Original Article
Development, diagnostic performance, and interobserver agreement of a 18F-flurpiridaz PET automated perfusion quantitation system

https://doi.org/10.1007/s12350-020-02335-6Get rights and content

Abstract

Background

Computerized methodologies standardize the myocardial perfusion imaging (MPI) interpretation process.

Methods

To develop an automated relative perfusion quantitation approach for 18F-flurpiridaz, PET MPI studies from all phase III trial participants of 18F-flurpiridaz were divided into 3 groups. Count distributions were obtained in N = 40 normal patients undergoing pharmacological or exercise stress. Then, N = 90 additional studies were selected in a derivation group. Following receiver operating characteristic curve analysis, various standard deviations below the mean normal were used as cutoffs for significant CAD, and interobserver variability determined. Finally, diagnostic performance was compared between blinded visual readers and blinded derivations of automated relative quantitation in the remaining N = 548 validation patients.

Results

Both approaches yielded comparable accuracies for the detection of global CAD, reaching 71% and 72% by visual reads, and 72% and 68% by automated relative quantitation, when using CAD ≥ 70% or ≥ 50% stenosis for significance, respectively. Similar results were observed when analyzing individual coronary territories. In both pharmacological and exercise stress, automated relative quantitation demonstrated significantly more interobserver agreement than visual reads.

Conclusions

Our automated method of 18F-flurpiridaz relative perfusion analysis provides a quantitative, objective, and highly reproducible assessment of PET MPI in normal and CAD subjects undergoing either pharmacological or exercise stress.

Introduction

In patients with apparently stable coronary artery disease (CAD), myocardial perfusion imaging (MPI) is a critical component in the determination of the hemodynamic significance of stenotic epicardial lesions and patient risk stratification.1,2 Despite multiple studies suggesting superior accuracy of positron emission tomography (PET) over single photon emission computed tomography (SPECT) MPI,3,4 particularly in female5,6 or obese7 patients, and in addition to lower radiation exposure,8 the widespread use of PET has been hampered by limitations of currently available radiopharmaceuticals.1 In this context, flurpiridaz, a novel 18F-based PET radiopharmaceutical with a half-life of 109 min available to imaging centers as unit doses, has undergone phase I-III multicenter clinical investigation.9, 10, 11

Relative perfusion assessed visually by radiotracer count distribution in regional myocardial tissue following image reconstruction remains the mainstay of MPI interpretation.12 To help standardize image interpretation and reduce intra- and interobserver variability, automated approaches to MPI have been developed for SPECT and PET radiotracers.13 Extraction of these quantitative parameters of cardiac radionuclide distribution assist the interpreting physician14 and have been derived for 13N-ammonia15,16 and 82Rb-chloride.17,18

In the present study, we developed an automated method for relative quantitation of 18F-flurpiridaz MPI. First, using a reference group we determined a normal database of flurpiridaz count distribution in individual coronary territories and in the global left ventricular myocardium. Then, using a derivation group we developed the lower limits of normal count distribution and the cutoffs for assessment of significant CAD in both pharmacological and exercise stress testing. Subsequently, we assessed the interobserver variability in the automated and visual approaches. Finally, we validated this method in a large group of patients from the phase III clinical trial of 18F-flurpiridaz by comparing performance metrics of automation with expert blinded visual reads using invasive coronary angiography as the reference standard.

Section snippets

Patient Population

PET MPI studies from all evaluable patients in the phase III trial of 18F-flurpiridaz11 (Lantheus Medical Imaging, ClinicalTrials.gov identifier: NCT01347710) suitable for quantitative analysis were included (N = 678) (Figure 1). All the patients signed an informed consent approved by the local institutional review boards. Patients were divided into 3 groups (Figure 1 and Table 1). Group 1: N = 40 normal patients to determine normal count distributions. Group 2: N = 90 derivation patients to

Normal Flurpiridaz Count Distribution and Determination of CAD Thresholds

In normal patients undergoing either pharmacological stress or treadmill exercise (Figure 2), flurpiridaz myocardial counts were reduced in the basal septum due to the membranous septum, and in the apex due to apical thinning. In the remaining LV segments, count distributions were uniform. In the derivation group, we then determined the standard deviation (SD) thresholds below mean normal counts leading to the highest ROC AUC for CAD detection in both pharmacological stress (Table 2) and

Discussion

We present the development and validation of an automated relative quantitation method of 18F-flurpiridaz PET MPI. To this end, we determined normal count distributions of flurpiridaz, established normal limits and criteria for abnormality, scrutinized interobserver agreement, and assessed the diagnostic performance of our computerized method in a large population of patients from the phase III clinical trial of 18F-flurpiridaz. To extract 3-dimensional myocardial maximal count distribution of

Limitations

Although the phase III clinical trial was a prospective study, we applied the automated relative perfusion quantitation method in a retrospective manner. Additionally, due to regulatory requirements, % stenosis by invasive coronary angiography was the “gold standard” comparator. Thus, other invasive findings such as fractional flow reserve (FFR) or instant wave-free ratio (iFR) were not measured as part of the trial. The known poor relationship of % stenosis with the functional significance of

Conclusion

The present comprehensive automated method of relative perfusion analysis provides a quantitative, accurate, objective, and highly reproducible, assessment of 18F-flurpiridaz PET MPI in normal and CAD subjects undergoing either pharmacological or exercise stress. These findings create the foundation for using an automated approach as a second on-site study interpretation with objective assessment.

New Knowledge Gained

The present study determined the diagnostic value of a newly developed automated relative quantitation system of 18F-flurpiridaz PET MPI in comparison to visual assessment. Whereas both techniques yield similar diagnostic accuracies, the software approach is significantly less prone to variability than visual reads. Our automated relative quantitation system of 18F-flurpiridaz PET MPI constitutes an objective, second interpretation available to the diagnosing physician who may decide to take

Acknowledgments

None.

Disclosures

Supported by NIH 1R43HL123069-01, 5R44HL123069-03, and VA Merit BX004558. Ernest Garcia and David Cooke receive royalties from the sales of the Emory Cardiac Toolbox related to the research described in this article. Kelly Champagne, Kenneth Van Train, and David Cooke are employees of Syntermed which commercially distributes the Emory Cardiac Toolbox. John Votaw is a consultant for the Emory Cardiac Toolbox. Joel Lazewatsky is a current and Cesare Orlandi a former employee of Lantheus

References (30)

Cited by (6)

  • Novel Tracers and Radionuclides in PET Imaging

    2021, Radiologic Clinics of North America

The authors of this article have provided a PowerPoint file, available for download at SpringerLink, which summarises the contents of the paper and is free for re-use at meetings and presentations. Search for the article DOI on SpringerLink.com.

View full text