- Email: [email protected]
Ischemic myocardial hysfunction assessed by temporal Fourier transform of regional myocardial wall thickening
January
Cogswell, Sagar, and Wann
23.
24.
25. 26. 27.
28.
Amatlcan
diographic features in 30 patienta. Am J Cardiol1979;44:401412. Keren G, Belhassen B, Sherez J, Miller HI, Megidish R, Berenfeld D, Laniado S: Apical-hypertrophic cardiomyopathy: Ev&ation by noninvasive and invasive techniques in 23 patienti. Circulation 1985;71:45-56. Harrison DC, Braunwald E, Glick G, Mason DR, Chidsey CA, Ross J Jr: Effects of beta adrenergic blockade on the circulation, with particular reference to observations in patients with hypertrophic subaortic stenosis. Circulation 1964;29:8498. Flamm MD, Harrison DC, Hancock EW: Muscular subaortic stenosis: Prevention of outflow obstruction with propranolol. Circulation 1968;38:846-858. Adelman AG, Shah PM, Graniak R, Wigle ED: Long-term propranolol therapy in muscular subaortic stenosis. Br Heart J 1970;32:804-811. Landmark K, Sire S, Thaulow E, Amlie JP, Nitter-Hauge S: Hemodynamic effects of nifedipine and propranolol in patients with hypertrophic obstructive cardiomyopathy. Br Heart J 1982;48:19-26. Kaltenbach M, Hopf R, Kober G, Bussmann WD, Keller M, Petersen Y: Treatment of hypertrophic obstructive cardiomyopathy with verapamil. I+ Heart J 1979;42:35-42.
Heart
1967 Journal
29. Rosing DR, Kent KM, Borer JS, Seides SF, Maron FJ, Epstein SE: Verapamil therapy: A new approach to the pharmacologic treatment of hypertrophic cardiomyopathy. I. Hemodvnamic effects. Circulation 1979:60:1201-1207. 30. Bonow RO, Roaing DR, Bacharach SL, direen MV, Kent KM, Lipson LC, Maron BJ, Leon MB, Epstein SE: Effect of verapamil on left ventricular systolic function and diastolic filling in patients with hypertrophic cardiomyopathy. Circulation 1981;64:787-795. 31. Levine RA, Weyman AE: Dynamic subaortie&&ruction in hypertrophic cardiomyopathy: Criteria and controversy. J Am Coll Cardiol 1985;6:16-18. 32. Kinoshita N, Nimura Y, Okmoto M, Miyatake K, Nagata S, Sakakibara H: Mitral regurgitation in hypertrophic cardiomyopathy. Non-invasive study by two-dimensional Doppler echocardiography. Br Heart J 1983;49:574-583. 33. Wigle ED, Adehnan AG, Auger P, Marquis Y: Mitral regurgitation in muscular subaortic stenosis. Am J Cardiol 1969; 24698-706. 34. Criley JM, Siegel IU Has “obstruction” hindered our understanding of hypertrophic cardiomyopathy? Circulation 1985; 721148-1154.
A Fourier analysis including the first 20 harmonics was performed on sonomkrometric measurements of regIonal myocardid wall thkkneas in eight consciour dogs under control conditions and at four levels of ischemia produced by a hydraulic occluder on the iefl clrcumfiex coronary artery. Systolc watt thickening was reduced from 28.47 + 6.20% (SD.) (control) to 28.05 rt 5.73% (mild atenoris), 17.W + 5.86% (moderate stenosis), 11.48 + 3.56% (severe aten+s), and 3.59 f 2.57% @O-second occlusion), values SlgMcantly dtfferent from each other @ < 0.01). The amplitude of the first harmonic decreased stepwke from 1.35 + 0.31 to 1.08 + 0.29 mm, 0.90 + 0.27 mm, 0.69 5 0.24 mm, and 0.43 + 0.12 mm, atl signifkantly dihereqt from each other (p < 0.05). These amplitude values correlated to percent systolic wall thickening (r= 0;8W, p = 0.001). A phase shift of the first harmonk from 13t + 11 to 139 + 14 egreer, 13 +‘I& degrees @ < 0.05 vs control), 161 5 21 degrees (p < 0.01 vs control), and 191 + 21 degrees (p < 0.01 vs control and severe stenosis) correlated with the increaes in time from end dlqtoh tb the point of maxhnum wall excursion (r = 0.662, p < 0.001). These data Indicate t&t the extent of ischem~c regional myocardial hypoktnesis can be adequately described by the am@tude of the first hermonk, and that the asynchrony of ventrkular contraction end relexstion can be detected from the phese of the first harmonic. (AM HEART J 198y; 113:lW.)
Gerd Hewh, M.D.,* Brian D. Guth, Ph.D., Thomas Widmann, M.D., Kirk L. Peterson, M.D., and John &as Jr., M.D. Sun Diego and La Jolla,
From the Division of Cardiology, Department of Medicine, University California, San Diego, School of Medicine. Supported hy a National Institute of Health (Specialized Center Research on Iachemic Heart Disease) Grant No. HL-17682. Received for publication Feb. 28, 1986; accepted May 26, 1986.
116
of of
C&f.
Reprint requests: John Ross, Jr., M.D., Division of Cardiology, Department of Medicine, M-013, University of Caliiornia, San Diego, La Jolla, CA 92093. *Dr. Gerd Heusch is the recipient of a research fellowship granted by the German Research Foundation.
Volume
113
Number
1
With the use of Fourier analysis, any periodic function can be represented as the sum of multiple sine and cosine waves of different frequencies. Each of these harmonics is characterized by a certain phase and amplitude. lr2 Fourier analysis can be applied to the cardiac cycle, which constitutes a periodic sequence of contraction and relaxation at the fundamental frequency of the heart rate. Application of Fourier analysis to the cardiac cycle permits a functional analysis of images by the use of gated blood pool radionuclide ventriculograms’s 3-7 and contrast angiographic ventriculograms.2T 8.g Contraction abnormalities can thereby be characterized by a phase delay,1-Q and a reduction in amplitude7q8 of the first harmonic of a temporal Fourier transform. A Fourier analysis employing only the first, or the first few harmonics, is an effective and objective high-frequency cut-off filter of noisy original timeactivity or time-intensity curves. However, it is unclear how many harmonics are essential for a precise functional analysis. A study by Bacharach et aI.12 suggests that two harmonics are sufficient for quantitative characterization of global systolic left ventricular function, but up to six harmonics may be necessary to describe global diastolic function adequately in radionuclide studies. Rankin et al.” found that sonomicrometric ventricular dimension data, including wall thickness; are of relatively low frequency and can be approximated by the first five harmonics. However, this problem has not up to now been addressed with respect to changes in regional myocardial function occuring during ischemia. Furthermore, there is some controversy as to whether phase shifts or amplitude changes of a Fourier transform better reflect changes in regional myocardial walI function occurring during regional ischemia. Most studies have examined only the phase distribution of the first harmonic for the assessment of regional myocardial dysfunction, Pavel et aL6 found good agreement between the contrast angiographic classification of regional dysfunction and the phase delay in a gated blood pool study. In a contrast angiographic study of patients with subtle regional myocardial dysfunction distal to completely occluded coronary arteries,12 phasic geometric analysis was even more sensitive than quantitative assessment of walI motion in identifying the coronary hypoperfusion. In contrast, a previous study from this institution8 found contrast angiographic amplitude images more sensitive to left ventricular dyssynergy than phase images. Therefore, the purpose of the present study was to determine the value of a temporal Fourier transform
Fourier analysis of regional LV function
117
for the characterization of ischemic regional myocardial dysfunction by means of a highly precise measurement of regional wall function. In conscious dogs, regional myocardial wall thickening was measured by sonomicrometry under control conditions and at four levels of ischemic dysfunction. The results of a temporal Fourier transform, including the first 20 harmonics, were then compared to conventional parameters of regional wall function. To test the sensitivity of phase shifts of the first harmonic for detecting asynchrony of ventricular contraction and relaxation, which is associated with regional myocardial ischemia in experimentaP3 and clinical studies,12v14 we also quantified the degree of synchrony between wall thickening in nonischemic and ischemic regions. METHODS
Eight mongrel dogs, weighing between 21 and 35 kg, were instrumented during sterile surgery. They were premeditated with acepromazine maleate (0.5 mg/kg intramuscularly) 30 minutes before induction of anesthesia with sodium pentobarbital (25 mg/kg intravenously); further analgesiawas provided by morphine (0.5 mg/kg intravenously). During artificial ventilation with a Harvard respirator, the heart was exposed through a left thoracotomy in the fifth intercostal space and was suspended in a pericardial cradle. For measurementof left ventricular pressure,a high-fidelity transducer (Konigsberg Instruments Inc., Pasadena,Calif., model P7) and a fluid-filled tube (Tygon, 1.27 mm inner diameter) were inserted through a stab wound in the apex. The proximal left circumflex coronary artery wasdissectedfor 5 mm and a hydraulic occluder wasplaced around the vessel.For the measurementof regional contractile function, two pairs of miniature ultrasonic crystals (5 MHz) were implanted in the left ventricular free wall to measurewall thickness.15 One crystal (2 mm diameter) of each pair was inserted obliquely through the ventricular wall to the subendocardium. The secondcrystal (6 mm diameter) was attached to a Dacron patch and wassewnon the epicardial surface opposite the subendocardial crystal where the ultrasonic transit time between the two crystals was shortest, as observed on an oscilloscope(Tektronix, Inc., Beaverton, Oreg., model 453). The distance between the two crystals wascontinuously recorded with the transit time technique (Sonomicrometer, Triton Technology, San Diego, Calif.). To verify appropriate crystal orientation, the received signalswere alsomonitored on an oscilloscope.One pair of crystals waspositioned in the left ventricular anterior wall perfused by the left anterior descendingcoronary artery and servedascontrol; the other pair of crystals wasplaced in the zone of the left ventricular posterior wall perfused by the left circumflex coronary artery to be rendered ischemic. The appropriate alignment of each pair of crystals was confirmed at autopsy in all animals reported.15Upon histologic examination, the crystals were surrounded by a fibrous rim (approximately 1 mm) of
Heusch et al.
118
fable I. The effects of three degreesof coronary artery stenosisand a 30-secondcomplete coronary occlusion on the phase(degrees)of a temporal Fourier transform of regional myocardial wall thickening; zero degreeswas defined at end diastole n=8
Mild stenosis
Control
Ischemic
Moderate stenosis
Severe stenosis
JO-second occlusion
wall
1. Harmonic
2. Harmonic 3. Harmonic 4. Harmonic 5. Harmonic Nonischemic
137 221 215 126 206
k f + f iz
11 21 60 76 63
139 216 213 172 162
+ 14 + 24 + 80 k 114 zk 138
150 221 211 208 139
+ + k + +
15* 31 91 137 131
161 247 231 214 213
f -t f f ?
21t 48 41 132 98
191 + 21tg 200264 209+60 160 f 130 227 + 120
137 216 180 200 162
f f + f f
12 24 68 89 74
134 212 211 142 139
* k + k f
137 214 166 199 201
+ 14 t- 28 rf: 112 + 124 f 115
139 220 147 194 160
* + + + k
18 42 85 137 150
147 231 117 227 121
wall
1. Harmonic
2. Harmonic 3. Harmonic 4. Harmonic 5. Harmonic
13 27 89 145 117
f + + f +
14*j 33 96 85 124
Dataare means f S.D. *p < 0.05,
tp < 0.05,
tp < 0.01 “S Control. §p < 0.01 ve next lower
level
of iechemia.
II. The effects of three degreesof coronary artery stenosisand a 30-secondcomplete coronary artery occlusion on the amplitude (mm) of a temporal Fourier transform of regional myocardial wall thickening
Table
n=8 Ischemic
2. Harmonic 3. Harmonic 4. Harmonic 5. Harmonic 2. Harmonic 3. Harmonic 4. Harmonic 5. Harmonic lp
Severe stenosis
30-second occlusion
1.35 0.66 0.13 0.09 0.06
f 0.31 + 0.16 * 0.10 + 0.05 I!z 0.03
1.08 0.59 0.10 0.10 0.09
zk 0.29t + 0.13* + 0.08 + 0.06 + 0.03
0.90 0.43 0.12 0.14 0.10
k 2 f f f
0.27tz 0.1378 0.06 0.06*$ 0.04
1.04 0.50 0.14 0.08 0.05
+ f + r +
1.01 0.52 0.14 0.07 0.05
f k f f Ik
1.08 0.52 0.12 0.07 0.05
k 0.26 -t 0.14 f 0.07 * 0.03 z!z0.02
0.69 0.30 0.12 0.14 0.08
f 0.24t$ + 0.12tj + 0.06 k 0.06* AZ0.03
0.43 0.21 0.17 0.21 0.11
+ + + f k
0.12ts O.llfq 0.11 o.lot$ 0.04
1.10 0.53 0.12 0.05 0.03
k f + * 2
1.15 0.50 0.11 0.07 0.06
ii k & f k
0.30 0.16 0.09 0.03 0.03
wall
1. Harmonic
Data
Moderate stenosis
wall
1. Harmonic
Nonischemic
Mild stenosis
Control
+ S.D. tp < 0.01 “8 Control. §p < 0.01 ve next lower
0.18 0.11 0.07 0.03 0.02
0.16 0.14 0.09 0.02 0.02
0.30 0.19 0.12 0.03 0.01
are means
< 0.05, $p < 0.05,
level
of Lchemia.
connective tissue, as previously described.16After instrumentation, the chest was closed, the pneumothorax was evacuated, and the wires and tubes were passedsubcutaneously to the back of the animal and were exteriorized betweenthe scapulae.Ampicillin (6.6 gm/day) wasadministered for 3 days after surgery. Dogswere studied no earlier than 7 days after surgery, when they had recovered and were afebrile. Each dog was studied when it was resting quietly on the table, as it had been previously trained to do. The Konigsberggaugewas calibrated with a Statham (P23 Db) transducer connected to the fluid-filled tube, with zero reference taken at the estimated level of the right atrium, asthe dog waslying on its right side. After a control recording, a level of ischemic myocardial dysfunction was produced by adjustment of
the hydraulic occluder. After steady-state dysfunction was recorded, the occluder was completely releasedand complete recovery of regional myocardial function wasallowed to occur. The occluder was then reinflated to produce a different degree of ischemic wall dysfunction, and three different degrees of a steady-state dysfunction due to partial coronary stenosiswere so obtained. In addition, a 30-second complete coronary occlusion was performed. These four levels of regional myocardial dysfunction (subsequentlytermed mild, moderate, and severestenosis, and coronary occlusion) were produced in random order. Recordings during each experiment were made on a forced-ink recorder (Brush, Cleveland, Ohio) and on magnetic tape for subsequentplayback and computation. The systemic hemodynamic parameters analyzed were
volume Number
113 1
Fourier analysis of regional LV function
119
III. The effects of three degreesof coronary artery stenosisand a 30-secondcomplete coronary artery occlusion on systemic hemodynamics
Table
Control
Mild stenosis
Moderate stenosis
Severe stenosis
HR @pm) PLVSP
98 + 15 136 f 7
93 rf: 15 132 + 8
99 f 12 134 f 14
102 * 17 132 + 9
103 + 17 140 + 7
(mmHg) LVEDP
13.6 k 4.7
13.7 f 4.9
15.6 k 7.9
14.8 I!Z3.7
20.8 + 7.2?$
3146 + 396
2917 f 451*
2875 f 407*
2844 f 351*
2732 + 319t
1718 rt 161
1561 + 215*
1598 + 250*
1423 + 184fJ
1359 * 201t
n=8
30 second occlusion
(mm Hg)
(+) dpldt (mmHg/sec) (-) dp/dt (mmHg/sec)
Data are means T S.D. HR = heart rate: PLVSP - Dealt left ventricular svetolic meeaure: l p < 0.05, tp < 0.01 VB corlt-rol. tp < 0.05, $p< 0.01 vs next lower level of iachemia.
LVRDP
heart rate, left ventricular end-diastolic and peak systolic pressures,and peak positive and negative left ventricular dp/dt. Regional myocardial wall thicknesswasdetermined at end diastole (defined asthe time dpldt crossedthe zero line) and at end systole (defined as the time of maximum excursion at end ejection within 20 msec before peak negative dp/dt, according to Theroux et al.‘? The maximum wall thickening excursion without reference to the time of its occurrence within the cardiac cycle was determined, since the amplitude values of the Fourier analysis are also without reference to the time within the cardiac cycle. The systolic wall thickening excursion (up to end ejection) and the extent of wall thickening occuring after end systole (post ejection thickening) were also determined. Systolic wall thickening wasalso calculated as the percent of end-diastolic wall thickness. In addition to these measures of the extent of thickening, the time between end diastole and the point of maximum wall thickening and the mean systolic wall thickening velocity were determined. Finally, the loop area of a plot of wall thickness vs left ventricular pressurewas calculated.‘* Hemodynamic and dimensiondata were digitized from magnetic tape with a computer system (PDP 11/03, Maynard, Masss.), and data from 20 consecutive cardiac cycles were collected at 5 msec intervals. Card& cycles, with an interval between sequential peak positive dp/dt values differing by lessthan 40 msec,were selected and averaged. Eight to 20 beats were averagedin this way for each lnterven$ion. In addition to standard measures of regional wall function describedabove, a temporal Fourier transform of regional wall thickening including the first 20 harmonics was performed. Zero degreeswas defined at end diastole. Data are presented as mean values with their standard deviations. Statistical analysiswasperformed by meansof an analysis of variance for repeated measurementsand, when a significant interaction was detected, single mean values were compared with Tukey’s range test.lg In addition to the analysis of differences between the arbitrarily defined degreesof myocardial ischemia,linear regression
= left ventricular
end-diastolic
preeeure.
analyseswere performed taking the data points of all dogs together as well asthose for eachindividual dog separately. A level of p < 0.05 was accepted as signillcant. RESULTS
The control conditions for hemodynamics, regional dimensions, and Fourier analysis values before each level of ischemic myocardial dysfunction were not different, and therefore they were combined. The sixth to twentieth harmonics of the temporal Fourier transform of regional wall thickening represented only “noise”; these harmonics were characterized by phase values with a high standard deviation (SD. > 50% of the mean value) and by low amplitude values (amplitudes < 5% of the first harmonic under nonischemic conditions). Therefore only phase and amplitude values of the first five harmonics are presented in Tables I and II. Hemodynamlcs. Regional myocardial ischemia induced only modest changes in systemic hemodynamics. Whereas heart rate and peak left ventricular systolic pressure were not significantly changed, left ventricular end-diastolic pressure was siguificantly increased during complete coronary occlusion, and peak positive and negative dp/dt were reduced with increasing severity of ischemia (Table III). Regional tunctlon: Conventional measures. A representative original recording of regional myocardial wall thickening in the ischemic and the nonischemic regions is shown in Fig. 1. Increases in the severity of regional myocardial ischemia were evidenced by progressive reductions in maximum wall thickening and systolic wall thickening, percent systolic wall thickening,
mean systolic wall thickening
velocity,
and the regional wall thickness-left ventricular pressure loop area (Table IV). End-diastolic wall thickness showed no significant changes and end-systolic
120
Heusch et al.
Am6rlcan
January 1887 ll66rt Journal
Anwi wall ThiaRerr. Cmm)
Fig. 1. Representative original recording showing changes in regional myocardial wall thickening occurring with increasing severity of myocardial ischemia in the ischemic and nonischemic region. Systolic wall thickening of the ischemic region progressively decreases and a significant post ejection thickening develops. Systolic thickening of the nonischemic region increases during complete coronary occlusion.
Table IV. The effects of three degrees of coronary on regional myocardial dimensions n-8
Control
Ischemic wall EDT (mm) EST (mm) MT (mm) ST (mm) PET (mm) ST (%I TMT (msec) MSTV bmhec) TLVPLA (mm x mm Hg) Nonischemic wall EDT (mm) EST (mm) MT (mm) ST (mm) PET (mm) ST (%) TMT (msec) MSTV (mm/set) TLVPLA (mm x mm Hg)
12.31 15.46 3.37 3.17 0.02 26.47 220 15.81
+ + + + + -t + k
1.58 1.72 0.65 0.66 0.03 6.20 23 3.39
340 + 65
11.24 13.59 2.79 2.51 0.01 22.08 219 12.02
+ f 2 + * + f +
2.03 2.37 0.58 0.65 0.03 3.88 25 2.23
259 * 31
artery
stenosis and a 30-second
Mild stenosis 12.25 14.90 2.91 2.64 0.05 22.05 232 12.56
+ + + + + i k IT
1.78 1.76 0.527 0.57t 0.08 5.73t 29 3.53t
289 + 56t
10.66 13.13 2.78 2.55 0.01 23.01 220 12.09
+ f k f + + f *
1.72 2.07 0.63 0.56 0.03 3.67 17 2.72
262 + 44
Data ara means 2 S.D. thickness; EST = end-systolic thickness; MT = maximum (% ) = systolic waU thickening expressed as percent of end-diastolic thickness; velocity; TLVPLA = thickness-left ventricular pressure loop area.
EDT = end-diastolic
*p< 0.05,tp < 0.01 $p< 0.05, §p< 0.01
“8 Control.
vs next lower level of ischemia.
Moderate stenosis 12.45 14.51 2.37 2.06 0.07 17.00 250 10.06
? * f * k r * *
1.09 1.99* 0.56t# 0.6ot~ 0.06* 5.8q§ 35t 3.27ts
220 k 66t§
11.58 14.09 2.82 2.51 0.01 21.95 218 12.07
f + + f * k r?r +
2.73 3.23 0.69 0.56 0.03 2.54 21 2.84
255 f 38
complete
coronary
Severe stenosis 11.96 13.32 1.91 1.39 0.20 11.46 278 6.34
+ 1.58 f 1.4879 + 0.56t9 AZ 0.49?$ f 0.15t & 3.56?§ + 29f$ + 2.4575
160 iz 48tfj
11.71 14.26 2.74 2.82 0.01 24.49 216 12.18
A 2.66 f 3.12 ZIZ 0.68 f 0.73 + 0.03 f 4.926 A 15 f 3.89
255 k 40
artery
occlusion
30-second occlusion
11.86 12.29 1.48 0.44 0.81 3.69 305 2.34
2 * f * f + + +
1.63 lAq§ 0.49?$ 0.32ts 0.28?§ 2.57t§ 16tj 1.7175
89 + 15t§
10.61 13.60 2.83 2.51 0.01 24.25 224 12.62
+ + + + + f + k
1.83 2.32 0.64 0.46 0.03 3.55* 17 3.15
264+24
thickening; ST = systolic thickening; PET = post ejection thickening; ST TMT = time to maximum thickening; MSTV = mean systolic thickening
Volume4 113 Numbr 1
Wall1
Fourier analysis of regional LV function
121
rhickner
Fig. 2. Representative reconstruction of the digitized wall thickening by the first five Fourier harmonics. The reduction in the extent of severity of ischemia is well reflected by the amplitude of only the first however, is only reflected after including the information of the third, With increasing severity of ischemia there is a delay in the phase of
wall thickness diminished (not significant for mild stenosis). The time from end-diastole to the point of maximum wall excursion increased progressively and a significant post ejection thickening developed (both not significant for mild stenosis). Only with the severe coronary stenosis and complete coronary occlusion was there a significant increase in percent systolic wall thickening of the nonischemic region, which was not reflected in other parameters of nonischemic wall function (Table IV). Regional function: Fourier analysis. A typical example of the reconstruction of ischemic wall thickening by the flrat few harmonics of a Fourier transform is shown in Fig. 2. Phase analysis. The phase values of only the first harmonic exhibited a phase delay with increasing severity of myocardial ischemia. This phase delay only became significant with moderate and severe stenosis and complete coronary occlusion. Furthermore, the phase delay did not distinguish between the different levels of ischemic dysfunction, except for complete occlusion (Table I). The phase of the first harmonic correlated to the time from end diastole to the point of maximum wall thickening: y = 0.301 X + 65.267, r = 0.662, p < 0.001 (by use of all data points), and r = 0.815 & 0.080 (by use of the average of the correlation coefficients from all eight dogs). The difference between the phases of the first harmonic of the posterior (ischemic) and the anterior (nonischemic) myocardium correlated closely to the difference in time from end diastole to the point of maximum wall thickening between both regions: y = 0.384 x + 1.431, r = 0.748, p < 0.001 (all data points), and r = 0.869 f 0.109 (average of eight dogs).
in the ischemic myocardial region systolic thickening with increasing harmonic. Post ejection thickening, fourth, and fifth harmonic as well. the first harmonic.
There also was a significant correlation between the phase of the first harmonic and percent systolic wall thickening: y = - 2.274 X + 173.2, r = -0.747, p < 0.001 (all data points), and r = -0.805 t 0.312 (average of eight dogs). The correlation between the phase of the first harmonic and the regional wall thickness-left ventricular pressure loop area, which contains the information not only about systolic function but the entire cardiac cycle, was not closer: Y = -0.145 x +171.6, r = -0.580, p < 0.005 (all data points), and r = -0.671 + 0.139 (average of eight dogs). Amplitude analysis. The amplitudes of both the first and the second harmonics were significantly reduced with increasing severity of myocardial ischemia, and there was a significant difference between each level of ischemia in the amplitude of the first two harmonics (Table II). Conversely, the reduction in the amplitude of the first two harmonics with increasing severity of myocardial ischemia was paralleled by an increase in the amplitude of the subsequent harmonics, which became sign&ant for the fourth harmonic except with mild stenosis (Table II). There was a good correlation between the amplitude of the first harmonic and maximum (r = 0.893), systolic (r = 0.939), and percent systolic wall thickening: y = 0.038 X + 0.274, r = 0.894, p < 0.001 (all data points) (Fig. 3), which was even closer for the individual animals: r = 0.963 f 0.048 (average of eight dogs). The correlation between the amplitude of the first harmonic and the mean systolic. wall thickening velocity was: y = 0.068 x +0.248, r = 0.959, p < 0.001 (all data points), and r = 0.988 f 0.041 (average of eight dogs). Again, the correlation between the amplitude of the first har-
January
122
Heusch et al.
Amarlcan
0
0
0
0 A 0 0
Q: F o1
10
20
A
o
Control Mild Stenosis Moderate Stenosis Severe Stenosis Occlusion
30
40
Systolic Wall Thjckening (%) Fig. 3. Graphic presentation of the correlation between
the amplitude of the first Fourier harmonic and percent systolic wall thickening with increasingseverity of regional myocardial ischemia.
manic and the regional wall thickness-left ventricular pressure loop area was not closer: y = 0.003 X + 0.182, r = 0.834, p < 0.001 (all data points), and r = 0.927 + 0.097 (average of eight dogs). The increase in the amplitude of the fourth harmonic correlated to the extent of post ejection thickening: y = 0.118 X + 0.106, r = 0.535, p < 0.01 (all data points), and r = 0.728 + 0.257 (average of eight dogs). There was no significant correlation between the extent of post ejection thickening and the amplitude of any other harmonic. DlSCUSSlON
This study was designed to compare Fourier analysis to conventional measures based on a highly accurate measurement of abnormal regional myocardial wall motion, and to determine how many harmonics are necessary for the quantitative assessment of regional myocardial wall function during ischemia. We also undertook to determine whether the phase or the amplitude of the Fourier analysis is more sensitive to regional ischemic dysfunction. Our data demonstrate that ischemic regional hypokinesis can be adequately quantified by only the first harmonic of a temporal Fourier transform, and that the amplitude of the tit harmonic correlates particularly closely to absolute and percent systolic wall thickening, which can be regarded as the reference standard for the assessment of ischemic regional dysfunction.
Heart
1987 Journal
Our experimental study is based on a different measurement of regional myocardial wall function (sonomicrometry) from that used in clinical studies employing radionuclide or contrast venticulography. This approach nevertheless confirms the value of Fourier analysis for the characterization of ischemic regional myocardial dysfunction.5~gIn accordance with the previous clinical study by Widmann et al.,8 our data suggest that the amplitude image should be more sensitive for the detection of regional myocardial ischemic dysfunction than the phase image. Moreover, a very close correlation between the amplitude of the first harmonic and percent systolic wall thickening and wall thickening velocity was evident within any individual dog. Thus the regional amplitude of the first harmonic may be particularly well suited for the evaluation of therapeutic procedures such as angioplasty. In addition, the progressive phase delay with increasing severity of myocardial &hernia described quantitatively the asynchrony of left ventricular contraction and relaxation which, in addition to the reduction in regional wall motion, contributes to the reduction in left ventricular pump function. Under control conditions at a heart rate of approximately 100 bpm, a phase value of the f&t harmonic of 137 degrees corresponds closely to the end of ejection into the aorta, which occurs about 220 msec after end diastole. The progressive phase delay with increasing severity of &hernia reflects regional post ejection wall thickening, which does not contribute to the left ventricular external work. In clinical studies with the use of phase analysis of the first harmonic of radionuclide ventriculograms, such loss of synchrony appeared to be relatively specific for ischemic heart disease as compared to wall motion abnormalities associated with valvular disease,5 although similar phase delays may be observed during abnormal electrical activation.2 Our experimental data confirm the value of the phase of the first harmonic for reflecting the degree of asynchrony of left ventricular contraction and relaxation, as evidenced by the close correlation to the difference in time from end diastole to the point of maximum wall excursion between the ischemic and the nonischemic myocardial regions. However, in our study this ischemia-related asynchrony was not quite as sensitive as the extent of hypokinesis in reflecting the degree of myocardial ischemia. This finding might be explained by the more sensitive analysis of mild regional hypokinesis by sonomicrometry compared to contrast ventriculography,12 rather than lack of value of phase analysis for the assessmentof regional myocardial ischemia. Thus, most ventricu-
Vohtmb Number
113 1
lographic analyses use minimum volume or maximum inward wall motion for defining end systole, whereas end ejection is identified in the present study and post ejection wall thickening is not included. The increasing extent of post ejection thickening with increasing severity of ischemia was also reflected by an increasing amplitude of the fourth harmonic. However, this correlation was not very close (r = 0.535), and a 40.5:fold ‘increase in post ejection thickening from control to complete coronary occlusion was only reflected by a 2.3-fold increase in the amplitude of the fourth harmonic. Presumably, major parts of the post ejection thickening information are included in other harmonics as well, although the amplitude of no single other harmonic correlated significantly to the extent of post ejection thickening. Amplitudes of the first harmonic in radionuclide or contrast ventriculograms are without reference to the time within the cardiac cycle and therefore may overestimate the extent of systolic excursion. However, since post ejection thickening was significantly correlated only to the amplitude of the fourth harmonic and, on the other hand, the amplitude of the first harmonic was closely correlated to systolic wall thickening, this concern appears to be negligible. Neither the amplitude nor the phase of the first harmonic correlated to the wall thickness-left ventricular pressure loop area better than to systolic wall thickening, although they similarly contain information from the entire cardiac cycle, whereas the calculation of systolic wall thickening is based on only two points in time. This lack of a closer correlation between the loop area and the Fourier analysis is most likely explained by the fact that the calculation of loop area is based on a regional wall thickness and a global left ventricular pressure, and the latter does not reflect the active regional pressure development or stress during ischemia, as pointed out previously.1s In conclusion, our experimental data emphasize the value of a Fourier analysis for the assessmentof acute regional myocardial ischemic dysfunction. The phase of the first harmonic reflects the degree of synchrony of left ventricular contraction and relaxation, but the amplitude of the first harmonic appears to be sufficient to quantify the degree of regional dysfunction adequately. Chronic wall motion abnormalities often resemble those during acute ischemia,20and it might be expected that this type of analysis would be applicable in a chronic setting as well. These data may be relevant to
Fourier analysis of regional LV function
123
clinical studies employing functional image analysis of radionuclide or contrast ventriculograms, since they justify the use of only the first Fourier harmonic and thereby allow a substantial reduction in computation time. Our data may futhermore be relevant to all other methods employing a dynamic analysis of myocardial dimensions, such as M-mode echocardiography and fast computer tomography. REFERENCES
1. Links JM, Douglass KH, Wagner HN. Patterns of ventricular emptying by Fourier analysis of gated blood-pool studies. J Nucl Med 1980,21:978. 2. Widmann TF, T&au JF, Ashburn WL, Bhargava V, Higgins CB, Peterson KL. Evaluation of regional wall motion by phase and amplitude analysis of intravenous contrast ventricular fluorangiography: Technical aspects and computation. In: Sigwart U, Heintsen PH, editors: Ventricular wall motion. Stuttgart-New York Georg Thieme Verlag, 19&1:24. 3. Adam WE, Tarkowska A, Bitter F, Stauch M, Geffers H. Equilibrium (gated) radionuclide ventriculography. Cardiovast Radio1 1979;2:161. 4. Botvinick E, Dunn R, Frais M, 0 Connell W, Shosa D, Her&ens R, Scheinmann M. The phase image: Its relationship to patterns of contraction and conduction. Circulation 1982;65:551. 5. Ratib 0, Henze E, Schoen H, Schelbert HR. Phase analysis of radionuclide ventriculograms for the detection of coronary artery disease. AM HEART J 1982;104:1. 6. Pave1 DG, Byrom E, Lam W, Meyer-Pave1 C, Swiryn S, Pietras R. Detection and quantification of regional wall motion abnormalities using phase analysis of equilibrium sated cardiac studies. Clin Nucl Med 1983:8:315. 7. Alcan KE, Robeson W, Graham MC, Pa&r0 C, Oliver FH, Benua RS. Fourier amplitude and phase analysis in the clinical evaluation of patients with cardiomyopathy. Clin Nucl Med 19&1;9:314. 8. Widmann TF, Favrot L, Smith SC, Ehmann RW, Peterson KL. Temporal Fourier analysis in quantitative assessment of cardiac digital images. Computers in Cardiology, Seattle: IEEE publication, 1984. 9. Ratib 0, Rutishauser W. Recent developments of cardiac digit& radiography. Int J Cardiac Imaging 1985;1:29. 10. Bacharach SL, Green MV, Vitale D, White G, Douglas MA, Bonow RO, Larson SM. Optimum Fourier filtering of cardiac data: A minimum error method: Concise communication. J Nucl Med 1983;24:1176. 11. Rankin JS, McHale PA, Arentzen CE, Ling D, Greenfield JC, Anderson RW. The three-dimensional dynamic geometry of the left ventricle in the conscious dog. Circ Res 1976; 39304.
12. Leighton RF, Nelson AD, Brewster P. Subtle left ventricular asyxiergy with completely obstructed coronary arteries. Am J Cardiol 1983;52:693. 13. Kumada T, Karlmer JS, Pouleur H, Gallagher KP, Shirato K, Ross J Jr. Effecta of coronary occlusion on early ventricular diastolic events in conscious dogs. Am J Physiol 1979; 237:H542. 14. Bonow RO, Vitale DF, Bacharach SL, Frederick TM, Rosing DR, Kent KM, Green MV. Heterogeneous left ventricular regional function and impaired global diastolic filling in coronary artery disease: Reversal after coronary angioplasty. -_ Circulation 1983;68:III-167. 15. Sasavama S. Franklin D. Ross J Jr. Kemner WS. McKnown D. &mu& changes in’left ventricular wall thickness and their use in analyzing cardiac function in the conscious dog. Am J Cardiol 1976;38:870. 16. Gallagher KP, Matauzaki M, Osakada G, Kemper WS, Ross J
Heusch et al.
American
Jr. Effect of exercise on the relationship between myocardial blood flow and systolic wall thickening in dogs with acute coronary stenosis. Circ Res 1983;52:716. 17. Theroux P, Ross J Jr, Franklin D, Kemper WS, Sasayama S. Regional myocardial function in the conscious dog during acute coronary occlusion and the response to morphine, propranolol, nitroglycerin and lidocaine. Circulation 1976; 53~302.
18. Guth BD, Tajimi
T, Seitelberger R, Lee JD, Matsuxaki
M,
January 1987 Heart Journal
Ross J Jr. Experimental exercise-induced ischemia: Drug therapy can eliminate regional dysfunction and oxygen supply-demand imbalance. J Am Coll Cardiol 1986;7:1686. 19. Winer BJ. Statistical .nrincinles . in exnerimental desian. New York: McGraw-Hit1 Book Co, Inc, 1971. 20. Theroux P, Ross J Jr, Franklin D, Cove11 JW, Bloor CM, Sasayama S. Regional myocardial function and dimensions early and late after myocardial infarction in the unanesthetized dog. Circ Res 1977;40:158.
The effects of cytochrome C, an electron carrier in the process of oxidatlve phoephoryiation, on infarct size and regional ieft ventrkuiar function after a coronary ertery occiuaion were investigated. Thus, in 30 dogs, 1 mtnute’after feft anterior dercending coronary artery occlusion, oo”Tc-i8beied albumin mkrosgheres (8 mCi) were injected into the left atrium for subsequent assessment of the hygop8rfused zone, that Is, the 8rea at risk of infarction. Fifteen mfnutes after coronary artery o&us&n, doga were randomtzed Into 8 control group (n = 15) and a cytochrome C-treated group (n = la). The I8tter immediately received cytochrome C, 2.5 mg/kg intravenousiy. Six hours after coron8ry artery OCCiuSiOn the dogs were sacrificed and their left ventrlcies were cut into 3 mm thfck s&es. tnf8rct Sire ~88 determined by trjphenyitetrazolium chkwide statning end me8sured by plantmetry. The 88me slices were then submitted to autoradiogr8phy 8nd the hypogerfused zone was then measured by pfanimetry. The hypoperfused zone was 22 + 2% and 23 + 2% of the left ventrkie in the control and treated groups, respecttveiy (MS), indtcattng that the extent of myocardtum at risk before treatment was simller. The extent of the hypeperfuaed zone which evolved to necrosis was 90 + 3% in the control group but only 50 * 7% in the treeted group @ < 0.001). MyoC8rdi8i s8ivage In the treated group was paraiieied by improvement in syatoiic W8ii thickness of the fschemic segment as measured by two-dimenstonai echocardiography. Thus, cytochrome C reduced the extent of myocardlai necrosis by 44% and tmproved systoiic function of the ischemic myocardtum. (AM HEM J lg87;113:124.)
Andrew Zalewski, M.D., Sheldon Goldberg, M.D., Richard Krol, M.D., and Peter R. Maroko, M.D. Browns Mills, N. J., and Philadelphia, Pa.
Several therapeutic interventions aimed at decreasing myocardial oxygen demand or increasing oxygen supply have been reported to limit ischemic injuW- lp2 The latter can be achieved by means of coronary reperfusion? by retroperfusion via cardiac From the Deborah Cardiovascular Research Institute, and Thomas Jefferson University Hospital. Supported by grants from: W.W. Smith Char&able Trust and Johnson & Johnson Family of Companies. Received for publication May 8, 1986; accepted June 3, 1986. Reprint requests: Andrew Zalewski, M.D., Cardiac Catheterization Laboratory, Thomas Jefferson University Hospital, 111 South 11th St., Suite 5611D, Philadelphia, PA 19107. 124
veinq3 or by increasing the utilization of oxygen, for example, changing the oxygen-hemoglobin dissociation curve.” It remains to be estatished, however, if improved utilixation of oxygen within the ischemic myocardium, caused by exogenous electron carriers, could result in myocardial salvage. Cytochrome C is an important electron carrier which is located in the inner mitochondrial membranes of aerobic cells with especially high concentrations in the heart5 This substance and others which transfer electrons to molecular oxygen in the mitochonclria allow for the generation of adenosine triphosphate (ATP) in
Cogswell, Sagar, and Wann
23.
24.
25. 26. 27.
28.
Amatlcan
diographic features in 30 patienta. Am J Cardiol1979;44:401412. Keren G, Belhassen B, Sherez J, Miller HI, Megidish R, Berenfeld D, Laniado S: Apical-hypertrophic cardiomyopathy: Ev&ation by noninvasive and invasive techniques in 23 patienti. Circulation 1985;71:45-56. Harrison DC, Braunwald E, Glick G, Mason DR, Chidsey CA, Ross J Jr: Effects of beta adrenergic blockade on the circulation, with particular reference to observations in patients with hypertrophic subaortic stenosis. Circulation 1964;29:8498. Flamm MD, Harrison DC, Hancock EW: Muscular subaortic stenosis: Prevention of outflow obstruction with propranolol. Circulation 1968;38:846-858. Adelman AG, Shah PM, Graniak R, Wigle ED: Long-term propranolol therapy in muscular subaortic stenosis. Br Heart J 1970;32:804-811. Landmark K, Sire S, Thaulow E, Amlie JP, Nitter-Hauge S: Hemodynamic effects of nifedipine and propranolol in patients with hypertrophic obstructive cardiomyopathy. Br Heart J 1982;48:19-26. Kaltenbach M, Hopf R, Kober G, Bussmann WD, Keller M, Petersen Y: Treatment of hypertrophic obstructive cardiomyopathy with verapamil. I+ Heart J 1979;42:35-42.
Heart
1967 Journal
29. Rosing DR, Kent KM, Borer JS, Seides SF, Maron FJ, Epstein SE: Verapamil therapy: A new approach to the pharmacologic treatment of hypertrophic cardiomyopathy. I. Hemodvnamic effects. Circulation 1979:60:1201-1207. 30. Bonow RO, Roaing DR, Bacharach SL, direen MV, Kent KM, Lipson LC, Maron BJ, Leon MB, Epstein SE: Effect of verapamil on left ventricular systolic function and diastolic filling in patients with hypertrophic cardiomyopathy. Circulation 1981;64:787-795. 31. Levine RA, Weyman AE: Dynamic subaortie&&ruction in hypertrophic cardiomyopathy: Criteria and controversy. J Am Coll Cardiol 1985;6:16-18. 32. Kinoshita N, Nimura Y, Okmoto M, Miyatake K, Nagata S, Sakakibara H: Mitral regurgitation in hypertrophic cardiomyopathy. Non-invasive study by two-dimensional Doppler echocardiography. Br Heart J 1983;49:574-583. 33. Wigle ED, Adehnan AG, Auger P, Marquis Y: Mitral regurgitation in muscular subaortic stenosis. Am J Cardiol 1969; 24698-706. 34. Criley JM, Siegel IU Has “obstruction” hindered our understanding of hypertrophic cardiomyopathy? Circulation 1985; 721148-1154.
A Fourier analysis including the first 20 harmonics was performed on sonomkrometric measurements of regIonal myocardid wall thkkneas in eight consciour dogs under control conditions and at four levels of ischemia produced by a hydraulic occluder on the iefl clrcumfiex coronary artery. Systolc watt thickening was reduced from 28.47 + 6.20% (SD.) (control) to 28.05 rt 5.73% (mild atenoris), 17.W + 5.86% (moderate stenosis), 11.48 + 3.56% (severe aten+s), and 3.59 f 2.57% @O-second occlusion), values SlgMcantly dtfferent from each other @ < 0.01). The amplitude of the first harmonic decreased stepwke from 1.35 + 0.31 to 1.08 + 0.29 mm, 0.90 + 0.27 mm, 0.69 5 0.24 mm, and 0.43 + 0.12 mm, atl signifkantly dihereqt from each other (p < 0.05). These amplitude values correlated to percent systolic wall thickening (r= 0;8W, p = 0.001). A phase shift of the first harmonk from 13t + 11 to 139 + 14 egreer, 13 +‘I& degrees @ < 0.05 vs control), 161 5 21 degrees (p < 0.01 vs control), and 191 + 21 degrees (p < 0.01 vs control and severe stenosis) correlated with the increaes in time from end dlqtoh tb the point of maxhnum wall excursion (r = 0.662, p < 0.001). These data Indicate t&t the extent of ischem~c regional myocardial hypoktnesis can be adequately described by the am@tude of the first hermonk, and that the asynchrony of ventrkular contraction end relexstion can be detected from the phese of the first harmonic. (AM HEART J 198y; 113:lW.)
Gerd Hewh, M.D.,* Brian D. Guth, Ph.D., Thomas Widmann, M.D., Kirk L. Peterson, M.D., and John &as Jr., M.D. Sun Diego and La Jolla,
From the Division of Cardiology, Department of Medicine, University California, San Diego, School of Medicine. Supported hy a National Institute of Health (Specialized Center Research on Iachemic Heart Disease) Grant No. HL-17682. Received for publication Feb. 28, 1986; accepted May 26, 1986.
116
of of
C&f.
Reprint requests: John Ross, Jr., M.D., Division of Cardiology, Department of Medicine, M-013, University of Caliiornia, San Diego, La Jolla, CA 92093. *Dr. Gerd Heusch is the recipient of a research fellowship granted by the German Research Foundation.
Volume
113
Number
1
With the use of Fourier analysis, any periodic function can be represented as the sum of multiple sine and cosine waves of different frequencies. Each of these harmonics is characterized by a certain phase and amplitude. lr2 Fourier analysis can be applied to the cardiac cycle, which constitutes a periodic sequence of contraction and relaxation at the fundamental frequency of the heart rate. Application of Fourier analysis to the cardiac cycle permits a functional analysis of images by the use of gated blood pool radionuclide ventriculograms’s 3-7 and contrast angiographic ventriculograms.2T 8.g Contraction abnormalities can thereby be characterized by a phase delay,1-Q and a reduction in amplitude7q8 of the first harmonic of a temporal Fourier transform. A Fourier analysis employing only the first, or the first few harmonics, is an effective and objective high-frequency cut-off filter of noisy original timeactivity or time-intensity curves. However, it is unclear how many harmonics are essential for a precise functional analysis. A study by Bacharach et aI.12 suggests that two harmonics are sufficient for quantitative characterization of global systolic left ventricular function, but up to six harmonics may be necessary to describe global diastolic function adequately in radionuclide studies. Rankin et al.” found that sonomicrometric ventricular dimension data, including wall thickness; are of relatively low frequency and can be approximated by the first five harmonics. However, this problem has not up to now been addressed with respect to changes in regional myocardial function occuring during ischemia. Furthermore, there is some controversy as to whether phase shifts or amplitude changes of a Fourier transform better reflect changes in regional myocardial walI function occurring during regional ischemia. Most studies have examined only the phase distribution of the first harmonic for the assessment of regional myocardial dysfunction, Pavel et aL6 found good agreement between the contrast angiographic classification of regional dysfunction and the phase delay in a gated blood pool study. In a contrast angiographic study of patients with subtle regional myocardial dysfunction distal to completely occluded coronary arteries,12 phasic geometric analysis was even more sensitive than quantitative assessment of walI motion in identifying the coronary hypoperfusion. In contrast, a previous study from this institution8 found contrast angiographic amplitude images more sensitive to left ventricular dyssynergy than phase images. Therefore, the purpose of the present study was to determine the value of a temporal Fourier transform
Fourier analysis of regional LV function
117
for the characterization of ischemic regional myocardial dysfunction by means of a highly precise measurement of regional wall function. In conscious dogs, regional myocardial wall thickening was measured by sonomicrometry under control conditions and at four levels of ischemic dysfunction. The results of a temporal Fourier transform, including the first 20 harmonics, were then compared to conventional parameters of regional wall function. To test the sensitivity of phase shifts of the first harmonic for detecting asynchrony of ventricular contraction and relaxation, which is associated with regional myocardial ischemia in experimentaP3 and clinical studies,12v14 we also quantified the degree of synchrony between wall thickening in nonischemic and ischemic regions. METHODS
Eight mongrel dogs, weighing between 21 and 35 kg, were instrumented during sterile surgery. They were premeditated with acepromazine maleate (0.5 mg/kg intramuscularly) 30 minutes before induction of anesthesia with sodium pentobarbital (25 mg/kg intravenously); further analgesiawas provided by morphine (0.5 mg/kg intravenously). During artificial ventilation with a Harvard respirator, the heart was exposed through a left thoracotomy in the fifth intercostal space and was suspended in a pericardial cradle. For measurementof left ventricular pressure,a high-fidelity transducer (Konigsberg Instruments Inc., Pasadena,Calif., model P7) and a fluid-filled tube (Tygon, 1.27 mm inner diameter) were inserted through a stab wound in the apex. The proximal left circumflex coronary artery wasdissectedfor 5 mm and a hydraulic occluder wasplaced around the vessel.For the measurementof regional contractile function, two pairs of miniature ultrasonic crystals (5 MHz) were implanted in the left ventricular free wall to measurewall thickness.15 One crystal (2 mm diameter) of each pair was inserted obliquely through the ventricular wall to the subendocardium. The secondcrystal (6 mm diameter) was attached to a Dacron patch and wassewnon the epicardial surface opposite the subendocardial crystal where the ultrasonic transit time between the two crystals was shortest, as observed on an oscilloscope(Tektronix, Inc., Beaverton, Oreg., model 453). The distance between the two crystals wascontinuously recorded with the transit time technique (Sonomicrometer, Triton Technology, San Diego, Calif.). To verify appropriate crystal orientation, the received signalswere alsomonitored on an oscilloscope.One pair of crystals waspositioned in the left ventricular anterior wall perfused by the left anterior descendingcoronary artery and servedascontrol; the other pair of crystals wasplaced in the zone of the left ventricular posterior wall perfused by the left circumflex coronary artery to be rendered ischemic. The appropriate alignment of each pair of crystals was confirmed at autopsy in all animals reported.15Upon histologic examination, the crystals were surrounded by a fibrous rim (approximately 1 mm) of
Heusch et al.
118
fable I. The effects of three degreesof coronary artery stenosisand a 30-secondcomplete coronary occlusion on the phase(degrees)of a temporal Fourier transform of regional myocardial wall thickening; zero degreeswas defined at end diastole n=8
Mild stenosis
Control
Ischemic
Moderate stenosis
Severe stenosis
JO-second occlusion
wall
1. Harmonic
2. Harmonic 3. Harmonic 4. Harmonic 5. Harmonic Nonischemic
137 221 215 126 206
k f + f iz
11 21 60 76 63
139 216 213 172 162
+ 14 + 24 + 80 k 114 zk 138
150 221 211 208 139
+ + k + +
15* 31 91 137 131
161 247 231 214 213
f -t f f ?
21t 48 41 132 98
191 + 21tg 200264 209+60 160 f 130 227 + 120
137 216 180 200 162
f f + f f
12 24 68 89 74
134 212 211 142 139
* k + k f
137 214 166 199 201
+ 14 t- 28 rf: 112 + 124 f 115
139 220 147 194 160
* + + + k
18 42 85 137 150
147 231 117 227 121
wall
1. Harmonic
2. Harmonic 3. Harmonic 4. Harmonic 5. Harmonic
13 27 89 145 117
f + + f +
14*j 33 96 85 124
Dataare means f S.D. *p < 0.05,
tp < 0.05,
tp < 0.01 “S Control. §p < 0.01 ve next lower
level
of iechemia.
II. The effects of three degreesof coronary artery stenosisand a 30-secondcomplete coronary artery occlusion on the amplitude (mm) of a temporal Fourier transform of regional myocardial wall thickening
Table
n=8 Ischemic
2. Harmonic 3. Harmonic 4. Harmonic 5. Harmonic 2. Harmonic 3. Harmonic 4. Harmonic 5. Harmonic lp
Severe stenosis
30-second occlusion
1.35 0.66 0.13 0.09 0.06
f 0.31 + 0.16 * 0.10 + 0.05 I!z 0.03
1.08 0.59 0.10 0.10 0.09
zk 0.29t + 0.13* + 0.08 + 0.06 + 0.03
0.90 0.43 0.12 0.14 0.10
k 2 f f f
0.27tz 0.1378 0.06 0.06*$ 0.04
1.04 0.50 0.14 0.08 0.05
+ f + r +
1.01 0.52 0.14 0.07 0.05
f k f f Ik
1.08 0.52 0.12 0.07 0.05
k 0.26 -t 0.14 f 0.07 * 0.03 z!z0.02
0.69 0.30 0.12 0.14 0.08
f 0.24t$ + 0.12tj + 0.06 k 0.06* AZ0.03
0.43 0.21 0.17 0.21 0.11
+ + + f k
0.12ts O.llfq 0.11 o.lot$ 0.04
1.10 0.53 0.12 0.05 0.03
k f + * 2
1.15 0.50 0.11 0.07 0.06
ii k & f k
0.30 0.16 0.09 0.03 0.03
wall
1. Harmonic
Data
Moderate stenosis
wall
1. Harmonic
Nonischemic
Mild stenosis
Control
+ S.D. tp < 0.01 “8 Control. §p < 0.01 ve next lower
0.18 0.11 0.07 0.03 0.02
0.16 0.14 0.09 0.02 0.02
0.30 0.19 0.12 0.03 0.01
are means
< 0.05, $p < 0.05,
level
of Lchemia.
connective tissue, as previously described.16After instrumentation, the chest was closed, the pneumothorax was evacuated, and the wires and tubes were passedsubcutaneously to the back of the animal and were exteriorized betweenthe scapulae.Ampicillin (6.6 gm/day) wasadministered for 3 days after surgery. Dogswere studied no earlier than 7 days after surgery, when they had recovered and were afebrile. Each dog was studied when it was resting quietly on the table, as it had been previously trained to do. The Konigsberggaugewas calibrated with a Statham (P23 Db) transducer connected to the fluid-filled tube, with zero reference taken at the estimated level of the right atrium, asthe dog waslying on its right side. After a control recording, a level of ischemic myocardial dysfunction was produced by adjustment of
the hydraulic occluder. After steady-state dysfunction was recorded, the occluder was completely releasedand complete recovery of regional myocardial function wasallowed to occur. The occluder was then reinflated to produce a different degree of ischemic wall dysfunction, and three different degrees of a steady-state dysfunction due to partial coronary stenosiswere so obtained. In addition, a 30-second complete coronary occlusion was performed. These four levels of regional myocardial dysfunction (subsequentlytermed mild, moderate, and severestenosis, and coronary occlusion) were produced in random order. Recordings during each experiment were made on a forced-ink recorder (Brush, Cleveland, Ohio) and on magnetic tape for subsequentplayback and computation. The systemic hemodynamic parameters analyzed were
volume Number
113 1
Fourier analysis of regional LV function
119
III. The effects of three degreesof coronary artery stenosisand a 30-secondcomplete coronary artery occlusion on systemic hemodynamics
Table
Control
Mild stenosis
Moderate stenosis
Severe stenosis
HR @pm) PLVSP
98 + 15 136 f 7
93 rf: 15 132 + 8
99 f 12 134 f 14
102 * 17 132 + 9
103 + 17 140 + 7
(mmHg) LVEDP
13.6 k 4.7
13.7 f 4.9
15.6 k 7.9
14.8 I!Z3.7
20.8 + 7.2?$
3146 + 396
2917 f 451*
2875 f 407*
2844 f 351*
2732 + 319t
1718 rt 161
1561 + 215*
1598 + 250*
1423 + 184fJ
1359 * 201t
n=8
30 second occlusion
(mm Hg)
(+) dpldt (mmHg/sec) (-) dp/dt (mmHg/sec)
Data are means T S.D. HR = heart rate: PLVSP - Dealt left ventricular svetolic meeaure: l p < 0.05, tp < 0.01 VB corlt-rol. tp < 0.05, $p< 0.01 vs next lower level of iachemia.
LVRDP
heart rate, left ventricular end-diastolic and peak systolic pressures,and peak positive and negative left ventricular dp/dt. Regional myocardial wall thicknesswasdetermined at end diastole (defined asthe time dpldt crossedthe zero line) and at end systole (defined as the time of maximum excursion at end ejection within 20 msec before peak negative dp/dt, according to Theroux et al.‘? The maximum wall thickening excursion without reference to the time of its occurrence within the cardiac cycle was determined, since the amplitude values of the Fourier analysis are also without reference to the time within the cardiac cycle. The systolic wall thickening excursion (up to end ejection) and the extent of wall thickening occuring after end systole (post ejection thickening) were also determined. Systolic wall thickening wasalso calculated as the percent of end-diastolic wall thickness. In addition to these measures of the extent of thickening, the time between end diastole and the point of maximum wall thickening and the mean systolic wall thickening velocity were determined. Finally, the loop area of a plot of wall thickness vs left ventricular pressurewas calculated.‘* Hemodynamic and dimensiondata were digitized from magnetic tape with a computer system (PDP 11/03, Maynard, Masss.), and data from 20 consecutive cardiac cycles were collected at 5 msec intervals. Card& cycles, with an interval between sequential peak positive dp/dt values differing by lessthan 40 msec,were selected and averaged. Eight to 20 beats were averagedin this way for each lnterven$ion. In addition to standard measures of regional wall function describedabove, a temporal Fourier transform of regional wall thickening including the first 20 harmonics was performed. Zero degreeswas defined at end diastole. Data are presented as mean values with their standard deviations. Statistical analysiswasperformed by meansof an analysis of variance for repeated measurementsand, when a significant interaction was detected, single mean values were compared with Tukey’s range test.lg In addition to the analysis of differences between the arbitrarily defined degreesof myocardial ischemia,linear regression
= left ventricular
end-diastolic
preeeure.
analyseswere performed taking the data points of all dogs together as well asthose for eachindividual dog separately. A level of p < 0.05 was accepted as signillcant. RESULTS
The control conditions for hemodynamics, regional dimensions, and Fourier analysis values before each level of ischemic myocardial dysfunction were not different, and therefore they were combined. The sixth to twentieth harmonics of the temporal Fourier transform of regional wall thickening represented only “noise”; these harmonics were characterized by phase values with a high standard deviation (SD. > 50% of the mean value) and by low amplitude values (amplitudes < 5% of the first harmonic under nonischemic conditions). Therefore only phase and amplitude values of the first five harmonics are presented in Tables I and II. Hemodynamlcs. Regional myocardial ischemia induced only modest changes in systemic hemodynamics. Whereas heart rate and peak left ventricular systolic pressure were not significantly changed, left ventricular end-diastolic pressure was siguificantly increased during complete coronary occlusion, and peak positive and negative dp/dt were reduced with increasing severity of ischemia (Table III). Regional tunctlon: Conventional measures. A representative original recording of regional myocardial wall thickening in the ischemic and the nonischemic regions is shown in Fig. 1. Increases in the severity of regional myocardial ischemia were evidenced by progressive reductions in maximum wall thickening and systolic wall thickening, percent systolic wall thickening,
mean systolic wall thickening
velocity,
and the regional wall thickness-left ventricular pressure loop area (Table IV). End-diastolic wall thickness showed no significant changes and end-systolic
120
Heusch et al.
Am6rlcan
January 1887 ll66rt Journal
Anwi wall ThiaRerr. Cmm)
Fig. 1. Representative original recording showing changes in regional myocardial wall thickening occurring with increasing severity of myocardial ischemia in the ischemic and nonischemic region. Systolic wall thickening of the ischemic region progressively decreases and a significant post ejection thickening develops. Systolic thickening of the nonischemic region increases during complete coronary occlusion.
Table IV. The effects of three degrees of coronary on regional myocardial dimensions n-8
Control
Ischemic wall EDT (mm) EST (mm) MT (mm) ST (mm) PET (mm) ST (%I TMT (msec) MSTV bmhec) TLVPLA (mm x mm Hg) Nonischemic wall EDT (mm) EST (mm) MT (mm) ST (mm) PET (mm) ST (%) TMT (msec) MSTV (mm/set) TLVPLA (mm x mm Hg)
12.31 15.46 3.37 3.17 0.02 26.47 220 15.81
+ + + + + -t + k
1.58 1.72 0.65 0.66 0.03 6.20 23 3.39
340 + 65
11.24 13.59 2.79 2.51 0.01 22.08 219 12.02
+ f 2 + * + f +
2.03 2.37 0.58 0.65 0.03 3.88 25 2.23
259 * 31
artery
stenosis and a 30-second
Mild stenosis 12.25 14.90 2.91 2.64 0.05 22.05 232 12.56
+ + + + + i k IT
1.78 1.76 0.527 0.57t 0.08 5.73t 29 3.53t
289 + 56t
10.66 13.13 2.78 2.55 0.01 23.01 220 12.09
+ f k f + + f *
1.72 2.07 0.63 0.56 0.03 3.67 17 2.72
262 + 44
Data ara means 2 S.D. thickness; EST = end-systolic thickness; MT = maximum (% ) = systolic waU thickening expressed as percent of end-diastolic thickness; velocity; TLVPLA = thickness-left ventricular pressure loop area.
EDT = end-diastolic
*p< 0.05,tp < 0.01 $p< 0.05, §p< 0.01
“8 Control.
vs next lower level of ischemia.
Moderate stenosis 12.45 14.51 2.37 2.06 0.07 17.00 250 10.06
? * f * k r * *
1.09 1.99* 0.56t# 0.6ot~ 0.06* 5.8q§ 35t 3.27ts
220 k 66t§
11.58 14.09 2.82 2.51 0.01 21.95 218 12.07
f + + f * k r?r +
2.73 3.23 0.69 0.56 0.03 2.54 21 2.84
255 f 38
complete
coronary
Severe stenosis 11.96 13.32 1.91 1.39 0.20 11.46 278 6.34
+ 1.58 f 1.4879 + 0.56t9 AZ 0.49?$ f 0.15t & 3.56?§ + 29f$ + 2.4575
160 iz 48tfj
11.71 14.26 2.74 2.82 0.01 24.49 216 12.18
A 2.66 f 3.12 ZIZ 0.68 f 0.73 + 0.03 f 4.926 A 15 f 3.89
255 k 40
artery
occlusion
30-second occlusion
11.86 12.29 1.48 0.44 0.81 3.69 305 2.34
2 * f * f + + +
1.63 lAq§ 0.49?$ 0.32ts 0.28?§ 2.57t§ 16tj 1.7175
89 + 15t§
10.61 13.60 2.83 2.51 0.01 24.25 224 12.62
+ + + + + f + k
1.83 2.32 0.64 0.46 0.03 3.55* 17 3.15
264+24
thickening; ST = systolic thickening; PET = post ejection thickening; ST TMT = time to maximum thickening; MSTV = mean systolic thickening
Volume4 113 Numbr 1
Wall1
Fourier analysis of regional LV function
121
rhickner
Fig. 2. Representative reconstruction of the digitized wall thickening by the first five Fourier harmonics. The reduction in the extent of severity of ischemia is well reflected by the amplitude of only the first however, is only reflected after including the information of the third, With increasing severity of ischemia there is a delay in the phase of
wall thickness diminished (not significant for mild stenosis). The time from end-diastole to the point of maximum wall excursion increased progressively and a significant post ejection thickening developed (both not significant for mild stenosis). Only with the severe coronary stenosis and complete coronary occlusion was there a significant increase in percent systolic wall thickening of the nonischemic region, which was not reflected in other parameters of nonischemic wall function (Table IV). Regional function: Fourier analysis. A typical example of the reconstruction of ischemic wall thickening by the flrat few harmonics of a Fourier transform is shown in Fig. 2. Phase analysis. The phase values of only the first harmonic exhibited a phase delay with increasing severity of myocardial ischemia. This phase delay only became significant with moderate and severe stenosis and complete coronary occlusion. Furthermore, the phase delay did not distinguish between the different levels of ischemic dysfunction, except for complete occlusion (Table I). The phase of the first harmonic correlated to the time from end diastole to the point of maximum wall thickening: y = 0.301 X + 65.267, r = 0.662, p < 0.001 (by use of all data points), and r = 0.815 & 0.080 (by use of the average of the correlation coefficients from all eight dogs). The difference between the phases of the first harmonic of the posterior (ischemic) and the anterior (nonischemic) myocardium correlated closely to the difference in time from end diastole to the point of maximum wall thickening between both regions: y = 0.384 x + 1.431, r = 0.748, p < 0.001 (all data points), and r = 0.869 f 0.109 (average of eight dogs).
in the ischemic myocardial region systolic thickening with increasing harmonic. Post ejection thickening, fourth, and fifth harmonic as well. the first harmonic.
There also was a significant correlation between the phase of the first harmonic and percent systolic wall thickening: y = - 2.274 X + 173.2, r = -0.747, p < 0.001 (all data points), and r = -0.805 t 0.312 (average of eight dogs). The correlation between the phase of the first harmonic and the regional wall thickness-left ventricular pressure loop area, which contains the information not only about systolic function but the entire cardiac cycle, was not closer: Y = -0.145 x +171.6, r = -0.580, p < 0.005 (all data points), and r = -0.671 + 0.139 (average of eight dogs). Amplitude analysis. The amplitudes of both the first and the second harmonics were significantly reduced with increasing severity of myocardial ischemia, and there was a significant difference between each level of ischemia in the amplitude of the first two harmonics (Table II). Conversely, the reduction in the amplitude of the first two harmonics with increasing severity of myocardial ischemia was paralleled by an increase in the amplitude of the subsequent harmonics, which became sign&ant for the fourth harmonic except with mild stenosis (Table II). There was a good correlation between the amplitude of the first harmonic and maximum (r = 0.893), systolic (r = 0.939), and percent systolic wall thickening: y = 0.038 X + 0.274, r = 0.894, p < 0.001 (all data points) (Fig. 3), which was even closer for the individual animals: r = 0.963 f 0.048 (average of eight dogs). The correlation between the amplitude of the first harmonic and the mean systolic. wall thickening velocity was: y = 0.068 x +0.248, r = 0.959, p < 0.001 (all data points), and r = 0.988 f 0.041 (average of eight dogs). Again, the correlation between the amplitude of the first har-
January
122
Heusch et al.
Amarlcan
0
0
0
0 A 0 0
Q: F o1
10
20
A
o
Control Mild Stenosis Moderate Stenosis Severe Stenosis Occlusion
30
40
Systolic Wall Thjckening (%) Fig. 3. Graphic presentation of the correlation between
the amplitude of the first Fourier harmonic and percent systolic wall thickening with increasingseverity of regional myocardial ischemia.
manic and the regional wall thickness-left ventricular pressure loop area was not closer: y = 0.003 X + 0.182, r = 0.834, p < 0.001 (all data points), and r = 0.927 + 0.097 (average of eight dogs). The increase in the amplitude of the fourth harmonic correlated to the extent of post ejection thickening: y = 0.118 X + 0.106, r = 0.535, p < 0.01 (all data points), and r = 0.728 + 0.257 (average of eight dogs). There was no significant correlation between the extent of post ejection thickening and the amplitude of any other harmonic. DlSCUSSlON
This study was designed to compare Fourier analysis to conventional measures based on a highly accurate measurement of abnormal regional myocardial wall motion, and to determine how many harmonics are necessary for the quantitative assessment of regional myocardial wall function during ischemia. We also undertook to determine whether the phase or the amplitude of the Fourier analysis is more sensitive to regional ischemic dysfunction. Our data demonstrate that ischemic regional hypokinesis can be adequately quantified by only the first harmonic of a temporal Fourier transform, and that the amplitude of the tit harmonic correlates particularly closely to absolute and percent systolic wall thickening, which can be regarded as the reference standard for the assessment of ischemic regional dysfunction.
Heart
1987 Journal
Our experimental study is based on a different measurement of regional myocardial wall function (sonomicrometry) from that used in clinical studies employing radionuclide or contrast venticulography. This approach nevertheless confirms the value of Fourier analysis for the characterization of ischemic regional myocardial dysfunction.5~gIn accordance with the previous clinical study by Widmann et al.,8 our data suggest that the amplitude image should be more sensitive for the detection of regional myocardial ischemic dysfunction than the phase image. Moreover, a very close correlation between the amplitude of the first harmonic and percent systolic wall thickening and wall thickening velocity was evident within any individual dog. Thus the regional amplitude of the first harmonic may be particularly well suited for the evaluation of therapeutic procedures such as angioplasty. In addition, the progressive phase delay with increasing severity of myocardial &hernia described quantitatively the asynchrony of left ventricular contraction and relaxation which, in addition to the reduction in regional wall motion, contributes to the reduction in left ventricular pump function. Under control conditions at a heart rate of approximately 100 bpm, a phase value of the f&t harmonic of 137 degrees corresponds closely to the end of ejection into the aorta, which occurs about 220 msec after end diastole. The progressive phase delay with increasing severity of &hernia reflects regional post ejection wall thickening, which does not contribute to the left ventricular external work. In clinical studies with the use of phase analysis of the first harmonic of radionuclide ventriculograms, such loss of synchrony appeared to be relatively specific for ischemic heart disease as compared to wall motion abnormalities associated with valvular disease,5 although similar phase delays may be observed during abnormal electrical activation.2 Our experimental data confirm the value of the phase of the first harmonic for reflecting the degree of asynchrony of left ventricular contraction and relaxation, as evidenced by the close correlation to the difference in time from end diastole to the point of maximum wall excursion between the ischemic and the nonischemic myocardial regions. However, in our study this ischemia-related asynchrony was not quite as sensitive as the extent of hypokinesis in reflecting the degree of myocardial ischemia. This finding might be explained by the more sensitive analysis of mild regional hypokinesis by sonomicrometry compared to contrast ventriculography,12 rather than lack of value of phase analysis for the assessmentof regional myocardial ischemia. Thus, most ventricu-
Vohtmb Number
113 1
lographic analyses use minimum volume or maximum inward wall motion for defining end systole, whereas end ejection is identified in the present study and post ejection wall thickening is not included. The increasing extent of post ejection thickening with increasing severity of ischemia was also reflected by an increasing amplitude of the fourth harmonic. However, this correlation was not very close (r = 0.535), and a 40.5:fold ‘increase in post ejection thickening from control to complete coronary occlusion was only reflected by a 2.3-fold increase in the amplitude of the fourth harmonic. Presumably, major parts of the post ejection thickening information are included in other harmonics as well, although the amplitude of no single other harmonic correlated significantly to the extent of post ejection thickening. Amplitudes of the first harmonic in radionuclide or contrast ventriculograms are without reference to the time within the cardiac cycle and therefore may overestimate the extent of systolic excursion. However, since post ejection thickening was significantly correlated only to the amplitude of the fourth harmonic and, on the other hand, the amplitude of the first harmonic was closely correlated to systolic wall thickening, this concern appears to be negligible. Neither the amplitude nor the phase of the first harmonic correlated to the wall thickness-left ventricular pressure loop area better than to systolic wall thickening, although they similarly contain information from the entire cardiac cycle, whereas the calculation of systolic wall thickening is based on only two points in time. This lack of a closer correlation between the loop area and the Fourier analysis is most likely explained by the fact that the calculation of loop area is based on a regional wall thickness and a global left ventricular pressure, and the latter does not reflect the active regional pressure development or stress during ischemia, as pointed out previously.1s In conclusion, our experimental data emphasize the value of a Fourier analysis for the assessmentof acute regional myocardial ischemic dysfunction. The phase of the first harmonic reflects the degree of synchrony of left ventricular contraction and relaxation, but the amplitude of the first harmonic appears to be sufficient to quantify the degree of regional dysfunction adequately. Chronic wall motion abnormalities often resemble those during acute ischemia,20and it might be expected that this type of analysis would be applicable in a chronic setting as well. These data may be relevant to
Fourier analysis of regional LV function
123
clinical studies employing functional image analysis of radionuclide or contrast ventriculograms, since they justify the use of only the first Fourier harmonic and thereby allow a substantial reduction in computation time. Our data may futhermore be relevant to all other methods employing a dynamic analysis of myocardial dimensions, such as M-mode echocardiography and fast computer tomography. REFERENCES
1. Links JM, Douglass KH, Wagner HN. Patterns of ventricular emptying by Fourier analysis of gated blood-pool studies. J Nucl Med 1980,21:978. 2. Widmann TF, T&au JF, Ashburn WL, Bhargava V, Higgins CB, Peterson KL. Evaluation of regional wall motion by phase and amplitude analysis of intravenous contrast ventricular fluorangiography: Technical aspects and computation. In: Sigwart U, Heintsen PH, editors: Ventricular wall motion. Stuttgart-New York Georg Thieme Verlag, 19&1:24. 3. Adam WE, Tarkowska A, Bitter F, Stauch M, Geffers H. Equilibrium (gated) radionuclide ventriculography. Cardiovast Radio1 1979;2:161. 4. Botvinick E, Dunn R, Frais M, 0 Connell W, Shosa D, Her&ens R, Scheinmann M. The phase image: Its relationship to patterns of contraction and conduction. Circulation 1982;65:551. 5. Ratib 0, Henze E, Schoen H, Schelbert HR. Phase analysis of radionuclide ventriculograms for the detection of coronary artery disease. AM HEART J 1982;104:1. 6. Pave1 DG, Byrom E, Lam W, Meyer-Pave1 C, Swiryn S, Pietras R. Detection and quantification of regional wall motion abnormalities using phase analysis of equilibrium sated cardiac studies. Clin Nucl Med 1983:8:315. 7. Alcan KE, Robeson W, Graham MC, Pa&r0 C, Oliver FH, Benua RS. Fourier amplitude and phase analysis in the clinical evaluation of patients with cardiomyopathy. Clin Nucl Med 19&1;9:314. 8. Widmann TF, Favrot L, Smith SC, Ehmann RW, Peterson KL. Temporal Fourier analysis in quantitative assessment of cardiac digital images. Computers in Cardiology, Seattle: IEEE publication, 1984. 9. Ratib 0, Rutishauser W. Recent developments of cardiac digit& radiography. Int J Cardiac Imaging 1985;1:29. 10. Bacharach SL, Green MV, Vitale D, White G, Douglas MA, Bonow RO, Larson SM. Optimum Fourier filtering of cardiac data: A minimum error method: Concise communication. J Nucl Med 1983;24:1176. 11. Rankin JS, McHale PA, Arentzen CE, Ling D, Greenfield JC, Anderson RW. The three-dimensional dynamic geometry of the left ventricle in the conscious dog. Circ Res 1976; 39304.
12. Leighton RF, Nelson AD, Brewster P. Subtle left ventricular asyxiergy with completely obstructed coronary arteries. Am J Cardiol 1983;52:693. 13. Kumada T, Karlmer JS, Pouleur H, Gallagher KP, Shirato K, Ross J Jr. Effecta of coronary occlusion on early ventricular diastolic events in conscious dogs. Am J Physiol 1979; 237:H542. 14. Bonow RO, Vitale DF, Bacharach SL, Frederick TM, Rosing DR, Kent KM, Green MV. Heterogeneous left ventricular regional function and impaired global diastolic filling in coronary artery disease: Reversal after coronary angioplasty. -_ Circulation 1983;68:III-167. 15. Sasavama S. Franklin D. Ross J Jr. Kemner WS. McKnown D. &mu& changes in’left ventricular wall thickness and their use in analyzing cardiac function in the conscious dog. Am J Cardiol 1976;38:870. 16. Gallagher KP, Matauzaki M, Osakada G, Kemper WS, Ross J
Heusch et al.
American
Jr. Effect of exercise on the relationship between myocardial blood flow and systolic wall thickening in dogs with acute coronary stenosis. Circ Res 1983;52:716. 17. Theroux P, Ross J Jr, Franklin D, Kemper WS, Sasayama S. Regional myocardial function in the conscious dog during acute coronary occlusion and the response to morphine, propranolol, nitroglycerin and lidocaine. Circulation 1976; 53~302.
18. Guth BD, Tajimi
T, Seitelberger R, Lee JD, Matsuxaki
M,
January 1987 Heart Journal
Ross J Jr. Experimental exercise-induced ischemia: Drug therapy can eliminate regional dysfunction and oxygen supply-demand imbalance. J Am Coll Cardiol 1986;7:1686. 19. Winer BJ. Statistical .nrincinles . in exnerimental desian. New York: McGraw-Hit1 Book Co, Inc, 1971. 20. Theroux P, Ross J Jr, Franklin D, Cove11 JW, Bloor CM, Sasayama S. Regional myocardial function and dimensions early and late after myocardial infarction in the unanesthetized dog. Circ Res 1977;40:158.
The effects of cytochrome C, an electron carrier in the process of oxidatlve phoephoryiation, on infarct size and regional ieft ventrkuiar function after a coronary ertery occiuaion were investigated. Thus, in 30 dogs, 1 mtnute’after feft anterior dercending coronary artery occlusion, oo”Tc-i8beied albumin mkrosgheres (8 mCi) were injected into the left atrium for subsequent assessment of the hygop8rfused zone, that Is, the 8rea at risk of infarction. Fifteen mfnutes after coronary artery o&us&n, doga were randomtzed Into 8 control group (n = 15) and a cytochrome C-treated group (n = la). The I8tter immediately received cytochrome C, 2.5 mg/kg intravenousiy. Six hours after coron8ry artery OCCiuSiOn the dogs were sacrificed and their left ventrlcies were cut into 3 mm thfck s&es. tnf8rct Sire ~88 determined by trjphenyitetrazolium chkwide statning end me8sured by plantmetry. The 88me slices were then submitted to autoradiogr8phy 8nd the hypogerfused zone was then measured by pfanimetry. The hypoperfused zone was 22 + 2% and 23 + 2% of the left ventrkie in the control and treated groups, respecttveiy (MS), indtcattng that the extent of myocardtum at risk before treatment was simller. The extent of the hypeperfuaed zone which evolved to necrosis was 90 + 3% in the control group but only 50 * 7% in the treeted group @ < 0.001). MyoC8rdi8i s8ivage In the treated group was paraiieied by improvement in syatoiic W8ii thickness of the fschemic segment as measured by two-dimenstonai echocardiography. Thus, cytochrome C reduced the extent of myocardlai necrosis by 44% and tmproved systoiic function of the ischemic myocardtum. (AM HEM J lg87;113:124.)
Andrew Zalewski, M.D., Sheldon Goldberg, M.D., Richard Krol, M.D., and Peter R. Maroko, M.D. Browns Mills, N. J., and Philadelphia, Pa.
Several therapeutic interventions aimed at decreasing myocardial oxygen demand or increasing oxygen supply have been reported to limit ischemic injuW- lp2 The latter can be achieved by means of coronary reperfusion? by retroperfusion via cardiac From the Deborah Cardiovascular Research Institute, and Thomas Jefferson University Hospital. Supported by grants from: W.W. Smith Char&able Trust and Johnson & Johnson Family of Companies. Received for publication May 8, 1986; accepted June 3, 1986. Reprint requests: Andrew Zalewski, M.D., Cardiac Catheterization Laboratory, Thomas Jefferson University Hospital, 111 South 11th St., Suite 5611D, Philadelphia, PA 19107. 124
veinq3 or by increasing the utilization of oxygen, for example, changing the oxygen-hemoglobin dissociation curve.” It remains to be estatished, however, if improved utilixation of oxygen within the ischemic myocardium, caused by exogenous electron carriers, could result in myocardial salvage. Cytochrome C is an important electron carrier which is located in the inner mitochondrial membranes of aerobic cells with especially high concentrations in the heart5 This substance and others which transfer electrons to molecular oxygen in the mitochonclria allow for the generation of adenosine triphosphate (ATP) in