1

Detection of occult paroxysmal atrial fibrilation by implantable long-term

2

electrocardiographic monitoring in cryptogenic stroke and transient ischemic attack

3

population: a study protocol for prospective matched cohort study.

4

5

Andrea Petrovičováa, Egon Kurčab, Miroslav Brozmanc, Jozef Hasillad, Pavel Vahalae, Peter

6

Blaškof,Andrea Andrášovág,Robert Hatalah, Luboš Urbani , Štefan Sivákj*

7

8

a

9

94901 Nitra, Slovak Republic, email: [email protected]

Department of Neurology, Faculty Hospital, Constantine Philosopher University, Špitálska 6,

10

b

11

03659 Martin, Slovak Republic, email: [email protected]

12

c

13

94901 Nitra, Slovak Republic, email: [email protected]

14

d

15

94901 Nitra, Slovak Republic, email: [email protected]

16

e

17

94901 Nitra, Slovak Republic, email: [email protected]

18

f

19

[email protected]

20

g

21

[email protected]

Clinic of Neurology, Jessenius Faculty of Medicine, Comenius University, Kollárova 2,

Department of Neurology, Faculty Hospital, Constantine Philosopher University, Špitálska 6,

Clinic of Cardiology, Faculty Hospital, Constantine Philosopher University, Špitálska 6,

Clinic of Cardiology, Faculty Hospital, Constantine Philosopher University, Špitálska 6,

Kardiocentrum Nitra s.r.o, Špitálska 1, 94901 Nitra, Slovak Republic, email:

Kardiocentrum Nitra s.r.o, Špitálska 1, 94901 Nitra, Slovak Republic, email:

1

22

h

23

Diseases, Pod Krásnou hôrkou 1, 83348 Bratislava, Slovak Republic, email: [email protected]

24

i

25

Diseases, Pod Krásnou hôrkou 1, 83348 Bratislava, Slovak Republic, email:

26

[email protected]

27

j*

28

University, Kollárova 2, 03659 Martin, Slovak Republic, email: [email protected]

Department of Arrhythmias and Cardiac Pacing, The National Institute of Cardiovascular

Department of Arrhythmias and Cardiac Pacing, The National Institute of Cardiovascular

Corresponding author, Clinic of Neurology, Jessenius Faculty of Medicine, Comenius

29

30

Abstract

31

Background: Cardio-embolic etiology is the most frequent predicted cause of cryptogenic

32

stroke/TIA. The detection of occult paroxysmal atrial fibrillation (AF) is crucial for the

33

selection of appropriate medication.

34

Methods: The eligible cryptogenic stroke and TIA patients are enrolled from 2014 to 2018.

35

They undergo a long-term (12 months) ECG monitoring (implantable loop recorder

36

“REVEAL XT, REVEAL LINQ”) and are investigated for PITX2 (chromosome 4q25) and

37

ZFHX3 (chromosome 16q22) gene mutations. Appropriate control group of age- and sex-

38

matched healthy volunteers is investigated, too. Summary descriptive statistics, statistical tests

39

for differences between groups and correlation analyses will be used for results analysis.

40

Discussion: In our study we focus on a possible correlation between detection of atrial

41

fibrillation by implantable ECG loop, and PITX2 and/or ZFHX3 gene mutations in

42

cryptogenic stroke/TIA patients. This would lead to potential implementation of this genomic

43

approach to cryptogenic stroke/TIA diagnostics and management. The results will be

44

published after data evaluation in 2018.

45

Trial registration: www.clinicaltrials.gov, NCT02216370

2

46

Key words: transient ischemic attack, cryptogenic stroke, atrial fibrillation, implantable loop

47

recorder, PITX2, ZFHX3

48

49

50 51

BACKGROUND Stroke has been consistently considered as the 2nd leading cause of death for people

52

aged 60 years or older worldwide.[1] Recently, stroke is the 4th leading cause of death and the

53

1st leading cause of serious long-term disability in the USA.[2] Accurate definition of stroke

54

etiology is crucial as this will guide the most effective care and therapy.

55

Cryptogenic stroke (CS) according the TOAST classification is defined as an ischemic

56

stroke whose etiology cannot be determined despite extensive evaluation.[3] Stroke might be

57

cryptogenic for several reasons: 1. evaluation is not carried out in time of event, 2.

58

investigation is not sufficiently extensive, or 3. some causes are really unknown. Therefore

59

the proportion of CS depends on the speed and range of diagnostic evaluation which exceeds

60

routine investigation. Epidemiological observational trials using standardized stroke

61

etiological classification report 20-40% and even up to 50% incidence of unrevealed

62

causes.[4,5,6]

63

Cardio-embolic stroke is the most frequently predicted cause of CS studies

64

(58%).[7,8] AF is the most common cardiac arrhythmia and the most frequent cause of

65

cerebral cardio-embolism. The incidence of stroke in people with non-valvular AF is

66

estimated to be 5 times higher than in people without AF[9]. It has been suggested that

67

paroxysmal AF (PAF) is more prevalent than chronic AF in stroke and TIA patients. Chronic

68

AF and PAF appear to carry very similar stroke risk.[10]

69 70

Episodes of PAF may recur with variable frequency and ECG monitoring plays an important role in their detection. Short and predominantly asymptomatic presentations of PAF 3

71

(often referred as occult PAF) may remain undetected by routine methods of arrhythmia

72

screening. The incidence of new identified PAF varies widely depending on the choice of

73

cardiac monitoring devices, interval of monitoring from stroke onset and duration of

74

monitoring. The detection rate of PAF with external ECG monitoring in patients after CS was

75

5 to 20%.[11,12,13] Insertable cardiac monitor (ICM) automatically provides non-stop long

76

term ECG recordings with very high sensitivity and specificity for PAF detection. Studies

77

using ICM in similar populations have suggested a detection rate of approximately

78

25%.[14,15,16]

79

In our study we focus on a possible correlation between detection of atrial fibrillation

80

by implantable ECG loop, and PITX2 and/or ZFHX3 gene mutations in cryptogenic

81

stroke/TIA patients. This would lead to potential implementation of this genomic approach to

82

cryptogenic stroke/TIA diagnostics and management.[17,18]

83

84

METHODS

85

Objective and endpoints

86

The objective of our study is to determinate the proportion of patients with CS or

87

cryptogenic TIA (CTIA) that have underlying PAF in comparison with control person group.

88

The PAF is defined as an episode of irregular heart rhythm, without detectable P waves,

89

lasting more than 30 seconds. Primary endpoints are the first documented PAF episode after

90

CS or CTIA and incidence of PITX2 and ZFHX3 genes mutation. Secondary endpoints

91

include recurrent stroke or TIA, new ischemic lesions on neuroimaging (CT and/or MRI), left

92

atrial size (TTE/TEE) and relevant medication changes (mainly starting of oral anticoagulant

93

drugs). The cases (patients) and controls follow-up will last 12 months.

94

Study patients 4

95

The recruitment of at least 100 consecutive patients with CS or CTIA admitted to the

96

Department of Neurology in Nitra between 2014 and 2018 as well as at least 30 age and sex

97

matched voluntary subjects without known cerebrovascular, cardiovascular or other vascular

98

diesease representing the control person group. The ethics committee of the Faculty Hospital

99

in Nitra (Slovak Republic) approved the study protocol. A written informed consent will be

100

obtained from study participants or their legal representatives.

101

Inclusion criteria

102

1. CS or CTIA within 72 hours of onset symptoms

103

2. Age ≥ 18 years

104

3. CS: modified Rankin Scale (mRS) 4 or less at discharge (because of patients

105

compliance for study purposes)

106

4. CTIA: ABCD2 score 4 and more

107

5. Ability and willingness of patients or their legal representatives to understand study

108

instructions, both verbal and written, in accordance with ICH, GCP and legislation of

109

the Slovak Republic .

110

Exclusion criteria

111

1. Known etiology of stroke/TIA

112

2. Untreated hyperthyroidism

113

3. Myocardial infarction within 1 month of stroke/TIA onset

114

4. Coronary bypass within 1 month of stroke/TIA onset

115

5. Valvular disease requiring urgent surgery

116

6. Documented AF or flutter

117

7. Patent foramen ovale 5

118

8. Permanent indication to oral anticoagulation therapy

119

9. Long-term steroid therapy > 30 days

120

10. Participation in another clinical trial aimed at experimental pharmacologic therapy

121

11. Chronic inflammatory disease (rheumatoid arthritis, inflammatory bowel diseases,

122

lupus erythematosus etc.)

123

12. Severe co-morbidity not compatible with 12 months follow-up

124

13. Pregnant and breastfeeding women

125

14. Indication to pacemaker implantation (also implantable cardioverter defibrillator,

126

implantable pulse generator)

127

Imaging and laboratory investigations

128

Stroke diagnosis must be supported by consistent findings on CT and/or MRI. Standard non-

129

contrast CT examination is performed initially after stroke symptoms onset for ischemic

130

infarction localization and size (also to exclude intracranial hemorrhage or other pathology).

131

All patients undergo brain MRI (including DWI) to confirm ischemic etiology when

132

convincing lesions are not visible on CT. Stroke is classified as cryptogenic according to

133

TOAST criteria after extensive imaging and laboratory testing. Patients with TIA are

134

considered only if symptoms at presentation are speech problems, limb weakness or

135

hemianopia (ABCD2 score) and enrolled when CTIA is confirmed.

136

PAF detection: patients undergo 48 hours ECG Holter monitoring within 7 days after

137

stroke/TIA onset. The ICM device (Reveal® XT, Medtronic Inc., Minneapolis, USA) will be

138

implanted in those without documented AF. “Reveal XT” allows a 12 months non-stop ECG

139

monitoring of both patients and control group.

140

Patients follow-up visits will be scheduled at 1, 6, and 12 months after stroke/TIA onset and

141

every 6 months thereafter until clinical study closure with unscheduled visits in case of new

6

142

symptom/sign occurrence. Disability of patients will be classified according to the mRS and

143

Barthel Index score, too. The control healthy participants undergo assessment at scheduled

144

visits with 12 months non-stop ECG monitoring. Unscheduled visits will be performed at the

145

discretion of the investigator.

146

Except standard blood laboratory analysis, complete vasculitis and thrombophilia testing

147

panels are investigated. Analysis of PITX2 (chromosome 4q25) and ZFHX3 (chromosome

148

16q22) gene polymorphisms will be evaluated by PCR with genetic variation analysis such as

149

restriction analysis (RFLP), high resolution melting (HRM), or allele-specific PCR. For

150

results analysis summary descriptive statistics, statistical tests for differences between groups

151

and correlation analyses will be used.

152

This prospective and observational study is registered at www.clinicaltrials.gov with study

153

identifier NCT02216370.

154

155

156

DISCUSSION The unrecognized PAF and the consequent lack of prophylactic anticoagulant

157

treatment may result in significant morbidity with the first clinical presentation as a heart

158

failure, stroke, systemic embolism, and occasionally death. The fact, that intermittent ECG

159

monitoring (Holter ECG) has low sensitivity to identify patients with PAF, is generally

160

accepted. The commonly used 24 hours period of Holter ECG monitoring reached a PAF

161

diagnostic yield 2.4% - 9.4%. [19,20,21] Longer or continuous non-stop ECG monitoring has

162

been shown to increase PAF detection. In a sub-study of TRENDS, a single 24 hours Holter

163

ECG would have diagnosed PAF in 3%, whereas 7, 21, or 30 days of monitoring would have

164

led to diagnosis of PAF in 6%, 9%, or 11%, respectively.[22]

7

165

In the recent study of EMBRACE (The 30-Day Cardiac Event Monitor Belt for Recording

166

Atrial Fibrillation After a Cerebral Ischemic Event) among patients with a cryptogenic stroke

167

or TIA noninvasive ambulatory ECG monitoring for a target of 30 days significantly

168

improved the detection of atrial fibrillation by a factor of more than five and nearly doubled

169

the rate of anticoagulant treatment, as compared with the standard practice of short-duration

170

ECG monitoring. AF was detected in 16.1% in the intervention group, compared with 3.2% in

171

the control group.[23]

172

ICM devices for long-term ECG monitoring have been primarily used for the investigation of

173

syncope.[24] Compared to classical Holter monitoring, the “Reveal XT” device has a

174

sensitivity and specificity for PAF detection of 96.1% and 85.4%, respectively (3 years non-

175

stop monitoring).[25] Cotter et al reported “Reveal XT” PAF identification in 25.5% (13

176

patients) of 51 patients with unexplained stroke.[14] A large randomized clinical trial The

177

Cryptogenic Stroke and Underlying Atrial Fibrillation trial (CRYSTAL AF) compared long-

178

term ECG monitoring by an ICM with the conventional ECG Holter follow-up in patients

179

with recent CS. Continuous ICM monitoring detected over 7 times more patients with PAF

180

within 12 months than ECG Holter. Three years follow-up detected 30% PAF occurrence in

181

the ICM group to 3% in the Holter group.[26]

182

Choe evaluated the sensitivity and negative predictive value (NPV) of various external

183

monitoring techniques within a cryptogenic stroke cohort. Simulated intermittent monitoring

184

strategies were compared to continuous rhythm monitoring in 168 ICM patients of the

185

CRYSTAL AF trial. Short-term monitoring included a single 24-hour, 48-hour, and 7-day

186

Holter and 21-day and 30-day event recorders. Periodic monitoring consisted of quarterly

187

monitoring through 24-hour, 48-hour, and 7-day Holters and monthly 24-hour Holters.

188

For a single monitoring period, the sensitivity for AF diagnosis was lowest with a 24-hour

189

Holter (1.3%) and highest with a 30-day event recorder (22.8%). The NPV ranged from

8

190

82.3% to 85.6% for all single external monitoring strategies. Quarterly monitoring with 24-

191

hour Holters had a sensitivity of 3.1%, whereas quarterly 7-day monitors increased the

192

sensitivity to 20.8%. The NPVs for repetitive periodic monitoring strategies were similar at

193

82.6% to 85.3%. Long-term continuous monitoring was superior in detecting AF compared to

194

all intermittent monitoring strategies evaluated (p