|Home | About | Journals | Submit | Contact Us | Français|
The systemic amyloidoses are a family of diseases induced by misfolded and/or misassembled proteins. Extracellular deposition of these proteins as soluble or insoluble cross β-sheets disrupts vital organ function.1 More than 27 different precursor proteins have the propensity to form amyloid fibrils.2 The particular precursor protein that misfolds to form amyloid fibrils defines the amyloid type and predicts the patient’s clinical course. Several types of amyloid can infiltrate the heart resulting in progressive diastolic and systolic dysfunction, congestive heart failure, and death. Treatment of cardiac amyloidosis is dictated by the amyloid type and degree of involvement. Consequently, early recognition and accurate classification are essential.3
The diagnosis of amyloidosis requires histologic identification of amyloid deposits. Congo Red staining renders amyloid deposits salmon pink by light microscopy, and a characteristic apple green birefringence under polarized light conditions (Figure 1). Additional immunohistochemical staining for precursor proteins identifies the type of amyloidosis (Figure 2).4 Ultimately, immunogold electron microscopy and mass spectrometry confer the greatest sensitivity and specificity for amyloid typing.5, 6
Two types of amyloid commonly infiltrate the heart: a) immunoglobulin light-chain (AL or primary systemic) amyloid, and b) transthyretin (TTR) amyloid. Transthyretin-related amyloidoses, in turn, encompass two forms of disease – familial disease arising from misfolding of a mutated or variant transthyretin (familial amyloid cardiomyopathy (FAC) or familial amyloidotic polyneuropathy (FAP)), and a sporadic, non-genetic disease due to misaggregation of wild-type transthyretin (senile systemic amyloidosis (SSA)). Cardiac amyloidosis can also be caused by other precursor proteins such as Apoliporotein A1, but the prevalence of disease is low and beyond the scope of this review.3 In contrast, AL amyloidosis has an estimated incidence approaching ~2500 new cases annually),7 with cardiac involvement in approximately 50% of cases.8–10 Untreated, prognosis with AL disease is poor, with median survival following diagnosis of < 1 year in the presence of heart failure symptoms.8 Unlike AL heart disease, transthyretin-related amyloid cardiomyopathy is slowly progressive and clinically well tolerated, often defying diagnosis until marked ventricular wall thickening, profound diastolic dysfunction, and conduction disease have occurred. Untreated, survival with transthyretin-related cardiac amyloidosis is measured in years to decades. Secondary or AA amyloidosis results from misfolding of serum amyloid A, an acute phase reactant induced by chronic inflammation. Echocardiographic evaluations in three studies involving 48, 30, and 224 patients with AA amyloidosis identified features of amyloid cardiomyopathy in only 1.3% of the aggregate cohort – and less than 1% of the population had recognized congestive heart failure.11–13 Although 30% of Finnish rheumatoid arthritis patients have histologic evidence of cardiac amyloid, heart failure is rarely reported in this population.14 These data indicate AA amyloid rarely infiltrates the myocardium or conduction system in a clinically meaningful fashion and that coronary infiltration is rarely reported.15
TTR, formerly known as prealbumin, is a 127-amino acid, 56 kDa transport protein primarily expressed by the liver. Under normal conditions TTR circulates as a homo-tetramer but, due to genetic mutation or aging, tetramers can dissociate to monomers that misassemble into amyloid fibrils.16 As two forms of TTR amyloid exist -- age-related or senile amyloidosis (wild-type TTR), and familial transthyretin amyloidosis (variant TTR) -- this review will compare the epidemiology, pathogenesis, diagnosis and treatment of senile systemic amyloidosis (SSA) and familial amyloid cardiomyopathy (FAC).
Soyka first described age-related cardiac amyloidosis in 1876,17 followed by numerous case reports and small case series published over the next 100 years. In 1965, Pomerance estimated the prevalence of senile cardiac amyloidosis to be approximately 10% over the age of 80 years, and 50% over the age of 90 years.18 A study of 85 consecutive autopsies in patients 80 or more years old found amyloid deposits with pre-albumin immunohistochemical staining (TTR) in the atria or left ventricle of 21 hearts -- a 25% prevalence for age-related TTR amyloid in this elderly population.19 On closer review, however, only 2/3 of the TTR-amyloid staining hearts had left ventricular involvement – described as “small and widely scattered” in over 50% of cases. These data suggest histologically significant cardiac TTR amyloid occurs in 8–16% of people over 80 years of age. A prospective study of Finnish octogenarians (the Vantaa 85+Study) also identified TTR amyloid deposits in 25% of hearts from 256 autopsies.20 Once again, moderate or severe cardiac amyloid deposition occurred in only 5.5% of the total autopsy population.
Prevalence of SSA indisputably increases with advancing age, and virtually all patients are over 60 years of age when diagnosed. SSA is a remarkably gender-specific disease, exhibiting approximately 25–50:1 male:female expression.21 While the Finnish autopsy study did not report a male predominance for amyloid deposition, male patients had more pronounced amyloid staining (greater amyloid burden) than did female SSA cases. By unclear aging mechanisms, aggregation and cardiac deposition of genetically normal TTR increase over time. Aggregate autopsy data suggest wild-type TTR, while histologically present in the hearts of 25–30% of septa- and octogenarians, drives cardiac dysfunction in a smaller but significant elderly cohort. Projections of octogenarian population growth over the next 20 years predict SSA will become the most common form of cardiac amyloidosis.
The TTR gene is located on chromosome 18q12.1, and spans 4 exons and 5 introns. There are over 100 single nucleotide polymorphisms encoding variant TTR, with 80 confirmed pathogenic mutations.22 These mutations tend to cluster into geographic and/or ethnic groupings and exhibit an autosomal dominant pattern of inheritance. The clinical phenotype of variant TTR amyloidosis varies greatly by mutation, the age of onset, disease penetrance, clinical course, and prognosis. Most FAP patients develop nervous system involvement with or without cardiac amyloidosis.23 The mutations that most commonly induce vTTR cardiac amyloidosis are summarized in Table 1. The most prevalent and best described mutation associated with amyloid neuropathy is V30M (Met 30), predominantly affecting patients originating from Japan, Portugal or Sweden. People with Met30 genotype have variable disease penetrance influenced by the place of origin, gender, parental gene transmission, and age affecting disease expression.24 The T60A (Ala 60) TTR mutation is also a frequent cause of amyloid neuropathy and cardiomyopathy – affecting 1% of the population in north-west Ireland (County Donegal).25 Although the UK National Amyloidosis Centre reports cardiac involvement in nearly 100% of T60A FAC patients referred with neuropathy,26 the US experience is more heterogeneous.
Among the variant TTR predominantly targeting the heart, the valine to isoleucine substitution at position 122 (V122I or Ile122) affects the greatest world population.27 First described in 1989 by Gorevic et al.,28 Jacobson et al. reported V122I TTR genopositivity in roughly 3–4% of the US African American population,29, 30 with virtually undetectable prevalence in the Caucasian population.31 The V122I founder appears to have originated in West Africa, explaining clinical expression of V122I TTR in the Caribbean Islands (Haiti, Jamaica, and Bermuda)(D. Jacobson, personal communication). While penetrance of this particular allele is unknown, there appears to be a strong association between carrier status, the development of heart failure (relative risk 2.6),32 and echocardiographic features of cardiac amyloidosis.33 Examining stored samples from the Beta-Blocker Evaluation in Survival Trial (BEST), Buxbaum et al., demonstrated that approximately 10% of African American study participants with heart failure over the age of 60 years carry V122I, suggesting that unrecognized cardiac amyloidosis may be a contributor to the development of heart failure.34 This trial enrolled patients with systolic dysfunction (LVEF < 35%), a cardiac profile more frequently observed in V122I than in other TTR mutations.35, 36 By recent US Census statistics, approximately 1.5 million African Americans carry the V122I mutation and are at risk for the development of ATTR cardiac amyloidosis.37 More specifically, African Americans older than 65 years of age carrying V122I constitute the population at immediate risk for clinical expression of TTR cardiomyopathy -- and number ~ 99,600 – 132,800 according to 2010 census figures (3–4% of 3,320,000 African Americans), rendering V122I a potentially important cause of heart failure in the elderly Black community.
Transthyretin, a 127 amino acid protein, is encoded by 7 kb of DNA spanning exons 1–4 of a single gene on chromosome 18.38 In its native state, transthyretin circulates as a tetramer with two C2 symmetric funnel-shaped thyroxine binding sites at its dimer-dimer interface.39 Elegant thermodynamic studies demonstrate that tetramer dissociation is the rate limiting step in misfolding of monomeric transthyretin to amyloid fibrils.16 In senile systemic amyloidosis (SSA), incompletely described age related events such as post-translational biochemical alterations in wild-type TTR or its chaperones appear to contribute to amyloid fibril formation. Data from transgenic mice overexpressing human SSA suggest that alterations in hepatic chaperone production or proteasome clearance of unfolded protein may determine which aging heart develops amyloidosis.40 In FAP, one amino acid substitution in native TTR (variant TTR) destabilizes the tetramer, promoting disaggregation of monomeric protein. Ultimately, the amyloidogenicity of a particular variant transthyretin is determined by the capacity of a specific amino acid substitution to destabilize circulating TTR tetramers, releasing monomeric TTR to permit misfolding to occur. To date, over 100 variant TTR have been described. Many amino acid substitutions associate with patterns of organ involvement, predicting distinct clinical courses. Typically variant transthyretin is induced by a single nucleotide substitution; however two nucleotide changes or complete codon deletion have been shown to produce one amino acid change in transthyretin.41
Affected organs invariably harbor extracellular amyloid deposits. Whether these deposits induce organ dysfunction or represent epiphenomena (disease markers) is debated. Review of sural nerve biopsies in 31 V30M ATTR patients described epineural amyloid deposits and nerve degeneration in half of the cohort. No correlation between the presence of amyloid deposits and histologic nerve damage could be assigned,42 suggesting an alternative mechanism of injury. In the kidney, the degree of functional disruption is not predicted by the extent of amyloid deposits.43, 44 Moreover serial kidney biopsies in patients with AL amyloidosis before and after clinically successful treatments reveal unchanging amyloid burden despite significant improvement in proteinuria.45, 46
Increasingly data support a central role for circulating or pre-fibillar amyloidogenic proteins in the disruption of cardiac function in AL and TTR mediated amyloid disease. Murine hearts perfused with clonal immunoglobulin light chain (AL-LC) isolated from patients with AL amyloid cardiomyopathy rapidly induced diastolic dysfunction. In contrast, light chain isolated from patients without AL amyloidosis did not alter ex vivo heart function.47 In vitro, exposure of cultured cardiomyocytes to physiologic levels of AL-LC stimulated production of reactive oxygen species and upregulated heme oxygenase-1, a redox-sensitive protein that identifies cell injury.48 In addition to altering the cardiomyocyte intracellular redox state, AL-LC reduced intracellular calcium levels and cardiomyocyte contractility – in the absence of amyloid fibril formation.48 Recent data indicate AL-LC alters cardiomyocyte ion fluxes, contractility and programmed cell death through a non-canonical p38α MAPK pathway.49
Similar evidence of end-organ damage by pre-fibrillar TTR amyloid exists in FAP. Nerve biopsies from asymptomatic V30M ATTR carriers revealed pre-fibrillar/Congo Red stain negative protein aggregates by anti-TTR immunohistochemistry and immunogold electron microscopy.50 Upregulation of NFκB and proinflammatory cytokines in these nerve biopsies signaled cell toxicity affected by these nonfibrillar TTR aggregates – well before the onset of clinical disease. Exposure of neuronal cell cultures to L55P ATTR nonfibrillar aggregates induced caspase-3 generation, and expression of programmed cell death. In contrast, mature L55P ATTR amyloid fibrils did not stimulate caspase-3 expression, suggesting that pre-fibrillar TTR aggregates are the neurotoxic mediator of disease. Additional data support a role for receptors of advanced glycosylation end products (RAGE) in mediating ATTR organ injury.51 Initial studies of V30M versus L55P ATTR transgenic mouse models of FAP do not detect differences in nerve toxicity mediated by the nonfibrillar forms of these variant TTR species. Taken together, data generated in AL and ATTR models of disease provide evidence that pre-fibrillar protein aggregates – and not mature amyloid fibrils – contribute to organ toxicity.
Diagnosis of systemic amyloidosis requires histologic identification of amyloid deposition by Congo red staining. Unlike light chain (AL) amyloid disease, kidneys and tongue are rarely involved in clinically significant fashion, thus biopsy of these organs is usually not pursued.36 Abdominal fat aspirate is a simple, office-based, biopsy procedure that identifies amyloid deposits in approximately 70% of those with vTTR such as V122I ATTR disease.35 Cardiac biopsy remains the gold-standard for amyloid cardiomyopathy, and is not complicated by sampling artifact (yielding false positive or negative findings) as occurs with other infiltrative processes such as sarcoidosis. Once amyloid is identified by Congo red staining, immunohistochemical (IHC) stains for κ and λ light chains, AA, and TTR can be performed to determine the precursor protein. Confirmation of histology and identification of the amyloidogenic protein may be aided by review by pathologists at international amyloidosis referral centers. IHC demonstrating TTR protein in amyloid deposits requires further analysis to distinguish variant from wild-type TTR. Isoelectric focusing (IEF) electrophoresis in most cases permits separation of variant from wild-type TTR by charge (Figure 3). In our experience, IEF reveals distinct electrophoretic mobility differences for ~95% of the variant TTRs tested. PCR amplification and sequencing of TTR exons 1–4 variant validates the IEF findings and definitively establishes variant from wild-type TTR genotype. Alternatively, biopsy tissue can be processed by laser dissection/liquid chromatography-tandem mass spectrometry (LC-MS/MS) to identify the precursor protein with 98% sensitivity.6 Occasionally, TTR cardiac amyloidosis can be diagnosed without cardiac biopsy – when variant TTR genopositivity is established, tissue biopsy from another site documents TTR amyloid deposits, and non-invasive data (see below) support cardiac involvement.
In many international amyloid centers, AL amyloidosis and monoclonal gammopathy of unknown significance (MGUS) dominate the clinic population, occasionally obscuring recognition of variant or wild-type TTR amyloidosis. Lachmann et al. identified TTR mutations among 4% of referrals to a national amyloid center for evaluation and treatment of presumed AL disease.52 None of these TTR amyloid patients had family histories suggesting genetic disease. Interestingly all of the ATTR patients presented with cardiomyopathy. Immunohistochemical staining of tissue sections ultimately identified TTR as the amyloid subunit protein.52 Similarly, Connors et al. reported biopsy-proven AL amyloidosis in 12% of V122I gene positive patients.35 These data from large amyloid referral centers illustrate the importance of establishing the correct subunit protein that forms the amyloid tissue deposits – particularly in patients with ATTR genopositivity.
Diagnosis of cardiac amyloidosis can be based on invasive heart biopsies or a non-invasive approach -- given the proper clinical context, supportive non-invasive testing, and identification of amyloid tissue deposits from a non-cardiac source such as abdominal fat aspirate (Figure 4). A contemporary approach to non-invasive diagnosis of TTR cardiac amyloidosis includes echocardiography with strain imaging, cardiac magnetic resonance (CMR), electrocardiography (ECG), and serum biomarker testing including B-type natriuretic peptide (BNP or nT-pro BNP) and cardiac troponin (T or I). Physical examination does not typically assist in differentiation of amyloid type, with the notable exceptions of macroglossia and peri-orbital eccymoses heralding AL amyloidosis.36 Physical findings vary significantly depending upon the severity of heart dysfunction, ranging from a relatively normal exam in early stage disease to extensive signs of congestive heart failure including pleural effusions, elevated jugular venous pressure, and peripheral edema. In many cases of TTR amyloid, isolated cardiac involvement or inconclusive biopsies from other sites (fat pad aspirate, gastric or rectal biopsies, extensor retinaculum sampling at carpal tunnel surgery, or salivary gland biopsies) warrant direct endomyocardial sampling.
Echocardiography remains the most useful imaging modality for identifying and monitoring cardiac amyloid disease. Ease of image acquisition and interpretation, relatively low cost, unparalleled diastolic functional assessment, and capacity for serial studies despite technical differences in data acquisition or disease progression make echocardiography the universal instrument for cardiac amyloid assessment. Recent reports validate detection of subtle systolic dysfunction by tissue Doppler imaging and speckle tracking technology.53
By classic echocardiographic teaching, the cardiac amyloid phenotype is a thick walled ventricle with “speckling” appearance of the myocardium, small LV chamber volume, valve thickening, atrial enlargement, and signs of elevated filling pressures (pericardial effusion, pleural effusions, dilated vena cava) due to restrictive diastolic filling (Figure 5, Supplemental video file 1).54 Although the preponderance of data on which these echo features are based is derived from the AL population, similar findings have been reported in TTR cardiomyopathy.21 Wall thickness increase remains the principal feature upon which cardiac amyloidosis is diagnosed. According to an international consensus panel of experts in amyloid disease, interventricular septal thickness of >12 mm – in the absence of aortic valve disease or significant systemic hypertension -- is the echocardiographic criterion that identifies cardiac involvement in patients with AL systemic amyloidosis (there are no established criteria for TTR disease).55 This single threshold fails to account for gender specific differences in normal wall thickness,56 and confers a high degree of specificity but low sensitivity for identification of cardiac involvement. The continuum of cardiac involvement makes early disease recognition challenging when wall thickness and diastolic function are only mildly abnormal. The perceived rarity of amyloid disease in comparison to other, more common, entities that produce ventricular thickening such as hypertensive remodeling and hypertrophic cardiomyopathy (HCM) likely lowers cardiologists’ recognition of new cases. Echocardiography alone is often unable to differentiate these very different processes, prompting multi-modality assessment. However, the presence of prominent right ventricular wall thickening, inter-atrial septal thickening, and restrictive (grade 3) diastolic dysfunction are uncommon in hypertensive remodeling or HCM, and can suggest that TTR amyloidosis may be present.24, 33
The challenges of amyloid diagnosis coupled with mimicry of hypertensive and hypertrophic cardiomyopathy result in late recognition of TTR cardiac disease. Consequently, advanced remodeling changes are more often present upon diagnosis of TTR cardiomyopathy then in AL heart disease. Patients with SSA cardiac amyloidosis tend to have the largest wall thickness and myocardial mass, as compared to AL21 and variant TTR disease.36 While systolic dysfunction is frequently a manifestation of more advanced disease in light-chain and variant TTR cardiac amyloidosis, it is fairly common in SSA disease, again likely owing to delayed recognition. Among patients with ATTR, lower LVEF (<50%) is associated with reduced survival.57
Longitudinal strain measurement by tissue Doppler and echo speckle tracking have emerged as useful clinical tools for the identification of cardiac involvement in AL disease53 and can assist in differentiation of cardiac amyloidosis from other causes of wall thickening, including hypertension and HCM.58 TTR cardiac amyloidosis also results in reduction in longitudinal shortening, although its prognostic significance has not been established.
In comparison to echocardiography, cardiac MR (CMR) offers superior myocardial border delineation and a three-dimensional approach to quantify ventricular volumes, wall thickness, and mass (Supplemental movie file 2). It is more precise and reproducible than echo, but also more expensive, less widely available, and limited by the inability to image patients with pacemaker or implanted cardio-defibrillator devices. At present, echocardiography provides the best imaging technique for assessment of ventricular diastolic function, conferring better temporal resolution than CMR. However, the principal advantage of CMR over echocardiography is the capacity to directly identify amyloid infiltration by means of late gadolinium enhancement (LGE) imaging. Gadolinium is an extracellular contrast agent, and under normal conditions is not retained in the myocardium after administration. Amyloid infiltration results in expansion of the extracellular space and abnormal myocardial gadolinium distribution kinetics, which result in contrast retained in the heart. Signal from normal myocardium is nulled or suppressed in LGE imaging, but due to diffusely retained contrast, this is difficult to achieve in cardiac amyloidosis. An important limitation of the application of contrast enhanced CMR in TTR cardiac amyloidosis is co-existent chronic kidney disease and the risk of nephrogenic systemic fibrosis (NSF).59 If the creatinine clearance is < 30 ml/min, gadolinium contrast cannot be safely administered and LGE imaging cannot be performed.
Maceira et al. first reported a CMR profile that identifies cardiac amyloidosis, but with an imaging technique that required an unusually short delay following contrast administration to obtain optimal LGE images.60 Subsequent studies, in mixed AL and TTR cohorts, have determined that the sensitivity and specificity of CMR for the identification of cardiac amyloidosis as compared to endomyocardial biopsy approaches 90%.61–63 Unlike LGE abnormalities associated with myocardial infarction, where focal regions of high signal intensity LGE are seen, patterns in cardiac amyloidosis are variable with global sub-endocardial, diffuse, and focal foci noted (Figure 6).64 Furthermore, the retained gadolinium greatly shortens myocardial T1,65 a fundamental MR characteristic upon which LGE contrast is founded. Due to retained contrast, myocardial T1 approaches that of the ventricular blood pool, rendering a distinct pattern of early myocardial signal suppression (co-incident to the blood pool) representative of diffuse amyloid infiltration.
The majority of published reports of CMR in cardiac amyloidosis involve mixed cohorts of patients with relatively advanced AL and TTR disease, with the notable exception of Di Bella et al., wherein CMR and nuclear scintigraphy were utilized to identify cardiac involvement in patients with FAP.66 While precursor proteins differ, the LGE findings in TTR and AL cardiac amyloidosis appear relatively similar. As reported in the echocardiographic literature, CMR determined wall thickness and mass are greater in TTR cardiac amyloidosis than light chain disease.67
Classically, cardiac amyloidosis is electrocardiographically typified by low-QRS voltage and a “pseudo-infarct” pattern of Q-wave or T wave changes on ECG. The presence of low-QRS voltage and increased LV wall thickness by echocardiography should prompt consideration of cardiac amyloidosis. Notably, certain ECG features support TTR cardiac disease more than AL heart disease. In different case series, low-QRS voltage has been reproducibly identified in approximately 46–60% of AL patients, but only 25–40% of TTR disease.36, 68, 69 However, a “pseudo-infarct” pattern is also equally seen in AL disease (47–69%) and TTR disease (66–69%).36, 68, 69 The presence of conduction system disease is more commonly seen in SSA disease, particularly left bundle branch block. Finally, while most patients present with sinus rhythm, atrial fibrillation is more commonly observed in SSA disease (approximately 30%) vs. variant TTR (<10%) and AL disease (typically <20%).36, 68
Nuclear pharmaceutical identification of cardiac amyloid deposition is a developing field. Three classes of nuclear tracers have application to cardiac amyloidosis: a) positron emission tomography (PET) agents, b) bone avid compounds, and c) amyloid directed molecules.
Design of the PET agents is based on the structure of thioflavin T, a benzothiazole dye that fluoresces when bound to β-rich amyloid fibrils. The clinical limitations of the short-lived original 11C compounds led to the development of 18F-tagged agents, with recent FDA market approval for 18F-aV-45 [florbetapir] (Amyvid™) as a tracer for β amyloid in Alzheimer’s dementia. No published data are available on the use of PET agents in amyloid cardiomyopathy.
Three different bone complexing molecules have varying avidity for cardiac amyloid deposits including 99m-technecium-labeled agents pyrophosphate (99mTc-PYP), methylene diphosphonate (99mTc-MDP), and 3,3-diphosphono-1,2-propanodicarboxylic acid (99mTc-DPD). Although the binding mechanism is debated, reports of bone scans with increased 99mTc-PYP uptake in the heart of patients with amyloid infiltration date to the early 1980’s. Testing in small cohorts of amyloid patients confirm 99mTc-PYP uptake in amyloid hearts, however low intensity signal and false positive results in hypertensive, sarcoid, and dilated cardiomyopathies limit use of this agent.70, 71
99mTc-MDP binds preferentially to cardiac amyloid, although with less avidity and sensitivity than 99mTc-PYP. Data comparing 99mTc-MDP to 99mTc-PYP in 7 patients with biopsy-proven cardiac amyloidosis reported 100% versus 56% sensitivity with the respective agents.72 Additionally, 99mTc-MDP uptake is less intense than 99mTc-PYP, diminishing its potential as the tracer of choice for nuclear study of cardiac amyloidosis.
99mTc-DPD, in contrast to the other bone avid molecules, displays preferential uptake in ATTR cardiomyopathy. In the first of two studies from Bolgona, Italy, 15 patients with ATTR and 10 patients with AL cardiomyopathy underwent 99mTc-DPD scanning. Radiotracer uptake was reported in all ATTR and none of the AL hearts.73 A second study involving 79 patients with amyloid cardiomyopathy (45 ATTR and 34 AL) and 15 control subjects identified mild 99mTc-DPD signal in AL hearts, lessening tracer selectivity for ATTR heart involvement.74 While not definitive, 99mTc-DPD uptake in the heart of a patient with systemic amyloidosis strongly suggests ATTR disease.75 Further studies will be needed to fully characterize the utility of 99mTc-DPD scanning in patients with suspected cardiac amyloidosis.
Serum amyloid P (SAP) is a stabilizing component of all amyloid deposits. Scans employing 123I-SAP can assess the amyloid burden affecting the liver, spleen and kidneys (reticuloendothelial system) in patients with amyloidosis. In contrast, 123I-SAP scans do not provide sufficient signal within affected hearts to permit its use as a diagnostic or serial measure of cardiac amyloidosis.76
Aprotinin is a serine protease inhibitor and constituent of amyloid matrix ground substance, prompting use of 99mTc-aprotinin as an amyloid marker. Two studies reported 99mTc-aprotinin uptake in 19 of 47 (40%) documented amyloid hearts.77, 78 In addition to relative insensitivity in patients with documented amyloid cardiomyopathy, 99mTc-aprotinin scans have low signal-to-background ratios for heart uptake, challenging the interpretation of studies.
123I-metaiodobenzylguanidine (MIBG) is a nuclear labeled tracer of neurotransmitter uptake by sympathetic neurons used in amyloid patients to document cardiac denervation.79 It has not proven as useful in disease detection as other modalities such as echo or CMR.
Cardiac serum biomarkers, specifically B-type natriuretic peptide or its n-terminal form (nT-pro-BNP) and cardiac troponins, are elevated in AL cardiac amyloidosis and associate with survival.80, 81 The mechanism of BNP elevation involves increased cardiac filling pressures from amyloid infiltration, whereas troponin elevation results from myocyte cell death. In AL disease, BNP elevations may reflect direct toxicity of light chains by oxidant stress48 versus protofilament-induced injury in TTR. Suhr et al. analyzed 2-dimensional, M-mode, Doppler, and strain echocardiography signs of amyloid cardiomyopathy versus troponin and BNP measures collected in 29 patients with ATTR.82 BNP elevations highly correlated with IVS thickness and basal septal stain pattern in 76% of the cohort. Troponin I and T, however, did not segregate with echo abnormalities. Despite extensive reporting in light-chain disease, few data exist regarding the prognostic utility of BNP or troponin in TTR cardiac amyloidosis although both appear elevated in advanced disease, and may track with disease progression.83 For diagnostic purposes, biomarker elevation can be viewed as supportive evidence of the presence of TTR amyloid disease.
The clinical progression of TTR cardiac amyloidosis depends upon fibril type (wild-type vs. variant), specific mutation, age of onset, and potentially, fragmented versus full-length fibrils.84 Untreated, TTR disease carries significantly longer median survival than AL amyloidosis, but progresses to intractable heart failure and death due to systolic heart failure or dysrhythmia. Rapezzi et al. reported 2-year and 5-year survivals of 98% and approximately 75% for vTTR disease.36 Notably the study cohort did not include V122I ATTR, the most common mutation in the US. In a retrospective analysis of V122I patients referred to our treatment center, median survival for V122I cardiac amyloidosis was only 27 months following diagnosis.35 Furthermore, the Transthyretin Amyloidosis Cardiac Study (TRACS),a prospective, multi-center observational study comparing V122I ATTR versus SSA found that survival following diagnosis was lower for V122I patients (26 months vs. 46 months, respectively).83 It is important to note that survival in both Connors et al. and TRACS was likely biased by the relatively advanced stage of cardiac amyloidosis at the time of clinical referral. The clinical course of Ala60 appears somewhat slower, with median survival reported to be 6.6 years following onset of symptoms and 3.4 years from diagnosis.26 Survival in wtTTR disease appears most favorable of the amyloid cardiomyopathies. Reported median survival varies from 43 months to 75 months.21, 83
There are limited data regarding the rate of disease progression in TTR cardiac amyloidosis. Benson et al., utilizing echocardiography and volumetric CMR (notably without LGE imaging reported) to study disease progression in ATTR disease, reported significant increases in myocardial mass (mean increase of 8% by CMR, 22% by echo) after 1 year of observation.85 The multicenter TRACS study observed declines in LV ejection fraction, 6 minute walk duration, and coincident rises in nT-proBNP for both SSA and ATTR patients over 18 months.83 Unlike AL disease, where successful treatment stabilizes myocardial amyloid infiltration or induces regression in wall thickness,86, 87 the effect of treatment on TTR amyloid heart disease is unknown.
The treatment approach in TTR cardiac amyloidosis is two-fold: first, directed towards alleviation of heart failure symptoms, and second, to slow or stop progressive amyloid deposition. In contrast to large data sets for other forms of cardiomyopathy, there are few data in TTR cardiac amyloidosis on which to base treatment decisions.
Like AL cardiac amyloidosis, TTR disease results in progressive bi-ventricular wall thickening, diastolic dysfunction resulting from loss of compliance, and symptoms of congestive heart failure from elevation in cardiac filling pressures. Due to the challenges of cardiac amyloid diagnosis, many patients are recognized later in the disease course, often presenting with heart failure symptoms.
Optimization of fluid status is the cardinal tenet in cardiac amyloidosis management. This is best accomplished by administration of loop diuretics and spironolactone. Frequently, acute or chronic renal failure complicate management of heart failure, and over diuresis precipitates hypotension. While Lobato et al. demonstrated amyloid deposition in the kidneys of V30M ATTR patients independent of clinical signs (proteinuria or decreased glomerular filtration rate),43 significant renal insufficiency occurs in only ~3% of this population.88 In general, chronic kidney disease is not a hallmark of ATTR. Comorbidities such as diabetes, hypertension, or age related reduction in nephron mass, on the other hand, can precipitate renal failure. Hemodynamic factors, including renal hypo-perfusion due to diastolic congestive heart failure, can contribute to reduced renal function. It is critical to dose reduce or discontinue agents that impair inotropic or chronotropic compensatory mechanisms. As TTR amyloid infiltration markedly impairs diastolic filling and reduces stroke volume, tachycardia is a critical compensatory mechanism that maintains cardiac output. Consequently, high doses of beta-adrenergic receptor blocking agents are often poorly tolerated as they blunt compensatory tachycardia drive and induce greater negative inotropic effects in amyloid infiltrated hearts. Calcium channel blockers89, 90 and digitalis91 are contra-indicated in cardiac amyloid disease due to binding of amyloid fibrils and potentiation of drug toxicity. Low doses of angiotensin receptor antagonists or converting enzyme inhibitors may be beneficial as afterload reducing agents, particularly in co-existing hypertension, to improve forward cardiac flow and renal perfusion.
Atrial fibrillation is the most common dysrhythmia associated with wtTTR cardiac affecting approximately 30% of cases. Given the atrial dilation from increased ventricular end-diastolic pressures, as well as atrial amyloid infiltration, restoration of sinus rhythm is challenging and frequently unsuccessful in the long term. It is reasonable, however, to attempt sinus rhythm restoration with DC cardioversion, provided no atrial thrombus is present by transesophageal echo. Atrial fibrillation recurs in most patients, and as such a rate-control and anticoagulation strategy is warranted in most circumstances. Patients with AL disease are at extremely high risk of thromboembolism in AF, and it is generally assumed that the risk in TTR disease is also elevated over and above that of non-amyloid affected hearts, with warfarin reducing embolic risk.92, 93 Low doses of beta-adrenergic receptor antagonists are useful as rate controlling agents, however care should be taken to not exceed doses where the negative chronotropic and inotropic affects of these agents overwhelm the benefit (typically 50–100 mg per day of metoprolol). Amiodarone can be useful as a rate controlling agent, and also may organize AF into a slow atrial flutter, rendering rate control easier. Amiodarone is presumed safe in TTR cardiac amyloidosis, although patients must naturally be monitored for the known toxicities, and the drug should be avoided in the presence of significant conduction disease (e.g., left bundle branch block) without pacemaker placement. Dronedarone is uncommonly used in TTR cardiac amyloidosis as it is contraindicated in advanced heart failure.
If rate control is AF is impossible to achieve without worsening heart failure or compromising cardiac output, then atrioventricular junctional ablation and placement of a permanent ventricular (or bi-ventricular) pacemaker is required. It is also our practice to recommend prophylactic pacemaker placement for TTR cardiac amyloidosis patients with symptoms of pre-syncope and high degree of heart block (for example, first degree AV block with right bundle and left anterior fascicular block), to avert possible complete heart block from progression of amyloid infiltration, an approach is consistent with the most recent ACC/AHA recommendations for device placement.94 There are no data regarding the utility of radiofrequency ablation for either AF or atrial flutter in TTR cardiac amyloidosis, however our experience is that typical flutter circuits, in particular, can be treated with this approach.
The role of cardio-defibrillator implantation for primary prevention of sudden cardiac death in TTR cardiac amyloidosis remains largely unexplored. In a small case series of high-risk AL patients, ICD implantation did not improve overall survival as the principal cause of arrhythmic death was pulseless electrical activity.95 Thus, ICD therapy is reserved for secondary prevention indications, i.e. those with aborted SCD, or those in whom a documented rhythm disturbance including sustained ventricular tachycardia is noted.
Transthyretin is predominantly produced by the liver with minor secretion by the choroid plexus (CNS) and retinal pigment epithelium.96 Orthotopic liver transplantation (OLT), a genetic experiment to replace the major organ synthesizing variant transthyretin with a producer of wild-type TTR, was undertaken by Sweden’s Karolinska Institute in 1990, and by Deaconess Hospital, Boston MA in 1991.97 Over the ensuing 20 years, transplant centers have voluntarily posted 1844 OLTs on the FAP World Transplant Registry (http://www.fapwtr.org), with 911 (49%) reported by Portugal, 235 (12.7%) France, 137 (7.4%) Sweden, 97 (5.3%) Brazil, 83 (4.5%) USA, and 80 (4.3%) UK. The vast majority of OLT have been performed in V30M ATTR patients (94.3%), with non-V30M ATTR limited to 5.7% of the transplanted population. Cumulative data on 579 OLT performed over the first 10 years of FAPWTR listing indicated a 5 year survival of 77%98 with high numbers of cardiac death (39%). Karolinska Institute subsequently reported a one center experience of 141 transplantations with 10 and 15 year survivals of 83% and 60%, significantly better than a medically treated control cohort (62% and 19%, respectively).99
Liver transplantation successfully eliminates circulating levels of variant transthyretin, however reports of progressive cardiac infiltration post-OLT began circulating 6 years after the first transplant in the USA, initially identifying non-V30M ATTR as the at-risk population.100 Later, V30M ATTR patients were noted to experience similar progressive interventricular septal thickening101– with pre-existing amyloid cardiomyopathy the apparent predisposing condition in both V30M and non-V30M cohorts. Biochemical analysis of heart biopsies collected post-OLT from patients with progressive cardiomyopathy revealed wild-type TTR fibrils deposited on variant ATTR-rich amyloid matrix 102– a phenomenon later described in patients with progressive neuropathy post-OLT.103
Recognition that established amyloid cardiomyopathy at OLT risk progressive infiltrative changes in the heart post-operatively prompted the first orthotopic heart transplantation (OHT). To date, the FAPWTR lists 26 patients undergoing combined orthotopic liver and heart procedures – including 16 simultaneous OLT/OHT procedures and 9 sequential organ transplants. Dubrey et al. reported the UK experience with OHT for amyloid cardiomyopathy.100 Among 24 OHT, 17 were performed in AL amyloid patients, 3 ATTR patients, 2 SSA and 2 other inherited amyloidoses. Five year survival post-OHT was 38% in AL patients, 67% at 2 year follow up in ATTR patients, and 100% at 3 years in SSA recipients.100 The literature includes 5 OHT for SSA, the oldest a 77 year old Korean man, with survival extending up to 4 years without biopsy evidence of amyloid recurrence in the transplanted heart.104, 105
The cumulative transplant experience in FAP defines the following clinical parameters as optimal determinants of survival post-OLT: a) age < 50 years, b) disease duration < 7 years, c) female gender, d) modified body mass index (mBMI) > 600, e) normal autonomic vasomotor regulation and bladder function, f) absence of amyloid cardiomyopathy, and g) V30M ATTR.99 Notably, the V30M ATTR survival advantage is based on a small non-V30M ATTR cohort experience. In 2004, Herlenius et al. published survival post-OLT in 449 V30M ATTR patients and62 non-V30M ATTR, reporting 85% 5 year survival in V30M and 60% 5 year survival in non-V30M patients.98 At the latest posting (31 December 2010), the FAPWTR data set remains limited to 106 non-V30M ATTR patients, including 46 different ATTR genotypes.
The frequency of cardiac deaths in ATTR patients post-OLT prompted consideration of prophylactic versus arrhythmia event-driven pacemaker insertion. Limited experience confounds the analysis. A retrospective one center (Karolinska Institute) experience involving 104 V30M ATTR patients documented 26 pacemaker insertions -- 7 placed pre-operatively and 19 inserted a median 5 years post-OLT.106 OLT did not decrease the rate of arrhythmias and pacemaker insertion did not improve survival.106 A second center (Johannes Gutenberg University) reported no difference in 5-year survival rates among ATTR patients undergoing OLT with (n=9) or without (n=7) a pacemaker.107 A definitive position on the survival impact of pacemaker in OLT for ATTR disease awaits a prospective randomized study.
Discovery and study of a Portuguese family with V30M ATTR genopositivity without clinical disease manifestations identified them as compound heterozygotes – possessing both V30M ATTR and a disease-inhibiting second mutation – T119M ATTR.108 Convinced that the intragenic trans-suppressor effect of T119M on V30M ATTR represented a key to understanding protein misfolding, Kelly’s laboratory at The Scripps Research Institute (La Jolla, CA) generated recombinant TTR tetramers expressing the spectrum of V30M/T119M combinations.109 Introducing T119M TTR into V30M homotetramers inhibited tetramer dissociation under a variety of denaturing stresses.109 Thermodynamic studies demonstrated that T119M expression raised the activation barrier of TTR tetramer dissociation, slowing the rate limiting step of TTR amyloid formation.109, 110 Noting that thyroxine binding also stabilized TTR tetramer, Kelly et al. characterized the thyroxine binding site, selected compounds with steric similarity to thyroxine, screened candidate molecules for TTR tetramer stabilizing effect, and identified two promising small ligands that inhibit ATTR amyloid fibril formation in vitro: diflunisal, an FDA-approved non-steroidal anti-inflammatory drug, and a novel agent, now known as Tafamadis®. By complexing the thyroxine binding site at the dimer-dimer interface, both diflunisal and Tafamadis tighten TTR tetramer associations, inhibiting TTR monomer release and suppressing TTR amyloid fibril formation.39, 110 Although Tafamadis binds the TTR tetramer more tightly, diflunisal overcomes weaker TTR binding coefficients with high serum drug concentrations. A proof-of-concept clinical trial involving Tafamadis in 126 V30M ATTR patients with early (stage I) peripheral neuropathy demonstrated significant slowing of disease progression in subjects completing the 18 months trial.110 The Diflunisal Trial (ClinicalTrials.govidentifierNCT00294671) adopted more inclusive entry criteria, enrolling 130 subjects with a broad range of neurologic disease and unrestricted variant TTR. Data collection over 24 months study participation will conclude in December 2012. The European Medicines Agency (EMA) has recently granted Tafamadis market approval to Pfizer Pharmaceuticals, Inc. for sale of the drug in the European Union. The US FDA recently announced (Feb 2012) that it will accept Pfizer’s new drug application (NDA) for Tafamidis, however at the present time the agent is not available commercially in the US.
Clinical trial evidence that small molecule TTR stabilizers slow neurologic disease progression in man fueled speculation that eliminating TTR expression might completely arrest ATTR disease (Table 2). As proof-of-concept, Benson et al. demonstrated that antisense oligonucleotides (ASO) directed against human I84S ATTR suppressed TTR transcription and RNA translation by up to 80% in a transgenic mouse model.111 Isis Pharmaceutical (Carlsbad, CA) and Alnylam Pharmaceutical (Cambridge, MA) subsequently reported at international meetings that both antisense RNA and RNA interference techniques inhibit TTR mRNA and protein expression by 40–90% in transgenic mice models, non-human primates, and phase I clinical trials. Initial accrual to phase II–III clinical trials examining both RNA technologies in FAP is anticipated in late 2012 or early 2013.
Epigallocatechin-3-gallate (EGCG), the predominant polyphenol in green tea, has been associated with decreasing interventricular septal thickness and left ventricular mass index, increasing left ventricular ejection fraction, and improved NYHA classification in a cohort of patients with AL amyloid cardiomyopathy.112 In FAP, EGCG stabilizes circulating TTR tetramers by binding a dimer-dimer site distinct from the thyroxine transport site through which small ligand stabilizers act.113 Unlike small molecule stabilizers, EGCG also disrupts mature amyloid TTR fibrils in vitro.114 In a human V30M TTR transgenic mouse model of FAP, EGCG inhibited pre-fibrillary TTR deposition, by immunohistochemical measures, while suppressing biomarkers of endoplasmic reticulum (ER) oxidative stress (BiP, Fas, 3-nitrotyrosine) that signal toxicity of those early pre-amyloid intermediates.114 EGCG treatment of the V30M TTR transgenic mice also decreased mature amyloid fibril matrix, confirming reports of amyloid fibril disruption in nerve cell culture.115
Resveratrol or 3,5,4’-trihydroxystilbene, a polyphenol present primarily in grape skins, stabilizes tetrameric TTR by binding the thyroxine transport pocket.39, 116 Studies in cultured human cardiac cells demonstrate that resveratrol limits the toxic effects of pre-fibrillary TTR moieties by promoting tetramer formation from free monomers.116 The bioavailability of orally administered resveratrol is limited due to rapid conjugation in the intestine, with < 5% of free drug ultimately circulating in blood plasma.117 Wine and grape ingestion are unlikely to clinically affect FAP (wine concentration <25 µmol/L).118 A phase II trial examining the effect of Resveratrol (500 mg to 2 grams daily) on biomarkers of Alzheimer’s Disease has been initiated, illustrating the magnitude of doses needed to influence the chemistry of neurodegenerative disease.
Doxycycline, a tetracycline antibiotic, inhibits amyloid fibril formation and disrupts deposited mature fibrils in FAP (human TTR-V30M/mouse TTR-KO)119 and AL transgenic mice (CMV-6).120 Notably, doxycycline administration in the AL transgenic mice experiments (15 mg/L) replicated serum concentrations achieved in humans with standard drug dosing (100 mg twice daily).121 Additionally, doxycycline inhibited matrix metalloproteinase-9 (MMP-9), a mediator of amyloid-induced organ injury, in both mouse models.122 Despite in vitro data demonstrating clearance of mature deposited amyloid fibrils, doxycycline failed to resolve toxic pre-fibrillar TTR moieties.119, 122 In contrast, tauroursodeoxycholic acid (TUDCA), a biliary acid with anti-oxidant and anti-apoptotic activities, decreases toxic TTR pre-fibrillar aggregates in transgenic TTR mice without significant effect on mature fibrils.122 When administered consecutively to TTR transgenic mice for 15 days and 30 days, respectively, doxycycline and TUDA induced significant reductions of fibrillar/congo red staining deposits and non-fibrillar oligomeric TTR.122 Measures of oxidant tissue injury -- BiP, Fas, and 2-nitrotyrosine levels -- all declined with treatment, supporting effective clearance of toxic TTR intermediates.122 To test these findings in an FAP cohort, a phase II open label single center study administering doxycycline (100 mg twice daily) and TUDCA (750 mg daily) for 12 months opened July 2010 at IRCCS Policlinico San Matteo, Pavia, Italy. The organizers anticipate final data collection in July 2012 with data analysis reported December 2012 (http://clinicaltrials.gov).
Transthyretin (TTR) cardiac amyloidosis is an under-appreciated contributor to heart failure in elderly patients. While diagnosis typically requires tissue biopsy and demonstration of amyloid by histologic techniques, cardiac amyloidosis can be also identified non-invasively by echocardiography and cardiac MR. Clinical management of TTR cardiac amyloidosis differs from other forms of heart failure, thus disease recognition is essential. Liver transplantation remains the established treatment for variant TTR-related amyloid neuropathy and cardiomyopathy, but small molecule pharmaceuticals may prove effective alternatives to surgery. In addition, small molecules may provide much needed treatment options for senile systemic amyloidosis, as only heart transplantation averts disease progression at present. The role of new and developing medical treatments for ATTR gene carriers remains to be established.
Drs. Martha Skinner, Lawreen H. Connors, David Seldin, Vaishali Sanchorawala and the Hematology/Oncology Clinical Trials staff.
Funding sources: This work was supported by the American Heart Association (10SDG2550011), the Amyloidosis Foundation (Junior Research Award) to FLR, the National Institutes of Health (R01NS051306) to JLB, and the Young Family Amyloid Research Fund.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Disclosures: Dr. Ruberg acknowledges prior support from FoldRx Pharmaceuticals, acquired by Pfizer Pharmaceuticals in 2010. Dr. Berk acknowledges support from Alnylam Pharmaceuticals, Isis Pharmaceuticals, and Pfizer Pharmaceuticals.