Electrocardiographic predictors of validated right ventricular outflow tract septal pacing for correct localization of transthoracic echocardiography

Abstract

Background

Electrocardiographic (ECG) characteristics of true right ventricular outflow tract (RVOT) septal pacing have not been clearly demonstrated.

Hypothesis

We hypothesized that ECG parameters would help operators differentiate true RVOT septum from non‐septal septum.

Methods

We analyzed 151 patients who underwent pacemaker implantation with a ventricular lead in the RVOT. Transthoracic echocardiographic (TTE) determination of pacing sites was applied in all patients after implantation. A 12‐lead ECG was recorded during forced ventricular pacing.

Results

According to TTE orientation, pacing at the RVOT septum was achieved in 94 patients (62.3%). Compared with nonseptal pacing, septal pacing had significantly shorter QRS duration (139.2 ± 18.5 ms vs 155.5 ± 14.7 ms; P < 0.001). More frequent negative or isoelectric QRS vector in lead I (76% vs 32%; P < 0.001), lead II/III R‐wave amplitude ratio < 1 (52% vs 25%; P = 0.001), and aVR/aVL QS‐wave amplitude ratio < 1 (59% vs 32%; P = 0.001) were observed in septal pacing. Transitional zone (TZ) score (3.8 ± 0.96 vs 4.2 ± 0.90; P = 0.004) and TZ index (0.3 ± 0.5 vs 0.6 ± 0.7; P = 0.008) were significantly lower in septal pacing than in nonseptal pacing, respectively. In multivariate analysis, paced QRS duration and negative or isoelectric QRS vector in lead I independently predicted RVOT septal pacing (P < 0.001). At ROC curve analysis, paced QRS duration ≤145 ms identified RVOT septal pacing with 85.1% sensitivity and 78.9% specificity.

Conclusions

This study reveals the heterogeneity of lead placement within the RVOT. Narrower paced QRS duration and negative or isoelectric QRS vector in lead I independently predict RVOT septal pacing.

1. INTRODUCTION

The right ventricular apex (RVA) has been a traditional pacing site due to its stable and reliable pacing parameters. However, chronic RVA pacing is associated with the onset of atrial fibrillation, impaired left ventricular (LV) function, and mortality.1, 2, 3, 4 This deleterious effect has led to an interest in selective pacing sites, so as to achieve a more physiological pattern of ventricular activation.

The most studied alternative site has been the right ventricular outflow tract (RVOT), with increasing focus on the septal aspect of this structure. RVOT septal pacing is theoretically associated with a more physiological LV activation pattern and contraction compared with RVA pacing.5 However, there are controversial data on its benefits.6, 7, 8 The RVOT is regarded as a spatial trapezoid‐shaped area. On the basis of the anatomy described by Mond et al.,5 the structure of the RVOT can be divided into 4 segments: septum, anterior, posterior, and free wall. So targeting the true RVOT septum may be technically challenging. Several studies have shown that, despite using the strict radiological criteria and a special preshaped stylet for placing the lead on the septum, the final position was heterogeneous.9, 10, 11 A percentage of “RVOT septal” leads may not have been on the true septum, but instead on possible detrimental pacing sites such as the anterior or free wall.12, 13

Electrocardiographic (ECG) criteria of paced QRS complex (in particular, a negative or isoelectric QRS vector in lead I) have been proposed to confirm the RVOT septal position. However, the ECG criteria have not been properly validated, as the true lead position was not confirmed by any imaging modalities other than fluoroscopy in these studies. Therefore, it is still unknown whether the ECG criteria could apply to the true RVOT septal position. The purpose of this study was to investigate the paced ECG characteristics using transthoracic echocardiography (TTE) for accurate anatomical localization after the procedure.

2. METHODS

2.1. Study population

The study retrospectively enrolled 160 consecutive adult patients (age > 18 years) undergoing permanent pacemaker implantation for accepted indications between July 2012 and July 2015. Exclusion criteria were patients with leads inserted in the RVA; indications for an implantable cardioverter‐defibrillator or cardiac resynchronization therapy; clinical manifestations of congestive heart failure, chronic atrial fibrillation, moderate or greater degree of valvulopathy, or chronic obstructive pulmonary disease; and poor echocardiographic windows and image qualities. All subjects gave written informed consent to the study. The institutional ethics review board of the hospital approved the study.

2.2. Implantation procedure

The pacemaker implantation was performed by experienced operators with patients under local anesthesia using conscious sedation. Prophylactic intravenous antibiotics were given half an hour before the procedure. The RV lead was placed via the left or right subclavian venous approach. A standard 58‐cm or 60‐cm bipolar steroid‐eluting active fixation lead (CapSureFix Novus MRI 5076, Medtronic Inc., Minneapolis, MN; Tendril ST 1888TC, St. Jude Medical Inc., St. Paul, MN; or Setrox S60, Biotronik GmbH, Berlin, Germany) was inserted into the RVOT septum using a hand‐shaped stylet with generous curve and a terminal straight bend with posterior angulation as previously described by Mond et al.5 With the stylet fully deployed, the lead was advanced into the pulmonary artery and then gradually withdrawn below the pulmonary valve into the RVOT guided by the posteroanterior position. The 40° right anterior oblique projection was used to prevent inadvertent positioning in the coronary sinus and great cardiac vein. The RVOT septal pacing site was confirmed once the 40° left anterior oblique fluoroscopic view showed the lead tip facing toward the spine (see Supporting Information, Figure S1, in the online version of this article). A 53‐cm passive fixation lead was inserted into the right atrial appendage.

The ventricular stimulation threshold at a 0.48‐ms pulse width, R‐wave amplitude, and lead impedance was measured several minutes after screw deployment. Perioperative complications requiring intervention were also recorded.

2.3. Determination of pacing sites by TTE

TTE was used for exact documentation of the anatomical location of pacing sites.9, 14 TTE was performed in all patients by 2 observers who were blinded to the results of fluoroscopy 3 days after pacemaker implantation. Disagreements between observers were resolved by consensus. Echocardiography was carried out in the left lateral decubitus position with a commercially available ultrasound transducer and equipment (iE33 S5‐1; Philips, Bothell, WA). Images were acquired from parasternal short‐axis (PSAX) views to detect the exact location of the tip of the lead as it screwed into the myocardium. The PSAX window at the level of the aortic valve was used primarily to identify the lead implanted within the RVOT. Then, the tip of the lead was actively tracked using all possible PSAX planes. Finally, the tip of the lead was completely exposed and verified. The position of the lead was attributed to the septal group if both the direction of the tip of the lead and its insertion site were seen to indicate the interventricular septum. If the previous criteria were not met, the position of the lead was attributed to the nonseptal group (Figure 1).

Figure 1.

Examples of the tip of lead (arrows) in parasternal short‐axis view: (A) RVOT septum and (B) RVOT nonseptum. Abbreviations: LV, left ventricle; RV, right ventricle; RVOT, right ventricular outflow tract

2.4. Paced ECG characteristics

Three days after pacemaker implantation, a 12‐lead ECG was recorded during forced ventricular pacing (VVI, 10 bpm above baseline ventricular rate) at a paper speed of 25 mm/s, with chest and limb leads placed in standard positions. The following parameters were obtained by an observer blinded to the results of the TTE: (1) QRS width; (2) QRS frontal axis; (3) QRS morphology in leads I, II, III, aVF, aVR, and aVL; (4) presence of notching of QRS complex in inferior leads; (5) R‐wave amplitude in inferior leads and R‐wave amplitude ratio of II/III; and (6) Q‐wave amplitude in leads aVR and aVL and the QS wave ratio of aVR/aVL.

The QRS duration was measured from the site of earliest initial deflection from the isoelectric line in any lead to the time of latest activation in any lead. The vector of the QRS complex was described in all limb leads as positive, negative, or isoelectric (defined as summed voltage of between –0.05 and 0.05 mV). The transitional zone (TZ) was defined as the position of the precordial leads in which the R‐wave amplitude and S‐wave amplitude were equal (ie, the point at which the predominantly negative QRS complex changes to a predominantly positive complex). The TZ score, as described by Yoshida et al.,15 was calculated. According to the site of the TZ in the precordial leads, TZ score was graded in 0.5‐point increments. If the R/S‐wave amplitude ratio was between 0.9 and 1.1 in any lead, the TZ score was the same as the lead number (Table 1). The TZ index was defined as TZ score of pacing minus the TZ score of sinus beat.

Table 1.

Value of TZ score based on site of R‐wave transition in precordial leads

Site of R‐Wave Transition in Precordial Leads	TZ Score
<V1	0.5
V1	1.0
V1< <V2	1.5
V2	2.0
V2< <V3	2.5
V3	3
V3< <V4	3.5
V4	4.0
V4< <V5	4.5
V5	5.0
V5< <V6	5.5
V6	6.0
>V6	6.5

Abbreviations: TZ, transitional zone.

2.5. Statistical analysis

Data are expressed as mean ± SD. The Mann–Whitney U test and the Student t test were used for comparison of continuous variables between groups. Categorical variables were compared by the χ² test between groups. Candidate factors predicting RVOT septal pacing were identified by a multivariate logistic regression model using a stepwise regression procedure (with P > 0.1 as the default criterion for eliminating all variables). Receiver operating characteristic curves corresponding to the selected logistic regression models were constructed and the area under the curve was calculated to offer a summary measure of the accuracy of the prediction model. A 2‐tailed value of P < 0.05 was considered statistically significant. All statistical analyses were performed using SPSS version 19.0 (IBM Corp., Armonk, NY) for Windows.

3. RESULTS

3.1. Patient demographics

Of the 160 enrolled patients, 9 patients were excluded due to echocardiographic images of insufficient quality. That left 151 patients eligible for recruitment in this project, with a mean age of 61.2 ± 15.2 years. Sixty‐three patients (42%) were males. The indications for pacemaker implantation were as follows: 99 for sick sinus syndrome and 52 for high‐degree atrioventricular block. Lead placement was successful in all cases, with no procedural complications.

Echocardiographic determination of pacing sites was available in 151 patients. A disagreement in lead‐tip positioning between 2 blinded echocardiography observers occurred in 10 patients (6.6%). Of these patients, the tip inserted in the RVOT septum was ultimately determined in 7 patients. The rest of patients were attributed to the nonseptal group. Finally, patients were divided into 2 groups according to the lead position determined by TTE: septal (n = 94) and nonseptal (n = 57) group. Pacing at the RVOT septum was achieved only in 62.3% of patients. There were no significant differences with regard to age, preoperative echocardiographic parameters, RV lead pacing threshold, lead impedance, and pacemaker indication between the groups (Table 2).

Table 2.

Patient basic clinical characteristics

	Septal Pacing, n = 94	Nonseptal Pacing, n = 57	P Value
Age, y	59.3 ± 16.0	64.2 ± 13.5	0.06
Male sex, %	40	44	0.78
SSS, %	68	61	0.4
Pre‐QRS duration, ms	97.6 ± 19.5	91.9 ± 24.8	0.17
LVEDD, mm	45.6 ± 4.5	46.1 ± 5.2	0.73
LVESD, mm	28.4 ± 4.8	29.3 ± 4.2	0.77
LVEF, %	65.6 ± 6.1	65.1 ± 5.3	0.49
RA diameter, mm	43.5 ± 5.6	44.3 ± 4.8	0.26
RV diameter, mm	47.5 ± 5.4	48.2 ± 4.6	0.34
RV threshold, V	0.53 ± 0.22	0.65 ± 0.19	0.20
RV impedance, Ω	546.8 ± 148.1	617.5 ± 218.2	0.56

Abbreviations: LVEDD, left ventricular end‐diastolic diameter; LVEF, left ventricular ejection fraction; LVESD, left ventricular end‐systolic diameter; RA, right atrial; RV, right ventricular; SD, standard deviation; SSS, sick sinus syndrome.

Data are presented as mean ± SD unless otherwise noted.

3.2. Paced ECG parameters

With forced ventricular pacing, septal pacing was associated with a shorter paced QRS duration compared with nonseptal pacing (139.2 ± 18.5 ms vs 155.5 ± 14.7 ms, respectively; P < 0.001). Negative or isoelectric QRS vector in lead I (76% vs 32%; P < 0.001), lead II/III R‐wave amplitude ratio < 1 (52% vs 25%; P = 0.001), and aVR/aVL QS‐wave amplitude ratio < 1 (59% vs 32%; P = 0.001) were more frequently observed in septal pacing compared with nonseptal pacing. The TZ score of septal pacing was lower, which meant earlier R‐wave transition, compared with nonseptal pacing (3.8 ± 0.96 vs 4.2 ± 0.90, respectively; P = 0.004). The TZ index was significantly lower in the septal pacing group than in the nonseptal pacing group (0.3 ± 0.5 vs 0.6 ± 0.7, respectively; P = 0.008). However, there was no difference in the presence of notching of QRS complex in inferior leads (Table 3).

Table 3.

Paced ECG characteristics of patients

	Septal Pacing, n = 94	Nonseptal Pacing, n = 57	P Value
QRS duration, ms	139.2 ± 18.5	155.5 ± 14.7	<0.001
Frontal axis, degrees	88.2 ± 13.4	77.6 ± 14.2	<0.001
TZ score	3.8 ± 0.96	4.2 ± 0.90	0.004
TZ index	0.3 ± 0.5	0.6 ± 0.7	0.008
Neg/iso in lead I	71 (76)	18 (32)	<0.001
Notching in inferior leads	9 (10)	10 (18)	0.15
II/III R‐wave amplitude ratio < 1	49 (52)	14 (25)	0.001
aVR/aVL QS‐wave amplitude ratio < 1	55 (59)	18 (32)	0.001

Abbreviations: ECG, electrocardiographic; Neg/iso, negative/isoelectric; SD, standard deviation; TZ, transitional zone.

Data are presented as n (%) or mean ± SD.

3.3. Predictors of RVOT septal pacing

In the univariate analysis, paced QRS duration, negative or isoelectric QRS vector in lead I, TZ score, and TZ index were significantly associated with RVOT septal pacing. Multivariate analysis confirmed paced QRS duration and negative or isoelectric QRS vector in lead I as predictors of RVOT septal pacing. ROC curve was constructed from paced QRS duration. Predictive performance increased with narrower paced QRS duration showing an area under the curve of 0.835. A paced QRS duration ≤145 ms yielded 85.1% sensitivity and 78.9% specificity (Figure 2).

Figure 2.

ROC curve of paced QRS duration in predicting RVOT septal pacing. Abbreviations: ROC, receiver operating characteristic; RVOT, right ventricular outflow tract

3.4. QRS patterns

All patients had a left bundle branch block pattern during forced ventricular pacing. Septal pacing produced significantly more QS, Qr, and rS complex compared with nonseptal pacing (77% vs 34%; P < 0.001). Conversely, nonseptal pacing demonstrated significantly more R, qR, Rs, and rsR′ compared with septal pacing (66% vs 23%, P < 0.001; see Supporting Information, Figure S2, in the online version of this article). No significant differences in QRS pattern were seen in inferior leads, lead aVR, and lead aVL.

4. DISCUSSION

To best of our knowledge, this is the largest study reporting ECG patterns of validated RVOT pacing sites. The primary findings of this study are that (1) in our cohort of pacemaker recipients paced from the RVOT, only 62% have a septal RVOT lead location, as documented by TTE; (2) paced QRS duration and negative or isoelectric QRS vector in lead I are independent predictors of true RVOT septal pacing; and (3) paced QRS duration ≤145 ms indicates true RVOT septal pacing with 85.1% sensitivity and 78.9% specificity.

Long‐term RVA pacing is related with adverse effects on LV function.1, 4 Theoretically, RVOT septal pacing produces more physiological LV activation and less dyssynchrony. Spatially, the walls of RVOT include 4 quadrants: the septal, anterior, posterior, and free walls. The septum is sited relatively posterior with the free wall in front, and separating them is the anterior wall. The conventional radiological criteria were relatively insensitive to document RVOT septal lead implantation.9 The final lead position may be heterogeneous, with the RV lead being sometimes unintentionally positioned on the RVOT anterior or free wall when determined by TTE, instead of on the true RVOT septum. In this study, we found that true RVOT septal pacing was achieved in only 94 of 151 patients (62.3%). This might be one reason why the majority of studies did not find improved outcomes for septal pacing in comparison with RV apical pacing.3, 14, 16

Several studies have attempted to describe ECG criteria to identify septal pacing. The criterion most often established was the negative or isoelectric QRS vector in lead I.5, 17, 18 However, these criteria were based on fluoroscopic images that were not validated against other imaging techniques. Only a few studies with small samples described the ECG characteristics of septal lead placement and confirmed the correct location of the lead exactly using techniques other than fluoroscopy. Burri et al19 studied ECG characteristics of septal and anterior (free) wall pacing in 31 patients using electroanatomical mapping. They showed that the presence of Q waves in lead I was only 30% sensitive and 74% specific for septal pacing. Osmancik et al..20 analyzed 51 patients in septal pacing. The exact location of the RV lead was verified using computed tomography. In the study, the tip of the lead was anchored in the septum in only 21 patients. The true septal pacing was associated with shorter QRS duration. There was no difference in the presence of Q waves in lead I or notching of QRS complex in inferior leads. The TZ in precordial leads was earlier in the septal group. In the study by Margulescu et al,21 with lead position confirmed by TTE, the ECG criteria showed low accuracy in identifying RVOT septal locations. In our study, we found that paced QRS duration and negative or isoelectric QRS vector in lead I are independent predictors of true RVOT septal pacing. A wider paced QRS duration was associated with an increased incidence of nonseptal pacing. Conversely, a narrower paced QRS duration seemed to be more frequently observed in true septal pacing. Paced QRS duration ≤145 ms indicates true RVOT septal pacing with 85.1% sensitivity and 78.9% specificity. The possible explanation for this observation is the presence of simultaneous vs sequential wavefronts activating RV and LV when pacing from the septal as opposed to anterior or free wall. A shorter total ventricular activation time and greater ventricular synchrony might help decrease adverse remodeling.3, 7, 22, 23, 24 In the present study, negative or isoelectric vector in lead I, which is the most common characteristic attributed to septal pacing, was useful in discriminating RVOT septal pacing from nonseptal pacing. RVOT true septal pacing also displayed more QS and Qr complexes in lead I than did nonseptal pacing. Furthermore, precordial QRS transition was earlier in septal pacing and later in nonseptal pacing. Lead II/III R‐wave amplitude ratio < 1 and aVR/aVL QS‐wave amplitude ratio < 1 were more frequently observed in septal pacing compared with nonseptal pacing. This may be explained by the orientation of the heart that the RVOT septum is more leftward and produces large vectors in lead III and against lead aVL.

In summary, the main findings of this analysis suggest that paced QRS duration ≤145 ms and negative or isoelectric QRS vector in lead I serve as the predictors of RVOT septal pacing. Hence, routine assessment of these parameters to guide RV lead implantation is encouraged.

4.1. Study limitations

This study was designed to summarize ECG characterizations in a cohort of patients whose leads were inserted into the RVOT, so we did not perform a clinical follow‐up. The study comparing true RVOT septal pacing with RVA pacing will be performed in the next step.

5. CONCLUSION

This study reveals the heterogeneity of lead placement within the RVOT. Narrower paced QRS duration and negative or isoelectric QRS vector in lead I independently predict the position of RVOT septal pacing. In addition, a 145‐ms cutoff for paced QRS duration indicates RVOT septal pacing with 85.1% sensitivity and 78.9% specificity. These parameters may help the operator differentiate true RVOT septum from non‐RVOT septum.

Conflicts of interest

The authors declare no potential conflicts of interest.

Supporting information

Figure S1 Supporting Information

Figure S2 Supporting Information

Wei H, Tang J, Chen D, et al. Electrocardiographic predictors of validated right ventricular outflow tract septal pacing for correct localization of transthoracic echocardiography. Clin Cardiol. 2018;41:354–359. 10.1002/clc.22873