The Impact of Corticosteroids on Human Airway Smooth Muscle Contractility and Airway Hyperresponsiveness: A Systematic Review

Classically, the effects elicited by corticosteroids (CS) are mediated by the binding and activation of cytosolic glucocorticoid receptors (GR). However, several of the non-genomic effects of CS seem to be mediated by putative non-classic membrane receptors characterized by pharmacological properties that are different from those of classic cytosolic GR. Since pre-clinical findings suggest that inhaled CS (ICS) may also regulate the bronchial contractile tone via putative CS membrane-associate receptors, the aim of this review was to systematically report and discuss the impact of CS on human airway smooth muscle (ASM) contractility and airway hyperresponsiveness (AHR). Current evidence indicates that CS have significant genomic/non-genomic beneficial effects on human ASM contractility and AHR, regardless of their anti-inflammatory effects. CS are effective in reducing either the expression, synthesis or activity of α-actin, CD38, inositol phosphate, myosin light chain kinase, and ras homolog family member A in response to several pro-contractile stimuli; overall these effects are mediated by the genomic action of CS. Moreover, CS elicited a strong bronchorelaxant effect via the rapid activation of the Gsα–cyclic-adenosine-monophosphate–protein-kinase-A pathway in hyperresponsive airways. The possibility of modulating the dose of the ICS in a triple ICS/long-acting β2-adrenoceptor agonist/long-acting muscarinic antagonist fixed-dose combination supports the use of a Triple MAintenance and Reliever Therapy (TriMART) in those asthmatic patients at Step 3–5 who may benefit from a sustained bronchodilation and have been suffering from an increased parasympathetic tone.


Introduction
Generally, the effects elicited by corticosteroids (CS) are generally mediated by the binding and activation of cytosolic glucocorticoid receptors (GR) that, in turn, translocate towards the nucleus, interact with glucocorticoid response elements (GRE), and ultimately elicit genomic effects that modulate protein expression [1,2]. Indeed, such a complex cascade requires a prolonged onset of action to activate/inhibit genomic processes for CS and also other steroid hormones [1,3].
However, a wide range of non-genomic effects of CS seems to be mediated by putative non-classic membrane receptors characterized by pharmacological properties that are different from those of classic cytosolic GR [3]. Interestingly, the rapid, almost-immediate, non-genomic effects of CS regulate several signaling processes leading to effects on intracellular calcium mobilization and homeostasis, reactive oxygen and nitrogen species, inflammatory and apoptotic pathways, and skeletal and smooth muscle function [1]. In this regard, current pre-clinical evidence suggests that in human isolated airways inhaled CS  The PICO (Patient problem, Intervention, Comparison, and Outcome) framework was applied to develop the literature search strategy and question, as previously reported [7]. Namely, the "Patient problem" included increased ASM contractility or AHR; the "Intervention" regarded CS; the "Comparison" was performed with respect to the controls or the baseline; and the assessed "Outcome" was human ASM contractility or AHR that was not related to the anti-inflammatory effects of CS.
Citations of previously published relevant and recently published reviews or editorials were examined to select further pertinent studies, if any [8]. Two reviewers independently checked the relevant studies identified from the literature search. The studies were selected in agreement with the previously mentioned criteria and any difference in opinion about eligibility was resolved by consensus.

Data Extraction
Data from the included studies were extracted in agreement with Data Extraction for Complex Meta-anALysis (DECiMAL) recommendations [9] and checked for study references and year of publication, type of study, type of cells and tissue donors, characteristics of analyzed patients, contractile stimuli, number of tissue donors or patients, age and sex, treatments, route of administration, outcome measurements to evaluate the impact on ASM contractility and AHR, Jadad score, and Cochrane Risk of Bias (RoB).

Endpoints
The endpoint of this systematic review was the impact of CS on human ASM contractility and AHR that was not related to the anti-inflammatory effects of CS. The effects of CS were assessed both alone and in combination with long-acting β 2 -adrenoceptor agonists (LABA) and long-acting muscarinic antagonists (LAMA).

Strategy for Data Analysis
Data from original papers were extracted and reported via qualitative synthesis.

Quality Score and RoB
The summary of the risk of bias for each included RCT was analyzed via the Jadad score [10] and Cochrane RoB 2 [11]. The Jadad score, with a scale of 1-5 (a score of 5 being the best quality), was used to assess the quality of the papers concerning the likelihood of bias related to randomization, double blinding, withdrawals and dropouts [10]. Studies were considered to be of low quality with a Jadad score <3, of medium quality with a Jadad score = 3, and of high quality with a Jadad score >3. The weighted assessment of the overall risk of bias was analyzed via the Cochrane RoB 2 tool [11] by using the robvis visualization software [12,13]. Two reviewers independently assessed the quality of individual studies, and any difference in opinion about the quality score was resolved by consensus.

DEX
An in vitro study on hASMC investigated whether CS transcriptionally regulated the expression of CD38, a~45-kDa glycosylated transmembrane protein having a central role in intracellular calcium homeostasis and AHR [18]. In hASMC transfected with a 3 kb human CD38 promoter containing a nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB), activator protein-1 (AP-1), and four GRE putative binding sites, 24 h treatment with DEX 1 µM completely reversed the two-fold activation of the promoter induced by tumor necrosis factor-alpha (TNF-α) 50 ng/mL [18].
A previous in vitro study [16] investigated the effect of DEX on inositol phosphate (IP) response produced by histamine (His) in primary cultured hASMC, considering the pivotal role of IP in inducing the release of calcium from intracellular stores with consequent ASM contraction [34]. When hASMC were preincubated with DEX at concentrations >10 nM for 22 h, the IP response to His 100 µM was significantly (p < 0.05) inhibited [16]. In hASMC stimulated with a range of His concentrations 1 µM-1 mM, pretreatment with DEX 1 µM significantly (p < 0.05) inhibited the formation of IP irrespective of His concentration. The effect of DEX was evident only after a preincubation of ≥6 h, and after 22 h (p < 0.001) the mean reduction observed was of 48.0 ± 5.0% [16].
Goldsmith et al. [14] investigated whether CS inhibit the expression of ASM contractile proteins in primary hASMC stimulated with transforming growth factor beta (TGFβ) 1 ng/mL. DEX 0.1 µM and 1 µM for 6 days significantly (p ≤ 0.01) reduced the overexpression of α-actin (from 7.12 ± 1.77-fold increase to 2.43 ± 0.49 and 2.47 ± 0.29-fold increase, respectively) and the short isoform of myosin light chain kinase (MLCK). Since DEX had no modulatory effect on the hASMC number, the authors concluded that a reduction in cell proliferation could not be the cause of the observed decrements in ASM protein abundance. The gene overexpression of α-actin, the rate of α-actin mRNA degradation, the synthesis of α-actin in presence of the transcriptional inhibitor actinomycin D, and the α-actin turnover were not modulated by DEX [14].
In primary hASMC stimulated with TGFβ 1 ng/mL, FP 10 nM and 100 nM administered for 6 days significantly (p > 0.05) reduced the overexpression of α-actin (from 7.12 ± 1.77-fold increase to 2.34 ± 0.37 and 2.32 ± 0.66-fold increase, respectively, with FP at 10 nM and 100 nM) and reversed the shift in MLCK expression from the long to the short isoform [14]. Incubation for 48 h with FP did not affect (p > 0.05) the TGFβ induced-α-actin gene overexpression and the rate of mRNA degradation following the addition of the transcriptional inhibitor actinomycin D. Considering that α-actin was reduced at the protein level rather than at the mRNA level, the authors suggested a posttranscriptional control exerted by the CS [14]. In presence of actinomycin D, hASMC exposed to TGFβ for 24 h and incubated with FP showed a significant (p < 0.05) reduction in α-actin protein synthesis and improved α-actin protein turnover only when administered at 100 nM. FP decreased the TGFβ-induced incorporation of α-actin into filaments and significantly (p < 0.001) reduced cell length and contractile function in response to stimulation with acetylcholine (aCh) and potassium chloride (KCL) when administered at 1-100 nM concentrations [14].
Lewis et al. [20] evaluated spontaneous contraction in hASMC incubated alone or co-cultured with human lung mast cells (LMC). Incubation of hASMC or hASMC-LMC co-culture embedded in collagen gels with FP 10 µM for 16 h did not significantly (p > 0.05) affect the spontaneous contraction.

BUD
In hASMC stimulated by IL-1β 10 U/mL to reproduce a cellular model of AHR in vitro, incubation with BUD 1 nM and 100 nM for 3 h numerically but not significantly (p > 0.05) reduced the gene overexpression of bradykinin B 2 receptors [17]. When hASMC were pretreated with BUD 100 nM for 6 h, the synthesis of IP induced by bradykinin 10 µM was significantly (p < 0.05) reduced (from 12874.74 ± 739 dpm/assay to 6776.18 ± 585 dpm/assay) [17].

DEX Plus LABA
In primary hASMC exposed to TGFβ 1 ng/mL, co-incubation with DEX 0.1-1 µM and salmeterol (SAL) 1 nM for 6 days significantly (p ≤ 0.01) reduced the overexpression of α-actin (from 7.12 ± 1.77 to a maximum of 3.05 ± 0.98) and the short isoform of MLCK [14]. Combining DEX with SAL did not result in a change in hASMC proliferation, therefore the authors argued against reduced cell number as the cause of the observed decrease in ASM proteins [14].

FP Plus LABA
In primary hASMC exposed to TGFβ 1 ng/mL, co-incubation with FP 10 nM and SAL 1 nM for 6 days significantly (p ≤ 0.01) reduced the overexpression of α-actin from 7.12 ± 1.77 to 3.46 ± 1.02-fold increase, while FP 100 nM plus SAL induced a numerical decrease [14]. Adding SAL to FP 10-100 nM significantly (p ≤ 0.01) reversed the TGFβinduced shift in MLCK expression from the long to the short isoform. Combining FP with SAL did not result in a change in hASMC proliferation [14].
In human bronchial myofibroblasts stimulated with TGFβ 5 ng/mL, co-incubation with FP 1 pM and SAL 10 nM for 24 h significantly (p ≤ 0.05) inhibited the protein overexpression of α-actin from 96.7 ± 1.5% to 2.3 ± 0.6% [21]. Within 30 min from FP and SAL administration, the contractile activity developed in single myofibroblasts disappeared in both young and aged cultures [21].

DEX
One study [23] evaluated the impact of CS in precision cut lung slices (PCLS; small airways characterized by an inner diameter <2 mm) collected from non-asthma donors, incubated overnight with human immunoglobulin E (IgE), and stimulated by carbachol (CCh) 100 µM. Following cross-linking with high-affinity IgE receptor (FcεRI), overnight treatment with DEX 1 µM did not significantly (p > 0.05) modulate the FcεRI-dependent ASM contractility [23].
Cazzola et al. [4] investigated the rapid non-genomic bronchorelaxant effect of BDP administered in medium bronchi and PCLS submaximally contracted with His. In non-sensitized medium bronchi, BDP modestly relaxed the histaminergic ASM tone in a concentrationdependent manner, reaching a maximal relaxant response (E max ) of 33.41 ± 3.47% and a potency (expressed as the negative logarithm of the half-maximal effective concentration [pEC 50 ]) of 4.79 ± 0.15. BDP was significantly (p < 0.001) more effective in passively sensitized medium bronchi, with an E max of 43.61 ± 2.06% and a pEC 50 of 5.09 ± 0.23. Pre-treatment with the Gsα subunit G protein antagonist NF449 and the cyclic-adenosinemonophosphate (cAMP)-dependent protein kinase A (PKA) inhibitor KT5720 significantly (p < 0.001) suppressed the non-genomic bronchorelaxant action of BDP in passively sensitized tissues, but not in non-sensitized ones, suggesting that the CS effect was dependent from the activation of Gsα-cAMP-PKA cascade [4]. In PCLS, the bronchorelaxant effect of BDP was significantly (p < 0.001) greater in passively sensitized tissues (E max 63.89 ± 5.09% and pEC 50 7.23 ± 0.27) than in non-sensitized ones (E max 31.94 ± 3.01% and pEC 50 7.27 ± 0.37) [4].
A recent ex vivo study [24] performed in medium bronchi contracted by transmural stimulation investigated whether pre-incubation with BDP for 1 h could abolish the AHR induced by cow's milk (CM) aspiration, which has been implicated in the etiology of various inflammatory lung diseases. In tissues not challenged with CM, BDP 1-10 µM did not significantly (p > 0.05) modulate the ASM contractile tone. In airways challenged with CM 1:10 v/v for 60 min, BDP administered at 1 µM and 10 µM, but not at 0.1 µM, significantly (p < 0.05) lowered the ASM contractility by −52.49 ± 10.97% and −66.98 ± 7.90%, respectively [24].

BDP Plus LABA
Combining cumulative concentrations of BDP with FOR at a 100:6 concentration ratio induced a significant (p < 0.01) synergistic relaxant effect in medium bronchi pre-contracted by His, both in non-sensitized and passively sensitized tissues [22]. In non-sensitized medium bronchi, the maximal synergistic bronchorelaxant response was achieved with BDP/FOR 10/0.6 mg/mL and it was +28.73 ± 7.25% greater than the expected additive effect as predicted by the Bliss Independence equation, while in passively sensitized tissues the synergism remained stable over the range of concentrations 1/0.06-100/6 ng/mL, with a maximal synergistic bronchorelaxant response of +12.74 ± 4.62% compared to the additive effect [22]. A synergistic interaction was already observed at low concentrations of BDP/FOR inducing ≤25% of the E max , whereas for concentrations eliciting ≥50% E max the extent of synergism was strong [22].
In non-sensitized and passively sensitized PCLS pre-contracted by His, increasing concentrations of BDP/FOR induced a significant (p < 0.001) synergistic bronchorelaxation. The maximal synergistic interaction was achieved by BDP/FOR 1/0.06 ng/mL in non-sensitized tissues (+20.41 ± 4.10% vs. expected additive effect) and by BDP/FOR 10/0.06 ng/mL in passively sensitized airways (+20.04 ± 2.18%). In non-sensitized PCLS, the synergistic interaction was greater at higher concentrations; the magnitude of interaction was strong at concentrations inducing 25-50% E max , and very strong at higher concentrations. In passively sensitized PCLS, BDP/FOR produced a greater synergistic interaction when administered at lower concentrations; the extent of synergism was very strong over the range of concentrations inducing 15-25% E max and strong for concentrations inducing 75% E max [22].
In medium bronchi collected from COPD donors, the concentration of BDP/FOR administered at a 100:6 concentration ratio that reduced by 50% the ASM contractility elicited by CCh (Ab EC 50 ) was 2.31 ng/mL, while in PCLS, it was 4.59 ng/mL [27]. The potency of BDP/FOR was 1.56 ng/mL in medium bronchi and 0.97 ng/mL in PCLS. The combination effectively decreased the CCh-induced contractile tone in medium bronchi, reaching an E max of 86.90%, whereas in PCLS the bronchorelaxant effect was only partial, with an E max of 51.83% [27].

MF Plus LABA
In passively sensitized medium bronchi pre-contracted by His, combining high, rather than medium concentrations of MF with the LABA indacaterol (IND) at a 100:90 molar ratio induced a significant (p < 0.05) bronchorelaxant effect and the maximal synergistic interaction was +17.61 ± 8.34% greater than the expected additive effect. The magnitude of synergistic interaction was strong to very strong [26]. In passively sensitized PCLS, combining medium concentrations of MF with IND at a 100:45 molar ratio elicited a significant (p < 0.05) bronchorelaxant effect on the histaminergic contractile tone, reaching an E max +20.97 ± 7.47% greater than the expected additive effect, and the magnitude of synergistic interaction was very strong. When high concentrations of MF were combined with IND, the synergism significantly (p < 0.001) increased, reaching an E max +27.36 ± 12.40% greater than the expected additive effect and the magnitude of interaction ranged from middle to very strong [26].

Triple Combinations Including BDP
In an ex vivo model of bronchial asthma, combining BDP with the LABA FOR, and the LAMA GLY at a 100:6:12.5 concentration ratio produced a significant (p < 0.05) synergistic bronchorelaxant effect in passively sensitized medium bronchi and PCLS submaximally contracted by His [25]. The maximal synergistic interaction was achieved with BDP/FOR/GLY 1/0.06/0.125 ng/mL (+43.57 ± 0.96% vs. expected additive effect). The extent of synergism was very strong and overall stable across the range of concentrations inducing 25-90% E max . In passively sensitized PCLS, the maximal synergistic interaction was achieved with BDP/FOR/GLY 10/0.6/1.25 ng/mL (+24.95 ± 7.85% vs. expected additive effect). When administered at concentrations inducing 25-75% E max , BDP/FOR/GLY produced a very strong synergism, while at concentrations inducing 90% E max , synergism was strong [25].
A more recent ex vivo study [27] confirmed that in medium bronchi collected from COPD donors and submaximally contracted by CCh, the triple BDP/FOR/GLY combination administered at a 100:6:12.5 concentration ratio produced a significant (p < 0.05) synergistic bronchorelaxant effect. Synergism was observed with BDP/FOR administered at concentrations of 0.318-31.8 ng/mL plus GLY at concentrations of 0.0375-3.75 ng/mL. The maximal synergistic interaction was reached when BDP/FOR 1.06 ng/mL was combined with GLY 0.125 ng/mL, reaching an improved bronchorelaxation of +32.00-35.00%, compared to the expected additive effect [27].
In PCLS collected from COPD donors and submaximally contracted with CCh, BDP/FOR/GLY administered at a 100:6:12.5 concentration ratio induced a supra-additive effect, as resulted from the analysis of interaction via Bliss model; according to the Loewe and HSA models, a significant (p < 0.05) synergistic bronchorelaxant response resulted when BDP/FOR administered at concentrations of 1.06-31.8 ng/mL was combined with GLY at concentrations of 0.125-3.75 ng/mL. Maximal synergism was detected for BDP/FOR 10.6 ng/mL combined with GLY 1.25 ng/mL, leading to an improved bronchorelaxant effect of +36.00-37.00%, compared to the expected additive effect [27].

Triple Combinations Including MF
In passively sensitized medium bronchi and PCLS pre-contracted by His, combining medium concentrations of MF with the LABA IND and the LAMA GLY at a 100:37:45 molar ratio produced a significant (p < 0.05) synergistic bronchorelaxant effect and the E max achieved was +22.94 ± 13.81% greater than the additive effect [26]. Synergism was further significantly (p < 0.001) increased when high MF concentrations were combined with IND and GLY at a 100:37:90 molar ratio, and the E max achieved was +28.73 ± 2.59% greater than the additive effect. The magnitude of synergistic interaction was always very strong, irrespective of MF concentration. In passively sensitized PCLS, treatment with medium concentrations of MF, IND, and GLY at a 100:37:45 molar ratio produced a significant (p < 0.001) synergistic bronchorelaxant response, reaching an E max +45.00 ± 4.41% greater than the additive effect. The synergistic interaction was further significantly (p < 0.001) enhanced when high concentrations of MF were combined with IND and GLY at the 100:37:90 molar ratio, and the E max was +53.72 ± 9.10% compared with the expected additive effect. The magnitude of synergistic interaction was always very strong, irrespective of MF concentration [26].

FP
Currie et al. [29] performed an RCT in mild asthmatic patients to characterize the impact of FP 250 µg twice daily (BID) on AHR, defined by the methacholine (MCh) challenge test. Treatment with FP for 3 weeks significantly (p < 0.05) increased the MCh provocative dose causing a 20% fall in FEV 1 (PD 20 ), by producing a doubling dose improvement of 1.6 (95% CI 0.8-2.3) vs. the baseline [29].

BUD
O'Connor et al. [31] investigated the effect of AMP and sodium metabisulfite (MBS) in mild asthmatic patients. Subjects underwent a bronchoprovocation challenge with inhaled AMP, MBS, and MCh before and after 2 weeks of treatment with BUD 0.8 mg BID [31]. Compared to the placebo, BUD significantly (p < 0.01) reduced the AHR to MBS and MCh to a similar extent, shifting the dose-response curve of each agonist to the right by 1.06-dd (0.34-1.78) and 1.17-dd (95% CI 0.34-2.00), respectively. BUD induced a further significant (p < 0.01) reduction in the AHR to AMP vs. placebo and the other challenges, shifting rightward the dose-response curve by 2.92-dd (2.12-3.72) [31].
Kelly et al. [30] conducted an RCT to evaluate the effect of 11 days of treatment with BUD on AHR in patients with mild atopic asthma, before and after an allergen inhalation challenge at day 9. Compared to the baseline, BUD 400 µg BID significantly (p < 0.05) increased the PC 20 to MCh at day 8 of treatment pre-allergen challenge and prevented the allergen-induced AHR at day 11 of treatment.
According to a clinical study [32] performed on patients with mild asthma, 4 weeks of treatment with BUD 200 µg BID significantly (p < 0.05) increased the PC 20 to MCh from 3.7 ± 2.7 mg/mL to >16 mg/mL vs. the baseline.

PSL
Yick et al. [33] evaluated the effect of 2 weeks of oral treatment with PSL 0.5 mg/kg/day on AHR and investigated changes in the ASM transcriptomic profile in endobronchial biopsies of patients with atopic asthma. PSL numerically but not significantly (p > 0.05) increased MCh PC 20 vs. the baseline and vs. the placebo. Across the 15 genes modulated by treatment with PSL, the FAM129A and SYNPO2 genes resulted to be significantly (p < 0.01) correlated with AHR (r = −0.740 and r = −0.746, respectively) [33].

RoB and Quality of Evidence
Of the six clinical studies [28][29][30][31][32][33] included in the systematic review, five RCTs [28][29][30][31]33] were assessable via the Cochrane RoB 2, whilst the study by Williams et al. [32] was neither randomized nor controlled, therefore it was not feasible for RoB judgement. All the studies (five, 100.0%) had a low risk of bias for deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported result. Three RCTs (60.0%) did not report information for the RoB in the randomization process, and the other two studies (40.0%) presented a low risk of bias. The overall RoB was low for all the included RCTs (five, 100.0%). Detailed information concerning the RoB assessment is reported in Figure 2. All the included RCTs were ranked as being of medium quality, in agreement with the Jadad score (Table 1).

Figure 2.
Traffic light plot for the assessment of the risk of bias of each included randomized trial (A) and weighted plot for the assessment of the overall risk of bias via the Cochrane RoB 2 tool (B) (n = 5 studies). Traffic light plot reports five risk-of-bias domains: D1, bias arising from the randomization process; D2, bias due to deviations from the intended intervention; D3, bias due to missing outcome data; D4, bias in measurement of the outcome; and D5, bias in the selection of the reported result; a green circle represents a low risk of bias and a blue circle indicates insufficient information on the risk of bias. RoB: risk of bias [28][29][30][31]33].

Discussion
The findings resulting from this systematic review indicate that, generally, CS have significant genomic/non-genomic beneficial effects on human ASM contractility and AHR, regardless of their anti-inflammatory effects (Figure 3). More specifically, CS are effective in vitro in reducing either the expression, synth or activity of -actin, CD38, IP, MLCK, and RhoA in response to several stimuli (i.e., A bradykinin, IL-13, KCL, TGF, and TNF-) that increase the contractile tone of hAS [35][36][37]; overall these effects were mediated by the genomic action of CS. Moreover, have been demonstrated to elicit strong bronchorelaxant effects in both small and med human isolated airways via the rapid activation of the Gs-cAMP-PKA pathway aft passive sensitization procedure, which is a validated model mimicking ex vivo the hyp More specifically, CS are effective in vitro in reducing either the expression, synthesis or activity of α-actin, CD38, IP, MLCK, and RhoA in response to several stimuli (i.e., ACh, bradykinin, IL-13, KCL, TGFβ, and TNF-α) that increase the contractile tone of hASMC [35][36][37]; overall these effects were mediated by the genomic action of CS. Moreover, CS have been demonstrated to elicit strong bronchorelaxant effects in both small and medium human isolated airways via the rapid activation of the Gsα-cAMP-PKA pathway after a passive sensitization procedure, which is a validated model mimicking ex vivo the hyperreactivity of asthmatic bronchial tissue [38]. Interestingly, data originating from ex vivo studies were corroborated by clinical studies [28][29][30][31][32] carried out in mild asthmatic patients, reporting that CS are effective in reducing or preventing AHR elicited by different pro-contractile stimuli (i.e., AMP, MBS, and MCh). A summary of the beneficial effects of specific CS, alone or in combination with bronchodilators, against human ASM contractility and AHR in vitro, ex vivo, and in clinical trials is reported in Tables 2 and 3. The evidence that, when administered in combination with a LABA, CS induced significant synergistic bronchorelaxant effects in passively sensitized small and medium human bronchi fully supports the current global initiative for asthma (GINA, 2022) approach in which an ICS are recommended for Steps 1-5 [39]. Effectively, in this range of treatments, ICS-FOR was suggested as the preferred controller and reliever therapy by modulating the dose of ICS in the fixed-dose combination (FDC) according to the disease severity [39]. To date ICS, always combined with FOR, represents the cornerstone of asthma treatment not only as a maintenance therapy, but also as-needed for the relief of symptoms and, if needed, before exercise [39]. Of note, this pharmacological approach reduces the risk of exacerbation compared with using a short-acting β 2 -adrenoceptor agonist (SABA) reliever [39]. Table 3. Statistically significant (p < 0.05) effects of specific CS administered in dual and triple combinations with bronchodilators against human ASM contractility and AHR in vitro and ex vivo studies as resulting from this systematic review. ICS-FOR as a MAintenance and Reliever Therapy (MART), currently recommended at Step 3-5 [39], resulted in acute and dose-related anti-inflammatory effect in symptomatic asthmatic patients [40]. Interestingly, in the same patients and in an acute setting, high-dose MART also exerted significant improvements in FEV 1 compared to the SABA terbutaline [40].

Corticosteroids Administered in Combination with
Triple ICS/LABA/LAMA FDC is currently considered an alternative treatment at Step 4 as well as a preferred therapy at Step 5 [39]. As a matter of fact, a strong to very strong synergistic interaction has been proved ex vivo among ICS, LABA, and LAMA [25,26]. Furthermore, triple ICS/LABA/LAMA FDC exerted ceiling bronchorelaxation at the level of small airways in humans, by improving hyperinflation in more severe patients and leading to substantial clinical benefits [27]. Therefore, there is the pharmacological rationale for combining an ICS with a LABA plus a LAMA, both characterized by a rapid bronchorelaxant onset and long duration such as FOR plus GLY [41][42][43], and administered as a Triple MAintenance and Reliever Therapy (TriMART). The possibility of modulating the dose of the ICS in the triple BDP/FOR/GLY FDC (lower dose 100/6/12.5 µg, higher dose 200/6/12.5 µg) [44] makes the novel TriMART approach a potential therapeutic strategy for asthmatic patients at Step 3-5, especially for those subjects who may benefit from a sustained bronchodilation and suffering from increased parasympathetic tone [45].
Another important point arising from this systematic review is that ICS can be ineffective in preventing AHR in smoking patients with asthma, a condition related to the presence of relative steroid resistance due to the impairment of histone deacetylase 2 (HDAC2) [28,46]. In these smoking asthmatics, perhaps adding drugs that are able to restore HDAC2 activity such as doxofylline may help support the therapeutic effect of ICS [47][48][49][50][51].
This systematic review has certainly some limitations, generally intrinsic to primary publications. First, although the large body of evidence resulting from in vitro and ex vivo research supported the acute direct effects of ICS against human ASM contractility and AHR, ad-hoc translational studies are still missing for most of the ICS. However, we can speculate that the beneficial effects of ICS on airway contractility is an effect of class not specific for each single molecule. Second, the trial by Williams et al. [32] was not an RCT, therefore it was not possible to assess the quality of this clinical study, whose results should be interpreted with caution. Third, the remaining RCTs [28][29][30][31]33] included in this systematic review were proof-of-concept studies that enrolled a small number of patients. Regrettably to date, it seems that excluded the evident interest raised from ex vivo research; the direct impact of ICS on ASM contractility and AHR is no longer a hot topic in clinical research in asthma and COPD.
In conclusion, while the genomic effects of CS have been well-characterized with respect to the expression, synthesis, and activity of pro-contractile factors, current evidence suggests that CS may also elicit rapid non-genomic effects on human airways via the Gsα-cAMP-PKA pathway, a cascade activated by one or more specific CS membraneassociated receptors [52,53]. However, since the existence of such distinct non-classic CS membrane receptors has been not yet proven [3], further studies are needed to demonstrate the expression of these putative receptors on hASMC.
Funding: This manuscript was not funded.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this study are available in the article.