174 Monday, 11 June 2007 Poster Display V. Pacing, resynchronisation and internal defibrillation associated with the device interrogation within 1 week. HRV and ACT were lower 1 week after Z events in both groups compared to baseline (p<0.001). After 3 weeks, HRV and ACT were significantly higher in Group ON compared to Group OFF (p<0.0001 for HRV; p<0.05 for ACT). Conclusions: Decreases of Z measured by an implantable device are associated with decreased ACT and HRV 1 week after the event. Both parameters recovered more quickly in pts seeking care in response to an audible alert than in those with the alarm programmed off. These results are in favor of the clinical usefulness of automatic patient alerts based on changes in Z. 765 Is dyssynchrony alone sufficient to identify responders to cardiac resynchronisation therapy C. Parsai 1 , B. Bijnens 2 , G.R. Sutherland 1 , A. Baltabaeva 1 , P. Claus 2 , G. Derumeaux 3 , M. Cannesson 1 , L. Anderson 2 1 St George’s Hospital, Cardiology Dept., London, United Kingdom; 2 University of Leuven, Leuven, Belgium; 3 Hopital Louis Pradel, Cardiology, Lyon, France Currently, the timing of long axis systolic events (regional velocity or strain indices) is used to identify left ventricular (LV) dysynchrony and thus predict response to CRT. However, overall passive LV rotational dis- placement induced by preserved right ventricular (RV) long axis short- ening may negate their use in “rocking hearts” which rotate around their long axis as a result of unbalanced myocardial forces. Furthermore, such a simplistic long axis velocity-based approach assumes a single mecha- nism of effect (i.e. dysynchrony) in both responders and non-responders. To determine if other mechanisms could underlie a CRT response, the fol- lowing study was performed. Data from 62 pts (pre- CRT, 66±1.4 years, ejection fraction (EF) 23±0.8%, QRS duration > 130 ms) were obtained at baseline (pre-CRT) and 6 months post-CRT. ↓NYHA class and LVESV ≥ 10% were defined as response markers. Dysynchrony was estimated by both regional longitudinal and radial velocity parameters and data com- pared with 2 established long axis Dyssynchrony Indexes. Four different pathophysiologic subgroups (3 responders/1 non-responders) were iden- tified: Group 1: 30 pts (48%) had radial septal/lateral wall dyssynchrony with an early peak septal radial velocity occurring in the isovolumic con- traction period -defined as a “septal flash”(SF). SF resolution with pacing was 100% predictive of long term response. 5 pts (16%) had SF persis- tence post- pacing and all failed as long-term responders. Group 2: 15 pts (24%), had no radial LV dyssynchrony but had abnormal LV filling with either an absent A wave or a long A-V delay+diastolic mitral re- gurgitation. 80% responded with only AV optimisation. Group 3: 5 pts (8%), had none of the above parameters, but exhibited abnormal RV - LV septal interaction with abnormal passive septal displacement causing dysynchrony. A marked clinical improvement (NYHA <1.5±0.3) was obtained in all after pacing-induced RV-LV optimisation, despite the lack of obvious remodelling. Group 4, 12 pts had none of the above response predictors and all (100%) failed to respond. Current long axis-based dysynchrony used to identify potential responders would have predicted that 75% should have responded. Conclusions: Within the CRT population there are 4 different sub groups of responders. The most important underlying pathophysiologic mecha- nism in responders is radial (and NOT long axis) intraventricular dysyn- chrony (RID). Immediate resolution of RID by LV pacing was 100% predictive of a long term response. However in 24% of pts, response may be achieved simply by AV optimisation. 766 Incidence of atrial and ventricular arrhythmias in heart failure patients treated by cardiac resynchronization therapy: a prospective, observational study S. Boveda 1 , E. Marijon 1 , A. Bulava 2 , J.P. Cebron 3 , P. Defaye 4 , N. Delarche 5 , M. Lambiez 6 , P. Chevalier 7 on behalf of MONA LISA 1 Clinique Pasteur, Toulouse, France; 2 Faculty Hospital, Olomouc, Czech Republic; 3 Nouvelles Cliniques Nantaises, Nantes, France; 4 University Hospital, Grenoble, France; 5 General Hospital, Pau, France; 6 Boston Scientific, Clinical Department, Rueil-Malmaison, France; 7 University Hospital, Lyon, France Purpose: Heart failure (HF) and arrhythmias often coexist and are re- sponsible for increased mortality and more frequent hospitalizations. MONA LISA study was designed to assess the incidence and the im- portance of sustained atrial tachycardia (sAT), and/or ventricular tachy- cardia (sVT) based on stored electrograms (EGM) during the follow-up of patients (pts) treated with CRT pacemaker (CRT-P). Methods: The MONA LISA study was observational, prospective, and carried-out in 24 international centers. 158 pts who remain symptomatic in NYHA Class III or IV HF despite optimal medical therapy were in- cluded and implanted with a CRT-P. Otherinclusion criteria were: intra and/or interventricular dyssynchrony, measured by echocardiography and tissue Doppler imaging, LVEF ≤ 40% and no indication for an ICD. Results: Mean age was 71±10 yrs. 68.2% were male. Mean QRS du- ration and LVEF were 163±28 ms and 26±8%, respectively. 39.2% of pts had an ischemic cardiomyopathy (ICM) and 88% were in NYHA Class III. 75% received ACE inhibitors, 17% ARB, 62% beta-blockers and 41% were under antiarrhythmic therapy. Atrial and ventricular ar- rhythmias were considered sustained if they lasted ≥5 min and ≥30 sec, respectively. During a mean follow-up of 10.7±4.8 months, the incidence of sAT and sVT based on at least one stored EGM was re- spectively 12% and 4.4%. Clinical characteristics and consequences are shown in teh table . One pt developed both sAT and sVT. Two out of 7 pts with sVT were upgraded with an ICD, 1 pt refused and received amiodarone, and for 1 pt the decision is pending. There was no signifi- cant difference between demographic data of pts with/without sustained arrhythmias. All-cause mortality was 8.2%. Two deaths were related to sVT, 3 were related to CHF, and the remaining were non-cardiac. The NYHA functional classification improved for 85% of the pts after im- plantation. Pts ICM Atrial arrhythmia Hosp* Drug Upgraded Deaths* history changes* to ICD* sAT %(n) 12 (19) 26 (5) 47 (9) 21 (4) 16 (3) NA 5 (1) sVT %(n) 4.4 (7) 43 (3) 57 (4) 43 (3) 29 (2) 29 (2) 29 (2) *due to arrhythmia. Conclusion: Preliminary results of the MONALISA study showed that the EGM-based prevalence of sAT and sVT was 12% and 4.4% respec- tively in CHF pts treated with a CRT-P. These findings may help tailor pharmacological and non-pharmacological treatment of CRT-P patients. 767 Prognostic role of ventricular dyssynchrony in patients with chronic heart failure O.Y. Khaleva, N.P. Nikitin, A.L.Clark, K. Good, A. Rigby, J.G.F. Cleland The University of Hull, Department of Academic Cardiology, Kingston Upon Hull, United Kingdom Objectives: QRS prolongation predicts a worse prognosis in patients with heart failure and this is thought to reflect the effects of dyssynchrony. However, the relationship between ventricular dyssynchrony measured by echocardiography and survival is not clearly established. To deter- mine the relationship between the severity of inter- and intra-ventricular