Commonalities of sepsis and severe SARS-CoV-2 infection/COVID-19

COVID-19 and sepsis show various similarities in immunopathogenesis, pathophysiology, and clinical manifestations [10, 17, 42]. Similar to sepsis, a life-threatening organ dysfunction resulting from a dysregulated host response to infection (e.g., bacteria, viruses, and fungi) [1, 2], severe SARS-CoV-2 infection/COVID-19 meets the current diagnostic criteria of sepsis (Sepsis-3) [1] and pre-disposes the infected host to sepsis and septic shock [17, 42]. However, the pathogenesis of disease-dependent coagulopathy differs somewhat between the two entities [10, 17].

The initial studies published on COVID-19 observed a correlation of serum levels of pro-inflammatory cytokines [e.g., interleukin (IL)-6, IL-1b, IL-2, IL-8, IL-17, tumor necrosis factor (TNF)-α, granulocyte-colony stimulating factor (G-CSF), granulocyte macrophage-colony stimulating factor (GM-CSF)] with the severity of COVID-19 suggesting a relevance of cytokines in disease progress (Table 1) [6, 43]. Therefore, it was natural to compare the alteration in serum levels in COVID-19 vs. other cytokine releasing syndromes, to identify resemblances and to classify the emerging disease. The comparisons yielded inconsistent results. One meta-analysis revealed lower cytokine level in COVID-19 compared to other cytokine releasing syndromes [e.g., ARDS, sepsis, chimeric antigen receptor (CAR) T cell-induced cytokine release syndrome (CRS)] [4]. Further studies indicated no apparent difference in plasma cytokine profile in severe COVID-19, sepsis, and ARDS [44] or demonstrated lower cytokine levels in patients with SARS-CoV-2 sepsis compared to bacterial sepsis. Thereby immunoglobulins (Igs) IgA and IgG were higher [45] and complement activation (e.g., C3, C4, C5a, sC5b-9, C5aR1, factor D, and factor B) was increased [3, 45, 46]. Blood (lymphocyte and monocytes counts) and inflammatory biomarkers [C-reactive protein (CRP), ferritin, lactate dehydrogenase (LDH)] hardly differ between SARS-CoV-2 sepsis and bacterial sepsis patients, except for neutrophil count and procalcitonin (PCT) in bacterial sepsis patients and D-dimer in critical COVID-19 patients [4, 45].

Long pentraxin 3 (PTX3) plasma concentrations, typically increased in infections of fungal, bacterial, and viral origin; severe inflammatory response syndrome; sepsis; and cardiovascular diseases, correlate with disease severity and mortality [9, 51, 69, 70]. In COVID-19 patients, PTX3 was likewise increased in plasma and emerged as a strong independent predictor of 28-d mortality [9] and in severe COVID-19 circulating PTX3 not differing from levels of patients with other pulmonary sepsis [71] indicating similarity.

To date, the precise immunopathogenesis of COVID-19 still remains unclear, as is whether the cytokine storm is the predominant driver of disease severity and organ dysfunction in COVID-19. Moreover, there is an ongoing discussion about the role of hyperinflammation following the cytokine storm and the role of immunosuppression by defected cellular host response [72, 73].

Sepsis, COVID-19, and the complement system

The serum-operating complement system is a complex defense mechanism against invading pathogens and an essential part of innate immune response that comprises of multiple signaling pathways that can be activated by various triggers [74–76]: firstly, the classic pathway after contact with IgG- and IgM-containing immune complexes, by PTXs, including CRP, serum amyloid P component (SAP), and PTX3; secondly, the alternative pathway, typically triggered by LPS-derived from gram-negative bacteria; thirdly, the lectin pathway that involves the interaction of mannose-binding lectin (MBL), a serum protein, with mannose residues on bacterial surfaces, and with ficolins; and finally, however, not commonly classified as a complement-activation pathway by proteases from neutrophils and macrophages that can cleave C5 [77, 78]. Anaphylatoxins C3a and C5a formed in the complement cascades bind to their corresponding receptors C3aR, C5aR1, and C5aR2, leading to downstream production of inflammatory mediators [74, 78]. In turn, C5a/C3a and C5aR/C3aR stimulate the activation of CD4⁺ and CD8⁺ T cells and bridge the innate and adaptive immune responses [74, 79] or C3d the B cell response [80, 81].

Complement activation and generation resulting in high levels of complement peptides C3a and C5a during sepsis are considered to be one of the hallmarks of sepsis [2], similar to those identified in severe and critical COVID-19 patients [11, 46]. In COVID-19 the classical pathway is activated in all patients, while hyperactivation of the lectin and alternative pathways is associated with disease severity [13] and patients requiring ICU admission C5a levels were significantly higher compared with those without [3].

This results in the traditional therapeutic approaches using antibodies against C3 (AMY101, Compstatin), C5 (Eculizumab, Ravulizumab), C5a (BDB-001, Vilobelimab), C5aR1 [Avdoralimab (IPH5401)]; antagonists C5aR1 [Avacopan (CCX168), PMX53]; and enzyme inhibitors of C3 cleavage (APL-2 and AMY-101) or C5 cleavage (Zilucoplan) [75, 78, 82–84].

Biologics and small molecules

Currently, immunomodulators, antivirals, and biologics (e.g., anti-SARS-CoV-2 antibodies) are the therapeutic pillars of COVID-19 treatment strategies (Table 2). Depending on the severity of the disease (pneumonia, ARDS) and associated organ dysfunction; however, they are substantially different from the therapy of typical sepsis patients [16, 103, 104]. In-depth knowledge about these drugs and immunomodulatory effectors in oncology and rheumatology suggested their use to COVID-19 [105, 106].

As of December 2021, more than 900 drug-related interventional clinical trials targeting treatment COVID-19 are registered at ClinicalTrials.gov. During the early pandemic period, the strength of the cytokine storm driven by COVID-19 infection was considered to figure prominently in predicting patient outcomes [6, 43]. Large-scale clinical trials with IL-6 (siltuximab) and IL-6R (sarilumab, tocilizumab) antibodies have yielded conflicting results and failed to demonstrate a significant benefit in COVID-19, except for IL-6R antibodies in severe COVID-19 (hospitalized patients who require supplemental oxygen, high-flow oxygen, non-invasive ventilation, or invasive mechanical ventilation) [16]. The non-response to blockade of IL-6/IL-6R argues against IL-6 as the assumed predominant driver of COVID-19 CRS/cytokine storm, disease severity, and organ dysfunction.

A number of studies focused on inhibiting the host humoral immune response yielded promising results, suggesting current immunomodulatory drugs (e.g., corticosteroids, biologics, and small molecules) as constituents of a successful therapeutic strategy for moderate and severe COVID-19 [16, 104]. In addition, biologics and small molecules provide effective tools to interrupt the relevant signaling pathways of cytokines, cytokine receptors, JAKs, and complement proteins [105, 107–109]. Some other small molecules and well-known drugs or their metabolites, most of them lysosomotropic, exhibit modulatory effects on sepsis and COVID-19 relevant signaling pathways and raise hope for the successful treatment of both diseases (Table 1, Figure 1) [20–22, 41, 110, 111].

Display full size

Figure 1. 

Targets of lysosomotropic drugs and metabolites in sepsis, SARS-CoV-2 infection, and severe SARS-CoV-2 infection/COVID-19. Lysosomotropic compounds aim various targets in host cells: lysosomal pH, enzymes, and metabolism (black); in response to LPS-induced inflammation/severe viral infection (dark red); in SARS-CoV-2 infection of host cells and viral replication (blue); and the expression of C3, C3aR, C5, and in particular of C5aR1, a highly probable promoter of sepsis and COVID-19 related vasculitides (green) [11, 112]. Effects are indicated as decreased (green arrow) and increased (red arrow). ^* TLR3 shedding and decreasing effects on TLR4 signaling after endocytosis are yet not confirmed, however, highly probable. MIP-1β: macrophage inflammatory protein-1β

In contrast to the humoral immune response, therapeutic option to intervene the hyperactivation of inflammatory response in sepsis failed in large multicenter clinical trials [113]. However, the use of hydrocortisone is the only immune-modulating drug recommended in septic shock, given the ongoing need requirement of vasoactive drugs [103].

The promising results in vitro as well as in TLR4-null mice targeting of the TLR4 signaling pathway and the production of pro-inflammatory cytokines failed to meet the high expectations in clinical trials. For example, resatorvid (TAK-242), a small molecule TLR4 antagonist, does not suppress cytokine levels in patients with sepsis and shock or respiratory failure/ARDS [114–117]. Other promising targets such as high mobility group box 1 (HMGB1) in sepsis still lack clinical trials [118].

Shortcoming of single targeting with biologics in signaling pathways

Antibodies are highly specific and highly selective tools for capturing peptide messengers, blocking the corresponding receptors or neutralizing the viral pathogen, so the natural choice was to apply antibodies for treating sepsis and SARS-CoV-2 infection/COVID-19. Contrary to expectations, however, the results of clinical trial demonstrate that the benefit is unincisive or very poor, except for the antibodies addressing the S protein receptor-binding domain (RBD) of SARS-CoV- (bamlanivimab plus etesevimab/casirivimab plus imdevimab) [16] or targeting an evolutionarily conserved epitope outside the rapidly evolving receptor binding motif to neutralize SARS-CoV-2, its variants (e.g., Omicron), and multiple other sarbecoviruses, including SARS-CoV-1 (sotrovimab) [119, 120].

In sepsis, inhibition of single downstream pro-inflammatory cytokines, such as TNF-α, IL-1b, and IL-6, have failed in clinical trials, not to be unexpected given the substantial amounts of mediators involved in the pathogenesis [78]. To improve outcome in SARS-CoV-2 infection/COVID-19, antibody cocktails consisting of anti-IL-6, IL-1 receptor blocker, IL-1 type 1 receptor, and TNF-α are suggested [121], irrespective of the risk of serious adverse effects (e.g., bacterial pneumonia) due to more pronounced interference with the immune defense.

Quest for multi target small molecules and drugs

Hydroxychloroquine is a well-known disease-modifying anti-rheumatic drug (DMARD) indirectly reducing the production of anti-inflammatory cytokines (e.g., IL-1, IL-6, TNF-α, and IFN-γ) by mononuclear cells and of TNF, IFN-α, IL-6, and CCL4 (MIP-1β) in plasmacytoid dendritic cells (pDCs), natural killer cell co-cultures stimulated with RNA-containing immune complexes, and other antigen-presenting cells [122]. Despite the successful in vitro inhibition of SARS-CoV-2 infection of cells in vitro [123, 124], hydroxychloroquine has failed to demonstrate benefits in clinical trials, leading to a dissuasion from the application [16]. Terminal elimination half-life of 41 ± 11 days, poor and/or delayed pulmonary accumulation (steady state on day 10) [125], severe adverse effects (e.g., dysrhythmias, prolonging the QTc interval in and beyond the therapeutic window, serious drug-drug interactions with various other drugs) [16, 126] rendered hydroxychloroquine unsuitable for widespread use in prophylaxis and therapy. However, the concept of multi-targeting pro-inflammatory cytokines, similar to the suggested antibody cocktail outlined above, can alternatively be implemented with small molecules that have DMARD characteristics and a suitable drug profile. In both sepsis and SARS-CoV-2 infection/COVID-19, it seems to be preferable to prevent the excessive emergence and release of pro-inflammatory cytokines than to trap them after excessive release. Small molecules have the advantage of being able to act within cells at the subcellular level.

Lysosomotropism of drugs is multi-targeting

Lysosomotropism is a characteristic of small molecules accumulating in lysosomes, usually by passive diffusion across the lysosomal membrane and trapped in the lysosome lumen [127]. Typically, lysosomotropic compounds are characterized by one or more easily protonatable aliphatic nitrogen atoms localized in side chains or saturated ring systems, possessing a ClogP > 2 (lipophilicity) and a basic pKa between 6.5 and 11 [25, 128]. Trapping and enrichment results in an increase in lysosomal pH from 4.5–5 to 6–6.5 [102] and an inactivation of lysosomal enzymes through a lysosomal pH beyond their optimum pH range (pH 4.5–5.5). Various well-known approved drugs such as ambroxol, amitriptyline, chlorpromazine, desipramine, desloratadine, fluvoxamine, hydroxychloroquine, and azelastine (Figure 2) or the small molecule model compound NB 06 have been classified as lysosomotropic compounds [22, 100, 128–134]. The pivotal involvement of the lysosome in metabolism, endocytosis, and exocytosis renders it an interesting therapeutic target. Raising the pH in (late) endosomes and lysosomes results in non-specific partial or complete inhibition of numerous lysosomal enzymes addressing SARS-CoV-2 infection, replication, and shedding; as well as TLR receptors and host response to infection/complement system (Table 1, Figure 1).

Display full size

Figure 2. 

Lysosomotropic drugs, lysosomotropic metabolites as candidates for (systemic) prophylaxis of sepsis, viral infection, and transition to severe SARS-CoV-2 infection/COVID-19. (A) Pairs of confirmed lysosomotropic and non lysosomotropic drugs and their lysosomotropic N-desmethyl metabolites [22, 100, 128–133] and their ratio (r) of pulmonary tissue/plasma concentration, if available. Accumulation in pulmonary tissue is given at a ratio r > 1 [135, 136]; (B) H₁-antihistamines and candidates for (systemic) prophylaxis of viral infection and transition to severe SARS-CoV-2 infection/COVID-19. Azelastine (^*, green) is a nasal and ophthalmic H₁-antihistamine over-the-counter (OTC) drug with confirmed lysosomotropism and anti-SARS-CoV(-2) efficacy [134]. Bromhexine, ambroxol, loratadine, desloratadine, azelastine, clemastine, dimethindene, and diphenhydramine are likewise OTC drugs and are easily accessible; (C) various commonly used cardiovascular drugs and the bronchodilator salmeterol. Salmeterol (ClogP 3.61) is administered by inhalation and decreases levels of LPS triggered pro-inflammatory cytokines TNF-α, IL-6, and IL-1β [137, 138] suggesting present lysosomotropism is indicated as confirmed (^*, blue), highly probable (black), but yet not confirmed, and no lysosomotropism (^#, red)

Impact of lysosomotropism in sepsis models and SARS-CoV-2 infection/COVID-19 observational studies and clinical trials

Some of the lysosomotropic drugs and metabolites in Figure 2, in particular desipramine, fluoxetine, amitriptyline, and ambroxol, have already been tested for their efficacy in LPS-induced septic shock or polymicrobial sepsis/peritoneal contamination and infection (PCI) induced sepsis in animal models. In the murine PCI model, desipramine significantly reduced sepsis-typical hallmarks such as sepsis-induced cardiomyopathy, cardiac dysfunction, endothelial stress response, hepatic biotransformation capacity, activation of hepatic stellate cell in the acute-phase, and development of hepatic sepsis sequelae (e.g., post sepsis liver fibrosis) [19, 26, 151, 162].

Likewise, amitriptyline tested in the CLP model decreased mortality, mitigates neutrophil/monocyte accumulation in pulmonary tissue, and pulmonary injury during sepsis associated with increasing IL-10 levels in peritoneal lavage and decreasing MCP-1/CCL2 [67]. Preventive treatment with desipramine and fluoxetine significantly reduced TNF-α production and mortality in the murine LPS-induced septic shock model [163]. Ambroxol, with recently confirmed lysosomotropism, decreases infiltration of inflammatory cells into lung tissue, IL-1β, TNF-α, and MCP-1/CCL2 expression after LPS-induced inflammation and lung injuries following inhalation [68, 132]. The inhaled bronchodilator salmeterol (ClogP 3.61) decreases levels of LPS triggered pro-inflammatory cytokines TNF-α, IL-6, and IL-1β [137, 138], however, still lacking experimental lysosomotropism confirmation. Although not considered a typical lysosomotropic drug due to its somewhat low lipophilicity (ClogP –0.81), in the LPS-induced septic shock model ciprofloxacin shows similar effects on mortality, TNF-α, IL-1β, and CXCL2 [164].

Chloroquine and hydroxychloroquine emerged as the first lysosomotropic drugs to be trialed in the setting of SARS-CoV-2 infections/COVID-19, unfortunately without the desired benefit [16]. Random observations in psychiatric patients in France with a lower prevalence (about 4%) of symptomatic and severe forms of SARS-CoV-2 infections/COVID-19 compared to healthcare professionals (about 14%) when taking chlorpromazine or decreased mortality when using hydroxyzine pioneered subsequent and wider observational study on lysosomotropic drugs [30, 165, 166]. There, for example, escitalopram, sertraline, amiodarone, and amitriptyline reduced the risk of intubation or death in severe forms of SARS-CoV-2 infections/COVID-19 [21]. At the same time, a placebo-controlled, randomized, adaptive platform clinical trial of the lysosomotropic selective serotonin reuptake inhibitor (SSRI) fluvoxamine was initiated with the goal of evaluating its benefit in patients at known risk for a severe progression of SARS-CoV-2 infection. Treatment with fluvoxamine (100 mg twice daily for 10 days) among high-risk outpatients with early diagnosed COVID-19 reduced the risk of emergency care and hospitalization [20, 130].

Another strategy of drug repurposing is the nasal application of azelastine to prevent viral infections and to reduce viral load of SARS-CoV-2 in the nasopharynx. Azelastine is lysosomotropic and holds a marketing authorization as a nasal spray for allergic rhinitis [167, 168]. Therefore, it is natural to test azelastine placebo-controlled for its antiviral and viral load-lowering characteristics. After the preliminary evaluation, 0.1% azelastine nasal spray treatment decreased viral load, in particular in patients with initial high viral load and turning negative much faster than in the placebo group [134]. In all these successfully applied small molecules with various indications, lysosomotropism is a common feature [22, 129, 130, 169] that was not necessarily a consideration in the selection process.

Rational drug repurposing: the quest for lysosomotropism and the perfect drug profile

To identify new, readily available drugs with existing marketing authorization, a rational repurposing strategy is needed to become independent of random observations. Various lysosomotropic drugs and metabolites (e.g., in Figure 2) demonstrated anti-SARS-CoV(-2) efficacy, and some of them in severe forms of SARS-CoV-2 infections/COVID-19 and murine models of sepsis [19, 20, 25, 98, 110, 163, 167, 170, 171]. Drugs such as desipramine, fluoxetine, fluvoxamine or chlorpromazine are, however, not suitable for the systemic prophylaxis in healthy individuals due to their intrinsic pharmacology as anti-depressants. Eligibility criteria for drugs could include (Figure 3): lysosomotropism of drugs or their metabolites is present within the therapeutic margin, no or acceptable undesirable systemic pharmacological effects at lysosomotropic drug concentrations, and favorable systemic drug profile (clinical pharmacology considerations). In the case of viral infection prophylaxis of the respiratory tract, attention must be paid to the accumulation of the drug or its active metabolites in the nasopharynx and the lung. However, local application is a fallback option for otherwise suitable drugs [41, 111]. Clinical pharmacology considerations for small molecule in treatments for severe forms of SARS-CoV-2 infections/COVID-19 and sepsis can be performed using the liberation, absorption, distribution, metabolism, and excretion (LADME) model [172]. Antihistamines such as azelastine, clemastine, dimethindene, diphenhydramine, and loratadine (active metabolite desloratadine) or the expectorants bromhexine/ambroxol could be such readily available drugs with broad therapeutic window and safe drug profile. The generic adverse effects of lysosomotropic prescription drugs (e.g., QTc prolongation, Torsade de Pointes, ventricular arrhythmia, bundle branch heart block, and cardiac deaths, in particular from overdosing) are very rare (loratadine/desloratadine) [173] or absent (ambroxol) [174].

Display full size

Figure 3. 

Rational repurposing of lysosomotropic drugs and metabolites for systemic prophylaxis of transition to sepsis and prophylaxis of (systemic) SARS-CoV-2 infection and prevention of transition to severe SARS-CoV-2 infection/COVID-19. Candidates (active compound and/or one or more of its metabolites) should meet the following criteria: weak protonatable nitrogen bases with lipophilicity (ClogP > 2) and a basic pKa between 6.5 and 11 [25, 128], lysosomotropism is present within the therapeutic margin, with or without acceptable undesirable systemic pharmacological effects as lysosomotropic drug or originating from its metabolites. Systemic prophylaxis of sepsis (top) and transition of viral infection to severe SARS-CoV-2 infection/COVID-19 (bottom) requires a favorable systemic drug profile; however, in the case of an unfavorable drug profile, in prophylaxis of viral infection local application can serve as a fallback (middle). Systemic prophylaxis of viral infection requires an accumulation in pulmonary tissue after systemic application of the drug

Multi-targeting of lysosomotropic drugs vs. antibodies and antibody cocktails

Antibodies are undoubtedly a highly specific tool to inhibit pro-inflammatory mediators such as IL-6, and, however, are lacking success in clinical sepsis trials and severe SARS-CoV-2 infection/COVID-19 [2, 16, 78]. IL-6 inhibitors (e.g., sarilumab, tocilizumab) are only recommended in hospitalized patients, and COVID-19 patients who require supplemental oxygen, high-flow oxygen, non-invasive ventilation, or invasive mechanical ventilation [16]. In an effort to inhibit more pro-inflammatory mediators simultaneously, antibody cocktails consisting of anti-IL-6, IL-1 receptor blocker, IL-1 type 1 receptor, and TNF-α are suggested [121] despite the risk of serious adverse effects due to massive interventions in the immune system (e.g., bacterial upper respiratory tract infection, pneumonia, and viral infections). Even the JAK-1 inhibitors tofacitinib and baricitinib have also failed to meet expectations and are only recommended for COVID-19 under defined conditions [16]. Although lysosomotropic compounds likewise interfere with prominent mediators of the immune defense, adverse effects of IL-6 and TNF-α antibodies are not reported. In contrast to antibodies, the resynthesis of IL-6 and TNF-α and thus the available amount of the mediators is reduced, but not completely inhibited, still allowing a moderate immune response. To reiterate the benefits of lysosomotropic drugs and metabolites, here are a handful of examples of their benefit in sepsis and SARS-CoV-2 infection/COVID-19.