Abstract
Background
Transient ischemic dilation (TID) of the LV is a specific marker of extensive coronary artery disease (CAD). Normal limits for TID using 18F-flurpiridaz have not been established.
Methods
We performed a secondary analysis of the multicenter phase III Flurpiridaz-301 trial, which included patients with suspected CAD who underwent rest–stress 18F-flurpiridaz positron emission tomography (PET), Tc-99m single-photon emission computed tomography (SPECT) myocardial perfusion imaging (MPI), and invasive coronary angiography (ICA). TID was calculated using ungated left ventricular volumes. Stress-specific reference thresholds were derived in patients without angiographic CAD. The effects of stress modality, CAD severity, and post-reconstruction Gaussian filtering on TID measurements were evaluated.
Results
Among 746 patients with evaluable PET, SPECT, and angiographic data, stress-specific TID reference thresholds (97.5th percentile) for 18F-flurpiridaz PET using a 5-mm Gaussian filter were 1.18 for exercise stress and 1.25 for pharmacologic stress. Corresponding Tc-99m SPECT thresholds were lower. PET-derived TID values increased progressively with angiographically defined CAD severity, demonstrating a graded association across disease categories. Application of post-reconstruction Gaussian filtering resulted in lower PET-derived TID values compared with unfiltered images, with modest differences between 5-mm and 8-mm filters.
Conclusion
This study suggests stress-specific reference thresholds for TID using 18F-flurpiridaz PET in a large multicenter cohort. PET-derived TID values were higher than those observed with Tc-99m SPECT, varied by stress modality, and were influenced by post-reconstruction filtering. These findings highlight the importance of tracer-, stress-, and protocol-specific reference standards for accurate clinical interpretation of TID.
Graphical Abstract
INTRODUCTION
Transient ischemic dilation (TID) of the LV is a well-established marker associated with extensive, multivessel, coronary artery disease (CAD).1–3 It can be assessed visually or quantitatively by comparing ungated left ventricular (LV) cavity size on stress and rest imaging. The underlying mechanism of TID remains debated and is likely multifactorial1. Proposed explanations include true LV cavity dilation, related to subendocardial ischemia and elevated end-diastolic pressure, particularly during exercise stress.1 Alternatively, apparent dilation may result from subendocardial hypoperfusion, with or without true ischemia, which may be accentuated during pharmacologic vasodilator stress through perfusion or coronary steal2,4.
Although TID demonstrates high specificity for severe CAD, its sensitivity is modest when used in isolation. Diagnostic performance improves when TID is interpreted alongside additional perfusion parameters such as total perfusion defect (TPD).5 Beyond its diagnostic role, elevated TID ratios have also been associated with adverse prognosis, including increased risk of myocardial infarction and all-cause mortality.2,4–7
Abnormal TID thresholds have been reported across multiple imaging modalities including planar scintigraphy, single-photon emission computed tomography (SPECT), and positron emission tomography (PET), as well as across different stress agents and radiotracers1,2. Reported values vary according to imaging protocol, tracer characteristics, quantitative software and post-processing methods, underscoring the need for modality and protocol-specific reference standards.
18F-flurpiridaz is a PET myocardial perfusion imaging tracer characterized by high myocardial extraction, a half-life of 109 minutes, allowing its performance with exercise, and high spatial resolution.9,10 In phase III clinical trials, 18F-flurpiridaz PET demonstrated diagnostic performance comparable or superior to technetium-99m SPECT for the detection of CAD.11 However, despite its growing clinical use, reference limits for abnormal TID using 18F-flurpiridaz PET have not yet been established. The objectives of this study were to derive reference limits for abnormal TID using 18F-flurpiridaz PET and to evaluate the relationship between TID and the extent of CAD.
METHODS
Patient population
This was a secondary analysis of data from a multicenter Phase III clinical trial11 (Lantheus Medical Imaging, NCT01347710) evaluating 18F-flurpiridaz PET myocardial perfusion imaging (MPI) in patients with suspected CAD who were referred for a clinically indicated invasive coronary angiography (ICA). The study was conducted in patients with known or suspected CAD across 72 centers across the United States, Canada, and Finland between 2011 and 2013. Exclusion criteria included unstable cardiovascular status (myocardial infarction or unstable angina within 6 months; transient ischemic attack or stroke within 3 months), symptomatic valvular heart disease, significant congenital heart disease, New York Heart Association (NYHA) class III or IV heart failure, prior coronary artery bypass grafting, nonischemic cardiomyopathy, and history of heart transplantation. Patients underwent both rest and stress PET and SPECT MPI, as well as invasive coronary angiography, and 746 patients had evaluable data from all three studies. Ethics committee approval was obtained at each participating site, and all patients provided written informed consent.
Flurpiridaz-301 Trial Design
All patients underwent rest-stress PET and SPECT myocardial perfusion imaging (MPI) within 60 days of invasive coronary angiography using standardized acquisition protocols. The same stress modality—either exercise or pharmacologic (regadenoson, dipyridamole, or adenosine)—was used for both modalities in each patient. PET MPI was performed using a 1-day protocol with 18F-flurpiridaz as the radiotracer. Dosing was prespecified as 2.5–3.0 mCi for rest, 9.0–9.5 mCi for exercise stress (with a minimum 60-minute rest-to-stress interval), and 6.0–6.5 mCi for pharmacologic stress (with a minimum 30-minute interval). SPECT MPI was performed using technetium-99m-labeled tracers (sestamibi or tetrofosmin) following either 1-day or 2-day protocols per site preference, with rest and stress doses of 8.0–12.0 mCi and 24.0–36.0 mCi, respectively, and a minimum 3:1 stress-to-rest dose ratio. All PET and SPECT cameras were pre-qualified for technical adequacy and image quality. Electrocardiographically gated imaging was acquired for both rest and stress PET studies and for stress SPECT to allow functional analysis.11 Invasive coronary angiography (ICA) was performed using local standards and interpreted by a blinded core laboratory (PERFUSE Core Laboratories, Boston, MA) for quantitative assessment of coronary stenosis.11
Data collection and processing
Experienced core laboratory technologists, blinded to the clinical data, conducted a thorough review of images. Automated left ventricular contours were reviewed for accuracy and manual adjustment was performed when necessary to ensure appropriate delineation of endocardial borders.12
Derivation of transient ischemic dilation (TID)
Transient ischemic dilation was calculated as the ratio of stress to rest left ventricular cavity volumes and was derived automatically from static (ungated) left ventricular volume measurements from rest and stress myocardial perfusion images using Quantitative Perfusion SPECT (QPS) (Cedars-Sinai Medical Center, Los Angeles, CA)13 and a quantitative PET package (QPET).
Image Reconstruction
PET and SPECT images were reconstructed according to site-specific and trial-approved reconstruction protocols. All reconstructed images met predefined image quality standards established by the core laboratory prior to inclusion in the analysis.
Gaussian filtering
Primary data was obtained from unfiltered PET images. In addition, to reflect standard clinical post-processing and to assess the effect of spatial smoothing on TID measurements, PET datasets were subsequently processed using Gaussian filters with kernel widths of 5 mm and 8 mm. TID values were derived from both unfiltered and filtered images.
Invasive coronary angiography
Invasive coronary angiography (ICA) was performed according to standard protocols at each investigational site. Following catheterization, imaging data were submitted to the core laboratory for quality control. Coronary artery stenosis was quantified using blinded quantitative coronary analysis performed by experienced core lab readers.11,14
CAD severity group classification
Patients were stratified into four mutually exclusive CAD severity groups based on quantitative coronary angiography. The Very Low-Risk group was used to define the limit for abnormality. It was defined as patients with no stenosis of ≥30% in any coronary artery and no prior CAD. The High-Risk CAD subgroup was defined as patients with stenosis ≥50% in the left main (LM) artery and/or ≥70% in the proximal left anterior descending (LAD) artery. The Intermediate-Risk CAD subgroup was defined as patients with stenosis of 30–49% in the LM or ≥70% in any other coronary artery other than the proximal LAD. The Low-Risk CAD subgroup was defined as patients not falling into the intermediate- or high-risk categories.
Statistical analysis
Continuous variables are summarized as means ± standard deviations, and categorical variables are presented as counts with percentages. Comparisons across CAD severity groups were performed using Kruskal-Wallis tests for continuous variables and Chi squared or Fisher’s exact tests for categorical variables, as appropriate. To establish reference thresholds for TID, analyses were performed in the Very Low-Risk subgroup. The distribution of TID values was assessed using the Shapiro-Wilk test. PET TID values deviated from strict normality, whereas SPECT TID values were consistent with normal distribution. Abnormal TID thresholds were derived using both a parametric approach (mean + 2 standard deviations) and a nonparametric approach based on the 97.5th percentile for comparison. All analyses were performed using R (R Foundation for Statistical Computing, Vienna, Austria). A two-sided p-value<0.05 was considered statistically significant where applicable.
RESULTS
Study cohort and baseline characteristics
The study cohort comprised a total of 746 patients, of whom 147 (19.7%) were classified as being Very Low-Risk, 385 (51.2%) were Low-Risk, 192 (25.7%) were Intermediate-Risk, and 22 (2.9%) were High-Risk. Baseline demographic and clinical characteristics stratified by CAD severity are shown in Table 1. The mean age among all patients was 62.4 years (SD 9.5). The Very Low-Risk subgroup had the highest proportion of females (52.4%). The Intermediate-Risk subgroup had the lowest proportion of females (13.5%). Mean body mass index (BMI) differed modestly between groups (p=0.016). Pharmacologic stress was used more frequently than exercise stress in all subgroups (p=0.929). The majority of patients were White (82.2%), followed by Black (13.3%) and race distribution varied across groups (p = 0.022). The prevalence of cardiovascular risk factors, including hypertension (p = 0.002), hyperlipidemia (p<0.001) and smoking (p<0.001) increased with CAD severity. The prevalence of stroke or TIA was similar across groups (p=0.207). Resting left ventricular ejection fraction decreased modestly with increasing CAD severity ( p = 0.04).
Table 1.
Baseline characteristics stratified by CAD severity
| All Patients (N=746) |
Very Low-Risk (N=147) |
Low-Risk CAD (N=385) |
Intermediate-Risk CAD (N=192) |
High-Risk CAD (N=22) |
P-Value | |
|---|---|---|---|---|---|---|
| Age (years) | 62.4 (9.5) | 58.4 (9.7) | 62.5 (9.3) | 65.0(8.5) | 63.7(11.5) | <0.001 |
| Female sex | 230 (30.8) | 77 (52.4) | 122 (31.7) | 26 (13.5) | 5 (22.7) | <0.001 |
| BMI, kg/m2 | 31.1(6.6) | 32.8 (6.9) | 30.7 (5.9) | 30.4 (6.8) | 31.7(8.1) | 0.016 |
| Stress Type | 0.929 | |||||
| Exercise | 219 (29.4) | 46 (31.3) | 113 (29.4) | 54 (28.1) | 6 (27.3) | |
| Pharmacologic | 527 (70.6) | 101 (68.7) | 272 (70.1) | 138 (71.9) | 16 (72.7) | |
| Race | 0.022 | |||||
| White | 613 (82.2) | 106 (72.1) | 328 (85.2) | 161 (83.9) | 18 (81.8) | |
| Black | 99 (13.3) | 28 (19.0) | 43 (11.2) | 25 (13.0) | 3 (13.6) | |
| Other | 34 (4.6) | 13 (8.8) | 14 (3.6) | 6 (3.1) | 1 (4.5) | |
| Comorbidities | ||||||
| Hypertension | 612 (82.0) | 107 (72.8) | 320 (83.1) | 169 (88.0) | 16 (72.7) | 0.002 |
| Hyperlipidemia | 644 (86.3) | 104 (70.7) | 344 (89.4) | 177 (92.2) | 19 (86.4) | <0.001 |
| Diabetes | 257 (34.5) | 33 (22.4) | 129 (33.5) | 85 (44.3) | 10 (45.5) | <0.001 |
| Tobacco use | 444 (59.5) | 70 (47.6) | 234 (60.8) | 131 (68.2) | 9 (40.9) | <0.001 |
| Stroke/TIA | 51 (6.8) | 7 (4.8) | 26 (6.8) | 18 (9.4) | 0 | 0.207 |
| Prior CAD | 459 (61.5) | 0 (0) | 301 (78.2) | 147 (76.6) | 11 (50.0) | - |
| Resting LVEF (%) | 58.4 (10.9) | 59.6 (11.2) | 58.8 (10.4) | 57 (10.9) | 55 (13.8) | 0.04 |
Baseline demographic and clinical variables are presented for all patients, Very Low-Risk, Low-Risk CAD, Intermediate-Risk CAD and High-Risk CAD subgroups. CAD subgroups were classified according to prespecified angiographic criteria, with refinement of Very Low-Risk and Low-Risk groups based on prior CAD. Continuous variables are reported as mean (standard deviation) and categorical variables as counts (percentage). Only the four CAD subgroups were used for statistical significance analysis.
CAD=coronary artery disease. BMI=body mass index. TIA=transient ischemic attack. LVEF=left ventricular ejection fraction.
Establishing TID reference thresholds
Among the reference group of patients classified as being Very Low-Risk, TID values were derived for both 18F-flurpiridaz PET using a 5-mm Gaussian post-reconstruction and for Tc-99m SPECT (Table 2). The upper normal limit for TID was defined as the 97.5th percentile of the reference population. Because PET TID values were not normally distributed, a nonparametric percentile-based threshold was preferred over a parametric mean + 2 SD approach, however, mean + 2 SD values were calculated and reported for comparison. The same framework was applied to SPECT to facilitate direct comparison between modalities.
Table 2.
TID reference thresholds for 18F flurpiridaz PET (5 mm Gaussian filter) and Tc-99m SPECT studies
| TID | 97.5th percentile | Mean+2SD | |
|---|---|---|---|
| PET (N=147) | |||
| Exercise (N=46) | 0.98 (0.11) | 1.18 | 1.20 |
| Pharmacologic (N=101) | 1.05 (0.10) | 1.25 | 1.26 |
| SPECT (N=147) | |||
| Exercise (N=46) | 0.97 (0.11) | 1.15 | 1.20 |
| Pharmacologic (N=101) | 1.02 (0.10) | 1.25 | 1.26 |
Mean (standard deviation) TID values are shown for the reference group classified as Very Low-Risk who underwent 18F flurpiridaz PET and Tc-99m SPECT exercise and pharmacologic stress studies. Post-reconstruction 5mm Gaussian filters were applied to 18F flurpiridaz PET perfusion images. The Very Low-Risk subgroup was defined according to prespecified angiographic criteria and absence of prior CAD. Thresholds for abnormal TID were derived using a percentile-based method because PET TID values deviated from normal distribution. Thresholds by parametric approach (mean + 2SD) are also reported for comparison.
TID=transient ischemic dilation. CAD=coronary artery disease. SD=Standard deviation.
Using this approach, TID thresholds were higher with pharmacologic compared with exercise stress for both PET and SPECT. For PET, the 97.5th thresholds were 1.18 with exercise and 1.25 with pharmacologic stress, while for SPECT the corresponding thresholds were 1.15 and 1.25, respectively.
TID increases with increasing CAD severity
Mean TID values for 18F-flurpiridaz PET and Tc-99m SPECT stratified by CAD severity are shown in Table 3 and plotted in Figure 1. For both imaging modalities, mean TID values increased progressively with increasing CAD severity, demonstrating a consistent stepwise relationship across disease categories. Across all CAD severity categories, PET-derived TID values were consistently higher than corresponding SPECT values. The magnitude of increase in TID from Very-Low Risk to High-Risk groups was greater for PET than for SPECT, with an absolute increase of 0.11 (10.7%) for PET compared with 0.07 (6.9%) for SPECT. When PET TID values were further stratified by stress modality, a similar increase in TID across severity categories was observed for both exercise and pharmacologic stress studies (Supplementary Figure S1), although confidence intervals were wider in the high-risk subgroups because of limited sample size.
Table 3.
TID values for 18F flurpiridaz PET (5 mm Gaussian filter) and Tc-99m SPECT studies stratified by CAD severity
| PET | SPECT | |
|---|---|---|
| Very Low-risk (N=147) | 1.03 (1.01–1.05) | 1.01 (0.99–1.03) |
| Low-Risk CAD (N=385) | 1.05 (1.03–1.06) | 1.01 (1.00–1.02) |
| Intermediate CAD (N=192) | 1.08 (1.07–1.11) | 1.04 (1.02–1.06) |
| High-risk CAD (N=22) | 1.14 (1.11–1.18) | 1.08 (1.02–1.14) |
Mean (95% CI) TID values are shown for participants who underwent 18F flurpiridaz PET and Tc-99m SPECT studies (both exercise and pharmacologic stress) stratified by CAD subgroup according to prespecified angiographic criteria, with further refinement of Very Low-Risk and Low-Risk groups based on prior CAD. Post-reconstruction 5mm Gaussian filters were applied to 18F flurpiridaz PET perfusion images.
CAD=coronary artery disease. TID=transient ischemic dilation; CI=confidence interval
Figure 1. TID values for 18F flurpiridaz PET (5 mm Gaussian filter) and Tc-99m SPECT studies stratified by CAD subgroups.
Plot of TID values (mean and 95% CI) for participants who underwent 18F flurpiridaz PET and Tc-99m SPECT studies (both exercise and pharmacologic stress) stratified by CAD subgroup according to prespecified angiographic criteria, with further refinement of Very Low-Risk and Low-Risk groups based on prior CAD. Post-reconstruction 5mm Gaussian filters were applied to 18F flurpiridaz PET perfusion images
CAD=coronary artery disease. TID=transient ischemic dilation; CI=confidence interval
Post-reconstruction Gaussian filtering results in lower PET-derived TID values
PET-derived TID reference values varied with the degree of post-reconstruction Gaussian filtering for both exercise and pharmacologic stress (Table 4, Figure 2). Compared with unfiltered images, filtered PET images yielded slightly lower mean TID values and lower TID thresholds. Differences between the 5-mm and 8-mm filters were modest for both stress modalities. Overall, the application of Gaussian filtering was associated with lower PET-derived TID thresholds. Representative short-axis images demonstrating the effect of increasing Gaussian filter width on left ventricular cavity definition are shown in Figure 3.
Table 4.
Effect of Gaussian filtering on TID reference thresholds for 18F flurpiridaz PET studies
| TID | Threshold for abnormal TID | |
|---|---|---|
| Exercise (N=46) | ||
| No filter | 1.00 (0.13) | 1.23 |
| 5 mm filter | 0.98 (0.11) | 1.18 |
| 8 mm filter | 0.97 (0.11) | 1.15 |
| Pharmacologic (N=101) | ||
| No filter | 1.06 (0.12) | 1.30 |
| 5 mm filter | 1.05 (0.10) | 1.25 |
| 8 mm filter | 1.06 (0.10) | 1.27 |
Mean (standard deviation) TID values are shown for the reference group classified as Very Low-Risk who underwent 18F flurpiridaz PET exercise and pharmacologic stress studies, followed by no post-reconstruction Gaussian filtering, application of 5 mm Gaussian filter, and application of 8 mm Gaussian filter. The Very Low-risk subgroup was defined according to prespecified angiographic criteria and absence of prior CAD. Threshold for abnormal TID in each category was set at greater than 97.5th percentile of the reference population.
TID=transient ischemic dilation. CAD=coronary artery disease.
Figure 2. Effect of Gaussian filtering on TID reference thresholds for 18F flurpiridaz PET studies.
Mean TID values in the Very Low-Risk reference group for patients undergoing 18F flurpiridaz PET with exercise or pharmacologic stress, following no post-reconstruction Gaussian filtering, application of a 5-mm Gaussian filter, and application of an 8-mm Gaussian filter. Dashed horizontal lines denote the threshold for abnormal TID, defined as greater than 97.5th percentile for each category
TID=transient ischemic dilation. CAD=coronary artery disease.
Figure 3. Effect of post-reconstruction Gaussian filtering on LV volume on short-axis 18F-flurpiridaz PET images.
Representative short-axis slices from a high-risk CAD patient with elevated TID demonstrating the effect of Gaussian filtering on LV cavity definition. Post-reconstruction without Gaussian filtering (a), with a 5 mm Gaussian filter (b) and with an 8 mm Gaussian filter (c)
DISCUSSION
In this large multicenter cohort from Flurpiridaz 301 trial, we suggest stress-specific reference thresholds for transient ischemic dilation using 18F-flurpiridaz PET and compared them to SPECT TID thresholds obtained in the same patients. SPECT was included to provide contextual comparison with an established clinical modality, as most prior TID thresholds have been derived from SPECT imaging. PET-derived TID values were consistently higher than those observed with Tc-99m SPECT across CAD severity groups, with higher thresholds observed under pharmacologic stress compared with exercise. TID demonstrated a graded association with angiographically defined CAD severity for both modalities, supporting its role as a marker of more extensive coronary disease. The greater magnitude of increase in PET-derived vs SPECT-derived TID suggests greater sensitivity of PET for diagnosis of high-risk CAD. Importantly, absolute TID values and abnormality thresholds were influenced by post-reconstruction Gaussian filtering, underscoring the need for tracer-, stress-, and protocol-specific reference standards when interpreting TID in clinical practice.
Our results also confirm that stress modality significantly influences TID. Pharmacologic vasodilator stress yielded higher PET TID values than exercise, consistent with earlier observations from SPECT and Rb-82 PET.5,15,16 Several physiological mechanisms may account for this difference. At rest, subendocardial perfusion is maintained by autoregulatory mechanisms despite higher wall stress. vasodilator agents induce marked coronary hyperemia without a corresponding increase in myocardial workload, potentially leading to relative diffuse subendocardial hypoperfusion (transmural steal) and reduced subendocardial tracer uptake. This results in apparent myocardial thinning – a partial-volume effect – and apparent LV cavity dilation. In contrast, exercise produces more physiologic increases in myocardial blood flow and oxygen demand, which better preserves subendocardial tracer activitiy and endocardial border definition, thereby not increasing the apparent LV cavity size as much as vasodilator stress17.
In this study, TID demonstrated a graded increase with worsening CAD severity across both PET and SPECT imaging, supporting its association with more extensive coronary disease. This stepwise pattern is consistent with the proposed pathophysiology of TID as a marker of subendocardial ischemia and stress-induced imbalance between myocardial demand and perfusion reserve and prior studies. Notably, the magnitude of TID and the separation between CAD subgroups were more pronounced with 18F-flurpiridaz PET compared with Tc-99m SPECT. This likely reflects differences in tracer characteristics and image quality, including higher myocardial extraction and improved myocardial-to-cavity contrast with PET, which may enhance sensitivity to subtle changes in LV cavity size under stress. Despite these modality-specific differences, the overall concordance in directional trends across PET and SPECT supports the robust relationship between increasing TID and angiographically defined CAD.
Gaussian filtering is a post-reconstruction smoothing technique routinely applied in clinical PET myocardial perfusion imaging to reduce image noise and stabilize myocardial and cavity contours.18,19 In this study, availability of both unfiltered and filtered PET images allowed direct evaluation of the impact of Gaussian filtering on TID measurements. Application of Gaussian filtering was associated with modestly lower TID thresholds and slightly lower TID threshold values compared with unfiltered images. This effect likely reflects smoothing of high-frequency noise and partial-volume effects at the myocardial-blood pool interface, which can influence automated cavity delineation and apparent cavity size on stress images. Although differences between 5 mm and 8 mm Gaussian filters were modest, filtering parameters nonetheless affected absolute TID values. These findings highlight the importance of standardized post-reconstruction filtering protocols in myocardial perfusion imaging (MPI) and underscore that filtering choices should be considered when defining and applying TID reference thresholds for clinical interpretation.
This study provides the first stress-specific reference thresholds for TID using 18F-flurpiridaz in a large, multicenter cohort. Static TID thresholds were 1.18 and 1.25 with exercise and pharmacologic stress, respectively. SPECT TID values in this cohort were concordant with established reference thresholds, serving as an internal benchmark5,15,16,20. Prior PET studies evaluating TID have reported variable thresholds values depending on tracer, software platform and quantification methodology. A study using 82Rb PET with dipyridamole stress reported a lower static TID threshold of 1.13, derived using a parametric approach based on a mean + 2.5 standard deviations20. In a study using 13N-ammonia PET with pharmacologic stress, the upper reference limit for static TID was reported as 1.17 with Cedars-Sinai QPS/QGS software, whereas higher values of 1.25 were observed with Corridor4DM21. While it is tempting to attribute the relatively higher TID values observed with 13N-ammonia and 18F-Flurpiridaz than with 82Rb PET to shared tracer characteristics such as high myocardial extraction and favorable myocardial-to-cavity contrast, that remains speculative. Methodological variability, including software platform, stress protocol, statistical approach and imaging processing likely also contribue substantially to differences in reported PET TID thresholds and underscores the need for tracer- and protocol-specific reference values.
Limitations
The Flurpiridaz-301 trial was primarily a diagnostic performance and safety study and patients with extensive multivessel disease were underrepresented, limiting the ability to robustly evaluate threshold-based prediction. Accordingly, the present findings should be interpreted as establishing reference values and demonstrating biologic association with disease severity rather than defining clinical decision thresholds for risk stratification.
Given the demonstrated influence of stress type on TID values, we performed stress-specific analysis to establish abnormal TID thresholds. However, analysis of TID across CAD severity (Table 3 and Figure 1) were not stratified by stress type due to the relatively small number of patients, especially in the Intermediate and High-Risk categories.
SPECT studies were acquired across multiple camera systems and reconstruction protocols, which may introduce variability in TID measurements. However, SPECT analysis was included primarily to provide contextual comparison with PET rather than to establish new SPECT-specific threshold. While we emphasize the variability of TID based on modality, stress-type and post-processing protocol, we did not stratify by pharmacologic stress agents. Differences in the hemodynamic effects of regadenoson, adenosine, and dipyridamole may contribute to variability in TID measurements under pharmacologic stress.
Conclusion
In a large multicenter cohort, we established stress-specific reference thresholds for transient ischemic dilation using 18F-flurpiridaz PET myocardial perfusion imaging and compared them to SPECT thresholds obtained in the same patients. PET-derived TID values were consistently higher than those observed with Tc-99m SPECT, varied by stress modality, and demonstrated a graded increase with angiographically defined CAD severity. Importantly, absolute TID values and abnormality thresholds were influenced by post-reconstruction Gaussian filtering, underscoring the sensitivity of TID measurements to post-processing methodology. Together, these findings highlight the need for tracer-stress and protocol-specific reference standards when interpreting TID in clinical practice and provide a framework for the appropriate application of TID using 18F-flurpiridaz PET.
Supplementary Material
NEW KNOWLEDGE GAINED AND CLINICAL IMPLICATIONS.
What is New?
This study provides stress-specific thresholds for transient ischemic dilation using 18F-flurpiridaz PET myocardial perfusion imaging in a very low-risk population.
Measured TID values were influenced by image filtering, demonstrating the importance of reconstruction and processing parameters when establishing normal reference ranges.
What Are the Clinical Implications?
Stress-specific TID thresholds may improve interpretation of 18F-flurpiridaz PET studies by accounting for physiologic differences between exercise and pharmacologic stress testing.
Recognition of effects of post-reconstruction filtering underscores the importance of establishing consistent processing protocols and reference limits specific to methodology.
ACKNOWLEDGEMENTS
The authors thank the investigators and study coordinators of the flurpiridaz phase III clinical trial for their contributions to data acquisition and study conduct.
Sources of funding:
This research was supported in part by grant R35HL161195 from the National Heart, Lung, and Blood Institute at the National Institutes of Health (PI: Slomka). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. This work was also supported in part by the Dr. Miriam and Sheldon G. Adelson Medical Research Foundation.
ABBREVIATIONS:
- TID
transient ischemic dilation
- CAD
coronary artery disease
- PET
positron emission tomography
- SPECT
single photon emission computed tomography
- MPI
myocardial perfusion imaging
- ICA
invasive coronary angiography
- LV
left ventricle
- LM
left main
- LAD
left anterior descending
Footnotes
Conflicts of interest: DB and PS participate in software royalties for QPS software at Cedars-Sinai Medical Center. PS received grant support from Siemens and Synektik. DB served as a consultant for GE Healthcare. RM has received research support from Alberta Innovates and consulting fees from Bayer and Alnylam. CB is an employee of GE Healthcare. All other authors have no disclosures.
REFERENCES
- 1.Teddy Weiss A, Berman DS, Lew AS, et al. Transient ischemic dilation of the left ventricle on stress thallium-201 scintigraphy: A marker of severe and extensive coronary artery disease. J Am Coll Cardiol. 1987;9(4):752–759. doi: 10.1016/S0735-1097(87)80228-0 [DOI] [PubMed] [Google Scholar]
- 2.Abidov A, Bax JJ, Hayes SW, 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. 2004;45(12):1999–2007. [PubMed] [Google Scholar]
- 3.McLaughlin MG, Danias PG. Transient ischemic dilation: A powerful diagnostic and prognostic finding of stress myocardial perfusion imaging. J Nucl Cardiol. 2002;9(6):663–667. doi: 10.1067/mnc.2002.124979 [DOI] [PubMed] [Google Scholar]
- 4.Slomka PJ, Berman DS, Germano G. Normal limits for transient ischemic dilation with (99m)Tc myocardial perfusion SPECT protocols. J Nucl Cardiol Off Publ Am Soc Nucl Cardiol. 2017;24(5):1709–1711. doi: 10.1007/s12350-016-0582-4 [DOI] [PubMed] [Google Scholar]
- 5.Xu Y, Arsanjani R, Clond M, et al. Transient ischemic dilation for coronary artery disease in quantitative analysis of same-day sestamibi myocardial perfusion SPECT. J Nucl Cardiol Off Publ Am Soc Nucl Cardiol. 2012;19(3):465–473. doi: 10.1007/s12350-012-9527-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.jia Y, Hu Y, Yang L, et al. Prognostic value of transient ischemic dilatation by 13N-ammonia PET MPI for short-term outcomes in patients with non-obstructive CAD. Ann Nucl Med. 2025;39(1):47–57. doi: 10.1007/s12149-024-01976-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bestetti A, Bigi R, Terranova P, Lombardi F, Fiorentini C. Prognostic implications of stress-induced transient ischemic dilation of the left ventricle in patients with systolic dysfunction and fixed perfusion defects. Int J Cardiol. 2010;140(3):323–327. doi: 10.1016/j.ijcard.2008.11.121 [DOI] [PubMed] [Google Scholar]
- 8.uz Zaman M, Fatima N, Samad A, et al. Predictive and prognostic values of transient ischemic dilatation of left ventricular cavity for coronary artery disease and impact of various managements on clinical outcome using technetium-99m sestamibi gated myocardial perfusion imaging. Ann Nucl Med. 2011;25(8):566–570. doi: 10.1007/s12149-011-0500-4 [DOI] [PubMed] [Google Scholar]
- 9.Maddahi J Properties of an ideal PET perfusion tracer: new PET tracer cases and data. J Nucl Cardiol Off Publ Am Soc Nucl Cardiol. 2012;19 Suppl 1:S30–7. doi: 10.1007/s12350-011-9491-8 [DOI] [PubMed] [Google Scholar]
- 10.Maddahi J, Agostini D, Bateman TM, et al. Flurpiridaz F-18 PET Myocardial Perfusion Imaging in Patients With Suspected Coronary Artery Disease. J Am Coll Cardiol. 2023;82(16):1598–1610. doi: 10.1016/j.jacc.2023.08.016 [DOI] [PubMed] [Google Scholar]
- 11.Maddahi J, Lazewatsky J, Udelson JE, et al. Phase-III Clinical Trial of Fluorine-18 Flurpiridaz Positron Emission Tomography for Evaluation of Coronary Artery Disease. J Am Coll Cardiol. 2020;76(4):391–401. doi: 10.1016/j.jacc.2020.05.063 [DOI] [PubMed] [Google Scholar]
- 12.Ramirez G, Lemley M, Shanbhag A, et al. The REgistry of Flow and Perfusion Imaging for Artificial Intelligence with positron emission tomography (REFINE PET):Rationale and design. J Nucl Cardiol Off Publ Am Soc Nucl Cardiol. 2025;52:102449. doi: 10.1016/j.nuclcard.2025.102449 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Hu LH, Sharir T, Miller RJH, et al. Upper reference limits of transient ischemic dilation ratio for different protocols on new-generation cadmium zinc telluride cameras: A report from REFINE SPECT registry. J Nucl Cardiol Off Publ Am Soc Nucl Cardiol. 2020;27(4):1180–1189. doi: 10.1007/s12350-019-01730-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Moody JB, Poitrasson-Rivière A, Hagio T, et al. Added value of myocardial blood flow using (18)F-flurpiridaz PET to diagnose coronary artery disease: The flurpiridaz 301 trial. J Nucl Cardiol Off Publ Am Soc Nucl Cardiol. 2021;28(5):2313–2329. doi: 10.1007/s12350-020-02034-2 [DOI] [PubMed] [Google Scholar]
- 15.Mandour Ali MA, Bourque JM, Allam AH, Beller GA, Watson DD. The prevalence and predictive accuracy of quantitatively defined transient ischemic dilation of the left ventricle on otherwise normal SPECT myocardial perfusion imaging studies. J Nucl Cardiol Off Publ Am Soc Nucl Cardiol. 2011;18(6):1036–1043. doi: 10.1007/s12350-011-9458-9 [DOI] [PubMed] [Google Scholar]
- 16.Jameria ZA, Abdallah M, Dwivedi A, et al. Computer derived transient ischemic dilation ratio for identifying extensive coronary artery disease using a CZT camera and imaging in the upright position. J Nucl Cardiol Off Publ Am Soc Nucl Cardiol. 2017;24(5):1702–1708. doi: 10.1007/s12350-016-0515-2 [DOI] [PubMed] [Google Scholar]
- 17.Chouraqui P, Rodrigues EA, Berman DS, Maddahi J. Significance of dipyridamole-induced transient dilation of the left ventricle during thallium-201 scintigraphy in suspected coronary artery disease. Am J Cardiol. 1990;66(7):689–694. doi: 10.1016/0002-9149(90)91131-o [DOI] [PubMed] [Google Scholar]
- 18.Kim HS, Cho SG, Kim JH, Kwon SY, Lee B, Bom HS. Effect of Post-Reconstruction Gaussian Filtering on Image Quality and Myocardial Blood Flow Measurement with N-13 Ammonia PET. Asia Ocean J Nucl Med Biol. 2014;2(2):104–110. https://aojnmb.mums.ac.ir/article_2841.html [PMC free article] [PubMed] [Google Scholar]
- 19.Chen GP, Branch KR, Alessio AM, et al. Effect of reconstruction algorithms on myocardial blood flow measurement with 13N-ammonia PET. J Nucl Med. 2007;48(8):1259–1265. doi: 10.2967/jnumed.106.038232 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Rischpler C, Higuchi T, Fukushima K, et al. Transient ischemic dilation ratio in 82Rb PET myocardial perfusion imaging: normal values and significance as a diagnostic and prognostic marker. J Nucl Med. 2012;53(5):723–730. doi: 10.2967/jnumed.111.097600 [DOI] [PubMed] [Google Scholar]
- 21.Martínez-Lucio TS, Lazarenko SV, Mendoza-Ibañez OI, et al. Transient ischemic dilation ratio by [(13)N]NH(3) PET/CT in patients with variable extent of myocardial ischemia: relevance of software package and quantification method. Int J Cardiovasc Imaging. 2025;41(8):1605–1617. doi: 10.1007/s10554-025-03468-w [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.