Download Assessment of wasted myocardial work

Articles in PresS. Am J Physiol Heart Circ Physiol (July 26, 2013). doi:10.1152/ajpheart.00191.2013
Russell et al.
1
Assessment of wasted myocardial work – A novel method to quantify
2
energy loss due to un-coordinated left ventricular contractions
3
4
Russell et al, a novel method to quantify wasted myocardial work
5
6
Kristoffer Russell, MD; Morten Eriksen, MD, PhD.; Lars Aaberge, MD, PhD; Nils
7
Wilhelmsen; Helge Skulstad, MD, PhD; Ola Gjesdal, MD, PhD; Thor Edvardsen, MD, PhD;
8
Otto A. Smiseth, MD, PhD, FACC.
9
Institute for Surgical Research and Department of Cardiology, Rikshospitalet, Oslo
10
University Hospital, Norway. University of Oslo, Medical faculty
11
12
Address for correspondence:
13
Otto A. Smiseth
14
Department of Cardiology
15
Rikshospitalet Oslo university Hospital
16
N-0027 Oslo, Norway
17
Phone: +47 23070000/+47 23073271
18
Fax: +47 23073917
19
E-mail: [email protected]
20
Total Word Count: 4479
21
Key words: Heart failure, myocardial function, dyssynchrony, cardiac resynchronization
22
therapy
23
24
1
Copyright © 2013 by the American Physiological Society.
Russell et al.
25
Background:
26
Left ventricular (LV) dyssynchrony reduces myocardial efficiency because work
27
performed by one segment is wasted by stretching other segments. We introduce a novel non-
28
invasive clinical method which quantifies wasted energy as the ratio between work consumed
29
during segmental lengthening (wasted work) divided by work during segmental shortening.
30
Methods:
31
The wasted work ratio (WWR) principle was studied in 6 anesthetized dogs with left
32
bundle branch block (LBBB), and in 28 patients with cardiomyopathy, including 12 with
33
LBBB and 10 with cardiac resynchronization therapy (CRT). Twenty healthy individuals
34
served as controls. Myocardial strain was measured by speckle tracking echocardiography
35
(STE), and left ventricular pressure (LVP) by micromanometer and a previously validated
36
non-invasive method. Segmental work was calculated by multiplying strain rate and LVP to
37
get instantaneous power, which was integrated to give work as a function of time. A global
38
WWR was also calculated.
39
Results:
40
In dogs WWR by estimated LVP and strain showed strong correlation (r =0.94) and
41
good agreement with WWR by LV micromanometer and myocardial segment length by
42
sonomicrometry. In patients, non-invasive WWR showed strong correlation (r =0.96) and
43
good agreement with WWR using LV micromanometer. In healthy controls global WWR
44
was 0.09±0.03, in patients with LBBB 0.36±0.16, and in cardiomyopathy patients without
45
LBBB 0.21±0.09. CRT reduced global WWR from 0.36±0.16 to 0.17±0.07, (p<0.001).
46
47
48
2
Russell et al.
49
50
Conclusions:
Energy loss due to in-coordinated contractions can be quantified non-invasively as LV
51
wasted work ratio. This method may be applied to evaluate the mechanical impact of
52
dyssynchrony.
53
3
Russell et al.
54
Introduction
55
In a normal left ventricle (LV) all segments contract in a synchronized fashion and
56
myocardial energy is utilized effectively to eject blood out of the ventricle. In ventricles with
57
regional differences in contractility as in ischemia or delay in electrical conduction as in
58
LBBB, some segments are stretched in systole while others contract, resulting in a
59
dyssychronous contraction pattern. The result of this dyssynchrony is that a substantial
60
amount of LV work is “wasted” on stretching opposing segments and therefore does not
61
contribute effectively to LV ejection. There is currently no clinical method to quantify how
62
much work is wasted due to dyssynchrony.
63
Recently we introduced a non-invasive clinical method to quantify regional LV work
64
(9). This method measures regional work in terms of LV pressure-strain loop areas which are
65
constructed from a non-invasive LV pressure curve in combination with strain by STE.
66
Analogous to the principle that the area of the LV pressure-volume loop reflects stroke work
67
and global myocardial oxygen consumption (11; 12), it has been shown by Delhaas et al (1)
68
and by Russell et al (9) that the pressure-strain loop area may serve as an index of regional
69
myocardial work and metabolism when comparing segments within a given ventricle. In the
70
present paper we propose to use segmental LV pressure-strain data acquired by a non-
71
invasive approach to quantify how much work is wasted in ventricles with dyssynchronous
72
contractions. This energy loss is expressed as work consumed during segmental lengthening
73
(negative work) as the ratio of work during segmental shortening (positive work), and we
74
named this the wasted work ratio (WWR).
75
76
The main objective of the present study was to validate WWR and investigate changes
in WWR to CRT in patients with heart failure and LBBB.
4
Russell et al.
77
First we investigated if LV WWR can be measured and quantified non-invasively.
78
Second, we studied if WWR can be measured by LV pressure-strain analysis in a dog model
79
using sonomicrometry as a refrence method for speckle tracking strain and in patients using
80
micromanometer as reference method for LV pressure. Finally, we investigated changes in
81
WWR to CRT in patients with heart failure and LBBB.
Methods
82
83
Experimental study
84
Animal preparation
85
Six mongrel dogs of either sex and body weight 35 ± 1 kg were anesthetized, ventilated,
86
and surgically prepared as previously described (13), including induction of LBBB (n=6) by
87
radiofrequency ablation (2). The study was approved by the National Animal
88
Experimentation Board. The laboratory animals were supplied by Centre for Comparative
89
Medicine, Oslo University Hospital, Rikshospitalet, Norway.
90
Sonomicrometry
91
Segment lengths were measured by pairs of 2 mm wide sonomicrometry crystals
92
(Sonometrics Corp., London, Ontario, Canada) implanted subendocardially. Four
93
circumferential segments incorporating all LV walls were analyzed. Strain was calculated as
94
percent length deviation from the end-diastolic length. Data were sampled at 200 Hz.
95
Hemodynamic measurements
96
Aortic, left atrial and LV pressures (LVP) were measured by micromanometers (MPC-
97
500, Millar Instruments Inc, Houston, Tex). A fluid-filled catheter placed in the left atrium
98
served as an absolute pressure reference for the LV micromanometer.
5
Russell et al.
99
100
Echocardiography
A Vivid 7 ultrasound scanner (GE Vingmed Ultrasound AS, Horten, Norway) was used
101
to record conventional 2-D grayscale images (frame rate 64±12 s-1) of the LV (at the level of
102
the papillary muscle) short-axis and apical 4 and 2 chamber views for STE.
103
104
105
106
Experimental protocol
Measurements were performed before and after induction of LBBB. Data were
recorded with the ventilator temporarily switched off.
107
108
Clinical study
109
Validation of non-invasive vs. invasively measured WWR
110
18 patients (mean age 66±8 years, 25% females) underwent LV catheterization,
111
including 11 with ischemic cardiomyopathy and 7 with non-ischemic dilated
112
cardiomyopathy. Twelve of these patients had LBBB (QRS 160±20 ms). Two of the patients
113
with LBBB had a CRT device allowing for measurements with CRT on and off.
114
Response in WWR to Cardiac Resynchronization Therapy
115
Ten additional patients with heart failure and NYHA II-IV and LBBB were included.
116
Coronary artery stenosis was ruled out by angiography prior to investigation.
117
Echocardiography was preformed prior to implantation of CRT device (n=10) and at follow
118
up 8±3 months later (n=9). To assess changes in wasted work after CRT we deliberately
119
included only responders in the patient population (>10% reduction in end systolic volume)
120
in order to demonstrate changes in work as a result of resynchronization treatment.
121
6
Russell et al.
122
123
WWR in healthy controls
Twenty age and sex matched healthy individuals were recruited from the hospital staff.
124
All had normal clinical examination, ECG and echocardiography.
125
The study was approved by the Regional Committee for Medical Research Ethics. All
126
subjects gave written informed consent.
127
128
129
Hemodynamic and Echocardiographic measurements
In all patients myocardial strain was measured by STE (Vivid 7) in apical long-axis, 2
130
and 4 chamber views (frame rate 65±6 s-1). Narrow sector 2-D imaging over the valves
131
(frame rate 87±22 s-1) was used to define opening and closure of the aortic and mitral valves.
132
In the invasive LV pressure group, LVP was measured by a micromanometer-tipped catheter
133
(Microtip, Millar Instruments) and a fluid-filled catheter connected to an external pressure
134
transducer served as absolute pressure reference. Pressure and strain data were recorded in a
135
synchronised fashion and stored for offline analysis. An average of 16±2 segments (from a
136
total of 18) were analyzed in each patient.
137
Data analysis
138
Estimation of LV pressure curve
139
The method for calculating the non-invasive LV pressure analogue has been described
140
in detail previously (9). In short, we utilized an empiric reference curve for the LV pressure
141
profile. This curve was obtained by pooling LVP data from a number of patients with
142
different pathologies, and for each patient the durations of IVC, LV ejection and IVR,
143
defined by echocardiography, were set to the same arbitrary value by stretching or
144
compressing the curves along the time axis. From this series of individual curves a common
145
average waveform was constructed. The profile of the averaged LV pressure waveform was
7
Russell et al.
146
used for predicting LVP in each specific subject by measuring the actual valvular timing for
147
the relevant cardiac cycle and adjusting the duration of time intervals of isovolumic
148
contraction (IVC), ejection and isovolumic relaxation (IVR) phases by stretching or
149
compressing the time axis of the averaged LV pressure curve to match the measured time
150
intervals. To scale the amplitude of the pressure curve peak arterial pressure was used in the
151
experimental study and arterial cuff pressure in patients and controls. Separate reference
152
curves were used for dogs and patients.
153
154
155
Calculation of segmental work
Work for individual segments as a function of time throughout the cardiac cycle was
156
calculated from the strain recordings and from measured or estimated LV pressure from a
157
representative heart beat. This was done by obtaining the rate of segmental shortening (strain
158
rate) by differentiation of the strain curve and multiplying this with instantaneous LVP. This
159
resulted in a measure of instantaneous power, which was integrated over time to give work as
160
a function of time (Figure 1). Performing this calculation for the IVC, ejection and IVR
161
phases is mathematically the same as calculation of the area between the x-axis and the
162
corresponding segment of the strain – pressure loop, however keeping in mind that areas
163
under sections of the loop that represent elongation are counted as negative (further details
164
are presented in the appendix). Work was calculated from mvc until mvo. The reason for this
165
is that work losses can occur by stretching of segments in the isovolumic phases, especially in
166
hearts with a dysynchroneous contraction pattern.
167
168
8
Russell et al.
169
170
Analysis of segmental work
Information was extracted from the segmental work curves over a time interval
171
spanning from mitral valve closure to mitral valve opening. The following variables were
172
calculated:
173
Wpos
174
175
Total positive work, sum of all work performed during shortening of the segment.
By definition a positive number.
Wneg
Total negative work, sum of all work performed during elongation of the
176
segment. Although a negative number by definition, Wneg is presented as a
177
positive number to facilitate understanding of the relative differences between
178
negative and positive work.
179
180
181
182
183
184
Global WWR was calculated from the sum of work for all segments:
WWRglobal =
W
W
neg
pos
The net work contributing to systolic ejection for each segment was calculated as:
Wnet = W pos − Wneg
A Wasted Work Ratio for each segment was also calculated:
Segmental Wasted Work Ratio =
Wneg
W pos
185
Figure 2 shows positive work (in black) and negative work (in red) during systole for a septal
186
and lateral wall segment before and during CRT in a representative patient.
187
9
Russell et al.
188
Work analysis
189
In the experimental study wasted work ratio calculated using strain by STE and
190
estimated LV pressure was compared to WWR calculated by strain by sonomicrometry and
191
invasively measured LVP.
192
In the patients with invasive LV pressure, work was analyzed by two different methods:
193
1. STE strain and measured LVP
194
2. STE strain and estimated LVP.
195
196
Statistical analysis
197
Values are expressed as mean±SD. Results from different methods of measurement
198
were compared using least-squares linear regression, Pearson correlation coefficients and
199
Bland-Altman plots with calculations of limits of agreement. For multiple comparisons we
200
used repeated measurements ANOVA with LSD post test (SPSS 15.0, SPSS Inc, Chicago,
201
IL). All post hoc tests are compared to baseline. P<0.05 was considered significant.
202
Results
203
204
Experimental study
205
Validation of WWR calculated from estimated LVP and echocardiography versus
206
measured LVP and sonomicrometry
207
The analysis which included measurements before and after induction of LBBB,
208
showed strong correlation (r = 0.94) and good agreement (SD = 0.07, mean difference = -
209
0.03) between global WWR calculated from estimated LVP and strain by STE versus WWR
210
by LV micromanometer and segment length by sonomicrometry.
10
Russell et al.
211
212
213
Changes in WWR from baseline to left bundle branch block
Introduction of LBBB increased global WWR measured by LVP and echocardiography
214
from 0.13±0.05 to 0.37±0.11 (p<0.001). Peak strain for septum and lateral wall were
215
moderately reduced after induction of LBBB, from -15±4 to -10±4% (p=0.014) and -14±4 to
216
-11±7 % (p=ns), respectively. However, the work done by the septum decreased
217
substantially, from 951±320 to 349±346 mmHg•% (p=0.002). In contrast net work done by
218
the lateral wall increased from 756±266 to 1059±516 mmHg•% (p=0.026) despite the
219
tendency of decrease in peak strain. Therefore, changes in peak strain after induction of
220
LBBB did not reflect changes in regional work.
221
222
Clinical study
223
Validation of WWR using estimated versus measured LVP
224
There was a strong correlation (r=0.96) and good agreement (SD=0.03, mean
225
difference= -0.01) between global WWR calculated by STE and invasive vs. non-invasive
226
LV pressure (Figures 3 and 4).
227
228
229
Wasted work ratio in healthy normals
Figure 5 shows a bull’s eye plot with segmental WWRs from a representative healthy
230
individual. The WWR was less than 0.14 in every segment, which implies that LV work was
231
dominated by positive work (i.e. shortening work) and there was very little negative work
232
(i.e. work wasted during systolic lengthening). The global WWR for the healthy normals was
233
0.09±0.03.
234
11
Russell et al.
235
236
Wasted work ratio in patients with cardiomyopathy
In the patients with LBBB peak strain values for the septum were on average half of the
237
value of the lateral wall. For the septum the segmental WWR was 1.57±0.97. Since WWR >
238
1 implies more negative than positive work, this implies that on average net septal work was
239
negative. In some LBBB patients net septal work was positive, although it was markedly
240
reduced. The lateral wall, however, showed increased positive work compared to the control
241
group and the segmental WWR was increased, but to a moderate degree compared to the
242
septum (0.18±0.11) (Figure 5 A).
243
In patients with cardiomyopathy and narrow QRS (96±18ms), without LBBB
244
segmental WWR was moderately increased to 0.17±0.08 for septum, 0.27±0.26 for the LV
245
lateral wall and global WWR was 0.21±0.09.
246
247
Changes in wasted work ratio with CRT
248
When CRT was introduced in patients with LBBB, the septum showed a dramatic
249
increase in positive work and a reduction in negative work, and the septal WWR decreased to
250
0.18 ± 0.09 (Figure 5 B). The lateral wall showed only minor changes with CRT (Figure 2
251
and 4, Table 2). These changes in segmental work with CRT were reflected in a reduction in
252
global WWR from 0.36±0.16 to 0.17±0.07
253
254
255
.
Discussion
In the present study we introduce a novel, non-invasive method which quantifies how
256
much work is wasted in ventricles with dyssynchronous contraction patterns. The method
257
gives segmental information about positive work (systolic shortening) and negative work
258
(systolic lengthening), allowing for assessment of local and global energetic efficacy as the
12
Russell et al.
259
ratio between negative and positive work. The validity of this method was confirmed in the
260
experimental part of the study. Furthermore, in the clinical study we showed that WWR
261
based on a non-invasive LV pressure estimate corresponded well with calculations using
262
invasive LV pressure. We showed that significant work was wasted in patients with LBBB
263
and was subsequently recruited when CRT was introduced, suggesting that the work analysis
264
may represent a means to explore the hemodynamic impact of electrical dyssynchrony.
265
266
267
Work analysis
As shown by Hisano (4), the area of the myocardial force-dimension loop represents
268
the work performed during a cardiac cycle. However, due to the complexities of measuring
269
force, LVP has been used as a surrogate to force. The use of pressure as a substitute for force
270
may be viewed as a limitation. However, previous studies from our group have shown that
271
pressure-lengths loops reflect regional changes in myocardial work when compared to force-
272
segment length loops (10). Furthermore, we have shown that stress-strain loop area,
273
calculated using curvature and area strain and showed quantitatively similar information
274
when compared to stress-segment length loop area when comparing regional differences
275
within the same heart. Another argument which supports that pressure strain loops reflect
276
regional work is our previous demonstration that pressure strain loop area reflects regional
277
glucose metabolism assessed my 18 flourodeoxyglucose positron emission tomography (9)
278
Since invasive LV pressure is often not available, we have introduced a non-invasive
279
estimate of the LV pressure curve, which is constructed by adjusting an empiric average
280
waveform according to duration of the isovolumic and ejection phases as defined by
281
echocardiography, and peak pressure is set equal to arterial cuff pressure. The validity of
282
calculating regional myocardial work from non-invasive LV pressure and strain by STE has
283
been confirmed both in an animal model and in patients (9).
13
Russell et al.
284
Using segment length for the assessment of regional work may be viewed as a
285
limitation as it depends on the relation of the crystal implantation site to myofiber orientation
286
which varies through the ventricular wall. Previous studies have assessed regional work using
287
wall thickness measurement as it does not dependent on fiber direction (3). Ideally regional
288
work should be calculated using area strain, however, in a clinical setting strain analysis are
289
commonly performed for longitudinal and circumferential segments and we therefore used
290
these measurements for work analysis in the present paper. Furthermore, we have previously
291
shown that work analysis by using segmental strain reflects work using area strain (9).
292
293
294
Wasted work
In a normal heart, all segments contract almost simultaneously and they therefore
295
contract against a similar LV pressure. However, in a ventricle with regional differences in
296
timing of contraction, segmental contractions occur against different pressures. Therefore,
297
information about both degree of shortening and LV pressure during shortening is needed to
298
compare work between the segments.
299
The same is true for calculation of wasted energy during elongation of segments, since
300
the energy loss depends on both elongation distance and LV pressure during elongation. In
301
ventricles with non-uniformity in timing of relaxation, some segments will start relaxing and
302
elongate in late systole, at high LV pressure while other segments are still contracting. This
303
means that a substantial amount work done by the contracting segments is wasted on
304
stretching other segments. Typically, during LBBB much of septal contraction occurs during
305
the pre-ejection phase at a time when the LVP is low, and therefore the septum performs little
306
work (8). As seen in the present study peak septal strain during LBBB in the animal model is
307
only marginally reduced compared to the lateral wall, but net work done by the septum is
14
Russell et al.
308
approximately one third relative to the lateral wall (Table 1). In the patients with LBBB peak
309
strain for the septum was on average half of that for the lateral wall, which is in keeping with
310
previous studies (5). The average segmental WWR, however, was above 1.5, indicating that
311
negative (wasted) work exceeded positive work. Therefore, instead of the septum
312
contributing to the hemodynamic performance of the LV, external work was done on the
313
septum which thereby “absorbed” work done by other segments.
314
Measuring wasted work is therefore a method for quantifying work that is being done
315
by the ventricle but does not contribute to LV ejection, and provided the myocardium is
316
viable, should represent a measure of contractile reserve. This was supported by our findings
317
in patients with LBBB who showed a decrease in WWR and an increase in ejection fraction
318
when CRT was introduced (Table 2).
319
In a previous study wasted work was estimated as the difference between strain at end
320
systole and peak strain during or after systole (1). The main limitation of this method is that it
321
does not take into account pressure, which is higher during lengthening in late systole than
322
during post-systolic shortening. Therefore stretching of a segment before AVC represents a
323
greater degree of wasted work compared to shortening after AVC when pressure is rapidly
324
declining. Although this is an interesting index of mechanical dyssynchrony, it does not
325
provide a measure of work since it does not incorporate load. Our method incorporates a
326
measure of regional load (LVP), thereby taking into account the energetic effects of
327
deformation at different pressures. The importance of taking into account pressure was
328
illustrated by an additional data analysis in which we considered strain only as a marker of
329
wasted work. In this analysis, which is detailed in the appendix, we calculated work by
330
assuming a constant LV pressure, set to an arbitrary value of 1 mmHg. The result from this
331
analysis was that the wasted work and WWR differed substantially from results when cyclic
332
LV pressure was used, as illustrated in Figure 5. The septum which had a wasted work ratio
15
Russell et al.
333
larger than one in most segments during LBBB, and therefore performed more negative than
334
positive work, appeared to do predominantly positive work when looking at strain only.
335
Similarly, the effect of CRT as reflected in the difference in WWR between the septum and
336
lateral became far less pronounced when ignoring pressure (Figure 5 B and C). We therefore
337
propose that our method which incorporates a LV pressure estimate, gives important
338
additional information about regional myocardial function.
339
Since the myocytes are unable to recycle chemical energy when the segments are
340
stretched, this energy will instead be converted to heat in the tissue. Measurements of WWR
341
will for this reason indicate the overall efficacy of the LV myocardium in converting
342
metabolic substrates into cardiac work. Quantification of regional work in ventricles with
343
uncoordinated contractions may provide important insights into mechanisms of remodelling
344
and progressive LV dysfunction (1; 6; 7; 14).
345
346
347
348
Limitations
When estimating the LV pressure curve by the non-invasive method, we use an empiric
349
standardized reference curve based on several invasively recorded pressure traces. One may
350
argue that a mathematically derived curve profile determined by regression analysis may be
351
used instead of a tabulated reference curve. A principal limitation of both concepts is that the
352
waveform might be inaccurate for some pathologies. The waveform found by regression
353
analysis will in addition contain errors caused by limitations in the number of degrees of
354
freedom used in the regression function. Using an empiric standardized reference curve based
355
on averaging of invasive measurements is for this reason expected to give a more accurate
356
profile, supporting its use over pressure traces defined as mathematical functions.
357
Furthermore, the method has been thoroughly validated and it performs well under a wide
16
Russell et al.
358
range of different hemodynamic conditions (9). However, for minor changes in regional work
359
the lack of a true LV pressure represents a limitation. The accuracy of the estimated LV
360
pressure curve may be improved by including data from arterial tonometry, and by using
361
specific reference curves for different patient populations.
362
The LV pressure-strain analysis does not provide a direct measure of work since
363
radius of curvature and wall thickness are not incorporated in the calculations. Therefore,
364
comparison of absolute positive or negative work estimates between different ventricles may
365
not be valid. Comparison of WWR between different hearts, however, should be valid
366
measure since it is a relative measure which compensates for limited information about local
367
geometry. The reproducibility for WWR was not assessed in the present paper, however,
368
interobserver variability for both strain and non-invasive pressure estimate (9) has previously
369
been evaluated.
370
The experimental dataset and the dataset from patients with simultaneous LVP
371
measurements and strain analysis has also been used in a previous publication from our group
372
(9).
373
The assessment of regional work proposed in the current study relies on strain rate
374
based on STE and relatively low time resolution may therefore represent a limitation. In the
375
current study we had a frame rate 64±12 s-1 in the experimental part and found this to be
376
sufficient when compared to results using sonomicrometry.
377
378
Conclusions
379
The present study introduces a non-invasive clinical method to quantify the negative impact
380
of dyssynchrony on myocardial work and energy utilization. Studies should be done to
381
determine if this new method will be useful in risk stratification and in the selection of
382
therapy for patients with dyssynchrony.
17
Russell et al.
383
384
Appendix
385
- Calculation of segmental work
386
Work of individual segments was calculated from the strain recordings and the left
387
ventricular pressure. Instead of calculating only one value of work in a completed cardiac
388
cycle for each segment, equal to the area of the strain – pressure loop, work was calculated as
389
a function of time throughout the cardiac cycle. This was done by obtaining the rate of
390
segmental shortening (strain rate) by differentiation of the strain curve (Figure 2A), and
391
multiplying this with instantaneous left ventricular pressure (Figure 2B). This resulted in a
392
measure of instantaneous power, which was integrated over time by the trapezoidal method to
393
give work as a function of time (Figure 2C). Performing this calculation for a complete cycle
394
is mathematically the same as calculation of the strain – pressure loop area according to
395
Green’s theorem.
396
In detail:
397
Let PLV(t) be the recorded or estimated pressure in the left ventricle as a function of
398
time, and ε(t) be the strain of a myocardial segment as a function of time.
399
The segment shortening rate (strain velocity) is:
400
401
402
1)
v(t ) = −
dε
(t )
dt
- and the power developed by the segment is:
2)
P(t ) = v(t ) PLV (t ) = −
dε
(t )PLV (t )
dt
18
Russell et al.
403
The work performed by the segment as a function of time is the integral of power from t’ = 0
404
to t’ = t:
405
3)
t
t
0
0
W (t ) =  P(t ′)dt ′ =  −
dε
(t′) PLV (t′) dt′
dt
406
- where t’ is an integration variable.
407
- Calculation of WWF using constant pressure.
408
To highlight the importance of changes in regional load when assissing work we performed
409
calculations by using the non-weighted sum of segment shortening as a surrogate for positive
410
work, and the sum of elongation as negative work. This is equivalent to using our calculation
411
method as described in the paper, but assuming a constant pressure of 1 instead of a time-
412
varying pressure. The calculation were performed for ejection and isovolumic contraction and
413
relaxation phases using raw data from the same patients before and after CRT. The bull eye
414
plots in figure 5 show that when WWR is calculated using a constant pressue (C) the
415
difference in wasted work ratio between the septum and lateral wall was significantly
416
reduced, and the effect of CRT became far less pronounced.
417
Sources of funding
418
419
Dr Russell was recipient of a clinical research fellowship from the University of Oslo.
Acknowledgements
420
421
We would like to thank Professor John V. Tyberg for his insightful comments.
Disclosures
422
423
None.
19
Russell et al.
424
Reference List
425
426
1. Delhaas T, Arts T, Prinzen FW and Reneman RS. Regional fibre stress-fibre
427
strain area as an estimate of regional blood flow and oxygen demand in the
428
canine heart. J Physiol 477 ( Pt 3): 481-496, 1994.
429
2. Gjesdal O, Remme EW, Opdahl A, Skulstad H, Russell K, Kongsgaard E,
430
Edvardsen T and Smiseth OA. Mechanisms of abnormal systolic motion of
431
the interventricular septum during left bundle-branch block. Circ Cardiovasc
432
Imaging 4: 264-273, 2011.
433
3. Heusch G, Rose J, Skyschally A, Post H and Schulz R. Calcium
434
responsiveness in regional myocardial short-term hibernation and stunning in
435
the in situ porcine heart. Inotropic responses to postextrasystolic potentiation
436
and intracoronary calcium. Circulation 93: 1556-1566, 1996.
437
4. Hisano R and Cooper G. Correlation of force-length area with oxygen
438
consumption in ferret papillary muscle. Circ Res 61: 318-328, 1987.
439
5. Klimusina J, De Boeck BW, Leenders GE, Faletra FF, Prinzen F, Averaimo
440
M, Pasotti E, Klersy C, Moccetti T and Auricchio A. Redistribution of left
441
ventricular strain by cardiac resynchronization therapy in heart failure patients.
442
Eur J Heart Fail 13: 186-194, 2011.
20
Russell et al.
443
6. Prinzen FW, Augustijn CH, Arts T, Allessie MA and Reneman RS.
444
Redistribution of myocardial fiber strain and blood flow by asynchronous
445
activation. Am J Physiol 259: H300-H308, 1990.
446
7. Prinzen FW, Cheriex EC, Delhaas T, van Oosterhout MFM, Arts T, Wellens
447
HJJ and Reneman RS. Asymmetric thickness of the left ventricular wall
448
resulting from asynchronous electric activation: A study in dogs with ventricular
449
pacing and in patients with left bundle branch block. American Heart Journal
450
130: 1045-1053, 1995.
451
8. Prinzen FW, Hunter WC, Wyman BT and McVeigh ER. Mapping of regional
452
myocardial strain and work during ventricular pacing: experimental study using
453
magnetic resonance imaging tagging. J Am Coll Cardiol 33: 1735-1742, 1999.
454
9. Russell K, Eriksen M, Aaberge L, Wilhelmsen N, Skulstad H, Remme EW,
455
Haugaa KH, Opdahl A, Fjeld JG, Gjesdal O, Edvardsen T and Smiseth OA.
456
A novel clinical method for quantification of regional left ventricular pressure-
457
strain loop area: a non-invasive index of myocardial work. Eur Heart J 33: 724-
458
733, 2012.
459
10. Skulstad H, Edvardsen T, Urheim S, Rabben SI, Stugaard M, Lyseggen E,
460
Ihlen H and Smiseth OA. Postsystolic shortening in ischemic myocardium:
461
active contraction or passive recoil? Circulation 106: 718-724, 2002.
21
Russell et al.
462
463
464
11. Suga H. Total mechanical energy of a ventricle model and cardiac oxygen
consumption. Am J Physiol 236: H498-H505, 1979.
12. Takaoka H, Takeuchi M, Odake M and Yokoyama M. Assessment of
465
myocardial oxygen consumption (Vo2) and systolic pressure-volume area
466
(PVA) in human hearts. Eur Heart J 13 Suppl E: 85-90, 1992.
467
13. Urheim S, Rabben SI, Skulstad H, Lyseggen E, Ihlen H and Smiseth OA.
468
Regional myocardial work by strain Doppler echocardiography and LV
469
pressure: a new method for quantifying myocardial function. Am J Physiol
470
Heart Circ Physiol 288: H2375-H2380, 2005.
471
14. van Oosterhout MF, Prinzen FW, Arts T, Schreuder JJ, Vanagt WY,
472
Cleutjens JP and Reneman RS. Asynchronous electrical activation induces
473
asymmetrical hypertrophy of the left ventricular wall. Circulation 98: 588-595,
474
1998.
475
476
477
478
22
Russell et al.
479
480
481
482
Table 1.
Hemodynamic variables and work calculations for experimental study
Baseline
(n=6)
LBBB
(n=6)
Heart rate, bpm
121±17
122±13
QRS width, ms
68±5
116±7*
1284±198
1098±189*
LV EDP, mmHg
10±3
10±4
Peak LVP, mmHg
96±12
94±9
Wasted work ratio
0.13±0.05
0.37±0.11*
Global measures
Hemodynamic variables and WWR
LV dP/dtmax, mmHg/s
Regional measures
Strain and work
Septum
Lateral
Wall
Septum
Lateral
wall
-15±4
-14±4
-10±4*
-11±7
Positive work, mmHg•%
1106±368
917±256
669±275
1295±615†
Negative work, mmHg•%
155±134
164±102
320±288
236±170
Net Work, mmHg•%
951±320
756±266
349±346*
1059±516*†
Segmental wasted work ratio
0.14±0.11
0.19±0.15
0.52±0.49*
0.18±0.09†
Peak strain %
Work analysis are calculated from longitudinal strain by speckle tracking echocardiography and invasively measured
LVP; Septum, mid septal and anterioseptal segments; Lateral wall, mid lateral and posteriolateral segments; Values are
mean±SD. LBBB, left bundle branch block; EDP, end diastolic pressure; *P<0.05 vs. baseline. † P<0.05 septum vs.
lateral wall
483
23
Russell et al.
484
485
486
487
488
489
Table 2.
Hemodynamic variables and non-invasive work calculation for CRT and
control group of the clinical study
Normals
(n=20)
LBBB
(n=10)
CRT
(n=9)
Heart rate, bpm
63±9
67±18
59±5
Age, years
63±8
65±7
65±6
Ejection fraction, %
60±4
29±6
42±14*
QRS width, ms
90±8
171±16
145±13
0.09±0.03
0.36±0.16
0.17±0.07*
Global measurements
Hemodynamic variables and
WWR
Global Wasted work ratio
Regional measurements
Strain and work
Septum
Lateral
wall
Septum
Lateral
wall
Septum
Lateral
wall
-20±4
-19±3
-6±3
-12±5
-12±5*
-12±5
Positive work, mmHg•%
2168+465
2084+327
434+202
1572+451†
1328+489*
1225+498
Negative work, mmHg•%
191+104
170+101
591+350
263+162†
212+120*
242+136
Net Work, mmHg•%
1978+483
1913+307
-158+384
1308+460†
1116+489*
983+505
Segmental wasted work ratio
0.09±0.06
0.08±0.05
1.57±0.97
0.18±0.11
0.18±0.09*
0.23±.014
Peak strain %
Work analyses are calculated from longitudinal strain by speckle tracking echocardiography and estimated LV pressure
curve. Septum, mid septal and anteroseptal segments; Lateral wall, mid lateral and posterolateral segments; Values are
mean±SD. LBBB, left bundle branch block; EDP, end diastolic pressure; *P<0.05 LBBB vs.BVP. † P<0.05 septum vs.
lateral wall
24
Russell et al.
490
491
492
Figure legends
493
494
Figure 1. Work of individual segments was calculated from the strain recordings and the left
495
ventricular pressure. Work was calculated as a function of time throughout the cardiac cycle.
496
This was done by obtaining the rate of segmental shortening (strain rate) by differentiation of
497
the strain curve (Figure 2A), and multiplying this with instantaneous left ventricular pressure
498
(Figure 2B). This resulted in a measure of instantaneous power, which was integrated over
499
time to give work as a function of time (Figure 2C).
500
501
Figure 2. Estimated LV pressure (upper panels), strain (middle panels) and calculation of
502
wasted work for septum and lateral wall (lower panels) in a representative patient with left
503
bundle branch block (LBBB) (left panel) and after treatment with cardiac resynchronization
504
therapy (CRT) (right panel). Vertical lines indicate valvular events measured by
505
echocardiography. Negative and positive work during systole is marked as red and black,
506
respectively. Wasted Work Ratios were calculated using cumulated work from mvc to mvo.
507
508
Figure 3. Representative traces from a patient with LBBB showing pressure-strain loops by
509
LVP and speckle tracking echocardiography (STE) (black line) vs. estimated LVP and STE
510
(red dotted line) for a later wall segment (A) and a septal segment (B). C. Correlation between
511
WWR by estimated LVP and STE vs. measured LVP and STE.
512
25
Russell et al.
513
Figure 4. Work analysis by LVP and speckle tracking echocardiography (STE) (solid line) vs.
514
the non-invasive method by estimated LVP and STE (dashed line), for a septal and lateral wall
515
segment in patient with the CRT device turned on and off.
516
517
Figure 5. Bulls eye plots showing segmental wasted work ratios. A: Plots from a normal
518
control and a patient with LBBB. B: Regional mean values for all patients before and after
519
CRT. WWR values > 1 implies more negative than positive work. C: Regional mean values
520
for all patients before and after CRT when assuming constant pressure and considering
521
changes in strain only.
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
26
Russell et al.
538
539
27
-100
0
100
-10
0
0
100
0
”Work”
1000
(mmHg·%)
2000
0
”Power” 5000
(mmHg·%/s)
Segment
shortening rate
(%/s)
Segment strain
(%)
LVP
(mmHg)
MVC
0.1
AVO
Time (s)
0.4
AVC
MVO
0.8
C.Integrate
by LVP
B.Multiply
change sign
A.Differentiate,
Calculation steps
LVP (mmHg)
Before CRT
150
100
50
Septum
Strain (%)
0
-5
Septum
-10
Lateral
wall
-15
Work (mmHg·%)
After CRT
Lateral
wall
Ratio 0.11
Lateral
Wall
1500
Lateral
Wall
Ratio 0.09
Ratio 0.08
1000
500
Septum
Septum
0
0
0.2
mvc
avo
Ratio 1.51
0.4
Time (s)
avc mvo
0.6
0
0.2
mvc avo
Time (s)
0.4
avc mvo
0.6
Regional Wasted Work Ratio
A.
0
0.05
0.13
0.04
0.10
0.05
0.25
0.04
0.78
0.06
1.32
0.06
3.72
0.00
4
0
Normal control
1.27
4
LBBB
Estimated LV pressure
B.
0
0.13
0.11
0.15
0.13
0.34
0.11
0.38
0.18
0.34
0.12
0.28
0.18
0.27
1.9
0
Before CRT
0.12
1.9
After CRT
Constant pressure
C.
0
0.29
0.36
0.29
0.36
0.50
0.39
0.48
0.40
0.43
0.30
0.39
0.34
0.43
Before CRT
1.9
0
0.30
After CRT
1.9