Introduction
Dermatomyositis (DM) is a systemic autoimmune and inflammatory disease that involves not only the muscle and skin but also several other organs, such as the lungs, heart and joints [
1]. Interstitial lung disease (ILD) is the most common internal organ manifestation, and it affects the prognosis of DM patients. ILD is classified into two subsets: acute/subacute interstitial pneumonia (A/SIP) and chronic interstitial pneumonia (CIP) [
2,
3]. A/SIP is often complicated with clinically amyopathic DM (CADM) [
4,
5], which shows the typical skin manifestations of DM but has no or little evidence of clinical myositis [
6]. A/SIP complicated with CADM is life-threatening and shows a rapidly progressive pattern, with a 6-month survival rate of 40.8%, irrespective of intensive therapy [
5]. Combination therapy with prednisolone (PSL), cyclosporine A (CsA) and intravenous pulse cyclophosphamide (IVCY) is more effective for A/SIP with DM than PSL plus one immunosuppressive agent [
7]. Therefore, diagnosis and evaluation of ILD is very important to determine the treatment strategy when DM is diagnosed [
8-
10]. Although several biomarkers that may reflect inflammatory activity of lung have been utilized, such as ferritin [
11-
14] and Krebs von den Lungen-6 (KL-6) [
15-
19], there are no established serum biomarkers for DM-associated ILD.
Progranulin (PGRN) [GenBank:NC_000017] is an extracellular glycoprotein containing seven and one-half repeats of cysteine-rich motifs. PGRN is proteolytically cleaved by extracellular proteases, such as proteinase 3 (PR3) and elastase, into granulin (GRN) [
20] that range from 6 to 25 kDa. PGRN is abundantly expressed in rapidly cycling epithelial cells, leukocytes, chondrocytes and neurons [
21], and its expression level is at steady state [
22]. PGRN plays a critical role in early embryogenesis [
22], wound healing [
23], maintenance of neuronal survival [
24] and tumorigenesis [
21]. Particularly, macrophage-derived PGRN is a key regulatory factor in the processes of inflammation and wound healing [
23,
25]. Recent mouse studies show that mice unable to convert PGRN into GRN due to lack of both elastase and PR3 cannot show inflammation in response to injection of immune complexes [
26]. These data indicate that PGRN is synthesized by macrophages and is cleaved into GRNs by elastase in tissues to enhance inflammation. Moreover, PGRN and/or GRN act as a soluble cofactor for Toll-like receptor 9 (TLR9) signaling and enhance it [
27]. We previously reported that serum PGRN levels are significantly elevated in systemic lupus erythematosus (SLE) patients in parallel with disease activities and that PGRN may have a role in the pathogenesis of SLE, partly by enhancing the TLR9 signaling and interleukin (IL)-6 production [
28]. Although macrophage activation is considered to underlie the pathogenesis of A/SIP with DM [
11-
14] and TLR9 and IL-6 is associated with DM [
29-
35], the roles of PGRN in DM are still unknown.
Here, we show that serum PGRN levels are significantly elevated in DM patients, particularly those with A/SIP, and are associated with disease activity and prognosis of DM patients with ILD. PGRN may have a role in the pathogenesis of DM.
Discussion
To our knowledge, this is the first study to show that PGRN levels were significantly elevated in sera of DM patients, in particular those with A/SIP, and these concentrations were associated with the disease activity and prognosis of DM patients with ILD.
Consistent with previous reports [
11,
12], the elevation of transaminases and hyperferritinemia were observed in DM with A/SIP in the present study. In addition, WBC and Hb tended to be lower in DM patients with A/SIP than in those without it. These facts suggest that macrophages are substantially activated and cause injury to the lungs and the liver in DM patients with A/SIP. Actually, the accumulation of ferritin-producing macrophages has been shown in a CADM-related acute interstitial pneumonia autopsy case [
40]. Because PGRN is highly expressed in macrophages [
22], elevation of serum PGRN levels in DM with A/SIP also reflects macrophage activation as that of ferritin.
In the present study, other than DM, we also measured concentrations of serum PGRN in patients with PM (n =21), including 6 with ILD (median: 59 ng/ml) and 15 with systemic sclerosis (median: 49 ng/ml) (same levels in NHCs), including 13 with ILD (median: 49 ng/ml). In our data, although sera from PM patients contained significantly elevated PGRN compared with NHCs, DM patients’ sera had much higher levels of PGRN, especially among DM patients who also had A/SIP. This difference may occur because activated macrophages appeared especially in DM patients with A/SIP. Although we could not recruit acute respiratory distress syndrome (ARDS) patients into this study, ARDS could be an additional control to determine whether PGRN is a more general marker of acute respiratory inflammation.
PGRN is reported to be important in the initiation of inflammation by recruiting fibroblasts, macrophages and neutrophils to the site of inflammation [
23]. PGRN is also converted to GRN by the elastase that is produced by leukocytes and other cells [
41], however, and recent mouse studies have shown that GRN possesses inflammatory functions [
25,
26,
42]. Thus, PGRN may be converted to GRN in the lungs and may be associated with the early phase of pathogenesis in DM with A/SIP. However, it is difficult to investigate the function of GRN in human studies, because we cannot measure GRNs at this moment.
Serum PGRN levels showed a significantly positive correlation with serum levels of LDH (
r
S =0.54,
P =0.0003) (Table
2) and ferritin (
r
S =0.77,
P <0.0001). However, there was no significant correlation between serum levels of PGRN and those of KL-6, another biomarker for ILD. Recent studies suggest serum ferritin levels as a marker for severity of acute progressive ILD in DM and CADM patients [
11-
14]. The serum ferritin concentrations in CADM patients with A/SIP are already elevated in early stages of the disease, before progression of ILD [
11]. LDH and surfactant protein D are released into the blood when pulmonary cells are damaged with inflammation [
43]. KL-6 rather reflects the regeneration and proliferation of pneumocytes after the damage of the pulmonary tissue [
44]. Thus, PGRN could also be an acute marker for ILD as well as ferritin and LDH.
When we compared clinical manifestations between DM patients with A/SIP and without A/SIP, the frequency of CADM was significantly higher in those with A/SIP (76.5%) than in those with CIP (37.5%) and in those without ILD (5.9%). CADM may be a useful indicator for the occurrence of A/SIP. The levels of CRP showed significant differences between DM patients with A/SIP and those without and correlated with serum PGRN levels (
r
S =0.48,
P =0.0015) (Table
2). We previously reported that serum PGRN levels are elevated in SLE patients and that PGRN stimulated IL-6 production via TLR9 [
28]. Upregulated IL-6 production stimulated by PGRN [
27,
28,
45] may be related to elevation of CRP in DM patients with A/SIP. Serum IL-6 levels were significantly higher in dead than alive CADM patients with ILD [
33] and were reported to be candidate biomarkers for disease activity in DM [
32] and SLE [
46]. Because TLR9 is highly expressed in DM [
29-
31] and SLE [
47], upregulated IL-6 production stimulated by PGRN via TLR9 may be a common pathogenesis of DM and SLE. However, the proposed role of PGRN signaling via TLR9 in DM must be directly demonstrated in future studies. Elevated serum PGRN levels have been noted for a number of other pathological conditions [48]. So, PGRN is most likely not a disease-specific marker, but rather is probably a marker of immune activation.
Most of the DM patients with CIP were treated with PSL and immunosuppressive agents. HSCT was performed for one DM patient with CIP and a refractory skin ulcer. In contrast, several patients with A/SIP required combination therapy with PSL, CNI and IVCY (and HSCTs). This distinction might be responsible for cellular phenotypes affecting the pathogenesis of ILD.
The cumulative survival rate was significantly lower in the group with serum PGRN levels ≥200 ng/ml than in the group with serum PGRN levels <200 ng/ml (
P =0.0009) (Figure
4). These data indicate that baseline serum PGRN levels can predict survival. The group with serum PGRN levels ≥200 ng/ml included two refractory patients with A/SIP treated with HSCTs, which saved their lives. If we had not been able to rescue them, the survival rate in the group with serum PGRN levels ≥200 ng/ml would have been lower. It is inferred from our findings that the intensive combination therapy with various immunosuppressive agents should be chosen for DM patients with ILD showing high serum PGRN levels, especially ≥200 ng/ml.
We must acknowledge some limitations of this study. The sample size of our study was small. Statistical tests usually require a larger sample size to justify that the effect did not happen by chance alone. Moreover, owing to the study’s cross-sectional design, it is difficult to establish the exact and definite causal relationships, except the association between PGRN and development of ILD with DM, on the basis of the collected data. There were potential confounding effects of medication use on serum PGRN levels in patients with inactive DM with CIP and without ILD. Last, to fully validate PGRN as a predictive biomarker, we need to show that elevated PGRN levels predate clinical respiratory decompensation.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
AT performed the ELISA experiments, the statistical analysis and drafted the manuscript. HT, HM, TH and KA designed the study and helped to draft and revise the manuscript. CK performed the important statistical analysis and helped to revise the manuscript. MAy and SO assisted in conducting the ELISA experiments, contributed to data analysis and helped to draft the manuscript. NU, YK, MAk, YA, HN and YT contributed to the acquisition, analysis and interpretation of clinical data and drafted the manuscript. All authors read and approved the final manuscript.