SUBMITTED 14 SEPT 22 1 REVISIONS REQ. 31 OCT 22; REVISION RECD. 20 NOV 22 2 ACCEPTED 15 DEC 22 3 ONLINE-FIRST: DECEMBER 2022 4 DOI: https://doi.org/10.18295/squmj.12.2022.070 5 6 The Yield of Cardiac Investigations in Patients Presenting with an Acute 7 Ischemic Stroke 8 A single tertiary centre experience 9 Azhaar Alshukri,1 *Sunil K. Nadar,2 Arunodaya R. Gujjar,2 Hatim Al 10 Lawati,2 Mohammed Al-Rawahi,2 Isra Al-Kindi,3 Maathar AlFarsi3 11 1Department of Medicine, Oman Medical Specialty Board, Muscat, Oman; 2Department of 12 Medicine, Sultan Qaboos University Hospital, Muscat, Oman; 3Department of Medicine, College 13 of Medicine, Sultan Qaboos University, Muscat, Oman. 14 *Corresponding Author’s e-mail: sunilnadar@gmail.com 15 16 Abstract 17 Objectives: Strokes are a major source of morbidity and mortality. The Aim of this study was to 18 evaluate the effectiveness of routine cardiac investigations in identifying a cardioembolic aetiology 19 for ischemic strokes. Methods: This was a retrospective study of patients who had presented with 20 a stroke to our institution between January and December 2019. Results: A total 183 patients 21 (mean age 66.2+13.5 years, 109 or 59.6% male) were included in the final analysis. The common 22 risk factors were hypertension (74.9%) diabetes (61.7%), and hyperlipidaemia (54.6%). The 23 Middle cerebral artery (MCA) was the commonest artery affected (44 or 24%). On admission, 24 14(7.6%) patients were in atrial fibrillation with the rest being in sinus rhythm. On 24-hour ECG 25 holter monitoring, no abnormalities were noted in 135 patients. Atrial fibrillation was seen in 26 15(8.1%) patients (inclusive of the 14 who had AF on resting ECG). 32 (17.4%) patients had 27 evidence of non-sustained atrial arrhythmia and 9(4.9%) had non sustained ventricular tachycardia. 28 Thirty patients(16.3%) were also noted to have frequent supraventricular ectopics (>30/hour) 29 while 5(2.7%) patients had a high ventricular ectopic burden (>10% burden). No significant 30 abnormalities were noted in the echocardiograms of the patients, but 10 out of 132 (7.5%) patients 31 had a positive bubble echo. Twenty-four (13.1%) patients had enlarged left atria. Conclusion: The 32 overall diagnostic yield of abnormalities from routine cardiac testing for patients with stroke 33 appears to be low. Targeted screening of patients with crytogenic stroke as suggested by newer 34 guidelines is recommended. 35 Keywords: Cerebrovascular Accidents; Echocardiography; Atrial Fibrillation; Oman. 36 37 Advances in knowledge 38  Routine cardiac testing in patients with stroke has a low yield rate in terms of identifying 39 cardiac source of embolism 40  However, these tests help identify patients at high risk for future atrial fibrillation 41 Application to patient care 42  Although 24 hour monitoring does not identify all patients with atrial fibrillation, it helps 43 identify those at risk of atrial fibrillation 44  Careful evaluation of these patients must be made to assess suitability for anticoagulation. 45 46 Introduction 47 Strokes are a leading cause of morbidity and mortality worldwide. According to figures from 2019, 48 the reported prevalence was 101 million cases worldwide with around 12.2 million incident cases 49 of stroke.1 They are the second leading cause of death and the third leading cause of death and 50 disability combined.1 This is associated with increased costs and loss of productivity and in the 51 United States alone, it is projected that the medical costs related to stroke will increase from around 52 75 billion dollars annually in 2010 to around 180 billion dollars by 2030, in addition to indirect 53 costs due to loss of productivity from around 30 billion dollars annually in 2010 to around 55 54 billion dollars annually during the same period.2 According to the latest data published by the 55 WHO, stroke related deaths accounted for 11.1% of the total deaths, with the age adjusted death 56 rate at 94.98 per 100,000 population in Oman. It is the third leading cause of mortality in Oman 57 after coronary heart disease and diabetes mellitus.3 Additionally, the prevalence of risk factors for 58 stroke is high among the Omani population, adding to the potential economic burden of the 59 disease.4 60 61 The aetiology of ischemic strokes can be divided into five categories based on the TOAST (Trial 62 of ORG 10172 in Acute Stroke Treatment) criteria, large artery atherosclerosis, cardioembolism, 63 small vessel occlusion, stroke of other determined aetiology, and stroke of undetermined 64 (“cryptogenic”) aetiology.5 Cardioembolic stroke accounts for about 20–25% of all ischemic 65 strokes and is associated with a worse prognosis with respect to disability, mortality and both early 66 and long -term recurrences compared to other aetiologies.6;7 67 68 Sources for cardioembolism are further classified into major- or minor-risk sources according to 69 their (thrombo-) embolic potential.8 The most common major-risk source of cardioembolism is 70 atrial fibrillation (AF).9 Less frequent major-risk sources are cardiomyopathies with left 71 ventricular (LV) dysfunction, intracardiac thrombi, cardiac tumours, prosthetic valves, and 72 endocarditis. Minor-risk sources include patent foramen ovale (PFO), atrial septum aneurysm 73 (ASA), and calcification of aortic and mitral valves.8 74 75 The overall purpose of such classification of strokes using TOAST system (or any other similar 76 ones) is that it has major implications for preventive management.6 Strokes have a high incidence 77 of recurrence with a five-year recurrence rates of around 30%.10 Identifying the aetiology of the 78 stroke is therefore important in preventing future strokes. While the large majority of larger or 79 small artery strokes are treated with antiplatelet agents and statin, their long term relative risk 80 reduction in the recurrence of stroke is about 20-25%,11 in contrast, using anticoagulants for 81 prevention of definitively recognized cardio-embolic stroke results in a long term relative risk 82 reduction of 60-65%.11 Further, cardio-embolic strokes are associated with a much higher 83 morbidity and mortality as compared to other forms of stroke by virtue of their tendency to cause 84 large infarcts with poor cardiovascular support for re-establishment of cerebral circulation. Thus 85 diagnosis and prevention of cardioembolic strokes is relatively more effective (with 86 anticoagulation- as compared to non-cardioembolic strokes) as well as more cost-effective (in that 87 it prevents higher morbidity and mortality). 88 89 In the above context, the American Heart Association/American Stroke Association (AHA/ASA) 90 and the European Stroke Organisation (ESO) had recommended routine cardiac testing such as a 91 minimum of 24 hours of cardiac ECG monitoring and a transthoracic echocardiogram for all 92 patients who have suffered a stroke.11;12 The diagnostic yield from these sets of investigations is 93 variable with many studies recommending longer durations of cardiac monitoring to improve the 94 diagnosis and identification of atrial fibrillation.9;13 Similarly, there are studies suggesting the 95 routine use of transoesophageal echo over transthoracic echo to improve the identification of 96 cardiac abnormalities that can predispose to strokes.14 While the older guidelines recommended 97 cardiac investigations in all patients presenting with a stroke, newer guidelines however 98 recommend these investigations in patients with cryptogenic strokes.15 99 100 In our institution, it is standard practice to perform 24-hour ECG holter monitoring and a 101 transthoracic echocardiogram (with a bubble study) for all patients presenting with an ischemic 102 stroke. The aim of this study was to assess the rate of identification of abnormalities with these 103 tests in our institution. 104 105 Methods 106 All patients aged 18 and above, who were admitted with a final diagnosis of stroke between 107 January and December 2019 at The Sultan Qaboos University Hospital (SQUH) were identified. 108 Their electronic case records were reviewed and only those who had full cardiac evaluation were 109 included in the final analysis. We excluded those whose scans had demonstrated an intracranial 110 bleed or who died before any cardiac investigations were performed or those on whom 111 investigations were not done due to presumed very poor prognosis or for other clinical reasons. 112 We also excluded those in whom the case records were incomplete. Patients were diagnosed to 113 have a stroke based on clinical findings and from a CT scan of the brain. Carotid artery imaging 114 and MRI scans of the brain were not routinely performed. Ethical approval was obtained from the 115 Medical research ethical committee of the Sultan Qaboos university Hospital, Muscat, Oman 116 (MREC number 1365). 117 118 All eligible patients had undergone 24-hour holter ECG monitoring and a standard transthoracic 119 echocardiogram with an agitated saline contrast study. The Holter data were analysed by SEER 120 1000 (GE Medical systems technologies Ltd., Boston MA, USA). The abnormalities were 121 classified as in other major studies. Atrial fibrillation was considered if the episode lasted more 122 than 30 seconds. Anything less was classified as atrial tachyarrhythmia. 16;17 The frequency of 123 premature ventricular contractions (PVC) was reported as high if the burden was greater than 10% 124 of the total QRS complexes in a 24-hour period. 18 We classified those from 1-10% as medium 125 and anything less was considered low. The frequency of premature atrial contractions (PAC) was 126 considered to be high if it exceeded 30 PACs/hour (720 PACs in 24 hours), medium if it was 127 between 10-30 PACs per hour (240-720 PACs per 24 hours) and low if less than 10 PACs per 128 hour.19 129 130 Echocardiograms were performed using a GE Vivid 7 machine as per standard protocol.(20) The 131 abnormalities that were examined for included valvular abnormalities such as mitral stenosis, 132 intracardiac masses such as vegetations, tumors or thrombi and evidence of intracardiac shunts. 133 An agitated saline study was also performed according to standard protocol with normal respiration 134 and post-Valsalva manoeuvre where possible.(20) The study was positive if agitated saline was 135 noted at the left sided cardiac chambers in less than 4 cardiac cycles. This could then identify 136 patients with a patent foramen ovale (PFO) or other causes of an intracardiac shunt. The left atrium 137 was considered to be enlarged if it measured more than 4cm in men and 3.9cm in women. 138 Pulmonary hypertension was defined as mild if the mean pressure was calculated to be 25-139 35mmHg, moderate if between 35-45 mmHg and severe greater than 45 mmHg. 16 140 141 The RoPE (Risk of Paradoxical Embolism) score has been developed and validated as an 142 assessment tool to determine the probability that a PFO is responsible for a cryptogenic stroke.21 143 It can be used when assessing patients with a PFO preceding closure. A high score correlates with 144 increased likelihood that a PFO is responsible for the index stroke. The PFO‐attributable fraction 145 of stroke for a score of 7, 8, and 9 is 72%, 84%, and 88%, respectively, and defines a subset of 146 patients who may benefit from PFO closure. We calculated the RoPE score for our patients as 147 described before. 21 148 149 The collected data were analysed using statistical software SPSS Statistics (SPSS Inc., Chicago, 150 US) version 22. A descriptive analysis of the categorized variables was presented as proportions, 151 and continuous variables were presented as the mean and standard deviation. Chi-square test was 152 used to check for differences among groups for categorical variables. 153 154 Results 155 A total of 215 patients were identified who were admitted with a diagnosis of stroke during the 156 specified time period. Of these 32 patients did not fulfil our inclusion criteria and 183 patients 157 (mean age 66.2+13.5 years, 109 (59.6%) male and 74 (40.4%) female) were included in the final 158 analysis of the study. The risk factors of the patients are summarized in Table 1. Hypertension was 159 the commonest risk factor (74.9%) followed by diabetes (61.7%), and hyperlipidiemia (54.6%). 160 79 (43.2%) of patients had three cardiovascular risk factors followed by 36 (19.7%) who had two 161 risk factors. Forty patients (21.9%) had one risk factor, while 28(15.3%) did not have any 162 cardiovascular risk factor. Data on smoking habits was incomplete. 26.2% of patients had a 163 previous myocardial infarction and in 52 patients (28.4%), this was a recurrent stroke. One patient 164 had a prosthetic cardiac valve, and one was already known to have atrial fibrillation and both 165 patients were on anticoagulation. At the time of admission with a stroke both these patients had 166 INR values that were in the therapeutic range. All the patients with previous myocardial infarction 167 were on a single antiplatelet agent. There were no differences between the demographics of the 168 patients who presented with a first stroke or a recurrent stroke. 169 170 As per the TOAST classification types, small vessel disease was the commonest mechanism of 171 stroke in this cohort accounting for 112 (61.2%) patients. Large artery stroke was the mechanism 172 in 56 (30.6%) patients, while in 15 (8.2%) it was of undetermined aetiology. Cardioembolism, on 173 its own, was not identified as a mechanism of stroke in any of these patients. The patients with 174 atrial fibrillation had multiple atherosclerotic risk factors and therefore classified as stroke of 175 undetermined aetiology. The Middle cerebral artery (MCA) was the commonest artery to be 176 affected (44 or 24%), followed by the posterior cerebral artery in 13 (7.1%) patients. Almost a 177 third of the patients (67 of 183 or 36.6%) had involvement of the left sided limbs, while 45 of 183 178 (24.6%) had involvement of right sided limbs and in another 71 patients (38.8%), the side of 179 hemiplegia was not documented or there was no focal hemiplegia as it was either symptoms of 180 posterior circulation (such as dizziness, cerebellar signs etc.) or a lacunar infarct or the patient was 181 comatose. 182 183 All the patients had undergone a resting 12-lead-ECG on arrival. This identified 14 patients to be 184 in atrial fibrillation on presentation with the rest being in sinus rhythm. All patients underwent a 185 24- hour holter ECG monitoring. The findings are summarized in table 2. The 24- hour recording 186 did not show any abnormality in 135 patients. Atrial fibrillation was seen in 15 patients (one newly 187 diagnosed in addition to the 14 who had AF on their resting ECG). 32 patients had evidence of 188 non-sustained atrial arrhythmia and 9 had non sustained ventricular tachycardia. Thirty patients 189 were also noted to have frequent supraventricular ectopics (defined as more than 30/hour) while 5 190 patients had a high ventricular ectopic burden (more than 10% burden). There were no differences 191 between the findings in patients with a first or recurrent stroke. 192 193 A total of 165 patients had an echocardiogram performed during their stay in the hospital [Table 194 3]. There is no documented reason why the remaining 18 did not have an echocardiogram. None 195 of the patients had any mass or vegetation or thrombus noted in the scan. Severe aortic stenosis 196 was identified in 3 patients, severe Mitral stenosis in one patient and severe mitral regurgitation in 197 three patients. Two were found to have severe pulmonary hypertension. 132 patients had an 198 agitated saline contrast scan done of which 10 were reported as positive for a left to right shunt 199 (positivity rate of 7.7%). The RoPE score of these patients were a median of 5 with a range from 200 2 to 8. Three patients had a score of 6 and one of 8. Twenty-four patients had a left atrial size that 201 was above the upper limits of normal (4cm in males and 3.9 cm in females). There were no 202 differences between patients presenting with a first or recurrent stroke. 203 204 The relationship between left atrial size and atrial arrhythmias is shown in table 4. There was a 205 higher proportion of patients having AF and atrial tachyarrhythmias in the group with enlarged left 206 atria. Although numerically different, there was no statistical difference between the PAC burden 207 in the two groups. 208 209 Discussion 210 The primary aim of cardiac investigations after a stroke is to identify any potential source of 211 cardioembolism with the intention to treat and prevent recurrence of strokes. These could be 212 obvious thrombi or masses within the cardiac chambers, cardiac valve abnormalities or 213 arrhythmias such as atrial fibrillation. Though the old stroke guidelines recommended routine 214 testing in all patients with strokes, the current guidelines recommend routine imaging in the form 215 of transthoracic echocardiography and monitoring for rhythm disturbances especially atrial 216 fibrillation for patients with cryptogenic strokes. Similarly, whilst earlier guidelines recommended 217 a minimum of 24 hours of ECG monitoring, subsequent studies have demonstrated that longer 218 monitoring increases the detection rate of atrial fibrillation17;22;23 The latest guidelines have 219 embraced this and recommend prolonged ECG monitoring of at least 48 to 72 hours or longer if 220 possible.24 None of the patients in our study had undergone prolonged monitoring. Besides the 14 221 patients who had AF on their resting ECG, 24-hour monitoring identified only one further patient. 222 223 Besides episodes of atrial fibrillation, there has been interest in atrial ectopics or premature atrial 224 contractions (PAC) as a precursor to AF. Studies have demonstrated conclusively that patients 225 with a high burden of PACs have a higher risk of developing AF.19;25 Binici et al have shown that 226 patients with a PAC rate of more than 30 per hour had a 2.7-fold increase in the risk of developing 227 AF and also had a greater than 60% increased risk of death or stroke.(19) They found that for each 228 increase of 10 PAC per hour, the risk of the primary end point of death or stroke increased by 27% 229 and the risk of atrial fibrillation by 50%. While the American guidelines on stroke prevention do 230 not specifically mention these patients, it has been previously recommended that patients with a 231 high PAC burden would benefit from anticoagulation especially if their CHADS2VASC score is 232 greater than 2. However, there are no studies to back this recommendation. In our study, 36 patients 233 had some form of atrial arrhythmia with a similar number having a high burden of PACs. Given 234 the propensity of these patients to develop AF in the future and in the absence of facilities to 235 perform long term ECG monitoring, careful consideration should be given regarding 236 anticoagulation of these patients. 237 238 The other abnormality that is commonly identified on holter monitoring is frequent ventricular 239 ectopics. These can predispose to impaired LV systolic function and formation of LV thrombus 240 which can embolise.(26) In our study, almost a fifth of the patients had at least moderate to high 241 burden of VEs. Although this can represent a high adrenergic stage in the immediate post stroke 242 state, these patients should be monitored for development of tachycardia related cardiomyopathy 243 or other abnormalities.(26) Conduction abnormalities are a commonly reported findings on holter 244 monitoring, but in our study, we did not detect any patients with significant conduction 245 abnormalities. 246 247 Echocardiography can detect many potential cardiac sources of embolism such as left atrial 248 thrombus, patent foramen ovale, atrial septum aneurysm, valvular or myocardial disease, 249 vegetations or cardiac tumors amongst others.(27) Furthermore, it can reveal other cardiac 250 pathologies of potential therapeutic consequences such as wall motion abnormalities or a reduced 251 left ventricular function. The mode of echocardiography could be either transthoracic (TTE), 252 which is widely available, non-invasive, less personnel-intensive and cheap or transoesophageal 253 echocardiography (TOE) which is otherwise superior for evaluation of the aortic arch, left atrium 254 and its appendage, and atrial septum.(14) 255 256 Agitated saline contrast study is often considered to be part of the protocol for echocardiography 257 in patients who have a stroke with the aim of identifying any left to right shunts notably PFOs. 258 The role of closing a PFO post stroke is still not clear with conflicting data. Earlier randomized 259 controlled trials failed to show a statistically significant benefit for a PFO closure;(28;29) thus, many 260 investigators believed that a PFO, being a not uncommon occurrence, was an incidental bystander 261 in patients with stroke. However, meta-analyses and more recent specific trials have eliminated 262 several confounding factors and possible biases and have demonstrated a benefit of the use of a 263 shunt closure over medical therapy in patients with cryptogenic stroke, particularly among those 264 below the age of 60 years with no obvious risk factors explaining the incident stroke.(30-32) 265 266 In our study 10 out of the 132 patients who had undergone an agitated saline contrast study were 267 positive for a PFO. This is lower than the reported incidence of PFOs in the general population 268 (20-30%) and the stroke population (around 50%).(33;34) We are unable to explain this discrepancy. 269 Except for one patient, all had a RoPE score of less than 7. The reason for the patient with a high 270 RoPE score not being referred for further TEE assessment was not documented. It could be argued 271 therefore that to be cost effective, the agitated saline study should be performed only in those with 272 a high RoPE score. 273 274 The TTE may also identify other surrogate markers of potential AF such as increased left atrial 275 volume and left atrial strain. Besides this, other anatomical variations that can favour the 276 promotion of thrombi can also be diagnosed. These include aneurysm of the interatrial septum 277 (defined as a septal protrusion of greater than 11 to 15 mm and which is often associated with 278 PFOs),(35) the persistence of eustachian valve (which directs fetal blood flow towards the PFO) (36) 279 and a prominent chiari network (persistence of a remnant of the fetal atrial development).(37) In 280 our study, the echo reports did not comment on the presence of any of these abnormalities in any 281 of the patients. The other abnormalities that were diagnosed on TTE in our study were minor valve 282 abnormalities. However, it is unlikely that this contributed to the stroke. 283 284 The overall yield of positive tests from routine cardiac investigations for all patients presenting 285 with a stroke appears low from our study. 14 patients had newly diagnosed AF on their presenting 286 ECG with only one extra patient identified on 24-hour holter monitoring. Echocardiography 287 identified 4 cases with significant valve disease and a doubtful link to the presenting stroke This 288 is similar to other reported studies, where the diagnostic yield of routine testing was low and where 289 the cost effectiveness of such routine tests was questioned.(38) However, if we include all the cases 290 with surrogate markers of potential atrial fibrillation such as high PAC burden and dilated left atria 291 on echocardiography, the diagnostic yield for all directly and indirectly linked pathologies is 292 higher. Additional studies are needed to systematically ascertain whether anticoagulation in such 293 patients carries any long-term benefits, especially as it relates to stroke prevention. Furthermore, 294 we observed that the results of the cardiac investigations (apart from those with AF) did not 295 appreciably alter the final management, thereby questioning the usefulness of these investigations 296 in all patients with a stroke. Indeed, the newer guidelines, recommend these tests only in patients 297 with cryptogenic strokes in contrast to the older guidelines where these tests were recommended 298 in all patients with a stroke and would suggest that we change the policy in our hospital. 299 300 There were a few limitations in our study. This was a retrospective study that involved examining 301 past electronic case records of the enrolled patients. The comprehensiveness of the case records 302 therefore was a limiting factor as some patients had incomplete data and these were not included 303 in the final diagnosis. We did not follow up the prognosis of these patients to see whether they had 304 a recurrent stroke. Another limitation was the fact that the left atrium size was measured in two-305 dimension, but newer guidelines suggest that the LA area is to be calculated and adjusted to body 306 surface area. This could have given us a more accurate account of the state of the LA. 307 308 Conclusion 309 The overall yield from cardiac investigations in patients presenting with a stroke is relatively low. 310 They are, however, useful in identifying surrogate markers that increase the future risk of atrial 311 fibrillation, though studies are required to investigate the effectiveness of long-term 312 anticoagulation in these patients. Further studies are also required to ascertain the cost 313 effectiveness of routine cardiovascular testing in all patients with a stroke and whether they 314 influence management and outcome beyond identifying patients with AF. 315 316 Author Contribution 317 AA, IA and MA collected the data. SKN, ARG, HA and MAR analysed the data and drafted the 318 manuscript. All authors approved the final version of the manuscript. 319 320 Conflict of Interest 321 The authors declare no conflicts of interest. 322 323 Funding 324 No funding was received for this study. 325 326 References 327 1. Global, regional, and national burden of stroke and its risk factors, 1990-2019: a systematic 328 analysis for the Global Burden of Disease Study 2019. Lancet Neurol 2021 Oct;20(10):795-329 820. doi: 10.1016/S1474-4422(21)00252-0 330 2. Ovbiagele B, Goldstein LB, Higashida RT, Howard VJ, Johnston SC, Khavjou OA, et al. 331 Forecasting the future of stroke in the United States: a policy statement from the American 332 Heart Association and American Stroke Association. Stroke 2013 Aug;44(8):2361-75. doi: 333 10.1161/STR.0b013e31829734f2 334 3. Incidence of non communicable diseases in Oman. 2022. Available at 335 https://www.worldlifeexpectancy.com/oman-stroke Accessed on 8 November 2022 336 4. Ganguly SS, Gujjar AR, Al HH, Al HA, Jaju S, Al-Mahrezi A, et al. Risk Factors for 337 Ischaemic Stroke in an Omani Community: A case-control study. Sultan Qaboos Univ Med J 338 2021 Nov;21(4):585-90. doi: 10.18295/squmj.4.2021.043 339 https://doi.org/10.18295%2Fsqumj.4.2021.043 5. Adams HP, Jr., Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, et al. 340 Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter 341 clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 1993 342 Jan;24(1):35-41. doi: 10.1161/01.str.24.1.35 343 6. Kolominsky-Rabas PL, Weber M, Gefeller O, Neundoerfer B, Heuschmann PU. 344 Epidemiology of ischemic stroke subtypes according to TOAST criteria: incidence, 345 recurrence, and long-term survival in ischemic stroke subtypes: a population-based study. 346 Stroke 2001 Dec 1;32(12):2735-40. doi: 10.1161/hs1201.100209 347 7. Perera KS, Swaminathan B, Veltkamp R, Arauz A, Ameriso S, Marti-Fabregas J, et al. 348 Frequency and features of embolic stroke of undetermined source in young adults. Eur Stroke 349 J 2018 Jun;3(2):110-6. doi: 10.1177/2396987318755585 350 8. Cohen A, Donal E, Delgado V, Pepi M, Tsang T, Gerber B, et al. EACVI recommendations 351 on cardiovascular imaging for the detection of embolic sources: endorsed by the Canadian 352 Society of Echocardiography. Eur Heart J Cardiovasc Imaging 2021 May 10;22(6):e24-e57. 353 doi: 10.1093/ehjci/jeab008 354 9. Haeusler KG, Tutuncu S, Schnabel RB. Detection of Atrial Fibrillation in Cryptogenic Stroke. 355 Curr Neurol Neurosci Rep 2018 Aug 8;18(10):66. doi: 10.1007/s11910-018-0871-1 356 10. Kolmos M, Christoffersen L, Kruuse C. Recurrent Ischemic Stroke - A Systematic Review 357 and Meta-Analysis. J Stroke Cerebrovasc Dis 2021 Aug;30(8):105935. DOI: 358 10.1016/j.jstrokecerebrovasdis.2021.105935 359 11. Jauch EC, Saver JL, Adams HP, Jr., Bruno A, Connors JJ, Demaerschalk BM, et al. 360 Guidelines for the early management of patients with acute ischemic stroke: a guideline for 361 healthcare professionals from the American Heart Association/American Stroke Association. 362 Stroke 2013 Mar;44(3):870-947. doi: 10.1161/STR.0b013e318284056a 363 12. European Stroke Organisation executive committee: Guidelines for management of 364 ischaemic stroke and transient ischaemic attack 2008. Cerebrovasc Dis 2008;25(5):457-507. 365 doi: 10.1159/000131083 366 13. Wachter R, Groschel K, Gelbrich G, Hamann GF, Kermer P, Liman J, et al. Holter-367 electrocardiogram-monitoring in patients with acute ischaemic stroke (Find-368 AFRANDOMISED): an open-label randomised controlled trial. Lancet Neurol 2017 369 Apr;16(4):282-90. doi: 10.1016/S1474-4422(17)30002-9 370 https://doi.org/10.1177%2F2396987318755585 14. de Bruijn SF, Agema WR, Lammers GJ, van der Wall EE, Wolterbeek R, Holman ER, et al. 371 Transesophageal echocardiography is superior to transthoracic echocardiography in 372 management of patients of any age with transient ischemic attack or stroke. Stroke 2006 373 Oct;37(10):2531-4.v 374 15. Kleindorfer DO, Towfighi A, Chaturvedi S, Cockroft KM, Gutierrez J, Lombardi-Hill D, et 375 al. 2021 Guideline for the Prevention of Stroke in Patients With Stroke and Transient 376 Ischemic Attack: A Guideline From the American Heart Association/American Stroke 377 Association. Stroke 2021 Jul;52(7):e364-e467. 378 16. Stahrenberg R, Weber-Kruger M, Seegers J, Edelmann F, Lahno R, Haase B, et al. Enhanced 379 detection of paroxysmal atrial fibrillation by early and prolonged continuous holter 380 monitoring in patients with cerebral ischemia presenting in sinus rhythm. Stroke 2010 381 Dec;41(12):2884-8. 382 17. Wasser K, Weber-Kruger M, Jurries F, Liman J, Hamann GF, Kermer P, et al. The cardiac 383 diagnostic work-up in stroke patients-A subanalysis of the Find-AFRANDOMISED trial. 384 PLoS One 2019;14(5):e0216530. 385 18. Baman TS, Lange DC, Ilg KJ, Gupta SK, Liu TY, Alguire C, et al. Relationship between 386 burden of premature ventricular complexes and left ventricular function. Heart Rhythm 2010 387 Jul;7(7):865-9. 388 19. Binici Z, Intzilakis T, Nielsen OW, Kober L, Sajadieh A. Excessive supraventricular ectopic 389 activity and increased risk of atrial fibrillation and stroke. Circulation 2010 May 390 4;121(17):1904-11. 391 20. Mitchell C, Rahko PS, Blauwet LA, Canaday B, Finstuen JA, Foster MC, et al. Guidelines 392 for Performing a Comprehensive Transthoracic Echocardiographic Examination in Adults: 393 Recommendations from the American Society of Echocardiography. J Am Soc Echocardiogr 394 2019 Jan;32(1):1-64. 395 21. Collado FMS, Poulin MF, Murphy JJ, Jneid H, Kavinsky CJ. Patent Foramen Ovale Closure 396 for Stroke Prevention and Other Disorders. J Am Heart Assoc 2018 Jun 17;7(12). 397 22. Sanna T, Diener HC, Passman RS, Di L, V, Bernstein RA, Morillo CA, et al. Cryptogenic 398 stroke and underlying atrial fibrillation. N Engl J Med 2014 Jun 26;370(26):2478-86. 399 23. Sposato LA, Cipriano LE, Saposnik G, Ruiz VE, Riccio PM, Hachinski V. Diagnosis of 400 atrial fibrillation after stroke and transient ischaemic attack: a systematic review and meta-401 analysis. Lancet Neurol 2015 Apr;14(4):377-87. 402 24. Prefasi D, Martinez-Sanchez P, Fuentes B, Diez-Tejedor E. The utility of the RoPE score in 403 cryptogenic stroke patients 10% 148(80.9%) 30(16.4%) 5(2.7%) 109 (83.2%) 20 (15.2%) 2 (1.5%) 39(75%) 10 (19.2%) 3(5.7%) 0.21 PAC burden <10/hr 10-30/hr >30/hr 143(78.1%) 10(5.5%) 30(16.4%) 102 (77.8%) 7 (5.3%) 21(16%) 41(78.8%) 2 (3.8%) 9(17.3%) 0.9 SVT- supraventricular tachycardia, VT- Ventricular tachycardia, PVC- premature ventricular 455 contraction, PAC- premature atrial contraction 456 Analysis by Chi-square test 457 458 Table 3: Echo findings 459 Findings Overall group (N=165) First stroke (n-120) Recurrent stroke (n=45) P value Ejection fraction (%) 54.2 +11.1 54.8+11.2 52.6+10.7 0.23 LA size (mm) 25.7 (23.8-33.7) 26.86(24-34) 25.0(22.8-29.9) 0.07 LVIDd (mm) 4.2+0.73 4.19+0.72 4.13+0.69 0.66 Aortic regurgition No/Mild Moderate Severe 157 (85.7%) 8 (4.4%) 0 114 (95%) 6(5%) 0 43 (95.5%) 2(4.4%) 0 0.64 Aortic stenosis No/mild Moderate Severe 161 (97.6%) 1(0.6%) 3(1.8%) 117(97.5%) 1(0.8%) 2 (1.7%) 44(97.7%) 0 1(2.3%) 0.68 Mitral stenosis No/Mild Moderate Severe 164(99.4) 0 1(0.6) 120 (100%) 0 0 44 (97.7%) 0 1(2.3%) 0.21 Mitral regurgitation No/mild Moderate Severe 150(90.9%) 12(7.3%) 3(1.8%) 108(90%) 9(7.5%) 3(2.5%) 42(93.3%) 3(6.7%) 0 0.26 Pulmonary hypertension No mild Moderate Severe 144(87.3%) 12(7.3%) 7(4.2%) 2(1.2%) 101(84.1%) 10(8.3%) 7(5.8%) 2(11.7%) 43(95.5%) 2(4.4%) 0 0 0.20 Agitated saline echo (n=132) positive 10 (7.6%) 6(5%) 4(8.9%) 0.38 Enlarged LA 24 (13.1%) 21(17.5%) 3(6.7%) 0.08 LA- Left atrium, LVIDd- Left ventricular internal diameter in diastole, Left atrium was 460 considered to be enlarged if it measured more than 4cm in men and 3.9cm in women. 461 Analysis by chi-square test, students t-test or Mann-Whitney U test as appropriate 462 463 Table 4: Left atrial size and atrial arrhythmias 464 Normal LA size (n=139) Enlarged LA (n=24) P value PAC burden Low Medium High 113 (81.2%) 7 (5.1%) 19 (13.7%) 15 (62.5%) 2 (8.3%) 7 (29.2%) 0.11 Atrial fibrillation 8 (5.7%) 5 (20.8%) 0.01 Atrial tachyarrhythmias 24 (17.2%) 5 (20.8%) 0.1 PAC count 6(0-58) 33 (80-2402) 0.06* Values are number (percentage) or Median(Interquartile range) Analysis by chi-square test, 465 *Analysis by Mann-Whitney test; PAC- Premature atrial contraction 466