Pacemaker Induced Cardiomyopathy: An Overview of Current Literature

Abstract

Pacemaker Induced Cardiomyopathy (PICM) is commonly defined as a reduction in left ventricular (LV) function in the setting of right ventricular (RV) pacing. This condition may be associated with the onset of clinical heart failure in those affected. Recent studies have focused on potential methods of identifying patients at risk of this condition, in addition to hypothesizing the most efficacious ways to manage these patients. Newer pacing options, such as His bundle pacing, may avoid the onset of PICM entirely.

1. INTRODUCTION

Over 1 million permanent pacemakers (PPM) are implanted annually worldwide [1]. Half of these devices are inserted in patients with high degree atrioventricular block who are dependent on a high percentage of right ventricular pacing. Most of these patients tolerate this without issue, however the condition of PICM is becoming increasingly recognised. This condition leads to impaired LV function and precipitates symptoms of heart failure (HF) [2, 3].

1.1. Definition

The definition of PICM varies significantly across studies due to the lack of a standardised definition. In general, this pathology is defined by a percentage decrease of LV ejection fraction (LVEF) from either the lower limit of normal or the pre-implantation LVEF for the patient. A number of studies have defined PICM as a post PPM LVEF of <50%, usually coupled with a requirement for an absolute drop of 5-10% LVEF from baseline [4-6].

Of note, however, when an absolute drop of LVEF >10% has been used without the requirement for the resultant LVEF to be < 50 %, a significantly higher percentage of patients may meet the definition for PICM [7]. This raises the question of whether LVEF is the best measure for assessing the detrimental effects of PICM. It may be that a number of patients who experience the negative effects of RV pacing are not recognised as having PICM as their LVEF remains >50%.

Recent studies have also included the use of global longitudinal strain (GLS) changes from baseline. In these studies, a reduction of 15% from baseline is considered to meet the definition of PICM [8]. Other measures of systolic and diastolic function, such as indices of isovolumic contraction and relaxation times, have been used also [9, 10].

To meet a diagnosis of PICM, patients are usually required to have a normal pre-PPM LVEF to exclude pre-existing LV failure as a cause of LVEF reduction post-implantation.

Previous research has suggested a high burden of RV pacing to accelerate a decline in cardiac function in patients with pre-existing LV failure [11].

HF symptoms and hospitalisation are considered a part of the definition of PICM in some studies [12]. Particularly due to concerns that patients with HF from RV pacing may have a relatively preserved LVEF. It has been suggested that the incidence of PICM is significantly underestimated when it is defined by a reduction in LVEF alone. This appears to be a late manifestation of this condition and patients potentially develop symptoms of PICM before LV function has dropped to an extent that can be reliably detected. For example, in the PACE study, which was undertaken to assess the superiority of cardiac resynchronization therapy (CRT) over RV pacing, all patients had an LVEF>45% at baseline [13]. In this study, the authors noted a significant difference in LVEF between these two groups after 1 year, despite the mean LVEF remaining over 50%.

Those who received RV pacing had significantly higher rates of hospitalization for HF, compared to those who received biventricular pacing, despite the mean LVEF remaining greater than 50% over a mean follow-up of 4.3 years [13].

The development of Atrial Fibrillation (AF) has been considered a potential manifestation of PICM [14]. A randomised trial of atrial vs. dual chamber pacing for sinus node dysfunction noted that patients with a high pacing burden were at a significantly increased risk of developing AF at follow-up, a mean of 2.9 years [15]. Of note, pacing burden also correlated with increased left atrial and ventricular volumes which may explain the association with the development of AF in these patients.

2. PATHOPHYSIOLOGY

The pathophysiology underlying the development of PICM is presumed to be related to the intraventricular dyssynchrony that occurs as a result of RV pacing. RV pacing leads to an altered ventricular activation pattern that is likely observed with the left bundle branch block. The electrical impulses traverse from cell to cell rather than through the Purkinje system, resulting in a prolonged QRS duration on the ECG. This disruption to the normal electrical and subsequent mechanical activation of the ventricles results in delayed activation of the basal and lateral left ventricle. This leads to inefficient contractile function, elevated filling pressures and ultimately maladaptive cardiac remodeling with the development of clinical heart failure [10, 16, 17].

The evidence suggests that RV pacing may result in alterations to the genetic expression of components of the sarcoplasmic reticulum which precede a decline in LV function [18]. Endomyocardial biopsies from patients subjected to long-term ventricular pacing have revealed alterations at the cellular and subcellular levels, including differences in fibrosis, fat deposition and mitochondrial morphological changes [19].

Studies in this area have reported that PICM is less likely to develop in patients that are RV paced from the septum or outflow tract positions when compared to patients paced from the RV apex [20, 21]. This finding, however, has not been consistent across all studies [22-25]. One potential explanation for this inconsistency is that the RVOT pacing site required to obtain a septal position of pacing is not standardized. There are inherent difficulties in standardization due to inter-individual variation in RVOT anatomy [26].

In the PROTECT-PACE study, 240 patients with high- grade atrio-ventricular block were randomised to receive PPM implantation with either RV apex pacing or high septal region pacing [27]. At 2 years, the LVEF was significantly reduced in both groups but there was no significant effect from lead placement on LVEF, HF hospitalisation or mortality.

The pathophysiology of PICM draws similarities to other cardiomyopathies that are associated with impaired electrical conduction, such as left bundle branch block cardiomyopathy and premature ventricle conduction induced cardiomyopathy. In all three of these conditions, the higher the frequency of dyssynchrony, the more likely patients are to develop cardiomyopathy. For PICM, this refers to the overall RV pacing burden. In these patients, pacing burden >20% is associated with a considerably higher incidence of PICM [4]. Likewise, there is evidence to suggest that if the frequency of dyssynchrony is reduced, then the cardiomyopathy may be reversible [28, 29]. This could be accomplished with left ventricular pacing in CRT systems.

3. INCIDENCE

The incidence of PICM varies greatly depending on the definition used, but overall it appears to occur in 10-20% of individuals within 3-4 years post PPM insertion [4-6]. One recent study demonstrated that depending on the definition used, the incidence of PICM in a single cohort varied from 5.9% up to 39% [7].

Estimating LVEF in the setting of high rates of RV pacing with ventricular dyssynchrony may be challenging and impact the ability to accurately detect patients with PICM. It is suggested that PICM is evident in affected patients within 1 year from implantation. The PACE study, for example, demonstrated a significant decrease in LVEF from 61.5% to 54.8% at 12 months post-implant [13].

There is evidence that the decrease in LV function with RV pacing may occur within hours to days of device implantation. One study demonstrated a significant decrease in LVEF within 2 hours, whilst another showed a significant reduction within the first 6 months [30, 31].

It remains unclear whether this early drop in LVEF, where LVEF remains >50%, is predictive of future clinical HF, or if patients may develop symptoms from ventricular dyssynchrony even at modest levels of reduced LV function. These relatively slight drops in LVEF in the early stages post-implantation may detect patients at risk for the development of significant PICM.

The ability to detect PICM at its earlier stages would, in theory, enable clinicians to take action to prevent the development of overt HF in these patients.

4. RISK FACTORS

Identifying risk factors for the development of PICM allows clinicians to monitor patients at higher risk more closely, such that action can be taken early in these patients to prevent the development of overt HF.

Baseline clinical variables that have been associated with an increased risk of PICM include older age, wider intrinsic QRS (pre-implantation) on ECG, wider paced QRS on ECG, male gender, history of AF, and impaired LVEF pre-PPM implantation [4-6, 31-35].

With regards to post PPM risk factors for PICM, the MOST trial identified that patients who had >40% RV pacing burden had a 2.5 fold higher risk of HF hospitalisation compared to those with a lower RV pacing burden [14]. A similar finding was also shown in the results of the DAVID trial where the incidence of death or HF hospitalisation was three times higher for pacing burden, >40%, compared to less than 40% [11].

Despite evidence to suggest that patients with a high burden of RV pacing are more likely to develop PICM, there remains a number of studies demonstrating a minimal reduction in LVEF despite a high burden of pacing [36-38]. In one of these studies, the burden of pacing was not predictive of reduced LVEF post device implantation [39].

Studies indicate that brain natriuretic peptide (BNP) levels highly correlate with the frequency of RV pacing in patients with dual-chamber pacemakers [9, 40]. Although not clear from current studies, it may be that BNP is an early marker predicting functional change in patients with a high frequency of RV pacing.

It has been suggested that an early decline in GLS post-PPM implantation may be predictive of a future decrease in LVEF [8]. In the study by Babu et al., a reduction of GLS by 15% from baseline was noted in 63.9% of patients with RV pacing at 12 months post-implantation.

A study by Khurshid and colleagues analysed data from 257 patients who underwent single or dual chamber PPM implantation with a pre-implantation LVEF of > 50 % and a pacing burden of > 20 % [4]. At approximately 1 year, 20% of patients were found to have an LVEF that had decreased >10% and was subsequently <50%. Male gender and a native QRS on the ECG of > 150 milliseconds were strongly correlated with the development of PICM. Interestingly, only half of the patients who met the definition for PICM had developed symptoms of HF.

5. TREATMENT

The most common approach to managing patients with PICM is to upgrade the single or dual lead pacemaker to a CRT device.

A number of studies have demonstrated that CRT device placement may avoid the occurrence of PICM [41, 42]. In the PACE study, the LVEF of patients with RV pacing was significantly reduced at 12 months, where the LVEF of those patients who received CRT was unchanged from baseline [13]. Patients who received CRT also had significantly lower incidences of HF compared to those with RV pacing.

CRT upgrade has been shown to be a highly effective treatment for this condition. One recent retrospective review demonstrated that 72.2% of patients with severe PICM (LVEF < 35 %) who were upgraded to CRT were found to have an LVEF that improved to > 35 % over a median of 7 months [29]. Much of this improvement in LVEF was seen at the first 3 months post-device upgrade.

Another study by Kiehl and colleagues demonstrated an 84% responder rate for patients with PICM who received an upgrade to a CRT device [5]. In those who responded to the CRT upgrade, the mean increase in LVEF was 18.5%.

Current guidelines do not address the role of CRT upgrade for those patients whose LVEF is >35%, but who have symptoms of HF consistent with PICM. There are, however, isolated reports of beneficial effects of CRT in these patients [43, 44].

There has been concern that the upgrade procedures ultimately needed for patients with PICM lead to a higher morbidity and mortality than the initial implantation procedure. A trend towards this has been noted in registry studies [45, 46]. Cheung and colleagues found, through analysis of registry data, that in-hospital mortality for patients undergoing a CRT upgrade was 1.8%, compared with 0.8% for de novo CRT implantation [45]. This difference was statistically significant (P < 0.001). This raises the implication that identifying patients at risk of PICM and choosing to implant a CRT as the initial device may be preferable to waiting for PICM to develop before offering a device upgrade.

6. PREVENTION

In order to reduce the burden of RV pacing, a strategy that has been trialed is the use of device-based algorithms designed to minimise ventricular pacing. An algorithm adopted by several manufacturers is an atrial-based dual-chamber mode that can provide atrial pacing alone unless a period of atrio venticular block (AVB) is detected, at which point the ventricle will be paced [47]. A meta-analysis was recently published by Shurrab and colleagues which consisted of 7 trials that enrolled a total of 4119 patients and compared the use of ventricular pacing reduction modalities with standard dual chamber pacing [38]. At the mean follow-up of 2.9 years, the RV pacing burden was significantly lower in the re-programmed group of patients, but there was no significant difference observed in the rates of hospitalization or death. These findings suggest that the rate of significant PICM is unlikely to be reduced overall.

One obvious strategy that may be used to prevent the development of PICM is to implant a CRT device from the outset. Kurshid and colleagues previously recommended that patients at higher risk of developing PICM should have a CRT device implanted [4]. This opinion remains controversial, however, and studies that have assessed this approach have provided results that are not convincingly in favor of CRT use as a routine device of choice [41, 48].

The BIOPACE study, which published preliminary results in 2014, is the largest study to date that has assessed the routine use of CRT compared to RV pacing, in patients with PR prolongation on ECG or complete heart block [48]. Among 1810 patients, over a mean of 5.6 years, there was no significant difference observed between CRT and RV pacing for primary endpoints of hospitalisation due to HF or death.

The BLOCK-HF study, which randomised 691 patients with an LVEF <50% to RVP or CRT, did, however, show an improvement in all-cause mortality, HF admissions and an increase in LV systolic function [41]. Of note, the outcomes in BLOCK-HF were all driven primarily by an improvement in LVEF. The authors noted that the overall risk of morbidity and mortality increased by approximately 1% for every 1 mL/m2 increase in LV end-systolic volume index.

A possible explanation for the differing results of the BLOCK-HF and BIOPACE studies is the different baseline characteristics of the participants of these studies. The prevalence of LBBB and AF in BLOCK -HF was nearly double to that in the BIOPACE study, suggesting that these patients may receive more benefit from CRT.

There remains concern regarding the complexity of CRT device insertion, with its higher procedural morbidity and potentially shorter device longevity. A large meta-analysis of CRT trials, which included 9,000 patients, demonstrated a 5% rate of failure to implant the LV lead, a 3.2% risk of mechanical complication and a 1.4% risk of infection [49]. Duray and colleagues compared the risk of complications requiring surgical revision in those with CRT devices and those with single or dual chamber pacemakers [50]. The authors found that patients with CRT devices were nearly three times as likely to require surgical intervention compared to dual chamber device patients (11.8% vs. 4.1%) over a median follow-up of 31 months.

Conduction system pacing, which includes His bundle pacing (HBP), has shown promise for the reversal of PICM. Permanent HBP engages the His-purkinje system and restores the native electromechanical sequence of cardiac tissue. HBP may be either selective or non-selective. Patients with selective His bundle capture will have a His-paced to QRS interval similar to their native His to QRS interval. Non-selective capture may occur when the lead is not fixed directly to the His Bundle. Non-selective HBP results in right ventricular septal pre-excitation, and as a result, these patients will have a wider QRS duration and His capture may only occur at higher pacing outputs [51].

The success rates for achieving selective HBP are limited by the technical difficulty of the procedure. Beer and colleagues found that in 350 consecutive patients who underwent HBP, selective pacing was achieved in only 33% [52]. Fortunately, initial studies have suggested that non-selective HBP may not be inferior to selective HBP in terms of patient outcomes, but more clinical studies are required to confirm this [52].

A recently published study showed a significant increase in LV function in patients with PICM who were successfully upgraded from RV pacing to HBP [53].

An observational study has demonstrated that routine use of HBP was associated with improvements in the endpoint of death, HF, or the need for CRT upgrade [54]. Challenges associated with the device implantation and concerns regarding sensing problems and increased pacing thresholds are some limitations.

Left bundle pacing (LBP), as may be achieved through a transventricular septal approach, has been suggested as an alternative that may avoid some of these challenges. This novel method of ventricular pacing may avoid the previously documented 30-40% non-response rate to CRT [55]. Compared to HBP, LBP pacing may offer lower pacing thresholds, large R waves on the ECG and a lower risk for the development of distal conduction block [56]. There is evidence showing improvement in LVEF in patients with PICM who are upgraded to LBP [57]. The success rate and safety of LBP in these early studies are promising, with one study reporting a success rate of over 90% [58]. It is important to note that, at present, patients with indications for an implantable cardioverter-defibrillator still require an RV defibrillator electrode to be placed.

CONCLUSION

Although several potential risk factors have been identified for PICM, the occurrence of this condition in the clinical setting remains difficult to predict. Current evidence suggests that risk factors, such as the high burden of RV pacing, do not universally lead to the development of PICM. This implies that there is a complex interplay of multiple biological and myocardial disease-associated factors not yet entirely appreciated. Newer methods of pacing, such as HBP, may avoid the development of PICM entirely. However, these methods of pacing are often associated with increased complexity and cost. Further studies may help elucidate the most effective way to identify and manage patients at risk of PICM.

CONSENT FOR PUBLICATION

Not applicable.

FUNDING

None.

CONFLICT OF INTEREST

The authors declare no conflict of interest, financial or otherwise.