Pulse wave velocity (PWV) in patients (pts) with moderate COPD and cardiovascular risk: The effect of an inhaled long-acting β2-agonist/corticosteroid (SUMMIT)

Abstract

Introduction: COPD is associated with cardiovascular disease (CVD). PWV can assess CVD risk and is elevated in COPD pts vs age/smoking matched controls. Some data suggest PWV may be reduced by COPD treatment. Methods: SUMMIT was a multicenter, randomized, double-blind trial in 16485 moderate COPD pts with, or at high risk of CVD. Pts were randomized to either inhaled placebo (n=4111), long-acting β 2 -agonist (vilanterol [VI] 25μg; n=4118), corticosteroid (fluticasone furoate [FF] 100μg; n=4135), or (VI 25μg + FF 100μg; n=4121). Carotid-femoral PWV was measured at baseline (n=1788) and over the next 2-3 yrs. Results: Pts were 65±8yrs, mostly men (75%); 47% active smokers. Most had overt CVD (66%) and were receiving anti-platelet, lipid-lowering, and renin-angiotensin system inhibitor therapies. All-cause mortality (primary endpoint) was not affected by FF/VI (HR 0.88, 95% CI 0.74-1.04; p=0.14), FF or VI. In a post hoc analysis of the subset with PWV measurements, probability of all-cause mortality [95% CI] at 2yrs increased as a function of baseline PWV: Quartile (Q) 1 ≤7.4m/s: 3.7% [2.0, 6.6]; Q2 >7.4- ≤8.8m/s: 2.3% [1.2, 4.3]; Q3 >8.8- ≤10.7m/s: 5.0% [2.8, 8.8]; Q4 >10.7m/s: 7.8% [5.2, 11.6]. A post hoc analysis of on-treatment CV composite (CV death, myocardial infarction, stroke, unstable angina, transient ischemic attack) showed a similar pattern. No treatment affected change in PWV at 1yr overall, or when baseline PWV was dichotomized (11m/s). Conclusion: In moderate COPD pts with or at high-risk of CVD, baseline PWV was predictive of mortality and CV events but unaffected by inhaled therapy. GSK Funded: 113782/NCT01313676.