Article Text

Download PDFPDF
Cardiac imaging and non-invasive testing
Triplane tissue Doppler imaging: a novel three-dimensional imaging modality that predicts reverse left ventricular remodelling after cardiac resynchronisation therapy
  1. N R Van de Veire1,2,
  2. C-M Yu3,
  3. N Ajmone-Marsan1,
  4. G B Bleeker1,
  5. C Ypenburg1,
  6. J De Sutter2,
  7. Q Zhang3,
  8. J W H Fung2,
  9. J Y S Chan2,
  10. E R Holman1,
  11. E E van der Wall1,
  12. M J Schalij1,
  13. J J Bax1
  1. 1
    Department of Cardiology, Leiden University Medical Centre, Leiden, The Netherlands
  2. 2
    Ghent University, Gent, Belgium
  3. 3
    Prince of Wales Hospital, Chinese University of Hong Kong, Hong Kong
  1. Dr N R Van de Veire, Department of Cardiology, Leiden University Medical Centre/Ghent University, Albinusdreef 2, 2333 ZA Leiden, The Netherlands; nico.vandeveire{at}ugent.be

Abstract

Background: Several two-dimensional (2-D) tissue Doppler imaging (TDI) echocardiographic techniques have proved useful to identify responders to cardiac resynchronisation therapy (CRT). Recently a 3-D probe allowing simultaneous acquisition of TDI data in three imaging planes became available.

Objective: To evaluate the value of triplane TDI to predict reverse left ventricular (LV) remodelling after CRT.

Methods: Sixty patients with heart failure, scheduled for CRT, underwent triplane echocardiography with simultaneous TDI acquisition before and 6 months after implantation. From the triplane dataset a 3-D LV volume was generated and LV volumes and ejection fraction were calculated. Intraventricular dyssynchrony was quantitatively analysed by evaluating time from onset of the QRS complex to peak myocardial systolic velocity in 12 LV segments from the triplane dataset and calculation of the standard deviation (Ts-SD-12). Clinical response was defined as an improvement of at least one New York Heart Association class. Reverse LV remodelling was defined as ⩾15% decrease of LV end-systolic volume at 6 months’ follow-up.

Results: Responders to CRT had significantly more LV dyssynchrony at baseline than non-responders (mean (SD) Ts-SD-12: 42 (14) vs 22 (12), p<0.001). A cut-off value of 33 ms for baseline Ts-SD-12, acquired from the triplane TDI dataset, yielded a sensitivity of 89% with a specificity of 82% to predict clinical response to CRT; sensitivity and specificity to predict reverse LV remodelling were 90% and 83%, respectively.

Conclusion: Triplane TDI echocardiography predicts clinical response and reverse LV remodelling 6 months after CRT implantation.

Statistics from Altmetric.com

Cardiac resynchronisation therapy (CRT) is now accepted as an alternative therapeutic option in patients with heart failure who remain highly symptomatic despite optimised medical treatment.1 However, using classic selection criteria for CRT including New York Heart Association (NYHA) class III or IV, impaired left ventricular (LV) systolic function and conduction delay on the surface electrocardiogram (ie, wide QRS complex >120 ms), up to 30% of patients undergoing CRT implantation do not respond favourably.2 The presence of LV dyssynchrony is considered a key factor in the identification of potential responders to CRT.3 4 Numerous two-dimensional (2-D) echocardiographic methods, often using tissue Doppler imaging (TDI), have been proposed to quantify LV dyssynchrony and predict response to CRT.510 The TDI method originally described by Yu and coworkers is a standard deviation of electromechanical activation times (Ts-SD-12) based on a 12-segment model of the left ventricle.7 11 To calculate this dyssynchrony measure, three different apical views need to be acquired separately, non-simultaneously, thus neglecting heart rate variability. Recently, a three-dimensional (3-D) TDI imaging modality, triplane TDI, became available which permits simultaneous acquisition of TDI from all LV segments during the same heartbeat. In addition, the triplane technique allows calculation of 3-D volumes and LV ejection fraction. No published data are available on the potential role of triplane TDI to predict response to CRT.

In this study, this new 3-D echocardiographic technique, triplane TDI, was applied to 60 patients with heart failure to quantify LV volumes, ejection fraction and LV dyssynchrony at baseline and at 6 months’ follow-up. The aim of the study was to assess the value of this triplane approach to predict clinical response and reverse LV remodelling, 6 months after CRT implantation.

PATIENTS AND METHODS

Study group

The study group comprised 60 consecutive patients with heart failure scheduled for CRT. Inclusion criteria were severely symptomatic heart failure despite optimal medical treatment (NYHA class III or IV), depressed LV ejection fraction and wide QRS complex (>120 ms, with left bundle branch block or interventricular conduction delay) on the surface electrocardiogram. Patients with atrial fibrillation or a previously implanted pacemaker were excluded.

Study protocol

Within 24 hours before CRT implantation, all patients underwent the following examinations: 12-lead electrocardiogram and extensive triplane transthoracic echocardiography as described below. NYHA functional class was assessed by a clinician who was unaware of all other data. Subjects completed the Minnesota Living with Heart Failure questionnaire, a 21-question self-administered instrument with scores ranging from 0 to 5 for each question; higher scores indicating poorer quality of life.12 Exercise capacity was assessed using the 6-minute hall walk test.13 Six months after implantation of the CRT device, all patients underwent the same examinations including triplane echocardiography.

Triplane TDI echocardiography

Acquisition of the three-dimensional dataset

Studies were performed with a commercially available echocardiographic platform (VIVID 7, GE Vingmed Ultrasound, Horten, Norway), equipped with a 3V-probe for triplane acquisition. Patients were scanned in the left lateral decubitus position, from the apical window. Care was taken to visualise the true LV apex allowing simultaneous acquisition of the apical four-, two- and three-chamber views. Sector size and depth were adjusted to achieve optimal visualisation of the six LV walls at the highest possible frame rate (ideally at least 100 frames/second). Colour-coded TDI was then applied to the triplane view to assess longitudinal myocardial regional function. Gain settings, filters and pulse repetition frequency were adjusted to optimise colour saturation. Patients were asked to hold their breath and at least three consecutive sinus beats were recorded; the images were digitally stored for offline analysis (EchoPac, GE Vingmed Ultrasound, Horten, Norway).

Quantification of LV volumes and ejection fraction using triplane technology

During post-processing, the triplane dataset was frozen in end diastole and the endocardial border was manually traced in the apical four-, two- and three-chamber, views respectively. To enhance endocardial border delineation, the software allows separate visualisation of the three apical views. Then, using the same heartbeat, the triplane dataset was frozen in end systole and again the endocardial border was manually traced in the apical four-, two- and three-chamber views. A 3-D LV end-diastolic and end-systolic volume was generated automatically by the software and LV volumes and ejection fraction were reported accordingly (fig 1).

Figure 1 Example of three dimensional left ventricular (LV) volumes, generated from a triplane dataset, at baseline (A) and at 6 months’ follow-up (B). Note the significant decrease of LV volumes and the increase in LV ejection fraction.

Quantitative analysis of LV dyssynchrony

During post-processing, the triplane colour-coded TDI dataset was used to analyse myocardial velocity curves from 12 LV segments derived from information acquired during the same heartbeat. Sample volumes were placed in the basal and mid-segments of the septal, lateral, inferior, anterior, posterior and anteroseptal LV walls of the triplane dataset. The software generates ECG gated myocardial velocity curves, allowing calculation of time from the beginning of the QRS complex to peak myocardial systolic velocity (Ts). From these 12 measurements a standard deviation was calculated (Ts-SD-12) as a measure of intraventricular dyssynchrony. Results are reported in a bulls’ eye plot. Figure 2 provides a patient example.

Figure 2 Example of myocardial velocity curves at the basal septum and lateral wall of the left ventricle derived from a triplane dataset. At baseline there is a significant delay between the septum and lateral wall (A). The standard deviation (B) of all 12 LV segments is 92, indicating severe LV dyssynchrony. At 6 months’ follow-up, the septum and lateral wall are activated simultaneously (C) and LV dyssynchrony has significantly improved (D).

Validation of the triplane technique

In 10 randomly chosen patients, LV dyssynchrony was measured as originally reported by Yu et al7, using separately recorded 2-D images from the apical four-, three- and two-chamber views. Ts was measured in 12 LV segments and Ts-SD-12 was calculated, allowing comparison with the triplane method to obtain these data.

Implantation of CRT device

The LV pacing lead was inserted transvenously through the subclavian route. First, a coronary sinus venogram was obtained during occlusion of the coronary sinus with a balloon catheter. Next, the LV pacing lead was inserted into the coronary sinus with the help of an 8F guiding catheter and positioned as far as possible in the venous system, preferably in the (postero-) lateral vein. The right atrial and ventricular leads were positioned conventionally. When a conventional indication for an internal defibrillator existed, a CRT-D device was implanted. At implantation, both the sensing and pacing thresholds (at pulse duration of 0.5 ms) of the LV pacing lead were measured.

Definition of response to CRT

Patients were subsequently divided into clinical responders and non-responders, based on an improvement in NYHA functional class by ⩾1 score, 6 months after implantation. Echocardiographic response was defined as a reduction of at least 15% in LV end-systolic volume (reverse LV remodelling) at 6 months’ follow-up.14

Statistical analysis

All analyses were performed with the statistical software program SPSS 12.0.1 (SPSS Inc, Chicago, IL, USA). Continuous data are presented as mean (SD). Categorical data are presented as absolute numbers or percentages. In the validation group, 2-D and triplane Ts data were compared using the Pearson correlation analysis. The Bland–Altman graph was used to compare Ts-SD-12 obtained with the conventional and triplane method. Comparisons between responders and non-responders were made using the independent-samples t-test; comparisons between pre- and postimplantation characteristics were performed using the paired-samples t-test. Correlations between changes in dyssynchrony and LV end-systolic volume and ejection fraction are graphically depicted using simple regression analysis. Receiver operating characteristic (ROC) curves were also analysed to determine the ability of Ts-SD-12 to predict clinical response and reverse LV remodelling after CRT. From these ROC curves, the area under the curve was calculated and presented with 95% confidence intervals (95% CIs). A p value <0.05 was considered significant.

RESULTS

Study group

A total of 60 consecutive patients (47 men, 13 women) undergoing CRT implantation were included in the study. Mean age was 66 (11) years, 37 patients (62%) had ischaemic cardiomyopathy, 23 (38%) had non-ischaemic cardiomyopathy. Moreover, 54 patients (90%) were in NYHA class III and 6 (10%) in NYHA class IV. All patients were receiving optimised medical treatment including β-blockers (62%), angiotensin converting enzyme inhibitors (85%), diuretics (84%) and aldactone (39%). Mean QRS width was 146 (32) ms. Using triplane technology, mean LV end-diastolic volume was 207 (74) ml, mean LV end-systolic volume 160 (71) ml and mean LV ejection fraction 25 (10)%.

Validation of the triplane technique

Consistent correlations were found between the Ts data measured using a conventional 2-D method and Ts data measured using triplane technology; r ranging between 0.94 and 0.98 (p<0.001). Bland–Altman analysis shows excellent limits of agreement between LV dyssynchrony (Ts-SD-12) measured by 2-D analysis and LV dyssynchrony measured by triplane technology (fig 3).

Figure 3 Bland–Altman plot comparing left ventricular dyssynchrony (Ts-SD-12) measured by a conventional two-dimensional method and triplane technology. SD, standard deviation.

Pacemaker implantation

CRT device and lead implantation was successful in all patients without major complications (Contak Renewal, Guidant, and Insync Sentry, Medtronic Inc). Two types of LV lead were used: Easytrack (Guidant, St Paul, Minnesota, USA) or Attain (Medtronic Inc, Minneapolis, Minnesota, USA). In 83% of cases a CRT-D device was implanted.

Clinical response 6 months after CRT implantation

At 6 months’ follow-up, 63% of the patients showed clinical improvement as defined by an increase of NYHA class by at least one score. The only pre-implantation characteristic differing significantly between responders and non-responders was the extent of LV dyssynchrony as assessed with triplane TDI (42 (14) vs 22 (12), p<0.001). Besides an improvement of NYHA class (3.1 (0.3) at baseline, 1.7 (0.5) at follow-up, p<0.001), clinical responders also showed improvement in 6-minute walking distance (335 (104) m at baseline vs 388 (79) m at follow-up, p = 0.001) and quality of life score (32 (19) at baseline vs 22 (17) at follow-up, p = 0.001). LV dyssynchrony Ts-SD-12 improved from 42 (14) to 23 (12) (p<0.001).

Reverse LV remodelling at 6 months’ follow-up

A total of 35 (58%) patients exhibited a reduction of LV end-systolic volume ⩾15% within 6 months after CRT implantation (fig 4). Patients with reverse LV remodelling showed not only an improvement of LV end-systolic volume (−48.6 ml on average) but also a significant decrease of LV end-diastolic diameter (−31.9 ml on average) and an increase of LV ejection fraction (+ 13.4% on average). Table 1 reports the baseline clinical and echocardiographic characteristics of patients without and with reverse LV remodelling.

Figure 4 Change in left ventricular (LV) volumes and LV ejection fraction at baseline and at 6 months’ follow-up in patients exhibiting reverse remodelling (A, C) and those without reverse remodelling (B, D).
Table 1 Baseline characteristics of patients without and with reverse left ventricular (LV) remodelling at 6 months’ follow-up

The only pre-implantation characteristic differing significantly between patients with and without reverse LV remodelling was the extent of LV dyssynchrony as assessed with triplane TDI. Patients with reverse LV remodelling at 6 months’ follow-up also exhibited a significant decrease of the degree of LV dyssynchrony (a mean reduction of 22 ms as compared to a mean increase of 1.7 ms in non-responders, p<0.001).

Figure 5 shows a relation between the decrease in LV dyssynchrony and the decrease in LV end-systolic volume or the increase in LV ejection fraction.

Figure 5 Scatter plots showing the relation between the decrease in left ventricular dyssynchrony (quantified as Ts-SD-12) and (A) decrease in left ventricular end-systolic volume (LVESV); (B) increase in left ventricular ejection fraction (LVEF), 6 months after cardiac resynchronisation therapy. A line is fitted using simple regression analysis and the r2 values are reported, respectively.

Prediction of clinical response after CRT implantation by triplane TDI

Overall, Ts-SD-12 measured with triplane technology was significantly larger in clinical responders than in non-responders, indicating severe LV dyssynchrony in the clinical responders. Figure 6A shows the ROC curve illustrating the excellent ability of Ts-SD-12, measured with triplane TDI technology, to predict clinical response. An ideal cut-off value for Ts-SD-12 to predict clinical response 6 months after CRT implantation, was defined at 33 ms based on the highest possible sensitivity (89%) and specificity (82%).

Figure 6 (A) Receiver operating characteristic (ROC) curve for Ts-SD-12 predicting clinical response 6 months after cardiac resynchronisation therapy (CRT) implantation; the area under the curve (AUC) is indicated with the 95% confidence interval. (B) ROC curve for Ts-SD-12 predicting reverse LV remodelling (defined as at least 15% reduction in LV end-systolic volume) 6 months after CRT implantation; the AUC is indicated with the 95% confidence interval.

Prediction of reverse LV remodelling after CRT implantation by triplane TDI

Overall, Ts-SD-12, measured with triplane technology, was significantly larger in patients with reverse LV remodelling than in those without reverse LV remodelling, indicating severe LV dyssynchrony in patients with reverse remodelling. Figure 6B shows the ROC curve illustrating the excellent value of Ts-SD-12, measured with triplane TDI technology, to predict reverse LV remodelling. An ideal cut-off value for Ts-SD-12 to predict reverse LV remodelling at 6 months’ follow-up, was defined at 33 ms based on the highest possible sensitivity (90%) and specificity (83%).

DISCUSSION

The findings in this study can be summarised as follows: (a) the triplane echo approach permits integrated assessment of LV volumes, LV ejection fraction and LV dyssynchrony; (b) in patients undergoing CRT, LV dyssynchrony was the only predictor of clinical response to CRT; (c) the standard deviation of time to peak myocardial systolic velocity in 12 LV segments (Ts-SD-12) ⩾33, calculated from the triplane TDI dataset, had a sensitivity of 90% with a specificity of 83% to predict reverse LV remodelling, at 6 months after CRT implantation.

Several authors have emphasised the importance of quantifying the degree of LV dyssynchrony in candidates for CRT, to identify potential responders. Earlier studies proposed methods using M-mode5 or pulsed-wave TDI,15 16 but recently, colour-coded TDI6 7 10 and techniques derived from colour-coded TDI such as tissue synchronisation imaging17 and strain rate imaging8 9 have gained interest. The majority of colour-coded TDI techniques are based on the measurement of time intervals between onset of the QRS complex and peak myocardial systolic velocity. Some authors have compared the activation times of the septal and lateral LV segments; for this approach TDI registration of the apical four-chamber view, using 2-D technology, is sufficient.6

Others used information derived from various LV segments. Yu and coworkers originally described the model (derived from 2-D data) using electromechanical activation times of 12 LV segments to quantify LV dyssynchrony.7 11 A potential problem with the 2-D approach if one uses the 12-segment model, is the cumbersome data acquisition. At least three separate colour-coded 2-D datasets have to be acquired, non-simultaneously. This implies that heart rate variability is often neglected, resulting in non-simultaneous comparison of segmental coordination. Moreover, in all CRT studies assessing LV remodelling, LV volumes and LV ejection fraction were estimated from 2-D datasets with Simpson’s biplane method.

Triplane imaging is a 3-D technique that integrates data from three conventional apical views. Validation studies have demonstrated that the triplane approach is better than the biplane method and comparable with magnetic resonance imaging for LV volume estimation.18 19 Recently, a 3-D probe became commercially available (GE Vingmed Ultrasound, Horten, Norway) allowing simultaneous acquisition of a triplane dataset and colour-coded TDI. In the present study, we have demonstrated the multipurpose of this triplane TDI dataset in 60 patients with heart failure scheduled for CRT implantation. A 3-D LV volume was generated, allowing evaluation of volumetric changes 6 months after implantation, showing reverse LV remodelling in responders.

Furthermore, an established 2-D TDI-based method7 11 to quantify LV dyssynchrony using a 12-segment LV model was successfully applied to the triplane dataset. Sample volumes were placed in 12 LV segments and myocardial peak systolic velocities or time to peak systolic velocities of any LV segment could be compared during the same heartbeat. From these data, the Ts-SD-12 was calculated.

Using 2-D technology, Yu et al applied the Ts-SD-12 in 30 patients undergoing CRT implantation and reported that a cut-off value >32.6 ms permitted separation of responders and non-responders to CRT.7 11 In a subsequent study the same authors used TSI in a 2-D dataset to quantify LV dyssynchrony.17 The Ts-SD-12 was most powerful to predict reverse LV remodelling; ROC curve analysis disclosed an area of 0.90 with a sensitivity of 87% and a specificity of 81% employing a cut-off value of 34.4 ms. In the present study, the Ts-SD-12 was calculated from the triplane dataset. An ROC curve area of 0.95 was shown with a sensitivity of 90% and a specificity of 87% for a cut-off value of 33 ms, well in line with the above-mentioned 2-D studies. The advantage of the triplane method is that acquisition of a single triplane dataset allows simultaneous comparison of 12 LV segments during the same heartbeat whereas the 2-D method requires at least three acquisitions. Applying the triplane technique reduces acquisition time, eliminates potential heart rate variability and offers a more reliable assessment of LV volumes, which is essential to evaluate LV reverse remodelling. A potential drawback of the triplane method is the higher frame rate, especially in extremely dilated hearts, making this a less suitable technique for TDI-derived strain and strain rate analysis. If strain analysis is the goal, than LV segments should be acquired separately, using conventional methods, in order to minimise angle dependency.

Experience with 3-D echocardiographic evaluation of patients undergoing CRT implantation is limited. Kapetanakis et al used real-time 3-D echocardiography and calculated a systolic dyssynchrony index derived from dispersion of time to minimal regional volume in 26 patients undergoing CRT implantation.20 The authors noted that patients with a large systolic dyssynchrony index before implantation experienced the greatest reduction in mechanical dyssynchrony after CRT, whereas patients without symptomatic improvement had a statistically lower systolic dyssynchrony index before CRT implantation. The method described by Kapetanakis et al20 is based on identification of the last LV region reaching minimum regional volume; the triplane method is based on the assessment of LV dyssynchrony from TDI velocity curves, currently the most frequently used echocardiographic method to predict benefit from CRT.4

Limitations

Patients with atrial fibrillation and previous pacemakers were excluded; further studies are needed in these specific patient groups to explore the value of triplane TDI.

CONCLUSION

In this study, a new 3-D echocardiographic application, allowing simultaneous acquisition of a triplane TDI dataset, was successfully applied in 60 patients with heart failure scheduled for CRT. From the triplane TDI dataset, quantitative analysis of LV dyssynchrony could be derived as well as LV volumes and ejection fraction. LV dyssynchrony, derived from the triplane TDI dataset, was highly predictive for clinical response and LV reverse remodelling after 6 months of CRT.

Obtained.

REFERENCES

Footnotes

  • Competing interests: None declared.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.