High levels of biomarkers of collagen remodeling are associated with increased mortality in COPD - results from the ECLIPSE study

Abstract

Background: There is a need to identify individuals with COPD at risk for disease progression and mortality. Lung tissue remodeling is associated with the release of extracellular matrix (ECM) fragments into the peripheral circulation. We hypothesized that ECM remodeling was associated with mortality in COPD and measured neo-epitopes originating from ECM proteins associated with lung tissue remodeling.

Methods: Biomarkers of ECM remodeling were assessed in a subpopulation (n = 1000) of the Evaluation of COPD Longitudinally to Identify Predictive Surrogate End-points (ECLIPSE) cohort. Validated immunoassays measuring serological neo-epitopes produced by proteolytic cleavage associated with degradation of collagen type I, III, IV, and VI, elastin, and biglycan, and formation of collagen type VI as well as fibrinogen and C-reactive protein were used. Multivariate models were used to assess the prognostic value of these biomarkers.

Results: Thirty subjects (3.0 %) died during follow-up. Non-survivors were older, had reduced exercise capacity, increased dyspnea score, and included fewer current smokers. All collagen biomarkers were significantly elevated in non-survivors compared to survivors. Mortality risk was significantly increased for subjects with collagen remodeling biomarkers in the upper quartile, especially for the degradation fragment of collagen type IV C6M (hazard ratio 6.6 [95 % confidence interval 2.9-15.2], P < 0.0001) after adjusting for relevant confounders.

Conclusions: Serological biomarkers of collagen remodeling were strongly associated with mortality in subjects with COPD indicating that assessment of tissue turnover in the parenchyma and small airways may be useful in the prognosis of COPD.

Trial registration: NCT00292552 , GSK Study No. SCO104960.

Keywords: Biomarker; COPD; Collagen; Elastin; Extracellular matrix; Mortality; Prognostic; Remodeling.

Figures

Fig. 1

Extracellular matrix composition in healthy and COPD lungs. In healthy lungs, the epithelial cells create a tight barrier thereby blocking entry of foreign particles from the inhaled air to the tissue. This is further enforced by the underlying basement membrane (BM) which mainly consists of collagen type IV. The interstitial matrix (IM) is below the BM and consists mainly of fibrillar collagens and elastin. In COPD, the continuous epithelial layer is disrupted and the underlying BM is exposed. The inflammatory response to repetitive tissue injury results in up-regulation of proteases and disruption of the BM, allowing for injury to the underlying IM. In response to this, fibroblasts are activated and converted to pro-fibrotic myofibroblasts that secrete collagens which accumulate in the IM of the airway wall. Both collagens and elastin undergoes proteolytic degradation in the airway and alveolar walls. The processes of synthesis and degradation release extracellular matrix (ECM) fragments which may enter the bloodstream from where they can be assessed as biomarkers of ECM remodeling

Fig. 2

Biomarker levels in survivors versus non-survivors. Biomarker levels were assessed at month six (BGM, C1M, EL-NE, Pro-C6) or year one (C3A, C3M, C4M, C6M, CRPM, ELM7, CRP, fibrinogen) in survivors (n = 970) and non-survivors (n = 30). All biomarkers, with the exception of BGM, EL-NE, and fibrinogen, were significantly elevated in non-survivors compared to survivors. Data are shown as geometric mean ± SEM. Statistical significance was determined using student’s t-test: *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001

Fig. 3

Cox proportional hazards ratios for death. Hazard ratios (HR) for death are shown per one standard deviation increase in biomarker (a + b) and for subjects in the upper vs. lower quartile (c + d). Cox proportional HR with 95 % confidence intervals are shown for crude analyses (a + c) and analyses adjusted for age, BODE, and previous exacerbations (b + d). All biomarkers, with the exception of BGM, ELM7, and EL-NE, were significantly related to mortality outcome in both crude and adjusted analyses. Statistical significant hazard ratios are indicated as *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001

Fig. 4

Kaplan-Meier survival curves for biomarker quartiles. Kaplan-Meier survival curves are shown for the biomarkers with crude hazard ratios for quartile 4 vs. 1 above 5. The relationship of biomarker quartiles with survival time from blood sampling are shown for plasma C3A (year one), plasma C3M (year one), plasma C6M (year one), serum ProC6 (month six), and plasma CRP (year one). Subjects in the upper biomarker quartiles (Q4) showed a higher number of deaths within the study period than the lower quartiles