A retrospective cohort study to evaluate the relationship of airway hyperresponsiveness to type 2 biomarkers in persistent asthma

A retrospective cohort study to evaluate the relationship of airway hyperresponsiveness to type 2 biomarkers in Abstract Airway hyperresponsiveness (AHR) is a hallmark of persistent asthma measured using direct or indirect airway bronchial challenge testing. The purpose of this study is to investigate the putative relationships between type 2 inﬂ ammatory biomarkers, airway geometry (FEV 1 and FEF 25-75 ) and speciﬁ c IgE (RAST or skin prick) to AHR. We performed a retrospective analysis of our database ( n = 131) of patients with asthma. Of these subjects, 75 had a histamine challenge and 56 had a mannitol challenge. Fractional exhaled nitric oxide (FeNO) and speciﬁ c immunoglobulin E (IgE) but not blood eosinophils were signiﬁ cantly higher in patients with AHR to either histamine or mannitol. FEV 1 % and FEF 25 - 75 % were signiﬁ cantly lower in patients with AHR. Elevated Type 2 biomarkers including FeNO and speciﬁ c IgE but not blood eosinophils were associated with AHR. Highlights: FeNO and speciﬁ c IgE but not blood eosinophils are raised in patients with airway hyperresponsiveness.


Background
Airway hyperresponsiveness (AHR) is a hallmark of persistent asthma and may be measured using direct histamine or indirect mannitol challenge. Direct airway challenges, using agents such as methacholine and histamine, are highly sensitive but not particularly speci ic for the detection of asthma [1]. They act directly by causing airway smooth muscle constriction resulting in reduced airway calibre. Indirect airway challenges such as mannitol and adenosine monophosphate (AMP) act by causing release of endogenous mediators such as leukotrienes and histamine which in turn stimulate airway smooth muscle constriction with or without microvascular leakage [2].
AHR is useful to clinicians as its severity correlates well with asthma severity and the amount of treatment required to control symptoms. AHR has traditionally been characterised by increased sensitivity to constrictor agents, a steeper slope of the dose response curve and a greater maximal response overall [3].
To better understand the concept of AHR, it can be useful to further divide its components into persistent and variable factors. Persistent factors are largely a consequence of structural airway changes including subendothelial thickening, smooth muscle hypertrophy, matrix deposition and vascular changes. These alterations in airway geometry are associated with a greater degree of airway constriction when stimulated by contractile agents. The more variable portion of AHR is thought to be related to and in luenced by environmental factors such as allergens, respiratory infections and treatment [4]. It is well recognised that airway in lammation and remodelling are associated with AHR [5].

Abstract
Airway hyperresponsiveness (AHR) is a hallmark of persistent asthma measured using direct or indirect airway bronchial challenge testing. The purpose of this study is to investigate the putative relationships between type 2 infl ammatory biomarkers, airway geometry (FEV 1 and FEF  ) and specifi c IgE (RAST or skin prick) to AHR. We performed a retrospective analysis of our database (n = 131) of patients with asthma. Of these subjects, 75 had a histamine challenge and 56 had a mannitol challenge. Fractional exhaled nitric oxide (FeNO) and specifi c immunoglobulin E (IgE) but not blood eosinophils were signifi cantly higher in patients with AHR to either histamine or mannitol. FEV 1 % and FEF 25 -75 % were signifi cantly lower in patients with AHR. Elevated Type 2 biomarkers including FeNO and specifi c IgE but not blood eosinophils were associated with AHR.
Highlights: FeNO and specifi c IgE but not blood eosinophils are raised in patients with airway hyperresponsiveness.
Additionally, AHR is associated with peripheral blood eosinophilia in asymptomatic individuals [6]. The degree of AHR correlates with sputum eosinophils, a steeper FEV 1 decline, disease severity and worse reported symptoms [7][8][9]. We have previously reported on the relationship between airway geometry and inhaled corticosteroid (ICS) dose to AHR using methacholine and AMP [10].
We therefore performed a retrospective analysis of our Scottish database of asthma patients with the aim to investigate the relationship between ICS dose, allergy and baseline spirometry to AHR. Furthermore, we explored the putative connection between type 2 in lammatory (T2) biomarkers and AHR. Instead of methacholine and AMP as in our previous study [10], we used histamine and mannitol for our challenge agents.

Database
The database consisted of 131 patients with known persistent asthma taking ICS, recruited retrospectively from the Scottish Centre for Respiratory Research who had previously attended for screening into clinical trials or had attended a National Health Service (NHS) specialist respiratory clinic. All patients had physician-diagnosed asthma based on history and objective testing and were all taking inhaled corticosteroid. As part of their clinical trial, patients were characterised according to spirometry, skin prick allergy testing, histamine or mannitol challenge testing, FeNO and blood eosinophil count. Instead of skin prick testing, NHS patients underwent blood sampling for allergen speci ic IgE testing i.e. radioallergosorbent testing (RAST). T2 biomarkers were obtained within 6 months of airway challenge testing.

Spirometry
Spirometry (Micromedical, Chatham, UK) was performed according to American Thoracic Society (ATS) and European Respiratory Society (ERS) guidelines [11]. Prior to attending the laboratory for spirometry and airway challenge, patients had been asked not to use their short acting beta-2 agonists for 6 hours, long acting beta-2 agonists and muscarinic antagonists, theophyllines and leukotriene receptor antagonists for 48 hours.

Allergy testing
Skin reactivity to common aeroallergens (grasses, trees, house dust mite, aspergillus, dog and cat) was determined with skin prick tests (Diagenics Ltd, Milton Keynes, UK) on the volar aspect of the forearm, using a standard puncture technique [12]. Saline solution (0.9%) and histamine (1 mg/ml) were used as negative and positive controls, respectively. Wheal and lare size were measured 15 minutes after administration of allergens and a positive reaction was deemed 2mm or larger than negative control.
Blood testing was performed to detect presence of circulating levels of speci ic IgE antibodies to de ined common allergens [Fluorescence enzyme linked immunoassay (Phadia Immunocap 250)]. In our NHS laboratory a speci ic IgE concentration greater than 0.35 kUA/L is considered a positive RAST test. We characterised allergy as the number of positive skin or RAST responses in an individual.

Bronchial challenge
Histamine was dispensed via nebuliser solution (Tayside Pharmaceuticals, Dundee, UK) and airway challenge was performed using a Mefar dosimeter with doubling concentrations up to a maximum of 32 mg/ml in accordance with ATS guidelines. The provocative concentration of histamine required to cause a 20% fall in FEV 1 (PC20) was calculated by logarithmic interpolation of the log doseresponse curve. A positive challenge was considered to be a PC20 < 8 mg/ml. Mannitol was given via dry powder inhaler (Aridol, Pharmaxis Ltd, Sydney, Australia) in dose increments up to a maximum cumulative dose of 635 mg until a fall in FEV 1 of 15% from baseline was achieved. The mannitol cumulative dose resulting in a 15% fall in FEV 1 (PD15) was calculated by linear interpolation of the log dose-response curve as previously described, with a value of < 635 mg being considered a positive test [14]. We elected to use the same threshold sensitivity values for mannitol and histamine challenges as previously reported [15].

Statistical analysis
Data were irst analysed for normality with Shapiro-Wilk tests and Boxplots. Data for FeNO were logarithmically transformed to normalise the distribution prior to analysis. Independent Student's T tests with alpha error set at 0.05 (2-tailed) were used to determine differences in FeNO, FEV 1 , FEF  and allergy according to AHR status i.e. positive vs. negative tests. Receiver operator characteristic (ROC) curves were also plotted to evaluate sensitivity and speci icity of FeNO for detecting pre-test probability of AHR. We calculated a beclomethasone equivalent daily ICS dose for the purposes of analysis. Analysis was performed using Statistical Products and Service Solutions (SPSS) for Windows Version 25 by International Business Machines Corporation (IBM).

Ethics
Caldicott Guardian approval was obtained to allow access to any National Health Service patient identi iable data including allergy, airway challenge testing, blood eosinophils, FeNO and spirometry.
All clinical trial patients consented to use of their data.

Results
131 patients taking ICS, presenting with a known diagnosis of persistent asthma over a 2-year period were entered into the database. Demographic data are shown in table 1. Of these, 75 had a histamine challenge and 56 underwent a mannitol challenge with 73/115 (63%) of subjects being characterised as allergic on the basis of at least one positive skin prick or RAST to a common aeroallergen panel. In the combined group, all patients were taking ICS at a mean daily dose of 858 μg; 55/131 (42%) were taking a long acting beta-2 receptor agonist; 16/131 (12%) were taking a long acting muscarinic antagonist; 32/131 (24%) were taking a leukotriene receptor antagonist and 3/131 (2%) were taking theophylline. Both FEV 1 % and FEF 25-75 % were signi icantly lower in patients with positive versus negative AHR (Table 2) when the challenges were combined (Figure 2), and also for histamine responders on their own. FeNO was shown in combined (Figure 2), histamine and mannitol groups to be signi icantly higher in patients with AHR compared to those without. Patients with AHR demonstrated a signi icantly higher allergic burden (Figure 2) in terms of number of positive skin or RAST responses. This difference was largely attributed by patients in the histamine subgroup as it was not signi icant in the mannitol subgroup.
In a combined ROC analysis of AHR responders vs. nonresponders (Figure 1a), the AUC value was 0.765 with p < 0.001. A FeNO threshold of > 14 ppb was associated with a sensitivity of 82% and speci icity of 51% in identifying patients with a positive histamine or mannitol challenge test. In the histamine subgroup (Figure 1b), AUC was 0.754, p < 0.001 while FeNO > 14 ppb resulted in a sensitivity of 80% and speci icity of 56%. For the mannitol subgroup (Figure 1c), AUC was 0.786; p < 0.001 and FeNO > 16 ppb had a sensitivity of 82% and speci icity of 50%. ROC analysis for eosinophils, allergic burden, FEV 1 or FEF  were not signi icant.
There were no signi icant differences in blood eosinophils for the combined group when comparing AHR positive vs.

Discussion
Our study demonstrated elevated levels of FeNO and speci ic IgE but not blood eosinophils in relation to AHR. Moreover, patients with AHR also had altered geometry as FEV 1 and FEF  . We observed that 57% of our patients had a positive challenge with either histamine or mannitol. Whether or not this infers that the remainder of our patients who were challenge negative did not have asthma is debatable. Ideally this would require a repeat challenge having had a washout period of at least 2 weeks without ICS. AHR non-responders received signi icantly higher ICS doses, which is in keeping with the known dose related suppressive effects of ICS on AHR and type 2 in lammation [16,17]. Non responders also had signi icantly better pulmonary function. In this regard improved airway calibre would attenuate AHR due to effects on airway geometry per se.
It has previously been shown that FeNO levels are higher in patients who exhibit AHR to histamine, methacholine or mannitol [7,18,19]. Pointedly our study con irmed this inding when looking at patients prescribed ICS. A prospective study showed that titrating ICS against mannitol challenge during a 1 year period was associated with a 1.52 doubling dose shift in AHR, which in turn was accompanied by improved symptom control and fewer exacerbations [20]. Our results found no association between blood eosinophils and AHR which is in contrast to previous observations [6]. This may well re lect our patients being on a relatively high dose of ICS (mean 858μg) which is known to suppress blood eosinophils [21]. FeNO is   primarily regulated by IL13 and therefore indings from our study raise the question whether dupilumab, an anti-IL4rα biologic therapy, might perhaps have more bene icial effects of attenuating AHR than anti-IL5 [22].
The present study also demonstrates that airway geometry expressed as FEV 1 and FEF 25-75 is signi icantly lower in patients who exhibit AHR. The presence of airway remodelling may offer an explanation for the relationship between altered airway geometry and AHR, although without performing biopsies this assertion is speculative. Indeed, airway remodelling can theoretically augment the degree of AHR, separate from its well established effect on airway caliber [23]. In a prospective study over 2 years when ICS was titrated against methacholine AHR, attenuation of basement membrane thickness and reduced mucosal eosinophils were accompanied by improved AHR along with better lung function and symptoms [24]. Evidence to support a disconnect between airway calibre and AHR emanates from a study where dose related improvements in FEV 1 with muscarinic antagonist were not accompanied by a shift in histamine AHR [25].
Previous studies examining the relationship between degree of allergy and AHR have yielded contradicting results [26]. We elected to use the number of positive skin prick or RAST tests to pragmatically assess speci ic IgE and hence the overall allergic load. Our study demonstrated a signi icant difference in allergic burden using speci ic IgE according to the presence of AHR.
We recognize the limitations of our retrospective analysis. We would have liked to correlate these indings with asthma control or number of exacerbations. Our study also raises the pertinent question as to whether the presence of a negative challenge indicates that the patient does not have asthma or whether this might be a false negative result as a consequence of treatment modi ication with ICS. We might have stopped ICS for at least 2 weeks before performing challenge testing in order to address this question. However, concurrent ICS therapy is more likely to in luence AHR using direct versus indirect challenge [10,20]. For example in one study we observed that 30% of unselected patients with community managed asthma were challenge negative to either mannitol or methacholine [15]. These patients had a high burden of treatment exposure with a median beclomethasone equivalent daily dose of 1,000 μg along with 68% also taking LABA. This in turn might suggest a need for a supervised step down protocol to un-diagnose asthma in such patients [27].

Conclusion
In conclusion, FeNO and allergy but not blood eosinophils were signi icantly different when comparing AHR responders and non-responders. FeNO had a high sensitivity but low speci icity in relation to the presence or absence of AHR.

Ethics Approval and consent to participate
Caldicott Guardian approval was obtained to allow access to any National Health Service patient identi iable data including allergy, airway challenge testing, blood eosinophils, FeNO and spirometry. All clinical trial patients consented to use of their data.

Consent for publication: Consent has been obtained.
Availability of data and materials: The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
Dr. Chan has no relevant con licts of interest.
Dr. Kuo reports personal fees (talks) from AstraZeneca, personal fees (advisory board) from Circassia, personal fees (talks) in relation to the submitted work, and other support from Chiesi (attending BTS) outside of the submitted work.
Dr. Lipworth reports non-inancial support (equipment) from GSK; grants, personal fees (consulting, talks and advisory board), other support (attending ATS and ERS) and from AstraZeneca, grants, personal fees (consulting, talks, advisory board), other support (attending ERS) from Teva, personal fees (consulting) from Sano i, personal fees (consulting, talks and advisory board) from Circassia in relation to the submitted work; personal fees (consulting) from Lupin, personal fees (consulting) from Glenmark, personal fees (consulting) from Vectura, personal fees (consulting) from Dr Reddy, personal fees (consulting) from Sandoz; grants, personal fees (consulting, talks, advisory board), other support (attending BTS) from Boehringer Ingelheim, grants and personal fees (advisory board and talks) from Mylan outside of the submitted work; and the son of BJL is presently an employee of AstraZeneca.