TY - JOUR
AU1 - Battaglia,, Alberto
AU2 - Odille,, Freddy
AU3 - Magnin-Poull,, Isabelle
AU4 - Sellal,, Jean-Marc
AU5 - Hoyland,, Philip
AU6 - Hooks,, Darren
AU7 - Voilliot,, Damien
AU8 - Felblinger,, Jacques
AU9 - , de Chillou, Christian
AB - Abstract Aims Our study assesses the value of electrograms (EGMs) characteristics to identify a ventricular tachycardia (VT) isthmus entrance in patients with post-infarct VT. Post-infarct VTs are mostly due to a re-entrant circuit. A pacemapping (PM) approach is able to localize the VT isthmus during sinus rhythm. Limited data are available about the role of local EGMs in defining VT isthmus location. Methods and results Twenty consecutive patients (70% male) referred for post-infarct VT catheter ablation were included in the present study. The VT isthmus was defined according to the PM method. At each recording site, 10 characteristics of the local EGM were assessed to predict the location of the VT isthmus entrance. In total, 924 EGMs were acquired, of which 127 were located in the VT isthmus entrance. Logistic regression analysis showed that bipolar voltage, number of EGM positive peaks, and sQRS interval were independently associated with VT isthmus entrance location. The ROC curve best fitted the model at the cut-off 0.1641 (sensitivity 72%, specificity 75.2%, positive predictive value 31.3%, negative predictive value 94.4%, area under the curve 0.78, P < 0.001). Based upon these results, we developed an algorithm implemented in an automatic calculator to determine the likelihood that an EGM is located at a VT isthmus entrance. Conclusion Our study suggests that three EGM characteristics: bipolar voltage, number of positive peaks, and sQRS interval can successfully identify a VT isthmus entrance in post-infarct patients. Ventricular tachycardia, Ischaemic cardiomyopathy, Myocardial infarction, Re-entry circuit, Pacemapping, VT isthmus entrance, Ablation What’s new? Between 10 easy detectable local electrograms (EGMs) characteristics, bipolar voltage, number of EGM positive peaks, and sQRS interval were independently associated with ventricular tachycardia (VT) isthmus entrance location. The major advantage is related to the independency of this approach to any 12 leads electrocardiogram VT morphology. The automatic calculator will help rapid adoption of this method. The likelihood of a point to be located in a VT isthmus entrance can be quickly assessed by all. Introduction In most cases, a re-entrant mechanism is responsible for post-infarct ventricular tachycardias (VTs).1 The relationship of these re-entrant circuits to the ventricular scar resulting from myocardial infarction was elegantly defined in early surgical studies.2 Pathologically, the essential component of the ventricular scar that promotes re-entry is its histologic heterogeneity, with varying degrees of subendocardial myocardial fibre preservation within dense zones of fibrosis. These zones are detectable as low voltage areas during electroanatomic mapping in sinus rhythm. The spatial extent of low voltage areas normally far exceeds that of the critical isthmus. The critical isthmus of the re-entrant circuit is a channel of viable myocardial cells delimitated by two lateral boundaries. These barriers are anatomical obstacles or simply areas of functional conduction block.3 Identifying these areas of residual conduction deep inside a larger area of fibrosis is key to successful VT ablation. Different techniques, such as entrainment mapping or activation mapping during VT, have been proposed4,5 to identify the VT isthmus. More recently, substrate-based approaches have been investigated,6–8 especially in cases of poorly-tolerated VTs, in order to ablate the VT substrate during sinus rhythm. Recently, we demonstrated that pacemapping (PM) during sinus rhythm is a powerful technique to unmask the protected isthmus of post-infarct VT circuits.9,10 This technique consists of comparing the paced electrocardiogram (ECG) to the clinical VT. Typically, VT exit pacing sites yield a highly matched QRS morphology. Conversely, the VT entrance channel is always located in a poor correlation zone. An abrupt transition between a paced-QRS that matches the clinical VT (exit site) and a non-matched paced-QRS (entrance site) identifies the core of a VT isthmus. The current PM technique is based on the availability of a 12-lead ECG during VT. Therefore the absence of an induced VT is a limitation which is inherent to the PM technique. In such cases analysis of electrogram (EGM) characteristics may be useful to identify a protected VT isthmus. In particular, it has been shown that the conduction velocity is reduced at VT isthmus entrance.11 Therefore, we hypothesized that the characteristics of the EGM recorded during sinus rhythm at VT isthmus entrance may have some specific features. Our study aims to evaluate this hypothesis. Methods Patients Twenty consecutive patients (14 men, mean age 65 ± 12 years) referred for post-infarct VT catheter ablation in our institution between January 2015 and January 2017 and fulfilling the following inclusion criteria were included in the present retrospective study: (i) presence of a spontaneous sustained monomorphic VT documented by a 12-lead ECG which was (ii) subsequently inducible during the electrophysiological study, (iii) identification of the index VT isthmus by the PM technique (with pacing cycle length at 600 ms), and (iv) confirmation that radiofrequency (RF) catheter ablation lesions applied across the VT isthmus prevented further VT induction. A consort diagram of the enrolled sample is available in the Supplementary material online, Figure S1. Procedural details Electrophysiological study and left ventricular anatomical mapping Electrophysiological study, mapping, and catheter ablation were performed as previously described.5,12 For each procedure, endocardial signal and surface ECG were recorded using the BARD LabSystem PRO (CR Bard Inc., Lowell, MA, USA) and CARTO® (Biosense Webster, Diamond Bar, CA, USA). In brief, a bipolar catheter was inserted via the femoral vein and positioned at the right ventricular apex and used primarily for VT induction with the application of up to three extrastimuli during spontaneous rhythm (600 ms and then 400 ms basic cycle length). Failure to induce a sustained VT promoted the same protocol from another site (alternatively right ventricle outflow tract or left ventricle). The 12-lead ECG of the induced VT was captured on the PASO software of the CARTO® system. Sinus rhythm was then restored by overdrive pacing or external cardioversion before mapping. Access to the left ventricle was achieved anterogradely through trans-septal puncture. Anticoagulation was maintained with unfractioned heparin, targeting an ACT 250–350 s. Sedation was obtained with 10 mg IV nalbuphine, with incremental doses at 5 mg as necessary. A 7 Fr, 3.5 mm-irrigated tip catheter (THERMOCOOL®, Biosense-Webster, Johnson & Johnson) was used for both left ventricular volume reconstruction and mapping/ablation of VT circuits, in a respiration-gated mode. All procedures were performed with the Niobe® robotic magnetic navigation system (Stereotaxis Inc., St. Louis, MO, USA). Radiofrequency delivery was performed with Stockert-Cordis RF generator (Stockert GmbH, Freiburg, Germany). Pacemapping points acquisition Bipolar pacing (600 ms cycle length) with an output at twice diastolic threshold was performed at each pacing site, tagged as ‘Pacing Site’. At each pacing site, the corresponding paced 12-lead ECG was stored in the CARTO® system. In order to accurately define the critical isthmus, points were acquired with highest density within zones of (i) low voltage (bipolar voltage <1.5 mV), (ii) diastolic potentials, and (iii) double potentials or fractionated EGMs The mapping procedure was terminated when a sufficient density of points was obtained to understand the VT circuit. For each pacing site, the PASO™ software calculated the percentage of morphology match between the paced-QRS and the 12-lead ECG during VT (range −100% to +100%). These values were displayed as spectrum of colour on the anatomical model [the so-called ‘pacemapping (PM) map’]. Ventricular tachycardia circuit and ventricular tachycardia isthmus definition The interpretation of a PM map starts with the identification of the VT exit zone which corresponds to the endocardial region where paced-QRS ECG matches (correlation values above 90%) the VT morphology. Dense PM in areas just adjacent to the VT exit zone shows a gradual decrease in match, except at the entrance of the VT isthmus where, within a few millimetres of distance, an abrupt transition between a good and a poor correlation (correlation values usually lower than 30%) is observed. The VT critical isthmus was defined in between a zone of good correlation (exit site, red coloured) and the poorest correlation zone (entrance site, purple coloured) surrounded by a zone of intermediate correlation (lateral boundaries, green-yellow-blue coloured).9,13 All portions of each VT isthmus (lateral boundaries, entrance and exit channel, mid-isthmus line) were manually drawn on the 3D map. Radiofrequency ablation and endpoint Identification of the ablation target was based on the analysis of the colour-coded PM map. Linear RF lesions were placed to transect the VT isthmus. Radiofrequency energy was delivered in a temperature-controlled mode for 60–120 s at each ablation site with a maximal temperature/power target of 45°C/40 W. A programmed ventricular stimulation protocol was again performed. Non-inducibility of sustained monomorphic VT <270/min was considered as an acute procedural success. After ablation, patients were monitored for 72 h by telemetry. Transthoracic echocardiography was performed within 2 days after ablation, prior to discharge. Post hoc analysis All EGMs recorded during sinus rhythm at each pacing site in the scar area (defined by bipolar voltage <1.5 mV) and 2 cm around were analysed (Figure 1). Each EGM was classified (Figure 2) according to the following published criteria.13 Electrograms with sharp, biphasic, or triphasic spikes, amplitude >1.5 mV and duration <70 ms were defined as normal. Fragmented EGMs were defined by the presence of more than 4 sharp spikes with amplitude of less than 1.5 mV. Double potentials EGMs were defined by two discrete deflections separated by an isoelectric interval. Electrograms with late potentials EGMs were defined by a sharp component occurring after the end of the QRS. Figure 1 Open in new tabDownload slide Example of area of interest analysed. (A) Isthmus channel classification according to published criteria. (B) Caudal postero-anterior view showing the 3D shell of the left ventricle (left panel) and the anatomical 3D reconstruction according to the CT scan performed the day before the procedure (right panel). The area of interest was all the myocardial scar (defined by bipolar cut-off value < 1.5 mv) and a surrounding area of 2 cm. CT, computed tomography. Figure 1 Open in new tabDownload slide Example of area of interest analysed. (A) Isthmus channel classification according to published criteria. (B) Caudal postero-anterior view showing the 3D shell of the left ventricle (left panel) and the anatomical 3D reconstruction according to the CT scan performed the day before the procedure (right panel). The area of interest was all the myocardial scar (defined by bipolar cut-off value < 1.5 mv) and a surrounding area of 2 cm. CT, computed tomography. Figure 2 Open in new tabDownload slide Local electrogram analysis. (A) Variables measured for each electrogram. (B) Examples of electrogram classification. Figure 2 Open in new tabDownload slide Local electrogram analysis. (A) Variables measured for each electrogram. (B) Examples of electrogram classification. The 10 following EGM characteristics were assessed for statistical analysis: (1) bipolar voltage, (2) unipolar voltage, (3) time interval from QRS onset to EGM onset, (4) time interval from EGM offset to QRS offset, (5) EGM duration, (6) number of positive peaks, (7) presence of fragmentation, (8) presence of double potentials, (9) presence of late potentials, and (10) time interval between the pacing spike and the resulting QRS complex (sQRS interval). Statistical analysis Categorical variables are reported as counts and percentages, while continuous variables as median and interquartile range. Correlations between baseline characteristics stratified for isthmus entrance location points were tested in cross-tabulation tables by means of the Pearson χ2 or Fisher’s exact test and by one-way ANOVA, respectively for categorical and continuous variables. To test the independent correlation of these parameters with entrance isthmus location, all variables reporting a significant correlation at univariate analysis were included in a stepwise forward multivariate logistic regression model. Reliability of the obtained model was tested by ROC curve. A two-sided P-value <0.05 was considered statistically significant. All analyses were performed with SPSS 21.0 (SPSS Inc., Chicago, IL, USA). Calculator development A web-based graphical user interface was developed in order to predict whether a pacing site was likely to be in the VT isthmus entrance based on the aforementioned multivariate logistic regression model (link: http://freddy.odille.free.fr/software/VTIsthmusEntranceCalculator/VTIsthmusEntranceCalculator.html). Results The clinical characteristics of the baseline population are shown in Table 1. Table 1 Baseline population characteristics Variables . Sample population (20 patients) . Age (years), mean ± SD 65 ± 12 Male, n (%) 14 (70) Cardiovascular risk factors, n (%)  Arterial hypertension 15 (75)  Diabetes 5 (25)  Dyslipidaemia 11 (80)  Active smoker 9 (45) Body mass index (kg/m2) 29.9 ± 7.0 History of dysthyroidism, n (%) 5 (32) History of stroke/TIA (%), n (%) 4 (20) Peripheral vasculopathy, n (%) 5 (20) History of atrial fibrillation, n (%) 3 (15) Ejection fraction, mean ± SD 31.1 ± 12.2 CAD history  Years of CAD history, mean ± SD 17.1 ± 11.5  At least two vessels involved 7 (35)  Presence of totally occluded coronary artery, n (%) 9 (45) Previously positioned ICD 14 (70)  Of which in secondary prevention 11 (55)  ICD shocks in the 30 days before the procedure (mean) 2.1  ATP delivered in the 30 days before the procedure (mean) 39.0 Antiarrhythmic drug treatment at admission, n (%)  Amiodarone 12 (60)  Beta-blockers 19(95)  ACE-inhibitors/ARB 15 (75)  Oral anticoagulation 11(55)  Antiaggregant therapy 16 (80) Variables . Sample population (20 patients) . Age (years), mean ± SD 65 ± 12 Male, n (%) 14 (70) Cardiovascular risk factors, n (%)  Arterial hypertension 15 (75)  Diabetes 5 (25)  Dyslipidaemia 11 (80)  Active smoker 9 (45) Body mass index (kg/m2) 29.9 ± 7.0 History of dysthyroidism, n (%) 5 (32) History of stroke/TIA (%), n (%) 4 (20) Peripheral vasculopathy, n (%) 5 (20) History of atrial fibrillation, n (%) 3 (15) Ejection fraction, mean ± SD 31.1 ± 12.2 CAD history  Years of CAD history, mean ± SD 17.1 ± 11.5  At least two vessels involved 7 (35)  Presence of totally occluded coronary artery, n (%) 9 (45) Previously positioned ICD 14 (70)  Of which in secondary prevention 11 (55)  ICD shocks in the 30 days before the procedure (mean) 2.1  ATP delivered in the 30 days before the procedure (mean) 39.0 Antiarrhythmic drug treatment at admission, n (%)  Amiodarone 12 (60)  Beta-blockers 19(95)  ACE-inhibitors/ARB 15 (75)  Oral anticoagulation 11(55)  Antiaggregant therapy 16 (80) ACE-inhibitors, angiotensin-converting enzyme inhibitors; ARB, angiotensin II receptor blockers; CAD, coronary artery disease; ICD, implantable cardioverter-defibrillator; SD, standard deviation; TIA, transient ischaemic attack. Open in new tab Table 1 Baseline population characteristics Variables . Sample population (20 patients) . Age (years), mean ± SD 65 ± 12 Male, n (%) 14 (70) Cardiovascular risk factors, n (%)  Arterial hypertension 15 (75)  Diabetes 5 (25)  Dyslipidaemia 11 (80)  Active smoker 9 (45) Body mass index (kg/m2) 29.9 ± 7.0 History of dysthyroidism, n (%) 5 (32) History of stroke/TIA (%), n (%) 4 (20) Peripheral vasculopathy, n (%) 5 (20) History of atrial fibrillation, n (%) 3 (15) Ejection fraction, mean ± SD 31.1 ± 12.2 CAD history  Years of CAD history, mean ± SD 17.1 ± 11.5  At least two vessels involved 7 (35)  Presence of totally occluded coronary artery, n (%) 9 (45) Previously positioned ICD 14 (70)  Of which in secondary prevention 11 (55)  ICD shocks in the 30 days before the procedure (mean) 2.1  ATP delivered in the 30 days before the procedure (mean) 39.0 Antiarrhythmic drug treatment at admission, n (%)  Amiodarone 12 (60)  Beta-blockers 19(95)  ACE-inhibitors/ARB 15 (75)  Oral anticoagulation 11(55)  Antiaggregant therapy 16 (80) Variables . Sample population (20 patients) . Age (years), mean ± SD 65 ± 12 Male, n (%) 14 (70) Cardiovascular risk factors, n (%)  Arterial hypertension 15 (75)  Diabetes 5 (25)  Dyslipidaemia 11 (80)  Active smoker 9 (45) Body mass index (kg/m2) 29.9 ± 7.0 History of dysthyroidism, n (%) 5 (32) History of stroke/TIA (%), n (%) 4 (20) Peripheral vasculopathy, n (%) 5 (20) History of atrial fibrillation, n (%) 3 (15) Ejection fraction, mean ± SD 31.1 ± 12.2 CAD history  Years of CAD history, mean ± SD 17.1 ± 11.5  At least two vessels involved 7 (35)  Presence of totally occluded coronary artery, n (%) 9 (45) Previously positioned ICD 14 (70)  Of which in secondary prevention 11 (55)  ICD shocks in the 30 days before the procedure (mean) 2.1  ATP delivered in the 30 days before the procedure (mean) 39.0 Antiarrhythmic drug treatment at admission, n (%)  Amiodarone 12 (60)  Beta-blockers 19(95)  ACE-inhibitors/ARB 15 (75)  Oral anticoagulation 11(55)  Antiaggregant therapy 16 (80) ACE-inhibitors, angiotensin-converting enzyme inhibitors; ARB, angiotensin II receptor blockers; CAD, coronary artery disease; ICD, implantable cardioverter-defibrillator; SD, standard deviation; TIA, transient ischaemic attack. Open in new tab The clinical VT was inducible in all patients. Inducted VTs were assessed, according to the PM technique, in each and every patient. Mean VTs cycle length was 377 ± 64 ms. As a result, 20 VT isthmuses were identified. Mean isthmus dimensions were 39.5 ± 10.6 mm (length) and 28.2 ± 10.3 mm (width). Radiofrequency application (mean 1276 ± 797 s) across the mid-isthmus line prevented clinical VT induction in all patients. Sixteen out of 20 (80%) patients were non-inducible at the end of the procedure. In the remaining four cases, a non-clinical non-sustained polymorphic VT (mean cycle length 276 ± 19 ms) was inducible. In total, 924 PM points were analysed, of which 127 (14%) were located at a VT isthmus entrance. As shown in Table 2, eight of the 10 EGMs characteristics were significantly different between VT isthmus entrance and other zones. Hence, EGMs located in the VT isthmus entrance showed lower bipolar voltage (0.52 ± 0.54 vs. 1.44 ± 1.64 mV, P < 0.001), lower unipolar voltage (3.18 ± 1.93 vs. 5.33 ± 3.76 mV, P < 0.001), and had more positive peaks (6.4 ± 3.6 vs. 4.4 ± 2.6, P < 0.001). Fragmented EGM and EGM with late potential were more frequently located in the VT isthmus entrance (64% vs. 35%, P < 0.001 and 14% vs. 6%, P = 0.002, respectively). Finally, the sQRS interval was significantly longer when pacing at the VT isthmus entrance (89.1 ± 34.9 ms vs. 49.9 ± 31.3 ms, P < 0.001). Table 2 Local electrograms characteristics analysed according to VT isthmus entrance location . Entrance isthmus site . Other site . P-value univariate . Entrance P-value multivariate . Odd ratio (95% CI) . Number (%) 127 797 Bipolar voltage (mV), mean ± SD 0.52 ± 0.54 1.44 ± 1.64 <0.001 0.005 0.581 (0.397–0.851) Unipolar voltage (mV), mean ± SD 3.18 ± 1.93 5.33 ± 3.76 <0.001 NS Stimulus to QRS interval (ms), mean ± SD 89.1 ± 34.9 49.9 ± 31.3 <0.001 <0.001 1.014 (1.009–1.020) EGM duration (ms), mean ± SD 144.3 ± 68.9 117.7 ± 68.9 <0.001 NS Delay QRS onset—EGM onset (ms), mean ± SD 26.0 ± 38.4 21.9 ± 36.2 0.24 Delay EGM offset—QRS offset (ms), mean ± SD 36.5 ± 63.9 12.8 ± 64.9 <0.001 NS Number of EGM positive deflections, mean ± SD 6.4 ± 3.6 4.4 ± 2.6 <0.001 0.021 1.089 (1.013–1.170) Fragmented EGM 81 (64%) 277 (35%) <0.001 NS Double potentials EGM 8 (6%) 30 (4%) 0.18 EGM with late potentials 18 (14%) 51 (6%) 0.002 NS . Entrance isthmus site . Other site . P-value univariate . Entrance P-value multivariate . Odd ratio (95% CI) . Number (%) 127 797 Bipolar voltage (mV), mean ± SD 0.52 ± 0.54 1.44 ± 1.64 <0.001 0.005 0.581 (0.397–0.851) Unipolar voltage (mV), mean ± SD 3.18 ± 1.93 5.33 ± 3.76 <0.001 NS Stimulus to QRS interval (ms), mean ± SD 89.1 ± 34.9 49.9 ± 31.3 <0.001 <0.001 1.014 (1.009–1.020) EGM duration (ms), mean ± SD 144.3 ± 68.9 117.7 ± 68.9 <0.001 NS Delay QRS onset—EGM onset (ms), mean ± SD 26.0 ± 38.4 21.9 ± 36.2 0.24 Delay EGM offset—QRS offset (ms), mean ± SD 36.5 ± 63.9 12.8 ± 64.9 <0.001 NS Number of EGM positive deflections, mean ± SD 6.4 ± 3.6 4.4 ± 2.6 <0.001 0.021 1.089 (1.013–1.170) Fragmented EGM 81 (64%) 277 (35%) <0.001 NS Double potentials EGM 8 (6%) 30 (4%) 0.18 EGM with late potentials 18 (14%) 51 (6%) 0.002 NS CI, confidence interval; EGM, electrogram; SD, standard deviation; NS, non significant. Open in new tab Table 2 Local electrograms characteristics analysed according to VT isthmus entrance location . Entrance isthmus site . Other site . P-value univariate . Entrance P-value multivariate . Odd ratio (95% CI) . Number (%) 127 797 Bipolar voltage (mV), mean ± SD 0.52 ± 0.54 1.44 ± 1.64 <0.001 0.005 0.581 (0.397–0.851) Unipolar voltage (mV), mean ± SD 3.18 ± 1.93 5.33 ± 3.76 <0.001 NS Stimulus to QRS interval (ms), mean ± SD 89.1 ± 34.9 49.9 ± 31.3 <0.001 <0.001 1.014 (1.009–1.020) EGM duration (ms), mean ± SD 144.3 ± 68.9 117.7 ± 68.9 <0.001 NS Delay QRS onset—EGM onset (ms), mean ± SD 26.0 ± 38.4 21.9 ± 36.2 0.24 Delay EGM offset—QRS offset (ms), mean ± SD 36.5 ± 63.9 12.8 ± 64.9 <0.001 NS Number of EGM positive deflections, mean ± SD 6.4 ± 3.6 4.4 ± 2.6 <0.001 0.021 1.089 (1.013–1.170) Fragmented EGM 81 (64%) 277 (35%) <0.001 NS Double potentials EGM 8 (6%) 30 (4%) 0.18 EGM with late potentials 18 (14%) 51 (6%) 0.002 NS . Entrance isthmus site . Other site . P-value univariate . Entrance P-value multivariate . Odd ratio (95% CI) . Number (%) 127 797 Bipolar voltage (mV), mean ± SD 0.52 ± 0.54 1.44 ± 1.64 <0.001 0.005 0.581 (0.397–0.851) Unipolar voltage (mV), mean ± SD 3.18 ± 1.93 5.33 ± 3.76 <0.001 NS Stimulus to QRS interval (ms), mean ± SD 89.1 ± 34.9 49.9 ± 31.3 <0.001 <0.001 1.014 (1.009–1.020) EGM duration (ms), mean ± SD 144.3 ± 68.9 117.7 ± 68.9 <0.001 NS Delay QRS onset—EGM onset (ms), mean ± SD 26.0 ± 38.4 21.9 ± 36.2 0.24 Delay EGM offset—QRS offset (ms), mean ± SD 36.5 ± 63.9 12.8 ± 64.9 <0.001 NS Number of EGM positive deflections, mean ± SD 6.4 ± 3.6 4.4 ± 2.6 <0.001 0.021 1.089 (1.013–1.170) Fragmented EGM 81 (64%) 277 (35%) <0.001 NS Double potentials EGM 8 (6%) 30 (4%) 0.18 EGM with late potentials 18 (14%) 51 (6%) 0.002 NS CI, confidence interval; EGM, electrogram; SD, standard deviation; NS, non significant. Open in new tab The multivariate model identified three EGM parameters as independent predictors of isthmus entrance: (i) EGM bipolar voltage, (ii) number of EGM positive peaks, and (iii) sQRS duration (Table 2). The reliability of this model was tested by a ROC curve analysis (Figure 3). The ROC curve best fitted the model at the cut-off 0.1641 (sensitivity 72%, specificity 75.2%, positive predictive value 31.3%, negative predictive value 94.4%, area under the curve 0.78, P < 0.001). Based upon the logistic regression analysis, we developed an automatic calculator (link: http://freddy.odille.free.fr/software/VTIsthmusEntranceCalculator/VTIsthmusEntranceCalculator.html). The user is asked to enter values of bipolar voltage, sQRS interval, and EGM number of positive deflections of a given EGM (Figure 4). Then the calculator returns the score given by the model along with the likelihood that this EGM is located in the VT isthmus entrance. Figure 5 shows two examples of maps with EGM assigned as ‘VT isthmus entrance points’ or ‘non-VT isthmus entrance points’ by the calculator. Figure 3 Open in new tabDownload slide ROC curve testing the logistic regression model. The ROC curve best fitted the model at the cut-off 0.1641 (sensibility 72%, specificity 75.2, positive predictive value 31.3%, negative predictive value 94.4%, AUC 0.78, P < 0.001). AUC, area under the curve; ROC, receiver operating characteristic. Figure 3 Open in new tabDownload slide ROC curve testing the logistic regression model. The ROC curve best fitted the model at the cut-off 0.1641 (sensibility 72%, specificity 75.2, positive predictive value 31.3%, negative predictive value 94.4%, AUC 0.78, P < 0.001). AUC, area under the curve; ROC, receiver operating characteristic. Figure 4 Open in new tabDownload slide Different EGM characteristics and sQRS intervals according to PM site location. (A) CT scan heart three-dimensional reconstruction. (B) Bipolar left ventricular voltage map showing wide infero-lateral myocardial scar (white line). (C) PM map showing exit site (red coloured) based on PM points with a VT correlation greater than 95%, next to the entrance channel (violet coloured) based on PM points with less than 30% VT correlation. Lateral borders delimitated by areas of intermediate correlation. Yellow stars highlight pacing points. (D) Exit site located PM point. Local EGM characteristics: bipolar voltage 1.36 mV, number of positive peaks 2, sQRS interval 0 ms. (E) Entrance site located PM point. Local EGM characteristics: bipolar voltage 0.14 mV, number of positive peaks 6, sQRS interval 182 ms. CT, computed tomography; EGM, electrogram; PM, pacemapping; VT, ventricular tachycardia. Figure 4 Open in new tabDownload slide Different EGM characteristics and sQRS intervals according to PM site location. (A) CT scan heart three-dimensional reconstruction. (B) Bipolar left ventricular voltage map showing wide infero-lateral myocardial scar (white line). (C) PM map showing exit site (red coloured) based on PM points with a VT correlation greater than 95%, next to the entrance channel (violet coloured) based on PM points with less than 30% VT correlation. Lateral borders delimitated by areas of intermediate correlation. Yellow stars highlight pacing points. (D) Exit site located PM point. Local EGM characteristics: bipolar voltage 1.36 mV, number of positive peaks 2, sQRS interval 0 ms. (E) Entrance site located PM point. Local EGM characteristics: bipolar voltage 0.14 mV, number of positive peaks 6, sQRS interval 182 ms. CT, computed tomography; EGM, electrogram; PM, pacemapping; VT, ventricular tachycardia. Figure 5 Open in new tabDownload slide Examples of calculator guided map. (A) PM map showing infero-lateral VT isthmus with exit channel located inferiorly. Yellow points show pacing points. (B) Calculator guided map showing entrance located points (red coloured). Pacing points not located in VT isthmus entrance are green coloured. Note that the calculator identify three PM sites in the proximal exit channel. (C) PM map showing an anterior wall VT isthmus with exit channel located inferiorly. Yellow points show pacing points. (D) Calculator guided map showing entrance located points (red coloured). Pacing points not located in VT isthmus entrance are green coloured. Note the identification of some points in the VT exit channel. PM, pacemapping; VT, ventricular tachycardia. Figure 5 Open in new tabDownload slide Examples of calculator guided map. (A) PM map showing infero-lateral VT isthmus with exit channel located inferiorly. Yellow points show pacing points. (B) Calculator guided map showing entrance located points (red coloured). Pacing points not located in VT isthmus entrance are green coloured. Note that the calculator identify three PM sites in the proximal exit channel. (C) PM map showing an anterior wall VT isthmus with exit channel located inferiorly. Yellow points show pacing points. (D) Calculator guided map showing entrance located points (red coloured). Pacing points not located in VT isthmus entrance are green coloured. Note the identification of some points in the VT exit channel. PM, pacemapping; VT, ventricular tachycardia. Discussion We assessed the value of several EGM parameters to identify a VT isthmus entrance in patients with post-infarct VT. Historically, EGM fragmentation was firstly targeted. Electrogram fractionation is generated by slowed and non-uniform conduction due to the transverse uncoupling of myocytes by fibrosis. Bogun et al.14 showed ‘late potentials’ (defined as EGM ending after the end of the QRS without diastolic activity) in 57 out of 94 ablation sites where concealed entrainment was performed during VT. Only 31 ‘late potential’ sites were observed out of 51 sites where RF application interrupted the VT (sensitivity 61%, positive predictive value 54%). Despite the selection bias that only concealed entrainment sites were assessed, the ‘late potential’ specificity and the negative predictive value were low (39% and 46%, respectively). Consequently, targeting fractionated signals alone is not an optimal strategy to correctly guide VT ablation. The effectiveness of the local bipolar voltage scanning to predict a VT isthmus location has also been investigated. In patients without known cardiovascular disease, 95% of EGMs have a bipolar voltage >1.55 mV. Accordingly, 1.5 mV has become the established bipolar voltage cut-off for identifying myocardial scar.15 Relative voltage preservation within denser regions of scar is a hallmark of central conducting channels that may form anatomically constrained diastolic isthmuses during VT. These zones can be seen by altering the voltage representation on colour isopotential maps. In the original study by Arenal et al.,7 electroanatomic maps from 26 patients with post-infarction cardiomyopathy were analysed for presence of voltage channels and their relationship with clinical or induced VTs. The authors found that 20 out of 23 (87%) identified channels were associated with a VT isthmus. They used five pre-specified voltage cut-offs with only a 0.01 mV difference between the upper and lower voltage limits. It is unclear whether such a small difference is truly important. These encouraging results were not confirmed thereafter. Mountantonakis et al.,16 in a larger sample size of patients, showed that voltage channels could be identified in 88% of patients with post-infarction VT. In his study, only 11 out of 25 induced VT were maintained by an identified channel (sensitivity 44%). Moreover, only 11 out of 37 channels contained a clinical VT isthmus (positive predicting value 29.7%). Nevertheless, the voltage channel ability to accurately guide ablation is poor. The sQRS interval may provide additional information about VT isthmus location. By pacing in low voltage areas the time required for the electrical impulse to reach the healthy surrounding myocardial tissue may be long. Thus slow conducting zones are often unmasked by the presence of a long sQRS interval. Stevenson et al.,17 in his sample of 13 post-infarct VT patients, considered 75 PM sites showing fragmented EGM during sinus rhythm. A sQRS >40 ms was observed in 41 out of 75 sites (55%). Participation of each site in a re-entry circuit was then evaluated by entrainment techniques during induced VT. Out of 28 points showing concealed entrainment during VT, 21 showed sQRS >40 ms in sinus rhythm (sensitivity 75%). Between 41 points with sQRS >40 ms only 21 were correctly located in a re-entry circuit (positive predictive value 51%). The resulting specificity and negative predictive values were 51% and 79%, respectively (considering the selection bias the only fragmented EGM sites were assessed in this study). A subsequent study,18 based on a larger sample of 829 PM points, confirmed the higher prevalence of a slow conducting zone in a re-entry circuit. By only considering the sQRS >40 ms the sensitivity/specificity of a point to be located in a VT isthmus were 68% and 53%, respectively. The positive/negative predictive value were 33% and 83%. Sites with prolonged sQRS intervals during sinus rhythm pacing were frequently associated with re-entry circuits. However, sQRS interval measurements are not sufficient to precisely identify the VT isthmus. None of these three variables (EGM fragmentation, local bipolar voltage, and the sQRS interval), taken correctly guide VT isthmus localization. Multiple EGM parameters evaluation is a promising approach to correctly localize slow conducting zones responsible for VT. The first multi-parameters EGM evaluation showed promising results. Brunckhorst et al.19 simultaneously analysed EGM duration, EGM bipolar voltage, and sQRS interval. This model, based on 931 PM points, showed a sensitivity of 80% and a specificity of 51% (positive and negative predictive values of this model were not disclosed). This proposed model showed good agreement with ours. Recently, Nayyar et al.,20 analysed 1292 PM points and reported that EGM activation time and EGM entropy accurately identified VT channels during sinus rhythm. This model was of impressive performance (sensitivity 86%, specificity 100%, positive predictive value 93%, and negative predictive value 100%). Electrogram parameters considered in both experiences are quite similar to ours. Main differences arise from the uniform pacing cycle length used in our study. Differences in reported model performance may be linked to different definitions of the target zones. In our study, VT isthmus was clearly anatomically defined, including mid-isthmus lines and lateral boundaries. Using this clear anatomical definition we were able to consider only PM points performed in the isthmus entrance: the anatomical area above the mid-isthmus line. In the Nayyar et al.20 experience a target area named ‘channel’ was used. Its relation with the VT isthmus was not clearly stated. A VT channel is usually located between the mid-isthmus line and the isthmus exit. It is not clearly demonstrated if a single VT channel ablation may effectively suppress all possible pathways inside a VT isthmus whose width may be wider. In fact, in their experience, several VT channels were identified in each patient leading to a wider ablation area. Patients enrolled in our experience showed only one VT isthmus enhancing our conclusion strength. Our study first demonstrated that three EGM parameters (local bipolar voltage, number of EGM positive peaks, and sQRS interval) are strong independent predictors of VT isthmus entrance. All 924 pacemap points were acquired at the same pacing cycle length of 600 ms. Each point was blindly reviewed by two independent Electrophysiologists. Each isthmus found was precisely anatomically defined. Mid-isthmus lines and lateral boundaries were clearly defined. To our knowledge, our study constitutes the largest sample of PM maps performed with a uniform pacing cycle length reported in the literature. Using an in-house calculator, the likelihood of a point to be located in a VT isthmus entrance can be quickly assessed. Our methodology is more rigorous and our performance statistics are as high as other published studies (sensitivity 72%, specificity 75%). The observed low positive predictive value of 31.3% is comparable with published studies considering an un-biased series of PM points. The low positive predictive value derives from the calculator inability to differentiate points located on the mid-isthmus line from those in the proximal VT isthmus exit channel. These points, statistically considered as false positive points, are anyway located inside the VT isthmus, the target area of each VT ablation. The negative predictive value of 94.4% of our model further strengthens its clinical utility. The calculator can be used in cases where clinical VT is non-inducible. Our approach is particularly easy and quick to apply, thus facilitating rapid adoption. Integration of this calculator in mapping software could refine the definition of the critical isthmus entrance and enhance the efficacy of VT ablation. This approach leads to significant clinical amendments. Limitations This is a non-randomized study. Consecutive cases were analysed to limit the consequences of selection bias. A routine LP or LAVA ablation was not performed in this series of patients. Electrogram information were acquired only in sinus rhythm. Further prospective studies need to be performed to validate the utility of the model derived from the current study. Conclusion The local bipolar voltage, EGM characteristics, and sQRS interval are strongly related with VT isthmus entrance. We have identified a selection criteria determining whether a point is located in the isthmus entrance. This approach does not rely on the induction of any VT. The criteria has been integrated in a calculator, thus facilitating adoption. Conflict of interest: none declared. References 1 Klein H , Karp RB , Kouchoukos NT , Zorn GL Jr , James TN , Waldo AL. Intra-operative electrophysiologic mapping of the ventricle during sinus rhythm in patients with a previous myocardial infarction: identification of the electrophysiologic substrate of ventricular arrhythmias . Circulation 1982 ; 66 : 847 – 53 . Google Scholar Crossref Search ADS PubMed WorldCat 2 de Bakker JM , van Capelle FJ , Janse MJ , Wilde AA , Coronel R , Becker AE et al. Reentry as a cause of ventricular tachycardia in patients with chronic ischemic heart disease: electrophysiologic and anatomic correlation . Circulation 1988 ; 77 : 589 – 606 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Stevenson WG , Khan H , Sager P , Saxon LA , Middlekauff HR , Natterson PD et al. Identification of reentry circuit sites during catheter mapping and radiofrequency ablation of ventricular tachycardia late after myocardial infarction . Circulation 1993 ; 88 : 1647 – 70 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Stevenson WG , Friedman PL , Sager PT , Saxon LA , Kocovic D , Harada T et al. Exploring postinfarction reentrant ventricular tachycardia with entrainment mapping . J Am Coll Cardiol 1997 ; 29 : 1180 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 5 de Chillou C , Lacroix D , Klug D , Magnin-Poull I , Marquié C , Messier M et al. Isthmus characteristics of reentrant ventricular tachycardia after myocardial infarction . Circulation 2002 ; 105 : 726 – 31 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Marchlinski FE , Callans DJ , Gottlieb CD , Zado E. Linear ablation lesions for control of unmappable ventricular tachycardia in patients with ischemic and nonsichemic cardiomyopathy . Circulation 2000 ; 101 : 1288 – 96 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Arenal A , del Castillo S , Gonzalez-Torrecilla E , Atienza F , Ortiz M , Jimenez J et al. Tachycardia-related channel in the scar tissue in patients with sustained monomorphic ventricular tachycardias influence of the voltage scar definition . Circulation 2004 ; 110 : 2568 – 74 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Jaïs P , Maury P , Khairy P , Sacher F , Nault I , Komatsu Y et al. Elimination of local abnormal ventricular activities: a new end point for substrate modification in patients with scar-related ventricular tachycardia . Circulation 2012 ; 125 : 2184 – 96 . Google Scholar Crossref Search ADS PubMed WorldCat 9 de Chillou C , Groben L , Magnin-Poull I , Andronache M , MagdiAbbas M , Zhang N et al. Localizing the critical isthmus of post-infarct ventricular tachycardia: the value of pace-mapping during sinus rhythm . Heart Rhythm 2014 ; 11 : 175 – 81 . Google Scholar Crossref Search ADS PubMed WorldCat 10 de Chillou C , Sellal J-M , Magnin-Poull I. Pace mapping to localize the critical isthmus of ventricular tachycardia . Card Electrophysiol Clin 2017 ; 9 : 71 – 80 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Anter E , Tschabrunn CM , Buxton AE , Josephson ME. High-resolution mapping of postinfarction reentrant ventricular tachycardia: electrophysiological characterization of the circuit . Circulation 2016 ; 134 : 314 – 27 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Pauriah M , Cismaru G , Magnin-Poull I , Andronache M , Sellal J-M , Schwartz J et al. A stepwise approach to the management of postinfarct ventricular tachycardia using catheter ablation as the first-line treatment: a single-center experience . Circ Arrhythm Electrophysiol 2013 ; 6 : 351 – 6 . Google Scholar Crossref Search ADS PubMed WorldCat 13 de Chillou C , Magnin-Poull I , Andronache M , Sacher F , Groben L , Abdelaal A et al. Showing up channels for postinfarct ventricular tachycardia ablation . Pacing Clin Electrophysiol 2012 ; 35 : 897 – 904 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Bogun F , Bender B , Li YG , Groenefeld G , Hohnloser SH , Pelosi F et al. Analysis during sinus rhythm of critical sites in reentry circuits of postinfarction ventricular tachycardia . J Interv Card Electrophysiol 2002 ; 7 : 95 – 103 . 7 Google Scholar Crossref Search ADS PubMed WorldCat 15 Koa-Wing M , Ho SY , Kojodjojo P , Peters NS , Davies DW , Kanagaratnam P. Radiofrequency ablation of infarct scar-related ventricular tachycardia: correlation of electroanatomical data with post-mortem histology . J Cardiovasc Electrophysiol 2007 ; 18 : 1330 – 3 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Mountantonakis SE , Park RE , Frankel DS , Hutchinson MD , Dixit S , Cooper J et al. Relationship between voltage map “channels” and the location of critical isthmus sites in patients with post-infarction cardiomyopathy and ventricular tachycardia . J Am Coll Cardiol 2013 ; 61 : 2088 – 95 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Stevenson WG , Sager PT , Natterson PD , Saxon LA , Middlekauff HR , Wiener I. Relation of pace mapping QRS configuration and conduction delay to ventricular tachycardia reentry circuits in human infarct scars . J Am Coll Cardiol 1995 ; 26 : 481 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Brunckhorst CB , Stevenson WG , Soejima K , Maisel WH , Delacretaz E , Friedman PL et al. Relationship of slow conduction detected by pace-mapping to ventricular tachycardia re-entry circuit sites after infarction . J Am Coll Cardiol 2003 ; 41 : 802 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Brunckhorst CB , Delacretaz E , Soejima K , Jackman WM , Nakagawa H , Kuck KH et al. Ventricular mapping during atrial and right ventricular pacing: relation of electrogram parameters to ventricular tachycardia reentry circuits after myocardial infarction . J Interv Card Electrophysiol 2004 ; 11 : 183 – 91 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Nayyar S , Kuklik P , Ganesan AN , Sullivan TR , Wilson L , Young GD et al. Development of time- and voltage-domain mapping (V-T-mapping) to localize ventricular tachycardia channels during sinus rhythm . Circ Arrhythm Electrophysiol. 2016 Dec;9(12). pii: e004050. OpenURL Placeholder Text WorldCat Author notes Alberto Battaglia, Freddy Odille, Jacques Felblinger and Christian de Chillou contributed equally to this work. Published on behalf of the European Society of Cardiology. All rights reserved. © The Author(s) 2019. For permissions, please email: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - An efficient algorithm based on electrograms characteristics to identify ventricular tachycardia isthmus entrance in post-infarct patients
JF - Europace
DO - 10.1093/europace/euz315
DA - 2020-01-01
UR - https://www.deepdyve.com/lp/oxford-university-press/an-efficient-algorithm-based-on-electrograms-characteristics-to-NCjhtZT5sa
SP - 109
VL - 22
IS - 1
DP - DeepDyve
ER -