Whole-vessel coronary ¹⁸F-sodium fluoride PET for assessment of the global coronary microcalcification burden

Abstract

Purpose

¹⁸F-sodium fluoride (¹⁸F-NaF) has shown promise in assessing disease activity in coronary arteries, but currently used measured of activity – such as maximum target to background ratio (TBRmax) -are defined by single pixel count values. We aimed to develop a novel coronary-specific measure of ¹⁸F-NaF PET reflecting activity throughout the entire coronary vasculature (coronary microcalcification activity [CMA]),

Methods

Patients with recent myocardial infarction and multi-vessel coronary artery disease underwent ¹⁸F-NaF PET and coronary CTA. We assessed the association between coronary ¹⁸F-NaF uptake (both TBRmax and CMA) and coronary artery calcium scores (CACS) as well as low attenuation plaque (LAP, attenuation <30 Hounsfield units) volume.

Results

In 50 patients (64% males, 63±7 years) CMA and TBRmax were higher in vessels with LAP compared to those without LAP (1.09 [0.02, 2.34] versus 0.0 [0.0, 0.0], p<0.001 and 1.23 [1.16, 1.37] versus 1.04 [0.93, 1.11], p<0.001). Compared to a TBRmax threshold of 1.25, CMA > 0 had a higher diagnostic accuracy for detection of LAP: sensitivity of 93.1 (83.3–98.1)% versus 58.6 (44.9–71.4)% and a specificity of 95.7 (88.0–99.1)% versus 80.0 (68.7–88.6)% (both p<0.001).

¹⁸F-NaF uptake assessed by CMA correlated more closely with LAP (r=0.86,p<0.001) than the CT calcium score (r=0.39,p<0.001), with these associations outperforming those observed for TBRmax values (LAP r=0.63,p<0.001; CT calcium score r=0.30,p<0.001).

Conclusions

Automated assessment of disease activity across the entire coronary vasculature is feasible using ¹⁸F-NaF CMA, providing a single measurement that has closer agreement with CT markers of plaque vulnerability than more traditional measures of plaque activity.

Introduction

Cardiovascular atherosclerotic disease is the leading cause of death globally [1]. In the past decade, enormous efforts have been undertaken to enable the detection of high-risk coronary artery plaques. While identification of high-risk lesions has traditionally relied on structural intravascular coronary artery imaging or computed tomography (CT) angiography, recently the assessment of plaque biological activity and vulnerability has gained major interest [2–8]. In particular, positron emission tomography (PET) has provided insight into coronary inflammation and micro-calcification activity [9–11]. To date, PET image analysis was typically performed using a maximum activity approach where the reader delineated volumes of interest only around areas with visually increased tracer activity, and measures of maximal activity recorded. However, this method is sub-optimal [12]. Joshi et al. showed that the difference in the ¹⁸F-sodium fluoride (¹⁸F-NaF) maximum target-to-background ratio (TBRmax) in culprit compared with non-culprit plaques is only 34% ¹¹. Additionally, the per plaque TBRmax approach relies on labor-intensive and subjective visual detection of lesions and does not truly reflect disease activity across the entire coronary vasculature. An alternative method for the assessment of ¹⁸F-NaF uptake is the global heart approach which measures activity across the whole heart region [13,14]. The utility of this method, however, is compromised by the inclusion of activity from foci of non-coronary uptake (including the aortic valve and mitral valve annulus) as well as blood pool noise [15].

We here propose a novel semiautomated approach investigating ¹⁸F-NaF uptake throught the entire coronary vasculature using centerlines defined by coronary CT angiography to build 3-dimensional tubular volumes of interest around each of the main epicardial coronary arteries. This method allows evaluation of coronary ¹⁸F-fluoride activity on a per-vessel and per-patient basis, providing more global assessments of disease activity in the coronary arteries. In the initial validation of this new measure, we compare it with the traditional TBRmax approach and assess its relationship with established anatomical quantitative risk indices from CT angiography including low attenuation plaque and coronary artery calcium scoring [2,3].

Methods

Patient population

The study was performed on subjects recruited for the ongoing PRE¹⁸FFIR (Prediction of Recurrent Events with ¹⁸F-Fluoride to Identify Ruptured and High-risk Coronary Artery Plaques in Patients with Myocardial Infarction) trial ( NCT02278211). Inclusion required the presence of multi-vessel coronary artery disease on invasive angiography, following recent (within 4 weeks) myocardial infarction. We excluded patients who had a known allergy to iodinated contrast or renal impairment (estimated glomerular filtration rate ≤30 mL/min/1.73 m²) and women of childbearing potential. The study was approved by the local institutional review board, the Scottish Research Ethics Committee (REC reference: 14/SS/0089 and 15/SS/0203), and the United Kingdom (UK) Administration of Radiation Substances Advisory Committee. It was performed in accordance with the Declaration of Helsinki. All patients provided written informed consent prior to any study procedures.

PET and CT acquisition and reconstruction

All patients underwent ¹⁸F-NaF PET on a hybrid PET-CT scanner (128-slice Biograph mCT, Siemens Medical Systems, Knoxville, USA). All imaging was performed after invasive angiography and revascularization. Prior to imaging, subjects were administered with a target dose of 250 MBq of ¹⁸F-NaF and rested in a quiet environment for 60 minutes. During a single imaging session, we acquired a non-contrast CT attenuation correction scan followed by a 30-min PET emission scan in list mode. The electrocardiography-gated list mode dataset was reconstructed using a standard ordered expectation maximization algorithm with time-of-flight, and point-spread-function correction. Using 10 cardiac gates, we reconstructed the data on a 256×256 matrix (109 slices, slice thickness 2.027mm) using 2 iterations, 21 subsets and 5-mm Gaussian smoothing.

Immediately after the PET scan, a low dose non-contrast (40mAs/rot, 100kV) CT for calculation of the coronary calcium score was performed. Subsequently, a contrast-enhanced coronary CT angiogram was obtained in the same PET/CT system without repositioning the patient. The CT angiography acquisition was conducted according to societal guidelines with the following settings: 330 ms rotation time, 100 or 120kV (depending on body mass index), 160–245 mAs and reconstructed with 0.6 slice thickness and a medium-soft convolution kernel [16].

To compensate for coronary motion associated with heart contraction, we performed cardiac motion correction of the PET/CT images. This technique enables alignment of all gates to the end-diastolic position and as a result allows for inclusion of all PET counts acquired [17].

Image analysis

CT

The Agatston coronary artery calcium score was measured using semiautomatic commercial software (NetraMD, ScImage, Los Altos, CA, USA)[18]. Quantitative plaque analysis of all coronary segments with a lumen diameter ≥2 mm was performed using semi-automated software (AutoPlaque version 2.0, Cedars-Sinai Medical Center, Los Angeles, USA)[19]. An experienced reader examined coronary CT angiography images in multiplanar format and manually determined the proximal and distal limits of lesions. Subsequent plaque quantification was automated using adaptive scan-specific thresholds. The quantitative presence of LAP was defined as a LAP volume > 0 mm³. The software quantified volumes of the total plaque (TPV), non-calcified plaque (NCP), low-density non-calcified plaque (low attenuation plaque - LAP, attenuation <30 Hounsfield units) and calcified plaque (CP).

PET

We used a dedicated software package for coronary PET image analysis (FusionQuant, Cedars Sinai Medical Center, Los Angeles)[20]. PET and CT angiography reconstructions were reoriented, fused and systematically co-registered in all 3 planes [15]. To evaluate ¹⁸F-NaF activity, we used the previously established TBRmax approach (standard quantification) which relies on visual detection of lesions with increased tracer uptake, as well as the newly developed per vessel (and per patient) whole-coronary total microcalcification activity method.

Standard TBRmax quantification

For a signal to be co-localized to a coronary artery, an atherosclerotic plaque had to be present on the CT angiography and the increased pattern of radiotracer had to arise from the coronary artery and follow its course over >5 mm in three dimensions on orthogonal views [21]. In all plaques which met these criteria, the maximum standardized uptake values (SUVmax) were measured within regions of interest drawn around these areas. To characterize ¹⁸F-NaF coronary uptake in the context of background activity we divided the maximum standard uptake values (SUVmax) by the blood pool activity (mean standard uptake values, SUVmean) measured in the center of the left atrium by delimiting spherical volumes of interest (radius=9–13mm depending on the cavity size) and produced the maximum target to background ratios (TBRmax). We used a 1.25 TBRmax threshold to categorise vessels/plaques as being positive or negative for ¹⁸F-NaF uptake [11].

Whole coronary ¹⁸F-NaF activity assessment

We devised a novel coronary ¹⁸F-NaF uptake measure based upon the Agatston method for quantifying coronary CT calcium scores and PET quantification techniques widely employed in oncology and cardiac sarcoidosis [22–24]. First, we automatically extracted whole-vessel tubular and tortuous 3D volumes of interest (VOIs) from CT angiography datasets (Figures 1 and S1). These encompass all the main epicardial coronary vessels and their immediate surroundings (4-mm radius) facilitating per-vessel and per-patient uptake quantification. For this study we evaluated ¹⁸F-NaF activity along the entire course of coronary arteries regardless of the presence of coronary stents and we included the left main in the left anterior descending artery VOI. Within such VOIs, we measured the whole vessel SUVmax, and used these to derive TBR values after correction for blood pool. In addition, we explored our novel parameter - the coronary microcalcification activity (CMA)— representing the overall disease activity in the vessel and based upon both the volume and intensity of ¹⁸F-fluoride PET activity within it. CMA was defined as the integrated activity in SUV units exceeding the background blood-pool SUVmean + 2 standard deviations (left atrium activity). The per patient CMA was defined as the sum of the per vessel CMA.

Figure 1.

3-dimensional rendering of coronary CT angiography with superimposed tubular whole vessel volumes of interest (light green) employed for evaluation of ¹⁸F-sodium fluoride uptake (blue and red). Despite the relatively lower tissue-to-background ratio maximum (TBRmax) due to multiple foci of increased ¹⁸F-NaF activity, the coronary microcalcification activity (CMA) in the right coronary artery (RCA) is only moderately lower than in the left anterior descending (LAD) coronary artery which presented with a very high TBRmax.

For comparisons between the standard TBRmax and novel whole-coronary quantification approach, we primarily used the maximal standard TBRmax value and CMA at both the per vessel and per patient levels. Analysis of the other quantification measures provided by the whole vessel technique and the standard approach are outlined in the Data Supplement.

Statistical analysis

Data were tested for normality using Shapiro Wilks test. Continuous data is expressed as mean (standard deviation) or median [interquartile range] dependent on the distribution. Skewed continuous variables were log transformed to achieve normal distribution. Data were compared with the two-sample t-test. We assessed the associations between two continuous variables using the Pearson’s or Spearman’s correlation coefficients where appropriate. Correlation coefficients were compared using Fisher’s z-transformation. We compared the diagnostic accuracy of uptake measures in detecting LAP with the McNemar test. We used the receiver-operating characteristic analysis and pairwise comparisons according to DeLong et al. to compare the performance of TBRmax and CMA in detecting LAP. A 2-sided p value <0.05 was regarded as significant. Statistical analysis was performed with SPSS software (version 24, SPSS, Inc., Chicago, Illinois).

Results

Patients

Fifty patients (64% males, mean age 63±7 years) with recent (range: 8–20 days) myocardial infarction comprised the study cohort (Table 1). All patients had multivessel coronary artery disease and a history of coronary revascularization. Twenty-six subjects (52%) underwent revascularization of more than one vessel.

Table 1.

Baseline Characteristics of Study Patients.

Baseline Characteristics
Age, years	63±7
Males, n (%)	32 (64%)
Diabetes, n (%)	4 (8%)
Hyperlipidemia, n (%)	40 (80%)
Hypertension, n (%)	39 (78%)
Tobacco use, n (%)	4 (8%)
Family history of CAD, n (%)	12 (24%)
Medications
Aspirin, n (%)	19 (38%)
Clopidogrel, n (%)	10 (20%)
Statin, n (%)	23 (46%)
ACEI/ARB, n (%)	17 (34%)
Beta Blocker, n (%)	8 (16%)
Coronary Computed Tomography Angiography
Segment involvement score	6 [5, 8]
Multivessel disease, n (%)	50 (100%)
Coronary artery calcium score	113 [90, 387]
Total plaque volume (mm3)	963 [670, 1120]
NCP volume (mm3)	858 [574, 1041]
Low attenuation plaque (mm3)	9 [0.0, 14]
PET/CT
SUVmax	1.5 [1.2, 1.6]
TBRmax	1.3 [1.1, 1.4]
CMA	1.2 [0.3, 1.7]
Background SUVmean	1.1 [1.0, 1.2]

CAD – coronary artery disease, CMA – coronary microcalcification activity, HDL – high-density lipoprotein, LDL – low-density lipoprotein, NCP – non-calcified plaque, SUVmaxSUVmean – maximum/mean standard uptake value, TBRmax – maximum target to background ratio

Coronary CT angiography

Of the quantitative coronary plaque composition indices, we observed a trend for higher non-calcified plaque (941.7 [632.6, 1552.8] mm² versus 839.8 [388.9, 1178.1] mm², p=0.17) and total-plaque volumes (1018.3 [822.6, 1670.3] mm² versus 897.1 [408.4, 1253.2] mm², p=0.13) in patients with TBRmax exceeding 1.25. These subjects also showed a trend for higher coronary artery calcium scores (374 [193, 820] versus 213 [88, 466], p=0.07) and had greater LAP volumes (17.9 [3.5, 33.3] mm² versus 1.4 [0.4, 4.7] mm², p<0.001) than those with TBRs that did not reach the predefined threshold.

Positron Emission Tomography

On standard quantification, 41 (82%) subjects had more than one plaque with visually detectable uptake (overall number of plaques with increased activity on visual assessment, range: 0 to 8). Twenty-eight (56%) patients showed uptake exceeding the 1.25 TBRmax threshold. CMA values were higher in patients who met the predefined TBRmax threshold (5.8 [3.5, 10.9] versus 0.2 [0.0, 2.0] p<0.001). There was a strong correlation between CMA and TBRmax values (r=0.76, p<0.001), but CMA had a wider range of values than TBRmax (0 to 12.13 versus 0.80 to 2.44; Figures 2 and 3)

Figure 2.

Coronary microcalcification activity (CMA) versus maximum target-to-background ratio (TBRmax) values of ¹⁸F-sodium fluoride uptake. Per vessel (n=142) TBRmax (A) and CMA (B) values in arteries with and without low attenuation plaque (LAP) on quantitative plaque analysis. Per patient (n=50) TBRmax (C) and CMA (D) values in patients with and without LAP on quantitative coronary plaque analysis.

Figure 3.

Case examples of coronary ¹⁸F-sodium fluoride uptake, low attenuation plaque (LAP) on CT angiography and LAP (orange arrows, II) on 3D CT rendering (III). Right (Patient 1) and left anterior descending (Patient 2) coronary arteries. Areas of LAP correspond with areas of increased ¹⁸F-NaF activity. While maximum target-to-background ratio (TBRmax) reflects single hot-spot activity coronary microcalcification activity (CMA) values represent the whole-vessel ¹⁸F-sodium fluoride burden.

Associations of CT derived plaque characteristics and ¹⁸F-NaF uptake

Vessels with LAP presented with higher CMA than those without LAP (1.09 [0.02, 2.34] versus 0.0 [0.0, 0.0], p<0.001). While vessel TBRmax was also different in arteries with and without LAP, we observed considerable overlap in uptake values: 1.23 [1.16, 1,37] versus 1.04 [0.93, 1.11], p<0.001 for TBRmax, (Figure 2). Similarly, at the patient level, subjects with LAP presented with higher CMA (3.5 [1.8, 6.5] versus 0.0 [0.0, 0.4], p<0.001) and TBRmax values (1.25 [1.11, 1.46] versus 1.05 [0.95, 1.10], p<0.001 than those without LAP.

In a per-vessel analysis, compared to the 1.25 TBRmax threshold, a CMA greater than 0 had a higher diagnostic accuracy for detection of LAP with a sensitivity of 93.1 (83.3–98.1)% versus 71.6 (56.9–79.4)% and a specificity of 95.7 (88.0–99.1)% versus 80.0 (68.7–88.6)% (both p<0.001; Tables 2 and 3, Figure 4). Compared to TBRmax and coronary artery calcium score, CMA also had a higher area under the ROC curve for detection of LAP (0.95, 95% CI: 0.90 to 0.99 versus 0.85, 95% CI: 0.80 to 0.90 and 0.67, 95% CI: 0.60 to 0.74 respectively; p<0.001).

Table 2.

The discriminatory performance of ¹⁸F-NaF uptake thresholds for distinguishing vessels with LAP on quantitative plaque analysis. TBRmax 1.25 threshold (A). CMA greater than 0 (B). (B).

A
	TBRmax>1.25	TBRmax<1.25
LAP>0	34	24	58
LAP=0	14	56	70
	48	80
B
	CMA>0	CMA=0
LAP>0	54	4	58
LAP=0	3	67	70
	57	71

Table 3.

The sensitivity, specificity, positive and negative predictive value of the 1.25 maximum target to background (TBRmax) threshold and coronary microcalcification activity (CMA) greater than 0 for predicting low attenuation plaque (LAP) on quantitative plaque analysis. CMA had a significantly higher sensitivity, specificity, positive and negative predict value than the TBRmax threshold (all p<0.001).

18F-NaF uptake measure	Sensitivity (95% CI)	Specificity (95% CI)	Positive predictive value (95% CI)	Negative predictive value (95% CI)
TBRmax	71.6 (56.9–79.4)	80.0 (68.7–88.6)	70.8 (59.2–80.3)	69.7 (62.7–76.4)
CMA	93.1 (83.3–98.1)	95.7 (88.0–99.1)	94.7 (85.6–98.2)	94.4 (86.7–97.7)

CI – confidence interval;

Figure 4.

The diagnostic performance of coronary microcalcification activity (CMA) and maximum target-to-background (TBRmax) in detecting low attenuation plaque (LAP). (A) The sensitivity, specificity of the 1.25 TBRmax threshold and CMA greater than 0 for predicting LAP on quantitative plaque analysis. CMA had a higher sensitivity, specificity than the TBRmax threshold (all p<0.001). (B) Receiver operator curves for detection of LAP using CMA, TBRmax and coronary artery calcium scores. CMA had a higher area under the curve (AUC) than TBRmax and CCS (p<0.001).

Of the quantitative plaque characteristics, only LAP volume and coronary artery calcium score correlated with ¹⁸F-NaF uptake. In particular, we observed a strong correlation between LAP volume and CMA at both the per vessel (r=0.84, p<0.001) and per patient (r=0.86, p<0.001) levels (Figure 5). In contrast, the associations between LAP volume and TBRmax were only moderate (per-vessel r=0.56, p<0.001; per-patient r=0.63, p<0.001 respectively). Coronary artery calcium score was only moderately associated with TBRmax and CMA both on the per vessel (r=0.30, p=0.001 versus r=0.39, p<0.001) and per patient level (r=0.38, p=0.01 versus r=0.43, p=0.003). Associations between LAP volume and, SUVmax, averaged-SUVmax, and averaged-TBRmax were all weaker than with CMA and TBRmax both at the per vessel and per patient levels (data supplement results and Figures S2–4).

Figure 5.

Coronary ¹⁸F-sodium fluoride uptake and low attenuation plaque volume (LAP) on quantitative coronary plaque analysis. Per vessel: (A) maximum target-to-background (TBRmax) and LAP volume. (B) Coronary microcalcification activity (CMA) and LAP volume. Per patient: (C) TBRmax and LAP volume. (D) CMA and LAP volume.

Discussion

In this study, we describe a novel, dedicated approach for measuring disease activity across the coronary vasculature. By leveraging CT angiography-derived whole-vessel 3-dimensional volumes of interest, we were able to express ¹⁸F-NaF PET tracer activity as the global CMA mimicking an approach successfully implemented in the field of oncology and cardiac sarcoidosis [22–24]. Indeed, CMA represents both the volume and intensity of coronary ¹⁸F-NaF activity, similar to the Agatston method for CT calcium scoring. Rather than being limited to reporting the highest SUV, TBR or the number of plaques with increased tracer activity, CMA reflects the whole-coronary ¹⁸F-NaF uptake burden which can be compared with total coronary burden of various plaque components on CT angiography. As a result, in patients and vessels with multiple foci of uptake, CMA provides a continuous measure of disease activity across the coronary vasculature. The proposed approach moves away from a single hot spot approach to a patient level total ¹⁸F-NaF activity burden assessment and hence does not rely on a single-pixel-value as traditional uptake measures (TBR or SUV). Moreover, our whole vessel method omits the need for time consuming and subjective manual delineation of lesions and only requires careful co-registration of the PET and CT angiography datasets [15].

We have previously shown that TBRmax of ¹⁸F-NaF is associated with LAP [25]. Here we show that CMA outperformed the hot-spot TBRmax approach for the detection of LAP. Similarly, we found a stronger correlation between LAP and CMA than LAP and conventional measurements of ¹⁸F-NaF activity, and better categorical separation. Given that LAP has repeatedly been shown to be associated with adverse outcome, the discrepancy in the correlations and diagnostic accuracy of LAP and CMA versus TBR/SUV supports the view that in characterizing disease severity, CMA is superior than traditional uptake measures [19].

The assessment of the whole-coronary disease burden is of great importance as over the past decade the ongoing pursuit of the vulnerable plaque has proved to be elusive [2]. To date, several studies have already shown that adverse plaque features (LAP, positive remodeling) observed on CT angiography are predictors of adverse events [6–8]. Despite initial enthusiasm, the low predictive value of individual plaques progressing to cause events has emerged as the major limitation of the vulnerable plaque strategy [4,5]. In view of these limitations, advanced invasive and non-invasive imaging is moving towards distinguishing vulnerable patients rather than detecting individual vulnerable plaques [2,3].

Since coronary ¹⁸F-NaF imaging is being potentially considered as a non-invasive tool for imaging high-risk coronary artery disease, we aimed to develop a quantification approach that would be in line with the vulnerable patient rather than the vulnerable plaque paradigm. Current coronary PET reporting does not provide insights into global coronary ¹⁸F-NaF burden. By reporting only the highest TBRmax values, overall coronary activity cannot be appreciated. Indeed, this approach can be misleading as it can underestimate the ¹⁸F-NaF burden in individuals with multiple foci of uptake in their coronary vasculature (Figure 3). In addition, by identifying the coronary artery boundaries from CT angiography and confining the assessment of ¹⁸F-NaF within these boundaries, the CMA approach has the potential to reduce the subjectivity and the time required for image analysis.

The global cardiac method is an alternative, previously proposed approach for cardiac ¹⁸F-NaF analysis [12]. This technique requires VOIs to be drawn around the cardiac silhouette on consecutive axial slices of fused PET/CT images and maximum radiotracer activity is recorded for each slice. However, the global cardiac uptake method relies on the assumption that any cardiac ¹⁸F-NaF activity detected as higher than blood pool originates from the coronary vasculature. There is accumulating evidence that cardiac valves and great vessels bind ¹⁸F-NaF [20,26]. Additionally, increased myocardial uptake of ¹⁸F-NaF is often observed in the infarcted myocardium, myocardial scar tissue and cardiac amyloidosis [27–29]. All of these may falsely elevate the global ¹⁸F-NaF reading leading to overestimation of coronary ¹⁸F-NaF activity. Compared to the global cardiac method, the approach we propose herein is highly specific (as only uptake originating from coronary arteries is being quantified) and therefore provides a global measure of coronary disease activity without contamination from surrounding structures.

In oncology and in the field of cardiac sarcoidosis, similar concerns surround SUVmax reporting. It was argued that SUV measurements are prone to several patient and scan protocol factors and are only accurate when there is homogeneous uptake in the tumor [25]. As a result, new quantitative approaches have been proposed. Rather than evaluating individual lesions, these approaches shifted towards metrics that better represent the patient’s total tumor load, such as the total metabolic active tumor volume [30–32]. Importantly these new measures have been found to predict disease progression, recurrence, and death [33,34]. Ahmadian and colleagues developed a quantitative measure of ¹⁸F-fluorodeoxyglucose volume intensity called Cardiac Metabolic Activity [24]. In patients with suspected cardiac sarcoidosis, this approach was independently associated with future events. In our study, we employed a method based on this oncological approach for quantification of coronary PET. We showed that compared to the local TBR or SUV based approach, it offers better in-depth characterization of the ¹⁸F-NaF burden and is more closely associated with established CT angiography derived markers of unfavorable prognosis (LAP) than the conventional approach.

Limitations

Our study lacks outcome data and thus we were unable to validate the whole-vessel uptake measurements with clinical endpoints. Instead, we employed established high-risk quantitative plaque features evaluated on CT angiography as a surrogate. Although mechanistically attractive, correlating ¹⁸F-NaF uptake measurements with quantitative coronary lesion analysis indices is suboptimal in patients who underwent coronary revascularization. While quantification of ¹⁸F-NaF uptake is not hampered by the presence of coronary stents, the assessment of plaque composition is not feasible in stented segments of arteries. Because the study cohort was comprised of subjects with recent myocardial infarction, all subjects had a medical history of coronary interventions (Figure S5). While we have shown that delayed ¹⁸F-NaF imaging facilitates image analysis, participants in this study underwent PET imaging 1 h after tracer injection and thus our whole vessel image uptake analysis remains to be tested with delayed imaging in the future [35].

Conclusions

Automated whole-vessel ¹⁸F-NaF coronary microcalcification activity assessment with CT angiography-guided 3-dimensional volumes of interest is feasible and provides a direct measure of the disease activity in each vessel and total disease burden for the patient. This approach correlates better than traditional measures of uptake with low attenuation plaque - an established anatomical marker of vulnerability.

Abbreviations and acronyms

CAD

Coronary Artery Disease

CMA

Coronary Microcalcification Activity

CI

Confidence Intervals

CT

Computed Tomography

PET

Positron Emission Tomography

SD

Standard Deviation

SUVmax

Maximum Standard Uptake Value

TBRmax

Maximum Target to Background Ratio

¹⁸F-NaF

¹⁸F-sodium Fluoride