Rule Out Myocardial Ischemia Infarction Using Computer Assisted Tomography

Rule Out Myocardial Ischemia/Infarction Using Computer Assisted Tomography A Multicenter, Randomized, Diagnostic Efficiency Trial ACRIN CV Committee October 2010 Udo Hoffmann, MD

Time Study Design Patients with Acute Chest Pain at Low to Intermediate Risk for ACS* Screening Consent & Randomization Standard of Care Discharge Admission Cardiac CT** Admission Discharge Patient Management: Diagnostic Testing , Interventions, Diagnosis, Discharge 48 -72 hour phone call Intervention Triage Decision Index Hospitalization 48 -72 hour phone call 28 -day Phone Interview Follow-Up 1 year Phone Interview 2 *low-intermediate risk for ACS • 1000 adults (≥ 40 -75 yrs) without known CAD • present with ACP (>5 min) to the ED at 7 sites • w/o ischemic ECG changes • further risk stratification required **Coronary morphology +/- LV function

Study Endpoints Length of Hospital Stay Primary Endpoint Rates of Direct ED Discharge Time to Diagnosis No. of invasive coronary angiograms and revascularizations Rates of MACE after ED discharge **, 28 days, and one year Secondary Endpoints Health care utilization after 28 days and one year Cost and Cost-Effectiveness Institutional and Caregiver Characteristics associated with primary and secondary outcomes Incremental Value of CTA over a CAC scan Incremental Value of LV function over a CTA Radiation Exposure during index hospitalization and follow-up 3 Tertiary Endpoints

Study Team Data Safety and Monitoring Board (DSMB) CLINICAL COORDINATING CENTER (CCC) Udo Hoffmann, MD MPH James Udelson, MD DATA COORDINATING AND STATISTICAL CENTER (DCSC) David Schoenfeld, Ph. D Clinical Sites Principal Investigator Beth Israel Deaconess Medical Center, Boston, MA (Thomas Hauser) Baystate Medical Center, Springfield, MA (J. Hector Pope) Center for Cost-Effectiveness and Decision Analysis (DACE) Kaiser Foundation Hospital – Fontana, CA (Eric Chou) Clinical Events Committee (CEC) Washington University, St. Louis, MI (Pamela Woodard) Steering Committee Tufts Medical Center, Boston, MA (Scott Weiner) Scott Gazelle, MD MPH Ph. D Stephen D. Wiviott, MD Jerome Fleg, MD (NIH – PO) Ruth Kirby (NIH) Quynh Truong, MD MPH External Advisory Committee Eugene Braunwald, MD - Chair 4 University of Maryland Medical Center, Baltimore, MD (Charles White) Massachusetts General Hospital, Boston, MA (J. Toby Nagurney)

Simulation of Primary Endpoint LOS Distribution of CT results and association with clinical outcomes within the study population – observed from ROMICAT I and predicted LOS 5

LOS - Power Evaluation - Standard of Care – observed from ROMICAT-I - CTA – predicted Sub Group Proportion LOS Mean ± SD Normal LV function without CAD 48. 6% 6 ± 6. 12 hours Normal LV function with non-obstructive CAD 28. 8%, 10 ± 8. 16 hours All other conditions 22. 6% 40 ± 8. 16 hours Powers to Detect Estimated Differences in Mean LOS depending on accuracy of assumptions ED LOS Accuracy Rate Assumption Estimated Mean (± SD) LOS Estimated Difference in Means (hours) CTA N=500 50% 40. 5 (± 43. 2) 37. 6 (± 50. 0) -2. 9 17% 60% 40. 5 (± 43. 2) 33. 7 (± 47. 4) -6. 8 66% 65% 40. 5 (± 43. 2) 31. 8 (± 46. 8) -8. 7 86% 70% 40. 5 (± 43. 2) 29. 8 (± 45. 5) -10. 7 97% 80% 40. 5 (± 43. 2) 25. 8 (± 42. 0) -14. 7 >99% 90% 6 Standard of Care N=500 40. 5 (± 43. 2) 22. 0 (± 38. 1) -18. 5 >99% 100% 40. 5 (± 43. 2) 18. 4 (± 33. 8) -22. 1 >99% Power

DSMB Recommendations • Approve length of stay (LOS) as the primary endpoint • Do not recommend the use of risk factors or risk scores as inclusion criterion • Physician-based assessment of “patient needs further risk stratification” as an inclusion criteria • Guidelines for patient management • Both prospective and retrospective CT imaging (lower dose) • Over-read CT for incidental findings and feedback to clinical sites 7

• Overview Data Collection • Update Enrollment/Patient population • Update Secondary Data Collection 8

Data Collection • Randomization - web-based RS 2 system • Data Capture - electronic medical record managed by web-based Research Electronic Data Capture (REDCap) database system 1. Main Record: 25 forms with 1271 fields 2. Screening Record: 1 form with 35 fields 3. CEC Adjudication: 5 forms with 108 fields 4. CT Core Lab Over read: 2 forms with 26 fields • Data Monitoring - automated weekly report including enrollment, screening, completeness, and accuracy – presented and reviewed by the Steering Committee 9

Overall Enrollment 10 Start of Enrollment: April 23 rd 2010

Overall Enrollment - Milestones Year 1 st. Quarter 2 nd Quarter 3 rd Quarter 4 th Quarter Total 2010 - 120 176 472 2011 176 184 168 - 528 Total 176 304 344 11 176 1000

Enrollment and Screening by Site and Week Average Weekly Enrollment Screened & Enrolled Total # of Subjects % Enrolled Site # Enrolled # Weeks Since 1 st Enrollment Baystate 32 22 1. 5 438 7. 3 BIDMC 15 15 1. 4 238 6. 3 Kaiser 54 22 2. 5 200 27. 0 MGH 37 22 1. 7 435 8. 5 Tufts 19 21 0. 9 219 8. 7 UMM 30 19 1. 6 457 6. 6 Wash U 32 22 1. 5 423 7. 6 Overall 219 22 10. 0 2410 9. 1 12

Study Population Demographics Ethnicity Age Male Female Total 40 s 53 24. 3% 29 13. 3% 82 37. 6% 50 s 51 23. 4% 37 17. 0% 18 8. 3% 21 9. 6% 39 17. 9% 70 s 4 1. 8% 5 2. 3% 9 4. 1% Total 126 59. 5% 92 42. 2% 218 100. 0% Percent NOT Hispanic or Latino 193 88. 1 Hispanic or Latino 16 7. 3 Unknown/not specified 10 4. 6 88 40. 4% 60 s Frequency Race Frequency Percent White 145 66. 2 Black or African Native 68 31. 1 Asian 4 2. 4 American Indian or Alaskan Native 1* 0. 5 Not reported 2 0. 9 Native Hawaiian or Pacific Islander 0 0 * This AI/AN subject is also counted as White 13

Activities for Enrollment • PI/SC visit all sites – Grand Rounds • Weekly PI/CRC calls • Website/Newsletter • Monitoring Visits (MGH, Baystate) 14

Secondary Aims - Data Collection • Cost Data • CT Reader Certification • CT Core Lab Over Reads • Discharge Diagnosis • Blood Biomarker Study 15

Cost Data Collection • Six of seven sites agreed on providing cost data • Pilots are initiated at these sites • Initial outpatient costs range from $1, 100 – $3, 300 per patient 16

CT Reader Certification • Why? To ensure uniformity and high quality of CT readers across the 7 clinical sites • How? Five instructor led cases, followed by 50 individual test cases with correlation of coronary CTA with invasive coronary angiography • Individual feedback provided after all readers certified 17

CTA Testing Software 18

Coronary CTA and Invasive Coronary Angio 19

CT Reader Certification Initial Results CT readers (n) % correct (range) % overcall % undercall Overall 24 76% (62 -84%) 14% 12% By Site MGH 3 79% (76 -80%) 13% 10% Wash U 5 78% (72 -84%) 11% Tufts 3 76% (68 -80%) 14% 10% Baystate 4 78% (64 -84%) 13% 10% UMM 3 75% (74 -76%) 14% 19% Kaiser 3 70% (62 -76%) 15% BIDMC 3 72% (64 -76%) 14% 20

Biomarker Study Methods: Hs –troponin at 0, 2, and 4 hours collection at sites, local storage, central measurement Primary Hypothesis: Hs-troponin followed by cardiac CTA will be more cost effective as compared to competing strategies in ED patients. Secondary Hypotheses: 1) Hs-troponin can accurately predict the presence of ACS much earlier than standard troponin. 2) Hs-troponin in combination with cardiac CT will predict one year MACE better than either strategy alone and better than standard troponin in combination with cardiac CT. 3) Elevated levels of hs-troponin will be associated with abnormal diagnostic test findings in both arms including presence and extent of CAD (CT), impaired regional LV function (CT or echo); myocardial perfusion defects (CT or SPECT), and ECG changes. Patient consented ? Site BIDMC MGH UMM Wash U Total No 3 25. 00 5 15. 63 14 37. 84 30 100. 00 16 51. 61 68 Yes 21 Baystate 9 75. 00 27 84. 38 23 62. 16 0 0. 00 15 48. 39 74 12 32 37 30 31 142 Total

ROMICAT II – Updated Timeline 10/09 start of NIH funding 09/09 -04/10 pre-enrollment period 04/10 -12/11 enrollment period 01/12 -06/12 follow-up and final database 06/12 -03/13 data analysis 01/13 -09/13 cost and cost effectiveness 22

Next Steps/Timeline • Add sites (October/November 2010) • Supplement for one year follow-up • Common CT database with other large studies (PROMISE, RESCUE, ISCHEMIA) • 23 DSMB meeting – review of mid enrollment period - 04/2011 3 nd

Thank you!!