Newby, David

Dweck, Marc

Tzolos, Evangelos

Miriam & Sheldon G. Adelson Medical Research Foundation

British Heart Foundation

2022-11-15T14:29:18Z

2022-11-15T14:29:18Z

2022-11-15

BACKGROUND: Calcification is one of the body’s primary responses to injury and a key pathological feature of cardiovascular disease. Calcification activity can now be imaged using 18F-sodium-fluoride positron emission tomography and computed tomography which targets microcalcification (diameter <50 μm), providing different information to established macroscopic calcification (minimum diameter 200 to 500 μm) imaged with computed tomography. Indeed 18F-sodium fluoride PET has now been explored in a wide range of conditions including coronary atherosclerosis, aortic stenosis and bioprosthetic valve disease, offering a marker of vascular injury and disease activity in each condition studied. At the moment, we lack reliable and reproducible methods for predicting myocardial infarction in patients with established coronary artery disease. Moreover, there remain major uncertainties regarding disease activity within the retained native aortic valve as well as bioprosthetic valve durability following transcatheter aortic valve implantation. In this thesis, I will address the following questions: 1) What is the optimal method for quantifying microcalcification activity and how reproducible is it? 2) Can we further refine it? 3) Can 18F-sodium fluoride PET/CT predict myocardial infarction in patients with advanced atherosclerosis? 4) Can machine learning improve the ability of 18F-sodium fluoride PET/CT to predict events? 5) Can 18F-sodium fluoride PET/CT predict native aortic valve disease progression and bioprosthetic valve degeneration in patients with transcatheter aortic valve implantation? METHODS AND RESULTS: Study one: 18F-sodium fluoride reproducibility We aimed to establish the observer repeatability and interscan reproducibility of coronary 18F-sodium fluoride positron emission tomography (PET) uptake using a novel semi-automated approach, coronary microcalcification activity (CMA). Nineteen patients with established coronary artery disease were included for analysis. CMA was assessed twice in 43 coronary vessels on two PET/CT scans performed 12±5 days apart. We demonstrated that CMA is a repeatable and reproducible global measure of coronary atherosclerotic activity. There was excellent intraclass correlation for intraobserver and interobserver repeatability as well as interscan reproducibility (all ≥0.991). There was 100% intraobserver, interobserver and interscan agreement for the presence (CMA>0) or absence (CMA=0) of coronary18F-sodium fluoride uptake. Mean CMA was 3.12± 0.62 with coefficients of repeatability of 0.24 and 0.22 for intraobserver, 0.30 and 0.29 for interobserver and 0.33 and 0.32 for interscan analysis at a per-vessel and per-patient level respectively. Study two: Refining 18F-sodium fluoride uptake We aimed to evaluate the impact of respiratory-averaged computed tomography attenuation correction (RACTAC) compared to standard single-phase computed tomography attenuation correction (CTAC) map, on the quantitative measures of coronary atherosclerotic lesions of 18F-sodium fluoride in PET/CT. Twenty-three patients with calcified plaques in the coronary arteries were included. We evaluated 34 coronary lesions using CTAC and RACTAC reconstructions. We demonstrated that respiratory-averaged and standard single-phase attenuation correction maps provide similar and reproducible methods of quantifying coronary 18F-sodium fluoride uptake on PET/CT. Assessments of the individual vessel microcalcification burden (CMA) revealed no major differences between PET images reconstructed using CTAC and RACTAC scans (median [IQR] CMA: CTAC = 0.10 [0-1.0], RACTAC = 0.15 [0-1.03], p=0.19). Bland-Altman plots of the CMA values revealed a high degree of agreement when comparing the per vessel burden, with coefficient of reproducibility of 0.17. Study three: Coronary 18F-sodium fluoride PET/CT In a multicentre study, we investigated whether coronary 18F-sodium fluoride PET uptake predicts future myocardial infarction in patients with established coronary artery disease, and whether it can provide additional prognostic information over and above current methods of risk stratification including clinical risk scores, coronary calcium scoring and the severity of obstructive coronary artery disease. We studied 293 participants (65±9 years; 84% male) with established coronary artery disease. Patients underwent 18F-sodium fluoride PET/CT and were followed-up for fatal or non-fatal myocardial infarction over 42 [31-49] months. Total coronary 18F-sodium fluoride uptake was determined using coronary microcalcification activity (CMA). Fatal or non-fatal myocardial infarction occurred only in patients with increased coronary 18F-sodium fluoride activity (20/203 CMA>0 versus 0/90 CMA=0; p<0.001). On receiver operator-curve analysis, fatal or non-fatal myocardial infarction prediction was highest for 18F-sodium fluoride CMA, outperforming coronary calcium scoring, modified Duke coronary artery disease index, cardiac risk scores (areas under curve: 0.76 versus 0.54, 0.62, 0.52 and 0.54; p<0.001 for all). Patients with CMA>1.56 had >7-fold increase in fatal or non-fatal myocardial infarction (hazard ratio 7.1, 95% confidence interval 2.2 to 25.1; p=0.003) independent of age, gender, risk factors, segment involvement and coronary calcium scores, presence of coronary stents, coronary stenosis, cardiac risk scores, the Duke coronary artery disease index and recent myocardial infarction. Study four: Artificial intelligence in 18F-sodium fluoride PET/CT We undertook a post-hoc analysis of the population of patients from study three. Coronary 18F-sodium fluoride PET/CT and CT angiography-based quantitative plaque analysis have shown promise in refining risk stratification in patients with coronary artery disease. We combined both novel imaging approaches to develop an optimal machine-learning model for the future risk of myocardial infarction in patients with stable coronary disease. We demonstrated that our machine learning approach has overcome the challenges posed by co-linearity of these variables and, for the first time, demonstrated that this information is complementary and additive with the combination of both providing the most robust outcome prediction. On univariable receiver-operator-curve analysis, only 18F-sodium fluoride coronary uptake emerged as a predictor of myocardial infarction (c- statistic 0.76, 95% confidence interval [CI] 0.68-0.83). When incorporated into machine-learning models, clinical characteristics showed limited predictive performance (c-statistic 0.64, 95% CI 0.53-0.76;) and were outperformed by a quantitative plaque analysis-based machine-learning model (c-statistic 0.72, 95% CI 0.60-0.84). After inclusion of all available data (clinical, quantitative plaque and 18F-sodium fluoride PET), we achieved a substantial improvement (p=0.008 versus 18F-sodium fluoride PET alone) in the model performance (c-statistic 0.85, 95% CI 0.79-0.91). Study five: 18F-sodium fluoride in transcatheter aortic valves We undertook a multicentre cross-sectional observational cohort study to determine whether the retained native aortic valves in patients undergoing transcatheter aortic valve implantation (TAVI) demonstrate evidence of ongoing disease progression. Additionally, since long-term durability of transcatheter aortic valves is yet to be established, we aimed to establish whether bioprosthetic valve durability or degeneration was appreciably different from patients with surgical aortic valve replacement (SAVR). Patients with TAVI or bioprosthetic SAVR underwent baseline echocardiography, CT angiography and 18F-sodium fluoride PET/CT. Subsequently patients underwent serial echocardiography to assess for changes in valve hemodynamic performance (change in peak aortic velocity) and evidence of structural valve dysfunction. Comparisons were made to matched patients with bioprosthetic SAVR who had undergone the same imaging protocol. We enrolled 47 patients (81±6 years old, 79% male) with TAVI from 3 high volume centres. We demonstrated that 18F-sodium fluoride uptake within the native aortic valve is higher with longer duration of implantation suggesting disease activity continues despite immobilisation of the valve leaflet. We have further shown using 3 complementary and distinct imaging modalities that the prevalence of valve degeneration within TAVI bioprostheses are similar to that of bioprosthetic SAVR valves for up to 7 years after valve replacement. Finally, we have confirmed that 18F-sodium fluoride PET of the bioprosthetic valve provides a powerful independent predictor of subsequent hemodynamic bioprosthetic valve degeneration that is applicable to both TAVI and SAVR and outperforms all other traditional risk factors. In patients with TAVI, native aortic valves demonstrated 18F-sodium fluoride uptake around the outside of the bioprostheses which showed a modest correlation with the time from TAVI (r=0.36, p=0.023). 18F-sodium fluoride uptake in the bioprosthetic leaflets was comparable between the SAVR and TAVI groups (target-to-background ratio 1.3 [1.2-1.7] versus 1.3 [1.2-1.5] respectively, p=0.27). The frequencies of imaging evidence of bioprosthetic valve degeneration at baseline were similar on echocardiography (6% versus 8% respectively, p=0.78), CT (15% versus 14% respectively, p=0.87) and PET (15% versus 29% respectively, p=0.09). Baseline 18F-sodium fluoride uptake was associated with subsequent change in peak aortic velocity for both TAVI (r=0.7, p<0.001) and SAVR (r=0.7, p<0.001). On multivariable analysis, 18F-sodium fluoride uptake was the only predictor of peak velocity progression (p<0.001). CONCLUSIONS: We have demonstrated that CMA is a repeatable and reproducible global measure of coronary atherosclerotic activity. Respiratory-averaged and standard single-phase attenuation correction maps provide similar and reproducible methods of quantifying coronary 18F-sodium fluoride uptake on PET/CT. In patients with established coronary artery disease, 18F-sodium fluoride PET provides powerful independent prediction of fatal or non-fatal myocardial infarction. Both 18F-sodium fluoride uptake and quantitative plaque analysis measures are additive and strong predictors of outcome in patients with established coronary artery disease. Optimal risk stratification can be achieved by combining clinical data with these approaches in a machine-learning model. Finally, in patients with TAVI, native aortic valves demonstrate evidence of ongoing active disease. Across imaging modalities, TAVI degeneration appears to be of similar magnitude to bioprosthetic SAVR suggesting comparable mid-term durability.

https://hdl.handle.net/1842/39478

http://dx.doi.org/10.7488/era/2728

en

The University of Edinburgh

Tzolos E, Andrews JP, Dweck MR. Aortic valve stenosis-multimodality assessment with PET/CT and PET/MRI. Br J Radiol. 2020 Sep 1;93(1113):20190688. doi: 10.1259/bjr.20190688

Tzolos E, Dweck MR Imaging Cardiovascular Calcification Activity with 18F-sodium fluoride PET, Cardiovascular Calcification and Bone Mineralisation, Springer 2020

Tzolos E, McElhinney P, Williams MC, et al. Repeatability of quantitative pericoronary adipose tissue attenuation and coronary plaque burden from coronary CT angiography. J Cardiovasc Comput Tomogr. 2021 Jan-Feb;15(1):81-84. doi: 10.1016/j.jcct.2020.03.007

Tzolos E, Dweck MR. 18F-Sodium Fluoride (18F-NaF) for Imaging Microcalcification Activity in the Cardiovascular System. Arterioscler Thromb Vasc Biol. 2020 Jul;40(7):1620-1626. doi: 10.1161/ATVBAHA.120.313785.

Tzolos E, Newby DE, Dweck MR. Aortic Stenosis & Bioprosthetic Valve Degeneration, Nuclear Cardiology and Multimodal Cardiovascular Imaging, E-Book: A Companion to Braunwald's Heart Disease, Elsevier 2022

Kwiecinski J, Tzolos E, Cartlidge TRG, Fletcher A, Doris MK, Bing R, et al. Native Aortic Valve Disease Progression and Bioprosthetic Valve Degeneration in Patients with Transcatheter Aortic Valve Implantation. Circulation. 2021

Kwiecinski J, Tzolos E, Meah M, Cadet S, Adamson PD, Grodecki K, et al. Machine-learning with (18)F-sodium fluoride PET and quantitative plaque analysis on CT angiography for the future risk of myocardial infarction. J Nucl Med. 2021

Zheng KH, Tzolos E, Dweck MR. Pathophysiology of Aortic Stenosis and Future Perspectives for Medical Therapy. Cardiol Clin. 2020;38(1):1-12

Tzolos E, Dweck MR. Threshold effect for lipoprotein(a) in aortic stenosis. Heart. 2021;107(17):1367-8

Tzolos E, Bing R, Newby DE, Dweck MR. Categorising myocardial infarction with advanced cardiovascular imaging. Lancet. 2021;398(10299):e9

Tzolos E, Bing R, Andrews J, Macaskill M, Tavares A, MacNaught G, et al. In vivo coronary artery thrombus imaging with 18F-GP1 PET-CT. European Heart Journal. 2021;42(Supplement_1)

18F-sodium fluoride

microcalcification

Positron emission tomography

breathing tracing

calcification activity

Novel positron emission tomography applications for imaging cardiovascular disease activity

Thesis or Dissertation

Doctoral

PhD Doctor of Philosophy