Several studies have demonstrated a relation between left ventricular (LV) dyssynchrony and response to cardiac resynchronization therapy (CRT). Many methods of determining LV dyssynchrony have been proposed, including a value of 65 ms as assessed by tissue Doppler imaging. The aim of the present validation study was to prospectively test the predictive accuracy of the 65-ms cutoff for LV dyssynchrony in a large cohort of patients with heart failure undergoing CRT. The study included 361 patients who had undergone CRT. The clinical and echocardiographic parameters were assessed at baseline and at 6 months of follow-up. The clinical response was defined as an improvement of ≥1 New York Heart Association class, and the echocardiographic response was defined as a reduction in LV end-systolic volume of ≥15%. At 6 months of follow-up, 259 patients (72%) had a clinical response and 187 patients (52%) had an echocardiographic response. The patients with a response had more LV dyssynchrony than did those without a response (91 ± 49 ms vs 50 ± 44 ms for the clinical response and 101 ± 46 ms vs 55 ± 45 ms for the echocardiographic response). On multivariate analysis, LV dyssynchrony remained predictive of the response, independent of other characteristics. In conclusion, LV dyssynchrony of ≥65 ms was an independent predictor of both the clinical and the echocardiographic response in patients with heart failure who underwent CRT in this validation study.
The Predictors of Response to Cardiac Resynchronization Therapy (PROSPECT) study was the first large-scale, multicenter clinical trial that evaluated the capability of several echocardiographic measures of dyssynchrony to predict the response to cardiac resynchronization therapy (CRT). Various measures of dyssynchrony were predictive of the clinical outcome and reverse remodeling at 6 months of follow-up. However, the sensitivity and specificity of these markers were modest. One of the dyssynchrony measures with the greatest predictive value in the PROSPECT study was the septal to lateral delay, defined as the delay between the interval to peak systolic velocity at the basal septal and basal lateral segments, measured using color-coded tissue Doppler imaging. This measure of dyssynchrony was proposed by our center, and the extent of left ventricular (LV) dyssynchrony needed to predict the response to CRT was defined as 60 ms. Subsequently, a more sensitive marker of LV dyssynchrony of ≥65 ms among the 4 basal walls within the left ventricle (most frequently observed between the interventricular septum and the lateral wall) was identified. This cutoff value, however, was derived by receiver operating characteristic curve analysis and thus needs validation. Therefore, that initial population should be considered a “learning population,” used to derive the optimal cutoff value for LV dyssynchrony, which then would require additional prospective testing in a so-called validation population. Accordingly, the aim of the present study was to prospectively test the predictive accuracy of the 65-ms cutoff value for LV dyssynchrony in a large cohort of patients with heart failure undergoing CRT.
Methods
A total of 361 patients who were scheduled for CRT were included in the present study. The patients were selected according to the current inclusion criteria: New York Heart Association (NYHA) class III and IV, LV ejection fraction (LVEF) of ≤35%, and QRS duration of ≥120 ms. The study protocol was as follows. In all patients, the clinical status was assessed before implantation, and 2-dimensional echocardiography was performed to measure the LVEF and LV volumes. Color-coded tissue Doppler imaging was used to measure the extent of LV dyssynchrony. The clinical status and changes in the LVEF and LV volumes were reassessed at 6 months of follow-up.
The clinical status was assessed at baseline and at 6 months of follow-up and included NYHA functional class, quality-of-life score (using the Minnesota Living with Heart Failure questionnaire), and the distance covered in the 6-minute walking test.
All patients underwent echocardiography in the left lateral decubitus position before and 6 months after CRT device implantation. The studies were performed using a commercially available echocardiographic system (VIVID 7, General Electric Vingmed Ultrasound, Milwaukee, Wisconsin). The images were obtained using a 3.5-MHz transducer, at a depth of 16 cm in the parasternal (long- and short-axis) and apical (2- and 4-chamber) views. Standard 2-dimensional and color Doppler data, triggered to the QRS complex, were saved in cineloop format. A minimum of 3 consecutive beats was recorded from each view, and the images were digitally stored for off-line analysis (EchoPac, version 7.0.0, General Electric Vingmed Ultrasound). The LV end-systolic volume (LVESV), LV end-diastolic volume, and LVEF were measured from the apical 2- and 4-chamber images, using the modified biplane Simpson’s rule.
For tissue Doppler imaging, the color Doppler frame rates were >80 frames/s in all subjects, and the pulse repetition frequencies were 500 Hz to 1 kHz, resulting in aliasing velocities of 16 to 32 cm/s. To assess LV dyssynchrony, the interval-to-peak systolic velocities was obtained by placing sample volumes in the basal portions of the septal and lateral walls. LV dyssynchrony was calculated as the maximum delay between the peak systolic velocities of these opposing basal walls. The inter- and intraobserver agreement for the assessment of the septal-to-lateral delay was 90% and 96%, respectively. On the basis of previous work, a delay of ≥65 ms was prospectively defined as substantial LV dyssynchrony.
The clinical response at 6 months after CRT was defined as an improvement of ≥1 in the NYHA class score. Additionally, the echocardiographic response at 6 months of follow-up was defined as a decrease in the LVESV of ≥15%. Patients who died before 6 months of follow-up were considered to have had both a clinical and an echocardiographic nonresponse.
Continuous data are presented as the mean ± SD and dichotomous data as numbers and percentages. A comparison of the data between the patient groups was performed using the independent samples t test for continuous data. Fisher’s exact tests or chi-square tests were used, as appropriate, to compare the dichotomous data. A comparison of data within patient groups (at baseline and at 6 months of follow-up) was performed using the paired-samples t test. Univariate and multivariate logistic regression analyses were performed to evaluate LV dyssynchrony as a predictive factor for the clinical and echocardiographic response at 6 months of follow-up. Variables that showed a statistically significant effect at the 0.50 level on the univariate analyses were entered into the multivariate model. Finally, linear regression analysis was performed to evaluate the relation between LV dyssynchrony at baseline and the echocardiographic response at 6 months of follow-up. All analyses were performed using the Statistical Package for Social Sciences, for Windows, version 16.0 (SPSS, Chicago, Illinois). A p value <0.05 was considered statistically significant. To find a significant difference as small as 25% of the group standard deviation between patient groups with a power of >0.90, the study required a minimum of 326 patients.
Results
A total of 361 consecutive patients were included in the present study. The baseline characteristics of these patients are summarized in Table 1 . The mean extent of LV dyssynchrony at baseline was 79 ± 51 ms, with 229 patients (63%) having predefined substantial LV dyssynchrony (≥65 ms). Of the 361 patients, 321 received a combined CRT–implantable cardioverter-defibrillator device, and 40 received a CRT device only.
| Variable | Value |
|---|---|
| Age (years) | 67 ± 10 |
| Gender | |
| Men | 272 |
| Women | 89 |
| New York Heart Association class | |
| III | 319 (88%) |
| IV | 42 (12%) |
| Heart failure etiology | |
| Ischemic | 221 (61%) |
| Nonischemic | 140 (39%) |
| QRS duration (ms) | 169 ± 24 |
| Left bundle branch block | 275 (76%) |
| Paced QRS | 54 (15%) |
| Other conduction disorder | 32 (9%) |
| Rhythm | |
| Sinus rhythm | 283 (78%) |
| Atrial fibrillation | 58 (16%) |
| Paced | 20 (6%) |
| Systolic blood pressure (mm Hg) | 120 ± 21 |
| Diastolic blood pressure (mm Hg) | 72 ± 11 |
| Serum creatinine (μmol/L) | 131 ± 69 |
| Left ventricular ejection fraction (%) | 23 ± 7 |
| Left ventricular end-diastolic volume (ml) | 239 ± 85 |
| Left ventricular end-systolic volume (ml) | 187 ± 75 |
| Left ventricular dyssynchrony (ms) | 79 ± 51 |
| Left ventricular dyssynchrony ≥65 ms (n) | 229 (63%) |
| Medication | |
| Diuretics | 329 (91%) |
| Angiotensin-converting enzyme inhibitors | 328 (91%) |
| β Blockers | 224 (62%) |
| Spironolactone | 188 (52%) |
| Digoxin | 88 (24%) |
| Calcium antagonists | 26 (7%) |
After 6 months of follow-up, the mean NYHA class had improved from 3.1 ± 0.3 to 2.2 ± 0.6 (p <0.001). In addition, the quality-of-life score had decreased from 42 ± 16 to 27 ± 18, and the 6-minute walking distance had increased from 268 ± 106 to 369 ± 128 m (both p <0.001). Also, echocardiographic improvement was noted after 6 months, as demonstrated by a decrease in the LV volumes (the LVESV decreased from 187 ± 75 to 150 ± 71 ml and the LV end-diastolic volume decreased from 239 ± 85 to 207 ± 80 ml, both p <0.001). LV systolic function had improved, as shown by an increase in the LVEF from 23 ± 7% at baseline to 30 ± 11% at follow-up (p <0.001). Of the 361 patients, 22 had died from worsening heart failure before reaching 6 months of follow-up; these patients were considered to have not had either a clinical or an echocardiographic response.
After 6 months of CRT, 259 patients (72%) were considered to have had a clinical response. The patients with and without a response had comparable baseline characteristics, except for more extensive LV dyssynchrony (91 ± 49 ms vs 50 ± 44 ms, p <0.001) in those responding ( Table 2 ). By definition, those with a response had a decrease in NYHA class from 3.1 ± 0.3 at baseline to 1.9 ± 0.4 at follow-up. Similarly, an increase in the 6-minute walking distance was observed from 275 ± 104 m at baseline to 403 ± 106 m at follow-up. Also, the quality-of-life score decreased in those with a response (from 41 ± 16 at baseline to 21 ± 14 at follow-up, p <0.001), but remained unchanged in those without a clinical response (46 ± 16 at baseline vs 46 ± 17 at follow-up, p = NS). As demonstrated in Figure 1 , those with a clinical response showed a significant increase in LVEF, with a significant reduction in LV volumes. These parameters did not improve in those without a clinical response. The percentage of biventricular pacing at follow-up was comparable between those with and without a response (98 ± 5% in those responding vs 98 ± 3% in those not responding; p = 0.415).
