Risk Scores for Stroke Risk and Bleeding in Atrial Fibrillation: - PowerPoint PPT Presentation

Risk Scores for Stroke Risk and Bleeding in Atrial Fibrillation: Accuracy and Utility in Trials and the Real World Christopher B. Granger

Disclosures  Research contracts: AstraZeneca, Bayer, BMS, Boehringer Ingelheim, Daiichi Sankyo, FDA, Janssen, Novartis, GSK, Medtronic Foundation, Pfizer, The Medicines Company, FDA, NIH  Consulting/Honoraria: AstraZeneca, Bayer, BMS, Boston Scientific, GSK, Pfizer, Lilly, Daiichi Sankyo, Novartis, Boehringer Ingelheim, Medtronic, Medtronic Foundation, The Medicines Company  For full listing see w ww.dcri.duke.edu/research/coi.jsp

Afib risk scores Success: they are simple and probably helpful to decide who we need not anticoagulate Failure: they are not very good at discriminating risk

Atrial Fibrillation Guidelines Risk Recommended Therapy ESC AHA/ACC/HRS 2014 2016 No risk factors No antithrombotic No antithrombotic therapy (III B) therapy (IIa) CHA 2 DS 2 -VASc= 0 OAC (IIa B) None or OAC CHA 2 DS 2 -VASc= 1 (NOAC > VKA) or ASA (IIb) OAC (I) OAC (I) CHA 2 DS 2 -VASc ≥ 2 (NOAC > VKA (IA)) (NOAC or VKA) Mechanical valve, VKA mitral stenosis ESC Guidelines. Eur Heart J 2016 AHA/ACC/HRS Guidelines. Circulation 2014

CHA 2 DS 2 -VASc Assessment of Thromboembolic Risk Annual stroke Score CHF/ LV dysfunction 1 rate, % Hypertension 1 n 1084 73 538 Age  75 2 0 0 0.78 Diabetes mellitus 1 1 1.3 2.01 Stroke/TIA/TE 2 2 2.2 3.71 3 3.2 5.92 Vascular disease 1 4 4.0 9.27 Age 65-74 1 5 6.7 15.26 Sex category (female) 1 6 9.8 19.78 Score 0 – 9 7 9.6 21.50 8 6.7 22.38 Validated in 1084 NVAF patients not on OAC with known TE status at 1 year in Euro Heart Survey 9 15.2 23.64 Olesen JB et al. Lip GYH, et al. BMJ 2011;342:124 Chest 2009

Danish Hospital Registry Data 1997-2006 14,572 patients CHADS-VASc 0 or 1 Olesen J. BMJ 2011;342:d124

Established Clinical Risk Factors (CHADS-VASc) Novel Clinical Echo Risk Factors Parameters Prior stroke/TIA Chronic kidney LA volume Age disease LA and LAA Hypertension Obstructive sleep Function apnea Diabetes AF burden Heart failure Female sex Vascular disease Serum Advanced Biomarkers Imaging Natriuretic peptides LA fibrosis LAA morphology Troponin Calenda B et al. Nat Rev Cardiol. 2016 Sep;13(9):549-59

Are bleeding scores helpful?

Important Changes Kirchhof P Eur Heart J 2016

Can we do better with use of biomarkers and improved models for risk assessment?

Biomarkers and Risk in AF By Quartiles of NT-proBNP and CHADS-VASc Hijazi Z. J Am Coll Cardiol 2013;61:2274-2284 Hijazi Z. Eur Heart J 2016;37:1582-1590

ABC (Age, Biomarker, Clinical factor) risk scores ABC-stroke score Based on 391 stroke or SE during 27,929 person yrs of follow-up from the ARISTOTLE trial ( ᵡ 2 -df) Hijazi Z et al. Eur Heart J 2016

ABC-risk within CHA 2 DS 2 -VASc scores Event rates by the three ABC-stroke risk classes (low, medium, and high) for the CHA 2 DS 2 -VASc score (panel): 0-1 points, 2 points, and ≥3 points. Hijazi Z, Lindbäck J, Alexander JH, et al. Eur Heart J 2016

ABC-bleeding score (age, biomarkers [GDF-15, cTnT-hs (or creat clearance), and hemoglobin], and clinical history [previous bleeding]) score yielded a higher c-index than HAS-BLED and ORBIT scores for major bleeding in both the derivation (0·68 vs 0·61 vs 0·65) and validation (0·71 vs 0·62 vs 0·68) cohorts Hijazi et al. Lancet 2016387: 2302 – 11

ESC Guideline 2016

Concluding thoughts  Determining who is at unacceptably high risk of bleeding is important in deciding who should receive LAA occlusion devices  Our current risk assessment tools for bleeding are not very good  Clinical factors like recurrent prior bleeding, non- treatable causes, age, hemoglobin, GDF-15 are probably important to incorporate