Introduction. CLINICAL RESEARCH Arrhythmia/electrophysiology

Transcription

1 European Heart Journal (2008) 29, doi: /eurheartj/ehn340 CLINICAL RESEARCH Arrhythmia/electrophysiology Rate-related changes in QRS morphology in patients with fixed bundle branch block: implications for differential diagnosis of wide QRS complex tachycardia Tomás Datino, Jesús Almendral*, Esteban González-Torrecilla, Felipe Atienza, Francisco J. García-Fernández, Ángel Arenal, Leonardo Atea, and Francisco Fernández-Avilés Cardiology Department, Hospital General Universitario, Gregorio Marañón, Doctor Esquerdo 46, Madrid 28007, Spain Received 3 December 2007; revised 15 June 2008; accepted 8 July 2008; online publish-ahead-of-print 29 July 2008 See page 2319 for the editorial comment on this article (doi: /eurheartj/ehn391) Aims To analyse QRS morphology in response to rapid atrial pacing (RAP) and supraventricular tachycardia (SVT) in patients with pre-existing bundle branch block (BBB).... Methods We prospectively studied 59 patients in sinus rhythm (SR), with QRS 120 ms, and no pre-excitation. Trains of RAP and results were introduced at increasing rates until atrioventricular block. QRS during SR and last QRS complex of each RAP train were compared on the 12-leads. Previously described criteria for minor and major configuration differences were used to identify QRS changes. During RAP minor QRS changes were seen in 22 (37%) and major changes in 23 (39%) subjects. One patient showed major axis shifts and no one showed a change to the contralateral BBB pattern. QRS changes were significantly and independently related to RAP rate and type of BBB (more frequent if right-bbb). Of 14 subjects (24%) with SVT, 13 displayed the same QRS changes during RAP.... Conclusion In patients with organic BBB, important changes in QRS morphology, except for a change in the contralateral BBB, can appear during RAP and SVT. Thus, in these patients, a change in QRS morphology during tachycardia does not necessarily imply that it is ventricular tachycardia Keywords Tachycardia Bundle branch block Diagnosis Wide QRS complex tachycardia Electrocardiography Introduction It has been widely held, 1 3 on the basis of quite limited information, 4 that the differentiation of wide QRS tachycardia in patients with preexisting wide QRS complex is easy: if the QRS complexes during tachycardia are identical to those in sinus rhythm (SR), the tachycardia is supraventricular tachycardia (SVT); if they are different, the arrhythmia is ventricular in origin. The basis for such a concept is simple: since in patients with bundle branch block (BBB) one bundle branch is already blocked, there is only one available pathway for conduction, and no aberrancy is possible in case of SVT; thus, if the QRS morphology changes during tachycardia, it must be ventricular in origin. However, this concept may be too simplistic, and has not been tested in a systematic way. In this study we simulated SVT at various rates by rapid atrial pacing (RAP) in patients with pre-existing BBB and analysed the resultant QRS morphology. Our purpose was to determine the precision of the diagnosis of SVT in these patients based on the absence of changes in QRS morphology, and to characterize QRS morphological changes if they occurred. Methods Study population and data collection We prospectively included patients referred for electrophysiological study, in SR with QRS 120 ms and no anterogradely conducting accessory pathways. Patients with QRS 120 ms due to ventricular * Corresponding author. Tel/Fax: þ , almendral@secardiologia.es Published on behalf of the European Society of Cardiology. All rights reserved. & The Author For permissions please journals.permissions@oxfordjournals.org.

2 2352 T. Datino et al. pacing were excluded from the study. Since the lack of previous studies did not allow us to establish a target sample size, we decided to include consecutive patients referred for electrophysiological study to the Hospital General Universitario Gregorio Marañón during a 3-year period. From March 2004 to March 2007 we performed 912 electrophysiological studies; 107 of these subjects met inclusion criteria and had no exclusion criteria. The atrial pacing protocol (vide infra) was performed in 75 of them (there were 32 patients in which either the clinical situation of the patient or the type of the study discouraged the protocol). Six additional patients were finally excluded from analysis because Wenckebach cycle length (WCL) was reached at a cycle length (CL) of.600 ms. As the aim was to analyse cases with fixed BBB, for the present study we excluded 10 patients with non-specific intraventricular conduction defects (IVCD). So, the final study population was 59 subjects. For the diagnosis of BBB and fascicular blocks we used the diagnostic criteria proposed by Willems et al. 5 All antiarrhythmic drugs were discontinued for at least five elimination half-lives prior to study. The protocol was approved by the Institutional Review Board of our Centre. Electrophysiological study After obtaining written informed consent, electrophysiological study was performed with patients in the fasting, unsedated state. For the purpose of the study, at least two quadripolar catheters were placed in the high right atrium and His bundle region. Tracings were recorded on a Prucka Cardiolab System (GE Medical Systems, Milwaukee, WI, USA) and filtered at Hz. Synchronized bipolar pacing was performed at twice diastolic threshold from the distal electrode pair of the catheter located at the right atrium. Incremental atrial pacing was performed as trains of at least 10 beats of rapid pacing at a constant rate, interrupted by at least five beats of SR after each pacing train. Pacing was started at a CL of 600 ms and then decreased by ms interval until atrioventricular (AV) block occurred. If WCL was reached at a CL of.500 ms, intravenous atropine was given (0.04 mg/kg) when possible. Electrocardiogram analysis Comparison of QRS complexes We analysed the electrocardiogram (ECG) during atrial pacing with the shortest CL that maintained 1:1 AV conduction (SAP-CL). We also analysed the ECG during atrial pacing with longer CL to determine at what CL changes began, and if changes were progressive while shortening atrial pacing CL. To avoid possible confusion due to superimposition of stimulus artefact, we used the last QRS complex of each atrial pacing train for analysis. It was compared with the QRS complex during SR. Criteria of configuration differences To minimize the subjectivity in the comparison of QRS configuration we used specific criteria previously described for the analysis of the spatial resolution of ventricular pace-mapping. 6,7 Differences were graded as: no differences, minor configuration differences or major configuration differences. Minor configuration differences consisted of: (i) appearance or disappearance of a notch; (ii) appearance or disappearance of a Q, R or S-wave that was,50% of the peak to peak QRS amplitude; (iii) a change.25% but,50% in the ratio of the amplitude of the individual component to the total QRS amplitude; and (iv) a change in the shape of a major component. Major configuration differences consisted of the following: (i) appearance or disappearance of a component that was.50% of the QRS amplitude; (ii) difference in the amplitude of a component of the QRS complex.50% of the total QRS amplitude; and (iii) two or more minor configuration differences. If a major configuration difference was noted, minor differences were ignored. Other analysis For descriptive purposes, we also searched for differences in QRS width, axis shift and BBB pattern. QRS duration during either atrial pacing or SR was always taken as the largest value of that parameter, measured in any ECG lead at a paper speed of 100 mm/s. QRS width differences of 10 ms or more, and axis shifts of.458, were taken into consideration. Statistical analysis Statistical analysis was performed using SPSS software (SPSS Inc., Chicago, IL, USA). Means (+SD) are reported for continuous variables. The x 2 test was used to assess the significance of the differences between proportions, and ANOVA or Student s t-tests were used for comparisons between means. To study independent predictors of QRS changes rates backward stepwise logistic regression analyses were performed, including variables with P, 0.25 on univariate analysis and plausibly related to the QRS changes. All the tests were twosided. The Holm method was used for correction of multiple testing. A P-value,0.05 was considered statistically significant. All ECGs were reviewed by two independent observers for interobserver agreement analysis; for the intra-observer analysis the time period that elapsed between the two readings by the observer was of at least 2 months. Agreement between data reported in a categorical way was appraised using a linearly weighted k index, 8 considering values,0.2, , , , and.0.81 to represent poor, fair, moderate, good, and excellent agreement. Results Our population consisted of 40 men and 19 women with an average age of years. Clinical and electrophysiological data for all patients are detailed in Table 1. Seventeen (59%) patients with right-bbb had bifascicular block, 14 had left anterior fascicular block and three left posterior fascicular block. Atropine was used in four subjects (7%) with WCL of.500 ms, to improve AV conduction. Configuration differences During atrial pacing minor configuration differences were seen in 22 (37%) patients, major differences in 23 (39%) subjects and no differences in 12 (24%) patients (Table 1). Table 1 also describes changes in QRS duration and axis shift. Figure 1 shows an illustrative example. No case was observed where QRS duration shortened to,120 ms, i.e. disappearance of BBB. No case was observed where the BBB pattern changed to the contralateral BBB block. Table 2 describes in detail the type of configuration change in each group of BBB. There was an excellent correlation, with an inter- and intra-observer agreement of 94% (linearly weighted kappa index of 0.81, 95% confidence interval ) and 95% (linearly weighted kappa index of 0.87, 95% CI ), respectively.

3 Rate-related changes in QRS morphology in patients with fixed BBB 2353 Table 1 Clinical and electrophysiological characteristics of patients (description of rate-related QRS changes) All patients (n 5 59) Clinical variables Age (mean + SD) Male sex, n (%) 40 (68) ICM, n (%) 22 (37) LVEF (mean + SD) a LVH, n (%) a 8 (15) Structural heart disease, n (%) 33 (61) QRS duration (mean + SD) Electrophysiological variables HV interval (mean + SD), ms WCL (mean + SD), ms QRS changes in relation to atrial pacing Major configuration changes 23 (39%) Only minor configuration changes 22 (37%) All configuration changes 45 (76%) Increase in QRS width 3 (5%) Decrease in QRS width 14 (24%) Axis shift.45º 1 (2%) Change to contralateral BBB 0 BBB, bundle branch block; ICM, ischaemic cardiomyopathy; LVEF, left ventricular ejection fraction; LVH, left ventricular hypertrophy; WCL, cycle length at which Wenckebach block was observed. a Evaluated in 55 (93%) patients with echocardiography. Variables related to configuration differences We analysed if the occurrence of QRS configuration changes during atrial pacing was related to a variety of variables (Table 3). Only the SAP-CL and the type of BBB were significantly related to the presence of changes, both in univariate (Table 3) and multivariable analysis (Table 4). The relationship of QRS changes to CL is further illustrated in Figure 2. As shown in the left panel, the shorter the SAP-CL the higher the probability of having QRS changes; the right panel depicts the likelihood of having QRS changes at each particular CL regardless of the final CL with 1:1 conduction in each particular patient. The likelihood of QRS changes increases from 17% at 100 b.p.m. to more than half at 120 b.p.m., and to.80% at 150 b.p.m. Furthermore, in the majority of patients (76%) changes were progressive, being more marked as paced CL was decreased (Figure 3). Figure 4 shows the relation between type of BBB and QRS changes. Patients with right-bbb virtually always developed QRS changes. Major QRS configuration differences were also related to SAP-CL (0% if SAP. 500, 18% if SAP-CL ms, and 53% if SAP-CL 400 ms; P, 0.01), but not to type of BBB (Table 2). Narrowing of QRS during atrial stimulation was also related to the SAP-CL (0% if SAP. 500, 17% if SAP-CL ms, and 30% if SAP-CL 400 ms; P ¼ 0.04), but not to the type of BBB (Table 2). In contrast, widening of QRS was only related to the left ventricular ejection fraction (LVEF); patients in whom QRS duration increased during atrial pacing had a mean LVEF of and subjects without widening of QRS had a mean Figure 1 Electrocardiogram tracings of a 48-year-old woman with severe systolic dysfunction due to non-ischaemic dilated cardiomyopathy, with left-bbb in sinus rhythm (SR); during atrial pacing (AP) major changes appear with a leftward axis shift (arrow), and a transition shift in pre-cordial leads (asterisk).

4 2354 T. Datino et al. Table 2 QRS changes in different types of intraventricular conduction defects Right-BBB Left-BBB P-value (n 5 29;49%) (n 5 30;51%) Major configuration changes Only minor configuration changes LVEF of (P ¼ 0.04). Axis shift of.458 occurred only in one subject with left-bbb and HV interval of 74 ms (Figure 1). Patients with supraventricular tachycardia Simultaneous 12-lead ECG during SVT was available in 14 patients (24%). ECG during tachycardia was recorded in the same body position as during atrial pacing. Mean age of these subjects ( years) was not different from the age of patients without SVT ( years; P ¼ 0.4). During SVT, QRS minor changes were observed in six cases (43%), major changes in seven additional patients (50%) and no changes in only one (6%). These changes were equal to those in ECG during atrial pacing at similar rates in all of them (Figure 3), except for one case. This patient had an SVT due to AV nodal re-entry with a CL of 420 ms and major QRS changes in comparison with the QRS complex in SR. SAP-CL was 460 ms, showing no QRS changes. Pattern of QRS complex changes We searched for a distinctive pattern of changes during atrial pacing. In patients with right-bbb typical changes included (Figures 3 and 5): the initial r in lead V1 (and/or V2), decreases or disappears along with an increase in amplitude and duration of R 0, that seems to move towards the initial r, even becoming a monophasic R-wave (with or without a notch), sometimes with a shortening of the QRS duration. In subjects with left-bbb, changes were less predictable. Typically, S-wave voltage increased in right pre-cordial leads (V1 V4) along with a slight shift to the left in the pre-cordial transition and, sometimes, an increase in the notch in V5 V6 (Figure 6). Discussion Main findings 12 (41%) 11 (37%) (52%) 7 (23%) 0.02 All configuration 27 (93%) 18 (60%) changes Increase in QRS width 2 (7%) 1 (3%) 0.61 Decrease in QRS width 9 (31%) 5 (17%) 0.20 BBB, bundle branch block. Our study shows that in patients with organic IVCD, QRS complex morphology can change in response to RAP. Although some of these changes are minor, we also observed important changes Table 3 Relation of QRS configuration changes to clinical and electrophysiological variables ECG configuration P-value Adjusted differences... P-value a No Yes (n 5 14) (n 5 45) Age (mean + SD) NS Male sex, n (%) 11 (77) 29 (64) 0.51 NS IVCD, n (%) Right-BBB, 2 (14) 27 (60), n ¼ 29 Left-BBB, n ¼ (86) 18 (40) ICM, n (%) 6 (43) 16 (37) 0.62 NS LVEF, n (%) b 50% 8 (62) 27 (64) 1.0 NS,50% 5 (39) 15 (36) LVH, n (%) b 1 (8) 7 (17) 0.67 NS Structural heart 9 (64) 27 (60) 0.77 NS disease QRS duration, ms, n (%) (29) 24 (53) 0.11 NS (71) 21 (47) HV interval NS (mean + SD), ms SAP-CL, ms, n (%) (43) 34 (76), (14) 9 (20) (43) 2 (4) BBB, bundle branch block; ICM, ischaemic cardiomyopathy; IVCD, intraventricular conduction defect; LVEF, left ventricular ejection fraction; LVH, left ventricular hypertrophy; SAP-CL, shortest atrial pacing cycle length that maintained 1:1 atrioventricular (AV) conduction. a P-value was adjusted by the Holm method. b Evaluated in 55 (93%) patients with echocardiography. Table 4 Independent predictors of QRS configuration changes Odds ratio (95% CI) P-value Right-BBB 7.5 ( ) SAP-CL a.500 ms ms 16.9 ( ) ms 18.1 ( ) BBB, bundle branch block; CI, confidence interval; SAP-CL, shortest atrial pacing cycle length that maintained 1:1 atrioventricular (AV) conduction. Variables included in the analyses were: QRS width, left ventricular ejection fraction, BBB and SAP-CL ( a codified as ms; 1 ¼ ms; ms).

5 Rate-related changes in QRS morphology in patients with fixed BBB 2355 Figure 2 Relation between pacing rate and QRS changes. Left panel: percentage of patients showing QRS changes at the shortest pacing cycle length with 1:1 atrioventricular conduction (SAP-CL). Patients are grouped according to whether the SAP-CL is.500, and,400 ms. Note that the shorter the SAP-CL the more frequent the appearance of electrocardiogram (ECG) changes. Right panel: percentage of ECG tracings showing QRS changes at each particular atrial pacing cycle length (AP-CL). Note that the shorter the AP-CL the more frequent the appearance of ECG changes. Figure 3 Rate-related progressive changes in a case with right bundle branch block (right-bbb). Electrocardiogram tracings of a 63-year-old woman without previous myocardial infarction, with right-bbb who showed progressive changes as atrial pacing (AP) rate was increased. Panels from left to right show sinus rhythm (SR), AP at a CL of 600, 500, 440, and 380 ms, and spontaneous termination of atrioventricular nodal reentrant tachycardia (AVNRT) followed by a premature ventricular contraction (PVC) and restoration of SR. Stimulus artefacts are marked with an asterisk. R-wave voltage increases and S-wave voltage decreases in leads V1, V2 and V3 as rate increases. During tachycardia (CL 390) QRS morphology is similar to QRS morphology during atrial pacing at a comparable rate. conventionally considered major changes (including significant shifts in the QRS-axis). In patients who had SVT, the QRS changes during tachycardia were identical to those during atrial pacing in most cases, confirming that these ECG changes are not laboratory artefacts. Thus, in patients with pre-existing BBB, QRS morphology during wide QRS complex tachycardia different to that during SR, does not necessarily imply that the tachycardia is ventricular.

6 2356 T. Datino et al. Previous studies Dongas et al., 4 based on observations during tachycardia in 18 subjects with IVCD (only nine had SVT), suggested that the presence of a pre-existing BBB makes differentiation between SVT and VT possible: if the QRS complexes during tachycardia are identical to Figure 4 Relation between type of intraventricular conduction defect and QRS changes. Patients with right bundle branch block (right-bbb) had electrocardiogram (ECG) changes more frequently. those during SR, the tachycardia is supraventricular; otherwise, the arrhythmia is ventricular. Our results appear to be at a variance with these observations, since major QRS changes appeared in.1/3 during supraventricular rhythms. Reasons to explain these differences include the following: Dongas et al. studied only a few patients with SVT and a case was included if they recorded at least three surface ECG leads during SR and tachycardia ; therefore, they did not have a 12-lead ECG of all the patients. Observers were asked to judge if QRS during SR and tachycardia were similar based on only two characteristics: BBB pattern in lead V1 and axis in leads I and II. Thus, they only looked for important changes and only in three leads, in a small series of patients. The methodology of the present study tries to overcome these limitations; by reproducing SVT with atrial pacing we could increase the sample size, we recorded 12-lead ECG in all cases, and used specific morphological criteria in the comparison of QRS configuration to minimize subjectivity. Other classical studies are in agreement with our observations but in a somewhat different setting. Using the atrial extrastimulus technique, several authors observed QRS changes in patients with wide QRS due to BBB Our study also shows QRS changes, but with continuous atrial pacing, closely resembling the situation during SVT. Observations with clinical implications Some of our observations may have clinical implications. QRS configuration changes were shown to be rate-dependent; they were more likely with shorter atrial pacing CL. These observations suggest that when a particular patient with pre-existing BB suffers SVT, the faster the tachycardia the more probable the appearance Figure 5 Electrocardiogram tracings of three patients with right bundle branch block who showed the typical pattern of changes during atrial pacing (AP). Left panel: 69-year-old man without prior myocardial infarction, central panel: 73-year-old woman with moderate aortic insufficiency without prior myocardial infarction, right panel: same patient as in Figure 3. Note: decrease in amplitude of the initial r with increase in amplitude of R 0 in lead V1 (and/or V2). See text for more details.

7 Rate-related changes in QRS morphology in patients with fixed BBB 2357 Figure 6 Electrocardiogram tracings of three patients with left bundle branch block without structural heart disease (left panel: 68-year-old man, central panel: 83-year-old woman, right panel: 44-year-old woman), who showed the typical pattern of changes during atrial pacing (AP). Note: an increase in voltage of the S-wave in the right pre-cordial leads. See text for more details. of QRS morphological changes. At a CL of 400 ms almost all subjects had QRS changes, and half have major QRS changes. The type of IVCD was an independent predictor of ECG changes during atrial pacing. Patients with right-bbb had ECG changes during atrial pacing more frequently than those with other IVCD. This suggests that QRS morphological changes during tachycardia should be interpreted more cautiously in right-bbb type tachycardias. An axis shift of.45º was observed in only one case (2%). May be this is why it was not observed in the nine cases of Dongas et al. However, this was described in response to extrastimuli; interestingly enough of this occurred in patients with a prolonged HV interval in SR, which was also the case in our patient. Contrary to what could be expected (rate-related fascicular block), it was never associated with right-bbb. Thus, important axis shifts during tachycardia in patients with pre-existent right-bbb may be specific for VT. Despite important QRS changes, we never saw a change in BBB pattern. This is in accordance with the observations of Dongas et al. and indicates that QRS morphology during tachycardia suggesting block of the contralateral bundle to that blocked during SR, may be specific for VT. Our finding of distinctive QRS morphological changes, particularly in cases with right-bbb, could also be of clinical help. When the changes during tachycardia are similar to those observed in our cases (see Results), SVT could be considered. Observations with mechanistic significance The criterion established by Dongas et al. 4 was based on the concept that in the presence of fixed complete BBB (and in the absence of accessory pathways) all beats of supraventricular (including sinus) origin conduct exclusively over the contralateral bundle and the QRS morphological pattern is expected to remain unaltered regardless of rate. Nevertheless, the electrocardiographic pattern of complete BBB, can appear with a mere slowing in conduction in a bundle branch. Classical studies using the atrial extrastimulus technique suggested the incomplete nature of the BBB. 9,13 More recently, it was observed that right bundle branch ablation in patients with bundle branch re-entry and left-bbb in SR, not always resulted in AV block, 14,15 but in a right-bbb pattern, demonstrating that the left bundle was not completely blocked. In such a context, rapid pacing could further deteriorate or improve conduction in the abnormal bundle. Furthermore, other phenomena such as appearance/disappearance of fascicular hemiblock in cases of right-bbb (whether complete or incomplete), or appearance/disappearance of distal blocks in the Purkinje network could be involved. Although our findings cannot fully differentiate among these mechanisms, some observations suggest the likely involvement of more than one. Since QRS changes were significantly more frequent in cases with right-bbb, fascicular block could be a priori the most likely mechanism. However, an axis shift in the presence of right-bbb would be the expected expression of a fascicular hemiblock, and, as previously mentioned, that was never seen, making this mechanism highly unlikely. In fact, since 59% of patients with right-bbb had fascicular hemiblock, additional abnormalities in the contralateral fascicle could have also resulted in infranodal block. Several findings suggest that an improvement in conduction in a bundle with depressed conduction, was a likely mechanism in some of our cases, particularly those with right-bbb. First, we observed

8 2358 T. Datino et al. rate-related QRS narrowing, significantly associated with right-bbb, and this is difficult to explain on the basis of complete block or further depression of already slowed conduction. Secondly, in some cases QRS changes appeared at a certain rate but were progressively more marked as atrial pacing rate increased, again difficult to explain on the basis of complete block. Thirdly, we were able to identify a common pattern of morphological changes, easier to explain by changes in an already depressed bundle than by distal Purkinje network block that would tend to be more variable. The physiologic rate-related shortening of refractoriness of the His Purkinje system, if preserved, could translate into an improvement in conduction, particularly if the right bundle branch is concerned. 19 On the other hand, further depression of conduction or distal Purkinje network block could be more likely in other cases, particularly in those with left-bbb. Widening of QRS complexes occurred in some cases, and was related to a decrease in LVEF, suggesting a severe myocardial disease with a severe depression of conduction in the His Purkinje system. This could represent the other end of the spectrum; cases with more localized and less severe abnormalities would tend to involve the right bundle with their physiologic relation to rate preserved, whereas cases with a more generalized abnormality would tend to relate to a more severe underlying disease, manifested by left-bbb and lose their physiologic relation to rate. Study limitations Our population could be biased in which we only evaluated cases that were referred for electrophysiological study. However, the clinical characteristics of our patients (wide range of ages, significant proportion of subjects with ventricular dysfunction) suggest that they represent a large spectrum of patients. We propose that atrial pacing at increasing rates mimics SVT, but we do not have tachycardia episodes of all the cases. Criteria that we used for comparison of QRS configuration differences, were described for the analysis of the spatial resolution of ventricular pace-mapping, and were not validated for our study. We did not analyse HV intervals during pacing, since several patients could not be consistently measured at all pacing rates. However, it has been shown that, although the HV intervals may increase at the initial unstable beat of the aberration phenomenon, they return to the values during SR after BBB is established. 20 Thus, it is unlikely that HV interval measurements during pacing could have shed further light on our study since in our cases BBB was already complete in SR. Conclusions In patients with wide QRS complex during SR, atrial pacing at increasing rates can result in important changes in QRS morphology, but neither in the disappearance of BBB nor in the development of the contralateral BBB pattern. Similar QRS differences appeared during SVT. Thus, if the 12-lead ECG morphology of a wide QRS tachycardia is different to that during SR, but with the same BBB pattern, it does not necessarily imply that the tachycardia is ventricular. Funding This study was funded by a Grant from Fundación para la Investigación Biomédica del Hospital Gregorio Marañon, Madrid, Spain. Conflict of interest: J.A. has received honoraria from Medtronic, Boston (former Guidant), Johnson & Johnson, and St Jude Medical for lectures, and has served as a consultant for Johnson & Johnson. The others report no conflict of interests. References 1. Akhtar M, Shenasa M, Jazayeri M, Caceres J, Tchou PJ. Wide QRS complex tachycardia. Reappraisal of a common clinical problem. Ann Intern Med 1988;109: Wellens HJ. Ventricular tachycardia: diagnosis of broad QRS complex tachycardia. Heart 2001;86: Drew BJ, Scheinman MM. Electrocardiogram criteria to distinguish between aberrantly conducted supraventricular tachycardia and ventricular tachycardia: practical aspects for the immediate care setting. Pacing Clin Electrophysiol 1995;18: Dongas J, Lehmann MH, Mahmud R, Denker S, Soni J, Akhtar M. Value of preexisting bundle branch block in the electrocardiographic differentiation of supraventricular from ventricular origin of wide QRS tachycardia. Am J Cardiol 1985;55: Willems JL, Robles de Medina EO, Bernard R, Coumel P, Fisch C, Krikler D, Mazur NA, Meijler FL, Mogensen L, Moret P. Criteria for intraventricular conduction disturbances and pre-excitation. J Am Coll Cardiol 1985;5: Kadish AH, Childs K, Schmaltz S, Morady F. Differences in QRS configuration during unipolar pacing from adjacent sites: implications for the spatial resolution of pace-mapping. J Am Coll Cardiol 1991;17: Kadish AH, Schmaltz S, Morady F. A comparison of QRS complexes resulting from unipolar and bipolar pacing: implications for pace-mapping. Pacing Clin Electrophysiol 1991;14: Brenner H, Kliebsch U. Dependence of weighted kappa coefficients on the number of categories. Epidemiology 1996;7: Wu D, Denes P, Dhingra R, Rosen KM. Bundle branch block. Demonstration of the incomplete nature of some complete bundle branch and fascicular blocks by the extrastimulus technique. Am J Cardiol 1974;33: Cannom DS, Goldreyer BN, Damato AN. Atrioventricular conduction system in left bundle-branch block with normal QRS axis. Circulation 1972;46: Barrett PA, Yamaguchi I, Jordan JL, Mandel WJ. Electrophysiological factors of left bundle-branch block. Br Heart J 1981;45: Moore EN, Knoebel SB, Spear JF. Concealed conduction. Cardiovasc Clin 1973;5: Surawicz B, Fisch C. Seminar on the changing role of electrocardiography in clinical practice-iv. Left bundle branch block: a continuously evolving concept. JAm Coll Cardiol 1987;9: Tchou P, Jazayeri M, Denker ST, Dongas J, Caceres J, Akhtar M. Transcatheter electrical ablation of right bundle branch: a method of treating macroreentrant ventricular tachycardia attributed to bundle branch reentry. Circulation 1988;78: Cohen TJ, Chien WW, Lurie KG, Young C, Goldberg HR, Wang YS, Langberg JJ, Lesh MD, Lee MA, Griffin JC. Radiofrequency catheter ablation for treatment of bundle branch reentrant ventricular tachycardia: results and long-term follow-up. J Am Coll Cardiol 1991;18: Denker S, Shenasa M, Gilbert CJ, Katar M. Effects of abrupt changes in cycle length on refractoriness of the His-Purkinje system in man. Circulation 1983;67: Miles WM, Prystowsky EN. Alteration of human right bundle branch refractoriness by changes in duration of the atrial drive train. Circulation 1986;73: Denes P, Wu D, Dhingra R, Pietras RJ, Rosen K. The effects of cycle length on cardiac refractory periods in man. Circulation 1974;49: Chilson DA, Zipes DP, Heger JJ, Browne KF, Prystowsky EN. Functional bundle branch block: discordant response of right and left bundle branches to changes in heart rate. Am J Cardiol 1984;54: Betts TR, Goldberger JJ, Kadish AH. Frequency and characteristics of progressive aberrancy during supraventricular tachycardia. Am J Cardiol 2003;92: