Importance of Bleeding in Patients with Cardiovascular Disease

Roxana Mehran, MD, FACC, FAHA, FESC, FSCAI Professor of Medicine Director, Interventional Cardiovascular Research and Clinical Trials Mount Sinai School of Medicine

Disclosure Statement of Financial Interest

Within the past 12 months, I or my spouse/partner have had a financial interest/arrangement or affiliation with the organization(s) listed below.

•

Affiliation/Financial Relationship Grant/Research Support

•

Consulting Fees/Honoraria

• •

Company

Sanofi/BMS- Significant Astra Zeneca, Cardiva, Cordis, The Medicines Company, Regado Biosciences

Impact of Therapies on Outcomes

Ischemic events: MI/CKMB↑ Stent Thrombosis

Bleeding

Bleeding and Mortality

Major Bleeding Hypotension Ischemia Cessation of ASA/Clop Stent Thrombosis Transfusion Inflammation Mortality Bhatt DL. In Braunwald EB, Harrison’s Online. 2005.

Impact of In-hospital Bleeding in ACS 34,146 Pts with ACS in the OASIS-1/2 and CURE Major bleeding occurred in 2.3% of pts First 30 days 14 12 12.8%

P<0.0001

10 8 6

Bleeding

4 2.5% 2

No bleeding

0 0 5 10 15 Days 20 25 30 No. at Risk No bleeding 33376 33419 Bleeding 470 459 33157 440 32990 32879 32769 430 420 410 32710 408 6 4 2 8 Landmark analysis, 1-6 mo 0 0 30

P=0.002

Bleeding No bleeding

4.6% 2.9% 60 90 Days 120 150 180 32634 560 32491 32161 31981 31166 30316 29238 559 554 548 533 519 489 Adj. HR [95%CI] = 5.37 [3.97, 7.26]

P<0.0001

Adj. HR [95%CI] = 1.54 [1.02, 2.36]

P=0.047

Eikelboom JW. Circulation 2006;114:774 –782

Impact of Major Bleed and MI after Elective and Urgent PCI

1-Year Mortality (N=6,012) With major bleed 8.8% With MI 5.7% Without major bleed 2.0% Without MI 1.9% Time from Randomization in Days Stone GW. J Inv Cardiol 2004;16(suppl G):12 –17.

Impact of 30 Day Adverse Events on 1-Year Mortality after PCI ISAR REACT-1, -2, -SWEET, -SMART-2 (n=5,384) Cox model relating 30 day events to mortality at 1 year Variable Bleeding w/i 30 days MI w/i 30 days Urgent revasc w/i 30 days Age (per 10 yrs) Diabetes MVD ↑ Troponin LVEF Creatinine (per 0.25 mg/dl ↑) HR (95% CI) 2.96 (1.96

–4.48) 2.29 (1.52

–3.46) 2.49 (1.16

–5.35) 2.27 (1.78

–2.89) 1.47 (1.11

–1.96) 2.72 (1.58

–4.67) 1.77 (1.27

–2.47) 0.71 (0.60

–0.85) 1.10 (1.06

–1.14) P 0.001

0.001

0.02

0.001

0.008

0.001

0.001

0.001

0.001

Ndrepepa et al. JACC 2008;51:690 –7

ACUITY:

Influence of Major Bleeding and MI in the First 30 Days on Risk of Death

Over 1 Year

Of 13,819 enrolled pts, 524 (3.8%) died within 1 year Cox model adjusted for 36 baseline predictors, with MI and major bleeding (non-CABG) as time-updated covariates Myocardial infarction HR

±

95% CI Major bleeding without or before transfusion Major bleeding after transfusion HR (95% CI) P-value 2.51 (1.95-3.25) <0.0001

2.00 (1.30-3.06) <0.0001

3.93 (2.95-5.24) <0.0001

Mehran RM et al. EHJ 2009;30:1457-66

ACUITY:

Influence of Major Bleeding and MI in the First 30 Days on Risk of Death

Over 1 Year

Of 13,819 enrolled pts, 524 (3.8%) died within 1 year Cox model adjusted for 36 baseline predictors, with MI and major bleeding (non-CABG) as time-updated covariates HR

±

95% CI HR (95% CI) P-value Attributable deaths Myocardial infarction 2.51 (1.95-3.25) <0.0001

51.5* Major bleeding without or before transfusion Major bleeding after transfusion 2.00 (1.30-3.06) <0.0001

3.93 (2.95-5.24) <0.0001

66.5** *9.8% of all deaths **12.7% of all deaths Attributable deaths = N deaths among pts with the time updated event (attribute) X (adj. HR – 1)/adj. HR Mehran RM et al. EHJ 2009;30:1457-66

Influence of MI, Major Bleed and Transfusion in the First 30 Days on the Risk of Death Over 1 Year MI Day 0-1 Days 2-7 Days 8-30 Days 31+ HR (95% CI) 17.6 (10.8 to 28.7) 8.2 (5.0 to 13.6) 2.9 (1.6 to 5.3) 1.4 (0.9 to 2.1) P-value <0.001

<0.001

0.001

0.12

Attributable deaths 21 19 12 25 Major bleed (non CABG) Day 0-1 Days 2-7 Days 8-30 Days 31+ 5.5 (2.7 to 11.0) 5.8 (3.5 to 9.7) 5.6 (3.5 to 8.8) 2.4 (1.7 to 3.3) Transfusion Day 0-1 Days 2-7 Days 8-30 Days 31+ 6.7 (3.1 to 14.7) 8.1 (4.6 to 14.1) 6.4 (3.7 to 10.9) 3.1 (2.1 to 4.5) Attributable deaths = N deaths among pts with the time updated event (attribute) X (adj. HR – 1)/adj. HR 0.5 1 2 4 8 16 32 Hazard ratio (95% CI) Mehran RM et al. EHJ 2009;30:1457-66 <0.001

<0.001

<0.001

<0.001

<0.001

<0.001

<0.001

<0.001

9 18 24 42 7 15 17 31

ACUITY (N=13,819) Impact of MI and Major Bleeding in the First 30 Days on Risk of Death Over 1 Year

25 20 15

Patients with Major Bleed (N=645) Patients with MI (N=705) Patients w/o Major Bleed (N=13,168) Patients w/o MI (N=13,108) 1 year Estimate 14.9% 11.4% 3.6% 3.8% 14.9% 11.4%

10 5 0 0 30 60 90 120 150 180 210 240 270 300 330 360 390

Days from Randomization 3.8% 3.6%

ACUITY

Costs of In-hospital Complications

Pinto DS et al. JACC 2008;52;1758-1768

H armonizing O utcomes with R evascular iz ati on and S tents in AMI 3602 pts with STEMI with symptom onset ≤12 hours Aspirin, thienopyridine R 1:1 UFH + GP IIb/IIIa inhibitor (abciximab or eptifibatide) Bivalirudin monotherapy (

±

provisional GP IIb/IIIa) Emergent angiography, followed by triage to… CABG – Primary PCI – Medical Rx 3006 pts eligible for stent randomization R 3:1 Paclitaxel-eluting TAXUS stent Bare metal EXPRESS stent Clinical FU at 30 days, 6 months, 1 year, and then yearly through 5 years; angio FU at 13 months

HORIZONS: Time-updated covariate adjusted Cox model relating MACE events to 1-year mortality - Complete model with MACE components and major bleeding Risk Factor HR (95% CI) HR [95% CI] P-value Attributable Deaths Reinfarction Incidence 138 (3.8%) 16 deaths after event 3.94

[1.73, 8.96] 0.001

12* [7, 14] Major bleeding (non CABG) Incidence 268 (7.4%) 44 deaths after event 3.39

[2.29, 5.03] <0.0001

31** [25, 35] 0.1

1 10 Hazard Ratio [95% CI] Attributable deaths = N deaths among pts with the time updated event (attribute) X (adj. HR – 1)/adj. HR *8.2% of 147 total deaths **21.1% of 147 total deaths

HORIZONS-AMI:

Influence of Non-CABG Major Bleed and MI on the Risk of Death Over 1 Year MI Day 0-2 Day 3-7 Day 8-30 Day >30 HR [95% CI] 20.3 [8.2,50.7] Attributable Deaths (%) P-value 5 <0.001

21.9 [7.8,61.0] 4 <0.001

4.8 [1.2,19.8] 1.3 [-0.2, 9.3] 2 1 0.03

0.80

Major Bleed (non-CABG) Day 0-2 Day 3-7 Day 8-30 Day >30 0.1

1.0

10.0

Hazard Ratio [95% CI] Attributable deaths = N deaths among pts with the time updated event (attribute) X (adj. HR – 1)/adj. HR 6.7 [3.0,15.1] 19.9 [9.3,42.4] 7.4 [3.7,14.9] 5.5 [3.0,10.2] 100.0

7 10 11 14 <0.001

<0.001

<0.001

<0.001

HORIZONS: 30 Day Adverse Events

2 0 6 4 12 10 8 Heparin + GPIIb/IIIa inhibitor (N=1802) Bivalirudin monotherapy (N=1800) P<0.001

8.3

P = 0.90

1.8

1.8

4.9

Reinfarction Major bleeding* *Not related to CABG ** Plat cnt <100,000 cells/mm 3 Stone GW et al. NEJM 2008;358:2218-30

HORIZONS: 1-Year All-Cause Mortality

5 4 Bivalirudin alone (n=1800) Heparin + GPIIb/IIIa (n=1802) 4.8% Δ = 1.4% 3.4% 3 3.1% 2 1 0 0 1 2.1% Δ = 1.0% P=0.049

HR [95%CI] = 0.69 [0.50, 0.97] P=0.029

2 3 4 5 6 7 Time in Months 8 9 10 11 12 Number at risk Bivalirudin alone Heparin+GPIIb/IIIa 1800 1802 1705 1678 1684 1663 1669 1646 1520 1486 Mehran R et al. Lancet 2009:on-line

30 Day Stent Thrombosis

(N=3,124 successfully stented pts) UFH + GP IIb/IIIa (N=1553) Bivalirudin (N=1571) ARC 30d definite or probable stent thrombosis* 1.9% 2.5% P Value 0.30

- definite 1.4% 2.2% 0.09

- probable 0.5% 0.3% acute (≤24 hrs) 0.3% 1.3% - subacute (>24 hrs – 30d) *Protocol definition of stent thrombosis, CEC adjudicated 1.7% Stone GW et al. NEJM 2008;358:2218-30 1.2% 0.24

0.0007

0.28

Time-updated covariate adjusted Cox model relating 30-day events to 30-day mortality

- Complete model in 3,124 pts with successfully implanted stents HR [95% CI] P-value Attributable Deaths Risk Factor Stent thrombosis (definite) Incidence 57 (1.8%) 5 deaths with event Major bleeding (non CABG) Incidence 195 (6.2%) 18 deaths with event 10.62

[3.96, 28.48] 6.22

[3.33, 11.60] <0.001

<0.001

4.5* [3.7, 4.8] 15.1** [12.6, 16.4]

0.01

0.1

1 10

Hazard Ratio [95% CI]

100

*8.3% of 54 total deaths **28.0% of 54 total deaths C-statistic = 0.87. Attributable deaths = N deaths among pts with the time updated event (attribute) X (adj. HR – 1)/adj. HR

Publication of Primary Results

NEJM 357: 2001-2015, 2007 www.NEJM.org

Wiviott SD, et al. N Engl J Med 2007;357:2001–15

TRITON TIMI 38: Study Design

ACS (STEMI or UA/NSTEMI) & Planned PCI ASA Double-blind N= 13,600 CLOPIDOGREL 300 mg LD/ 75 mg MD PRASUGREL 60 mg LD/ 10 mg MD Median duration of therapy – 12 months 1 2 o o endpoint: endpoints: Safety endpoints: Key Substudies: CV death, MI, Stroke CV death, MI, Stroke, Rehosp-Rec Isch CV death, MI, UTVR Stent Thrombosis (ARC definite/prob.) TIMI major bleeds, Life-threatening bleeds Pharmacokinetic, Genomic Wiviott SD, et al. N Engl J Med 2007;357:2001–15

Primary Endpoint CV Death,MI,Stroke

15 10 5 HR 0.77

P=0.0001

HR 0.80

P=0.0003

Clopidogrel 12.1

(781) 9.9 (643) Prasugrel HR 0.81

(0.73-0.90)

P=0.0004

NNT= 46 0 0 30 60 90 ITT= 13,608 LTFU = 14 (0.1%) 180 Days 270 360 450 Wiviott SD, et al. N Engl J Med 2007;357:2001–15

1 2

Stent Thrombosis (ARC Definite + Probable)

3 Any Stent at Index PCI N= 12,844 Clopidogrel 0 0 30 60 90 2.4

(142) Prasugrel 1.1 (68) HR 0.48

P <0.0001

180 NNT= 77 Days 270 360 450 Wiviott SD, et al. N Engl J Med 2007;357:2001–15

15 10

Balance of Efficacy and Safety

CV Death / MI / Stroke 138 events Clopidogrel 12.1

9.9

HR 0.81

(0.73-0.90)

P=0.0004

NNT = 46 Prasugrel 5 NonCABG Bleeds 0 0 30 60 90 TIMI Major 180 Days 270 Prasugrel Clopidogrel 2.4

1.8

35 events HR 1.32

(1.03-1.68)

P=0.03

360 450 NNH = 167 Wiviott SD, et al. N Engl J Med 2007;357:2001–15

PLATO: K-M estimate of time to first primary efficacy event

(Composite of CV death, MI or stroke) Completeness of follow-up 99.97% = five patients lost to follow-up 6 5 4 3 2 1 0 13 12 11 10 9 8 7 Clopidogrel HR 0.84 (95% CI 0.77

Ticagrelor 11.7

9.8

–0.92), p=0.0003

No. at risk Ticagrelor Clopidogrel 0 9,333 9,291 60 8,628 8,521 120 180 240 Days after randomisation 8,460 8,362 8,219 8,124 6,743 6,743 300 5,161 5,096 360 4,147 4,047 K-M = Kaplan-Meier; HR = hazard ratio; CI = confidence interval

Wallentin et al. N Engl J Med. 2009 Sep 10;361(11):1045-57

PLATO: K-M estimates of time to primary safety event

(Major Bleeding) Completeness of follow-up 99.97% = five patients lost to follow-up 15 10 Ticagrelor Clopidogrel 11.58

11.20

5 HR 1.04 (95% CI 0.95

–1.13), p=0.434

0 No. at risk Ticagrelor Clopidogrel 0 9,235 9,186 60 7,246 7,305 120 180 240 Days from first IP dose 6,826 6,930 6,545 6,670 5,129 5,209 300 3,783 3,841 360 3,433 3,479

Wallentin et al. N Engl J Med. 2009 Sep 10;361(11):1045-57

Definitions of Major Bleeding in Clinical Trials: Main Components

       

Clinical Events Intracranial / intracerebral bleeding Intraocular bleeding Bleeding causing hemodynamic compromise Cardiac tamponade Retroperitoneal hematoma Hematoma Surgical intervention for bleeding Blood product transfusion Laboratory Parameters

   

Decrease in Hgb bleeding ≥3 g/dL with overt source of Decrease in Hgb bleeding ≥4 g/dL w/o overt source of Decrease in Hgb bleeding ≥5 g/dL with or w/o overt source of Decrease in Hct ≥15% with overt source of bleeding

Definitions of Major/Severe Bleeding in Randomized Controlled Clinical Trials Type of bleeding Intracranial/intracerebral Intraocular Retroperitoneal Bleeding causing hemodynamic compromise Cardiac tamponade Bleeding requiring surgical intervention Hematoma >5cm at the puncture site Transfusion, units Decrease in Hgb with overt bleeding, g/dL Decrease in Hgb without overt bleeding, g/dL GUSTO + TIMI phase I + + ≥1 + ≥1 ≥5.0* TIMI phase II REPLACE-2 OASIS-5 ESSENCE + + + + + + + CURE + + + + ≥1 ≥3.0

≥2 ≥3.0

≥4.0

≥2 ≥3.0

+ + ≥2 ≥5.0

STEEPLE ACUITY HORIZONS PLATO + + + + + + + + + + ≥1 ≥3.0

+ + ≥1 ≥3.0

≥4.0

+ + + ≥4 ≥5.0

-

*Or decrease in Hct ≥15%

TIMI Major TIMI Minor

Bleeding Definitions

Bleeding with >5 g/dL fall in hgb Intracranial bleeding Intraocular bleeding Access site bleed requiring intervention ≥ 5 cm hematoma at puncture site Reoperation for bleeding ACUITY and HORIZONS Major Bleeding Blood product transfusion Hgb  ≥3g/dL with an overt source Hgb  ≥4g/dL w/o overt source Retroperitoneal bleeding Gross hematuria or hematemesis

Rao AK et al. JACC 1988;11:1-11; Stone GW et al. NEJM 2006;355:2203-16

Hierarchical Incidence of Major Bleeding

†

Within 30 Days After PCI

REPLACE-2 (N=5894) ACUITY (N=7760) HORIZONS (N=3348) Total (n=17,002*) TIMI major bleed ACUITY major (non TIMI major) bleed with blood transfusion** ACUITY major (non TIMI major) bleed without blood transfusion** 73 (1.2%) 125 (1.6%) 95 (2.8%) 293 (1.7%) Large hematoma only 100 (1.7%) 82 (1.1%) 13 (0.4%) 195 (1.1%) Total 289 (4.9%) 462 (6.0%) 228 (6.8%) * Excluding patients with any bleed prior to the PCI ; ** Excluding hematomas if the only criteria † Not related to CABG. Each patient is represented only once according to their most severe bleed 979 (5.8%)

Influence of Bleeding Severity within 30 Days After PCI on the Risk of Death Over 1 Year Baseline covariate-adjusted time-updated Cox multivariable model Type of Bleed TIMI major bleed ACUITY major (non TIMI major) bleed with transfusion* ACUITY major (non TIMI major) bleed without transfusion* Hematoma ≥5 cm only HR (95% CI) 4.85 (3.56-6.60) P value <0.001

Attributable deaths within 1 yr 53 2.98 (2.10-4.24) <0.001

1.79 (1.09-2.93) 0.02

1.30 (0.58-2.92) 0.53

40 17 6 HR (95%CI)

Mehran, et al. JACC Int 2011- In-Press

* Excluding hematomas if the only criteria Each patient is represented only once according to their most severe bleed

How Does Access Site Impact Major Bleeding Rates in PCI Patients?

•

Meta-analysis of 18 randomized trials (5 had no bleeding events) of femoral versus radial access involving 4,458 patients undergoing angiography or PCI Major Bleeding

Radial access reduced major bleeding by 73%, with a trend for reductions in the composite of death, MI, or stroke (2.5% vs 3.8%, P = .058)

Jolly SS. Am Heart J 2009;157:132-40.

Non-CABG Major Bleeding in PCI-Treated ACS Patients

7 6 5 4 1 0 3 2

88% Femoral Access 84% Radial Access 5.9

ACUITY 30 Days 3.7

3.7

3.1

0.67

1.2

TRITON 3 Days EARLY ACS 120 hours SYNERGY 30 Days OASIS 5 9 Days ABOARD 30 Days

6 5 4 1 0 3 2

Sources and Incidence of Bleeding Among 17,393 PCI Patients

No Location Both 5.2% 1.5

0.5

0.7

2.5

Protocol Major Non-Access Site Only Access Site Only 5.3% (n=925) 1.6 (281) 0.8 (142) 0.8 (145) 3.3% Non- access site bleeds are 61.4% of TIMI bleeding events 1.6% 0.5

0.7

2.1

2.1% Access site only accounts for 38.6% TIMI Major TIMI Major + Minor Verheugt JACC Cardio Interv 2011;4:191-7:

Incidence and source of bleeding excluding access site

25 20 15 10 5 0 50 45 40 35 30 45.2

17.9

15 10.1

6.9

3.7

0.9

GU GI Head/Neck Pulmonary ICH Other No site Axis Title

Verheugt JACC Cardio Interv 2011;4:191-7:

Relative Risk of 1-year Mortality Associated

6.0

with Bleeding and Source (unadjusted)

P<0.0001 for all bleeding versus none

5.7

5.5

5.0

5.4

5.2

4.0

3.0

2.3

2.0

1.0

0.0

Access Only Both Non-Access Only No Location All Non-Access

Verheugt JACC Cardio Interv 2011;4:191-7:

Risk for 1 year mortality

● 1-year mortality risk from

non-access

HR 2.27 (95%CI 1.42-3.64), p=0.0007

site bleeding vs

access

site = Relative Risk P-Value

Unadjusted

Access site Non-access site 2.33 (1.53 – 3.53) <0.0001

5.40 (4.32 – 6.74) <0.0001

Hazard ratio

Adjusted

Access site Non-access site 1.82 (1.17

–2.83) 0.008

3.94 (3.07

–5.15) <0.0001

0 1 2 3 4 5 6 7

No Bleed TIMI Major + Minor Bleed Verheugt JACC Cardio Interv 2011;4:191-7:

Impact of Antithrombotic Therapy on Bleeding by Source

Relative Risk P-Value

TIMI Major + Minor Bleeding

Access Only Both Non Access Only No Location 0.45 (0.35-0.59) 0.31 (0.19-0.49) 0.70 (0.47-1.05) 0.75 (0.58-0.96) <0.0001

<0.0001

0.08

0.02

All non-access 0.62 (0.51-0.75) <0.0001

0 0.5

Bivalirudin better

1 1.5

2

Hep + GPI better Verheugt JACC Cardio Interv 2011;4:191-7:

Impact of Randomized Antithrombotic Therapy on TIMI Bleeding by Location

Intracranial GI GU NEENT Pulmonary Other No location bleed All Non-Access

0 0.5

2

Hep + GPI (%) 0.04

0.64

0.64

0.33

0.18

0.3

2.82

3.66

Bivalirudi n (%) 0.03

0.28

0.28

0.22

0.05

0.15

1.83

2.27

Relative Risk 0.66 (0.11-3.97) 0.44 (0.26-0.74) 0.44 (0.26-0.74) 0.66 (0.35-1.24) 0.31 (0.10-0.94) 0.49 (0.24-1.01) 0.65 (0.52-0.80) 0.62 (0.51-0.75)

1 1.5

Bivalirudin better Hep + GPI better Verheugt JACC Cardio Interv 2011;4:191-7:

Standardized Bleeding Definitions for Cardiovascular Clinical Trials: A Consensus Report from the Bleeding Academic Research Consortium (BARC)

•

Roxana Mehran, MD, Sunil V. Rao, MD, Deepak L. Bhatt, MD, MPH, C. Michael Gibson, MS, MD, Adriano Caixeta, MD, PhD, John Eikelboom, MD, MBBS, Sanjay Kaul, MD, Stephen D. Wiviott, MD, Venu Menon, MD, Eugenia Nikolsky, MD, PhD, Victor Serebruany, MD, PhD, Marco Valgimigli, MD, PhD, Pascal Vranckx, MD, David Taggart, MD, PhD, Joseph F. Sabik, MD, Donald E. Cutlip, MD, Mitchell W. Krucoff, MD, E. Magnus Ohman, MD, Philippe Gabriel Steg, MD, and Harvey White, MB ChB DSc

Circulation 2011 In-press

BARC

• •

Type 0: No evidence of bleeding.

Type 1: Bleeding that is not actionable and does not cause the patient to seek unscheduled performance of studies, hospitalization, or treatment by a health care professional. Examples include, but are not limited to, bruising, hematoma, nosebleeds, or hemorrhoidal bleeding for which the patient does not seek medical attention. Type I bleeding may include episodes that lead to discontinuation of medications by the patient because of bleeding without visiting a health care provider.

Circulation 2011 In-press

BARC

•

Type 2: Any clinically overt sign of hemorrhage (e.g., more bleeding than would be expected for a clinical circumstance; including bleeding found by imaging alone) that is actionable, but does not meet criteria for Type 3 BARC bleeding,

•

The bleeding must require diagnostic studies, hospitalization or treatment by a health care professional. In particular, the bleeding must meet at least one of the following criteria:

  

1) Requiring intervention: defined as a health care professional-guided medical treatment or percutaneous intervention to stop or treat bleeding, including temporarily or permanently discontinuing a medication or study drug.

2) Leading to hospitalization or an increased level of care: defined as leading to or prolonging hospitalization or transfer to a hospital unit capable of providing a higher level of care; or 3) Prompting evaluation: defined as leading to an unscheduled visit to a healthcare professional resulting in diagnostic testing (laboratory or imaging).

Circulation 2011 In-press

BARC

•

Type 3: Clinical, laboratory, and/or imaging evidence of bleeding with specific healthcare provider responses, as listed below:

• •

a. BARC Type 3a Bleeding

Any transfusion with overt bleeding; Overt bleeding plus hemoglobin drop ≥3 to <5 g/dL* (provided hemoglobin drop is related to bleeding)

• •

b. BARC Type 3b Bleeding

Overt bleeding plus hemoglobin drop ≥ 5 g/dL* (provided hemoglobin drop is related to bleed), Cardiac tamponade, Bleeding requiring surgical intervention for control (excluding dental/nasal/skin/hemorrhoid), Bleeding requiring intravenous vasoactive drugs

• •

c. BARC Type 3c Bleeding

Intracranial hemorrhage (does not include microbleeds or hemorrhagic transformation; does include intraspinal). Intra ocular bleed compromising vision

Circulation 2011 In-press

BARC Type 4: CABG-Related Bleeding.

• • • • •

Perioperative intracranial bleeding within 48 hours Reoperation following closure of sternotomy for the purpose of controlling bleeding Transfusion of ≥ 5 units of whole blood or packed red blood cells within a 48 hour period* Chest tube output ≥ 2L within a 24 hour period. Notes: If a CABG-related bleed is not adjudicated as at least a Type 3 severity event, it will be classified as ‘not a bleeding event.’ If a bleeding event occurs with a clear temporal relationship to CABG (i.e. within a 48 hour timeframe) but does not meet Type 4 severity criteria, it will be classified as ‘not a bleeding event’. * only allogenic transfusions are considered as transfusions for CABG-related bleeds

Circulation 2011 In-press

BARC Type 5: Fatal Bleeding.

• • •

Fatal bleeding is bleeding that directly causes death with no other explainable cause. BARC Fatal Bleeding is categorized as either definite or probable as follows: a) Probable fatal bleeding (Type 5a) is bleeding that is clinically suspicious as the cause of death, but the bleeding is not directly observed and there is no autopsy or confirmatory imaging. b) Definite fatal bleeding (Type 5b) is bleeding that is directly observed (either by clinical specimen – blood, emesis, stool, etc.- or by imaging) or confirmed on autopsy.

Circulation 2011 In-press

Possible Mechanisms Linking Hemorrhagic Complications to Mortality

1.

Fatal hemorrhage (e.g. intracranial bleed) 2.

Vol. depletion



Hypotension, ischemia, arrhythmias 3.

Complications from procedures to manage bleeding 4.

Discontinuation of lifesaving medications (antiplatelet agents, beta blockers, statins) 5.

Blood transfusions depleted in NO



systemic vasoconstriction, inflammation, apoptosis 6.

Unmeasured confounders

Impact of Transfusion in ACS

30-Day Survival by Transfusion Group

GUSTO IIb, PURSUIT, PARAGON B trials

N=24,111 Rao SV et. al. JAMA 2004;292:1555 –1562

PRBC Transfusion in NSTE ACS

Time-Updated Cox Model for 30-day Death N=24,111

Adjusted for transfusion propensity Adjusted for baseline characteristics Adjusted for baseline characteristics, bleeding propensity, transfusion propensity, and nadir HCT

-4.0

1.0 Rao SV et. al. JAMA 2004;292:1555 –1562 3.77

(3.13, 4.52) 3.54

(2.96, 4.23) 3.94

(3.26, 4.75) 10.0

Impact of the Age of PRBC Transfusion After Cardiac Surgery on Outcomes Cleveland Clinic, June 30, 1998 – January 30, 2006

2,872 pts transfused with 8,872 U of blood stored ≤14 days (mean 11d ; “newer blood”) and 3,130 pts transfused with 10,782 U stored 15 days – 42 days (mean 20d; “older blood”)

% P=0.004

P<0.001

P=0.003

P=0.01

P<0.001

Koch CG et al. NEJM 2008;358:1229-1239

Discharge Medication Use in Patients who Bleed: PREMIER Registry (STEMI)

1433 STEMI pts treated with primary stenting P=0.001

P=0.002

P<0.001

P=0.05

Wang TY et. al. Circulation 2008;118:2139-2145

Discharge Medication Use in Patients who Bleed: HORIZONS-AMI (STEMI)

3,345 STEMI pts in whom primary PCI was performed P=0.12

P=0.05

P<0.0001

P<0.0001

Balancing Safety and Efficacy

High risk of ischemic events “Sweet spot” High risk of bleeding events

Inhibition of platelet aggregation

Ischemic risk Bleeding risk Ferreiro & Angiolillo. Thromb Haemost 2010 (in press)

Challenges Facing American Medicine

•

Costs of Care

•

Disparities in Care

•

Safety of Care

•

Evidence-Based care

•

Personalization of Care

How do we use evidence to improve medical decision making?

Traditional EBM Approach

Good Outcome Intermediate Outcome Bad Outcome • Clinical trials and EBM provide answers for “average” patients • In real life, however, there are no average patients

National Cardiovascular Data Registry

Imaging Registry Timeline and growth…

ICD Long CathPCI Registry 1998…..

2004 ICD Registry 2005 CARE Registry ACTION Registry 2006 2007 IMPACT Registry PINNACLE 2008 AF Abl Registry PAD Registry Valve Registry beyond

National Data Repository for Comparative Effectiveness Research Pharm STS Registry NDI UPI NCDR CATHPCI CLAIMS

Risk-Treatment Paradox

Bleeding Avoidance Rx in the Cath Lab

• Highest risk patients get the lowest rate of intervention

50%

• Lowest risk patients get the

40% Medium Risk High Risk 30% 20% 10% 0% N=1,522,935 None Closure Bivalirudin Closure + Bival

Marso et al.

JAMA

2010; 303: 2156-2164

ePRISM system at MAHI

•

John Spertus, M.D., M.P.H.

•

David Cohen, MD

•

Adam Salisbury, M.D.

•

MAHI cardiologists, nurses, and administration

e

PRISM

:

Clinical Risk Modeling at the Point-of-Care

Projected Symptoms at 6 Months

0   1

x T

 1   

n x

1

x T

 1   

n x



i

 

i

 0  

i

 1

x T

 1   

x



i

     

i



i

   1

j

 1  1 

i j

 1  1 

j j i

= 2 , ,

s

=

s

 1 where 

i

( 

LO

, 

HI

)   

F

  1   2 2    1 ( 

i

)  2       

x x T T

1

n

1          

c c n

0

c c n

     1

x T

1

x T n

Bedside Decision Support

Traditional Informed Consent Form

Personalized Medicine: ePRISM Consent

Risk Models Currently Implemented

Model Mortality Major bleeding Restenosis/TVR Data Source ACC-NCDR ACC-NCDR MASS-DAC Interventions

•

Transfer to main campus

•

Surgical consultation

•

Bivalirudin

•

Radial Access

•

Closure Device

•

DES vs. BMS

Personalized PCI Care: Case Examples

Mrs. Jones A 77-year old woman with HTN and stable CAD

Mr. Wilson

A 46-year old man with HTN, diabetes, and NSTEMI

Thienopyridine Selection Model

•

Model based on individual-level data from TRITON-TIMI 38 database (n=12,579)

•

Separate models for major ischemia (death, MI, stroke) and bleeding (TIMI major + minor)



Periprocedural MIs excluded (prognostic importance less certain)

 

Models include only variables known at time of procedure C-statistics ~ 0.70



reasonable discrimination

P2Y12 Selection Model: Potential Application

6 independent RF for non-CABG bleeding (n=17421, from HORIZONS and ACUITY)

1.

female sex 2.

advanced age 3.

elevated serum creatinine 4.

white blood cell count 5.

anemia 6.

non-ST-segment elevation MI or ST segment elevation MI

Mehran R. JACC 2010;55:2556-66.

ACS Risk Score

Mehran R. JACC 2010;55:2556-66.

A new era of man vs. machine competition is dawning

Garry Kasparov, the World No. 1 chess player from Russia, makes his move against Deep Junior, the world computer chess champion. Amir Ban (right), Deep Blue's operator, physically makes the move when told to do so by the computer.

Conclusions



Pharmacologic treatment of patients with CVD has improved over the years to decrease ischemic and bleeding complications



As most drugs which ↓ ischemia also ↑ bleeding, the offsetting impact of adverse ischemic and hemorrhagic events must be carefully examined



The net balance of ischemia and bleeding may vary tremendously with the risk profile of the individual pt for each complication, and the follow-up duration



Models are needed to assess a patient’s risk for bleeding as well as ischemic complications to further enhance treatment of the patients with therapies that are efficacious and also safe.