European Journal of Radiology, 10 (1990)
19-27
Elsevier
19
EURRAD
00016
Quantitative thallium-201 myocardial exercise scintigraphy in normal subjects and patients with normal coronary arteries M.G. Niemeyer ‘p2, G.J. Laarman3, S. Lelbach2, M.J. Cramer2, L.T. Go4, J. F. Verzijlbergen 4, E.E. van der Wal15, A.H. Zwinderman6, C.A.P.L. Ascoop and E.K.J. Pauwels ’ ‘Department
of Diagnostic Radiology. University Hospital, Leiden. 2Department of Cardiology, St Antonius Hospital, Nieuwegein. ‘Department
of
Cardiology. Thorax Center, University Hospital, Rotterdam, 4Department of Nuclear Cardiology, St Antonius Hospital, Nieuwegein, 5Department of Cardiology, University Hospital, Leiden and 6Department of Medical Statistics, University of Leiden. Leiden,
(Received
8 August 1989; revised version received 10 October
Key words: Myocardium,
radionuclide
study; Radionuclide
1989; accepted
imaging, technology;
15 October
The Netherlands
1989)
Coronary vessel, disease
Abstract Quantitative thallium-201 myocardial exercise scintigraphy was tested in two patient populations representing alternative standards for cardiac normality: group I comprised 18 male uncatheterized patients with a low likelihood of coronary artery disease (CAD); group II contained 41 patients with normal coronary arteriograms. Group I patients were younger, they achieved a higher rate-pressure product than group II patients; all had normal findings by physical examination and electrocardiography at rest and exercise. Group II patients comprised 21 females, 11 patients showed abnormal electrocardiography at rest, and five patients showed ischemic ST depression during exercise. Twelve patients had signs of minimal CAD. Twelve patients revealed abnormal visual and quantitative thallium findings, three of these patients had minimal CAD. Profiles of uptake and washout of thallium-201 were derived from both patient groups, and compared with normal limits developed by Maddahi et al. Furthermore, low likelihood and angiographically normal patients may differ substantially, and both sets of normal patients should be considered when establishing criteria for abnormality in exercise thallium imaging. When commercial software containing normal limits for quantitative analysis of exercise thallium-201 imaging is used in clinical practice, it is mandatory to compare these with normal limits of uptake and washout of thallium-201, derived from the less heterogeneous group of low-likelihood subjects, which should be used in selecting a normal population to define normality.
Introduction For the quantitative interpretation of diagnostic test results used to evaluate coronary artery disease (CAD), it is necessary to define the normal range. Healthy volunteers [l-3], patients with a low probability of CAD [4,5] and patients with normal coronary angiograms [6] were used to provide standards in assessing the accuracy of diagnostic tests in cardiology. In the context of tests used for evaluation of CAD, the defmition of normal is usually based on findings in patients Address for reprints: Professor Dr. E.K.J. Pauwels, Department of Diagnostic Radiology, University Hospital, Rijnsburgerweg 10, 2333 AA Leiden, The Netherlands. 0720-048X/90/$03.50
0 1990 Elsevier Science Publishers
with no significant CAD at coronary arteriography [ 7-91. Alternatively, clinically normal subjects with an extremely low probability of CAD may be a purer group from which to define the normal range [4,5,10,11]. Quantitative methods for evaluation of myocardial uptake and washout of thallium-201 have increased the diagnostic efficacy of exercise thallium scintigraphy by using diagnostic criteria based on thallium kinetics in normal subjects [5,7,12]. Despite the differences between the normal populations, commercial software containing normal limits is used for quantifying thallium stress-redistribution scintigrams and applied at many institutions using a variety of scintillation cameras in different patient populations. There have been few reports on normal limits for quantitative analysis of up-
B.V. (Biomedical
Division)
20
TABLE I
group I, normal values were determined in 18 lowlikelihood patients referred for exercise thallium imaging. There were 18 men and no women, with a mean age of 42 k 8 years. Normality was defined by a normal physical examination, normal electrocardiogram at rest, greater than 85 % of maximal predicted heart rate with exercise and no chest pain or electrocardiographic alterations during exercise, and no hypertension [ 7, lo]. These patients with an extremely low probability of CAD are presumed to be normal without undergoing coronary arteriography. In group II, normal values were determined in 41 patients who had exercise thallium imaging and cardiac catheterization. In this group there were 20 men and 21 women, with a mean age of 52 & 11 years. None had CAD, defined as 50% or more luminal diameter coronary artery narrowing of one or more major coronary arteries as assessed by two independent observers who where blinded to the clinical findings and scintigraphy.
Clinical characteristics and data of thallium stress testing in normal subjects (group I) and patients with ‘normal’ coronary arteries
Clinical history
take and washout of thallium-201 after exercise from low-likelihood patients and normal catheterized patients [3,7,13]. The purpose of his study was (1) to compare the findings of quantitative exercise thallium imaging in 18 subjects with a low probability of CAD with those in 41 patients who were found to have normal coronary angiograms, (2) to compare results of computer profiles of uptake and thallium-201 washout curves of these patient groups with profiles defined by Maddahi et al.
[51. Methods Patient population
This study consisted of two groups (I and II). The patient characteristics are presented in Table I. In
(group II) Group I
Group II
Total
18
41
Age (years)
42.2 + 8.2
51.9 f 10.7
Males/females
18/O
2012 1
Chest pain No pain Nonanginal Atypical AP NYHA II AP NYHA III AP during the test Myocardial
infarction (QRS)
Exercise test results Maximal HR % predicted heart rate Maximal BP Double product/1000 X-ECG positive X-ECG negative Nondiagnostic test Scintigraphic findings VA positive VA negative QA positive QA negative
0
3 16 13
18 0 0 0 0 0
165 101 189.2 31.2 0 18 0
0
18 0 18
3
a 8 +_ 2.8 t 18.6 + 3.4
151.7 f 12.9 96.7 +_ 6.8 177.5 + 14.9 23.7 _+ 4.3 5 25 11
9 32 5 36
AP, angina pectoris; NYHA, New York Heart Association; HR, heart rate; BP, blood pressure; X-ECG, exercise electrocardiogram; VA, visual analysis; QA, quantitative analysis.
All patients were interviewed before testing. The clinical history was characterized in terms of age, sex and symptoms (classified with respect to substernal location, exertional precipitation and prompt relief by rest or nitroglycerin). Typical angina was defined by the presence of all three characteristics, atypical angina by any two and nonanginal discomfort by fewer than two; patients who denied discomfort were termed as asymptomatic [ 141. Exercise
electrocardiography
Exercise electrocardiography was carried out in the upright position on a calibrated bicycle ergometer (Lode) using a test protocol with a stepwise increased work load. The initial external workload was 60 W during 3 min. Thereafter, the load was increased every 3 min by 30 W until one of the stop criteria was fulfilled. The following criteria for exercise termination were used: target heart rate, angina, exhaustion, dyspnea, dizziness, disturbances of the heart rhythm (repetitive extrasystoles or tachycardias) or conduction disturbances (bundle branch block or atrioventricular conduction problems) fall in systolic blood pressure compared with an earlier stage of effort or an ischemic ST segment depression of at least 0.2 mV and a duration of 80 ms (leads CMS-CCS). In all patients antianginal drugs, such as B-blocking agents and calcium antagonists, were stopped 24 h before the test, and digitalis derivates were withheld 2 weeks before the test. Thallium scintigraphy was performed in conjunction with the exercise test.
21
Thallium-201 imaging Our method for obtaining the exercise thallium-201 scintigrams has previously been described [ 151. A dose of 74 MBq (2 mCi) of thallium-201 was injected through an indwelling intravenous cannula at maximal exercise, which was continued for 1 min after injection. After termination of exercise, multiple view myocardial scintigrams were obtained at approx. 5 min and 4 h after injection of thallium. At each interval, imaging was performed in the anterior, 30’ and 70’ left anterior oblique (LAO) views for 10 min per view. The patient’s physical activities and food consumption were restricted between the two recordings. For imaging a General Electric small field-of-view camera (26 cm) was used equipped with 37 photo-multiplier tubes, $-inch thick sodium iodide crystal and a general purpose, low energy, hexagonal parallel hole collimator was used. A 20% energy window centered on the 80-keV photopeak was used. There was a minimum of 300 000 counts per image obtained. All images were stored by the computer on magnetic disc in a 128 x 128 x 8-bit matrix (MDS-A2, Nucletron). Visual analysis The visual assessment was performed by means of a semi-quantitative analysis [ 161. Each image was divided into five roughly equal segments for a total of 15 segmentsnumbered 1-15 (Fig. la). The thallium-201 scintigrams of the initial uptake and the delayed interval or redistribution phase were inspected on a computer gray scale without smoothing or background subtraction. Matching views from initial uptake and (1
b
ANTERIOR
ANTERIOR
6
LEFT
ANTERIOR
RIGHT
CORONARY
LEFT
CIRCUMFLEX
LAO 30
LAO
LAO 30
LAO 70
60
DESCENDING
70
. o
CORONARY
15
ARTERY
ARTERY ARTERY
Fig. 1. Flow regions of the major coronary arteries as used in (a) semiquantitative (visual) analysis and (b) quantitative analysis.
delayed images were displayed side by side for comparison. Computer processing and analysis We have previously described details of the technique for computer processing and quantitative analysis of thallium-201 images [ 151. Briefly, the computer algorithm involved the following steps: (1) after ninepoint weighted image smoothing each image (in the initial and redistribution phase) was compensated for tissue cross-talk using a interpolative background subtraction technique described by Goris et al., modified using a proximity weighting function described by Watson et al. [ 17,181. (2) An ellipse was centered closely around the left ventricle; the first radius was defined by the operator from the center of the ellipse accross the apex. (3) One hundred and twenty radii spaced at 3 ’ intervals were constructed from the center of the left ventricle. (4) Along each radius the point with maximal thallium-201 activity was chosen and plotted as number of counts versus angle (sample point). Combination of all 120 points resulted in circumferential profiles of myocardial thallium-201 activity [5,19,20]. (5) The three profiles were then aligned so that the 180 ’ point in each view corresponded to the visually located apex. (6) These curves in the postexercise phase were normalized to the maximum count ( = 100%) in the profile of each view. (7) In addition to the initial profile, washout circumferential profiles (after 4 h) were generated by calculating percent washout for each point from the time of stress to the delayed imaging ((counts initial - counts delayed/counts initial) x 100). The results of both groups were used separately to establish ‘normal limits’. The mean value and standard deviation were established based on the pooled data for each of the 120 angular segments of the anterior, 30” LAO and 70” LAO views of the initial and delayed images. For each segment the mean minus two standard deviations was considered the lower limit of normal. This was displayed as a profile of normal distribution. The normal limits for washout at 4 h were determined in a similar way. The uptake and washout rate circumferential profiles were compared to the limits of the profiles of normal population available from a study by Maddahi and colleagues [5]. In addition to the calculations for the entire myocardium according to the three views, flow regions were calculated for the left anterior descending artery (LAD) (anterior), right coronary artery (RCA) (inferior) and the left circumflex artery (LCX) (lateral-posterior) as shown in Fig. lb. The angle in each view designated to the apex was considered as not belonging to a specific artery.
22
Coronary arteriography and left ventricular angiography
Coronary arteriography was performed in the patients of group II in multiple left anterior and right anterior oblique projections and with crania-caudal projections [21] according to the Sones or Judkins technique [ 22,231. The angiograms were interpreted independently by two cardiologists; disagreement was resolved by a third independent cardiologist. Patients were considered to have significant coronary artery disease if they had 50% or more luminal diameter coronary narrowing. The segmental wall motion was interpreted as normal, hypokinetic, akinetic or dyskinetic. Results The characteristics ofclinical, hemodynamic, electrocardiographic and scintigraphic findings of both groups are shown in Table I. Group I. The patients in group I (n = 18) were younger, all were male, and they achieved higher maximal heart rate and higher maximal blood pressure during exercise than the patients of group II (n = 41). In group I all patients experienced nonanginal chest pain. All had normal findings on physical examination and electrocardiography at rest and exercise. No cardiovascular symptoms developed during the test. Based on these findings, the patients of group I revealed a low probability of CAD [ 11,241. Visual interpretation and quantitative analysis of the scintigrams showed normal test results in all patients. Group II. This group consisted of 20 males and 21 females. Three patients experienced no chest pain, 16 nonanginal pain, 13 atypical chest pain and eight typical anginal pain (NYHA II or III). One patient had anginal pain during the test. Three patients had signs of previous myocardial infarction. Eleven patients had abnormal electrocardiography at rest, and five patients showed ischemic ST segment depression during exercise. Twelve patients had signs of minimal coronary artery disease (narrowings O-49% in diameter). All patients revealed normal findings on left ventricular angiography. Visual analysis revealed nine (22%) and quantitative analysis five (12%) patients in group II with abnormal thallium images. Among the 12 patients with abnormal visual and/or quantitative thallium findings, eight were male and four showed minimal coronary artery disease. Of the nine patients with visually analyzed initial uptake defects, eight patients showed redistribution. Of the 12 ‘false-positive’ scintigrams in group II there was discordance in ten between visual interpretation and quantitative analysis. Three patients showed abnormalities in one flow region (anterior, two
and inferior, one, respectively). Seven patients revealed two abnormal flow regions (anterior/lateral, one, anterior/inferior, four and inferior/lateral, three, respectively). One patient had scintigraphic abnormalities in three flow regions. Initial uptake profiles and washout curves were derived from both patient groups. Mean regional percentage washout at 4 h in group I ranged from 47.43 + 6.62% to 53.43 + 8.62% (mean f S.D.: 50.31 + 7.53%) and in group II from 48.21 k 6.28% to (mean + S.D.: 50.15 & 11.19%) 52.86 + 14.03% (Fig. 2). Comparing these profiles with normal limits developed by Maddahi et al. [5] (Fig. 3), it is shown that the relative uptake is higher in the posterolateral region. Comparing the profiles of group I with group II, the uptake in the anterior region (LAO 70) is higher (Fig. 3~).
Patients with low probability 80
% Washcut
80-160
160-210
Myocardial a
m
Anterior
210-300
Regions (in degree) m
L/0
30
0
LAO
70
Patients with ‘normal’ coronary arteries 60
% Washcut
60-160
160-210
Myocardial
b m
Anterior
210-300
Regions (in degree) m
L/O
30
m
LAO
70
Fig. 2. Display of the mean regional washout of thallium-201 after 4 h for (a) group I and (b) group II. LAO, left anterior oblique.
23
II[
STRESS
it&-
4h
fi
r
‘&$+q
E
_&.,c””
R
ioi
60 150 no m barsal teptul apex wt. lattraul
360 bsrsal rcptufi
amx
wt.
laterah
IO&-
51
.-Y-J=
;...-...-~ .,..-““‘~~...,“,.~~. _..._.,,.,.. .”
‘0
. . . . .._..
+
300 apex
Post.
3(1
lattraal
P
150 zlo apex post.
m lateraal
50
1Wl
anterior
WPCX
inf.
post.
Fig. 3. (a) Comparison ofthallium uptake (stress) and washout profiles (mean - 2 S.D.) ofgroup I (normal subjects) with normal profiles (solid lines) according to Maddahi et al. [5] in three views. (b, see page 24) Comparison of thallium uptake (stress) and washout profiles (mean - 2 SD.) of group II (normal catheterized patients) with normal profiles according to Maddahi et al. (c, see page 25) Comparison of thallium uptake (stress) and washout profiles (mean - 2 S.D.) of group II (normal catheterized patients) with normal profiles (solid lines) of group I (normal subjects).
Discussion The probability of having CAD may be calculated from several clinical variables including age, sex, type of chest pain and risk factors for CAD [ 10,251. Patients with an extremely low probability of CAD are then presumed to be normal without undergoing coronary angiography. Although this presumption may be valid,
the true prevalence of coronary artery disease in this population is unknown. Two different normal populations were evaluated in this study: (1) patients with a low likelihood of CAD and (2) patients with normal coronary arteries. The groups differed in several important clinical characteristics (Table I). The patients in the low probability group were quite similar to one another; they were all
24
nW%,
STRESS I I
,
4h WRSHOUr
l((x-
I
s( ..a
19-27
Elsevier
19
EURRAD
00016
Quantitative thallium-201 myocardial exercise scintigraphy in normal subjects and patients with normal coronary arteries M.G. Niemeyer ‘p2, G.J. Laarman3, S. Lelbach2, M.J. Cramer2, L.T. Go4, J. F. Verzijlbergen 4, E.E. van der Wal15, A.H. Zwinderman6, C.A.P.L. Ascoop and E.K.J. Pauwels ’ ‘Department
of Diagnostic Radiology. University Hospital, Leiden. 2Department of Cardiology, St Antonius Hospital, Nieuwegein. ‘Department
of
Cardiology. Thorax Center, University Hospital, Rotterdam, 4Department of Nuclear Cardiology, St Antonius Hospital, Nieuwegein, 5Department of Cardiology, University Hospital, Leiden and 6Department of Medical Statistics, University of Leiden. Leiden,
(Received
8 August 1989; revised version received 10 October
Key words: Myocardium,
radionuclide
study; Radionuclide
1989; accepted
imaging, technology;
15 October
The Netherlands
1989)
Coronary vessel, disease
Abstract Quantitative thallium-201 myocardial exercise scintigraphy was tested in two patient populations representing alternative standards for cardiac normality: group I comprised 18 male uncatheterized patients with a low likelihood of coronary artery disease (CAD); group II contained 41 patients with normal coronary arteriograms. Group I patients were younger, they achieved a higher rate-pressure product than group II patients; all had normal findings by physical examination and electrocardiography at rest and exercise. Group II patients comprised 21 females, 11 patients showed abnormal electrocardiography at rest, and five patients showed ischemic ST depression during exercise. Twelve patients had signs of minimal CAD. Twelve patients revealed abnormal visual and quantitative thallium findings, three of these patients had minimal CAD. Profiles of uptake and washout of thallium-201 were derived from both patient groups, and compared with normal limits developed by Maddahi et al. Furthermore, low likelihood and angiographically normal patients may differ substantially, and both sets of normal patients should be considered when establishing criteria for abnormality in exercise thallium imaging. When commercial software containing normal limits for quantitative analysis of exercise thallium-201 imaging is used in clinical practice, it is mandatory to compare these with normal limits of uptake and washout of thallium-201, derived from the less heterogeneous group of low-likelihood subjects, which should be used in selecting a normal population to define normality.
Introduction For the quantitative interpretation of diagnostic test results used to evaluate coronary artery disease (CAD), it is necessary to define the normal range. Healthy volunteers [l-3], patients with a low probability of CAD [4,5] and patients with normal coronary angiograms [6] were used to provide standards in assessing the accuracy of diagnostic tests in cardiology. In the context of tests used for evaluation of CAD, the defmition of normal is usually based on findings in patients Address for reprints: Professor Dr. E.K.J. Pauwels, Department of Diagnostic Radiology, University Hospital, Rijnsburgerweg 10, 2333 AA Leiden, The Netherlands. 0720-048X/90/$03.50
0 1990 Elsevier Science Publishers
with no significant CAD at coronary arteriography [ 7-91. Alternatively, clinically normal subjects with an extremely low probability of CAD may be a purer group from which to define the normal range [4,5,10,11]. Quantitative methods for evaluation of myocardial uptake and washout of thallium-201 have increased the diagnostic efficacy of exercise thallium scintigraphy by using diagnostic criteria based on thallium kinetics in normal subjects [5,7,12]. Despite the differences between the normal populations, commercial software containing normal limits is used for quantifying thallium stress-redistribution scintigrams and applied at many institutions using a variety of scintillation cameras in different patient populations. There have been few reports on normal limits for quantitative analysis of up-
B.V. (Biomedical
Division)
20
TABLE I
group I, normal values were determined in 18 lowlikelihood patients referred for exercise thallium imaging. There were 18 men and no women, with a mean age of 42 k 8 years. Normality was defined by a normal physical examination, normal electrocardiogram at rest, greater than 85 % of maximal predicted heart rate with exercise and no chest pain or electrocardiographic alterations during exercise, and no hypertension [ 7, lo]. These patients with an extremely low probability of CAD are presumed to be normal without undergoing coronary arteriography. In group II, normal values were determined in 41 patients who had exercise thallium imaging and cardiac catheterization. In this group there were 20 men and 21 women, with a mean age of 52 & 11 years. None had CAD, defined as 50% or more luminal diameter coronary artery narrowing of one or more major coronary arteries as assessed by two independent observers who where blinded to the clinical findings and scintigraphy.
Clinical characteristics and data of thallium stress testing in normal subjects (group I) and patients with ‘normal’ coronary arteries
Clinical history
take and washout of thallium-201 after exercise from low-likelihood patients and normal catheterized patients [3,7,13]. The purpose of his study was (1) to compare the findings of quantitative exercise thallium imaging in 18 subjects with a low probability of CAD with those in 41 patients who were found to have normal coronary angiograms, (2) to compare results of computer profiles of uptake and thallium-201 washout curves of these patient groups with profiles defined by Maddahi et al.
[51. Methods Patient population
This study consisted of two groups (I and II). The patient characteristics are presented in Table I. In
(group II) Group I
Group II
Total
18
41
Age (years)
42.2 + 8.2
51.9 f 10.7
Males/females
18/O
2012 1
Chest pain No pain Nonanginal Atypical AP NYHA II AP NYHA III AP during the test Myocardial
infarction (QRS)
Exercise test results Maximal HR % predicted heart rate Maximal BP Double product/1000 X-ECG positive X-ECG negative Nondiagnostic test Scintigraphic findings VA positive VA negative QA positive QA negative
0
3 16 13
18 0 0 0 0 0
165 101 189.2 31.2 0 18 0
0
18 0 18
3
a 8 +_ 2.8 t 18.6 + 3.4
151.7 f 12.9 96.7 +_ 6.8 177.5 + 14.9 23.7 _+ 4.3 5 25 11
9 32 5 36
AP, angina pectoris; NYHA, New York Heart Association; HR, heart rate; BP, blood pressure; X-ECG, exercise electrocardiogram; VA, visual analysis; QA, quantitative analysis.
All patients were interviewed before testing. The clinical history was characterized in terms of age, sex and symptoms (classified with respect to substernal location, exertional precipitation and prompt relief by rest or nitroglycerin). Typical angina was defined by the presence of all three characteristics, atypical angina by any two and nonanginal discomfort by fewer than two; patients who denied discomfort were termed as asymptomatic [ 141. Exercise
electrocardiography
Exercise electrocardiography was carried out in the upright position on a calibrated bicycle ergometer (Lode) using a test protocol with a stepwise increased work load. The initial external workload was 60 W during 3 min. Thereafter, the load was increased every 3 min by 30 W until one of the stop criteria was fulfilled. The following criteria for exercise termination were used: target heart rate, angina, exhaustion, dyspnea, dizziness, disturbances of the heart rhythm (repetitive extrasystoles or tachycardias) or conduction disturbances (bundle branch block or atrioventricular conduction problems) fall in systolic blood pressure compared with an earlier stage of effort or an ischemic ST segment depression of at least 0.2 mV and a duration of 80 ms (leads CMS-CCS). In all patients antianginal drugs, such as B-blocking agents and calcium antagonists, were stopped 24 h before the test, and digitalis derivates were withheld 2 weeks before the test. Thallium scintigraphy was performed in conjunction with the exercise test.
21
Thallium-201 imaging Our method for obtaining the exercise thallium-201 scintigrams has previously been described [ 151. A dose of 74 MBq (2 mCi) of thallium-201 was injected through an indwelling intravenous cannula at maximal exercise, which was continued for 1 min after injection. After termination of exercise, multiple view myocardial scintigrams were obtained at approx. 5 min and 4 h after injection of thallium. At each interval, imaging was performed in the anterior, 30’ and 70’ left anterior oblique (LAO) views for 10 min per view. The patient’s physical activities and food consumption were restricted between the two recordings. For imaging a General Electric small field-of-view camera (26 cm) was used equipped with 37 photo-multiplier tubes, $-inch thick sodium iodide crystal and a general purpose, low energy, hexagonal parallel hole collimator was used. A 20% energy window centered on the 80-keV photopeak was used. There was a minimum of 300 000 counts per image obtained. All images were stored by the computer on magnetic disc in a 128 x 128 x 8-bit matrix (MDS-A2, Nucletron). Visual analysis The visual assessment was performed by means of a semi-quantitative analysis [ 161. Each image was divided into five roughly equal segments for a total of 15 segmentsnumbered 1-15 (Fig. la). The thallium-201 scintigrams of the initial uptake and the delayed interval or redistribution phase were inspected on a computer gray scale without smoothing or background subtraction. Matching views from initial uptake and (1
b
ANTERIOR
ANTERIOR
6
LEFT
ANTERIOR
RIGHT
CORONARY
LEFT
CIRCUMFLEX
LAO 30
LAO
LAO 30
LAO 70
60
DESCENDING
70
. o
CORONARY
15
ARTERY
ARTERY ARTERY
Fig. 1. Flow regions of the major coronary arteries as used in (a) semiquantitative (visual) analysis and (b) quantitative analysis.
delayed images were displayed side by side for comparison. Computer processing and analysis We have previously described details of the technique for computer processing and quantitative analysis of thallium-201 images [ 151. Briefly, the computer algorithm involved the following steps: (1) after ninepoint weighted image smoothing each image (in the initial and redistribution phase) was compensated for tissue cross-talk using a interpolative background subtraction technique described by Goris et al., modified using a proximity weighting function described by Watson et al. [ 17,181. (2) An ellipse was centered closely around the left ventricle; the first radius was defined by the operator from the center of the ellipse accross the apex. (3) One hundred and twenty radii spaced at 3 ’ intervals were constructed from the center of the left ventricle. (4) Along each radius the point with maximal thallium-201 activity was chosen and plotted as number of counts versus angle (sample point). Combination of all 120 points resulted in circumferential profiles of myocardial thallium-201 activity [5,19,20]. (5) The three profiles were then aligned so that the 180 ’ point in each view corresponded to the visually located apex. (6) These curves in the postexercise phase were normalized to the maximum count ( = 100%) in the profile of each view. (7) In addition to the initial profile, washout circumferential profiles (after 4 h) were generated by calculating percent washout for each point from the time of stress to the delayed imaging ((counts initial - counts delayed/counts initial) x 100). The results of both groups were used separately to establish ‘normal limits’. The mean value and standard deviation were established based on the pooled data for each of the 120 angular segments of the anterior, 30” LAO and 70” LAO views of the initial and delayed images. For each segment the mean minus two standard deviations was considered the lower limit of normal. This was displayed as a profile of normal distribution. The normal limits for washout at 4 h were determined in a similar way. The uptake and washout rate circumferential profiles were compared to the limits of the profiles of normal population available from a study by Maddahi and colleagues [5]. In addition to the calculations for the entire myocardium according to the three views, flow regions were calculated for the left anterior descending artery (LAD) (anterior), right coronary artery (RCA) (inferior) and the left circumflex artery (LCX) (lateral-posterior) as shown in Fig. lb. The angle in each view designated to the apex was considered as not belonging to a specific artery.
22
Coronary arteriography and left ventricular angiography
Coronary arteriography was performed in the patients of group II in multiple left anterior and right anterior oblique projections and with crania-caudal projections [21] according to the Sones or Judkins technique [ 22,231. The angiograms were interpreted independently by two cardiologists; disagreement was resolved by a third independent cardiologist. Patients were considered to have significant coronary artery disease if they had 50% or more luminal diameter coronary narrowing. The segmental wall motion was interpreted as normal, hypokinetic, akinetic or dyskinetic. Results The characteristics ofclinical, hemodynamic, electrocardiographic and scintigraphic findings of both groups are shown in Table I. Group I. The patients in group I (n = 18) were younger, all were male, and they achieved higher maximal heart rate and higher maximal blood pressure during exercise than the patients of group II (n = 41). In group I all patients experienced nonanginal chest pain. All had normal findings on physical examination and electrocardiography at rest and exercise. No cardiovascular symptoms developed during the test. Based on these findings, the patients of group I revealed a low probability of CAD [ 11,241. Visual interpretation and quantitative analysis of the scintigrams showed normal test results in all patients. Group II. This group consisted of 20 males and 21 females. Three patients experienced no chest pain, 16 nonanginal pain, 13 atypical chest pain and eight typical anginal pain (NYHA II or III). One patient had anginal pain during the test. Three patients had signs of previous myocardial infarction. Eleven patients had abnormal electrocardiography at rest, and five patients showed ischemic ST segment depression during exercise. Twelve patients had signs of minimal coronary artery disease (narrowings O-49% in diameter). All patients revealed normal findings on left ventricular angiography. Visual analysis revealed nine (22%) and quantitative analysis five (12%) patients in group II with abnormal thallium images. Among the 12 patients with abnormal visual and/or quantitative thallium findings, eight were male and four showed minimal coronary artery disease. Of the nine patients with visually analyzed initial uptake defects, eight patients showed redistribution. Of the 12 ‘false-positive’ scintigrams in group II there was discordance in ten between visual interpretation and quantitative analysis. Three patients showed abnormalities in one flow region (anterior, two
and inferior, one, respectively). Seven patients revealed two abnormal flow regions (anterior/lateral, one, anterior/inferior, four and inferior/lateral, three, respectively). One patient had scintigraphic abnormalities in three flow regions. Initial uptake profiles and washout curves were derived from both patient groups. Mean regional percentage washout at 4 h in group I ranged from 47.43 + 6.62% to 53.43 + 8.62% (mean f S.D.: 50.31 + 7.53%) and in group II from 48.21 k 6.28% to (mean + S.D.: 50.15 & 11.19%) 52.86 + 14.03% (Fig. 2). Comparing these profiles with normal limits developed by Maddahi et al. [5] (Fig. 3), it is shown that the relative uptake is higher in the posterolateral region. Comparing the profiles of group I with group II, the uptake in the anterior region (LAO 70) is higher (Fig. 3~).
Patients with low probability 80
% Washcut
80-160
160-210
Myocardial a
m
Anterior
210-300
Regions (in degree) m
L/0
30
0
LAO
70
Patients with ‘normal’ coronary arteries 60
% Washcut
60-160
160-210
Myocardial
b m
Anterior
210-300
Regions (in degree) m
L/O
30
m
LAO
70
Fig. 2. Display of the mean regional washout of thallium-201 after 4 h for (a) group I and (b) group II. LAO, left anterior oblique.
23
II[
STRESS
it&-
4h
fi
r
‘&$+q
E
_&.,c””
R
ioi
60 150 no m barsal teptul apex wt. lattraul
360 bsrsal rcptufi
amx
wt.
laterah
IO&-
51
.-Y-J=
;...-...-~ .,..-““‘~~...,“,.~~. _..._.,,.,.. .”
‘0
. . . . .._..
+
300 apex
Post.
3(1
lattraal
P
150 zlo apex post.
m lateraal
50
1Wl
anterior
WPCX
inf.
post.
Fig. 3. (a) Comparison ofthallium uptake (stress) and washout profiles (mean - 2 S.D.) ofgroup I (normal subjects) with normal profiles (solid lines) according to Maddahi et al. [5] in three views. (b, see page 24) Comparison of thallium uptake (stress) and washout profiles (mean - 2 SD.) of group II (normal catheterized patients) with normal profiles according to Maddahi et al. (c, see page 25) Comparison of thallium uptake (stress) and washout profiles (mean - 2 S.D.) of group II (normal catheterized patients) with normal profiles (solid lines) of group I (normal subjects).
Discussion The probability of having CAD may be calculated from several clinical variables including age, sex, type of chest pain and risk factors for CAD [ 10,251. Patients with an extremely low probability of CAD are then presumed to be normal without undergoing coronary angiography. Although this presumption may be valid,
the true prevalence of coronary artery disease in this population is unknown. Two different normal populations were evaluated in this study: (1) patients with a low likelihood of CAD and (2) patients with normal coronary arteries. The groups differed in several important clinical characteristics (Table I). The patients in the low probability group were quite similar to one another; they were all
24
nW%,
STRESS I I
,
4h WRSHOUr
l((x-
I
s( ..a
