Author + information
- ↵∗Reprint requests and correspondence:
Dr. Jeanne E. Poole, Division of Cardiology, University of Washington, 1959 NE Pacific Street, Box 356422, Seattle, Washington 98195.
The benefit of cardiac resynchronization therapy (CRT) for most heart failure (HF) patients with low ejection fraction (EF) and widened QRS is undisputed. Mortality and HF hospitalization are reduced, and multiple measures of cardiac function are improved. It remains of clinical importance that between 30% and 50% of patients are poor responders to CRT. The obligate requirement to achieve resynchronization requires appropriate candidate substrate and effective CRT pacing.
In this issue of the Journal, Ruwald et al. (1) present a report of baseline ectopic beats as a new marker of nonresponse from the rich database of the MADIT-CRT trial (Multicenter Automatic Defibrillator Implantation Trial with Cardiac Resynchronization Therapy), which has yielded an abundance of interesting substudy observations (1,2). The original study was designed to test the hypothesis in a randomized clinical trial that less symptomatic HF patients (mostly New York Heart Association functional class II) with low EF (≤35%) and widened QRS (≥130 ms) would benefit from CRT added to an implantable cardiac-defibrillator. Several substudies from this trial have made contributions to influence clinical practice, mainly the post-hoc analysis that patients with left bundle branch block patterns derived the greatest benefit from CRT.
This substudy focuses on the results of baseline 24-h Holter monitor recordings before CRT implantation. The density of atrial premature contractions (APCs) and ventricular premature contractions (PVCs) was assessed for their association with subsequent clinical and left ventricular (LV) reverse remodeling response.
Overall, the authors identified a low incidence of baseline ectopy with 3 predefined groups of patients based on quintiles of the percentage of ectopic beats identified on the 24-h Holter recording: <0.1% (n = 160), 0.1% to 1.5% (n = 321), and >1.5% (n = 320). The mean percentage of ectopic beats was relatively low at 3.2 ± 0.8% of all beats, and only 9.5% of patients had ≥10% ectopic beats.
Two important clinically relevant points were made by this study: 1) It reconfirmed that CRT needs to be delivered almost all of the time (>97%, the mean percent CRT pacing in MADIT-CRT) for the benefit to be realized; and 2) it revealed that the presence of a strikingly small amount of baseline recorded ectopic beats (0.1% to 1.5%) on Holter monitoring was associated with an unfavorably low amount of CRT pacing, increased clinical HF hospitalization and death events, and less LV reverse remodeling response as quantified by end-systolic volume reductions at 1 year. An impressive 13-fold increase in the risk of achieving low CRT pacing (<97%) was observed in patients with baseline ectopic beats >1.5%.
Although the authors combined APCs and PVCs in their analysis, the major differences between groups in this report appear related to ventricular ectopy. Specifically, poor outcomes were driven by the presence of PVCs, not APCs, despite both varieties of ectopy predicting a lower percentage of CRT pacing. The reason for this may simply be that the majority of ectopic beats identified on the 24-h Holter monitor were ventricular, and the associations of APCs with less CRT pacing and poor outcomes were less convincing.
As might be expected, patients with higher levels of ectopy were more ill and importantly were less likely to have a left bundle branch block or wider QRS, both of which were potent predictors of improved outcomes with CRT in MADIT-CRT. The observation that patients with higher levels of ectopic beats were less likely to experience LV reverse remodeling is not surprising purely from the perspective of patients already known to have a lower likelihood of CRT response by clinical and electrocardiographic features. The striking finding in this study was that patients with as low as ≥0.1% ectopy had worse outcomes. In fact, a burden of >1.5% ectopic beats added negligibly to the long-term risk of HF hospitalization or death already observed at a level of 0.1% ectopy.
What is the explanation for these findings? Previous literature has reported that frequent APCs, PVCs, and nonsustained atrial or ventricular tachyarrhythmias interfere with the delivery of effective CRT pacing. However, the mechanism whereby ectopy such as PVCs at a frequency of only 0.1% over 24 hours can decrease CRT pacing to <97% and interfere with LV remodeling is not entirely clear. Rapid intrinsic atrioventricular conduction during atrial fibrillation is a frequently encountered reason for loss of effective CRT. It is possible that ectopy identified by Holter recordings, even in small amounts, was associated with a greater chance of a higher ectopic beat burden over time, which may be in the form of new sustained or nonsustained arrhythmias. The authors postulated that PVCs may create havoc due to the known interplay of device-induced refractory periods in response to sensed ventricular beats resulting in loss of synchrony in subsequent beats. Although it is well known that ventricular ectopy is a marker of a higher-risk patient population, the observed association of worse outcomes in patients with a very small percentage of ectopy in this study is quite remarkable.
Hayes et al. reported that achieving CRT pacing at a target of >98.5% discriminated best for mortality when patients enrolled in the LATITUDE database were examined (Boston Scientific, Natick, Massachusetts) (3). This was true regardless of whether the patient’s dominant rhythm was atrial fibrillation or sinus rhythm. An important observation was that patients with atrial fibrillation and >98.5% CRT pacing had a mortality outcome only as good as sinus rhythm patients who did not achieve the target of >98.5% pacing. Nevertheless, their mortality was significantly reduced compared with patients with atrial fibrillation who did not achieve the target of >98.5% pacing. This study demonstrated the interaction of comorbid risk factors and the importance of achieving a very high percentage of CRT pacing for beneficial effects to be realized. Similar effects of both intermittent atrial beats and nonsustained ventricular tachycardia on percent CRT pacing, measures of remodeling (left atrial volume/left ventricular dimensions), and HF or death have been demonstrated previously by the MADIT-CRT investigators.
Like all post-hoc subgroup analyses from clinical trials with patients enrolled and randomized using different criteria, great caution should be used when one extrapolates cause-and-effect associations with clinical outcomes. In particular, PVCs have long been associated with poorer clinical outcomes among patients with low EF, regardless of CRT. Important baseline factors associated with PVCs and poorer clinical outcomes included markers of ischemic heart disease, such as prior myocardial infarction and history of bypass surgery. Accordingly, these associations with poorer clinical outcomes may be related in part to the substrate of myocardial scar and natural history of ischemic heart disease, which is comparatively different from nonischemic forms of cardiomyopathy.
The question we are left with is whether a very low level of ectopy (>0.1%) is enough to preclude patients from consideration of CRT. We would suggest it is not. The finding of even small amounts of ectopy identified on 24-h Holter monitoring appears to be an important identifier of a more ill patient and thus poorer outcome, which along with other unfavorable clinical factors may be predictive of poor CRT response. However, ectopy on a Holter monitor does not exclude patients from CRT on the basis of current guidelines. The options for treatment in the individual patient will dictate the choice of appropriate therapy. The knowledge that frequent ectopy will interfere with achievement of target levels of CRT pacing may suggest a specific treatment plan. It remains unclear whether attempts at suppression of PVCs with antiarrhythmic medications in these HF patients with low EF are warranted. In the case of frequent monomorphic PVCs, ablation therapy may be considered reasonable on the basis of recent supportive data (4,5). Notably, neither antiarrhythmic medication use nor ablation was directly addressed in the present study.
In conclusion, a little bit of premature beats and a little bit of incomplete biventricular pacing can mean a lot to CRT patients. This report reaffirms that nearly continuous CRT pacing (>97%) is necessary for optimal CRT response and that even a very small percentage of ectopic beats (>0.1%) is associated with less CRT pacing and worse patient outcomes after CRT. The report by Ruwald et al. (1) underscores the fact that more must be learned about the complexities of CRT response.
↵∗ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
Dr. Gorcsan has received research grant support from GE, Toshiba, Medtronichttp://dx.doi.org/10.13039/100004374, and Biotronik; and is a consultant for CardioInsight and Covis Pharmaceuticals. Dr. Poole has received research funding from the National Institutes of Healthhttp://dx.doi.org/10.13039/100000002/National Heart, Lung, and Blood Institutehttp://dx.doi.org/10.13039/100000050 (NIH/NHLBI NCT00911508) and Sanofi-Aventis; educational speaking fees from Boston Scientific, Biotronik, Medtronic, and St. Jude Medical; advisory board/consulting fees from Boston Scientific and Physio-Control; and has equity in Cameron Health.
- American College of Cardiology Foundation
- Ruwald M.H.,
- Mittal S.,
- Ruwald A.-C.,
- et al.
- Lakkireddy D.,
- Di Biase L.,
- Ryschon K.,
- et al.