Personalised Ablation Strategies in AF

This study is a single centre prospective cohort study conducted at the Barts Heart Centre, St Bartholomew's Hospital. It will include patients that are undergoing and have provided informed consent for catheter for AF. These patients will be selected from outpatient clinics and referring physicians.

Patients undergoing catheter ablation for persistent AF will be included (\<24 months AF duration and no previous left atrial ablation). Patients in whom AF ablation is contraindicated or unable to consent for the procedure will be excluded. All patients included in these studies will have to provide informed consent for study participation.

For all the studies performed all patients will have a bipolar voltage map created in sinus rhythm. If patients are not in sinus rhythm at the start of the procedure, they will undergo DC cardioversion (DCCV) to achieve sinus rhythm. Low voltage zones (LVZs) will be defined as sites with a voltage \<0.5mV on bipolar voltage map. If patients have ≥30% of LVZs in the LA body excluding the PVs and mitral valve annulus, they will be classified as patients with underlying LVZs whilst those with a proportion of LVZs of \<30% will be defined as those without LVZs. The project study protocol undertaken will be decided based on whether the patient has underlying LVZs or not. Patients will therefore be in sinus rhythm at the start of the procedural protocol.

Patients with underlying LVZs (≥30% of LVZs in the LA body)

1. Hypothesis 1- Study 1- Develop and establish a method that will be used to evaluate CV dynamics in the left atrium (LA) and its relationship with LVZs utilising pacing protocols and multipolar catheters that are applicable to those routinely used in conventional ablation procedures. The method will be established to allow prospective identification of RDCV slowing sites. To achieve this, 30-seconds of unipolar electrograms will be obtained using multipolar catheters throughout the LA endocardially +/-epicardially to achieve optimal coverage using different pacing protocols in sinus rhythm at different LA sites. Twenty patients will be included. Sequential mapping with multipolar catheters will be employed to develop a method that can be utilised to prospectively create CV maps in the LA. These patients will then undergo conventional AF ablation and follow-up as per clinical grounds.
2. Hypothesis 2 Study 2- Assess the mechanistic importance of RDCV slowing sites. Once the CV methodology has been established, twenty patients will have CV maps created in sinus rhythm to identify RDCV slowing sites endocardially +/- epicardially. These sites will be tagged on the geometry created with the 3D mapping system. Patients will then have AF induced with atrial pacing using an anterograde curve and sensed extras +/- Isoprenaline and the AF inducibility score will be determined. Following a 5-minute waiting period to ensure rhythm stabilisation, unipolar electrogram recordings for 30-seconds will be obtained sequentially throughout the LA body with a multipolar catheter to ensure optimal LA coverage. Five minutes of coronary sinus (CS) unipolar signals will also be recorded simultaneously. The 30-seconds unipolar recordings will be used to performed spectral analysis using a custom written Matlab script and using a novel methodology that have shown to more accurately predict sites with an ablation response. Sites of highest dominant frequency (DF), fastest cycle length (CL) and regional DF gradients will be identified. The five minutes of CS unipolar signals will be used to determine CS CL variability and CS activation pattern stability: both novel markers that I have shown to be predictive of achieving a pre-defined ablation response and AF termination on ablation. RDCV slowing sites will then be ablated and the ablation response will be monitored including CL slowing and AF termination. Following ablation of all RDCV slowing sites if AF persists the unipolar recordings will be repeated. The unipolar recordings will be used to obtain spectral analysis parameters and analyse CS electrogram characteristics pre- and post-RDCV site ablation. Patients will then undergo DCCV. Attempts will then be made to re-induce AF as above to re-assess the AF inducibility score following RDCV slowing site ablation. Following this all patients will undergo conventional ablation and follow-up on clinical grounds.
3. Hypothesis 3 Study 3- Assess the impact autonomic modulation has on CV dynamics and RDCV slowing sites. Twenty patients will have CV maps and restitution curves created in sinus rhythm and RDCV slowing sites identified endocardially +/- epicarrdially as per the methodology developed in study 1. Patients will then undergo autonomic modulation with ganglionated plexi (GP) site stimulation, internal jugular vein stimulation and pharmacological means with Isoprenaline. With Isoprenaline, an isoprenaline infusion will be used to achieve ≥30% heart rate increase. GP site stimulation will be achieved through delivering high frequency stimulation through an ablation catheter in the LA. Each site tested will be tagged depending on the impact on the atrioventricular (AV) conduction. Sites will be tagged as GP sites if stimulation results in ventricular asystole or bradycardia. Internal jugular vein stimulation will be achieved through delivering high frequency stimulation through an ablation catheter CV maps and restitution curves will be re-created with autonomic modulation and the impact assessed particularly on the distribution of RDCV slowing sites. Patients will then undergo conventional ablation and follow-up on clinical grounds.

4. Hypothesis 4 Study 4- Prospectively perform GP site ablation and substrate modification guided by RDCV slowing sites whereby substrate ablation is limited to substrate with these electrical properties and evaluate freedom from AF/atrial tachycardia (AT) during 12 months follow-up. Forty patients will be included in this study. This is compatible to that of other proof of concept studies. All patients will have GP sites mapped as per methodology described earlier and CV maps created utilising the developed methodology to identify RDCV slowing sites. Patients will then have substrate modification as guided by RDCV slowing sites and GP site ablation. Patients will then have PV isolation using wide area circumferential ablations (WACAs). All patients will undergo clinical follow-up at 3, 6, 9 and 12 months, with 48-hour ambulatory Holter monitoring at 6 and 12 months. Clinical success will be defined as freedom from AF/AT lasting >30 seconds off anti-arrhythmic drugs.

   Patients without underlying LVZs (\<30% of LVZs in the LA body)

5. Hypothesis 5- Study 5- Determine the impact GP site ablation has on mechanisms in persistent AF in patients without underlying LVZs through utilising novel markers that have shown to predict ablation response and procedural outcomes. Twenty patients will be included in this study. AF will be induced with atrial pacing using an anterograde curve and sensed extras +/- Isoprenaline and the AF inducibility score will be determined. Following a 5-minute waiting period to ensure rhythm stabilisation unipolar sequential LA recordings and CS unipolar recordings will be performed to determine spectral analysis parameters and CS electrogram characteristics as per the methodology described earlier. Patients will then undergo GP site ablation and the ablation response monitored including CL slowing and AF termination. Monitoring the ablation response will also allow for better characterization of GP sites to determine if certain sites are mechanistically more important in AF. Following ablation of all GP sites if AF persists the unipolar recordings will be repeated. The unipolar recordings will be used to obtain spectral analysis parameters and analyse CS electrogram characteristics pre- and post-GP site ablation. Patients will then undergo DCCV. Attempts will then be made to re-induce AF as above to re-assess the AF inducibility score following GP site ablation. Following this all patients will undergo conventional ablation and follow-up on clinical grounds.

d) Hypothesis 6- Study 6- Prospectively targeting GP sites in addition to PV isolation and determine if therapeutically targeting sites of autonomic innervation results in improvement in procedural outcomes in this cohort of patients. Forty patients will undergo prospective ablation of GP sites followed by PV isolation with WACAs. All patients will undergo clinical follow-up at 3, 6, 9 and 12 months, with 48-hour ambulatory Holter monitoring at 6 and 12 months. Clinical success will be defined as freedom from AF/AT lasting >30 seconds off anti-arrhythmic drugs.

Patients with and without underlying LVZs f) Hypothesis 7 Study 7- In both cohorts of patients, with and without LVZs a subgroup of patients undergoing prospective guided ablation as per study ablation strategy will undergo cardiac magnetic resonance imaging (MRI) prior to their ablation. The aim is to ensure 20 patients in each cohort, with a total of 40 patients undergoing cardiac MRI. In patients with LVZs, the aim is to obtain 3D late gadolinium enhancement (LGE) MRI of the LA whilst in patients without LVZs the aim is to evaluate atrial EAT. To achieve this, one of the sequences that will be used is the 3D Dixon-LGE pulse sequence which has shown to allow simultaneous visualisation of LA fibrosis and atrial epicardial adipose tissue (EAT). In the patient cohort with LVZs, the 3D LGE cardiac MRI sequences will be imported into ADAS 3D software (ADAS 3D medical). This will be used to create a 3D shape of the segmented LA with the associated fibrosis. The segmented LA image will be derived through tracing the border of the LA. This 3D shape will be compatible to that obtained using the 3D mapping system. This will be imported into the 3D mapping system and co-registered to the existing bipolar voltage map. This will allow the characteristics of sites on the MRI derived LA map that correlate to RDCV slowing sites tagged on the bipolar voltage LA map created with the 3D mapping system to be made. The aim is to evaluate different MRI sequences and image intensity ratios to effectively evaluate the use of cardiac MRI to identify RDCV slowing sites. GP site mapping is time consuming and relay on additional mapping equipment. Therefore, I will evaluate whether in patients without underlying LVZs, atrial EAT identified on cardiac MRI could predict GP sites and thereby enable the use of a non-invasive modality to identify GP sites and aid in planning the ablation strategy and procedure. The 3D LGE cardiac MRI will also be reviewed to elicit for the presence of scar that has been missed on the bipolar voltage map. This will be used to enhance our understanding of the structural remodelling in these patients and relationship to AF.