Author + information
- Received July 24, 2017
- Revision received October 2, 2017
- Accepted October 4, 2017
- Published online November 15, 2017.
- Shaan Khurshid, MDa,
- Edmond Obeng-Gyimah, MDb,
- Gregory E. Supple, MDb,
- Robert Schaller, DOb,
- David Lin, MDb,
- Anjali T. Owens, MDc,
- Andrew E. Epstein, MDb,
- Sanjay Dixit, MDb,
- Francis E. Marchlinski, MDb and
- David S. Frankel, MDb,∗ ()
- aCardiology Division, Massachusetts General Hospital, Boston, Massachusetts
- bElectrophysiology Section, Cardiovascular Division, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania
- cHeart Failure Section, Cardiovascular Division, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania
- ↵∗Address for correspondence:
Dr. David S. Frankel, Cardiovascular Division, Electrophysiology Section, Perelman School of Medicine at the University of Pennsylvania, 9 Founders Pavilion, 3400 Spruce Street, Philadelphia, Pennsylvania 19104.
Objectives This study sought to determine the extent, time course, and predictors of improvement following cardiac resynchronization therapy (CRT) upgrade among pacing-induced cardiomyopathy (PICM) patients.
Background PICM is an important cause of heart failure in patients exposed to frequent right ventricular (RV) pacing. CRT may reverse PICM.
Methods We retrospectively studied 1,279 consecutive patients undergoing CRT procedures between 2003 and 2016. Patients undergoing CRT upgrade from a dual-chamber or single-chamber ventricular pacemaker for PICM were included. PICM was defined as decrease of ≥10% in left ventricular ejection fraction (LVEF), resulting in LVEF <50% among patients experiencing ≥20% RV pacing without an alternative cause of cardiomyopathy. Severe PICM was defined as pre-upgrade LVEF ≤35%. Clinical, electrocardiographic, and echocardiographic characteristics associated with both the extent of LVEF recovery and with post-upgrade LVEF of >35% among those with severe PICM were identified.
Results Of 69 PICM patients, LVEF improved from 29.3% to 45.3% over a median 7.0 months. Of 54 patients with severe PICM, 39 (72.2%) improved to LVEF >35% over a median 7.0 months. Most improvement occurred within the first 3 months, although improvement continued over the remainder of the first year. In linear regression, narrower native QRS was associated with greater LVEF improvement following CRT upgrade (+2.00% per 10-ms decrease; p = 0.05).
Conclusions CRT is highly efficacious in reversing PICM, with 72% of severe PICM patients achieving LVEF >35% and most of the improvement occurring within 1 year. These data support initial upgrade to a CRT pacemaker with consideration of further upgrade to CRT defibrillator after 1 year if LVEF remains ≤35%.
- biventricular pacing
- cardiac resynchronization therapy
- heart failure
- pacing-induced cardiomyopathy
Right ventricular (RV) pacing can lead to progressive left ventricular (LV) systolic dysfunction through chronic exposure to electrical and mechanical dyssynchrony (1–3). This RV pacing-induced cardiomyopathy (PICM) is increasingly recognized as a common and important cause of heart failure in patients with atrioventricular block (4–6).
Cardiac resynchronization therapy (CRT), or biventricular pacing, has been shown to reverse PICM in small cohorts (7–9). However, the extent and time course of improvement, as well as clinical factors predicting the degree of response, remain poorly characterized. Among patients with PICM resulting in left ventricular ejection fraction (LVEF) ≤35%, predictors of improvement to LVEF >35% following CRT upgrade are of particular interest to guide the decision of whether to upgrade to a biventricular pacemaker or implantable cardioverter-defibrillator (ICD) (10). Patients who are expected to improve to LVEF >35% could be treated with a biventricular pacemaker alone.
In this study, we drew upon a large single-center experience of consecutive CRT upgrades for patients with PICM over more than a decade and sought to characterize the time course and degree of improvement in LVEF. Furthermore, we analyzed clinical factors at the time of upgrade associated with extent of LVEF response over long-term follow-up.
We retrospectively studied consecutive patients with PICM undergoing CRT upgrade at the Hospital of the University of Pennsylvania between January 2003 and November 2016. Patients were considered to have PICM if they had an echocardiogram documenting LVEF ≥50% either prior to or shortly after pacemaker implantation and subsequent exposure to frequent (≥20%) RV pacing with ≥10% decrease in LVEF resulting in LVEF of <50% (6). Patients were considered to have severe PICM if the nadir LVEF was ≤35%. Patients with alternative causes of LV dysfunction, including myocardial infarction, significant myocardial ischemia on stress testing, severe valvular heart disease, atrial arrhythmias with rapid ventricular response, frequent (>15%) ventricular premature depolarizations, and uncontrolled hypertension (>160/100 mm Hg), were not considered to have PICM and therefore were excluded. The onset of PICM was considered the date of the first echocardiogram documenting decreased systolic function.
Patients with PICM were included in the present study if they underwent upgrade of a single chamber ventricular or dual chamber pacemaker to biventricular pacemaker or ICD and repeat echocardiogram was performed following CRT upgrade. Patients undergoing pulse generator change, lead revision, initial CRT implantation, or unsuccessful LV lead insertion were excluded. All patients provided informed consent for the implantation procedure and for their anonymized medical information to be used for research.
Pre-upgrade clinical, echocardiographic, and electrophysiologic measurements
Demographics, medical comorbidities, and medication use were extracted from the electronic medical record. Clinical heart failure was defined by manual review of cardiology notes, assessing for the presence of at least 2 of the following criteria adapted from the Framingham Heart Study (11): unexplained weight gain, dyspnea on exertion, paroxysmal nocturnal dyspnea, elevated jugular venous pressure, auscultatory crackles, S3 gallop, ascites or lower extremity edema, radiographic pulmonary edema or pleural effusion, and need for initiation or up-titration of diuretic therapy. The LV end diastolic diameter, end systolic diameter, and EF were measured by Hospital of the University of Pennsylvania echocardiograms, using standard techniques. In the 5 patients (7.2%) who did not have echocardiograms preformed at the Hospital of the University of Pennsylvania, these measurements were abstracted from their outside reports. Pre-upgrade native (unpaced) QRS duration, as well as presence of left or right bundle branch block were measured from the electrocardiogram (ECG) most closely preceding CRT upgrade demonstrating conducted QRS complexes. Patients with left or right bundle branch block were excluded from primary analyses involving native QRS duration. In secondary analyses, all native QRS complexes were included. Left and right bundle branch blocks were defined according to standard criteria established by the American Heart Association, American College of Cardiology, and Heart Rhythm Society (12). Pre-upgrade paced QRS duration was measured from the ECG immediately preceding CRT upgrade. Pre-upgrade ventricular pacing percentage was obtained from the pacemaker interrogation just prior to CRT upgrade. All pacemakers were programmed to favor intrinsic atrioventricular conduction whenever possible.
Post-upgrade echocardiographic and electrophysiologic measurements
Left ventricular end diastolic diameter, end systolic diameter, and EF were measured during all echocardiograms following CRT upgrade, performed in the context of routine clinical care. CRT response was prospectively defined as LVEF improvement of ≥5%. Post-upgrade heart rate, biventricular paced QRS duration, and biventricular paced QRS morphology were determined from the ECG immediately following CRT upgrade. The percentage of biventricular pacing was obtained from interrogation performed 4 to 6 weeks following upgrade. The incidence of sustained ventricular arrhythmias (defined as lasting >30 s or requiring ICD therapy) was determined by review of all device interrogations following CRT upgrade, performed at least once every 6 months. All events recorded as ventricular arrhythmias were adjudicated by Dr. Frankel.
Continuous variables are expressed as mean ± SD, and categorical variables are expressed as percentages. The Student’s paired t-test was used to compare mean pre-upgrade LVEF with mean peak LVEF following CRT. To determine independent predictors of LVEF improvement following CRT upgrade, multivariate linear regression was performed. Variables subjected to univariate screening included sex, age, body mass index, coronary artery disease, atrial fibrillation or flutter, hypertension, diabetes, clinical heart failure, heart failure admissions, medication use, pre-upgrade ECG and echocardiographic variables, pre-upgrade ventricular pacing percentage, time from PICM diagnosis to CRT upgrade, upgrade to CRT defibrillator as opposed to CRT pacemaker, post-upgrade heart rate, biventricular paced QRS duration, biventricular pacing percentage, presence of initial R-wave in lead V1, and presence of initial Q-wave in lead I. Variables showing significant (p < 0.10) associations with LVEF improvement on univariate testing were assessed in a multivariate model.
To assess factors associated with improvement in LVEF >35% post-upgrade, this analysis was repeated among the subset of patients with severe PICM using a binary logistic regression model with post-upgrade LVEF >35% as the outcome of interest. Analyses were performed using SPSS software (version 22, SPSS, Inc., Chicago, Illinois) and R version 3.2.2 software (R Project, Vienna, Austria) (13). We considered p values ≤0.05 to indicate statistical significance.
Response to CRT UPGRADE
Of the 1,279 patients undergoing CRT-related procedures during the study period, 472 underwent CRT upgrade of an existing pacemaker. Of those 472 patients, 376 had an alternative potential cause of cardiomyopathy, resulting in 69 patients who underwent CRT upgrade for suspected PICM, the current study cohort (Figure 1). Baseline characteristics of these patients are displayed in Table 1. Among the 69 PICM patients who underwent CRT upgrade, a median of 2 echocardiograms (interquartile range: 1 to 3) were performed over median 7.0 months of follow-up, with LVEF increasing from 29.3% to 45.2% (p < 0.01) (Central Illustration). Fifty-nine patients (85.5%) had LVEF improvement of ≥5% post-upgrade, and 49 patients (71.0%) had improvement of ≥10%. Among the 54 patients with severe PICM, 39 (72.2%) achieved LVEF >35% post upgrade, with a median time to LVEF >35% of 7.0 months (interquartile range: 2.7 to 13.2 months). Among those 39 severe PICM patients whose LVEF eventually improved to >35%, 24 (61.5%) had undergone initial upgrade to CRT defibrillator. Most LVEF improvement occurred within the first 3 months, with mean LVEF improving from 29.3% to 39.1% in the overall PICM group and from 26.4% to 36.8% in the severe PICM group (p < 0.01 for both comparisons), although more modest improvement continued through the first year (Central Illustration). Individual LVEF responses to CRT upgrade are listed in the Online Table and illustrated in Figure 2. Importantly, the rate of heart failure medication use was similar before CRT upgrade and at the time of last follow-up (ACE inhibitor/angiotensin receptor blocker 71.0% before CRT vs. 72.5% at last follow-up, beta blocker 68.1% vs. 81.2%, aldosterone antagonist 14.5% vs. 17.4%, p = NS for all comparisons).
Among the entire study cohort, 4 patients (5.8%) had sustained, monomorphic ventricular tachycardia during follow-up. Three of those patients had severe PICM at baseline and responded to CRT with normal LVEF at the time of arrhythmia. The fourth patient did not have severe PICM at baseline, and that patient’s LVEF did not improve following CRT upgrade (37.5% at the time of arrhythmia). Ventricular tachycardia rates were 162 to 202 beats/min, and all occurred >1 year following CRT upgrade. Two of those episodes (2.9%) terminated with antitachycardia pacing, and 2 (2.9%) terminated spontaneously. Ventricular fibrillation did not occur.
Predictors of LVEF Improvement
In univariable analysis of the entire cohort with PICM, age, beta blocker use, narrower native QRS, lower pre-upgrade LVEF, upgrade to CRT-defibrillator (as opposed to CRT-pacemaker), and increased time with PICM were associated with larger improvement in LVEF post-upgrade (Table 2). In multivariate analysis, narrower native QRS (+2.00% per 10-ms decrease; 95% confidence interval [CI]: 0.01 to 3.95; p = 0.05; bundle branch blocks excluded) remained associated with greater improvement in LVEF post upgrade. In a secondary multivariate analysis including all native QRS complexes, narrower native QRS was no longer significantly associated with LVEF improvement (+0.53% per 10-ms decrease; 95% CI: −0.62 to 1.68; p = 0.40).
Predictors of LVEF Improvement TO >35% Among Those with Severe PICM
In univariable analysis among the 54 patients with severe PICM, age, BMI, loop diuretic use, upgrade to CRT-defibrillator (as opposed to CRT-pacemaker), and narrower native QRS were associated with improvement to LVEF >35% post-upgrade (Table 3). In multivariable analysis, narrower native QRS (odds ratio [OR] 1.92 per 10-ms decrease, 95% CI: 1.09 to 5.00, p = 0.06, bundle branch blocks excluded) remained marginally associated with improvement to LVEF >35% post-upgrade. In a secondary multivariate analysis including all native QRS complexes, narrower native QRS was not significantly associated with LVEF >35% post upgrade (OR: 1.21 per 10-ms decrease; 95% CI: 0.94 to 1.61; p = 0.10).
In a sizeable cohort of PICM patients identified through systematic inclusion and careful screening for alternative causes of cardiomyopathy, CRT upgrade was associated with marked improvement in systolic function, with mean LVEF increasing from 29.3% to 45.2%. Among patients with severe PICM, defined as an LVEF ≤35% prior to CRT upgrade, more than 70% achieved LVEF >35% following upgrade. Those with narrower native QRS duration experienced the greatest improvement. Most patients with severe PICM who achieved LVEF >35% post upgrade did so within 1 year. Rates of ventricular arrhythmias were low, and none occurred within the first year following CRT upgrade.
Our study supports and extends previous findings demonstrating robust improvement in LVEF after CRT upgrade for patients with PICM. Our observed improvement in mean LVEF from 29% to 45% and overall response rate of 86% compares favorably with the mean absolute LVEF improvement of 9% observed in 2 small cohorts of patients with frequent RV pacing and systolic dysfunction upgraded to CRT (8,9). The smaller improvement in LVEF observed in these studies is likely the result of including patients with concomitant structural heart disease, such as obstructive coronary artery disease, and sarcoidosis, whom we excluded. Our findings are more consistent with those recently reported by Kiehl et al. (14), who found a CRT response rate of 84% among 29 patients with PICM. In aggregate, our findings and those of others suggest that, in patients carefully screened for alternative causes of cardiomyopathy and who therefore have a more confident diagnosis of PICM, a marked improvement in LVEF can be expected the majority of the time following CRT upgrade.
Importantly, our findings further demonstrate that a robust CRT response can still be expected despite the presence of severe PICM. Even among patients with severe PICM, defined as LVEF ≤35%, more than 70% of our patients achieved a post-upgrade LVEF greater than 35%. This is similar to the findings of Nazeri et al. (7), who reported LVEF improvement in 76% of patients with severe PICM undergoing CRT upgrade. Furthermore, a lower pre-upgrade LVEF was associated with increased magnitude of LVEF improvement in linear regression, suggesting that patients with more severe PICM have greater room for LVEF improvement following CRT upgrade. Taken together, these findings support the notion that the systolic dysfunction characteristic of PICM is due to the electromechanical dyssynchrony induced by RV pacing and that withdrawal of this deleterious stimulus by instituting CRT can lead to nearly complete resolution of RV pacing-induced abnormalities regardless of the extent of initial dysfunction (2,7).
Furthermore, our study demonstrates that substantial improvement can still be expected despite relatively longstanding PICM. The average patient in our study carried a diagnosis of PICM for nearly 1.5 years prior to CRT upgrade, and 25% of patients with severe PICM who ultimately achieved LVEF >35% had PICM for >2 years. The finding that time between PICM diagnosis and CRT upgrade was not a predictor of extent of improvement following CRT upgrade suggests that the window of opportunity to treat PICM remains open. It is possible that PICM is less likely to result in myocardial fibrosis or other permanent structural abnormalities even with prolonged duration, although this hypothesis requires further study. Overall, these findings suggest that regardless of duration of exposure to RV pacing, and regardless of nadir LVEF, clinicians can generally expect a robust improvement following CRT upgrade.
Although other studies have primarily examined short-term response to CRT upgrade, our study provides a more detailed characterization of the time course of LVEF improvement. In both the overall and severe PICM cohorts, the bulk of LVEF improvement occurred within the first 3 months following upgrade, although more modest improvement continued through the first year. Among patients with severe PICM who ultimately achieved LVEF >35% during follow-up, 70% achieved this level of improvement within the first year. This has important implications regarding the decision to upgrade to CRT pacemaker versus CRT defibrillator, as illustrated by the fact that nearly two-thirds of those in the severe PICM group who achieved LVEF >35% underwent initial upgrade to CRT defibrillator and later lost their guideline-based indication for primary prevention ICD (10). Furthermore, no patients had sustained ventricular arrhythmias within the first year following upgrade and all detected arrhythmias during follow-up were slow, monomorphic ventricular tachycardias unlikely to be life-threatening. Similarly, Barra et al. (15) found the rate of sudden arrhythmic death to be <1 per 100 patient years among those with PICM upgraded to CRT pacemaker. The combination of a high proportion of responders, rapid rate of improvement, low incidence of arrhythmic events, and simpler upgrade procedure suggests that a reasonable strategy would be to initially upgrade patients suspected of having PICM, regardless of duration and nadir LVEF, to a biventricular pacemaker (Central Illustration). If after 1 year the LVEF remains ≤35%, upgrade to CRT defibrillator could then be considered. Although this approach has potential to limit the costs and harms associated with unnecessary ICD implantation, it must be balanced with the risks of reopening a pocket for defibrillator upgrade, and therefore requires cost-effectiveness analysis and prospective validation (16).
This study adds to the growing body of knowledge highlighting the importance of the native QRS duration as a marker for myocardial vulnerability. We have previously demonstrated that a prolonged native QRS duration (in the absence of bundle branch block) is a strong predictor of future development of PICM in patients exposed to RV pacing (6). Similarly, prolonged native QRS duration has been associated with an increased risk of death in patients undergoing RV pacing (17). Our current study adds the novel finding that a prolonged native QRS duration predicts not only greater risk of developing PICM but also less favorable response following CRT upgrade. It is likely that a prolonged native QRS duration is indicative of subtle degrees of myocardial fibrosis and substrate vulnerability.
Our findings should be interpreted in the context of study design. First, because echocardiograms were performed for clinical purposes rather than at protocol-defined intervals, the rate of LVEF recovery cannot be precisely determined. Second, because most of our patients were >80% RV-paced prior to CRT upgrade, our power is limited to determine whether patients exposed to less frequent RV pacing derive similar benefit from CRT upgrade. Of note, we previously reported that percentage of RV pacing (when >20%) was not a significant predictor of incident PICM (6). Third, although we made maximal effort to exclude alternative causes of cardiomyopathy, it is possible that some patients classified as having PICM might have had concomitant structural heart disease. For example, not all patients underwent genetic testing. It is possible that some patients who presented with heart block carried a lamin A/C mutation, and thus the cause of cardiomyopathy was not RV pacing but rather an inherited cardiomyopathy. Nevertheless, this is similar to clinical practice, where PICM is diagnosed after a reasonable attempt to exclude alternative causes. Fourth, given that we carefully excluded alternative causes of cardiomyopathy, our findings reflect CRT response in a “pure” PICM population. A lesser degree of improvement in LVEF following CRT upgrade would be anticipated in patients with coexistent structural heart disease, such as obstructive coronary artery or valvular heart disease. Fifth, the robust response to CRT we observed among those with PICM does not argue for initial CRT in all patients with anticipated frequent ventricular pacing, as most patients exposed to frequent RV pacing never develop PICM.
CRT is highly effective in reversing PICM, with 86% of PICM patients demonstrating improvement in LVEF of ≥5% and 72% of severe PICM patients achieving LVEF >35% following upgrade. The benefits of upgrade appear to be rapid, with most improvement occurring within 3 months and more gradual improvement over the remainder of the first year. Rates of ventricular arrhythmia in our population were low, and none occurred within the first year after CRT upgrade. Given that most PICM patients respond to CRT, the bulk of response occurs within the first year, serious arrhythmic events are uncommon, and the upgrade procedure is simpler, a reasonable approach would be to initially upgrade all PICM patients to CRT pacemaker. If LVEF remains ≤35% after 1 year, further upgrade to CRT defibrillator could then be considered.
COMPETENCY IN MEDICAL KNOWLEDGE: CRT has been shown to improve morbidity and mortality in selected patients with symptomatic, systolic heart failure and conduction abnormalities. CRT can also be used to treat RV pacing-induced cardiomyopathy, although its role in this disease is not well characterized.
COMPETENCY IN CLINICAL CARE: We demonstrate that, after a careful effort to exclude alternative causes of cardiomyopathy, most patients with PICM experience substantial improvement in LVEF after CRT upgrade and that the pace of improvement is rapid.
TRANSLATIONAL OUTLOOK: We therefore propose that patients with PICM be upgraded to biventricular pacemaker initially. If the LVEF remains ≤35% after 1 year, upgrade to CRT defibrillator can be considered. Cost-effectiveness analysis will be needed to validate this strategy. The role of cardiac MRI as a predictor of response to CRT merits investigation.
This work was supported by the Koegel Family Electrophysiology Fund. Drs. Epstein and Marchlinski are consultants and compensated speakers for St. Jude Medical, Medtronic, Boston Scientific, and Biotronik. Dr. Lin is a consultant and compensated speaker for St. Jude Medical. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- cardiac resynchronization therapy
- implantable cardioverter-defibrillator
- left ventricular ejection fraction
- pacing-induced cardiomyopathy
- right ventricle
- Received July 24, 2017.
- Revision received October 2, 2017.
- Accepted October 4, 2017.
- 2017 American College of Cardiology Foundation
- Nielsen J.C.,
- Andersen H.R.,
- Thomsen P.E.B.,
- et al.
- Nielsen J.C.,
- Kristensen L.,
- Andersen H.R.,
- Mortensen P.T.,
- Pedersen O.L.,
- Pedersen A.K.
- Leon A.R.,
- Greenberg J.M.,
- Kanuru N.,
- et al.
- Epstein A.E.,
- DiMarco J.P.,
- Ellenbogen K.A.,
- et al.
- Surawicz B.,
- Childers R.,
- Deal B.J.,
- Gettes L.S.
- ↵R Core Team. R: a language and environment for statistical computing. R Foundation for Statistical Computing. Vienna, Austria. Available at: https://www.R-project.org/. Accessed January 6, 2017.
- Kiehl E.L.,
- Makki T.,
- Kumar R.,
- et al.
- Barra S.,
- Duehmke R.,
- Providencia R.,
- et al.
- Poole J.E.,
- Gleva M.J.,
- Mela T.,
- et al.
- Sweeney M.O.,
- Hellkamp A.S.,
- Lee K.L.,
- Lamas G.A.,
- Investigators M.