Quantitative 99mTc-PYP SPECT/CT at 90 minutes improves diagnostic stratification in transthyretin cardiac amyloidosis Authors: Tina Nazerani-Zemann, Susanne Stanzel, Friedrich Weitzer, Thomas Kuenzer, Verena Maria Stangl, Jan Bucerius Published: 2026 ABSTRACT PurposeScintigraphy using 99mTc-labeled bone-seeking tracers is a cornerstone in the non-invasive diagnosis of transthyretin amyloid cardiomyopathy (ATTR-CM). However, planar imaging alone, particularly in patients with Perugini score 1, often lacks diagnostic specificity. This study evaluates the incremental diagnostic value of quantitative SPECT/CT at 90 min post-injection (p.i.) when added to standard planar imaging at 60 and 180 min.MethodsIn this retrospective single-center study, 170 patients with suspected cardiac amyloidosis underwent 99mTc-PYP scintigraphy, including planar imaging and SPECT/CT. Quantitative myocardial uptake was assessed using myocardium-to-vertebra and myocardium-to-rib ratios across predefined left ventricular segments. Imaging findings were correlated with electrocardiographic, echocardiographic, cardiac magnetic resonance (CMR), biomarker, and histopathological data where available.ResultsQuantitative SPECT/CT correlated strongly with Perugini scores and improved diagnostic certainty, particularly in patients with equivocal planar findings. The addition of SPECT/CT led to diagnostic reclassification in more than one-third of patients, with the greatest impact observed in the intermediate Perugini group. Myocardial uptake ratios were significantly associated with NT-proBNP, troponin T, echocardiographic markers of diastolic dysfunction, and CMR parameters including native T1 and extracellular volume. Blood pool activity showed an inverse relationship with myocardial uptake and was associated with both upgrade and downgrade, indicating its role as a confounding factor. Correlation analysis demonstrated only weak associations with renal function and no meaningful correlations with biomarkers or structural parameters. Histopathology (7%) was concordant with imaging findings.ConclusionQuantitative SPECT/CT at 90 min p.i. provides significant incremental diagnostic value, particularly in equivocal cases. Its ability to improve diagnostic classification and distinguish true myocardial uptake from blood pool activity supports its integration into routine diagnostic algorithms for ATTR-CM.Supplementary InformationThe online version contains supplementary material available at 10.1007/s00259-026-07969-1. Introduction Cardiac amyloidosis (CA) is a progressive infiltrative cardiomyopathy characterized by the deposition of misfolded proteins within the myocardial interstitium, leading to diastolic dysfunction and heart failure. The two most common forms are immunoglobulin light-chain amyloidosis (AL) and transthyretin amyloidosis (ATTR), the latter resulting from the deposition of misfolded transthyretin protein. When transthyretin deposition leads to clinically manifest cardiomyopathy, the condition is referred to as transthyretin amyloid cardiomyopathy (ATTR-CM). Previously considered a rare and frequently underdiagnosed condition, CA is now increasingly identified, particularly in patients who were earlier misclassified as undifferentiated heart failure with preserved ejection fraction (HFpEF) or unexplained left ventricular hypertrophy [1–3]. This shift is largely due to advancements in non-invasive imaging techniques and the emergence of targeted therapies, which have renewed clinical interest and emphasized the importance of early diagnosis and intervention [1–3]. Cardiac scintigraphy using bone-seeking radiotracers, while limited in its ability to provide structural or hemodynamic data, has proven highly valuable in differentiating ATTR from AL amyloidosis. Although it cannot definitively differentiate between ATTR and AL subtypes on its own, particularly since AL amyloidosis still requires light-chain assessment via serum and urine testing - the high specificity of bone scintigraphy, especially in patients without evidence of a plasma cell dyscrasia, makes it a powerful diagnostic tool. Nevertheless, current guidelines recommend endomyocardial biopsy for definitive differentiation between AL and ATTR amyloidosis in selected cases. In such constellations, a negative scan combined with negative light-chain results may reduce the need for invasive myocardial biopsy in selected cases; however, in patients with high clinical suspicion, further evaluation (e.g., cardiac MRI) may still be warranted [2, 3]. A landmark study by Perugini et al. introduced a semiquantitative approach using planar 99mTechnetium-3,3-diphosphono-1,2-propanodicarboxylic acid (99mTc-DPD) scintigraphy for this purpose [4]. However, the method’s reliance on operator interpretation and comparison with extracardiac uptake presents challenges, particularly in cases with atypical soft tissue distribution [5]. Furthermore, as already mentioned above, the results of that study are based solely on planar imaging and do not consider the additional diagnostic value of single-photon emission computed tomography (SPECT). In that regard, SPECT more recently has become central to the non-invasive evaluation of CA, offering enhanced spatial resolution and allowing for improved localization and characterization of tracer uptake. However, it is important to note that SPECT alone does not permit definitive differentiation between ATTR and AL amyloidosis. This distinction requires integration with laboratory biomarkers or, in some cases, advanced imaging such as positron emission tomography (PET), which may help differentiate subtypes based on distinct myocardial uptake patterns [6, 7]. Under normal physiological conditions, 99mTechnetium-pyrophosphate (99mTc-PYP) is rapidly cleared from the blood circulation and primarily taken up by bone tissue, with approximately 40–50% skeletal uptake occurring within 1–2 h post-injection and peaking at around 2–3 h [8, 9]. In the absence of amyloid infiltration or acute myocardial injury, myocardial retention of 99mTc-PYP is minimal. However, in patients with transthyretin amyloid cardiomyopathy (ATTR-CM), the altered biodistribution leads to characteristic myocardial uptake, serving as a key imaging biomarker for diagnosis [10]. The combination of SPECT and CT (as nowadays established in hybrid SPECT/CT systems) has further expanded the diagnostic potential of bone scintigraphy, also enabling absolute myocardial quantification that may improve diagnostic accuracy, facilitate disease staging, and support longitudinal treatment monitoring [7, 11]. Furthermore, early acquisition protocols - such as imaging at 60 to 90 min post-injection - have gained attention for their logistical and clinical benefits compared to traditional 3-hour protocols [9, 12, 13]. The present study focuses on early-phase SPECT/CT imaging using 99mTc-PYP acquired 90 min post-injection; however, imaging protocols for cardiac amyloidosis vary according to the radiotracer used. Whereas early imaging is well established for PYP, delayed acquisition at approximately 3 h post-injection remains standard for 99mTc-DPD and hydroxymethylene diphosphonate (HMDP) [14]. While prior studies have explored semiquantitative and delayed imaging approaches [5, 9, 11, 15], robust data integrating fully quantitative early-phase 99mTc-PYP SPECT/CT with multimodal imaging parameters and histopathological validation in large consecutive clinical cohorts remain limited. In particular, objective discrimination of equivocal Perugini grade 1 cases continues to represent a major diagnostic challenge [9, 13]. This study therefore evaluates the clinical and diagnostic value of quantitative 90-minute SPECT/CT imaging in a large patient cohort with suspected cardiac amyloidosis. Additionally, we explored the integration of complementary diagnostic modalities, including echocardiography, cardiac magnetic resonance imaging, and laboratory biomarkers, to provide a comprehensive assessment of disease severity and imaging-guided patient management [16–19]. Materials and methods Patient population This study was conducted in accordance with the Declaration of Helsinki and approved by the local institutional ethics committee (Ethikkommission der Medizinischen Universität Graz, approval number: 34–138 ex 21/22). Written informed consent was obtained from all participants prior to inclusion. A total of 170 patients referred to our department between January 2022 and August 2024 for evaluation of suspected cardiac amyloidosis were included. All patients underwent imaging after intravenous administration of 370–555 MBq (10–15 mCi) of 99mTc - PYP. Planar thoracic scintigraphy (anterior, posterior, and left anterior oblique views) was performed at 60 min and 3-hours post-injection. Additionally, a hybrid SPECT/CT scan of the thorax was acquired at 90 min post-injection to assess early-phase myocardial uptake. Image acquisition and quantification Imaging was conducted using either a Symbia or Pro.specta (Siemens, Erlangen, Germany) or a DiscoveryPro (GE Healthcare, Chicago, IL, USA) SPECT/CT system. Standardized acquisition settings included a 256 × 256 matrix, zoom factor 1.0, and low-dose CT for attenuation correction (130 kVp, variable mAs). Scanner assignment was randomized. Qualitative interpretation of planar images employed the Perugini visual scoring system (0–3) as proposed by Perugini et al. in 2005 [4], independently rated by two blinded nuclear medicine specialists with > 10 years of combined experience. Quantification of myocardial, vertebral, and rib uptake was performed using a dedicated commercially available software (HERMES SMART Workstation Version 2.18.0.93). Quantitative imaging findings Quantitative myocardial uptake was assessed using 99mTc-PYP SPECT/CT acquired 90 min post-injection. The myocardium was segmented into anatomically defined volumes of interest (VOIs), including the septum, apex, and lateral wall. VOIs were manually placed by experienced readers using dedicated software (HERMES SMART Workstation, Version 2.18.0.93), ensuring consistent positioning while avoiding spillover from adjacent structures. Reference VOIs were manually placed in the contralateral rib and vertebral body at the corresponding thoracic level. Counts were normalized to VOI volume and expressed as counts per volume (c/v). Myocardial uptake ratios were calculated by dividing myocardial counts by counts in the respective reference regions (rib and vertebra), allowing for standardized quantitative assessment across patients. In addition to visual Perugini scoring, a semiquantitative analysis of planar images was performed using the heart-to-contralateral lung (H/CL) ratio. Regions of interest (ROIs) were manually placed over the cardiac region and the contralateral chest on anterior planar images. Mean counts per pixel were obtained for both regions, and the H/CL ratio was calculated by dividing myocardial counts by counts in the contralateral region. This approach is widely used to support the differentiation of transthyretin amyloid cardiomyopathy and to complement visual assessment. Segmental analysis was performed to account for potential regional heterogeneity of tracer uptake, particularly in early or equivocal disease stages. This approach allows for more detailed characterization of myocardial involvement and reduces the risk of partial volume effects or contamination from adjacent structures that may occur when using a single global left ventricular VOI (Fig. 1). Fig. 1Segmentation schema for quantitative 99mTc-PYP SPECT/CT analysis. VOIs were defined in the septum, apex, lateral wall, vertebral body, and contralateral rib. Counts were normalized by VOI volume and reported as c/v. Interregional ratios were calculated to assess myocardial uptake relative to background structures Segmentation schema for quantitative 99mTc-PYP SPECT/CT analysis. VOIs were defined in the septum, apex, lateral wall, vertebral body, and contralateral rib. Counts were normalized by VOI volume and reported as c/v. Interregional ratios were calculated to assess myocardial uptake relative to background structures Echocardiography and cardiac MRI protocols Transthoracic echocardiography was performed using standard parasternal and apical views, with measurements including left ventricular wall thickness, diastolic function grading, and strain imaging when available. Cardiac MRI was conducted on a 1.5T system using standard sequences including native T1 mapping, late gadolinium enhancement (LGE), and extracellular volume (ECV) quantification. Ejection fraction was calculated via cine Steady-State Free Precession (SSFP) imaging. Elevated extracellular volume (ECV) and native T1 values were defined according to institutional reference ranges and established literature thresholds. Concentric hypertrophy and global wall thickening were assessed based on standard echocardiographic criteria. Concentric hypertrophy was defined according to established geometric parameters (e.g., increased relative wall thickness), whereas global wall thickening was recorded as a qualitative parameter reflecting more diffuse and pronounced myocardial thickening. Comparative analysis Quantitative tracer uptake values and regional myocardial-to-reference ratios were systematically compared across clinical, imaging, and laboratory parameters. Although several previous studies have recommended planar imaging as early as 60 min post-injection - particularly in workflows optimized for patient throughput - there is growing evidence that acquisition at 90 min p.i. offers a superior balance between blood pool clearance and myocardial retention, particularly for 99mTc-PYP. In this study, we therefore performed SPECT/CT with quantitative analysis at 90 min p.i., reflecting current protocol optimization. To contextualize the 90-minute SPECT/CT data and allow comparison with established semiquantitative methods such as Perugini scoring, additional planar images were acquired at 60 and 180 min post-injection. The Perugini score was derived from the 3-hour planar scan, which is known to enhance diagnostic specificity by allowing for tracer washout. In contrast, blood pool activity was assessed exclusively using the 90-minute SPECT/CT images. This separation of imaging time points enables accurate interpretation of tracer kinetics and supports cross-validation between visual and quantitative assessment, particularly in diagnostically challenging cases (e.g., Perugini Score 1). All comparative data, imaging and laboratory, were collected retrospectively within a two-month window of the SPECT/CT acquisition to ensure clinical relevance and temporal proximity. Methodological details, variable definitions, and grouping criteria are presented in the main tables, Tables 1 and 2, while extended datasets and detailed parameter distributions are provided in the Supplementary Material (Supplementary Tables S1–S4). Table 1Clinical, imaging and laboratory characteristics stratified by Perugini score at 60 min and 3 h post-injection. Patients were grouped into Perugini score 0, 1, and 2–3 to reflect increasing degrees of cardiac tracer uptake and suspected amyloid burden. ATTR, transthyretin amyloidosis; ECG, electrocardiogram; ECV, extracellular volume; EF, ejection fraction; ESV, end-systolic volume; GFR, glomerular filtration rate; H/CL, heart-to-contralateral thorax ratio; IQR, interquartile range; IVS, interventricular septum; LVM, left ventricular mass; MRI, magnetic resonance imaging; NT-proBNP, N-terminal pro–B-type natriuretic peptide; QfKa/fLa, ratio of serum free kappa to free lambda light chainsOverallPerugini score 60 min. p.i.Perugini score 3 h p.i. Characteristic012–3012–3N = 170N = 27N = 48N = 95N = 64N = 46N = 60Demographics Age [y], median (IQR)75 (66, 81)76 (56, 82)70 (61, 76)79 (70, 83)72 (62, 79)68 (60, 76)82 (75, 84) Male sex111 (65.3%)14 (51.9%)28 (58.3%)69 (72.6%)36 (56.3%)32 (69.6%)43 (71.7%) H/CL ratio1.62 (1.40, 1.95)1.40 (1.29, 1.59)1.44 (1.26, 1.68)1.90 (1.57, 2.23)1.42 (1.26, 1.61)1.52 (1.43, 1.65)2.12 (1.91, 2.31) Blood pool detected99 (58.2%)17 (63.0%)42 (87.5%)40 (42.1%)53 (82.8%)36 (78.3%)10 (16.7%) Echocardiography Hypertrophy none6 (3.8%)0 (0.0%)4 (9.3%)2 (2.2%)2 (3.6%)3 (7.1%)1 (1.7%) concentric128 (82.1%)20 (87.0%)34 (79.1%)74 (82.2%)47 (83.9%)27 (64.3%)54 (93.1%) global22 (14.1%)3 (13.0%)5 (11.6%)14 (15.6%)7 (12.5%)12 (28.6%)3 (5.2%) (missing)14455842Diastolic dysfunction none72 (46.2%)10 (43.5%)26 (60.5%)36 (40.0%)29 (51.8%)22 (52.4%)21 (36.2%) Grade 140 (25.6%)8 (34.8%)9 (20.9%)23 (25.6%)16 (28.6%)10 (23.8%)14 (24.1%) Grade 230 (19.2%)4 (17.4%)4 (9.3%)22 (24.4%)7 (12.5%)8 (19.0%)15 (25.9%) Grade 314 (9.0%)1 (4.3%)4 (9.3%)9 (10.0%)4 (7.1%)2 (4.8%)8 (13.8%) (missing)14455842 Apical sparing27 (17.3%)2 (8.7%)6 (14.0%)19 (21.1%)6 (10.7%)5 (11.9%)16 (27.6%) (missing)14455842 Elevated filling pressure37 (23.7%)7 (30.4%)5 (11.6%)25 (27.8%)11 (19.6%)6 (14.3%)20 (34.5%) (missing)14455842 EF [%], median (IQR)55 (48, 60)55 (50, 59)55 (45, 60)55 (48, 60)55 (49, 60)55 (50, 58)55 (45, 62) > 50%100 (64.1%)16 (69.6%)28 (65.1%)56 (62.2%)37 (66.1%)30 (71.4%)33 (56.9%) (missing)14455842 IVS, median (IQR)16.0 (14.0, 18.2)16.0 (15.0, 19.0)15.0 (12.0, 17.0)17.0 (15.0, 19.0)16.0 (14.0, 18.0)15.0 (11.7, 17.0)17.4 (15.2, 19.0) (missing)53121427221615 MRI Elevated ECV37 (44.0%)5 (45.5%)7 (26.9%)25 (53.2%)10 (31.3%)10 (45.5%)17 (56.7%) (missing)86162248322430 Diffuse late enhancement33 (39.3%)4 (36.4%)7 (26.9%)22 (46.8%)10 (31.3%)4 (18.2%)19 (63.3%) (missing)86162248322430Hypertrophy none14 (16.7%)1 (9.1%)7 (26.9%)6 (12.8%)7 (21.9%)4 (18.2%)3 (10.0%) concentric52 (61.9%)9 (81.8%)14 (53.8%)29 (61.7%)20 (62.5%)11 (50.0%)21 (70.0%) global18 (21.4%)1 (9.1%)5 (19.2%)12 (25.5%)5 (15.6%)7 (31.8%)6 (20.0%) (missing)86162248322430 T1 elevated46 (54.8%)7 (63.6%)10 (38.5%)29 (61.7%)14 (43.8%)11 (50.0%)21 (70.0%) (missing)86162248322430 EF, median (IQR)55 (45, 64)60 (52, 70)59 (43, 66)52 (43, 59)59 (49, 66)52 (43, 59)52 (45, 61) (missing)85152050302431 LVM [g], median (IQR)152 (104, 200)107 (104, 175)123 (101, 212)158 (119, 200)113 (104, 175)153 (101, 200)186 (128, 205) (missing)135213876513648 Laboratory Findings QfKa/fLa, median (IQR)1.34 (1.06, 1.66)1.38 (0.82, 1.52)1.46 (1.02, 1.73)1.33 (1.07, 1.62)1.46 (1.02, 1.68)1.22 (0.96, 1.64)1.32 (1.12, 1.61) <= 0.319 (6.4%)2 (9.1%)2 (5.7%)5 (6.0%)2 (4.1%)5 (13.2%)2 (3.7%) > 1.5640 (28.4%)4 (18.2%)12 (34.3%)24 (28.6%)14 (28.6%)10 (26.3%)16 (29.6%) (missing)29513111586 Nt-proBNP [pg/mL], median (IQR)1 678 (566, 5 007)1 812 (631, 5 080)1 026 (271, 3 619)2 417 (715, 5 260)1 804 (275, 5 370)809 (383, 2 332)2 891 (1 338, 5 355) > 125 pg/mL138 (92.0%)19 (90.5%)40 (90.9%)79 (92.9%)48 (88.9%)37 (90.2%)53 (96.4%) (missing)2064101055 Troponin T [pg/mL], median (IQR)36 (17, 60)30 (19, 79)26 (12, 42)43 (18, 67)29 (14, 40)21 (9, 48)51 (30, 73) > 14 pg/mL115 (79.9%)19 (86.4%)29 (74.4%)67 (80.7%)40 (78.4%)25 (62.5%)50 (94.3%) (missing)2659121367 Creatinine, median (IQR)1.05 (0.90, 1.42)0.91 (0.78, 1.19)1.13 (0.85, 1.41)1.11 (0.94, 1.42)1.01 (0.84, 1.37)0.96 (0.85, 1.33)1.17 (0.97, 1.46) (missing)14437734 GFR, median (IQR)61 (45, 76)67 (56, 94)59 (43, 75)60 (45, 74)62 (51, 78)68 (54, 85)54 (41, 70) (missing)14437734 Table 1Clinical, imaging and laboratory characteristics stratified by Perugini score at 60 min and 3 h post-injection. Patients were grouped into Perugini score 0, 1, and 2–3 to reflect increasing degrees of cardiac tracer uptake and suspected amyloid burden. ATTR, transthyretin amyloidosis; ECG, electrocardiogram; ECV, extracellular volume; EF, ejection fraction; ESV, end-systolic volume; GFR, glomerular filtration rate; H/CL, heart-to-contralateral thorax ratio; IQR, interquartile range; IVS, interventricular septum; LVM, left ventricular mass; MRI, magnetic resonance imaging; NT-proBNP, N-terminal pro–B-type natriuretic peptide; QfKa/fLa, ratio of serum free kappa to free lambda light chainsOverallPerugini score 60 min. p.i.Perugini score 3 h p.i. Characteristic012–3012–3N = 170N = 27N = 48N = 95N = 64N = 46N = 60Demographics Age [y], median (IQR)75 (66, 81)76 (56, 82)70 (61, 76)79 (70, 83)72 (62, 79)68 (60, 76)82 (75, 84) Male sex111 (65.3%)14 (51.9%)28 (58.3%)69 (72.6%)36 (56.3%)32 (69.6%)43 (71.7%) H/CL ratio1.62 (1.40, 1.95)1.40 (1.29, 1.59)1.44 (1.26, 1.68)1.90 (1.57, 2.23)1.42 (1.26, 1.61)1.52 (1.43, 1.65)2.12 (1.91, 2.31) Blood pool detected99 (58.2%)17 (63.0%)42 (87.5%)40 (42.1%)53 (82.8%)36 (78.3%)10 (16.7%) Echocardiography Hypertrophy none6 (3.8%)0 (0.0%)4 (9.3%)2 (2.2%)2 (3.6%)3 (7.1%)1 (1.7%) concentric128 (82.1%)20 (87.0%)34 (79.1%)74 (82.2%)47 (83.9%)27 (64.3%)54 (93.1%) global22 (14.1%)3 (13.0%)5 (11.6%)14 (15.6%)7 (12.5%)12 (28.6%)3 (5.2%) (missing)14455842Diastolic dysfunction none72 (46.2%)10 (43.5%)26 (60.5%)36 (40.0%)29 (51.8%)22 (52.4%)21 (36.2%) Grade 140 (25.6%)8 (34.8%)9 (20.9%)23 (25.6%)16 (28.6%)10 (23.8%)14 (24.1%) Grade 230 (19.2%)4 (17.4%)4 (9.3%)22 (24.4%)7 (12.5%)8 (19.0%)15 (25.9%) Grade 314 (9.0%)1 (4.3%)4 (9.3%)9 (10.0%)4 (7.1%)2 (4.8%)8 (13.8%) (missing)14455842 Apical sparing27 (17.3%)2 (8.7%)6 (14.0%)19 (21.1%)6 (10.7%)5 (11.9%)16 (27.6%) (missing)14455842 Elevated filling pressure37 (23.7%)7 (30.4%)5 (11.6%)25 (27.8%)11 (19.6%)6 (14.3%)20 (34.5%) (missing)14455842 EF [%], median (IQR)55 (48, 60)55 (50, 59)55 (45, 60)55 (48, 60)55 (49, 60)55 (50, 58)55 (45, 62) > 50%100 (64.1%)16 (69.6%)28 (65.1%)56 (62.2%)37 (66.1%)30 (71.4%)33 (56.9%) (missing)14455842 IVS, median (IQR)16.0 (14.0, 18.2)16.0 (15.0, 19.0)15.0 (12.0, 17.0)17.0 (15.0, 19.0)16.0 (14.0, 18.0)15.0 (11.7, 17.0)17.4 (15.2, 19.0) (missing)53121427221615 MRI Elevated ECV37 (44.0%)5 (45.5%)7 (26.9%)25 (53.2%)10 (31.3%)10 (45.5%)17 (56.7%) (missing)86162248322430 Diffuse late enhancement33 (39.3%)4 (36.4%)7 (26.9%)22 (46.8%)10 (31.3%)4 (18.2%)19 (63.3%) (missing)86162248322430Hypertrophy none14 (16.7%)1 (9.1%)7 (26.9%)6 (12.8%)7 (21.9%)4 (18.2%)3 (10.0%) concentric52 (61.9%)9 (81.8%)14 (53.8%)29 (61.7%)20 (62.5%)11 (50.0%)21 (70.0%) global18 (21.4%)1 (9.1%)5 (19.2%)12 (25.5%)5 (15.6%)7 (31.8%)6 (20.0%) (missing)86162248322430 T1 elevated46 (54.8%)7 (63.6%)10 (38.5%)29 (61.7%)14 (43.8%)11 (50.0%)21 (70.0%) (missing)86162248322430 EF, median (IQR)55 (45, 64)60 (52, 70)59 (43, 66)52 (43, 59)59 (49, 66)52 (43, 59)52 (45, 61) (missing)85152050302431 LVM [g], median (IQR)152 (104, 200)107 (104, 175)123 (101, 212)158 (119, 200)113 (104, 175)153 (101, 200)186 (128, 205) (missing)135213876513648 Laboratory Findings QfKa/fLa, median (IQR)1.34 (1.06, 1.66)1.38 (0.82, 1.52)1.46 (1.02, 1.73)1.33 (1.07, 1.62)1.46 (1.02, 1.68)1.22 (0.96, 1.64)1.32 (1.12, 1.61) <= 0.319 (6.4%)2 (9.1%)2 (5.7%)5 (6.0%)2 (4.1%)5 (13.2%)2 (3.7%) > 1.5640 (28.4%)4 (18.2%)12 (34.3%)24 (28.6%)14 (28.6%)10 (26.3%)16 (29.6%) (missing)29513111586 Nt-proBNP [pg/mL], median (IQR)1 678 (566, 5 007)1 812 (631, 5 080)1 026 (271, 3 619)2 417 (715, 5 260)1 804 (275, 5 370)809 (383, 2 332)2 891 (1 338, 5 355) > 125 pg/mL138 (92.0%)19 (90.5%)40 (90.9%)79 (92.9%)48 (88.9%)37 (90.2%)53 (96.4%) (missing)2064101055 Troponin T [pg/mL], median (IQR)36 (17, 60)30 (19, 79)26 (12, 42)43 (18, 67)29 (14, 40)21 (9, 48)51 (30, 73) > 14 pg/mL115 (79.9%)19 (86.4%)29 (74.4%)67 (80.7%)40 (78.4%)25 (62.5%)50 (94.3%) (missing)2659121367 Creatinine, median (IQR)1.05 (0.90, 1.42)0.91 (0.78, 1.19)1.13 (0.85, 1.41)1.11 (0.94, 1.42)1.01 (0.84, 1.37)0.96 (0.85, 1.33)1.17 (0.97, 1.46) (missing)14437734 GFR, median (IQR)61 (45, 76)67 (56, 94)59 (43, 75)60 (45, 74)62 (51, 78)68 (54, 85)54 (41, 70) (missing)14437734 Clinical, imaging and laboratory characteristics stratified by Perugini score at 60 min and 3 h post-injection. Patients were grouped into Perugini score 0, 1, and 2–3 to reflect increasing degrees of cardiac tracer uptake and suspected amyloid burden. ATTR, transthyretin amyloidosis; ECG, electrocardiogram; ECV, extracellular volume; EF, ejection fraction; ESV, end-systolic volume; GFR, glomerular filtration rate; H/CL, heart-to-contralateral thorax ratio; IQR, interquartile range; IVS, interventricular septum; LVM, left ventricular mass; MRI, magnetic resonance imaging; NT-proBNP, N-terminal pro–B-type natriuretic peptide; QfKa/fLa, ratio of serum free kappa to free lambda light chains Table 2Histopathology stratified by Perugini score at 60 min and 3 h post-injectionOverallPerugini score 60 min. p.i.Perugini score 3 h p.i.Characteristic012–3012–3 Histopathology Myocardial biopsyN = 12N = 1N = 4N = 7N = 5N = 2N = 5 negative7 (58.3%)1 (100%)4 (100%)2 (28.6%)5 (100%)1 (50.0%)1 (20.0%) ATTR5 (41.7%)0 (0.0%)0 (0.0%)5 (71.4%)0 (0.0%)1 (50.0%)4 (80.0%) Iliac crest biopsyN = 17N = 5N = 5N = 7N = 7N = 8N = 2 negative11 (64.7%)2 (40.0%)3 (60.0%)6 (85.7%)4 (57.1%)5 (62.5%)2 (100%) AL6 (35.3%)3 (60.0%)2 (40.0%)1 (14.3%)3 (42.9%)3 (37.5%)0 (0.0%) Table 2Histopathology stratified by Perugini score at 60 min and 3 h post-injectionOverallPerugini score 60 min. p.i.Perugini score 3 h p.i.Characteristic012–3012–3 Histopathology Myocardial biopsyN = 12N = 1N = 4N = 7N = 5N = 2N = 5 negative7 (58.3%)1 (100%)4 (100%)2 (28.6%)5 (100%)1 (50.0%)1 (20.0%) ATTR5 (41.7%)0 (0.0%)0 (0.0%)5 (71.4%)0 (0.0%)1 (50.0%)4 (80.0%) Iliac crest biopsyN = 17N = 5N = 5N = 7N = 7N = 8N = 2 negative11 (64.7%)2 (40.0%)3 (60.0%)6 (85.7%)4 (57.1%)5 (62.5%)2 (100%) AL6 (35.3%)3 (60.0%)2 (40.0%)1 (14.3%)3 (42.9%)3 (37.5%)0 (0.0%) Histopathology stratified by Perugini score at 60 min and 3 h post-injection Statistical analysis Descriptive statistics are presented as median (interquartile range) for continuous variables and frequencies (n, %) for categorical variables. Group comparisons across Perugini scores were performed using Kruskal–Wallis tests (for non-parametric continuous variables), and Pearson’s Chi square (χ²) or Fisher’s exact test where appropriate. Correlations were analyzed using Spearman’s rho with 95% confidence intervals. Analyses were conducted using R version 4.2.3, with results considered significant at the nominal significance level p < 0.05. In addition, subgroup analyses were performed stratifying patients according to the presence or absence of blood pool activity on 90-minute SPECT/CT. Furthermore, reclassification analyses were conducted to assess changes in diagnostic categorization after incorporation of SPECT/CT findings compared to planar imaging. Reclassification was categorized as upgrade, downgrade, or no change, and analyzed in relation to Perugini score categories and myocardial tracer uptake. Given the exploratory nature of the analysis and the focus on clinically relevant variables, no formal correction for multiple testing was applied. To minimize the risk of spurious associations, the number of variables included in the correlation analysis was reduced to those of established clinical relevance. Results Perugini score stratification Among 170 patients (median age: 75 [66–81]; 65.3% male), stratification by Perugini score at 60 min post-injection was: score 0 (15.9%), score 1 (28.2%), and pooled scores 2–3 (55.9%). Pooling of scores 2 and 3 was justified by similar quantitative uptake profiles and low number 36 for score 2. At 180 min, the distribution shifted, with an increased frequency of score 0 (37.6%) and decreased scores 2–3 (35.3%), reflecting tracer washout and improved specificity of delayed imaging. Myocardial uptake across Perugini score categories At 3 h post-injection, myocardial tracer uptake increased markedly across Perugini score categories. Myocardial uptake was observed in 36.1% of patients with Perugini score 0, 71.7% of patients with score 1, and 100% of patients with scores 2–3. These findings confirm the expected relationship between visual grading and quantitative tracer retention, while also highlighting the heterogeneity within the intermediate Perugini score 1 group. A comprehensive overview of myocardial uptake, reclassification patterns, and blood pool activity stratified by Perugini score at 3 h post-injection is provided in Table 3. Table 3Higher Perugini scores are associated with increased myocardial uptake and stable classification, while lower scores show greater blood pool interference and diagnostic variabilityCharacteristicPerugini Score 3 p.i.OverallN = 17010N = 6411N = 4612-3N = 601Myocardial uptake112 (68.3%)22 (36.1%)33 (71.7%)57 (100.0%)Unknown6303Reclassification Downgrade13 (7.9%)0 (0.0%)13 (28.3%)0 (0.0%) No change96 (58.5%)39 (63.9%)0 (0.0%)57 (100.0%) Upgrade55 (33.5%)22 (36.1%)33 (71.7%)0 (0.0%) Unknown6303 Blood pool99 (58.2%)53 (82.8%)36 (78.3%)10 (16.7%)1n (%) Table 3Higher Perugini scores are associated with increased myocardial uptake and stable classification, while lower scores show greater blood pool interference and diagnostic variabilityCharacteristicPerugini Score 3 p.i.OverallN = 17010N = 6411N = 4612-3N = 601Myocardial uptake112 (68.3%)22 (36.1%)33 (71.7%)57 (100.0%)Unknown6303Reclassification Downgrade13 (7.9%)0 (0.0%)13 (28.3%)0 (0.0%) No change96 (58.5%)39 (63.9%)0 (0.0%)57 (100.0%) Upgrade55 (33.5%)22 (36.1%)33 (71.7%)0 (0.0%) Unknown6303 Blood pool99 (58.2%)53 (82.8%)36 (78.3%)10 (16.7%)1n (%) Higher Perugini scores are associated with increased myocardial uptake and stable classification, while lower scores show greater blood pool interference and diagnostic variability 1n (%) Blood pool activity and myocardial uptake Blood pool activity on 90-minute SPECT/CT was present in a substantial proportion of patients and showed a strong inverse association with myocardial tracer uptake. In patients without blood pool activity, myocardial uptake was observed in 80.9% overall and reached 100% in those with Perugini scores 2–3. In contrast, among patients with blood pool activity, myocardial uptake was less frequent overall (59.4%), particularly in lower Perugini categories. Specifically, in the presence of blood pool activity, myocardial uptake was detected in 40.0% of patients with Perugini score 0, 75.0% with score 1, and 100% with scores 2–3. These findings suggest that blood pool activity may confound interpretation of early imaging and underscore the importance of SPECT/CT for differentiating true myocardial retention from residual intravascular tracer. Detailed distributions of myocardial uptake and reclassification patterns according to the presence or absence of blood pool activity are shown in Table 4. Table 4Blood pool activity reduces specificity of myocardial uptake assessment and drives higher rates of reclassification on SPECT/CTCharacteristicNo blood pool detectedBlood pool detectedOverallN = 7110N = 1111N = 1012-3N = 501OverallN = 9910N = 5311N = 3612-3N = 101Myocardial uptake55 (80.9%)2 (18.2%)6 (60.0%)47 (100.0%)57 (59.4%)20 (40.0%)27 (75.0%)10 (100.0%)Unknown30033300Reclassification Downgrade4 (5.9%)0 (0.0%)4 (40.0%)0 (0.0%)9 (9.4%)0 (0.0%)9 (25.0%)0 (0.0%) No change56 (82.4%)9 (81.8%)0 (0.0%)47 (100.0%)40 (41.7%)30 (60.0%)0 (0.0%)10 (100.0%) Upgrade8 (11.8%)2 (18.2%)6 (60.0%)0 (0.0%)47 (49.0%)20 (40.0%)27 (75.0%)0 (0.0%) Unknown300333001n (%) Table 4Blood pool activity reduces specificity of myocardial uptake assessment and drives higher rates of reclassification on SPECT/CTCharacteristicNo blood pool detectedBlood pool detectedOverallN = 7110N = 1111N = 1012-3N = 501OverallN = 9910N = 5311N = 3612-3N = 101Myocardial uptake55 (80.9%)2 (18.2%)6 (60.0%)47 (100.0%)57 (59.4%)20 (40.0%)27 (75.0%)10 (100.0%)Unknown30033300Reclassification Downgrade4 (5.9%)0 (0.0%)4 (40.0%)0 (0.0%)9 (9.4%)0 (0.0%)9 (25.0%)0 (0.0%) No change56 (82.4%)9 (81.8%)0 (0.0%)47 (100.0%)40 (41.7%)30 (60.0%)0 (0.0%)10 (100.0%) Upgrade8 (11.8%)2 (18.2%)6 (60.0%)0 (0.0%)47 (49.0%)20 (40.0%)27 (75.0%)0 (0.0%) Unknown300333001n (%) Blood pool activity reduces specificity of myocardial uptake assessment and drives higher rates of reclassification on SPECT/CT 1n (%) Interaction between Perugini score, myocardial uptake, and blood pool activity The distribution of myocardial uptake across Perugini score categories differed substantially depending on the presence of blood pool activity. In particular, patients with Perugini score 0 and 1 demonstrated heterogeneous uptake patterns, especially in the presence of blood pool signal. This interaction between visual grading, tracer distribution, and blood pool activity is summarized in Table 5. Table 5Heterogeneity of myocardial uptake in Perugini scores 0–1 is partly explained by blood pool activityCharacteristicNo blood pool detectedBlood pool detectedNo myocardial uptakeN = 131Myocardial uptakeN = 551No myocardial uptakeN = 391Myocardial uptakeN = 571Perugini Score 3 Stunden p.i.09230201469272–30470101n Table 5Heterogeneity of myocardial uptake in Perugini scores 0–1 is partly explained by blood pool activityCharacteristicNo blood pool detectedBlood pool detectedNo myocardial uptakeN = 131Myocardial uptakeN = 551No myocardial uptakeN = 391Myocardial uptakeN = 571Perugini Score 3 Stunden p.i.09230201469272–30470101n Heterogeneity of myocardial uptake in Perugini scores 0–1 is partly explained by blood pool activity 1n Impact of SPECT/CT on diagnostic reclassification The addition of 90-minute SPECT/CT resulted in clinically relevant reclassification of patients compared to planar imaging alone. Overall, 33.5% of patients were reclassified to a higher diagnostic category (upgrade), while 7.9% were downgraded, and 58.5% remained unchanged. Reclassification was most pronounced in patients with intermediate findings (Perugini score 1), where 71.7% were upgraded and 28.3% were downgraded, indicating substantial diagnostic uncertainty in this group when relying on planar imaging alone. Importantly, reclassification patterns differed according to blood pool status. In patients with blood pool activity, upgrades were observed in 49.0% of cases compared to 11.8% in patients without blood pool, suggesting that SPECT/CT provides particular benefit in cases with ambiguous tracer distribution. Conversely, downgrades were more frequent in the presence of blood pool activity (9.4% vs. 5.9%), further supporting the role of SPECT/CT in avoiding false-positive interpretations. These reclassification patterns in relation to Perugini score and blood pool activity are summarized in Tables 3 and 4. Quantitative SPECT/CT analysis Blood pool activity on SPECT/CT at 90 min post-injection was present in 58.2% of patients overall, most commonly in those with Perugini score 1 (87.5%) and least in scores 2–3 (42.1%). It is important to emphasize that blood pool assessment was performed exclusively on the 90-minute SPECT/CT images, not on delayed planar imaging. In contrast, Perugini scoring was based on the 3-hour planar acquisitions, in line with established protocols that favor later imaging for improved specificity due to tracer washout. For the 3-hour Perugini scores, the contrast in blood pool activity became even more pronounced, with persistent activity seen in 82.8% of score 0 and only 16.7% of scores 2–3 (Fig. 2). This differentiation between the imaging time points is essential to avoid confusion and ensure consistency in interpreting the uptake patterns and their diagnostic implications. Fig. 2 Imaging findings in a patient with suspected cardiac amyloidosis. (A) Planar static acquisition at 60 min post-injection demonstrates tracer uptake in the cardiac region approximately equal to that of the contralateral thoracic wall, consistent with a Perugini Score 1. (B) Delayed planar imaging at 180 min shows residual tracer uptake in the left thoracic region, but with reduced intensity relative to the contralateral thorax, indicating washout and suggesting a lower likelihood of transthyretin amyloid deposition. (C) SPECT/CT acquisition at 90 min post-injection reveals prominent blood pool activity without corresponding myocardial uptake, supporting the interpretation of blood pool retention rather than true myocardial tracer binding Imaging findings in a patient with suspected cardiac amyloidosis. (A) Planar static acquisition at 60 min post-injection demonstrates tracer uptake in the cardiac region approximately equal to that of the contralateral thoracic wall, consistent with a Perugini Score 1. (B) Delayed planar imaging at 180 min shows residual tracer uptake in the left thoracic region, but with reduced intensity relative to the contralateral thorax, indicating washout and suggesting a lower likelihood of transthyretin amyloid deposition. (C) SPECT/CT acquisition at 90 min post-injection reveals prominent blood pool activity without corresponding myocardial uptake, supporting the interpretation of blood pool retention rather than true myocardial tracer binding The semiquantitative heart-to-contralateral thorax (H/CL) ratio progressively increased with Perugini grade at both time points (60 min: median 1.40 → 1.90; 3 h: 1.42 → 2.12) (Table 1). Quantitative tracer uptake, normalized to vertebral and rib reference tissues, showed a clear stepwise increase across all myocardial segments in accordance with the Perugini score. At 60 min post-injection, the septum-to-vertebra ratio rose from 0.26 in patients with Perugini score 0 to 1.05 in those with scores 2–3 (p < 0.001). This difference became even more pronounced at 3 h, with septum-to-vertebra ratios reaching 1.39 in advanced cases (p < 0.001). Rib-normalized uptake followed a similar pattern, with even steeper relative increases - for example, the septum-to-rib ratio increased from 0.91 in score 0 to 3.26 in scores 2–3 at 3 h. Table 6 and Table 7 summarize the quantitative uptake ratios by myocardial segment for both vertebral and rib references at 60 min and 3 h post-injection. Figures 3 illustrates these distributions with boxplots stratified by Perugini group. Fig. 3Quantitative ratios stratified by Perugini score 3 h p.i Quantitative ratios stratified by Perugini score 3 h p.i Table 6The ratios of counts-per-volume (c/v) myocardium and different segments to c/v vertebra in Perugini score 60 minutes and 3 hours p.i., p-values for comparison of the groups using Kruskal-Wallis testsPerugini score 60 min p.i.Perugini score 3 h p.i.012–3p-value012–3p-valueN = 25N = 44N = 92N = 58N = 46N = 57Myocardium, median (IQR)0.29 (0.23–0.35)0.34 (0.28–0.38)0.71 (0.40–1.00)< 0.0010.31 (0.26–0.37)0.37 (0.31–0.45)0.90 (0.74–1.17)< 0.001Septum, median (IQR)0.26 (0.19–0.33)0.33 (0.28–0.42)1.05 (0.46–1.62)< 0.0010.30 (0.26–0.38)0.39 (0.33–0.53)1.39 (1.10–1.77)< 0.001Apex, median (IQR)0.24 (0.20–0.35)0.33 (0.26–0.40)0.65 (0.35–1.02)< 0.0010.29 (0.22–0.36)0.33 (0.29–0.44)0.89 (0.69–1.13)< 0.001Lateral wall, median (IQR)0.21 (0.17–0.26)0.27 (0.21–0.32)0.64 (0.33–1.24)< 0.0010.25 (0.18–0.29)0.28 (0.23–0.41)1.12 (0.65–1.40)< 0.001 Table 6The ratios of counts-per-volume (c/v) myocardium and different segments to c/v vertebra in Perugini score 60 minutes and 3 hours p.i., p-values for comparison of the groups using Kruskal-Wallis testsPerugini score 60 min p.i.Perugini score 3 h p.i.012–3p-value012–3p-valueN = 25N = 44N = 92N = 58N = 46N = 57Myocardium, median (IQR)0.29 (0.23–0.35)0.34 (0.28–0.38)0.71 (0.40–1.00)< 0.0010.31 (0.26–0.37)0.37 (0.31–0.45)0.90 (0.74–1.17)< 0.001Septum, median (IQR)0.26 (0.19–0.33)0.33 (0.28–0.42)1.05 (0.46–1.62)< 0.0010.30 (0.26–0.38)0.39 (0.33–0.53)1.39 (1.10–1.77)< 0.001Apex, median (IQR)0.24 (0.20–0.35)0.33 (0.26–0.40)0.65 (0.35–1.02)< 0.0010.29 (0.22–0.36)0.33 (0.29–0.44)0.89 (0.69–1.13)< 0.001Lateral wall, median (IQR)0.21 (0.17–0.26)0.27 (0.21–0.32)0.64 (0.33–1.24)< 0.0010.25 (0.18–0.29)0.28 (0.23–0.41)1.12 (0.65–1.40)< 0.001 The ratios of counts-per-volume (c/v) myocardium and different segments to c/v vertebra in Perugini score 60 minutes and 3 hours p.i., p-values for comparison of the groups using Kruskal-Wallis tests Table 7The ratios of c/v myocardium and different segments to c/v rib in Perugini score 60 minutes p.i., p-values for comparison of the groups using Kruskal-Wallis testsPerugini score 60 min p.i.Perugini score 3 h p.i.012–3p-value012–3p-valueN = 25N = 44N = 92N = 58N = 46N = 57Myocardium, median (IQR)0.94 (0.76–1.02)0.86 (0.66–1.16)1.56 (1.17–2.28)< 0.0010.94 (0.71–1.11)1.08 (0.84–1.32)2.10 (1.42–2.56)< 0.001Septum, median (IQR)0.84 (0.63–1.17)0.91 (0.66–1.16)2.21 (1.26–3.68)< 0.0010.91 (0.65–1.17)1.12 (0.88–1.53)3.26 (2.21–4.17)< 0.001Apex, median (IQR)0.77 (0.64–1.05)0.83 (0.66–1.03)1.49 (1.03–2.35)< 0.0010.89 (0.67–1.05)0.98 (0.72–1.25)2.14 (1.49–2.73)< 0.001Lateral wall, median (IQR)0.77 (0.56–0.84)0.65 (0.53–0.91)1.43 (0.97–2.75)< 0.0010.68 (0.56–0.89)0.82 (0.60–1.06)2.22 (1.42–3.26)< 0.001 Table 7The ratios of c/v myocardium and different segments to c/v rib in Perugini score 60 minutes p.i., p-values for comparison of the groups using Kruskal-Wallis testsPerugini score 60 min p.i.Perugini score 3 h p.i.012–3p-value012–3p-valueN = 25N = 44N = 92N = 58N = 46N = 57Myocardium, median (IQR)0.94 (0.76–1.02)0.86 (0.66–1.16)1.56 (1.17–2.28)< 0.0010.94 (0.71–1.11)1.08 (0.84–1.32)2.10 (1.42–2.56)< 0.001Septum, median (IQR)0.84 (0.63–1.17)0.91 (0.66–1.16)2.21 (1.26–3.68)< 0.0010.91 (0.65–1.17)1.12 (0.88–1.53)3.26 (2.21–4.17)< 0.001Apex, median (IQR)0.77 (0.64–1.05)0.83 (0.66–1.03)1.49 (1.03–2.35)< 0.0010.89 (0.67–1.05)0.98 (0.72–1.25)2.14 (1.49–2.73)< 0.001Lateral wall, median (IQR)0.77 (0.56–0.84)0.65 (0.53–0.91)1.43 (0.97–2.75)< 0.0010.68 (0.56–0.89)0.82 (0.60–1.06)2.22 (1.42–3.26)< 0.001 The ratios of c/v myocardium and different segments to c/v rib in Perugini score 60 minutes p.i., p-values for comparison of the groups using Kruskal-Wallis tests Correlation analysis Correlation analyses, performed using Spearman’s rho, revealed robust associations between Perugini scores (at 60 and 180 min) and myocardial uptake ratios (e.g., myocardium-to-vertebra and myocardium-to-rib). Negative correlations with blood pool activity on 90-minute SPECT/CT further supported the diagnostic specificity of these quantitative markers. Age showed a weak to moderate positive correlation with both Perugini score and uptake ratios. Blood pool activity and clinical parameters Blood pool activity on 90-minute SPECT/CT showed an inverse correlation with myocardial uptake ratios and Perugini score. A weak inverse correlation was observed with renal function, as reflected by glomerular filtration rate (GFR), while correlations with biomarkers (NT-proBNP, troponin T) and structural imaging parameters were negligible. Electrocardiographic findings Sinus rhythm was present in 67.6% of patients, but less frequently in those with Perugini scores 2–3 (58.9%). Ventricular high-frequency activity (28.2%) and conduction abnormalities - such as bundle branch block (34.7%) and AV node block (11.2%) - were more common in higher Perugini groups and positively correlated with myocardial uptake ratios. Atrial fibrillation and pacemaker rhythms also showed weak to moderate positive correlations with both Perugini score and tracer retention indices, while AV node block showed no significant association. Echocardiographic findings Among 156 patients with echocardiographic data, concentric hypertrophy was observed in 82.1% and global wall thickening in 14.1%. Apical sparing, a hallmark of ATTR-CM, was present in 17.3% overall and 21.1% of patients with Perugini scores 2–3. Although no correlation was found between left ventricular hypertrophy and uptake metrics, diastolic dysfunction and elevated filling pressures correlated moderately with myocardial-to-vertebra and apex-to-rib ratios. Notably, Left Ventricular Outflow Tract (LVOT) Vmax, pressure gradients, and velocity time integral (VTI) exhibited strong correlations with both Perugini score and tracer uptake, especially in the apex and lateral wall. No meaningful associations were identified for ejection fraction, aortic valve velocities, gradients, or valve area. Cardiac MRI findings Of 84 patients with MRI data, elevated extracellular volume (ECV) was found in 44.0%, diffuse late gadolinium enhancement (LGE) in 39.3%, and native T1 times were elevated in most cases. Native T1 values showed strong correlations with vertebra-based uptake ratios across all myocardial segments but not with rib-based metrics. ECV and LGE showed weak correlations with uptake values and Perugini scores. MRI-derived ejection fraction was only associated with the Perugini score at 60 min; other volumetric MRI parameters did not show significant correlations. Laboratory findings NT-proBNP and troponin T were significantly elevated in patients with Perugini scores 2–3 (median NT-proBNP: 2,417 pg/mL; troponin T: 43 pg/mL) and showed weak to moderate correlation with all myocardial uptake ratios, particularly those normalized to the vertebral body. Renal function markers such as creatinine and GFR demonstrated weak correlations with vertebra-based ratios, particularly in the apex and lateral wall. Apical-to-rib and lateral-to-rib ratios showed moderate associations with renal function. While urinary creatinine and light chains were not significantly correlated, total urinary protein showed positive associations with Perugini scores and several uptake ratios. Histopathological findings Out of 12 myocardial biopsies, 5 confirmed ATTR - all in patients with Perugini 2–3 - while 7 were negative, including 2 from the same group. Among 17 iliac crest biopsies, 6 revealed AL amyloidosis, but only one of these cases had a Perugini score of 2–3. This highlights the limited sensitivity of extra-cardiac tissue biopsy in detecting cardiac involvement. Histopathological findings were strongly aligned with both Perugini score and tracer uptake values, reinforcing their diagnostic value. To complement the quantitative imaging findings, representative histopathological specimens were analyzed in patients who underwent myocardial or iliac crest biopsy. Standard hematoxylin–eosin (HE) staining demonstrated extracellular, amorphous, eosinophilic deposits within the myocardial interstitium, consistent with amyloid infiltration. In contrast to fibrotic tissue, which typically appears more collagenous and structured, amyloid deposits showed a homogeneous, acellular, and glassy appearance. Trichrome staining further facilitated differentiation between amyloid and fibrosis, highlighting the distinct extracellular accumulation pattern. While collagen fibers stained intensely, amyloid deposits remained relatively pale to grayish, emphasizing their unique histochemical characteristics compared to replacement fibrosis. Immunohistochemical staining using antibodies against amyloid P component (AP) and transthyretin (ATTR) confirmed the presence and subtype of amyloid deposition in positive cases. In biopsy-proven ATTR cases, strong immunoreactivity was observed within the interstitial deposits, correlating closely with elevated myocardial tracer uptake on 99mTc-PYP SPECT/CT and higher Perugini scores. These histopathological findings support the imaging-based assessment and demonstrate concordance between quantitative tracer retention and true amyloid infiltration. Importantly, negative myocardial biopsies in selected Perugini 2–3 patients highlight the potential for sampling error and further underscore the diagnostic value of whole-heart quantitative SPECT/CT imaging. Figure 4 represents histological findings in cardiac amyloidosis. Results of the correlation analyses are summarized in Fig. 5, which illustrates the relationships between key clinically relevant parameters. An extended correlation matrix including additional variables is provided in the Supplementary Material. Fig. 4(A) Hematoxylin–eosin (HE) staining demonstrates extracellular, amorphous eosinophilic amyloid deposits within the myocardial interstitium. (B) Trichrome staining differentiates amyloid from fibrosis, highlighting collagen deposition while amyloid appears pale gray. (C) Immunohistochemical staining for transthyretin (ATTR) shows strong positivity within interstitial deposits. (D) Immunohistochemistry for amyloid P component confirms the presence of amyloid fibrils (A) Hematoxylin–eosin (HE) staining demonstrates extracellular, amorphous eosinophilic amyloid deposits within the myocardial interstitium. (B) Trichrome staining differentiates amyloid from fibrosis, highlighting collagen deposition while amyloid appears pale gray. (C) Immunohistochemical staining for transthyretin (ATTR) shows strong positivity within interstitial deposits. (D) Immunohistochemistry for amyloid P component confirms the presence of amyloid fibrils Fig. 5Correlation heat map for the key clinically patient, imaging, laboratory and SPECT-CT parameters Correlation heat map for the key clinically patient, imaging, laboratory and SPECT-CT parameters Discussion This study evaluated the diagnostic and clinical value of quantitative single-photon emission computed tomography/computed tomography (SPECT/CT) using technetium-99m pyrophosphate (99mTc-PYP) acquired 90 min post-injection in a large cohort with suspected cardiac amyloidosis (CA). Our findings underscore the potential of quantitative metrics, including counts and myocardial-to-contralateral structure ratios, in refining diagnostic accuracy and stratifying patient risk in clinical practice. Beyond overall diagnostic performance, this analysis provides mechanistic insight into the improved accuracy of SPECT/CT, particularly through its impact on diagnostic reclassification and the influence of blood pool activity. More than one-third of patients were reclassified, with the greatest effect observed in those with intermediate planar findings (Perugini score 1), underscoring the limited specificity and heterogeneity of this category on planar imaging alone. These findings are consistent with prior reports highlighting the limitations of planar scintigraphy in equivocal cases and the added value of tomographic imaging for refining diagnosis [4, 5, 9, 11, 13]. Blood pool activity emerged as a key confounder. Residual intravascular tracer was associated with reduced apparent myocardial uptake and increased rates of both upgrade and downgrade, indicating that planar imaging may both underestimate and overestimate myocardial involvement. This aligns with previous observations in early imaging protocols, where blood pool interference can compromise specificity, and further supports the role of SPECT/CT for accurate localization of tracer distribution [9, 12, 13]. Correlation analysis further supports this interpretation, demonstrating an inverse relationship between blood pool activity and myocardial uptake as well as only weak associations with renal function and no meaningful correlations with biomarkers or structural cardiac parameters, suggesting that blood pool activity primarily reflects tracer kinetics rather than disease severity. Importantly, the interaction between Perugini score, myocardial uptake, and blood pool activity reinforces these findings. While patients with Perugini scores 2–3 consistently showed myocardial uptake, those with scores 0–1 demonstrated marked variability, particularly in the presence of blood pool activity. This suggests that intermediate Perugini scores reflect heterogeneous tracer distribution patterns rather than a uniform disease state, highlighting the limitations of visual grading alone and the value of quantitative SPECT/CT for improved characterization [7, 11, 13]. Consistent with prior literature, our quantitative assessment using 99mTc-PYP SPECT/CT exhibited strong concordance with traditional visual grading methods such as the Perugini score. Elevated myocardium-to-vertebra and myocardium-to-rib ratios significantly correlated with higher Perugini scores, demonstrating the efficacy of quantitative imaging in objectively capturing amyloid burden and severity, in line with findings by Avalon et al. [20]. Furthermore, the significant differences in myocardial uptake values across Perugini score categories highlight both the strengths and the inherent limitations of visual grading alone and support the complementary role of quantitative approaches in routine interpretation. A noteworthy observation in our study was the enhanced diagnostic correlation when myocardial uptake was normalized to vertebral structures rather than contralateral ribs. This finding supports the work of Nichols et al., who emphasized the anatomical and physiological stability of vertebrae as reference tissues, contributing to more reliable background normalization [21]. This approach mitigates variability related to respiratory motion, rib positioning, or extracardiac tracer heterogeneity, thereby improving the robustness and reproducibility of quantitative SPECT/CT metrics. Our results also confirm the value of integrating SPECT/CT findings with echocardiographic and cardiac magnetic resonance (CMR) biomarkers. Significant associations were observed between myocardial tracer uptake and echocardiographic hallmarks of CA, including diastolic dysfunction and increased interventricular septal thickness, consistent with prior literature on amyloid cardiomyopathy [22, 23]. Tissue Doppler and strain parameters, such as reduced global longitudinal strain and apical sparing, demonstrated parallel trends with radiotracer uptake, reinforcing their complementary diagnostic relevance [22, 24]. Quantitative SPECT/CT values additionally showed strong correlations with CMR-derived markers such as extracellular volume (ECV), late gadolinium enhancement (LGE), and native T1 times parameters known to reflect amyloid infiltration and myocardial fibrosis, mirroring the multimodal diagnostic framework advocated by Dorbala et al. [13]. The observed positive correlation between LVOT parameters (Vmax and VTI) and myocardial tracer uptake should be interpreted with caution, as it is not necessarily consistent with the expected hemodynamic changes in advanced cardiac amyloidosis. This finding may be influenced by patient heterogeneity, preserved cardiac output in earlier disease stages, or measurement variability, and does not imply a direct physiological relationship. Biomarker analyses further reinforced the imaging findings. Patients with elevated myocardial tracer retention exhibited significantly higher NT-proBNP, troponin T, and serum creatinine levels, consistent with increased myocardial stress and impaired renal clearance commonly associated with transthyretin amyloid cardiomyopathy (ATTR-CM). Histopathological validation demonstrated that patients with high myocardial uptake were reliably diagnosed with ATTR, whereas AL amyloidosis was not associated with elevated Perugini scores, findings that align with the landmark multicenter study by Gillmore et al., which reported > 99% sensitivity and 86% specificity for ATTR using bone-seeking tracers in appropriate clinical constellations [25]. The close agreement between quantitative SPECT/CT metrics and histopathological confirmation further reinforces the biological validity of vertebra-normalized uptake ratios as surrogate markers of true myocardial amyloid burden. Quantitative analysis proved especially beneficial in patients with a Perugini score of 1—a subgroup traditionally associated with diagnostic uncertainty due to intermediate myocardial tracer uptake. In our cohort, these borderline cases frequently exhibited persistent blood-pool activity on planar imaging at both 60 and 180 min post-injection, which can mimic low-grade myocardial retention and substantially limit the specificity of visual interpretation. This diagnostic ambiguity is further compounded by evidence from Quarta et al., who demonstrated that a significant proportion of patients with immunoglobulin light-chain (AL) cardiac amyloidosis show myocardial uptake on bone scintigraphy using 99mTc-DPD, including moderate-to-high uptake in a minority of cases, despite the absence of transthyretin amyloid [26]. Importantly, in that study, myocardial tracer uptake in AL amyloidosis correlated more closely with cardiac dysfunction and biomarker elevation than with amyloid burden on cardiac magnetic resonance imaging, underscoring that myocardial uptake on bone scintigraphy is not pathognomonic for ATTR cardiomyopathy and reinforcing the need for careful etiologic discrimination in low-grade uptake scenarios. By implementing quantitative SPECT/CT imaging at 90 min post-injection, we were able to overcome these limitations through objective measurement of myocardial uptake relative to reference structures such as vertebrae and contralateral ribs. These ratios enabled reliable differentiation between true myocardial tracer retention and residual intravascular activity, with vertebra-normalized uptake values showing strong correlations with histopathology, NT-proBNP and troponin levels, and echocardiographic markers of cardiac dysfunction. Importantly, these findings are concordant with recent evidence by Poledniczek et al., who demonstrated that adjunct thoracic SPECT/CT in Perugini grade 1 patients undergoing bone scintigraphy allows clear discrimination between true myocardial uptake and diffuse non-myocardial or blood-pool activity, thereby refining final diagnosis and improving selection for invasive procedures such as endomyocardial biopsy [27]. However, whereas Poledniczek et al. primarily focused on qualitative and semi-quantitative assessment using delayed 99mTc-DPD imaging in a Perugini grade 1 population, our study extends these observations by applying fully quantitative SPECT/CT metrics acquired at an earlier time point using 99mTc-PYP and by integrating tracer uptake with multimodal imaging, biomarker, and histopathological data across the full spectrum of Perugini scores. Taken together, the findings from Quarta et al., Poledniczek et al., and the present study collectively highlight the limited reliability of planar Perugini scoring alone in low-grade uptake scenarios and underscore the critical role of SPECT/CT, particularly when quantitative metrics are incorporated, in resolving this diagnostic grey zone, minimizing the risk of misclassification between ATTR and AL amyloidosis, and supporting more informed clinical decision-making. Our findings are also in line with those of Régis et al., who demonstrated that quantitative SPECT techniques outperform visual scoring alone in differentiating ATTR-CM from light-chain (AL) amyloidosis, especially in equivocal cases [28]. Taken together, these studies support the enhanced discriminatory power of quantitative SPECT/CT when conventional planar imaging yields inconclusive findings. Given the inherent limitations of myocardial biopsy—including procedural risk, limited availability, and frequent clinical reluctance, particularly in elderly or multi-morbid patients, quantitative SPECT/CT represents a robust, non-invasive alternative for diagnostic clarification, with Perugini score 1 patients emerging as a subgroup that particularly benefits from objective quantification. These results advocate for the routine use of quantitative imaging, especially when biopsy is not feasible. Furthermore, our findings echo the call by Desy et al. for careful attention to platform-specific calibration, reconstruction parameters, and dead-time correction when interpreting quantitative SPECT/CT values across centers [29]. While this study focused on 99mTc-PYP imaging at 90 min post-injection, it is important to emphasize that this acquisition time point applies primarily to PYP. For other radiotracers, such as 99mTc-DPD or hydroxymethylene diphosphonate (HMDP), delayed imaging at approximately 3 h post-injection remains standard practice. Awareness of tracer-specific timing recommendations is essential to avoid misinterpretation of myocardial uptake patterns in amyloidosis imaging protocols [14]. From a clinical workflow perspective, implementation of 90-minute quantitative SPECT/CT may reduce unnecessary biopsies in borderline uptake cases, streamline diagnostic pathways, and improve diagnostic confidence in elderly patients with multimorbidity. Adoption of vertebra-normalized thresholds could facilitate standardization across centers. From a clinical perspective, an important question is whether early 90-minute SPECT/CT imaging could replace delayed planar imaging or should be used selectively in specific patient groups. Based on our findings, SPECT/CT provides clear incremental diagnostic value, particularly in patients with equivocal planar findings (e.g., Perugini Score 1), where it enables differentiation between true myocardial tracer uptake and residual blood pool activity. While early SPECT/CT offers advantages in terms of workflow efficiency and diagnostic confidence, it cannot yet be considered a full replacement for delayed planar imaging, as standardized acquisition protocols and validated quantitative thresholds are still evolving. Instead, a combined or stepwise imaging approach may currently be most appropriate in clinical practice. In particular, SPECT/CT may be especially valuable in cases where planar imaging suggests myocardial uptake but uncertainty remains regarding the presence of true myocardial tracer binding versus intravascular signal. Future prospective studies are warranted to determine whether early quantitative SPECT/CT could serve as a standalone diagnostic modality in selected patient populations. Limitations This study has several limitations that warrant consideration. First, the study was conducted at a single tertiary care center using a standardized imaging platform from a single vendor. While this ensured technical consistency, it may limit the external validity and generalizability of the quantitative thresholds and uptake ratios to other institutions or equipment configurations. Second, although strong correlations were observed between nuclear imaging findings and histopathological confirmation of cardiac amyloidosis, biopsy was performed in only a small subset of patients, approximately 7% of the total cohort. This low biopsy rate was particularly evident among patients with lower Perugini scores or suspected non-ATTR etiologies. As a result, the sensitivity and specificity estimate of the imaging-based diagnosis must be interpreted with caution, as the limited number of histological confirmations may introduce bias and restrict the generalizability of diagnostic performance metrics. Third, echocardiographic and cardiac MRI data were not uniformly available for all patients due to the retrospective nature of the study and heterogeneous clinical work-up. As a result, several clinical, imaging, and laboratory variables contained a substantial proportion of missing values (Tables 2 and 3), leading to reduced sample sizes in certain subgroup and correlation analyses. This may have limited statistical power and introduced potential selection bias. Consequently, associations between quantitative SPECT/CT metrics and advanced imaging parameters, such as native T1 or extracellular volume, as well as selected biomarkers, should be interpreted with caution, particularly in subgroup analyses. Fourth, quantitative SPECT/CT analysis was based on normalization to osseous reference regions (vertebra and rib) without inclusion of soft-tissue background correction. While bone structures provide stable and reproducible reference values with low interpatient variability, the use of additional soft-tissue reference regions may further refine quantitative assessment and reduce potential confounding effects. This aspect was not addressed in the present study and should be considered in future investigations. Fifth, interobserver agreement for quantitative SPECT/CT measurements was not assessed. This omission may impact on the reproducibility of the proposed semiquantitative parameters in routine clinical settings and underscores the need for validation in prospective studies. Lastly, while 90-minute SPECT/CT acquisition improves workflow efficiency and facilitates quantification, early-phase imaging remains susceptible to blood pool activity. This can confound interpretation in intermediate uptake cases, particularly when planar imaging is used in isolation. As highlighted by Gerber and Miller, this represents a known diagnostic pitfall, reinforcing the need for SPECT/CT to reliably differentiate true myocardial retention from intravascular signal [30]. Future prospective, multicenter studies with harmonized imaging protocols and complete multimodal assessments are essential to validate these findings and to standardize quantitative thresholds for broader clinical application. Conclusion Quantitative 99mTc-PYP SPECT/CT at 90 min post-injection provides a robust and clinically meaningful extension of conventional planar Perugini grading. Vertebra-normalized uptake ratios demonstrated strong concordance with multimodal imaging markers, biomarkers, and histopathology, particularly in diagnostically challenging Perugini grade 1 cases. Integration of quantitative SPECT/CT into structured diagnostic algorithms may improve diagnostic confidence, reduce misclassification between ATTR and AL amyloidosis, and support risk stratification and longitudinal disease monitoring. Further prospective validation across centres is essential to establish standardized quantitative thresholds and facilitate broader clinical implementation. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 (DOCX 140 KB) Supplementary Material 1 (DOCX 140 KB)