A Serological Biomarker of Versican Degradation is Associated with Mortality Following Acute Exacerbations of Idiopathic Interstitial Pneumonia

Idiopathic interstitial pneumonia (IIP) is a subgroup of the diffuse parenchymal lung diseases (or interstitial lung diseases) characterized by inflammation and fibrosis of the alveolar structures [1]. This definition includes idiopathic pulmonary fibrosis (IPF), which has received special attention due to its severe nature and poor prognosis [2]. An acute exacerbation of IIP is an acute, clinically significant respiratory deterioration that is associated with disease progression and high risk of mortality [1, 3]. In IPF, up to 46% of deaths are preceded by an acute exacerbation and median survival time is as low as 3–4 months [3]. Acute exacerbations may accelerate the underlying fibrotic process, i.e. increase disease activity, and in that way contribute to disease progression. Acute exacerbations of IIP are hard to define and impossible to predict, thus diagnostic and prognostic biomarkers are needed.

Fibrosis, a hallmark of IIP, is characterized by extracellular matrix (ECM) deposition in the parenchyma (i.e. the interstitial matrix) and uncontrolled tissue remodeling. It results from an increased rate of ECM remodeling and a shift in the balance between protein synthesis and degradation, leading to ECM accumulation [4, 5]. ECM remodeling results in the release of protein fragments or neo-epitopes from the tissue. These may enter the circulation where levels can be assessed by specialized ELISAs as a measure of tissue turnover [6]. We have previously shown that serological levels of neo-epitope biomarkers are associated with the presence of an acute exacerbation of chronic obstructive pulmonary disease (COPD) [7, 8]. This indicates that rate of ECM remodeling is significantly altered during acute exacerbations of a chronic lung disease and that the group of serological neo-epitope biomarkers may be reflecting acute changes in disease activity. The association with disease activity is supported by the significant association of neo-epitope biomarkers with progression and mortality of IPF shown by Jenkins et al. in the PROFILE study [9].

Versican is a member of the chondroitin sulfate/dermatan sulfate proteoglycan family and is found in the ECM of soft tissues. It consists of a core protein with attached glycosaminoglycan (GAG) chains that can be sulfated to various degrees, with the pattern affecting cellular events and chemokine binding. Versican plays an important role in tissue homeostasis and is involved in tissue repair, ECM remodeling, inflammation, infection, and cell differentiation, migration, adhesion, and proliferation. Alterations in versican expression influences ECM compliance and the permeability of vessels and airways, and thus, it may be involved in various fibroproliferative disorders. The involvement of versican in chronic obstructive pulmonary disease, asthma, and bronchiolitis obliterans syndrome have been described (reviewed in [10]), underlining the importance of versican for upholding lung structure and function.

In this study, we wanted to investigate the changes in ECM remodeling during an acute exacerbation of IIP and hypothesized that the increased disease activity associated with acute exacerbations would lead to a change in the rate of ECM degradation. This was investigated by assessing ECM remodeling by measuring neo-epitopes in the circulation and evaluating their potential as biomarkers of acute exacerbations of IIP.

Here we have demonstrated for the first time that acute exacerbations of IIP are associated with acute changes in remodeling of the lung ECM and that these are related to outcome. Our main findings were: degradation of collagen type IV (C4M) and VI (C6M) was significantly elevated during an acute exacerbation of IIP as compared with stable disease, while degradation of elastin (ELM7) and versican (VCANM) was significantly decreased; and lower levels of versican degradation (VCANM) during an AE-IIP was significantly associated with increased risk of mortality.

Increased degree of epithelial injury during AE-IIP has previously been suggested [22] and is supported by the finding of elevated KL-6 and SP-D plasma levels in AE-IPF as compared with stable IPF [23]. In the current study, we found increased levels of C4M and C6M, indicating an increased rate of degradation of collagen type IV and VI during acute exacerbations. Collagen type IV is the main component of the basement membrane, located beneath the epithelial layer and collagen type VI is closely associated with the basement membrane, linking it to the interstitial matrix. Injury to the protective epithelial layer is likely to disrupt the underlying basement membrane, resulting in release of degradation fragments from collagen type IV and VI. Thus, our data support the hypothesis that epithelial injury is increased during acute exacerbations of IIP.

Serum levels of VCANM were significantly decreased during AE-IIP as compared to the stable state, indicating a potential build-up of versican in the interstitial lung tissue during an acute exacerbation. Previous results showed decreased serum levels of another proteoglycan, syndecan-4, during AE-IIP [24]. To assess syndecan-4 in serum, the transmembrane protein needs to be shed from the cell surface by MMPs and move to the bloodstream. For this to happen, the protein must undergo proteolytic processing, which could indicate that the degradation of both syndecan-4 and versican is decreased during an acute exacerbation of IIP. Increased versican deposition has been demonstrated in fibroblastic foci of IPF patients [25, 26] as well as in the lungs of patients with COPD [27, 28] or asthma [29], suggesting a role in the faulty repair processes related to chronic lung diseases. Serum VCANM levels at AE-IIP are comparable to the reference level seen in a healthy, unmatched population. One could speculate that an increased rate of degradation during stable disease is an attempt to normalize the increased versican levels found in the lungs. This response might be halted temporarily while the body fights off the acute effects of an exacerbation, resulting in normal levels of versican turnover. Versican has been demonstrated to inhibit the function of elastin-binding protein, resulting in impaired formation of elastic fibers [27, 30]. Thus, the accumulation of versican may have a negative impact on lung and vessel structure. The poorer quality of elastic fibers may lead to a decreased degradation of elastin, as indicated by lower levels of ELM7, as an attempt to prevent further loss of functioning elastic fibers.

Versican is involved in the inflammatory response, and the macrophage-derived versican expression has been found increased in response to acute inflammation [31]. IIP is a chronic condition that may show altered protein expression, whereas an acute exacerbation is an acute state of inflammation leading to a different protein expression pattern. A mouse model of acute lung injury induced by polyinosine-polycytidylic acid to mimic a viral infection has shown accumulation of versican in the lung. Furthermore, versican knock-out mice had an attenuated proinflammatory response [32]. These data could indicate that versican may build up in the lungs during acute exacerbations of chronic lung diseases, which is in line with a decrease in its degradation seen in the present study. Additionally, a decrease in versican degradation has also been associated with acute exacerbations of COPD [7].

The major limitations to this study are related to the retrospective nature of the study design and the heterogeneous study population. The diagnosis of IIP includes different fibrotic lung diseases, which may introduce a greater variation in the present data. We have diagnosed patients with an UIP pattern on HRCT as having IPF using the definition in the 2011/ATS/ERS/JRS/ALAT joint statement. Because surgical lung biopsy was not performed, a histopathological analysis to confirm the UIP pattern was not possible, and the non-IPF patients in the current study represents a heterogeneous population. It has been debated whether a possible UIP pattern on HRCT is sufficient to make a diagnosis of IPF [33‐35], and one study found 23% of patients with HRCT findings inconsistent with the UIP pattern to have a definite/probable UIP pattern on biopsy. Although idiopathic non-specific interstitial pneumonia (NSIP) might be the major type of IIP in the non-IPF group in the present study, it is possible that other IIPs such as IPF and unclassifiable IIPs were included. As a definite diagnosis could not be made in the non-IPF patients, subgroup analysis was only performed for the IPF group. Future studies should include more patients and perform subgroup analysis on IPF and NSIP to evaluate if ECM remodeling differs between these patient groups. Another limitation to this study is the inclusions of patients with unpaired samples or paired samples taken far apart. We have compared baseline characteristics for all patients to investigate the effect of inter-patient variation in the total population, IPF only as well as for patients with paired samples. Only lung function differed significantly between stable disease and time of exacerbations, indicating that the included patients were comparable as lung function is expected to be affected by an acute exacerbation. However, it cannot be ruled out that a lower lung function during acute exacerbation may be related to a permanent change and thus disease progression. We compared unpaired samples and subsequently confirmed the findings for C4M, ELM7 and VCANM using paired samples. Only the association with C6M could not be confirmed using paired samples. For patients with paired samples, the follow-up time between stable disease and acute exacerbation was long which may have influenced the disease progression and activity.

In conclusion, our data indicate that acute exacerbations of IIP are associated with degradation of the basement membrane and a build-up of versican and elastin in the interstitial matrix of the lung, as determined by serological neo-epitope specific biomarkers. The difference in degradation profile for the proteins studied is intriguing and indicate activation of different processes contributing to AE-IIP. Furthermore, the potential build-up of versican indicated by lower levels of VCANM, a biomarker reflecting versican degradation, is associated with increased risk of mortality. Thus, serological neo-epitope biomarkers reflecting lung ECM remodeling may be useful tools in the non-invasive assessment of acute exacerbations of IIP, and VCANM may represent a novel prognostic biomarker. However, the current data should be validated in separate cohorts to evaluate the potential use.