Progression of Kidney Disease in Non-Diabetic Patients with Coronary Artery Disease: Predictive Role of Circulating Matrix Metalloproteinase-2, -3, and -9

BACKGROUND: Circulating matrix metalloproteinase (MMP)-2, -3 and -9 are well recognized in predicting cardiovascular outcome in coronary artery disease (CAD), but their risks for chronic kidney disease (CKD) are lacking. Therefore, the present study aimed to investigate whether circulating MMP levels could independently predict future kidney disease progression in non-diabetic CAD patients. METHODS: The prospective study enrolled 251 non-diabetic subjects referred for coronary angiography, containing normal coronary artery (n = 30) and CAD with insignificant (n = 95) and significant (n = 126) stenosis. Estimated glomerular filtration rate (eGFR) was calculated using the CKD-EPI formula. eGFR decline rate was calculated and the primary endpoint was a decline in eGFR over 25% from baseline. RESULTS: The eGFR decline rate (ml/min/1.73 m(2) per year) in patients with CAD (1.22 [−1.27, 1.05]) was greater than that in those with normal coronary artery (0.21 [−2.63, 0.47], P<0.01). The circulating MMP-2, -3 and -9 were independently associated with faster eGFR decline among CAD patients. The mean follow-up period was 8.5±2.4 years, and 39 patients reached the primary endpoint. In multivariate Cox regression model, the adjusted hazard ratios of MMP-2 ≥861 ng/mL, MMP-3 ≥227 ng/mL and MMP-9 ≥49 ng/mL for predicting CKD progression were 2.47 (95% CI, 1.21 to 5.07), 2.15 (1.12 to 4.18), and 4.71 (2.14 to 10.4), respectively. While added to a model of conventional risk factors and baseline eGFR, MMP-2, -3 and -9 further significantly improved the model predictability for CKD progression (c statistic, 0.817). In the sensitivity analyses, the results were similar no matter if we changed the endpoints of a decline of >20% in eGFR from baseline or final eGFR < 60 mL/min/1.73 m(2). CONCLUSION: Circulating MMP-2, -3 and -9 are independently associated with kidney disease progression in non-diabetic CAD patients and add incremental predictive power to conventional risk factors.