Atrial fibrillation (AF) is the most common heart arrhythmia. Many people do not have symptoms and are not aware they have AF. Others may feel dizzy, short of breath, feel very tired and become aware of a fast and irregular heart beat (palpitations). The main complication of AF is an increased risk of stroke and incidence of heart failure. There are two key aspects of treatment for AF. The first is protection from stroke, treated with oral anticoagulants. Treatment of AF is either by controlling the rate (frequency of contraction) or controlling the rhythm (restoring regular contraction). Rate-control is generally employed first with an intent to reduce the rate at which the lower pumping chambers contract and improve their efficiency. Appropriate medication is used and with this treatment strategy it is accepted that AF will be present as the long term heart rhythm. If symptoms persist despite medication the preferred strategy is to restore sinus rhythm (SR) and regular contraction in all pumping chambers of the heart. This can be done with electric shock treatment (DC cardioversion) together with long-term tablet medication, or by a more definitive 'cauterisation' therapy (catheter or thoracoscopic surgical ablation). In this study the investigators will study patients with symptomatic long standing persistent AF (continuous AF for more than 1 year) who have tried and failed drug and/or electrical therapy. At present the investigators do not know what the best ablation technique is for treating symptomatic, long-standing persistent AF (LSPAF). Catheter ablation (CA) is the most widely available invasive treatment available for AF. Thoracoscopic surgical ablation (SA) is not widely available but our hospitals have the expertise to conduct this procedure. CA has been shown to achieve modest degrees of success in restoring normal SR with the caveat that most patients do require 'multiple' procedures (usually two or three). SA offers patients an alternative choice of therapy with a keyhole surgical thoracoscopic) approach. It may have a higher single procedure success rate although there is the potential for greater complication rates. The investigators aim to examine this in detail to help us understand which approach might be better for managing LSPAF.

This is a multi-center randomised controlled study of catheter ablation compared with totally thoracoscopic surgical ablation.

The study population will be patients above the age of 18 with symptomatic long-standing persistent atrial fibrillation (≥1≤5 years) and good left ventricular function where at least one anti-arrhythmic drug (AAD) has failed, or where such drugs are contraindicated or not tolerated.

Subjects randomised to thoracoscopic Surgical Ablation will undergo minimally invasive, thoracoscopically assisted, surgical ablation to isolate the pulmonary veins (PVI) using a radiofrequency (RF) clamp device. Posterior wall will be isolated in a box fashion with cool rail bipolar RF device. This will include ganglionated plexi ablation +/- LAA excision/exclusion.

Catheter Ablation Group Patients will undergo pulmonary vein isolation and linear ablations in the left and right atrium.

There will be a 3 month blanking period and symptomatic atrial arrhythmia may have catheter ablation during the period of 12 month follow up. The primary end point of the study will be assessed by continuous cardiac recording through an internal loop recorder that will be inserted at the end of the index procedure. The analysis and reporting of the recordings will be performed by a blinded core lab.

Inclusion Criteria:

1. Age≥ 18 yrs.
2. LSPAF (> 12 months' duration)
3. EHRA>2
4. Left ventricular ejection fraction ≥ 40%
5. Suitable for either ablation procedure

Exclusion Criteria:

1. Left sided valvular heart disease with severity greater than mild
2. Contraindication to anticoagulation
3. Thrombus in the left atrium despite anticoagulation in therapeutic range
4. Cerebrovascular accident within the previous 6 months
5. Previous thoracic or cardiac surgery (including surgical interventions for AF)
6. Prior left atrial catheter ablation for AF
7. Unable to provide informed written consent
8. Active malignancy, another severe concomitant condition or presence of implanted intracardiac devices that would preclude patient undergoing study specific procedures
9. Pregnant or breast-feeding, or women of childbearing age not using a reliable contraceptive method.
10. Implanted non MRI compatible cardiac devices