Revascularization First !! Jeffrey W. Moses, MD Professor of Medicine Columbia University Medical Center

Disclosure Statement of Financial Interest I, Jeffrey Moses, have a financial interest/arrangement or affiliation with one or more organizations that could be perceived as a real or apparent conflict of interest in the context of the subject of this presentation. Consultant: BSC (minor)

CASE 1 CCS Class II Stress test: Exercise ECG: 6:50 Minutes of Bruce Protocol, 2mm ST Depression, Stopped Because of Chest Pain Duke Treadmill Score: -11 Echo: EF: 55-60%, No Significant Valvular Disease Clinical Presentation New York, NY Patient Demographics Age: 60 Gender: Male Family History ↑ Lipids Hypertension RiskFactors Nephrolithiasis Past Medical History Treatment: Med Rx, No Further Investigation

CASE 2 81 y/o man with colon ca s/p chemo and resection 15 y/a, now presents with 3 months progressive L shoulder pain with exertion ASA 81 mg daily, doxazosin 1 mg daily Nuclear Stress test: small, mild, reversible inferolateral perfusion defect, LVEF 60%

Symptoms and QOL

Months COURAGE: Freedom from Angina Weintraub et al, N Engl J Med 2008;359: PCI + OMT OMT Angina-free (%) P=0.35 P<0.001 P=0.005 P=0.010P= % Crossover at a Median of 10.8 months

Top Tercile (weekly) COURAGE: Angina Frequency PCI plus OMT OMT P Value First Baseline month < months < months months months months Weintraub et al, N Engl J Med 2008;359:

PCI plus OMT OMT P Value Second Baseline month months < months months months months Middle Tercile (monthly) COURAGE: Angina Frequency Weintraub et al, N Engl J Med 2008;359:

COURAGE: QOL PCI + OMT OMT 0 Months from Baseline Mean Score * * * * Weintraub et al, N Engl J Med 2008;359:

Safety Endpoint at 4.6 Years % of PatientsPCIOMTPCIOMTPCIOMT DeathSpontaneous MIRevascularization At mean 10 mos At mean 10.8 mos 40%  P< %  P=NS 11%  P=NS NEJM 2007;356: ; AHJ 2006;151:1173-9

BARI 2D: Prompt Revascularization Reduces Ischemic Symptoms Over 3 Years Analysis of the BARI 2D trial Conclusion: In diabetic patients with stable ischemic heart disease, a strategy of early revascularization and optimal medical therapy achieves better angina outcomes than does optimal medical therapy alone. Dagenais GR, et al. Circulation. 2011;Epub ahead of print. 3-year Follow-Up Prompt Revascularization (n = 1,173) Medical Therapy (n = 1,191)P Value Worsening Angina8%13%<0.001 New Angina37%51%0.001 Subsequent Revascularization 18%33%0.001

BARI 2D: QOL by Revasc. Status DASISelf-rated Health Rev vs Med: Delta = 1.34, p = Delta = 1.70, p = DASI Self-rated Health Courtesy of Frederick Feit, MD TCT: September 23, 2010 Med Rev BaselineYear 1Year 2Year 3Year BaselineYear 1Year 2Year 3Year

OAT Costs of OMT Mark et al NEJM 2009;360:774 PCIMED Hospital $22,800$12,700 Year 1 $3400$5300 Year 2 $1500$2700 BARI 2D: Lifetime costs PCI vs OMT $200 less Dagenais GR, et al. Circulation. 2011; Epub ahead of print.

PCI plus OMT OMT P Value Third Baseline month months months months months months Bottom Tercile (none) Weintraub et al, N Engl J Med 2008;359: COURAGE: Meds Can Make You Feels Worse! Angina Frequency

>5 Years 0.85 (0.67, 1.07) Pursnani et al, Circ Cardiovasc Interv. 2012;5: Meta-analysis of PCI for Chronic CAD Freedom From Angina Study or SubgroupEvents OMT M-H, Random, 95% CI ACME-1 (3 years) (0.93, 1.63) PCI Total Events TotalWeight Risk Ratio 4390 Risk Ratio 9.6% ACME-2 (5 years) (0.94, 2.31) % ALKK (5 years) (1.09, 1.48) % AVERT (1.5 years) (1.04, 1.65) % BARI 2D (5 years) (0.95, 1.12) % COURAGE (5 years) (0.92, 1.21) % DEFER (5 years) % MASS-1 (5 years) MASS-2 (5 years) (1.72, 4.24) % (1.06, 1.55) % TOTAL (95% CI) % <1 Year Favors OMT Favors PCI (1.06, 1.37) ACME-1 (3 years) (1.06, 1.79) % ACME-2 (5 years) (0.94, 2.31) % ALKK (5 years) (1.00, 1.20) % (1.43, 1.93) % BARI 2D (5 years) (1.05, 1.22) % COURAGE (5 years) (1.14, 1.79) % MASS-1 (5 years) 1995 Subtotal (95% CI) %1.32 (1.13, 1.54) Total events ACME-1 (3 years) (0.93, 1.63) % AVERT (1.5 years) (1.04, 1.65) % MASS-1 (5 years) (1.77, 3.60) % Subtotal (95% CI) %1.57 (1.06, 2.32) Total events 1-5 Years ALKK (5 years) (1.09, 1.48) % BARI 2D (5 years) (0.95, 1.12) % COURAGE (5 years) (0.92, 1.21) % (0.67, 1.07) % DEFER (5 years) (1.72, 4.24) % MASS-1 (5 years) (1.06, 1.55) % MASS-2 (5 years) 2004 Subtotal (95% CI) %1.17 (1.00, 1.38) Total events

Does PCI reduce Death and MI? Probably YES! (in the right pt)

MASS II 10-year Follow-up MED 203 CABG 205 PCI 203 CD/MI/CVA48%30%33% D/MI/CVA58.5%43.8% Angina Free 43%64%59% Hueb et al, Circ 2010 August 58% 3VD

Cumulative incidence (%) Registry PCI+MT MT No. at risk FAME 2: Primary Outcomes MT vs. Registry: HR 4.32 ( ); p<0.001 PCI+MT vs. Registry:HR 1.29 ( ); p=0.61 PCI+MT vs. MT: HR 0.32 ( ); p<0.001 Months after randomization De Bruyne B et al. NEJM 2012:on-line

Cumulative incidence (%) 0 7days Months after randomization p-interaction = <. 003 > 8 days: HR 0.42 ( ); p=0.053 ≤ 7 days: HR 7.99 ( ); p=0.038 MT alone PCI plus MT MT alone PCI plus MT ≤7 days >8 days FAME 2: Kaplan-Meier Plots of Landmark Analysis of Death or MI Cumulative incidence (%) Days after randomization

JSAP:PCI In Stable Angina 365 Patients Randomized PCI vs OMT Kazuhiko N et al, J. AM. Coll. Cardiol. Intv. 2008;1; Years 0.9 Hazard ratio, 0.541; 95% CI ( ) P= Years0.9 Hazard ratio, 0.664; 95% CI ( ) P= PCI plus medical therapy Initial medical therapy only PCI plus medical therapy Initial medical therapy only Initial-Medical* PCI+Medical* Initial-Medical* PCI+Medical* Death+ACS+CVADeath+ACS+CVA+hospitalization

Rates of Spontaneous MI: 12 RCTs 37,000 Patient-years Test for interaction P=0.53 Bangalore et al. Circ 2013;on-line.11 Favors PCI 10 Trial PCI MT EventIRR (95% CI)% Weight ACME-1 ACME-2 ALKK-1 AVERT DEFER MASS-1 RITA-2 SWISS Favors Medical Therapy N No Stents IRR (95% CI)EventN D+L Subtotal (I-squared = 46.3%, P=0.071) I-V Subtotal Random Effects Poisson Regression (0.13, 2.04) 0.98 (0.28, 3.39) 0.76 (0.32, 1.80) 0.92 (0.23, 3.70) 5.06 (0.24, ) 1.00 (0.20, 4.95) 1.11 (0.63, 1.95) 0.25 (0.12, 0.51) 0.72 (0.43, 1.22) 0.71 (0.50, 1.00) (0.58, 0.99) I-V Overall D+L Overall (I-squared = 31.6%, P=0.138) 0.77 (0.60, 0.99) 0.82 (0.69, 0.97) Stents D+L Subtotal (I-squared = 0.0%, P=0.556) I-V Subtotal BARI 2D COURAGE JSAP MASS (0.65, 1.33) 0.90 (0.69, 1.17) 0.43 (0.11, 1.66) 0.67 (0.39, 1.17) 0.86 (0.71, 1.05)

Cardiac Mortality Test for interaction P=0.03 Bangalore et al. Circ 2013;on-line.11 Favors PCI 10 Trial PCI MT Event% Weight ALKK-1 AVERT DEFER MASS-1 RITA-2 SWISS Favors Medical Therapy N No Stents IRR (95% CI)EventN D+L Subtotal (I-squared = 36.6%, P=0.162) I-V Subtotal Random Effects Poisson Regression I-V Overall D+L Overall (I-squared = 49.7%, P=0.036) Stents D+L Subtotal (I-squared = 0.0%, P=0.625) I-V Subtotal BARI 2D COURAGE JSAP MASS (0.10, 0.88) 0.92 (0.06, 14.79) 0.67 (0.11, 4.03) 2.00 (0.37, 10.92) 0.60 (0.30, 1.20) 0.15 (0.04, 0.50) 0.47 (0.24, 0.93) 0.48 (0.30, 0.77) (0.44, 1.09) 0.74 (0.49, 1.11) 0.86 (0.67, 1.11) (0.86, 2.12) 0.91 (0.52, 1.61) 0.67 (0.11, 3.99) 0.95 (0.54, 1.66) 1.08 (0.80, 1.45) IRR (95% CI)

Rate of MI / Rate of Crossover Bangalore et al. Circ 2013;on-line Crossovers (%) Log Incident Rate Ratio (Death) MT Better PCI Better

Odds Ratio (95% Confidence Interval) Overall Trial Sievers et al. Dakik et al. ACIP ACME-1 TIME ALKK AVERT Bech et al. MASS ACME-2 RITA-2 Year of Publication /3675 PCI 0/44 1/21 2/192 16/115 45/153 6/149 1/177 2/90 6/72 9/51 43/ /3838 Medical 1/44 1/23 20/366 15/112 40/148 17/151 1/164 4/91 6/72 10/50 43/514 Deaths/Total SWISSI II DANAMI COURAGE INSPIRE Hambrecht et al. MASS II /96 19/503 85/1149 2/104 28/205 22/105 24/505 95/1138 1/101 35/203 0/500/ Random effects model Fixed effects model P heterogeneity =0.263; I 2 =17% 17 RCTs: PCI vs. Medical Treatment Primary End Point: All-Cause Death Kastrati et al; (0.64 to 0.99) 0.80 (0.68 to 0.95)

PCI vs. Med Rx-21Trials: Mortality Jeremias et al, Am Jour Med 2009;122: ACME TOPS Sievers et al MASS I (PCI) ACIP RITA ACME DANAMI Dakik et al Horie et al AVERT TOAT Bech et al TIME ALKK MASS Il (PCI) DECOPI OAT INSPIRE SWISS II COURAGE StudyOR LowerLimitUpperLimit Odds Ratio and 95% Combined 0.82 Favors Revasc.Favors Med

Safety Endpoint at 4.6 Years % of PatientsPCIOMTPCIOMTPCIOMT DeathSpontaneous MIRevascularization At mean 10 mos At mean 10.8 mos 40%  P< %  P=NS 11%  P=NS NEJM 2007;356: ; AHJ 2006;151:1173-9

Cardiac Mortality in Medically Treated Patients According to Ischemic Risk – CSMC database Hachamovitch et al Circulation. 2003;107: % Total Myocardial Ischemia 0%1- 5%5-10%11-20%>20% Cardiac Death Rate (%) (1.9 yr FU) N=7110N=1331N=718N=545N=252 N=9,956 pts 5.4% cardiac mortality in 1.9 years - Is this “stable” angina?

MPS % Ischemic Myocardium (95% CI) Pre-Rx & 6-18 Months Pre-Rx6-18m 8.2% 5.5% (4.7%-6.3%) PCI + OMT (n=159) OMT (n=155) Pre-Rx6-18m (6.9%-9.4%) 8.6% 8.1% Mean = -2.7% (95% CI = -3.8% to -1.7%) Mean = -2.7% (95% CI = -3.8% to -1.7%) Mean = -0.5% (95% CI = -1.6% to 0.6%) Mean = -0.5% (95% CI = -1.6% to 0.6%) p<0.0001

RCTs with Ischemia OMT vs PCI Identified 4 randomized controlled trials (RCT) Identified 4 randomized controlled trials (RCT)  ACME  COURAGE AHJ (baseline ischemia substudy)  FAME-2  SWISSI-II Comprised a total of 1769 patients Comprised a total of 1769 patients  871 randomized to PCI  898 randomized to MT alone Length of follow-up ranged from ~7 mo – 10 yrs Length of follow-up ranged from ~7 mo – 10 yrs  Weighted average of 4.5 years

All-Cause Mortality: HR Analysis PCI Better Study ID ACME 1 All-cause Mortality in Randomized Trials of PCI vs. MT (Patients with Ischemia or Equivalent) ES (95% CI) % Weight (I-V).110 MT Better 0.73 (0.24, 2.22)20.42 COURAGE AHJ 0.62 (0.30, 1.28)48.00 FAME (0.03, 3.17) 4.65 SWISSI-II 0.42 (0.16, 1.11) Fixed Effects (I 2 =0.0%, p=0.84) 0.56 (0.34, 0.93) Random Effects 0.56 (0.34, 0.93)

Summary of Results: Mortality Significant 44% reduction in all-cause mortality was observed with PCI vs. MT Significant 44% reduction in all-cause mortality was observed with PCI vs. MT  HR 0.56; 95% CI [ ], p=0.02 Point estimate of the HR for mortality favoring PCI vs. MT varied from ; I 2 =0% Point estimate of the HR for mortality favoring PCI vs. MT varied from ; I 2 =0%  Baseline ischemia cohort of the COURAGE trial comprised 48% weight of the studies Analysis Using Count Data: Analysis Using Count Data:  28/871 (3.2%) deaths with PCI  54/898 (6.0%) deaths with MT  RR 0.56; 95% CI [ ], p=0.01

PCI Better Study ID ACME 1 ES (95% CI) % Weight (I-V).110 MT Better 0.73 (0.24, 2.22)11.13 COURAGE AHJ 0.62 (0.30, 1.28)26.16 FAME (0.03, 3.17) 2.54 SWISSI-II 0.42 (0.16, 1.11) Fixed Effects (I 2 =0.0%, p=0.76) 0.61 (0.42, 0.89) Random Effects 0.61 (0.42, 0.89) ACIP* 0.24 (0.05, 1.11)5.73 DANAMI Symptomatic* 0.68 (0.30, 1.55)20.42 DANAMI Silent* 0.93 (0.40, 2.16)19.36 All-Cause Mortality: HR Analysis All-cause Mortality in Randomized Trials of PCI vs. MT* (Patients with Ischemia or Equivalent) *Secondary analysis includes studies with small no. of CABG pts

ACC Appropriateness Categories Underuse and Adverse Outcomes Ko et al, JACC 2012; in press CABG Proportion of Cardiac Catherization (%) HR: 0.99 HR: 0.57 (p=0.12) n=311n=326n=991 PCI HR: 0.61 (p=0.009) Medical 1625 pts with Chronic CAD and Cath: 3 year risk : Death /ACS

Hazards of Underutilization 9300 Patients with recent onset chest pains 57% appropriate patients did not get angio median follow-up: three years Hemingway et al, Annals of Int Med 2008;248:221 Angio +Angio – 11%22%Death or ACS HR : 2.5

Symptoms Med. Rx Class llI or lV Max Rx UAAAA Class I or lI Max Rx UUAAA Asympto- matic Max Rx IIUUU Class llI or lV No/min Rx IUAAA Class I or lI No/min Rx IIUUU Asympto- matic No/min Rx IIUUU Coronary Anatomy CTO of 1 vz. no other disease 1-2 vz. disease no prox. LAD 1 vz. disease of prox. LAD 2 vz. disease with prox. LAD 3 vz. disease no Left Main Low-Risk Findings on Non-invasive Study Patel et al JACC (February): Asymptomatic Stress Test Med. Rx High Risk Max Rx UAAAA High Risk No/min Rx UUAAA Int. Risk Max Rx UUUUA Int. Risk No/min Rx IIUUA Low Risk Max Rx IIUUU Low Risk No/min Rx IIUUU Coronary Anatomy CTO of 1 vz. no other disease 1-2 vz. disease no prox. LAD 1 vz. disease of prox. LAD 2 vz. disease with prox. LAD 3 vz. disease no Left Main Appropriateness Ratings by Low-Risk Findings on Noninvasive Imaging Study and Asymptomatic

If the goal was really best outcomes why aren’t physicians monitored for potential underuse?   Overuse may cost money   Underuse costs lives

We have hoodwinked the patients into denying themselves treatment   If you have a treatment that may help and doesn’t harm you take it   We have convinced the patient to do the opposite

Conclusions In cases of moderate symptoms PCI provides immediate and superior symptomatic relief vs “OMT” (i.e., crossovers are Rx failure) with no cost of MI, death or CABG In cases of moderate symptoms PCI provides immediate and superior symptomatic relief vs “OMT” (i.e., crossovers are Rx failure) with no cost of MI, death or CABG In cases of demonstrated ischemia or critical anatomy the overwhelming evidence favors revascularization to reduce death and MI In cases of demonstrated ischemia or critical anatomy the overwhelming evidence favors revascularization to reduce death and MI “OMT” alone is safe for mild symptoms, little or no ischemia and noncritical anatomy …..Period “OMT” alone is safe for mild symptoms, little or no ischemia and noncritical anatomy …..Period

8000 Ischemia-Eligible Stable Patient (Stable CAD, Moderate-Severe Ischemia) Blinded Coronary CTA Eligible Anatomy? RANDOMIZE Invasive Strategy (Cath with Optimal Revasc + OMT) CT Exclusion Ancillary Study OMT Strategy (OMT Alone) YES NO ISCHEMIA Trial Design J. Hochman, TCT 2010 This is an Untested Hypothesis