1. Cardiac Resynchronization Therapy
Alena Goldman, MD
11/7/07
Harvard Medical School

2. Rationale for CRT
Bundle branch block or other intraventricular conduction delay can worsen HF due to systolic dysfunction
Electrical ventricular dyssynchrony common in advanced HF; correlated with increased mortality
Initial theory behind use of CRT was an idea that hemodynamic benefits follow the correction of dyssynchrony with CRT
CRT was developed in the early 90s and was FDA approved as an adjunctive therapy for severe systolic HF in 2001

3. CRT: Moderate to severe systolic heart failure with wide QRS
Jessup M, Brozena S. Medical Progress--Heart Failure. N Eng J Med 2003; 348: Copyright 2002 Massachusetts Medical Society. All rights reserved.

4. Prevalence of Electrical Ventricular Dyssynchrony in Heart Failure
Masoudi and colleagues used retrospective medical chart data of 19,710 pts Medicare beneficiaries hospitalized w/ HF and for whom LV systolic function was confirmed. LBBB present in 8% of those with preserved LV systolic function (diastolic HF) and in 24% of those with EF < 50% (p<0.001).
Aaronson developed and validated a multivariable survival model for ambulatory advanced heart failure patients wait listed for a heart transplant. IVCD (QRS > 120 ms) present in 27% of the 268 pts in derivation sample, and in 53% of the 199 pts in validation sample. IVCD identified as contributing risk factor.
Other studies have shown that fro the entire HF population about 15% have a wide QRS.
1. Masoudi, et al. JACC 2003;41:217-23
2. Aaronson, et al. Circ 1997;95:2660-7

5. Types of Dyssynchrony Mechanical: contractile dyscoordination
Electrical: QRS width
Cause and effect relationship: Electrical dyssynchrony leads to inefficient contraction (exception when mechanical dyssynchrony is present despite normal QRS width)

6. Mechanisms of Mechanical Dyssynchrony
Interventricular dyssynchrony: RV contracts before LV; affects septal contribution to LV stroke volume
Intraventricular dyssynchrony: septum contracts before the lateral wall (lateral wall can contract in early diastole); early contraction is ineffective and late contraction stretches early contracting segments
Atrioventricular dyssynchrony
Negative LV remodeling: increased LVESV/increased wall stress/increased demand/ reduced contractility worsening LV systolic function
Impaired relaxation: LV diastolic dysfunction
Mitral regurgitation

7. Discoordinate Motion
Adverse Effects on Global Function From RV-Pacing–Induced Dyssynchrony
Normal Sinus Rhythm
Acute Dyssynchrony (RV Pace) 80 40
LV Pressure (mm Hg)
This figure depicts the effect of chamber dyssynchrony induced by acute right ventricular (RV) pacing in a normal canine heart. Left ventricular (LV) pressure-volume (P-V) loops and end-systolic P-V relations are displayed for synchronous and acutely dyssynchronous conditions. With RV pacing to induce dyssynchrony, there is a right shift of the P-V loop, with reduced width (stroke volume) and area (stroke work), and a right shift in the end-systolic P-V point (green dot; increased end-systolic stress).1 30 60 90
LV Volume (mL)
Adapted from Kass DA. Rev Cardiovasc Med. 2003;4(suppl 2):S3-S13.
Reference:
Kass DA. Ventricular resynchronization: pathophysiology and identification of responders. Rev Cardiovasc Med. 2003;4(suppl 2):S3-S13.

8. CRT: Rationale CRT resynchronizes contraction
Improves contractile LV function
Is associated with reverse ventricular remodeling
Improves CO/CI; reduces PCWP
Improves diastolic function
Reduces frequency of ventricular arrythmias and ICD therapies
Increases HRV
Improves NYHA Class symptoms: QOL, exercise capacity, functional capacity
Reduces mortality, due to both HF and SCD (Care-HF)

9. Achieving Cardiac Resynchronization
Goal: Atrial synchronous
biventricular pacing
Transvenous approach for left ventricular lead via coronary sinus
Back-up epicardial approach
Right Atrial Lead
Right Ventricular Lead
Left Ventricular Lead
Main purpose: Illustrate for referral clinicians how the leads are placed to achieve cardiac resynchronization. Many outside the implant world may not be entirely aware of how the device is placed.
Key messages:
The implant procedure, while typically of longer duration, is similar to that of a standard pacemaker or implantable defibrillator implantation.
A key difference is the placement of a left ventricular lead via the coronary sinus opening.
Coronary venous anatomy varies significantly between patients. In a small percentage of cases it may not be possible to place the left ventricular lead transvenously. Some centers are opting for an epicardial approach if the transvenous approach is unsuccessful.
Additional information:
Standard pacing leads are placed in the right atrium and right ventricle. The LV lead is placed via the coronary sinus in a cardiac vein, preferably a lateral or postero-lateral vein in the mid part of the LV. The successful deployment of this lead to physician-guided development of left-heart delivery systems, and new LV leads to meet varying patient

10. Regional Wall Motion With CRT: Improved LVEF
Septum
Seconds
0.4
Regional Fractional Area Change
Lateral
These echocardiogram (ECHO) and radial displacement tracings of the septal and lateral walls show how regional wall motion is improved by CRT. With pacing off, radial septal motion is initially inward but then shifts toward the right ventricle as the lateral wall contracts (paradoxic motion). CRT converts this to a more consistent inward motion. In the lateral wall, there is initial stretch followed by delayed contraction. CRT influences the phase, but not amplitude of motion; this stimulates contraction earlier, resulting in increased cardiac efficiency.1
Seconds
0.4
Pacing Off
Pacing On
Adapted from Kass DA. Rev Cardiovasc Med. 2003;4(suppl 2):S3-S13.
Adapted from Kawaguchi M, et al. J Am Coll Cardiol. 2002;39:
References:
Kass DA. Ventricular resynchronization: pathophysiology and identification of responders. Rev Cardiovasc Med. 2003;4(suppl 2):S3-S13.
Kawaguchi M, et al. Quantitation of basal dyssynchrony and acute resynchronization from left or biventricular pacing by novel ECHO-contrast variability imaging. J Am Coll Cardiol. 2002;39:

11. Ventricular Reverse Remodeling With Resynchronization
7.5
P<0.001
P<0.001 30
End-Diastolic Dimension (mm)
6.5
Ejection Fraction (%) 20
Chronic CRT significantly improves chamber function and results in a modest decline in cardiac volumes in patients with moderate-to-severe HF. The data above are from the MIRACLE trial (Multicenter InSync Randomized Clinical Evaluation), demonstrating significant reduction of end-diastolic dimension and increase in ejection fraction in patients randomized to active CRT therapy during the initial 6-month period. In contrast, patients receiving the implant but in a nontherapeutic atrial pacing mode displayed no changes. This is consistent with positive reverse remodeling of the LV resulting from CRT.1
6.0 10
Placebo
n=63
CRT
n=61
Placebo
n=81
CRT
n=63
CRT 6-month
Control
6-month
CRT
Adapted from Abraham WT, et al. N Engl J Med. 2002;346:
Reference:
Abraham WT, Fisher WG, Smith AL, et al. Multicenter InSync randomized clinical evaluation. Cardiac resynchronization in chronic heart failure. N Engl J Med. 2002;346:

12. Improvement with CRT - MR

13. AV Interval Optimization
AV delay (0 to PR – 30 msec)
LV BV
Change in Aortic PP (%)
Change in dP/dtmax (%) 24 18 12 6
-12 -6 16 8 4 -8 -4 1
Adapted from Auricchio A, et al. Circulation. 1999;99:

14. Synchronous vs Non-Synchronous BV Pacing: Is RV-LV Delay Important?6 * * 5 4
Systolic Function (Echo Index) 3
The figure shows average global systolic contraction amplitude (GSCA), a measure of chamber systolic function, and suggests that optimal CRT can be achieved by slightly advancing the pacing of the LV versus the RV lead. Marked prestimulation of the LV lead in this study appeared to lower the efficacy of CRT, although this has not yet been confirmed by other studies. In this regard, the decline associated with very premature LV stimulation is somewhat in conflict with data showing that LV-only pacing is as good as or better than BV pacing on systolic performance.1 2 1
RV Preactivation S LV Preactivation
* P<0.01 vs. Simultaneous (s) Sogaard P, et al. Circulation. 2002;106:
Reference:
Sogaard P, Egeblad H, Pedersen AK, et al. Sequential versus simultaneous biventricular resynchronization for severe heart failure: evaluation by tissue Doppler imaging. Circulation. 2002;106:

15. Mortality/Morbidity From Published Randomized, Controlled Trials
Risk reduction with CRT
Study
(n random.)
Follow-up
Mor-tality & Hosp.
Mortal. & HF Hosp.
Mor-tality
HF Mort.
HF Hosp.
MIRACLE1 (n=453)
6 Mo NR
39%*
27%
50%*
MIRACLE ICD2 (n=369) 2% 0%
Contak CD3 (n=490)
3-6 Mo
30%
18%
Meta-analysis4 (n=1634)
23%
51%*
29%*
Main purpose: Illustrate improvements in mortality and hospitalization with CRT and CRT + ICD on top of optimal medical therapy
Key messages:
Neither MIRACLE, nor MIRACLE ICD, nor Contak CD were powered to detect differences in mortality or hospitalizations.
Additional information:
Reprints of MIRACLE: UC EN
Reprints of JAMA meta-analysis: UC EN
1. Abraham WT, et al. N Engl J Med 2002;346:
2. Young JB, et al. JAMA 2003;289:
3. Higgins SL, et al. JACC 2003;
4. Bradley DJ, et al. JAMA 2003;289: [Includes MIRACLE, MIRACLE ICD, Contak CD, and MUSTIC studies]
* P < 0.05
NR = Not reported in publication
Individual trials were not powered for mortality or hospitalization

16. Cumulative Enrollment in Cardiac Resynchronization Randomized Trials
Main purpose: Show that a large number of patients have been studied in completed and ongoing randomized controlled studies of CRT. Use in conjunction with previous slide.
Key messages:
Over 3000 patients have been enrolled in randomized controlled clinical trials presented to date.
When CARE-HF, another landmark trial assessing mortality and hospitalization, is reported, close to 4,000 patients will have been studied.

17. Patient selection
Current recommendations for bi-ventricular pacing are based on evidence of electrical (NOT mechanical) dyssynchrony

18. Can We Predict Responders?
Electrical dyssynchrony/Wide QRS complex
Widely used, but only broadly correlates with acute response
Weak predictor of chronic response
Mechanical dyssynchrony
More direct target of CRT
Used to follow responce
Measures of wall dyssynchrony (MRI, ECHO, TDI) best correlate with acute and chronic responsiveness
Several different criteria are currently used to help identify patients most likely to respond to CRT. The marker most widely used to select candidates for CRT is a widened QRS complex on the surface electrocardiogram (ECG). While there is a general correlation between basal QRS duration and treatment efficacy, even acutely this measure has poor predictive value for identifying responders versus nonresponders, and chronic predictive value appears to be even less. Mechanical dyssynchrony can be quantified more directly with MRI, echocardiography, or tissue Doppler imaging (TDI), and all of these measures correlate well with both acute and chronic responses to CRT. Low contractile state and marked P-R delay are additional features that are likely to predict a positive response to CRT.1
Kass DA. Rev Cardiovasc Med. 2003;4(suppl 2):S3-S13.
Reference:
Kass DA. Ventricular resynchronization: pathophysiology and identification of responders. Rev Cardiovasc Med. 2003;4(suppl 2):S3-S13.

19. Who Responds to Cardiac Resynchronization?
Responder Parameter(s)
Finding
Limitation(s)
NYHA III/IV, QRS 120 ms, EF 35%, LVEDD 55 mm
Confirmed in RCTs of over 2,500 patients
~ 70% respond favorably
QRS  150/155 and/or dP/dt  700 mm Hg/s
Correlated with improved dP/dt 1,2
Small studies, < 30 pts;
No clinical endpoint
not confirmed by MIRACLE
Difference in time to peak systolic contraction
Correlated with 
volumes 3,4,5
Small studies,  30 pts;
Varying techniques
No MI, significant mitral regurgitation
Correlated with improved NYHA6
Observational study;
Main purpose: Summarize clinical results of studies done to identify responders to CRT
Key messages:
The 2,500 + patients included in randomized controlled studies of CRT have already been sub-selected from the heart failure population to include those with systolic dysfunction, moderate to severe symptoms, dilated left ventricles, and importantly, a wide QRS as a measure of ventricular dysynchrony.
Studies using other methods to further sub-select responders have been small in scope and used varying techniques.
A multivariate analysis of MIRACLE results shows that only the existence of CRT predicts response.
A larger scale multi-center trial is underway to assess the potential of echo based measures of dysynchrony to better identify those most likely to respond.
1. Circulation. 2000;101:
2. Circulation 1999;99:
3. Am J Cardiol 2002;91:684–688
4. J Am Coll Cardiol 2002;40:
5. J Am Coll Cardiol 2002;40:723–730
6. Am J Cardiol 2002;89:

20. Summary of Major Trials
Significant clinical benefit of CRT in patients with class III-IV HF, low EF, and QRS > 120
Improvement in symptoms
Improvement in objective standards of HF
Meta-analysis
29% decrease in HF hospitalization (13% vs. 17.4%)
51% decrease in deaths from HF (1.7% vs. 3.5%)
Trend toward decrease in overall mortality (4.9% vs 6.3%)
BUT: >30% non-responders consistent through most trials
Bradley et al. JAMA 2003;289:730

21. Targeting Electrical Dyssynchrony: QRS Duration
Pros:
QRS >120 ms
LBBB>RBBB
Correlation between QRS and response to CRT modest (r2 = 0.6)
Cons:
Evidence of LV dyssynchrony with QRS < 120
Small trial in patients with QRS < 120 suggest these patients may also benefit from CRT

22. Imaging Measures of Mechanical Dyssynchrony:
20-30% of patients with evidence of electrical dyssynchrony do not benefit from CRT regardless of baseline QRS duration and QRS narrowing with CRT
Imaging allows direct visualization of mechanical dyssynchrony

23. Imaging Techniques M Mode and 2D TDI with echo
Myocardial strain imaging
3D Echo
CMR

24. Other Modalities
Electrical activation pattern during bi-V pacing by EP mapping
Delta QRS during bi-V pacing

25. M-mode Echo Interventricular dyssynchrony/motion delay – IVMD
Time difference between left and right pre-ejection intervals
IVMD ≥ 50 ms

26. M Mode Echo Intraventricular Dyssynchrony
SPWMD predicts improvement with CRT
(in 25 patients)
Intraventricular Dyssynchrony
Septal-to-posterior wall motion delay (SPWMD)
SPWMD ≥ 130 ms
+20
r =-.70 P=.001
-20
-40
 LVESVI (mL/m2)
-60
-80
-100 D 20 60
140
220
300
380
SPWMD (msec)
Adapted from Pitzalis MV, et al. J Am Coll Cardiol. 2002;40:

27. TDI Imaging PW Doppler Reflects regional systolic velocity
Timed to the QRS
Dyssynchrony criteria:
12 sample volume model (any 2 > 100 ms, SD > 33ms)
2 sample volume – basal septum and lateral wall delay ≥ 50ms
Interventricular delay ≥ 50ms

28. TDI Assessment for Predicting Responders
Adapted from Sogaard P, et al. J Am Coll Cardiol. 2002;40:

29. 85 patients with severe HF, LBBB, QRS duration > 120 ms
Patients with Intraventricular LV Dyssynchrony of ≥ 65 ms Have an Excellent Response to CRT
85 patients with severe HF, LBBB, QRS duration > 120 ms
TDI prior to CRT
Dyssynchrony was defined as the maximum delay between the time to peak systolic contraction velocity among four ventricular walls (anterior, inferior, septal and lateral)
Bax et. Al., JACC 2004:

30. TDI as Predictor of Response to CRT, Cont’
Bax et. Al., JACC 2004:

31. TDI as Predictor of Response to CRT, Cont’
Bax et. Al., JACC 2004:

32. TDI as Predictor of Response to CRT, Cont’
ROC curve analysis
Sensitivity and specificity of 80% to predict CRT response at a cut-off level of 65 ms of LV dyssynchrony
Response defined as improvement in NYHA class and 6 min walk
Bax et. Al., JACC 2004:

33. TDI as Predictor of Response to CRT, Cont’
Sensitivity and specificity of 92% to predict reverse LV remodeling
Defined as improvement of LVESV of ≥ 15%
Bax et. Al., JACC 2004:

34. Limitations of TDI
Technical limitations: multiple peaks (can be seen even in structurally normal hearts), artifact, experience of the operator
Examines motion, not contraction per se
Interpretation difficult in the setting of akinetic wall/scar

35. Strain Rate Analysis
Differentiates between tethering or passive motion of non-contractile myocardium of TDI alone and active contraction
Limitations: technical factors, artifacts, low signal-to-noise ratio, difficult image acquisition
Radial strain is not well reproduced in multiple studies

36. Strain Rate Imaging: Normal Heart
Breithardt et. Al, Eur Heart J, 2004: D16-24

37. Strain Rate Imaging: Patient with LBBB
Onset of radial motion and strain in inferoseptal, inferior and inferolateral walls
Interregional delay in onset of regional thickening

38. 3D Echo Better spatial resolution High level post processing
Evaluate all walls simultaneously
Need more data

39. 3D Echo

40. PROSPECT Study Predictors of Response to CRT ESC Congress Reports 2007
Prospective study evaluating role of echo in predicting response to CRT
Primary end-point: clinical composite score (CCS) and LVESV
Ghio, et. al

41. PROSPECT Study, Cont’ Echo prior to CRT or CRT-D
Echo post with AV delay optimization
Training of participating sites
Repeat echo in 6 months
Baseline characteristics: 426 patients, avearage QRS 160 ms, LVEF 24%, most with LBBB, NYHA class III and IV sxs
Ghio, et. al

42. PROSPECT Study, Cont’ At 6 months
Overall CCS improvement rate is 75.6% for non-ischemic and 63.7% for ischemic patients
Overall LVESV improvement rate is 63% for non-ischemic and 50.3% for ischemic patients
Ghio, et. al

43. PROSPECT Study, Cont’
Substantial inter-core lab variability in all TDI based dyssynchrony measures
At the same time, the presence of a single mechanical delay (MD) measure added 11-13% response to CCS and 13-23% to LVESV
Ghio, et. al

44. PROSPECT Study: Conclusion
No single measure of mechanical dyssynchrony may be recommended to improve patient selection for CRT
Methodology to determine mechanical dyssynchrony needs further elaboration
Ghio, et. al

45. Conclusions
CRT is an effective adjunctive non-pharmocological therapy for patients with advanced heart failure due to systolic left ventricular dysfunction with evidence of electrical and mechanical dyssynchrony
Many imaging modalities exist to evaluate for mechanical LV dyssynchrony
TDI based measures do not appear to be good predictors that could improve patient selection for CRT
Up to 30% of patients, selected based on current guidelines, are non-responders

46. Conclusions, Cont’
TDI based measures are helpful in following/optimizing patients post bi-V implant (AV delay optimization, V-V optimization)
More studies required to evaluate TDI modalities in patients with narrow QRS and RBBB with evidence of mechanical dyssynchrony
Echo guided LV (and maybe RV) lead placement, especially in patients with prior transmural infarct
3D echo
CMR data (especially with development of CMR compatible leads)