Background
Granulomatosis with polyangiitis (GPA) is a type of antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis characterized by necrotizing granulomatous inflammation and small to medium-sized vessel vasculitis, which may involve multiple organs [
1,
2]. In GPA, granulomatous inflammation usually involves the upper and lower respiratory tract and the orbits and meninges, leading to mass formation (e.g. pulmonary nodules or masses, orbital masses, subglottic and/or bronchial stenosis, hypertrophic pachimeningitis) and destructive lesions such as bony erosions of the sinuses, saddle nose, septal perforation or cavitation of pulmonary nodules/masses [
3,
4]. Once regarded as a relatively harmless phenotype of GPA due to its expression mainly in patients with localized disease, recent studies have shown that necrotizing granulomatous inflammation may lead to significant morbidity and even death in GPA, and it is often refractory to conventional treatment [
5,
6].
Macrophages are cells of the innate immune system that play an important role in inflammation, host defense, tissue remodeling and regulation of metabolism [
7]. The plasticity of macrophages is highlighted by their capacity to change their functional phenotypes in response to distinct stimuli. Stimulation by toll-like receptor (TLR) ligands or by interferon (IFN)-γ triggers M1 polarization, while the exposure to cytokines like interleukin (IL)-4 and/or IL-13 leads to M2 polarization, mirroring T helper (Th) Th1/Th2 polarization of T cells [
8]. M1 macrophages usually express IL-12 and IL-23, produce effector molecules (e.g. reactive oxygen and nitrogen species) and inflammatory cytokines such as IL-1β, tumor necrosis factor α (TNFα) and IL-6. These M1 macrophages initiate an inflammatory response and mediate tissue damage, and they are responsible for resistance to intracellular pathogens and tumor cells. In contrast, M2 macrophages express IL-10 and a high level of scavenger-like, mannose-like and galactose-like receptors. M2 macrophages are involved in inhibiting inflammation, in tissue repair and remodeling and in angiogenesis and tumor progression [
7,
8]. Nonetheless, M2 macrophages are further sub classified into four different additional phenotypes as follows: M2a, M2b, M2c and M2d, with distinct functions. M2a is a wound-healing macrophage that promotes tissue repair and remodeling while M2c is a suppressor macrophage that is responsible for downregulating immune response. M2b macrophage is a pro-inflammatory M2 macrophage that is induced by exposure to immune complexes, TLR ligands or IL-1 receptor antagonist. M2d macrophages produce high levels of IL-10 and vascular endothelial growth factor (VEGF), inhibiting immune response and promoting angiogenesis [
9].
Macrophages are key cells in granulomatous inflammation and may be present as epithelioid cells, multinucleated giant cells and in some instances as foam cells [
10]. Granulomatous inflammation in GPA is characterized by poorly formed granulomas usually associated with geographic necrosis, micro abscesses and scattered giant cells [
11,
12].
In the early stages of ANCA-associated pauci-immune necrotizing glomerulonephritis, CD163
+ M2 macrophages predominate at sites of fibrinoid necrosis, where they exceed the numbers of neutrophils and T cells. CD68
+ cells and CD163
+ macrophages are more frequently found in normal-appearing glomeruli in this condition than in glomeruli of controls with thin basement membrane nephropathy [
13]. However, to the best of our knowledge, it is not known whether macrophages found in granulomatous lesions in the upper airways of patients with GPA display an M1 or M2 phenotype. Assessment of their distribution would help to uncover whether granulomatous inflammation in the airways of patients with GPA displays a pro-inflammatory phenotype of macrophages in line with a Th1 response or a more reparative and regulatory phenotype mirroring a predominant Th2 response. We hypothesize that macrophages found in granulomatous lesions in the airways of patients with GPA are predominantly of the M1 phenotype, due to the highly inflammatory nature of these lesions and the Th1-polarized T cell response involved in airway inflammation in patients with GPA. This study aimed to investigate the distribution of macrophage phenotypes found in active granulomatous lesions in the upper airways of patients with GPA and to analyze associations with infiltrating T and B cells from these inflammatory sites and with chronic carriage of
Staphylococcus aureus, disease parameters and therapy in GPA.
Methods
Patients
Thirty-five consecutive patients who were classified as having GPA according to the 1990 American College of Rheumatology (ACR) criteria for GPA (Wegener’s granulomatosis) or according to the European Medicines Agency (EMEA) classification algorithm [
14,
15], and underwent a biopsy from the respiratory tract were selected for this study. Disease extension was ascertained according to European Vasculitis Study (EUVAS) disease categorization of ANCA-associated vasculitis [
16]. Table
1 describes disease features, biopsy sites, and therapy in those patients with GPA. All biopsies were performed for diagnostic purposes by nasofibroscopy, bronchoscopic procedures or thoracotomy. The majority of patients with GPA underwent nose biopsies; other biopsy sites (one patient per site) included ethmoidal sinus, Eustachian tube, nasopharynx, subglottic area, trachea, bronchus and lung.
Table 1
Disease features and biopsy sites in patients with granulomatosis with polyangiitis
Age, years | 48.9 ± 12.6 |
Female, n (%) | 22 (62.9) |
Disease status
|
Onset, n (%) | 9 (25.7) |
Relapse, n (%) | 25 (71.4) |
Persistent disease, n (%) | 1 (2.8) |
Years since diagnosisa
| 8.7 ± 5.6 |
Disease extension
|
Localized disease, n (%) | 29 (82.9) |
Generalized disease, n (%) | 6 (17.1) |
Disease manifestations
|
ENT involvement, n (%) | 34 (97.1) |
Eye disease, n (%) | 8 (22.8) |
Lung involvement, n (%) | 7 (20.0) |
Glomerulonephritis, n (%) | 4 (11.4) |
Cutaneous vasculitis, n (%) | 1 (2.8) |
ANCA positivity
|
cANCA by IIF, n (%) | 33 (94.3) |
PR3-ANCA by ELISA, n (%) | 32 (91.4) |
Biopsy sites
|
Nose, n (%) | 28 (80.0) |
Other sites, n (%) | 7 (20.0) |
Staphylococcus aureus
|
Positive culture, n (%)b
| 19 (55.9) |
Therapy
|
No therapy, n (%) | 13 (37.1) |
Prednisolone, n (%) | 10 (28.6) |
Prednisolone daily dose, mg | 11.2 (9.3–20.0) |
Daily co-trimoxazole, n (%) | 14 (40.0) |
Cyclophosphamide, n (%) | 4 (11.4) |
Azathioprine, n (%) | 4 (11.4) |
Methotrexate, n (%) | 1 (2.8) |
Mycophenolate mofetil, n (%) | 4 (11.4) |
Cyclosporin A, n (%) | 1 (2.8) |
The ANCA test was performed by in-house indirect immunofluorescence (IIF) on ethanol-fixed neutrophil slides and results were displayed as the pattern (i.e. cytoplasmic, perinuclear or atypical ANCA) and ANCA titers. ANCA specificity for anti-proteinase 3 (anti-PR3) and anti-myeloperoxidase (anti-MPO) antibodies was confirmed by an in-house capture enzyme-linked immunosorbent assay (ELISA) as described earlier [
17].
Immunohistochemical evaluation
Biopsy specimens were processed for immunohistochemical evaluation. Paraffin-embedded specimens and formalin-fixed 4-μm sections were deparaffinized and were incubated with 10 mM Tris/HCl and 1 mM EDTA (Ethylenediaminetetraacetic acid) buffer (pH 9.0) at 90 °C for 60 minutes for antigen retrieval. Endogenous peroxidase was blocked with 0.3% H2O2 in phosphate-buffered saline (PBS) for 30 minutes. Slides were pre-incubated with 20% bovine serum albumin (BSA) and incubated with the following primary antibodies diluted in PBS with 1% BSA: anti-CD3 (Abcam, Bristol, UK), anti-CD20 (Abcam, Bristol, UK), anti-CD68 (Abcam, Bristol, UK), anti-CD86 (Abcam, Bristol, UK) and anti-CD163 (Abcam, Bristol, UK) for 1 hour at room temperature. Binding was detected by appropriate horseradish-peroxidase-labeled secondary antibodies and peroxidase activity was developed using 3, 3’-diaminobenzidine tetrachloride (10 minutes at room temperature). To obtain nuclear staining for the transcription factors Tbet and GATA-3 for Th1 and Th2 cells, respectively, 0.5% Triton was added to wash and incubation buffer and anti-T-bet/Tbx21 clone 4B10 (Abcam, Bristol, UK) and anti-GATA-3 (Abcam, Bristol, UK) antibodies were used. For GATA-3 staining, antigen retrieval was performed with citrate buffer (10 mM, pH 6.0).
All sections were stored digitally after examination using a Nanozoomer Digital Pathology Scanner (NDP Scan U10074-01; Hamamatsu Photonics K.K.) and quantified using the positive pixel count algorithm with ImageScope Viewer software (V11.2.0.780 Aperio; e-Pathology Solution). The percentage of strong positive brown and positive pixels was determined in comparison to the total number of pixels (positive and negative).
Statistical analysis
Data were analyzed using IBM SPSS Statistics for Windows, version 20.0 (Armonk, USA) and graphs were built using GraphPad Prism version 5.0 for Windows (La Jolla, USA). Categorical variables were presented as total number and percentage while continuous data were presented as mean ± standard deviation or as median and interquartile range as appropriate. Student’s t test or the Mann-Whitney U test was used for comparison of continuous data between two groups, while comparisons between three or more groups were performed using the Kruskal-Wallis test. Correlation between variables was analyzed using Spearman’s correlation coefficient. Univariate and multivariate models of linear regression were built to analyze associations between clinical and therapy variables with frequencies of CD3, CD20, CD68, CD86 and CD163 in airway biopsies from patients with GPA. Results are expressed as β coefficient, p value and R
2 as appropriate. Statistical significance was set at 5% (p < 0.05).
Discussion
In this study, we found that macrophages and T cells are more frequent than B cells in airway lesions of patients with GPA. M2 macrophages were the predominant phenotype, mainly in airway sites other than the nose, and this increase in M2 macrophages was further associated with the use of immunosuppressive agents. Frequencies of macrophages and macrophage subsets were significantly correlated with the daily dose of prednisolone whereas B and T cells were not. Patients with GPA with a positive nose culture for Staphylococcus aureus had higher frequencies of T cells infiltrating the nose mucosa.
As macrophage phenotypes mirror Th1 and Th2 functionality [
8], and airway inflammation in GPA may be predominantly from a Th1 response, we originally hypothesized that the M1 phenotype would be more frequently found in active lesions in the airways of patients with GPA. However, this hypothesis was proven to be wrong and the expression of the M2 marker CD163 and the Th2 marker GATA-3 was significantly higher in airway lesions from patients with GPA. This unexpected finding may be due to the ongoing Th2 response of patients with GPA evaluated in this study, reflected by the frequent lymphoid follicles observed in parallel with granulomatous airway inflammation and by the high frequency (94.3%) of a positive ANCA test, indicating a shift towards a Th2 response [
18]. Indeed, it is worth mentioning that sera from patients with AAV had a tendency to shift towards an M2c phenotype in macrophages regardless of disease activity, whereas sera from health controls led to an M0 phenotype [
19]. Alternatively, another mechanism to explain the predominant M2 phenotype in GPA is the influence of therapy on our results; M2c macrophages may be induced by exposure to glucocorticoids [
20]. Although in this study only one third of patients with GPA were under therapy with prednisolone at a relatively low mean daily dose, there was a trend toward higher scores of M2 macrophages and lower scores of T cells in airway biopsies from patients with GPA on prednisolone (Additional file
2: Table S2).
A recent study showed an increase in CD163
+ M2 macrophages in early lesions of ANCA-associated pauci-immune necrotizing glomerulonephritis, mostly patients with either PR3- or MPO-ANCA. However, it is not possible to associate this finding with therapy, because no information was provided about the use of glucocorticoids or immunosuppressive agents [
13]. In addition Park et al. found abundant expression of CD163
+ cells in granulomatous lesions in the lungs in patients with GPA [
21]. Together with our findings, these results indicate that M2 polarization is observed in different tissues affected by GPA, including the upper airways, lungs and kidneys.
Macrophage polarization has been evaluated in tissues from patients with different disorders. Every disease shows a unique pattern of macrophage polarization that may be influenced by the underlying pathophysiologic process [
8,
22]. The M1 phenotype is predominantly found in adipose tissue in obese patients, in synovial tissue in patients with rheumatoid arthritis, and in atherosclerotic plaques [
9,
22‐
28]. M2 macrophages are predominantly found in muscle biopsies from patients with neuromuscular sarcoidosis, in lung tissue from patients with idiopathic pulmonary fibrosis and from patients with asthma, and in synovial tissue from patients with osteoarthritis and patients with spondyloarthritis [
23,
24,
29‐
32]. In cancer, M1-polarized macrophages have anti-tumoral activity while infiltration of M2 macrophages in tumoral lesions is associated with tumor progression and with a worse prognosis [
32]. Surface markers have been used to identify M1 and M2 macrophages in different studies. The most common markers used for M1 macrophages are CD80 and CD86, whereas CD163 and mannose receptor (also known as CD206) are used to characterize M2 macrophages [
9].
In this study, we found higher percentages of CD86 and CD163, specific M1 and M2 markers, respectively, compared with the pan-macrophage marker (CD68). The possible reason for this may be the cytoplasmic distribution of CD68 in contrast with the membrane staining of CD86 and CD163 [
33,
34]. Moreover, it is worth mentioning that this was not a cellular count but a pixel count of brown staining from immunohistochemistry slides. In fact, some studies evaluating CD68 and CD163 by immunohistochemical evaluation also found higher scores for CD163 compared with CD68 [
12,
35‐
39].
Macrophages are polarized toward the M1 phenotype in the early phases of bacterial infections, but when infection becomes chronic or upon resolution and convalescence, M2 macrophages prevail in order to prevent excessive tissue damage [
21]. Chronic carriage of
Staphylococcus aureus was included in this study due to its high prevalence in patients with GPA and its impact on the risk of relapse [
40]. Moreover, it would influence macrophage polarization in inflammatory lesions of the airways in GPA. Although in chronic rhinosinusitis with nasal polyps an increased number of M2 macrophages and decreased phagocytosis is associated with the persistence of
Staphylococcus aureus [
41], we did not find any association between an increased number of M2 macrophages in nasal mucosa from patients with GPA and nasal carriage of
Staphylococcus aureus. In the present study, the association found between
Staphylococcus aureus and T cells in nasal mucosa could indicate that these bacteria trigger local T cell expansion, as staphylococcal super-antigens could be responsible for polyclonal T cell activation and an increased risk of relapse, especially strains of
Staphylococcus aureus that produce the tsst-1 (toxic shock syndrome toxin-1) super-antigen [
40,
42].
Amongst lymphocytes and macrophages, B cells were the cells least present in granulomatous lesions in the airways of patients with GPA. It is known in GPA that granulomas harbor B cells in germinal center-like lymphoid structures, and these cells are considered to be pathogenic ANCA-secreting cells [
43]. Activated B cells are found alongside PR3-expressing cells and cells expressing B cell activating factor (BAFF) and a proliferation-inducing ligand (APRIL) in nasal mucosa from patients with GPA [
44]. Nonetheless, in this study we did not find any associations between B cells and disease parameters.
Limitations of this study include mainly its cross-sectional nature. Longitudinal evaluation of macrophage phenotypes in airway infiltrates from patients with GPA (i.e. repeated biopsies of nasal mucosa) prior to commencing therapy and after remission is obtained would yield better understanding about the influence of disease activity and therapy. Another limitation is the heterogeneity of biopsy sites evaluated in this study. Even though most biopsies were performed from nasal mucosa, other airway sites could behave in a different way under the influence of different local factors.
Acknowledgements
The authors would like to thank Paulina Sosicka and Johan Bijzet for their contributions in standardizing the immunohistochemistry techniques used in this study.