Resynchronization therapy in heart failure

Introduction Cardiac resynchronization therapy (CRT) received a class I-A recommendation for patients: with symptomatic chronic heart failure associated with left bundle branch block (LBBB) (QRS ≥150 ms) and severe left ventricular (LV) dysfunction (EF ≤35%), persisting after an adequate period of optimal medical therapy, and who are not affected with correctable underlying cardiac disease (myocardial ischemia, tachycardia-induced ventricular dysfunction, significant valvular problems). For patients with a QRS between 130 and 149 ms the indications for the implant are weaker, and for patients without LBBB the support for this treatment is scarce.1,2 For patients with a standard implantable cardioverter defibrillator the recommendation for CRT is class I-B if they present an EF 35% or less and worsening heart failure while experiencing a high proportion of right ventricular pacing, regardless of the width of the QRS. On the other hand CRT is not recommended for patients with QRS less than 130 ms, in whom it can actually contribute to an unfavorable outcome, as outlined in the recent guidelines (class III-A).1 (Table 1).Table 1: Summary of guidelines recommendationsIn selected patients CRT is effective in achieving symptomatic improvement, and better LV function, lengthening survival with better quality of life; the treatment is important for patients in NYHA class I–II to prevent disease progression, and in patients with NYHA class III–IV for the reduction of symptoms and morbidity. Between 20 and 30% of the patients are ‘nonresponders’, not experiencing significant benefits from CTR, in terms of neither symptomatic improvement nor betterment of LV function. A variety of pre, peri, or postprocedural factors can negatively affect the outcome and need to be addressed through a structured, multidisciplinary approach. The patients benefitting the most from this treatment have a reverse remodeling of the geometry of the LV, with reduction in diameters and volumes as well as an increase in the EF. These patients have usually a wider QRS, with LBBB, and are women with nonischemic cardiomyopathy.3 Imaging studies are not very helpful in representing LV mechanical dyssynchrony, and are thus not generally useful in selecting patients for CRT.2 Currently, we do not have a uniform agreed-upon definition for CRT ‘responders’ or ‘nonresponders’. ‘Responders’ are those patients experiencing a sustained improvement in quality of life, as well as in the clinical and instrumental parameters of heart failure, or no disease progression for those with less advanced heart failure.4 ‘Non responders’ are all the other patients. A multiparametric approach is necessary to correctly evaluate the efficacy of resynchronization therapy, and it should be based on the evaluation of the clinical status, symptoms, and quality of life, as well as the objective data derived from echocardiography and cardiopulmonary exercise test. Mortality and the number of repeated hospital admissions for heart failure are the best indicators of the efficacy of CRT, albeit not always strictly dependent on CRT itself. Echocardiographic measurements of cardiac dimensions and possible reverse remodeling are in favor of a positive effect of CRT, even though not always accompanied by symptoms improvement. In general, the evaluation of the effect of CRT requires at least 6 months of optimal medical and device therapy.5 Cardiopulmonary exercise test is useful in the evaluation and stratification of patients after CRT, despite the fact that the results are operator dependent and the test is time consuming. The number of ‘responders’ reported by various studies changes according to the end point selected (echocardiography evidence of remodeling, other instrumental parameters, mortality/recurrent episodes of heart failure, or a combination of clinical variables), with higher proportion of ‘nonresponders’ when the selected end point include a combination of quality of life/subjective symptoms.6 Optimizing CRT, by limiting the number of ‘nonresponders’, is a three phase process: accurate patient selection; correct implant technique and lead positioning; and correct device programming. Patients selection The first step toward a correct approach to CRT is the selection of patients with heart failure, severe LV dysfunction (EF ≤35%), and dyssynchrony caused by delay in left interventricular conduction (spontaneous or induced by right ventricular pacing). Patients affected with ischemic heart disease, or with valvular or arrhythmic problem should have their underlying condition treated before considering CRT.7 Optimizing medical treatment as well as the complete information and education of the patients are the necessary first steps toward the selection of the candidates for the treatment. It is important to consider that CRT often allows a positive modulation of bradycardia treatment by increasing the tolerance to the drugs, particularly to β-blockers. Furthermore, there are factors associated with a proven lesser efficacy of CRT treatment that must be considered, such as: severe heart failure (NYHA class IV), male sex, atrial fibrillation, nonsignificant interventricular conduction delay, severe left atrial enlargement, and severe mitral insufficiency. Apart from these factors among the main causes of ineffective CRT are the presence of supraventricular arrhythmias and frequent ectopic ventricular beats (PVC), causing loss of capture in more than 10% of patients after CRT. Accordingly, a better control of the ‘arrhythmic burden’, either pharmacological or nonpharmacological, could lead, per se, to an improvement of the LV function,7 and allows more adequate use of biventricular pacing after the implant of the device. Severe functional mitral insufficiency is common in patients candidates for CRT. After the procedure, mitral regurgitation improves by upto 40%,8 whether the persistent, at least moderate, mitral regurgitation is an unfavorable prognostic sign.8 Should important residual mitral regurgitation persist after the CTR, the MitraClip procedure could represent a valid alternative for ‘nonresponders’ with advanced NYHA class at high risk for surgical correction.9 An added value of cardiac magnetic resonance, apart from the assessment of LV function and the presence of dyssynchrony, is the tissue characterization, namely the extent and location of myocardial fibrosis. When no significant posterolateral fibrosis (<15% of myocardial tissue) is demonstrated, CRT is more effective; on the other hand, the number of ‘nonresponders’ is particularly high when pacing occurs in scar tissue areas. Accordingly, magnetic resonance has been suggested as guiding tool for lead positioning.10 Atrial fibrillation represents a peculiar situation. In fact, the effectiveness of CRT is inferior as compared with patients in sinus rhythm, but numerous studies, as well as recent guidelines, confirmed a role for CRT in patients with atrial fibrillation (recommendation class IIa-B),2 as long as adequate biventricular pacing is achieved, as close as possible to 100%. Lacking reliable echocardiographic criteria for selecting the best patients for CRT, ECG morphology remains the cornerstone for the selection of patients with the best chance for improvement after CRT: a ‘true’ LBBB (QS or rS morphology in V1–V2, duration ≥140 ms in men and ≥130 ms in women, notched QRS in ≥2 leads among I, aVL, V1,V2, V5, and V6), as compared with other patterns of conduction delay, is associated with a higher efficacy of CRT.11 Implantation of cardiac resynchronization therapy device Close attention should be given, at the time of the implant, to the positioning of the ventricular leads; in general the wider the distance between the right and the left lead, the better the resynchronization. When venous coronary anatomy allows, the LV lead should be placed in the posterolateral area, not too close to the apex; in any case the optimal site is where the longer delay between QRS beginning and LV signal is found (or the longer interval between two ventricular signals). A more recent solution is the quadripolar electrode LV lead (four pacing locations along the distal end of the lead), which can pace via transvenous access, the LV wall, at several locations and multiple vectors along the lead; optimal pacing sites can more easily be reached and suboptimal sites can be avoided, without compromising lead stability, as the tip can be wedged in a distal part of the vein while other electrodes are placed near the optimal pacing site. Moreover, quadripolar leads offer the possibility of timing optimization, because a single quadripolar lead can offer multiple pacing options. Accordingly, the number of ‘nonresponders’ has been reduced, as well as those patients in whom phrenic nerve stimulation prevents appropriate resynchronization from occurring.4,12 Finally, quadripolar electrode LV lead, allowing multipoint pacing (pacing the LV with two of four electrodes), could further enhance response.13 Programming and device follow-up Optimal response to CRT is also dependent on appropriate device programming both at the time of the implant procedure and during the follow-up. In the first place, pacing as close to 100% as possible should be targeted. Commercially available devices provide data regarding the pacing delivered, which could overestimate the actual effective pacing, considering the fusion and pseudofusion beats (where the ventricular activity is almost exclusively spontaneous). Although some device companies are developing specifically designed algorithms, at present the distinction between ‘effective’ pacing and fusion/pseudofusion beats, is the domain of surface ECG analysis (e.g. a short paced P-QRS interval and a tall R in V1). In patients with atrial fibrillation achieving an adequate pacing rate could be difficult, such that an A-V node ablation procedure is often required after the implant. The suboptimal programming of the atrioventricular and ventricular–ventricular interval is one of the common causes of lack of response to CRT. Echocardiography is not reliable as a guide to optimization of these intervals, and is also time consuming and of difficult implementation, particularly in high volume centers. Multicenter studies suggested the noninferiority of empiric (with atrioventricular interval 100–120 ms and synchronous pacing of the right and LV) as compared with ‘echo-guided’ programming. The optimal atrioventricular and ventricular–ventricular interval could change over time as the heart rate varies. Accordingly, a further reduction of the number of ‘nonresponders’ has been achieved with the use of algorithms, periodically and in automated fashion, able to modify the device programming, employing also hemodynamic sensors for detection of atrioventricular and ventricular–ventricular optimal intervals. Finally atrial pacing could lead to an atrioventricular dyssynchrony, determining an intraatrial conduction delay. The unnecessary atrial pacing should be avoided by reducing the minimal pacing rate. Conclusion To improve the risk/benefit ratio and to optimize the results of an invasive and expensive treatment such as CRT, it is paramount to accurately select the patients, provide a correct implantation technique, choose the most appropriate device, and organize an attentive follow-up. To accomplish this goal a close multidisciplinary cooperation is necessary (cardiologist, electrophysiologist, nursing staff, device companies, and so on), which will also include a case by case evaluation of the possible causes of less than expected advantages of CRT. Acknowledgements Conflicts of interest There are no conflicts of interest.