Brian Olshansky, MD, FHRS,* John D. Day, MD, FHRS, Renee M. Sullivan, MD,* Patrick Yong, MSEE, Elizabeth Galle, MS, Jonathan S. Steinberg, MD, FHRS

Does cardiac resynchronization therapy provide unrecognized benefit in patients with prolonged PR intervals? The impact of restoring atrioventricular synchrony: An analysis from the COMPANION Trial Brian Olshansky, MD, FHRS,* John D. Day, MD, FHRS, Renee M. Sullivan, MD,* Patrick Yong, MSEE, Elizabeth Galle, MS, Jonathan S. Steinberg, MD, FHRS From *University of Iowa Hospitals and Clinics, Iowa City, Iowa; Intermountain Heart Rhythm Specialists, Salt Lake City, Utah; Boston Scientific Corporation, St. Paul, Minnesota; Valley Heart and Vascular Institute, Columbia University College of Physicians & Surgeons, New York, New York. BACKGROUND The influence of PR prolongation on outcomes after cardiac resynchronization therapy (CRT) is uncertain. OBJECTIVE To determine whether PR prolongation predicts outcomes in potential CRT candidates and whether CRT benefits these candidates regardless of baseline PR interval. METHODS A database of 1520 patients fulfilling criteria for CRT implant (the COMPANION Trial) was examined. Patients assigned to normal (PR 200 ms) or prolonged (PR 200 ms) cohorts were compared within the optimal pharmacologic therapy (OPT) and CRT groups regarding an endpoint of all-cause mortality or heart failure hospitalization. CRT was compared with OPT in normal and prolonged PR interval groups. An interaction test was performed to determine whether CRT influenced outcome as a function of PR interval. RESULTS PR prolongation was present in 52% of COMPANION subjects. Randomization to CRT was associated with a reduction in the endpoint, but the strength of the association was greater for those with prolonged PR (hazard ratio 0.54; P.01) versus normal PR (hazard ratio 0.71; P.02) intervals. CRT (vs OPT) was associated with reduction in the endpoint for subjects with normal or prolonged PR intervals. Reduction in relative risk (CRT vs OPT) was 29% (P.02) for those with normal PR intervals but was 46% (P.01) for those with PR prolongation. No interaction was detected between PR interval cohort and treatment (P.17). CONCLUSIONS PR prolongation may affect mortality and heart failure hospitalizations in patients with systolic dysfunction, heart failure, and wide QRS complexes. The effect of PR prolongation may be attenuated by CRT. KEYWORDS First-degree AV block; Heart failure; Cardiac resynchronization therapy ABBREVIATIONS AV atrioventricular; CARE-HF CArdiac REsynchronization-Heart Failure; COMPANION Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure; CRT cardiac resynchronization therapy; CRT-D cardiac resynchronization therapy with a defibrillator; CRT-P cardiac resynchronization therapy alone; HR hazard ratio; LVEF left ventricular ejection fraction; NYHA New York Heart Association; OPT optimal pharmacological therapy (Heart Rhythm 2012;9:34 39) 2012 Heart Rhythm Society. All rights reserved. Dr Olshansky works as a consultant for Boston Scientific Corporation. Dr Day, consultant for Boston Scientific Corporation, has received honoraria from St. Jude Medical. Patrick Yong and Elizabeth Galle are full-time employees of Boston Scientific Corporation. Dr Steinberg works as a consultant for Medtronic and St. Jude Medical and has received research support from Boston Scientific Corporation, Biotronik, and Medtronic and fellowship support from Boston Scientific Corporation, Medtronic, St. Jude Medical, and Biotronik. Address for reprints and correspondence: Dr Brian Olshansky, MD, FHRS, Division of Cardiology, University of Iowa Hospitals, 200 Hawkins Drive, Iowa City, IA 52242. E-mail address: brian-olshansky@uiowa.edu. Introduction The weight of evidence indicates that cardiac resynchronization therapy (CRT) can reduce mortality and the risk for heart failure hospitalizations and may improve several qualityof-life measures for patients with New York Heart Association (NYHA) functional classes III and IV congestive heart failure who have a wide QRS complex and left ventricular dysfunction. 1 3 Current guidelines support the use of CRT devices in patients with these clinical findings. 4 Recent data indicate that PR interval prolongation is independently associated with increased risk for mortality and atrial fibrillation in the general population. 5 PR interval prolongation is independently associated with increased risk for mortality, heart failure hospitalization, and the combined endpoint of heart failure hospitalization and mortality in patients with coronary artery disease. 6 In patients with congestive heart failure and a wide QRS complex, the PR interval may also have a predictive value. In the CArdiac REsynchronization-Heart Failure (CARE-HF) trial, patients 1547-5271/$ -see front matter 2012 Heart Rhythm Society. All rights reserved. doi:10.1016/j.hrthm.2011.07.038

Olshansky et al Long PR in COMPANION Trial 35 with a wide QRS complex and ventricular dysfunction were randomized to CRT or no pacing therapy. Independent predictors of adverse outcome in a multivariable analysis included prolongation of the PR interval ( 2 12.3; P.01) 7 despite small variations in this interval in this study. Why PR prolongation may be associated with adverse outcomes is uncertain. Perhaps, ventricular filling, influenced adversely by PR prolongation, can be attenuated by CRT. The aim of the present study was to determine the relationship of PR interval prolongation with the combined endpoint of heart failure hospitalization or all-cause mortality in a population of patients with cardiomyopathy and QRS complex widening who were being treated with optimal pharmacological therapy (OPT) alone or CRT with OPT. We hypothesized that the PR interval, a measure of atrioventricular (AV) synchrony, is an important and modifiable predictor of outcomes in this population. Methods This analysis used the database from the Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure (COMPANION) trial, a multicenter, randomized controlled clinical trial comparing CRT with OPT. 2 Patients with either ischemic or nonischemic cardiomyopathy were enrolled and randomized to one of three arms if they had NYHA functional class III or IV congestive heart failure, QRS duration of 120 ms or more, and a left ventricular ejection fraction (LVEF) of 0.35 or less. All patients enrolled in the COMPANION trial had baseline PR intervals of 150 ms or more in an attempt to avoid enrolling patients who may have fusion or pseudofusion with CRT pacing. The primary endpoint for this analysis is all-cause mortality or heart failure hospitalization. Patients in all randomization arms received OPT, including beta blockers and angiotensin-converting enzyme inhibitors as tolerated. In one arm, patients received only OPT; in the other two arms, patients received OPT and either cardiac resynchronization therapy alone (CRT-P) or cardiac resynchronization therapy with a defibrillator (CRT-D). Patients were randomized on a 2:2:1 basis to CRT-D, CRT-P, or OPT, respectively. For purposes of this analysis, the CRT arms were pooled since there was no measurable difference in outcome between the two CRT arms. Patients were assigned to one of two cohorts by using the intrinsic PR interval at baseline. A normal PR interval was considered less than 200 ms. The PR interval was considered prolonged if it was 200 ms or more. Device programming data were not stored in a central database. All programmed parameters were stored on disk, kept at each individual centers, and not forwarded. Statistics Demographic data that were continuous were presented as mean and standard deviation and compared with a Student s t test. Demographic data that were categorical were presented as proportions and compared with a 2 test. For the primary analysis, Cox proportional hazard models and Kaplan Meier survival curves were used to estimate the outcome as a composite of all-cause mortality or heart failure hospitalization, whichever came first. A Cox proportional hazard model was used to evaluate the interaction between baseline PR ( 200 vs 200 ms) and treatment, adjusted for CRT and baseline PR cohort. Interaction testing was also performed by using the PR interval as a continuous variable. All hazard ratios (HRs) were adjusted for significant baseline characteristics by using the Cox proportional hazard model. The following baseline covariates were entered into the stepwise stepwise model at a P value of less than.30 and remained in the model if the associated P value was less than.05: age, gender, NYHA functional class, LVEF, left bundle branch block, QRS (dichotomized at 150 ms), heart rate, systolic and diastolic blood pressure, ischemic status, baseline comorbidities (diabetes, renal disease, peripheral vascular disease, paroxysmal atrial fibrillation, hypertension), and baseline medication (beta blockers, angiotensinconverting enzyme inhibitors, angiotensin receptor blockers, digoxin, diuretics, spironolactone). All analyses were performed in a manner consistent with the intention-to-treat principle with SAS version 9 software (SAS Institute, Inc, Cary, NC). P values of less than.05 were considered statistically significant. Results All 1520 patients enrolled in the COMPANION trial were included in this analysis. There were 728 patients in the normal PR interval group, accounting for 48% of the population. There were 792 patients, or 52% of the population, in the prolonged PR interval group. Baseline characteristics of the two groups are presented in Table 1. CRT was the randomized therapy in 79% of the entire population with a normal PR interval and 81% of those with a prolonged PR interval (P ns). The mean PR interval in the normal PR interval group was 173 ms, and the mean PR interval in the prolonged PR interval group was 230 ms (P.01). The mean age in the normal PR interval group was 64 years, whereas it was 67 years in the prolonged PR interval group (P.01). Patients in the prolonged PR interval group were more likely to have an ischemic cardiomyopathy, have a wider QRS complex, and were more likely to be male (all P.01). Notably, there was no difference in the mean LVEF between the groups (23% in the normal PR interval group vs 22% in the prolonged PR interval group [P.12]). In addition, the proportion of NYHA functional class III congestive heart failure was similar between the two groups (P.14). Paroxysms of atrial fibrillation were present in 133 of 728 (18%) patients with a normal PR interval and 208 of 792 (26%) patients with a prolonged PR interval (P.01). There was no difference in beta-blocker use when it was considered as a categorical variable or by group (PR or 200 ms, OPT or CRT). With carvedilol dosing as standard, metoprolol dosing was calculated as carvedilol equiva-

36 Heart Rhythm, Vol 9, No 1, January 2012 Table 1 Baseline characteristics of participants in the COMPANION trial Baseline characteristic Normal PR (n 728) Prolonged PR (n 792) P value PR interval (ms) 175 13 230 34.01 QRS width (ms) 155 23 162 25.01 Age (y) 64 12 67 11.01 Male sex, n (%) 440 (60) 587 (74).01 Ischemic etiology, n (%) 376 (52) 462 (58).01 LVEF (%) 23 7 22 7.12 LVEDD (mm) 68 8 70 9.01 NYHA class III, n (%) 634 (87) 669 (84).14 Renal disease, n (%) 136 (19) 203 (26).01 Beta-blocker use, n (%) 505 (69) 521 (66).14 ACE/ARB use, n (%) 648 (89) 705 (89) 1.00 ACE/ARB angiotensin-converting enzyme/angiotensin receptor blocker; COMPANION Comparison of Medical Therapy, Pacing, and Defibrillation in Heart Failure; LVEDD left ventricular end diastolic dimension; LVEF left ventricular ejection fraction. lents. The median dose for all groups was 25 mg of carvedilol or its equivalent. This was unchanged at one year for OPT and CRT groups. Split into PR 200 and 200 ms the median dose for all groups at all time points (from baseline to one year) and for all randomized therapies was 25 mg/d. As a continuous variable, the PR interval did not predict outcomes. Analysis of the PR interval as a continuous variable was found to exhibit a nonlinear relationship with outcomes such that greater degrees of PR prolongation within the prolonged PR cohort were associated with disproportionally worsened outcomes. This was not the case within the normal PR cohort. Efforts to transform the PR variable did not produce a linear model. So, a clinically relevant cut point of 200 ms was selected to separate the population into normal and prolonged cohorts ( 200 and 200 ms, respectively) for analyses. By using a stepwise multivariable analysis and adjusting for significant baseline variables, the PR interval cohort was found to be one of the variables associated with the composite outcome but only for those patients randomized to OPT (Table 2). NYHA functional class, renal disease, LVEF, and systolic blood pressure were other variables associated with the endpoint for the OPT and the CRT groups. Randomization to CRT was associated with a reduction in the endpoint for patients with a normal or a prolonged PR interval, but the strength of the association was greater (HR 0.54; P.01) for those in the prolonged PR interval group than those in the normal PR interval group (HR 0.71; P.02) (Table 3). When considering the OPT group alone, a prolonged PR interval (compared with a normal PR interval) was associated significantly with a 41% increase in the risk of all-cause mortality or heart failure hospitalization (P.04). Kaplan Meier curves (Figure 1A) show the outcome by PR interval cohort. CRT was associated with a statistically significant reduction in all-cause mortality or heart failure hospitalization Table 2 Multivariate predictors of outcomes Covariate Value HR P value Patients randomized to OPT (n 308) NYHA class IV 2.06.01 Renal disease Yes 1.88.01 LVEF Per 1% decrease 1.03.01 Systolic blood pressure Per 1 mm Hg drop 1.01.02 PR interval 200 ms 1.41.04 Patients randomized to CRT (n 1212) NYHA class IV 1.89.01 Renal disease Yes 1.64.01 ACE/ARB BB No 1.46.01 QRS 150 ms 1.43.01 LVEF Per 1% decrease 1.02.01 History of AF Yes 1.41.01 Etiology Ischemic 1.32.01 Spironolactone use No 1.38.01 Systolic blood pressure Per 1 mm Hg drop 1.01.01 Peripheral vascular disease Yes 1.38.01 Age Per 1 year increase 1.01.01 PR interval 200 ms 1.14.19 Hazard ratio (HR) 1.0 indicates that the covariate is associated with worsened outcome. ACE/ARB angiotensin-converting enzyme/angiotensin receptor blocker; AF atrial fibrillation; BB beta blocker; LVEF left ventricular ejection fraction; OPT optimal pharmacological therapy.

Olshansky et al Long PR in COMPANION Trial 37 Table 3 Multivariate predictors of outcomes by PR interval Covariate Value HR P value Patients with prolonged PR (n 792) NYHA class IV 2.30.01 Renal disease Yes 1.51.01 Systolic blood pressure Per 1 mm Hg drop 1.01.01 Spironolactone use No 1.46.01 History of AF Yes 1.36.01 ACE/ARB BB No 1.33.01 Etiology Ischemic 1.31.03 Age Per 1 year increase 1.01.01 Randomized therapy CRT 0.54.01 Patients with normal PR (n 728) Renal disease Yes 1.92.01 Beta blocker use No 1.48.01 LVEF Per 1% decrease 1.05.01 QRS 150 ms 1.57.01 NYHA class IV 1.67.01 Bundle branch morphology Non-LBBB 1.47.01 Diuretic use Yes 2.58.02 Randomized therapy CRT 0.71.02 ACE/ARB angiotensin-converting enzyme/angiotensin receptor blocker; AF atrial fibrillation; BB beta blocker; HR hazard ratio; LBBB left bundle branch block; LVEF left ventricular ejection fraction; OPT optimal pharmacological therapy. compared with OPT for those with a normal PR interval and those with a prolonged PR interval (Figure 1B and 1C). The reduction in relative risk was 29% (P.02) for patients with normal PR intervals (Table 2), whereas the corresponding reduction in risk was 46% (P.01) for patients with prolonged PR intervals. Within the CRT group, the PR interval did not appear to exert any appreciable influence on outcome (Figure 1D). There was no statistically significant interaction detected for PR interval cohort and treatment (P.17). We evaluated the data based on CRT-D or CRT-P implantation. There was no specific advantage of pacing alone or pacing with backup defibrillation for patients with normal or prolonged PR intervals (P ns for both). Discussion Our results, based on data from the COMPANION trial, demonstrate that patients randomized to OPT with a prolonged PR interval fared worse than patients in the OPT arm with a normal PR interval. Importantly, patients randomized to CRT had a better outcome regardless of the baseline PR interval. Outcomes after CRT were similar in the groups with normal and prolonged PR intervals. Those with a prolonged PR interval appeared to derive the most benefit from CRT. CRT attenuated the adverse influence of PR prolongation so that outcomes in the CRT-treated patients were independent of the baseline PR interval. These data suggest that a prolonged PR interval is a modifiable marker of risk in this population. This issue becomes relevant in clinical practice as PR interval prolongation occurs commonly in patients with severe congestive heart failure. Little data highlight the frequency of PR prolongation in this population of patients with heart failure. If the database of the COMPANION trial is any indication of the frequency of prolonged PR intervals in this population, it is common; 50% of the patients had a PR interval of at least 200 ms. The PR interval, considered as a continuous variable, was not associated with outcomes, but the relationship of the PR interval and outcome was nonlinear such that only greater degrees of PR prolongation were associated with disproportionally worsened outcomes. A PR interval at or near the normal range made no difference in outcomes. The effect of PR prolongation on hemodynamics and symptoms may be hard to determine with certainty since adjustment of this interval generally requires some degree of ventricular pacing. Unopposed right ventricular pacing itself can have adverse hemodynamic effects and worsen prognosis in patients with left ventricular dysfunction 8 so that modifying the PR interval alone becomes impossible. Resynchronization pacing therapy is one way to offset the adverse consequences of unopposed right ventricular pacing, but even this is not identical to normal His-Purkinje activation. The strengths of this study relate to the hard outcome endpoints in a large randomized population. This analysis is the first evidence that PR prolongation in this population is associated with adverse outcomes and that CRT improves outcomes in both groups, but perhaps more so in those with PR prolongation. While the test for interaction was not significant, these data may be underpowered to detect an interaction in the CRT groups. There may be differences in outcomes after CRT in those with normal and prolonged PR intervals. Other data support potential benefits of some right ventricular pacing in populations of patients who have prolonged PR intervals but these are patients without NYHA functional classes III and IV heart failure and systolic left

38 Heart Rhythm, Vol 9, No 1, January 2012 Figure 1 A: Differences in outcome by PR interval in the OPT group. B: Outcome in the normal PR interval group: OPT versus CRT. C: Outcome in the prolonged PR interval group: OPT versus CRT. D: Differences in outcome by PR interval in the CRT group. CRT cardiac resynchronization therapy; HF heart failure; HR hazard ratio; OPT optimal pharmacological therapy. ventricular dysfunction. 9 12 No data define the risk of PR prolongation in a heart failure population and support the value of CRT pacing for these patients. In some ways, these data are remarkable since the patients who had prolonged PR intervals in the COMPANION trial had additional evidence for increased risk of the endpoint of heart failure hospitalization or all-cause mortality. Those patients with a prolonged PR had wider QRS complexes, were more likely to be men, had more ischemic cardiomyopathy, and were older. It is known that females, younger patients, and those with nonischemic dilated cardiomyopathy who undergo resynchronization therapy are at lower risk. Nevertheless, this high-risk group fared as well as the group with a normal PR interval. The mechanism by which CRT benefits patients with normal or prolonged PR intervals may be similar, but the exact cause for the greater improvement in outcomes in patients with a prolonged PR interval remains in question. Some acute data suggest that the AV interval is not all that important. 13 Other data support shortening of the AV interval to improve hemodynamics in patients with heart failure, 14 but the programmed value can vary significantly. 15 While it is clear that patients with a prolonged PR interval appear to be at greater risk, CRT seems to have a greater effect on this group, perhaps by correction of ventricular dyssynchrony as well as by restoring AV synchrony. The potential hemodynamic mechanisms of benefit with shortening the PR interval include restoration of left and/or right AV synchrony, which can help reduce diastolic AV valve regurgitation and improvement in left (and perhaps right) ventricular diastolic filling to affect stroke volume. 16 Perhaps the benefit of CRT was attenuated in those with normal PR intervals as they were more susceptible to fusion and pseudofusion. This possibility was minimized by including patients in the COMPANION trial with PR intervals of 150 ms or more. There are many reasons for PR prolongation. The PR interval may be prolonged because of a delay in atrial activation or AV nodal or His-Purkinje conduction. If because of prolongation in atrial activation, right and left AV synchrony may differ and, in fact, the AV interval may be too short, at least with respect to the left atrial/left ventricular mechanical activation. In this case, it is

Olshansky et al Long PR in COMPANION Trial 39 possible, and could be argued, that patients with a more prolonged PR interval could derive greater benefit since some of the benefit may be lost by a pacemaker syndrome because of shortening of the AV interval on the left side of the heart. 17 Pacemaker syndrome, of course, could occur with a very prolonged AV interval as well. In this study, we were not able to determine the optimal AV interval at different pacing rates and there was no specific mandate to determine the optimal AV interval. Furthermore, atrial pacing may change atrial activation and the AV interval, so that the AV interval settings in short and long PR intervals may differ during atrial pacing and atrial sensing. We did not necessarily determine that the same AV settings should be utilized in patients with prolonged versus normal PR intervals. The best AV interval and the best methodology to measure it remains a challenge. 18 As has been shown in a study by Gervais et al, 7 the PR interval is one of the only independent predictors of outcome in patients undergoing CRT in the CARE-HF study. In the CARE-HF study, the differences in the PR interval were less than those present in the COMPANION trial. As in the CARE-HF study, the QRS duration was not a strong predictor of the primary outcome in the COMPANION trial as the QRS was already wide. It is possible that improvement in the PR interval is one of the key mechanisms responsible for improvement in outcomes in patients undergoing CRT. Data from the Heart and Soul Study 6 emphasize the importance of the PR interval. In this study of patients with stable coronary artery disease, PR intervals of 220 ms or more were independently associated with a greater risk for heart failure hospitalization, mortality, and the combined endpoint of heart failure hospitalization or cardiovascular mortality (age-adjusted HR 2.43; P.01). Limitations This is a retrospective analysis of the COMPANION trial data. While prospective studies may be required to ascertain with certainty the importance of the PR interval, studies that evaluate the benefit of CRT in patients with PR interval prolongation (ie, comparing prospectively CRT in this group against no CRT) are problematic, difficult, and likely impossible in this specific patient population. There was no mandate regarding optimization of the PR interval, and there may be important differences in the appropriate settings of the AV interval that could have more strongly influenced outcomes in this population. The programming of the AV interval is not known. Nevertheless, recent data suggest that the optimal PR (AV) interval can be difficult to predict for those with or without PR interval prolongation. AV delay programming was formalized by the protocol but not recorded or compiled into a centralized database. The percent biventricular pacing was not reported and could not be analyzed by group. The presence and degree of fusion and pseudofusion in CRT patients with prolonged or normal PR intervals are not known. Conclusions PR prolongation, when present in patients with impaired left ventricular function and widened QRS complexes, is independently associated with an increased risk for all-cause mortality or heart failure hospitalizations. CRT appears to provide benefit regardless of the baseline PR interval but appears to provide greater benefit for those with a prolonged PR interval. References 1. Abraham WT, Fisher WG, Smith AL, et al. Cardiac resynchronization in chronic heart failure. N Engl J Med 2002;346:1845 1853. 2. Bristow MR, Saxon LA, Boehmer J, et al. Cardiac-resynchronization therapy with or without an implantable defibrillator in advanced chronic heart failure. N Engl J Med 2004;350:2140 2150. 3. Cleland JG, Daubert JC, Erdmann E, et al. The effect of cardiac resynchronization on morbidity and mortality in heart failure. N Engl J Med 2005;352: 1539 1549. 4. Epstein AE, DiMarco JP, Ellenbogen KA, et al. ACC/AHA/HRS 2008 Guidelines for Device-Based Therapy of Cardiac Rhythm Abnormalities: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the ACC/AHA/NASPE 2002 Guideline Update for Implantation of Cardiac Pacemakers and Antiarrhythmia Devices): developed in collaboration with the American Association for Thoracic Surgery and Society of Thoracic Surgeons. Circulation 2008;117:e350 e408. 5. Cheng S, Keyes MJ, Larson MG, et al. Long-term outcomes in individuals with prolonged PR interval or first-degree atrioventricular block. JAMA 2009;301: 2571 2577. 6. Crisel RK, Farzaneh-Far R, Na B, Whooley MA. First-degree atrioventricular block is associated with heart failure and death in persons with stable coronary artery disease: data from the Heart and Soul Study. Eur Heart J 2011;32:1875 1880. 7. Gervais R, Leclercq C, Shankar A, et al. Surface electrocardiogram to predict outcome in candidates for cardiac resynchronization therapy: a sub-analysis of the CARE-HF trial. Eur J Heart Fail 2009;11:699 705. 8. Steinberg JS, Fischer A, Wang P, et al. The clinical implications of cumulative right ventricular pacing in the multicenter automatic defibrillator trial II. J Cardiovasc Electrophysiol 2005;16:359 365. 9. Barold SS, Ilercil A, Leonelli F, Herweg B. First-degree atrioventricular block. Clinical manifestations, indications for pacing, pacemaker management & consequences during cardiac resynchronization. J Interv Card Electrophysiol 2006; 17:139 152. 10. Barold SS. Optimal pacing in first-degree AV block. Pacing Clin Electrophysiol 1999;22:1423 1424. 11. Capucci A, Romano S, Puglisi A, et al. Dual chamber pacing with optimal AV delay in congestive heart failure: a randomized study. Europace 1999;1:174 178. 12. Olshansky B, Day JD, Lerew DR, Brown S, Stolen KQ. Eliminating right ventricular pacing may not be best for patients requiring implantable cardioverter-defibrillators. Heart Rhythm 2007;4:886 891. 13. Gold MR, Feliciano Z, Gottlieb SS, Fisher ML. Dual-chamber pacing with a short atrioventricular delay in congestive heart failure: a randomized study. J Am Coll Cardiol 1995;26:967 973. 14. Auricchio A, Sommariva L, Salo RW, Scafuri A, Chiariello L. Improvement of cardiac function in patients with severe congestive heart failure and coronary artery disease by dual chamber pacing with shortened AV delay. Pacing Clin Electrophysiol 1993;16:2034 2043. 15. Auricchio A, Stellbrink C, Block M, et al. Effect of pacing chamber and atrioventricular delay on acute systolic function of paced patients with congestive heart failure. The Pacing Therapies for Congestive Heart Failure Study Group. The Guidant Congestive Heart Failure Research Group. Circulation 1999;99:2993 3001. 16. Maloney J, Khoury D, Simmons T, et al. Effect of atrioventricular synchrony on stroke volume during ventricular tachycardia in man. Am Heart J 1992;123: 1561 1568. 17. Auricchio A, Ding J, Spinelli JC, et al. Cardiac resynchronization therapy restores optimal atrioventricular mechanical timing in heart failure patients with ventricular conduction delay. J Am Coll Cardiol 2002;39:1163 1169. 18. Valeur N, Fritz-Hansen T, Risum N, Mogelvang R, Bloch Thomsen PE, Sogaard P. Echocardiographic effects of changing atrioventricular delay in cardiac resynchronization therapy based on displacement. J Am Soc Echocardiogr 2010; 23:621 627.