? NIH/NHLBI (U01HL089458 & U01HL089145) ZOLL Jeffrey Olgin, MD - - PDF document

9/14/2019 1

A WCD Should be Placed in Post-MI Patients with EF≤35%

?

Jeffrey Olgin, MD Division of Cardiology, UCSF

Disclosures

• This is a debate and meant to be fun, not

personal

• Polarizing views are artificial and no issue is

black and white

• Actual position is more nuanced than presented
• I’ll be presenting data funded by:

– NIH/NHLBI (U01HL089458 & U01HL089145) – ZOLL

Background

• SCD high early post MI in patients with low EF
• Guidelines suggest waiting 90 days post MI

before assessing for ICD implantation

– IRIS and DINAMIT showed early implant of ICD does not decrease long-term mortality – EF will improve to >35% in over half of the patients – Competing risks of death

• A wearable cardioverter defibrillator (WCD) can serve

as a bridge to evaluation for ICD

9/14/2019 2

Guideline recommendations

2017 ACC/AHA/HRS Guideline for Management of Patients With Ventricular

• Arrhythmias. JACC 2017

VEST Study design

• Multi-center, randomized, open-label trial
• Participants within 7 days of hospital d/c for acute

MI & EF≤35%

• Randomized 2:1 to receive:

– Wearable cardioverter defibrillator (WCD) + GDMT or – Guideline-directed medical therapy alone

• Primary Outcome: 90 day arrhythmic death

(changed from total mortality)

• Secondary Outcomes: Total Mortality, Non-SD, others
• Analyses: Intention-to-treat & as-treated

Olgin, et al. NEJM 2018

0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Results: Outcomes, proportions analysis

WCD

Sudden Death (1°) Total mortality Non-sudden death

Relative Risk

0.67 Uncorrected P=0.18 0.64 Uncorrected P=0.04 0.63 Uncorrected P=0.15 Sudden Death Death, any cause Non- sudden death Control 2.4% 4.9% 2.2% WCD 1.6% 3.1% 1.4% 0% 1% 2% 3% 4% 5% 6% Control WCD

90-day Event Rates Worse Better

ARR 0.8% 1.8% 0.8% NNT 125 56 125

(1°)

P=0.18 P=0.04 P=0.15

0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Results: Outcomes, proportions analysis

WCD

Sudden Death (1°) Total mortality Non-sudden death

Relative Risk

0.67 Uncorrected P=0.18 0.64 Uncorrected P=0.04 0.63 Uncorrected P=0.15 Sudden Death Death, any cause Non- sudden death Control 2.4% 4.9% 2.2% WCD 1.6% 3.1% 1.4% 0% 1% 2% 3% 4% 5% 6% Control WCD

90-day Event Rates Worse Better

ARR 0.8% 1.8% 0.8% NNT 125 56 125

(1°)

P=0.18 P=0.04 P=0.15

Of the 48 deaths in the WCD arm, only 12 were wearing the WCD at the time of death.

9/14/2019 3

Hours worn per day 6 12 24 18

Hours worn (incl non-users)

Patient-Days (%) 10 20 30 40

Results: WCD wear-time

Olgin, et al. NEJM 2018

Characteristic WCD (N=1524) Control (N=778) WCD received, n (%) 1481 (97.2%) 20 (2.6%)* Median hours/day WCD worn [IQR] 18 [3.8-22.7] 0 [0-0]* Average hours/day WCD worn ± SD 14.0 ± 9.3 0.4 ± 2.7*

*P <0.001

Results: On treatment analysis

0.5 1 1.5 2 2.5 SD Rate per 100 person-Months Event Rate 0.2 0.4 0.6 0.8 1 1.2 1.4 Rate Ratio 0.5 1 1.5 2 2.5 Death Rate per 100 Person-Months Event Rate 0.2 0.4 0.6 0.8 1 1.2 1.4 Rate Ratio Rate Ratio

Sudden Death Total Mortality

P=0.02 P<0.001 0.5 1 1.5 2 2.5 Non-Sudden Death Rate per 100 … Event Rate 0.2 0.4 0.6 0.8 1 1.2 1.4 Rate Ratio Rate Ratio

Non-Sudden Death

P<0.001

Wearing WCD Not wearing WCD Rate Ratio

Worse Better Worse Better Worse Better

What do you call someone who posts a lot

• n Twitter?

What do you call someone who posts a lot

• n Twitter?
• A. A Donald

9/14/2019 4

What do you call someone who posts a lot

• n Twitter?
• A. A Donald
• B. Someone with too much free time

What do you call someone who posts a lot

• n Twitter?
• A. A Donald
• B. Someone with too much free time
• C. A Twit

Science is like an onion. It has layers, Jackass.

Ogres with PhDs Ogres with MAs Ogres with BAs Jackasses Dumb Jackasses Trump Jackasses EP Jackasses

Discussion: How to interpret VEST

• The WCD did not significantly reduce sudden

death and thus did not meet 1°outcome

9/14/2019 5

Discussion: Why primary outcome wasn’t met

• Power insufficient to conclude WCD was not

effective

– Broad confidence intervals

• Rate of sudden death lower than expected
• Possible misclassification of sudden deaths

– 5% of death adjudicated as indeterminate – 4 of 9 SD’s wearing WCD had no VT/VF – Reducing power for SD outcome but not total mortality

• Lower WCD wear time further reduced power

0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3

Treatment <-- Better — Worse -->

Spectrum of negative trials

VEST HERS WHI COURAGE CAST Inconclusive with suggestion of benefit Inconclusive with suggestion of harm & confirmatory study Clear harm

Relative Risk

MADIT CABANA

Discussion: How to interpret VEST

• The WCD did not significantly reduce sudden

death and thus did not meet 1°outcome

– Inconclusive result due to insufficient power – Low WCD wear-time

• The WCD was associated with a 36% reduction
• f total mortality (nominal p=0.04)

Discussion: Correction for multiplicity

• Bonferroni correction assumes complete

independence of hypotheses testing

– Mortality outcomes are NOT independent but rather subsets of total mortality – Remains significant with less conservative corrections, taking into account outcome correlations and importance of total mortality

• Not all secondary outcomes are created equally

– Total mortality is a different outcome than a combined

• utcome of mortality + hospitalizations + status decline

9/14/2019 6

Discussion: Sudden death etiology

Tseng, Z. et al Circulation, 2018

Cardiac, Arrhythmic (SAD) N = 78 (65%) Cardiac, Non Arrhythmic N = 10 (8%) Non-Cardiac N = 32 (27%)

Witnessed N = 120

Cardiac, Arrhythmic (SAD) N = 215 (53%) Cardiac, Non Arrhythmic N = 12 (3%) Non-Cardiac N = 178 (44%)

Unwitnessed N = 405

• Autopsy-proven cause of

death

• 525 consecutive ”sudden

deaths” in SF County

• 98% autopsy rate.

Discussion: How to interpret VEST

• The WCD did not significantly reduce sudden

death and thus did not meet 1°outcome

– Inconclusive result due to insufficient power – Low WCD wear-time

• The WCD was associated with a 36% reduction
• f total mortality (nominal p=0.04)
• As-treated analysis showed a significant

reduction in sudden death and total mortality

Discussion: How to interpret VEST

• The WCD did not significantly reduce sudden

death and thus did not meet 1°outcome

– Inconclusive result due to insufficient power – Low WCD wear-time

• The WCD was associated with a 36% reduction
• f total mortality (nominal p=0.04)
• As-treated analysis showed a significant

reduction in sudden death and total mortality

• Very few serious adverse events
• High WCD shock success rate, consistent with
• ther studies

Concluding Remarks

9/14/2019 7

Concluding Remarks

How many electrophysiologists have stopped doing AF ablations as a result of CABANA?

Concluding Remarks

• Mortality remains high immediately post-MI in

patients with EF≤35%, despite PCI, optimal medical therapy and EF recovery at 3 months in ~60% of patients

• Given the totality of the results of VEST and

prior series, it may be reasonable to prescribe the WCD post-MI in patients at highest risk for sudden death and in whom a high compliance is expected – Shared decision making seems reasonable

Concluding Remarks

• Mortality remains high immediately post-MI in

patients with EF≤35%, despite PCI, optimal medical therapy and EF recovery at 3 months in ~60% of patients

• Given the totality of the results of VEST and

prior series, it may be reasonable to prescribe the WCD post-MI in patients at highest risk for sudden death and in whom a high compliance is expected – Shared decision making seems reasonable

TWITTER

1min ago

New message in feed

9/14/2019 8

Concluding Remarks

• Mortality remains high immediately post-MI in

patients with EF≤35%, despite PCI, optimal medical therapy and EF recovery at 3 months in ~60% of patients

• Given the totality of the results of VEST and

prior series, it may be reasonable to prescribe the WCD post-MI in patients at highest risk for sudden death and in whom a high compliance is expected – Shared decision making seems reasonable

12:15 PM – 14 Sep 2019

And now, the counter argument…