1. Does Cyclosporine ImpRove Clinical oUtcome in ST-elevation myocardial infarction patients ? (the CIRCUS study) Michel OVIZE, MD, PhD Louis Pradel Hospital and Claude Bernard University, Lyon, France

2. Reperfusion injury contributes to MI Reperfusion injury Ca 2+ Transition pore REPERFUSION NECROSIS Ischemia / Reperfusion - ATP Pi Ca2+ ROS Cyclosporine Mitochondria Onset of chest pain CORONARY ARTERY OCCLUSION - TIMI flow grade =0-1 WINDOW TO START TREATMENT OF REPERFUSION INJURY 30 minutes to 12 hours First medical care Cath lab admission AMBULANCE PCI CARDIOLOGIST PCI Reperfusion Ischemia injury

3. Reperfusion injury is reduced by cyclosporine 0 1000 2000 3000 4000 5000 6000 CK release Adm. 4h 8h 12h 16h20h 24h 30h 36h42h48h 54h 60h 66h72h Control Cyclo (UI/L) Piot et al. NEJM 2008 Direct stenting Day 1-3 CK / TnI release Day 5 MRI Infarct size STEMI < 12 hrs PCI treatment TIMI flow grade 0-1 No visible collateral Cyclosporine (or saline) (2.5 mg/kg, IV bolus) 0 control cyclosporine area of hyperenhancement (g) 10 20 30 40 50 60 70 120 * CMR infarct size

4.  OBJECTIVE To determine whether cyclosporine improves clinical outcomes in STEMI patients  PRIMARY ENDPOINT Combined incidence within 1 year of : - all-cause mortality - worsening of heart failure during initial hospitalization or re-hospitalisation for heart failure - adverse LV remodeling* * (adverse LV remodeling (echo): increase > 15% of LVEDV at 1 year versus baseline) Objective & Primary Endpoint

5. LAD occluded (TIMI 0-1) Coronary angiography e-randomization CicloMulsion® (Neurovive Pharmaceuticals): lipid emulsion of cyclosporine (no cremophor content) Initial Echo Final Echo Discharge PCI CicloMulsion® : (2.5 mg/kg, IV bolus) 1 year Anterior STEMI Experimental design > 18 years symptom onset < 12 hrs ST shift ≥ 0.2 mV in two contiguous anterior leads LAD as culprit artery with TIMI flow grade : 0 – 1

6. Results Study coordination: Centre for Clinical Investigation (Hospices Civils de Lyon (HCL), France) Independent, blinded corelabs (HCL): Coronary angiography / Echocardiography / ECG Independent Statistical CRO (Itech, France) First patient included : 2011 April, 19 Last patient included : 2014 February, 16 Last visit last patient : 2015 April 2 nd Cumulative inclusions Target enrollment 42 investigation centres in 3 countries Recruitment rate

7. ITT Analysis Primary endpoint Analysis Anterior STEMIs Randomized (n= 970) Control (n=495) No informed content (n=1) Imprisoned and therefore ineligible (n=1) Did not receive any treatment (n=4) Missing or poor echographic data (n=74) Did not receive any treatment (n=4) Missing or poor echographic data (n=95) Cyclosporine (n = 475) Cyclosporine (n = 395) Control (n=396) modified as treated analysis Consort Diagram (80% power to detect a 20% relative risk reduction)

8. Baseline characteristics Cyclosporine (n=474) Control (n=495) Age (years) 60.4 ± 13.159.5 ± 12.7 Male sex84.2 %80.0 % Body Mass Index (kg/m2) 26.9 ± 4.326.8 ± 4.1 Killip class I at admission87.4 %87.2 % Current smoker 39.0 % *45.7 % Hypertension37.6 %37.0 % Diabetes mellitus13.7 %11.7 % Dyslipidemia39.2 %37.8 % Previous ischemic heart disease6.5 % Previous Heart Failure 0.2 %1.0 %

9. PCI procedure Procedure Cyclosporine (n=474) Control (n=495) Total ischemic time 4.4 ± 3.0 hrs4.5 ± 2.9 hrs Door-to-treatment time 1.0 ± 1.3 hrs1.1 ± 1.7 hrs Time from symptom onset to treatment : 2-6 hours70.8 %68.8 % Medication from 1 st medical care to PCI Glycoprotein IIb/IIIa inhibitors38.2 %37.5 % P2Y 12 inhibitors loading dose (clopidogrel, prasugrel, ticagrelor)90.3 %88.2 % LAD site of occlusion: Proximal / Main left45.1 %41.0 % Multi-vessel disease ≥ 2 40.9 % *33.1 % Area at risk size 37 ± 8 %36 ± 9 % Thrombus aspiration75.7 %76.2 % Stenting89.0 %87.7 % Final TIMI flow = 2 or 396.0 %

10. Discharge medication Cyclosporine (n=474) Control (n=495) Antiplatelet agents96.6 %97.2 % Beta-blockers90.4 %93.8 % Statins95.5 %95.9 % ACE inhibitors or ARBs86.4 %88.0 % Calcium Antagonists 2.3 % 2.9 % Nitrates13.0 %13.2 % Diuretic21.5 %22.0 % Soludactone / Eplerenone24.9 %23.3 %

11. Primary outcome at 1 year Combined incidence of [all-cause mortality; worsening of heart failure during initial hospitalization or re-hospitalisation for heart failure ; LV remodeling] within 1 year after acute MI (LV remodeling (echo): increase > 15% of LVEDV at 1 year versus initial discharge) Cyclosporine (n=395) Control (n=396) Odds Ratio (95% CI) P value (Death / HF / LV remodeling)233 (59.0 %)230 (58.1%)1.04 [0.78; 1.39]0.77

12. Secondary outcomes at 1 year Cyclosporine (n=395) Control (n=396) Odds Ratio (95% CI) P value Death: all-cause 7.1 % 6.6 % 1.09 [0.63 ; 1.90] 0.76 Death: cardiovascular 6.1 % 1.01 [0.56 ; 1.81] 0.98 HF worsening or re-hospitalization for HF22.8 %22.7 % 1.01 [0.72 ; 1.41] 0.97 HF worsening15.7 %16.9 % 0.92 [0.63 ; 1.34] 0.65 Re-hospitalization for HF10.6 %10.4 % 1.03 [0.65 ; 1.63] 0.89 Cardiogenic shock 6.6 % 6.1 % 1.09 [0.61 ; 1.94] 0.77 Recurrent Myocardial infarction 2.3 % 3.8 % 0.59 [0.26 ; 1.37] 0.22 Stroke 1.8 % 3.0 % 0.58 [0.22 ; 1.48] 0.25 Major bleeding 1.8 % 2.3 % 0.73 [0.27 ; 2.00] 0.54 HF: heart failure

13. Prespecified subgroup analysis Cyclosporine Better Control Better

14. LV volumes and function

15. Estimation of infarct size Plasma total CK (IU/L) Cyclosporine (n=474) Control (n=495) Baseline 531 [103; 445]626 [114; 541] H4 4183 [1992; 5744]3920 [1780; 5691] H12 3191 [1855; 4182]3220 [1840; 4224] H24 1765 [1004; 2276]1837 [1019; 2368] Peak 3992 [1910; 5447]3917 [1878; 5608]

16. Discussion Limitations: LV remodeling contributed to a large proportion of event rate in the primary endpoint But, there was no evidence of effect of cyclosporine on any individual clinical endpoints Discrepancy with our previous phase II trial: -Different PCI procedures or conditions (stenting, thrombus aspiration, P 2 Y 12 inhibitors, anterior infarcts) -Different formulation of cyclosporine (Cremophor EL vers lipid emulsion) -Phase II versus phase III trial (type I error)

17. Conclusion In anterior STEMI, cyclosporine did not reduce the risk of the composite primary outcome -One out of four patients died or experienced heart failure despite receiving state- of-the-art medical care. -Despite the results of CIRCUS, reperfusion injury is clinically important. The impact on clinical outcomes of recent encouraging phase II trials remains to be determined.

18. Acknowledgements Committees Steering Com: D.Garcia-Dorado, M.Claeys, F.Pinto, PG Steg, G.Derumeaux, C.Piot, L.Belle, P.Croisille, M.Ovize Event-validation Com: N.Mewton, JF.Aupetit, I.Sanchez, G.Rioufol, H.Thibault Data Monitoring & Safety: JL.Bonnet, A.Pathak, R.Porcher Corelabs Coronary angiography (G.Rioufol) Echocardiography (C.Bergerot, M.Altman, H.Thibault, G.Derumeaux) ECG (S.Pichot, M.Schaaf, N.Mewton) Lyon Center for Clinical Investigation (CIC) Jossan C., Boussaha I, Mewton N. Investigators Thien-Tri Cung, Olivier Morel, Guillaume Cayla, Gilles Rioufol, David Garcia Dorado Denis Angoulvant, Eric Bonnefoy-Cudraz, Patrice Guérin, Meier Elbaz, Nicolas Delarche, Pierre Coste, Gerald Vanzetto, Marc Metge, Jean-François Aupetit, Bernard Jouve, Pascal Motreff, Christophe Tron, Jean-Noel Labeque, Philippe Gabriel Steg, Yves Cottin, Grégoire Range, Jérome Clerc, Marc Jérome Claeys, Patrick Coussement, Fabrice Prunier, Frédéric Moulin, Olivier Roth, Loïc Belle, Philippe Dubois, Paul Barragan,Martine Gilard, Christophe Piot, Patrice Colin, Fabien De Poli, Marie-Claude Morice, Omar Ider, Jean-Luc Dubois-Randé, Thierry Unterseeh, Hervé Le Breton, Thierry Béard, Didier Blanchard, Gilles Grollier, Vincent Malquarti, Patrick Staat, Arnaud Sudre, Eskil Elmer, Magnus J. Hansson, Cyrille Bergerot, Michel Ovize.

19. Publication available online at nejm.org (http://www.nejm.org/)