Article Text
PDF
Abstract
Objective The natural history of frequent premature ventricular complexes (PVCs) in association with preserved left ventricular ejection fraction (LVEF) is uncertain. The optimal management of this population is thus undefined. We studied the outcomes of untreated patients with frequent PVCs and preserved LVEF.
Methods This cohort study prospectively evaluated consecutive patients from 2012 to 2017, with asymptomatic or minimally symptomatic frequent idiopathic PVCs (≥5% PVCs in 24 hours; normal LVEF; no cause identified on comprehensive evaluation). No suppressive therapy (ablation or antiarrhythmic drugs) were used and patients were followed with serial ambulatory ECG monitoring and echocardiography. The primary arrhythmic outcome was reduction in PVC burden to <1% on serial ambulatory monitoring. The primary echocardiographic outcome was a reduction of LVEF to <50%.
Results One hundred patients met inclusion criteria (mean age 51.8 years, 57% female) with a median PVC burden of 18.4%. Reduction to <1% PVCs occurred in 44 of 100 patients (44.0%) at a median of 15.4 months (range 2.6 to 64.3). Recurrence was uncommon (4/44, 9.1%). Four patients (4.3%) with a persistently elevated PVC burden developed left ventricular dysfunction (LVEF <50%) during the follow-up period at a range of 53–71 months. The initial PVC burden did not predict subsequent resolution (HR 1.00(0.97, 1.03); p=0.86).
Conclusions A strategy of active surveillance is appropriate for the majority of patients with frequent idiopathic PVCs in association with preserved LVEF, owing to the low risk of developing left ventricular systolic dysfunction and the high rate of spontaneous resolution.
- premature ventricular beats
- epidemiology
Statistics from Altmetric.com
Introduction
Frequent premature ventricular complexes (PVCs) have historically been considered benign in the absence of other cardiac conditions, such as myocardial infarction (MI).1 2 More recently, frequent PVCs have been implicated in the development of left ventricular (LV) systolic dysfunction.3 4 However, the pathophysiology and epidemiology of PVC-induced LV dysfunction (termed ‘PVC-induced cardiomyopathy’) remains poorly understood.
Specifically, there are limited natural history and outcome data, which create uncertainty regarding the optimal management of patients with frequent idiopathic PVCs and preserved LV function. While a persistently high PVC burden predisposes to cardiomyopathy, it is uncertain which patients warrant invasive or pharmacological treatment.5–7
Defining the natural history of frequent idiopathic PVCs has important management implications. We prospectively evaluated the longitudinal PVC burden and clinical outcomes in patients with frequent idiopathic PVCs and preserved LV function.
Methods
Study population
Consecutive adults with frequent idiopathic PVCs referred to the Heart Rhythm Service at St Paul’s and Vancouver General Hospitals from 1 March 2012 to 31 March 2017 were prospectively enrolled in the British Columbia PVC Registry. Subjects underwent standardised evaluation and follow-up (supplementary appendix 1). Subjects were followed from the time of first ambulatory ECG documentation of frequent (≥5%) PVCs until 30 June 2018. The work was conducted according to the Declaration of Helsinki.
Supplementary file 1
Frequent idiopathic PVCs were defined as: (1) ≥5% PVCs in 24 hours on at least one ambulatory ECG (Holter) monitor. (2) Left ventricular ejection fraction (LVEF) ≥50% by echocardiography. (3) Idiopathic/unknown PVC aetiology despite comprehensive clinical evaluation including laboratory and imaging. Cardiac MRI was recommended for all patients in the absence of contraindications and all patients in this cohort showed no delayed enhancement or wall motion abnormalities. Further imaging was at the discretion of the treating physician. Of note, the inclusion threshold of ≥5% PVC burden was chosen to capture a broad population with frequent PVCs, while also recognising there exists an uncertain threshold of PVC burden that increases risk of LV dysfunction.8 9
All subjects underwent comprehensive baseline evaluation and follow-up in dedicated heart rhythm clinics. Exclusion criteria were age <19 years, LVEF <50% at time of initial assessment, previous MI or history of other structural heart disease (eg, dilated cardiomyopathy, sarcoidosis, arrhythmogenic right ventricular cardiomyopathy), or active treatment with class I or class III antiarrhythmic therapy. Of note, patients who failed ablation therapy or discontinued class I or class III antiarrhythmic drug (AAD) therapy were eligible for enrolment if the frequent PVCs persisted in the absence of therapy. Likewise, patients with persistent frequent PVCs despite class II or class IV antiarrhythmic therapy at baseline (β blockers or calcium blockers) were included, given the limited efficacy of these agents in reducing PVC burden;10 however these medications were discontinued on follow-up. Following enrolment, patients were followed longitudinally with serial 24 hours ECG monitoring and echocardiography.
Ambulatory ECG monitoring
All patients had serial 24 hours ambulatory ECG (Holter) recordings. Once enrolled, ambulatory ECG monitoring was recommended at a minimum of every 6 months for the first 24 months after diagnosis, then every 12 months thereafter. Given that a high proportion of patients with frequent PVCs are minimally symptomatic or asymptomatic,11 time 0 for the study was defined as the date of the first 24-hour ECG.
ECG data
The 12-lead ECGs were manually analysed with respect to the characteristics of the native beat and the PVC. The native beat and PVC QRS durations were measured, using an average over three beats. The PVC origin was determined using the 12-lead ECG for those patients with complete recordings (PVC morphology captured in all 12 leads). The PVC site of origin was classified into: (1) Right ventricular outflow tract (inferior axis, left bundle branch morphology and a V2 transition ratio <0.60).2 12 (2) Left ventricular outflow tract (inferior axis, left or right bundle branch block morphology and a V2 transition ratio ≥0.60). (3) RV non-outflow tract (non-inferior axis, left bundle branch block morphology). (4) LV non-outflow tract (non-inferior axis, right bundle branch block morphology).
LV function assessment
LVEF and dimensions (LV end-systolic and end-diastolic dimensions) were assessed using echocardiography and/or cardiac MRI. Most cardiac MRI protocols rely on summed imaging which may underestimate LV function with frequent PVCs. Therefore, any impaired LVEF by MRI was reassessed by echocardiography, and the echocardiographic estimate of LVEF was used for analysis. Given the similar, but more profound, limitations of radionuclide angiography, this modality was not used for reassessment of LV function. Echocardiography was recommended at a minimum of every 12 months.
For patients with a persistently elevated PVC burden (≥5% in 24 hours), serial echocardiography was performed at least annually. Where possible, the echocardiographic LVEF was determined using the Simpson’s biplane method on consecutive sinus beats.
Outcomes
The primary arrhythmic outcome was PVC resolution, defined as a reduction in PVC burden to <1% per 24 hours. This cut-off reflects approximately the 99th centile in the normal population.2 It was recommended that confirmatory ambulatory ECG monitoring be performed at least 6 months after demonstration of a reduction to <1% PVCs.
The secondary arrhythmic outcome was ≥80% reduction from the initial PVC burden, a threshold suggested to improve LVEF in PVC-induced cardiomyopathy.13
The primary non-arrhythmic outcome was reduction in LVEF to <50% as assessed by echocardiography.
Statistical analysis
Time to the primary arrhythmic outcome (reduction to <1% PVCs in 24 hours) was estimated using the Kaplan-Meier method. Cox proportional hazards models were used to examine univariate baseline predictors of spontaneous resolution. Statistical analysis was performed using STATA V.13 (StataCorp, USA).
Results
Baseline characteristics
We identified 194 patients with PVC ≥5% in 24 hours, between 1 March 2012 and 31 March 2017 (figure 1). Of those, 100 patients met inclusion criteria. Reasons for exclusion included baseline impaired LV function (45 patients with idiopathic LV dysfunction and 9 with prior MI), or active treatment with class I/III AADs or ablation (25) due to severe symptoms but with preserved LV function.
Flow chart demonstrating the design of the study population. LVEF, left ventricular ejection fraction; MI, myocardial infarction; PVC, premature ventricular complex.
For the 100 patients meeting inclusion criteria, the mean age was 51.8±16.5 years and 57 were female. The most common symptom was palpitations (table 1). Median follow-up from the first to last 24 hours monitor was 29.3 months (range 6 to 123). A median of four 24-hour monitors were performed per patient (range two to seven). The median initial PVC burden was 18.4% (range 5.4% to 49.8%). The PVC location was of right ventricular origin in 34%, with two-thirds originating from the outflow tract (table 2).
Baseline characteristics and comparison of clinical characteristics between patients with persistent PVCs and spontaneous PVC reduction
Characteristics of PVCs
Primary arrhythmic outcome
The primary outcome occurred in 44% of patients (n=44). The median time to PVC resolution (reduction to <1% in 24 hours) was 15.4 months (range 2.6 to 64.3, figure 2). Among those with spontaneous resolution, 75% of patients experienced resolution within 30.9 months.
Time to spontaneous resolution of PVCs to <1% in the overall cohort. The median time to spontaneous reduction was 10.9 months. PVC, premature ventricular complex.
There were few differences in the baseline characteristics between patients with a persistently elevated PVC burden, and those with reduction to <1% PVCs (table 1). There was a borderline significant (univariate) increased chance of resolution in patients with a baseline right bundle branch block and there was also an increased chance of resolution with increasing age. Baseline PVC burden had no bearing on PVC resolution and there was a borderline significant relationship between higher maximum PVC burden and persistence of PVCs (table 2 and figure 3). The majority of those with reduction to <1% PVC burden sustained the improvement. Only 9 of the 44 patients (20.5%) encountered a subsequent increase in PVC burden to ≥1%. Of those nine, only four exhibited a PVC burden of >5% (10.5% overall).
Secondary arrhythmic outcomes
Fifty-two patients (52.0%) had a ≥80% reduction in PVC burden. The median time to ≥80% reduction in PVCs was 14.1 months (range 1.2 to 64.3 months). The majority of patients experienced a persistent PVC burden reduction, with only 6/52 (11.5%) manifesting a subsequent increase in PVC burden to >20% of the initial burden. Among those six, four patients had a subsequent absolute PVC burden above 5%.
LV function outcomes
During follow-up, 4/100 (4.0%) developed reduced LVEF <50%, all of whom had persistently frequent PVCs. The median time to recorded LVEF <50% was 60.9 months (range 52.7 months to 74.8 months). The clinical characteristics and outcomes of these four patients are detailed in table 3. Only one patient developed clinical heart failure and severe LV dysfunction. This patient had a biscuspid aortic valve with mild regurgitation, baseline right bundle branch block, multimorphic PVCs and a progressive increase in PVC burden to 49% just prior to the onset of heart failure. The remaining patients developed asymptomatic LV dysfunction, with LVEF >40%.
Characteristics of patients experiencing a reduction in LVEF to <50%
Discussion
This study demonstrates a high rate of spontaneous reduction in PVC burden to normal levels among patients with normal LV function presenting with frequent, minimally symptomatic PVCs. This resolution predominantly occurred within the first 3 years of follow-up, and the majority demonstrated a persistent improvement. This study also demonstrated a relatively low rate of development of LV systolic dysfunction in this population, over 5 years of follow-up. However, we cannot exclude a higher risk of development of LV dysfunction over a longer term, as there may be a long latent period required.
It is well established that frequent PVCs may cause, or at least contribute to, LV dysfunction.11 13 This study does not invalidate those findings and PVC suppressive therapy should be considered in patients in whom frequent PVCs may be contributing to LV dysfunction. It is also important to note that this was an observational study that cannot answer the question of whether early or prophylactic treatment of frequent PVCs is warranted. However, given the high rate of spontaneous reduction observed and the lack of evidence supporting prophylactic therapy, we support a cautious approach.
To our knowledge, there are only two contemporary studies (in which structural heart disease was excluded) examining the natural history of idiopathic PVCs. Our study is consistent with a prior study of children with PVCs without underlying heart disease (n=163) in which the prognosis has been found to be favourable.14 The PVC group in that study had an average burden of 10.0%±8.9%, which was slightly lower than the current study. Similar to our study, the PVCs resolved during follow-up in 28% of the patients with the estimated mean time to disappearance of PVCs in the PVC group of 115.2±6.4 months. A recent study by Niwano et al looked at the prognostic significance of frequent PVCs (defined as >1000 beats/day) in 239 asymptomatic patients with normal LV function over 5.6 years.9 A burden of >20 000 PVCs/24 hours was associated with a subclinical deterioration in LVEF and LV dilation over 5 years. The rate of development of LV dysfunction was low (13/239, 5.4%) and comparable to the current study, despite the fact that the current study had an older population (mean 52 years vs 43 years) and a higher initial PVC burden (mean 18% vs 12%). Interestingly, there was a remarkable lack of reported PVC burden variability over time in the study, with no patients experiencing spontaneous resolution, which is in stark contrast to the findings of this study.
We showed that spontaneous reduction predominantly occurred within the first 3 years of follow-up, and the majority of these patients demonstrated a persistent reduction in PVCs. Several studies in the 1980s and early 1990s had noted a significant amount of ‘variation’ in PVC burden among untreated patients15–18 when PVCs were measured serially. However, most of these studies enrolled patients with newly discovered PVCs (often in hospitalised patients), with a relatively low burden (mean PVC burden ranged from 2% to 18% PVCs), who had a variety of underlying cardiac conditions and with limited longitudinal follow-up.
Limitations and future research
We monitored longitudinal PVC burden using serial 24-hour ECG monitoring which cannot account for day-to-day variation in PVC burden. Continuous or longer-term monitoring would be preferable but many technologies to accomplish this have no validated PVC-detection algorithm (implantable cardiac monitors, for example) or cannot be justified in otherwise healthy patients (dual chamber pacemakers). Nonetheless, longer-term monitoring of PVC burden may yield more insights into which patients may be at risk of persistence of the frequent PVCs and be at risk of development of LV dysfunction.
Referral bias to a tertiary referral centre may also have resulted in underestimation of the rate of spontaneous reduction to <1%. The echocardiographic follow-up in this study was relatively short and longer follow-up will be required to determine a more accurate estimate of the rate of development of LV dysfunction over time, particularly in light of the demonstration of a long latent period.
While no clinical predictors of spontaneous resolution were found in this study, we had a limited ability (low statistical power) to detect potentially important predictors, given the sample size of the current study. Further investigation into the pathophysiology of frequent idiopathic PVCs may identify clinical predictors and potentially previously unrecognised factors that may enhance our ability to prognosticate both spontaneous reduction and the development of LV dysfunction.
Conclusion
This study demonstrated a high rate of resolution of frequent PVCs among untreated patients with normal LV function and minimal symptoms. It also demonstrated a low rate of subsequent development of LV dysfunction. Given these findings, a conservative approach consisting of observation with serial echocardiographic and ambulatory ECG evaluation is reasonable for patients with minimal or no symptoms, normal LV function a negative initial work-up.
Time to reduction of PVCs to <1%, by initial PVC burden. PVC, premature ventricular complex.
Key messages
What is already known on this subject?
Frequent premature ventricular complexes (PVCs) have been associated with the development of left ventricular systolic dysfunction. It is unclear whether patients with frequent idiopathic PVCs warrant proactive therapy to prevent the development of left ventricular systolic dysfunction. The optimal management of this population is unclear.
What might this study add?
This prospective cohort study demonstrates that a significant proportion of patients will have a spontaneous resolution of their frequent PVCs, even in the absence of treatment. It also demonstrates that these patients are at low risk of developing cardiac dysfunction.
How might this impact on clinical practice?
The findings of this study support conservative management, with active surveillance, for most patients with frequent idiopathic PVCs in association with normal left ventricular function. Treatment targeting PVCs, such as antiarrhythmic drug therapy or catheter ablation, may not be required in the majority of these patients.
References
Footnotes
Contributors AKYL and MWD were responsible for the design of the project. AKYL, GA, MWD, SC, JA, MTB were responsible for data gathering. AKYL and MWD were responsible for data analysis. All authors critically revised the manuscript and all authors approved the final manuscript.
Funding This research was supported by funding from the Division of Cardiology of the University of British Columbia, in partnership with the Heart and Stroke Foundation of Canada. AKYL, MWD and ZWL are recipients of Career Scholar awards from the Michael Smith Foundation for Health Research. AK receives support from the Heart and Stroke Foundation of Canada, the Sauder Family and Heart and Stroke Foundation Chair in Cardiology and the Paul Brunes Chair in Heart Rhythm Disorders.
Competing interests None declared.
Ethics approval Providence Health Care/University of British Columbia ethics committee.
Provenance and peer review Not commissioned; externally peer reviewed.
Patient consent for publication Not required.