Electrocardiographic patterns of ventricular pre-excitation in dogs with right-sided accessory pathways

The aim of the study was to describe the electrocardiographic features of ventricular pre-excitation (VPE) patterns characterized by the presence of delta (δ) wave, short P-δQRS interval, wide δQRS complexes in dogs with right-sided accessory pathways. Twenty-six dogs with a confirmed accessory pathways (AP) via electrophysiological mapping were included. All dogs underwent a complete physical examination, 12-lead ECG, thoracic radiography, echocardiographic examination and electrophysiologic mapping. The AP were located in the following regions: right anterior, right posteroseptal, right posterior. The following parameters were determined: P-δQRS interval, δQRS duration, δQRS axis, δQRS morphology, δ-wave polarity, Q-wave, R-wave, R’-wave, S-wave amplitude, and R/S ratio. In lead II, the median δQRS complex duration was 82.4 (IQR 7.2) and the median P-δQRS interval duration was 54.6 (IQR 4.2) msec. The median δQRS complex axis in the frontal plane was: + 68° (IQR 52.5) for right anterior APs, − 24 ° (IQR 24) for right postero-septal APs, − 43.5 ° (IQR 27.25) for right posterior APs (P = 0.007). In lead II, the polarity of the δ wave was positive in 5/5 right anterior APs and negative in 7/11 postero-septal APs and 8/10 in right posterior APs. In precordial leads of all dogs, R/S was ≤ 1 in V1 and > 1 in all leads from V2 to V6. Surface electrocardiogram can be used to distinguish right anterior APs from right posterior and right postero-septal ahead of an invasive electrophysiological study.


Introduction
In 1913, Stanley Kent first described the presence of extra connections between the right atrium and right ventricle in the human heart (Kent 1893). Since this first report, atrioventricular accessory pathways (AP) have been described both in people (Wang et al. 1991;Kalarus et al. 2003;Sriratanasathavort et al. 2016;Beson and Choen 2017;Ho 2020) and in dogs (Hill and Tilley 1985;Thomas et al. 2006;Santilli et al. 2006, 2007, 2018a, b, Wright at al. 2018. Accessory atrioventricular pathways represent anomalous muscular bundles that directly connect the atrial and the ventricular myocardium co-existing with the His-Purkinje system (Wang et al. 1991;Kalarus et al. 2003;Sriratanasathavort et al. 2016;Beson and Choen 2017;Ho 2020). Both in humans and dogs, APs may show retrograde, antegrade or bidirectional conduction. Retrograde conduction is related to AP mediated tachycardia, a rhythm disturbance characterized by narrow QRS complexes, regular RR intervals and short RP' interval (Santilli et al. 2006, 2007, 2018a, b, Wright et al. 2018. Bidirectional conduction is reported to be less frequent in dogs compared to humans (Wang et al. 1991;Kalarus et al. 2003;Sriratanasathavort et al. 2016;Beson and Choen 2017;Ho 2020). In dogs, antegrade conduction is reported to occur in 31.3-43.8% (Santilli et al. 2018a, b;Wright et al. 2018) of the cases, and it is associated with ventricular pre-excitation (VPE) syndrome. Ventricular pre-excitation is present when, concerning atrial events, the whole or a portion of the ventricular muscle is activated earlier than it would be if the atrial impulse only reached the ventricles via the atrioventricular node and the specialized conduction system (Kent 1893). The term syndrome is used when the electrocardiographic or endocardial mapping evidence of VPE is associated with an accessory pathway-mediated tachycardia (Kent 1893). The electrocardiographic features of VPE are a short PQ interval, a wide QRS complex, the presence of a δ wave, and abnormalities of repolarization (ST segment and T wave) (Rogel and Kaplinsky 1963;Ferrer 1977;Gallagher et al. 1978;Hill and Tilley 1985;Bathia et al. 2016;Santilli et al. 2018a, b).
Radiofrequency catheter ablation is considered the gold standard therapy treating APs and associated arrhythmias both in humans (Wang et al. 1991;Sriratanasathavorn et al. 2016) and dogs (Santilli et al. 2006(Santilli et al. , 2010(Santilli et al. , 2012. This technique carries a high success rate in veterinary medicine and is associated with a low complication rate (Santilli et al. 2006, 2018a, b, Wright et al. 2018). The accurate mapping required for radiofrequency catheter ablation of APs offers the unique opportunity to compare the electrocardiographic patterns of pre-excitation with the precise anatomical position APs (Rosenbaum et al. 1945). It is of clinical importance because differentiating anterior right-sided APs from posterior right-sided APs is crucial to plan a procedure and defining the risks related to the ablation technique (Santilli et al. 2018a, b;Wright et al. 2018). Rosenbaum et al. first attempted to localize the APs using surface electrocardiogram (ECG) tracings and identified two types of VPE patterns: type A corresponds to a left AP and a VPE pattern characterized by a positive QRS in the right precordial leads; type B corresponds to a right AP and a VPE pattern characterized by a negative QRS in the right precordial leads (Rosenbaum et al. 1945). In the recent years, more accurate localization of the APs based on several algorithms have been developed based upon the analysis of the delta wave and QRS deflection in the frontal and horizontal planes and the analysis of the duration of the PQ interval (Milstein et al. 1987;Lemery et al 1987;Xie et al. 1994;Fitzpatrick et al. 1994, Chiang et al. 1995d'Avila et al. 1995;Iturralde et al. 1996;Arruda et al. 1998;Boesma et al. 2002;Ratner at al. 2012;Pambrum et al. 2018;Jamal et al. 2019;Crinion and Baranchuk 2020).
Since no data is available in veterinary medicine regarding the localization of APs according to the surface ECG, the aim of this study was to describe the electrocardiographic features of right-sided VPE patterns with the goal of using the standard 12-lead surface ECG as a tool to help determine the anatomic location of the APs in the dog.

Animals
Twenty-six dogs were included in the study between December 2006 and December 2018 at Clinica Veterinaria Malpensa with the owner consent. All dogs underwent a complete physical examination, 5-minute 12-lead ECG, thoracic radiography, echocardiographic examination and electrophysiologic mapping. Dogs presenting abnormalities on radiographic and echocardiographic examination were excluded from the study.

Design of the study
The five-minute 12-lead ECG was recorded in unsedated dogs gently restrained in right lateral recumbency. Limb leads (I, II, III), unipolar augmented limb leads (aVR, aVL, aVF), and precordial leads with V1 placed at the costochondral junction of the right first intercostal space were recorded. The sixth intercostal space was used for all the left-sided leads (V2 through V6) with lead V2 positioned adjacent to the sternum, V3 at a midway between V2 and V4, V4 at the level of the costochondral junction, V5 and V5 sequentially positioned dorsal to V4 at a distance equal to that between V3 and V4 (Santilli et al. 2019).
All dogs selected for the study had a single AP, electrocardiographic signs compatible with VPE, including a PQ interval < 60 msec, a QRS complex > 70 msec, and a presence of a delta wave, normal thoracic radiography, normal echocardiography and no obvious causes of intraventricular conduction delay or blocks (Hill and Tilley 1985).
Dogs included in the study were anesthetized for electrophysiologic mapping as previously described (Santilli et al. 2018a, b;Wright et al. 2018). They were placed in right lateral recumbency, and two 7-Fr introducers were positioned in the right femoral vein and one introducer in the left external jugular vein using the modified Seldinger technique. Under fluoroscopic guidance, a decapolar catheter 1 was placed guided from the left external jugular vein into the coronary sinus (CS) ostium and great cardiac vein. Its correct position was confirmed by intracardiac electrograms (Santilli et al 2018a, b). A quadripolar catheter 2 was inserted into the right femoral vein and positioned at the level of the atrioventricular junction to record His bundle potentials. A third catheter, the ablation catheter with deflectable curve 3 , was inserted into the right femoral vein and positioned at the level of the right atrium, right ventricular apex or tricuspid annulus to perform atrial or ventricular programmed electrical stimulations for unipolar or bipolar endocardial mapping. Twelve-lead surface ECG and 10 intracardiac recordings (5 CS signals, 2 His signals, 2 ablation site signals, and 1 unipolar recording from the distal pole of the ablation catheter) were displayed on the electrophysiologic system at a speed ranging from 100 to 300 mm/s. Surface ECG and intracardiac electrograms were recorded with filter settings from 50 to 500 Hz. 4 The presence of VPE during sinus rhythm or atrial pacing was based on a short or negative His-to-ventricle (HV) interval and a wide QRS complex (Santilli et al. 2018a, b).
Accessory pathways were precisely localized using fluoroscopic and intracardiac ECG guidance by recording AP activation potentials from the closely spaced electrodes positioned near the site of earliest ventricular activation during antegrade AP conduction (Santilli et al. 2018a, b). At these sites, unipolar recordings from the distal tip of the ablation catheter resulted in a sharp and negative electrocardiographic deflection with a QS morphology (Santilli et al. 2018a, b).
On the basis of their position along the tricuspid annulus, APs were classified as: right anterior APs (including anteroseptal and mid-septal APs), right postero-septal APs, right posterior APs (including posterior, postero-lateral and lateral APs).
For each tracing the following parameters on three non-consecutive beats were obtained by a board-certified cardiologist (MP). Traces were anonymous and the cardiologist was blinded with respect to the results of the EP study. Delta wave is defined as the first deflection of the QRS complex due to the early depolarization of a portion of the ventricular myocardium activated by the accessory pathway (Beson and Choen 2017). The first positive deflection of the δQRS complex is defined as R wave and any subsequent positive wave is defined as R' wave. Negative electrocardiographic waves are identified as Q waves if they precede the R wave, and S waves if they follow the R wave. P-δQRS complex interval (msec), δQRS duration (msec), δ wave polarity, Q-wave, R-wave, R'-wave, S-wave amplitude (mV) were measured using a digital caliper in all standard limb and precordial leads, using a 50mm/s speed and 1 mV/10mm or 2 mV/10mm calibration depending on the amplitude of the electrical signal (Fig. 1). In standard limb leads the δQRS complex axis in the frontal plane was calculated using the following formula: arctan (I amp , aVF amp ) x 180 / π. In lead II the duration of the RR interval and QT interval (msec) was measured. In standard limb leads and precordial leads the morphology of the δQRS complex was described. In all precordial leads, R/S ratio was evaluated. Depending on whether pre-excitation was a fixed or a transient phenomenon, it has been classified as intermittent or persistent.
Statistical analysis was performed using a statistical software (JMP pro 16.0). The power of statistical model was evaluated before the analysis and resulted appropriate even in the small number of cases (α = 0.05; β = 0.95). Data were tested for normal distribution with a Shapiro-Wilk test. ANOVA and Kruskal-Wallis test were, respectively, applied for normally and not normally distributed parameters. Normally distributed data were expressed as mean ± standard error, while, nonnormal distributed data were expressed as median and interquartile range. When a significant alpha (P ≤ 0.05) was detected, pairwise comparisons were performed using Wilcoxon each pair test or Student's t-test for each pair test for nonnormally and normally distributed data respectively.
All the examined parameters were non-normally distributed except the R wave amplitude in lead V3 and V4. In Table 1 the measurements and morphology of the δQRS complex in standard limb leads and precordial leads are reported.
The median δQRS complex axis in the frontal plane was: + 68° (IQR 52.5) for right anterior APs, − 43.5° (IQR 27.25) for right posterior APs, and − 24° (IQR 24) for right postero-septal APs for differences among groups. (Figs. 4A,4B and 5) In lead II, the polarity of the δ wave was positive in 5/5 right anterior APs and negative in 7/11 postero-septal APs and 8/10 in right posterior APs (Fig. 2) . Table 1 presents the data and the P values across the 3 AP locations and Fig. 6 shows the significant differences between groups. In 26/26 dogs, delta wave was negative in lead aVR.
In precordial leads, V1 presented a R/S < 1 in 5/5 right anterior, 11/11 right postero-septal and 10/10 right posterior APs, while all leads from V2 to V6 presented Fig. 1 In all limb and precordial leads, P-δQRS complex interval (msec), δQRS complex duration, Q-wave, R-wave, S-wave amplitude (mV) were measured as shown in the figure   Fig. 2 Electrocardiographic tracings in lead II show persistent ventricular pre-excitation (A, in 76.9% of the cases) and intermittent ventricular pre-excitation (B, in 23.1% of the cases). Ventricular pre-excitation is characterized by a short P-δQRS complex interval (< 60 msec), wide δQRS complex (> 70 msec) and the presence of a δ wave. Note in tracing A negative δ wave (arrow). Note in tracing B electrocardiographic signs of ventricular pre-excitation in the 1 st and 4 th complexes with a positive δ wave (arrow) and a normal atrioventricular conduction with normal PQ interval, narrow QRS complex and absence of δ wave in the other complexes. Paper speed 50 mm/ sec -Amplitude: 10 mm = 1 mV R/S > 1 in 5/5 right anterior, 11/11 right postero-septal and 10/10 right posterior APs. (Fig. 4C).

Discussion
As previously reported in veterinary literature (Hill and Tilley 1985), our data confirmed that VPE is characterized by wide QRS complexes and short P-δQRS interval. A short P-δQRS complex interval depends on several variables such as the interatrial conduction times, the distance of the atrial origin of the bypass tract from the sinus node, the refractory period of the atrioventricular node and the bypass tract and the autonomic nervous system tone (Steurer et al. 1994). Other than wide QRS complexes and short P-δQRS interval, a reported approach to diagnose VPE is to test the presence of an initial R wave in lead aVR owing to the activation of the interventricular septum. In VPE, the first part of ventricular activation is usually from the basal aspect towards the apex, away from aVR, obscuring the usual R The results of median, interquartile difference (IQR) or mean ± standard deviation (SD) and the p value are shown according to localization of the accessory pathway (right anterior, right posterior and right posteroseptal) Numbers in bold are values statistically significant (P < 0.05)  (Eisenberger et al. 2010). In our sample, 26/26 dogs presented no initial positive delta wave in lead aVR, so probably this electrocardiographic criterion is suggestive of VPE in the dog. Ventricular pre-excitation is a dynamic phenomenon and can be intermittent or manifest. In human medicine, VPE is intermittent in 50% of cases. Intermittent VPE is defined by the loss of the δ wave with concomitant prolongation in the PQ interval documented on at least one beat (Kein and Gulamhusein 1983). In our study, intermittent pre-excitation occurred in a lower percentage of cases (23.1%) (Fig. 2). Human patients with intermittent VPE present a longer anterograde refractory period of the AP and poorer conduction over the pathway than patients with constant pre-excitation. These electrophysiological findings support the hypothesis that intermittent pre-excitation is a manifestation of an AP with "precarious" conduction (Kein and Gulamhusein 1983). The anterograde effective Fig. 4 The analysis of the δQRS complex morphology and axis on the standard limb lead and the analysis of the δQRS complex morphology on the precordial leads can suggest the localization of the AP. Panels A and B display six limb leads and main ventricular vector analysis in the frontal plane in a dog with an anterior AP (A) and a dog with a posterior AP (B), respectively. Pre-excitation along an anterior AP presents a wide δQRS complex with a normal mean electrical axis mimicking a complete left bundle branch block (A) (notice the black portion of the arrow that starts from the Hissian area and then moves anteriorly becoming transparent and dotted). In contrast, VPE along a posterior (lateral, posterior and postero-septal) AP presents a leftward deviation of the mean electrical axis in the frontal plane I and aVL mimicking a left anterior fascicular block (B) (notice the black portion of the arrow that starts from coronary sinus ostium area, moves posteriorly and then anteriorly becoming transparent and dotted), Panel C displays precordial leads and main ventricular vector analysis in the horizontal plane in a dog with a right-sided AP. Considering the morphology of the δQRS complex in precordial leads, pre-excitation along a right accessory pathway presents a wide δQRS complex characterized by rS morphology with R/S < 1 in lead V1 and R/S > 1 from V2 to V6. A and B, paper speed 50 mm/sec -Amplitude: 10 mm = 1 mV; C, paper speed 50 mm/sec -Amplitude: 5 mm = 1 mV. In the lower part of the figure, the fontal plane with respective leads (I, II, III, aVR, aVL, aVF) and horizontal plane (V1-V6) are represented. The arrows show the direction of the ventricular activation wavefront in case of anterior (A) and posterior (B) located AP refractory period is not influenced by the AP location in humans (de Chillou et al. 1992) although in our study, none of the right anterior APs were correlated with intermittent VPE.
Concerning the treatment, radiofrequency catheter ablation has become the treatment of choice for patients with symptomatic and asymptomatic VPE and correlated supraventricular arrhythmias (Morady 2004;Santilli et al. 2018a, b;Wright et al. 2018). During recent years criteria for the localization of the bypass tract from the conventional 12-lead ECG have become increasingly apparent (Arruda et al. 1998;Boesma et al. 2002;Ratner et al. 2012;Pambrun et al. 2018;Jamal et al. 2019;Crinion and Baranchuk 2020). A non-invasive method, such as 12-lead surface ECG, to guide accurate localization of the anatomic substrate of the supraventricular arrhythmia would represent a significant adjunct to the electrophysiologic mapping (Silka et al. 1993).
More accurate localization of the bypass tracts could be possible and is based upon the following principles: analysis of the main QRS deflection and analysis of the delta wave.
Our study showed that the polarity of the QRS complexes in standard limb and precordial leads and the mean electrical axis in the frontal plane can help in the localization of an accessory pathway based on the analysis of the 12-lead surface ECG (Fig. 4). Pre-excitation along a right anterior AP presents a wide δQRS complex with a normal mean electrical axis mimicking a complete left bundle branch block (median electrical axis in the frontal plane + 68°), while in most cases VPE associated with right posterior AP was characterized by QRS complex axis in the frontal plane deviated to the left (median electrical axis in the frontal plane − 43.5°). On the other hand, postero-septal pathways, based on their position, usually have a less pronounced left axis deviation both in human medicine and in this study (-24°) (Steuer et al. 1994;Xie et al. 1994;Okamura et al. 1980) ( Table 2).
The analysis of the polarity of the QRS complex in precordial leads seems to be at least as important as the analysis of the QRS mean electrical axis in the frontal plane. In human medicine, the right or left bypass tracts are easily distinguished by the main polarity in lead V1 and in leads V2-V6. Right-sided APs presented an rS morphology with R/S < 1 in V1 and an Rs morphology with R/S > 1 in V2 to V6 (Lemery et al. 1987;Haghjoo et al. 2008) (Fig. 3). In our study, all dogs had right-sided accessory pathway, V1 had rS morphology with R/S < 1 and from V2 to V6 R/S > 1 in all the cases ( Table 2).
The polarity of the δ wave in the frontal plane can be an additional factor that can help to localize the accessory pathway, although in human medicine it has been reported that the classification considering δ wave polarity has significant limitations in predicting APs site (Kamakura et al. 1986). In human medicine, negative δ waves in lead I and aVL are never associated with right posterior or posteroseptal APs (Haghjoo et al. 2008). This data can be suggestive also in veterinary medicine since in our sample δ waves in lead I and aVL were positive in all right posterior and right postero-septal APs. Furthermore, in human medicine right anterior APs are usually associated with frontal plane δ wave axis in the region of + 30°/+60° resulting in a positive δ wave in inferior leads II, III, aVF (Fitzpatrick et al. 1994;Chiang et al. 1995). We observed a similar pattern in our study in which the δ wave was positive in leads II, III, aVF in 5/5 cases presenting anterior Aps (Table 2).
In human medicine, different algorithms were able to differentiate right postero-septal from right posterior and right lateral APs (Xie et al. 1994;Fitzpatrick et al. 1994;Chiang et al. 1995;d'Avila et al. 1995;Iturralde et al. 1996;Arruda et al. 1998), while in our study, probably because of the smaller dimensions of the tricuspid valve annulus and the proximity of the different posterior locations, this differentiation was not possible. Furthermore, in this study, the electrocardiographic feature of VPE was studied only considering a right-sided position, since in our population no dog presenting a left-sided AP was represented.
Besides the limited number of dogs in this study and the small number of animals in each group, any electrocardiographic classification of the VPE syndrome has important limitations. Limitations included variable degrees of fusion between the normal and AP-derived wavefronts traveling to the ventricles due to the dynamic phenomenon of VPE, superimposition of the terminal portion of the P wave on the initial δ wave, and the effect of endocardial versus epicardial location of the Boxplots comparing the amplitude of electrocardiographic waves in relation to the AP position in different leads. A: the graph shows R wave in lead I, which was significantly taller in right posterior APs for differences among medians within groups. B: the graph shows R' wave in lead II, which was significantly taller in right anterior APs for differences among medians within groups. C: the graph shows R' wave in lead III, which was significantly taller in right anterior APs for differences among medians within groups Intermittent VPE 0/26 (0%) 3/26 (11.5%) 3/26 (11.5%) Positive polarity of δ wave in lead II 5/5 (100%) 4/11 (36.4%) 2/10 (20%) R/S < 1 in V1 5/5 (100%) 11/11 (100%) 10/10 (100%) R/S > 1 in V2-V6 5/5 (100%) 11/11 (100%) 10/10 (100%) AP are considered (Gallagher et al. 1978;Bathia et al. 2016). The δ wave and δ-QRS complex may further be modified by anatomic shift of the heart inside the thorax resulting from extracardiac factors (Hill and Tilley 1985;Wright at al. 1995). Furthermore, the interpretation of the VPE patterns can be complicated by the geometry of the chest in different morphotypes and the conductivity of the body tissue in relation to the body condition score (De Ambroggi et al. 1976). Finally, a variability and repeatability assessment on electrocardiographic measurement was not done.
In conclusion, in humans, during recent years criteria for the localization of the bypass tract from the conventional 12-lead ECG have become increasingly apparent (Arruda et al. 1998;Boesma et al. 2002;Ratner et al. 2012;Pambrun et al. 2018;Jamal et al. 2019;Crinion and Baranchuk 2020). Based on the results of the present study, the localization of the rightsided AP based on electrocardiographic criteria in dogs is also possible. Considering the mean electrical axis of the QRS complex in the frontal plane, pre-excitation along an anterior AP presents a wide δQR complex with a normal mean electrical axis mimicking a complete left bundle branch block. In contrast, pre-excitation along a posterior (lateral, posterior, and posteroseptal) AP presents a leftward deviation of the mean electrical axis in the frontal plane mimicking a left anterior fascicular block. Delta wave in the frontal plan was positive in lead I and aVL in right posterior and right postero-septal APs and positive in the inferior leads in right anterior APs. Furthermore, all right APs are characterized by an rS morphology with R/S < 1 in lead V1 and R/S > 1 from V2 to V6.