JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2020 ª 2020 PUBLISHED BY ELSEVIER ON BEHALF OF THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION Long-Term Outcomes in Patients With a Left Ejection Fraction #15% Undergoing Cardiac Resynchronization Therapy John Rickard, MD, MPH,a Divyang Patel, MD,a Carolyn Park, MD,b Joseph E. Marine, MD,c Sunil Sinha, MD,c W.H. Wilson Tang, MD,c Niraj Varma, MD, PHD,a Bruce L. Wilkoff, MD,c David Spragg, MDc ABSTRACT OBJECTIVES This study aimed to determine the long-term outcomes and predictors of left ventricular (LV) ejection fraction (LVEF) improvement in patients with severe cardiomyopathies undergoing cardiac resynchronization therapy (CRT). BACKGROUND Whether patients with severe LV dysfunction benefit from CRT or have reached a point in disease severity past the point at which CRT is beneficial is unknown. METHODS We collected clinical and echocardiographic data on 420 patients with an LVEF of #15% and a QRS duration of $120 ms undergoing CRT at the Cleveland Clinic and 2 hospitals in the Johns Hopkins Health System between April 2003 and May 2014. Multivariate models were created to determine factors associated with response to CRT, defined as an absolute improvement in LVEF of >5% and survival free of LVAD and heart transplant. Procedure-related deaths were also collected. RESULTS A total of 298 patients had pre- and appropriately timed post-CRT echocardiograms, of whom 145 (48.7%) met the criteria for response. In multivariate analysis, LV size and left bundle branch block (LBBB) were associated with response. Among the most dilated quintile (LV end-diastolic diameter [LVEDD] of >7.8 cm), 30.4% met the criteria for response. In multivariate analysis, smaller LV end-diastolic dysfunction and presence of LBBB were associated with improved survival free of heart failure and LVAD over a mean follow-up period of 5.2 years. There were no procedurerelated deaths. CONCLUSIONS Patients with severe LV dysfunction respond to CRT, although at a lower rate compared to traditional CRT candidates. Smaller LV size and LBBB are important predictors of positive outcomes in this population. Even among the most dilated patients, 30.4% realized a meaningful improvement in LVEF with CRT. The CRT implant procedure itself appears well tolerated. (J Am Coll Cardiol EP 2020;-:-–-) © 2020 Published by Elsevier on behalf of the American College of Cardiology Foundation. C ardiac resynchronization therapy (CRT) rep- In clinical practice, however, patients with signifi- resents one of the most important advances cantly more advanced LV dysfunction are occasion- in the treatment of patients with electrical ally encountered. The outcomes of such patients dyssynchrony and significant left ventricular (LV) with CRT and the safety of the procedure itself have dysfunction over the past 20 years (1–4). In clinical not been reported. In clinical practice, when a CRT trials of CRT, the mean LV ejection fraction (LVEF) candidate with an LVEF of #15% presents, the ques- before implant was approximately 21% to 24% (1–4). tion often arises of whether a meaningful response From the aDivision of Cardiology, Heart and Vascular Institute Cleveland Clinic, Cleveland, Ohio; bJohns Hopkins Bayview Medical Center, Baltimore, Maryland; and the cDivision of Cardiology/Electrophysiology, Johns Hopkins University School of Medicine, Baltimore, Maryland. The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Clinical Electrophysiology author instructions page. Manuscript received March 30, 2020; revised manuscript received July 27, 2020, accepted July 27, 2020. ISSN 2405-500X/$36.00 https://doi.org/10.1016/j.jacep.2020.07.025 2 Rickard et al. JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2020 - 2020:-–- Severe Cardiomyopathy and CRT ABBREVIATIONS to CRT can be expected given the severity of implantation of a CRT device at the Cleveland Clinic AND ACRONYMS the cardiomyopathy. In addition, it is unclear (Cleveland, Ohio) and the Johns Hopkins Hospital and whether improvement in LVEF correlates Johns Hopkins Bayview Medical Center (both in Bal- with improved outcomes in this population. timore, Maryland) between April 2003 and May 2014. The question of whether to proceed with The study was approved by the Institutional Review resynchronization therapy CRT becomes even more of a challenge Boards of the Cleveland Clinic and John Hopkins ICD = implantable cardioverter when the patient already has an implantable Medicine for retrospective medical records review defibrillator cardioverter-defibrillator (ICD) and presents and was performed according to institutional guide- LBBB = left bundle branch for consideration for a cardiac implantable lines. Clinical, electrocardiographic, and echocardio- block electrical device (CIED) revision to CRT. As graphic data were gathered via review of electronic LV = left ventricular such, the chances of improvement must be and physical medical records. For inclusion in the LVAD = left ventricular assist weighed against the potential procedural final cohort, all patients had an LVEF of #15% and a device risks, which, given the level of LV dysfunc- QRS duration of $120 ms. An assessment of all-cause LVEDD = left ventricular end- tion, may be significant. In the current study, mortality was made using the U.S. Social Security we sought to characterize procedural sur- Death Index, Ohio death records, chart review, and vival, long-term outcomes, and reverse ven- Internet obituary searches. The subsequent implan- tricular remodeling in patients undergoing tation of a left ventricular assist device (LVAD) or CIED = cardiac implantable electrical device CRT = cardiac diastolic diameter LVEF = left ventricular ejection fraction CRT with extremely severe LV dysfunction. heart transplant was assessed by using current chart review at both the Cleveland Clinic and Johns Hop- METHODS kins Medicine as well as a search of Cleveland Clinic advanced heart failure therapy registry data. This retrospective study involved the analysis of a cohort of patients who underwent the new Kaplan-Meier curves were constructed, and a multivariate model was created to determine T A B L E 1 Baseline Characteristics of the Total Cohort* Total (N ¼ 420) Ischemic Cardiomyopathy (n ¼ 211) Nonischemic Cardiomyopathy (n ¼ 209) Age, yrs 63.7 12.3 67.5 10.1 59.8 13.1 0.001 Men 306 (72.9) 172 (81.5) 134 (64.1) <0.001 Baseline left ventricular ejection fraction, % 13.2 2.4 13.4 2.3 13.0 2.5 0.07 Baseline left ventricular end-diastolic diameter, cm 6.8 1.0 6.6 0.8 6.9 1.2 0.003 <0.001 p Value 6.0 1.5 5.7 1.0 6.3 1.8 CRT-D device, % 409 (97.4) 208 (98.6) 201 (96.2) 0.14 Serum hemoglobin, g/dl 12.7 1.9 12.4 1.9 13.1 1.8 <0.001 <0.001 Baseline left ventricular end-systolic diameter, cm 1.4 0.9 1.6 1.2 1.1 0.5 163.7 26.3 162.4 24.3 165.0 28.1 0.28 Left bundle branch block 222 (52.9) 88 (41.7) 134 (64.1) <0.001 Paced upgrade 60 (14.3) 32 (15.2) 28 (13.4) 0.68 31 (7.4) 23 (10.9) 8 (3.8) 0.008 0.002 Serum creatinine, mg/dl QRS duration, ms Right bundle branch block Nonspecific intraventricular conduction delay 106 (25.2) 67 (31.8) 39 (18.6) History of atrial fibrillation, any type 202 (48.1) 107 (50.7) 95 (45.5) 0.29 58 (13.8) 37 (17.5) 21 (10.0) 0.03 Chronic obstructive pulmonary disease Hypertension 257 (61.2) 131 (62.1) 126 (60.3) 0.76 Hyperlipidemia 234 (55.7) 148 (70.1) 86 (41.1) <0.001 History of malignancy 44 (10.5) 17 (8.1) 27 (12.9) 0.10 Diabetes mellitus 163 (38.8) 92 (43.6) 71 (34.0) 0.05 0.47 History of stroke or transient ischemic attack 56 (13.3) 31 (14.7) 25 (12.0) Beta-adrenergic blocker 339 (83.9) 162 (81.0) 177 (86.8) 0.14 ACE inhibitor or angiotensin II receptor blocker 320 (79.2) 144 (72.0) 176 (86.3) <0.001 Aldosterone antagonist 169 (41.8) 82 (41.0) 87 (42.6) 0.76 Diuretic 351 (86.9) 172 (86.0) 179 (87.7) 0.66 87 (21.5) 53 (26.5) 34 (16.7) 0.02 Anti-arrhythmic medications Values are mean SD or n (%). *Medications were available in 404 patients: 204 with nonischemic cardiomyopathy and 200 with ischemic cardiomyopathy. ACE ¼ angiotensin-converting enzyme; CRT-D ¼ cardiac resynchronization therapy defibrillator. JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2020 Rickard et al. - 2020:-–- 3 Severe Cardiomyopathy and CRT predictors of survival free of LVAD and heart transplant. In patients in whom an appropriately timed T A B L E 2 Multivariate Cox Proportional Hazards Model for Survival Free of LVAD or Heart Transplant follow-up echocardiogram was available, an assessHazard Ratio (95% Confidence Interval) ment of reverse remodeling was made. The echocar- p Value diogram closest to 9 months and at least >3 months Baseline left ventricular end-diastolic diameter 1.19 (1.03–1.36) 0.01 post-procedure was chosen. Response was defined as Age 1.02 (1.00–1.03) <0.001 an absolute improvement in LVEF of >5%. Variables Male 1.44 (1.06–2.00) 0.02 selected for the models were selected based on a History of atrial fibrillation 1.15 (0.89–1.48) 0.277 priori data obtained from review of past studies. In Ischemic cardiomyopathy 1.45 (1.11–1.90) 0.07 the cohort as a whole, CRT device implantations were Serum creatinine 1.06 (0.96–1.17) 0.25 performed transvenously in the vast majority of patients by electrophysiologists targeting a lateral or posterolateral vein for the LV lead position. Medica- QRS duration 0.996 (0.991–1.001) 0.63 (04.7–0.85) Paced upgrade 0.88 (0.59–1.30) 0.51 NYHA functional class 2.10 (1.51–2.86) <0.001 tions were recorded immediately before implantation of the CRT device, with subsequent titration of LVAD ¼ left ventricular assist device; NYHA ¼ New York Heart Association. medications made at the discretion of patients’ outpatient physicians. STATISTICAL ANALYSIS. Continuous variables were was 13.2 2.4%, and patients were more commonly presented as a mean SD and dichotomous variables men (72.1%) with left bundle branch block (LBBB) as an absolute number with percentage. Comparisons (53.1%). Over a mean follow-up period of 5.2 years, between 2 continuous variables were made using there were 291 endpoints (17 LVADs, 20 heart trans- Student’s t-test for parametric variables and a Mann- plants, and 254 deaths). In multivariate analysis, Whitney test for nonparametric variables. Compari- younger age, female sex, smaller baseline LV end- sons among more than 2 continuous variables were diastolic diameter (LVEDD), and presence of LBBB made using an analysis of variance test for parametric were independently associated with improved sur- variables and the Kruskal-Wallis test for nonpara- vival free of LVAD and heart transplant (Table 2, metric were Central Illustration). There was a nonsignificant trend compared using the chi-square test. Kaplan-Meier toward improved survival in patients with non- curves using the log-rank test were created to assess ischemic cardiomyopathy compared with ischemic survival free of LVAD or heart transplant over the cardiomyopathy (Table 2, Figure 1). variables. Dichotomous variables duration of follow-up. A multivariate Cox propor- A total of 298 patients had an available pre-CRT tional hazards model was created to determine fac- and appropriately timed follow-up echocardiogram, tors associated with long-term survival free of LVAD of whom 145 met the criteria for response (48.7%) and heart transplant, controlling for multiple vari- (Table 3). Responders tended to have a smaller LVEDD ables selected based on a priori knowledge. To test and more often have LBBB (Table 3). There were 66 the Cox assumption that the hazard ratio between patients who were revised from ICD to CRT-D, of individuals is constant, a time-varying covariate was whom 30 met the criteria for response (45.5%; entered into the model for each variable, with a p p ¼ 0.58). In multivariate logistic regression, a smaller value of >0.05 needed to satisfy this assumption. In LV size was associated with improved response those patients in whom an appropriately timed (Table 4). There was a nonsignificant trend toward follow-up echocardiogram was available, a multivar- presence of LBBB and improved response to CRT. iate logistic regression analysis was performed to Patients whose LVEF improved demonstrated a sig- determine factors associated with response. All ana- nificant survival advantage over those whose LVEF lyses were performed with SPSS Inc. (Chicago, Illi- failed to improve (Figure 2). When stratified by nois). A procedure-related death was defined as a LVEDD quintile, patients in the highest stratum (7.8 death before discharge from the hospital. to 11 cm) had a 30.4% chance of response, which was lower than the other strata (Table 5), When response RESULTS 0.14 Left bundle branch block was defined by an LVEF improvement of $10%, patients in the highest stratum (7.8 to 11 cm) had a 19.6% Between April 2003 and May 2014, 420 patients met chance of response, which was lower than the other inclusion criteria. The baseline characteristics of the strata (Table 6). There were no procedure-related cohort are shown in Table 1. The mean baseline LVEF deaths. There were 122 patients in whom complete 0.002 4 Rickard et al. JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2020 - 2020:-–- Severe Cardiomyopathy and CRT C E N T R A L IL L U ST R A T I O N Kaplan-Meier Curves for Survival Free of LV Assist Device or Heart Transplant Based on Baseline LV End-Diastolic Diameter Tertile Rickard, J. et al. J Am Coll Cardiol EP. 2020;-(-):-–-. Blue indicates LV end-diastolic diameter (LVEDD) of $7.2 cm, green indicates LV end-diastolic diameter of 6.4 to 7.2 cm, and tan indicates LV end-diastolic diameter of <6.4 cm. Log-rank p < 0.01. LV ¼ left ventricular. echocardiographic unavailable Of the 122 patients, 32 (24.2%) died within 1 year of (Table 7). Such patients had a higher baseline serum follow-up was CRT implant. Baseline characteristics based on QRS creatinine (1.6 1.3 mg/dL vs. 1.3 0.7 mg/dL; morphology are characterized in Table 8. Patients p < 0.001), more ischemic cardiomyopathy (68.0% vs. with LBBB had improved survival compared to the 42.6%; p < 0.001), and more non-LBBB than those other QRS morphologies (Figure 3). Response was with complete echocardiographic follow-up (Table 7). noted in 54.3% of patients with LBBB, 20.0% with JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2020 Rickard et al. - 2020:-–- Severe Cardiomyopathy and CRT F I G U R E 1 Kaplan-Meier Curves for Survival Free of LVAD or Heart Transplant Based on Cardiomyopathy Subtype Blue indicates nonischemic cardiomyopathy, and green indicates ischemic cardiomyopathy. Log-rank p < 0.001. LVAD ¼ left ventricular assist device. right bundle branch block, 46.8% with a paced those who did not. Finally, the procedure itself morphology, and 45.1% with a nonspecific intraven- appears safe, with no procedural deaths. tricular conduction delay. Patients with a severely depressed LVEF who present for CRT implant present a treatment challenge. DISCUSSION Current implant guidelines offer no guidance as to whether CRT is futile or potentially harmful in pa- Patients with an extremely reduced LVEF (defined tients with an extremely poor ejection fraction (5–7). here as an LVEF of #15%) have been underrepre- First, it is unclear how much benefit such patients sented in clinical trials of CRT. Therefore, outcomes derive from CRT once their LV function has dropped and safety of CRT implantation in such patients are to such an extreme level. The mean LVEF in clinical unknown. The current study has multiple note- trials has been in the range of 20% to 30% (1–4). In the worthy findings. First, 48.9% of patients with se- MADIT-CRT (Cardiac-Resynchronization Therapy for vere significant the Prevention of Heart-Failure Events) trial, the improvement in LVEF with CRT, a number some- baseline LVEF was 24%; in the RAFT (Cardiac- what lower than that previously reported in pa- Resynchronization tients with better baseline LV function. Second, LV Heart Failure) trial, it was 23%; in the COMPANION size is an important predictor of both ventricular (Cardiac-Resynchronization Therapy With or Without remodeling and long-term outcomes in this patient an Implantable Defibrillator in Advanced Chronic population. Nevertheless, 30% of patients in the Heart Failure) trial, it was 21%; and in the MIRACLE highest quintile of LV dilatation still realized an (Cardiac Resynchronization in Chronic Heart failure echocardiographic benefit. Third, patients whose trial) trial, it was 22% (1–4). In subgroup analysis from LVEF improved had better long-term outcomes than MADIT-CRT, patients with an LVEF of #25% were LV dysfunction realized a Therapy for Mild-to-Moderate 5 6 Rickard et al. JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2020 - 2020:-–- Severe Cardiomyopathy and CRT T A B L E 3 Baseline Characteristics Based on Response* Total (n ¼ 298) Responders (n ¼ 145) Nonresponders (n ¼ 152) Age, yrs 63.2 12.9 63.4 12.1 63.1 13.6 0.87 Men 209 (70.1) 97 (66.9) 113 (74.3) 0.20 Baseline left ventricular ejection fraction, % 13.1 2.4 13.1 2.4 13.1 2.4 0.95 Baseline left ventricular end-diastolic diameter, cm 6.9 1.1 6.7 1.0 7.1 1.1 0.001 0.001 p Value Baseline left ventricular end systolic diameter, cm 6.0 1.2 5.8 1.2 6.2 1.2 Upgrade from ICD to CRT-D 66 (22.1) 30 (23.4) 36 (23.7) 0.58 Ischemic cardiomyopathy 127 (42.6) 59 (40.7) 68 (44.7) 0.49 Serum hemoglobin, g/dl 12.9 1.8 12.8 1.7 12.9 1.9 0.60 Serum creatinine, mg/dl 1.3 0.7 1.3 0.7 1.4 0.7 0.40 QRS duration, ms 165.3 26.7 165.6 25.1 165.0 28.3 0.86 QRS post, ms 160.6 23.6 158.8 23.1 162.4 24.0 0.19 QRS change, ms 4.6 27.6 6.7 26.8 2.6 28.3 0.20 Left bundle branch block 0.048 164 (42.6) 89 (61.4) 75 (49.3) Paced rhythm 47 (15.8) 22 (15.2) 25 (16.4) 0.75 Right bundle branch block 15 (5.0) 3 (2.1) 12 (7.9) 0.03 0.50 Nonspecific intraventricular conduction delay 71 (23.8) 32 (22.1) 39 (25.7) History of atrial fibrillation (any type) 151 (50.7) 80 (55.2) 71 (46.7) 0.17 CRT-D (vs CRT-P) 290 (97.3) 144 (99.3) 146 (96.0) 0.28 II 23 (7.7) 13 (9.0) 10 (6.6) III 124 (81.6) NYHA functional class 0.66 245 (82.2) 121 (83.4) III-IV 11 (3.7) 4 (2.8) 7 (4.6) IV 19 (6.4) 8 (5.5) 11 (7.2) Chronic obstructive pulmonary disease 43 (14.4) 17 (11.7) 26 (17.1) 0.19 Hypertension 188 (63.1) 89 (61.4) 90 (59.2) 0.19 Hyperlipidemia 167 (56.0) 76 (52.4) 91 (60.0) 0.20 35 (11.7) 14 (9.7) 21 (13.8) 0.28 116 (38.9) 65 (44.8) 51 (33.6) 0.06 0.50 History of malignancy Diabetes mellitus History of stroke or transient ischemic attack 39 (13.1) 17 (11.7) 22 (15.2) Beta-adrenergic blocker 245 (85.1) 122 (87.1) 123 (83.1) 0.41 ACE inhibitor or angiotensin II receptor blocker 236 (81.9) 119 (85.0) 117 (79.1) 0.20 Diuretic 253 (87.8) 126 (90.0) 127 (85.8) 0.29 Antiarrhythmic medications 58 (20.1) 29 (20.7) 29 (19.6) 0.90 Nitrates 66 (22.9) 32 (22.9) 34 (23.0) 1.00 Hydralazine 37 (12.8) 16 (11.4) 21 (14.2) 0.60 Values are mean SD or n (%). *Medications were available in 288 patients: 140 responders and 148 nonresponders. CRT-P ¼ cardiac resynchronization therapy pacemaker; ICD ¼ implantable cardioverter defibrillator; other abbreviations as in Tables 1 and 2. noted T A B L E 4 Multivariate Logistic Regression Model for Response to have improved outcomes with CRT compared to ICD alone (8). This benefit, however, was less than patients with a higher LVEF (8). The number Odds Ratio (95% Confidence Interval p Value Left ventricular end-diastolic diameter 0.68 (0.53–0.97) 0.002 Male 0.95 (0.55–1.65) 0.85 of patients in the MADIT-CRT trial with an LVEF of #15% was very small (8). A question that has often been raised is whether History of atrial fibrillation 1.34 (0.82–2.90) 0.24 there is a degree of LV dysfunction that is so severe Ischemic cardiomyopathy 0.88 (0.53–1.46) 0.88 that the cardiomyopathic process has surpassed the QRS duration 1.001 (0.99–1.01) 0.80 ability of CRT to demonstrate a meaningful benefit 1.70 (0.92–3.11) 0.09 (9). Although improvement in LVEF has been shown Paced rhythm 1.14 (0.477–2.88) 0.77 NYHA functional class 0.86 (0.45–1.64) 0.86 Left bundle branch block NYHA ¼ New York Heart Association. to translate into a survival benefit in a traditional CRT population, whether such improvement translates into a survival benefit in patients with extreme LV dysfunction is unclear (10–12). Finally, the safety of JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2020 Rickard et al. - 2020:-–- Severe Cardiomyopathy and CRT F I G U R E 2 Kaplan-Meier Curves for Survival Free of LVAD or Heart Transplant Based on Change in LVEF Yellow indicates LVEF of >10%, green indicates LVEF of 5% to 10%, and blue indicates LVEF of <5%. Log-rank p < 0.001. LVAD ¼ left ventricular assist device; LVEF ¼ left ventricular ejection fraction. the procedure itself has been called into question. tion or general anesthesia in patients with extremely CRT has been shown to be a safe procedure in more poor LV function adds to the considerations of traditional candidates (13). In some patients, CRT whether CRT is appropriate. When patients already implantation can be difficult and time consuming, have an ICD and the indication is a revision to CRT-D, especially in patients with LV dilatation and severe the decision is even more challenging. We found that LV dysfunction, in whom coronary venous anatomy patients with an LVEF of #15% had a 48.9% chance of can be distorted. Coupling a more challenging improvement in LVEF of at least >5% and that such implant with the risks of prolonged conscious seda- improvement translated into a long-term survival T A B L E 5 Percent Responders Based on LVEDD Quintile T A B L E 6 Improvement in Left Ventricular Ejection Fraction $10% Based on LVEDD Quintile LVEDD (cm) by Quintile Response 11–7.8 17/56 (30.4) LVEDD (cm) by Quintile 7.7–7.1 27/56 (48.2) 11–7.8 7.0–6.6 28/59 (47.5) 7.7–7.1 23/56 (41.1) 6.5–6.1 34/52 (65.4) 7.0–6.6 27/59 (45.8) #6.0 35/63 (55.6) 6.5–6.1 31/52 (59.6) #6.0 30/63 (47.6) Values are n/N (%). LVEDD ¼ left ventricular end-diastolic diameter. Values are n/N (%). LVEDD ¼ left ventricular end-diastolic diameter. Response 11/56 (19.6) 7 8 Rickard et al. JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2020 - 2020:-–- Severe Cardiomyopathy and CRT T A B L E 7 Baseline Characteristics of the Total Cohort* No Follow-Up Echocardiogram (n ¼ 122) Follow-Up Echocardiogram (n ¼ 298) 64.7 10.8 63.2 12.9 0.25 Men 88 (72.1) 210 (70.5) 0.09 Baseline left ventricular ejection fraction 13.4 2.3 13.1 2.4 0.59 Baseline left ventricular end-diastolic diameter, cm 6.6 1.0 6.9 1.1 0.29 Ischemic cardiomyopathy 83 (68.0) 127 (42.6) <0.001 CRT-D device 119 (97.5) 290 (97.3) >0.99 Serum hemoglobin, g/dl 12.3 2.0 12.9 1.8 0.22 Serum creatinine, mg/dl 1.57 1.3 1.34 0.7 <0.001 0.07 Age, yrs p Value 160.3 24.9 165.3 26.7 Left bundle branch block 58 (47.5) 164 (55.0) 0.16 Paced upgrade 13 (10.7) 47 (15.8) 0.22 Non–left bundle branch block 49 (40.2) 84 (28.2) 0.02 History of atrial fibrillation, any type 51 (41.9) 151 (50.7) 0.11 Chronic obstructive pulmonary disease 15 (12.3) 43 (14.4) 0.64 0.23 QRS duration, ms Hypertension 69 (56.6) 188 (63.1) Hyperlipidemia 67 (54.9) 167 (56.0) 0.91 9 (7.4) 35 (11.7) 0.22 History of malignancy Diabetes mellitus 48 (39.3) 16 (5.4) >0.99 History of stroke or transient ischemic attack 17 (13.9) 39 (13.1) 0.88 Beta-adrenergic blocker 95 (77.9) 245 (85.1) 0.37 ACE inhibitor or angiotensin II receptor blocker 85 (73.3) 236 (81.9) 0.06 Aldosterone antagonist 53 (45.7) 117 (40.6) 0.37 Diuretic 98 (84.5) 253 (87.8) 0.42 Antiarrhythmic medications 28 (24.1) 58 (20.1) 0.42 Values are mean SD or n (%). *Medications were available in 404 patients: 116 with no follow-up echocardiogram and 288 with follow-up echocardiogram. Abbreviations as in Table 1. benefit. In addition, no procedure-related mortalities extreme LV dilatation (LVEDD of 7.8 to 11 cm) had a were noted. Patients revised from a previous ICD to 30.4% chance of demonstrating significant LVEF CRT-D had similar rates of response. These data improvement. This suggests that although the chan- suggest that CRT in patients with extremely severe ces of benefit are lower compared to a traditional CRT LV dysfunction is safe and that these patients have a population, benefit is still possible (15). This refutes reasonable chance to improve cardiac function and the notion that there may be a point at which LV survival free of LVAD and heart transplant. In addi- function has become so poor and LV diameter so tion, we found that many of the same predictors of dilated that the disease process has moved past the outcomes seen in a more traditional CRT population point of at least partial mitigation with CRT. (e.g., cardiomyopathy morphology) were subtype, similarly gender, predictive in QRS this population. STUDY LIMITATIONS. First, this is a cohort from 3 large tertiary care referral centers and, therefore, may Patients with extreme LV dysfunction typically not be indicative of other cohorts. The retrospective have significant LV dilatation, which was indeed nature cannot account for all confounders, despite noted in the current study. In a more traditional pa- our best effort to do so, although the large number of tient population undergoing CRT, LV size has been patients reported in this study allowed us to control shown to correlate with long-term outcomes (9). for many potential confounders. Echocardiographic Some investigators have suggested normalizing the information was obtained via chart review, and the QRS duration to LV dimension as a means of pre- imaging studies could not be readjudicated. However, dicting outcomes (14). We found that among patients both the Cleveland Clinic and John Hopkins Medicine with severe LV dysfunction, even patients with have served as echocardiographic core labs, with JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2020 Rickard et al. - 2020:-–- Severe Cardiomyopathy and CRT T A B L E 8 Baseline Characteristics Based on Bundle Branch Block Morphology* Paced (n ¼ 67) Right Bundle Branch Block (n ¼ 32) Nonspecific Intraventricular Conduction Delay (n ¼ 107) p Value 64.6 12.5 66.4 12.2 64.4 12.7 59.9 11.0 0.002 Men 139 (65.0) 52 (77.6) 29 (90.6) 86 (80.4) 0.001 Ischemic cardiomyopathy, % 86 (40.2) 33 (49.3) 24 (75.0) 68 (63.6) <0.001 Baseline left ventricular ejection fraction, % 13.1 2.4 13.3 2.3 13.8 1.9 13.0 2.5 0.32 Baseline left ventricular end diastolic diameter, cm 6.7 1.0 6.9 1.1 6.3 0.9 7.0 1.0 0.01 Age, yrs Left Bundle Branch Block (n ¼ 214) Baseline left ventricular end systolic diameter, cm 6.0 1.7 5.9 1.3 5.6 0.9 6.1 1.2 0.36 CRT-D device 209 (97.7) 64 (95.5) 32 (100.0) 104 (97.2) 0.60 Serum hemoglobin, g/dl 12.7 1.9 12.8 1.7 12.2 2.0 12.8 2.1 0.51 Serum creatinine, mg/dl 1.4 1.0 1.6 1.0 1.4 0.8 1.4 0.7 0.42 <0.001 164.8 21.1 189.7 28.6 162.6 14.7 146.1 22.7 History of atrial fibrillation, any type 97 (45.3) 39 (58.2) 16 (50.0) 50 (46.7) 0.32 Chronic obstructive pulmonary disease 30 (14.0) 9 (13.4) 3 (9.4) 16 (15.0) 0.88 Hypertension 137 (64.0) 42 (62.7) 18 (56.3) 60 (56.1) 0.51 Hyperlipidemia 113 (52.8) 36 (53.7) 24 (75.0) 61 (57.0) 0.13 History of malignancy 21 (9.9) 10 (14.9) 5 (15.6) 8 (7.5) 0.33 Diabetes mellitus 81 (37.9) 30 (44.8) 10 (31.3) 42 (39.3) 0.60 History of stroke or transient ischemic attack 22 (10.3) 11 (16.4) 3 (9.4) 20 (18.7) 0.15 Beta-adrenergic blocker 178 (86.4) 54 (85.7) 22 (71.0) 85 (81.7) 0.15 ACE inhibitor or angiotensin II receptor blocker 168 (81.6) 52 (82.5) 21 (67.7) 79 (76.0) 0.24 Aldosterone antagonist 81 (39.3) 28 (44.4) 15 (48.4) 45 (43.3) 0.72 Diuretic 177 (85.9) 55 (87.3) 25 (80.6) 94 (90.4) 0.50 32 (15.5) 17 (5.9) 3 (9.7) 35 (33.7) 0.001 QRS duration, ms Antiarrhythmic medications Values are mean SD or n (%). *Medications were available in 404 patients: 206 with left bundle branch block, 63 paced, 31 with right bundle branch block, and 104 with nonspecific intraventricular conduction delay. Abbreviations as in Table 1. standard methods of LVEF and LVEDD reporting. severely dysfunctional that mitigation with CRT is Patients echocardiogram completely precluded. Improvement in LVEF translates appeared to have more comorbidities than those with into improved survival in this population. LV size and complete Overall, bundlebranchblockmorphologyareimportantfactorsin 24.2% of patients without an echocardiographic predicting outcomes, and the CRT implant procedure it- follow-up met the primary endpoint within 1 year. As selfinthesepatientsappearswelltolerated. without a follow-up echocardiographic follow-up. such, some selection bias may be present. There was no non-CRT control arm; therefore, a direct compar- AUTHOR RELATIONSHIP WITH INDUSTRY ison with or without CRT could not be made. Finally, Dr. Rickard has performed consulting for Medtronic; and has per- the number of patients who underwent defibrillation formed research for Abbott. Dr. Varma has performed research for threshold testing or who had other types of proce- St. Jude. Dr. Wilkoff is a speaker for Boston Scientific and Con- dural complications was not collected. vaTec; consultant to Phillips, Medtronic, and Abbott; and provides research support to Phillips. All other authors have reported that they have no relationships relevant to the contents of this paper to CONCLUSIONS disclose. Patients with severe LV dysfunction derive benefit from ADDRESS FOR CORRESPONDENCE: Dr. John Rick- CRT, although at lower rates than traditionally studied ard, Heart and Vascular Clinic, Cleveland Clinic, 9500 patients with higher LVEFs. These data argue against a Euclid specific point at which cardiac function becomes so rickarj2@ccf.org. Avenue, Cleveland, Ohio 44195. E-mail: 9 10 Rickard et al. JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2020 - 2020:-–- Severe Cardiomyopathy and CRT F I G U R E 3 Kaplan-Meier Curves for Survival Free of LVAD or Heart Transplant Based on Baseline QRS Morphology Blue indicates left bundle branch block, green indicates paced, yellow indicates right bundle branch block, and purple indicates nonspecific intraventricular conduction delay. Log-rank p < 0.001. LVAD ¼ left ventricular assist device. PERSPECTIVES COMPETENCY IN MEDICAL KNOWLEDGE: CRT re- gauge the presence of electrical dyssynchrony, the pri- mains an important tool for many patients with systolic mary target for CRT. Better tools are needed to measure heart failure and evidence of electrical dyssynchrony. electrical dyssynchrony noninvasively to gauge CRT can- Occasionally in clinical practice, patients with extreme LV didacy and CRT effect. Such tools are especially pertinent dysfunction despite good medical therapy are encoun- in patients with severe LV dysfunction who often have a tered, and the utility and safety of CRT are questioned. concomitantly elevated LV mass. Determining how much The results of this study suggest that CRT is a reasonable QRS widening is due to true electrical dyssynchrony and well-tolerated procedure in patients with severe LV versus slow conduction due to a combination of scar and functional impairment. mass effect will be useful to guide therapy. TRANSLATIONAL OUTLOOK: CRT candidacy continues to hinge on the use of 12-lead echocardiography to JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2020 - 2020:-–- Rickard et al. Severe Cardiomyopathy and CRT REFERENCES 1. Moss AJ, Hall WJ, Cannom DS, et al. Cardiacresynchronization therapy for the prevention of heart-failure events. N Engl J Med 2009;361: 1329–38. of the European Society of Cardiology (ESC). Developed in collaboration with the European Heart Rhythm Association (EHRA). Eur Heart J 2013;29:2281–329. 2. Tang AS, Wells GA, Talajic M, et al. Cardiac resynchronization therapy for mild to moderate 7. Exner DV, Birnie DH, Moe G, et al. Canadian heart failure. N Engl J Med 2010;36:2385–95. 3. Abraham WT, Fisher WG, Smith AL, et al. Cardiac resynchronization in chronic heart failure. N Engl J Med 2002;24:1845–53. 4. 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;21:2140–50. 5. Tracy CM, Epstein AE, Darbar D, et al. 2012 ACCF/AHA/HRS focused update of the 2008 guidelines for device-based therapy of cardiac rhythm abnormalities: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Heart Rhythm 2012;10:1737–53. 6. Brignole M, Auricchio A, Baron-Esquivias G, et al. 2013 ESC guidelines on cardiac pacing and cardiac resynchronization therapy: the Task Force on Cardiac Pacing and resynchronization Therapy cardiac resynchronization therapy. Circulation 2005;112:1580–6. Cardiovascular Society guidelines on the use of cardiac resynchronization therapy: evidence and patient selection. Can J Cardiol 2013;2:182–95. 12. Ypenburg C, van Bommel RJ, Borleffs JW, et al. Long-term prognosis after cardiac resynchronization therapy is related to the extent of left ventricular reverse remodeling at midterm follow up. J Am Coll Cardiol 2009;53: 483–90. 8. Kutyifa V, Kloppe A, Zareba, et al. The influence of left ventricular ejection fraction on the effectiveness of cardiac resynchronization therapy. J Am Coll Cardiol 2013;61:936–44. 13. Rickard J, Michtalik H, Sharma R, et al. Use of Cardiac Resynchronization Therapy in the Medicare Population. Rockville, MD: Agency for Healthcare Research and Quality, 2015. 9. Rickard J, Brennan DM, Martin DO, et al. The impact of left ventricular size on response to cardiac resynchronization therapy. Am Heart J 2011;162:646–53. 10. Rickard J, Cheng A, Spragg DD, et al. Durability of the survival effect of cardiac resynchronization therapy by level of left ventricular functional improvement: fate of “nonresponders”. Heart Rhythm 2014;11:412–6. 11. Yu CM, Bleeker GB, Fung JW, et al. Left ventricular reverse remodeling but not clinical improvement predicts long-term survival after 14. Zweerink A, Friedman DJ, Klem I, et al. Normalization of QRS duration to left ventricular dimension improves prediction of long-term cardiac resynchronization therapy outcomes. Circ Arrythm Electrophysiol 2018;11:e006767. 15. Birnie DH, Tang AS. The problem of nonresponse to cardiac resynchronization therapy. Curr Opin Cardiol 2006;219:20–6. KEY WORDS cardiac resynchronization therapy, reverse remodeling, severe cardiomyopathy, survival 11