Clinical Cardiac Electrophysiology: Techniques and Interpretations
3rd Edition

SITE OF “BLOCK” OR CONDUCTION DELAY DURING BUNDLE BRANCH BLOCK
Interest in the site of block and/or conduction delay in the fascicles stems from the recognition that bifascicular block, especially right bundle branch block with left anterior hemiblock, is the most common ECG pattern preceding the development of A-V block. Because these ECG patterns are common, we need to elucidate factors that can predict who will develop A-V block. Determining the site of bundle branch block is the first step in this process.
Despite the ECG-anatomic correlates that have been made, the exact site of block or conduction delay producing bundle branch block patterns is not certain in all cases. Longitudinal dissociation with asynchronous conduction in the His bundle may give rise to abnormal patterns of ventricular activation (6); hence, the conduction problem may not necessarily lie in the individual bundle branch. This phenomenon has been substantiated in the case of both left bundle branch block and left anterior hemiblock by normalization of the QRS complex, H-V interval, and ventricular activation times during His bundle stimulation (Fig. 5-3) (7,8). These findings suggest that fibers to the left and right ventricles are already predestined within the His bundle and that lesions in the His bundle may produce characteristic bundle branch block patterns. Moreover, it is not uncommon to observe intra-His disease (widened or split His potentials) accompanying bundle branch block, particularly left bundle branch block.
FIG. 5-3. Normalization of left bundle branch block and left anterior fascicular block by distal His bundle (BH) stimulation. A. Sinus rhythm with an H-V interval of 55 msec. B. QRS complex normalization for the first three beats during distal BH stimulation with a PI (pacing impulse)–R interval of 40 msec, which is 15 msec shorter than the basal H-V interval. The last three beats during proximal BH pacing demonstrate a QRS complex identical to those in panel A and a PI-R interval (55 msec) equivalent to the basal H-V interval. (From: Narula OS. Longitudinal dissociation in the His bundle: bundle branch block due to asynchronous conduction within the His bundle in man. Circ 1977;56:996.)
The frequency with which conduction disturbances in the His bundle are responsible for the fascicular and bundle branch blocks is not known. The use of multipolar catheters to record distal, mid-, and proximal His bundle potentials or proximal right bundle and proximal His bundle potentials are of great use in delineating how frequently very proximal lesions result in a particular bundle branch block. It is theoretically appealing to postulate that such longitudinal dissociation in the His bundle causes the conduction abnormalities that ultimately result in complete A-V block in either the setting of acute anteroseptal infarction or sclerodegenerative diseases of the conducting system. The sudden simultaneous failure of conduction through all peripheral fascicles would appear much less likely than failure at a proximal site in the His bundle or at the truncal bifurcation. An example supporting this concept is shown in Figure 5-4, in which atrial pacing in a patient with left bundle branch block and split His potentials produces progressive prolongation of conduction and ultimate block between the two His potentials, that is, in the His bundle.
FIG. 5-4. Site of block in left bundle branch block. ECG leads I, II, III, V1 and V6 are shown with electrograms from the high-right atrium (HRA) and proximal (HISp) and distal (HISd) His bundle. On the left, during right atrial pacing at a cycle length of 380 msec, there is an intra-His conduction helay H-H’ of 85 msec but the H’-V is normal. With an abrupt to a paced cycle length of 360 msec, the H-H’ increases to 140 msec, then block occurs between the H and H’. Delay and block is therefore intra-His.
The site of transient bundle branch block may differ from that of chronic or permanent bundle branch block. Studies in the clinical electrophysiology laboratory in which distal His bundle or proximal right bundle branch recordings are used along with right and left ventricular endocardial mapping and/or intraoperative studies have both proved useful in better defining the sites of block in patients with either chronic or transient bundle branch block (i.e., aberration). Because the likelihood of developing complete A-V block may depend on the site of conduction or block in individual fascicles, obtaining such
P.113

data is critical to predicting risk of A-V block. Failure to do so may explain variability of published data in predicting progression of bifascicular block to complete A-V block. (See section on predicting risk of heart block.) Results of activation studies during bundle branch block are described in the following paragraphs.
Chronic Right Bundle Branch Block
Intraoperative studies have clearly shown that the electrocardiographic pattern of right bundle branch block can result from lesions at different levels of the conducting system in the right ventricle (9,10). These studies were performed during operative procedures for congenital heart
P.114

disease, the most common of which was repair of tetralogy of Fallot. Endocardial mapping was performed from the His bundle along the length of the right bundle branch, including the base of the moderator band, where it separated into the septal divisions, the anterior free wall, and the outflow tract. Epicardial mapping was also performed. The use of both endocardial and epicardial mapping clearly demonstrated three potential levels of block in the right ventricular conduction system that could lead to the electrocardiographic pattern of right bundle branch block: (a) proximal right bundle branch block, (b) distal right bundle branch block at the level of the moderator band, and (c) terminal right bundle branch block involving the distal conducting system of the right bundle branch or, more likely, the muscle itself (10).
Proximal right bundle branch block was the most common form noted after transatrial repair of tetralogy of Fallot (9,10), and similar studies in adults suggest it is also the most common spontaneously occurring type of chronic right bundle branch block (11,12 and 13). In proximal right bundle branch block, loss of the right bundle potential occurs where it is typically recorded (Fig. 5-5). Activation at these right ventricular septal sites is via transseptal spread from the left ventricle. This results in delayed activation of the mid- and apical right ventricular septum, the right ventricular anterior wall at the insertion of the moderator band, and the right ventricular outflow tract (RVOT) (Fig. 5-6). Epicardial mapping in patients preoperatively
P.115

and postoperatively shows a change from the normal right ventricular breakthrough at the midanterior wall with concentric spead thereafter, to a pattern showing no distinct right ventricular breakthrough but right ventricular activation occurs via transseptal spread following left ventricular activation. In Figure 5-7, this transseptal spread begins at the apex and then sequentially activates the midanterior wall and base of the heart. Left ventricular activation remains normal. In cases of proximal right bundle branch block, the mid- and apical septum are activated at least 30 msec after the onset of the QRS.
FIG. 5-5. Specialized conduction system map in proximal right bundle branch block (RBBB). ECG lead 3 is shown with electrograms recorded from the distal His bundle (H), proximal right bundle branch (PRB), distal right bundle branch (DRB), and anterior wall Purkinje fiber (PJ). Intervals (in milliseconds) measured from the conduction system electrograms to the onset of ventricular activation are shown. Before repair, sequential electrograms were recorded along the length of the right bundle branch. The ventricular septal electrograms recorded at each site occur early in the QRS complex. After repair, RBBB is present. A distal His bundle potential was recorded, but no other electrograms of the specialized conduction system could be recorded. The septal electrograms in the mid-RV and apical areas (PRB, DRB) are recorded later in the QRS. (From: Horowitz LN, Alexander JA, Edmunds LH Jr. Postoperative right bundle branch block: identification of three levels of block. Circ 1980;62:319.)
FIG. 5-6. Selected epicardial electrograms in proximal right bundle branch block. ECG leads 1, 2, and 3 are shown with a reference electrogram recorded in the left ventricle (LV) and mapping electrograms recorded at the right ventricular apex (RVA), right ventricular anterior wall (RVAW), and right ventricular outflow tract (RVOT) before and after transatrial repair of tetralogy of Fallot. The vertical lines indicate the onset of the QRS, and the numbers indicate time (in milliseconds) from the vertical line to the local electrogram. Before repair, the RVAW was the earliest site in the right ventricle. After repair, the earliest right ventricular site was the apex, which was activated 25 msec later than before repair. The latest ventricular epicardial activation occurred on the RVOT at 144 msec (cf. Fig. 5-7). (From: Horowitz LN, Alexander JA, Edmunds LH Jr. Postoperative right bundle branch block: identification of three levels of block. Circ 1980;62:319.)
FIG. 5-7. Epicardial activation in proximal right bundle branch block. ECG leads 1, 2, and 3 are shown with a schematic representation of the epicardial surface before and after transatrial repair of ventricular septal defect. Anterior and posterior projections of the heart are shown. (The lateral projection is omitted because no changes occurred in that segment of the left ventricle.) Activation times at selected epicardial sites are shown with 20 msec, and the right ventricular activation pattern was normal. After repair, the QRS duration was 92 msec, and activation at all right ventricular sites was delayed. Right ventricular activation began along the anterior interventricular groove and spread radially to the base. (From: Horowitz LN, Alexander JA, Edmunds LH Jr. Postoperative right bundle branch block: identification of three levels of block. Circ 1980;62:319.)
In distal right bundle branch block, activation of the His bundle and proximal right bundle branch is normal. Right bundle branch potentials persist at the base of the moderator band and are absent at the midanterior wall, where the moderator band normally inserts (Fig. 5-8). This form of bundle branch block only occurred when the moderator band was cut during surgery. Thus, this form of right bundle branch block is extremely rare as a spontaneous form of bundle branch block. In distal right bundle branch block, the apical and midseptum are normally activated, but activation of the free wall at the level of the moderator band is delayed, as is the subsequent activation of the RVOT (Fig. 5-9 and Fig. 5-10). Epicardial activation shows a small area of apical right ventricular activation that is the same as preoperatively. Activation at the midanterior wall, which was the site of epicardial breakthrough in the right ventricle preoperatively, was delayed,
P.116

as was subsequent activation of the remaining right ventricle (Fig. 5-10).
FIG. 5-8. Specialized conduction system map in distal right bundle branch block. ECG lead 3 is shown with electrograms recorded from the distal His bundle (H), proximal right bundle branch (PRB), two sites in the distal right bundle branch at the septal base of the moderator band (DRB1) and at the midportion of the moderator band (DRB2), and an anterior wall Purkinje fiber (PJ). Intervals (in milliseconds) measured from the conduction system electrograms to the onset of ventricular activation are shown. Sequential recordings along the course of the right bundle branch were obtained before the repair. After sectioning of the moderator band, conduction system recordings were obtained from PRB and DRB1. Electrograms recorded beyond this level showed no specialized conduction system potentials. The DRB2 electrogram recorded after repair is not shown because that area was resected. (From: Horowitz LN, Alexander JA, Edmunds LH Jr. Postoperative right bundle branch block: identification of three levels of block. Circ 1980;62:319.)
FIG. 5-9. Selected epicardial electrograms in distal right bundle branch block. ECG leads 1, 2, and 3 are shown with a reference electrogram recorded in the left ventricle (LV) and mapping electrograms recorded at the right ventricular apex (RVA), right ventricular anterior wall (RVAW), and right ventricle outflow tract (RVOT) before and after transatrial repair of tetralogy of Fallot. The vertical lines indicate the onset of the QRS, and numbers indicate time (in milliseconds) from the vertical line to the local electrograms. Before repair, the RVAW was the earliest right ventricular site. After repair, right bundle branch block (RBBB) was present, and the earliest activation was at the RVA. Despite the presence of RBBB, the RVA activation time after repair did not significantly differ from the value before repair. Activation of the RVAW and RVOT were delayed (cf. Fig. 5-10). (From: Horowitz LN, Alexander JA, Edmunds LH Jr. Postoperative right bundle branch block: identification of three levels of block. Circ 1980;62:319.)
FIG. 5-10. Epicardial activation in distal right branch block. ECG leads 1, 2, and 3 are shown with a schematic representation of the epicardial surface before and after transatrial repair of tetralogy of Fallot. Anterior and posterior projections of the heart are shown. Activation times at selected epicardial sites are shown with 20 msec isochrones. Before repair, the right ventricular activation pattern was normal. After repair, the QRS duration increased and a terminal slurred S wave appeared. The earliest right ventricular activation occurred at the right ventricular apex and along the midanterior interventricular groove. Right ventricular apical activation times before and after repair were similar. Activation of the remainder of the right ventricle was delayed; left ventricular activation did not change. (From: Horowitz LN, Alexander JA, Edmunds LH Jr. Postoperative right bundle branch block: identification of three levels of block. Circ 1980;62:319.)
Terminal right bundle branch block was the second most common form of right bundle branch block observed following either transatrial or transventricular repair of tetralogy of Fallot. Activation along the specialized conducting system (i.e., His bundle and and right bundle branches) remains normal up to the Purkinje-myocardial junction (Fig. 5-11). In contrast to proximal and distal right bundle branch block, activation of the midanterior wall is unchanged from normal, and only the right ventricular outflow tract shows delayed activation. Epicardial mapping demonstrates slowly inscribed isochrones from the infundibulum to the base of the heart (Fig. 5-12 and Fig. 5-13). Of note, terminal right bundle branch block, which is probably due to interruption of the terminal Purkinje network and/or intramyocardial delay can have two quantitatively similar but qualitatively different patterns (10).
FIG. 5-11. Specialized conduction system map in terminal right bundle branch block. ECG lead 3 is shown with electrograms recorded from the distal His bundle (H), proximal right bundle branch (PRB), distal right bundle branch (DRB), and anterior wall Purkinje fiber (PJ) before and after transatrial repair of tetralogy of Fallot. Intervals (in milliseconds) measured from the conduction system electrograms to the onset of ventricular activation are shown. Sequential electrograms were recorded along the length of the right bundle branch before and after repair despite the presence of right bundle branch block after repair. (From: Horowitz LN, Alexander JA, Edmunds LH Jr. Postoperative right bundle branch block: identification of three levels of block. Circ 1980;62:319.)
FIG. 5-12. Selected epicardial electrograms in terminal right bundle branch block (RBBB). ECG leads 1, 2, and 3 are shown with a left ventricular (LV) reference electrogram and with mapping electrograms recorded at the right ventricular apex (RVA), right ventricular anterior wall (RVAW), and right ventricular outflow tract (RVOT) before and after transatrial repair of tetralogy of Fallot. The vertical lines indicate the beginning of the QRS, and the numbers indicate time (in milliseconds) from the vertical line to the local electrograms. Before repair, right ventricular activation was earliest in the RVAW and latest in the RVOT, 67 msec after the onset of the QRS. After repair, RBBB was present and RVAW and RVA activation were not changed, but activation of the RVOT was delayed 45 msec (cf. Fig. 5-13). (From: Horowitz LN, Alexander JA, Edmunds LH Jr. Postoperative right bundle branch block: identification of three levels of block. Circ 1980;62:319.)
FIG. 5-13. Epicardial activation in terminal right bundle branch block. ECG leads 1, 2, and 3 are shown with a schematic representation of the epicardial surface before and after transatrial repair of tetralogy of Fallot. Anterior and posterior projections of the heart are shown. Activation times at selected epicardial sites are shown with 20-msec isochrones. Before repair, the right ventricular activation pattern was normal. After repair, right bundle branch block appeared, and outflow tract activation was delayed. Activation of the right ventricular anterior wall with a normal epicardial breakthrough site was unchanged after repair. (From: Horowitz LN, Alexander JA, Edmunds LH Jr. Postoperative right bundle branch block: identification of three levels of block. Circ 1980;62:319.)
During terminal bundle branch block produced by ventriculotomy, the latest activation is adjacent to the ventriculotomy scar. In contrast, when terminal bundle branch block is produced by transatrial resection, then the delayed activation appears as a smooth homogenous slow spread from the anterior infundibulum to the posterobasal aspects of the outflow tract. These data help resolve previously reported experimental work and clinical studies by a variety of authors (14,15,16,17,18 and 19). Our findings were similar to those of Wyndham (11), Wyndham et al. (12), and Van Dam (13).
Clinically, these findings (9,10,11,12 and 13) are relevant because terminal or distal bundle branch block, when accompanied by disorders of left ventricular conduction, may not indicate an increased risk of heart block. Such distal or terminal block has been seen in atrial septal defect, where
P.117

conduction in the right bundle branch is normal but stretching of terminal Purkinje fibers and/or muscle causes delayed activation of the right ventricular outflow tract (17). I have also seen terminal bundle branch block in patients with cardiomyopathy and chronic lung disease. Finally, I have also seen this pattern in rare patients with right ventricular infarction associated with the development of an RBBB pattern. In this latter case the duration of the QRS does not exceed 120 msec. In my opinion, these data favor a delay in intramyocardial conduction as the primary cause of terminal conduction delay. Determination of either proximal or terminal bundle branch block can easily be made in the clinical electrophysiology laboratory by demonstrating normal activation time of the mid- and apical septum in the presence of right bundle branch block. If right bundle branch block is proximal, activation of the mid- and apical septum will be delayed, producing a V (onset of the QRS) to local RV activation time exceeding 30 msec (20,21). If the right bundle branch block is the rare distal type or, more commonly, the terminal right bundle branch block, the local activation time at the mid- and apical septum will be normal (i.e., <30 msec), whereas that of the anterior wall (in the case of distal right bundle branch block) and/or the outflow tract (either distal or terminal right bundle branch block) will be delayed.
Left Bundle Branch Block
Far fewer data are available to evaluate the site of conduction abnormalities in left bundle branch block. No human studies have traced conduction from the His bundle down
P.118

the left bundle system, as done in right bundle branch block (9,10). Epicardial mapping data in a few patients with left bundle branch block (11,12 and 13) demonstrate (a) that right ventricular activation is normal but occurs relatively earlier in the QRS; (b) that a discrete left ventricular breakthrough site is absent, in contrast to normal, in which two or three breakthrough sites may be observed; (c) that transseptal conduction is slow, as manifested by crowded isochrones in the interventricular sulcus with more rapid isochrones along the left ventricular free wall. In left bundle branch block with normal axis, Wyndham et al. (11) found that the latest left ventricular site activated was not the A-V sulcus, as it is with normal intraventricular conduction. These studies, however, not only were limted by the small number of patients but by the fact that the authors did not consider myocardial disease and did not adequately address the effect of axis deviation. Furthermore, the extent to which intramural conduction delay contributed to QRS widening has never been assessed.
We therefore decided to evaluate endocardial activation during left bundle branch block in a heterogenous group of patients, including four with no organic heart disease, six with congestive cardiomyopathy, and eight with coronary artery disease and previous infarction (22). All but one patient had a QRS>140 msec. Only three patients had normal H-V times. Unfortunately, we did not record right bundle potentials or distal His potentials to localize the source of H-V prolongation (see following discussion).
We performed catheter mapping studies, as described in Chapter 2. We recorded standard activation sites (Fig. 5-14) in the right and left ventricle in sinus rhythm with fixed and variable-gain electrograms using both a 1 cm and 5 mm interelectrode bipolar recording. There was no difference in activation times between them. We therefore defined local activation as the point on the 1-cm variable-gain electrogram at which the largest rapid deflection crossed the baseline. When a fractionated electrogram was present without a surface discrete deflection >1 mV in amplitude, we used the rapid deflection of highest amplitude as local activation time. In addition, we measured the onset and offset of local activation from the fixed gain electrogram from the time the electrical signal reached 0.1 mV from baseline to the time of the amplifier decay signal (see Fig. 2-11). We defined transseptal conduction time as the difference between local activation time at the right ventricular septum (usually near the apex) and the earliest left ventricular activation time. We also evaluated the total left ventricular activation time, which was the difference in time from the earliest to the latest left ventricular endocardial activation. The average number of left ventricular sites mapped was 14 ± 3 per patient (range, 8 to 19). There was no difference in a number of sites mapped among any of the three groups.
FIG. 5-14. Schema of mapping sites in the right and left ventricles. (From Josephson ME, et al. Role of catheter mapping in the preoperative evaluation of ventricular tachycardia. Am J Cardiol 1982;49:207.)
P.119

Twelve of 18 patients had only one site of a left ventricular endocardial breakthrough. In nine patients, this was in the middle third of the left ventricular septum, and in three patients, it was at the apical third of the septum. In the remaining six patients, we observed simultaneous early activation at two left ventricular sites; in two patients, two sites were on the septum (one in the middle third and one at the apical septum), and in one, the apical septum and superior basal free wall. In contrast to the studies of Wyndham et al. (11), we found that the latest site of left ventricular activation was frequently at the base of the heart in patients with normal axis, while it was more variable in those with left axis (Table 5-1) (22).
TABLE 5-1. Relationship Between QRS Axis and Activation Sequence
Left ventricular endocardial activation began a mean of 52 ± 17 msec after the onset of the surface QRS (Table 5-2). We observed no difference in any of the three groups. In the normal patients, left ventricular activation began 44 ± 13 msec after the QRS, in patients with cardiomyopathy, left activation began 58 ± 13 msec after the onset of the QRS, and in the patients with prior infarction, left ventricular septal activation began at 51 ± 20 msec after the QRS.
TABLE 5-2. Results of Left Ventricular Mapping
The earliest activity in local anteroseptal sites was similar in the normal and cardiomyopathic groups, at 23 ± 9 and 23 ± 19 msec. However, earliest activation recorded in the septum in patients with prior anteroseptal infarction was only 11 ± 11 msec and was significantly shorter than that reported in the other groups (p < .05).
Using the difference from the earliest to latest activation times, total left ventricular endocardial activation was also much greater in the group with prior infarction (119 ± 32 msec) than in the other two groups (81 ± 26 and 61 ± 15, respectively) (p < .05). Total left ventricular activation time, as measured by the earliest onset to the latest offset of the fixed gain electrograms, was also much greater than in the group with prior infarction: 219 ± 77 msec versus
P.120

126 ± 37 msec in the normal patients and 125 ± 22 msec in the patients with cardiomyopathy. This difference was significant, p < .05. Of note, comparison of total left ventricular endocardial activity to the total QRS complex indicated that the duration of left ventricular endocardial activity in patients with prior infarction was 113 ± 34% of the QRS duration, while in the normal patients and cardiomyopathic patients, endocardial activation approximated 80% of the QRS duration. The latest site of left ventricular activation was most often near the base of the heart in the normal and cardiomyopathy group, just as it is in patients with normal QRS. However, the latest site of endocardial activation in patients with left bundle branch block associated with myocardial infarction was variable and was related to the site of previous infarction. Frequently, the latest site to be activated was within the site of prior infarction.
The interval between local activation at the right ventricular apex and the earliest rapid deflection noted in the left ventricle (i.e., transseptal activation) was similar in our groups of patients and averaged 33 msec. However, if one measured the interval between local activation at the right ventricular apex and the rapid deflection at the corresponding left ventricular site (site 2), it was longer, averaging 46 ± 50 msec; we noted no differences in any of the three groups. As noted earlier, when we used the fixed high-gain recording for earliest activation, the patients with prior anteroseptal infarction had the earliest activation recorded from onset of the QRS to the high-gain septal recording. We believe that this represents activation within the septum, which is thinner, and probably represents the right and medial part of the intraventricular septum. Patients with cardiomyopathy and normal hearts have thicker septums and therefore do not record right and intramural septal recordings from the endocardial surface of the left ventricle.
We performed right ventricular endocardial mapping in seven patients with left bundle branch block (22). One patient had a normal heart, three had cardiomyopathy, and three had prior infarctions. We mapped an average of six right ventricular sites. Right ventricular endocardial breakthrough occurred 8 ± 9 msec after the onset of the QRS and was usually at the midseptum. This corresponded with the initial 5% of the QRS complex. Activation then spread concentrically to the midanterior wall and the remainder of the septum, with latest activation at the RVOT. Total right ventricular endocardial activation using the beginning of the first to the last rapid deflection was completed in 36 ± 13 msec, which corresponded to the first 21 ± 7% of the surface QRS complex. We noted no difference in right ventricular activation in any of the groups. When measurements were made using high-gain electrograms, earliest activity typically preceded the onset of the QRS and was at the midseptal site similar to using the rapid deflection.
Thus, although rapid delay of transseptal activation is common to all forms of left bundle branch block, the type of heart disease markedly influences the subsequent pattern of left ventricular activation. Patients with prior and extensive infarction had the longer left ventricular activation times than those patients with no heart disease or cardiomyopathy. Of note, patients with cardiomyopathy and no heart disease had rapid left endocardial activation comparable to endocardial activation in patients with a normal QRS.
Examples of isochronic maps during left bundle branch block in patients with normal left ventricular endocardial activation and cardiomyopathy are contrasted with that of a patient with delayed activation associated with anterior infarction in Figure 5-15 and Figure 5-16. Analog recordings in comparable patients are shown in Figure 5-17 and Figure 5-18. In patients with cardiomyopathy, left ventricular endocardial activation is rapid and smooth. In contrast, in patients with infarction, left ventricular endocardial activation is markedly delayed and associated with abnormal conduction, manifested by fractionated electrograms and narrowed isochromes. Higher density mapping (60–200 sites) using the Carto System (Biosense) has confirmed these data.
FIG. 5-15. Isochronic map of left ventricular activation in a patient with a cardiomyopathy and two sites of left ventricular breakthrough. Numbers represent local activation times, and lines represent 10-msec isochrones. Note early breakthrough at apical septum (site 2, 65 msec) and basal superior free wall (site 12, 64 msec). The isochrones are widely spaced, demonstrating a normal left ventricular endocardial activation. (From: Vassallo JA, Cassidy DM, Marchlinski FE, et al. Endocardial activation of left bundle branch block. Circ 1984;69:914.)
FIG. 5-16. Isochronic map of left ventricular activation of a patient with coronary artery disease and one site of left ventricular breakthrough. Numbers represent local activation times, and lines represent 10-msec isochrones. Earliest left ventricular breakthrough is at midseptum (site 3) 45 msec after the onset of the QRS. Note closely aligned isochrones and total endocardial activation, which is prolonged, ending 158 msec after the onset of the QRS. (From: Vassallo JA, Cassidy DM, Marchlinski FE, et al. Endocardial activation of left bundle branch block. Circ 1984;69:914.)
FIG. 5-17. Analog map of patient with cardiomyopathy. Surface leads 1, aVF, and V1 are displayed with local electrograms from the right ventricular apex (RVA) and designated left ventricular (LV) sites. The duration of total LV endocardial activation is 40 msec. (From: Vassallo JA, Cassidy DM, Marchlinski FE, et al. Endocardial activation of left bundle branch block. Circ 1984;69:914.)
FIG. 5-18. Analog map of a patient with coronary artery disease. Surface leads 1, aVF, and V1 with local electrograms from the right ventricular apex (RVA) and designated left ventricular (LV) sites. The duration of total LV endocardial activation is 143 msec. (From: Vassallo JA, Cassidy DM, Marchlinski FE, et al. Endocardial activation of left bundle branch block. Circ 1984;69:914.)
We believe that the heterogeneity of endocardial activation in patients with left bundle branch block is a manifestation of the integrity of the distal specialized conducting system. Patients with normal hearts and cardiomyopathies appear to have intact distal conducting system and, hence,
P.121

early engagement and rapid spread through the rest of the intramural myocardium. In those patients with large anterior infarctions, the bulk of their distal specialized conducting system has been destroyed. As a consequence, their endocardial activation is via muscle-to-muscle conduction and thus is much slower. That hypothesis is strongly supported by our analysis of left ventricular endocardial maps in 40 patients during right ventricular pacing, which, we have shown, mimics left bundle branch block (23).
We compared patients with no heart disease, those with inferior infarction, and those with anterior infarction (23). The data again demonstrated that left ventricular endocardial activation patterns and conduction times were markedly influenced by the site and extent of prior infarction. We always observed longer endocardial activation times in patients with large anterior infarctions. Left ventricular activation times in patients with inferior infarction were intermediate between those without heart disease and those with anterior infarction. We believe this may be due to a lower density of His-Purkinje fibers, which contribute less to activation of the basal inferior wall, which is normally activated late in the QRS. Thus, inferior infarction would have less of an effect on total endocardial activation.
Analog records and isochronic maps in a patient with no infarction and one with an anterior infarction are shown in Figure 5-19, Figure 5-20 and Figure 5-21. They are remarkably similar to those comparable patterns recorded in spontaneous left bundle branch block. Thus, the only common bond in patients with left bundle branch block is a delay in transseptal activation.
FIG. 5-19. Analog map during right ventricular pacing. Surface leads 1, aVF, and V1 are displayed with local electrograms from the right ventricular apex (RVA) and 12 designated left ventricular (LV sites). Duration of LV endocardial activation is shorter in the group I patient without infarction (left) than in the group III patient with anterior myocardial infarction (AMI). See text for explanation. (From: Vassallo JA, Cassidy DM, Miller JM, et al. Left ventricular endocardial activation during right ventriuclar pacing: Effect of underlying heart disease. J Am Coll Cardiol 1986;7:1228.)
FIG. 5-20. Isochronic map during right ventricular (RV) pacing in a group I patient: no infarct, QRS duration, 180 msec. Left ventricular local activation times (in milliseconds) are indicated with 10-msec isochrones. Note one septal breakthrough site at 70 msec and complete activation in 58 msec. (From: Vassallo JA, Cassidy DM, Miller JM, et al. Left ventricular endocardial activation during right ventriuclar pacing: effect of underlying heart disease. J Am Coll Cardiol 1986;7:1228.)
FIG. 5-21. Isochronic map during right ventricular (RV) pacing in a group III patient: anterior myocardial infarction (AMI) QRS duration, 200 msec. Local activation times are indicated with 10-msec isochrones. Note one septal breakthrough site at 65 msec and complete activation in 103 msec. See text for explanation. (From: Vassallo JA, Cassidy DM, Miller JM, et al. Left ventricular endocardial activation during right ventriuclar pacing: effect of underlying heart disease. J Am Coll Cardiol 1986;7:1228.)
The pattern in which the left ventricle is activated initially (i.e., site of breakthrough), as well as the remainder of left ventricular endocardial and transmural activation, depends critically on the nature of the underlying cardiac disease. Thus, the bizarreness of the QRS, that is, the greater QRS width, is more a reflection of left ventricular
P.122

pathologic condition than of primary conduction disturbance. Our data (22,23) suggest that whenever the pattern of left bundle branch block is present, regardless of variability of QRS patterns, a similar degree of “block” in the left bundle branch is present in all groups of patients, at least as regards transseptal activation. This conclusion is at odds with others who suggest that left axis deviation is required for “complete” left bundle branch block (28). Thus, the risks of A-V block associated with the appearance of left bundle branch block should be similar in all groups of patients. However, the underlying cardiac disease is the major determinant of QRS width and morphology, left ventricular endocardial activation times, and overall mortality.
Transient Bundle Branch Block
Intermittent or transient bundle branch block (aberration) may have several mechanisms. These include (a) phase 3 block in which the initial aberrant complex is caused by encroachment on the refractory period (phase 3 of the action potential); (b) acceleration-dependent block in which at critical increasing rates (but well below the action potential duration) block occurs; (c) phase 4 or bradycardic-dependent
P.123

block, which is due to a loss of resting membrane potential owing to disease and/or phase 4 depolarization; and (d) retrograde concealment in which retrograde penetration of a bundle branch renders it refractory to subsequent beats. Both acceleration-dependent and bradycardic-dependent block are manifestations of a diseased His-Purkinje system and should be thought of as abnormal. Phase 3 block, however, is physiologic.
These proposed mechanisms of aberration can occur anywhere in the specialized conducting system. Unlike chronic bundle branch block, the site of block during aberration can shift. The use of intracardiac recordings that contain proximal and distal His bundle recordings or proximal His bundle and proximal right bundle branch recordings has been helpful in demonstrating this (25).
Proximal right bundle branch block has been defined by disappearance of a right bundle recording that was previously present. However, absence of a right bundle branch block potential may represent block proximal to the right bundle potential, slow conduction proximal to the right bundle potential so that it is activated during the QRS, or such slow decremental conduction in the right bundle branch that is not recordable as a “spike.” Distal block in the right bundle branch is said to exist when the proximal right bundle branch recording remains present during intermittent bundle branch block.
Most of the data suggest, however, that proximal block is responsible for at least the initial appearance of right bundle branch block during phase 3 block. An observation that supports this concept is that during increasing “degrees” of right bundle branch block in response to atrial premature stimuli, an increasing His to right bundle branch potential is observed before the development of complete block, when absence of the right bundle recording is observed (Fig. 5-22 and Fig. 5-23) (24).
FIG. 5-22. Right bundle branch potentials in incomplete and complete right bundle branch block (RBBB). A and B. The basic atrial cycle length is constant at 700 msec, and progressively shorter atrial coupling intervals (A1A2) are shown. The HV and RB-V interval during sinus beats (last beat in both panels) measure 50 and 25 msec, respectively, with an H-RB interval of 25 msec. Shortening in the S1S2 interval to 350 msec, A, results in an incomplete RBBB pattern, and a further increase in the H2–RB2 occurs (RB1–RB2 exceeds H1–H2 by 15 msec). B, During the complete RBBB pattern, no identifiable RB2 potential is recorded. Although not labeled, H1V1 and H2–V2 values measure the same in all panels, and the H1–H2 therefore equals V1V2. Tracings from top to bottom are ECG leads I, II, VI, high-right atrial (HRA) electrogram, His bundle electrogram (HBE), right bundle electrogram (RBE), and time lines (T). All measurements are in milliseconds. Pertinent deflections and intervals are labeled and for the most part self-explanatory. (From: Akhtar M, Gilbert C, Al-Nouri M, Denker S. Site of conduction delay during functional block in the His-Purkinje system in man. Circ 1980;61:1239.)
FIG. 5-23. Disappearance of right bundle potential during right bundle branch block (RBBB) pattern in two patients. A. Two recordings are from the His bundle (HB) region with HV intervals of 50 and 40 msec, respectively, and one is right bundle (RB) recording with an RB-V interval of 20 msec. With premature stimulation (S) at comparable atrial coupling intervals, the RB potential is not identifiable when the QRS complex shows an RBBB pattern (second and sixth beats) but is clearly recognizable when the QRS complex has a normal morphology (fourth beat). The HV interval measures the same for all beats. B. The HV and RB intervals are 40 and 10 msec, respectively. During premature atrial stimulation from the coronary sinus (CS, second and fifth beats), the QRS complex shows an RBBB pattern without concomitant change in the HV interval. The RB potential, however, disappears from its expected location and can be recognized within the local ventricular electrogram (small arrows), signifying a marked increase in H-RB interval after the atrial premature beats. All measurements are in milliseconds. (From: Akhtar M, Gilbert C, Al-Nouri M, Denker S. Site of conduction delay during functional block in the His-Purkinje system in man. Circ 1980;61:1239.)
P.124

In addition to the absence of the proximal right bundle branch potential, the time from the onset of QRS to the midseptal right ventricular recording site exceeds 30 msec, as noted previously, in chronic right bundle branch block. Almost invariably, when right bundle block aberration occurs at long cycle lengths, the delay and block are proximal to the right bundle branch recording. Distal delay (i.e., right bundle branch block in the presence of a normal His to right bundle branch potential recording) is rare but may occur at shorter cycle lengths (Fig. 5-24). Thus, there appears to be some cycle length dependency of the site of block, and shifts can occur. This is particularly so during rapid rhythms, whether paced or spontaneous, in which the initial site of block is almost always proximal with loss of a right bundle recording, but on subsequent complexes, the right bundle branch potential may reappear, suggesting a shift from proximal to distal site of block.
FIG. 5-24. Distal block in the right bundle (RB) branch. The basic atrial cycle length is 700 msec, and the H1V1 and RB1V1 measure 45 and 25 msec, respectively. A. The A1 conducts with a right bundle branch block (RBBB) pattern. B. A2 shows a block in the His-Purkinje system. In both panels, the H-RB interval is the same as sinus beats, and conduction delay and block are distal to the RB recording site. All measurements are in milliseconds. (From: Akhtar M, Gilbert C, Al-Nouri M, Denker S. Site of conduction delay during functional block in the His-Purkinje system in man. Circ 1980;61:1239.)
Akhtar (25) has clearly demonstrated a shift in the site of block during a rapid pacing producing 2:1 block below the His (Fig. 5-25). Although the initial site of block is below the His and above the recorded right bundle potential, on subsequent blocked atrial complexes, a right bundle potential reappears, suggesting that during trifascicular block the site of right bundle branch block is distal to the right bundle recording. Persistence of right bundle branch block during 1:1 conduction with a normal H-RB interval may be due to a shift to a distal site; however, I believe this more likely is due to retrograde concealed conduction with collision below the right bundle recording site (Fig. 5-25B).
FIG. 5-25. Migration of block in the right bundle (RB) branch during rapid pacing. A. A period of stable function 2:1 atrioventricular block in the His-Purkinje system before resumption of 1:1 conduction. The second, fourth, and sixth impulses are followed by His (H) but no RB potentials. The eighth atrial impulse, however, is followed by both H and RB potentials, and the H-RB interval measures the same as sinus beats. The tenth atrial impulse conducts with a left bundle branch block (LBBB) pattern and is preceded by prolonged H-V and RB-V intervals and a normal H-RB interval. The association of delay distal to RB recording in association with LBBB at this moment is because the site of delay had already shifted from a proximal to a distal location (beyond the RB potential) before resumption of 1:1 conduction. Perpendiculars are drawn in appropriate places to show the timing of the H-RB activation. B. Atrial pacing at a constant cycle length of 380 msec and two missed atrial captures (seventh and eight stimuli). The second atrial impulse is followed by H but no RB deflection, while the third atrial impulse conducts normally. The next two beats conduct with an LBBB pattern; the H-V and H-RB intervals preceding the first aberrant beat are prolonged but return to normal with the second aberrant beat. The RB-V interval preceding both beats with LBBB pattern measure the same as sinus beats (not labeled). During the first beat with an RBBB pattern (seventh QRS complex), the H-V interval is prolonged and no RB potential is recorded; however, the H-V and RB-V intervals preceding the second beat with RBBB measure the same as normal beats and are shown by perpendiculars. (From: Akhtar M, Gilbert C, Al-Nouri M, Denker S. Site of conduction delay during functional block in the His-Purkinje system in man. Circ 1980;61:1239.)
The persistence of aberration at longer cycle lengths than that at which aberration was initially noted strongly favors retrograde concealment as the mechanism of persistent aberration. A shifting site of block should ultimately lead to resumption of normal conduction unless persistence of retrograde concealment is also present. Furthermore, in the presence of a shift, one would expect changing His-to-RB and RB-V intervals, because these intervals should change as the site of block changes. The sudden resumption of normal H-RB and RB-V times (Fig. 5-26) is more consistent, I believe, with retrograde concealment. The fact that ventricular premature complex (VPCs) delivered during aberration can suddenly normalize conduction is further evidence that retrograde concealment is the responsible mechanism (see Chap. 6) (26).
FIG. 5-26. Site of block during perpetuation of right bundle branch block aberration. Same patient as Figure 5-22. A and B. The induction of right bundle (RB) branch block with two, (A) and three (B) successive premature atrial beats is shown. The H-V intervals following all premature beats in both panels measure the same as sinus beats (not labeled). After A2, no RB potentials can be identified in eight of the panels; however, the RB potentials with subsequent premature beats (RB3 and RB4) are clearly recognizable. Although not labeled, the H3RB3, H4RB4, and corresponding RB-V intervals measure the same as sinus beats. (From: Akhtar M, Gilbert C, Al-Nouri M, Denker S. Site of conduction delay during functional block in the His-Purkinje system in man. Circ 1980;61:1239.)
Our experience is consistent with Akhtar’s (25) in that the initial beat of long-short-induced right bundle branch block aberration almost always is proximal. The shift in site of block at short cycle lengths from proximal to distal may merely represent the difference in shortening of refractoriness at different levels in the right bundle branch, which may be complicated by retrograde invasion of the right bundle following this development of RBBB. Chilson et al. (27) showed that the refractory period of the right bundle branch shortens to a greater degree than that of the left bundle branch system at increasing heart rates. This leads to the greater ability to demonstrate left bundle branch block at short drive cycle lengths and right bundle branch block at long drive cycle lengths. This is illustrated
P.125

in Figure 5-25B where long H-H intervals precede RBBB and shorter H-H intervals are associated with LBBB. The shortening of right bundle branch refractoriness at shorter drive cycle lengths might also allow the impulse to penetrate distally to the proximal site of block noted at long drive cycle lengths.
Transient left bundle branch block aberration is less common than right bundle branch block. In our laboratory, approximately 25% of phase 3 type aberration is of the left bundle branch block variety. This is approximately that seen in prior studies (27). Use of multiple electrode recordings along the His bundle are particularly useful in demonstrating that left bundle branch conduction delays are very proximal and, I believe, are in the His bundle itself. As demonstrated previously, in Figure 5-4, an H-H’ often can be noted, and when block appears, it frequently
P.126

occurs between His spikes (Fig. 5-4). These data, along with prior data in chronic left bundle branch block demonstrating normalization of the QRS by His bundle pacing, further support the His bundle origin of the conduction defect. Longitudinal dissociation in His bundle may cause individual fascicular block, which has been suggested by the observation that catheter manipulation in the His bundle region can produce left anterior hemiblock. Furthermore, if one records multiple His potentials, or even a His and proximal right bundle potential, the appearance of left anterior hemiblock in a patient with left bundle branch is often associated with an intra-His conduction delay, or proximal His to proximal right bundle delay, as seen in Figure 5-27. In this figure, a leftward shift in axis is accompanied by H-V prolongation, which is caused by a 70-msec increment between proximal and distal His bundle recordings.
FIG. 5-27. Longitudinal dissociation in the His bundle causing left anterior hemiblock. (Same arrangement as Figure 5-4.) Left bundle branch block with a 10° axis is shown. A premature atrial complex, A1, results in marked left axis deviation and is associated with an increase between proximal (HISp) and distal (Hisd) His deflections. No change of distal HiS-V occurs during the premature complex.
Regardless of axis deviation, the initiation of left bundle branch block aberration by an atrial premature beat is almost always accompanied by an increase in H-V interval. In the majority (approximately 75%) of cases, multiple recordings from the His bundle or between the His and
P.127

right bundle branch show incremental delays within the His bundle or, when a right bundle branch potential is measured, proximal to that right bundle branch potential (Fig. 5-28).
FIG. 5-28. Site of conduction delay in left bundle branch block. Two His bundle electrograms (HBE) and a right bundle electrogram (RBE) are shown. The H1–V1 and RV1–V1 are 40 and 20 msec, respectively. An atrial premature complex, A2, produces a marked increase in H2RB2 interval (320 msec), which is associated with left bundle branch block aberration. Normal conduction beyond the RBE is evident by a persistent RB-V of 20 msec. This points to the His bundle as the location of the site of left bundle branch delay. T = Time line. (From: Akhtar M, Gilbert C, Al-Nouri M, Denker S. Site of conduction delay during functional block in the His-Purkinje system in man. Circ 1980;61:1239.)
In a smaller number of cases, the H-V interval is prolonged by 5 to 15 msec without a change in H-RB or H-RV, suggesting that the increase in H-V is due to a relative difference in ventricular activation over the left and right bundle branches and is not due to conduction delay in the His bundle or contralateral bundle (Fig. 5-29). In the latter instances, the H-RB or H-RV would increase. These findings strongly suggest that the His bundle is probably longitudinally dissociated into fibers predestined to serve the right ventricle (right bundle branch) and left ventricle (anterior and posterior fascicle). Distinguishing between the intra-His site and the truncal site just proximal to the
P.128

division of the right and left bundle is impossible; however, I believe that in the majority of cases block at a very proximal site is responsible for both transient and permanent bundle branch block. In fact, when bifascicular block involving both bundle branches is observed, I believe it is only when the site of conduction delay and/or block is proximal that the risk of developing spontaneous heart block is increased. This is an important concept when one uses intracardiac recordings to predict risk of A-V conduction disturbances (see following discussion).
FIG. 5-29. Effect of LBBB on the H-V interval. During sinus rhythm (first two complexes) the H-V interval is 45 msec and the H-RV interval is 75 msec. An atrial extrastimulus (arrow) results in LBBB aberration and an increase in H-V to 60 msec. However, the H-RV remains the same. Thus despite the increase in H-V interval, the conduction down the right bundle branch is unaltered. Therefore, in the presence of LBBB, the H-V interval may be prolonged to 60 msec in the absence of trifascicular delay. See text for discussion.
Bradycardia or phase 4 block almost always manifests a left bundle branch block pattern. I have not yet seen a case of isolated bradycardia-dependent right bundle branch block, although there is no particular reason why this should not occur. A possible explanation for this observation is that the left ventricular conducting system is more
P.129

P.130

susceptible to ischemic damage and has a higher rate of spontaneous phase 4 depolarization than the right; therefore, it is more likely to be the site of bradycardia-dependent block. This may be one reason why some cases of bradycardia-dependent left bundle branch block demonstrate an H-V interval no different from the normal H-V interval. That is, the His-to-right bundle branch potential, or intra-His delay, is not the source of the conduction defect; but that delay in the distal conducting system is responsible of bradycardia-dependent left bundle branch block. Moreover, it has a higher rate of underlying automatically than the right bundle branch system. An example of bradycardia-dependent left bundle branch block is shown in Figure 5-30; a change from 1:1 conduction to 2:1 conduction with block in the A-V node resulted in a slower rate of engagement of the His-Purkinje system and the development of left bundle branch block. Acceleration-dependent block in contrast is observed in both the RBBB and LBBB.
FIG. 5-30. Bradycardia-dependent left bundle branch block. A. Atrial pacing at a cycle length of 800 msec with 1:1 A-V conduction and normal intraventricular conduction. B. Atrial pacing at a cycle length of 545 msec, 2:1 block in the A-V node, and an effective cycle length in the His-Purkinje system of 1090 msec. A widened QRS complex with a left bundle branch block configuration is evident.
Clinically, both tachycardia or acceleration-dependent and bradycardia-dependent bundle branch blocks are often seen in the same patient with an intermediate range of cycle lengths and normal conduction. The range of normal conduction may be quite broad, with rapid pacing and/or prolonged carotid sinus pressure required to produce the extremes of cycle length that precipitate the bundle branch block (28). In some cases, however, the range of cycle lengths resulting in normal conduction may be very narrow, and to some extent variable and affected by drugs and other factors. However, the isolated appearance of bradycardia-dependent block is almost always associated with left bundle branch block. The clinical implications of rate-dependent bundle branch block are not clear. Acceleration-dependent and bradycardia-dependent block usually occur in diseased tissue and, in the setting of infarction, usually an inferior infarction. The prognosis for such dependent blocks is purely related to the underlying cardiac disorder. On the other hand, phase 3 block is a physiologic normal response to encroachment on the refractory period of the bundle branch. Therefore, in and of itself, it should have no adverse prognostic implications.