Asthma and leukotrienes: antileukotrienes as novel anti-asthmatic drugs

Phospholipase A₂

Phospholipase A₂ (PLA₂) comprises a family of enzymes that can be divided into several groups depending on molecular weight, localization, structure, fatty acid chain specificity and calcium dependency [ 2, 4, 5]. Several low molecular mass PLA₂s (14–18 kDa) are widely distributed in mammalian tissues. These PLA₂s display essentially no fatty acid chain specificity and are dependent on millimolar calcium concentrations for activity. They are described as secretory PLA₂ (sPLA₂) since they are stored in granules to be secreted upon activation of the cell. In recent years much interest has focused on a high molecular weight calcium-dependent cytosolic PLA₂ (85 kDa cPLA₂), which is expressed in a variety of cells found, for example, in lung tissue [ 2, 6, 7]. This particular enzyme is described as cPLA₂, and preferentially hydrolyses phospholipids with arachidonic acid bound at the sn-2 position. During activation, cPLA₂ translocates to the nuclear envelope and endoplasmatic reticulum in a calcium-dependent manner [ 8, 9]. Intracellular calcium-independent PLA₂ (iPLA₂) has also been isolated from leucocytes [ 10]. Subsequently, cloning and functional expression of hamster, murine and human iPLA₂ have been demonstrated [ 11–14].

Since PLA₂ initiates the synthesis of eicosanoids, much work has also been directed towards understanding which PLA₂(s) is involved in the synthesis of eicosanoids. Recently, studies in transgenic mice deficient in cPLA₂ suggested that this enzyme played a pivotal role in eicosanoid synthesis in that particular species [ 15, 16]. However, this particular strain of mice is also congenitally defective in sPLA₂ activity [ 17], which makes it difficult to draw general conclusions about the relative role of cPLA₂ in eicosanoid synthesis. There is apparently evidence for species and cell variation with respect to the participation of distinct PLA₂s in leukotriene and prostaglandin synthesis. For example, antisense-oligonucleotides directed against cPLA₂ were found to inhibit prostaglandin synthesis but did not affect LTC₄ synthesis in human monocytes [ 18]. Studies on PLA₂ and prostaglandin synthesis also suggest that cPLA₂ and sPLA₂ may act in concert to release arachidonic acid in murine mast cells [ 19]. Furthermore, iPLA₂ appears also to be involved in leukotriene synthesis in human granulocytes [ 20]. It is therefore possible that the release of arachidonic acid, destined for leukotriene synthesis, is regulated by ‘cross-talk’ between different cellular phospholipases.

5-Lipoxygenase and FLAP

The enzyme 5-lipoxygenase is restrictively expressed in the human body and predominantly found in granulocytes, monocytes, macrophages, mast cells and B-lymphocytes [ 1, 21, 22]. The active site of 5-lipoxygenase contains a non-heme binding iron that is essential for the catalytic activity and the enzyme is calcium-dependent and requires ATP for maximal activity [ 23, 24]. Complementary DNA for 5-lipoxygenase has been cloned from human placenta and HL-60 cells as well as from various murine sources [ 21, 24]. All cDNAs coded for a protein with a calculated molecular weight of 78 kDa. However, no defined ATP-binding motif or calcium-binding domains were identified in the amino acid sequences. The human 5-lipoxygenase gene is located on chromosome 10 and comprises more than 82 kb DNA, consisting of 14 exons and 13 introns [ 25]. The 5-lipoxygenase promoter contains multiple GC boxes, potential Sp1 and Egr-1 binding sites, but lacks TATA and CCAAT elements [ 26, 27]. These transcriptional elements do not explain the restricted expression of 5-lipoxygenase since the enzyme is primarily found only in bone marrow-derived cells. Therefore, other unidentified transcriptional elements are in operation. Recently, naturally occurring mutations in the 5-lipoxygenase promoter region have been reported [ 28]. Genomic DNA was isolated from both healthy subjects and patients with asthma, including six patients with aspirin-intolerant asthma. The prevalence of mutations was similar in both groups. Reporter gene activity directed by any of the mutant promoter forms was less effective than the wild type. It is therefore uncertain if these observed mutations in the promoter region of 5-lipoxygenase in asthma patients could explain or contribute to the enhanced biosynthesis of leukotrienes found in certain patients with asthma.

The cellular biosynthesis of leukotrienes involves an 18 kDa membrane-bound protein named FLAP (five-lipoxygenase activating protein) [ 24]. It is an arachidonate-binding protein, which stimulates the conversion of cellular arachidonic acid by 5-lipoxygenase [ 29–31]. The human FLAP gene spans > 31 kb of DNA and consists of five exons divided by four introns, and is located on chromosome 13q12. The FLAP promoter region contains a possible TATA box and AP-2 and glucocorticoid receptor binding sites [ 32]. The transcriptional regulation of 5-lipoxygenase and FLAP are different since U937 cells and resting T-lymphocytes express FLAP but not 5-lipoxygenase [ 33, 34].

Leukotriene A₄ hydrolase

Leukotriene A₄ hydrolase is a cytosolic 69 kDa enzyme that catalyses the conversion of LTA₄ to LTB₄ [ 35]. The enzyme is a zinc metalloenzyme that also exhibits peptide cleaving activity [ 36–38], but the physiological relevance of this is at present unclear. Several reports suggest the presence of isoenzymes of LTA₄ hydrolase [ 39, 40]. The gene contains > 35 kb DNA and is localized to region 22q of chromosome 12 [ 41]. The promoter showed consensus sequences for AP-2 and two xenobiotic-response elements. In contrast to 5-lipoxygenase, LTA₄ hydrolase is widely distributed and has been found in almost all mammalian cells. One possible explanation for this discrepancy could be that LTA₄ hydrolase primarily is involved in leukotriene synthesis in cells expressing 5-lipoxygenase, whereas peptide cleavage occurs in cells which do not produce leukotrienes. Alternatively, LTA₄ hydrolase might play a role in the transcellular metabolism of leukotrienes (see below).

Leukotriene C₄ synthase

Leukotriene C₄ synthase is an integral membrane 18 kDa protein, catalysing the final step in LTC₄ formation [ 42–44]. The enzyme catalyses the conjugation of glutathione with LTA₄ and is active as a homodimer. Leukotriene C₄ synthase activity is also detected in cells which apparently do not express 5-lipoxygenase, such as endothelial cells and platelets [ 45–48]. The enzyme has also been characterized and isolated from platelets [ 49]. A profound overexpression of LTC₄ synthase has been found in bronchial biopsies from aspirin-intolerant asthmatic patients [ 50]. The cloning of the cDNA for human LTC₄ synthase revealed that the deduced amino acid sequences had 31% identity with FLAP [ 51, 52]. There was no homology with microsomal glutathione S-transferase 1 (MGST 1) but the enzyme possesses 44% identity with MGST 2 and 27% identity with MGST 3, which also possessed some LTC₄ synthase activity [ 53, 54]. The LTC₄ synthase gene is 2.5 kb in length and contains five exons and four introns [ 55, 56]. The promoter contains consensus sequences for Sp 1, AP-1 and AP-2 binding sites, where the latter two are phorbol-ester response elements. A polymorphism in the promoter region has been demonstrated [ 57]. This observed polymorphism was more common among patients with aspirin-induced asthma, but the consequence of this finding for the expression of LTC₄ synthase remains to be determined. The gene for LTC₄ synthase has been localized to the 35 q region of chromosome 5 [ 55, 56]. This region appears to play a pivotal role in allergic and inflammatory disorders since it is also close to the locus for genes encoding pro-inflammatory cytokines and growth factor receptors [ 58, 59]. Linkage analysis studies also suggest that certain familial elevations of IgE and bronchial hyperresponsiveness are located to this region [ 60, 61].

Cellular formation of leukotrienes

The mechanisms involved in the cellular biosynthesis of leukotrienes are still incompletely understood [ 24]. The subcellular distribution of the 5-lipoxygenase, for example, varies between different cells. In neutrophils, the 5-lipoxygenase is located in the cytosol of non-activated cells, but after cell adherence or upon stimulation with calcium ionophore A23187 or IgE/antigen [ 62–64], the enzyme translocates to the nuclear envelope ( Fig. 2). In contrast, the 5-lipoxygenase is primarily localized to the nuclear envelope in non-activated rat alveolar macrophages and basophilic leukaemia cells [ 65, 66]. In resting alveolar macrophages, the 5-lipoxygenase appears to be associated with the euchromatin of the nucleus and moves to the nuclear envelope upon activation [ 67]. It has also been suggested that the 5-lipoxygenase interacts with the SH3 domain of cellular proteins such as certain cytoskeletal proteins and the growth factor receptor-bound protein 2 (Grb2) [ 68], an adapter protein for cellular tyrosine kinases. The translocation of the 5-lipoxygenase to the nuclear envelope is FLAP-independent since the 5-lipoxygenase is also translocated in cells lacking FLAP [ 69]. The physiological relevance of the localization of the 5-lipoxygenase and FLAP in the nuclear envelope is not known, but it is possible that 5-lipoxygenase and its products might also have a function in transcriptional regulation of genes.

The regulation of leukotriene synthesis appears to be different in myeloid cells and B-lymphocytes, since B-lymphocytes do not produce leukotrienes after stimulation with calcium ionophore A23187 alone [ 22, 70, 71]. However, activation of B-lymphocytes with ionophore A23187 in the presence of exogenous arachidonic acid leads to the formation of LTB₄. The biosynthesis of LTB₄ in B-lymphocytes is greatly enhanced in the presence of thiol-reactive substances [ 22, 71]. The physiological condition and ligand(s) that induce leukotriene synthesis in B-lymphocytes are at present not known.

Immunohistochemical analysis of human lung showed that LTC₄ synthase was localized to perinuclear membrane in alveolar macrophages [ 72]. Recently, it was reported that 5-lipoxygenase and LTC₄ synthase were assembled in lipid bodies in platelet activating factor (PAF) stimulated eosinophils [ 73], demonstrating cell variation in the localization of the enzymes involved in leukotriene synthesis.

The inflammatory response in asthma is characterized by accumulation and activation of T-lymphocytes, eosinophils and mast cells in the airway wall. Eosinophils and mast cells have high capacity to generate LTC₄ from the endogenous pool of arachidonic acid. Leukotriene C₄ can also be produced in the lung through interaction between different types of cells. Several lines of evidence suggest that stimulated leucocytes can release arachidonic acid or LTA₄, which can be further metabolized by other surrounding cells (so-called transcellular metabolism) [ 74–78]. A possible scenario in the lung is that activated myeloid cells release LTA₄ which is subsequently metabolized by various types of leucocytes, endothelial cells and platelets to LTC₄ [ 46, 77, 79]. It has been shown that adenosine is an inhibitor of transcellular metabolism of LTA₄, released by leucocytes upon activation with physiological agonists [ 79]. The relative role of transcellular metabolism in the formation of LTC₄ in asthma and other pathophysiological conditions remains to be clarified.

Export of leukotrienes to the extracellular space

After intracellular formation of LTC₄, the compound is released to the extracellular space ( Fig. 2). Carrier-mediated export of LTC₄ has been demonstrated in various cells, such as human eosinophils, myeloid leukaemia cell lines and mastocytoma cells [ 80–82]. ATP-dependent transport of LTC₄ across plasma membranes has been demonstrated in different type of cells [ 82, 83]. Subsequently, this carrier protein was identified as the multidrug-resistance associated protein (MRP) [ 84, 85]. The export of conjugated cytostatic drugs and several naturally occurring glutathione conjugates is known to be mediated by the MRP. The transport of LTC₄ in membrane vesicles from MRP-transfected HeLa cells can be inhibited by MRP antibodies [ 86], indicating that MRP is the principal transporter for LTC₄ in certain cells. Mice lacking MRP also have a diminished export of LTC₄ and an attenuated response to inflammatory stimuli [ 87]. After the carrier-mediated export of LTC₄, the sequential metabolism of LTC₄ provides the extracellular LTD₄ and LTE₄. Intracellular LTB₄ is also secreted via a specific transporter, but the carrier protein has not yet been characterized [ 88]. These carrier proteins are a potential future target for antileukotriene drugs.

Mechanism of action of leukotriene biosynthesis inhibitors

Two mechanisms have successfully been employed to inhibit the cellular synthesis of leukotrienes, either direct inhibition of the 5-lipoxygenase or indirect inhibition of leukotriene biosynthesis by interference with FLAP.

The directly acting 5-lipoxygenase inhibitors can be divided into three different groups:

1 Redox-active agents such as N-hydroxyureas, N-alkylhydroxamid acids, selenite, hydroxybenzofurans, hydroxylamines and catechols [ 24, 89–91].

2 Alkylating agents and compounds which react with SH groups have also been found to inhibit leukotriene synthesis in vitro [ 92]. Such compounds are however, relatively non-selective and may interfere with a number of other cellular reactions.

3 Competitive inhibitors of the 5-lipoxygenase, which are based on thiopyranoindole and methoxyalkyl thiazole structures, and are more selective and may act as reversible non-redox inhibitors of 5-lipoxygenase [ 24, 93].

The other main class of leukotriene synthesis inhibitors is represented by compounds which bind to FLAP [ 94, 95] (and thereby block the utilization of the endogenous pool of arachidonic acid (see above). The compounds which are indirectly acting 5-lipoxygenase inhibitors have been synthesized from indole and quinoline structures [ 24].

Effects of leukotriene B₄

Leukotriene B₄ has been found to have primarily leucocytes as targets for its biological activity ( Fig. 3). It is a potent stimulus for activation of leucocytes, eliciting chemokinetic and chemotactic responses in vitro [ 96]. In vivo, LTB₄ increases leucocyte rolling and adhesion to the venular endothelium [ 97], followed by their emigration into the extravascular space [ 8]. During a short-lasting exposure to LTB₄, polymorphonuclear leucocytes are mainly recruited. With prolonged exposure to LTB₄, as presumably occurs when LTB₄ is formed in vivo, other granulucytes, including eosinophils, are found in tissues or exudates after challenge with LTB₄ [ 98]. It has been shown that LTB₄ is a chemoattractant for interleukin-5 primed eosinophils [ 99], and LTB₄ may stimulate production of interleukin-5 in T-lymphocytes [ 100].

In addition to effects on leucocyte adhesion and migration, LTB₄ stimulates secretion of superoxide anion and release of different granulae constituents from leucocytes [ 101, 102]. Among effects of LTB₄ on inflammatory cells, it has been observed that LTB₄ may stimulate expression of low-affinity receptors for IgE on B-lymphocytes [ 103], and the secretion of IgM, IgG and IgE [ 103–105]. Possibly relating to these effects of LTB₄ on B-cells, it was recently observed that sensitized 5-lipoxygenase-deficient mice produced lower levels of IgG and IgE upon ovalbumin challenge than wild-type mice [ 106]. Furthermore, the observation that LTB₄ is an agonist for the nuclear transcription factor PPAR_α (peroxisome proliferator-activated receptor α) has created considerable interest [ 107]. The finding may implicate a role for LTB₄ in the control of central events in lipid metabolism and inflammation, but also the possible existence of a feedback loop which may respond to increased leukotriene production by enhanced leukotriene catabolism. The observations additionally point towards the possibility that LTB₄ also has intracellular and nuclear targets, which may participate in long-term control of gene expression.

With respect to effects of LTB₄ on the lung, it is also established that LTB₄ has contractile activity in the guinea pig lung parenchyma [ 108–111]. The response is indirect, involving release of TXA₂ [ 109–111], and most likely also histamine [ 111]. The response has so far not been observed in other smooth muscles, including human bronchi, but nevertheless indicates that certain tissues may contain elements with the ability to release spasmogenic mediators when exposed to LTB₄. In a dog model, LTB₄ was found to increase airway reactivity to acetylcholine [ 112], whereas among non-asthmatic humans there was no change in bronchial hyperresponsiveness to histamine following the inhalation of LTB₄ alone or in combination with PGD₂ [ 113]; nor was there any direct bronchoconstrictor effect of inhaled LTB₄ [ 113]. A subsequent study observed that inhalation of LTB₄ by healthy human volunteers was followed by distinct cellular changes in the airways, and possibly also some plasma exudation [ 114]. In a study where LTB₄ was inhaled by a group of asthmatics, the lack of immediate bronchoconstrictive properties was confirmed, as well as prominent acute effects on leucocyte traffic in the lung and blood [ 115].

Receptors for leukotriene B₄

The different profiles of biological activities for LTB₄ and cysteinyl-leukotrienes ( Fig. 3) suggested that the two main classes of leukotrienes possessed distinct receptors. The experimental data have indeed established that LTB₄ acts at a specific receptor, which now is designated the BLT-receptor [ 116]. Thus, specific [³H]-LTB₄ binding has been demonstrated in many tissues, including human polymorphonuclear leucocytes (PMNs) [ 117, 118]. Pharmacological evidence suggested that BLT-receptors were G-protein coupled [ 119–121], and the receptor was recently cloned from retinoic acid differentiated HL-60 cells and functionally expressed in CHO cells [ 122]. Incidentally, the BLT-receptor was originally cloned as an orphan receptor from a human B-lymphoblast cDNA library [ 123]. The cDNA encoded a 352 amino acid cell-surface protein which was G-protein coupled and mediated chemotaxis. Northern blotting experiments of human tissues displayed a preferential expression of mRNA for the BLT-receptor in PMNs [ 122]. There was also some expression in the spleen and thymus, whereas most other examined tissues, including the lung, showed no or insignificant expression of mRNA for the BLT-receptor [ 122].

Some selective and relatively potent antagonists of LTB₄ have been developed. A few compounds have entered into early clinical testing in humans. The compound LY-293 111 (VML 295) was recently found to inhibit LTB₄-induced neutrophil responses in vivo and allergen-induced neutrophil activation, but it had no effect on allergen-induced early- or late-phase airway obstruction in asthmatics [ 124]. The results with LY-293 111 in asthmatics argue against an important role for LTB₄ as a mediator in asthma, but do not exclude the possibility that LTB₄ may be involved in other pulmonary reactions, e.g. syndromes characterized by neutrophil accumulation and activation.

Biological effects of the cysteinyl-leukotrienes

Soon after their discovery, it was documented that LTC₄ and LTD₄ were potent inducers of bronchoconstriction in guinea pig airways in vitro and in vivo [ 125, 126] and caused contractions of isolated human bronchi [ 127–129]. Local injection of these two cysteinyl-leukotrienes also increased accumulation of i.v. injected Evans blue in the skin [ 125, 126], suggesting an increase in microvascular permeability. In the hamster cheek pouch, it could be established that LTC₄ and LTD₄ indeed caused exudation of plasma proteins in postcapillary venules [ 97]. In the guinea pig, it was shown that each of LTC₄, LTD₄ and LTE₄ was capable of causing accumulation of the plasma protein tracer Evans blue in the airways [ 130]. The plasma exudation occurred in all airway segments, ranging from the most peripheral small bronchi to trachea, and there was evidence of Evans blue accumulation in superficial as well as deep layers of the airway mucosa.

The biological effects of LTE₄ have been studied less, perhaps because this leukotriene was found to be an incomplete and less potent agonist than LTC₄ and LTD₄ in some smooth muscle preparations such the guinea pig ileum [ 131–133]. However, LTE₄ has been documented to possess a bronchoconstrictor activity in vitro and in vivo which is very similar to that of LTC₄ and LTD₄ [ 134, 135]. It has also been observed that prolonged exposure to LTE₄ may produce enhancement of the responsiveness of smooth muscle to histamine [ 136, 137]. Moreover, LTE₄ is a full agonist for contraction of human bronchi in vitro [ 138], and it is not significantly less potent than LTC₄ and LTD₄ [ 138, 139]. It is therefore evident that with regard to bronchoconstrictive activity, the three different cysteinyl-leukotrienes have very similar potencies.

It has been observed that LTC₄ and LTD₄ may stimulate mucus secretion in isolated animal and human airways [ 140–142] ( Fig. 3). Experiments in isolated perfused hearts also disclosed a depressive effect on cardiac contractility [ 143, 144]. The effect correlated with coronary vasoconstriction [ 145], but a direct negative inotropic effect on the myocardium may also be involved [ 54, 146]. More recently, additional effects with potential relevance to the role of cysteinyl-leukotrienes in asthma and pulmonary inflammation have been reported. Thus, increased infiltration of eosinophils into the airway mucosa of asthmatics was observed following inhalation of LTE₄ [ 147], and inhalation of LTD₄ increased the number of eosinophils in induced sputum samples from asthmatics [ 148]. The capacity of cysteinyl-leukotrienes to promote eosinophil recruitment has been confirmed in experimental models [ 149, 150], although the mechanisms involved remain to be defined. The response may involve leukotriene-induced release of IL-5 from epithelial cells [ 149], but direct chemotactic effects are also possible [ 150]. There are also experimental data in vitro [ 151–154] and in vivo [ 155] supporting the notion that cysteinyl-leukotrienes may be involved in airway smooth muscle proliferation and remodelling.

The exquisite spasmogenic potency of cysteinyl-leukotrienes on isolated human bronchi [ 126–129, 138, 139] was soon confirmed in bronchoprovocation studies of normal subjects [ 156–158]. The cysteinyl-leukotrienes are in fact the most potent endogenous bronchoconstrictors known to date. Accordingly, cysteinyl-leukotrienes have generally been found to produce bronchoconstriction in doses that are 100– 10000 times lower than those required with histamine or methacholine. Asthmatics have also been found to be hyperresponsive to inhalation of LTC₄, LTD₄ and LTE₄ [ 159–161]. There are indications that LTD₄ may, in addition, cause greater airway narrowing than methacholine [ 162]. On the other hand, at least for LTC₄ and LTD₄, there are observations indicating that the hyperresponsiveness of asthmatics to these leukotrienes relative to methacholine was lower than in normals [ 160, 163]. This raises the interesting hypothesis that chronically elevated production of leukotrienes in asthmatics induces adaptive changes in the airway effector cells or at the receptor level. However, concerning LTE₄, the opposite finding has been reported, namely that asthmatics were especially hyperresponsive to this particular leukotriene [ 164]. In this context, there are also indications that repeated challenge with cysteinyl-leukotrienes is associated with tachyphylaxis mediated by local generation of a prostanoid, presumably PGE₂ [ 165].

Receptors for cysteinyl-leukotrienes

Despite the current introduction of receptor antagonists for cysteinyl-leukotrienes as a new therapy in asthma, the structure of the receptors for cysteinyl-leukotrienes remain unknown. On the basis of primarily functional studies in smooth muscle assays, two main classes of receptors have been outlined [ 116] ( Fig. 3). The CysLT₁ receptor is blocked by the class of drugs which is currently being introduced in the clinic. It appears that most effects of cysteinyl-leukotrienes in human airways are mediated by this CysLT₁ receptor ( Fig. 3). Certain responses to cysteinyl-leukotrienes in vascular tissues from humans and in some animal smooth muscle assays are, however, resistant to available CysLT₁ antagonists. These responses are currently classified as being mediated by a CysLT₂ receptor ( Fig. 3), but there is no selective antagonist available and the CysLT₂ receptor is currently considered as an operational concept. In addition, there are indications that subclasses may exist among both CysLT₁ and CysLT₂ receptors.

Allergen

Soon after the discovery of the leukotrienes and the structure elucidation of SRS-A, it was possible to obtain in vitro evidence that cysteinyl-leukotrienes were generated in asthmatic lung tissue and that they mediated a major part of the Schultz–Dale contraction to allergen in airways from atopic asthmatics in vitro [ 134]. Using passively sensitized tissues from non-asthmatics or challenge with anti-IgE in non-atopic human airways [ 188–190], it has consistently been possible to document that a major part of the allergen-induced bronchial contraction in vitro is due to release of cysteinyl-leukotrienes.

Before pharmacological interventions were possible, it was furthermore established that allergen-induced airway obstruction was associated with increased urinary excretion of LTE₄, at least during the early phase [ 191, 192]. The findings that bronchoconstriction induced by histamine [ 192] or methacholine [ 193] had no influence on urinary LTE₄ supported the idea that the leukotrienes were released as a primary and early consequence of the allergen challenge. For the early reaction, it appears most likely that the leukotrienes originate from the pulmonary mast cells. Since prostaglandin D₂ (PGD₂) in humans is almost exclusively generated in activated mast cells, the recent observation of an apparent correlation between increased urinary excretion of LTE₄ and the PGD₂-metabolite 9α,11β-PGF₂ following allergen bronchoprovocation supports this hypothesis [ 194]. More recently, it has been possible to demonstrate increased urinary LTE₄ also during the late phase [ 195]. It is also known that the eosinophil, which often is increased in the tissue and activated during the late phase, is a very proficient producer of LTC₄ [ 196].

From studies with different antileukotrienes, it is now established that cysteinyl-leukotrienes mediate the predominant component of allergen-induced airway obstruction. In particular, the early fall in FEV₁ after challenge with a fixed dose of antigen has consistently been found to be blocked by between 50 and 75% after pretreatment with either potent CysLT₁-receptor antagonists [ 197–200] or effective inhibitors of leukotriene biosynthesis [ 171–173, 201]. Likewise, in studies using cumulative challenge with allergen and PD₂₀ as the end-point, a highly significant increase in the PD₂₀ for allergen was obtained after leukotriene antagonism [ 202, 203].

The late phase of allergen-induced airway obstruction is usually considered to have most bearing on the chronic airway inflammation of asthmatics, and to be most predictive of treatment sucess. Concerning the late phase, the findings with antileukotrienes were initially more variable with respect to the extent of protection afforded by this class of drugs. However, some of the first studies involved single-dose administration of drugs in comparatively low doses, and sometimes the half-life of the drug was shorter than the duration of the late reaction. In studies where the drugs have been given repeatedly and in doses sufficient to be potent and bioavailable, it has been evident that cysteinyl-leukotrienes also contribute significantly to the late phase [ 197, 198, 200, 201]. For example, after 1 week of pretreatment with zafirlukast 80 mg b.i.d., the late phase was inhibited by more than 55% [ 200], and after 3 days' pretreatment with the FLAP antagonist BAY × 1005 it was inhibited by 46% [ 201].

Exercise

Different antileukotriene drugs have consistently been shown to provide a major inhibition of exercise-induced bronchoconstriction (EIB) [ 204–208]. Antileukotrienes also protect against airway obstruction induced by inhalation of dry cold air [ 209], which is a closely related challenge since the mechanisms that trigger the response to exercise involve changes in airway temperature or osmolarity, or both. It has, with a few exceptions [ 210, 211], generally been difficult to detect increased levels of urinary LTE₄ during EIB. This presumably relates to the short duration of the reaction, since the effect of the pharmacological antagonists convincingly shows that the local levels of cysteinyl-leukotrienes are sufficiently increased to produce the airway response. It should also be acknowledged that subjects with strong responses to exercise challenge demonstrate a high degree of bronchial hyperresponsiveness, which means that comparatively small elevations in the local levels of mediators may be sufficient to produce significant bronchoconstriction.

Aspirin-induced bronchoconstriction

Using the method of bronchial challenge with lysine-aspirin in aspirin-intolerant asthmatics [ 212, 213], it was observed that lysine-aspirin-induced bronchoconstriction was associated with increased urinary excretion of LTE₄ [ 192]. In contrast, the urinary levels of LTE₄ remained unchanged during the course of bronchoprovocation in five asthmatics who did not develop an airway obstruction to inhaled lysine-aspirin [ 192]. This observation was in line with concurrent reports that oral aspirin challenge increased urinary LTE₄ [ 214, 215].

The first pharmacological study intended to test whether or not leukotrienes mediated aspirin-induced bronchoconstriction produced some support for leukotriene involvement, but the inhibition of the response to oral challenge with aspirin provided by the inhaled antagonist SKF 104,353 was incomplete and not observed among all the subjects [ 216]. Using bronchial challenge with lysine-aspirin and a more potent antagonist (MK-0679), it was subsequently possible to provide strong evidence for a major leukotriene component in aspirin-induced bronchoconstriction [ 217]. Thus, one single oral dose of MK-0679 produced a marked inhibition of the aspirin-induced bronchoconstriction in all of eight subjects. In fact, three subjects failed to produce a 20% drop in FEV₁ despite a cumulative challenge protocol was used. Israel et al. [ 218] were able to show that pretreatment with the 5-lipoxygenase inhibitor zileuton inhibited the bronchoconstriction induced by oral aspirin challenge. Interestingly, they also reported that several of the extrapulmonary symptoms were blocked by aspirin. Nasser et al. [ 219] confirmed that another 5-lipoxygenase inhibitor, ZD2138, was able to inhibit the bronchoconstriction induced by oral aspirin provocatation. Finally, using another leukotriene antagonist (ONO-1078) and another NSAID (dipyrone), Yamamoto et al. [ 220] have also provided evidence that the response to bronchial challenge has a major leukotriene component. The published studies have thus established that treatment with antileukotriene drugs can prevent aspirin/NSAID-induced bronchoconstriction.

Acute bronchodilation

When antileukotrienes were first studied in normal subjects or mild asthmatics subjected to bronchoprovocations with allergen [ 197, 198, 203] or LTD₄ [ 167, 168, 233], there were no significant effects on baseline pumonary function. However, when antileukotrienes were administered to asthmatics with baseline asthma symptoms and compromised baseline pulmonary function, an acute bronchodilator response was consistently observed [ 234, 235]. The magnitude of the effect have varied between 10 and 20% improvement in FEV₁, and in one of the studies there was a direct relation between the magnitude of the response and the reduction of baseline FEV₁% of predicted [ 235].

Moreover, the bronchodilator response to the CysLT1 antagonists zafirlukast [ 234] and MK-571 [ 235] was found to be additive to the effects of inhaled and nebulized β-agonists, suggesting that the response may be due to inhibition also of components of the airway obstruction which are relatively less sensitive to smooth muscle relaxant drugs. In a more recent study with montelukast [ 236], the bronchodilator response to a single oral dose of montelukast showed, in addition, a trend to be greater in the subgroup which received baseline treatment with inhaled glucocorticosteroids. The findings suggest that ongoing leukotriene formation in inflamed airways is a prerequisite for a bronchodilator response to antileukotrienes. The finding that the 5-lipoxygenase inhibitor zileuton also causes a prompt bronchodilator response in asthmatics with decreased baseline pulmonary function [ 222, 225] is consistent with this view and illustrates the rapid turnover of leukotrienes in the airways.

Effects of chronic administration

With repeated administration of antileukotrienes to different groups of asthmatics, chronic improvements in pulmonary function have been demonstrated, both as increased morning and evening peak expiratory flow rates (PEFR) and as increased FEV₁ recorded at clinic visits. In some studies, the long-term improvements in pulmonary function have been of a similar magnitude as the acute bronchodilator response to the drugs, but in several studies there have been, during the treatment period, further improvements over and above the level obtained with the first dose of drug. Such a progressive increase in the therapeutic response over time suggests that the treatment has long-term effects on the underlying inflammatory process in the airways.

The improvements of pulmonary function have generally been associated with favourable changes in other asthma outcome measures. Thus, asthma symptoms have diminished both during the day and at night, and the rescue use of bronchodilators has been reduced. Some recent studies have also shown improvements in asthma-specific quality of life scores [ 230, 237].

Prevention of asthma exacerbations is one of the main goals in asthma management. Therefore, it is encouraging that the few studies of a size and design which have permitted evaluation of the number of asthma exacerbations have observed reductions of this particular measure. For example, zileuton caused a highly significant reduction in the number of exacerbations requiring rescue with an oral steroid cure [ 224]. Furthermore, addition of pranlukast prevented the asthma deterioration which was provoked in a group of severe asthmatics by a 6-week reduction of the inhaled dose of beclomethasone (at baseline 1500 μg or more) to half its original value [ 228]. Interestingly enough, the prevention of asthma exacerbations in that particular study was also associated with prevention of the rise in serum ECP and exhaled NO which was precipitated in the group that received placebo and deteriorated. Such effects on surrogate markers of inflammation would also indicate that antileukotriene may affect the cellular events which are believed to cause persistent airway inflammation in asthma. In line with these effects on inflammatory markers, recent trials in children [ 238] and adults [ 239] have documented a decrease in the total number of circulating eosinophils during treatment with montelukast. The reductions in eosinophil numbers were associated with clinical improvements of the asthma and increased pulmonary function. In the adult study [ 239] the reduction in circulating eosinophils in subjects treated with montelukast was parallelled by a significant reduction of sputum eosinophils. Moreover, treatment with zileuton or zafirleukast inhibited the influx of eosinophil in the airways induced in asthmatics by segmental bronchial challenge with antigen [ 240, 240a]. In view of the ability of inhaled LTE₄ [ 147] and LTD₄ [ 148] to recruit eosinophils into the airway mucosa and sputum, respectively, it appears likely that the cysteinyl-leukotrienes, rather than LTB₄, are promoting eosinophil migration and activation in the lung.

From a review of the larger published trials with antileukotrienes in mild-to-moderate asthma, it appears as if the response to treatment with antileukotrienes during 4–13 weeks is fairly similar to the effect of introducing low doses of inhaled glucocorticosteroids during similar or longer time periods in comparable groups of patients [ 241–244]. Direct comparisons between antileukotrienes and other asthma medications, and in particular inhaled glucocorticosteroids, have, however, not yet been published.

The initial studies with antileukotrienes were conducted in subjects who not were treated with glucocorticosteroids, and who sometimes also suffered from relatively mild asthma. More recently, antileukotrienes have been studied in more severe patients, and as an add-on therapy to patients not being sufficiently controlled with conventional therapy, including glucocorticosteroids [ 229, 230, 237]. The results of such studies suggest that treatment with antileukotrienes is at least as effective in more severe asthmatics, and possibly relatively more effective in those not controlled by inhaled glucocorticosteroids.

Antileukotrienes in aspirin-intolerant asthma

It has consistently been observed that the baseline urinary excretion of LTE₄ is significantly higher in aspirin-intolerant asthmatics than in aspirin-tolerant asthmatics [ 192, 214, 215]. In addition, it was recently observed that there was a marked overexpression of the LTC₄ synthase in bronchial biopsies of aspirin-intolerant asthmatics [ 50]. In contrast, the basal urinary excretion of the mast cell marker 9α,11β-PGF₂ in aspirin-intolerant asthmatics was not different from that of aspirin-tolerant asthmatics [ 194], suggesting that mast cell activation was not involved in the basal overproduction of cysteinyl-leukotrienes. On the basis of the well established association between eosinophils and aspirin-intolerant asthma [ 245], and the indications that circulating eosinophils have enhanced capacity to generate leukotrienes, the findings together would seem to fit with the hypothesis that the eosinophils are one major source of the baseline production of leukotrienes. This adds to the clinical indications that aspirin-induced bronchoconstriction and the persistent airflow obstruction may be different and mechanistically unrelated components in the syndrome. There are also reports that aspirin-intolerant asthmatics are particularly sensitive to inhalation of cysteinyl-leukotrienes [ 246], and in particular LTE₄ [ 247]. Therefore, it may be that aspirin-intolerant asthmatics display both a basal overproduction of leukotrienes and an increased bronchial responsiveness to these compounds.

In support of such indications that even when not exposed to NSAIDs, airflow obstruction in aspirin-intolerant subjects has a leukotriene component, it was observed that one single dose of treatment with the leukotriene antagonist MK-0679 induced a prompt improvement in pulmonary function [ 248]. The response to this single dose of the leukotriene antagonist was about 80% of the maximal reserve for bronchodilation. The response was found to correlate strongly with the severity of asthma expressed as the sum of each subject's rank order score for FEV₁% of predicted and asthma medication. There was also a good correlation between the bronchodilator response to the drug and the individual's sensitivity to aspirin, expressed as pre-study PD₂₀ for ASA. The findings again serve to illustrate that ongoing leukotriene formation in the airways is a prerequisite for a bronchodilator response to drugs which block the action or release of leukotrienes.

Finally, a recently concluded treatment trial with the 5-lipoxygenase inhibitor zileuton in 40 subjects with aspirin-intolerant asthma supports the notion that leukotrienes indeed mediate persistent airway obstruction and other symptoms in these particular asthmatics [ 229]. Zilueton (600 mg tablets four times daily) or placebo was administered during 6 weeks in a cross-over double-blind design with 6 weeks wash-out period in-between. The majority of subjects had suffered from asthma for 5 years or more, and all subjects had demonstrated reversibility of bronchoconstriction following inhalation of a beta-stimulant. This was an add-on study and all subjects were kept on their regular asthma therapy according to existing guidelines. It is noteworthy that 97.5% of the subjects in this study were already being treated with glucocorticosteroids, and most often in high doses of inhaled preparations (> 1200 μg budesonide or beclomethasone) or with prednisone orally. When the outcome of the treatment periods was evaluated, it was evident that zileuton caused both acute and chronic improvement in pulmonary function, at the same time as there was a significant decrease in the use of beta-agonists. The bronchial hyperresponsiveness to histamine was evaluated before and at the end of each treatment period. Despite the fact that the patients had been treated with glucocorticosteroids for long periods, there was a further significant reduction of bronchial hyperresponsiveness when zileuton was added. In addition, consistent with previous preliminary findings, the treatment with zileuton caused an improvement in nasal function and, in particular, a return of the sensation of smell. Therefore, the study supported the hypothesis that inhibition of leukotrienes may provide a new therapeutic alternative in aspirin-intolerant asthma. The improvements observed were particularly encouraging because the 5-lipoxygenase inhibitor zileuton at the employed dose level only caused partial inhibition of leukotriene biosynthesis, measured as urinary excretion of leukotriene E₄. Furthermore, the finding that addition of the 5-lipoxygenase inhibitor zileuton caused improvement over and above that provided by treatment with glucocorticoids lends further support to the concept that antileukotrienes and glucocorticoids treat different parts of the airway inflammation. It is also noteworthy that the degree of improvement with zileuton was as great or greater than those obtained in recent trials where long-acting β₂-stimulants [ 249–251], or theophylline [ 252] have been added to glucocorticosteroids.

Evaluation of antileukotrienes as anti-asthmatics

The published and reviewed studies support the hypothesis that antileukotrienes may be tried at every junction where existing therapy fails to provide the desired level of asthma control. It is most clearly documented that mild asthmatics who are insufficiently controlled by ‘as needed’ use of β₂-agonists may achieve better control by administration of an antileukotriene. Likewise, subjects with mild-to-moderate asthma but insufficient control with low or moderate doses of inhaled glucocorticosteroids have been found to improve on addition of antileukotrienes. As discussed, more recent trials have documented clinically important improvements by antileukotrienes in more severe asthmatics, including aspirin-intolerant asthmatics. In addition to these possible indications for regular therapy with antileukotrienes, the prevention of trigger-factor-evoked or night asthma may constitute reasons for sole use of antileukotrienes occasionally in certain patients. For example, prevention of exercise or allergen-induced reactions in children by the use of an oral medication which may be taken once or twice daily appears attractive.

In addition to prevention of trigger-factor-evoked bronchoconstriction, the antileukotrienes have been shown to improve asthma control, measured, for example, as a reduction in the number of asthma exacerbations. According to our current understanding, such an improved asthma control would seem to provide circumstantial evidence that antileukotrienes inhibit central components of asthmatic airway inflammation. This would also be in line with the acute and chronic inflammatory effects that may be evoked by leukotrienes, including recruitment of eosinophils in asthmatics. The studies which indicate progressive improvement in pulmonary function and other outcome measures with prolonged treatment also argue in favour of beneficial long-term effects of antileukotrienes on structure and function of the asthmatic airways. The introduction of the antileukotrienes therefore provides a new class of drugs which both control asthma and prevent bronchoconstriction. The observations that glucocorticosteroids do not inhibit leukotriene production in asthmatics highlight the current awareness that antileukotrienes possess therapeutic properties that are different and complementary to those of glucocorticosteroids.

Several of our current asthma drugs have been developed to be given by the inhaled route in order to minimize the rather pronounced side-effects which occur when they are given systemically. Despite the safety of low-to-moderate doses of inhaled glucocorticosteroids, their use in children remains an area of concern and debate [ 253]. All antileukotrienes introduced for clinical use are given orally. The oral route of administration appears to carry some possible advantages over conventional therapies. Apart from the potentially favourable effect of oral admininistration on patient compliance with asthma therapy, there are several indications that antileukotrienes also prevent some of the symptoms in rhinitis (254–256]. It is well established that a large proportion of asthmatics suffer from rhinitis [ 257] and it would therefore seem to be an added benefit for this new group of asthma medication to alleviate some of the rhinitic symptoms as well.

Received 13 May 1998; accepted 4 June 1998.