Mapping of the LV was performed using three different high-density three-dimensional electroanatomic mapping (EAM) systems and respective multipolar catheters: 1) CARTO 3 and a 20 electrode PentaRay® catheter (Biosense Webster); 2) Ensite Precision and a 16 electrode HDGrid® catheter (Abbot); 3) Rhythmia and a 64 electrode Intellamap Orion® basket catheter (Boston Scientific). The choice of EAM was based on availability at our center. During transmitral mapping, the catheters were supported by an Agilis steerable sheath (Abbot, St Paul, MN).

Substrate mapping was performed during sinus rhythm or while pacing at 600 milliseconds from the proximal electrode of the coronary sinus catheter or from the right ventricle apex, according to operator discretion. Areas of scar (defined as bipolar voltage <1.5 mV) and dense scar (defined as bipolar voltage <0.5 mV) were automatically annotated through adjustment of the voltage thresholds. In patients mapped with the Intellamap Orion® basket catheter an adjusted voltage range of 0.2-0.8 mV was used.20 LAVAs were manually annotated on each substrate map. LAVAs were defined as sharp high-frequency ventricular potentials, distinct from the far-field ventricular electrogram occurring anytime during or after the ventricular electrogram.21 Ripple mapping®, SparkleMap® and Lumipoint® were employed as adjunctive mapping tools to improve the visualization of the scar channels, when the CARTO 3, Ensite Precision or Rhythmia EAM systems were used, respectively.16,22

Use of real-time image integration (RTII) was included into the workflow, if available. Cardiac multidetector computed tomography (MDCT) or magnetic resonance (MRI) images were processed using dedicated software (ADAS 3D, Galgo Medical S.L, Barcelona, Spain). Structural substrate was segmented on imaging as areas of wall thinning ≤5 mm on MDCT or delayed gadolinium enhancement on MRI (Figure 1).

Figure 1.

High-density multielectrode mapping and real-time image integration in a case of post-anterior myocardial infarction ventricular tachycardia. (A) Real-time image integration using ADAS 3D LV software, demonstrating an anterior and apical left ventricle thickening (≤5 mm). (B) A large anterior and apical scar (<1.5 mV) is shown on endocardial bipolar voltage map, revealing good correlation with RTII. (C) Local abnormal ventricular activities were recorded with a multielectrode catheter (Pentaray). (D) Voltage map during sinus rhythm. (E) Activation map during sinus rhythm depicting late activation on de left ventricle anterior medial segment. (F) Clinical ventricular tachycardia activation map, depicting a slow conducting isthmus on the LV anterior medial segment. (G) Mid-diastolic electrograms recorded with multielectrode catheter (Pentaray) on the left ventricle anterior medial segment, during ventricular tachycardia.

ECG: Electrogram; LV: Left ventricle; LAVAs: Local abnormal ventricular activities; RTII: Real time integration imaging.

(0,16MB).

After substrate mapping, programmed stimulation from the right ventricle apex at cycle lengths of 600 or 400 ms and up to three extrastimuli down to the ventricular effective refractory period or 200 ms was performed. Isoprenaline was not used for the purpose of inducibility. Induced VTs were classified as clinical or nonclinical based on comparison to the 12-lead electrocardiogram of the clinical VTs. When no 12-lead tracings were available, classification was based on the tachycardia cycle length on ICD tracings. An activation map using EAM and multielectrode catheters was obtained of any induced sustained and hemodynamically stable monomorphic VT. Activation maps during VT were automatically annotated. In regions of low amplitude signal where the far-field signal had been annotated, the annotation was manually adjusted for the local electrogram. Wherever possible, the VT protected isthmus location was confirmed by entrainment techniques. Also, pace mapping was allowed according to operator discretion.

Radiofrequency ablation was delivered with either a sensor enabled open-irrigated catheter with contact force monitoring Thermocool Smarttouch® (Biosense Webster) or a TactiCath Quartz® (Abbot) or without contact force monitoring Intellanav MIFI OI® (Boston Scientific) at 30-50 W endocardial and 25-35 W epicardial.

In the presence of sustained and tolerated monomorphic VT circuits, the VT protected isthmus was the first target for ablation, followed by other regions of LAVAs. In the event of noninducible or hemodynamically not tolerated VT, regions where LAVAs were identified were ablated. The earliest LAVAs were first ablated. After ablation, the LV was remapped with the original mapping catheter, and additional substrate ablation was performed if residual LAVAs were still identified.

Procedural endpoints were VT non-inducibility and LAVAs elimination. Inducibility testing after ablation was avoided if patients were unstable or if they had required ≥2 cardioversions for hemodynamically unstable VT during the procedure.

Venous and arterial sheaths were withdrawn after the procedure. Pericardial sheaths were retracted after obtaining a dry pericardial aspirate and after intra-pericardial corticoid instillation. ICD therapies were reactivated at the end of the procedure. Patients were monitored in hospital for at least 2 days. Antiarrhythmic-drug therapy was continued according to operator discretion.

Ventricular tachycardia in IHD is scar-related and results from an underlying re-entrant circuit around areas of myocardial patchy scar or at scar borders. Early coronary intervention for acute MI has led to smaller infarct size and as a consequence, smaller re-entrant circuits and shorter VT cycles. Also, patients with structural heart disease have LV systolic dysfunction, increasing hemodynamic instability during VT. As such, activation and entrainment mapping are only applicable to selected cases and a substrate-based approach using 3D EAM system data should be standard.11,30 In our study, all patients underwent a substrate-based approach. This is in contrast to the majority of previous RCTs, where mapping strategies were underreported (Table 3). A recent meta-analysis found that in patients with structural heart disease related mainly to IHD, there was a significantly lower risk of the composite primary outcome of VT recurrence and all-cause mortality among those undergoing substrate-based ablation.12

The substrate-based approach focuses on identification of low voltage areas and elimination of LAVAs in sinus rhythm or pacing. As such, this approach depends on factors affecting size and shape of electrograms, namely catheter electrodes size and interelectrode spacing. Using very small electrodes with small interelectrode distance in areas of very low voltage, previously considered “dense scar” when mapped using standard point-by-point mapping catheters, surviving myocardial bundles originate sharp and high-voltage signals that can be identified.13–15 In our experience, LAVAs were identified in all patients undergoing mapping using multielectrode catheters. In previous RCTs, the type of catheter used for map collection is underreported (Table 3). When reported, the majority of patients underwent a point-by-point mapping strategy solely with the ablation catheter. Multielectrode catheters enable better mapping resolution within areas of low voltage and improve identification of scar channels, thereby increasing ablation success.16,31 A prior study, adopting a functional substrate mapping strategy based on wavefront discontinuities assessed with these catheters, reported an 80% freedom from VT recurrence at 12±10 months in patients with IHD.18

Real-time image integration, including wall thinning on MDCT and delayed gadolinium enhancement on MRI, correlate with low voltage areas on EAM and the presence of LAVAs.32,33 Also, it has been reported that MRI-aided scar dechanneling is associated with a lower need for radio frequency delivery, higher non-inducibility rates after substrate ablation, and a higher VT-recurrence-free survival.34 In our study, RTII was not systematically used, but it was included into the workflow when available (11 patients, 17%).

Clear and specific procedural endpoints must be established. A meta-analysis of VT ablation in IHD found that the absence of inducible VT at the end of the procedure was associated with a substantially lower risk of recurrent VT.35 Elimination of LAVAs is feasible and safe and is associated with improved survival free from recurrent VT.21 Also, scar dechanneling results in high event-free survival despite the limited ablation extent required.36 Thus, in addition to achieving non-inducibility, elimination of LAVAs and/or scar dechanneling should be pursued. In our series, non-inducibility after ablation was achieved in 28 of the 47 tested patients (60%) and LAVAs elimination was achieved in 93.8% of patients. In previously published RCT there is a clear underreporting of procedural endpoints (Table 3).

Complications are a major concern in RCA of VT in IHD patients. In our series, procedure-related complications were observed in 8 patients (12.5%), while in previous RCT the percentage varied between 3.8% and 11.6% - Table 3. In our study, two patients (3.1%) with advanced HF died within 48 hours after the procedure, as a result of low-flow state. The majority of our patients (60.9%) were at intermediate risk of acute hemodynamic decompensation, as predicted by an increased baseline PAINESD risk score. Periprocedural acute hemodynamic decompensation may be related to the prolonged low-output state related do VT induction and mapping, as well to volume overload due to irrigated catheter ablation and general anesthesia and is associated with an increased risk of mortality. Proper identification of patients at higher risk using appropriate scores such has the PAINESD risk score, and prophylactic mechanical heart support may be valuable to reduce the risk of post procedural adverse events.23,37

In our study, after a mean follow-up of 25±18 months, 90% and 85% of patients were free from appropriate ICD shocks at one and two years, respectively. The proportion of patients experiencing VT storm decreased from 39% to 1.6%. Overall survival was 89% and 84% at one and two years, respectively. Although all-cause mortality was high in our population (23.4%), most deaths were related to advanced HF. A recent systematic review and meta-analysis of RCTs reported similar results, with a reduced risk of VT storm, ICD therapies and, in particular, ICD shocks, albeit no significant difference in all-cause or cardiovascular mortality with RCA for VT in patients with IHD.10 After the index RCA procedure, the mean follow-up ranged from 13–28 months, appropriate ICD shocks and overall mortality occurred in 91 of 375 (24.3%) and in 64 of 391 (16.4%) of the patients included in the ablation arm, respectively (Table 3). RCA procedural characteristics varied between the studies, namely LV access, substrate mapping, VT Inducibility and mapping, ablation and procedural outcomes.

Previous studies found that age ≥70, chronic obstructive pulmonary disease, LVEF <30%, the non-use of multielectrode catheters, non-use of RTII and incomplete substrate ablation were independent predictors of recurrent ICD shocks and overall mortality.31,38 We observed higher mortality in older individuals and among patients with renal failure, albeit these variables did not remain significant in multivariate analysis.

Certainly it will be desirable for future studies to have clear and defined procedural endpoints and outcomes to enable valid comparisons.