Voltar
Carta Acesso aberto Revisado por pares
BIBTEX RIS

Eosinophil peroxidase, GATA3, and T-bet as tissue biomarkers in chronic rhinosinusitis

2019; Elsevier BV; Volume: 143; Issue: 6 Linguagem: Inglês

10.1016/j.jaci.2019.01.038

ISSN

1097-6825

Autores

Devyani Lal, Benjamin L. Wright, Kelly P. Shim, Matthew A. Zarka, James J. Lee, Yu‐Hui Chang, Sergei I. Ochkur, Rohit Divekar, Alfred D. Doyle, Elizabeth A. Jacobsen, Hirohito Kita, Matthew A. Rank,

Tópico(s)

Asthma and respiratory diseases

Resumo

Sinonasal inflammation is the result of diverse pathogenetic mechanisms that are poorly understood, but important to elucidate because therapeutic strategies and outcomes of chronic rhinosinusitis (CRS) subtypes differ significantly.1Orlandi R.R. Kingdom T.T. Hwang P.H. Smith T.L. Alt J.A. Baroody F.M. et al.International consensus statement on allergy and rhinology: rhinosinusitis.Int Forum Allergy Rhinol. 2016; : S22-S209PubMed Google Scholar CRS with nasal polyps has been classically associated with TH2 patterns of inflammation, whereas CRS without nasal polyps has been associated with TH1 inflammatory patterns.1Orlandi R.R. Kingdom T.T. Hwang P.H. Smith T.L. Alt J.A. Baroody F.M. et al.International consensus statement on allergy and rhinology: rhinosinusitis.Int Forum Allergy Rhinol. 2016; : S22-S209PubMed Google Scholar However, recent research illustrates that these classical patterns are overly simplistic. CRS without nasal polyps may have eosinophilic and TH2 inflammatory signatures, and CRS with nasal polyps may also be associated with noneosinophilic, neutrophilic, and TH1 cytokine patterns (eg, in Asian populations and cystic fibrosis, etc). Indeed, many clusters can be identified in CRS that cross over the conventional eosinophilic/noneosinophilic-TH1/TH2 paradigms.2Tomassen P. Vandeplas G. Van Zele T. Cardell L.O. Arebro J. Olze H. et al.Inflammatory endotypes of chronic rhinosinusitis based on cluster analysis of biomarkers.J Allergy Clin Immunol. 2016; 137: 1449-1456.e4Abstract Full Text Full Text PDF PubMed Scopus (608) Google Scholar Clinical characteristics are unable to reliably predict inflammatory underpinnings3Steinke J.W. Smith A.R. Carpenter D.J. Patrie J. Payne S.C. Borish L. Lack of efficacy of symptoms and medical history in distinguishing degree of eosinophilia in nasal polyps.J Allergy Clin Immunol Pract. 2017; 5: 1582-1588Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar; clinicians therefore need biomarkers that can be used to subtype CRS so as to best identify inflammatory targets and personalize therapy. Eosinophil peroxidase (EPX) is a biomarker for eosinophilic inflammation in patients with asthma and eosinophilic esophagitis.4Wright B.L. Nguyen N. Shim K.P. Masteron J.C. Jacobsen E.A. Ochkur S.I. et al.Increased GATA-3 and T-bet expression in eosinophilic esophagitis versus gastrointestinal reflux disease.J Allergy Clin Immunol. 2018; 141: 1919-1921Abstract Full Text Full Text PDF PubMed Scopus (7) Google Scholar, 5Rank M.A. Ochkur S.I. Lewis J.C. Teaford III, H.G. Wesselius L.J. Helmers R.A. et al.Nasal and pharyngeal eosinophil peroxidase levels in adults with poorly controlled asthma correlate with sputum eosinophilia.Allergy. 2016; 71: 567-570Crossref PubMed Scopus (27) Google Scholar Some CRS subtypes have pathophysiologic underpinnings similar to allergic rhinitis and allergic asthma, which are disorders characterized by eosinophilic inflammation and TH2-dominated inflammation. The transcription factor guanine adenine thymine adenine sequence-binding protein 3 (GATA3) has been characterized as a master regulator of TH2 inflammation. T-box expressed in T cells (T-bet) is a master regulator for TH1 inflammation. In this study, we hypothesized that EPX, and the GATA3:T-bet (G:T) ratio, would be associated with clinical phenotypes previously linked to tissue eosinophilia and TH2 inflammation in patients undergoing sinus surgery. Institutional review board approval was obtained to perform this retrospective study. Surgical tissue of adult patients who underwent sinonasal surgery with the lead author (D.L.) and had structured histopathology results were studied further with immunohistochemical staining for EPX, GATA3, and T-bet. These methods and scoring system have been previously described by our group.4Wright B.L. Nguyen N. Shim K.P. Masteron J.C. Jacobsen E.A. Ochkur S.I. et al.Increased GATA-3 and T-bet expression in eosinophilic esophagitis versus gastrointestinal reflux disease.J Allergy Clin Immunol. 2018; 141: 1919-1921Abstract Full Text Full Text PDF PubMed Scopus (7) Google Scholar An automated pixel algorithm was used to quantify the number of EPX-positive pixels. A nuclear algorithm was used to identify cells that stained positive for GATA3 and T-bet. The biospecimens from this study were linked to demographic data (age and sex), clinical data (patient-reported 22-item sinonasal outcome test), sinus computed tomography (CT) scan data (Lund-Mackay scores), and standardized, structured histopathologic reports. Wilcoxon rank-sum test, ANOVA, or Kruskal Wallis test was used for continuous variables. Fisher exact test was used for categorical variables. Pearson correlation coefficient was used to present the relationship between GATA3, T-bet, and G:T and the other clinical, radiographic, and histopathologic variables. Statistical tests were performed using SAS 9.4 (SAS Institute, Minneapolis, Minn) and R 3.4.4 (R Core Team, Vienna, Austria). Demographic characteristics, clinical data, and radiographic data are presented in Table E1 in this article's Online Repository at www.jacionline.org. Ethmoidal tissue of 93 patients with CRS was studied. Controls included 8 patients without CRS who underwent sinonasal surgery to address recurrent sinusitis, fungal ball, and patients with allergic rhinitis undergoing surgery for nasal obstruction due to deviated septum and hypertrophied turbinates. Previous sinus surgery (P = .02) and the Lund-Mackay sinus CT score (P < .001) were significantly higher in patients with CRS than in controls. Histologic findings from hematoxylin and eosin stains are reported in Table E2 in this article's Online Repository at www.jacionline.org. The only significant difference between the CRS and control groups was the tissue eosinophil count; 52 (56.5%) patients with CRS had more than 10 eosinophils/hpf, whereas more than 10 eosinophils/hpf were not noted in any control subject (P = .002). Immunohistochemistry results for GATA3, T-bet, and EPX are presented in Table I; no significant difference between CRS and control tissue could be detected for GATA3, T-bet, and G:T ratio. However, EPX staining was significantly higher in CRS tissue compared with controls: 334 (290) versus 28 (30) (P < .001). Representative slides for T-bet, GATA3, and EPX stains are displayed in Fig E1 in this article's Online Repository at www.jacionline.org.Table ISpecial sinus immunohistochemical stainsTissue stainCRS (N = 93)Control (N = 8)Total (N = 101)P value∗Wilcoxon rank-sum test.EPX, +pixels/mm2, mean ± SD334 ± 29028 ± 30310 ± 291<.001GATA3 (positive cells/total cells)4.9 (4.2)4.1 (4.3)4.9 (4.2).45T-bet (positive cells/total cells)4.1 (3.2)4.3 (4.3)4.1 (3.3).85G:T (positive cells/total cells)1.4 (1.1)1.2 (0.8)1.4 (1.1).90∗ Wilcoxon rank-sum test. Open table in a new tab There were no significant between-group differences for inflammatory predominance in tissue (eg, eosinophilic and neutrophilic) based on staining for GATA3, T-bet, or G:T ratio (data not shown). However, there were significant differences when EPX tissue staining was compared across patients with CRS. As expected, patients with eosinophilic inflammatory predominance had higher EPX levels compared with other inflammatory categories: 544.2 (95% CI, 456.3-632.1) versus 185.5 (95% CI, 130.3-240.7; P < .001). In accordance with this result, patients with lymphoplasmocytic inflammatory predominance had lower EPX levels: 224.0 (95% CI, 152.9-295.1) versus 408.3 (95% CI, 321.8-494.9; P = .001). GATA3, T-bet, and G:T ratio were not significantly correlated with EPX or tissue eosinophil levels (Fig 1, A-G). However, GATA3 and T-bet were strongly and positively correlated: Pearson correlation coefficient = 0.69 (95% CI, 0.56-0.78; see Fig E2 in this article's Online Repository at www.jacionline.org). EPX positivity was significantly associated with the presence of nasal polyps (see Fig E3, A, in this article's Online Repository at www.jacionline.org), asthma status (Fig E3, B), aspirin intolerance (Fig E3, C), higher sinus CT scan score (Fig E3, D), and basement membrane thickening (Fig E3, E).Fig 1A, EPX is compared with the G:T ratio in subjects with CRS and controls. B, G:T by eosinophil count categorized as 10 or less/hpf or more than 10/hpf. C, G:T by nasal polyp status. D, G:T by asthma status. E, G:T by aspirin tolerance status—aspirin-exacerbated respiratory disease (AERD) or not. F, G:T by sinus CT score using the Lund-Mackay scoring system. G, G:T by basement membrane thickening categories.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Our findings favor the role of EPX as a biomarker for subtyping CRS. Regardless of the clinical presentation, some of the most recalcitrant forms of CRS are associated with tissue eosinophilia.1Orlandi R.R. Kingdom T.T. Hwang P.H. Smith T.L. Alt J.A. Baroody F.M. et al.International consensus statement on allergy and rhinology: rhinosinusitis.Int Forum Allergy Rhinol. 2016; : S22-S209PubMed Google Scholar EPX staining correlated with tissue eosinophilia in our study of 101 subjects. In addition, although EPX staining is higher in patients with CRS than in controls, EPX was specific in not staining CRS tissue with lymphoplasmacytic-predominant inflammation. These findings have clinical relevance because symptoms fail to distinguish the degree of eosinophilia in nasal polyposis3Steinke J.W. Smith A.R. Carpenter D.J. Patrie J. Payne S.C. Borish L. Lack of efficacy of symptoms and medical history in distinguishing degree of eosinophilia in nasal polyps.J Allergy Clin Immunol Pract. 2017; 5: 1582-1588Abstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar while EPX in tissue correlates with objective disease severity on CT. EPX can be used in patients with CRS with nasal polyposis, asthma, aspirin intolerance, or radiographically severe disease as a surrogate and quick marker for tissue eosinophilia, potentially bypassing the need for structured histopathology. Furthermore, EPX is able to measure both intact and degranulated eosinophils in tissue and in secretions. Clinical studies will need to validate whether EPX can be used to direct and titrate therapy with corticosteroids or antieosinophilic biologic drugs in recalcitrant patients with CRS. In the future, use of EPX (which is currently not commercially available) in nasal secretions may further bypass the need for tissue studies altogether.5Rank M.A. Ochkur S.I. Lewis J.C. Teaford III, H.G. Wesselius L.J. Helmers R.A. et al.Nasal and pharyngeal eosinophil peroxidase levels in adults with poorly controlled asthma correlate with sputum eosinophilia.Allergy. 2016; 71: 567-570Crossref PubMed Scopus (27) Google Scholar We were unable to find a use for GATA3 and T-bet staining to categorize patients with CRS. A study from Thailand found no significant association between GATA3 measured by RT-PCR and tissue eosinophilia.6Tantilipikorn P. Sookrung N. Muangsomboon S. Lumyongsatien J. Bedavenija A. Suwanwech T. Endotyping of chronic rhinosinusitis with and without polyp using transcription factor analysis.Front Cell Infect Microbiol. 2018; 8: 82Crossref PubMed Scopus (8) Google Scholar This is consistent with our finding that GATA3 and eosinophil activity appear to be independent in CRS. The same study also found no increase in T-bet in patients with CRS relative to controls, which is also consistent with our finding that T-bet did not readily distinguish CRS from non-CRS. In contrast, a Korean study found that GATA3 measured by RT-PCR was increased in patients with eosinophilic nasal polyps.7Shin S.H. Kim Y.H. Ye M.K. Choi S.Y. Immunopathologic characteristiscs of nasal polyps in adult Koreans: a single-center study.Am J Rhinol Allergy. 2017; 31: 168-173Crossref PubMed Scopus (8) Google Scholar A Chinese study also reported an increase in GATA-3 measured by RT-PCR and by immunohistochemistry in patients with CRS.8Xu R. Xu G. Shi J.B. Wen W.P. Chen H.X. Feng L.Q. Relationship between transcription factor GATA-3 and cytokines expression in chronic sinusitis [in Chinese].Zhonghua Er Bi Yan Hou Tou Jing Wai Ke Za Zhi. 2007; 42: 250-254PubMed Google Scholar The differences in our study compared with the study in China and Korea could be related to known differences among individuals of Asian and European descent in inflammatory patterns. Because GATA-3 is necessary but not sufficient for TH2 differentiation, it is possible that a factor besides GATA3 is important for TH2 differentiation in CRS.9Paul W.E. Zhu J. How are Th2 type immune responses initiated and amplified?.Nat Rev Immunol. 2010; 10: 225-235Crossref PubMed Scopus (697) Google Scholar Finally, the small number of controls (n = 8) in our study may have limited our ability to find important associations of GATA3 and T-bet. Our study has other limitations. Although samples in patients with CRS were from ethmoidal tissue, this was not consistently the case in controls if ethmoidectomy was not performed. We also did not measure cytokine expression profiles to verify classification into TH2/TH1 categories. We did not perform concurrent venipuncture to measure blood eosinophils, another potentially important tissue eosinophil biomarker. However, we were able to compare the nearest blood eosinophil levels with eosinophil and EPX levels in tissue. Mean blood eosinophils levels were 0.42 (95% CI, 0.33-0.51) in individuals with more than 10 eosinophils/hpf and 0.16 (95% CI, 0.12-0.20) in individuals with 10 or less eosinophils/hpf. Blood eosinophil counts were modestly correlated with EPX levels (Pearson correlation coefficient = 0.44; 95% CI, 0.27-0.59). In addition, we did not stain for master regulators of TH17 or regulatory T cells, both of which may be important in CRS. Finally, we were also unable to categorically determine the extent and timing of exposure to corticosteroid therapy. Therefore, prospective studies are needed to confirm the findings from this retrospective study. In conclusion, tissue EPX is a useful biomarker for eosinophilic inflammation in CRS. We were unable to find differences in GATA3 and T-bet staining to categorize patients with CRS. We acknowledge the expert assistance of Shirley “Charlie” Kern. We dedicate the science in this article to the late James J. Lee, PhD, who devoted his career to clarifying the role of eosinophils in human health. Fig E2GATA3 compared with T-BET in subjects with CRS and controls.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E3EPX compared with (A) nasal polyp status, (B) asthma status, (C) aspirin tolerance status—aspirin-exacerbated respiratory disease (AERD) or not, (D) sinus CT scan score (Lund-Mackay), and (E) basement membrane thickening.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Table E1Patient clinical characteristicsCharacteristicCRS (N = 93)Control (N = 8)Total (N = 101)P value∗Wilcoxon rank-sum test for continuous variables and Fisher exact test for categorical variables.Age at surgery (y), mean ± SD54.0 ± 15.462.0 ± 15.154.6 ± 15.5.29Sex: female47 (50.5)2 (25.0)49 (48.5).27Previous sinus surgery38 (40.9)0 (0.0)38 (37.6).02Nasal polyps46 (49.5)(all missing)46 (49.5)Aspirin intolerance12 (13.0)1 (12.5)13 (13.0)>.99Asthma43 (46.7)1 (12.5)44 (44.0).08Sinus CT scan total score (Lund-Mackay score), mean ± SD11.5 ± 4.84.5 ± 3.511.0 ± 5.1<.001SNOT-22 total score (before surgery), mean ± SD43.6 ± 21.042.6 ± 23.243.6 ± 21.1.94Values represent n (%) unless otherwise indicated.SNOT-22, Sino-Nasal Outcome Test 22.∗ Wilcoxon rank-sum test for continuous variables and Fisher exact test for categorical variables. Open table in a new tab Table E2Patient histopathology characteristics (per Snidvongs et alE1Snidvongs K. Lam M. Sacks R. Earls P. Kalish L. et al.Structured histopathology profiling of chronic rhinosinusitis in routine practice.Int Forum Allergy Rhinol. 2012; 2: 376-385Crossref PubMed Scopus (138) Google Scholar)CharacteristicCRS (N = 93)Control (N = 8)Total (N = 101)P value∗Fisher exact test.Overall inflammation.19 Mild36 (39.1)6 (75.0)42 (42.0) Moderate44 (47.8)2 (25.0)46 (46.0) Severe12 (13.0)0 (0.0)12 (12.0)Eosinophil count.002 10/hpf52 (56.5)0 (0.0)52 (52.0)Neutrophilic infiltrate.19 Absent59 (65.6)8 (100.0)67 (68.4) Focal29 (32.2)0 (0.0)29 (29.6) Diffuse2 (2.2)0 (0.0)2 (2.0)Inflammatory predominance Lymphocytic23 (24.7)2 (25.0)25 (24.8)>.99 Lymphoplasmocytic48 (51.6)6 (75.0)54 (53.5).28 Eosinophilic35 (37.6)0 (0.0)35 (34.7).05 Lymphohistiocytic000 Neutrophilic2 (2.2)0 (0.0)2 (2.0)>.99Basement membrane thickening.38 Absent12 (13.0)1 (12.5)13 (13.0) 15 μm32 (34.8)2 (25.0)34 (34.0)Subepithelial edema.93 Absent22 (24.2)2 (25.0)24 (24.2) Mild (focal or perivascular only)43 (47.3)5 (62.5)48 (48.5) Moderate (distortion of mucosal structure)21 (23.1)1 (12.5)22 (22.2) Severe (diffuse/polypoid change)5 (5.5)0 (0.0)5 (5.1)Hyperplastic/papillary change.59 Absent80 (87.0)8 (100.0)88 (88.0) Present12 (13.0)0 (0.0)12 (12.0)Mucosal ulceration>.99 Absent86 (93.5)8 (100.0)94 (94.0) Present6 (6.5)0 (0.0)6 (6.0)Squamous metaplasia.21 Absent83 (90.2)6 (75.0)89 (89.0) Present9 (9.8)2 (25.0)11 (11.0)Fibrosis.30 Absent16 (17.4)3 (37.5)19 (19.0) Partial70 (76.1)5 (62.5)75 (75.0) Extensive6 (6.5)0 (0.0)6 (6.0)Fungal elements.34 Absent80 (86.0)6 (75.0)86 (85.1) Present13 (14.0)2 (25.0)15 (14.9)Charcot-Leyden crystals.34 Absent75 (81.5)8 (100.0)83 (83.0) Present17 (18.5)0 (0.0)17 (17.0)Eosinophil aggregates.10 Absent63 (69.2)8 (100.0)71 (71.7) Present28 (30.8)0 (0.0)28 (28.3)Values represent n (%).∗ Fisher exact test. Open table in a new tab Values represent n (%) unless otherwise indicated. SNOT-22, Sino-Nasal Outcome Test 22. Values represent n (%).

Referência(s)