Post-operative atrial fibrillation (POAF) is frequently observed after open-heart surgery. Despite numerous attempts to predict POAF, it remains a challenge to correctly identify the patients at risk. New technologies are available but not yet in cooperated in clinical guidelines and prediction models. We aim at identifying patients at risk of AF occurrence/recurrence after open-heart surgery by use of signal processed surface ECG (wavECG).

The study is an explorative, prospective cohort study of 100 patients undergoing first-time elective open-heart surgery in the Left Atrial Appendage Closure by Surgery-2 (LAACS-2) trial (NCT03724318). Patients are examined by electrocardiogram (ECG), wavECG, transthoracic echocardiography and continuous heart rhythm monitoring. Primary endpoint is occurrence/recurrence of AF. Left ventricular diastolic dysfunction as evaluated by wavECG will be used to predict incidence of primary endpoint.

Introduction: Post-operative atrial fibrillation (POAF) is frequently observed after open-heart surgery,1-3 and atrial fibrillation (AF) is frequently complicated by stroke.4-6 During AF, the electrical activity is uncoordinated and the atrial contraction is reduced, increasing the risk of thrombi formation. Subsequently, the thrombi can enter the general circulation, reach the small intracerebral arteries and block blood flow resulting in ischaemic damage.5,7 Anti-coagulant medicines efficiently prevent ischaemic strokes,8 however many patients are asymptomatic or have vague symptoms why diagnosis can be difficult to obtain.5 Furthermore, AF that occurs secondarily to surgery is currently considered self-limiting and continuous anti-coagulation is not necessarily recommended.4,9 However, cumulating evidence suggest that secondary AF in patients with known heart disease increases the long-term risk of stroke and transitory ischaemic attack, recurrent AF and mortality.10-15 In patients undergoing coronary artery bypass grafting (CABG) new-onset AF is associated with increased long-term risk of stroke and mortality compared to patients without AF.14,15 Though there are well-established risk factors for AF, such as age, arterial hypertension, heart failure, myocardial infarction, heart valve disease and diabetes mellitus, the pathophysiology of AF is complex and not fully understood.6,16,17 Both structural remodelling of the atria and electrical pathways, as well as oxidative stress, calcium overload, myofibroblast activation, microRNAs and inflammation, are mentioned as possible factors involved in initiation and progression of AF.6,16-18 Attempts to create and improve risk stratification models for AF and stroke has previously been done, both in community based cohorts and patients undergoing cardiac surgery, e.g. CHA2DS2-VASC, FHS AF risk score, ARIC risk prediction tool, CHARGE-AF risk model, HAVOC score, SYNTAX score, POAF score and POAF prediction model.19-27 However, it remains challenging to accurately predict occurrence of AF both perioperatively and in the subsequent years.28,29

A recent systematic review and meta-analysis concluded that p-wave terminal force in lead V1 (PTFV1), p-wave duration and maximum p-wave area, are all correlated with stroke,30 and a recent review similarly suggest that short and prolonged p-wave and PTFV1 is correlated with AF.31 sp-ECG by Myovista (Southlake, Texas, USA) is a novel advanced ECG utilizing continues wavelet transform signal processing and can detect left ventricular diastolic dysfunction (LVDD) which is associated with AF and post-operative heart failure in other studies.17,32-35 LA strain and E/LA strain indices predicted new cardiovascular events including strokes, in a stroke population36 In patients undergoing catheter ablation for AF, LA strain can identify patients at risk of AF recurrence37,38 POAF has furthermore been predicted in a cohort of patients with severe aortic stenosis by LA strain indexes.39

Aim: We aim at characterizing patients with AF occurrence by novel, easy, and accessible methods such as electrocardiogram (ECG), signal processed surface ECG (Myovista, wavECG) and transthoracic echocardiography (TTE).

Methods: Explorative, prospective cohort study of adult patients undergoing first-time elective open-heart surgery at Department of Cardiothoracic Surgery at Copenhagen University Hospital, Rigshospitalet. One hundred consecutive patients enrolled in the Left Atrial Appendage Closure by Surgery-2 (LAACS-2) trial (NCT03724318) fulfilling the inclusion/exclusion criteria are included in the current study.

The study is approved by the Regional Committee on Health Research Ethics and all procedures followed are in accordance with the ethical standards of the responsible committee on human experimentation and with the Helsinki Declaration. All patients have signed informed consent prior to study enrolment.

Study timeline: One day prior to surgery; patient charts are reviewed, and ECG, wavECG and TTE are obtained. Anaesthesiologists will report AF occurred during intra-operative cardiac monitoring. Post-surgery cardiac rhythm monitoring is applied for at least 48 hours. At discharge patient charts and monitoring are reviewed for any occurrence of new-onset AF. Three months after surgery, patient charts are reviewed, and patients are invited for a single visit at Department of Cardiology at Frederiksberg hospital, where a control ECG, wavECG and TTE are performed. Deviation of the examination programme is accepted, i.e. patients can decline parts of the pre- or post-surgery examinations and remain in the study for follow-up.

Statistics: Precise sample size calculations are not applicable in this explorative study with LVDD in sp-ECG as primary end-point. One previous study has been performed where wavECG identified patients with LVDD in a population (n=188) referred for computed tomography coronary angiography.32 LA strain measured by TTE predicted POAF in 26 of 60 patients undergoing open-heart surgery for severe aortic stenosis.39 P-wave indices from pre-operative ECGs have predicted POAF in a population of 105 cardiac surgery patients.40 The study will include 100 consecutive patients undergoing open-heart surgery. Expected distribution of patients are n=10 for pre-surgery AF (corresponding to the incidence in the background population 1%6), POAF n= 35-50 (corresponding to the incidence in previous cardiac surgery populations; 35-50%1-3), and non-AF n=40-55 (remaining population). Follow-up will be three months.

Parametric statistics is used as normal distribution is expected (e.g. two-sample unpaired Student's t-test) and data is presented with mean ± standard deviation. All tests are two-sided and a P-value < 0.05 is considered statistically significant. IBM SPSS version 22 will be used for statistical analyses.

Inclusion Criteria:

• Age >18 years
• First time open-heart surgery
• Coronary artery bypass grafting and/or valve surgery
• Signed informed consent

Exclusion Criteria:

• Planned closure of the left atrial appendage
• Planned ablation for atrial fibrillation during surgery
• Ongoing inflammation or infection (including endocarditis)
• Connective tissue disease
• Ongoing cancer (not control)
• Pregnancy
• Follow-up not possible

Consecutive adult patients undergoing first-time elective open-heart surgery at Department of Cardiothoracic Surgery at Copenhagen University Hospital not fulfilling any exclusion criteria and who have signed informed consent

De-identified individual patient data for primary and secondary outcomes measures will be made available upon reasonable request.