Algorithm for differentiation of left and right posterior paraseptal accessory pathway

Journal of Electrocardiology Vol. 37 No. 2 2004

Algorithm for Differentiation of Left and Right Posterior Paraseptal Accessory Pathway

So Takenaka, MD,* San-Jou Yeh, MD,† Ming-Shien Wen, MD,† Kuan-Hung Yeh, MD,† Chun-Chieh Wang, MD,† Fun-Chung Lin, MD,† and Delon Wu, MD†

Abstract: We studied 196 consecutive patients with posterior paraseptal accessory pathway (AP); 124 showed manifest preexcitation and 72 were concealed AP. Successful ablation was obtained from left-sided approach in 134 patients (left posterior pasaseptal [LPS] group) and from right sided approach in 62 patients (right posterior paraseptal [RPS] group). A ventriculoatrial (VA) interval of ⬍50 ms recorded at LPS region (VALPS) during right ventricular pacing identified 95 of the 134 patients (71%) with LPS AP with 100% specificity and positive predictive value. In the 101 patients with VALPS ⱖ50 ms, a difference in VA interval of ⬍20 ms recorded at the His bundle region and LPS region, ⌬VA(H-LPS), during right ventricular pacing predicted RPS AP with a sensitivity of 97%, a specificity of 85% and a positive predictive value of 91%. When these 2 parameters were used together for prediction of LPS or RPS AP, the sensitivity, specificity, and positive predictive value were 96%, 97%, and 98% for LPS AP, and 97%, 96%, and 91% for RPS AP, respectively. This simple new algorithm using VALPS and ⌬VA (H-LPS) during right ventricular pacing successfully discriminates LPS and RPS AP with high sensitivity, specificity, and positive predictive value and could facilitate radiofrequency ablation in patients with posterior paraseptal AP. Key words: Ablation, electrophysiology, supraventricular tachycardia.

The posterior paraseptal area is a complex anatomic structure where 4 cardiac chambers and the coronary sinus come in close proximity (1–3) (Fig.

1). Previous works have suggested that the connections between the atria and ventricles in this region are through a “right atrial-left ventricular connection” in the right posterior paraseptal (RPS) region and a “left atrial-left ventricular connection” in the left posterior paraseptal (LPS) region (2,4). Thus, patients with a RPS accessory pathway, retrograde atrial activation at the LPS region is propagated from the right atrium through the right atrial-left ventricular connection, whereas patients with a LPS accessory pathway, retrograde atrial activation at the LPS region is directly from the left ventricle

From the *Hiroshima University Graduate School of Biomedical Sciences, Hiroshima, Japan; and †Department of Medicine, Second Section of Cardiology, Chang Gung Memorial Hospital, Chang Gung University College of Medicine, Taoyuan, Taiwan. Reprint requests: Delon Wu, MD, Second Section of Cardiology, Chang Gung Memorial Hospital, 199 Tung Hwa North Rd, Taipei, Taiwan; e-mail: [email protected] © 2004 Elsevier Inc. All rights reserved. 0022-0736/04/3702-0001$30.00/0 doi:10.1016/j.jelectrocard.2004.01.004

75

76 Journal of Electrocardiology Vol. 37 No. 2 April 2004 this institution between June 1991 and December 2002. Of the 1,508 patients, 269 were found to have a posterior paraseptal accessory pathway. A posterior paraseptal accessory pathway is defined as a pathway which was successfully ablated at a site near around the coronary sinus over the inferoposterior one third of Koch’s triangle, either within 1 cm adjacent to the ostium of the coronary sinus over the inferomedial aspect of the right atrium or 2 cm to the left of the coronary sinus ostium over the posteromedial mitral annulus (5–7). To be included in this study, the patient should fulfill the following criteria: 1) the accessory pathway was successfully ablated at the posterior paraseptal site, 2) there was only a single retrograde accessory pathway, and 3) the coronary sinus was successfully cannulated during basic electrophysiologic study. Seventy three of the 269 patients were excluded, 4 with unsuccessful ablation, 46 with multiple accessory pathways, 20 with no retrograde conduction of the accessory pathway, and 3 without a coronary sinus catheter. Thus, the study population consists of 196 patients, 120 men and 76 women, with a mean age of 45 ⫾ 17 years. Fig. 1. Diagram showing left posterior paraseptal accessory pathway running along the “left atrial-left ventricular connection” and (A) retrograde impulse activation and right posterior paraseptal accessory pathway running along the “right atrial-left ventricular connection” and (B) retrograde impulse activation. Asterisk indicates left posterior paraseptal region. CS, coronary sinus; LA, left atrium; LV, left ventricle; RA, right atrium; RV, right ventricle.

through the left atrial-left ventricular connection. Chiang et al. has previously reported an algorithm based on electrophysiologic parameters for differentiation of RPS and LPS accessory pathway, which could facilitate the ablation procedure (5). In this study, we developed a simple new algorithm based on anatomic structure of the posterior paraseptal area and the electrophysiologic parameters for discrimination of LPS and RPS accessory pathway with high sensitivity, specificity and positive predictive value.

Materials and Methods Study Population A total of 1,508 patients with 1,620 accessory pathways underwent radiofrequency ablation in

Electrophysiologic Study Four electrode catheters were positioned respectively in the high right atrium, the His bundle region, the coronary sinus, and the right ventricular apex. The coronary sinus catheter was a 6F quadripolar catheter with 1 cm interelectrode spacing, and was introduced through the right internal jugular vein. The proximal electrode (the 4th electrode) was positioned just outside the ostium of the coronary sinus for recording of unipolar electrogram from the right posterior paraseptal area, the 3rd electrode was inside the ostium of the coronary sinus for recording of the unipolar electrogram from the left posterior paraseptal area and the distal pair of electrodes (the 1st & the 2nd electrode) was positioned inside the coronary sinus for recording of electrograms from the left posterior region. Programmed stimulation was performed from the high right atrium and the right ventricle as previously described (6,8 –10). The following parameters were analyzed: 1) the shortest cycle length sustaining 1:1 conduction over the accessory pathway, 2) the effective refractory period of the accessory pathway, 3) the ventriculo-atrial (VA) and/or atrioventricular (AV) interval at the LPS region (VALPS or AVLPS) and RPS region (VARPS or AVRPS) from the coronary sinus catheter at a paced cycle length of 500 ms, and 4) the difference of stimulus to atrial

Posterior Paraseptal Accessory Pathway •

interval between that recorded at LPS region and the His bundle region, ⌬VA(H-LPS), and between that recorded at the RPS region and the His bundle region, ⌬VA(H-RPS). The cycle length during induced orthodromic atrioventricular reentrant tachycardia, the AH and HA interval recorded at the His bundle region and at the ostium of the coronary sinus, as well as VALPS and VARPS during tachycardia were also analyzed. Radiofrequency Ablation The site for delivery of radiofrequency current was selected by anterograde mapping during sinus rhythm (patients with manifest pre-excitation), and/or retrograde mapping during atrioventricular reentrant tachycardia or right ventricular pacing (6,8,9). Left-sided approach was attempted first if VA interval was the shortest at LPS region, while right-sided approach was attempted first if VA interval was the shortest at RPS region. For delivery of radiofrequency current to the left posterior paraseptal area, the ablation catheter was introduced percutaneously into the right femoral artery, advanced to the left ventricle, and positioned at the lower third of the mitral inflow tract or, alternatively, at the atrial aspect of the mitral annulus. For delivery of radiofrequency current to the right posterior paraseptal area, the ablation catheter was introduced percutaneously into the right femoral vein and advanced to the atrial aspect of the tricuspid annulus. Fine adjustments of the catheter were made by previously described technique to obtain the ideal ablation site (6). When the initial attempt was unsuccessful, the contralateral side was tried. If both approaches were unsuccessful, the ablation catheter was advanced into the cardiac venous system for further ablation. The anterograde or retrograde local activation time (AV or VA) was recorded from the ablation catheter at the successful ablation site. Radiogram of the successful ablation site was recorded in the 30° right anterior oblique and 30° left anterior oblique projections. Statistical Analysis Continuous variables were expressed as mean ⫾ SD and were compared by unpaired Student’s t test. Univariate analysis of nonparametric data was performed by the table analysis. Multivariate analyses were performed by use of the multiple logistic regression technique with the successful ablation site as the dependent variable and the electrophysiologic parameters as the independent variables. A value of P ⬍ .05 was considered significant.

Takenaka et al. 77

Results Of the 196 patients with posterior paraseptal accessory pathway, 124 showed manifest preexcitation and 72 concealed accessory pathway. Successful ablation was achieved from left sided approach in 134 patients (LPS group) and from right sided approach in 62 patients (RPS group). There were no difference in age (45 ⫾ 16 vs. 45 ⫾ 19 years) and sex (male to female ratio: 84/50 vs. 36/26) between the 2 groups. During baseline electrophysiologic study, orthodromic atrioventricular re-entrant tachycardia was induced in 184 patients. Uni-variance analysis of the electrophysiologic data revealed difference in 11 parameters (Table 1): manifest pre-excitation was more common in LPS group, AVLPS during high right atrial pacing was shorter in LPS group. The retrograde effective refractory period of the accessory pathway was longer in LPS group, VALPS during right ventricular pacing was shorter in LPS group, ⌬VA (H-LPS) was longer in LPS group, ⌬VA (H-RPS) was longer in LPS group, both A–H interval and CSos-H interval during tachycardia was longer in LPS group, H-CSos interval during tachycardia were shorter in LPS group, both VALPS and VARPS during tachycardia were shorter in LPS group, and both anterograde and retrograde local activation time recorded from the ablation catheter were shorter in LPS group. With multivariance analysis, 2 independent parameters able to discriminate the 2 groups were identified. These were VALPS (P ⫽ .033; 95% confidence interval ⫽ 1.178 – 1.875) and ⌬VA (H-LPS) (P ⫽ .001; 95% confidence interval ⫽ 0.794 – 0.990) during right ventricular pacing. With VALPS during right ventricular pacing, 95 of the 134 patients (71%) in LPS group had an interval shorter than 50 ms, while all patients in RPS group had an interval ⱖ50 ms (Fig. 2A, 3, and 4A). The specificity and positive predictive value for LPS accessory pathway were 100% when VALPS was less than 50 ms. In the 101 patients with a VALPS ⱖ50 ms, a ⌬VA (H-LPS) ⬍20 ms predicted RPS accessory pathway with a sensitivity of 97%, specificity of 85%, and a positive predictive value of 91% (Fig. 2B, 3, and 4B). Figure 3 is an algorithm for differentiation of LPS or RPS accessory pathway. When VALPS ⬍50 ms was examined in the 196 patients with posterior paraseptal accessory pathway, 95 patients fulfilled this criteria and all 95 patients were successfully ablated from LPS region. The other 101 patients with VALPS ⱖ50 ms, 66 patients had a ⌬VA (H-LPS) ⬍20 ms and 60 of these 66 patients were successfully ablated from the RPS region; 35 patients had a

78 Journal of Electrocardiology Vol. 37 No. 2 April 2004

Fig. 2. (A) Density dot chart showing VALPS (A), and (B) ⌬VA (H-LPS) during right ventricular pacing in patients with left posterior paraseptal accessory pathway and right posterior paraseptal accessory pathway. See text for discussion.

⌬VA (H-LPS) ⱖ20 ms and 33 were successfully ablated from the LPS region (Fig. 4C). Two patients with a VALPS ⬎50 ms and a ⌬VA (H-LPS) ⱖ20 ms during right ventricular pacing were ablated successfully from the middle cardiac

vein with a diverticulum. Of the 6 patients with VALPS ⱖ50 ms and a ⌬VA (H-LPS) ⬍20 ms, a transient loss of accessory pathway conduction was noted in 3 patients with right sided approach. In the other 3 patients, the retrograde conduction time of the accessory pathway was equal to or longer than the normal pathway. Successful ablation of the accessory pathway in these 6 patients was achieved from the left side. The anterograde and retrograde local activation time recorded from the ablation catheter at the successful ablation site was shorter in LPS group as compared to the RPS group (32.1 ⫾ 10.3 vs. 38.7 ⫾ 12.0 ms, P ⫽ .015 and 35.7 ⫾ 11.8 vs. 50.1 ⫾ 14.8 ms, P ⬍ .001, respectively).

Discussion Anatomic Consideration

Fig. 3. Stepwise algorithm for selection of successful ablation site between left posterior paraseptal and right posterior paraseptal region. See text for discussion.

The posterior paraseptal area is a region where four cardiac chambers and the coronary sinus come in close proximity (1–3) (Fig. 1). It encompasses an area corresponding to the AV sulcus between the posterior superior process of the left ventricle and the inferior wall of the right atrium and, thus, could be designated as “right atrial-left ventricular sulcus.” The “right atrial-left ventricular sulcus” meets the right AV sulcus anteriorly and to the right and the left AV sulcus posteriorly and to the left. There are “right atrial-left ventricular connections” over the right and the middle portion of the posterior paraseptal region and “left atrial- left ventricular

Posterior Paraseptal Accessory Pathway •

Takenaka et al. 79

Fig. 4. Recordings from 3 patients showing left posterior paraseptal accessoary pathway with a (A) short VALPS of 40 ms, right posterior paraseptal accessory pathway with a long VALPS of 75 ms and a (B) ⌬VA (H-LPS) of 15 ms, and left posterior paraseptal accessory pathway with a (C) long VALPS of 80 ms and a ⌬VA (H-LPS) of 30 ms. I, aVF & V1 ⫽ electrocardiographic leads I, aVF & V1; HRA, CS1, CS2, CS3 & CS4 ⫽ bipolar atrial electrogram recorded from high right atrium and unipolar electrogram recorded from the distal electrode, the 2nd, the 3rd & the proximal electrode of the coronary sinus catheter; HBEd, HBEm & HBEp ⫽ bipolar electrogram recorded from the distal, middle and proximal pair of electrodes of the His bundle catheter; RV ⫽ bipolar electrogram recorded from the right ventricular apex. * VALPS & VARPS were measured from CS3 & CS4, respectively.

connection” over the left posterior paraseptal region. A right-sided posterior paraseptal accessory pathway may run along the “right atrial-left ventricular connection,” while the left-sided posterior paraseptal accessory pathway may run along the “left atrial-left ventricular connection” (2,4,11). If so, electrophysiologic parameters may be identified for discrimination of right and left posterior posterior paraseptal accessory pathway.

Comparison With Previous Study Electrocardiographic algorithm has been developed for prediction of right or left posterior paraseptal accessory pathway with analysis of the initial delta wave vector and the R/S ratio in the inferior and right precordial leads (12,13,14). However, analysis of electrocardiographic parameters is dependent on the presence of manifest pre-excitation

80 Journal of Electrocardiology Vol. 37 No. 2 April 2004 Table 1. Baseline Characteristics and Electrophysiological Parameters

Patient number Age Male/Female Manifest Preexcitation Antegrade conduction (ms) 1:1 AP ERP AP AVLPS AVRPS Retrograde conduction (ms) 1:1 AP ERP AP VALPS VARPS ⌬VA (H-LPS) ⌬VA (H-RPS) During tachycardia CL A-H H-A CSos-H H-CSos VALPS VARPS Local activation time (ms) Antegrade Retrograde

LPS

RPS

P

134 45 ⫾ 16 84/50

62 45 ⫾ 19 36/26

.9 .53

100/34

24/38

⬍.001

n ⫽ 100 367 ⫾ 129 313 ⫾ 115 54.7 ⫾ 20.0 58.9 ⫾ 22.0

n ⫽ 24 328 ⫾ 91 296 ⫾ 72 66.0 ⫾ 19.3 64.0 ⫾ 18.7

n ⫽ 134 330 ⫾ 52 275 ⫾ 48 47.0 ⫾ 14.3 65.5 ⫾ 17.6 31.2 ⫾ 8.3 18.7 ⫾ 9.3 n ⫽ 129 337 ⫾ 59 177 ⫾ 60 160 ⫾ 24 208 ⫾ 59 129 ⫾ 29 66.8 ⫾ 19.3 77.4 ⫾ 17.0

n ⫽ 62 290 ⫾ 38 257 ⫾ 37 68.9 ⫾ 10.6 66.1 ⫾ 14.7 6.0 ⫾ 15.1 12.1 ⫾ 9.6 n ⫽ 55 323 ⫾ 55 157 ⫾ 54 166 ⫾ 30 180 ⫾ 64 143 ⫾ 30 88.8 ⫾ 21.8 88.1 ⫾ 22.8

32.1 ⫾ 10.3 35.7 ⫾ 11.8

38.7 ⫾ 12.0 50.1 ⫾ 14.8

.15 .45 .016 .28 .14 .006 ⬍.001 .83 ⬍.001 ⬍.001 .13 .02 .18 .004 .006 ⬍.001 .007 .015 ⬍.001

LPS and RPS, left and right posterior paraseptal accessory pathway; 1:1 AP, shortest paced cycle length sustaining 1:1 conduction over accessory pathway; AVLPS or AVRPS, atrioventricular interval recorded at left or right posterior paraseptal region; VALPS or VARPS, ventriculoatrial interval recorded at left or right posterior paraseptal region; CL, cycle length; A-H or CSos-H, interval between low septal right atrium or ostium of coronary sinus and His bundle electrogram; H-A or H-CSos, interval between His bundle electrogram & low septal right atrium or ostium of coronary sinus.

as well as the degree of pre-excitation and, therefore, is limited in clinical practice. Chiang et al. recently reported an algorithm for discrimination of concealed right and left posterior paraseptal accessory pathway (5). They recorded atrial electrogram at the right posterior paraseptal region from the ostium of the coronary sinus and the left posterior paraseptal region inside the coronary sinus using a orthogonal catheter with circumferential electrode positioned in the coronary sinus as well as the atrial electrogram from the His bundle recording catheter during induced atrioventricular reentrant tachycardia. They found that the earliest atrial activation occurring at the left posterior paraseptal region recorded from the orthogonal catheter identified left posterior paraseptal accessory pathway with 100% specificity and 100% positive predictive

value. They also measured the ⌬VA during induced tachycardia as defined by the difference in VA intervals that recorded at the His bundle catheter and that recorded from the orthogonal catheter with the earliest atrial activation. A ⌬VA ⱖ25 ms identified left posterior paraseptal accessory pathway with a sensitivity of 92%, a specificity of 89%, and a positive predictive value of 79%. The presence of long RP’ tachycardia identified right posterior paraseptal accessory pathway with 100% specificity. The present study excluded patients with long RP’ interval during tachycardia and also those with Mahaim type of ventricular preexcitation because the conduction property of the anomalous pathway in these patients is different from the ordinary accessory pathway. In addition, we simplified the algorithm by using the parameters during right ventricular pacing. We found that a VALPS ⬍50 ms identified patients with left posterior paraseptal accessory pathway. This parameter concurs with the hypothesis that the left posterior paraseptal accessory pathway is a pathway running along the “left atrial-left ventricular connection” (Fig. 1A), and doesn’t conflict with the algorithm of Chiang et al. in which the earliest atrial activation during tachycardia at the left posterior paraseptal region identifies a left posterior paraseptal accessory pathway. The VARPS has no value in discrimination of right or left posterior paraseptal accessory pathway because a right posterior paraseptal accessory pathway is a pathway running along the “right atrial-left ventricular connection” and, thus, is a longer pathway having a longer retrograde conduction time (Fig. 1B). The finding that both the anterograde and retrograde local activation time recorded from the ablation catheter at the successful ablation site was shorter in patients with a left posterior paraseptal accessory is also consistent with this hypothesis. A recent study by Sun et al. showed that the coronary sinus myocardial coat, present in all individuals, is anatomically and electrically connected to both atria and that the coronary sinus myocardial coat may have extensions to the ventricle in some patients (15). This “coronary sinusventricular accessory connection” may produce posterior paraseptal and left posterior accessory pathway. Under such circumstances, a left posterior paraseptal accessory pathway may have a VALPS longer than 50 ms. In patients with a retrogradely conducting accessory pathway, atrial activation during ventricular pacing may result from 2 wave fronts, one from the accessory pathway and the other from the normal pathway. However, it could also come solely from the accessory pathway if the retrograde normal

Posterior Paraseptal Accessory Pathway •

pathway conduction is slow and the impulse is collided with the returning anterograde impulse from the accessory pathway. Thus, retrograde activation at the low septal right atrium recorded from the His bundle catheter may reflect retrograde impulse activation from the normal pathway and/or anterograde impulse activation returning from the accessory pathway. The left atrial activation at the left posterior paraseptal region in a patient with a right posterior paraseptal accessory pathway would be later than that of the right atrial activation at the ostium of the coronary sinus, and the ⌬VA (H-LPS) would be shorter than that of the left posterior paraseptal accessory pathway, regardless whether activation of the low septal right atrium at the His bundle recording site is from retrograde or anterograde impulse. Because there are no data in the literature concerning the conduction between the right or left posterior paraseptal atrium and the right low septal atrium at the His bundle recording site of the impulse returning from a posterior paraseptal accessory pathway, nor are there data concerning local activation time in patients with left posterior paraseptal accessory pathway using a “coronary sinus-ventricular connection,” we could not provide an explanation for patients with left posterior paraseptal accessory pathway who showed a VALPS ⱖ50 ms yet with a ⌬VA ⱖ20 ms.

Conclusion A simple new algorithm using VA interval recorded at the left posterior paraseptal region from the coronary sinus catheter and the difference of interval between the stimulus and the atrial activation recorded at the His bundle region and that at the left posterior paraseptal region during ventricular pacing successfully discriminates left and right posterior paraseptal accessory pathway with a high sensitivity, specificity, and positive predictive value and, thus, could facilitate radiofrequency procedure in patients with posterior paraseptal accessory pathway.

References 1. Sealy WC, Mikat EM: Anatomical problems with identification and interruption of posterior septal Kent bundles. Ann Thorac Surg 36:584, 1983

Takenaka et al. 81

2. Guiraudon GM, Klein GJ, Sharma AD, et al: Surgical ablation of posterior septal accessory pathways in the Wolff-Parkinson-White syndrome by a closed heart technique. J Thorac Cardiovasc Surg 92:406, 1986 3. Dean JW, Ho SY, Rowland E, et al: Clinical anatomy of the atrioventricular junctions. J Am Coll Cardiol 24:1725, 1994 4. Guiraudon GM, Klein GJ, Yee R, et al: Surgery for the Wolff-Parkinson-White syndrome, in Zipes PD, Jalife J, (eds): Cardiac Electrophysiology-From Cell to Bedside (ed 2). Philadelphia, PA, Saunders, 1995, p 1553 5. Chiang CE, Chen SA, Tai CT, et al: Prediction of successful ablation site of concealed posteroseptal accessory pathways by a novel algorithm using baseline electrophysiological parameters. Implication for an abbreviated ablation procedure. Circulation 93: 982, 1996 6. Wen MS, Yeh SJ, Wang CC, et al: Radiofrequency ablation therapy of the posteroseptal accessory pathway. Am Heart J 132:612, 1996 7. Davis LM, Byth K, Ellis P, et al: Dimensions of the human posterior septal space and coronary sinus. Am J Cardiol 68:621, 1991 8. Yeh SJ, Wang CC, Wen MS, et al: Radiofrequency ablation in multiple accessory pathways and physiologic implications. Am J Cardiol 71:1174, 1993 9. Yeh SJ, Wang CC, Wen MS, et al: Characteristics and radiofrequency ablation therapy of intermediate septal accessory pathway. Am J Cardiol 73:50, 1994 10. Wu D, Denes P, Amat-y-Leon F, et al: Clinical, electrocardiographic, electrophysiologic observations in patients with paroxysmal supraventricular tachycardia. Am J Cardiol 41:1045, 1978 11. Guiraudon GM, Klein G, Sharma AD, et al: “Atypical” posteroseptal accessory pathway in Wolff-Parkinson-White syndrome. J Am Coll Cardiol 12:1605, 1988 12. Arruda MS, McClelland JH, Wang X, et al: Development and validation of an ECG algorithm for identifying accessory pathway ablation site in Wolff-Parkinson-White syndrome. J Cardiovasc Electrophysiol 9:2, 1998 13. Chiang CE, Chen SA, Teo WS, et al: An accurate stepwise electrocardiographic algorithm for localization of accessory pathways in patients with WolffParkinson-White syndrome from a comprehensive analysis of delta waves and R/S ratio during sinus rhythm. Am J Cardiol 76:40, 1995 14. Fitzpatric AP, Gonzales RP, Lesh MD, et al: New algorithm for the localization of accessory atrioventricular connections using a baseline electrocardiogram. J Am Coll Cardiol 23:107, 1994 15. Sun Y, Arruda M, Otomo K, et al: Coronary sinusventricular accessory connections producing posteroseptal and left posterior accessory pathways. Incidence and electrophysiological identification. Circulation 106:1362, 2002