Early diagnosis of CA is crucial for initiating timely and appropriate treatment, as the prognosis worsens with the advancement of heart disease.¹ When clinical suspicion of amyloidosis arises, imaging methods become invaluable for enabling a non-invasive diagnosis.² These CA diagnostic techniques include electrocardiography, echocardiography, CMR, and nuclear imaging methods like pyrophosphate scintigraphy with technetium (PYP).^8,9

ECG patterns that raise suspicion for CA include (1) low voltage QRS complexes, (2) pseudo infarction criteria in the septal wall, and (3) conduction abnormalities, such as bundle branch block and atrioventricular conduction defects.¹⁰ However, the diagnosis of amyloidosis by ECG is challenging because it has patterns similar to those of other myocardial infiltrative diseases.¹⁰ Algorithms that are easy to apply in clinical practice can be employed to improve sensitivity, such as that proposed by Schrutka et al., who described patterns that increase the chance of diagnosing CA with an odds ratio of 7.66 (95% CI: 5,47 – 10,72).¹¹ However, it is important to consider that typical CA patterns on the ECG are not conclusive for diagnostic and are considered to be red flags. These findings are often the starting point for continuing to investigate the disease with other imaging methods.

Newer approaches apply artificial intelligence (AI) to evaluate the ECG, which may be relevant for detecting patterns not easily identified by the human eye. Grogan et al. developed an AI method trained with data from the Mayo Clinic and obtained an area under the receiver-operating-characteristics (ROC) curve of 0.91 (95% CI: 0.90 – 0.93), in addition to predicting CA more than six months in advance of the clinical diagnosis in 59% of patients with CA or suspected CA.¹²

Echocardiography is often one of the initial tests requested when CA is suspected, given its widespread availability and ability to estimate various parameters of interest that may exhibit suggestive patterns in patients with amyloidosis. These findings typically manifest in the disease's advanced stages and primarily resemble restrictive cardiomyopathy of the infiltrative type.¹³ Among the parameters assessed in echocardiography, several are associated with an amyloidosis pattern, including thickening of the left ventricle (LV) walls, thickening of the interatrial septum and mitral valve leaflets, reduction in the LV cavity's diastolic diameter, and the LV apical sparing pattern in longitudinal strain (also known as the "cherry on top" sign, indicating exclusive strain preservation at the apex).^13,14

Although many studies indicate an association between apical sparing and CA, this pattern is not always found. A study by Wali et al. sought to evaluate the prevalence of this pattern in patients with CA, concluding that approximately 1/3 of patients had apical sparing.¹⁵ Although some echocardiography parameters indicate the possibility of CA, they are not specific and may relate to other cardiomyopathies with ventricular hypertrophy. In this context, some studies suggest the use of AI to differentiate between types of amyloidosis. For example, Zhang et al. proposed an AI method based on texture characteristics of images to distinguish CA from other conditions like hypertrophic cardiomyopathy, uremic cardiomyopathy, and hypertensive heart disease.¹⁶

Magnetic resonance imaging plays a role in diagnosing CA by assessing myocardial tissue characteristics and cardiac involvement. CMR can provide valuable diagnostic insights by evaluating both cardiac morphology and function, thus facilitating the differentiation of CA from other cardiomyopathies.¹³ Various CMR parameters, such as delayed gadolinium enhancement, extracellular volume, and the R1 myocardium-light ratio on post-contrast T1 maps, have been studied for their diagnostic accuracy in detecting CA. In the study by Kidoh et al., the techniques of extracellular volume and R1 myocardium-light ratio on T1 map were evaluated for detection of CA, in which the area values under the ROC curve were respectively 0.99 (95% CI: 0.97 – 1.00) and 0.98 (95% CI: 0.95 – 0.99), with no difference between the two techniques for this objective (p = 0.10).¹⁷ Dohy et al. evaluated the use of CMR for diagnosing AL-type CA and found a specificity of 88% and 100% sensitivity for late gadolinium enhancement. In comparison, global strain reduction with apical sparing showed low sensitivity (6%) but very high specificity (100%). Another highly diagnostic finding is a septal wall thickness > 15 mm, which can differentiate CA from the AL type of hypertensive heart disease with a sensitivity of 91% and a specificity of 89%.¹⁸

One limitation of CMR is its inability to differentiate between TTR and AL types of CA. In this regard, Germain et al. proposed the use of an AI model to attempt this classification, comparing its performance with that of human readers.¹⁹ While one of the models showed slightly better results than human evaluation (75% accuracy vs. 67.5%), neither achieved sufficient accuracy for clinical use. Therefore, other modalities are needed for this differentiation.

Endomyocardial biopsy is the most sensitive method for detecting amyloid deposits and allows for the detection of concomitant diseases and classification by type.²⁰ Congo red staining is used for characterization, showing an anomalous apple-green birefringence under polarized light.²⁰ For AL amyloidosis, less invasive procedures such as subcutaneous fat or salivary gland biopsies can confirm the diagnosis in 50%-80% of cases. However, for ATTR, endomyocardial biopsy is the preferred method for histopathological diagnosis despite being riskier and requiring greater expertise.²¹

Based on studies that have explored the use of tracers, mainly bisphosphonates, to evaluate CA, Perugini et al. conducted a pilot study aiming to differentiate between ATTR and AL types of CA using scintigraphy with 3,3-diphosphate-1,2-propanedicarboxylic-99mTc acid (DPD-Tc^99m).²¹ The study indicated a higher uptake of the tracer by ATTR patients, suggesting a potential correlation with this type of amyloid deposit. Since then, other studies have expanded the knowledge base of this technique, as this pilot study was conducted with only 25 patients. It is worth mentioning that in the original study by Perugini et al.,²¹ there were only 10 patients with AL amyloidosis and none exhibited accumulation of bone radiotracer in the myocardium. However, subsequent studies have shown that a significant portion of patients with AL amyloidosis also exhibit an accumulation of bone radiotracers, which shows that cardiac scintigraphy with bone radiotracer alone is not enough to diagnose CA-ATTR and is necessary to rule out the presence of AL amyloidosis by investigating the presence of free ALs.²²

A multicenter study (USA and Europe) with 1,217 patients with suspected CA was conducted to evaluate the diagnostic power of this scintigraphy method with bone markers (DPD, pyrophosphate and Hydroxymethylene diphosphonate labeled with Tc^99m), in which there was prior evidence of its craving for amyloid deposits. However, the molecular basis of this process is not known.²³ The results reported by Gillmore et al. indicated a positive predictive value of 100% (95% CI: 98% - 100%) for the diagnosis of ATTR-type amyloidosis, if monoclonal gammopathy has been ruled out (through testing for free ALs in plasma and urine).²³ Therefore, for patients in whom AL-type amyloidosis has been ruled out, pyrophosphate scintigraphy is a diagnostic method with very high specificity, meaning there is no need for endomyocardial biopsy for confirmation and consequently reducing the risks for the patient and improving their prognosis.

The 99mTc-Pyrophosphate (PYP) is a radiopharmaceutical discovered in the 70s. PYP was initially used for skeletal imaging due to its ability to fixate in regions of osteogenesis, being soon replaced by more stable markers for bone studies. However, researchers soon noted its application in the detection of necrotic cardiac areas, especially in the acute phase of acute myocardial infarction.²⁴ The uptake of compounds labeled with metastable technetium 99 in calcium deposits in patients with Amyloidosis was observed in the late 70s, allowing the use of 99mTc-PYP and other phosphate-derived compounds in the evaluation of this disease.²⁵

Wizenber et al.²⁶and Falker et al.²⁷demonstrated the sensitivity of 99mTc-PYP in the diagnosis of CA, correlating uptake with positive biopsies and morphological changes in the LV, as assessed by echocardiography. However, heterogeneous results in other studies raised doubts about this correlation. As previously mentioned, Perugini et al. used a visual scoring method to analyze scintigraphic images. Their study demonstrated that TTR amyloidosis cases exhibited abnormal cardiac uptake of the radioactive tracer, with intensity surpassing that of the costal arches.²¹The imaging protocol with 99mTc-PYP involves administering 10 to 20 mCi of the radiopharmaceutical and acquiring images at 1 and 3 hours post-administration. Planar thoracic images are obtained to quantify myocardial uptake. In contrast, tomographic sections obtained through SPECT are crucial for confirming that the radiotracer accumulation occurs in the walls of the LV and not in the blood within the heart cavities (blood pool), which are common causes of false positives when analyzing planar images in isolation.²⁸

Figure 1

The Perugini score classifies myocardial 99mTc-PYP uptake compared to the ribs in a planar image of the chest 3 hours after the injection on a semiquantitative scale from 0 to 3. The extensive study by Gillmore et al. revealed that Grades 2 and 3 uptakes, in the absence of monoclonal protein, exhibited 100% specificity and positive predictive value for ATTR-CA, obviating the need for invasive biopsy for diagnosis.²³

While the mechanism behind 99mTc-PYP uptake in ATTR CA remains uncertain, the prevailing theory implicates calcium in amyloid fibrils. Despite this uncertainty, pyrophosphate scintigraphy proves to be a non-invasive method with high diagnostic accuracy, facilitating appropriate treatment. In addition to pyrophosphate scintigraphy, other diagnostic methods play crucial roles in the investigative process. Proper imaging protocol is vital to avoid false positives, as uptake observed in planar images should always be confirmed in tomographic scintigraphy (SPECT) images to distinguish uptake in the myocardium from uptake in the blood pool, which can lead to false positives. Furthermore, it is imperative to rule out monoclonal gammopathy, as up to 40% of AL amyloidosis cases may exhibit pyrophosphate uptake in the myocardium. Therefore, adherence to investigation algorithms is essential for accurately diagnosing the type of amyloidosis.¹³

Several imaging methods have been proposed for quantitatively assessing cardiac amyloid overload, which can help in prognostic stratification and monitoring disease progression.

Reviews of the literature have consolidated the use of multimodality imaging in evaluating CA.^29-31 For example, a systematic review by Cai et al., involving 955 patients across nine studies, suggests using this imaging modality to assess prognosis. They found relative risks for mortality of 1.33 (95% CI = 1.10; 1.60) for each 60 ms increase in T1 time, 1.16 (95% CI = 1.09; 1.23) for each 3% increase in ECV, and 5.23 (95% CI = 2.27; 12.02) for a below-average myocardial-skeletal ratio.³¹ Pyrophosphate scintigraphy has been extensively studied to explore its ability to predict clinical outcomes and monitor disease progression.

Quantitative assessment of myocardial bone radiotracer activity appears to be a non-invasive measure that correlates with cardiac amyloid protein overload and may be useful for assessing disease progression or response to therapy. In a study by Miller et al., increased cardiac pyrophosphate activity in patients with ATTR cardiomyopathy was associated with a lower left ventricular ejection fraction (OR of 1.28) and a higher rate of hospitalizations for heart failure (hazard ratio of 1.29).³² Similarly, Roshankar et al. found a significant hazard ratio for cardiac events and death due to increased tracer activity and volume of involvement.³³

Further studies using pyrophosphate scintigraphy are necessary to validate its effectiveness in monitoring disease progression and response to targeted treatments, such as AMYTRE.⁴ Additionally, quantitative assessment of bone radiotracer activity in the myocardium seems to be a non-invasive and promising approach for monitoring disease evolution and therapeutic responses. This potential was demonstrated in studies by Miller et al.³² and Roshankar et al.³³ This evidence underscores the significance of integrating various imaging modalities for the optimal management of CA. Figure 2 contains an exemplary case in which the application of quantitative analysis tools for myocardial radiotracer uptake in sequential cardiac scintigraphy exams with 99mTc-Pyrophosphate, planar and SPECT showed a reduction in uptake over time.

Figure 2

Porcari et al. with a study of 1422 patients demonstrated diffuse, as opposed to focal, right ventricle uptake of bone tracer on SPECT imaging was associated with poor outcomes in patients with ATTR-CM and is an independent prognostic marker at diagnosis.34 So, besides informing the presence and intensity of myocardial uptake of bone tracer agents it is very useful to discriminate the uptake in the right ventricle and its pattern to obtain the most prognostic information.

Pyrophosphate scintigraphy has been integrated into AI models to enhance diagnostic precision and mitigate the subjectivity of human eye interpretation in image analysis. A recent study examined over 19,000 scintigraphic examinations using various tracers and protocols. The findings showed that AI-driven screening for uptake indicative of CA in scintigraphy patients was dependable, eliminating inter-rater differences, and offered prognostic value. These results have potential implications for identifying, referring, and managing patients clinically.³⁵