Cardiac Amyloid Quantification Using ¹²⁴I-Evuzamitide (¹²⁴I-P5+14) Versus ¹⁸F-Florbetapir

Abstract

BACKGROUND

Cardiac amyloid quantification could advance early diagnosis of amyloid cardiomyopathy (CMP) and treatment monitoring. However, current imaging tools are based on indirect measurements. ¹²⁴I-evuzamitide is a novel pan-amyloid radiotracer binding to amyloid deposits from multiple amyloidogenic proteins. Its ability to quantify cardiac amyloid has not yet been investigated.

OBJECTIVES

The objectives of this pilot study were to quantify myocardial ¹²⁴I-evuzamitide uptake and to compare its diagnostic value to ¹⁸F-florbetapir in participants with amyloid CMP and control subjects.

METHODS

This study included 46 participants: 12 with light-chain (AL) CMP, 12 with wild-type transthyretin (ATTRwt) CMP, 2 with hereditary amyloidosis, and 20 control subjects. All amyloidosis participants underwent positron emission tomography/computed tomography with ¹²⁴I-evuzamitide and ¹⁸F-florbetapir. Control subjects underwent ¹²⁴I-evuzamitide (n = 10) or ¹⁸F-florbetapir (n = 8) positron emission tomography/computed tomography. Left ventricular percent injected dose (LV% ID) was measured as mean activity concentration × myocardial volume/injected activity. High LV %ID was defined using Youden’s index.

RESULTS

In CMP participants, median age was 74 years and 92% were men. ¹²⁴I-evuzamitide LV %ID differed across groups: median AL-CMP 1.48 (IQR: 1.12-1.89), ATTRwt-CMP 2.12 (IQR: 1.66-2.47), and control subjects 0.00 (IQR: 0.00-0.01; overall P < 0.001). High LV %ID perfectly discriminated CMP from control subjects. Discrimination performance was similar for ¹⁸F-florbetapir LV %ID. Notably, for ATTRwt-CMP, LV %ID was higher with ¹²⁴I-evuzamitide than ¹⁸F-florbetapir (P = 0.002). ¹²⁴I-evuzamitide LV %ID was correlated with interventricular septum thickness (Spearman’s ρ = 0.78) and LV global longitudinal strain (ρ = 0.54) from echocardiography, and with LV mass index (ρ = 0.82) and extracellular volume (ρ = 0.51) from cardiac magnetic resonance.

CONCLUSIONS

¹²⁴I-evuzamitide demonstrates uptake by cardiac amyloid and accurately discriminates amyloid CMP from control subjects. In AL-CMP, discrimination performance is similar to ¹⁸F-florbetapir. In ATTRwt-CMP, performance may be better with ¹²⁴I-evuzamitide. Moderate-to-strong correlations of ¹²⁴I-evuzamitide uptake with cardiac structural and functional metrics suggest valid amyloid quantification. Hence, ¹²⁴I-evuzamitide is a promising novel radiotracer to detect and quantify cardiac amyloid.

Cardiac amyloidosis develops when amyloidogenic proteins misfold and aggregate as fibrils in the myocardial extracellular compartment, leading to restrictive cardiomyopathy (CMP), heart failure, and death.^1,2 The most frequent forms are transthyretin amyloidosis (ATTR) and light-chain amyloidosis (AL) CMP.^1,2 Wild-type transthyretin amyloidosis (ATTRwt) CMP is increasingly diagnosed as the cause of heart failure in older adults,³ and several effective therapies able to slow disease progression are now clinically available or are in clinical trials.² Although transthyretin stabilizers, silencers, and drugs preventing light-chain production do not directly remove amyloid fibrils already deposited, novel antifibril therapies are now being developed.^4,5 Thus, amyloid quantification has become the key next step to understand the effects of therapy and to potentially advance early diagnosis and treatment of cardiac amyloidosis.⁶ Currently, myocardial amyloid burden is evaluated using relatively indirect measurements.^6,7 Echocardiography assesses the structural and functional consequences of amyloid deposition, but not amyloid burden itself.^6,7 Bone-avid radiotracers with single-photon emission computed tomography (SPECT)/computed tomography (CT) are taken up in the amyloid-infiltrated heart, but uptake assessment is only semiquantitative,^8,9 and only useful for ATTR amyloidosis, not for AL or rarer forms of cardiac amyloidosis.^6,7,10 Extracellular volume (ECV) by cardiac magnetic resonance (CMR) is frequently used to estimate cardiac amyloid burden, but it cannot differentiate it from fibrosis, inflammation, or edema.⁶ More specific methods have been developed, such as beta-amyloid-binding radiotracers utilizing positron emission tomography (PET). Initially used for Alzheimer’s disease, these radiotracers have shown promise for cardiac amyloidosis.^11–17 Among them¹⁸F-florbetapir can detect cardiac amyloid even in early disease,¹⁵ but has lower myocardial uptake in ATTR-CMP than in AL-CMP,¹² and cannot reliably assess hepatic and renal amyloid.¹⁸

The novel pan-amyloid-reactive peptide p5+14 was specifically developed to bind to electronegative surfaces of glycosaminoglycans and amyloid fibrils in all types of amyloid deposits.¹⁹ This peptide was able to image visceral amyloid deposits in a murine model of serum amyloid protein A amyloidosis using ¹²⁵I-p5+14 or ^99mTc-p5+14 for SPECT,^20–22 and in humans with multiple types of systemic amyloidosis using ¹²⁴I-p5+14 for PET (¹²⁴I-evuzamitide).^23,24 Unlike ¹⁸F-based radiotracers, ¹²⁴I-evuzamitide can quantify hepatic and renal amyloid.²⁰ Moreover, a therapeutic monoclonal antibody fusion protein with a similar amyloid-binding peptide is currently under Phase I investigation for amyloid fibril removal. However, the ability of ¹²⁴I-evuzamitide to detect and quantify myocardial amyloid is unknown.

The primary aims of this pilot study were as follows: 1) to quantify ¹²⁴I-evuzamitide cardiac uptake in participants with amyloid CMP and control subjects; 2) to compare the diagnostic value of ¹²⁴I-evuzamitide and ¹⁸F-florbetapir in AL-CMP and ATTRwt-CMP; and 3) to measure correlations between ¹²⁴I-evuzamitide myocardial uptake and metrics of cardiac structure and function, serum biomarkers, functional status, and quality of life as markers for amyloid burden.

METHODS

PARTICIPANT INCLUSION.

The study was approved by the Mass General Brigham Human Research Committee, and each participant provided written informed consent (protocol number 2021P000085). The use of ¹²⁴I-evuzamitide was approved by the U.S. Food and Drug Administration under an Investigational New Drug application (155808). Between July 2021 and June 2023, 46 participants were included: 12 with AL-CMP, 12 with ATTRwt-CMP, 2 with variant amyloid CMP (1 variant transthyretin amyloidosis [ATTRv] with Thr60Ala [p.T80A], 1 apolipoprotein A-IV amyloidosis [AApoAIV]), and 12 control participants without amyloidosis. Additionally, 8 control participants from previous studies with ¹⁸F-florbetapir PET/CT were included.^12,15 One additional AL-CMP participant was excluded because of an acquisition error during ¹²⁴I-evuzamitide imaging. Variant amyloidosis participants were not included in statistical analyses because of their small numbers (n = 2), but were presented visually for comparison. AL-CMP was diagnosed using standard criteria for systemic AL amyloidosis, including biopsy with confirmation of amyloid type by immunohistochemistry or mass spectrometry, and proof of cardiac involvement by imaging or endomyocardial biopsy.²⁵ ATTRwt amyloidosis was diagnosed by a grade 2/3 cardiac ^99mTc-pyrophosphate (^99mTc-PYP) SPECT/CT scan and exclusion of AL amyloidosis by serum free light chain assay and serum and urine immunofixation electrophoresis, or by endomyocardial biopsy with immunohistochemistry or mass spectrometry.²⁵ Variant ATTR amyloid CMP was diagnosed by biopsy with immunohistochemistry or mass spectrometry, the finding of a variant TTR by genetic sequencing if appropriate, and proof of cardiac involvement by imaging or endomyocardial biopsy (see detailed inclusion and exclusion criteria in the Supplemental Methods).

PET/CT ACQUISITION.

Participants with amyloid CMP underwent molecular imaging of amyloidosis using ¹²⁴I-evuzamitide and ¹⁸F-florbetapir PET/CT (except 1 AL-CMP participant in whom only ¹²⁴I-evuzamitide PET/CT could be performed). Control subjects underwent 1 PET/CT: 12 with ¹²⁴I-evuzamitide and 8 with ¹⁸F-florbetapir. Imaging was performed using a Discovery MI PET/CT scanner (GE Healthcare) with CT scout for patient positioning and low-dose chest CT scan for attenuation correction and image fusion. Acquisition and quantification methods were specific to each radiotracer, given differences in radiotracer characteristics. ¹²⁴I-evuzamitide was injected intravenously over 5 minutes. Cardiac PET acquisition was performed for 30 minutes in static mode only, 5 hours postinjection to allow for clearance of blood pool activity, based on previous data on radiotracer kinetics.²⁰ The median net injected activity, limited by concerns about exposure to participants and staff, was 1.00 mCi (IQR: 0.94-1.07 mCi), with a median effective dose of 9.01 mSv (IQR: 8.55-9.47 mSv, based on 0.23 mSv/MBq, including 0.5 mSv for the low-dose CT scan).²³ We protected the thyroid gland using potassium iodide 130 mg for 7 days, starting 1 day before radiotracer injection. ¹⁸F-florbetapir was injected as a bolus 1 minute after starting the PET acquisition in list mode. Static images of the heart were reconstructed using data from 4 to 30 minutes after radiotracer injection, based on a previous publication.¹² The median net injected activity was 7.15 mCi (IQR: 6.33-8.40 mCi), with a median effective dose of 5.58 mSv (IQR: 4.99-6.46 mSv, including 0.5 mSv for low-dose CT scan). The ¹²⁴I-evuzamitide scan was performed a median of 1.5 days (IQR: 1.0-4.5 days, range 1-12 days) after the ¹⁸F-florbetapir scan.

PET/CT QUANTIFICATION.

Myocardial uptake was measured volumetrically on static images using PMOD software (PMOD Technologies LLC). Because manual tracing of the myocardial contour on each slice lacks reproducibility and is inaccurate in control subjects, we used automatic iso-contouring in PMOD to exclude ventricular blood pool and delineate the myocardium in a reproducible manner. First, volumes of interest (VOIs) were manually traced on PET emission images guided by fused CT images to define the left ventricular (LV) and right ventricular (RV) contours including cavity blood pool, and attributing the interventricular septum to LV tracings. Given differences in radiotracer kinetics and target-to-background ratio, specific iso-contouring thresholds were chosen for each radiotracer. Blood pool activity concentration was measured in a 10-mm-diameter left atrial (LA) spherical VOI. For ¹⁸F-florbetapir, we used a threshold of 2 × mean blood pool activity concentration, as in our previous study.¹² For ¹²⁴I-evuzamitide, we used a threshold of mean + 2 SDs of blood pool activity concentration, which better delineated myocardial volumes, particularly in mild cases (Supplemental Figure 1). Our primary LV and RV uptake metric was percent injected dose (%ID), calculated as VOI mean activity concentration × VOI volume/injected activity.²⁶ This metric adjusts for injected activity, but not for body weight, because the latter is unnecessary for a radiotracer accumulating in the heart and specific organs, not in the whole body. We additionally analyzed standardized uptake (SUV_mean, SUV_max) as mean or maximal VOI activity concentration/(injected activity/body weight), cardiac amyloid activity (CAA) as: VOI SUV_mean × VOI volume, and target-to background ratio (TBR) as: VOI mean activity concentration/blood pool mean activity concentration (identical to an SUV_ratio, where corrections for injected activity and weight in SUV cancel out in the ratio).²⁶ Thresholds for abnormally high uptake values were defined in receiver-operating characteristic (ROC) analysis by maximization of Youden’s index (sensitivity + specificity − 1) to identify cutoffs with optimal sensitivity and specificity to classify CMP cases vs control subjects, giving equal weight to false positive and false negative cases.

STRUCTURAL AND FUNCTIONAL MARKERS OF AMYLOID BURDEN.

Reports and, when available, images from clinically performed echocardiograms (n = 23), CMR (n = 13), and ^99mTc-PYP SPECT (n = 11) within 1 year of PET/CT were retrieved. Older ^99mTc-PYP SPECT results were included if Grade 3, because no further change was expected. Median time differences to PET/CT were 1.7 months (IQR: 0.7-4.1 months) for echocardiogram, 2.7 months (IQR: 0.7-5.0 months) for CMR, and 4.5 months (IQR: 2.5-13.8 months) for SPECT. Interventricular septum thickness, LV mass index, ejection fraction, myocardial contraction fraction (MCF) (stroke volume/myocardial volume), global longitudinal strain (GLS), late gadolinium enhancement, ECV, and ^99mTc-PYP grade were evaluated.

LABORATORY TESTS, FUNCTIONAL STATUS, AND QUALITY OF LIFE.

In all participants, serum levels of troponin T, N-terminal pro-B-type natriuretic peptide, and creatinine were measured. Functional status was assessed using the NYHA functional class and the 6-minute walk test distance. Quality of life was measured using the Minnesota Living With Heart Failure Questionnaire (21 questions, total score 0-105 points, higher values indicating worse symptoms), the Kansas City Cardiomyopathy Questionnaire (23 questions, overall score 0-100 points, higher values indicating less symptoms), and the 36-Item Short Form Survey (SF-36) (36 questions, transformed T-scores for normalized population 0-100, higher values indicating better quality of life).

STATISTICAL ANALYSIS.

Continuous variables were presented as median with IQR and compared using Kruskal-Wallis test, followed by Dunn’s test for pairwise testing. We included an adjustment for multiple testing using the Benjamini-Hochberg procedure. Categorical variables were displayed as frequency with percentage and assessed using the Fisher exact test. Paired variables were compared using Wilcoxon signed-rank test or McNemar’s test, as appropriate. Correlations were quantified using Spearman’s ρ with 95% CI using the Fieller correction. We presented 2-sided P values and considered them to be statistically significant if <0.05. Data were analyzed using R version 4.3.0 (R Core Team, R Foundation for Statistical Computing), using the packages tidyverse, DescTools, gtsummary, rstatix, pROC, and correlation.

RESULTS

PATIENT CHARACTERISTICS.

In amyloid CMP participants (n = 24), median age was 74 years (IQR: 69-78 years) and 22 were men (92%). ATTRwt-CMP participants (n = 12) were older, had a lower heart rate, and had worse cardiac metrics, such as LV mass index, ejection fraction, and GLS, than AL-CMP participants (n = 12). But, other demographics, clinical characteristics, quality of life, functional status, and biomarkers were similar (Table 1). At study inclusion, 10 AL-CMP participants were in hematological remission after plasma cell-directed therapy (83%), and 2 were undergoing therapy and not in remission (17%), whereas 10 ATTRwt-CMP participants received tafamidis (83%). In control subjects (n = 20), median age was 62 years (IQR: 58-67 years) and 13 were men (65%).

TABLE 1.

Participant Characteristics

	AL-CMP	ATTRwt-CMP	Control Subjectsa	P Value
Demographics	(n = 12)	(n = 12)	(n = 20)
Age, y	71 (63-74)	78 (73-80)	62 (58-67)	<0.001
Male	10 (83)	12 (100)	13 (65)	0.062
Race and ethnicity				0.054
White non-Hispanic	11 (92)	12 (100)	14 (70)
Black or African American	1 (8.3)	0 (0)	6 (30)
Other	0 (0)	0 (0)	0 (0)
Body mass index, kg/m2	26.5 (23.8-29.4)	26.9 (25.1-29.1)	27.8 (24.3-32.1)	0.629
Clinical parameters	(n = 12)	(n = 12)	(n = 12)
Heart rate, beats/min	82 (76-92)	66 (63-70)	72 (68-82)	0.028
Systolic BP, mm Hg	133 (116-150)	133 (127-141)	128 (119-138)	0.808
Diastolic BP, mm Hg	72 (62-81)	68 (67-74)	72 (67-81)	0.823
Arterial hypertension	5 (42)	7 (58)	6 (50)	0.913
Diabetes mellitus	2 (17)	2 (17)	2 (17)	>0.999
Current smoking	0 (0)	0 (0)	0 (0)	–
Functional status and quality of life	(n = 12)	(n = 12)	(n = 12)
NYHA functional class				<0.001
I	4 (33)	3 (25)	12 (100)
II	7 (58)	6 (50)	0 (0)
III	1 (8.3)	3 (25)	0 (0)
IV	0 (0)	0 (0)	0 (0)
MLWHFQ total, lower better	26 (5-30; 0-52)	22 (3-39; 0-53)	0 (0-0; 0-3)	<0.001
KCCQ overall, higher better	77 (66-96; 48-99)	87 (58-93; 52-100)	100 (100-100; 94-100)	<0.001
SF-36 physical, higher better	51 (44-55; 35-60)	51 (42-60; 31-64)	63 (61-64; 48-66)	<0.001
SF-36 mental, higher better	54 (50-57; 40-61)	52 (47-60; 31-63)	61 (57-62; 54-65)	0.007
6-min walk test distance, m	403 (372-468)	345 (312-395)	420 (392-422)	0.173
Laboratory tests	(n = 12)	(n = 12)	(n = 12)
Troponin T, ng/mL	34 (17-50)	38 (27-56)	9 (7-10)	<0.001
NT-proBNP, pg/mL	1,344 (924-1,975)	1,290 (506-1,530)	39 (28-140)	<0.001
eGFR, mL/min/1.73 m2	65 (42-73)	55 (40-64)	88 (73-95)	0.005
Echocardiogram, n = 23	(n = 11)	(n = 12)
Interventricular septum, mm	14 (12-15)	17 (14-17)	–	0.047
LV mass index, g/m2	113 (88-119)	144 (128-169)	–	0.011
LV ejection fraction, %	58 (55-60)	45 (40-56)	–	0.033
LV MCF, %	23 (20-25)	17 (14-20)	–	0.091
LV GLS, %	−15.4 (−16.3 to −12.9)	−11.7 (−13.1 to −8.7)	–	0.025
CMR, n = 13	(n = 6)	(n = 7)
Interventricular septum, mm	15 (14-17)	21 (16-23)	–	0.248
LV mass index, g/m2	68 (60-80)	107 (76-115)	–	0.065
LV ejection fraction, %	56 (55-64)	48 (42-54)	–	0.051
LV MCF, %	59 (53-86)	42 (42-50)	–	0.065
Typical LGE	4 (67)	7 (100)	–	0.192
LV ECV, %	54.0 (48.3-55.4)	53.0 (48.3-56.3)	–	0.810
99mTc-PYP SPECT, n = 11	(n = 0)	(n = 11)		–
Grade 0	–	0 (0)	–
Grade 1	–	1 (9)	–
Grade 2	–	3 (27)	–
Grade 3	–	7 (64)	–
Biopsy for amyloidosis	12 (100)	1 (8)	–	<0.001

Values are median (IQR) (with P values from the Kruskal-Wallis test) or n (%) (with P values from the Fisher exact test). For the Minnesota Living With Heart Failure Questionnaire (MLWHFQ), Kansas City Cardiomyopathy Questionnaire (KCCQ), and 36-Item Short Form Survey (SF-36), the data range is also presented. Values are median (IQR) (with P values from the Kruskal-Wallis test for 3 groups or Wilcoxon rank-sum test for 2 groups) or n (%) (with P values from the Fisher exact test).

^aOnly demographical data were available for the ¹⁸F-florbetapir control cohort.

AL = light-chain amyloidosis; ATTRwt = wild-type transthyretin amyloidosis; BP = blood pressure; CMP = cardiomyopathy; CMR = cardiac magnetic resonance; ECV = extra cellular volume; eGFR = estimated glomerular filtration rate; GLS = global longitudinal strain; LGE = lategadolinium enhancement; MCF = myocardial contraction fraction (stroke volume/myocardial volume); NT-pro BNP = N-terminal pro-B-type natriuretic peptide; ^99mTc-PYP SPECT = ^99mTc-pyrophosphate single-photon emission computed tomography.

¹²⁴I-EVUZAMITIDE PET/CT LV AND RV MYOCARDIAL UPTAKE.

Radiotracer injection was well tolerated in all participants and no adverse effects were noted. Substantial cardiac uptake of ¹²⁴I-evuzamitide was observed in all amyloid CMP participants regardless of amyloid type, while all control participants had no visual cardiac uptake (Central Illustration). Moreover, in a participant with equivocal Grade 1 ^99mTc-PYP SPECT/CT, ¹²⁴I-evuzamitide PET/CT demonstrated definite myocardial uptake, confirmed by endomyocardial biopsy showing TTR amyloid deposits (Figure 1). ¹²⁴I-evuzamitide LV %ID differed across groups (overall P < 0.001): AL-CMP median 1.48 (IQR: 1.12-1.89), ATTRwt-CMP 2.12 (IQR: 1.66-2.47), and control subjects 0.00 (IQR: 0.00-0.01) (Table 2). Several uptake metrics tended to be higher in ATTRwt-CMP than in AL-CMP, particularly LV SUV_mean and RV metrics (Figure 2, Table 2). Based on Youden’s index, high LV %ID, LV CAA, and LV TBR perfectly discriminated amyloid CMP cases from control subjects, whereas LV SUV_mean, SUV_max and RV metrics showed some overlap (Figure 2, Table 2). Indeed, a few control participants presented heterogeneous background activity with small spots of activity concentration above the automatic iso-contouring threshold based on blood pool. This led to SUV_mean in the range of amyloid CMP, but it was not visually suggestive of amyloid deposition. By contrast, %ID and CAA were close to zero in control participants, because these metrics included VOI volume, and therefore better differentiated amyloid deposits in CMP cases from heterogeneous background activity in control participants.

CENTRAL ILLUSTRATION. ¹²⁴I-Evuzamitide Positron Emission Tomography/Computed Tomography Imaging in Multiple Types of Amyloid Cardiomyopathy.

¹²⁴I-evuzamitide is a pan-amyloid imaging agent with visually apparent myocardial uptake in participants with cardiac amyloidosis from Light-chain amyloidosis (AL), wild-type transthyretin amyloidosis (ATTRwt), variant transthyretin amyloidosis (Thr60Ala [p.T80A]) (ATTRv), and apolipoprotein A-IV amyloidosis (AApoAIV), compared with no myocardial uptake in a control participant. Esophageal activity is caused by salivary excretion of radioiodide. PYP 1/3 = ^99mTc-pyrophosphate single-photon emission computed tomography grade 1 or 3; %ID = percent injected dose.

FIGURE 1. ¹²⁴I-Evuzamitide PET/CT in Wild-Type Transthyretin-CMP With Equivocal ^99mPYP SPECT/CT.

This participant presented with heart failure, preserved Left ventricular ejection fraction, diastolic dysfunction grade 2, asymmetric hypertrophy, and impaired global Longitudinal strain on echocardiogram. He had a history of surgery for bilateral carpal tunnel syndrome and lumbar stenosis, and had no evidence for abnormal immunoglobulin light chains. ^99mTc-pyrophosphate (^99mTc-PYP) single-photon emission computed tomography (SPECT)/computed tomography (CT) was equivocal with Grade 1 uptake. Then, cardiac magnetic resonance (CMR) showed diffuse late gadolinium enhancement. Wild-type transthyretin-cardiomyopathy (CMP) (National Amyloidosis Center stage 2, Mayo stage 2) was confirmed by endomyocardial biopsy with mass spectrometry and normal TTR gene sequencing. ¹²⁴I-evuzamitide positron emission tomography (PET)/CT performed within 6 months showed definite myocardial uptake. We thank Dr Bobby Padera from BWH Pathology Department for providing the histology images.

TABLE 2.

¹²⁴I-Evuzamitide and ¹⁸F-Florbetapir Myocardial Uptake Metrics

	AL-CMP	ATTRwt-CMP	Control Subjects	P Value
124I-evuzamitide	(n = 12)	(n = 12)	(n = 12)
Left ventricle
%ID	1.48 (1.12-1.89)	2.12 (1.66-2.47)	0.00 (0.00-0.01)	<0.001
SUVmean	3.67 (3.20-4.49)	4.78 (4.25-5.47)	3.04 (2.97-3.46)	0.004
SUVmax	6.08 (5.16-7.20)	8.03 (6.84-8.82)	3.48 (3.21-3.97)	<0.001
CAA	1,236 (856-1,424)	1,908 (1,340-2,046)	2 (1-8)	<0.001
TBR	1.98 (1.75-2.16)	1.89 (1.77-2.06)	1.20 (1.17-1.27)	<0.001
Volume, mL	300 (246-350)	375 (334-426)	1 (0-3)	<0.001
High %ID (>0.38)	12 (100)	12 (100)	0 (0)	<0.001
Right ventricle
%ID	0.28 (0.08-0.52)	0.44 (0.35-0.75)	0.00 (0.00-0.01)	<0.001
SUVmean	2.83 (2.39-3.12)	3.92 (3.37-4.26)	3.01 (2.60-3.23)	0.008
SUVmax	4.23 (3.67-5.10)	6.22 (5.56-7.16)	3.36 (3.04-3.93)	<0.001
CAA	173 (55-481)	346 (301-633)	1 (0-8)	<0.001
TBR	1.49 (1.37-1.62)	1.48 (1.40-1.59)	1.19 (1.16-1.25)	<0.001
Volume, mL	64 (24-147)	123 (74-160)	0 (0-2)	<0.001
High %ID (>0.02)	10 (83)	12 (100)	0 (0)	<0.001
18F-florbetapir	(n = 11)	(n = 12)	(n = 8)
Left ventricle
%ID	1.03 (0.78-2.03)	1.10 (0.89-1.44)a	0.04 (0.00-0.29)	<0.001
SUVmean	3.04 (2.06-5.06)	2.59 (2.31-2.87)a	0.98 (0.57-1.32)a	<0.001
SUVmax	5.22 (4.08-8.77)	4.06 (3.69-4.37)a	1.82 (1.43-2.88)a	<0.001
CAA	953 (604-1,277)	913 (731-1,173)a	32 (1-185)	<0.001
TBR	6.24 (3.90-7.73)b	4.05 (3.74-4.59)b	2.18 (1.37-2.73)	<0.001
Volume, mL	282 (269-328)	392 (308-420)	36 (0-147)	<0.001
High %ID (>0.40)	11 (100)	12 (100)	0 (0)	<0.001
Right ventricle
%ID	0.24 (0.18-0.40)	0.25 (0.20-0.40)a	0.01 (0.00-0.10)	<0.001
SUVmean	2.07 (1.43-2.82)a	1.92 (1.73-2.10)a	1.08 (0.78-1.67)a	0.024
SUVmax	4.27 (2.67-5.80)	3.43 (2.96-3.71)a	1.86 (1.48-2.65)a	0.004
CAA	182 (115-318)	201 (186-330)a	13 (1-69)	<0.001
TBR	3.55 (2.80-4.43)b	3.04 (2.83-3.53)b	2.12 (1.97-2.73)b	0.003
Volume, mL	115 (59-126)	122 (101-154)	15 (1-66)b	0.001
High %ID (>0.17)	8 (73)	11 (92)	0 (0)	<0.001

Values are median (IQR) (with P values from the Kruskal-Wallis test) or n (%) (with P values from the Fisher exact test). Measurement procedures and calculations for all metrics are explained in the Methods. High %ID was defined using Youden’s index, and thresholds are indicated. Marked ¹⁸F-florbetapir values were significantly (P < 0.05)

^alower or

^bhigher than their ¹²⁴I-evuzamitide counterparts by Wilcoxon signed-rank test or McNemar’s test for AL-CMP and ATTRwt-CMP participants(paired), by Wilcoxon rank-sum test or Fisher exact test for control participants (unpaired). The percentage of high %ID in AL-CMP and ATTRwt-CMP corresponds to sensitivity, and 100 - percentage of high %ID in control subjects corresponds to specificity.

SUV = standardized uptake value; other abbreviations as is Figure 2.

FIGURE 2. ¹²⁴I-Evuzamitide Uptake Metrics.

¹²⁴I-evuzamitide Left ventricular (LV) percent injected dose (%ID), cardiac amyloid activity (CAA) (SUV units × mL), and target-to-background ratio (TBR) perfectly discriminated amyloid CMP cases from control subjects, but mean standardized uptake value (SUV_mean) did not. Uptake measurement and calculations are explained in the Methods section. Thresholds shown for high metrics were defined using Youden’s index. Between-group comparisons were made using Dunn’s test.

¹⁸F-FLORBETAPIR PET/CT LV AND RV MYOCARDIAL UPTAKE.

In the same amyloid CMP participants (except 1 AL-CMP participant without ¹⁸F-florbetapir PET/CT), and in different control participants, we found similar results with ¹⁸F-florbetapir as with ¹²⁴I-evuzamitide (Figure 3, Table 2). ¹⁸F-florbetapir LV %ID also differed across groups (overall P < 0.001): AL-CMP median 1.03 (IQR: 0.78-2.03), ATTRwt-CMP 1.10 (IQR: 0.89-1.44), and control subjects 0.04 (IQR: 0.00-0.29). LV myocardial uptake metrics did not differ between AL-CMP and ATTRwt-CMP participants, but the range of values was higher with AL-CMP. High LV %ID and LV CAA perfectly discriminated amyloid CMP cases from control subjects, while LV SUV_mean, LV TBR, and RV metrics showed some overlap.

FIGURE 3. ¹⁸F-Florbetapir Uptake Metrics.

¹⁸F-Florbetapir LV %ID and CAA perfectly discriminated amyloid CMP cases from control subjects, but SUV_mean and TBR did not. Uptake measurement and calculations are explained in the Methods section. Thresholds shown for high metrics were defined using Youden’s index. Between-group comparisons were made using Dunn’s test. Abbreviations as in Figure 2.

MYOCARDIAL RADIOTRACER UPTAKE METRICS WITH ¹²⁴I-EVUZAMITIDE VS ¹⁸F-FLORBETAPIR.

When comparing LV myocardial radiotracer uptake within the same participants, LV and RV %ID, SUV_mean, SUV_max, and CAA were similar with ¹²⁴I-evuzamitide and ¹⁸F-florbetapir in AL-CMP, but higher with ¹²⁴I-evuzamitide in ATTRwt-CMP (all P ≤ 0.002) (Figure 4, Table 2). LV and RV TBR were higher with ¹⁸F-florbetapir than with ¹²⁴I-evuzamitide in AL-CMP and ATTRwt-CMP (all P ≤ 0.002), reflecting the differing blood and tissue kinetics of the 2 tracers.

FIGURE 4. ¹²⁴I-Evuzamitide and ¹⁸F-Florbetapir Uptake Metrics in the Same Participants.

¹²⁴I-evuzamitide and ¹⁸F-florbetapir imaging were obtained in the same participants (1 Light-chain amyloidosis [AL]-cardiomyopathy [CMP] participant without ¹⁸F-florbetapir PET/CT is not included). LV %ID, SUV_mean, and CAA were higher in wild-type transthyretin (ATTRwt)-CMP with ¹²⁴I-evuzamitide. But, LV TBR was higher with ¹⁸F-florbetapir in AL-CMP and ATTRwt-CMP. Uptake measurement and calculations are explained in the Methods section. P values were obtained from paired Wilcoxon signed-rank test. Abbreviations as in Figure 2.

CORRELATION BETWEEN ¹²⁴I-EVUZAMITIDE UPTAKE AND DIRECT AND INDIRECT MARKERS OF AMYLOID BURDEN, FUNCTIONAL STATUS AND QUALITY OF LIFE.

In the entire cohort, we found moderate correlations between ¹²⁴I-evuzamitide LV % ID and NYHA functional class (ρ = 0.47), Minnesota Living With Heart Failure Questionnaire total score (ρ = 0.59), Kansas City Cardiomyopathy Questionnaire overall score (ρ = −0.62), SF-36 physical T-score (ρ = −0.50), SF-36 mental T-score (ρ = −0.41), as well as strong correlations with troponin T (ρ = 0.73), and N-terminal pro-B-type natriuretic peptide (ρ = 0.72) (Table 3). Moreover, in the AL-CMP and ATTRwt-CMP participants, we found moderate to strong correlations between ¹²⁴I-evuzamitide LV %ID and ¹⁸F-florbetapir LV %ID (ρ = 0.50), interventricular septum thickness (echocardiogram ρ = 0.78, CMR ρ = 0.58), LV mass index (echocardiogram ρ = 0.70, CMR ρ = 0.82), LV MCF (echocardiogram ρ = −0.66, CMR ρ = −0.62), LV GLS (ρ = 0.54), and LV ECV (ρ = 0.51).

TABLE 3.

Correlation Between ¹²⁴I-Evuzamitide Uptake and Direct and Indirect Markers of Amyloid Burden, Functional Status and Quality of Life

	Correlation With 124I–Evuzamitide LV %ID	P Value
Quality of life and functional status, n = 36
NYHA functional class	0.47 (0.16 to 0.70)	0.004
MLWHFQ total score	0.59 (0.32 to 0.78)	<0.001
KCCQ overall score	−0.62 (−0.79 to −0.36)	<0.001
SF-36 physical summary T-score	−0.50 (−0.72 to −0.20)	0.002
SF-36 mental summary T score	−0.41 (−0.65 to −0.08)	0.014
6-min walk test distance, m	−0.13 (−0.45 to 0.22)	0.458
Laboratory tests, n = 36
Troponin T, ng/mL	0.73 (0.53 to 0.86)	<0.001
NT-proBNP, pg/mL	0.72 (0.50 to 0.85)	<0.001
Echocardiogram, n = 23
Interventricular septum, mm	0.78 (0.53 to 0.90)	<0.001
LV mass index, g/m2	0.70 (0.39 to 0.87)	<0.001
LV ejection fraction, %	−0.37 (−0.68 to 0.06)	0.084
LV MCF, %	−0.66 (−0.85 to −0.33)	<0.001
LV GLS, %	0.54 (0.10 to 0.80)	0.017
CMR, n = 13
Interventricular septum, mm	0.58 (−0.10 to 0.89)	0.079
LV mass index, g/m2	0.82 (0.45 to 0.95)	0.001
LV ejection fraction, %	−0.32 (−0.75 to 0.30)	0.289
LV MCF, %	−0.62 (−0.88 to −0.04)	0.033
LV extracellular volume, %	0.51 (−0.11 to 0.84)	0.092
Radionuclide imaging
99mTc-PYP SPECT grade, n = 11	0.45 (−0.23 to 0.83)	0.167
18F-florbetapir LV %ID, n = 23	0.50 (0.10 to 0.76)	0.015

Values are Spearman’s ρ (95% CI) unless otherwise indicated. Correlations among ¹²⁴I-evuzamitide LV %ID and quality of life, functional status, and laboratory tests were assessed in participants undergoing ¹²⁴I-evuzamitide PET/CT (12 AL-CMP, 12 ATTRwt-CMP, and 12 control subjects). Correlations with echocardiogram, CMR and radionuclide imaging were evaluated in participants with amyloid cardiomyopathy and available imaging metrics.

Abbreviations as in Tables 1 and 2.

DISCUSSION

This pilot study quantified myocardial uptake of ¹²⁴I-evuzamitide, a novel pan-amyloid-binding radiotracer, in participants with amyloid CMP, compared its diagnostic value to that of ¹⁸F-florbetapir, and measured correlations between ¹²⁴I-evuzamitide myocardial uptake and markers of amyloid burden in amyloid CMP. Visual interpretation of ¹²⁴I-evuzamitide PET/CT images identified cardiac amyloidosis in all participants with AL, ATTRwt, ATTRv, or AApoAIV CMP. All ¹²⁴I-evuzamitide LV and RV myocardial quantitative uptake metrics were significantly higher in participants with either AL-CMP or ATTRwt-CMP than in control participants. Importantly, ¹²⁴I-evuzamitide LV %ID, CAA, and TBR perfectly discriminated all cases of known AL-CMP and ATTRwt-CMP from control subjects. Similar results were observed with ¹⁸F-florbetapir in the same amyloid CMP participants, except for TBR, supporting the validity of ¹²⁴I-evuzamitide findings. Thus, LV %ID and CAA were the most accurate metrics to identify amyloid CMP with both radiotracers. When comparing radiotracers, ¹²⁴I-evuzamitide consistently showed higher myocardial values in ATTRwt-CMP, except for TBR. Furthermore, among AL-CMP and ATTRwt-CMP participants, ¹²⁴I-evuzamitide LV %ID was moderately to strongly correlated with indirect metrics of amyloid burden from echocardiogram and CMR (interventricular septum thickness, LV mass index, MCF, GLS, ECV), suggesting valid amyloid burden quantitation. Moreover, ¹²⁴I-evuzamitide LV %ID was moderately correlated with ¹⁸F-florbetapir LV %ID, suggesting different binding sites, different binding affinities and/or quantification of different components of amyloid deposits. Indeed, ¹²⁴I-evuzamitide was specifically designed to bind to electronegative surfaces of glycosaminoglycans and amyloid fibrils,¹⁹ whereas ¹⁸F-florbetapir is a stilbene, similar in structure to thioflavin T, which is known to bind to the beta-sheet surface along channels formed by cross-strand ladders within amyloid fibrils.²⁷ Correlation between ¹²⁴I-evuzamitide LV %ID and ECV was also moderate, which may be caused by ECV measuring fibrosis, inflammation, and/or edema in addition to amyloid deposits.⁶ Correlations with functional status and quality of life were moderate, and correlations with cardiac biomarkers were strong, which further supports the validity of cardiac amyloid quantitation with ¹²⁴I-evuzamitide PET/CT. Finally, ¹²⁴I-evuzamitide injection was well tolerated in all participants, and no adverse effects were noted.

¹²⁴I-evuzamitide is an innovative radiotracer based on the synthetic polypeptide p5+14, which was developed to bind to electronegative surfaces of glycosaminoglycans and amyloid fibrils in all types of amyloid deposits.¹⁹ This polypeptide was successfully tested in diverse radiotracers in a murine model of serum amyloid A amyloidosis^20–22 and in humans with multiple types of systemic amyloidosis.^23,24 The present study is one of the first to systematically evaluate myocardial uptake with ¹²⁴I-evuzamitide in a cohort of participants with amyloid CMP from multiple precursor proteins, and the first to compare it to ^l8F-florbetapir. Our findings support the use of ¹²⁴I-evuzamitide PET/CT imaging to accurately and safely diagnose AL-CMP and ATTRwt-CMP, and to at least contribute to identify hereditary amyloid CMP. These results suggest that this pan-amyloid binding radiotracer has the potential to reduce the need for endomyocardial biopsy to diagnose cardiac amyloidosis, limiting biopsy to patients with positive scans for whom amyloid type confirmation is needed. Our results also suggested that ¹²⁴I-evuzamitide can quantify cardiac amyloid burden. Indeed, LV ¹²⁴I-evuzamitide uptake was correlated with LV structural and functional measures of amyloid burden, as well as with cardiac biomarkers and heart-failure–related quality of life. With further validation, quantitative ¹²⁴I-evuzamitide uptake may be particularly useful to diagnose early disease or to monitor disease progression in cardiac amyloidosis. Such validation would be timely, because several therapies aiming to reduce amyloid deposition or to remove amyloid deposits are currently available or emerging,^2,4,5 making the assessment of response to therapy a key next step in amyloidosis imaging. Moreover, the ability of ¹²⁴I-evuzamitide to specifically bind to amyloid deposits and the favorable safety profile are promising for the ongoing Phase I trial investigating a monoclonal antibody fusion protein with a similar amyloid-binding peptide, which targets amyloid deposits to trigger their removal. In this context, imaging with ¹²⁴I-evuzamitide before therapy could identify the presence and quantify the amount of amyloid in the heart and other organs, specifically predicting potential main effects and organ-specific side effects, while imaging after therapy could measure organ response. Although ECV may be an alternative approach to quantify amyloid burden in the heart, it is not specific for amyloidosis and may also represent inflammation or fibrosis,⁶ processes that may be triggered by macrophage-mediated amyloid removal. Therefore, targeted imaging of myocardial amyloid using a specific molecular radiotracer may offer relevant advantages over other methods.

Recently, beta-amyloid-targeting PET radiotracers (¹¹C-Pittsburgh-B-compound, ¹⁸F-florbetapir, ¹⁸F-florbetaben, and ¹⁸F-flutemetamol) have emerged as useful tools to evaluate AL-CMP.^11–17 However, these radiotracers have a limited ability to detect ATTR-CMP, particularly ¹⁸F-florbetapir and ¹⁸F-florbetaben.^12,17 Noninvasive diagnosis of ATTR-CMP is usually based on SPECT bone-avid radiotracers (^99mTc-PYP, ^99mTc-3-diphosphono-1,2-propanodicarboxylic acid, ^99mTc-hydroxymethylene diphosphonate).^6,7 Grade 2/3 uptake of these radiotracers was reported as nearly 100% specific, but only 74% sensitive for ATTR-CMP.¹⁰ Indeed, bone-avid SPECT radiotracers may lack the sensitivity to detect early disease in ATTR-CMP, and are unreliable to diagnose AL-CMP, some forms of ATTRv-CMP, and other rare forms of amyloid CMP.^6,7,10 Furthermore, these data originate from high-volume centers assessing selected patients with advanced disease or high pretest probability, and may therefore overestimate the diagnostic performance of SPECT in nonspecialized centers. Moreover, serial quantitative testing with bone-avid radiotracers is limited by the relatively low sensitivity of SPECT compared with PET, and by the early stage of development of quantitative SPECT methods. Importantly, the mechanism of myocardial uptake of bone-avid radiotracers is not known, compromising the interpretation of changes following therapy.^6,7 By contrast, ¹²⁴I-evuzamitide, a PET pan-amyloid-binding radiotracer, might offer higher quantitative accuracy than SPECT bone-avid radiotracers and may be more sensitive that beta-amyloid-specific PET radiotracers for ATTR-CMP. Notably, in the present study, we found significantly higher myocardial uptake values with ¹²⁴I-evuzamitide than with ¹⁸F-florbetapir among ATTRwt-CMP participants. These participants also tended to show higher ¹²⁴I-evuzamitide uptake than AL participants, while uptake was more similar across amyloid types with ¹⁸F-florbetapir. In parallel, ATTRwt-CMP participants had worse structural and functional metrics on echocardiogram and CMR than AL-CMP participants, such as LV mass index and GLPS. Thus, the higher myocardial uptake values with ¹²⁴I-evuzamitide in ATTRwt-CMP might better reflect the severity of the cardiomyopathy and might be more sensitive to diagnose early disease than ¹⁸F-florbetapir. Moreover, the ability to quantify hepatic and renal amyloid was previously shown as another advantage of ¹²⁴I-evuzamitide over ¹⁸F-based beta-amyloid-specific radiotracers.²⁰ Extracardiac uptake was observed in some of our participants, and will be the focus of an upcoming publication. Finally, techniques that provide specific and quantitative estimates of both cardiac AL and ATTR amyloid burden are much needed to evaluate organ response to the novel amyloid-depleting therapies.^{4,5 124}I-evuzamitide may fill that gap.

STUDY LIMITATIONS.

First, we included a small number of participants with AL-CMP and ATTRwt-CMP and 2 participants with hereditary amyloidosis. This limited our statistical power for certain analyses. However, statistically significant results for most analyses despite the small sample size support high effect sizes. Second, because we evaluated a novel amyloid-binding radiotracer, our participants were selected as definite amyloid CMP or as healthy control subjects. Therefore, the perfect discrimination found with ¹²⁴I-evuzamitide might not be generalizable to real-world settings with patients having various stages of cardiac amyloidosis, including early disease. Hence, we have not reported sensitivity, specificity, or accuracy in this pilot study, because the results could be potentially skewed and misleading. Further validation of diagnostic accuracy in larger cohorts of patients with suspected cardiac amyloidosis (amyloid phenocopies) and various stages of amyloidosis including early disease is warranted, as well as validation of the ability to quantify amyloid burden. Healthy control subjects were chosen for this initial pilot study to define the physiological distribution of this novel radiotracer in the heart and other organs. Third, endomyocardial biopsy confirmation of radiotracer binding to amyloid was lacking. However, the specific binding of ¹²⁴I-evuzamitide to amyloid is supported by extensive preclinical data^19–24 and by the correlation observed with ¹⁸F-florbetapir, which has a distinct amyloid-binding mechanism. Fourth, no participant with ATTRv caused by the Val122Ile (p.V142I) variant could be included during the enrollment period. This limits the generalizability of our findings to this form of cardiac amyloidosis particularly affecting individuals of West African ancestry. Fifth, imaging with ¹²⁴I-evuzamitide requires a PET scan, which has lower availability than SPECT scans. However, PET scans are increasingly available, and the radiotracer also exists as ^99mTc-evuzamitide for SPECT.

CONCLUSIONS

¹²⁴I-evuzamitide PET/CT imaging is a promising novel radiotracer able to accurately detect cardiac amyloid deposits of multiple types, to discriminate between presence and absence of cardiac amyloidosis, and to quantify cardiac amyloid burden. In ATTRwt-CMP, amyloid quantification with ¹²⁴I-evuzamitide might be more accurate than with ¹⁸F-florbetapir. Larger, multicenter studies are needed to validate these findings and to further evaluate the clinical utility of ¹²⁴I-evuzamitide PET/CT imaging for early diagnosis and therapeutic monitoring of cardiac amyloidosis.

PERSPECTIVES.

COMPETENCY IN MEDICAL KNOWLEDGE:

¹²⁴I-evuzamitide PET/CT imaging can accurately and noninvasively detect cardiac amyloid deposits of multiple types, discriminate between presence and absence of cardiac amyloidosis, and quantify cardiac amyloid burden. In ATTRwt-CMP, amyloid quantification with ¹²⁴I-evuzamitide PET/CT might be more accurate than with ¹⁸F-florbetapir PET/CT.

COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS:

¹²⁴I-evuzamitide may provide accurate diagnosis of cardiac amyloidosis of multiple types and might offer advantages over other radiotracers for ATTRwt-CMP.

TRANSLATIONAL OUTLOOK:

Our initial findings that ¹²⁴I-evuzamitide is able to detect, discriminate, and quantify cardiac amyloidosis of different types paves the way for future definitive research to validate this imaging agent for detection of advanced cardiac amyloidosis, early cardiac amyloidosis, and for monitoring of therapeutic response.

ABBREVIATIONS AND ACRONYMS

AL

light-chain amyloidosis

ATTR

transthyretin amyloidosis

ATTRv

variant (hereditary) transthyretin amyloidosis

ATTRwt

wild-type transthyretin amyloidosis

CAA

cardiac amyloid activity

CMP

cardiomyopathy

LV

left ventricular

RV

right ventricular

SUV

standardized uptake

TBR

target-to-background

VOI

volume of interest

%ID

percent injected dose