PO-02-020 CHARACTERIZING PREDICTORS AND OUTCOMES OF LATE RESPONDERS TO CARDIAC RESYNCHRONIZATION THERAPY

Abstract

Typically, response to cardiac resynchronization therapy (CRT) is evaluated at 6 months. There is a subset of patients who do not fully respond to CRT at this point, however, do demonstrate improvement at 12 months. Characterizing these patients and their outcomes has not been widely explored. We examined in the MORE-CRT MPP study, the predictors of patients who do not demonstrate a reduction in left ventricular end-systolic volume (LVESV) >15% at 6 months, who however do reach that endpoint at 12 months. The MORE-CRT MPP study is a prospective, randomized, multicenter study. All patients (n = 5803) initially received conventional biventricular (BiV) CRT for 6 months. At this stage, echocardiographic non-responders were randomised to either continued conventional BiV pacing, or MultiPoint pacing. This post-hoc analysis evaluated the BiV pacing group only. This group was split into: i) non-responders (<15% reduction in LVESV at 6 and 12 months), ii) early responders (>15% reduction in LVESV at 6 months), and iii) delayed responders (<15% reduction in LVESV at 6 months, but >15% improvement at 12 months). Logistic regression models were used to assess the baseline characteristics of delayed responders. Heart failure hospitalization was the clinical endpoint. Of the 2,920 patients in the analysis 16.7% (n=486) were non-responders, 78% were early responders (n=2263), and 5.8% (n=171) were delayed responders. On multivariable analysis, only patients with ischaemic aetiology and a reduced QRS duration/left ventricular end-diastolic volume (LVEDV) ratio at baseline were statistically significant for predicting delayed response (see table). Freedom from heart failure hospitalization was similar in the early and late CRT responders, and significantly worse in the non-responder group. Patients with increased scar burden and those with a lower QRS/LVEDV ratio (i.e., reduced relative degree of dyssynchrony) were likely to be delayed responders to conventional BiV CRT. Early and delayed CRT responders have a similar risk of heart failure hospitalization despite the slower echo remodelling.Tabled 1Univariate and multivariate Cox regression model to predict likelihood of delayed response to cardiac resynchronization therapy (CRT).ParameterUnivariate AnalysisMultivariate AnalysisOR [95% CI]p-valueSample SizeOR [95% CI]p-valueAge1.004 [0.990, 1.019]0.5724351.01 [0.99, 1.02]0.46COPD (Yes vs No)0.804 [0.466, 1.388]0.432436Diabetes (Yes vs No)1.106 [0.795, 1.539]0.552436Hypercholesterolaemia (Yes vs No)1.099 [0.800, 1.509]0.562436Hypertension (Yes vs No)0.857 [0.624, 1.176]0.342436Ischaemic vs Non-Ischaemic1.638 [1.198, 2.240]0.00224361.64 [1.13, 2.37]0.009LBBB vs Non-LBBB0.650 [0.448, 0.944]0.02418640.75 [0.51, 1.10]0.14LVEF0.983 [0.961, 1.006]0.152436NYHA I/II vs NYA III/IV1.066 [0.779, 1.459]0.692427QRS duration/LVEDV ratio1.000 [0.993, 1.006]0.9320670.47 [0.22, 0.99]0.047Renal disease (Yes vs No)1.244 [0.799, 1.937]0.332436Female vs Male0.786 [0.555, 1.112]0.1724360.96 [0.63, 1.46]0.84 Open table in a new tab