Background
Polymyositis (PM) and dermatomyositis (DM) are systemic inflammatory diseases affecting the muscles, skin, and several other organs including the lungs [
1,
2]. Interstitial lung disease (ILD) is a common complication seen in 20–78% of patients with PM/DM and can be a prognostic determinant [
3‐
8]. Particularly, patients with DM and clinically amyopathic DM (CADM) frequently develop an acute and severe form of ILD [
4,
5]. Thus, management of ILD is essential for the care of patients with PM/DM, and appropriate biomarkers reflecting the disease severity are required.
CD163, a scavenger receptor specifically expressed on the surface of activated macrophages and monocytes, is involved in the uptake of hemoglobin-haptoglobin and mediates the clearance of hemoglobin [
9]. CD163 expression is upregulated by several inflammatory cytokines including interleukin (IL)-6 [
10,
11]. Inflammatory infiltrates containing CD163-positive macrophages are frequently found in the diseased muscles in PM/DM [
12,
13], where activated macrophages themselves directly induce inflammation and also mediate inflammatory responses via lymphocyte activation [
14,
15]. Indeed, PM/DM is one of the diseases that occasionally cause macrophage activation syndrome (MAS). In contrast, macrophages could produce transforming growth factor-β, the most prominent fibrosis-inducing molecule, in chronic lung inflammation [
16]. Additionally, activated macrophages expressing CD163 have been demonstrated to play a key role during lung fibrosis [
17]. Macrophage activation would be involved in the pathogenesis of PM/DM-related ILD.
Upon acute and chronic inflammations, CD163 is shed from surface of activated macrophages and released into peripheral blood in the soluble form [
11]. Therefore, soluble CD163 (sCD163) in the serum has been highlighted as a representative of macrophage activation and as a biomarker in several inflammatory diseases [
18‐
23]. However, the clinical significance of this molecule in PM/DM-related ILD is not fully understood. The aim of this retrospective study was to evaluate the potential of serum sCD163 as a biomarker in PM/DM-related ILD.
Methods
Patient selection
A retrospective review of 52 patients with PM/DM-related ILD, consecutively admitted to Hamamatsu University Hospital between 1995 and 2014, was conducted. Subsequently, four patients (one with PM, one with CADM, and two with classic DM) who received immunosuppressive treatment with corticosteroids on diagnosis of ILD were excluded. Thus, this study included the remaining 48 patients with PM/DM-related ILD. Among them, 45 patients had never received immunosuppressive treatment, whereas the other 3 patients had received treatment but terminated it at least 1 year before the ILD diagnosis. In addition, 10 patients with PM/DM without ILD, who had not received immunosuppressive treatment, and 20 age-matched and gender-matched healthy volunteers were enrolled as control groups (Additional file
1: Table S1).
Classic PM/DM was diagnosed according to the criteria of Bohan and Peter [
2]. Patients with definite or probable PM/DM according to these criteria were included in the study. CADM was diagnosed when a patient had a skin rash typical of DM without clinical evidence of myositis and with little or no increase in serum creatine kinase (CK) [
5,
24]. The diagnosis of classic PM/DM and CADM had been confirmed by the dermatologists and/or rheumatologists and the pulmonologists in our institution. The diagnosis of ILD was based on the radiological assessment of high-resolution computed tomography (HRCT) of the chest, which included appearances of bilateral reticulation, ground-glass attenuation, or consolidation.
Data collection
The patients’ clinical data at the time of PM/DM-related ILD diagnosis were obtained through medical record reviews. These data included symptoms; laboratory findings; pulmonary function test results (forced vital capacity (FVC); diffusing capacity of the lung for carbon monoxide (DLco), measured using a single breath technique); analyses of bronchoalveolar lavage (BAL) fluid; treatment; and prognosis.
Serum sCD163 levels were quantified by enzyme-linked immunosorbent assay in accordance with the manufacturer's protocol (eBioscience, USA). The samples were obtained at the time of PM/DM-related ILD diagnosis and stored at −80 °C for future analysis. All measurements were done in duplicate and the mean value of serum sCD163 was recorded. Written informed consent for serum conservation was obtained from all the subjects.
Immunohistochemical analysis
Lung specimens were obtained from two patients with classic DM-related ILD (on surgical lung biopsy and autopsy, respectively) and one patient with early-stage lung cancer as a normal control (on lung resection). These specimens had been fixed in 10% formalin and embedded in paraffin. Deparaffinized sections (5 μm thick) were preheated to 120 °C for 20 minutes. After blocking endogenous peroxidase activity with 0.3% H2O2 for 20 minutes, the slides were incubated with the rabbit anti-human CD163 monoclonal antibody (1:200; Leica Biosystems, Germany) or isotype control antibody (Jackson ImmunoResearch Laboratories, USA) for 1 h at 20 °C. Subsequently, the sections were incubated with biotin-conjugated gout anti-rabbit IgG antibody (Nichirei, Japan) for 15 minutes, and the immunoreaction was visualized using a 3,3-diaminobenzidine (DAB) chromogen solution (DAB substrate kit; Vector Laboratories, USA), and then counterstained with hematoxylin. Additionally, hematoxylin–eosin staining was also done on another slide from each patient.
Statistical analysis
Data were described as numbers (percentages) or medians with observed ranges. Spearman's rank correlation coefficients were evaluated for correlation between serum sCD163 and the other clinical factors with continuous variables. The Mann–Whitney
U test was performed for two-group comparisons. One-way analysis of variance was performed followed by post-hoc analysis (Dunn's multiple comparisons test) for multi-group comparisons. Overall survival time was defined as the number of months from PM/DM-related ILD diagnosis (the date of the first HRCT evaluation) until all-cause death or at censoring. Patients were censored if alive on 30 June 2015. Cox proportional hazard modeling was used to identify prognostic factors. To estimate the cutoff value of serum sCD163 level, time-dependent receiver operating characteristic analysis was performed, as described previously [
25,
26]. The Kaplan–Meier survival method and the log-rank test were employed to compare the survival rate between groups. Values of
P < 0.05 were considered significant. Statistical analyses were performed by using R software version 2.15.1 (The R Foundation for Statistical Computing, Austria) and SPSS software version 13.0 (SPSS, USA).
Discussion
In this study, we evaluated the clinical significance of sCD163 in patients with PM/DM-related ILD. Serum sCD163 levels at ILD diagnosis were associated with several clinical factors and even the prognosis, suggesting that this molecule might be a potential biomarker of the disease.
The expression of CD163, a macrophage-specific marker within the macrophage/monocyte lineage, is intricately regulated by various factors, such as IL-6, IL-10, interferon-γ, tumor necrosis factor (TNF)-α, and corticosteroids. Cleavage of CD163 from activated macrophages, which is performed by a disintegrin and metalloproteinase 17/TNF-α converting enzyme in response to several inflammatory mediators, leads to an increase in sCD163 in the serum [
10,
11]. In fact, the serum sCD163 levels in our patients were correlated with the presence of fever and serum CRP levels, suggesting that elevated serum sCD163 is associated with systemic inflammation. Notably, among macrophages, CD163 is mainly expressed on alternatively activated macrophages, rather than classically activated macrophages [
11]. Collectively, our results indicate the significance of alternatively activated macrophages in the inflammation seen in PM/DM-related ILD.
Serum sCD163 levels were significantly higher in patients with PM/DM-related ILD than those in healthy volunteers, which is consistent with the results of the recent study by Peng et al. [
13]. They measured serum sCD163 levels in 108 patients with PM/DM, consisted of 65 patients without ILD and 43 patients with ILD, and found that serum sCD163 levels were correlated with the activity of myositis. However, the clinical implication of sCD163 in PM/DM-related ILD remains unclear. In the present study, we focused on the ILD and further advanced the understandings of sCD163 physiology in this disease entity, particularly the relevance to pulmonary function and prognosis.
In the Cox proportional hazard modeling, higher serum sCD163 was significantly associated with worse prognosis. This result implies that macrophage activation in PM/DM-related ILD plays an important role in advancing the disease. In contrast, several studies have evaluated the clinical significance of serum ferritin, which is a key biomarker of MAS, in PM/DM-related ILD [
27‐
29]. Indeed, PM/DM is known to cause MAS, in which the lack of T cell regulation and excessive production of cytokines, including TNF-α, IL-1β, IL-6, and IL-18, results in macrophage activation. Those previous studies reported that hyperferritinemia was associated with high mortality in patients with PM/DM-related (particularly CADM-related) ILD, which was reproduced in our cohort. Interestingly, there was a weak, but not significant correlation between serum sCD163 levels and ferritin levels in the present study. The total value of serum ferritin might be affected by the storage of iron (in the form of L-subunit ferritin). Meanwhile, sCD163 is secreted from activated macrophages and elevated for days in peripheral blood, suggesting that this molecule specifically reflects the degree of macrophage activation. Although the superiority of either of the two biomarkers could not be evaluated in our small cohort, our results suggest that serum sCD163 may have value in the assessment of disease severity in PM/DM-related ILD.
Human CD163 expression has been confirmed in several tissues, such as red pulp macrophages, bone marrow macrophages, liver macrophages (Kupffer cells), and alveolar macrophages [
10,
11]. In fact, on the basis of immunohistochemical analysis, one of the sources of elevated serum sCD163 seemed to be the diseased lungs. Additionally, more severe infiltration of CD163-positive macrophages was found in the lung of the non-surviving patient with DM-related ILD, compared with that in the lung of the survivor (Fig.
3). Although we are speculating on the assessment of only two lung specimens, the amount of infiltrating CD163-positive macrophage may reflect the severity of PM/DM-related ILD. The following results may support this: serum sCD163 was negatively correlated with percent predicted FVC. Previous studies demonstrate that alveolar macrophages produce several profibrotic cytokines including transforming growth factor-β, platelet-derived growth factor, and fibroblast growth factor-2 [
16,
30,
31]. Intriguingly, Gibbons et al. recently demonstrated the paradoxical and essential role played by macrophages in the pathogenesis of lung fibrosis, in which activated macrophages had both profibrotic and resolution-promoting roles during fibrogenesis and the reversible phase, respectively [
17]. Thus, the exact functions of CD163-positive macrophages in the pathogenesis of PM/DM-related ILD remain unclear, but the infiltration to the lungs may have some impact not only on inflammation but also on fibrosis.
The present study has several limitations. First, this was a small, retrospective study, which may have involved various biases. Validation in larger and prospective cohorts is needed to confirm our preliminary results. Second, there are several subtypes in the development of PM/DM-related ILD, such as ILD-preceding, concomitant, and myositis-preceding patterns, which indicates the difficulty in determining the exact ILD onset, particularly in patients with no symptoms or only minor symptoms. Our definition of ILD onset might be not accurate in some patients. Third, the patients included in the present study were not treated according to a consistent regimen. Although most of the patients were treated with corticosteroids and additional immunosuppressants, regimen variations, such as the doses and selection of drugs, may have affected the clinical disease course and prognosis. Additionally, we could not follow serum sCD163 levels chronologically. It is unclear how changes correlated with the clinical course of PM/DM-related ILD. In future studies, clinical factors important in the management of ILD, such as changes in pulmonary function and oxygenation, and treatment response, should be compared with changes in sCD163 levels. These would explore the usefulness of this biomarker.
Acknowledgements
We thank Dr Hajime Yamakage (SATISTA Inc.) and Dr Masaaki Kaneta (Department of Clinical Pharmacology and Therapeutics, Hamamatsu University School of Medicine) for the statistical advice and immunohistochemistry procedures, respectively.