Abhinav Sharma1,2, Yinggan Zheng3, Justin A Ezekowitz2,3, Cynthia M Westerhout3, Jacob A Udell4, Shaun G Goodman3,5, Paul W Armstrong3, John B Buse6, Jennifer B Green7, Robert G Josse5, Keith D Kaufman8, Darren K McGuire9, Giuseppe Ambrosio10, Lee-Ming Chuang11, Renato D Lopes7, Eric D Peterson7, Rury R Holman12. 1. 1Division of Cardiology, McGill University, Montreal, Quebec, Canada. 2. 2Mazankowski Alberta Heart Institute, University of Alberta, Edmonton, Alberta, Canada. 3. 3Canadian VIGOUR Centre, University of Alberta, Edmonton, Alberta, Canada. 4. 4Peter Munk Cardiac Centre, University Health Network and Women's College Hospital, University of Toronto, Toronto, Ontario, Canada. 5. 5St. Michael's Hospital, University of Toronto, Toronto, Ontario, Canada. 6. 6School of Medicine, University of North Carolina at Chapel Hill, Chapel Hill, NC. 7. 7Duke Clinical Research Institute, Duke University School of Medicine, Durham, NC. 8. 8Merck & Co., Inc., Kenilworth, NJ. 9. 9Division of Cardiology, University of Texas Southwestern Medical Center, Dallas, TX. 10. 10School of Medicine, University of Perugia, Perugia, Italy. 11. 11Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan. 12. 12Radcliffe Department of Medicine, University of Oxford, Oxford, U.K.
Abstract
OBJECTIVE: Phenotypic heterogeneity among patients with type 2 diabetes mellitus (T2DM) and atherosclerotic cardiovascular disease (ASCVD) is ill defined. We used cluster analysis machine-learning algorithms to identify phenotypes among trial participants with T2DM and ASCVD. RESEARCH DESIGN AND METHODS: We used data from the Trial Evaluating Cardiovascular Outcomes with Sitagliptin (TECOS) study (n = 14,671), a cardiovascular outcome safety trial comparing sitagliptin with placebo in patients with T2DM and ASCVD (median follow-up 3.0 years). Cluster analysis using 40 baseline variables was conducted, with associations between clusters and the primary composite outcome (cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, or hospitalization for unstable angina) assessed by Cox proportional hazards models. We replicated the results using the Exenatide Study of Cardiovascular Event Lowering (EXSCEL) trial. RESULTS: Four distinct phenotypes were identified: cluster I included Caucasian men with a high prevalence of coronary artery disease; cluster II included Asian patients with a low BMI; cluster III included women with noncoronary ASCVD disease; and cluster IV included patients with heart failure and kidney dysfunction. The primary outcome occurred, respectively, in 11.6%, 8.6%, 10.3%, and 16.8% of patients in clusters I to IV. The crude difference in cardiovascular risk for the highest versus lowest risk cluster (cluster IV vs. II) was statistically significant (hazard ratio 2.74 [95% CI 2.29-3.29]). Similar phenotypes and outcomes were identified in EXSCEL. CONCLUSIONS: In patients with T2DM and ASCVD, cluster analysis identified four clinically distinct groups. Further cardiovascular phenotyping is warranted to inform patient care and optimize clinical trial designs.
OBJECTIVE: Phenotypic heterogeneity among patients with type 2 diabetes mellitus (T2DM) and atherosclerotic cardiovascular disease (ASCVD) is ill defined. We used cluster analysis machine-learning algorithms to identify phenotypes among trial participants with T2DM and ASCVD. RESEARCH DESIGN AND METHODS: We used data from the Trial Evaluating Cardiovascular Outcomes with Sitagliptin (TECOS) study (n = 14,671), a cardiovascular outcome safety trial comparing sitagliptin with placebo in patients with T2DM and ASCVD (median follow-up 3.0 years). Cluster analysis using 40 baseline variables was conducted, with associations between clusters and the primary composite outcome (cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, or hospitalization for unstable angina) assessed by Cox proportional hazards models. We replicated the results using the Exenatide Study of Cardiovascular Event Lowering (EXSCEL) trial. RESULTS: Four distinct phenotypes were identified: cluster I included Caucasian men with a high prevalence of coronary artery disease; cluster II included Asian patients with a low BMI; cluster III included women with noncoronary ASCVD disease; and cluster IV included patients with heart failure and kidney dysfunction. The primary outcome occurred, respectively, in 11.6%, 8.6%, 10.3%, and 16.8% of patients in clusters I to IV. The crude difference in cardiovascular risk for the highest versus lowest risk cluster (cluster IV vs. II) was statistically significant (hazard ratio 2.74 [95% CI 2.29-3.29]). Similar phenotypes and outcomes were identified in EXSCEL. CONCLUSIONS: In patients with T2DM and ASCVD, cluster analysis identified four clinically distinct groups. Further cardiovascular phenotyping is warranted to inform patient care and optimize clinical trial designs.
Authors: Matthew A Cavender; Ph Gabriel Steg; Sidney C Smith; Kim Eagle; E Magnus Ohman; Shinya Goto; Julia Kuder; Kyungah Im; Peter W F Wilson; Deepak L Bhatt Journal: Circulation Date: 2015-07-07 Impact factor: 29.690
Authors: Abhinav Sharma; Lauren B Cooper; Mona Fiuzat; Robert J Mentz; João Pedro Ferreira; Javed Butler; David Fitchett; Alan Charles Moses; Christopher O'Connor; Faiez Zannad Journal: JACC Heart Fail Date: 2018-08-08 Impact factor: 12.035
Authors: Jasper Tromp; Adriaan A Voors; Abhinav Sharma; João P Ferreira; Wouter Ouwerkerk; Hans L Hillege; Karla A Gomez; Kenneth Dickstein; Stefan D Anker; Marco Metra; Chim C Lang; Leong L Ng; Pim van der Harst; Dirk J van Veldhuisen; Peter van der Meer; Carolyn S P Lam; Faiez Zannad; Iziah E Sama Journal: JACC Heart Fail Date: 2020-02-05 Impact factor: 12.035
Authors: Marat Fudim; Jennifer White; Neha J Pagidipati; Yuliya Lokhnygina; Julio Wainstein; Jan Murin; Nayyar Iqbal; Peter Öhman; Renato D Lopes; Barry Reicher; Rury R Holman; Adrian F Hernandez; Robert J Mentz Journal: Circulation Date: 2019-09-23 Impact factor: 29.690
Authors: David P Kao; James D Lewsey; Inder S Anand; Barry M Massie; Michael R Zile; Peter E Carson; Robert S McKelvie; Michel Komajda; John J V McMurray; JoAnn Lindenfeld Journal: Eur J Heart Fail Date: 2015-08-06 Impact factor: 15.534
Authors: Jennifer B Green; M Angelyn Bethel; Sanjoy K Paul; Arne Ring; Keith D Kaufman; Deborah R Shapiro; Robert M Califf; Rury R Holman Journal: Am Heart J Date: 2013-10-23 Impact factor: 4.749
Authors: Subhashish Agarwal; David R Jacobs; Dhananjay Vaidya; Christopher T Sibley; Neal W Jorgensen; Jerome I Rotter; Yii-Der Ida Chen; Yongmei Liu; Jeanette S Andrews; Stephen Kritchevsky; Bret Goodpaster; Alka Kanaya; Anne B Newman; Eleanor M Simonsick; David M Herrington Journal: Cardiol Res Pract Date: 2012-03-21 Impact factor: 1.866
Authors: Steven P Marso; Gilbert H Daniels; Kirstine Brown-Frandsen; Peter Kristensen; Johannes F E Mann; Michael A Nauck; Steven E Nissen; Stuart Pocock; Neil R Poulter; Lasse S Ravn; William M Steinberg; Mette Stockner; Bernard Zinman; Richard M Bergenstal; John B Buse Journal: N Engl J Med Date: 2016-06-13 Impact factor: 176.079
Authors: Ana María Sánchez de la Nava; Lidia Gómez-Cid; Gonzalo Ricardo Ríos-Muñoz; María Eugenia Fernández-Santos; Ana I Fernández; Ángel Arenal; Ricardo Sanz-Ruiz; Lilian Grigorian-Shamagian; Felipe Atienza; Francisco Fernández-Avilés Journal: BioTech (Basel) Date: 2022-06-30