|
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |
|
|
|
|||||||||
|
|||||||||||
|
J Am Coll Cardiol, 2003; 42:1818-1825, doi:10.1016/j.jacc.2003.07.010 © 2003 by the American College of Cardiology Foundation |
* Department of Imaging (Division of Nuclear Medicine) and Department of Medicine (Division of Cardiology), Cedars-Sinai Medical Center, Los Angeles, California, USA
Leiden University Medical Center, Leiden, The Netherlands
Keck School of Medicine, University of Southern California, Los Angeles, California, USA
Manuscript received April 8, 2003; revised manuscript received July 14, 2003, accepted July 17, 2003.
* Reprint requests and correspondence: Dr. Daniel S. Berman, Department of Imaging, Cedars-Sinai Medical Center, 8700 Beverly Boulevard, Taper Building, A1258, Los Angeles, California 90048, USA.
bermand{at}cshs.org
 |
Abstract |
|---|
 Top Abstract MethodsResultsClinical characteristics of...Risk of TE by...Prediction of TE by...Prediction of TE in...Angiographic correlations in...DiscussionReferences |
|---|
BACKGROUND: The prognostic value of TID in patients with an otherwise normal MPS has not been defined.
METHODS: We identified 1,560 patients who had normal stress MPS (436 vasodilator and 1,124 exercise stress), and no rest LV enlargement (Population 1) and followed up for 2.30 ± 0.67 years for hard events (HE) (cardiac death or myocardial infarction) and soft events (SE) (revascularization). Prediction of first HE or SE (total events [TE]) was evaluated by multivariable Cox analysis, which was also applied to a broader group of 2,037 patients (including patients with minimal defects (Population 2).
RESULTS: In Population 1, there were 13 HE, 36 SE, and 42 TE. Patients in the highest TID quartile (TID 
1.21) had a higher TE rate than others, regardless of stress type. By multivariable analysis, highest TID quartile was predictive of TE (p = 0.008). Other independent predictors of TE were age, typical angina, and diabetes. In Population 2, TID was also predictive of TE.
CONCLUSIONS: An entirely normal stress MPS study does not always imply an excellent prognosis. In patients with otherwise normal MPS, TID is an independent and incremental prognostic marker of TE even after significant clinical variables—age, typical angina, and diabetes—are accounted for. When TID is present, caution in making low-risk prognostic statements may be warranted, especially in patients with typical angina, the elderly, and diabetics. Our findings also appear to apply to the broader population of "normal" MPS, which included patients with minimal perfusion defects.
| ||||||||||||||||||||||||||||||||
Our study was performed to evaluate the hypothesis that TID may have independent prognostic value in patients with normal MPS. To selectively evaluate the prognostic impact of TID, we studied patients with otherwise "perfectly" normal MPS and normal LV cavity size at rest; we also evaluated other independent predictors of increased cardiac risk in patients with "perfectly" normal MPS. Further, we assessed whether the findings regarding TID also applied to patients with the standard definition of normal MPS.
|
Methods |
|---|
|
TopAbstractMethodsResultsClinical characteristics of...Risk of TE by...Prediction of TE by...Prediction of TE in...Angiographic correlations in...DiscussionReferences |
|---|
We also evaluated a broader group of 2,037 patients in whom complete follow-up information was available (Population 2) (48.7% men; mean age 63.2 ± 12.8 years; 618 patients had vasodilator stress [556 adenosine, 62 dipyridamole] and 1,419 had exercise) defined as having normal MPS using our standard definition of normal scan, which includes 477 patients with minimal defects, considered too small to be called abnormal (1–7).
Imaging procedure. All patients underwent rest 201Tl/stress 99mTc-sestamibi MPS as previously described (19). Whenever possible, beta-blockers and calcium channel antagonists were terminated 48 h before testing and nitrates at least 6 h before testing. Patients performed a symptom-limited exercise treadmill test or vasodilator (dipyridamole or adenosine) stress using standard protocols (20,21). Patients were instructed not to consume coffee or other products containing caffeine for 24 h before the test. During both types of stress, heart rate, blood pressure, and a 12-lead electrocardiogram (ECG) were recorded at baseline and every minute thereafter for at least 5 min. The ECG was monitored continuously (leads aVF, V1, and V5) for development of arrhythmia or ischemic ST-segment deviation (22). Blood pressure was measured and recorded at rest, at the end of each stress stage, and at peak stress.
Acquisition protocol. All patients underwent separate acquisition, dual isotope MPS (23). Myocardial perfusion SPECT was started 10 min after 201Tl injection and 15 to 60 min after 99mTc-sestamibi injection. Myocardial perfusion SPECT employed circular or elliptical 180° acquisition for 64 projections at 20 to 25 s/projection for 201Tl and at 15 to 25 s/projection for 99mTc-sestamibi. All images were subject to quality control measures. The projection data were reconstructed into tomographic transaxial images using filtered back projection and automatic reorientation (24). No attenuation or scatter correction was used.
Image interpretation. Semiquantitative visual interpretation was performed using the 20-segment model (1,19). Each segment was scored using a 5-point scoring system (0 = normal, 1 = equivocal, 2 = moderate, 3 = severe reduction of uptake, and 4 = absence of detectable tracer uptake). Summed stress scores (SSS) were obtained by summing the individual stress scores of the 20 segments. Only patients with MPS interpreted as normal with SSS = 0 and no visual evidence of resting LV enlargement were included in Population 1. Population 2 consisted of all patients with follow-up information who had a normal scan defined as SSS = 0 to 3 (13).
Measurement of TID. For calculation of TID, we used a commercially available automated program (QPS, Cedars-Sinai, Los Angeles, California), which estimates three-dimensional image volumes from gated or ungated SPECT studies (25). The software is completely automatic but does allow manual operator interaction if needed. The algorithm operates in the three-dimensional space and uses the stress and rest short-axis image sets. After calculation of the endocardial volumes (bounded by the endocardial surface and the valve plane), it derives the TID ratio as the ratio of LV volumes at stress and rest (16). For purposes of this study, the TID ratio calculated from the ungated SPECT studies was employed.
Patient follow-up. Patient follow-up (all >1 year) was performed as previously described (1) by individuals blinded to the patient's test results. All patients were followed up for HE (CD or nonfatal MI) (26,27); soft events (SE), defined by revascularization after an index MPS; and total events (TE), defined by the first HE or SE. Any revascularization after MPS was used rather than our usual late revascularization occurring >60 days (28,29) after MPS because, when normal, the results of the scan would not be expected to increase the rate of early coronary angiography. Cardiac death was defined as death due to any cardiovascular cause (26); MI was documented by a consistent history accompanied by elevation of cardiac enzymes and/or new Q waves on the ECG.
Statistical analysis. Comparisons between patient groups were performed using a t test for continuous variables or analysis of variance and chi-square for categorical variables. All continuous variables were described as a mean ± SD. A p value <0.05 was considered significant.
Due to few observed HE, we studied the prediction of TE using multivariable Cox proportional hazards analysis. Selection of variables for consideration for entry was based on univariable statistical significance and clinical judgment (21). A significant increase in global chi-square of a model after addition of a variable indicated incremental prognostic value. Variables in the final model were tested for confounding, interactions, and linearity.
|
Results |
|---|
|
TopAbstractMethodsResultsClinical characteristics of...Risk of TE by...Prediction of TE by...Prediction of TE in...Angiographic correlations in...DiscussionReferences |
|---|
In Population 2, during follow-up of 2.26 ± 0.69 years, there were 19 HE (10 CD and 9 nonfatal MI) and 52 SE; 61 patients (3.0%) had TE. Of the 52 SE, 7 (13.5%) occurred <60 days after SPECT. The annualized event rate for Population 2 was as low as for Population 1: 0.2%/year for CD, 0.2%/year for MI, 0.4%/year for HE, and 1.3%/year for TE. Hard events and TE rates were higher in the Population 2 with vasodilator than with exercise stress: (HE and TE rates 0.9% and 2.0%/year for vasodilator stress vs. 0.2% and 1.0%/year for exercise stress, respectively, p < 0.01 for both). When only the 477 patients with SSS = 1 to 3 were considered, the annualized event rates were 0.6%/year for HE and 1.4%/year for TE. These rates were not different from those observed in the patients with SSS = 0 (Population 1).
Prediction of events in Population 1: (SSS = 0). cardiac risk of patients with different values of TID ratio.
Population 1 was separated into quartiles based on the TID ratio (Table 1). Patients in the highest quartile had a TID ratio that was similar in patients with the exercise and vasodilator stress (1.36 ± 0.16 vs. 1.35 ± 0.13, p = 0.505) and was well above the previously validated normal TID limit of 1.22 (16). The risks for both TE and HE were higher in the highest TID ratio quartile (Fig. 1); patients in the three lower quartiles all had the same low risk of future cardiac events. Accordingly, the patients in the three lower TID ratio quartiles were combined as a control group for comparison with the patients in the highest TID ratio quartile. Patients in the control group had a low risk of TE (
1%/year both in exercise and vasodilator stress subgroups); however, patients in the highest TID quartile had an intermediate risk of TE with exercise (1.8%/year), and high risk of TE with vasodilator stress (3.2%/year) (Fig. 2).
|
|
|
|
Clinical characteristics of patients in the highest TID quartile versus controls |
|---|
|
TopAbstractMethodsResultsClinical characteristics of...Risk of TE by...Prediction of TE by...Prediction of TE in...Angiographic correlations in...DiscussionReferences |
|---|
|
|
Risk of TE by univariable analysis |
|---|
|
TopAbstractMethodsResultsClinical characteristics of...Risk of TE by...Prediction of TE by...Prediction of TE in...Angiographic correlations in...DiscussionReferences |
|---|
|
|
Prediction of TE by multivariable analysis |
|---|
|
TopAbstractMethodsResultsClinical characteristics of...Risk of TE by...Prediction of TE by...Prediction of TE in...Angiographic correlations in...DiscussionReferences |
|---|
|
|
|
Prediction of TE in Population 2: (SSS = 0 to 3) |
|---|
|
TopAbstractMethodsResultsClinical characteristics of...Risk of TE by...Prediction of TE by...Prediction of TE in...Angiographic correlations in...DiscussionReferences |
|---|
|
Angiographic correlations in Population 1 |
|---|
|
TopAbstractMethodsResultsClinical characteristics of...Risk of TE by...Prediction of TE by...Prediction of TE in...Angiographic correlations in...DiscussionReferences |
|---|
70% luminal stenosis) in at least one of the three major coronary arteries than patients in the control group (12/20 [60%] vs. 11/33 [33%], p = 0.058); only patients with TID had severe and extensive CAD as defined by >90% stenosis of the proximal LAD or of multiple vessels (16) (5/20 [25%] vs. 0/33 [0%], p = 0.003). |
Discussion |
|---|
|
TopAbstractMethodsResultsClinical characteristics of...Risk of TE by...Prediction of TE by...Prediction of TE in...Angiographic correlations in...DiscussionReferences |
|---|
Comparison with previous studies. Patients with normal MPS have a very low risk of HE (usually <1%/year) (6,10,11,27); nonetheless, a small proportion of patients with normal MPS do develop future cardiac events. In a recent report, we found that age, gender, diabetes, type of stress, and a history of known CAD are multivariable predictors of cardiac events in this low-risk population (12). We also reported significant interactions between stress type and previous CAD (lower risk in patients without previous CAD undergoing exercise stress vs. all others) and between diabetes and gender (higher risk in diabetic females) (12), but we did not investigate TID in this previous report. Furthermore, most previous prognostic reports of patients with normal MPS have not distinguished those with minimal perfusion abnormalities or LV enlargement from those with "perfectly" normal scans. In this study we examined the prognostic value of TID and other variables in a large population of patients with no perfusion abnormality on dual isotope MPS (SSS = 0) and no resting LV enlargement as well as in the broader group of patients with normal MPS (SSS = 1 to 3). Previous work from our group has indicated that automatically measured dual isotope TID using QPS software is sensitive and highly specific for detection of severe and extensive CAD (16). Peace et al. (28) found that TID in Tc-99m tetrofosmin MPS, measured by six different algorithms, is an indicator of severe CAD, and all existing algorithms provided a repeatable, quantitative measure of the TID. These measures have not been previously assessed for prognostic value in patients with normal MPS.
Which patients with normal MPS are at increased risk?. We have previously shown that increased age, noncardiac comorbidities (in particular diabetes), prior CAD, and vasodilator stress are associated with a higher event rate in patients with normal MPS (12,21,29), suggesting that underlying clinical conditions influence their cardiac prognosis after a normal perfusion study (3,17).
A new finding in our study is the incremental prognostic value of an abnormal TID ratio in patients with normal perfusion in a multivariable model including all other significant markers. Clinical characteristics that were predictive in patients with entirely normal scans (SSS = 0, normal LV cavity at rest) included age, typical angina, and diabetes. Other clinical variables previously associated with an increased cardiac event risk (21) did not emerge in the final prognostic model; in patients with entirely normal MPS, prior MI and history of revascularization or the composite variable, history of known CAD, were not predictive of future cardiac events. These findings are likely to be related to characteristics of the present study population and to the smaller number of events in the current population than in our previous work (12).
In the broader population of patients with normal scans defined by our standard criteria (SSS = 0 to 3), follow-up TE and HE rates were as low as in patients in the main study population (SSS = 0); TID in patients with SSS = 0 to 3 also was found to be predictive of TE by both univariable and multivariable analysis. Of interest, the HE rate was as low in this group as in the group with the entirely normal scans. When the small group of patients with SSS = 1 to 3 was separately analyzed, there were no differences noted in HE and TE rates compared with the patients with SSS = 0.
Possible mechanisms of prognostic impact of the TID in patients with normal MPS. Transient ischemic dilation of the LV and increased lung-heart ratio (LHR) are both associated with severe CAD (13–16,30–32). Increased LHR with 201Tl has been shown to be valuable in evaluation of prognosis in patients with myocardial perfusion abnormalities (33,34), but does not appear to risk-stratify patients with normal MPS (33,35). A number of clinical studies have demonstrated in multiple settings that the presence of TID on MPS predicts left main or severe multivessel CAD (13,14,36–40) or adverse cardiac outcome (18,41). The underlying mechanisms of TID are likely to be varied, with some pathologic and others physiologic. Probably the most common pathologic mechanism is "apparent" (as opposed to "true") TID caused by nonvisualization of the extensive amount of the subendocardial myocardium after stress in the presence of severe stress-induced ischemia/hypoperfusion (15,37,38,42). The measurement of TID employed in this manuscript relies on the automatic definition of endocardial boundaries of the LV. The presence of stress-induced subendocardial ischemia can result in an apparent thinning of the myocardium and a consequent overestimation of true LV cavity size. Because the subendocardium in this setting would generally have resting perfusion that is either normal or not as hypoperfused as during stress, the measured resting LV volume would appear smaller than the post-stress volume, even if the true volumes were the same.
Another pathologic mechanism is true TID in which the LV dilates during stress and remains dilated through the post-stress MPS imaging (43,44). This mechanism would imply the presence of stress-induced stunning of enough of the LV to cause overall enlargement. This stress-induced stunning is a well-described phenomenon, associated with critical coronary stenosis (45). Evidence of the presence of true transient dilation in many of the patients with TID was provided in the initial description of the abnormal TID ratio in which the TID defined by the epicardial edges of the LV on planar 201Tl images was predictive of severe and extensive CAD (13). The presence of this mechanism was further supported by the findings in the preliminary report of the automatically measured SPECT TID ratio, not included in the final manuscript (16). When this ratio was derived from the epicardial edges of the LV, it was predictive of severe and extensive CAD, although less so than that derived from the endocardial edges (16).
Regarding the finding of an abnormal TID ratio in patients with otherwise normal MPS scans, either of these mechanisms would be expected to result in increased risk of events. In the case of subendocardial ischemia, the ischemia would most likely be diffuse and extensive in order to explain the absence of an apparent relative perfusion defect. Similarly, true transient dilation persisting through the post-stress scanning time would be expected to be associated with severe and diffuse stress-induced ischemia. The current findings demonstrate that abnormal automatically measured TID ratio has prognostic value in patients with a normal perfusion and should, thus, be considered in the prognostic assessment of these patients. Both of these mechanisms would imply that an increased TID ratio could be related to diffuse, "balanced ischemia" due to severe CAD (46).
Thus, our results showing increased event rates in patients with normal MPS and TID may be explained by severe underlying CAD in these patients, missed by perfusion defect analysis alone. The limited data from coronary angiography in our Population 1 support this postulation. Obstructive CAD was more common in the patients with TID than in the controls (p = 0.058), and severe and extensive CAD was found only in patients with TID.
Recent published data (47) demonstrated that, in patients with classical syndrome X (with normal MPS in the majority), magnetic resonance spectroscopy revealed subendocardial hypoperfusion following intravenous administration of adenosine, implying another possible pathologic mechanism for apparent TID without extensive coronary atherosclerosis. However, it is likely that in many patients TID is simply physiologic, occurring as a variant of normal for reasons that are not yet understood.
Study limitations. The present study used different radionuclides to produce the rest and stress images adding complexity to the comparison of stress and rest volumes; as previously shown, however, the automatic algorithm used in our study for measuring the TID ratio has been shown to be effective in identifying severe and extensive CAD in dual-isotope studies (16). If applied to single radionuclide studies, the cut-off for an abnormal TID ratio might be different, but the mechanisms likely to be operative in this study should still apply. Although it has been shown that automated TID calculations using different software methods are similar (28), the data presented in this paper apply to the TID ratio derived using QPS software; further studies would be required to determine if these results apply to other software programs. Due to the small number of HE in the study population, the multivariable prognostic assessment analysis was limited to the prediction of TE. The current results are based on a population referred for nuclear testing and, therefore, may not be applicable to a broader population; however, the patients in this study are typical of those referred to a university-affiliated community hospital in a major urban area, and the results of this study should be applicable to this setting. This study is retrospective in nature. Although all data were collected and entered prospectively, further confirmation of our findings in other populations is needed. Finally, the study is based on data of a single nuclear cardiology center with unique technical characteristics, which may not be applicable in other nuclear laboratories.
Clinical implications. In patients with elevated TID ratio and otherwise normal MPS, the total event rates are increased, but only to intermediate levels. Given this, we do not believe that patients need to be considered as candidates for coronary angiography on the basis of an increased TID alone. However, when integrated with all other available information (e.g., typical angina, advanced age, and diabetes), the finding of increased TID may be useful in selecting of patients at sufficiently increased risk that coronary angiography or further noninvasive testing (e.g., noninvasive coronary angiography) would be appropriate.
Conclusions. An entirely normal stress MPS study does not always imply an excellent prognosis. In patients with otherwise normal MPS, TID is an independent and incremental prognostic marker of TE even after significant clinical variables—age, typical angina, and diabetes—are accounted for. When TID is present, caution in making low-risk prognostic statements may be warranted, especially in patients with typical angina, the elderly, and diabetics. Our findings also appear to apply to the broader population of "normal" MPS, which included patients with minimal perfusion defects.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractMethodsResultsClinical characteristics of...Risk of TE by...Prediction of TE by...Prediction of TE in...Angiographic correlations in...DiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. D. Metz, M. Beattie, R. Hom, R. F. Redberg, D. Grady, and K. E. Fleischmann The Prognostic Value of Normal Exercise Myocardial Perfusion Imaging and Exercise Echocardiography: A Meta-Analysis J. Am. Coll. Cardiol., January 16, 2007; 49(2): 227 - 237. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. S. Berman, R. Hachamovitch, L. J. Shaw, J. D. Friedman, S. W. Hayes, L. E.J. Thomson, D. S. Fieno, G. Germano, N. D. Wong, X. Kang, et al. Roles of Nuclear Cardiology, Cardiac Computed Tomography, and Cardiac Magnetic Resonance: Noninvasive Risk Stratification and a Conceptual Framework for the Selection of Noninvasive Imaging Tests in Patients with Known or Suspected Coronary Artery Disease J. Nucl. Med., July 1, 2006; 47(7): 1107 - 1118. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. K. Min, K. A. Williams, T. M. Okwuosa, G. W. Bell, M. S. Panutich, and R. P. Ward Prediction of Coronary Heart Disease by Erectile Dysfunction in Men Referred for Nuclear Stress Testing Arch Intern Med, January 23, 2006; 166(2): 201 - 206. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. F. Heston and D. M. Sigg Quantifying Transient Ischemic Dilation Using Gated SPECT J. Nucl. Med., December 1, 2005; 46(12): 1990 - 1996. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Emmett, M. Magee, S. B. Freedman, H. Van der Wall, V. Bush, J. Trieu, W. Van Gaal, K. C. Allman, and L. Kritharides The Role of Left Ventricular Hypertrophy and Diabetes in the Presence of Transient Ischemic Dilation of the Left Ventricle on Myocardial Perfusion SPECT Images J. Nucl. Med., October 1, 2005; 46(10): 1596 - 1601. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Abidov, J. J. Bax, S. W. Hayes, I. Cohen, H. Nishina, S. Yoda, X. Kang, F. Aboul-Enein, J. Gerlach, J. D. Friedman, et al. Integration of Automatically Measured Transient Ischemic Dilation Ratio into Interpretation of Adenosine Stress Myocardial Perfusion SPECT for Detection of Severe and Extensive CAD J. Nucl. Med., December 1, 2004; 45(12): 1999 - 2007. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J. Gibbons and P. A. Araoz The year in cardiac imaging J. Am. Coll. Cardiol., November 16, 2004; 44(10): 1937 - 1944. [Full Text] [PDF]
|
|
|
|
|
|
C. J. Hardebeck Transient ischemic dilation J. Am. Coll. Cardiol., July 7, 2004; 44(1): 211 - 212. [Full Text] [PDF]
|
|
|
|
|
|
D. S. Berman, G. Germano, and A. Abidov Reply J. Am. Coll. Cardiol., July 7, 2004; 44(1): 212 - 213. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
