The Summit Score Stratifies Mortality and Morbidity in Chronic Obstructive Pulmonary Disease

Benjamin D Horne,1,2 Matthew J Hegewald,3 Courtney Crim,4 Susan Rea,5 Tami L Bair,1 Denitza P Blagev3 1Intermountain Medical Center Heart Institute, Salt Lake City, UT, USA; 2Division of Cardiovascular Medicine, Department of Medicine, Stanford University, Stanford, CA, USA; 3Division of Pulmonary Medicine, Department of Internal Medicine, Intermountain Medical Center, Salt Lake City, UT, USA; 4Research and Development, GlaxoSmithKline, Research Triangle Park, NC, USA; 5Care Transformation, Intermountain Healthcare, Salt Lake City, UT, USACorrespondence: Benjamin D HorneIntermountain Medical Center Heart Institute, 5121 S. Cottonwood Street, Salt Lake City, UT 84107, USATel +1801-507-4701Email benjamin.horne@imail.org Twitter@DrBenjaminHorneIntroduction: Tobacco use and other cardiovascular risk factors often accompany chronic obstructive pulmonary disease (COPD). This study derived and validated the Summit Score to predict mortality in people with COPD and cardiovascular risks.Methods: SUMMIT trial subjects (N=16,485) ages 40– 80 years with COPD were randomly assigned 50%/50% to derivation (N=8181) and internal validation (N=8304). Three external COPD validations from Intermountain Healthcare included outpatients with cardiovascular risks (N=9251), outpatients without cardiovascular risks (N=8551), and inpatients (N=26,170). Cox regression evaluated 40 predictors of all-cause mortality. SUMMIT treatments including combined fluticasone furoate (FF) 100μg/vilanterol 25μg (VI) were not included in the score.Results: Mortality predictors were FEV1, heart rate, systolic blood pressure, body mass index, age, smoking pack-years, prior COPD hospitalizations, myocardial infarction, heart failure, diabetes, anti-thrombotics, anti-arrhythmics, and xanthines. Combined in the Summit Score (derivation: c=0.668), quartile 4 vs 1 had HR=4.43 in SUMMIT validation (p< 0.001, 95% CI=3.27, 6.01, c=0.662) and HR=8.15 in Intermountain cardiovascular risk COPD outpatients (p< 0.001, 95% CI=5.86, 11.34, c=0.736), and strongly predicted mortality in the other Intermountain COPD populations. Among all SUMMIT subjects with scores 14– 19, FF 100μg/VI 25μg vs placebo had HR=0.76 (p=0.0158, 95% CI=0.61, 0.95), but FF 100μg/VI 25μg was not different from placebo for scores < 14 or > 19.Conclusion: In this post hoc analysis of SUMMIT trial data, the Summit Score was derived and validated in multiple Intermountain COPD populations. The score was used to identify a subpopulation in which mortality risk was lower for FF 100μg/VI 25μg treatment.Trial Registration: The SUMMIT trial is registered at ClinicalTrials.gov as number NCT01313676.Keywords: clinical decision tool, risk score, randomized controlled trial, Intermountain risk score, IMRS