Presenter Disclosure Information Robert E. Michler, M.D. Influence of Left Ventricular Volume Reduction on Outcome After Coronary Artery Bypass Grafting.

Presenter Disclosure InformationPresenter Disclosure Information

Robert E. Michler, M.D.Robert E. Michler, M.D.

Influence of Left Ventricular VolumeInfluence of Left Ventricular Volume Reduction on Outcome After Coronary Reduction on Outcome After Coronary Artery Bypass Grafting With or Without Artery Bypass Grafting With or Without Surgical Ventricular ReconstructionSurgical Ventricular Reconstruction

Financial Disclosure: NIH STICH GrantsFinancial Disclosure: NIH STICH Grants NHLBI CT Surgery NetworkNHLBI CT Surgery Network

Unlabeled/Unapproved Uses Disclosure: NoUnlabeled/Unapproved Uses Disclosure: No

Influence of Left Ventricular Volume Reduction on Outcome After Coronary Artery Bypass Grafting With or Without

Surgical Ventricular Reconstruction

Robert E. Michler, Gerald M. Pohost, Krzysztof Wrobel, Robert O. Bonow, Jan Pirk, Jae K. Oh,

Carmelo A. Milano, Patricia A. Pellikka, Francois Dagenais, Thomas A. Holly,

Anne S. Hellkamp, Kerry L. Lee, Marisa Di Donato, on behalf of the STICH Investigators

Late-Breaking Clinical Trial UpdateAmerican College of Cardiology

March 16, 2010

IntroductionIntroduction

Ischemic cardiomyopathy resulting from progressive LV volume increase due to CAD and anterior-apical myocardial scar compromises clinical outcome.

The objective of surgical ventricular reconstruction (SVR) is to create a smaller left ventricle with a more natural shape.

One STICH (Surgical Treatment for Ischemic Heart Failure) Trial specific aim was to determine if adding SVR to CABG provided patient benefit beyond that of CABG alone.

SVR Hypothesis QuestionSVR Hypothesis Question

Does addingDoes adding SVR to CABG in medically-treated SVR to CABG in medically-treated ischemic heart failure patients decrease death/cardiac ischemic heart failure patients decrease death/cardiac rehospitalization?rehospitalization?

1000 patients randomized 2002-20061000 patients randomized 2002-2006CAD, EF ≤ 35% CAD, EF ≤ 35%

Anterior akinesia/dyskinesia Anterior akinesia/dyskinesia amenable to SVRamenable to SVR

499499CABG onlyCABG only

501501CABG + SVRCABG + SVR

Follow-up 99% complete at 48-months

Randomized

PurposePurpose

To determine whether any magnitude of postoperative change in LV ESVI identified a subgroup of CABG + SVR patients who have increased survival when compared to patients undergoing CABG alone.

Analysis DesignAnalysis Design

Cohort identified with Cohort identified with pairedpaired core laboratory core laboratory studies of fair to excellent quality permitting studies of fair to excellent quality permitting accurate assessment of end-systolic volume index accurate assessment of end-systolic volume index (ESVI).(ESVI).

Individual preoperative and postoperative ESVI Individual preoperative and postoperative ESVI illustrated for patient groups with:illustrated for patient groups with:

1. ESVI <60 mL/m1. ESVI <60 mL/m22

2. ESVI 60-90 mL/m2. ESVI 60-90 mL/m22

3. ESVI >90 mL/m3. ESVI >90 mL/m22

Kaplan-Meier curves and hazard ratios calculated Kaplan-Meier curves and hazard ratios calculated on cohorts to examine for a differential effect of on cohorts to examine for a differential effect of adding SVR to CABG.adding SVR to CABG.

Paired Left Ventricular Studies Before and After Paired Left Ventricular Studies Before and After Operation in 979 SVR Hypothesis PatientsOperation in 979 SVR Hypothesis Patients

Operated PatientsN = 979

86 died before 4 months

Patients eligible for paired study analysis

N = 893

13 Specific reasons provided for missing study 3 Unexplained

Patients with baseline studies sent to core lab

N = 877

64 Telephone follow-up only22 Technical problem21 Unexplained

Patients with paired baseline and 4-month

studies, N = 770

15 Pts with CMR assessment 14 Pts with SPECT assessment146 Pts with ECHO assessment

Patients analyzed in this report, N = 595

16 NO baseline study

sent to core lab

107 NO 4-month study

sent to core lab

both paired studies offair to excellent quality

175 patients without

86 (9%) of operated patients

Observational cohort defined by ESVI taken from a Randomized population

Cumulative Distribution of 595 SVR Hypothesis Patients Cumulative Distribution of 595 SVR Hypothesis Patients by Baseline ESVIby Baseline ESVI

Preoperative to Postoperative Change in ESVI by Operation ReceivedPreoperative to Postoperative Change in ESVI by Operation Received160 Patients with Baseline ESVI < 60 mL/m160 Patients with Baseline ESVI < 60 mL/m22

GROUP 1GROUP 1

Preoperative to Postoperative Change in ESVI by Operation ReceivedPreoperative to Postoperative Change in ESVI by Operation Received200 Patients with Baseline ESVI 60–90 mL/m200 Patients with Baseline ESVI 60–90 mL/m22

GROUP 2GROUP 2

Preoperative to Postoperative Change in ESVI by Operation ReceivedPreoperative to Postoperative Change in ESVI by Operation Received235 Patients with Baseline ESVI > 90 mL/m235 Patients with Baseline ESVI > 90 mL/m22

GROUP 3GROUP 3

ESVI Change

Lowest to Highest ESVI

Group 1<60 ml/m2

(N=160)

Group 260-90 ml/m2

(N=200)

Group 3>90 ml/m2

(N=235)

Operation CABG(N=91)

CABG+SVR

(N=69)CABG

(N=111)

CABG+SVR

(N=89)CABG

(N=118)

CABG+SVR

(N=117)

% No Reduction 55% 43% 48% 20% 32% 21%

% >30% Reduction 23% 28% 20% 38% 24% 44%

Preop LVEFMedian

0.37 0.28 0.22 0.36 0.30 0.21

Postop LVEFMedian

0.38 0.31 0.25 0.40 0.35 0.27

Hazard Ratios and 95% Confidence IntervalsHazard Ratios and 95% Confidence IntervalsAll-Cause MortalityAll-Cause Mortality

All Patients (n=1000)(as randomized)

All Patients (n=979)(by operation received)

Patients with Pre & Post Surgery Studies

(n=595)

Patients Excluded (n=384)

CABG+SVR Better CABG Only Better

Hazard Ratios and 95% Confidence IntervalsHazard Ratios and 95% Confidence IntervalsAll-Cause MortalityAll-Cause Mortality

Patients with Pre & Post Surgery Studies (n=595)

Baseline ESVI > 90 ml/m2 (n=235)

Baseline ESVI 60- 90 ml/m2 (n=200)

Baseline ESVI < 60 ml/m2 (n=160)

CABG+SVR Better CABG Only Better

Kaplan-Meier Estimates: Cumulative Risk of DeathKaplan-Meier Estimates: Cumulative Risk of Death Patients with Pre & Post Surgery Studies : Baseline ESVI Patients with Pre & Post Surgery Studies : Baseline ESVI ≤ 90 mL/m≤ 90 mL/m2 2

(n=360)(n=360)

HR=0.59 (95% CI: 0.35 – 1.00; rank test: p=0.0475)

Total events (63): 20 in CABG+SVR and 43 in CABG Only

Kaplan-Meier Estimates: Cumulative Risk of DeathKaplan-Meier Estimates: Cumulative Risk of Death Patients with Pre & Post Surgery Studies : Baseline ESVI Patients with Pre & Post Surgery Studies : Baseline ESVI > 90 mL/m> 90 mL/m2 2

(n=235)(n=235)

HR=1.24 (95% CI: 0.75 – 2.06; rank test: p=0.4071)

Total events (60): 33 in CABG+SVR and 27 in CABG Only

Hazard Ratios and 95% Confidence IntervalsHazard Ratios and 95% Confidence IntervalsAll-Cause MortalityAll-Cause Mortality

Baseline ESVI ≤ 90 mL/m2 with Small or no Reduction

(n=180)

Baseline ESVI ≤ 90 mL/m2 with Large Reduction

(n=180)

Baseline ESVI > 90 mL/m2 with Small or no Reduction

(n=117)

Baseline ESVI > 90 mL/m2 with Large Reduction

(n=118)

CABG+SVR Better CABG Only Better

Kaplan-Meier Estimates: Cumulative Risk of DeathKaplan-Meier Estimates: Cumulative Risk of Death Patients with Pre & Post Surgery Studies Patients with Pre & Post Surgery Studies

Baseline ESVI Baseline ESVI > 90 mL/m> 90 mL/m22 and Small/or No Reduction in Post_Op ESVI and Small/or No Reduction in Post_Op ESVI (n=117)(n=117)

HR=1.54 (95% CI: 0.80 – 2.98; rank test: p=0.1950)

Total events (36): 17 in CABG+SVR and 19 in CABG Only

Small/or no reduction = change from baseline ESVI ≤ -23.7 mL/mmL/m22

Kaplan-Meier Estimates: Cumulative Risk of DeathKaplan-Meier Estimates: Cumulative Risk of Death Patients with Pre & Post Surgery Studies Patients with Pre & Post Surgery Studies

Baseline ESVI Baseline ESVI > 90 mL/m> 90 mL/m22 and Large Reduction in Post_Op ESVI and Large Reduction in Post_Op ESVI (n=118)(n=118)

HR=1.20 (95% CI: 0.51 – 2.80; rank test: p=0.6777)

Total events (23): 16 in CABG+SVR and 8 in CABG Only

Large reduction = change from baseline ESVI > -23.7 mL/mmL/m22

Limitations of StudyLimitations of Study

Baseline LV volume and regional function data Baseline LV volume and regional function data were not available in every STICH patient.were not available in every STICH patient.

Secondary structural and hemodynamic variables Secondary structural and hemodynamic variables related to LV function, such as sphericity index or related to LV function, such as sphericity index or mitral regurgitation, were not considered in this mitral regurgitation, were not considered in this analysis.analysis.

Bias of investigators towards not sending Bias of investigators towards not sending suboptimal postoperative studies cannot be suboptimal postoperative studies cannot be excluded.excluded.

ConclusionsConclusions

1. A broad range of baseline ESVI is represented among STICH patients.

2. The postoperative ESVI decrease is significantly larger for CABG+SVR patients.

3. In patients with larger volumes, ESVI > 90 ml/m2, CABG alone resulted in a substantial reduction in ESVI.

4. Patients with preoperative ESVI values ≤ 90 ml/m2 trended toward benefit from CABG + SVR, whereas patients with preoperative ESVI values > 90 ml/m2 trended toward benefit from CABG alone.

5. No threshold of ESVI at baseline, ESVI at 4 months postoperative or ESVI volume change identified a patient group that benefited from adding SVR to CABG.

6. The post-op ESVI is the most important prognostic mortality measure. Its prognostic importance is significant even after accounting for the baseline ESVI or the pre-to-post change in ESVI.

Discussion Slides

4

1

2 6

3 510

15

17

13

7

8 12

119

14 16

4

1

2 6

3 510

15

17

13

7

8 12

119

14 16

4

1

2 6

3 510

15

17

13

7

8 12

119

14 16

4

1

2 6

3 510

15

17

13

7

8 12

119

14 16

PostoperativePostoperativePreoperativePreoperative

Group 1: LVESVI < 60 mL/mGroup 1: LVESVI < 60 mL/m22 (N = 129) (N = 129)

CABGCABGN = 70N = 70

CABG + SVRCABG + SVRN = 59N = 59

Change in Regional Cardiac Function in 504 SVR Hypothesis PtsChange in Regional Cardiac Function in 504 SVR Hypothesis Pts

Dyskinesia

Normal

Group 2: LVESVI 60–90 mL/mGroup 2: LVESVI 60–90 mL/m22 (N = 170) (N = 170)

CABGCABGN = 96N = 96

CABG + SVRCABG + SVRN = 74N = 74

Change in Regional Cardiac Function in 504 SVR Hypothesis PtsChange in Regional Cardiac Function in 504 SVR Hypothesis Pts

4

1

2 6

3 510

15

17

13

7

8 12

119

14 16

4

1

2 6

3 510

15

17

13

7

8 12

119

14 16

4

1

2 6

3 510

15

17

13

7

8 12

119

14 16

4

1

2 6

3 510

15

17

13

7

8 12

119

14 16

PostoperativePostoperativePreoperativePreoperative

Dyskinesia

Normal

Group 3: LVESVI > 90 mL/mGroup 3: LVESVI > 90 mL/m22 (N = 205) (N = 205)

CABGCABGN = 105N = 105

CABG + SVRCABG + SVRN = 100N = 100

Change in Regional Cardiac Function in 504 SVR Hypothesis PtsChange in Regional Cardiac Function in 504 SVR Hypothesis Pts

4

1

2 6

3 510

15

17

13

7

8 12

119

14 16

4

1

2 6

3 510

15

17

13

7

8 12

119

14 16

4

1

2 6

3 510

15

17

13

7

8 12

119

14 16

4

1

2 6

3 510

15

17

13

7

8 12

119

14 16

PostoperativePostoperativePreoperativePreoperative

Dyskinesia

Normal

Hazard Ratios and 95% Confidence IntervalsHazard Ratios and 95% Confidence IntervalsAll-Cause MortalityAll-Cause Mortality

Patients with Pre & Post Surgery Studies

(n=595)

Post-Op ESVI > 90 mL/m2 (n=150)

Post-Op ESVI 60- 90 mL/m2 (n=198)

Post-Op ESVI < 60 mL/m2 (n=247)

CABG+SVR Better CABG Only Better