Sputum periostin is a biomarker of type 2 inflammation but not airway dysfunction in asthma

Accurate phenotypic and endotypic characterization of asthma is critical to inform treatment decisions, particularly for patients with eosinophilic or type-2 (T2) high asthma who are more responsive to inhaled corticosteroids and are candidates for biologic therapies targeting T2 inflammation.1 Current biomarkers of T2 inflammation, including blood and sputum eosinophilia, and fractional exhaled nitric oxide (FENO), have limitations. Thus there is an unmet need for additional, more accurate biomarkers and to understand the relationship between such biomarkers and airway immunopathology.1 Periostin, a matrix protein secreted by epithelial and stromal cells and a biomarker of T2 inflammation in clinical research, can be measured in sputum and has been associated with the T2-high endotype and persistent airflow obstruction in patients with severe asthma.2 The relationship of sputum periostin to airway dysfunction in the form of airway hyperresponsiveness (AHR) and airway inflammation was examined in individuals with mild-to-moderate asthma who were not using controller therapies. In this cross-sectional assessment, individuals with and without asthma were recruited based on rigorous diagnostic testing, with further characterization of participants for endogenous AHR in the form of exercise-induced bronchoconstriction (EIB).3, 4 Briefly, individuals underwent methacholine challenge testing, dry air exercise challenge, induced sputum collection and research bronchoscopy. Periostin levels in induced sputum supernatant were measured at a 1:2 dilution using a sandwich ELISA assay (Genentech, Inc., South San Francisco, CA; lower limit of quantitation [LLOQ]: 37 pg/mL).5 We assessed induced sputum cell differentials and expression of selected genes (IL4, IL5, IL13, ARG2, INFG, TPSAB1, CMA1, and CPA3) by qPCR. Individuals were categorized for T2 inflammation using the sputum Type-2 Gene Mean (T2GM), with T2-high asthma defined by a T2GM ≥2 standard deviations above that of healthy controls within the study population.3, 4, 6 Endobronchial biopsies were assessed by immunohistochemistry and quantified by design-based stereology, to precisely quantify the numerical density of mast cells and eosinophils per reference volume within different compartments of the airway wall.3, 4 Nonparametric statistics were used. Correlations between continuous variables were assessed using Spearman's rho (r). Forty-three study participants had sputum periostin measured, and all but two had periostin levels above the LLOQ including 10 healthy controls (median age [IQR], 24.5 [23–34.3]; % females, 80%) and 33 individuals with asthma (median [IQR] for age, 23 [21–28.5]; FEV1% predicted, 92 [83–95] and FEV1/FVC ratio, 0.77 [0.71–0.84]; % females, 72.7%). There was no difference in sputum periostin concentration between healthy controls and subjects with asthma (median [IQR], 0.30 [0.09–0.69] vs. 0.18 [0.06–0.45]; p = 0.51) and there was no significant association between sputum periostin levels and baseline lung function or the severity of AHR to methacholine (Table 1). There was also no association between sputum periostin concentration and baseline lung function amongst individuals with asthma (Table 1). Amongst individuals with asthma, 20 (60.6%) had a positive exercise challenge test (EIB+). Sputum periostin concentration was similar between EIB− and EIB+ asthmatics (median [IQR], 0.14 [0.06 to 0.23] vs. 0.20 [0.12 to 0.74]; p = 0.16) and did not correlate with the severity of endogenous AHR (Table 1). Airway T2 gene expression was available from 31 individuals, with 14/31 classified as T2-high according to their T2GM. Our prior analysis of this cohort found that the T2GM and individual T2 cytokines correlate with the severity of AHR.3, 4 Here, sputum periostin was non-significantly higher amongst T2-high individuals, across the full study population (median [IQR], 0.36 [0.18–0.68] vs. 0.19 [0.06–0.59], p = 0.14) and after restricting analysis to individuals with asthma (median [IQR], 0.36 [0.18–0.68] vs. 0.15 [0.05–0.38], p = 0.08). Further correlation analyses revealed a significant association between sputum periostin and the sputum T2GM (Table 1). Of the genes included in the T2GM, sputum periostin was most strongly correlated with the expression of IL4 and IL5, and to a lesser extent with IL13 in induced sputum cells (Table 1). Conversely, neither the expression of ARG2 nor INFG in sputum correlated with sputum periostin concentration (Table 1). Subjects with ‘eosinophilic asthma’, defined by sputum eosinophil percentage ≥3%, had a higher sputum periostin concentration compared to non-eosinophilic asthma (median [IQR], 0.50 [0.20–1.09] vs. 0.16 [0.06–0.37], p = 0.03). Sputum periostin differentiated eosinophilic from non-eosinophilic asthma with an area under the ROC curve of 0.8 (95% CI, 0.63–0.97; p = 0.04). Across the full study population and amongst the individuals with asthma, sputum periostin correlated significantly with sputum eosinophil concentration (Table 1). We also further refined the relationship between sputum periostin and the precise location of eosinophils in the airway wall, revealing that sputum periostin concentration correlated with the density of subepithelial eosinophils, but not intraepithelial eosinophils (Table 1). Prior analysis of this cohort revealed that intraepithelial mast cell density and the expression of TPSAB1 and CPA3 in induced sputum were associated with the severity of EIB.4 No association was identified between sputum periostin and the density of mast cells in different compartments of the airway wall. However, expression of genes encoding key mast cell proteases implicated in T2 inflammation (TPSAB1, CPA3) correlated with sputum periostin concentration across the full study population (Table 1). From this pilot assessment, we show that sputum periostin is positively correlated with airway T2 gene expression and airway eosinophilia, but did not differ between individuals with and without asthma and was not associated with reduced lung function or with non-specific AHR to methacholine challenge. As the precise location of eosinophils and mast cells in the airway wall was characterized in this cohort, we demonstrated that sputum periostin correlated with the overall eosinophil density in the airway wall and eosinophils in the subepithelial compartment, the eosinophil location that is most closely associated with T2 airway inflammation.3 However, sputum periostin was not related to mast cells or eosinophils within the epithelial compartment, which are increased in subjects with asthma and tightly associated with airway dysfunction in the form of endogenous AHR, the most specific form of AHR in asthma.3, 4 Periostin has been proposed as a potential biomarker for asthma and for identifying individuals with T2 inflammation. However, serum periostin levels do not reliably differentiate adults with asthma from healthy controls7 but are consistently associated with classic peripheral T2 biomarkers2 and may predict responsiveness to T2-directed biologic therapies.1 Studies evaluating airway periostin levels in induced sputum have been limited with mixed results. One study found similar levels between individuals with mild-to-moderate asthma and healthy controls8 and another identified significantly higher sputum periostin levels in individuals with severe asthma.9 No prior studies have evaluated the details of the components of airway dysfunction with airway periostin levels, but serum periostin levels have been associated with AHR in children, particularly endogenous AHR.10 In children, asthma tends to be T2 predominant, so there may be less discordance between T2 inflammation and AHR, and periostin levels vary widely with age in children.11 We used a direct assessment of T2 gene expression in the airways (rather than peripheral biomarkers such as blood eosinophils or FENO) to demonstrate an association between the level of periostin in induced sputum and the T2 endotype in a well-defined cohort of individuals with mild-to-moderate asthma who were not using controller therapy at the time of assessment. We present evidence that sputum periostin is more closely associated with the T2 endotype and less of a marker of asthma or airway dysfunction, with additional details shown about the relationship of periostin to the precise location of mast cells and eosinophils in the airway wall. As few studies have characterized airway physiology and tissue infiltration with immune cells in relation to airway levels of periostin amongst individuals with asthma, these results add to the growing body of evidence of discordance between T2 biomarkers and airway dysfunction in the form of AHR in asthma. Taha Al-Shaikhly: Data curation (equal); formal analysis (lead); investigation (equal); writing – original draft (lead); writing – review and editing (equal). Ryan C. Murphy: Data curation (equal); formal analysis (equal); validation (equal); writing – original draft (equal); writing – review and editing (equal). Ying Lai: Investigation (equal); writing – review and editing (equal). Charles W. Frevert: Conceptualization (equal); formal analysis (equal); investigation (equal); methodology (equal); writing – review and editing (equal). Jason S. Debley: Formal analysis (equal); investigation (equal); validation (equal); writing – review and editing (equal). Steven F. Ziegler: Funding acquisition (equal); writing – review and editing (equal). Kit Wong: Formal analysis (equal); writing – review and editing (equal). Guiquan Jia: Formal analysis (equal); investigation (equal); writing – review and editing (equal). Cecile T. J. Holweg: Formal analysis (equal); writing – review and editing (equal). Michael C. Peters: Formal analysis (equal); investigation (equal); writing – review and editing (equal). Teal S. Hallstrand: Conceptualization (lead); data curation (equal); formal analysis (equal); funding acquisition (lead); investigation (lead); methodology (lead); project administration (lead); resources (lead); supervision (lead); validation (lead); writing – original draft (equal); writing – review and editing (equal). The work was supported by the National Institutes of Health (grant numbers: U19AI125378, K24AI130263 and R01HL153979). Taha Al-Shaikhly has a patent MicroRNAs as Predictors of Response to Anti-IgE Therapies in Chronic Spontaneous Urticaria pending. Charles W. Frevert, Jason S. Debley, Steven F. Ziegler, and Teal S. Hallstrand report grants from the National Institute of Health during the conduct of the study. Michael C. Peters reports grants from National Institute of Health-NHLBI, Boeringer-Ingelheim, Astrazeneca, Boehringer-Ingelheim, Genentech, GlaxoSmithKline, Sanofi-Genzyme-Regeneron, and Teva, outside the submitted work. Kit Wong, Guiquan Jia, and Cecile T. J. Holweg are employees of Genentech. All other authors declared no conflict of interest. This study was performed in accordance with the Declaration of Helsinki. This human study was approved by Institutional Review Board at the University of Washington (Seattle, Washington). All adult participants provided written informed consent to participate in this study. Data available on request from the authors. Visual Abstract Sputum periostin is a biomarker of type 2 inflammation but not airway dysfunction in asthma Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.