Panel A: Shows one patient with S-HBP where a transient but complete correction of acute RBBB appears to occur with S-HBP with recurrence as voltage was lowered to 3.5 V. However, if corrected and uncorrected beats are compared (first and second beats) there are subtle differences (L1 voltage and initial r in V1) suggesting that septal activation may have been different in corrected beat and an alternate pathway may have been activated.
Panel B: In remaining 3 patients with RBBB (one acute and 2 chronic) did not resolve with selective His bundle pacing at any pacing voltage.
Panel C: In one patient, chronic RBBB resolves partially, however with transient NS-HBP (3rd and 4th beats) but recurs abruptly as selective HBP occurs with voltage still at 5V.
In patients in S-HBP group, the site of block was distal to pacing lead and was not resolved with S-HBP.
Note HVAT intervals in NS-HBP and S-HBP of 93msec and 97 msec respectively.
Discussion
The observation that both acute and chronic right bundle branch block (RBBB) remained uncorrected with selective His bundle pacing (S-HBP) clearly suggests that site of RBBB was distal to pacing lead tip (Fig. 1-4). Intra-operative study of conduction times and velocity would indicate that the mean H-V intervals of 53±11 ms obtained in our study would place the lead tip in proximal His bundle; 2-3 cm from the bifurcation of His bundle into bundle branches and the ventricular septum (6). Even so, the RBBB was partially or completely corrected in all patients with non-selective (NS)-HBP where a characteristic ‘delta’ wave indicating that a parallel pathway, in addition to the His bundle, is also being activated (Fig. 1-4). From same pacing site, even in acute RBBB ostensibly located in close proximity to pacing lead, conduction block was rarely corrected with S-HBP.
Correction of RBBB appeared to correlate with pacing voltage, as higher pacing voltage (5V) resulted in complete correction of RBBB in 22 patients while only 10 showed complete correction at lower voltage.The voltage effect including a unique pattern of progressive beat by beat correction of RBBB with higher voltage was seen only during NS-HBP . Complete recurrence of RBBB was seen only with S-HBP never with NS-HBP (Fig. 1,4).
NS-HBP group showing lower capture threshold of the delta wave (1.3±0.5V,) had more patients with complete resolution of RBBB (9/14 patients) as compared to NS-S HBP group where the delta wave had higher capture threshold (2.4±0.8V) and only 3/11 patients had complete correction of RBBB.
Voltage related resolution of slow conduction has been observed in injured Purkinje fibers where higher voltage is required to generate a propagating impulse (7-10) and our observations during NS-HBP suggests that the parallel pathway may have properties similar to specialized conduction fibers. Increased pacing threshold is also predicted by virtual electrode polarization (VEP) theory and is felt to occur as a result of tissue damage (11,12). Based on planar model of cardiac muscle, the VEP theory does not explain how higher voltage would result in improvement of conduction particularly in His Purkinje system. In our study, in acute RBBB in which injured His bundle fibers are in close proximity to pacing lead tip a clinically significant VEP effect was not readily apparent in S-HBP where conduction is exclusively via His bundle.
The explanation why the parallel pathway in NS-HBP demonstrates voltage dependent phenomenon, may lie in the work of Effimov group which shows evidence of parallel pathways in the His region (13). The His bundle and surrounding area shows a rich density of gap junction proteins (Cx43, Cx40, and Cx45) extending from right lower extension (RLE) of the AV node to the interatrial septum and the latter has more than twice the expression of CX-43 compared to the His bundle (13,14). Thus, a molecular compartmentalization exists in the peri- Hisian region which connects directly to a specific Cx43‐positive domain of the His bundle (13). Hucker et al. have suggested that as these molecular pathways are not encased in fibrous tissue, pacing the RLE would not only activate the His bundle but may require lower voltage than direct pacing of His bundle itself (13).
Thus, one may speculate that the pacing lead may disrupt and activate subendocardial tissues rich in gap junction proteins which may behave as a parallel specialized conduction pathway injured by the active fixation process.
Thus, If one assumes as our observations suggest, that conduction through normal pathway (His bundle) results in RBBB and conduction through parallel pathway results in correction of RBBB then robust conduction through the parallel pathway at higher voltage may completely resolve RBBB while at lower pacing voltage, slower conduction in the parallel pathway would allow conduction through His bundle to manifest the RBBB and progressive slowing in the parallel pathway would result in progressively greater RBBB (Fig.1,4).
Two possible mechanisms whereby interaction between the two wavefronts may resolve RBBB is depicted in Figure 6. In one possibility, early excitation of the RV free wall may decrease or abolish ECG evidence of RBBB. This mechanism is suggested by the counterclockwise rotation of the NS-HBP complex (Table 1, Fig. 3), a consistent feature of NS-HBP (15) it may result from an early rightward wavefront.
The work of Durrer et al (16,17) and more recently of Almeida et al (18,19) lays out the specialized conduction pathway to early RV free wall pre-excitation in NS-HBP. In normal QRS activation (Fig.6), following left to right septal activation, the wavefront proceeds to the septal border of the crista supraventricularis (CSV) and meets the RV free wall wavefront exiting from right bundle branch (16-19)
(Fig. 6, left upper panel). During NS-HBP early activation of septum would also result in activation proceeding anteriorly to the crista supraventricularis (CSV) which alone connects the septum to RV free wall (20). This early activation wavefront, upon finding the RV free wall unexcited because of RBBB, would proceed to pre-excite the RV free wall (fig. 6, right upper panel), thus decreasing the duration of the S wave in Lead 1 without actually resolving the conduction delay at the distal site. This mechanism also explains why allpatients, particularly those in whom conduction block may be located more distally in the right bundle branch, showed at least partial ‘correction’ of RBBB .
The right lower panel in figure 6 shows another mechanism where the parallel pathway bypasses a more proximally related site of block. This mechanism may be a more plausible explanation of complete correction of RBBB. The transition from actively bypassing a more proximal RBBB to just passive pre-excitation of RV free wall at lower voltage may explain the change from complete resolution of RBBB to partial resolution of RBBB.
More than one parallel pathway with different capture thresholds would also explain the voltage dependent phenomenon.
The novel effect of NS-HBP in correcting a distal RBBB may seemingly be at odds with previous reports which suggest that given the longitudinal dissociation of the His bundle (21,22) only pacing the His bundle distal to site of block would correct bundle branch block (1,2). However, it appears that the narrowed QRS they described as ‘normal range, not normal activation’ (1), ‘with stim-Q interval shorter than H-V interval’ (1,2), would be similar to the NS-HBP QRS complex we observed with high voltage pacing. It is also likely that our use of 12 lead ECG with filter setting of 0-100 Hz allowed easier visualization of the delta wave as opposed to fewer leads with filter setting of 0-20 Hz in previous studies (1).
El Sherif et al also reported that in some cases higher pacing voltage was required to narrow the QRS complex and in discussing their results state that longitudinal dissociation in His bundle itself would only explain their conclusions if the transverse interconnections between the longitudinally dissociated fibers (23,24) become functionally inoperative and the conduction block was indeed located in proximal His bundle in all patients (2).
In summary, our previous observation that ventricular activation time decreases with increased pacing voltage in NS-HBP (5,15) (manuscript in review), is extended in this study with the novel finding of greater reduction of RBBB with higher voltage and there appears to be a stark difference between selective and non-selective His bundle pacing on correction of distal block. While we have attempted to explain the mechanisms involved with our current understanding of the A-V conduction system, the answers may well lie in the less well understood molecular biology of the His region (13, 14, 25). Further studies are needed to understand non-selective His bundle pacing, which appears to be distinctly different from the septal paced complex with HVAT of 146 ±26 ms reported by Vassallo et al (26), or the description of the ‘wide paced septal complex’ to which the term ‘non-selective His bundle pacing’ was originally applied. (27).
Fig. 6: Myocardial activation models in normal QRS, with non-selective His bundle pacing.