Author + information
- Received November 15, 2002
- Revision received March 23, 2003
- Accepted April 10, 2003
- Published online July 2, 2003.
- Ronaldo S.L Lima, MD, PhD*,§,
- Denny D Watson, PhD‡,
- Allen R Goode, MS*,
- Mir S Siadaty, MD, MS†,
- Michael Ragosta, MD, FACC*,
- George A Beller, MD, MACC* and
- Habib Samady, MD, FACC*,* ()
- ↵*Reprint requests and correspondence:
Dr. Habib Samady, Cardiovascular Division, Department of Medicine, P.O. Box 800158, University of Virginia Health System, Charlottesville, Virginia 22908-0158, USA.
Objectives We hypothesized that combining functional assessment to perfusion enhances the ability of electrocardiographic gating Tc-99m sestamibi single photon emission computed tomography (gated SPECT) myocardial perfusion imaging (MPI) to detect defects in multiple vascular territories in patients with severe three-vessel coronary artery disease (3VD).
Background In patients with 3VD, perfusion defects in multiple vascular territories may not always be evident due to globally reduced perfusion.
Methods Gated SPECT MPIs were interpreted sequentially with perfusion first, followed by combined perfusion/function, in 143 patients with angiographic 3VD and a control group of 112 non-3VD patients. All patients underwent coronary arteriography within one month of MPI.
Results In 3VD patients, combined perfusion/function analysis yielded significantly greater numbers of abnormal segments/patient (6.2 ± 4.7 vs. 4.1 ± 2.8, p < 0.001) and more defects in multiple vascular territories (60% vs. 46%, p < 0.05) than perfusion alone. In the control group, there were no differences between the combined perfusion/function and perfusion alone interpretations. Multivariate analysis of 15 different clinical, stress, and scintigraphic variables in all patients revealed age (p < 0.0001) and number of abnormal vascular territories by combined perfusion/function (p < 0.0001) to be the most powerful predictors of 3VD. Addition of functional data to clinical, stress, and perfusion yielded a significant increase in the predictive value of 3VD (global chi-square: 131.7 vs. 89.8, p < 0.00001). Specificity of combined perfusion/function analysis was not lower than perfusion alone (72% vs. 69%, p = NS).
Conclusions Adjunctive assessment of function with perfusion by gated SPECT MPI enhances the detection of defects in multiple vascular territories in patients with severe 3VD, without adversely affecting its specificity.
Among patients with coronary artery disease (CAD), those with left main and three-vessel disease (3VD) are at highest risk for death (1,2). A primary goal of noninvasive testing is to identify this high-risk group in whom revascularization results in improvement in symptoms and survival. Although in the pooled literature the overall sensitivity of myocardial perfusion imaging (MPI) to detect multivessel disease is 80% to 95% (3,4), only 29% of patients with angiographic 3VD show defects in all three territories (5). Similarly, in patients with left main disease, multivessel perfusion defects on thallium-201 imaging is found in only 49% of patients (6). A possible mechanism for this underestimation of disease, in patients with critically diseased vessels in all vascular beds, is balanced ischemia yielding few or no relative perfusion defects. In this situation, there is no normal scintigraphic reference segment with which to compare segments supplied by stenotic vessels.
Electrocardiographic gating Tc-99m sestamibi single photon emission computed tomography (SPECT) perfusion imaging (gated SPECT MPI) has been shown to provide added diagnostic and prognostic value to perfusion imaging (7–9). Persistent post-stress wall motion abnormalities on gated SPECT has been shown to increase the accuracy of identifying severe and extensive CAD in patients representing the spectrum of coronary disease (10).
In patients with severe 3VD and/or left main CAD, little information is available on whether combined perfusion/function analysis on gated SPECT would have incremental value over perfusion alone for detection of disease in multiple vascular territories, without adversely affecting the specificity of disease detection. Accordingly, the purpose of this study was to determine if post-stress gated SPECT MPI can more accurately identify patients with 3VD, compared with quantitative perfusion SPECT MPI, without a corresponding decrease in specificity in patients without 3VD.
Between July 1999 and March 2002, all patients who underwent stress gated Tc-99m sestamibi SPECT and coronary arteriography within 30 days who had significant 3VD or its equivalent formed the study cohort. Patients with ischemic events between MPI and coronary angiography were excluded. The angiographic criteria for inclusion into the 3VD group were: 1) presence of ≥50% narrowing of the internal diameter of left main coronary artery plus an ≥70% narrowing of the right coronary artery; or 2) an ≥70% narrowing of the left anterior descending artery, right coronary artery, and left circumflex artery, or their major branches. Patients who had undergone coronary bypass grafting were excluded. A total of 143 patients fulfilled these criteria and were classified as the 3VD group. The control group constituted 112 patients without 3VD who underwent both stress gated Tc-99m sestamibi SPECT and coronary arteriography within 30 days. This cohort included 43 patients with no CAD (0VD), 43 with one-vessel disease (1VD), and 26 with two-vessel disease (2VD).
Patients deemed to be relatively low risk were encouraged to come off medications for 48 h before the study; SPECT imaging protocol was performed as previously described (9,11–12). A same-day rest-stress Tc-99m sestamibi gated SPECT MPI protocol was performed, with a weight-adjusted dose of 240 to 300 MBq of Tc-99m sestamibi at rest and 750 to 900 MBq at peak stress at least 3 h after the first injection. Images were obtained 60 min after tracer injection. Briefly, gated images were acquired on a Picker Prism 3000 triple-headed gamma camera using a low energy, high resolution, and a parallel-hole collimator (Picker, Cleveland, Ohio). Sixty projection images were acquired using the step-and-shoot method. An electrocardiogram R-wave detector provided a gate to acquire eight frames per cardiac cycle during the post-stress acquisition. The University of Virginia quantitative gated-SPECT software was used for image analysis (14 segments per patient). Three sets of SPECT slices representing summed, end-diastolic, and end-systolic images were formed for computer display. The summed image set was normalized conventionally for comparison of stress and rest images. The gated images were normalized to the region of highest activity on the end-systolic image set.
Studies of controls were mixed randomly with the 3VD studies and read blindly by two readers: perfusion alone followed by combined perfusion/function. Quantitative values of relative tracer activity in each of 14 segments are compared to a gender-specific normal database and flagged if abnormal using the same criteria as with clinical readings (lower limits of normal set to a normalcy rate >90% on the normal database).
Wall-motion abnormalities were determined based on the quantitative thickening fractions (TF). Gated SPECT function studies in our system are displayed as end-diastolic and end-systolic images juxtaposed and scaled to the end-systolic image set. We use the partial volume effect to compute segmental TFs in each of the same 14 segments used for quantitative perfusion analysis (11,12). Regional quantitative TFs are compared to a normal database and flagged as abnormal if they are more than 2 SDs below the normal average. The use of quantitative values based on the partial volume effect (ratio of end-diastolic counts over end-systolic counts) does not require endocardial edge detection.
For the purposes of this study, we additionally graded scans as showing defects in single or multiple vascular regions (anterior, lateral, or inferior). If a wall motion abnormality (significantly reduced TF) was indicated in the region of a perfusion abnormality or in an adjacent segment, it was considered to be in the same vascular territory. Wall motion abnormalities can sometimes show in segments rotationally displaced from the perfusion abnormality as a result of cardiac torsional rotation during contraction.
The key to this study was the identification of abnormal wall motion in segments for which there were no concordant perfusion abnormalities. As stated, these were identified only if the computed value of TF was significantly reduced in a region remote from another identifiable perfusion abnormality. Because there are 14 segments, a 2 SD error in computed TF in at least one of 14 segments can occur with significant statistical frequency. Consequently, if the perfusion scan was entirely normal except for a single flagged segment of reduced TF, the wall motion was still graded as entirely normal. However, if there was an identifiable perfusion abnormality (with or without an associated TF abnormality) and additional segment(s) of reduced TF in a different vascular territory where there was not a relative perfusion defect, then this was interpreted as abnormal function in a remote vascular territory.
All patients underwent coronary angiography within 30 days of MPI (7 ± 8 days), and no event or procedure occurred between the two exams. Coronary angiography was performed in multiple oblique projections. Coronary artery narrowing was assessed by visual analysis and reported as the percentage luminal diameter stenosis. Left main stenosis was considered significant when there was ≥50% lesion. In all other vessels, a luminal narrowing of ≥70% was considered significant.
Data are presented as mean ± SD when appropriate. Univariate comparison of continuous variables between 3VD and control groups was performed using the ttest and discrete variables using the chi-square test. Multivariate analysis of 15 variables was performed using a logistic model where the outcome was the presence of 3VD.
The number of abnormal SPECT segments per patient, abnormal SPECT vascular territories per patient, and abnormal SPECT MPI per group were compared by perfusion alone versus combined perfusion/function using ttest for continuous variables and chi-square tests for discrete variables. Global chi-square statistic by likelihood ratio test was then used in the whole cohort to compare the overall performance of perfusion versus combined perfusion/function for detection of angiographic 3VD.
To test the hypothesis that the incremental value of combined perfusion/function was greater in patients with 3VD compared with control patients, an ordinal logistic regression model was used. A score, with values from 3 to −3, was ascribed to each patient based on whether the number of abnormal SPECT vascular territories on combined perfusion/function versus perfusion alone analysis matched the number of angiographically diseased vessels (e.g., a score of +3 was ascribed to a patient with 3VD, if perfusion showed no defects and combined perfusion/function showed defects in three vascular territories). Thus, the mathematical sum of all these scores provides a relative value of the combined perfusion/function analysis versus perfusion alone for accurate diagnosis of patients with angiographic 0VD, 1VD, 2VD, and 3VD. Other clinical, stress, perfusion, and function variables were entered into this model to adjust for possible confounding effects and for differentially distributed background variables. From this regression model, an odds ratio for correctly matching number of angiographically diseased vessels (zero, one, two, or three) was calculated for combined perfusion/function versus perfusion alone. A p value of <0.05 was considered statistically significant. Analyses were done by SPSS v.10 and R v.1.5.1, both for Windows (SPSS Inc., Chicago, Illinois).
Demographics and SPECT results of 3VD versus control patients
Demographic characteristics of 3VD and control groups are shown in Table 1. Compared with the control group, patients in the 3VD group were older (64 ± 10 vs. 57 ± 12, p < 0.001), more frequently male (71% vs. 57%, p < 0.05), and more likely to undergo pharmacologic stress testing (53% vs. 38%, p < 0.05). There were no differences in history of diabetes, prior myocardial infarction, angina score, heart failure score, stress-induced chest pain, ischemic ST-segment depression, or exercise workload between the two groups.
On SPECT MPI, compared with the control group, patients in the 3VD group had more reversibility (53% vs. 38%, p < 0.05); greater prevalence of fixed left ventricular enlargement (37% vs. 21%, p < 0.01); greater frequency of transient ischemic dilation of the left ventricle (20% vs. 11%, p = 0.05); more abnormal segments per patient by perfusion analysis alone (4.1 ± 2.8 vs. 2.5 ± 2.5, p < 0.001); more abnormal segments by combined perfusion/function analysis (6.2 ± 4.7 vs. 2.8 ± 3.6, p < 0.001); and lower post-stress left ventricular ejection fraction (LVEF) (45 ± 12% vs. 53 ± 10%, p < 0.001).
Multivariate analysis of six clinical, three stress, four perfusion, and two combined perfusion/function variables in the whole cohort (3VD and non-3VD groups) identified four variables as independent predictors of the presence of angiographic 3VD. These were age (p < 0.0001), male gender (p < 0.05), combined perfusion/function analysis (p < 0.0001), and post-stress LVEF (p < 0.05) (Table 2).
Combined perfusion/function versus perfusion analysis
In patients with 3VD, compared with perfusion alone, combined perfusion/function analysis resulted in significantly greater number of abnormal segments per patient compared with perfusion alone (6.2 ± 4.7 vs. 4.1 ± 2.8, p < 0.001). Interestingly, in 3VD patients, pharmacological stress resulted in more segments with combined perfusion/function abnormalities than exercise stress (7.2 ± 4.5 vs. 5.1 ± 4.7, p < 0.001) despite similar number of segments with abnormal perfusion (4.4 ± 3.7 vs. 3.8 ± 2.9, p = NS). In contrast, in the control group, there were no differences in number of abnormal segments per patient with abnormal perfusion/function compared with perfusion (2.8 ± 3.6 vs. 2.5 ± 2.5, p = NS).
Figure 1shows the distribution of number of abnormal vascular territories by perfusion versus combined perfusion/function in patients with 3VD. The overall sensitivity for identifying any SPECT abnormality of the combined perfusion/function assessment was higher than that of perfusion imaging alone (88% vs. 83%, p = NS). Importantly, the prevalence of multivessel defect pattern was significantly greater in the combined perfusion/function than perfusion analysis (60% vs. 46%, p < 0.05) due to higher detection of 3VD pattern (25% vs. 10%, p < 0.001) (Fig. 2).
In control patients with 2VD and 1VD, the overall sensitivities of combined perfusion/function versus perfusion were identical, 92% versus 92% for 2VD, and 86% vs. 86% for 1VD. Similarly, in this cohort, there were no differences in prevalence of multivessel defect pattern between combined perfusion/function and perfusion alone (42% vs. 38% for 2VD and 28% vs. 21% for 1VD, p = NS for both comparisons).
The specificity of Gated SPECT MPI was not adversely affected by combined perfusion/function versus perfusion (72% vs. 67%, p = NS) among all patients in the control group. When the definition of no angiographic CAD was made less stringent (from <70% to <50% stenoses in all three vessels), the specificity of gated SPECT imaging improved, but remained similar between combined perfusion/function and perfusion (79% vs. 79%, p = NS).
Figure 3shows that the addition of perfusion variables to clinical and stress variables resulted in a significant increase in the global chi-square in the prediction of 3VD (89.8 vs. 51.0, p < 0.00001). Subsequent addition of functional analysis by gated MPI to the combined model of clinical, stress, and perfusion data yielded a further significant increase in the global chi-square (131.7 vs. 89.8, p < 0.00001).
Table 3shows that the odds that combined perfusion/function analysis versus perfusion alone correctly matched the number of angiographically diseased vessels was 8.8 times higher in 3VD than in control patients. Interestingly, when the 3VD group is compared individually with 0VD, 1VD, or 2 VD patients, combined perfusion/function assessment has the greatest added value over perfusion in 1VD patients (6.2, 14.7, and 6.2 times higher, respectively).
The present study assessed the incremental diagnostic value of segmental post-stress thickening abnormalities on gated sestamibi SPECT MPI to better identify patients with severe angiographic 3VD. Adding function to perfusion yielded significantly greater detection of multivessel defects (60% vs. 46%, p < 0.05) and 3VD pattern (25% vs. 10%, p < 0.001) compared with perfusion imaging alone in patients with severe 3VD. This was achieved without a loss in specificity in the control (72% vs. 69%, p = NS). Age and combined perfusion/functional analysis were the strongest predictors of angiographic 3VD. Addition of functional analysis to the combined model of clinical, stress, and perfusion data yielded a further significant increase in predictive value for 3VD (global chi square 131.7 vs. 89.8, p < 0.00001). Combining functional assessment with perfusion was 8.8 times more likely than perfusion alone to yield a scintigraphic pattern identical to the angiographic extent of disease in patients with 3VD compared with patients without 3VD.
Although the sensitivity for the detection of CAD in patients with 3VD or left main disease is high (3,4), perfusion defects are often seen in only one vascular territory (5,6). In patients undergoing exercise stress, variables such as magnitude of the ST depression, rate-pressure product, or hypotensive blood pressure response may assist in identification of these high-risk patients (5).
Johnson et al. (13)have previously highlighted the value of global post-stress functional analysis by demonstrating that a significant decrease in post-stress LVEF compared with the rest LVEF correlates with more extensive ischemic burden. Similarly, Yamagishi et al. (14)have shown that post-exercise worsening of LVEF is more likely to detect multivessel CAD than reversible perfusion defects alone.
Segmental analysis of post-stress images for a regional dysfunction could also be useful to enhance the detection rate of high-risk CAD patients. Sharir et al. (10)demonstrated the incremental value of combined wall motion/perfusion over perfusion alone in identifying severe disease. The additional information was more relevant in the presence of multivessel disease or proximal left anterior descending coronary artery. Emmet et al. (15)also showed that the reversible wall motion abnormalities correlated quite well with the angiographic severity of the coronary lesions.
The results of the present study support our hypothesis that, in the presence of three severe coronary lesions, when tracer uptake may be reduced in all vascular territories, detection of segmental wall motion abnormalities on post-stress images could be extremely useful for detecting additional stenotic lesions. We found that 3VD patients were almost 14.7 times more likely than 1VD patients, 6.8 times more likely than 2VD, and 6.2 times more likely than 0VD patients to derive incremental diagnostic value from combined function and perfusion than perfusion assessment alone.
Previous studies were not specifically designed to study the population with severe 3VD. Indeed, the percentage of patients with 3VD were 30% of the total patients in the study by Sharir et al. (10), 28% in the report by Emmett et al. (15), and 17% in the study by Yamagishi et al. (14). Thus, the number of patients with 3VD was too small to test the hypothesis that functional data regarding abnormal wall motion or thickening provided supplementary diagnostic value to perfusion data in patients who might have balanced ischemia.
An important aspect of our data is the demonstration that the incremental diagnostic value of gated SPECT imaging over perfusion alone did not result in loss of specificity for detection of CAD in lower risk patients. In fact, the combined perfusion and function analysis resulted in a slight increase in specificity in the control group (72% vs. 67%, p = NS). These specificity values are lower than those reported in the literature (4–6,9), which likely reflects our more stringent criteria for “no significant” CAD (<70%) in our 0VD patients, compared with the usual <50% stenosis used for most studies. When the specificity of gated SPECT imaging was based on <50%, our specificity increased to 79% by both gated SPECT interpretations.
Potential mechanisms for wall motion abnormalities in patients with 3VD
Wall motion in this study was determined about 1 h post-stress. Wall motion abnormality at this time could be the result of prolonged post-stress stunning. However, in our experience, true post-stress stunning is infrequent in stable patients. It is likely that the majority of wall motion abnormalities we detected were the result of subendocardial scar or of chronic repetitive myocardial stunning or hibernation in a heart where every vascular region was compromised. We cannot conclude from our data the mechanism of wall motion abnormality. However, we can conclude that observation of wall motion abnormality is an important indicator of the extent of CAD and may be the only clue when diffuse 3VD has caused diffuse myocardial flow limitations that will not be shown by regional relative myocardial uptake measurements.
Pharmacologic stress is less likely than exercise to cause frank ischemia and, therefore, regional wall motion abnormalities. Surprisingly, we found the combined perfusion/functional reading to identify more myocardial segments as abnormal than perfusion alone, when stress was achieved by pharmacologic means than by exercise. Recent studies (16,17)suggest that wall motion abnormalities induced by dipyridamole occur commonly in flow-limiting stenoses. In addition to the well-described mechanism of blood flow diversion from collaterals by vasodilation (horizontal steal), “steal” of blood from the endocardium to the epicardium (vertical steal) might result in regional ischemia and subsequent stunning.
It is also possible that patients who undergo dipyridamole stress are simply a selected group with more extensive CAD associated with greater wall motion abnormalities than patients undergoing exercise stress.
Although angiography is frequently used as the gold standard for validation of noninvasive testing, its limitations are well known. Indeed, even in the presence of angiographic CAD, a normal SPECT scan portends a good prognosis. We did not specifically account for presence of collateral circulation, which could have contributed to some discrepancy between angiography and SPECT MPI. However, we believe our stringent criterion of >70% stenosis resulted in severe flow-limiting lesions (mean and median percent stenosis were 88 ± 11% and 92%, respectively). Furthermore, angiographic disease burden remains the predominant means for ultimate clinical decision-making regarding revascularization.
Combined assessment of global and regional ventricular function with perfusion on stress Tc-99m sestamibi SPECT MPI enhances the sensitivity for detection of high-risk variables such as defects in multiple vascular territories or reduced LVEF in patients with severe 3VD. The number of abnormal vascular territories by combined perfusion and function analysis has strong incremental predictive value for the presence of 3VD when compared with other clinical, stress, and scintigraphic variables, without adversely affecting the specificity of disease detection.
- coronary artery disease
- gated SPECT
- electrocardiographic gating Tc-99m sestamibi single photon emission computed tomography
- left ventricular ejection fraction
- myocardial perfusion imaging
- single photon emission computerized tomography
- thickening fractions
- zero-vessel coronary artery disease
- one-vessel coronary artery disease
- two-vessel coronary artery disease
- three-vessel coronary artery disease
- Received November 15, 2002.
- Revision received March 23, 2003.
- Accepted April 10, 2003.
- American College of Cardiology Foundation
- CASS Principal Investigators and Associates
- Mahmarian J.J.,
- Boyce T.M.,
- Goldberg R.K.,
- Cocanougher M.K.,
- Roberts R.,
- Verani M.S.
- DePasquale E.E.,
- Nody A.C.,
- DePuey E.G.,
- et al.
- Nygaard T.W.,
- Gibson R.S.,
- Ryan J.M.,
- Gascho J.A.,
- Watson D.D.,
- Beller G.A.
- Smanio P.E.,
- Watson D.D.,
- Segalla D.L.,
- Vinson E.L.,
- Smith W.H.,
- Beller G.A.
- Sharir T.,
- Germano G.,
- Kavanagh P.B.,
- et al.
- Taillefer R.,
- DePuey E.G.,
- Udelson J.E.,
- Beller G.A.,
- Latour Y.,
- Reeves F.
- Calnon D.A.,
- Kastner R.J.,
- Smith W.H.,
- Segalla D.,
- Beller G.A.,
- Watson D.D.
- Smith W.H.,
- Kastner R.J.,
- Calnon D.A.,
- Segalla D.,
- Beller G.A.,
- Watson D.D.
- Johnson L.L.,
- Verdesca S.A.,
- Aude W.Y.,
- et al.
- Yamagishi H.,
- Shirai N.,
- Yoshiyama M.,
- et al.
- Emmett L.,
- Iwanochko R.M.,
- Freeman M.R.,
- Barolet A.,
- Lee D.S.,
- Husain M.
- Bin J.P.,
- Le E.,
- Pelberg R.A.,
- Coggins M.P.,
- Wei K.,
- Kaul S.
- Lima RL, Hatada K, Riou LM, et al. Dipyridamole stress-induced functional abnormalities can detect the presence of ischemia even in the absence of a myocardial perfusion defect on images (abstr). J Nucl Cardiol 2003;10:54