| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Clinical Investigation |
1 The Weatherhead PET Center For Preventing and Reversing Atherosclerosis, University of Texas Medical School at Houston, Houston, Texas; 2 Division of Cardiology, Department of Medicine, University of Texas Medical School at Houston, Houston, Texas; and 3 Memorial Hermann Hospital, Houston, Texas
Correspondence: For correspondence or reprints contact: K. Lance Gould, MD, The Weatherhead PET Center, University of Texas Medical School at Houston, 6431 Fannin St., Room 4.256 MSB, Houston, TX 77030. E-mail: gould{at}pet.med.uth.tmc.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONReferences |
|---|
Key Words: coronary artery disease • PET • risk factors • myocardial perfusion • progression–regression CAD
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONReferences |
|---|
Similarly, regression of coronary atherosclerosis is characterized by modestly decreased stenosis severity and absence of new lesions in association with reduced cardiac events proportionately greater than the modest reduction in stenosis severity (3,4,10–13). The essential concept is that worsening of a flow-limiting stenosis over time most likely reflects the mechanism of gradual accumulation of lipids, whereas the appearance of new flow-limiting stenosis not present at baseline most likely reflects the mechanism of plaque rupture. Because active, progressive coronary atherosclerosis is multicentric (14–17), assessing changes in the entire coronary artery tree is essential for identifying and quantifying progression or regression of coronary artery disease (CAD), its response to risk factor treatment, and for predicting risk of clinical events.
We have previously reported that combined intense lifestyle and pharmacologic lipid treatment to specific risk factor goals improves myocardial perfusion and markedly reduces coronary events, more than either lifestyle or pharmacologic treatment alone over 5-y follow-up (18). However, despite the severity of the baseline perfusion defect being a good predictor of clinical events, the changes in the most severe perfusion defect from baseline–to–follow-up PET over 2.6 ± 1.4 y were poorly predictive of coronary events over the next 5 y, a surprising finding.
Therefore, to explain this unexpected finding based on the literature above, we hypothesized that changes in myocardial perfusion throughout the entire coronary vascular tree in regions other than the baseline worst perfusion abnormalities by sequential dipyridamole PET perfusion imaging would (i) predict clinical outcomes at long-term follow-up better than the changes of the most severe perfusion defect due to the single worst flow-limiting coronary stenosis at baseline, (ii) correlate with intensity of risk factor treatment, (iii) identify new flow-limiting stenosis in new regions without a significant perfusion defect at baseline, and (iv) demonstrate worsening or improvement of the worst baseline flow-limiting stenosis consistent with a mechanism of gradual lipid accumulation or removal depending on risk factor treatment.
In principle, precise myocardial perfusion mapping should noninvasively provide objective quantification of the net balance of mixed progression or regression of diffuse disease and localized stenosis for each coronary artery and for the entire heart that is not possible with coronary arteriography or intracoronary ultrasounds. For testing these new hypotheses, we developed automated software for objectively quantifying PET perfusion defects for changes in the same region as the most severe perfusion defect at baseline or changes in other baseline–to–follow-up paired regions that are not the same as the region with the most severe defect at baseline.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONReferences |
|---|
At the follow-up PET study, patients were categorized by blinded observers using predefined objective criteria into 3 groups based on the intensity of medical therapy during the interval between the 2 PET studies. As previously described in detail (18), patients in the poor treatment group were not on a diet or a lipid active medication or were actively smoking. Patients in the moderate therapy group were instructed in an American Heart Association diet and took lipid active drugs not dosed to specific goals or adhered to a very low fat diet with 
10% of calories as fat without lipid-lowering drugs. Patients in the maximal treatment group adhered to a strict diet (<10% fat calories), regular exercise, and lipid active medications dosed to target low-density lipoproteins (LDLs) <90 mg/dL, high-density lipoproteins (HDLs) >40 mg/dL, and triglycerides <100 mg/dL.
Patients with only 12 mo or less between baseline and follow-up PET, defined as inadequate for predicting long-term events, were analyzed as a separate group for PET endpoints. Accordingly, the minimum treatment period was 1 y. Patients with percutaneous coronary intervention (PCI) or coronary artery bypass graft surgery (CABG) in the interval between the 2 PET studies were considered separately as an internal comparison because the intervention would alter myocardial perfusion separately from risk factor treatment. Follow-up for clinical outcome was obtained at a mean interval of 5 y after the final PET study for death, nonfatal myocardial infarction, PCI, CABG, or stroke.
We did not exclude patients with resting defects for several reasons: (i) such selection might introduce a possible bias; (ii) even if the defect were a fixed scar, the changes in the rest of the heart provide important data on progression–regression in borderzone regions or new stenosis in other myocardial regions; and (iii) the patients with no improvement due to a fixed scar despite optimal, intense medical therapy would reduce the changes we observed and be against our conclusions that remained significant despite their inclusion. Therefore, including patients with resting defects is the most conservative statistically correct approach but patients with significant defects on the resting perfusion scan were also analyzed separately. Patients with severe liver or renal dysfunction or with coronary revascularization within 6 mo from the baseline PET study were excluded.
PET
PET was performed using the University of Texas–designed, Posicam, bismuth germanate multislice tomograph with a reconstructed resolution of 10-mm full width at half maximum as previously described (18–24). Transmission images to correct for photon attenuation contained 100–150 million counts. Emission images obtained after intravenous injection of 666 MBq (18 mCi) of cyclotron-produced 13N contained 20–40 million counts.
At 40 min after administration of the first dose of ammonia, dipyridamole (0.142 mg/kg/min) was infused for 4 min. Four minutes after infusion was complete, a second dose of 666 MBq (18 mCi) of 13N-ammonia was injected intravenously. Four minutes later to allow blood-pool clearing, PET was repeated by the same protocol as for the resting study. For angina, aminophylline (125 mg) was given intravenously. The follow-up PET scan was obtained at a mean of 2.6 y after the baseline PET.
Automated Quantitative Analysis of PET Images
Completely automated analysis of severity–size of PET abnormalities was performed by previously described software (18,21,23,24). A 3-dimensional (3D) restructuring algorithm generates true short- and long-axis views from PET transaxial cardiac images, perpendicular to and parallel to the long axis of the left ventricle (LV). Circumferential profiles are used to reconstruct 3D topographic views of the LV showing relative regional activity distribution divided into lateral, inferior, septal, and anterior quadrant views of the 3D topographic display in Figure 1.
|
PET Definitions
Severity of a perfusion defect is quantified as the lowest quadrant average relative activity—that is, the average relative activity for the quadrant having the lowest average activity of anterior, septal, lateral, and inferior quadrants for each subject, expressed as the percentage of the highest 2% of activity in the image dataset. Size of perfusion defects is quantified as the percentage of the whole cardiac image outside the 97.5% CI or 2.5 SD of healthy controls. Combined size and severity of perfusion defects is defined as the percentage of the whole cardiac image with relative activity of <60% of maximum activity (100%) that is 3.0 SD below the mean maximum activity of healthy controls. At follow-up study, a change in activity within ± 2.5 SD of our PET reproducibility data is considered not significant. Table 1 lists measurement values for the 20 healthy subjects and for the 10 patients with serial PET for reproducibility. Changes greater than this ± 2.5 SD of repeated scans were considered significantly changed as better or worse.
|
Baseline Worst Quadrant (BW).
BW is the quadrant with the minimum average quadrant activity after dipyridamole on the baseline PET, indicating the quadrant with the most severe defect representing the most severe flow-limiting stenosis at baseline. For changes at follow-up PET, this BW is compared with the same quadrant on follow-up dipyridamole PET images.
Follow-up Worst Quadrant (FuW).
FuW is the quadrant with the minimum average activity on the follow-up dipyridamole PET. FuW may be different from or the same as the BW.
Maximal Change Quadrant (MCh).
MCh is any quadrant with the greatest change from baseline to follow-up regardless of the presence or severity of baseline perfusion defects.
Combined Quadrants Change.
Combined Quadrants change is based on objectively measured changes in the BW, the FuW, and the MCh altogether. The Combined Quadrants change was defined as disease progression if all of these 3 baseline–to–follow-up comparisons—BW, FuW, and MCh—showed worsening. The Combined Quadrants change was defined as disease regression if all 3 baseline–to–follow-up comparisons showed improvement. The Combined Quadrants change was defined as mixed if neither progression nor regression criteria were met (specifically, if the change in the BW was opposite to the change in the FuW or MCh unless the FuW was the same quadrant as the BW; in that case, improvement or worsening in the MCh determined the final categorization as regression or progression, respectively). These quadrant pairings are summarized in Table 2.
|
|
2 testing. Continuous variables are reported as mean ± 1 SD with significance of differences among groups determined by ANOVA with Bonferroni–Dunn post hoc correction. Correlations were calculated using a Spearman rank correlation coefficient. Statistica for Windows 5.1 (Statsoft Inc.) was used for all statistical calculations. Stepwise multivariate logistic regression analyses adjusted for severity of the baseline PET defects were performed on all treated patients with the dependent variable being cardiovascular events—that is, cardiac death, nonfatal myocardial infarction, PCI, CABG, or stroke, over 5-y follow-up. The independent variables were as follows: (i) 15 clinical variables involving risk factors and intensity of treatment; (ii) 5 continuous values of the changes in severity on dipyridamole perfusion images at baseline and follow-up of the above 3 categories of quadrant pairs; (iii) change in size; (iv) change in combined size–severity of the perfusion defects; and (v) the discrete Combined Quadrants change of progression, regression, or mixed as defined above.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONReferences |
|---|
|
|
To avoid any potential bias due to patients with resting myocardial scar, all 409 patients were divided into those with and those without severe resting perfusion defects. A severe resting perfusion defect indicating myocardial scar was defined as >15% of the LV outside ± 2 SD of normal on the resting image. Of the 409 patients, 99 or 24% had a perfusion defect on the resting image that was outside ± 2 SD of healthy subjects. For these patients with a myocardial scar, the MCh, for either better or worse, was different from the BW in 75% of patients with myocardial scar as defined above. For the remaining 310 patients of the 409 with <15% of the LV outside ± 2 SD of normal on the resting image, the MCh was different from the BW in 78% of patients without myocardial scar, comparable to the patients with myocardial scar. The results are the same if myocardial scar at baseline was defined as >15% of the LV with <60% of maximum activity. Therefore, the presence of a baseline myocardial scar did not bias our results.
To examine these changes in a different way, we also compared the location of the FuW with the location of the BW. FuW was different from BW in 162 of 409 or 40% patients and coincided with BW in 247 of 409 or 60%. When FuW was different from BW, it had a more severe defect than the BW in 84 or 21% of patients. This observation illustrates that disease progression on follow-up PET frequently involved a quadrant different from baseline and, for 21% of patients, the new defects were worse than the worst defect at baseline.
Of the 409 patients, 45 had a revascularization procedure in the interval between the first and second study and 38 had a follow-up PET scan within 12 mo from baseline. The remaining 326 patients were assigned to 1 of the 3 medical treatment groups with comparable baseline clinical characteristics (18).
In Figure 4, the MCh, compared with its baseline, improved in 70% of patients on maximal treatment, in 48% of patients on moderate treatment, in 39% on poor treatment and in 47% of the patients undergoing revascularization procedures (P < 0.001).
|
|
|
|
The changes in severity of the BW from baseline to follow-up (BW) first reported here in this study parallel the changes in combined size–severity of perfusion defects for each of the treatment groups previously reported, except for the revascularization group. In the revascularization group of 45 patients, the quantitative severity of the BW decreased (improved) by 5.2% from baseline values, whereas the combined size–severity integrated throughout the whole heart increased (worsened) by 12.6%. Therefore, the expected improvement in severity of the worst baseline defect after revascularization was not paralleled by improvement throughout the whole heart. The revascularization group of 45 people is not large enough for analysis by intensity of medical treatment as maximal, moderate, or poor. Such analysis would also be biased by the effects of revascularization, irrespective of the intensity of medical treatment, if analyzed as a single or combined group. The differences among the treatment groups, including the revascularization group, for all continuous quantitative baseline–to–follow-up comparisons were significant (P 0.001).
In Table 4, multivariate logistic regression analysis adjusted for baseline size and severity of the perfusion defects of 15 clinical variables and 6 PET endpoints identified the following independent variables to be predictors of cardiovascular events over 5-y follow up: diabetes, family history of CAD, LDL cholesterol, HDL cholesterol, combined intense lifestyle and pharmacologic treatment, statin treatment, niacin or fibrate treatment, exercise, baseline–to–follow-up change in combined size–severity (odds ratio, 3.3; CI, 2.1–5.5; P = 0.01) and the Combined Quadrant changes (odds ratio, 2.6; CI, 1.7–3.1; P = 0.02), both of which accounted for the integrated overall changes throughout the heart. The change in the BW or any single quadrant pairing did not predict outcomes (Table 4). By Pearson correlation analysis, size or severity alone was not significantly correlated with either combined size–severity or Combined Quadrant changes.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONReferences |
|---|
Measures of the net overall perfusion changes throughout the entire coronary vascular tree, including regions other than the worst baseline stenosis, were predictors of coronary events but the change in the worst flow-limiting stenosis at baseline or any one segment of myocardium were not. Finally, patients undergoing intense lifestyle and pharmacologic treatment had more frequent regression of the baseline worst flow-limiting stenoses, fewer new perfusion defects and more frequent perfusion improvement in areas other than the baseline worst perfusion defect. The perfusion images reflect the balance of changes in anatomic stenosis, diffuse disease, and vasomotion due to endothelial dysfunction or their improvement with differing intensities of risk factor treatment not accounted for by arteriography or intravascular ultrasonography.
The change in severity of the BW was not significantly correlated with either the combined size–severity or Combined Quadrant changes by Pearson correlation analysis, consistent with the failure of changes in the BW or of any single quadrant pairing to predict clinical events. Thus, the changes accounting for the net balance of improving or worsening perfusion throughout the heart were better independent predictors of events than the change of the worst baseline flow-limiting stenosis. As expected and previously demonstrated (18), clinical risk factors and treatment intensity were also predictors of outcomes paralleling the integrated PET changes.
The regression or progression of coronary atherosclerosis throughout the entire coronary vascular tree as measured by changes in myocardial perfusion over a wide range of risk factor treatment provides direct evidence on the mechanism for the disproportionately greater reduction in cardiac events than changes in single stenosis severity in prior lipid-lowering trials using arteriographic endpoints (3,4,10–13). Inadequate risk factor treatment is associated with modest worsening in severity of the region with the worst baseline flow-limiting stenosis and a proportionately much greater frequency of new perfusion defects due to new stenosis in regions other than the worst region at baseline in association with more coronary events. This progression in areas other than the worst baseline stenosis also explains why the culprit artery in myocardial infarction often shows only mild stenosis on a prior arteriogram (7).
The group with revascularization procedures provides interesting information for the following reasons: (i) this group shows changes in regions other than the quadrant with the most severe baseline perfusion defect that subsequently underwent revascularization; (ii) the improved perfusion after revascularization in the quadrant with the most severe perfusion defect was no greater than the improvement in patients undergoing intense medical treatment; and (iii) the revascularization group serves as an internal control that tests whether PET perfusion images show improvement as they should after revascularization following the baseline PET.
Although the BW improved with intense risk factor treatment and worsened in its absence, the changes in the BW did not predict outcomes. In contrast, combined perfusion changes throughout the heart did predict outcomes, thereby indicating that the benefits of intense risk factor treatment are largely due to prevention of progression and plaque rupture in regions of the heart other than the baseline worst flow-limiting stenosis.
The current article provides new data on the mechanism for the disproportionately greater effects of lipid management on clinical outcomes than changes in single stenosis severity as assessed by coronary arteriography. Randomization of subjects to treatment groups is not necessary for this study on the mechanisms of progression or regression associated with clinical outcomes where a wide range of treatment intensity is essential to the analysis. Absolute myocardial perfusion in mL/min/g was not used in our study because quantification of relative myocardial uptake is optimal for the hypothesis tested without the assumptions and complexity of serial uptake images, arterial input imaging, and calculated absolute perfusion (25).
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONReferences |
|---|
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. L. Gould Coronary flow reserve and pharmacologic stress perfusion imaging beginnings and evolution. J. Am. Coll. Cardiol. Img., May 1, 2009; 2(5): 664 - 669. [Full Text] [PDF]
|
|
|
|
 |
|
 K. L. Gould, T. Pan, C. Loghin, N. P. Johnson, and S. Sdringola Reducing Radiation Dose in Rest-Stress Cardiac PET/CT by Single Poststress Cine CT for Attenuation Correction: Quantitative Validation J. Nucl. Med., May 1, 2008; 49(5): 738 - 745. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. L. Gould Not All Randomized Trials Are Equal J. Am. Coll. Cardiol., November 13, 2007; 50(20): 2013 - 2015. [Full Text] [PDF]
|
|
|
|
|
|
K. L. Gould, T. Pan, C. Loghin, N. P. Johnson, A. Guha, and S. Sdringola Frequent Diagnostic Errors in Cardiac PET/CT Due to Misregistration of CT Attenuation and Emission PET Images: A Definitive Analysis of Causes, Consequences, and Corrections J. Nucl. Med., July 1, 2007; 48(7): 1112 - 1121. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. S. Hochman and P. G. Steg Does Preventive PCI Work? N. Engl. J. Med., April 12, 2007; 356(15): 1572 - 1574. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| JOURNAL OF NUCLEAR MEDICINE TECHNOLOGY | THE JOURNAL OF NUCLEAR MEDICINE |
