Famotidine is an off-patent drug available as either branded (“PEPCID®”) or generic medicines in tablet, capsule or intravenous forms. The general pharmacology of famotidine is well-characterized, with an excellent absorption, distribution, metabolism, excretion and toxicology profile 53. Famotidine is unique among the drugs currently being tested for treatment of COVID-19, in that it is an H2 receptor antagonist (and inverse agonist). Famotidine is currently being tested for treating COVID-19 in a double blind randomized clinical trial at high intravenous doses in combination with either hydroxychloroquine or remdesivir (ClinicalTrials.gov Identifier: NCT04370262). A recent retrospective cohort study of 1,620 hospitalized COVID-19 patients indicates that 84 propensity score matched patients receiving famotidine during hospitalization (oral or IV, 20mg or 40mg daily) had a statistically significant reduced risk for death or intubation (adjusted hazard ratio (aHR) 0.42, 95% CI 0.21-0.85) and also a reduced risk for death alone (aHR 0.30, 95% CI 0.11-0.80) 20. In contrast, proton pump inhibitor use was not associated with reduced risk for these outcomes. Anecdotal reports and undisclosed data indicating that famotidine provided protection from COVID-19 mortality while neither cimetidine nor proton pump inhibitors were similarly protective lead to an initial inference that the beneficial effects of famotidine were not related to the known on-target activity of the drug 21. Studies detailed in this report and others, however, indicate that famotidine does not act by directly inhibiting either of the principal SARS-CoV-2 proteases (PLpro or Mpro) 27. Vero E6-based cell assays also indicate that famotidine has no direct antiviral activity in this cell line, although antiviral activity in cells that express H2 has not been tested. Additional hypotheses that famotidine may act via binding either the sigma-1 or -2 receptors have not been supported by the studies summarized herein.
The most straightforward explanation of the apparent famotidine activity as a COVID-19 therapy is that the drug acts via its antagonism or inverse-agonism of histamine signaling and via its arrestin biased activation—all a result of its binding to H2. If true, then it is reasonable to infer that a SARS-CoV-2 infection that results in COVID-19 is at least partially mediated by pathologic histamine release. The anecdotal lack of protection provided by oral administration of the H2 antagonist cimetidine can be accounted for by insufficient systemic drug levels after oral administration and does not contradict potential benefit provided by famotidine H2 binding. Intravenous cimetidine at sufficient doses may achieve levels high enough for clinical benefit and would further support this hypothesis. Failure to achieve clinical COVID-19 responses with cimetidine may indicate that inverse agonism or other GPCR-mediated effects of famotidine may play an important role in the (preliminary) observed clinical benefits. Analysis of famotidine activity in histamine receptor competition assays indicate that, over the range of clinical steady state famotidine drug levels being tested, famotidine is specific for H2. Therefore, off-target antagonism of histamine H1 receptor, H3 receptor, or H4 receptor is unlikely to contribute to famotidine-mediated effects.
Steady state famotidine concentrations sufficient to elicit H2 antagonism (and inverse agonism) are readily achieved using inexpensive oral tablets and safe dosage levels. The famotidine dosage employed in the only retrospective hospital study currently available examining famotidine effects on COVID-19 outcomes appears to have employed dosages (20mg to 40mg daily) which are unlikely to fully inhibit histamine-mediated effects at the H2 receptor 20. In contrast, study NCT04370262 administers intravascular famotidine doses that are more than 20-fold greater than the IC50 for antagonism of H2. The data presented herein provides a rationale for famotidine dose selection to maintain a steady state concentration at a reasonable multiple of the IC50 for systemic antagonism of H2, and indicate that oral tablet dosages of between 40mg every eight hours to 60mg every eight hours should be sufficient to insure maximal H2 target effects. As famotidine is primarily cleared by the kidney, adequate renal function is required for higher dosages 53.
In addition to H2 antagonism, famotidine may also act as an inverse agonist thereby lowering the concentration of cyclic-Adenosine Monophosphate (c-AMP) 32. Endothelial cell permeability has been attributed to histamine H2 activation and is blunted by famotidine pretreatment 54. Histamine, bradykinin and des-arg-bradykinin receptor engagements can lead to increased endothelial permeability through a common pathway that results in AKT-1 activation 55. The H2 receptor also signals through Gq/11 proteins, resulting in inositol phosphate formation and increases in cytosolic Ca2+ concentrations which may account for increased the increased endothelial cell fluid permeability 56.
One alternative hypothesis is that famotidine may not only inhibit signaling through the H2 receptor but may also engage in cross talk with the kinin B1 receptor, which moderates the response of endothelial cells to DABK and DAKD ligands. Data provided here in are not consistent with this hypothesis; no activation of bradykinin receptor B1 or B2 were observed in quadruplicate replicate TANGO assay.
While COVID-19 symptoms affect multiple organ systems, respiratory failure due to acute respiratory distress syndrome (ARDS) is the most common cause of death. Examination of RNA expression profiles of the cells which contribute to lung anatomy and function demonstrate the presence of multiple ACE2/TMPRSS2 positive cell types susceptible to SARS-CoV-2 infection in the lung. In addition, these and other associated lung cells that are positive for histamine receptors H1 and H2 could respond to local histamine release following mast cell degranulation 57, and therefore those cells positive for H2 may be responsive to famotidine effects.
To understand how famotidine may act to reduce pulmonary COVID-19 symptoms requires an understanding of COVID-19 lung pathophysiology, which appears to have two principal disease phases. In turn, this requires an appreciation of pulmonary tissue and cell types. Pulmonary edema results from loss of a regulation of fluid transfer that occurs at several levels in the alveolus, as diagrammed in Figure 7. In the capillary wall, there are the glycocalyx, the endothelial cell with associated tight junctions, and the basement membrane. In the epithelium there is a surfactant layer on the alveolar lining fluid, manufactured and secreted by the Type II pneumocyte, and the Type I pneumocyte itself with its tight junctions and negatively charged basement membrane which restricts albumin. The pulmonary pericytes located in the terminal conducting airway region play a critical role in synthesizing the endothelial basement membrane and regulating blood flow in the precapillary arteriole, the capillary and the postcapillary venule. Disruption of any of these cells or layers can lead to edema. This edema fluid may be a transudate in milder dysfunctions or an exudate when inflammation or necrosis develop. Two possible pathologies that could result in edema of the alveolar wall and space include infection of cells by SARS-CoV-2 and mast cell degranulation with release of hundreds of compounds that can impact on cellular and basement membrane functions, glycocalyx and tight junction integrity. These compounds include histamine, bradykinin, heparin, tryptase and cytokines.
Gene expression patterns of these pulmonary cells provide insight into which cells are likely to be infected, and which express the H2 receptor that could be directly impacted by famotidine treatment and resulting H2 antagonism or inverse agonism (Figure 8). These patterns suggest that epithelial cells and endothelial cells are more likely to be infected based on ACE2 and TMPRSS2 expression patterns in those cell types. The cells most likely to show a famotidine effect include Type 2 pneumocytes, smooth muscle cells, pericytes, and myeloid granulocytes (which includes mast cells, neutrophils and eosinophils).
The limited tissue pathology available from early COVID-19 cases seems to support both viral infection as well as histamine effects in the lung. In a singular study of early COVID-19, Sufang Tian et al 59 describe the viral lung pathology of early COVID-19 in tissue resected for cancer. Their photomicrographs show two different patterns of disease. As shown in Figure 9 panel B, some samples of this lung tissue demonstrate the usual mononuclear inflammatory pattern of interstitial pneumonitis and fibrinous exudate that one would associate with a viral infection. It is striking that no neutrophils or eosinophils are observed in the inflammatory infiltrate. One explanation is that H2 activation of neutrophils inhibits neutrophil effector functions including O2- release 42,43, platelet-activating-factor induced chemotaxis 44 and leukotriene biosynthesis 45. Eosinophil functions are also inhibited by H2 activation; histamine binding diminishes eosinophil peroxidase release 46 and, at high concentrations, inhibits eosinophil chemotaxis 47,48.
The report of Tian et al 59 also includes an image in which there is interstitial and alveolar edema while the alveolar septae retain normal architecture (Figure 9 panel A). This is not a pattern typically observed in viral infection, as there is no inflammation, and the fluid appears to be a transudate. It is consistent with dysregulation of the fluid barrier due to the effect of histamine or other mast cell products on endothelial cells, pericytes or Type II pneumocytes. Increased endothelial permeability due to histamine is driven by H1 receptor activation, and so if any potential famotidine treatment effect on these cells occurs it would most likely be indirect by inhibition of mast cell degranulation. Forskolin activates the enzyme adenylyl cyclase and increases intracellular levels of cAMP, and can be used to inhibit the release of histamine from human basophils and mast cells 60. Histamine may act as an autocrine regulator of mast cell cytokine and TNF-a release in a PGE2-dependent fashion. Based on in vitro studies, this autocrine feedback appears to be mediated by H2 and H3 (88). Endothelial cells are also susceptible to infection by SARS-CoV-2. Mast cell degranulation-related pulmonary edema could correlate with the early phase silent hypoxia and the high compliance non-ARDS ventilation pattern associated with shortness of breath 61. The image in Figure 9 panel B does not permit evaluation for microvascular thrombi.
These findings are supported in a separate autopsy case report of a patient dying 5 days after onset of COVID-19 symptoms. In this case, photomicrographs also show a non-inflammatory transudative-type edema 62. In both of these studies, the observed non-inflammatory edema in early-stage COVID-19 pulmonary disease is consistent with histamine release by mast cells.
Mast cell degranulation correlates with the COVID-19 natural history that progresses through functionally and clinically different early and later phases. Most SARS-CoV-2 infections follow the typical early phase pattern of any lower respiratory virus, in which a majority of patients have asymptomatic or minimal disease, while a minority go on to later phase acute respiratory distress syndrome (ARDS). Within this spectrum typical of any severe viral disease, COVID-19 has a number of distinctive features. In the out-patient setting, early COVID-19 is usually indistinguishable from other “influenza-like illnesses”, presenting with various non-specific symptoms ranging from sore throat, headache and diarrhea to fever, cough, and myalgias. In these first few days however, COVID-2 may also be associated with anosmia, a unique feature 63. It is towards the end of the first week of symptoms that COVID-19 patients develop shortness of breath (SOB). This follows cough and fever by several days, a feature not typical of other viruses 64. On physical examination of COVID-19 patients with SOB, the oxygen saturation drops dramatically on exertion. CT scan will usually show bilateral bibasilar ground glass opacifications consistent with pulmonary edema. Nasopharyngeal swabs test positive for SARS-CoV-19. This SOB correlates with a distinctive clinical phenotype of hypoxia with near normal compliance (i.e. >50 mLcmH2O). Some authors attribute this to a loss of pulmonary vasoconstriction, one cause of which could be histamine effect on the H2 receptors of pericytes and/or vascular smooth muscle. H1-related edema and microthrombosis of lung vessels could also be causes. These are the patients that PEEP ventilation will not help, as there are no recruitable alveoli. These patients are helped by lying prone 65. It is at this stage that the patient is at greatest risk to progress onto the serious complications of later disease, especially ARDS with its 60-80% mortality if ventilation is required. Patients may also present with additional neurological symptoms and complications including ischemic stroke 66-68. Cardiac complications of later COVID-19 include myocarditis, acute myocardial infarction, heart failure, dysrhythmias, and venous thromboembolic events 69,70.
Multiple studies have demonstrated a hypercoagulable state in COVID-19 patients requiring hospitalization. Results from a recent large autopsy study suggests that there is also a novel lung-centric coagulopathy that manifests as a small vessel microthrombosis. Based on this study, there are indications that over 50% of patients dying of COVID-19 have pulmonary microthrombosis 71. This thrombosis is not only in arterial vessels, but also can be found in alveolar capillaries in the absence of inflammation and ARDS, as seen in Figure 10 72.
Capillary wall disruption accompanied by fibrin deposition and red cell extravasation, with neutrophils in the septa and within the alveolar spaces. (Hematoxylin and eosin, 1000x). For further discussion of microvascular coagulation associated with COVID-19, see 72.
Because small microthrombi are difficult to identify on CT scan even with iodinated contrast 73, pre-mortem diagnosis is difficult. Laboratory coagulation tests have typically shown normal or mildly prolonged Prothrombin time (PT) and activated partial thromboplastin time (aPTT), normal to increased or slightly decreased platelet counts, elevated fibrinogen levels and very elevated D-dimers 74. Although referred to by some authors as a DIC-like state, this pulmonary microthrombosis does not appear as a typical coagulation factor consumptive bleeding condition typical of overt DIC, but instead more closely resembles hypercoagulable thrombosis. This coagulopathy appears to be a core pathophysiology of COVID-19 as rising D-dimer levels, correlate with a poor prognosis, as do rising levels of IL-6 and CRP. IL-6 levels have been correlated to fibrinogen levels in one study, possibly through the acute phase reactant response 75. The pathogenesis of microthrombosis of the lung in COVID-19 is not known. There are multiple working hypotheses concerning this finding currently being assessed 76. Damage to the vascular endothelial glycocalyx can be caused by TNF‐α, ischemia and bacterial lipopolysaccharide. As well, activated mast cells release cytokines, proteases, histamine, and heparinase, which degrade the glycocalyx 77 and may thereby contribute to microthrombosis. Disruption of the glycocalyx exposes endothelial cell adhesion molecules, triggering further inflammation, rolling and adhesion of white blood cells and platelets 78. Glycocalyx components measured in serum positively correlate with increased mortality in septic patients 79. Other causes of hypercoagulability include direct damage to ACE2 positive endothelial cells by viral invasion or secondary damage from the inflammatory response to the infection. Mast cells release heparin which activates the contact system, producing plasmin and bradykinin. Plasmin activation could account for the singular rise in D-dimer levels. Activation of platelets also seems likely as part of the thrombo-inflammatory response but their precise role in thrombus formation remains to be elucidated 80. A more complete understanding awaits further study.
In addition to the usual features of a viral infection, early COVID-19 often presents with anosmia, ageusia, skin rashes including pruritis and urticaria, neuropsychiatric symptoms (including altered dream states), and silent hypoxia. These symptoms are all consistent with histamine signaling. Anosmia, ageusia, and other symptoms relating to cachexia are often reported in both COVID-19 and mast cell degranulation syndrome, and the potential role of histamine signaling in driving the pathophysiology of cachexia has been reviewed 81,82. As summarized in Figure 11, the distinctive later findings of abnormal coagulation, involvement of other organ systems and ARDS occur in the second week after the appearance of symptoms. This is coincidental with a rising immunoglobulin response to SARS-CoV-2 antigens. For a subset of patients, disease progress may suddenly worsen at days 7-10, and this correlates with the onset of SARS-CoV-2 spike protein neutralizing antibody titers 83. In this study, it was shown that IgG starts to rise within 4 days post-symptoms, inconsistent with a first antigenic exposure 83. Rapid onset of specific neutralizing antibody responses beginning less than seven days after exposure to SARS-CoV-2 implies a recall rather than primary B cell response, and therefore the response is being driven by a pre-existing memory cell population. These memory cells may have been educated by prior exposure to another coronavirus (e.g. circulating alphanumeric coronaviruses), raising concerns that this second phase of COVID-19 disease progression could share an immunologic basis with Dengue hemorrhagic fever 84. Antibodies produced from this early rapid humoral response may drive further mast cell degranulation. During this phase rising D-dimer levels correlate with poor prognosis, as do measured levels of CRP and IL-6.
Current reviews seek to explain COVID-19 clinical and pathologic findings based on standard models of antiviral innate and adaptive immune responses which do not consider the potential role of mast cell activation and degranulation. Reviews emphasize the inflammatory cell response cascade associated with monocytes, macrophages 85, and adaptive T and B cell helper and effector responses 86. These types of immune responses are also invoked to explain the novel microvascular pulmonary intravascular coagulopathy associated with COVID-19 87.
We propose an alternative paradigm; SARS-CoV-2 infection-induced mast cell activation could account for some of the core pathologic cascade and much of the unusual symptomatology associated with COVID-19. Many of the unique clinical symptoms observed during the early phase of COVID-19 are consistent with known effects of histamine release. Histamine may act as an autocrine regulator of mast cell cytokine and TNF-a release in a PGE2-dependent fashion, and based on in vitro studies the autocrine feedback appears to be mediated by H2 and H3 88. This model is consistent with the histopathologic findings seen at surgery, autopsies, and is supported by clinical pharmacologic findings suggesting potential benefits of histamine H2 receptor blockade using famotidine. This model is also supported by the significant overlap in the clinical signs and symptoms of the initial phase of COVID-19 disease and those of mast cell activation syndrome (MCAS) 89-92 as well similarities to Dengue hemorrhagic fever and shock syndrome (including T cell depletion) during the later phase of COVID-19 84,93,94. The cardiac events, stroke, and related outcomes associated with COVID-19 also appear consistent with the Kounis syndrome 95-97.
If COVID-19 is partially driven by dysfunctional mast cell degranulation, then a variety of medical interventions employing marketed drugs useful for treating mast cell-related disorders may help to reduce death and disease associated with SARS-CoV-2 infection. Examples include drugs with mast cell stabilizing activity, other histamine antagonists (for example H1 and H4 types), leukotriene antagonists and leukotriene receptor antagonists 98, anti-inflammatory agents such as those developed for inflammatory bowel diseases, and mast cell activation inhibitors 99. If such repurposed drugs are used in combination with pharmaceuticals that directly inhibit SARS-CoV-2 infection or replication, it may be possible to rapidly develop potent, safe and effective outpatient treatments for preventing or treating COVID-19 until such time as a safe and effective SARS-CoV-2 vaccine becomes available.