Gregg W. Stone MD For the AMIHOT II Investigators

1. A Prospective, Randomized Evaluation of Supersaturated Oxygen Therapy After Percutaneous Coronary Intervention in Acute Anterior Myocardial Infarction Gregg W. Stone MD For the AMIHOT II Investigators

2. Disclosures • Gregg W. Stone MD • Research support from TherOx Inc.

3. Background • Despite successful reperfusion in AMI, myocardial recovery is often suboptimal, resulting in extensive infarction. • In experimental infarct models, hyperbaric oxygen reduces myocardial tissue damage, in part by reducing reperfusion injury and improving microcirculatory perfusion. • Regional hyperoxemia in the infarct zone can be achieved by infusion of supersaturated blood into the infarct artery after successful primary PCI. • This concept was tested in the AMIHOT I trial.

4. The AMIHOT I Trial 269 pts with anterior or large inferior AMI and TIMI 0-2 flow undergoing primary or rescue PCI within 24 hours from symptom onset were randomized after successful PCI to intracoronary supersaturated oxygen therapy (SSO2; PO2 760-1000 mmHg) for 90’ versus control. Therox, Inc.

5. AMIHOT I Results 3 Co-Primary Efficacy Endpoints O’Neill WW et al. JACC 2007;50:397-405.

6. AMIHOT II Trial Design Anterior AMI* with TIMI 0-2 flow reperfused by PCI with stenting within 6 hrs TIMI 2-3 flow achieved Randomize** Standard therapy SSO2 for 90 mins 2 Primary Endpoints Efficacy: Infarct size (superiority) (tc=99m sestamibi SPECT @14 days) Safety: 30 day MACE (noninferiority) *STE ≥1 mm in ≥2 contiguous leads V1-V4 or LBBB with LAD infarct **Stratified by time to reperfusion (<3 vs. 3-6 hrs) and prox vs. non prox lesion

7. Endpoints and Statistical Methodology • Objective 1 - Efficacy: To demonstrate that compared to control, SSO2 results in reduced infarct size as measured by tc-99m-sestamibi SPECT imaging at 14 (±7) days in pts with anterior MI reperfused within 6 hours • Objective 2 - Safety: To demonstrate that compared to control, SSO2 has noninferior rates of major adverse cardiac events (MACE – death, reinfarction, TVR or stroke) at 30 days • Bayesian hierarchical modeling: To allow pooling of data from AMIHOT I, with the amount of pooling determined by the similarity of the AMIHOT II results to the AMIHOT I data, while still preserving type I error to <5% (as per FDA “Draft Guidance for the Use of Bayesian Statistics in Medical Device Clinical Trials”)* *http://www.fda.gov/cdrh/osb/guidance/1601.pdf

8. Patient Enrollment 304 patients randomized at 20 sites in 4 countries (US, Canada, Netherlands, Italy) between September 13, 2005 and May 26, 2007 3 randomization errors 301 ITT patients Randomize 2.8:1 SSO2 N=222 Control N=79 SPECT endpoint N=175 (78.8%) N=69 (87.3%) 30 day FU complete N=222 (100%) N=79 (100%)

9. Primary Efficacy EndpointInfarct Size by Tc-99m-sestamibi SPECT Pooled, adjusted N=382 Difference of medians -6.5% PWilcoxon=0.023  Bayesian Posterior Probability = 98.0%* Infarct size, %LV Control N=124 SSO2 N=258 Median [IQR] 25 [7, 42] Median [IQR] 18.5 [3.5, 34.5] *Imputed; 95.6% using only non imputed data

10. Immeasurable Infarcts P = 0.11 RR [95%CI] = 1.76 (1.04, 3.00) P = 0.03 P = 0.20 Proportion with “0% LV” infarcts (%)

11. Primary Safety Endpoint: 30 Day MACE  Bayesian Posterior ProbNI = 99.8% 1 = using all pts from AMIHOT I 2 = using only anterior MI reperfused <6 from AMIHOT I

12. Conclusion Among high risk patients with acute anterior MI undergoing successful PCI within 6 hours of symptom onset, infusion of SSO2 into the myocardial infarct territory results in a significant reduction in infarct size