Non responders to BIV pacing
The most important question to address about non response to CRT is how to define non responders. Some studies define non responders based on absence of LV remodeling, whereas other studies define non responders based on lack of symptom improvement.
Moreover, achievement of hemodynamic stability, even in the presence of persistent dyspnea or absence of significant reverse remodeling, as demonstrated by Mullens et al38, might be also a desirable goal of CRT. Definition of response to CRT can also be challenged by patients’ expectations. Patients in advanced heart failure are looking for symptoms’ relief and as they feel significantly better after CRT, their attention would be directed toward less frequent admission to the hospital, greater need for social life and activities, and finally prolongation of life39.
Assessment of CRT response
Yu et al40 first demonstrated that a better outcome is achieved with a reduction of left ventricular (LV) end-systolic volume (ESV) of at least 10% after CRT; patients who showed a significant reverse remodeling had survival rate close to 90% at 3 year follow-up compared to a survival rate of about 50% of the remaining patients. Similarly, Ypenburg et al41,showed that both hospitalization and mortality were related to the magnitude of change in LVESV; patients who showed further dilatation of LVESV had the highest event rate (about 70% at 3 years follow-up) compared to those who showed impressive reduction (near normalization) of LVESV (about 6%). It needs to be kept in mind that measuring reverse remodeling has several technical problems. The assessment of LV volume and LVEF using 2-dimensional echocardiography (the most frequently used technique) has a relatively large inter-observer variability39. The Predictors of Response to CRT (PROSPECT) study used 2-dimensional echocardiography for quantifying LV volumes42; the coefficient of variation for LVESV measurement was 3.8% for intra-observer variability but was as high as 14.5% for inter-observer variability.
Predictors of CRT Response
Patients with LBBB morphology have demonstrated the best response to CRT, whereas those with non-LBBB morphology generally have responded poorly. In a meta-analysis of COMPANION, CARE-HF, RAFT, and MADIT-CRT, Sipahi et al.43 found that patients with RBBB or IVCD did not benefit from CRT (relative risk for the composite primary outcome: 0.97; 95% confidence interval: 0.82 to 1.15; p = 0.75).
Birnie et al.44 analysed data from the RAFT trial. A total of 1,483patients in sinus rhythm with QRS durations >120 ms were examined. A LBBB was present in 1,175 patients (79.2%), 141 patients had an RBBB (9.5%), and the remainder (11.3%) had an IVCD. Patients with RBBB and IVCD were more likely to have an ischemic etiology of HF. Among the patients with LBBB, the benefit of CRT increased directly as QRS duration increased. In contrast, the benefit of CRT only began to emerge in the non-LBBB patients once the QRS duration was >160 ms and only after 2 years of follow-up. Sundaram V et al.45 has shown from the analysis of the Medicare registry that the improvements in both survival and HF hospitalization with CRT-D were greatest in patients with QRSD ≥180 ms with or without LBBB. The presence of electrical resynchronization leading eventually to mechanical resynchronization predicts a good response to CRT. This can be determined by observing the reduction in QRS duration46 and by a change in the shape of the QRS complex (indicating fusion of right ventricular and LV originating activation waves)47. A sequence of right-to-left electrical and mechanical activation that is present in LBBB or RV pacing is detrimental11,17,48. Thus, proper fusion of RV and LV activation waves requires proper positioning of the pacing leads. Also, the amount of scar tissue appears to be a predictor of CRT-non-response49. Bleeker et al. have shown that CRT does not reduce LV dyssynchrony in patients with transmural scar tissue in the posterolateral LV segments, resulting in clinical and echocardiographic nonresponse to CRT.
Considering that location of the pacing lead is of paramount importance, it is preferred for the LV lead to be placed in non-apical areas50. A posterolateral scar should be avoided and the areas of maximal electrical delay should be targeted51. Patients are more likely to be responders if the qLV at implant is more than 95 ms52. Alternatively, , LV can be paced from more than one site either by multipoint pacing (pacing from multiple sites from the same lead) or from multisite LV pacing 53. In multisite LV pacing, the first LV lead is inserted into a postero-lateral or lateral vein. The second LV lead is placed as far as possible from the first lead, in the anterior vein, a high antero-lateral vein, or the middle cardiac vein. Another modality for LV pacing is LV endocardial pacing with transventricular delivery of an active-fixation lead to the endocardial wall of the lateral LV 54. Advantage of LV endocardial pacing is easy access to all LV endocardial surface but the disadvantages are requirement for lifelong anticoagulation and the risk for lead extraction. One interesting new technique is the LV pellet that can be delivered retrogradely via the aorta into an LV site. In the select LV study, Reddy et al. applied this technology to non-responders and showed an improvement of EF at 6 months55.