Idiopathic pulmonary fibrosis (IPF) is characterized by myofibroblast accumulation and progressive lung scarring. Mesenchymal cells are responsible for deposition of extracellular matrix proteins such as collagen and fibronectin that fill the lung and compromise gas exchange (1). These mesenchymal cells may arise from three distinct sources. First, resident lung fibroblasts proliferate and differentiate into myofibroblasts, prodigious producers of collagen and contractile contributors to alveolar collapse and traction bronchiectasis (2). The second source may involve epithelial to mesenchymal transition, a process whereby epithelial cells release from basement membrane and undergo reprogramming that allows them to acquire a mesenchymal phenotype (3). The third source involves recruitment of circulating bone marrow–derived precursors, known as fibrocytes, which share mesenchymal and leukocyte markers (4). Animal models have indicated that circulating progenitors can augment airway and interstitial lung fibrosis (5–9). However, verification in human IPF is limited to two studies. One study analyzed four patients with the usual interstitial pneumonia form of IPF as well as one patient with nonspecific interstitial pneumonia and found that fibrocytes comprised between 6 and 10% of the leukocytes in the buffy coat compared with approximately 0.5% in normal controls (10). A histological study demonstrated the presence of cells that coexpressed mesenchymal and leukocyte markers in IPF lung (11).

In this issue of the Journal, Moeller and colleagues (pp. 588–594) report on a cohort of fifty-eight patients with IPF (12). Within this cohort, seven patients were identified as experiencing an acute exacerbation, whereas fifty-one were characterized as stable. Significant increases in percentages of circulating fibrocytes were found in both stable and acute cohorts with IPF when compared with healthy volunteers, but the numbers were most striking in patients with IPF acute exacerbations. One surprise in this study was that significantly increased percentages of circulating fibrocytes were not present in patients with acute respiratory distress syndrome. These results may reflect the fact that none of the ten patients with acute respiratory distress that were examined went on to develop fibroproliferative lung disease. These studies suggest that the percentage of fibrocytes in circulation may serve as a biomarker for the presence of fibrotic lung disease and may be a useful biomarker for acute exacerbations. Additionally, the current study indicates that fibrocyte percentages greater than 5% are independent predictors of early mortality in IPF. The fact that circulating fibrocytes are increased in the setting of IPF acute exacerbation, but not in the setting of acute respiratory distress syndrome, also suggests that the biological processes occurring in the setting of IPF acute exacerbations are fundamentally different from those occurring with diffuse alveolar damage during acute lung injury.

Now that the study by Moeller and colleagues (12) provides evidence for increased fibrocyte numbers on a robust cohort of patients with IPF, what are we to do with this information? I believe that these data beg for new areas of both basic science and clinical investigation. What causes the influx of fibrocytes into the blood? Are fibrocytes in stable patients with IPF fundamentally different from fibrocytes in patients experiencing acute exacerbations?

How do fibrocytes augment fibrosis? Adoptive transfer of fibrocytes worsens bleomycin-induced fibrosis in mice suggesting that fibrocytes are pathogenic (7), but their mechanism of action is unknown. Studies in skin wounding and airway fibrosis models suggest fibrocytes differentiate into myofibroblasts in vivo (4), but whether this occurs in the lung interstitium is unclear. One study suggests bone marrow–derived mesenchymal cells do not differentiate into myofibroblasts (6). Thus, fibrocytes may have paracrine effects that activate resident lung fibroblasts via the secretion of profibrotic mediators, such as transforming growth factor β (13). Fibrocytes may also induce epithelial to mesenchymal transition via their secretion or induction of proteases (e.g., MMP9) and profibrotic factors (14). These scenarios should be addressable in animal models.

Do increases in fibrocyte numbers predict the onset of acute exacerbations? Perhaps the most intriguing data in the Moeller study is the observation that circulating fibrocyte percentages decreased in three patients who recovered following an acute exacerbation (12). Although the numbers of patients with acute exacerbations were small, the data suggest frequent, serial measurements of fibrocyte percentages may predict acute exacerbations, so larger studies are needed.

Are fibrocyte numbers a useful predictor of mortality for clinical trials? Currently, therapeutic trials in IPF are hampered by the lack of robust easily obtained outcome variables for analysis and response to therapy. Changes in forced vital capacity are used to predict disease severity but may not reflect pathologic changes in the disease process (15). Serial analyses of fibrocytes may predict stability, deterioration, and/or acute exacerbation. In the Moeller study (12), fifteen stable patients with IPF showed no difference in circulating fibrocyte percentages when measured at two different timepoints between 4 and 12 months after the initial analysis. In contrast, elevated fibrocyte percentages predicted mortality. Thus, it is possible that a relative increase in fibrocyte numbers may precede rapid clinical and functional deteriorations in IPF.

Can alterations in fibrocyte numbers predict response to therapy and do efficacious drugs diminish fibrocyte percentages in circulation? This question is best addressed in prospective placebo-controlled clinical trials such as those organized through the IPF Network.

Do subsets of fibrocytes better predict disease outcome? Subsets of fibrocytes are readily identified on the basis of chemokine receptor expression (4, 8, 16). However, it is not known whether these subsets display functional differences in recruitment, extracellular matrix secretion, proliferation, or mediator release. The first step in addressing this question is to analyze fibrocyte subsets serially in patients. In fact, an ancillary study for the IPF Network “Prednisone, Azathioprine and N-acetylcysteine: a Three arm study that Evaluates Responses in IPF,” or PANTHER trial, will include an analysis of fibrocyte numbers and chemokine receptor profiling at baseline, midpoint, and at the conclusion of the study (Weeks 0, 32, and 60) for up to 130 patients in each of three study arms. Future animal studies should provide mechanistic insights into the pathobiology of fibrocytes, whereas clinical investigations focusing on longitudinal analyses should provide clarification on the usefulness of fibrocyte measurements as biomarkers in IPF.

References

Section:

ChooseTop of pageReferences <<CITING ARTICLES