Keys to early diagnosis of cardiac amyloidosis: red flags from clinical, laboratory and imaging findings

Abstract

Cardiac involvement in systemic amyloidosis, due either to immunoglobulin light-chain or transthyretin amyloidosis, influences clinical presentation and is a strong predictor of unfavourable outcome. Until recently considered as a rare, incurable disease, cardiac amyloidosis, is still mis/underdiagnosed, although treatments effective in improving patient survival are now available for both subtypes, including chemotherapy regimens for immunoglobulin light-chain amyloidosis and tetramer stabiliser for transthyretin amyloidosis. Achieving a timely diagnosis allows initiating life-saving therapies and requires the early recognition of clinical, laboratory and imaging signs of cardiac involvement, some of them may be apparent well before the disease becomes clinically manifest. Given the systemic nature of amyloidosis, a close interaction among experts in multiple specialties is also required, including cardiologists, nephrologists, haematologists, neurologists, radiologists, nuclear medicine specialists and internists. As an increased awareness about disease presentation is required to ameliorate diagnostic performance, we aim to provide the clinician with a guide to the screening and early diagnosis of cardiac amyloidosis, and to review the clinical, biohumoral and instrumental ‘red flags’ that should raise the suspicion of cardiac amyloidosis.

Cardiac amyloidosis, diagnosis, red flags

Cardiac amyloidosis: why to suspect it

Amyloidosis is a systemic disease characterised by extracellular deposition of insoluble fibrils, deriving from proteins encoded by mutated genes or normal, misfolded proteins. Cardiac involvement in amyloidosis is a major determinant of clinical presentation and outcome, and is common in immunoglobulin light-chain (AL) amyloidosis and transthyretin amyloidosis (ATTR).

AL amyloidosis is the most common type of systemic amyloidosis, affecting approximately 10 people per million per year.¹ Monoclonal light chains undergo extracellular misfolding and aggregation into proteotoxic soluble oligomers and, finally, into amyloid fibrils. The outcome of patients with AL amyloidosis is poor, especially when cardiac involvement is present.² Nonetheless, given recent advances in haematological therapy and better support therapy for cardiac amyloidosis (CA), patient survival is improving.³

TTR is a highly conserved protein rich in beta strands which is synthesised mostly by the liver and involved in the transportation of thyroxine and retinol-binding protein. Wild-type (wt) TTR has an intrinsic propensity to aggregate into insoluble amyloid fibres. The process of fibrillogenesis requires the dissociation of TTR homotetrameric structure into misfolded monomers that self-assemble in soluble oligomeric species, protofibrils and finally mature amyloid fibres, which deposit within tissues. ATTR can follow the deposition of either mutant (variant ATTR, vATTR) or wild-type TTR (ATTRwt). Single-base substitutions resulting in missense mutations represent the majority of genetic alterations in vATTR; the clinical presentation of vATTR is largely influenced by the underlying mutations, ranging from pure polyneuropathy (as in familial amyloid polyneuropathy), to the coexistence of neurological and cardiac disease, to isolated cardiomyopathy. The most common mutation in vATTR is a point mutation causing the replacement of a valine with a methionine at position 30 (V30M).⁴ The epidemiology of vATTR is difficult to define as its prevalence displays significant regional variations, although in a recent study 5% of patients initially diagnosed with left ventricular (LV) hypertrophy had TTR gene mutations.⁵ Previously termed senile CA, as the heart is usually the most severely affected organ, ATTRwt is a sporadic disorder, most often affecting men. Several reports have challenged the common view of ATTRwt as a rare disease. Autopsy studies suggest that the prevalence of ATTRwt may be as high as 25% of individuals older than 85 years,⁶ and 13% among patients with heart failure and preserved ejection fraction (HFpEF).⁷ Tafamidis, a tetramer stabiliser, has recently been proven to be the first therapy able to modify the natural history of ATTR cardiomyopathy, reducing all-cause mortality and cardiovascular-related hospitalisations,⁸ and many other molecules are currently being tested.⁹

As the therapeutic armamentarium for both AL amyloidosis and ATTR-related cardiomyopathy is increasing, and the management of CA is substantially different from other forms of heart failure (HF)¹⁰ or hypertrophic cardiomyopathy (HCM),¹¹ clinicians possibly dealing with amyloidosis patients (including, but not limited to, cardiologists, internal and nuclear medicine specialists, nephrologists, neurologists and general practitioners) should be aware of this condition, which remains a highly underdiagnosed disorder, with often long times to diagnosis.¹² In the present paper, we summarise the red flags, i.e. those clinical, biohumoral and imaging findings prompting the diagnosis of a specific disease, that should raise a suspicion of CA (Figure 1).

Figure 1.

Clinical, electrocardiographic, biohumoral and imaging findings raising the suspicion of cardiac amyloidosis. Darker areas indicate higher diagnostic specificity. AV: atrio-ventricular; BNP: B-type natriuretic peptide; CMP: cardiomyopathy; CTS: carpal tunnel syndrome; GFR: glomerular filtration rate; HFmr/pEF: heart failure with mid-range/preserved ejection fraction; LGE: late gadolinium enhancement; NT-proBNP: N-terminal fraction of B-type natriuretic peptide; RV: right ventricular.

Clinical features

AL amyloidosis or ATTR-related CA is typically identified when signs and symptoms of restrictive cardiomyopathy and/or HFpEF develop, including dyspnoea, reduced exercise tolerance, peripheral oedema, hepatomegaly, ascites and elevated jugular pressure. Nonetheless, such a presentation is usually observed in advanced stages of disease, while cardiac involvement may be more subtle or even asymptomatic in earlier phases.¹³

AL amyloidosis is a systemic disease, with half of the patients having renal involvement, 16% having liver disease and 10% having neuropathy.¹⁴ Information about the coexistence of such conditions should be acquired as a part of routine patient interview. Phenotypic heterogeneity is also significant in the case of vATTR, following differences in the responsible mutation, penetrance, ethnicity and geographical area. Some mutations are associated with combined neurological and cardiac manifestations, whereas others have exclusively either a neurological or, less frequently, cardiological presentation.¹⁵ Especially in endemic areas, a family history of hereditary amyloidosis or, more commonly, cardiac hypertrophy presenting in aged individuals, should prompt the initiation of screening measures even in asymptomatic individuals. Extra-cardiac disease is less frequent in ATTRwt, although clues to systemic amyloid deposition come, for example, from carpal tunnel syndrome (which is observed in half of the patients and often precedes cardiac manifestations by several years),¹⁶ or from rupture of the long head of the biceps tendon, producing the ‘Popeye sign’, both representing clinically relevant red flags.¹⁷ Although generally affecting the elderly male population, recent reports have shown that women account for up to 20% of ATTRwt cases.¹³ In addition, ATTRwt is increasingly being recognised as a cause of HFpEF, accounting for a considerable number of cases in recent series.⁷ Notably, other possible causes of myocardial hypertrophy do not exclude the diagnosis of CA. For example, hypertension is present in more than half of patients with ATTRwt, and significant mitral or aortic diseases (possibly even requiring transcatheter aortic valve replacement) may be encountered.¹³

Electrocardiogram: common and less common findings

The electrocardiogram (ECG) in CA may show low-voltage QRS complexes. This can be attributed to the accumulation of non-conducting amyloid fibres, and possibly also to myocardial oedema (which might explain why the low-voltage pattern is more common in AL than ATTR amyloidosis despite a greater amyloid burden in the latter condition).¹⁸ A decrease in QRS voltages is a red flag for the disease, often preceding a significant increase in LV wall thickness. A characteristic discrepancy between peripheral and precordial leads is commonly observed, with peripheral leads showing low voltages while voltages in precordial leads are normal or occasionally high. This discrepancy is not observed with other causes of low-voltage ECG patterns, namely pericardial or pleural effusion, obesity, emphysema, pneumothorax, or myxoedema.^19,20 Among 337 consecutive treatment-naive AL amyloidosis patients (233 with cardiomyopathy and 104 without), the estimated prevalence of low-voltage patterns in AL amyloidosis ranged from 84% when the Sokolow–Lyon index of 15 mm or less was used as a cut-off, to 27% when low voltages were defined as QRS amplitude of 5 mm or less in each peripheral and 10 mm or less in each precordial lead.²¹ In other cohort studies on AL amyloidosis, using different diagnostic criteria, the prevalence of low QRS voltages ranged from 42% to 71%.^18,,–21 Finally, among patients with biopsy-confirmed CA, only 60% (AL) and 35% (ATTR) displayed low peripheral voltages (<5 mV in all peripheral leads).¹⁸ Overall, despite the multiplicity of diagnostic criteria, the absence of a low-voltage ECG pattern does not rule out AL or ATTR-type CA and the discrepancy between QRS voltages and LV mass as measured by echocardiography could be even more relevant than QRS voltages per se.

Conduction disturbances are also common in patients with CA. Among 344 patients diagnosed with AL amyloidosis, those with cardiac involvement (n = 240) displayed prolonged PQ, QRS, QT and QTc intervals (all P < 0.05) and a higher prevalence of intraventricular blocks (28% vs. 17%, P < 0.05).²² Other ECG features described in patients with AL-type CA are a pseudo-infarction pattern with QS complexes in the anterior leads, an unusual axis of the QRS complex (particularly extreme right axis deviation),¹⁹ abnormal P wave morphology and duration (reflecting slow atrial conduction). In AL-type CA, a fragmentation of QRS complexes has been reported, consisting of abnormalities such as notches and a RsR′ pattern in the absence of QRS prolongation, reasonably due to areas of myocardial necrosis and fibrosis.²³ Atrial fibrillation is also rather frequent in patients with amyloidosis. For example, more than half of patients with ATTRwt had any form of atrial fibrillation in a recently published series.¹³

Laboratory findings

Sensitive biomarkers are required to identify cardiac damage at an early, asymptomatic stage. Such markers might be particularly useful in patients considered at higher risk for the development of amyloidosis, such as those with monoclonal gammopathy of unknown significance (MGUS), at risk of AL CA especially when an abnormal free light chain κ/λ ratio is present,²⁴ and carriers of TTR gene mutations, searchable by genetic screening when needed to differentiate vATTR and ATTwt amyloidosis.

With regard to the confirmation of the suspicion of AL CA the search for the κ/λ monoclonal protein either in serum or urine, although not detectable in 50% of patients, is necessary, while the increase in free light chain blood concentration is essential to diagnosis, useful for risk stratification and evaluation of treatment efficacy.¹⁴

Natriuretic peptides and troponins are invariably elevated in patients with CA, following interstitial amyloid deposition or direct toxic effects of light chains on cardiomyocytes.²⁵ In patients with AL-type CA, B-type natriuretic peptide (BNP) and N-terminal proBNP (NT-proBNP) are produced and released into the bloodstream following volume overload and direct toxicity by immunoglobulin light chains. Among 152 subjects with AL amyloidosis, those with cardiac disease never had NT-proBNP levels lower than the 97.5 percentile for normal individuals, indicating 100% sensitivity; furthermore, a 1285 ng/L NT-proBNP cut-off had 92% accuracy for the detection of cardiac involvement.²⁶ As NT-proBNP increase predicts HF development in AL amyloidosis,²⁷ routine NT-proBNP testing is recommended during the follow-up of patients with MGUS and an abnormal κ/λ ratio.²⁴

Unexplained or disproportionate circulating levels of natriuretic peptides may also support the clinical and instrumental suspicion of ATTR cardiomyopathy, in particular in high-risk populations, such as carriers of TTR mutations (either asymptomatic or with only neuropathic symptoms),⁵ or in patients with conditions mimicking or associated with ATTR, such as aortic stenosis or HFpEF.⁷ Natriuretic peptides also reflect the severity of cardiac amyloid burden,²⁸ and patients with elevated BNP and NT-proBNP at diagnosis are at higher risk of mortality.²⁹

Cardiac troponins (cTn) are specific markers of myocardial injury. Newer, high-sensitivity assays allow us to detect much lower concentrations, thus refining diagnosis and risk stratification in both acute and chronic settings, as in HF and cardiomyopathies. The elevation of plasma highly sensitive troponin T (hs-TnT) is observed in the vast majority of patients with AL amyloidosis, including those free from overt cardiac involvement, and represents a red flag for the disease. Moreover, hs-TnT is associated with clinical indices of HF severity, LV systolic dysfunction, and wall thickness in patients with AL-type CA.³⁰ The exact mechanism of cTn release in patients with amyloidosis has to be elucidated, but the proinflammatory or toxic effects of the amyloid fibres or their precursors and/or microvascular ischaemia may lead to cardiomyocyte damage and troponin release. cTn is commonly elevated in patients with ATTR cardiomyopathy, although to a lesser extent than in AL-type CA, despite a more severe increase in wall thickness and worse LV systolic function.³¹ Nonetheless, evidence on the diagnostic properties of circulating biomarkers in CA is still limited.^32,33 Notably, natriuretic peptide and cTn levels must be interpreted based on renal function and/or atrial fibrillation, which are commonly encountered in these patients and affect circulating levels of these biomarkers.

Confirmation and refinement of the diagnosis may include the search for amyloid deposits either at the periumbilical fat level, or in selected cases with the suspicion of AL CA, by endomyocardial biopsy.²⁴

Echocardiography

Echocardiography is the most widely used non-invasive test in patients with HF or abnormal cardiac findings on examination. The most common two-dimensional (2D) echocardiographic findings observed in CA are biatrial dilation, increased LV and right ventricular (RV) wall thickening, and normal or small LV cavity size; occasionally, a pericardial effusion can be found.³⁴ Concentric LV pseudohypertrophy is seen in over 90% of cases of CA, although asymmetric septal hypertrophy, LV outflow tract obstruction, or even normal wall thickness, non-dilated or minimally dilated LV, and reduced ejection fraction may be encountered. The myocardium can acquire a granular sparkling appearance, which is attributed to the increased echogenicity of the amyloid protein.³⁴ In addition, amyloid deposition accounts for the mismatch between QRS voltages and the degree of LV hypertrophy, as described above.

Doppler imaging usually reveals only mild valvular dysfunction, while diastolic function is often significantly impaired.³⁴ At the initial stages, an abnormal relaxation pattern is observed, possibly progressing to a restrictive pattern. Peak early diastolic velocity (E′) assessed by tissue Doppler imaging is decreased in the earliest stages of the disease, and further decreases with the disease progression, which helps differentiate CA from disorders such as constrictive pericarditis or HCM, in which E′ is normal or mildly reduced.³⁴

Strain imaging demonstrates an impairment of longitudinal ventricular contraction before the decline of ejection fraction and HF development. In a study of 40 patients with biopsy-confirmed CA and 60 patients with HCM or hypertensive heart disease, LV global longitudinal strain (GLS) showed the best performance to discriminate CA, both in the whole population and in the subgroups with maximum wall thickness of 16 mm or less or left ventricular ejection fraction (LVEF) greater than 55%.³⁵ As GLS is impaired much earlier than LVEF, the LVEF strain ratio (LVEF/absolute value of global longitudinal strain – GLS) was proposed, and found to be significantly higher among CA patients than other individuals with myocardial hypertrophy, with a cut-off value of 4.1 being associated with an area under the curve of 0.91 for the differentiation of CA from HCM or healthy controls.³⁶ Together with GLS, circumferential and radial deformations were reported to be significantly decreased in patients with CA.³⁷ A severe impairment of basal longitudinal strain with relative preservation of apical strain is typically observed. This ‘apical sparing’ is found in both AL and ATTR cardiomyopathy, and is both sensitive and specific for the diagnosis of CA.³⁸

RV dilation may occur in late disease stages, probably because of a combination of pulmonary hypertension and RV systolic dysfunction due to infiltration. LA function is very often impaired in CA, as assessed through conventional and strain-derived parameters of LA function, reflecting reservoir, conduit and contractile LA functions.³⁹ The prevalence of LA thrombi is remarkably high, even among patients in sinus rhythm.⁴⁰ Peak LA longitudinal strain and active LA emptying fraction were reported to be more altered in patients with ATTRwt than AL and vATTR.³⁹

Cardiac magnetic resonance

Cardiac magnetic resonance (CMR) allows us to evaluate biventricular morphology and size, functional parameters (particularly systolic function) and myocardial tissue composition. CMR may reveal almost pathognomonic findings in advanced amyloidosis, and may lead to a non-invasive diagnosis of ATTR cardiomyopathy when combined with clinical and electrocardiographic data, diphosphonate scintigraphy and the exclusion of a monoclonal protein.⁴¹

CA is frequently associated with extra-cardiac abnormalities, most notably pleural effusion (possibly accompanied by pericardial effusion) and ascites.⁴² Asymmetrical hypertrophy has been traditionally associated with HCM, but emerged as the commonest pattern of ventricular remodelling in ATTR-type CA (79%), followed by symmetrical concentric hypertrophy (18%) and the absence of LV hypertrophy (3%). Asymmetrical hypertrophy was much less frequent in AL amyloidosis (14%), while concentric hypertrophy was found in 68%, and no hypertrophy in 18%.⁴³ Hypertrophy is typically more pronounced than in hypertensive or valvular heart disease, and is more prominent in ATTR than AL amyloidosis.⁴² Contrary to HCM or other forms of secondary hypertrophy, changes over months in wall thickness and function are quite typical of AL amyloidosis.⁴² The LVEF may be normal even into the late phase of disease, while indexed stroke volume is usually severely reduced.⁴² Atrial wall thickening, dilation and thrombosis in the LA appendage can be observed. Finally, an increase in right ventricular mass is quite common.⁴²

At myocardial tissue characterisation, the typical amyloid pattern consists of global subendocardial late gadolinium enhancement (LGE), which may be combined (particularly in AL amyloidosis) with blood pool early darkening due to the rapid contrast washout from the blood pool into the systemic interstitial space, which is extremely enlarged by amyloid deposition. The typical LGE pattern is virtually pathognomonic and significantly more specific and sensitive than echo or CMR functional assessment, and may be an early finding, i.e. presenting before a significant increase in mass develops.⁴² In a meta-analysis of seven studies, the diagnostic accuracy of LGE CMR, partially deriving from the detection of this typical LGE pattern, was reported as high (sensitivity 85%, specificity 92%).⁴⁴ Notably, different LGE patterns can be encountered, such as localised enhancement or diffuse transmural or patchy LGE.⁴⁵ Overall, non-typical LGE patterns do not allow us to exclude CA. Furthermore, as LGE is typically more prominent in ATTR compared with AL amyloidosis, LGE analysis has been proposed as a tool to differentiate ATTR from AL cardiomyopathy. An LGE scoring system taking into account the presence of circumferential and/or transmural LGE in basal, mid-wall and apical slices, as well as the presence of RV LGE, was proposed. A score of 13 or greater differentiated ATTR from AL type with 82% sensitivity and 76% specificity.⁴⁵

LGE quantification and assessment of LGE changes over time may still prove challenging. The use of contrast medium is also limited in patients with an estimated glomerular filtration rate less than 30 mL/min/1.73 m², which is quite common in systemic AL amyloidosis. Native (i.e. pre-contrast) T1 mapping may help overcome these limitations. Myocardial native T1 values increase progressively from diffuse fibrosis to scar tissue and finally to amyloid and oedema (the latter being an important element of AL-type CA), and allow us to differentiate with high accuracy AL and ATTR cardiomyopathy from HCM.⁴² An increase in native T1 may represent a red flag for the disease, as it is observed before the onset of LV pseudohypertrophy, LGE development or elevation in blood biomarkers, and is also able to track cardiac amyloid burden with the potential to become a useful tool to quantify cardiac amyloid and track changes over time.⁴² At 1.5 T, specific native myocardial T1 cut-off values have been found (<1036 ms to exclude CA with 98% negative predictive value; T1 > 1164 ms to confirm CA with 98% positive predictive value), and a diagnostic algorithm reserving gadolinium administration only for patients with intermediate values (native T1 between 1036 and 1164 ms, representing 58% of patients in the study population) has been proposed.⁴⁶

Interstitial expansion due to amyloid infiltration manifests as extracellular volume increase, which can be calculated at CMR from postcontrast T1 mapping, after correction for native T1 and haematocrit.⁴⁷ Extracellular volume has been demonstrated to be not only an important early diagnostic tool, being generally higher in TTR than AL amyloidosis,⁴⁸ but also a prognostic marker in both AL⁴⁹ and TTR amyloidosis,⁴³ and a potential tool to monitor response to treatment.

Diphosphonate scintigraphy

Radiolabelled phosphate derivatives, initially developed as bone tracers, were first noted to localise to amyloid deposits in 1977 in two patients receiving ^99mTc-diphosphonate.⁵⁰ This finding led to the development of several tracers, including ^99mTc-3,3-diphosphono-1,2-propanodicarboxylic acid (DPD), ^99mTc-pyrophosphate (PYP), ^99mTc-methylene diphosphonate (MDP) and ^99mTc-hydroxymethylene diphosphonate (HMDP).⁵¹ In 2005, Perugini et al.⁵² performed ^99mTc-DPD imaging on 25 patients with CA (10 vATTR, five wtATTR, 10 AL). All 15 ATTR patients displayed a strong myocardial uptake of ^99mTc-DPD while no uptake was observed in AL patients. Therefore, the sensitivity and specificity of ^99mTc-DPD scintigraphy to diagnose ATTR cardiomyopathy were both 100%.⁵² Several other studies confirmed these findings, allowing us to conclude that, in patients with CA, intense ^99mTc-DPD retention is basically pathognomonic of ATTR cardiomyopathy, while absent uptake tends to exclude this diagnosis. Moderate ^99mTc-DPD myocardial uptake was reported to be of indeterminate significance, with a prevalence in AL and ATTR cardiomyopathy of 18% and 36%, respectively.^52,–54

The correlation between the intensity of tracer uptake and cardiac disease severity, evaluated in terms of septal thickness or impaired longitudinal function, provides the conceptual framework for the visual Perugini scoring system: grade 0, no cardiac uptake; grade 1, cardiac uptake present but less intense than the bone signal; grade 2, cardiac uptake with intensity similar to or greater than bone signal; grade 3, cardiac uptake with much attenuated or absent bone signal.⁵² This system has been incorporated in the algorithm for the non-biopsy diagnosis of ATTR-type CA, in which diphosphonate scintigraphy is advised when CA is suspected based on a combination of clinical, biohumoral and/or imaging data. When the Perugini grade is 2 or 3, and no monoclonal protein is found, ATTR cardiomyopathy can be diagnosed without the need for histological confirmation. By contrast, when a monoclonal protein is found or Perugini grade is 1, histological confirmation and typing of amyloid (preferably through an endomyocardial biopsy) is suggested. Finally, when Perugini grade is 0, CMR should be performed or reviewed when a monoclonal protein is found, while CA is unlikely if no monoclonal protein is detected (Figure 2).⁴¹

Figure 2.

Diagnostic algorithm for cardiac amyloidosis. A schematic representation of the diagnostic algorithm for cardiac amyloidosis, revised from that proposed by Gillmore et al.,⁴¹ is proposed. Consultation with a haematologist is warranted when a monoclonal component is detected. AL: immunoglobulin light-chain amyloidosis; ATTR: transthyretin amyloidosis (ATTRv): variant transthyretin amyloidosis; ATTRwt: wild-type transthyretin amyloidosis; CMR: cardiac magnetic resonance; HF: heart failure. Modified from Emdin et al.⁹

In addition to the semiquantitative Perugini score, a quantitative analysis can be performed by drawing a region of interest over the heart corrected for contralateral counts and calculating a heart-to-contralateral ratio (H/CL). The degree of tracer uptake in the heart was reported to correlate with LV wall thickness and mass, and patients with ATTR-type CA displayed a higher quantitative score than those with AL-type CA. A H/CL ratio greater than 1.5 had a 97% sensitivity and 100% specificity for identifying ATTR cardiac amyloidosis.⁵⁴

There are a few reports of the use of single-photon emission computed tomography (SPECT) in patients with ATTR-type CA.^55,56 Compared with planar imaging, tomographic acquisitions could allow us to diagnose cardiac involvement at an earlier stage, to quantify the global and regional amyloid burden, and to provide insight into disease progression and the response to treatment.

The detection of cardiac uptake during a bone scan performed for other reasons has been reported.⁵⁷ We are not aware of studies assessing the frequency of this incidental finding, which should prompt a diagnostic work-up for CA.

Gatekeepers in cardiac amyloidosis

Amyloidosis is a systemic disease and, although cardiac involvement is one of the major determinants of outcome, the disease may present with a functional impairment of other organs. Moreover, some non-cardiac conditions define populations at risk of the development of CA. Consequently, there is a need for collaboration among neurology, nephrology, haematology and cardiology specialists in order to achieve a timely diagnosis and improve prognosis (Figure 3).

Figure 3.

A multidisciplinary approach as a key to the early diagnosis of cardiac involvement in amyloidosis. GFR: glomerular filtration rate; HFpEF: heart failure with preserved ejection fraction; HFrEF: heart failure with reduced ejection fraction; NPs: natriuretic peptides; LV: left ventricular.

Haematologists following patients with plasma cell disorders should raise the suspicion of CA when nephrotic range proteinuria, HFpEF, non-diabetic peripheral neuropathy, unexplained hepatomegaly or diarrhoea develop. Physical signs of amyloidosis, such as macroglossia or periorbital purpura, are highly specific but poorly sensitive, and should never be used to rule out the disease.⁵⁸ Patients with MGUS or smouldering myeloma should also be monitored to check for the development of AL amyloidosis, which is 8.8-fold more likely than in the general population.⁵⁹ Although patients with MGUS or multiple myeloma typically have a monoclonal peak on serum protein electrophoresis, patients with AL amyloidosis often have little intact monoclonal Ig; furthermore, most of them have only light chains, and about half the patients are missed if only serum protein electrophoresis is used for screening, which underlies the importance of immunofixation electrophoresis to search for κ or λ light chains.⁶⁰ The latest recommendations suggest that cardiac involvement should be diagnosed when mean wall thickness is greater than 12 mm at echocardiography (in the absence of other causes of hypertrophy), and/or NT-proBNP greater than 332 ng/L when renal failure or atrial fibrillation are not present.⁶¹ On the other hand, lower circulating concentrations do not allow us to rule out CA, especially when clinical suspicion is high and other causes of natriuretic peptide elevation are excluded.

Kidney involvement is found in two-thirds of patients with AL amyloidosis at the time of diagnosis, and may present with proteinuria in the nephrotic range (composed mainly of albumin, often with detectable urine monoclonal light chains) and, in some cases, decreased glomerular filtration rate.⁶² A relevant number of cases is then diagnosed by nephrologists. Although the screening for cardiac involvement is mandatory once the diagnosis has been clarified, most early signs of cardiac disease are to be considered with caution; for example, circulating concentrations of markers of myocardial damage are influenced by renal function,⁶³ and long-term reduction in the glomerular filtration rate can lead to LV hypertrophy by itself.

Autonomic neuropathy, either alone or associated with peripheral neuropathy, is observed in both AL and ATTR amyloidosis. Such symptoms, when not explained by concomitant disorders, should prompt a screening for amyloidosis and for cardiac involvement, particularly in high-risk subsets, e.g. carriers of TTR gene mutations,⁵ those with a family history of ‘hypertrophic’ cardiomyopathy, or signs of systemic disease.

Conclusions and future directions

The diagnosis of CA is challenging because of its phenotypic heterogeneity, multi-organ involvement often requiring the interaction among experts in different specialties and subspecialties, lack of a single non-invasive diagnostic tool, and limited awareness in the medical community. Recent studies have challenged the dogma of CA as a rare, incurable disease, and have redefined the epidemiology and therapeutic options of this condition. Missing or delaying the diagnosis of CA may have a profound impact on patient outcome, as potentially life-saving treatments (in particular chemotherapy in the case of AL amyloidosis) may be omitted or delayed. To obtain a timely identification, physicians potentially encountering CA must be able to recognise the ‘red flags’ prompting a dedicated diagnostic work-up.

Clinical signs of CA often appear in late-stage disease, thus effort is required to identify novel tools for an early diagnosis. Biohumoral markers of cardiac involvement will be likely to be of help, although there is currently scarce evidence on their use for screening purposes in at-risk individuals. Future research should also be directed at identifying imaging features holding high specificity for the differential diagnosis of subtypes of CA. In this view, the use of [18F]-Florbetaben positron emission tomography/computed tomography seems promising in the non-invasive diagnosis of AL CA.⁶⁴

Author contribution

GV, CP and ME contributed to the conception and design of the work. GV, AA, AB, DG and GB contributed to the literature review and to the drafting of the manuscript. ME and CP critically revised the manuscript. All authors gave final approval and agree to be accountable for all aspects of the work ensuring integrity and accuracy.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.