Rationale: Matrix metalloprotease (MMP)-9 couples cigarette smoke to inflammation, airway remodeling, and emphysema development – all hallmarks of COPD. However, no study has characterized individuals with COPD and elevated MMP-9 nor evaluated its impact on outcomes including COPD exacerbations (AECOPD).
Methods: We used demographic, clinical, physiologic, and CT data from the Subpopulations and Intermediate Outcomes in COPD Study (SPIROMICS) and Evaluation of COPD Longitudinally to Identify Predictive Surrogate End-Points (ECLIPSE) studies. We included individuals with GOLD Stage 2-4 COPD that had plasma MMP-9 values and prospective AECOPD data recorded. High MMP-9 was defined as plasma MMP-9 459ng/m1 (>90th percentile value measured in a control population) measured by RBM-Myriad’s multiplex-platform at baseline.
An AECOPD was defined as an unscheduled healthcare visit or by antibiotic and/or systemic glucocorticoid treatment. We used logistic regression models to identify associations between clinical characteristics and high MMP-9, with or without prospective AECOPDs; zero-inflated negative binomial regression to assess associations with AECOPD frequency; and Cox proportional hazards models to determine time-to-first-AECOPD.
Results: We analyzed data from 200 ECLIPSE participants and 796 individuals from SPIROMICS. Participants from SPIROMICS had more current smokers (33% versus 2.5%) and higher FEVi (53±17% versus 44±17%) compared to ECLIPSE. Baseline plasma MMP-9 levels were 279±191 ng/mI and 223±183 ng/ml in the ECLIPSE and SPIROMICS cohorts, respectively. High MMP-9 was present in 20 (10%) of ECLIPSE and 78 (9.7%) of SPIROMICS subjects.
AECOPD occurred during follow-up in 162 (81%) in ECLIPSE when followed for 1085 (IQR=720-1115) days and in 453 (58%) in SPIROMICS when followed for 1022 (IQR=739-1119) days. In both cohorts, we found positive associations between high MMP-9 and FEVi as well as high MMP-9 and plasma white blood cell count. Using logistic regression models adjusting for age, sex, smoking status, pack-year, FEV.I, chronic bronchitis, asthma, GERD, and AECOPD within the last year, high MMP-9 (OR 3.2, 95%Cl 1.0-10.0, P=0.04 – ECLIPSE; OR 2.27, 95%CI 1.27-4.05, P=0.006 – SPIROMICS) was associated with at least one AECOPD during follow-up.
In SPIROMICS, the high MMP-9 subgroup had a higher rate of AECOPD (rate ratio 1.86, 95%CI 1.44-5.05, P=0.002) and a shorter time to first event (577 vs 734 days, HR 1.36, 95%CI 1.01-1.84, P=0.043) compared to the non-elevated-MMP-9 subgroup.
Conclusions: Systemic MMP-9 is reliably measured in two well-characterized COPD cohorts, is associated with FEVi and plasma WBC count, and independently predicts future exacerbations. These findings provide strong evidence for plasma MMP-9 as a COPD biomarker.