Taming cardiac disorders:
…with some help from applied mathematics…
Alain Pumir
Laboratoire de Physique, ENS Lyon, France
Cardiac arrhythmia :
meet the monster…
Life-threatening cardiac arrhythmias result from
spiral wave breakup leading to
electrical turbulence
in the
heart, and to lack of proper functioning of the blood pump.
Pacing and control of chaos
For lack of a better strategy, chaos control in the heart rests on electric fields.
- Defibrillation shock : one shock, with a very large currents delivered through the heart.
* High success rate but numerous undesirable side effects -- including pain.
- Anti-Tachycardia-Pacing is used to control an incipient arrhythmia
(tachycardia), by sending through one electrode several small stimuli, at frequency larger than the main frequency of the ECG.
* Success rate : ok (70-90%).
How to terminate turbulent activity in
the heart ?
An old problem for applied mathematics !
Pacing of the heart : role of
heterogeneities.
Wiener and Rosenblueth (1946)
Topological conservation :
(Wiener and Rosenblueth, 1946)
# of waves rotating counterclockwise, clockwise around the obstacle.
In the figure,
Impossible to remove the rotating wave by pacing in the case of the
cellular automata model of Wiener and Rosenblueth !
n
n
cst.
n
,
n
n
n
1
Control of vortices by pacing.
Pumir, Sinha et al. (2010)
The topological
conservation law of
Wiener and Rosenblueth
does not apply to small
obstacles in the case of
realistic models !
Possibility of pacing a wave from a small obstacle
nb : generic phenomenon in
excitable media; experimental possibilities...
Taming a large turbulent system??
• Suggestion : use many pacing sites –
Ingredients for better pacing.
Pacing from multiple sites
has been noticed many times to
be more efficient than from just one site (e.g., Allessie et al,
1991; Pande et al., 2001).
Heterogeneities
in the heart can act as
wave sources
when
an external electric field is applied (e.g., R. Plonsey, 1982,
Pumir and Krinsky, 1999).
Combine the two ideas =>
Far Field Pacing
~ use heterogeneities as ‘pacing electrodes’.
n.b. : there are
many heterogeneities
in a heart !!
Action potential propagation
In the tissue, the
electric potential
propagates due to
the resistive
coupling between
points
Cm
e
t Istim (INa IK IL) 1
2e
x2Shaw, R. M. et al. Circ Res 1997;81:727-741
Dynamical model of the heart : the dynamic
Luo-Rudy ventricular cell model (LRd)
The minimal model : FitzHugh
• The simplest model that describes the
class of excitable system is the
FitzHugh system :
dw dt (ew) de dt Ae(1e)(eu)w de dt 0 dw dt 0The horizontal motion is fast compared to the vertical motion :
Two time scale problem.
Interaction of an external
Obstacle/electric field interaction
Model problem
:
one hollow circular obstacle in a 2-dimensional piece of cardiac tissue.
Use the monodomain model (simplest generalization of cable theory).
V
m = membrane potentiale=(V
m-V
m,rest)
=
deviation from the resting membrane potential.In the linear regime (weak deviations from resting potential) :
with the b.c. :
when
and at the obstacle’s boundary ( ).
2e
e
2
0
e
0
r
ˆ
n
.
(
e
E
.
r
)
0
r
R
Obstacle/electric field interaction
The solution is obtained in 2d as a standard Bessel function :
( = angle between E and r)
e
(
r
)
E
K
1(
r
/
)
K
'
1(
R
/
)
cos
=> Part of the tissue is
depolarized (e > 0).
If
depolarization
exceeds a
threshold
,
an Action Potential
Obstacle/electric field interaction
Maximum depolarization
:
e
max
E
K
1(
R
/
)
K
'
1(
R
/
)
The stimulation threshold is reached when is
larger than a threshold: .
The field needed to reach the stimulation threshold
decreases with the size
R
of the obstacle:
e
max
e
max
(
t
rest)
E
E
(
t
rest)
K
1'(
R
/
)
K
1(
R
/
)
Obstacle/electric field interaction
Dependence of the electric field needed to start a
wave as a function of the size of the obstacle:
Consequence
:
When an electric field of
strength
E=e,
is applied,
all obstacles with a size
emit waves.
n.b.:
when
R
minis small.
R
R
min(
e
)
E
(
t
rest)
K
1'(
R
/
)
K
1(
R
/
)
E
~ 1/
R
min(
E
)
Obstacle/electric field interaction
Consequence
: depolarization reaches the stimulation
threshold provided the obstacle is large enough, or the field
is strong enough.
By increasing the external field, one may increase the
number of (virtual) pacing electrodes.
Obstacle/electric field interaction
Numerical illustration:
Integrate numerically the Luo-Rudy 1 model, with several
circular obstacles
of various sizes :
As the externally applied
electric field increases
,
obstacles of
decreasing sizes
trigger AP propagation
see: Pumir et al, Phys. Rev. Lett. 2007.
Far field pacing :
Far field pacing : numerical example
Heterogeneities:
collagenous septa
randomly
distributed
throughout the
medium.
From Fenton et al, Circulation 2009.
E=0.8 V/cm
E=1.14 V/cm
Far field pacing : numerical example
Apply 8
low-intensity shocks:
S3: T=970ms
S8: T=1905ms
From Fenton et al, Circulation 2009.
E=0.8 V/cm
E=1.14 V/cm
Far field pacing: numerical example
The success of the method is based on the presence of
obstacles close to the core of the spiral.
=> one needs precise information on the
distribution of heterogeneities !!
Activation time and distribution
of heterogeneities
Response time in the presence of many obstacles
V
4
3
N
v
3 define density as = N/V
(3/(4
))
1/ 3/
v
assumptions: • 3d • homogenoeus tissue • propagation velocity vResponse time in the presence of many obstacles
V
4
3
N
v
3 define density as = N/V
(3/(4
))
1/ 3/
v
assumptions: • 3d • homogenoeus tissue • propagation velocity vResponse time in the presence of many obstacles
(3/(4
))
1/ 3/
v
tissue: distribution function of sizes p(R)
E
p R
Rmin
E Rmax
d
R
E 1/R MRIObstacle Size Distributions
p
1/
R
2.520.07
p
1/
R
E
p R
Rmin E Rmax
d
R
n
Obstacle Size Distributions
p
1/
R
2.520.07
p
1/
R
E
p R
Rmin E Rmax
d
R
E 1/R
E
1
(3/(4
))
1/ 3/
v
with
E
(1) 3 predictionOptical Mapping
• simultaneous epi/endo mapping • EMCDD (128 x 128, 16 bit) • 500 fps
•Use voltage-sensitive dyes (di-4-anepps).
Relating the density of obstacle size and the response time: Ventricular tissue (3d) MRI result: p(R) ~ R-2.75 Response time: τ ~ E-0.58 Comparison: Response time vs. MRI measurment.
Protocol:
5 pulses, 1.4 V/cm
[5 ms duration, 90 ms period]
Summary and Conclusion
• virtual electrodes exist; they reflect the fine structure of the tissue, which can be described in terms of power laws.
• Virtual electrode mechanism is consistent with observation.
• Virtual electrodes can be used to defibrillate (== synchronize) in an optimal way.
Thanks to:
• V. Krinsky (Nice and Göttingen)
• S. Sinha, S. Sridhar (Chennai)
• E. Bodenschatz, S. Luther, C. Richter, A.
Squires, D. Hornung, P. Bithin (Göttingen)
• R.Gilmour, F.Fenton, E.Cherry and N.Otani
(Cornell)
Thank You !
Ref :
A. Pumir et al., Phys. Rev. Lett. 99, 208101 (2007)
F. Fenton et al., Circulation, 120, 467 (2009)
Experimental results
The main issues addressed here :
In a real tissue :
• Possibility of terminating VT/VF by applying a small intensity
electric pulse ?
• Does increasing the electric field strength lead to more
heterogenities acting as wave sources ?
Experimental preparation (Fenton et al, 2009):
o
in-vitro
experiments with Mongrel dog hearts.
o
optical mapping of the atrium surface
o
Deliver shocks of duration 5-10ms, and of amplitude up to E
= 4.6 V/cm.
Recruitment of wave sources
A: pace with a single electrode B: apply a shock E=0.32 V/cm C: E=0.46 V/cm D: E=0.93 V/cm E: E=1.4 V/cm
Activation time
The stronger the applied Electric field, the shorter it takes to excite
completely the tissue.
Indirect evidence that when the electric field is increased,
the number of wave sources increases !
Successful termination of AF
Use 5 shocks of 5ms; Period : 45ms; E=1.4 V/cm. « Electrical turbulence» Apply 5 shocks Restoration of the normal rhythm.Unsuccessful termination of AF
Use 5 shocks of 5ms; Period : 45ms; E=0.9 V/cm. « Electrical turbulence» Apply 5 shocks Arrhythmia continuesSuccess of multishock therapy
Ambrosi et al, 2011:
Proof of success of ‘multiple shock
Conclusions
Heterogeneities are known to potentially arrhythmogenic.
They can also help in getting rid of arrhythmias !
Main observations in this work :
- Heterogeneities can be used as Wave Emitting Sites, the number of WES increasing with the electric field.
- Properly placed virtual electrodes can pace spiral waves away.
- Experimental evidence : suggests that it does work… still work to be done to fully understand the mechanisms...
Obstacle/electric field interaction
3-dimensional case.
For a simple spherical obstacle of radius R, with a value of the electrotonic length : Emax E 1 R 1 2 R 22 R2 E R R2 2 22R R2 2 Emax ER/2 in 3d ER in 2d
Obstacle/electric field interaction
In cardiac tissue, heterogeneities may have a noncircular
structure (e.g., clefts experiments on tissue culture; cf Fast et
al, 1998)
Obstacle/electric field interaction
Effect of the angle :
Integrate numerically the Luo-Rudy 1 model, with several
elliptical obstacles
of various sizes, either horinzontal or
vertical :
By
flipping the direction
of the electric field, waves
start from an entirely
different set of virtual
Obstacle/electric field interaction
A simple model problem
: field generated by an ellipse :
on the obstacle.
at infinity
E a b
2e
e
2
0
ˆ
n
.
(
e
E
.
r
)
0
e
0
Obstacle/electric field interaction
Solution
: In the case where solve Laplace
equation : by using conformal transformation :
with
Solution for the membrane potential :
2
2e
0
z
2
1
2
a
2
b
2
e
(
z
)
Re
(
a
b
)
2e
i
(
a
2
b
2)
e
i4
Obstacle/electric field interaction
Maximum depolarization
:
e
Max= b E
e
Max= a E
[Hard to emit
waves]
[Easy to emit
waves]
e
Max
E
(
a
b
)
2
(
b
2
a
2)cos(2
)
2
1/ 2
(
2
)
A more realistic bidomain study (Takagi et al, 2004)
Reentry “unpinning” by a low-energy
shock: Mathematical 2D bidomain model
Far field pacing : a simple
numerical example
It can happen that pacing from one single site is enough to get rid of free vortices.
… ATP Consider again the tissue with a stripe of tissue with a long refractory time, where ATP was illustrated to fail.
Assume that there exists relatively big obstacle (virtual electrode) not too far away from the cores of the vortices.
Turn on and off the external electric field :
• field on for ~5ms, then off for ~100ms, several times; • typical values of the field : ~0.5-1 V/cm)
Application:
Unpinning of vortices attached
to obstacles.
Theoretical idea: What can be done with a single pulse, using the interaction with the obstacle ?
The depolarization induced by a (relatively) moderate electric field, if properly timed, can detach a wave.
Proof of concept obtained by studying over-simplified models (Pumir and Krinsky, JTB 1999).
How to get rid of spiral waves “attached”
to obstacles (scar tissue, etc) ?
Condition of successful “unpinning”:
low intensity, but precise timing !
Successful “unpining” of waves attached to an obstacle in a rabbit heart preparation (Ripplinger et al, 2006).
Controlling electric turbulence
in the heart with low energy far
field stimulation.
Alain Pumir
(ENS Lyon, France).
Collaborators:
V.Krinsky
(Nice and Göttingen),E.Bodenschatz and S.Luther (et al.)
(MPI Göttingen),
R.Gilmour, F.Fenton, E.Cherry and N.Otani (et al.)
(Cornell),
K.Yoshikawa, A.Isomura, K.Agladze & M.Hörning
(Kyoto),Low field defibrillation of cardiac muscle via wave
emission from heterogeneities
Alain Pumir
Laboratoire de Physique, ENS Lyon, France
S. Luther, E. Bodenschatz, V. Krinsky, G. Luther, C. Richter, A. Squires, D. Hornung, P. Bithin
Max Planck Institute for Dynamics and Self-Organization &
Inst. for Nonlinear Dynamics, Georg August Universität, Göttingen, Germany
F. Fenton, R. Gilmour, E. Cherry, N. Otani.
Outline
-
Electric waves in the heart
-
Interaction of an external electric field with an obstacle.
-
Far Field Pacing : numerical examples.
-
Activation time and distribution of heterogeneities.
Pacing vortices: how many wave
sources does one need ?
“One” can be good enough…
Anti-Tachycardia Pacing : involves pacing waves from an electrode at high enough a frequency (higher than the frequency of the rotating waves in the heart).
Gottwald, Pumir and Krinsky, 2001
Recent observation of the phenomenon in cardiac cell cultures : Agladze et al., Am. J. Physiology, 2007
Far field pacing : a simple
numerical example
Pumir et al., Phys. Rev. Lett. 2007. Stripe of longer Refractory period Obstacles (virtual electrode) Final time : The pacing waves have swept away the vortices Initial time : Two vortices
Failure or success of pacing ?
key to success: Wave sources should be close enough to the spiral cores. A remote wave source can be shielded from the vortex, e.g. by a region with a longer refractory period (=> unsusccesful pacing.)
P
: Pacing siteStripe of tissue with a longer refractory period
Far field pacing : a simple
numerical example
Lessons:
•
The study of very simple models suggests that success
can be achieved with a relatively weak external field
amplitude, provided the density of potential wave sources
is large enough.
• The success or failure of Far Field Pacing does
NOT
depend
much on the precise phase of the pacing w.r.t. the phase(s) of
the spiral(s) provided the wave sources are properly located.
Effect of Increasing Field Strength
Multisite excitation
Single Wave propagation (n = 1)
Cardioversion (n )
Relating the density of obstacle size and the response time: Atrial tissue (2d) MRI result: p(R) ~ R-2.54 Response time: τ ~ E-0.87 Comparison: Response time vs. MRI measurment.
Obstacle/electric field interaction
The simplest problem:
one fiber with an externally applied electric field (Weidmann 1953).
Mathematical description: Use cable theory. Vm : membrane potential
e = (Vm – Vm,rest) : deviation from the resting potential.
In the linear regime ( ~ weak deviations from resting potential) :
λ = electrotonic length (~1mm)
with: far away from the boundary, and de/dx = E at the boundary.
Solution: e(x) = λ E exp(-x/λ) for x > 0;
Depolarization: + or - λ E
d2e/dx2 e/
2 0