5/12/2014

1

CRYPTOGENIC STROKECRYPTOGENIC STROKE

Victor Barredo, M.D.,Neurologist

Baptist and South Miami Hospitals

DisclosuresI have no relevant commercial relationships to disclose.

Ischemic Artery

Occlusion (85%)

Hemorrhagic Vessel Rupture (15%)

Breakdown of Ischemic Strokes:

Atherothrombotic (25-30%)Stenotic artery feeding area of infarction

Cardioembolic (20%)A thrombus or other material dislodges from theA thrombus or other material dislodges from the heart or aortic arch

Lacunar/Small Vessel (15-20%)Small, deep infarct

Other/Uncommon (5-10%)

Cryptogenic (25-40%)Unknown cause

Adams HP Jr, Stroke. Jan 1993; 24; 35-41Data from NINCDS Stroke Data Bank: Foulkes et al. Stroke. 1988;19:547

Small vessel strokes

Generally due to metabolic disease Hypertension Hyperlipidemia Diabetes Smoking

Not due to embolic disease The arteries are microscopic, and too small to hold an

embolus Workup and treatment generally consists of

Carotid dopplers Lipids Glucose monitoring Blood pressure control Antiplatelet agents

3

5/12/2014

2

Thalamic Lacunar stroke

4

Embolic Strokes

Strokes often line up at the watershed distribution or at the cortex

May be a small shower representing a lysed clot May be in one or multiple vascular territories Multiple vascular territories is pathognomonic for

emboli or vasculitisemboli or vasculitis Infarcts tend to be larger, with more severe

symptoms and deficits Aphasia Neglect Field cuts seizures

Can cause edema and death

5

Watershed Distributions

6

5/12/2014

3

Embolic stroke

7

Why MRI Matters

• Delayed diagnosis

• Missed diagnosis

8 8

MRICT

diagnosis

• Increased risk from  alternative imaging   modalities

Patient presents with mild left sided weakness

• Examination shows no cortical signs, only slight Left sided weakness.• Most consistent with lacunar syndrome.

• CT at left is normal.• MRI at right shows embolic, cortical infarct.

Copyright © American Heart Association

9

5/12/2014

4

Embolic Stroke Workup

Carotid Dopplers to rule out artery to Artery Embolus Fasting Lipids Blood Sugar MRA or CTA of intracranial vessels to rule out stenosis

The SAMPRIS trial shows that intracranial angioplasty does not reduce the risk of recurrent stroke

Medical management Does give a culprit

Transesophageal Echocardiogram Rule out thrombus in the Left Atrial Appendage

Cannot be seen on 2D surface Echo PFO is of no clinical importance most of the time given

CLOSURE study.

10

Cardioembolic Stroke Causes

Atrial Fibrillation/Flutter Low EF Cardiac Thrombus or “Smoke” Rare Causes: Rare Causes: Cardiac Tumor Vegetation on valve Mobile aortic arch atheroma (>3mm) Endocarditis PFO with venous source of embolism

11

The dilemma of Cryptogenic Stroke

If all tests are negative (and the patient has no history of Atrial Fibrillation) we now have a cryptogenic stroke

The patient is at high risk for a recurrent stroke and hence further disability and stroke, and hence further disability and death

The stroke obviously looks embolic in nature, but how to prove it? Paroxysmal atrial fibrillation may only be

present for a few minutes a month, but that is enough to cause a devastating stroke.

12

5/12/2014

5

Cryptogenic StrokeWhy AF Matters

AF equals 5 fold increase for stroke risk¹

Up to 90% of Paroxysmal Atrial Fibrillation (PAF) episodes may be asymptomatic.² 

Risk of stroke annually is equal for PAF and permanent AF ³

Detection of AF in Cryptogenic Stroke Patients changes treatment

Guidelines state change from antiplatelet to OAC4

1. Wolf PA, et al. Stroke 1991;22:983‐988 ; 2. Isreal et al, J AM Coll Cardiol. 2004;43:47‐52; 3. Page, RL, et al Circulation. 1994;89:224‐227 

Hart RG, et al Coll Cardiol. 2000; 35:183‐187 4. Camm et al, European Heart Journal . 2012; 

33, 2719‐2747

Importance of AF and Stroke

• AF is frequently paroxysmal and asymptomatic, making detection of AF difficult

• 25% of those with AF-associated stroke have no known prior history of AF

• Even in stroke patients with known PAF, 50-70% are in p ,sinus rhythm at time of stroke

• AF is one of the only reasons to use anticoagulation for secondary stroke prevention

• For almost all other reasons antiplatelet agents are used

• Anticoagulants have significantly high risk associated with use.

14

• 163 patients with previous ischemic stroke/TIA, no known AF, were continuous monitored via pacemaker or ICD

• NDAF > 5 minute

from AT/AF

0.8

0.9

1.089% of NDAF patients identified beyond 1 day

78% of NDAF patients identified beyond 7 days

60% of NDAF patients identified beyond 30 days

Ziegler P. et al. Stroke. 2010;41

duration were found in 28% patients.

• 73% of patients had newly detected AT/AF on <10% of follow-up days

Time from Device Implant (months)

0 3 mo. 6 mo. 9 mo. 12 mo.

Freedom f

0.5

0.6

0.7

Number at Risk:

163 127 111 106 67

5/12/2014

6

AT/AF burden >5.5 hours on any of 30 prior days appeared to double thromboembolic (TE) risk2

“35% of all strokes and systemic emboli were preceded by device detected atrial tachyarrhythmia's”2

1. Glotzer T. et al. The TRENDS Study Circ Arrhythmia Electrophysiol. 2009.

2. http://m.theheart.org/article/1153461.do

16

Clinical Impact of NDAF in Stroke Patients

Risk of stroke recurrence is 4 times greater among prior stroke patients with NDAF (15.5%) compared to those with either known AF or no AF (3.9%).

Emphasizes the need for timely detection of AF

17

p<0.0001

Kamel et al. J Stroke Cerebrolvasc Dis ;2009 Nov‐Dec; 18(6):453‐7

Ziegler PaperTopic: Intermittent and symptom-based monitoring

• Data from 574 AT500 IPG

patients were analyzed

retrospectively over 1 year,

with intermittent monitoring

simulated by analyzing data

from randomly selected

days

1. Ziegler P. Comparison of continuous versus intermittent monitoring of atrial arrhythmias. Heart Rhythm 2006;3:1445–1452).

days

• “Intermittent and

symptom-based

monitoring is highly

inaccurate for identifying

patients with any or long-

duration AT/AF and for

assessing AT/AF burden.”

Example: Quarterly Holter recording detects AF in

54% of the patients with AF, and is correct 29% of

the time in ruling out AF in patients. 18

5/12/2014

7

How AF is Detected in Cryptogenic Stroke PatientsThe more you look, the more AF you find

20%

30%

•N = 1491

•Acute stroke or TIA and 

no history of AF 

•ECG monitoring in 

Incremental % AF Detection

Acute Workup

After CS Diagnosis

a

0%

10%

2.7%4.1% 5.0% 5.7%

Hospital

•24‐hour Holter 

recording if normal ECG

•7‐day event monitor if 

normal Holter

Cotter / Ritter Data

1. Jabaudon D. Et al. Usefulness of Ambulatory 7‐Day ECG Monitoring for the Detection of Atrial Fibrillation and Flutter After Acute Stroke and Transient Ischemic Attack Stroke 2004;35:1647‐1651.2. Cotter, P.E., et al., Incidence of atrial fibrillation detected by implantable loop recorders in unexplained stroke. Neurology, 2013 Apr 23;80(17):1546‐503. Ritter, M.A., et al., Occult Atrial Fibrillation in Cryptogenic Stroke: Detection by 7‐Day Electrocardiogram Versus Implantable Cardiac Monitors. Stroke, 2013 May;44(5):1449‐52.

SURPRISE Study Updated Results

Methods: 85 patients with cryptogenic stroke/TIA and no AF on 24‐hour telemetry 

were implanted with Reveal XT All patients had a minimum of 6 months of monitoring

Results 14 of 85 (16.5%) of patients diagnosed with AF

Median time from stroke onset to first recorded event ‐ 98 days Average AF burden was 2 hours per day monitored CHADS2Vasc in AF group was 4.14 vs. 3.24 in no AF (p=0.03)

Christensen LM. et al. Paroxystic Atrial Fibrillation (PAF) in Patients with minor ischemic stroke or transient ischemic attack. 2013 EuroStroke

Incidence of Atrial Fibrillation detected by Implantable Loop Recorders in Unexplained Stroke

Methods: 51 patients with cryptogenic stroke

implanted Patient workup included vascular &

cardiac imaging, at least 24 hours of cardiac rhythm monitoring

l

21

Results: AF was identified in 25.5% (13)

patients Median time to AF detection after

implant was 48 days Median duration of first detected AF

episode was 6 minutes AF was associate with increasing age (p

= 0.018), interarterial conduction block (p = 0.02), left atrial volume (p = 0.025) and the occurrence of atrial premature contractions on preceding external monitoring (p = 0.004)

Clinical action (OAC) was taken on all patients where AF was detected

Cotter et al, Neurology, 2013; 80 (14)

5/12/2014

8

Occult Atrial Fibrillation in Cryptogenic StrokeDetection by 7‐day ECG vs. ICM

Methods: 60 patients with cryptogenic stroke implanted Compared ICM to 7-day Holter monitor• Patient workup included Cerebral imagining, ECG,

72 hour telemetry, 24-hour Holter, TEE

Results: AF was identified in 17% (10) patients Average time to detection 64 days post-stroke Yield of ICM (17%) vs. 7-day ECG

(1.7%) significantly higher p=0.0077

Ritter et al, Stroke, 2013; Volume 44, Issue 5

22

Insertable Cardiac Event Recorder in Detection of AtrialFibrillation After Cryptogenic Stroke: An Audit Report

Etgen et al., 2013

MethodsPatient work‐up included MRI, 12‐lead ECG, 24‐72 hr. bedside telemetry, 24‐hr Holter, TEE, computed tomography/MRI angiography

22 patients with cryptogenic stroke and eligible for oral anticoagulation were implanted with Reveal XT

AF defined as episode ≥ 6 minutes

ResultsAF detected in 27.3% (6) patientsAverage time to detection post‐stroke was 161 days

Etgen et al, Insertable Cardiac Event Recorder in Detection of Atrial Fibrillation After Cryptogenic Stroke: An Audit Report.  Stroke. 2013;44:2007‐2009

Patient Pathway for Acute Stroke

24

5/12/2014

9

Patient Pathway for Acute Stroke

25

ReimbursementCryptogenic Stroke – Appropriate Diagnosis  Coding

“Cryptogenic stroke” usually refers to strokes with no clearly definable cause even after extensive workup.

Although atrial fibrillation (AF) may be suspected in these patients, it can’t be assigned as a diagnosis code until the diagnosis is established. 

Diagnosis coding for cardiac monitoring insertion depends on the circumstances: 

Inpatient (during the same admission as the acute stroke)

The acute stroke is the principal diagnosis code

Outpatient (after discharge from the acute stroke admission)

If a specific symptom or sign is present that necessitated the device insertion test/procedure, that should be used as the principal diagnosis code; however, these patients are usually asymptomatic 

If the patient has residuals from the stroke, a code from category 438.XX (late effects of cerebrovascular disease) is appropriate as the principal diagnosis code

If the patient has no residuals from the stroke, code V12.54 (personal history of stroke or TIA w/o residual deficits) is appropriate as the principal diagnosis

Code V12.54 can also be used as secondary diagnosis codes if a specific symptom or sign is coded as the principal diagnosis

2013 coding analysis & recommendations provided by Clarity Coding; www.claritycoding.com

ICMs are UnderutilizedICMs are recommended by clinical guidelines; yet significantly underutilized

Up to 3 in 4 patients who met appropriate criteria

for ICM implantation 

did not receive one1

5/12/2014

10

Cardiac Diagnostics Landscape

24-Hour Holter

2.5 Million

Event Recorder

1.5 Million

14-30 Day “MCOT”

250,000

Insertable Cardiac Monitor

25,000

Source:  Frost & Sullivan report:  North American Cardiac Monitoring and Diagnostic Services Markets

Symptoms and Intermittent MonitoringThe Tip of the AF Iceberg

Symptoms/Intermittent monitoring

Continuous monitoring

29

Conclusion

Embolic strokes have a specific appearance on a diagnostic MRI

With no known hx of AF, the workup is usually negative, leaving us with a diagnosis of cryptogenic strokediagnosis of cryptogenic stroke

The yield of diagnosing AF in these patients can be quite high if proper protocol is followed

An enthusiastic and cooperative partnership between a neurologist and cardiologist is imperative

30

5/12/2014

11

ThankThank YouYou

Thank you

31

Baptist Children’s HospitalBaptist Hospital Doctors Hospital

Baptist Cardiac &Vascular Institute

Homestead Hospital Mariners Hospital

West KendallBaptist Hospital

Baptist Outpatient Services

South Miami Hospital

32