Right ventricular apical lead position is associated with prolonged signal-averaged P-wave duration

Introduction 

The right ventricular apex (RVA) has been the traditional site for pacemaker lead placement because of its accessibility, stability, and low failure rates.1 However, pacing at this site has recently been associated with adverse clinical outcomes; both long-term and short-term RVA pacing has been associated with left ventricular dysfunction,2, 3, 4 and RVA pacing has been associated with increased risk of atrial fibrillation (AF) and death in sick sinus syndrome (SSS).5, 6 The incidence of new-onset AF is higher in ventricular-paced patients, particularly with a diagnosis of SSS.7 Such findings are believed to be due in part to the ventricular dyssynchrony associated with apical pacing.8 Long-term asynchronous ventricular pacing has been shown to cause atrial electrical remodeling and increased atrial diameters resembling that associated with chronic AF.9, 10

Given the known impact of RVA pacing on the potential for atrial arrhythmia, one might expect to observe changes in atrial substrate, such as dimensions and electrophysiologic properties of conduction and refractory period. The effect of ventricular dyssynchrony may manifest as atrial arrhythmia after atrial remodeling has occurred. Noninvasive assessment of atrial substrate is possible using contemporary signal averaging technology.11 Parameters derived from the signal-averaged P wave (SAPW) have consistently demonstrated the ability to correlate with clinical and electrophysiologic variables.12, 13, 14, 15 The presence of prolonged P-wave duration (PWD) is well associated with paroxysmal atrial fibrillation (PAF); however, prolonged PWD is also associated with hypertension, acute pulmonary edema, and heart failure. It is speculated that parameters derived from the SAPW may also reflect conventional parameters of atrial electrophysiology, such as conduction velocity and refractory period.14 The SAPW affords a convenient and accessible method for assessment of atrial substrate.

To date, little is known about the incidence of new-onset AF in patients paced for atrioventricular (AV) block with apical or septal pacing. We hypothesized that any adverse effect on atrial substrate with increased ventricular dyssynchrony from RVA pacing would manifest as change in the SAPW duration and/or increase incidence of atrial high-rate events when compared with a matched cohort with right ventricular septal pacing.

Methods 

Candidate patients were identified from a retrospective chart and database review of patients under follow-up at a cardiac rhythm device clinic. Septal pacing is performed routinely by some physicians in an unselected patient population. We reviewed the charts from all patients, with dual-chamber devices and septal ventricular leads with more than 3 months of follow-up, expected to return to the clinic within the 3-month period allocated for this study. From this group, we identified patients with a cumulative percentage of ventricular pacing (CVP) of 70% or higher from stored pacemaker data that formed the primary study cohort; they were matched with a group of patients with RVA placement and with a CVP of 70% or higher. Previously obtained SAPW recordings taken from 49 healthy patients with no history of cardiac disease and 73 patients with documented PAF were used as negative and positive controls, respectively. All controls were screened with history and physical and had a normal-resting 12-lead electrocardiogram (ECG).

Patients were excluded if they had a single-chamber device, antiarrhythmic medication, a history of atrial arrhythmia or SSS, previous cardiac surgery, or were pregnant.

Patients in each cohort were contacted and invited to return for clinic follow-up or were approached before their normally scheduled checkup. During this appointment, device diagnostics were interrogated for incidence and duration of atrial high-rate episodes. Other device and lead parameters were recorded, such as lead pacing threshold and impedance. After interrogation, a high-resolution recording (1000 Hz sampling frequency) was collected during sinus rhythm and according to the following methodology.

Careful skin preparation and positioning of silver-silver chloride electrodes in an orthogonal manner (modified Frank lead position) preceded recording. The ECG leads were attached, and the subject was then asked to lie still for 10 minutes during data acquisition using a commercial Holter system (SpiderView, ELA Medical, Le Plessis-Robinson, France). Derived analogue signals were amplified 10 000 times and band pass filtered between 1 and 300 Hz. One lead exhibiting the most obvious P wave was then further band pass filtered between 20 and 50 Hz and was used as a trigger to align subsequent P waves for signal averaging. The analogue data were sampled at 1 kHz with 12-bit resolution. Approximately 600 beats were stored for subsequent offline analysis, according to previously validated methodology.16

The paced QRS duration was measured in each patient from the beginning of the ventricular pacing spike to the end of the QRS complex with calipers on a standard surface ECG recording at 25 mm/s. The maximum-paced QRS duration in any of the limb or precordial leads was used for measurement.

Results 

A total of 131 candidate patients with dual-chamber pacemakers were identified to return within the 3-month study window; 74 were apically paced and 57 were paced at the right ventricular septum. On chart review, 22 and 23 patients were paced in the ventricle more than 70% of the time for apical and septal patients, respectively.

The results from this cohort are shown in Table 1. There was no significant difference in burden of ventricular pacing, programed AV delay, or echocardiographic parameters. The groups were well matched in terms of mean follow-up and sex. There was no significant difference in the number of mode switches between the 2 groups.

Table 1. Characteristics of patients with CVP of 70% or higher on chart review

	Variables	Septal	RVA	P
	No. of patients	23	22
	Age	71.0 ± 10.7	73.5 ± 7.3	.36
	Sex (male/female)	14/9	13/9	NS
	Pacing duration	15.7 ± 6.9	16.5 ± 5.6	NS
	% A pacing	17.7 ± 19.4	26.5 ± 29.4	.25
	% V pacing	94.1 ± 9.3	95.1 ± 8.8	.71
	AV delay (S)	134.3 ± 30	156.4 ± 50	.084
	AV delay (P)	173.9 ± 45	186.8 ± 50	.37
	LA size	39.8 ± 6.1	39.5 ± 9.1	.92
	LV EF	58.4 ± 16.7	54.5 ± 13.8	.44
	Mode switches	0.5 [0.0-5.4]	0.40 [0.1-6.2]	.78

Septal refers to patients with septal positioning of the right ventricular lead; RVA refers to the cohort with apical placement of the lead. Age, age on inclusion (mean ± SD, years); pacing duration, months since implantation on inclusion (mean ± SD); % A pacing, percentage of apical beats that are paced (mean ± SD); % V pacing, percentage of ventricular beats that are paced (mean ± SD); AV delay (S), sensed AV delay (mean ± SD, milliseconds); AV delay (P), paced AV delay (mean ± SD, milliseconds); LA size, left atrial size by echocardiography (mean ± SD, mm); LV EF, left ventricular ejection fraction (mean ± SD, %); mode switches, mode switches per month recorded (median [IQR]).

Of the 45 patients selected for examination, high-resolution recordings were obtained in 21 patients; 2 patients from the apical group were paced in the atrium at the time of the recording, leaving 19 for analysis (10 septal and 9 apical; see Table 2 for demographics). Reasons recording were not made varied and were not prospectively collected. The predominant reasons were missed appointments and late appointments, resulting in reluctance to stay for the high-resolution recording.

Table 2. Characteristics of patients who underwent SAPW analysis

	Variables	Septal	RVA	P
	No. patients	10	9
	Age (y)	71.9 ± 12.1	71.9 ± 5.7	.99
	Sex (male/female)	5/5	5/4	NS
	Pacing duration	10.5 ± 3.2	11.4 ± 6.4	NS
	% A pacing	11.7 ± 15.5	27 ± 28.2	.20
	AV delay (S)	130.5 ± 15.0	165.6 ± 50.5	.18
	LA size	37.5 ± 7.35	39.0 ± 7.96	.71
	QRS duration	160 ± 14	191 ± 23	.004
	CHB	7 (70%)	5 (56%)
	2:1 AVB	1 (10%)	2 (22%)
	Second-degree AVB (II)	2 (20%)	2 (22%)
	Mode switches	0.49[0-3.0]	0.34[0.2-1.2]	.36
	Threshold	0.56 ± 0.11	0.69 ± 0.27	.19
	Impedance	598 ± 80	642 ± 102	.31

Septal refers to patients with septal positioning of the right ventricular lead; RVA refers to the cohort with apical placement of the lead. Age, age on inclusion (mean ± SD, years); pacing duration, months since implantation on inclusion (mean ± SD); % A pacing, percentage of apical beats that are paced (mean ± SD); AV delay (S), sensed AV delay (mean ± SD, milliseconds); LA size, left atrial size by echocardiography (mean ± SD; mm); QRS duration, paced QRS duration (mean ± SD, milliseconds); mode switches, mode switches per month recorded (median [IQR]); threshold, latest threshold voltage recorded in ventricular lead (mean ± SD, V in 0.4 milliseconds); impedance, latest ventricular lead resistance recorded (mean ± SD, Ω).

Of these 19 patients with high-resolution recordings suitable for analysis, the clinical characteristics of both study cohorts were similar, whereas the RVA group had a significantly longer-paced QRS duration (septal, 160 ± 14 milliseconds; RVA, 191 ± 23 milliseconds; P = .004; Table 2). There were no significant differences between the septal and RVA groups with regard to number of mode switches per month recorded and the latest ventricular lead threshold and impedance (P = .356, .193, and .314, respectively; Table 2).

The SAPW duration differed significantly between septal and RVA cohorts (septal, 133.0 ± 5.5 milliseconds; RVA, 144.8 ± 6.9 milliseconds; P = .001), whereas there was no significant difference between septal and normal cohorts (septal, 133.0 ± 5.5 milliseconds; normal, 129.3 ± 7.1 milliseconds; P = .08; Table 3). There was a significant difference between the PAF and RVA cohorts (RVA, 144.8 ± 6.9 milliseconds; PAF, 152.3 ± 16.3 milliseconds; P = .02).

Table 3. Signal-averaged PWDs

	Events	n	PWD
	Controls	49	129.3 ± 7.1
	Septal	10	133.0 ± 5.5a
	RVA	9	144.8 ± 6.9b
	PAF	73	152.3 ± 16.3

PWD indicates signal-averaged PWD (mean ± SD, milliseconds); controls, PAF, septal, and RVA as defined in text.

a Refers to a nonsignificant difference when compared with controls, 3.71 milliseconds (confidence interval [CI], −0.56 to 7.99; P = .08). b Denotes a difference of 11.78 milliseconds (CI, 5.65 to 17.91; P = .001) when compared with the septal cohort, and a difference of −7.48 (CI, −13.68 to −1.28; P = .02) when compared with the PAF cohort.

Discussion 

These pilot data have shown that when matched cohorts of patients are examined according to the location of the right ventricular lead, a significant difference in SAPW duration can be appreciated, implying a difference in atrial substrate. Right ventricular apical pacing was associated with a significantly longer PWD relative to septal pacing and normal controls (P < .001); PWD in septal pacing was not different than that in normal controls.

Alternatives to RVA pacing are left ventricular pacing and right ventricular septal pacing; the latter is the preferred alternative because of ease of access and long-term lead performance. It is suggested that pacing at the septum leads to a more synchronous contraction of the ventricles, reducing delays to the free wall of the left ventricle compared with apical pacing. This may reduce wall stress and left ventricular remodeling compared with RVA pacing. Shortening of the QRS width has been associated with better LV function.16 It has also been suggested that septal pacing offers improved hemodynamic function relative to RVA pacing.1 Studies, to date, have presented inconsistent data on the relative merits of right ventricular septal pacing when compared with apical pacing.17, 18, 19 However, a recent retrospective analysis has demonstrated survival benefit at 5 years with septal pacing.20

Our data show a difference in SAPW duration in cohorts retrospectively matched other than right ventricular lead location. We did not record baseline SAPW duration before pacing and, thus, could not demonstrate a change in SAPW duration in the RVA group from implant. However, we speculate that the groups were the same before implant and that apical pacing adversely affected atrial substrate in some way. The alternative is that both groups began with abnormal PWD, and the septal cohort normalized over time; this would seem less likely. If we assume the former, RVA pacing may be increasing PWD relative to septal pacing by its inherent ventricular dyssynchrony. This might be because of the abnormal mitral valve leaflet apposition leading to regurgitation and, ultimately, left atrial enlargement.21 Septal pacing minimizes ventricular dyssynchrony, possibly reducing or eliminating the regurgitation. Our data showed that septal pacing was associated with a reduced QRS duration on 12-lead ECG as compared with apical pacing (160 ± 14 and 191 ± 23 milliseconds, respectively; P = .004). Similar pathology was suggested to occur where the progression of tricuspid valve regurgitation was observed in apically paced patients but not in those paced septally over a 7-year period.22 Future studies including echocardiographic monitoring may clarify the mechanism of the observed electrocardiographic changes.

Longer PWD both from surface measurement and signal averaging has been observed in conditions where atrial abnormalities are present. Furthermore, such electrocardiographic abnormalities have been strongly associated with atrial arrhythmia, in particular, AF. The abnormal conduction time observed from a long PWD reflects abnormal atrial conduction and biatrial conduction time.11, 12 This is often associated with increased left atrial size on echocardiography. The long PWD in our RVA cohort may imply an increased risk of atrial arrhythmia compared with the relatively preserved P wave in the septal group. If this proves to be true, this would represent a novel and hitherto unforeseen benefit of septal pacing over apical placement in heart block.

The number of mode switches seen in each group, a surrogate marker for cumulative incidents of atrial tachycardia, was not significantly different; however, the cohort is small, and follow-up is limited to a mean of 10 months. The incidence of atrial tachycardias may not be expected to be observed in AV block until the duration of pacing is longer. In the paper by Cheung et al,7 the mean follow-up was 596 days; future studies including larger cohorts and increased observation time may have the necessary power to observe any difference. There was no observed difference in ventricular lead threshold or impedance, which suggests that septal lead placement is not inferior to apical placement in these regards. No difference between RVA and septal pacing in terms of dislodgement rates, lead positioning time, and thresholds during implantation or follow-up was seen in previous studies.23, 24

This pilot study suggests that a larger and more prolonged study to confirm and expand on the differences seen between septal and apical pacing is warranted. Signal averaging before pacing and during follow-up to prospectively examine for change would improve our understanding; moreover, echocardiography during follow-up would help to link any electrocardiographic and hypothesized physical changes.

Conclusions 

This small study matched patients for cumulative pacing and diagnosis that differed only in the location of the right ventricular pacing lead. A significant difference in SAPW duration was observed with prolonged PWD in the apical group and normal values in the septal group. This finding warrants further investigation and may allude to a novel benefit of septal pacing over RV apical placement.