Fibrotic lung diseases have recently been given a new classification as idiopathic interstitial pneumonias (IIPs) and have been categorized into different subtypes, such as usual (UIP), nonspecific (NSIP), and desquamative (DIP) interstitial pneumonias as well as cryptogenic organizing pneumonia (COP) (1). However, fibrotic remodeling of lung parenchyma is also a main feature of lung involvement in patients with rheumatic diseases. Approximately half of all diagnosed cases with the typical histologic pattern of IIPs are associated with a rheumatic disease (2). Therefore, lung involvement in patients with rheumatic diseases can be identified by the same histologic and radiologic patterns as described in patients with IIPs, and is not distinguishable from the idiopathic variants. For example, in patients with systemic sclerosis (SSc), lung involvement may occur in a pattern typical of NSIP, but could also resemble that of UIP (2).
In patients whose disease is associated with fibrotic lung remodeling, including patients with SSc as well as patients with radiologic types III and IV sarcoidosis, hypersensitivity pneumonitis, or idiopathic pulmonary fibrosis (IPF) (UIP pattern), we recently found that a common feature is shared: the phenotype of the alveolar macrophages in these patients is characterized by alternative activation with exaggerated production of the CC chemokine ligand CCL18 (3). In addition, increased expression of CCL18 messenger RNA (mRNA) has been described in bronchoalveolar lavage (BAL) cells from patients with SSc (4). Recently, Kodera et al (5) observed increasing CCL18 concentrations in the serum of patients with SSc, indicating the presence of pulmonary fibrotic remodeling. Furthermore, their findings suggest that CCL18 might serve as an indicator of disease activity, since the authors observed decreasing CCL18 concentrations in 20 patients with SSc after treatment with immunosuppressive agents.
Our previous studies, along with those by Atamas et al (6), have shown that the increased production of CCL18 by alternatively activated alveolar macrophages in patients with pulmonary fibrosis promotes the production of collagen by lung fibroblasts (3). In these patients, we observed a vicious circle in which the interaction of alternatively activated alveolar macrophages and lung fibroblasts perpetuates the fibrotic process (3). CCL18 is a chemokine that has thus far been described only in primates and is abundantly produced within the lung (7–9). Its receptor is not known, and only myeloid-derived cells have been found to produce this mediator, which was previously termed pulmonary and activation-regulated chemokine (7–9).
Following the revised classification of the IIPs, several large studies exploring new therapeutic strategies for IPF and SSc have recently been published (10, 11) or are in progress. Very often, the results of the various studies have been contradictory with regard to the influence of the tested treatment schedule on pulmonary function, mortality, or the 6-minute walk test (12, 13). At present there is no established biomarker that can adequately reflect pulmonary fibrotic activity.
Given this background, we wondered whether CCL18 can serve as a biomarker of disease activity in patients with pulmonary fibrosis. We examined a total of 43 patients with either IPF, DIP, NSIP, or COP and 12 patients with SSc, as well as 23 healthy control subjects. Our results reveal that CCL18 is abundantly expressed in the culture supernatants of BAL cells and the BAL fluid and sera from patients whose pulmonary fibrosis is histologically proven. Our results also show that concentrations of CCL18 in BAL fluid are highly correlated with known markers of disease severity. In addition, in the course of disease in 40 patients with different IIPs or SSc, we observed a close negative correlation between changes in the total lung capacity and changes in the serum CCL18 concentrations. It therefore appears that the concentration of CCL18 in the serum may serve as an indicator of pulmonary disease activity in patients with pulmonary fibrosis, in both its idiopathic and its rheumatic forms.
- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
Pulmonary fibrosis represents a final pathway of various pulmonary diseases, including lung involvement in rheumatic diseases. In a recent report we showed that different interstitial lung diseases, such as radiologic types III and IV sarcoidosis, hypersensitivity pneumonitis, and IPF, share the up-regulation of CCL18 production as a common mechanism, and that a vicious circle involving macrophages and fibroblasts and their products, CCL18 and collagen, perpetuates pulmonary fibrosis (3). In this study we have shown that CCL18 is abundantly produced by BAL cells from patients with IIPs and patients with SSc. Production of CCL18 by BAL cells was correlated with pulmonary function results and BAL neutrophil cell counts, which are both known markers of disease progression. In addition, in another group of 40 patients with fibrotic lung disease, we observed an extraordinarily close correlation between the course of total lung capacity and changes in serum CCL18 levels.
Our data clearly demonstrate that CCL18 production by BAL cells is markedly enhanced in patients with fibrotic lung disease in the IIPs or SSc. BAL cells from patients with IPF (UIP pattern), patients with NSIP, or patients with SSc produced up to 100-fold more CCL18 than did the BAL cells of healthy control subjects. In addition, BAL cells from patients with other diseases associated with pulmonary fibrosis, such as COP or DIP, produced a higher level of CCL18 than was found in controls. We have recently shown that there is a strong link between CCL18 production and the evolution of pulmonary fibrosis in such different diseases as sarcoidosis, IPF, and hypersensitivity pneumonitis. We herein show that there is also an increase in CCL18 protein production in IIPs such as DIP, NSIP, COP, and IPF (UIP pattern), as well as in SSc. An increase in CCL18 mRNA expression in patients with SSc (4) and patients with IPF (18) has been described previously. However, we are the first to comprehensively describe CCL18 production by BAL cells in a broad spectrum of fibrotic lung diseases. In a recent study, we found that CCL18 produced by alveolar macrophages enhances collagen production by lung fibroblasts (3), and conversely, fibroblast production of collagen up-regulates CCL18 production by alveolar macrophages. The present study demonstrates that this vicious circle is a common mechanism in all tested fibrotic lung diseases. Thus, our data suggest that an increase in CCL18 production by BAL cells indicates the evolution of pulmonary fibrotic remodeling.
CCL18 concentrations were also highly increased in the BAL fluid from all tested patient cohorts with fibrotic lung disease. In patients with IPF, SSc, or NSIP, CCL18 concentrations in BAL fluid were up to 100-fold higher than is considered normal, similar to the results obtained on CCL18 production by BAL cells. Consistent with our previous findings and those by Pardo et al (3, 18), we were able to show, by in situ hybridization and immunohistochemistry and by assessment of isolated cells, that CCL18 is mainly produced by alveolar macrophages. The close correlation between the levels of CCL18 in BAL fluid and those in BAL cell cultures suggests that the main CCL18-producing cells can be recovered by BAL. According to previously published results from our group (3), flow cytometric analysis of BAL cells revealed that only alveolar macrophages, and not BAL lymphocytes, produce CCL18. The difference in CCL18 production between patients and controls is caused by an increase in the percentage of CCL18-producing cells as well as an increase in the CCL18 production per cell in patients with fibrotic lung diseases.
Several factors could have an influence on the observed differences in CCL18 production by BAL cells between different patient cohorts. We have previously shown that collagen itself up-regulates CCL18 production by alveolar macrophages, in a concentration-dependent manner. Therefore, differences in fibrotic activity, which is tantamount to increased collagen production and fibroblast proliferation, might be responsible for the observed differences in CCL18 production between patient cohorts. Since there was heterogeneity in disease severity, particularly in the cohort of patients with SSc, this might also explain the broad range of CCL18 production within the patient cohorts.
In this context we wondered whether CCL18 levels in BAL cell supernatants and in serum were indicators of disease severity. To test this, we analyzed the relationship between parameters of pulmonary function and differential BAL cell counts. We observed a negative correlation between CCL18 concentrations in BAL cell supernatants or serum and the total lung capacity, and a negative correlation between CCL18 concentrations and the DLCO. Both of these pulmonary function markers are known to reflect disease severity and are associated with the risk of death in patients with pulmonary fibrosis (19–22). In addition, there was a correlation between CCL18 concentrations in BAL cell supernatants or serum and BAL neutrophil and eosinophil cell counts, which have been described as markers of disease progression and fibrotic activity in patients with pulmonary fibrosis and SSc (23–27). The possibility that CCL18 has a direct influence on peripheral blood neutrophils via its chemotactic activity was recently excluded (28). CCL18 is described as a chemokine capable of attracting naive T cells, activated T cells, monocytes, eosinophils, and dendritic cells (9, 28, 29). Nevertheless, despite the correlations with BAL eosinophil cell counts, there was no correlation between BAL cell lymphocyte counts and concentrations of CCL18 in BAL cell supernatants, and there was only a weak correlation with BAL macrophage cell counts.
We showed a close correlation between the concentrations of CCL18 in BAL fluid and those in serum. Interestingly, CCL18 levels in the serum were similarly high, and detectable within the 100-ng range. This correlation between CCL18 levels in BAL fluid and those in the serum from all examined disease groups suggests that the level of CCL18 production within the lung is reflected in the serum production of CCL18. However, we cannot exclude the possibility that other factors might influence CCL18 levels in the serum. In our experiments, alveolar macrophages produced ∼100-fold more CCL18 compared with freshly isolated monocytes. None of our patients had a disease, such as atopic dermatitis (30) or Gaucher disease, that is known to be associated with increased CCL18 serum levels (31). The increase in serum CCL18 concentrations in patients with fibrotic lung diseases might be associated with multiple mechanisms. First, there might be spillover from the lung. Second, results from several studies have suggested that there is also an increase in serum collagen concentrations in patients with fibrotic lung disease (32, 33). We have shown that macrophages produce more CCL18 following stimulation with collagen (3). Therefore, it might be possible that the increase in serum collagen concentrations will cause an increase in CCL18 production by extrapulmonary macrophages that are exposed to serum collagen. Furthermore, it is tempting to speculate that there is systemic aberrant CCL18 production by macrophages in patients with fibrotic lung diseases. This issue requires further study.
In other experiments, we analyzed the relationship between serum CCL18 concentrations and disease progression. We observed a close negative correlation between the changes in serum CCL18 concentrations and the changes in total lung capacity. This effect was observed regardless of the given therapy. More precisely, in patients in whom pulmonary function improved with the use of steroids (with or without further immunosuppressive therapy), CCL18 levels decreased. In contrast, in patients with IPF or SSc whose pulmonary function declined despite the use of therapy with immunosuppressants such as steroids, CCL18 serum concentrations increased. This finding suggests that changes in serum CCL18 concentrations reflect disease activity rather than the severity of the disease. In patients in whom disease activity was obviously ameliorated by steroids, such as in patients with COP, the CCL18 concentrations dramatically decreased. Therefore, serum CCL18 levels could serve as an indicator of disease activity in patients with pulmonary fibrosis who are being treated with a distinct drug regimen.
The observed correlation between the course of predicted total lung capacity and changes in serum CCL18 concentrations was extraordinarily close, with a correlation coefficient of −0.78. To our knowledge, the correlation between other known serum markers, such as KL-6, interleukin-8, or surfactant D, and pulmonary function data is far lower. Recently, Kodera et al (5) showed that serum CCL18 concentrations in patients with SSc were indicative of the evolution of pulmonary fibrosis, and that serum CCL18 levels better reflected the pulmonary outcome in patients with SSc than did, for instance, the levels of KL-6. In 21 patients with SSc who showed an improvement in their fibrosis activity score on HRCT following immunosuppressive therapy, Kodera et al (5) found decreased serum CCL18 concentrations in parallel with a change in pulmonary fibrotic activity in all but 1 of the patients. Combining these data with our findings, we propose that increasing serum CCL18 concentrations can be used as an indicator of the evolution of fibrotic lung remodeling, and that serum CCL18 concentrations reflect the fibrotic lung activity in patients with pulmonary fibrosis independent of the cause of the disease.
In conclusion, our studies show that the production of CCL18 within the lung is dramatically up-regulated in patients with IIPs and in those with SSc, both of which are diseases known to be associated with fibrotic lung tissue remodeling. We demonstrate that there is a close correlation between CCL18 production and other known factors of disease progression in patients with pulmonary fibrosis. Serum CCL18 concentrations are strongly correlated with CCL18 production by BAL cells. In contrast to an approach utilizing BAL neutrophil cell counts, serum CCL18 concentrations are easily determined and repeatable. Furthermore, in contrast to pulmonary function tests, no cooperation of the patient is needed for determinations of CCL18. We believe that repeated measurements of serum CCL18 concentrations can provide additional and valid information. Monitoring serum CCL18 concentrations in patients with rheumatic diseases might reveal the evolution of pulmonary fibrotic remodeling. Furthermore, monitoring serum CCL18 concentrations during the course of the disease and at the time when pulmonary fibrosis is already established might be an extraordinarily useful tool in clinical practice and in studies evaluating new antifibrotic approaches to treatment.
- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
Dr. Prasse had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study design. Prasse, Pechkovsky, Toews, Müller-Quernheim.
Acquisition of data. Prasse, Pechkovsky, Eggeling, Kollert, Germann, Schäfer, Ludwig, Zissel.
Analysis and interpretation of data. Prasse, Pechkovsky, Toews, Eggeling, Kollert, Zissel, Müller-Quernheim.
Manuscript preparation. Prasse, Toews, Eggeling, Müller-Quernheim.
Statistical analysis. Prasse, Zissel.
Recruitment of study cohort. Prasse, Müller-Quernheim.