Transcript Cordeiro et al. Am J Physiol Heart Circ Physiol

Ventricular Mechanics of Asynchronous
Activation
Ben Coppola
Senate Proposal
01/09/07
Motivation
• Heart failure leads to:
– Reduced pump function
– Structural remodeling
• Ventricular pacing in normal hearts leads to:
– Reduced pump function
– Structural remodeling
• Common link: Asynchronous Activation
• 170,000/yr receive pacemakers in the U.S. alone.
– Some resynchronize while others dyssynchronize
Background I: Prestretch
• Pacing from the ventricle results in areas of:
– early shortening (close to stimulus)
– prestretch (far from stimulus)
Pacing site
*
Activation Time
early
shortening
prestretch
*
Strain
Wyman et al. Am J Physiol
Heart Circ Physiol 276: H881891, 1999.
Background II: Fiber Anatomy
• Myofiber angle varies across the wall in a nearly linear fashion
Sample Data (Anterior Wall)
80
R2 = 0.9773
60
Fiber Angle (º)
40
20
0
-20
0
5
10
-40
-60
-80
Depth from Epicardium (mm)
Streeter et al. Circ Res 24: 339-347, 1969.
15
20
Background II: Sheet Anatomy
sheet
fibers
• Fibers are arranged in sheets
– Approximately 4 cells thick
– Loosely coupled by collagen
– Shearing allows for large changes
in wall thickness
– Complex distribution
base
5 mm slice perpendicular to fibers
apex
LeGrice et al. Circ Res 1995;77:182-193
Background III: Cellular Heterogeneity
• Normal transmural gradients of:
– Ion channel densities (Na+, K+)
– Ryanodine receptors
– SERCA2a concentration/SR Ca2+ content
– phospholamban
+ others
Can manifest in electrophysiological
differences such as action potential
duration (APD)
Action
Potential
Cordeiro et al. Am J Physiol Heart Circ
Physiol 286: H1471-1479, 2004.
Many altered in heart failure
Specific Aims
1. To test the hypothesis that prestretch is a diastolic
phenomenon and that its functional consequences are the
result of length-dependent activation.
2. To test the hypothesis that the deformation of the laminar
sheets is dependent on the pattern of activation, not just
their anatomical orientation.
3. To test the hypothesis that observed patterns of transmural
strains during normal and altered activation are influenced
by the presence of cellular heterogeneity.
Aim 1
• To test the hypothesis that prestretch is a diastolic
phenomenon and that its functional consequences are the
result of length-dependent activation.
• Experimental measurements in this aim:
–
–
–
–
–
Midwall bead arrays
Transmural bead arrays
Bipolar electrode recordings
LV Pressure
Aortic Flow
Midwall bead & electrode array
Aim 1 Methods – Midwall Array
Base
Base
Apex
Apex
Anterior/Posterior View
LV Pressure
Lateral View
Aim 1 Methods – Midwall Array
Anterior/Posterior View
LV Pressure
Lateral View
Blue = shortening
Fiber Strain
Aim 1 Methods – Transmural Array
• Same general approach but now:
– 3 columns of 4-6 beads are implanted across the wall.
– Can get 3D strain tensor in one small area instead of 2D strains
over a broader area.
– Non-homogeneous
Fit
Omens et al. Circulation 84: 1235-1245
McCulloch et al. J Biomech 24: 539-548, 1991.
Aim 1a: Timing
• To test the hypothesis that prestretch occurs prior to ventricular
end diastole, due to early-activated regions generating
intracavitary pressure and passively inflating late-activated
regions before they become electrically active.
• Rationale:
– ‘Diastole’ continuing in areas not yet active
• Experimental Methods:
–
–
–
–
–
5 open-chest dogs
Bead array in anterior midwall
Pacing from posterior wall
Bipolar electrode array
LV Pressure & Aortic Flow
Example, Expt. #1
Timing
*: Ventricular
Stimulus
Local
Activation
Time
(ms)
44
Fiber
Strain
.016
End Diastole
Aortic Valve
Opening
N=5
Local Activation
Example, Expt. #1
Timing
*: Ventricular
Stimulus
Local
Activation
End Diastole
Time
(ms)
44
55
Fiber
Strain
.016
.019
Aortic Valve
Opening
N=5
Local Activation
End Diastole
Example, Expt. #1
Timing
*: Ventricular
Stimulus
Local
Activation
End Diastole
Aortic Valve
Opening
Time
(ms)
44
55
134
Fiber
Strain
.016
.019
.060
N=5
Local Activation
End Diastole
Aortic Valve Opening
Timing
* P < 0.05 vs. Local Activation
† P < 0.05 vs. End Diastole.
Aim 1b: Initial Fiber Length
• To test the hypothesis that myocardial fibers in the late-activated
regions are longer at the onset of ejection during ventricular pacing
than during atrial pacing, resulting in increasing ejection shortening in
these late-activated regions.
• Rationale: Compensatory work done by late-activated regions
Aortic Valve Opening Strain
Ejection Strain
Eff
0.10
0.00
0.00
-0.05
-0.05
Eff
Ecc
Ecc
Efc
Efc
0.08
Strain
0.04
0.02
-0.10
-0.15
Strain
0.06
Strain
Ref: Atrial Pacing, onset of ejection
Def: Vent. Pacing, onset of ejection
-0.10
-0.15
*
0.00
Eff, Fiber Strain
Ecc, Cross-fiber Strain
-0.20
Efc,-0.20
Shear Strain
Atrial
Ventricular Control
Atrial
Ventricula
Aim 1c: AV Delay
• To test the hypothesis that the length of AV delay
modulates the prestretch response by changing the preload.
• Rationale:
– 35 ms AV delay resulted in same fiber length as atrial pacing
– Hemodynamic function varies with AV delay
Aortic Valve Opening Strain
1.5
0.05
1.4
0.04
1.3
0.03
1.2
0.02
1.1
Fiber Strain
Normalized Stroke Volume
Stroke Volume
1
0.9
0.8
0.01
0
-0.01 0
10
20
30
40
-0.02
0.7
-0.03
0.6
-0.04
0.5
0
10
20
30
40
50
60
70
-0.05
AV Delay
• Need to create AV block in order to get long AV delays
AV Delay (ms)
50
60
70
Aim 1c: AV Delay
Stroke Volume
Normalized Stroke Volume
1.5
1.4
1.3
1.2
1.1
1
0.9
0.8
0.7
0.6
0.5
0
50
100
150
200
250
AV Delay (ms)
Truncated atrial kick
Filling while ventricle
not fully relaxed
Stroke Volume follows expected pattern; strains not yet analyzed
Aim 1d: Conduction Velocity
• To test the hypothesis that decreasing the conduction
velocity (without changing activation sequence) will
increase prestretch in a finite element model of ventricular
pacing.
Activation Time
Approach:
Use same activation pattern but
scale activation times
Aim 1d: Conduction Velocity
• Suppose that all areas activated X ms after stimulus are
prestretched
– Increase in volume of late-activated tissue
– Decrease in stroke volume
• Alternative outcome:
– Reduced rate of pressure development
– No change in pattern of prestretch relative to cardiac cycle
– No change in stroke volume
Aim 1 Summary
• Most of aims 1a, 1b, 1c complete
• 1d may already have been looked at by Roy
– If so, could look at other factors such as contractility
Aim 2
• To test the hypothesis that the deformation of the laminar
sheets is dependent on the pattern of activation, not fixed
by their anatomical orientation.
Fibers in plane of screen
Fibers go into screen
Yellow lines demarcate a sheet in each electron micrograph (LeGrice, I. J. et al. Circ Res 1995;77:182-193)
Aim 2
•
•
•
•
Sheet angle (b) varies through the wall
Multiple sheet families can exist at one location
Strains can be expressed in a fiber-sheet coordinate system
Changes in sheet angle can be derived from strains
Ashikaga et al. Am J Physiol Heart Circ
Physiol 286: H640-647, 2004.
Large S.D.’s probably due
to averaging of 2 sheet
populations
Aim 2a: Sheet Angle Change
• To test the hypothesis that pacing from the LV epicardium
alters the systolic sheet angle change compared to normal
activation.
• Rationale:
– Magnitude and gradient of sheet extension (ESS) reduced by
epicardial pacing
Atrial pacing
LV epicardial pacing
Ashikaga et al. Am J Physiol Heart Circ Physiol 286: H2401-2407, 2004.
Aim 2a: Sheet Angle Change
• Multiple sheet families can exist at one location
• Direction of approach of activation wavefront may preferential
activate one of these families
Methods:
•
•
•
•
Open-chest dogs
Transmural bead array
4 Pacing Sites
Histology
X
X
X
X
X: Epicardial Pacing Site
Expected Results: Sheet angle change dependent on pacing site
(only in areas of multiple sheet families?)
Aim 2b: Transmural Activation
• To determine the relative contribution of sheet deformation
to wall thickening with normal and reversed transmural
activation.
• Rationale:
– Preliminary data suggests epicardial pacing is functionally
inferior to endocardial pacing (though not significantly)
– The order of activation of the sheets may be important
– For example, earlier activation of endocardial sheets may
cause them to deform first, and through tissue tethering
epicardial sheets may deform even before being activated.
Aim 2b: Transmural Activation
• Can calculate contributions of fiber-sheet strains to wall
thickening (ERR = ESScos2b + ENNsin2b + ESNsinbcosb)
• During systole, primary mechanism is sheet extension, ESS.
Contribution to wall thickening
ESS
ENN
ESN
66%
-5%
41%
(Anterior wall, ¼ of
distance from apex
to base)
Costa et al. Am J Physiol 276: H595-607, 1999.
Expected Results (Epi pacing compared to Endo):
• Decrease in overall wall thickening
• Change in contributions to wall thickening, sheet angle change
Aim 2c: Maximum Shear
Predicted
Measured
• To test the hypothesis that sheets are oriented along the plane of
maximum shear (constrained to contain the myofibers) only when
normally activated.
• Rationale:
– The sheet angles can be predicted based on measured
deformations, using the assumption that they are aligned along the
direction of maximum shear that contains the fibers.
Arts et al. Am J Physiol Heart Circ Physiol 280: H2222-2229, 2001.
Aim 2c: Maximum Shear
Rationale(cont.)
• Altered activation changes the overall pattern of deformation
• Direction of maximum shear is likely changed as well such that it
no longer coincides with the sheet direction
• Could be a stimulus for the remodeling of sheets that is seen with
chronic ventricular pacing (Helms, 2006)
Expected Results (based on experiments of 2a & 2b):
• Pacing at the bead set will change the direction of maximum shear,
making it closer to the radial direction
• This is the direction in which the sheet angle remodels chronically
Summary of Aim 2
• Use existing techniques to measure deformation
• Make histological measurements of fibers and sheets
• Determine role of sheets in abnormal activation
Aim 3
• To test the hypothesis that observed patterns of transmural
strains during normal and altered activation are influenced
by the presence of cellular heterogeneity.
Action Potential
Unloaded Cell Shortening
Cordeiro et al. Am J Physiol Heart Circ Physiol 286: H1471-1479, 2004.
• Methods
– Computational modeling, no new experiments
Aim 3a: Normal Gradients
• To test the hypothesis that cellular heterogeneity influences
observed strain patterns with normal activation.
• Rationale:
Transmural conduction time
– Preliminary data
shows that a finite element model with
homogeneous cellular properties incorrectly predicts epicardial
prestretch (due to endocardial contraction)
Endo
AP –
Transmural conduction time is thought to be balanced byEpi
differences in electrical-mechanical delays, resulting in a synchronous
contraction (Cordeiro, 2004)
Cell
Shortening
Schematic representation based on data from Cordeiro et al.
Am J Physiol Heart Circ Physiol 286: H1471-1479, 2004.
Aim 3a: Normal Gradients
• Methods:
– Incorporate cellular model which is transmurally heterogeneous
(Flaim, 2006) into current finite element model
– Compare experimentally measured strains to:
• Simulations with homogeneous cellular properties
• Simulations with heterogeneous cellular properties
• Expected Results:
– Eliminate erroneous prestretch
– Improve end-systolic strains? (radial or transverse shear)
Aim 3b: Abnormal Activation
• To test the hypothesis that cellular heterogeneity creates
transmural dyssynchrony during epicardial pacing.
• Rationale:
Retrograde conduction time
– If the transmurally heterogeneous properties are “designed”
to make the wall contract more synchronously, then reversing
Endo
the direction of activation should have a dramatic effect.
AP
Cell
Shortening
Epi
Aim 3b: Abnormal Activation
• Methods
– Compare Epi & Endo pacing experimentally to:
• Simulation with homogeneous cellular properties
• Simulation with heterogeneous cellular properties
• Expected Results
– Endocardial pacing will be improved w/heterogeneity
– Epicardial pacing will generate endocardial prestretch (may
not be reflected experimentally
– Likely to have effects on relaxation as well but this may be
difficult to address with model
Summary of Aims
To look at the significance of the following during asynchronous activation:
1. Late Activated Tissue (3 months)
2. Fiber-Sheet Architecture (6 months)
3. Cellular Heterogeneity (3-6 months)
Acknowledgments:
Hiroshi Ashikaga
Roy Kerckhoffs
Rachel Alexander
Aundrea Graves
Katrina Go
Jay Yee
Leonard Lee