Electrophysiological study prior to planned pulmonary valve replacement in patients with repaired tetralogy of Fallot

Ventricular arrhythmias (VAs) are the most common cause of death in patients with repaired Tetralogy of Fallot (rTOF). However, risk stratifying remains challenging. We examined outcomes following programmed ventricular stimulation (PVS) with or without subsequent ablation in patients with rTOF planned for pulmonary valve replacement (PVR).


| INTRODUCTION
Patients with repaired tetralogy of Fallot (rTOF) now commonly survive into mid-adulthood as a testament to the progress made in their medical and surgical management. 1 However, their risk of sudden cardiac death (SCD) due to ventricular arrhythmias (VAs) remains high (estimated at approximately around 0.1%-0.2% per year), 2,3 and accurately estimating their arrhythmic risk is difficult. 4 Historical markers, such as QRS complex duration and patients' clinical histories, have limited predictive value. 5 Given that VAs are the leading cause of death in patients with rTOF, 5 developing tailored strategies to predict-and ideally to modify-their arrhythmic risk has the potential to markedly improve their clinical outcomes.

Invasive programmed ventricular stimulation (PVS) in patients
with rTOF was first assessed by Khairy et al., suggesting its potential diagnostic and prognostic value, as well as improving our understanding of the underlying mechanisms of VA in this population. 6,7 VA are typically monomorphic, characterized by macro-reentrant circuits arising around patches and valve annuli with areas of scar and slow conduction contributing to the arrhythmic substrate and anatomical isthmuses (AI). 8,9 As the general understanding of these circuits improved, ablation strategies were used to create lines of block along critical parts of these AIs. 10 Pulmonary valve replacement (PVR) is commonly required for pulmonic valve insufficiency in patients with rTOF and may represent an opportunity to perform an electrophysiologic study (EPS) with PVS as it could prompt additional interventions, including surgical cryoablation, implantable cardioverter-defibrillator (ICD) implantation, and correction of residual lesions. 11 Indeed, this rationale led to our center adopting a protocol of systematically performing EPS and surgical or percutaneous ventricular ablation 12 in patients with rTOF planned for PVR. We herein report the procedural and clinical outcomes of this approach.

| METHODS
The present research adhered to the guiding principles of the Declaration of Helsinki and has been approved by our ethics committee on human research. Nonopposition for personal data used information was addressed to the patients (CNIL, MR004).
Patients did not object to the use of their personal data (regulatory authorization MR-004, CNIL). Consecutive patients aged ≥18 years with rTOF who were planned to undergo PVR and who were referred to our center between January 1, 2010, and December 12, 2018, were included. This period was selected to allow for at least 5 years of follow-up. All patients referred for PVR beneficiated from an EPS.

| Data collection
Medical records were reviewed for clinical and surgical details. ECGs were analyzed to measure QRS duration. Rhythm assessments were based on ECGs, Holter monitors, and ICD interrogations, when applicable. Cardiac magnetic resonance (CMR) and transthoracic echocardiography (TTE) studies were reviewed to determine biventricular function and volumes. The presence of established risk factors for VA were assessed, including QRS duration >180 ms, left ventricular ejection fraction (LVEF) <45%, right ventricular ejection fraction (RVEF) <35%, prior nonsustained ventricular tachycardia (NSVT), history of syncope, and prior palliative shunts. 13 These factors were considered along with the results of EPS and acute success of ablation, if applicable, in decisions regarding ICD implantation.
PREVENTION-ACHD risk score model: We decided to incorporate the risk score model developed by Vehmeijer et al. 14 to stratify the risk of SCD in a noninvasive manner.

| Cardiac magnetic resonance
CMR was performed during the electrophysiological study (EPS) with chamber volume assessments as well as a high-resolution late gadolinium enhancement (LGE) sequence. The processing of highresolution LGE was performed using MUSIC software (IHU Liryc, University of Bordeaux and Inria Sophia Antipolis, France).
In-plane and through-plane interpolation were applied to resample the imaging volume at a voxel size of 0.625 × 0.625 × 1 mm. Semiautomated tools were used to segment endocardial and epicardial contours of the right ventricle. Iterative histogram thresholding was applied to segment myocardial scar, the threshold being set three standard deviations above the mean intensity of normal myocardium, as measured in a normal septal area. Endocardial segmentation was used to compute a 3D surface mesh at high density ( >50 000 triangles), onto which myocardial scars were projected and scar areas were measured.
The ratio of total right ventricular (RV) to scar surface area was calculated. 15

| Electrophysiologic study and right ventricular mapping
All consecutive patients addressed for PVR underwent EPS. RV endocardial map was performed to identify scar and potential isthmuses as reported previously. 6,16 Procedures were performed under local anesthesia and two introducers were placed inside the right femoral vein. We used either the Carto 3 mapping system (Biosense Webster) or the RHYTHMIA system (Boston Scientific). In patients who underwent surgical cryoablation alone or who still had inducible VAs after percutaneous ablation, an EPS with PVS and electro-anatomical mapping was repeated 3 months post-PVR to guide decisions regarding ICD implantation.

| Definition of anatomical isthmuses (AI)
Four types of AIs have been described previously in rTOF. 8,17 Type 1 is bound by the tricuspid annulus and the RV incision, RV patch, or large transannular patch. Type 2 is bound superiorly by the pulmonary valve annulus and inferiorly by the infundibular patch or RV incision. Type 3 is bordered by the ventricular septal defect patch and the pulmonary valve annulus. Type 4 is defined between the tricuspid annulus and a larger VSD patch, usually when the defect is located in the muscular region.

| Pulmonary valve replacement
In most cases, PVR was performed surgically using a prosthetic valve or a homograft sized to the annulus. 17 When anatomical and technical conditions were appropriate, transcatheter valve implantation with a Melody transcatheter pulmonary valve (Medtronic) was performed.

| Ablation
Percutaneous catheter ablation was performed in patients with inducible VAs during EPS or with slow conduction within their AIs. 13 Before induction, a mapping catheter was placed on the isthmus 3. In case of VT induction, activation mapping was attempted when possible, with entrainment performed at isthmus 3. When mapping was not achievable, ablation targeted the isthmus involved deter-  Holter monitor recordings, ICD interrogations, or 12 lead ECG documentations were used to asses VAs. Mortality was evaluated from hospital records. The primary outcome of interest was a combined event including VT, appropriate ICD therapy, and SCD. Transcatheter pulmonary valve implantation was performed in 11% in both inducible and non-inducible groups. Additional clinical, imaging, and procedural details are described in Table 1. 3.1 | Noninvasive stratification using the PREVENTION ACHD score model

| ICD implantation
ICDs were implanted in five patients after EPS (7%). Four of these for primary prevention and one for secondary prevention. Patient 1 was inducible at baseline but was not inducible after ablation, however, he was implanted in primary prevention because of risk factors (important PVC and history of syncope for pre-syncope, which prompted their physicians to recommend an ICD for primary prevention (see Table 2). During follow-up, Patient 5 experienced inappropriate therapy and ultimately underwent device extraction because of a lead infection.
Since the device extraction, the patient has not experienced any syncope or documented arrhythmia thus far (i.e., 3 years). with a cycle length of 400 ms. He benefited from an ablation and ICD implantation. He has now been free from any arrhythmia for 2 years.

| Clinical outcomes
One noninducible patient with an ICD experienced inappropriate anti-tachycardia pacing for atrial flutter (Patient 5 in the ICD implantation section).

| DISCUSSION
In this study of 77 consecutive patients with rTOF who systematically underwent EPS before planned PVR, 18 (24%) had inducible VAs. VT was monomorphic except from one patient who exhibited a F I G U R E 2 Flow chart of all patients undergoing electrophysiological study before pulmonary valve replacement. EPS, electrophysiological study.
T A B L E 2 Summary of ICD implantation indications.   factors. Systematic EPS before PVR seems to be an efficient approach to selecting patients at risk for VAs. Some authors report that PVR in itself reduces the burden of arrhythmia by decreasing the volume and pressure of the right ventricle. However, we must remain cautious, since the substrate is still present in these patients, and it is, therefore, important to associate PVR with preemptive ablation. 22,23 It is important to specify that only patients in the inducible group experienced VT despite ablation and acute isthmus block. Therefore, a systematic control assessing isthmus block 3-6 months after the procedure might be recommended, especially if the patient was inducible before ablation and no ICD was implanted.

| LIMITATIONS
Our study has important limitations. The small number of events limit the strength of our conclusion. Referral bias could also have influenced our results. In our population, the average age at repair of TOF is quite old compared to the current population, which exhibits a lower risk. However, all patients referred for PVR had an EPS before, which helps reduce selection bias.

| CONCLUSION
Performing systematically EPS in patients with rTOF planned for PVR may help identifying patients at higher risk of VA. Ablation of culprit or slow-conducting AIs, either by catheter ablation at the time of EPS or surgical cryoablation at the time of PVR, is feasible and safe. In our study, inducibility after ablation was associated with medium-term arrhythmic complications, suggesting it may be a helpful marker of patients who may benefit from ICD implantation.