- Email: [email protected]
Prominent R wave in ECG lead V1 predicts improvement of left ventricular ejection fraction after cardiac resynchronization therapy in patients with or without left bundle branch block
Prominent R wave in ECG lead V1 predicts improvement of left ventricular ejection fraction after cardiac resynchronization therapy in patients with or without left bundle branch block Weeranun D. Bode, MD, Michael F. Bode, MD, Leonard Gettes, MD, Brian C. Jensen, MD, John P. Mounsey, BMBCh, PhD, Eugene H. Chung, MD, FHRS From the Division of Cardiology, University of North Carolina Hospitals, Chapel Hill, North Carolina. BACKGROUND QRS morphology on postprocedural ECG indicating posterolateral left ventricular pacing may be predictive of response to cardiac resynchronization therapy (CRT). OBJECTIVE The purpose of this study was to assess whether a positive vector in V1 and/or negative vector in lead I on the first postprocedural ECG, suggesting posterolateral capture from CRT, correlates with improvement in left ventricular ejection fraction (LVEF). METHODS A retrospective chart review was conducted on all patients who underwent CRT implantation at our institution between April 2008 and December 2011. Biventricular (BiV) paced QRS morphology was defined as R/S Z1 in V1 and/or R/S r 1 in lead I. The primary outcome was improvement of LVEF Z7.5%. The χ2 and t tests were used for analysis. RESULTS Of 68 patients, 49 (72%) met our BiV paced QRS morphology criteria. Thirty-four of these 49 patients (69%) had improvement in LVEF. Of the 19 patients who did not meet our criteria, 17 (89%) did not have an improvement in LVEF (sensitivity 94%, specificity 53%, χ2 ¼ 19.04, P o .0001). The average LVEF improvement in patients who met our BiV paced QRS morphology criteria was significantly greater than in those who did not (14.27%
Introduction Cardiac resynchronization therapy (CRT) is associated with reduced mortality, improved quality of life, increased functional capacity, and reduction in hospitalizations for heart failure in patients with impaired left ventricular (LV) systolic function.1,2 However, only 70% of patients who receive CRT experience such benefit.3 Thus, it remains a challenge to identify those patients who will respond to CRT. It has been shown that heart failure patients with left bundle branch block (LBBB) and QRS duration 4150 ms respond best to Dr. Mounsey is a paid speaker and consultant to Boston Scientific, Medtronic, and St. Jude Medical. Address reprint requests and correspondence: Dr. Eugene H. Chung, Division of Cardiology, Cardiac Electrophysiology, UNC School of Medicine, 160 Dental Circle, CB 7075, Chapel Hill, NC 27599. E-mail address: [email protected].
1547-5271/$-see front matter B 2015 Heart Rhythm Society. All rights reserved.
vs 2.63%, P ¼ .0001). Preprocedural left bundle branch block was not a predictor of echocardiographic response. CONCLUSION Our results highlight the importance of periprocedural ECG analysis to optimize response to CRT. Moreover, patients without left bundle branch block still benefited from CRT if they met our BiV paced morphology criteria. This suggests that postprocedural left ventricular activation as reflected on the ECG may supersede the baseline conduction delay. KEYWORDS Cardiac resynchronization therapy; Biventricular pacing; QRS morphology; Electrocardiographic criteria; Left ventricular ejection fraction; Left bundle branch block ABBREVIATIONS BiV ¼ biventricular; CRT ¼ cardiac resynchronization therapy; ECG ¼ electrocardiogram; ISFC ¼ International Society and Federation for Cardiology; LBBB ¼ left bundle branch block; LV ¼ left ventricle; LVEF ¼ left ventricular ejection fraction; MADIT ¼ Multicenter Automatic Defibrillator Implantation Trial; RBBB ¼ right bundle branch block; RV ¼ right ventricle; WHO ¼ World Health Organization (Heart Rhythm 2015;12:2141–2147) I 2015 Heart Rhythm Society. All rights reserved.
CRT.4 This suggests that electrical dyssynchrony as well as systolic dysfunction are required for CRT to be effective.5 Because electrical dyssynchrony is important in predicting the response to CRT, multiple studies have focused on identifying preprocedural ECG morphologic characteristics that predict CRT response.6,7 In a subgroup analysis of the Multicenter Automatic Defibrillator Implantation Trial–Cardiac Resynchronization Therapy (MADIT–CRT), the group with QRS 4130 ms and LBBB meeting World Health Organization (WHO) and International Society and Federation for Cardiology (ISFC) criteria derived the most benefit.8,9 Other factors such as the electrical position of the LV lead measured at CRT implantation seem to be important as well.10 However, few efforts have been made to assess the response of patients to CRT by interpreting the paced QRS complex.11 http://dx.doi.org/10.1016/j.hrthm.2015.06.001
2142 Several algorithms have been described for diagnosing loss of LV capture during CRT, but these algorithms have not been used to predict functional response.12–14 Based on these algorithms, we sought to test an ECG-based algorithm that would predict CRT response and hypothesized that a positive vector in lead V1 and/or negative vector in lead I on the first postprocedural ECG, suggesting posterolateral capture from CRT, would indicate the reversal of electrical dyssynchrony and correlate with an improvement in left ventricular ejection fraction (LVEF). We also sought to determine the relative importance of the pre- and postbiventricular (BiV) pacing ECGs in predicting the response to CRT.
Methods A retrospective chart review was conducted between April 2008 and December 2011. All patients older than 18 years who had a new CRT device implanted and had available preand postpacing ECGs and echocardiograms at our institution were included. Patients who did not have pre- and postimplantation ECGs as well as pre- and postimplantation echocardiograms recorded at least 3 months to 3 years after the implantation were excluded. ECG analysis was performed by a board-certified electrophysiologist, and echocardiographic analysis was performed by a board-certified echocardiographer. They were each blinded to the results of the other. The data collection was approved by the institution’s review board and complies with the Declaration of Helsinki. We defined the BiV paced QRS morphology as positive when there was a prominent R wave in lead V1 (R/S ratio Z1 in lead V1) and/or a prominent S wave in lead I (R/S ratio r1 in lead I) on the first postprocedural ECG. For the purposes of this study, we defined the predominant component of the QRS complex above the isoelectric line as the R wave and the predominant component of the QRS complex below the isoelectric line as the S wave (Figure 1). The primary outcome was defined as the change in LVEF between the preprocedural echocardiogram and the first postprocedural echocardiogram, which was obtained between 3 months to 3 years after the CRT. Responders were defined as patients who had a net improvement of LVEF Z7.5%. This was based on the findings of Hsu et al,15 who defined hyporesponders as those with LVEF change o7.9% or the lowest quartile in that study. Echocardiographic analysis was based on the biplane Simpson method. Chi-square and t tests were used for statistical analyses. P r.05 was considered significant.
Results A total of 130 patients underwent initial CRT implantation from April 2008 to December 2011. Sixty-two patients were excluded. Of those patients, 58 did not undergo postprocedural echocardiography at our institution, and 4 received epicardial left ventricular leads because of inability to place the lead in the coronary sinus. The remaining 68 patients
Heart Rhythm, Vol 12, No 10, October 2015 formed our study group (Figure 2). Average age was 64.8 years, and 35% of the patients were female (Table 1). Average baseline LVEF was 23.5%. Nonischemic cardiomyopathy was present in 59.1% of the patients. LBBB, as defined by the World Health Organization criteria,16 was present on the preprocedural ECG in 50% of the patients. Of the 50% with non-LBBB at baseline, 1 was right ventricular (RV) paced, 5 had right bundle branch block (RBBB), and 28 had an interventricular conduction delay. The position of the coronary sinus (LV) pacing lead was determined by a 2view chest X-ray film: posterolateral in 66.2%, anterolateral in 25%, and posterior in 8.8% of the patients. BiV paced QRS morphology criteria were met in 49 of the 68 included patients (72%). Thirty-four of the 49 patients (69%) who met the BiV paced QRS morphology criteria had improvement of LVEF, but 15 patients (31%) did not. Conversely, of the 19 patients who did not meet the BiV paced QRS morphology criteria, 2 (11%) had improvement of LVEF; the remaining 17 patients (89%) did not. There was a statistically significant association between improvement of LVEF and the desired BiV paced QRS morphology (χ2 ¼ 19.04, P o .0001). The sensitivity and specificity of the ECG criteria to predict LVEF improvement were 94% and 53%, respectively. The positive predictive value and negative predictive values were 0.69 and 0.89, respectively (Figure 3). Average LVEF improvement was 14.27% (95% confidence interval 10.91–17.62) in the group with positive BiV paced QRS morphology. In the group with negative BiV paced QRS morphology, average LVEF improvement was 2.63 % (95% confidence interval 0.78– 4.49). This difference was statistically significant (P ¼ .0001; Figure 4). There were no significant differences in the demographic characteristics between responders and nonresponders. Analysis of detailed patient characteristics (lead position, type of cardiomyopathy, preprocedural LBBB, pre- and postprocedural average QRS duration, change in QRS duration, preand postprocedural average QRS axis, and change in QRS axis) did not yield any significant differences between groups (Table 1). We conducted several further analyses. First, in our cohort, 50% of all patients had LBBB on the preprocedural ECG. Fifty-eight percent of the responders and 41% of the nonresponders had LBBB. The presence of LBBB on the preprocedural ECG was not a predictor of LVEF improvement in our cohort (P ¼ .113; Figure 5). We further analyzed whether the duration of the LBBB QRS complex influenced these results but found no significant contribution of QRS duration greater or less than 130 ms. We then evaluated the ability of our algorithm to predict LVEF improvement in patients with preprocedural nonLBBB (Figure 6). One of these patients was chronically RV paced before upgrade to a BiV implantable cardioverterdefibrillator and thus was excluded from this analysis. Of the 20 patients with non-LBBB at baseline but positive postprocedural BiV ECG morphology, 13 (65%) had a positive response in LVEF. Of the 13 patients with non-LBBB at
Bode et al
Prominent R Wave in V1 Predicts LVEF Improvement After CRT
2143
Figure 1 A–C: Examples of positive biventricular (BiV) paced ECG morphology (R/S Z1 in lead V1 and/or R/S r1 in lead I). A: V1: R ¼ 8 mm, S ¼ 5 mm, R/S ¼ 1.6; lead I: R ¼ 0 mm, S ¼ 7 mm, R/S ¼ 0. B: V1: R ¼ 6 mm, S ¼ 5 mm, R/S 1.2; lead I: R ¼ 3 mm, S ¼ 3 mm, R/S ¼ 1. C: V1: R ¼ 4 mm, S ¼ 12 mm, R/S ¼ 0.33; lead I: R ¼ 0 mm, S ¼ 7 mm, R/S ¼ 0. D, E: Examples of negative BiV paced ECG morphology. D: V1: R ¼ 4 mm, S ¼ 12 mm, R/S ¼ 0.33; lead I: R ¼ 6 mm, S ¼ 3 mm, R/S ¼ 2. E: V1: R ¼ 0 mm, S ¼ 11 mm, R/S ¼ 0; lead I: R ¼ 5 mm, S ¼ 1 mm, R/S ¼ 5
baseline but negative postprocedural BiV ECG morphology, 12 (92%) had no response in LVEF, demonstrating the ability of our algorithm to predict LVEF response independently of LBBB. Second, we compared the baseline demographics of the subgroup of nonresponders who had positive BiV paced QRS morphology with the subgroup of responders who had positive BiV paced QRS morphology (Table 2). This subgroup caught our attention because they could potentially provide the explanation to the question as to why patients who fulfilled the BiV paced QRS morphology criteria, and therefore had presumably good lead placement, did not respond to CRT. However, we found no significant differences between the 2 subgroups. There was a trend of more female patients in the group that had improvement of LVEF (P ¼ .081; Table 2).
Third, we assessed whether better specificity or sensitivity could be achieved by increasing the prespecified R/S ratio of lead V1 or decreasing the prespecified R/S ratio of lead I from 1 (ie, R/S Z2 in lead V1 and/or R/S r0.5 in lead I; R/S Z3 in lead V1 and/or R/S r0.33 in lead I). This analysis showed slightly improved specificity (from 53% to 64%) but decreased sensitivity (from 94% to 76%) of the algorithm. The results were similar in all subgroups.
Discussion Several characteristics have been shown to predict CRT responders. Female gender, NICM, LBBB, QRS duration Z150 ms, previous hospitalization for heart failure, LV enddiastolic volume Z125 mL/m2, and left atrial volume o40 mL/m2 all are known to be independent factors associated
2144
Heart Rhythm, Vol 12, No 10, October 2015
Figure 2
Study algorithm. BiV ¼ biventricular; LVEF ¼ left ventricular ejection fraction.
with CRT response.4 Several studies have tried to predict CRT response based on the ECG by analyzing QRS morphology and duration.6,7,11 More recently, attempts have been made to predict LV reverse remodeling (defined as Z10% reduction in LV end-systolic volume) with an algorithm based on the post CRT surface ECG, which combined BiV QRS morphology in leads V1 and V2 with already known predictors such as shortening of QRS duration after CRT, less LV scar, and wide QRS complex with LBBB.17 This approach was able to predict LV reverse Table 1
remodeling, but it is unclear how much impact the morphology criteria had on the algorithm because it was a combined algorithm. A recent retrospective analysis that used QRS morphology criteria showed that a postimplantation ECG with a tall (44 mm) R wave in lead V1 and a predominant negative deflection in lead I predicts a lower risk for unplanned hospitalization, but there was no significant change in LVEF.18 Several possible explanations for the failure to show improvement in LVEF include: (1) in some cases immediate postprocedural ECGs were not available;
Demographics and patient characteristics
Total no. of patients Average age (years) Female (%) Caucasian (%) African-American (%) Nonischemic cardiomyopathy (%) Left bundle branch block (%) Posterolateral CS lead position (%) Anterolateral CS lead position (%) QRS duration (pre) (ms) QRS duration (post) (ms) QRS axis (pre) (degrees) QRS axis (post) (degrees) Baseline LVEF (%)
All patients
Responder ΔLVEF Z7.5%
Nonresponder ΔLVEF o7.5%
P value
68 64.8 (59.7–69.9) 35.3 (25.0–47.2) 57.4 (45.5–68.4) 35.3 (25.0–47.2) 59.1 (47.1–70.1) 50.0 (41.2–58.8) 66.2 (54.3–76.3) 25.0 (16.2–36.4) 153 (145–162) 155 (148–163) -6 (-28–17) 111 (65–157) 23.5 (19.2–27.8)
36 64.8 (60.0–69.5) 36.1 (22.5–52.4) 58.3 (42.2–72.9) 33.3 (20.2–49.7) 58.8 (42.2–73.6) 58.3 (45.8–70.8) 66.7 (50.3–79.8) 27.8 (15.9–44.0) 157 (150–164) 155 (148–162) -11 (-28–6) 129 (85–173) 22.8 (18.7–26.9)
32 64.8 (59.7–70.0) 34.4 (20.4–51.7) 56.3 (39.3–71.8) 37.5 (22.9–54.8) 59.4 (42.3–74.5) 40.6 (30.4–50.8) 65.6 (48.3–79.6) 21.9 (11.0–38.8) 149 (140–158) 155 (148–163) 0 (-27–27) 91 (47–135) 24.5 (19.9–29.0)
.98 .88 .86 .72 .96 .11 .93 .57 .20 .95 .47 .24 .28
P values were calculated using χ2 test for categorical data and Student’s 2-tailed t test for noncategorical data. CS ¼ coronary sinus; LVEF ¼ left ventricular ejection fraction.
Bode et al
Prominent R Wave in V1 Predicts LVEF Improvement After CRT
Figure 3 Results. Thirty-six patients were responders (left ventricular ejection fraction [LVEF] improvement Z7.5%), and 32 patients were nonresponders (LVEF improvement o7.5%). Of the responders, 34 (94%) showed positive biventricular (BiV) paced ECG morphology, and 2 (6%) showed negative BiV paced ECG morphology. Of the nonresponders, 15 (47%) showed positive BiV paced ECG morphology, and 17 (53%) showed negative BiV paced ECG morphology (χ2 ¼ 19.04, P o .0001).
(2) post-CRT echocardiography was performed too soon after the procedure; and (3) the ECG criteria used the absolute value of the R wave in lead V1 rather than the R/S ratio. We hypothesized that using morphologic criteria on the post-CRT surface ECG could provide a simple approach to predict which patients will respond favorably to CRT.19 Ideally, BiV pacing reverses electrical dyssynchrony by promoting activation of the left ventricle before activation of the RV. This generates a RBBB pattern on the BiV paced ECG and creates a more prominent positive R wave in lead V1 and a more negative S wave in lead I. Therefore, we hypothesized that the vectors in leads V1 and I would provide important clues regarding the efficacy of BiV pacing. To this end, we created the morphology criteria to predict CRT responders by using R/S ratio Z1 in lead V1 and/or R/S ratio r1 in lead I and defined CRT response as an increase in LVEF Z7.5%. Our results show that a positive BiV paced QRS morphology on the postprocedural ECG was associated with improvement of LVEF (sensitivity 94%, specificity 53%, P ¼ .0001). Although an ECG with a positive BiV paced QRS morphology did not guarantee a positive response in LVEF (positive predictive value 69%), an ECG not fulfilling the BiV paced QRS morphology criteria had a high negative predictive value (89%). In addition, we found that among the
Figure 4 Left ventricular ejection fraction (LVEF) improvement (%) for patients with positive biventricular (BiV) paced ECG morphology (14.3%; 95% confidence interval 10.91–17.62) (white) and negative BiV paced ECG morphology (2.6%; 95% confidence interval 0.78–4.49) (black). *P ¼ .0001.
2145
Figure 5 Left bundle branch block (LBBB) and left ventricular ejection fraction (LVEF) improvement. Of the responders, 21 (60%) were diagnosed with LBBB, and 14 (40%) were diagnosed with non-LBBB. Of the nonresponders, 13 (41%) were diagnosed with LBBB, and 19 (59%) were diagnosed with non-LBBB (χ2 ¼ 2.51, P ¼ .113).
36 patients with improvement in LVEF Z7.5% after BiV pacing, the 34 with a positive BiV paced QRS morphology had a significantly greater improvement of LVEF than the 2 patients in whom the BiV paced QRS morphology did not meet our criteria. Based on these findings, we believe it is desirable to have a postprocedure BiV paced ECG complex that meets our morphology criteria. Furthermore, it may be beneficial for the patient if measures are taken at the time of implantation to achieve these criteria by checking the BiV paced 12-lead ECG (before removing the guiding catheter) and, if our positive criteria are not present, attempting to reposition the coronary sinus lead. Because our BiV ECG criteria are based on leads V1 and I, we would also like to emphasize the importance of meticulous attention to precise ECG lead placement. We previously showed that high placement of anterior precordial leads can create an RSR0 pattern and alter the ST segment in leads V1 and V2.20 Our criteria also could be useful in the outpatient setting, because patients who do not fulfill the criteria could be referred back to the electrophysiologist to adjust the device settings or attempt to change the lead position (to achieve a more favorable ECG response). Of note, only 50% of patients in our cohort had LBBB when using the WHO criteria to diagnose LBBB. LBBB was not a predictor of LVEF response in our cohort. We were surprised by these findings because we had expected a higher percentage of LBBB due to selection by the electrophysiologist before CRT implantation. The discrepancy may be explained by the strict WHO criteria we used to diagnose LBBB in this study. Compared to LBBB as a criterion to predict response in this relatively unselected cohort, our BiV ECG criteria performed much better in predicting response in LVEF (compare Figures 3 and 5). Moreover, of the 20 patients with non-LBBB at baseline but a positive postprocedural BiV ECG, a relatively high number of patients (n = 13 [65%]) had a positive response in LVEF. This implies not only that patients with LBBB can benefit from CRT but also that a high percentage of non-LBBB patients may respond to CRT if our BiV ECG criteria are present after the procedure.
2146
Heart Rhythm, Vol 12, No 10, October 2015
Figure 6 Left bundle branch block (LBBB) and biventricular (BiV) ECG morphology criteria. Patients by LBBB status on their preprocedural ECG and prediction of response by BiV ECG morphology criteria on their postprocedural ECG. One patient with chronic right ventricular pacing was not included in this analysis, thus reducing the number of patients with non-LBBB to 33. Asterisk denotes presence of right bundle branch block (RBBB) in nonresponders: 1 nonresponder with non-LBBB /þBiV ECG had RBBB, and 4 nonresponders with non-LBBB/–BIV ECG had RBBB. þBiV ECG ¼ positive BiV paced ECG morphology; –BiV ECG ¼ negative BiV paced ECG morphology; RBBB ¼ right bundle branch block.
It appears that the ratio R/S Z1 in lead V1 and/or R/S r1 in lead I achieves the most favorable balance between the sensitivity and specificity of our algorithm. Increasing the prespecified R/S ratio of lead V1 from 1 to higher ratios and by decreasing the prespecified R/S ratio of lead I from 1 to lower ratios decreases the sensitivity of the algorithm. In addition, we were not able to predict LVEF response based Table 2 Comparison of patient characteristics between responders and nonresponders with positive BiV paced QRS morphology criteria
Total no. of patients Average age (years) Female (%) Caucasian (%) Nonischemic cardiomyopathy (%) Left bundle branch block (%) Posterolateral CS lead position (%) Anterolateral CS lead position (%) QRS duration (pre) (ms) QRS duration (post) (ms) QRS axis (pre) (1) QRS axis (post) (1)
LVEF Z7.5% BiV ECGþ
LVEF o7.5% BiV ECGþ
P value
34 65.40 38.24 58.82 54.29
15 65.18 13.33 66.67 56.25
.960 .081 .604 .968
58.82
53.33
.354
64.71
66.67
.894
29.41
26.67
.727
157 155 –9.91 139.12
154 166 –12.27 161.53
.759 .116 .896 .572
Selected patient characteristics of responders (LVEF Z7.5%) with positive BiV paced QRS morphology criteria (BiV ECGþ) and nonresponders (LVEF o7.5%) with positive BiV paced QRS morphology criteria. BiV ¼ biventricular; CS ¼ coronary sinus; LVEF ¼ left ventricular ejection fraction.
on QRS axis, but the QRS axis postimplantation was significantly shifted toward the right after CRT implantation, which has been shown previously.7 Responders experienced on average a slightly larger rightward shift than did nonresponders (Table 1).
Study limitations Our study is not without limitations. It is a retrospective analysis involving the relatively small number of patients who received CRT and underwent postprocedural echocardiography, the timing of which was not constant at our institution. Therefore, it needs to be validated in a larger prospective trial. Subgroup analyses were done on a small number of patients, and there could have been a selection bias because this was a retrospective study involving only patients who presented for follow-up echocardiography. We used LVEF improvement as a marker for CRT response, but it would be desirable to also assess other markers such as heart failure hospitalizations, morbidity and mortality, 6-minute walk test, and quality-of-life surveys.
Conclusion We identified a simple approach to predict improvement after CRT based on analysis of QRS morphology in leads V1 and I on the first postprocedural ECG. This gives the clinician an easy to use and cost-effective tool to assess the success of the procedure, and it highlights the importance of intraprocedural ECG analysis of LV wall capture to optimize response to BiV pacing. In addition, we found that a high proportion of patients without LBBB at baseline but a positive BiV paced QRS benefited from CRT which may
Bode et al
Prominent R Wave in V1 Predicts LVEF Improvement After CRT
expand the indications for CRT. Our future work will continue to focus on identifying patients pre- and postprocedure most likely to benefit from CRT.
References
11.
12.
1. Auricchio A, Prinzen FW. Non-responders to cardiac resynchronization therapy: the magnitude of the problem and the issues. Circ J 2011;75:521–527. 2. 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. J Am Coll Cardiol 2012;60: 1297–1313. 3. Lecoq G, Leclercq C, Leray E, Crocq C, Alonso C, de Place C, Mabo P, Daubert C. Clinical and electrocardiographic predictors of a positive response to cardiac resynchronization therapy in advanced heart failure. Eur Heart J 2005;26: 1094–1100. 4. Goldenberg I, Moss AJ, Hall WJ, et al. Predictors of response to cardiac resynchronization therapy in the Multicenter Automatic Defibrillator Implantation Trial with Cardiac Resynchronization Therapy (MADIT-CRT). Circulation 2011;124:1527–1536. 5. Ruschitzka F, Abraham WT, Singh JP, et al. Cardiac-resynchronization therapy in heart failure with a narrow QRS complex. N Engl J Med 2013;369:1395–1405. 6. Birnie DH, Ha A, Higginson L, Sidhu K, Green M, Philippon F, Thibault B, Wells G, Tang A. Impact of QRS morphology and duration on outcomes after cardiac resynchronization therapy: Results from the ResynchronizationDefibrillation for Ambulatory Heart Failure Trial (RAFT). Circ Heart Fail 2013;6: 1190–1198. 7. Takaya Y, Noda T, Nakajima I, et al. Electrocardiographic predictors of response to cardiac resynchronization therapy in patients with intraventricular conduction delay. Circ J 2014;78:71–77. 8. Gettes LS, Kligfield P. Should electrocardiogram criteria for the diagnosis of left bundle-branch block be revised? J Electrocardiol 2012;45:500–504. 9. Zareba W, Klein H, Cygankiewicz I, et al. Effectiveness of cardiac resynchronization therapy by QRS morphology in the Multicenter Automatic Defibrillator Implantation Trial-Cardiac Resynchronization Therapy (MADIT-CRT). Circulation 2011;123:1061–1072. 10. Gold MR, Niazi I, Giudici M, Leman RB, Sturdivant JL, Kim MH, Yu Y. A prospective, randomized comparison of the acute hemodynamic effects of
13.
14.
15.
16.
17.
18.
19.
20.
2147
biventricular and left ventricular pacing with cardiac resynchronization therapy. Heart Rhythm 2011;8:685–691. Mollo R, Cosenza A, Coviello I, Stazi A, Russo G, Villano A, Sestito A, Bencardino G, Lanza GA, Crea F. A novel electrocardiographic predictor of clinical response to cardiac resynchronization therapy. Europace 2013;15: 1615–1621. Ammann P, Sticherling C, Kalusche D, Eckstein J, Bernheim A, Schaer B, Osswald S. An electrocardiogram-based algorithm to detect loss of left ventricular capture during cardiac resynchronization therapy. Ann Intern Med 2005;142(12 Pt 1):968–973. Ganière V, Domenichini G, Niculescu V, Cassagneau R, Defaye P, Burri H. A new electrocardiogram algorithm for diagnosing loss of ventricular capture during cardiac resynchronisation therapy. Europace 2013;15:376–381. Yong P, Duby C. A new and reliable method of individual ventricular capture identification during biventricular pacing threshold testing. Pacing Clin Electrophysiol 2000;23(11 Pt 2):1735–1737. Hsu JC, Solomon SD, Bourgoun M, McNitt S, Goldenberg I, Klein H, Moss AJ, Foster E. Predictors of super-response to cardiac resynchronization therapy and associated improvement in clinical outcome: the MADIT-CRT (Multicenter Automatic Defibrillator Implantation Trial with Cardiac Resynchronization Therapy) study. J Am Coll Cardiol 2012;59:2366–2373. Willems JL, Robles de Medina EO, Bernard R, Coumel P, Fisch C, Krikler D, Mazur NA, Meijler FL, Mogensen L, Moret P. Criteria for intraventricular conduction disturbances and pre-excitation. World Health Organizational/International Society and Federation for Cardiology Task Force Ad Hoc. J Am Coll Cardiol 1985;5:1261–1275. Sweeney MO, Hellkamp AS, van Bommel RJ, Schalij MJ, Borleffs CJW, Bax JJ. QRS fusion complex analysis using wave interference to predict reverse remodeling during cardiac resynchronization therapy. Heart Rhythm 2014;11: 806–813. Coverstone E, Sheehy J, Kleiger R, Smith T. The post-implantation electrocardiogram predicts clinical response to cardiac resynchronization therapy. Pacing Clin Electrophysiol 2015;38:572–580. Bode W, Bode M, Simpson R, Gehi A, Schwartz J, Mounsey J, Chung E. Improvement of left ventricular ejection fraction after cardiac resynchronization therapy can be predicted based on biventricular capture on electrocardiogram. Circulation 2013;128:A16364. Chung EH, McNeely DE, Gehi AK, et al. Brugada-type patterns are easily observed in high precordial lead ECGs in collegiate athletes. J Electrocardiol 2014;47:1–6.
CLINICAL PERSPECTIVES We report a simple approach for predicting improvement from cardiac resynchronization therapy (CRT) based on the QRS morphology in leads V1 and I on the12-lead ECG. Analysis depends on proper placement of ECG leads. Intraprocedural ECG review of biventricular (BiV) pacing should be considered, and improving the QRS morphology by coronary sinus lead repositioning might result in better response. In addition, we found that a high proportion of patients without left bundle branch block (LBBB) at baseline but with a positive BiV paced QRS morphology postprocedure responded to CRT. Our findings suggest that optimizing posterolateral wall capture is crucial to optimizing response to CRT, even in patients without true LBBB.
Introduction Cardiac resynchronization therapy (CRT) is associated with reduced mortality, improved quality of life, increased functional capacity, and reduction in hospitalizations for heart failure in patients with impaired left ventricular (LV) systolic function.1,2 However, only 70% of patients who receive CRT experience such benefit.3 Thus, it remains a challenge to identify those patients who will respond to CRT. It has been shown that heart failure patients with left bundle branch block (LBBB) and QRS duration 4150 ms respond best to Dr. Mounsey is a paid speaker and consultant to Boston Scientific, Medtronic, and St. Jude Medical. Address reprint requests and correspondence: Dr. Eugene H. Chung, Division of Cardiology, Cardiac Electrophysiology, UNC School of Medicine, 160 Dental Circle, CB 7075, Chapel Hill, NC 27599. E-mail address: [email protected].
1547-5271/$-see front matter B 2015 Heart Rhythm Society. All rights reserved.
vs 2.63%, P ¼ .0001). Preprocedural left bundle branch block was not a predictor of echocardiographic response. CONCLUSION Our results highlight the importance of periprocedural ECG analysis to optimize response to CRT. Moreover, patients without left bundle branch block still benefited from CRT if they met our BiV paced morphology criteria. This suggests that postprocedural left ventricular activation as reflected on the ECG may supersede the baseline conduction delay. KEYWORDS Cardiac resynchronization therapy; Biventricular pacing; QRS morphology; Electrocardiographic criteria; Left ventricular ejection fraction; Left bundle branch block ABBREVIATIONS BiV ¼ biventricular; CRT ¼ cardiac resynchronization therapy; ECG ¼ electrocardiogram; ISFC ¼ International Society and Federation for Cardiology; LBBB ¼ left bundle branch block; LV ¼ left ventricle; LVEF ¼ left ventricular ejection fraction; MADIT ¼ Multicenter Automatic Defibrillator Implantation Trial; RBBB ¼ right bundle branch block; RV ¼ right ventricle; WHO ¼ World Health Organization (Heart Rhythm 2015;12:2141–2147) I 2015 Heart Rhythm Society. All rights reserved.
CRT.4 This suggests that electrical dyssynchrony as well as systolic dysfunction are required for CRT to be effective.5 Because electrical dyssynchrony is important in predicting the response to CRT, multiple studies have focused on identifying preprocedural ECG morphologic characteristics that predict CRT response.6,7 In a subgroup analysis of the Multicenter Automatic Defibrillator Implantation Trial–Cardiac Resynchronization Therapy (MADIT–CRT), the group with QRS 4130 ms and LBBB meeting World Health Organization (WHO) and International Society and Federation for Cardiology (ISFC) criteria derived the most benefit.8,9 Other factors such as the electrical position of the LV lead measured at CRT implantation seem to be important as well.10 However, few efforts have been made to assess the response of patients to CRT by interpreting the paced QRS complex.11 http://dx.doi.org/10.1016/j.hrthm.2015.06.001
2142 Several algorithms have been described for diagnosing loss of LV capture during CRT, but these algorithms have not been used to predict functional response.12–14 Based on these algorithms, we sought to test an ECG-based algorithm that would predict CRT response and hypothesized that a positive vector in lead V1 and/or negative vector in lead I on the first postprocedural ECG, suggesting posterolateral capture from CRT, would indicate the reversal of electrical dyssynchrony and correlate with an improvement in left ventricular ejection fraction (LVEF). We also sought to determine the relative importance of the pre- and postbiventricular (BiV) pacing ECGs in predicting the response to CRT.
Methods A retrospective chart review was conducted between April 2008 and December 2011. All patients older than 18 years who had a new CRT device implanted and had available preand postpacing ECGs and echocardiograms at our institution were included. Patients who did not have pre- and postimplantation ECGs as well as pre- and postimplantation echocardiograms recorded at least 3 months to 3 years after the implantation were excluded. ECG analysis was performed by a board-certified electrophysiologist, and echocardiographic analysis was performed by a board-certified echocardiographer. They were each blinded to the results of the other. The data collection was approved by the institution’s review board and complies with the Declaration of Helsinki. We defined the BiV paced QRS morphology as positive when there was a prominent R wave in lead V1 (R/S ratio Z1 in lead V1) and/or a prominent S wave in lead I (R/S ratio r1 in lead I) on the first postprocedural ECG. For the purposes of this study, we defined the predominant component of the QRS complex above the isoelectric line as the R wave and the predominant component of the QRS complex below the isoelectric line as the S wave (Figure 1). The primary outcome was defined as the change in LVEF between the preprocedural echocardiogram and the first postprocedural echocardiogram, which was obtained between 3 months to 3 years after the CRT. Responders were defined as patients who had a net improvement of LVEF Z7.5%. This was based on the findings of Hsu et al,15 who defined hyporesponders as those with LVEF change o7.9% or the lowest quartile in that study. Echocardiographic analysis was based on the biplane Simpson method. Chi-square and t tests were used for statistical analyses. P r.05 was considered significant.
Results A total of 130 patients underwent initial CRT implantation from April 2008 to December 2011. Sixty-two patients were excluded. Of those patients, 58 did not undergo postprocedural echocardiography at our institution, and 4 received epicardial left ventricular leads because of inability to place the lead in the coronary sinus. The remaining 68 patients
Heart Rhythm, Vol 12, No 10, October 2015 formed our study group (Figure 2). Average age was 64.8 years, and 35% of the patients were female (Table 1). Average baseline LVEF was 23.5%. Nonischemic cardiomyopathy was present in 59.1% of the patients. LBBB, as defined by the World Health Organization criteria,16 was present on the preprocedural ECG in 50% of the patients. Of the 50% with non-LBBB at baseline, 1 was right ventricular (RV) paced, 5 had right bundle branch block (RBBB), and 28 had an interventricular conduction delay. The position of the coronary sinus (LV) pacing lead was determined by a 2view chest X-ray film: posterolateral in 66.2%, anterolateral in 25%, and posterior in 8.8% of the patients. BiV paced QRS morphology criteria were met in 49 of the 68 included patients (72%). Thirty-four of the 49 patients (69%) who met the BiV paced QRS morphology criteria had improvement of LVEF, but 15 patients (31%) did not. Conversely, of the 19 patients who did not meet the BiV paced QRS morphology criteria, 2 (11%) had improvement of LVEF; the remaining 17 patients (89%) did not. There was a statistically significant association between improvement of LVEF and the desired BiV paced QRS morphology (χ2 ¼ 19.04, P o .0001). The sensitivity and specificity of the ECG criteria to predict LVEF improvement were 94% and 53%, respectively. The positive predictive value and negative predictive values were 0.69 and 0.89, respectively (Figure 3). Average LVEF improvement was 14.27% (95% confidence interval 10.91–17.62) in the group with positive BiV paced QRS morphology. In the group with negative BiV paced QRS morphology, average LVEF improvement was 2.63 % (95% confidence interval 0.78– 4.49). This difference was statistically significant (P ¼ .0001; Figure 4). There were no significant differences in the demographic characteristics between responders and nonresponders. Analysis of detailed patient characteristics (lead position, type of cardiomyopathy, preprocedural LBBB, pre- and postprocedural average QRS duration, change in QRS duration, preand postprocedural average QRS axis, and change in QRS axis) did not yield any significant differences between groups (Table 1). We conducted several further analyses. First, in our cohort, 50% of all patients had LBBB on the preprocedural ECG. Fifty-eight percent of the responders and 41% of the nonresponders had LBBB. The presence of LBBB on the preprocedural ECG was not a predictor of LVEF improvement in our cohort (P ¼ .113; Figure 5). We further analyzed whether the duration of the LBBB QRS complex influenced these results but found no significant contribution of QRS duration greater or less than 130 ms. We then evaluated the ability of our algorithm to predict LVEF improvement in patients with preprocedural nonLBBB (Figure 6). One of these patients was chronically RV paced before upgrade to a BiV implantable cardioverterdefibrillator and thus was excluded from this analysis. Of the 20 patients with non-LBBB at baseline but positive postprocedural BiV ECG morphology, 13 (65%) had a positive response in LVEF. Of the 13 patients with non-LBBB at
Bode et al
Prominent R Wave in V1 Predicts LVEF Improvement After CRT
2143
Figure 1 A–C: Examples of positive biventricular (BiV) paced ECG morphology (R/S Z1 in lead V1 and/or R/S r1 in lead I). A: V1: R ¼ 8 mm, S ¼ 5 mm, R/S ¼ 1.6; lead I: R ¼ 0 mm, S ¼ 7 mm, R/S ¼ 0. B: V1: R ¼ 6 mm, S ¼ 5 mm, R/S 1.2; lead I: R ¼ 3 mm, S ¼ 3 mm, R/S ¼ 1. C: V1: R ¼ 4 mm, S ¼ 12 mm, R/S ¼ 0.33; lead I: R ¼ 0 mm, S ¼ 7 mm, R/S ¼ 0. D, E: Examples of negative BiV paced ECG morphology. D: V1: R ¼ 4 mm, S ¼ 12 mm, R/S ¼ 0.33; lead I: R ¼ 6 mm, S ¼ 3 mm, R/S ¼ 2. E: V1: R ¼ 0 mm, S ¼ 11 mm, R/S ¼ 0; lead I: R ¼ 5 mm, S ¼ 1 mm, R/S ¼ 5
baseline but negative postprocedural BiV ECG morphology, 12 (92%) had no response in LVEF, demonstrating the ability of our algorithm to predict LVEF response independently of LBBB. Second, we compared the baseline demographics of the subgroup of nonresponders who had positive BiV paced QRS morphology with the subgroup of responders who had positive BiV paced QRS morphology (Table 2). This subgroup caught our attention because they could potentially provide the explanation to the question as to why patients who fulfilled the BiV paced QRS morphology criteria, and therefore had presumably good lead placement, did not respond to CRT. However, we found no significant differences between the 2 subgroups. There was a trend of more female patients in the group that had improvement of LVEF (P ¼ .081; Table 2).
Third, we assessed whether better specificity or sensitivity could be achieved by increasing the prespecified R/S ratio of lead V1 or decreasing the prespecified R/S ratio of lead I from 1 (ie, R/S Z2 in lead V1 and/or R/S r0.5 in lead I; R/S Z3 in lead V1 and/or R/S r0.33 in lead I). This analysis showed slightly improved specificity (from 53% to 64%) but decreased sensitivity (from 94% to 76%) of the algorithm. The results were similar in all subgroups.
Discussion Several characteristics have been shown to predict CRT responders. Female gender, NICM, LBBB, QRS duration Z150 ms, previous hospitalization for heart failure, LV enddiastolic volume Z125 mL/m2, and left atrial volume o40 mL/m2 all are known to be independent factors associated
2144
Heart Rhythm, Vol 12, No 10, October 2015
Figure 2
Study algorithm. BiV ¼ biventricular; LVEF ¼ left ventricular ejection fraction.
with CRT response.4 Several studies have tried to predict CRT response based on the ECG by analyzing QRS morphology and duration.6,7,11 More recently, attempts have been made to predict LV reverse remodeling (defined as Z10% reduction in LV end-systolic volume) with an algorithm based on the post CRT surface ECG, which combined BiV QRS morphology in leads V1 and V2 with already known predictors such as shortening of QRS duration after CRT, less LV scar, and wide QRS complex with LBBB.17 This approach was able to predict LV reverse Table 1
remodeling, but it is unclear how much impact the morphology criteria had on the algorithm because it was a combined algorithm. A recent retrospective analysis that used QRS morphology criteria showed that a postimplantation ECG with a tall (44 mm) R wave in lead V1 and a predominant negative deflection in lead I predicts a lower risk for unplanned hospitalization, but there was no significant change in LVEF.18 Several possible explanations for the failure to show improvement in LVEF include: (1) in some cases immediate postprocedural ECGs were not available;
Demographics and patient characteristics
Total no. of patients Average age (years) Female (%) Caucasian (%) African-American (%) Nonischemic cardiomyopathy (%) Left bundle branch block (%) Posterolateral CS lead position (%) Anterolateral CS lead position (%) QRS duration (pre) (ms) QRS duration (post) (ms) QRS axis (pre) (degrees) QRS axis (post) (degrees) Baseline LVEF (%)
All patients
Responder ΔLVEF Z7.5%
Nonresponder ΔLVEF o7.5%
P value
68 64.8 (59.7–69.9) 35.3 (25.0–47.2) 57.4 (45.5–68.4) 35.3 (25.0–47.2) 59.1 (47.1–70.1) 50.0 (41.2–58.8) 66.2 (54.3–76.3) 25.0 (16.2–36.4) 153 (145–162) 155 (148–163) -6 (-28–17) 111 (65–157) 23.5 (19.2–27.8)
36 64.8 (60.0–69.5) 36.1 (22.5–52.4) 58.3 (42.2–72.9) 33.3 (20.2–49.7) 58.8 (42.2–73.6) 58.3 (45.8–70.8) 66.7 (50.3–79.8) 27.8 (15.9–44.0) 157 (150–164) 155 (148–162) -11 (-28–6) 129 (85–173) 22.8 (18.7–26.9)
32 64.8 (59.7–70.0) 34.4 (20.4–51.7) 56.3 (39.3–71.8) 37.5 (22.9–54.8) 59.4 (42.3–74.5) 40.6 (30.4–50.8) 65.6 (48.3–79.6) 21.9 (11.0–38.8) 149 (140–158) 155 (148–163) 0 (-27–27) 91 (47–135) 24.5 (19.9–29.0)
.98 .88 .86 .72 .96 .11 .93 .57 .20 .95 .47 .24 .28
P values were calculated using χ2 test for categorical data and Student’s 2-tailed t test for noncategorical data. CS ¼ coronary sinus; LVEF ¼ left ventricular ejection fraction.
Bode et al
Prominent R Wave in V1 Predicts LVEF Improvement After CRT
Figure 3 Results. Thirty-six patients were responders (left ventricular ejection fraction [LVEF] improvement Z7.5%), and 32 patients were nonresponders (LVEF improvement o7.5%). Of the responders, 34 (94%) showed positive biventricular (BiV) paced ECG morphology, and 2 (6%) showed negative BiV paced ECG morphology. Of the nonresponders, 15 (47%) showed positive BiV paced ECG morphology, and 17 (53%) showed negative BiV paced ECG morphology (χ2 ¼ 19.04, P o .0001).
(2) post-CRT echocardiography was performed too soon after the procedure; and (3) the ECG criteria used the absolute value of the R wave in lead V1 rather than the R/S ratio. We hypothesized that using morphologic criteria on the post-CRT surface ECG could provide a simple approach to predict which patients will respond favorably to CRT.19 Ideally, BiV pacing reverses electrical dyssynchrony by promoting activation of the left ventricle before activation of the RV. This generates a RBBB pattern on the BiV paced ECG and creates a more prominent positive R wave in lead V1 and a more negative S wave in lead I. Therefore, we hypothesized that the vectors in leads V1 and I would provide important clues regarding the efficacy of BiV pacing. To this end, we created the morphology criteria to predict CRT responders by using R/S ratio Z1 in lead V1 and/or R/S ratio r1 in lead I and defined CRT response as an increase in LVEF Z7.5%. Our results show that a positive BiV paced QRS morphology on the postprocedural ECG was associated with improvement of LVEF (sensitivity 94%, specificity 53%, P ¼ .0001). Although an ECG with a positive BiV paced QRS morphology did not guarantee a positive response in LVEF (positive predictive value 69%), an ECG not fulfilling the BiV paced QRS morphology criteria had a high negative predictive value (89%). In addition, we found that among the
Figure 4 Left ventricular ejection fraction (LVEF) improvement (%) for patients with positive biventricular (BiV) paced ECG morphology (14.3%; 95% confidence interval 10.91–17.62) (white) and negative BiV paced ECG morphology (2.6%; 95% confidence interval 0.78–4.49) (black). *P ¼ .0001.
2145
Figure 5 Left bundle branch block (LBBB) and left ventricular ejection fraction (LVEF) improvement. Of the responders, 21 (60%) were diagnosed with LBBB, and 14 (40%) were diagnosed with non-LBBB. Of the nonresponders, 13 (41%) were diagnosed with LBBB, and 19 (59%) were diagnosed with non-LBBB (χ2 ¼ 2.51, P ¼ .113).
36 patients with improvement in LVEF Z7.5% after BiV pacing, the 34 with a positive BiV paced QRS morphology had a significantly greater improvement of LVEF than the 2 patients in whom the BiV paced QRS morphology did not meet our criteria. Based on these findings, we believe it is desirable to have a postprocedure BiV paced ECG complex that meets our morphology criteria. Furthermore, it may be beneficial for the patient if measures are taken at the time of implantation to achieve these criteria by checking the BiV paced 12-lead ECG (before removing the guiding catheter) and, if our positive criteria are not present, attempting to reposition the coronary sinus lead. Because our BiV ECG criteria are based on leads V1 and I, we would also like to emphasize the importance of meticulous attention to precise ECG lead placement. We previously showed that high placement of anterior precordial leads can create an RSR0 pattern and alter the ST segment in leads V1 and V2.20 Our criteria also could be useful in the outpatient setting, because patients who do not fulfill the criteria could be referred back to the electrophysiologist to adjust the device settings or attempt to change the lead position (to achieve a more favorable ECG response). Of note, only 50% of patients in our cohort had LBBB when using the WHO criteria to diagnose LBBB. LBBB was not a predictor of LVEF response in our cohort. We were surprised by these findings because we had expected a higher percentage of LBBB due to selection by the electrophysiologist before CRT implantation. The discrepancy may be explained by the strict WHO criteria we used to diagnose LBBB in this study. Compared to LBBB as a criterion to predict response in this relatively unselected cohort, our BiV ECG criteria performed much better in predicting response in LVEF (compare Figures 3 and 5). Moreover, of the 20 patients with non-LBBB at baseline but a positive postprocedural BiV ECG, a relatively high number of patients (n = 13 [65%]) had a positive response in LVEF. This implies not only that patients with LBBB can benefit from CRT but also that a high percentage of non-LBBB patients may respond to CRT if our BiV ECG criteria are present after the procedure.
2146
Heart Rhythm, Vol 12, No 10, October 2015
Figure 6 Left bundle branch block (LBBB) and biventricular (BiV) ECG morphology criteria. Patients by LBBB status on their preprocedural ECG and prediction of response by BiV ECG morphology criteria on their postprocedural ECG. One patient with chronic right ventricular pacing was not included in this analysis, thus reducing the number of patients with non-LBBB to 33. Asterisk denotes presence of right bundle branch block (RBBB) in nonresponders: 1 nonresponder with non-LBBB /þBiV ECG had RBBB, and 4 nonresponders with non-LBBB/–BIV ECG had RBBB. þBiV ECG ¼ positive BiV paced ECG morphology; –BiV ECG ¼ negative BiV paced ECG morphology; RBBB ¼ right bundle branch block.
It appears that the ratio R/S Z1 in lead V1 and/or R/S r1 in lead I achieves the most favorable balance between the sensitivity and specificity of our algorithm. Increasing the prespecified R/S ratio of lead V1 from 1 to higher ratios and by decreasing the prespecified R/S ratio of lead I from 1 to lower ratios decreases the sensitivity of the algorithm. In addition, we were not able to predict LVEF response based Table 2 Comparison of patient characteristics between responders and nonresponders with positive BiV paced QRS morphology criteria
Total no. of patients Average age (years) Female (%) Caucasian (%) Nonischemic cardiomyopathy (%) Left bundle branch block (%) Posterolateral CS lead position (%) Anterolateral CS lead position (%) QRS duration (pre) (ms) QRS duration (post) (ms) QRS axis (pre) (1) QRS axis (post) (1)
LVEF Z7.5% BiV ECGþ
LVEF o7.5% BiV ECGþ
P value
34 65.40 38.24 58.82 54.29
15 65.18 13.33 66.67 56.25
.960 .081 .604 .968
58.82
53.33
.354
64.71
66.67
.894
29.41
26.67
.727
157 155 –9.91 139.12
154 166 –12.27 161.53
.759 .116 .896 .572
Selected patient characteristics of responders (LVEF Z7.5%) with positive BiV paced QRS morphology criteria (BiV ECGþ) and nonresponders (LVEF o7.5%) with positive BiV paced QRS morphology criteria. BiV ¼ biventricular; CS ¼ coronary sinus; LVEF ¼ left ventricular ejection fraction.
on QRS axis, but the QRS axis postimplantation was significantly shifted toward the right after CRT implantation, which has been shown previously.7 Responders experienced on average a slightly larger rightward shift than did nonresponders (Table 1).
Study limitations Our study is not without limitations. It is a retrospective analysis involving the relatively small number of patients who received CRT and underwent postprocedural echocardiography, the timing of which was not constant at our institution. Therefore, it needs to be validated in a larger prospective trial. Subgroup analyses were done on a small number of patients, and there could have been a selection bias because this was a retrospective study involving only patients who presented for follow-up echocardiography. We used LVEF improvement as a marker for CRT response, but it would be desirable to also assess other markers such as heart failure hospitalizations, morbidity and mortality, 6-minute walk test, and quality-of-life surveys.
Conclusion We identified a simple approach to predict improvement after CRT based on analysis of QRS morphology in leads V1 and I on the first postprocedural ECG. This gives the clinician an easy to use and cost-effective tool to assess the success of the procedure, and it highlights the importance of intraprocedural ECG analysis of LV wall capture to optimize response to BiV pacing. In addition, we found that a high proportion of patients without LBBB at baseline but a positive BiV paced QRS benefited from CRT which may
Bode et al
Prominent R Wave in V1 Predicts LVEF Improvement After CRT
expand the indications for CRT. Our future work will continue to focus on identifying patients pre- and postprocedure most likely to benefit from CRT.
References
11.
12.
1. Auricchio A, Prinzen FW. Non-responders to cardiac resynchronization therapy: the magnitude of the problem and the issues. Circ J 2011;75:521–527. 2. 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. J Am Coll Cardiol 2012;60: 1297–1313. 3. Lecoq G, Leclercq C, Leray E, Crocq C, Alonso C, de Place C, Mabo P, Daubert C. Clinical and electrocardiographic predictors of a positive response to cardiac resynchronization therapy in advanced heart failure. Eur Heart J 2005;26: 1094–1100. 4. Goldenberg I, Moss AJ, Hall WJ, et al. Predictors of response to cardiac resynchronization therapy in the Multicenter Automatic Defibrillator Implantation Trial with Cardiac Resynchronization Therapy (MADIT-CRT). Circulation 2011;124:1527–1536. 5. Ruschitzka F, Abraham WT, Singh JP, et al. Cardiac-resynchronization therapy in heart failure with a narrow QRS complex. N Engl J Med 2013;369:1395–1405. 6. Birnie DH, Ha A, Higginson L, Sidhu K, Green M, Philippon F, Thibault B, Wells G, Tang A. Impact of QRS morphology and duration on outcomes after cardiac resynchronization therapy: Results from the ResynchronizationDefibrillation for Ambulatory Heart Failure Trial (RAFT). Circ Heart Fail 2013;6: 1190–1198. 7. Takaya Y, Noda T, Nakajima I, et al. Electrocardiographic predictors of response to cardiac resynchronization therapy in patients with intraventricular conduction delay. Circ J 2014;78:71–77. 8. Gettes LS, Kligfield P. Should electrocardiogram criteria for the diagnosis of left bundle-branch block be revised? J Electrocardiol 2012;45:500–504. 9. Zareba W, Klein H, Cygankiewicz I, et al. Effectiveness of cardiac resynchronization therapy by QRS morphology in the Multicenter Automatic Defibrillator Implantation Trial-Cardiac Resynchronization Therapy (MADIT-CRT). Circulation 2011;123:1061–1072. 10. Gold MR, Niazi I, Giudici M, Leman RB, Sturdivant JL, Kim MH, Yu Y. A prospective, randomized comparison of the acute hemodynamic effects of
13.
14.
15.
16.
17.
18.
19.
20.
2147
biventricular and left ventricular pacing with cardiac resynchronization therapy. Heart Rhythm 2011;8:685–691. Mollo R, Cosenza A, Coviello I, Stazi A, Russo G, Villano A, Sestito A, Bencardino G, Lanza GA, Crea F. A novel electrocardiographic predictor of clinical response to cardiac resynchronization therapy. Europace 2013;15: 1615–1621. Ammann P, Sticherling C, Kalusche D, Eckstein J, Bernheim A, Schaer B, Osswald S. An electrocardiogram-based algorithm to detect loss of left ventricular capture during cardiac resynchronization therapy. Ann Intern Med 2005;142(12 Pt 1):968–973. Ganière V, Domenichini G, Niculescu V, Cassagneau R, Defaye P, Burri H. A new electrocardiogram algorithm for diagnosing loss of ventricular capture during cardiac resynchronisation therapy. Europace 2013;15:376–381. Yong P, Duby C. A new and reliable method of individual ventricular capture identification during biventricular pacing threshold testing. Pacing Clin Electrophysiol 2000;23(11 Pt 2):1735–1737. Hsu JC, Solomon SD, Bourgoun M, McNitt S, Goldenberg I, Klein H, Moss AJ, Foster E. Predictors of super-response to cardiac resynchronization therapy and associated improvement in clinical outcome: the MADIT-CRT (Multicenter Automatic Defibrillator Implantation Trial with Cardiac Resynchronization Therapy) study. J Am Coll Cardiol 2012;59:2366–2373. Willems JL, Robles de Medina EO, Bernard R, Coumel P, Fisch C, Krikler D, Mazur NA, Meijler FL, Mogensen L, Moret P. Criteria for intraventricular conduction disturbances and pre-excitation. World Health Organizational/International Society and Federation for Cardiology Task Force Ad Hoc. J Am Coll Cardiol 1985;5:1261–1275. Sweeney MO, Hellkamp AS, van Bommel RJ, Schalij MJ, Borleffs CJW, Bax JJ. QRS fusion complex analysis using wave interference to predict reverse remodeling during cardiac resynchronization therapy. Heart Rhythm 2014;11: 806–813. Coverstone E, Sheehy J, Kleiger R, Smith T. The post-implantation electrocardiogram predicts clinical response to cardiac resynchronization therapy. Pacing Clin Electrophysiol 2015;38:572–580. Bode W, Bode M, Simpson R, Gehi A, Schwartz J, Mounsey J, Chung E. Improvement of left ventricular ejection fraction after cardiac resynchronization therapy can be predicted based on biventricular capture on electrocardiogram. Circulation 2013;128:A16364. Chung EH, McNeely DE, Gehi AK, et al. Brugada-type patterns are easily observed in high precordial lead ECGs in collegiate athletes. J Electrocardiol 2014;47:1–6.
CLINICAL PERSPECTIVES We report a simple approach for predicting improvement from cardiac resynchronization therapy (CRT) based on the QRS morphology in leads V1 and I on the12-lead ECG. Analysis depends on proper placement of ECG leads. Intraprocedural ECG review of biventricular (BiV) pacing should be considered, and improving the QRS morphology by coronary sinus lead repositioning might result in better response. In addition, we found that a high proportion of patients without left bundle branch block (LBBB) at baseline but with a positive BiV paced QRS morphology postprocedure responded to CRT. Our findings suggest that optimizing posterolateral wall capture is crucial to optimizing response to CRT, even in patients without true LBBB.