Targeted Molecular Imaging Sheds Light on Bioprosthetic Aortic Valve Thrombosis

Molecular imaging with positron emission tomography (PET) has a growing number of applications in cardiovascular diseases. One such area of recent progress has involved the development of targeted radiotracers to detect thrombosis.^1-3 A novel tracer derived from elarofiban, ¹⁸F-GP1, to target the glycoprotein (GP) IIb/IIIa receptor on activated platelets has identified in vivo thrombosis in case studies.^2,3 Although early results are promising, the range of potential clinical applications for ¹⁸F-GP1 and other tracers targeting thrombosis continues to be elucidated.

Bioprosthetic aortic valve (BAV) thrombosis is relatively common and contributes to both valvular degeneration and dysfunction.⁴ It has increasingly gained recognition with the rapidly expanding use of transcatheter BAVs in clinical practice.⁵ Currently, the diagnosis of BAV thrombosis relies on the identification of morphologic changes in valve leaflets and hemodynamic changes caused by thrombus with the use of computed tomography (CT) and echocardiography rather than direct visualization of thrombus itself.⁶ Furthermore, the impact of subclinical BAV thrombosis on valve durability and function remains unknown.⁷ Given its utility in identifying in vivo thrombosis, ¹⁸F-GP1 PET may provide a more specific means of distinguishing the presence of thrombosis from other causes of BAV degeneration and dysfunction as well as a novel means of determining the impact of subclinical thrombosis on BAV durability.

In this issue of JACC: Cardiovascular Imaging, Bing et al⁸ evaluated the utility of ¹⁸F-GP1 for the assessment of BAV thrombosis by assessing: 1) ex vivo ¹⁸F-GP1 uptake by activated platelets in explanted BAVs; 2) in vivo ¹⁸F-GP1 uptake in aortic valves on PET imaging in a prospective cross-sectional cohort of 75 patients with BAVs (n = 53; median duration of 37 months since implantation) and native valves (n = 22 with recent myocardial infarction but no aortic valve disease) who are not on anticoagulation; and 3) a proof-of-concept in vivo case study of 2 individuals with symptomatic BAV thrombosis. The ex vivo experiments demonstrated highly selective binding of ¹⁸F-GP1 to activated platelet GP IIb/IIIa receptors and adherent thrombus on all explanted BAVs. This finding was confirmed by histology, immunohistochemistry, and autoradiography, which are elegantly presented by the authors. Notably, there was no uptake in control bioprosthetic valves or in areas of valvular fibrosis or calcification.

The in vivo study expanded on these findings. Those in the cross-sectional study were largely asymptomatic aside from 1 individual with suspected BAV thrombosis. All patients with BAVs demonstrated ¹⁸F-GP1 uptake that was inversely related to time since implantation, whereas no uptake was observed in native aortic valves. Furthermore, ¹⁸F-GP1 uptake largely localized to BAV leaflets (rather than frames and stents) and positively associated with confirmed or suspected valve thrombosis, all showing hypoattenuated leaflet thickening on CT angiography, a nonspecific finding of valve thickening that is currently used to assess for BAV thrombosis.⁶ An additional patient with thrombus external to the valve frame was identified with the use of ¹⁸F-GP1 PET and CT angiography. Notably, ¹⁸F-GP1 uptake was not associated with transthoracic echocardiography (TTE) findings, and TTE identified valve thrombosis in only 2 of the 3 patients identified by means of ¹⁸F-GP1 PET and CT angiography (ie, those in the proof-of-concept cohort). Moreover, the patients who were diagnosed as having BAV thrombosis with abnormal ¹⁸F-GP1 uptake and were treated with anticoagulation for 3 months had improvements in valve gradients, resolution of hypoattenuated leaflet thickening, and reduced radiotracer uptake on reevaluation, suggesting an additional role for ¹⁸F-GP1 in therapeutic monitoring.

The in vivo study also demonstrated nonvalvular uptake of ¹⁸F-GP1 that identified other sites of thrombus. In individuals within the study population, uptake was observed in ascending aortic grafts, a retained guidewire from a complication of a coronary intervention, pacemaker wires, a vascular plug that remained after repair of a paravalvular leak, and within the myocardium in an area of recent infarct. These findings suggest a range of applications in which ¹⁸F-GP1 may prove to be useful for the identification of thrombosis in which other imaging modalities are contraindicated or not definitive.⁹

The results of the study offer mechanistic insights into the biology of BAV endothelialization, degeneration, and thrombosis. The inverse relationship between the presence of ¹⁸F-GP1 uptake and the length of time after surgery illustrates that low-level platelet activation during endothelialization is quite common and appears to affect all models of BAV. Over time, this uptake should decline; however, in the presence of valvular dysfunction, ¹⁸F-GP1 uptake suggests BAV thrombosis rather than downstream fibrosis or calcification as a contributor.¹⁰ Individuals with prolonged BAV ¹⁸F-GP1 uptake could be studied to identify valve- and patient-related factors that are associated with enhanced thrombogenicity. Unfortunately, the cross-sectional study design does not offer insights into the natural history of ¹⁸F-GP1 uptake in BAVs. Is the low-level ¹⁸F-GP1 uptake a universal indolent finding in BAVs without serious clinical implications, or is it an early predictor of future complications that may compromise the durability and function of the BAV? Would systemic anticoagulation in individuals with prolonged ¹⁸F-GP1 uptake after valve implantation be beneficial? The answers to these questions require additional studies in larger cohorts with longer follow-up and various types of BAV degeneration to fully determine the clinical role of ¹⁸F-GP1 PET in BAV thrombosis and degeneration and its impact on clinical end points.

The study has several other important limitations. It was observational in design. The sample size of the prospective cohort was modest and included a variety of different bioprosthetic valves (eg, both transcatheter and surgical versions of various models). Although the tracer has improved target-to-background ratio and clearance compared with other radiotracers targeting thrombosis, the study design did not allow for true validation with in vivo imaging compared with ex vivo evaluation of the same BAV.³ Furthermore, normative values for BAV ¹⁸F-GP1 uptake remain unknown. Therefore, the values observed in this study cannot be compared with an established standard.

This study suggests a potential adjunctive role for ¹⁸F-GP1 PET imaging for the assessment of BAV thrombosis beyond standard imaging with the use of echocardiography and CT angiography. However, because ¹⁸F-GP1 uptake is common even in patients without evidence of valve dysfunction, it is necessary to interpret the findings in the context of the clinical presentation until there is further evaluation of the clinical implications. The results also suggest that ¹⁸F-GP1 PET may be useful for determining the role of thrombosis in BAV degeneration and monitoring the impact of anticoagulation in the setting of BAV thrombosis. Finally, the recognition of uptake in other regions suggests a potential role for ¹⁸F-GP1 PET in the assessment of thrombotic events when other imaging modalities are not definitive or are contraindicated.