Authors: Dominik C. Benz (1Division of Nuclear Medicine, Department of Radiology, Brigham and Women’s Hospital, Boston, Massachusetts; 2Cardiovascular Imaging Program, Cardiovascular Division, Brigham and Women’s Hospital, Boston, Massachusetts; 3Cardiac Amyloidosis Program, Division of Cardiology, Department of Medicine, Brigham and Women’s Hospital, Boston, Massachusetts; 4Cardiovascular Imaging, Department of Nuclear Medicine, University Hospital Zurich, Zurich, Switzerland.), Sharmila Dorbala (1Division of Nuclear Medicine, Department of Radiology, Brigham and Women’s Hospital, Boston, Massachusetts; 2Cardiovascular Imaging Program, Cardiovascular Division, Brigham and Women’s Hospital, Boston, Massachusetts; 3Cardiac Amyloidosis Program, Division of Cardiology, Department of Medicine, Brigham and Women’s Hospital, Boston, Massachusetts)
Categories: Article
Source: Heart (British Cardiac Society)
Authors: Dominik C. Benz, Sharmila Dorbala
Systemic amyloidosis is characterized by misfolding of proteins which deposit as insoluble amyloid fibrils in various organs, including the heart. Most cases of cardiac amyloidosis (CA) result from misfolding of immunoglobulin light chains produced by a clonal plasma cell disorder (AL amyloidosis), or transthyretin (TTR) protein produced predominantly in the liver (ATTR amyloidosis). ATTR amyloidosis frequently results from age-related misfolding of the wild-type TTR (ATTRwt), and less commonly, from misfolding of a variant TTR from an autosomal dominant mutation of the TTR gene (ATTRv). The misfolded proteins deposit as insoluble amyloid fibrils in the myocardial interstitial space, disrupting its architecture and causing a prototypical restrictive cardiomyopathy with high mortality. Without treatment, median survival from diagnosis ranges from < 6 months for AL-CA to 3–5 years for ATTR-CA ^1^. Diagnosis is often delayed because of low disease awareness, leading to worse clinical outcomes ^2^. Fortunately, highly effective treatments are available for amyloidosis. Daratumumab, a CD38 monoclonal antibody directed against plasma cells, when added to standard of care therapy with CyBORD (cyclophosphamide, bortezomib, and dexamethasone), substantially improves survival in AL amyloidosis compared to CyBORD alone; daratumumab was approved for treatment of AL amyloidosis in 2022 ^3^. Novel therapies that stabilize (tafamidis, acoramidis) or silence (inotersen, patisiran, vutrisiran) ATTR protein have improved survival and heart failure hospitalizations compared to placebo ^4^. Currently, tafamidis is the only approved therapy for ATTR-CA. Remarkably, a recent report described 3 patients with spontaneous auto antibody-associated reversal of ATTR-CA; also, several therapies targeting the removal of amyloid fibril are currently under investigation^5^. Because of the development of these highly effective drugs, there is an urgent need for accurate diagnosis of CA.
In this Education in Heart section, we discuss the role of imaging, outline clinical features that should raise the suspicion of CA, list imaging features that suggest CA, discuss multisocietal criteria to diagnose CA, and provide an algorithmic approach to the evaluation of CA using multimodality imaging.
Until recently, the diagnosis of CA was made by endomyocardial biopsy in ATTR-CA or AL-CA, and extracardiac biopsy with typical imaging features in AL-CA. Congo red staining of amyloid deposits show apple-green birefringence under polarized light and confirm amyloidosis, while immunohistochemistry stains or mass spectrometry analysis of the amyloid deposits identifies the type of amyloidosis^6^. But biopsy as a diagnostic standard has some limitations. First, endomyocardial biopsy is not widely available and, because it is invasive, many cardiologists order biopsy only in patients with classical imaging features, typically present only in advanced disease. Second, the amyloid content of blind samples obtained from the right side of the interventricular septum may not reflect amyloid burden in the entire heart ^7^. Third, the diagnostic yield of extracardiac biopsy from fat pad is low in ATTR (ATTRv: 45%, ATTRwt: 15%, AL: 84%) ^8^. Finally, with emerging disease modifying therapies, assessment of response to therapy is gaining importance; biopsy is not suitable for this purpose.
Imaging offers accessible and non-invasive estimation of amyloid burden throughout the heart, as well as other affected organs. Quantitative imaging provides assessment of response to therapy. Because each imaging test provides distinct insights into the pathogenesis and the functional effects of amyloid deposits, comprehensive assessment often requires the use of multiple imaging modalities.
Throughout this article, two cases from our clinical practice will be Mr. Abadi (not the real name) with ATTR-CA and Ms. Miller (not the real name) with AL-CA.
Mr. Abadi is a male patient in his mid 60s who presented for mildly progressive exertional shortness of breath. He has a 5-year history of well-controlled hypertension and he has had surgery for lumbar spinal stenosis. His blood pressure was 139/83 mm Hg and his clinical examination was unremarkable. His ECG revealed sinus rhythm, left anterior fascicular block, and first-degree atrioventricular block.
Ms. Miller is a female patient in her mid 50s who presented with progressive exertional shortness of breath, angina, and jaw claudication. Her medical history was unremarkable. On clinical examination, her blood pressure was 90/60 mm Hg and bilateral pedal edema was noted. Her ECG revealed a low voltage signal in the standard leads. Left heart catheterization excluded obstructive coronary artery disease.
Patients with AL or ATTR-CA typically present with manifestations of heart failure with preserved ejection fraction (HFpEF). AL-CA is a rare disease with estimated 4000 new cases diagnosed each year in the United States (https://www.cancer.net/cancer-types/amyloidosis/statistics). The prevalence of ATTR-CA in patients with HFpEF and increased wall thickness ranges from 2% in female and 10% in male outpatient to 13% in hospitalized patients ^9^. Another clinical phenotype is paradoxical low-flow, low-gradient aortic stenosis, typically with mid-range EF; ATTR-CA co-exists in 16% of these patients ^10^. Because both AL and ATTR amyloidosis are systemic diseases, multiple non-cardiac clinical, imaging, and laboratory clues should alert clinicians of the possibility of amyloidosis. These “red flags” are listed in Table 1 ^11^. Musculoskeletal manifestations, fatigue, and gastrointestinal disturbances (constipation and feeling of fullness) are common to both AL and ATTR amyloidosis. Periorbital bruising, macroglossia with changes in speech, jaw or buttock claudication, or peripheral sensory or autonomic neuropathy suggest AL amyloidosis, because they indicate systemic/multiorgan involvement, but may also be seen in ATTR amyloidosis ^12^. The kidney is affected in 70% of patients with AL amyloidosis. A low voltage ECG is neither a sensitive nor a specific marker of CA; normal or increased voltage are seen in half of the patients. However, the presence of conduction disease abnormalities – like left anterior fascicular block and first-degree AV block in Mr. Abadi – are characteristic for CA, with 49% of ATTRwt having 1^st^ degree atrioventricular (AV) block, 51% having wide QRS complex, 10% having high-grade AV block ^13^.
Mr. Abadi had a significantly thickened left ventricle with an interventricular septal thickness of 17 mm (normal <12 mm) on echocardiography (see Figure 1). His septal thickness had increased substantially from 10 mm on an echocardiogram from 5 years prior. The echocardiogram also revealed diastolic dysfunction with a restrictive filling pattern and a severely dilated left atrium (52 ml/m2, normal 34 ml/m2). Global longitudinal strain was reduced (−12%, normal values −18% or lower) with an apical sparing pattern.
Ms. Miller presented for a second opinion to our hospital about 1 year after onset of her initial symptoms. Her echocardiogram showed mild asymmetric LV wall thickening of 14 mm as well as RV thickening and diffuse thickening of the valves, with a restrictive filling pattern and mildly dilated left atrium (38 ml/m2). Global longitudinal strain was reduced (−9%) with apical sparing pattern.
Echocardiography reveals changes in cardiac structure and function induced by amyloid infiltration. Figure 2 illustrates the typical echocardiographic features of advanced CA. Myocardial infiltration with amyloid fibrils increases wall thickness; this is distinct from myocyte hypertrophy due to hypertensive heart disease or increased left ventricular afterload. Generally, LV wall thickness is higher in ATTRwt-CA than in AL or ATTRv-CA, and probably reflects the longer duration of amyloid accumulation ^14^. Myocardial stiffness, mass, and echogenicity are increased in CA. Increased myocardial echogenicity and granular sparkling appearance are characteristic of advanced CA, but they are neither sensitive nor specific for diagnosis ^15^. Additional structural abnormalities include unexplained thickening of the right ventricle (RV), the valves, or the interatrial septum and pericardial or pleural effusions. In addition, echocardiography provides exclusive insights on diastolic function. Advanced amyloidosis is characterized by a small A wave, a high E/A ratio, and a rapid deceleration time on transmitral inflow Doppler and a small S-wave on pulmonary venous inflow Doppler. Annular e’ velocities are markedly reduced on tissue Doppler imaging, with a high E/e’ ratio indicating high filling pressures. Left ventricular ejection fraction is often preserved until late stages, but global longitudinal strain is impaired early in the disease ^14^. Furthermore, the characteristic regional pattern of longitudinal strain, impairment at the mid and basal segments and relative sparing of the apex (i.e., a ratio of apical longitudinal strain / average of longitudinal strain in the mid and basal segments >1) is sensitive for diagnosis (93%), but has limited specificity (82%) to distinguish CA from LV hypertrophy ^16^. Indeed, none of the echocardiographic features can definitively distinguish hypertrophic heart disease from CA, or AL-CA from ATTR-CA. Beyond its diagnostic value, global longitudinal strain is an independent predictor of poor survival in both forms of CA ^14^. For this reason, global longitudinal strain has been proposed to refine the clinical- and biomarker-based Mayo staging scheme for AL amyloidosis ^17^. Furthermore, improvement in strain in response to plasma cell therapy provides incremental prognostic value; an absolute improvement in strain of ≥2% indicates favorable survival ^17^. In patients with ATTR-CA, global longitudinal strain has proven useful to track treatment response; tafamidis stabilized the progression of global longitudinal strain from a median absolute increase of +1.1% per year in untreated patients to +0.3% per year in treated patients ^18^.
Ms. Miller was referred to cardiac MRI, which confirmed the echocardiographic findings. In addition, diffuse, predominantly subendocardial LGE and bilateral pleural effusions were observed (Figure 3).
The characteristic MRI features of CA are shown in Figure 4. Cardiac MRI provides incremental information to echocardiography findings, as well as high-resolution structural and functional images of all cardiac chambers. The major strength of cardiac MRI, however, is the characterization of myocardial tissue by late gadolinium enhancement (LGE) imaging and T1 mapping, as well as estimation of extracellular volume (ECV). In LGE imaging, the operator determines a null point, which is the inversion recovery time at which the normal myocardium is black or “nulled.” In CA, the nulling of the myocardium can be challenging. Furthermore, the myocardial signal nulls prior to the blood pool signal (on an inversion time, TI scout sequence), and this finding is strongly suggestive of CA, with a sensitivity of 100% ^19^. LGE is typically diffuse (but can also be patchy), can be either subendocardial or transmural, and has a high diagnostic accuracy (sensitivity 85%, specificity 92%) for CA. Although LGE is more prevalent and more extensive in ATTR-CA, neither the presence nor the pattern of LGE can differentiate ATTR-CA from AL-CA ^20^. In native T1 mapping, the T1 time of each pixel, which reflects changes in the myocyte and/or the extracellular matrix, is measured prior to gadolinium administration. Compared to normal tissues, native T1 values are elevated in areas of amyloid deposition and abnormal T1 times are scanner-specific, and normal ranges must be established for each scanner. Native T1 mapping has 92% sensitivity and 91% specificity to detect CA ^21^. After contrast administration, post-contrast T1 mapping can be performed to estimate ECV and characterize the extracellular volume expansion. Patients with CA typically have markedly elevated ECV values, often >40% (normal ≤32%) ^22^. Both native T1 mapping and ECV are correlated with other markers of amyloidosis severity ^21, 23^ but neither differentiates the subtypes of CA. Furthermore, overlap in values of native T1 times and ECV with other cardiomyopathies limits the specificity of these metrics for diagnosis of early amyloidosis. Beyond their diagnostic value, LGE, native T1 mapping, and ECV are markers of risk; of these, ECV is as the most robust predictor of survival in AL and ATTR-CA ^24^. More importantly, changes in ECV with chemotherapy are prognostic in AL amyloidosis; survival is worse in patients with >5% increase, and better in patients with <5% increase, in ECV at 6 months ^25^.
Mr. Abadi had normal levels of serum-free light chains (kappa, 12.9 mg/L; lambda, 8.7 mg/L; ratio, 1.5) and normal immunofixation studies in the serum and urine.
In contrast, Ms. Miller had abnormally elevated levels of serum free lambda light chains (kappa, 13.7 mg/L; lambda, 164.0 mg/L; ratio, 0.08) and evidence of a monoclonal protein spike on immunofixation in the serum and urine. She was referred for fat pad biopsy, which confirmed extracardiac amyloidosis (Figure 3). Bone marrow biopsy revealed overgrowth of clonal plasma cells (8%). Her cardiac biomarkers were elevated (NT-proBNP of 38000 pg/ml and high-sensitivity troponin of 89 ng/L). She was diagnosed with AL-CA and started on chemotherapy with daratumumab, cyclophosphamide, bortezomib and dexamethasone.
When CA is suspected, an important initial step is to exclude a monoclonal pathology/AL amyloidosis through hematological testing. This is crucial, as AL amyloidosis is a rapidly progressive and fatal disease. The monoclonal protein that misfolds and accumulates as amyloid fibrils originates in abnormal plasma cells stemming from a family of diseases that include monoclonal gammopathy of undetermined significance (MGUS), smoldering myeloma, and multiple myeloma. The most efficient and effective approach to exclude a monoclonal process is 3 simple laboratory tests to determine 1) serum free light chain levels, 2) immunofixation of the serum, and 3) immunofixation of the urine. If all 3 tests are normal, the negative predictive value for excluding AL amyloidosis approaches 99% ^26^. Normal values for serum free light chains vary by assay. Using the Freelite assay (The Binding Site, Birmingham, United Kingdom), the normal kappa/lambda ratio ranges from 0.26 to 1.65. However, because the kappa light chains are partially cleared by the kidneys, a disproportionate rise in serum-free kappa light chain may occur with renal dysfunction, leading to an elevated ratio, even without a circulating monoclonal protein. Recently, renally-adjusted kappa/lambda ratios in CA have been validated by a large multicenter study ^27^. Abnormal kappa/lambda ratio or monoclonal protein on serum/urine immunofixation is seen in about one in three patients with ATTR-CA, because of the high prevalence of unrelated MGUS in the elderly cohort ^28^. For patients with abnormal light chain studies, a hematology consultation is recommended^29^ and, in that context, a diagnosis of CA requires a cardiac or extracardiac tissue biopsy for amyloid typing.
Mr. Abadi was referred for single-photon emission computed tomography (SPECT)/CT with ^99m^Technetium (Tc)-labeled pyrophosphate (PYP), which was strongly positive (Figure 1). He was diagnosed with ATTR-CA and started on treatment with tafamidis.
For over 40 years, the uptake of bone-avid SPECT radiotracers such as 3-diphosphono-1,2-propanodicarboxylicacid (DPD), PYP, or hydroxymethylene diphosphonate (HMDP) in the myocardium has been known to signify presence of CA. In recent years, researchers have recognized that this imaging modality has high diagnostic accuracy for ATTR-CA, while being less sensitive for AL-CA. The exact mechanism of bone avid tracer uptake in the myocardium is yet to be elucidated. Studies in animal models of myocardial infarction by Buja et al., have shown binding of ^99m^Tc-PYP to tissue calcium stores including, calcium bound to macromolecules, myofibrils, as well as amorphous calcium phosphate and crystalline hydroxyapatite^1^.Consequently, bone-avid tracer SPECT is currently used primarily for diagnosing suspected cases of ATTR-CA. This modality provides additional diagnostic value over echocardiography and cardiac MRI by differentiating ATTR amyloidosis from other forms of heart disease with LV thickening. Because these tracers remain in the blood pool for prolonged period, current practice is to obtain images 2–3 hours post radiotracer injection using SPECT imaging; SPECT/CT is preferred, if available, for confirming myocardial uptake of radiotracers. Bone-avid tracer SPECT images are interpreted visually using a grading scale of 0–3, indicating uptake less than, equal to, and greater than rib uptake, respectively (Figure 5). A large multi-center study has demonstrated the high specificity (nearly 100%) and positive predictive value of Grade 2/3 myocardial uptake for ATTR-CA in symptomatic patients with typical imaging features and no monoclonal protein. However, the sensitivity of this finding was only 72%, underscoring the need for continued evaluation of ATTR-CA, particularly in patients with clinical or imaging features of infiltrative cardiomyopathy. Other causes of a positive cardiac bone-avid tracer SPECT include AL-CA, acute or subacute myocardial infarction, hydroxychloroquine cardiotoxicity, and certain rare forms of CA ^29^. False-negative results may occur in certain hereditary forms of ATTRv, including Phe64Leu and Val30Met, or at very early stages of ATTR-CA. Currently, quantitative SPECT/CT imaging, yielding standardized uptake value (SUV) and other metrics of cardiac amyloid burden, is under investigation for early diagnosis and assessment of treatment response.
There are several methods for diagnosing CA. A recent expert consensus document outlined the imaging features of CA and provided diagnostic criteria for both AL-CA and ATTR-CA^6^. The three main approaches to diagnosis of CA are endomyocardial biopsy, extracardiac biopsy, or imaging using cardiac SPECT with bone-avid radiotracers (as shown in Table 2).
The evaluation algorithms for CA are constantly evolving with increasing knowledge about this disease. We suggest a step-by-step approach, as shown in Figure 6, to aid in the early and accurate diagnosis of CA.
In everyday practice, physicians can quickly perform the correct diagnostic workup, including hematological tests and bone-avid tracer cardiac SPECT, when they are aware of the clinical and imaging red flags discussed in this review (Table 1). A recent position statement from the European Society of Cardiology Working Group on Myocardial and Pericardial Diseases has defined the patients at risk of CA who benefit from screening (Table 3) ^30^.
Cardiac amyloidosis imaging is rapidly advancing. Intrinsic cardiac elastography, myocardial work indices, measures of lung water on cardiac MRI, exercise imaging, and quantitative bone-avid tracer cardiac SPECT are emerging novel techniques that provide additional information beyond traditional methods. Exciting new developments in molecular imaging of amyloidosis using amyloid-binding PET radiotracers (^11^C-Pittsburgh compound-B, ^18^F-florbetapir, ^18^F-florbetaben, ^124^I-evuzamitide) provide visualization of patterns of amyloid fibril deposition and quantification of cardiac and whole-body amyloid burden.
Early diagnosis of CA is crucial for effective management. Identifying clinical, echocardiography, and cardiac MRI red flags can raise suspicion of CA and prompt further diagnostic evaluation, including hematological testing, cardiac SPECT with bone-avid radiotracers, and biopsy when appropriate. Fortunately, the most prevalent form of CA, ATTR-CA, can be diagnosed noninvasively using cardiac SPECT with bone-avid radiotracers. Early diagnosis and initiation of therapy directed against ATTR or AL protein can stabilize the disease, reduce heart failure hospitalizations, and enhance survival rates for patients with CA.