Non‐steroidal anti‐inflammatory drugs (NSAIDs) and other eicosanoid pathway modifiers are one of the most frequently used anti‐inflammatory medications worldwide against symptoms of infections, other acute and chronic inflammatory diseases, and pain. Eicosanoids, including prostaglandins (PGs), leukotrienes (LTs), thromboxanes (TXs), hydroxyeicosatetraenoic acids (HETEs), lipoxins (LXs), and many recently proposed pro‐resolving mediators constitute a wide range of active lipid mediators possessing pro‐ and anti‐inflammatory, as well as pro‐resolution properties. ¹  They are products of the major unsaturated fatty acids: arachidonic acid (AA), dihomo‐γ‐linolenic acid (DHGLA), eicosapentaenoic acid (EPA), and docosahexaenoic acid (DHA), metabolized in three main pathways: cyclooxygenase (COX), lipoxygenase (LO or LOX), and cytochrome P450 (Figure 1). Those active lipid mediators play substantial roles in the development and resolution of inflammation, including allergic and viral inflammation, which we have reviewed extensively in the previous report. ¹ Even though NSAIDs and other eicosanoid pathway modifiers are so commonly consumed and are relatively safe for the majority of people, the current pandemic of the severe acute respiratory syndrome coronavirus‐2 (SARS‐CoV‐2) revealed substantial knowledge gaps in understanding their modes of action, benefits, and risks related to their use in patients with respiratory and allergic diseases. Unfortunately, this resulted in the conflicting messages sent to the public from the scientific community. Therefore, we, the European Academy of Allergy and Clinical Immunology Task Force (EAACI TF) on Eicosanoids, here critically review the most recent findings on the roles of NSAIDs, leukotriene antagonists (LTRAs), prostaglandin D₂ (PGD₂) receptor antagonists, and cannabinoids, as well as we summarize their selectivity, and additional modes of actions in allergic airway diseases, drug allergy, and respiratory virus infections. In addition, we also describe here the effects of novel T2 biologicals used in allergic diseases on eicosanoid pathways.

Aside from NSAIDs‐exacerbated respiratory disease (NERD), NSAIDs are usually well tolerated by patients with asthma, and therefore, they are often used in real life against symptoms of respiratory infection such as fever and pain, even during asthma exacerbations together with intensified asthma treatment in children and adults. ² Current Global Initiative for Asthma (GINA) guidelines are stating with evidence level A, that aspirin and other NSAIDs are generally not contraindicated in patients with asthma, unless there is a history of previous reactions to those medications. Nevertheless, it is still not very well studied whether NSAIDs facilitate or inhibit achievement of asthma control following exacerbations, and if they affect the speed of resolution of airway inflammation. Exacerbations of asthma are most often induced by common respiratory viruses including rhinovirus (RV), respiratory syncytial virus (RSV), bocavirus, influenza viruses, adenovirus, and others. ³ , ⁴ , ⁵ RV is responsible for up to 76% of exacerbations of wheeze in children and up to 83% of asthma attacks in adults. ³ , ⁴ , ⁵ Recurrent viral infections do not only cause acute disease and exacerbations of established disease, but they also contribute to the pathophysiology of early wheezing in children and the development of asthma. Prophylaxis of RSV‐induced bronchiolitis with palivizumab, an anti‐RSV monoclonal antibody, in late pre‐term infants decreased the risk of recurrent infant wheeze and the rate of parent‐reported asthma symptoms at 6 years of age, however, without any effect on lung function or doctor‐diagnosed asthma. ⁶ , ⁷  The majority of respiratory viruses are known to modify several major eicosanoid pathways, including the COX and the LOX pathways ⁸ (Figure 2). RV infection increases expression of 5‐lipoxygenase (5‐LOX), 5‐lipoxygenase activating protein (FLAP), and cyclooxygenase‐2 (COX‐2), as well as the production of prostaglandins E₂ (PGE₂) and PGD₂ by the respective isomerases in human bronchial epithelial cells, with higher levels in asthmatic patients than in controls. ⁹ In addition, cysteinyl leukotriene (cysteinyl‐LT) levels, 5‐LOX‐positive cells, and FLAP‐positive cells in bronchoalveolar lavage fluid are increased in humans upon RV infection and correlate with the emergence of upper respiratory symptoms. ¹⁰ Infection with RV affects airway mucosal barriers and also the peripheral blood and distant tissues. PGE₂ plays an important role in optimal antibody synthesis, as COX inhibitors reduce antibody release by plasma cells, also in case of viral infections. ¹¹ , ¹² Healthy individuals experimentally infected with RV showed a suppressed serum neutralizing antibody response when treated with aspirin or acetaminophen. ¹³ Significant increase in COX‐2 (PTGS2) expression and in COX‐derived metabolites is a hallmark of RSV ¹⁴ and influenza virus infection. ¹⁵ Pharmacologic inhibition of the COX pathway decreased RSV‐induced lung pathology, although this was not linked to a specific metabolite. ¹⁴ , ¹⁶ At a later stage of RSV infection there is an increase in LOX metabolites, which might promote appropriate resolution of infection‐induced inflammation. ¹⁷  This resolution is impaired in 5‐LOX and 15‐LOX knockout mice upon RSV or pathogenic influenza strain infection. Moreover, in mice lacking 5‐LOX, there is an upregulation of COX2 expression and aggravation of infection‐induced lung pathology. ¹⁴ , ¹⁶ During influenza A infection, newly generated PGE₂ leads to the inhibition of type I interferon (IFN) production, inhibition of macrophage apoptosis, and subsequent increase in virus replication. Ptges ^−/− knockout mice, which do not produce PGE₂, or wild‐type mice treated with PGE₂ type 2 (EP2) and type 4 receptor (EP4) antagonists demonstrated enhanced protection against lethal influenza infection. ¹⁸ PGE₂ also inhibits activation of the Nod‐like receptor family pyrin domain containing 3 (NLRP3) inflammasome in human monocytes and macrophages, and COX pathway blocking increases inflammasome activation and mature Il‐1β release. NLRP3 inflammasome activation can contribute to limiting viral replication at the early stages of infection, but in some instances, it may also lead to harmful hyperinflammation during late‐stage infection. ¹⁹  Much less is known about PGD₂ and antiviral responses. However, it was shown that in aging mice there is an increase in PGD₂ in the lungs which correlates with the impaired migration of respiratory dendritic cells (DC) to lymph nodes, diminished T‐cell responses, and more severe clinical disease in older mice infected with respiratory viruses, such SARS‐CoV‐1 and influenza. ²⁰ Blocking PGD₂ function enhances DC migration, T‐cell responses, and survival in the infected animals. ²⁰ In summary, prostaglandins and other COX‐dependent metabolites are involved in a complex way in the pathogenesis of respiratory viral infections and thus in virus‐induced exacerbation of asthma. Therefore, the use of NSAIDs to alleviate symptoms of viral infections in general population and in patients with asthma should be re‐evaluated with assessment of the effects of the timing of the administration, their selectivity, and the long‐term effects.

NSAIDs are used worldwide to alleviate symptoms of viral infections and inflammation, such as fever, cough, and pain. Since NSAIDs inhibit COX‐1 and COX‐2 and thus decrease the release of many downstream lipid mediators, such as various PGs, prostacyclin, and TXs, they have very broad effects on inflammation and immune responses, ranging from anti‐inflammatory, immunosuppressive, and anti‐thrombotic to pro‐resolving (Figure 2). ¹ , ²¹  Therefore, at the beginning of the COVID‐19 pandemic, there were several concerns and uncertainties about the effects of NSAIDs on SARS‐CoV‐2 infection and the course of COVID‐19. ²²  They were suspected to alter the expression of angiotensin‐converting enzyme 2 (ACE2), the main entry receptor for SARS‐CoV‐2 and/or modify viral replication. ²³ , ²⁴ , ²⁵ , ²⁶ In addition, they could be either harmful by impairing antiviral response and delaying resolution of inflammation ¹⁵ , ¹⁸ , ²⁷ , ²⁸ or be beneficial by dampening of hyperinflammation and cytokine storm ²⁹ , ³⁰ and preventing thrombosis ³¹ , ³² (Figure 3). Some of these concerns have now been addressed experimentally and epidemiologically, and several clinical trials have been initiated. Indeed, SARS‐CoV‐2 increases PTGS2 (COX‐2) gene expression in variety of cell lines, in mouse lungs and in primary human bronchial epithelial cells as well many eicosanoids and docosanoids are increased in the lungs of severe COVID‐19 patients. ³³ , ³⁴ , ³⁵ However, inhibition of the COX pathway by either ibuprofen (non‐selective COX1/COX2 inhibitor) or meloxicam (more selective COX‐2 inhibitor) did not change the expression of ACE2 in human cell lines (Calu‐3 or Huh7.5) in vitro or in lungs, kidney, heart, or ileum of mice in vivo. ³³ Similarly, both NSAIDs did not affect SARS‐CoV‐2 entry or its replication in the same human cell lines. ³³ Aspirin also did not affect ACE2 or transmembrane serine protease 2 (TMPRSS2) expression in human nasal epithelium. ³⁶  Meloxicam also did not prevent SARS‐CoV‐2‐infection‐induced weight loss in mice and did not change frequencies or activations status of alveolar macrophages, neutrophils, NK cells, Ly6C+ Mo/Mθ, CD4⁺ T cells, CD8⁺ T cells, and γδ T cells. However, meloxicam treatment decreased the amount of spike‐specific IgM and IgG antibodies and their neutralizing capacities as well as decreased infection‐induced levels of IL‐6, CCL2, GM‐CSF, CXCL10, IL‐2, and TNF‐α, suggesting that while meloxicam can impair humoral immune response against SARS‐CoV‐2 to some extent, it might also limit levels of proinflammatory cytokines. ³³ In contrast, naproxen, which is a non‐selective COX‐1/COX‐2 inhibitor, has been shown to bind to the nucleocapsid protein N of SARS‐CoV‐2, which led to inhibition of SARS‐CoV‐2 replication in VeroE6 cells and primary human bronchial epithelial cells and protected epithelium against SARS‐CoV‐2‐induced barrier damage. ³⁷  There were no analogous effects in similar experiments with paracetamol (acetaminophen, which may affect PG production in the brain or may act via its metabolite on the cannabinoid receptors), ³⁸ , ³⁹ or celecoxib (selective COX‐2 inhibitor). ³⁷  Naproxen is currently examined in the clinical trial in COVID‐19 (eudract_number:2020‐001301‐23; accessed 11.06.2021). So far, in various patient groups, it has been shown that usage of NSAIDs does not lead to the worse COVID‐19 outcomes. ⁴⁰ In a retrospective study of 403 confirmed cases of COVID‐19, there were no differences in terms of mortality rate or need for respiratory support between patients who were taking ibuprofen or those who did not take any NSAIDs. ⁴¹ It was also confirmed in the large prospective cohorts that either acute or chronic use of NSAIDs was not associated with worse COVID‐19 outcomes. ⁴² , ⁴³ , ⁴⁴ It was even shown that in patients, who were treated with aspirin or other NSAIDs due to the cardiovascular diseases, positive aspects of such therapies have been noted, including reduction of COVID‐19 mortality. ³¹  This clinical observation is further supported by a study on COX‐2 induction and PGE₂ overproduction in the human lung infected by SARS‐CoV‐2. ⁴⁵ However, further basic in vitro, in vivo, and large clinical studies assessing the influence of NSAIDs on the pathogenesis and treatment of COVID‐19 are still greatly needed.

All clinically available LTRAs (montelukast, zafirlukast, and pranlukast) act on the cysteinyl leukotriene type 1 receptor (CysLT₁) and by competitive antagonism at this receptor are believed to be responsible for the control of airway inflammation, bronchoconstriction, and remodeling. ⁴⁶ , ⁴⁷ , ⁴⁸ However, anti‐inflammatory activity of LTRAs independent of CysLT₁ antagonism has been suggested. LTRAs reduced the eosinophil protease activity ⁴⁹ and inhibited TNFα‐ ⁵⁰ or UDP‐mediated ⁵¹ , ⁵² cytokine expression, as well as NF‐κB activation in human mononuclear ⁵³ or epithelial cells ⁵⁴ through processes that appear to be distinct from CysLT₁ antagonism. Although mechanisms of these non‐CysLT₁‐related LTRA activities are not fully understood, concentration‐dependent inhibition of distinct receptors such as P2Y1, P2Y2, P2Y6, and GPR17 by LTRAs have been reported, ⁵¹ , ⁵² , ⁵⁵  suggesting, at least for P2Y receptor, a non‐competitive mechanism of action. Interestingly, it was also shown that LTRAs may have a potent inhibitory effect on 5‐LOX activity (i.e., LT production) ⁵² , ⁵⁶ and transport of LTs by the multidrug resistance protein ABCC4, ⁵⁷  suggesting a much broader mechanism of action for these drugs than previously suspected. Indeed, non‐CysLT₁‐related mechanisms of LTRA might represent another level of variability in the response to treatment in patients with asthma and allergy. Some of these activities may be compound‐specific or may depend on drug concentration (most non‐CysLT₁‐related effects required micromolar drug concentrations ⁵¹ , ⁵² , ⁵³ in contrast to nanomolar levels needed for CysLT₁ antagonism ⁵⁸ , ⁵⁹ , ⁶⁰ ) or may depend on the presence of a particular inflammatory pathway in patients with asthma (allergy), and therefore, clinically significant effects of treatment may be observed in some, but not all, treated patients. It should be emphasized that initial clinical interventional studies of montelukast in asthma used doses up to 200 mg a day ⁶¹ , ⁶²  showing greater lung function improvement than in subsequent studies using recommended dose of 10 mg, ⁶³  suggesting that higher doses of currently known LTRAs or new compounds derived from this class of drugs may represent a novel strategy for finding more efficient therapy. The demonstration that the bronchoconstrictive actions of LTE₄ in asthma are solely mediated by the CysLT₁ receptor further supports that effects on other targets than the CysLT receptors may take place ⁶⁴

Heterogeneous effects of LTRA therapy in asthma and allergic diseases have been reported in many studies. Although some genetic ⁶⁵ , ⁶⁶ and acquired factors have been suggested, ⁶⁷ other reasons for this heterogeneity remain unclear. While currently no clinical characteristics or laboratory assay can reliably predict responses to LTRAs, the most plausible biomarker that could potentially serve as response predictor to LTRAs seems LTE₄ production. Urinary LTE₄ (uLTE₄) is a biomarker of total body cysteinyl‐LT production, ¹ associated with Type 2 asthma, asthma severity, exacerbations, and NERD. ⁶⁸ , ⁶⁹ Increased uLTE₄ to fractional exhaled nitric oxide (LTE₄: Fe_NO) ratio has been suggested to predict favorable response to LTRA therapy (montelukast) in asthmatic children, ⁷⁰ , ⁷¹ but these observations have not been confirmed in adult patients. There is a considerable amount of evidence supporting the concept that some patients or clinical phenotypes seem sensitive to LTRAs, especially in a real‐life setting, due to enhanced cysteinyl‐LT production, better adherence to oral therapy or oral drug delivery. LTRAs have proven to be particularly effective in exercise‐induced asthma, ⁷² asthma associated with allergic rhinitis, ⁷³  NERD, ⁷⁴ viral‐induced wheezing episodes, ⁷⁵ and patients having difficulties with inhaled therapy such as children and elderly. ⁷⁶ , ⁷⁷  Cigarette smoking while inhibiting steroid anti‐inflammatory responses ⁷⁸ increases cysteinyl‐LT production, ⁷⁹  leading to a greater response to montelukast in smokers with asthma, suggesting that LTRA could be more effective in treating such individuals. ⁸⁰ In fact, asthmatic patients with smoking history above 11 pack‐years showed more benefit with montelukast treatment than inhaled steroids. ⁸¹ Obesity is another potential risk factor for asthma development and efficacy of treatment. Interestingly, higher body mass index (BMI) is associated with increased LT production in asthmatics ⁸² and as therapeutic response to inhaled corticosteroids decreases with increasing BMI, response to montelukast remains unaffected, ⁸³  suggesting LTRA therapy to be more effective in obese patients. The response to LTRA may also be associated with sex differences. The existence of a sex bias in LT biology is already suggested by the fact that many LT‐related diseases including asthma, allergic rhinitis, rheumatoid arthritis, or NERD have a higher occurrence in women compared to men, pointing to more pronounced pathophysiological roles of LTs in females. ⁸⁴ , ⁸⁵ Furthermore, several observations suggest that female sex is associated with higher LT biosynthesis, while androgens seem to exert a suppressing role on LT formation both in vitro and in vivo. ⁸⁶ , ⁸⁷ , ⁸⁸ , ⁸⁹ , ⁹⁰ Although the clinical significance of these data is still to be confirmed, in a small prospective cohort study, montelukast showed superior effects on symptoms and lung function in women compared to men, ⁹¹ while a tendency for a better response to montelukast was evident in girls exposed to tobacco smoke. ⁹²

Due to the involvement of complement, coagulation, and inflammation in COVID‐19, ³² , ⁹³ , ⁹⁴ anti‐inflammatory drugs have gained great interest as disease modifiers (Figure 3). ⁹⁵ Already at the beginning of the COVID‐19 pandemic, researchers suggested the use of the LTRA montelukast for treatment of COVID‐19. ⁹⁶ , ⁹⁷ , ⁹⁸  The reason for this early interest in LT‐modifying drugs was on the one hand related to the viral cell entry via ACE2 receptors and the known inhibitory effect of montelukast on bradykinin‐related airway response ⁹⁹ and on the other hand to the fact that patients with severe COVID‐19 develop an overwhelming state of inflammation that has been labeled COVID‐19 cytokine storm syndrome (CSS). ³² , ¹⁰⁰ , ¹⁰¹  Moreover, the most important cause of death in COVID‐19 was recognized as the progressive respiratory failure with limited response to treatment together with hyperinflammation and hypoxia, quite similar to a severe acute respiratory distress syndrome (ARDS), which has been demonstrated to be characterized by an elevated level of LTs. ¹⁰² Of note, high levels of LTE₄ have been detected in bronchoalveolar lavage (BAL) of hospitalized patients with severe COVID‐19, ³⁴ as well as there is a shift in serum eicosanoids into the increase of 5‐LOX products in such patients. ³⁵ Indeed, specific benefits of montelukast, or other LTRAs, have been suggested in the situation of hyperinflammation and massive cytokine release ¹⁰³ to reduce elevated levels of LPS‐induced IL‐6, TNF‐α, and MCP‐1 production in the peripheral blood MNCs of patients with asthma, ¹⁰⁴ as well as to reduce levels of many cytokines and chemokines (IL‐4, IL‐5, IL‐1β, TNF‐α, RANTES, and IL‐8) in nasal mucosa ¹⁰⁵ possibly due to modulation of TNF‐α‐stimulated IL‐8 expression through changes in NF‐κB p65‐associated histone acetyltransferase activity. ⁵⁰ In addition to its anti‐inflammatory properties in humans, in silico studies also suggested, but still to be demonstrated, a direct antiviral effect by showing a high‐affinity binding of montelukast to the terminal end of the virus’ main protease enzyme needed for viral protein assembly. ¹⁰⁶

Thus, with increasing understanding of disease mechanisms, LTRAs have been also considered for treatment of COVID‐19. Indeed, in a small retrospective study on COVID‐19‐hospitalized subjects, patients receiving montelukast had fewer episodes of confirmed COVID‐19 or experienced significantly fewer events of clinical deterioration compared to patients not receiving montelukast. ¹⁰⁷ , ¹⁰⁸  These lipid mediators might not only contribute to inflammation and lung pathologies associated with COVID‐19, but can also be involved in thrombosis, fibrosis, neuronal damage, and cardiovascular disease. ⁹⁷ , ¹⁰⁹ , ¹¹⁰ Interestingly, we have recently demonstrated that montelukast inhibits platelet activation and microvesicles release induced by plasma from COVID‐19 patients, as well as the formation of circulating monocyte‐ and granulocyte‐platelet aggregates. ¹¹¹ All these data suggest the repurposing of montelukast as a possible auxiliary treatment for COVID‐19 syndrome. Accordingly, since May 2020 a series of clinical trials involving montelukast have been registered (https://clinicaltrials.gov). However, not only antagonism of the CysLT receptors could be beneficial for patients with COVID‐19, but interventions targeting LT biosynthesis, using, for example, Zileuton, might represent promising targets, specifically at the turning point from a mild to critical disease course. ¹¹²

In sensitized subjects, PGD₂ is initially released by allergen‐triggered mast cells and plays a key role in the sequelae of the allergic response. Its proinflammatory effects are mediated through the interaction with G‐protein‐coupled receptors (GPCR): DP1, thromboxane (TP), and chemoattractant‐homologous receptors (CRTH2 or DP2). ¹¹³ Apart from its broncho‐ and vaso‐active properties in allergic airway disease, PGD₂ also acts as an important link between the allergen‐induced early (EAR) and late phase allergic response (LAR) through the interaction with the DP2‐receptors on key effector cells. DP2‐receptors are expressed on immune (ILCs, Th2), inflammatory (eosinophils, basophils), and structural (epithelial) cells and involved in the recruitment and activation of these cells as well as the subsequent release of Th2‐cytokines during the LAR. ¹¹³ , ¹¹⁴ , ¹¹⁵ , ¹¹⁶  Therefore, DP2 (CRTH2) antagonists have been initially aimed for the treatment of allergic airway disease (allergic rhinitis, asthma). ¹¹⁷ , ¹¹⁸

In two proof‐of‐concept studies in (unphenotyped) allergic asthmatics, DP2 (CRTH2) antagonists (timapiprant and setipiprant, respectively) showed only modest reduction (approx. 25%) in the allergen‐induced LAR ¹¹⁹ , ¹²⁰ while no convincing effects were observed on the allergen‐induced changes in T2 biomarkers (blood eosinophils, FeNO) ¹²⁰ with only a minimal reduction in sputum eosinophils post‐allergen. ¹¹⁹ In addition, there was no decrease in the EAR in either study. The (relative) lack of protection against allergen‐induced airway responses may (partly) consist with the fact that even with effective DP2‐blockade, an allergen‐triggered mast cell (lacking DP2) ¹²¹  mediator release (histamine, PGD₂, cysteinyl‐LTs) may still occur which is capable of causing an EAR and/or an LAR ¹²² , ¹²³ and therefore, especially in allergic asthma, a combined blockade of, for example, DP2 ± DP1 ± TP ± cysteinyl‐LT‐R might provide a superior protection.

In line with this reasoning—and despite prior evidence of superior efficacy in phase 2B studies of patients with an allergic (T2‐) profile (atopy ± eosinophils ≥250/mcL) ¹²⁴ , ¹²⁵ —several DP2 (CRTH2) antagonists (e.g., setipiprant, fevipiprant) failed in phase 3 clinical trials of allergic airway disease. ¹²⁵  More recently, DP2‐blockade has been associated with the reduction in airway smooth muscle mass by decreasing airway eosinophilia and the recruitment of myofibroblasts and fibrocytes. ¹²⁶  Therefore, with several clinical trials still ongoing, (add‐on) DP2‐blockade may show efficacy in more severe T2 asthma ¹²⁷ and related conditions based on its anti‐inflammatory and disease‐modifying potential. ¹²⁶ , ¹²⁸

Respiratory viruses (e.g., RSV) represent other important triggers of chronic inflammatory airway disease capable of activating the PGD₂/DP2 receptor‐mediated pathway, thereby eliciting a “non‐allergen‐induced” T2‐immune response through airway epithelial cells and innate immune cells. ¹²⁹ Indeed, RSV has been associated with upregulation of the PGD₂/DP2 pathway and increased PGD₂ levels both in experimental and in clinical studies, while DP2‐blockade alone or combined with DP1 agonism showed protective potential in preclinical studies. ¹³⁰  Therefore, selective targeting of PGD₂ receptors has been postulated to protect against respiratory viral infections, and more recently, including SARS‐CoV‐2. ¹³¹ Presently, this hypothesis awaits clinical evidence. In addition, the potent bronchoconstrictive actions of PGD₂ and other constrictive prostanoids in human airways call for trials with TP receptor antagonists in patients with asthma ¹³²

The human endogenous cannabinoid system (ECS) is involved in many physiological processes. It consists of the cannabinoid receptors (CBRs), the endogenous ligands (anandamide (AEA) and 2‐arachidonoylglycerol (2‐AG)), and the proteins related to their synthesis and degradation. ¹³³ Cannabinoid receptor 1 (CB1) and 2 (CB2) are the main CBRs. CB1 is largely expressed in the central nervous system but also in peripheral tissues and immune cells. CB2 is mainly expressed in immune cells but also in other cell types such as progenitor neurons. ¹³⁴ , ¹³⁵  The biosynthesis and inactivation of endocannabinoids involve several enzymes. AEA can be synthetized by the hydrolysis of its precursor N‐acyl‐phosphatidylethanolamine (NAPE) by NAPE‐hydrolyzing phospholipase D (NAPE‐PLD). Alternative pathways involving other phospholipases have been described. AEA can be also obtained from the reverse reaction of fatty acid amide hydrolase (FAAH) by the conjugation of ethanolamine and AA. The main pathway for the synthesis of 2‐AG is the hydrolysis of AA‐containing 1,2‐diacylglycerol (DAG) species by DAG lipase‐α or β. ¹³³ , ¹³⁶ Endocannabinoids are rapidly metabolized. The hydrolysis of AEA is mediated by FAAH, whereas 2‐AG is hydrolyzed by monoacylglycerol lipase (MGL). The resulting products of AEA and 2‐AG degradation are AA, and ethanolamine and glycerol, respectively (Figure 4).. ¹³³ , ¹³⁶ Moreover, endocannabinoids are susceptible to be metabolized by eicosanoid biosynthetic enzymes including COX2, 12‐LOX, 15‐LOX, and P450. ¹³⁶ FAAH and MGL are considered promising therapeutic targets for the treatment of several disorders, including inflammation. Pharmacological inhibitors of FAAH and MGL increase the levels of AEA and 2‐AG, prolonging their anti‐inflammatory effects, and decrease AA levels. ¹³⁷ , ¹³⁸  The NSAIDs that inhibit COX2 also enhance endocannabinoid levels and reduce AEA‐ and 2‐AG‐derived prostaglandins. ¹³⁹ , ¹⁴⁰ In addition, some NSAIDs such as ibuprofen can influence the endocannabinoid levels by inhibiting FAAH (Figure 4).

The role of cannabinoids in allergic diseases is still a bit controversial. ¹⁴¹ Sukawara et al demonstrated that endocannabinoids limited mast cell maturation and activation in human airway mucosa and skin through CB1. ¹⁴² , ¹⁴³  Tetrahydrocannabinol (THC) and cannabidiol (CBD) attenuated airway allergic inflammation, decreased cytokine production, cell infiltration, mucus secretion, and bronchial hyperresponsiveness in mice. ¹⁴⁴ , ¹⁴⁵ , ¹⁴⁶ Similarly, the synthetic agonist CP55,940 induced lung protection in ovalbumine (OVA)‐induced asthma guinea pig models via CB1 and CB2. ¹⁴⁷ In keratinocytes, CB1 prevented transepithelial water loss and skin inflammation, cell infiltration and cytokine production in atopic dermatitis mouse model. ¹⁴⁸ Anandamide and different CB1 agonists also accelerated skin barrier recovery and reduced proinflammatory cytokine production and cell recruitment. ¹⁴⁹ , ¹⁵⁰ Several cannabinoids have also shown a protective role in allergic contact dermatitis by reducing inflammatory responses. ¹⁵¹ , ¹⁵² , ¹⁵³ CB1 activation may also induce bronchodilation in the airways. ¹⁴⁷ , ¹⁵⁴ In human bronchial epithelial cells, the synthetic agonist WIN55212‐2 restored the epithelial barrier disruption induced by RV. ¹⁵⁵ In addition, WIN55212‐2 decreased the immediate anaphylactic reaction in a mouse model of peanut allergy and promoted the generation of allergen‐specific regulatory T cells. ¹⁵⁶ Currently, different studies suggest the therapeutic potential of cannabinoids in COVID‐19 pandemic. ¹⁵⁷ , ¹⁵⁸ , ¹⁵⁹ In contrast, Frei et al showed that CB2 activation enhanced migratory responsiveness of eosinophils in an OVA‐asthma mouse models. ¹⁶⁰ Accordingly, the lack of CB2 decreased allergic inflammation in asthma and dermatitis mouse model. ¹⁶¹  This result correlated with increased number of NK cells and reduced number of ILC2s in the lung of CB2 knockout mice, demonstrating that NK cells are negative regulators of ILC2s. ¹⁶² Interestingly, it has been described that mRNA expression levels of CB1 are upregulated in tonsils and peripheral blood of patients with allergic rhinitis, atopic dermatitis, and food allergy, but the functional relevance remains unknown. ¹⁶³  These studies suggest that the ECS could be explored as a potential therapeutic target in the treatment of asthma, allergic and skin diseases, and viral infections.

Ample evidence from clinical trials showing effectiveness of drugs targeting T2‐inflammation (targets include IgE and the cytokines IL5, IL4, and IL13) on asthma exacerbations, as well as improvements in symptoms and disease severity in chronic rhinosinusitis with nasal polyps (CRSwNP), ¹⁶⁴ , ¹⁶⁵ underscored the involvement of T2‐inflammation in these conditions. ¹⁶⁶ , ¹⁶⁷ As mentioned above, the majority of asthma exacerbations are precipitated by respiratory viruses (esp. RSV and RV), ¹⁶⁸ while in sensitized subjects, allergen exposure may enhance virally triggered exacerbations due to synergistic interaction through joint mechanisms including the T2‐inflammatory pathway. ¹⁶⁸ , ¹⁶⁹ , ¹⁷⁰ , ¹⁷¹ Both viral and allergen‐triggered pathways include several inflammatory and immune (effector) cells, such as mast cells, basophils, Th2 cells, ILCs, macrophages, neutrophils, and eosinophils. Many of these cells are capable of releasing eicosanoids upon activation and/or possess one or more eicosanoid receptors, ¹⁷² thus contributing to the exacerbation and its sequelae (e.g., bronchoconstriction, airway inflammation, and bronchial hyperresponsiveness). ¹⁷³ In CRSwNP, the T2‐inflammatory pathway is also triggered by several stimuli such as viruses, bacteria, and allergens, which stimulate inflammatory cell‐ and cytokine‐mediated pathomechanisms in the nasal and paranasal mucosa. ¹⁶⁷

Although in vitro data indicate that biologicals may influence eicosanoid pathways in mast cells and basophils, ¹⁷⁴  so far there are no published data on direct effects of T2‐targeted biologicals on the synthesis or release of eicosanoids in humans in vivo (Figure 5). However, it makes sense that, by blocking pathways and cells (esp. mast cells, basophils, eosinophils, and neutrophils) responsible for the release of these proinflammatory mediators, may consequently also reduce eicosanoid levels. In addition, previous evidence from clinical studies in asthma showed (partial) reduction of both allergen‐ and virus‐induced airway responses and asthma exacerbations by selective eicosanoid antagonists. ⁴⁸ , ¹¹⁹ , ¹²⁰ , ¹⁷⁵ , ¹⁷⁶ , ¹⁷⁷ Besides, clinical studies on biologicals in CRSwNP also included a representative cohort of patients with NERD and also found a good clinical response ¹⁶⁴ and a reduced T2‐biomarker profile in this subpopulation. ¹⁷⁸ However, so far there are no data on the direct effect of T2 biologicals on the individual eicosanoids nor head‐to‐head studies comparing biologicals with selective eicosanoid blockers or combinations.

NERD, also called AERD‐aspirin‐exacerbated respiratory disease or AIA‐aspirin‐intolerant asthma, is a phenotype of asthma recognized in 5 to 25% asthmatics. It is characterized by a non‐immunological hypersensitivity to low doses of NSAIDs and a cross‐reactivity (a multi‐responder phenotype). Profound changes in biosynthesis of eicosanoids comprise overproduction of cysteinyl‐LTs, excreted in urine as LTE₄. ¹⁷⁹ , ¹⁸⁰ Some patients have higher excretion of LTE₄ also during a stable period of NERD. ⁷⁴ , ¹⁸¹ , ¹⁸² , ¹⁸³ It is debatable, which cells produce cysteinyl‐LTs in NERD. Since overproduction of PGD₂ and increase of histamine concentration accompany symptoms of NERD, these could be mast cells. However, eosinophils in NERD overexpress leukotriene C4 synthase (LTC4S), thus can contribute to the symptoms and concurrent release of eosinophils cationic protein was observed. PGE₂ plays a key role in NERD, where both decreased production of PGE₂ and reduced EP₂ expression were observed. ¹⁸⁰ , ¹⁸⁴  When PGE₂ is further decreased, it leads to mast cell activation and bronchoconstriction because it removes the stabilizing effect of PGE₂ on mast cell mediator release. ¹³² Accordingly, inhalation of PGE₂ before aspirin challenge prevented reduction in pulmonary function and mast cell activation. ¹⁸⁵ However, inhibition of PGE₂ biosynthesis by NSAIDs is difficult to measure, since this prostaglandin is produced by most cells of the body. Interestingly, patients with NERD have also an imbalance in pro‐resolving lipoxin A₄ (LXA₄) that may contribute to the increased severity of this particular asthma endotype. ¹⁸⁶  The minimal dose triggering bronchial constriction and extra bronchial symptoms (cutaneous flush, nasal obstruction, irritations of conjunctiva) varies across patients, but generally it reflects NSAID potency to inhibit cyclooxygenase‐1 isoenzyme (COX‐1). ¹⁸⁷ , ¹⁸⁸ Highly selective inhibitors of COX‐2 like coxibs (e.g., celecoxib, etoricoxib) ¹⁸⁹ are well tolerated in most NERD patients, whereas preferential COX‐2 inhibitors (nimesulide, meloxicam) can trigger symptoms at high doses ¹⁹⁰ (Table 1). Diclophenac, ketorolac, ibuprofen, naproxen, indomethacin, or pyrazolone derivatives inhibit both COX‐2 and COX‐1, ¹⁹¹ therefore are contraindicated in NERD. Acetylsalicylic acid is more potent inhibitor of COX‐1 than COX‐2. ¹⁸⁹ , ¹⁹²  This was the first NSAID ever reported to trigger symptoms in asthmatics. Paracetamol (acetaminophen), with an unclear effects on prostanoids biosynthesis including possible inactivation of brain cyclooxygenases by a non‐substrate mechanism, is tolerated by the vast majority of NERD patients unless given in very high doses. ³⁹

Most of the information available on the role of eicosanoids in allergy and related diseases concerns NERD. ¹⁹³ , ¹⁹⁴  This fact can be explained because it was the first clinical phenotype in which a link between NSAIDs pharmacological activity and the inhibition of PGE₂ synthesis by blocking COX‐1 and the subsequent increase in cysteinyl‐LTs release was established. ¹⁹⁵  Nevertheless, some data are also available for cutaneous NSAID‐induced cross‐hypersensitivity. Thus, increased LTE₄ and 9α,11β‐PGF₂ urinary levels have been described for NERD ¹⁹⁶ , ¹⁹⁷ , ¹⁹⁸ , ¹⁹⁹ and for NSAID‐induced acute urticaria/angioedema (NIUA). ¹⁹⁹

For NSAID‐exacerbated cutaneous disease (NECD), contrasting results have been found regarding eicosanoids levels at basal state. Thus, Di Lorenzo et al. did not report baseline differences for LTE₄ in patients with chronic urticaria and hypersensitivity to acetylsalicylic acid (ASA, aspirin) or food additives, ²⁰⁰ and no variations at basal state were reported for LTE₄ and 9α,11β‐PGF₂ by two other independent studies. ¹⁹⁷ , ¹⁹⁹ However, Mastalerz et al. reported increased LTE₄ levels in NECD patients with a positive aspirin challenge with respect to those with a negative aspirin challenge, and with no changes found for 9α,11β‐PGF₂. ²⁰¹ It has been recently published that NIUA and NECD showed similar increased levels in both LTE₄ and 9a,11b‐PGF₂ within the first 3 hours following a positive aspirin challenge; however, after this time interval, these mediators showed different behaviors, being such levels long‐lasting in NECD. ¹⁹⁹ In spite of these differences being not statistically significant, the reasons explaining the existence of these particular profiles are at present unknown although they may be due to the presence of additional factors in NECD, which could include sensitization to autoantibodies or the existence of histamine‐releasing factors. ¹⁹⁹

Data on the role of eicosanoids beyond NSAIDs‐hypersensitivity are scarce. However, a potential role for cysteinyl‐LTs was proposed in adverse reactions to non‐ionic contrast media. Thus, iopromide and iotrolan induced a significant increase of cysteinyl‐LTs in vivo, with no changes in preformed mediators levels. ²⁰² However, a previous study showed the heterogeneity of the effects of contrast media on mediator release, showing an increase in histamine and tryptase release from different human cells without changes in LTE₄ or PGD₂ levels. ²⁰³

NSAIDs, LT modifiers, and biologicals are used every day in clinical practice in treatment of viral infections and common respiratory or allergic diseases. Although a significant progress has been made in our understanding how these medications act and how they affect eicosanoid pathways, there are still no sufficient data available to fully address all issues important for prediction of their activities affecting immune response and estimation of their clinical efficacy. This consensus report summarizes up to date knowledge in this complex area and identifies major knowledge gaps and unmet needs to be addressed in the future.

MiSo has received research grants from the Swiss National Science Foundation, GlaxoSmithKline, Novartis, and AstraZeneca speaker fee. ZD acted as a director respiratory/allergy for QPS‐NL (2012‐2020); she has received consulting fees from ALK, Antabio, GSK, Sanofi‐Genzyme, QPS‐NL; has participated in the speaker's bureaus of Boehringer Ingelheim, Sanofi‐Genzyme; she serves in EUFOREA as an Asthma Expert Panel Chair. OsPa has received fees for lectures and/or participation in Advisory Boards from Allergy Therapeutics, Amgen, AstraZeneca, Diater, GSK, Inmunotek SL, Novartis, Sanofi‐Genezyme, Regeneron, and Stallergenes. OP has received research grants from Inmunotek SL and Novartis SL. CA has received research grants from the Swiss National Science Foundation, Christine Kühne‐Center for Allergy Research and Education, European Commission Horizon's 2020 Framework Programme “CURE,” Novartis Research Institutes, GlaxoSmithKline and AstraZeneca. He took part in the advisory board and received research grants from GlaxoSmithKline, Sanofi/Regeneron, SciBase, and Novartis. He is the editor in chief of Allergy. LOM is a consultant to PrecisionBiotics and has received research funding from GSK and Chiesi. LOM has participated in speaker's bureau for Nestle, Nutricia, Reckitt, and Abbott. MJ has received consulting fees (ALK‐Abello, Stallergenes‐Greer, Takeda, Zentiva); honoraria for lectures, presentations (ALK‐Abello, Stallergenes‐Greer, Takeda, Zentiva, Mundipharma, AstraZeneca, SOBI, Chiesi, CSL Behring, Novartis, Benela, Pfizer, Viatris); support for attending meetings and/or travel (ALK‐Abello, Stallergenes‐Greer, Takeda, Novartis, Sanofi Pasteur) and honoraria for participation on Advisory Boards (ALK‐Abello, Stallergenes‐Greer, Chiesi, Novartis, SOBI, Pfizer, Sanofi Genzyme/Pasteur). OlPh reports grants and personal fees from ALK‐Abelló, grants and personal fees from Allergopharma, grants and personal fees from Stallergenes Greer, grants and personal fees from HAL Allergy Holding B.V./HAL Allergie GmbH, grants and personal fees from Bencard Allergie GmbH/Allergy Therapeutics, grants and personal fees from Lofarma, grants from Biomay, grants from Circassia, grants and personal fees from ASIT Biotech Tools S.A., grants and personal fees from Laboratorios LETI/LETI Pharma, personal fees from MEDA Pharma/MYLAN, grants and personal fees from Anergis S.A., personal fees from Mobile Chamber Experts (a GA2LEN Partner), personal fees from Indoor Biotechnologies, grants and personal fees from GlaxoSmithKline, personal fees from Astellas Pharma Global, personal fees from EUFOREA, personal fees from ROXALL Medizin, personal fees from Novartis, personal fees from Sanofi‐Aventis and Sanofi‐Genzyme, personal fees from Med Update Europe GmbH, personal fees from streamedup! GmbH, grants from Pohl‐Boskamp, grants from Inmunotek S.L., personal fees from John Wiley and Sons, AS, personal fees from Paul‐Martini‐Stiftung (PMS), personal fees from Regeneron Pharmaceuticals Inc., personal fees from RG Aerztefortbildung, personal fees from Institut für Disease Management, personal fees from Springer GmbH, personal fees from AstraZeneca, personal fees from IQVIA Commercial, personal fees from Ingress Health, outside the submitted work; and member of EAACI Excom, member of ext. board of directors DGAKI; coordinator, main or co‐author of different position papers and guidelines in allergology and allergen‐immunotherapy. SED has received consultation fees from AZ, GSK, Merck, Novartis, Regeneron, Sanofi, Teva; has participated in the speaker's bureaus of AZ, GSK, Sanofi. Other authors do not have any COI to declare.

MiSo and GW led the Task Force and supervised the project. MiSo, GER, ZD, EU, JS, ZL, FS, AA, OP, CA, LOM, MJ, OP, MJT, MS, SED, and GW conceptualized the project, reviewed the literature, wrote different sections of the manuscript, and edited the manuscript.