactions with membrane-bound GR(41,46). Non-genomic mechanisms of GC activity may provide an explanation for the beneficial therapeutic effect of pulse GC therapy for many inflammatory or immune-mediated diseases, including severe forms of rheumatoid arthritis, immune thrombocytopenia, juvenile dermatomyositis, juvenile chronic arthritis, optic neuritis, rapidly progressive glomerulonephritis, and pemphigus vulgaris(44). The genomic and non-genomic effects of commonly used GCs are summarized in Table 1.

Cushing’s syndrome and Covid -19 infection

Patients with active Cushing’s syndrome (CS) are immunosuppressed and at risk of bacterial, viral, and opportunistic infections due to immune system dysregulation(47). Chronic endogenous glucocorticoid excess alters the innate immune system response, due to suppressed natural killer cytotoxic activity and an adaptive immune system reaction resulting from the reduction of T-helper1and B lymphocyte counts, as well as increased apoptosis in the early phase of lymphocyte development(47,48).

It was hypothesized that CS patients with severe COVID-19 infection complicated with ARDS might have more severe outcomes due to a preexisting inflammatory state and deepening lymphopenia, because of T and B cell maturation impairment resulting from CS and concomitant lymphopenia caused by COVID infection with reduced number CD4T lymphocytes. At the same time, impairment of the cytokine response in CS may also lead to a more stable clinical course, avoiding ARDS development(48). The main features of the case reports on patients with COVID-19 and CS are summarized in Table 2 (49-52). Notably, the most severe course of COVID -19 infection occurred in a patient severely affected with florid, uncontrolled CS. These findings are in concordance with the previous data that showed that the predisposition to bacterial and opportunistic severe infections seems to be directly and positively correlated with cortisol levels and is more frequent in ectopic CS (49).

Table 2

Patients with COVID-19 and Cushing syndrome.

	Patient 1(49)	Patient 2 (49)	Patient 3 (49)	Patient 4 (50)	Patient 5 (51)	Patient 6 (52)
Age, years	71	38	66	67	27	71
Gender	F	F	F	M	F	M
Comorbidities
				N/A
• BMI>30 kg/m2	Yes	Yes	Yes		No	Yes
• Hypertension	Yes	No	Yes		Yes	Yes
• Diabetes mellitus	Yes	No	Yes		No	Not reported
• Cardiovascular disease	Yes	No	Yes		No	Not reported
Duration of CS	<3months	5 years	4 years	7 years	New diagnosis	N/A
Treatment of CS before COVID-19	No	No	No	Pasireotide	Metyrapone	Metyrapone
				Cabergoline
				Metyrapone
Max. morning ACTH level pg/ml	445.8	74.4	25.6	425	37.8	N/A
Reference range	7.2-63.3	7.2-63.3	7.2-63.3	10-50	N/A
Max Urine cortisol	N/A	959.7	286.2	315	1300	N/A
Reference range	100-379 nmol/24 h	100-379 nmol/24 h	100-379 nmol/24 h	16-168 nmol/24h	27-137 nmol/24h
Severity of COVID-19	Critical	Moderate	Mild	Moderate	Severe	Mild
O2 supplementation/mechanical ventilation	Yes	Yes	No	Yes	Yes	No

Role of Glucocorticoids and other immunomodulatory drugs in the treatment of severe COVID-19 infection and the connection to ARDS and systemic hyperinflammatory syndrome.

Macrophages play an important role in the development and progression of inflammatory responses associated with COVID-19 infection complicated by ARDS as well as in other acute and chronic inflammatory pulmonary diseases such as COPD and idiopathic pulmonary fibrosis(13). Significant similarity between COVID-19 associated macrophages and profibrotic macrophage populations (CD 163+) identified from lung tissue of patients with idiopathic pulmonary fibrosis (IPF) was demonstrated(14). Noticeably, acute exacerbations of interstitial pulmonary fibrosis closely resemble symptoms of ARDS(53). Beneficial effects of GCs in ARDS were found in a few large clinical trials (20, 21, 22), albeit patients treated with high-dose corticosteroids (>500 mg /day methylprednisolone) had a higher mortality rate(54,55).

The role of GCs in management of interstitial pulmonary fibrosis is highly controversial, either for maintenance or for acute exacerbations. Moreover, harmful effects probably outweigh benefits of anti-inflammatory GCs action. A possible explanation is related to pathogenesis of interstitial pulmonary fibrosis, in which damage to alveolar tissue and poor wound healing predominates over chronic inflammation(53).

Elevated Il-6 level is a sensitive biomarker of hyperinflammatory syndrome in patients infected with COVID-19, as well as for other hyperinflammatory syndromes, and it was also found to be an early predictor of severe COVID-19 disease, requirement for mechanical ventilation and worse prognosis in ARDS patients (56, 57, 58). Recently, additional treatment with an anti-IL-6 receptor monoclonal antibody tocilizumab, on top of standard treatment including antiviral drugs, systemic GCs in dose of ≤1 mg/kg/day methylprednisolone or its equivalent, and the nucleotide analogue remdesivir, was found to reduce the progression to mechanical ventilation and death in patients with moderate to severe COVID-19 pneumonia (59).

Recently, tofacitinib, a small-molecule Janus-activated kinase (JAK (inhibitor (preferentially inhibits JAK1 and JAK3) which dampens Janus kinase/ signal transducers and activators of transcription)JAK-STAT(pathways related to the synthesis of inflammatory mediators, was found to be effective to reduce respiratory failure and death in patients with COVID-19 pneumonia(60).

Several prospective controlled studies were conducted to evaluate the role of GCs in patients hospitalized due to COVID-19(61, 62, 63, 64, 65) (Table 3 ) . Among them, a large prospective RCT that included over 6000 patients clearly demonstrated the beneficial effect of dexamethasone treatment on 28-day mortality, when the patient needed invasive mechanical ventilation (64).The results of this study are in accordance with a recent meta- analysis review that reported a significant reduction of hospital mortality (RR 0.79; 95% CI 0.64–0.98) and ICU mortality (RR 0.64; 95% CI 0.42–0.97; P = 0.04) especially with early (<7 days) rather than late GC administration(5).

Table 3

RCT trials evaluating the role of GCs in patients hospitalized due to COVID-19.

Name	Patients number	Dose and duration	Cumulative dose dexamethasone equivalent	Main outcome
	Allocation
Glucocovid study, Corral-Gudino L et al (ref 61)	N=61(35 GT, 29- SOC)	Methylprednisolone	72 mg	Primary outcome∗ in 48%
		80 mg 3 days		SOC vs 40% in GT group(p=0.04)
		40 mg 3 days
Effect of hydrocortisone on 21-day mortality.	N=149 (76 GT, 73 SOC)	Hydrocortisone (2 regimens)	44 mg	Primary outcome∗∗
Dequin P-F et al (ref 63)		200 mg 4 days		SOC 50.7% vs 42.1% in GT
	ICU 100%	100 mg 2 days		group (p=0.29)
		50 mg 2 days
		Or
		200 mg 7 days
		100 mg 4days
		50 mg 3 days	78 mg
The RECOVERY trial (ref 64).	N=6425(2104 GT, 4321 SOC)	Dexamethasone 6 mg 10 days	60 mg	Primary outcome∗∗∗
				22.9% in SOC vs 25.7% in GT
				group (p<0.001)
				40.9% in SOC vs 29% in GT
				group (mechanically ventilated)
				(RR 0.64; CI 0.5-0.8).
				26.2% in SOC vs 23.3% in GT
				group (RR 0.82; CI 0.72-0.94)
The efficacy of corticosteroids therapy in patients with moderate to severe SARS-CoV-2 infection.	N=336	Prednisolone 25 mg 5 days	20 mg	Primary outcome:
Ghanei M et al(ref 65)	120 GT			Number of deaths,
	116 azithromycin			percentage of ICU admission,
	116 lopinavir/ritonavir			mechanical ventilation (NS)
				Shorten LOS (p=0.028,
				p=0.0007)

GT- glucocorticoid treatment

SOC- standard of care.

The primary outcome was a composite of death, admission to the intensive care unit, or requirement for non-invasive ventilation.

The primary outcome was death or persistent dependency on mechanical ventilation or high-flow oxygen therapy.

The primary outcome was 28 -day mortality.

Tocilizumab, tofacitinib, and dexamethasone were found to reduce mortality, due to respiratory failure, in patients with COVID-19 pneumonia (59, 60, 64). It’s worth noting that these drugs alter multiple signaling transduction pathways, downregulate a variety of inflammatory mediators and inhibit synthesis of inflammatory cytokines.

According to NIH Coronavirus Disease 2019 treatment guidelines, dexamethasone treatment, alone or in combination with other immunomodulators, such as tocilizumab baricitinib and tofacitinib have demonstrated clinical benefit in patients with early severe disease and/or systemic hyperinflammation (66).

The roles of dexamethasone in severe COVID-19 disease may be summarized as follows:

Inhibits activation of the transcription factor, NF-κB in various cells, such as the lung macrophages; prevents inflammatory cells infiltration and diffuse pulmonary alveolar injury (67,68).

Decreases IL-6 activity, thus reducing activation of multiple cytokines(67).

Represses AP-1 signal transduction pathway, thereby increasing gene transcription of anti-inflammatory cytokines(69).

Increases expression of annexin-1 and as such modulates anti-inflammatory response(29).

Inhibits endothelial expression of adhesion molecules like endothelial-leukocyte adhesion molecule 1 and intercellular adhesion molecule 1 and consequently probably prevents thrombotic complications(70, 71, 72).

Promotes macrophage differentiation toward anti-inflammatory phenotype, improving, in this way, macrophage survival and healing of damaged tissue(73).

Influences T-helper anti-inflammatory activity, through inhibition of IFN-γ response and IL-12 synthesis(74).

Replaces relative cortisol deficiency, due to inappropriate HPA axis response related to severe illness(20).

Synergistically acts with other immunomodulatory treatments (75,76).

Although WHO recommend GCs for patients with severe or critical COVID-19(77), several aspects of this treatment (i.e., dosage, adequate timing, duration, proper weaning, and choice of the right patient) are still a matter of debate and need more clinical trials to be clarified.

In Figure 1 . we summarized the anti-inflammatory versus profibrotic effects of GCs related to late stage of ARDS.

Figure 1

. Hypothetical dual anti-inflammatory and profibrotic action of glucocorticoids in severe Covid-19 infection. Glucocorticoids (GCs) inhibit NFκB expression and downregulates the expression of pro-inflammatory cytokines, chemokines, and adhesion molecules (a) (28). GCs prevent hyperinflammatory response by dampening activation of monocyte, macrophage, and dendritic cells (b) (34,35,37). GCs restrain macrophage differentiation in the M1 type inflammatory macrophages and mediates alternatively polarization into M2 type anti-inflammatory macrophages (c) (35). Glucocorticoids are not effective, and probably harmful in the fibrotic phase of ARDS related to severe lung involvement, related to Covid-19 (d) (54).

Conclusion

Recent studies showed the dramatic effects of dexamethasone in COVID-19 in reducing mortality, particularly in the subset of critically ill mechanically ventilated patients whereas there was no benefit and potentially harm in hospitalized patients not receiving oxygen support. In our review we focused on the current understanding of the unique anti-inflammatory and immunomodulating effects of GCs on the pathophysiologic processes induced by SARS-CoV-2 infection, that underlie the present recommendations on GC use in COVID-19 management. We review the effects of exogenous GCs on different immune cells and mediators of innate and adaptive immune system and discuss the alterations of the HPA axis activity during severe COVID-19. We also refer to the question of the value of GCs administration in those patients with relative adrenal insufficiency which has been described recently in severely ill COVID-19 patients. The efficacy of highly potent GCs in ARDS was proven in large RCT trials. Moreover, recently acquired data showed that patients with persistent severe ARDS syndrome related to SARS-CoV-2 infection suffered from pulmonary fibrotic changes like those seen in patients with IPF. In such, patients GC treatment is not effective, at least, and probably harmful. Although the right time to administer GCs after onset of symptoms is a subject of debate, it seems that in well-assessed patients that include inflammatory markers, respiratory parameters, and chest imaging, GCs alone or in combination with other immunomodulatory treatments could shorten hospital lengths of stay and reduce mortality.

Uncited reference

1., 2., 62..