Disclosures Redefining Sudden Cardiac Death: Insights from the San - - PDF document

1

Redefining Sudden Cardiac Death: Insights from the San Francisco POstmortem Systematic invesTigation of Sudden Cardiac Death Study

14 September 2019 10th Annual California Heart Rhythm Symposium

Zian H. Tseng, M.D., M.A.S. Professor of Medicine in Residence Murray Davis Endowed Professor Cardiac Electrophysiology Section University of California, San Francisco

Disclosures

• Industry
• None
• Grants
• NIH/NHLBI: R01 HL 102090
• NIH/NHLBI: R01 HL 126555
• CDC: 6 NU38DP000019-01-01
• CDC: 1 NU38DP000019-01-00

U.S. Mortality by Death Certificates

100,000 200,000 300,000 400,000 500,000 AIDS Breast Cancer Lung Cancer Stroke SCD # deaths/year

1 U.S. Census Bureau, Statistical Abstract of the United States: 2001. 2 American Cancer Society, Inc., Surveillance Research, Cancer Facts and Figures 2001. 3 2002 Heart and Stroke Statistical Update, American Heart Association. 4 Circulation. 2001;104:2158-2163.

Etiology of Sudden Cardiac Death

Huikuri et al. N Engl J Med, Vol. 345 2001

2

Background

• Traditional SCD definitions are based on:
• Death certificates
• EMS records (CARES)
• Epidemiologic criteria (ACC/AHA/HRS, WHO,

Hinkle-Thaler) which all presume cardiac cause

• Therefore, incidence of SCD estimates vary widely
• WHO criteria:
• Witnessed: unexpected death within 1 h symptom
• nset
• Unwitnessed: unexpected death within 24 h of

having been observed alive and free of symptoms

RCTs, Cohorts: Sudden Cardiac Death

• ARIC, CHS:

– “A sudden pulseless condition from a cardiac origin in a previously stable individual occurring OOH or in the ED. For unwitnessed deaths, the participant must have been seen within 24h of arrest in a stable condition and without evidence of a noncardiac cause.”

• MERIT-HF trial: Metoprolol for Heart Failure

– “Witnessed instantaneous death in the absence of progressive circulatory failure lasting for 60 min or more, unwitnessed death in the absence of pre-existence progressive circulatory failure or other causes of death”

• Hinkle-Thaler, 1982:

– “No evidence of circulatory impairment until they collapsed and the pulse disappeared”

• SCD-HeFT and MADIT 2

– No definition of SCD

Emperor’s New Clothes?

• ~ 90% of SCDs occur out of hospital (OOH)1 in

jurisdiction of coroner or medical examiner (ME)

• Investigation after such natural deaths is not routine ME
• practice. Autopsy rates
• OOH deaths: ~10% (U.S.),2 23% (Finland)3
• Hinkle-Thaler: 27%4
• Exemplar autopsy studies demonstrating CAD as cause
• f > 80% of SCDs5,6 therefore limited by referral bias
• 1. Nichol JAMA 2008
• 2. Shojania NEJM 2008
• 3. Lunetta Inj Prev 2007
• 4. Hinkle, Thaler Circulation 65,1982.
• 5. Davies Circulation 85, 1992
• 6. Burke….Virmani NEJM1997

14 February 2016

3

Comprehensive Surveillance of SCD

▪ Oregon–SUDS (Chugh, JACC, 2004)

▪ WHO criteria ▪ Portland, OR: population 1,000,000 ▪ Track dozens of ambulance companies and area hospitals ▪ Review of all available records ▪ SCA+SCD: 53/100,000

Chugh SS et al JACC 2004

Comprehensive Surveillance of SCD

▪ Oregon–SUDS (Chugh, JACC, 2004)

▪ WHO criteria ▪ Portland, OR: population 1,000,000 ▪ Track dozens of ambulance companies and area hospitals ▪ Review of all available records ▪ SCA+SCD: 53/100,000

Chugh SS et al JACC 2004

Autopsy rate: 11%

Sudden Cardiac Arrest vs. Sudden Cardiac Death

SCA SCD?

Sudden “Cardiac” Death

SCA Valvular CAD DCM HCM 1o electrical disease Tamponade Neurologic Ao Dissection Hemorrhage

4

Sudden Arrhythmic Death

SCA

Valvular

CAD DCM HCM 1o electrical disease

Tamponade Neurologic Hemorrhage Ao Dissection

San Francisco POST SCD Study

POstmortem Systematic InvesTigation of Sudden Cardiac Death SF Medical Examiner SF Cardiology UC

All out of hospital and ER deaths reported by law to ME Every Incident SCD 2011-

Case Adjudication

Data reviewed at adjudication

• PMH (active problems, prescriptions, recent visits)
• Medications (Rx, QT-prolonging, methadone)
• EMS runsheets and rhythms
• Witness/family interviews
• Autopsy, tox, histology findings

Adjudication panel

• Dr. Phil Ursell

Chief Cardiac Pathologist UCSF

• Dr. Ellen Moffatt

Medical Examiner, City and County of San Francisco

• Dr. Zian H. Tseng

Study PI Cardiac Electrophysiologist, UCSF

• Dr. Jeff Olgin

Chief of Cardiology, UCSF

• Dr. Anthony Kim

Neurologist Director of UCSF Stroke Center Reported to ME All OOH, ED, Unexpected Inpatient Deaths N = 12,671 Attended Deaths Ineligible for Autopsy Active MD care < 3 wk, MD Signed DC N = 1,120 Non-Sudden Deaths N = 1,031 Non-Sudden Deaths N = 3,115 Adjudicated Autopsied WHO-Defined SCDs N = 525 OHCA Deaths Referred for Autopsy N = 912 Adjudication EMS + ME Scene Investigation Review Comprehensive Medical Records Review

N = 89 Declined Autopsy N = 16

Autopsied OHCA Deaths N = 896 Deaths in San Francisco County 2/1/2011-3/1/2014 N = 20,440 WHO-Defined SCDs w/o Autopsy N = 105 Death Certificate Review

N = 371 N = 2,744

Not Reported to ME Inpatient/Hospice Deaths N = 7,769 Inpatient / Nursing Home / Hospice Deaths N = 3,862 Non-Natural Deaths N = 2,021 Age <18 or >90 N = 2,012

OHCA Deaths Excluded Deaths (Suppl Table I) WHO-Defined SCDs (Suppl Table II) Tseng ZH…Moffatt E Circulation 2018

5

SCD Case Study #1

• 74 yo Filipino gentleman

– 4 V CABG 2002 – EF 22%, fixed defect anterior, inferior walls – Diabetes

• Admitted for fever and bronchitis, receiving IV

antibiotics

• Troponin negative, slightly fluid overloaded
• Called to consult on several asymptomatic

runs of NSVT (5-7 beats) and to consider primary prevention ICD

SCD Case Study #1

• Recommended uptitration of ß blocker, ICD

implant as an outpatient after completing antibiotic treatment

• ICD scheduled for 1 month after discharge
• 2 weeks later patient found dead in the

morning by wife

• Pt had returned to usual state of health, no

complaints the night before

“Non-Cardiac SCD”

• Autopsy: 2.5 L fresh blood

in stomach and duodenum

• Heart: no acute coronary

lesions

• Cause of death:

exsanguination

• ICD would not have

prevented SCD, pt may not have survived procedure

Adjudicated Etiologies of SCD

Tseng, ZH….Moffatt E. AHA Late-breaking 2016

6

T r a u m a D e a t h P u l m

• n

a r y E m b

• l

i s m O t h e r N

• n
• C

a r d i a c C a r d i a c , N

• n
• A

r r h y t h m i c H y p e r t e n s i v e H e a r t D i s e a s e N e u r

• l
• g

i c P r i m a r y E l e c t r i c a l D i s e a s e O c c u l t O v e r d

• s

e C h r

• n

i c C A D A c u t e C A D A

• r

t i c D i s s e c t i

• n

N

• n
• I

s c h e m i c / D C M O t h e r A r r h y t h m i c R e n a l F a i l u r e H C M

• 2
• 1

1 2 3 4

Z-Score

Median IQR Trauma Death Non-SAD Autopsy-Defined SAD Tseng, ZH….Moffatt E. In review 57.4% 45% 63.1% 61.8% 45% 54.6% 59.6% 56.7% N = 1001 630 194 436 147 89 48 323 23 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 Total Total Female Male Asian Black Hispanic White Other

Adjusted Incidence Rates per 100,000 Person-Years

WHO-Defined SCD Weighted Autopsy-Defined SAD IRR P-value IRR P-value Male vs. Female 2.37 <.0001 3.33 <.0001 Asian vs. White 0.48 <.0001 0.44 <.0001 Black vs. White 2.04 <.0001 1.52 .05 Hispanic vs. White 0.48 <.0001 0.44 <.0001 Other vs.White 1.11 .34 1.03 .91 WHO-Defined SCDs OHCA Deaths

Key: SAD Weighted SCD Observed SAD Unweighted

SCD and SAD Incidence Rates per 100,000 person-years

Tseng, ZH….Moffatt E. In review

SAD in Women vs. Men

P=0.0006

SCD Case Study #2

• 78 yo Asian man

– Dilated cardiomyopathy, stable EF 25% – Paroxysmal AF, on warfarin – Primary prevention ICD implanted 3 years ago, no shocks

• In usual state of health when wife went

shopping

• 3 hours later wife found him unresponsive
• Paramedics called, asystole on arrival, no

resuscitation attempted

7

ICD Interrogation

x 30

VF ≠ Sudden Arrhythmic Death

• At autopsy

– Massive subarachnoid hemorrhage (requires perfusing rhythm) – Heart 760 g

• Neurocardiogenic injury

– VF due to acute adrenergic surge

• Despite rhythm documentation of

VF, cause of death was neurologic

Acute CAD N=54 (10%, 18%) Chronic CAD N=115 (22%, 39%) Cardiomyopathy N=53 (10%, 18%) Hypertrophy N=45 (9%, 15%) Primary Electrical Disease N=7 (1%, 2%) Other Cardiac, Arrhythmic N=20 (4%, 7%) Cardiac, Non Arrythmic N=22* (4%, 100%) Acute Renal Failure N=6 (1%, 3%) Aortic Dissection N=14 (3%, 7%) Aspiration/Asphyxia N=5 (1%, 2%) Chemical Overdose N=71 (14%, 34%) GI Hemorrhage/Other GI N=11 (2%, 5%) Hypo/Hyperglycemia/DKA N=9 (2%, 4%) Infection N=24 (2%, 11%)

Neurologic N=28 (5%, 13%)

Pulmonary Embolism N=19 (4%, 9%) Other Non-Cardiac N=22 (4%, 11%)

Sudden Neurologic Death: 2nd largest Non-cardiac Cause

Cardiac, Arrhythmic N=294 (56%) Cardiac, Non-Arrhythmic N=22 (4%) Non-Cardiac N=209 (40%)

N=525

Kim AS….Tseng ZH. Neurology 2016

Sudden Neurologic Death Masquerading as SCD

• Risk of SND higher in women and non-whites

(p<0.01)

• Mostly intracranial hemorrhages but also stroke

and SUDEP

• Antiplatelet or AC associated with ICH among

noncardiac deaths (OR 6.0 [95% CI 1.5–24.8], p = 0.01) and overall (OR 3.9 [95% CI 1.01–15.5, p = 0.05)

• Up to 25,000 SNDs missed annually in U.S.
• 50% increase in fatal ICH incidence in U.S.

annually

Kim AS….Tseng ZH. Neurology 2016

8

SCD Case Study #3

• 46 yo Hispanic man with well-controlled HIV

disease (CD4 1000, VL <50)

• Compliant with HAART and statins
• Without complaint, found dead by his

roommate 1 hour after last seen well

HIV Increases Risk of Fibrosis and SCD

• 760 g heart
• Pulmonary edema
• 2V CAD
• 50% LAD, 60% LCx
• Dense transmural and

interstitial fibrosis

Total Male Female Asian Black Hispanic White Other Presumed SCD IRR

HIV+ vs. HIV–

1.86 1.38 2.31 n/a 1.05 0.81 2.00 n/a

95% CI; p* 1.39-2.50; <0.0005 1.02-1.88; 0.04 0.74-7.21; 0.15

• 0.53-2.09;

0.89 0.20-3.33; 0.77 1.42-2.82; <0.0005

• SAD

IRR HIV+ vs. HIV–

1.58 1.16 n/a n/a 1.01 n/a 1.75 n/a

95% CI; p* 1.02-2.43; 0.039 0.75-1.79; 0.51

• 0.36-2.83;

0.76

• 1.08-2.85;

0.023

• Adjusted Incidence Rates of Presumed SCD and SAD

10 20 30 40 50 60 70 80 90 100

+ - + - + - + - + - + -

46% 44%

Adjusted Incidence Rates per 100,000 py

HIV status N=

+ -

47 505

+ -

44 344 3 161 0 110 9 75 2 38 36 267 0 21

Total Male Female Asian Black Hispanic White Other Key:

Presumed SCD % SAD SAD 54% 46% 50% 58% 45% 0% 55% 52% 0% 50% 56% 60%

9

SCD Case Study #4

• 74 yo Caucasian man with CAD, PPM for

CHB

• Gen change scheduled for 5 weeks after ERI
• Did not show up to UCSF EP lab morning of

procedure: died in sleep

SCD Case Study #4

• Autopsy negative (no acute MI, PE, or bleed)

Current Postmarket Surveillance of CIEDs

• >3 million in U.S. alone have a PPM or ICD
• Manufacturer and User Facility Device

Experience (MAUDE) is mandatory for manufacturers, voluntary for healthcare providers

• Major limitation is that it generally captures patients under

active care who by definition are alive

• No mandatory surveillance of devices exists for patients who

die to determine the role of CIED failure

• Vast majority of SCDs with CIEDs do not get investigated

Tseng ZH, Hayward R, Clark N et al JAMA-IM 2015

Sudden Death in Patients with CIEDs

SCD with CIED (22, 4.3%)

Tseng ZH, Hayward R, Clark N et al JAMA-IM 2015

10

Sudden Death in Patients with CIEDs

SCD with CIED (22, 4.3%)

~1% SCDs w/ device malfunction

Tseng ZH, Hayward R, Clark N et al JAMA-IM 2015

• Improper device selection
• VT programming: delayed VF detection, ATP in VF zone
• Opportunities for MD practice improvement: device

selection, programming

Tseng ZH, Hayward R, Clark N et al JAMA-IM 2015

Missed VF on Postmortem ICD Interrogation

Tseng ZH, Hayward R, Clark N et al JAMA-IM 2015

76 M ischemic CM, AF Autopsy: Anterior LV scar, no acute MI

ICD Lead Fracture During Shock

• Improper device selection
• VT programming: delayed VF detection, ATP in VF zone
• Opportunities for MD practice improvement: device

selection, programming

Tseng ZH, Hayward R, Clark N et al JAMA-IM 2015

11

Presumed SCD (WHO) N = 615

Autopsy-Defined SAD (N = 348) Non-SAD (N = 267) Tseng, ZH….Vittinghoff E.. Circ A&E 2019

Presumed SCD (WHO) N = 615

ETIOLOGIES PREDICTORS

Pulmonary Embolus

Neurologic Occult Overdose

Dissectio n Acute Renal Failure Tamponade Acute Heart Failure

Hypertrophy

Infectious

Gastrointestinal

CAD

Cardiomyopathy Primary Electrical

Witnessed (N = 144)

SAD: 93 Non-SAD: 51

Unwitnessed (N = 471)

SAD: 255 Non-SAD: 216 ↑ Time since last seen normal

SSRI

Opiates

Illicit drug use

PEA

Depression

↑ Age

Mal e

Beta blocker Dyslipidemia

VT/VF

Autopsy-Defined SAD (N = 348) Non-SAD (N = 267) Tseng, ZH….Vittinghoff E.. Circ A&E 2019

Presumed SCD (WHO) N = 615

Tseng, ZH….Vittinghoff E.. Circ A&E 2019

https://ucsfhealthcardiology.ucsf.edu/sites/ucsfhealthcardiology.ucsf.edu/files/2018-08/SADCalculator.xlsx

Magnitude of Sudden Cardiac Death in the U.S.

100,000 200,000 300,000 400,000 500,000 AIDS Breast Cancer Lung Cancer Stroke SCD # deaths/year

1 U.S. Census Bureau, Statistical Abstract of the United States: 2001. 2 American Cancer Society, Inc., Surveillance Research, Cancer Facts and Figures 2001. 3 2002 Heart and Stroke Statistical Update, American Heart Association. 4 Circulation. 2001;104:2158-2163.

> 2 fold

• verestimate

30-50% underestimate

12

SCA ≠ SCD

SCA Valvular CAD DCM HCM

1o electrical disease, Renal channelopathies?

Tamponade Neurologic HIV CIEDs

Etiology of Sudden Cardiac Death

Adapted from Huikuri et al. N Engl J Med, Vol. 345 2001

2-fold

• verestimate

2-fold underestimate

Conclusions

• ~Half of conventionally defined SCDs in an entire metro

area were autopsy-defined SAD

• PPV WHO criteria for SAD: 55.8%, OHCA: 48.6%
• <50% SAD in women
• >50% had no cardiac history
• 98% of SADs had structural heart disease
• SAD incidence rates vary widely by sex and race
• Black males highest risk, Hispanic females lowest
• Most common non-SADs: Occult OD, neurologic, CIEDs
• CAD 1/3 overall SCD and 58% of SADs
• Cardiomyopathy and hypertrophy: 1/3 of SADs
• Cardiac mass increased risk for non-SAD and SAD

Implications

• The emperor is half naked: Conventionally defined

SCDs should be considered “presumed SCDs”

• Further investigation in minority groups and women
• To reduce overall public health burden of SCD, in

addition to CAD, efforts also should be directed towards screening, treating, and preventing non- arrhythmic causes (OD, neurologic diseases), hypertrophy, cardiomyopathy, and CIED problems

• SCD cohorts for genetic and molecular association

studies need refinement of phenotype

13

Acknowledgements

• Office of the Chief Medical

Examiner, San Francisco County

• Ellen Moffatt
• Amy Hart
• Michael Hunter
• ME investigators
• ZSFG Cardiology
• Priscilla Hsue
• UCSF Pathology
• Phil Ursell
• Andy Connolly
• UCSF Electrophysiology
• Jeffrey Olgin
• SF VAMC
• Joseph Wong
• UCSF Neurology
• Anthony Kim
• UCSF Epidemiology/Biostatistics
• Eric Vittinghoff
• SFFD Emergency Medical

Services/SFGH Emergency Medicine

• Karl Sporer
• Clement Yeh
• SF EMS personnel
• All victims of SCD and their

families in San Francisco County

R01 HL102090 (NIH) R01 HL126555 (NIH) DP14-1403 (CDC) R56 1067039 (NIH)

Next Steps

• Molecular association studies and family risk

stratification

• Precision EMS protocols
• SND: SUDEP, hemorrhage risk with anticoagulants,

anti-platelets

• HIV tissue reservoirs for cure?
• FDA postmortem postmarket surveillance: all deaths,
• ther municipalities
• Workup occult renal failure in infants

Incident SCDs in the San Francisco POST SCD Study, 1/1/11-11/30/13 N=517 SCDs with CIEDs* N=22 (22/517, 4.3%) PPM SCDs N=14 ICD SCDs N=8 Device concern N=7 (7/8, 87.5%) Device concern N=4 (4/14, 28.6%) No device concern N=10 Terminal rhythms: VF (N=3) PEA (N=2) Unknown (N=2) Other (N=3) Hardware Failure N=3 Improper Device Selection N=1 No device concern N=1 Patient with an Indication for an ICD died from VF

• Rapid battery depletion with presumed

asystole or profound bradycardia (N=1)

• Rise in RV lead impedance prior to death

with polymorphic VT/VF on device interrogation (N=1)

• Rapid rise in lead impedance prior to death

in a patient with pneumonia. Lead fracture

• r a global pacemaker circuit issue could not

be excluded (N=1). Terminal rhythm: VF; COD: Subarachnoid hemorrhage N=1 Hardware Failure N=1 Undersensing/failure to detect VF N=5 Programming issue N=1 RV lead fracture while delivering shock for VF

• Undersensing/failure to

detect VF only(N=3)

• Delay to shock due to

unsuccessful ATP in VF zone (device algorithm issue) and undersensing of VF (N=2) VT slower than lower limit of VT zone SCDs without CIEDs N=495 (495/517, 95.7%) San Francisco ICD Population, 1/1/11- 11/30/13 N=712 ICD Deaths N=109 (109/712, 15.3%) Survived N=603 (603/712, 84.7%) Non-Sudden Deaths N=101

Tseng ZH, Hayward R, Clark N et al JAMA-IM 2015

Sudden Cardiac Death in Patients with HIV Infection

14

Mortality Rates by Cause and Year

• 230 deaths over 3.7 median years’ follow-up
• 13% SCDs, 86% (30/35) of all cardiac deaths
• Mean HIV SCD rate: 2.6/1,000 PY (95% CI 1.8-3.8), 4.5-fold

higher than background HIV- SCD rate

Tseng ZH et al. JACC 2012 59(21):1891-6