Introduction
Amyloidosis is a rare disorder where abnormal proteins are first misfolded and then
deposited in tissues as highly ordered aggregates forming amyloid fibrils, leading
to progressive organ failure. Amyloid deposition can occur in any organ but the commonly
involved organs are heart, lungs, liver, kidneys, bones, skin, or nerves. While there
are over 30 different types of amyloid fibril proteins, cardiac amyloidosis (CA) is
commonly associated with two types of proteins namely, monoclonal light chain (AL)
and transthyretin (ATTR) amyloid proteins ([Table 1]).[1] ATTR CA is further classified as hereditary or acquired. The acquired form is known
as wild type ATTR (ATTRwt) that was previously thought to be confined to elderly men;
however, higher prevalence in women has been recently reported in a study by González-López
et al.[2] ATTRwt almost exclusively affects the heart and is often associated with carpal
tunnel syndrome and lumbar canal stenosis.[3]
[4]
[5] In contrast, hereditary ATTR (ATTRv) presents in younger patients and can be associated
with either cardiomyopathy (familial amyloidotic cardiomyopathy, FAC) or neuropathy
(familial amyloidotic polyneuropathy, FAP). In contrast, extracardiac manifestations
are more common in AL type of CA.[6]
Table 1
Common types of amyloidosis (original)
|
Type
|
Precursor
|
Distribution
|
Acquired/Hereditary
|
Management
|
|
AL
|
Immunoglobulin light chain
|
Systemic/Localized
|
Acquired/hereditary
|
Chemotherapy, stem cell transplantation
|
|
ATTR
|
Transthyretin wild type
|
Systemic
|
Acquired
|
Liver transplantation
TTR stabilizers
|
|
Transthyretin variants
|
Systemic
|
Hereditary
|
TTR stabilizers
|
Abbreviations: AL, light chain; ATTR, transthyretin amyloidosis.
Cardiac involvement is the main cause of morbidity and mortality associated with amyloidosis.
Diagnosis is often delayed due to the indolent nature of the disease in some forms.
The cardiac presentation can range from asymptomatic increase in ventricular wall
thickness to severe rapidly progressive heart failure.[6]
[7]
[8]
[9] Other clues for cardiac involvement include persistent low-level elevation in serum
troponin, discordance between QRS voltage on electrocardiogram and wall thickness
on imaging ([Fig. 1]), unexplained atrioventricular block or prior pacemaker insertion, unexplained atrial
or ventricular wall thickening, and family history of cardiomyopathy.[8]
Fig. 1 A 64-year-old woman with amyloid deposition on gastric biopsy. (A) Echocardiography shows biventricular hypertrophy (solid white arrows) and atrial fibrillation. (B) Axial contrast-enhanced computed tomography shows left ventricular hypertrophy (solid black arrows). (C and D) cardiovascular magnetic resonance reveals concentric left ventricular hypertrophy
(dashed white arrows in C) and diffuse transmural myocardial enhancement of ventricles (dashed yellow arrows in D).
The gold standard for establishing CA deposition is an endomyocardial biopsy (EMB)
with Congo red histology demonstrating the pathognomonic apple-green birefringence
under polarized light. The limitations of EMB are availability of expertise, complications,
errors in tissue processing and unsuitability for quantification. The main advantage
of histology is typing of the precursor protein that is essential prior to treatment.
With newer treatment options being available, especially for TTR subtype, there is
a need to provide information on the CA subtype, extent of disease involvement for
prognostication, and assessment of treatment response.[8]
[9]
[10] Advances in cardiac imaging have led to early identification and improved diagnostic
confidence of CA for ATTR and have a role in disease quantification and prognostication.
This review aims to describe imaging assessment and approach to diagnosis in clinically
suspected CA.
Echocardiography
Echocardiography is the first line of screening in patients with suspected CA. CA
is characterized by biventricular wall thickening with left ventricular (LV) thickening
measuring more than 12 mm in the setting of a nondilated LV ([Fig. 1A]).[11] The ventricular wall thickening tends to be symmetric in AL and asymmetric in ATTR.[2] The ventricular wall thickening is more prominent in ATTR at diagnosis than in patients
with AL, which may reflect the early clinical presentation of AL.[12] Abnormal septal morphology is also noted in ATTR.[2] In addition to increased LV wall thickness presence of granular sparkling appearance
of the ventricular myocardium, increased thickness of atrioventricular valves or interatrial
septum and pericardial effusion are additional diagnostic clues but are not specific
for CA.[11] Diastolic dysfunction is the earliest abnormality and may occur before the onset
of symptoms. The LV systolic function is often preserved until the advanced stage
of the disease and is measured by ejection fraction reflecting radial thickening.
Besides radial thickening, strain imaging also enables assessment of other parameters
of ventricular contraction, namely circumferential shortening (circumferential strain
[CS]) and longitudinal shortening (longitudinal strain [LS]), of which the longitudinal
function is typically involved earlier in CA than the radial contraction. Myocardial
strain measures deformation of the myocardium as a percentage change in myocardial
length from relaxed state to contractile state produced by application of a force.
Strain ratio is the rate by which the deformation occurs (deformation or strain per
unit time). Myocardial strain imaging in echocardiography is performed by tissue Doppler
imaging, two-dimensional speckle tracking echocardiography (STE), and three-dimensional
STE. STE is the currently the widest available technique to quantify myocardial deformation.[13]
[14] The LS of the LV is severely impaired in CA and can provide an indirect evaluation
of the extent of amyloid infiltration.[15] Patterns in LS have been reported as increasing diagnostic accuracy for amyloidosis
such as ratio of regional LS values (apical:basal strain ratio or relative apical
sparing ratio), or left ventricular ejection fraction to global LS (GLS) ratio.[15]
[16] Abnormal GLS in the mid and basal walls of the LV with normal values in the apex,
known as the apical sparing pattern (or “cherry on top” pattern) ([Fig. 2]), is 93% sensitive and 82% specific in identifying patients with CA.[17] These indices aid in differentiating CA from other causes of LV wall thickening
but are not definitively diagnostic of CA. Additionally, although echocardiographic
features of AL-CA and ATTR-CA have been described, echocardiography does not reliably
distinguish between the two subtypes, and additional imaging with another modality
is often required. Nonetheless, echocardiography remains the superior imaging modality,
especially to assess cardiac hemodynamics and diastolic functional assessment that
is critical in CA.
Fig. 2 A 64-year-old patient genopositive for mutant ATTR c.349G > T (p.Ala117Ser). (A) Four-chamber view with longitudinal strain map. (B) The bullseye longitudinal map of all myocardial segments. Note the reduced global
longitudinal strain at −10.6% and apical sparing (> 2:1 apical/basal ratio or “cherry
on top”) pattern. ATTR, transthyretin amyloidosis.
Nuclear Medicine
Scintigraphy with Bone Seeking Tracers
Bone seeking tracers are the mainstay of diagnosis of CA. The localization of 99mTc-disphosphonate to amyloid deposits was first noted by Kula et al in 1977.[18] Currently used bone imaging tracers are 99mTc pyrophosphate (PYP), 99mTc 3,3-diphosphono-1,2-propanodicarboxylic acid (99mTc DPD), and 99mTc hydroxymethylene diphosphonate (HMDP). There is limited access to 99mTc DPD and 99mTc HMDP in the United States and 99mTc PYP in Europe.[19] Although there is no direct comparison between these tracers, available information
suggests that these can be used interchangeably.[19]
Bone seeking tracers can definitively diagnose amyloid type when a plasma cell dyscrasia
is excluded and may reliably distinguish CA from its mimics.[20]
[21] Bone seeking tracers are highly sensitive in detecting early or asymptomatic ATTR
type of CA and enable reliable diagnosis of ATTR CA without the need for invasive
EMB.[10]
[21]
[22]
[23]
[24] The ability to diagnose ATTR CA noninvasively by bone seeking tracers has rapidly
increased the diagnostic utilization of 99mTc PYP as ATTR CA has become a treatable chronic disease with the approval of tamafidis.[9]
[25] As treatment options comes at a significant expense, it is crucial to accurately
interpret 99mTc PYP imaging.
Imaging Technique
99mTc PYP scan needs no specific patient preparation and has no known contraindication
to tracer administration. The 99mTc PYP doses are either readily available as unit doses or as kits for preparation.
The radio-pharmaceutical (15–20 mCi) is intravenously injected followed by cardiac
or chest (and optional whole body) planar imaging and single-photon positive emission
computed tomography (SPECT) after 1 hour of tracer administration.[26] The planar images are obtained in anterior, lateral, and left anterior oblique views
acquired for 750,000 counts using a 90 degrees detector configuration with 1.46 zoom
factor.[26] The acquisition parameters for SPECT imaging are angular range of 180 degrees, detector
configuration of 90 degree with 40 detectors at a zoom factor of one.[26] Delayed SPECT or planar imaging at 3 hours is optional and is recommended if there
is persistent blood pool activity on 1 hour images.[26] Recently, Masri et al showed that if 99mTc PYP positivity is based on the presence of diffuse myocardial tracer uptake on
SPECT, a time-efficient 1 hour protocol has the same diagnostic efficacy as a 3 hours
protocol.[27]
While planar images can be used for visual interpretation and quantification of myocardial
uptake, SPECT gives further information on the distribution of the tracer uptake.
SPECT is particularly useful in differentiating blood pool uptake from myocardial
uptake and assessing myocardial uptake distribution in individuals with positive planar
scans. SPECT can also identify regional myocardial differences in uptake and identify
uptake in the interventricular septum that is commonly involved in CA. The other advantages
of SPECT are its ability to avoid overlap of bone uptake and to quantify the degree
of myocardial uptake by comparing to the rib uptake. Whole body planar imaging has
the advantage of identifying uptake in other areas of the body beyond the field of
view used in cardiac/chest planar imaging ([Fig. 3]).[28]
Fig. 3 Spectrum of 99mTc-3,3-diphosphono-1,2- propanodicarboxylic acid uptake. (A) Unaffected control subject without visually detectable uptake. (B) Patient with light chain amyloidosis without visually detectable myocardial uptake;
mild uptake is visible only at the soft tissue level. (C and D) Two patients with TTR-related amyloidosis showing strong myocardial uptake with
absent bone uptake (reprinted from Perugini et al[28]).
Image Interpretation
The patterns of myocardial uptake can be categorized as absent, focal, diffuse, or
focal on diffuse. The uptake can be further quantified either by comparing the myocardial
uptake with contralateral lung (H/CL ratio) at 1 hour ([Fig. 4]) or semiquantitatively by comparing to the ribs at 3 hours.
Fig. 4 Raw images of a representative negative (A) and positive subject (B) are shown 1 hour after radiotracer infusion. ROI circles are depicted in red and
the contralateral comparison circle is depicted in blue. C/L, contralateral; Cts,
counts; ROI, region of interest; Std Dev, standard deviation (reprinted from from
Bokhari et al[23]).
For H/CL assessments, a region of interest (ROI) is drawn over the heart that is then
copied and mirrored over the contralateral chest to measure total and absolute mean
counts in each ROI. The fraction of heart ROI mean counts to contralateral chest ROI
mean counts gives the H/CL ratio. In the presence of visually detectable myocardial
uptake on SPECT images, H/CL ratios of more than 1.5 are classified as ATTR positive
and ratios below 1.5 as ATTR negative.[23] Semiquantitative method is a visual comparison of myocardial and bone uptake and
is scored as grade 0-no uptake, grade 1-less than rib, grade 2-equal to rib, and grade
3-more than rib.[28] A grade of 2 or above is considered positive for semiquantitative assessment, a
grade of 1 may either be AL amyloidosis or early ATTR and a negative study can be
seen in normal individuals or with AL amyloidosis. Thus, semiquantitative visual score
of 2 and above or H/CL ratio more than 1.5 is highly suggestive of ATTR type of CA
and can clinch the diagnosis without noninvasive biopsy in the absence of detectable
monoclonal protein.[22]
[23]
Limitations
In some individuals, AL type of CA may be superimposed by ATTR type, especially in
elderly individuals, resulting in positive 99mTc PYP and evidence of plasma cell dyscrasias. EMB would still be required in these
cases to exclude AL type of CA. A negative or equivocal 99mTc PYP scan despite high clinical suspicion of ATTR may also require an EMB to confirm
ATTR type of CA. Although highly sensitive for diagnosing ATTR, bone seeking tracers
have limited role in disease follow-up and monitoring treatment response.
Sympathetic Innervation
Amyloid deposits impair the function of myocardial sympathetic nerve endings and may
even lead to sudden death due to fatal arrythmias. The impaired cardiac sympathetic
function can be detected in early stages of amyloidosis on I-123 metaiodobenzylguanidine
(I-123 MIBG) scintigraphy and in conjunction with cardiac function indices can predict
lethal cardiac events and identify the set of patients that can benefit from prophylactic
implantable cardioverter-defibrillator implantation.[29]
[30]
[31] I-123 MIBG scintigraphy is analyzed by heart-to-mediastinum activity ratio (HMR)
and by cardiac washout ratio that is defined as the percentage change in activity
ratio from early to late images. Patients with FAP show significant reduction of MIBG
uptake despite preserved LV function and cardiac perfusion. Severity of polyneuropathy
correlates inversely with MIBG uptake. Cardiac washout is high in both ATTR and AL
patients, whereas the HMR is significantly lower in ATTR.[29] It is not known if patients with ATTRwt or ATTRv FAC without predominant autonomic
system dysfunction would benefit from sympathetic innervation imaging and the current
role in CA imaging is limited.
Positron Emission Tomography
Fluorodeoxyglucose has limited use in detecting CA as it is difficult to differentiate
pathological uptake from physiological uptake.[32] Direct imaging of the amyloid plaques is possible through positron emission tomography
imaging using Pittsburgh compound B(11C-PIB), 18F-Florbetaben, and 18F-Florbetapir
tracers.[33]
[34]
[35]
[36] However, the high cost and limited availability limit the use of these tracers in
routine imaging.
I-123 Serum Amyloid P Compound and Technetium-Labeled Aprotinin
I-123 serum amyloid P compound and technetium-labeled aprotinin scintigraphy have
role in assessing the whole-body burden of certain types of amyloidosis but their
use is limited in CA.[37]
[38]
[39]
[40]
Cardiac Magnetic Resonance Imaging
Late Gadolinium Enhancement
The extracellular gadolinium-based contrast agents (GBCA) distribute in the extracellular
space of the myocardium but are excluded from the normal myocyte. In CA, the expanded
extracellular space retains contrast on delayed imaging that commonly imparts a global
pattern of late gadolinium enhancement (LGE) ([Fig. 1D]).[41] Global LGE is associated with high amyloid burden and markers of systolic and diastolic
dysfunction and provides incremental prognostic information.[42]
[43] The distinct pattern in CA is diffuse global subendocardial LGE, common in AL type
([Fig. 5]).[44]
[45] It is also a strong predictor of 1-year mortality in patients with suspected CA.[46] In ATTR, the LGE is more extensive and transmural ([Fig. 5]).[47] The degree of LGE can also be quantified by query amyloid late enhancement (QALE)
score that is an independent predictor of amyloid type ([Fig. 6]).[47] The total score ranges from 0 (no LGE) to 18 (global transmural LV LGE and RV involvement).
The QALE score is higher in ATTR, a score of 13 or above is 82% sensitive and 76%
sensitive in differentiating ATTR from AL type.[47] In AL type, a score of 9 or below is associated with better prognosis.[48]
Fig. 5 Schematic representation of global subendocardial and transmural patterns of late
gadolinium enhancement (LGE) in amyloidosis (enhancement depicted in white).
Fig. 6 Schematic representation of query amyloid late enhancement (QALE) score. The QALE
score is assessed on late gadolinium enhancement (LGE) images at the level of base,
mid, and apex of the ventricles. The highest score is 4 at each level or 6 if right
ventricle is involved. The total score ranges from 0 (no LGE) to 18 (global transmural
left ventricle (LV) LGE and right ventricle (RV) involvement).
Suboptimal Nulling of the Myocardium
Inversion recovery technique is employed after 10 minutes postcontrast administration
to detect LGE.[49] Prior to LGE, incremental inversion times (TI) are applied at mid-ventricular level
(TI scout sequence) to determine the optimal TI that nulls the myocardium completely.
Normally, after contrast administration the blood pool nulls before myocardium giving
good contrast between the myocardium and the blood pool. In CA, increased gadolinium
retention in the myocardium shortens the T1 and the myocardium nulls before the blood
pool. This suboptimal nulling of the myocardium limits the selection of optimal TI
and is a strong predictor of mortality.[50] Loss of temporal nulling of blood pool, myocardium, and spleen in that order in
such patients indicates that the suboptimal nulling of the myocardium is related to
CA rather than technical failure.[51]
Myocardial T1 Mapping
Native T1 mapping quantifies the longitudinal relaxation of tissue without GBCA and
can be applied even in patients where GBCA is contraindicated such as in renal failure.
A series of images at various inversion times are acquired to derive a T1 recovery
curve from which a parametric image that displays T1 relaxation is obtained.[52] T1 values can then be quantified by placing a ROI on the myocardium on the T1 map.[52] Native T1 mapping requires measurement of normal values locally for each scanner
configuration.[53]
[54] On the other hand, postcontrast T1 mapping can be readily incorporated into standard
delayed gadolinium enhancement-cardiac magnetic resonance protocols without significant
prolongation of study duration.[55] However, it is variable and dependent on various factors such as dosage of contrast
administered, time between contrast administration, and image acquisition and renal
clearance.[56]
Partition coefficient (λ) is the relationship between changes in pre- and postcontrast
myocardium and blood T1 and is calculated by the ratio of the difference of reciprocal
values of postcontrast and native T1 of myocardium to that of the blood pool.
Extracellular volume (ECV) fraction is obtained from partition coefficient by correcting
for hematocrit (λ x [100-hematocrit]) and gives a direct measure of extracellular
space reflecting interstitial disease without myocytes.[53]
[54]
[57] The native myocardial T1 and the ECV are increased in CA.[58]
[59]
[60] Amyloid deposition is more extensive in ATTR that is reflected as higher ECV than
in AL.[61] Despite the higher amyloid volume in ATTR, the native myocardial T1 is lower than
in AL. This discordance is probably due to edema in AL that also contributes to the
T1 value.[58]
[59]
[60] While both the native myocardial T1 and ECV correlate with indexes of systolic and
diastolic function, indexed LV mass, and known prognostic biomarkers, the correlation
is stronger with ECV.[61]
Myocardial T2 Mapping
T2 mapping employs similar principle used in T1 mapping where a series of T2-weighted
images are applied to obtain T2 decay curve with sequences such as turbo spin echo
(TSE) with varying echo time, T2 preparation steady-state free precession (SSFP) sequence,
or a sequence scheme that combines spin echo excitation with gradient echo readout.[54] T2 mapping is also a noncontrast sequence and hence can be used in patients with
renal failure. However, T2 mapping is prone to confounders such as sensitivity to
T1 and off resonance effects.[54] T2 relaxation time increases with increasing water content and is elevated in CA
owing to myocardial edema. While myocardial T1 is a measure of combination of amyloid
burden and myocardial edema, T2 is more specific for myocardial edema. T2 values are
higher in AL than in ATTR and within AL, the values are higher in treatment naïve
patients than during or after treatment.[62] It is well known that AL portrays poor survival compared with ATTR and hence myocardial
edema is a possible additional mechanism in myocardial damage and predictor for outcome
in AL patients.[62]
Myocardial Tagging
Myocardial tagging analyzes deformation or strain of myocardium to quantify LV wall
motion abnormalities ([Fig. 7]). Lines or grids are superimposed on the myocardium at the beginning of a cine sequence
and the subsequent deformation of the lines throughout the cardiac cycle is noted
that is reduced in CA.[63] CS is shortening in short-axis plane in a direction tangential to epicardial surface.
LS is base to apex shortening tangential to the myocardial wall in the long-axis plane.
The LS is significantly decreased in CA when compared with hypertrophic cardiomyopathy
that is an imaging mimicker of CA, but further studies are needed to establish differentiation
from other causes of LV hypertrophy.[64] Prominent base to apex gradient higher at the apical level suggests apical sparing
that is known to be specific for CA on echocardiography.[64] Good correlation of CS and LS with positive CA and the quantitative nature of the
parameters have a potential for disease detection, quantification of amyloid load,
and prognostication.[65]
Fig. 7 Cardiovascular magnetic resonance short-axis (A) and long-axis (B) images through the mid-left ventricle demonstrate application of myocardial tagging
at end-diastole. The tag lines are seen as dark lines that deform along with the myocardium
during systole.
Cardiac Computed Tomography
The morphological manifestation of myocardial thickening in CA can be appreciated
on CT ([Fig. 1B]).[66] Studies have shown that ECV can also be estimated on conventional as well as dual-energy
CT.[57]
[67] However, poor signal to noise ratio compared with cardiac magnetic resonance (CMR)
and the use of ionizing radiation limit wider application CT for myocardial tissue
characterization.[68] CT may be of value in the subset of patients where magnetic resonance imaging may
not be feasible such as in patients with pace makers and claustrophobia.
Diagnostic Approach
In patients with clinically suspected amyloidosis, echocardiography is often the first
line of investigation. If echocardiography shows features indicative or suggestive
of CA, CMR should be considered as the next line of investigation followed by bone
scan. When combined with monoclonal protein detection, there is high diagnostic confidence
of bone scan for diagnosing ATTR CA or in ruling out CA. In the absence of monoclonal
proteins, a high grade (2 or 3) visual uptake or H/CL ratio more than 1.5 increases
the specificity of ATTR amyloidosis, while CA can be deemed unlikely when there is
no visual uptake (grade 0). In the former subset of patients, combination of clinical
features and echocardiography or CMR findings, diagnosis of ATTR can be established
without histological confirmation.[22] Further genotyping is required to differentiate between wild type and variant amyloid,
the latter showing TTR gene mutation.[22] For the cases that do not fall into either category, further review or assessment
with CMR and/or histological confirmation and amyloid typing may be required for further
evaluation. There has been significant progress in the treatment of CA and the therapies
are specific for the type of CA. Hence, there is need to diagnose CA at an early stage
and to identify the amyloid type. Thus, the goal of the diagnostic algorithm ([Fig. 8]) is to direct therapeutic options in treating CA.
Fig. 8 Diagnostic algorithm in suspected cardiac amyloidosis (CA). Echocardiography is the
first line of investigation in clinically suspected patients with CA. Cardiac magnetic
resonance imaging (CMR) is not diagnostic but can exclude other causes or direct further
biopsy. ATTR is more likely if the monoclonal assay is negative and the bone scintigraphy
shows grade 2–3 uptake. A combination of positive monoclonal assay and positive CMR
findings can indicate light chain type of CA. ATTR, transthyretin amyloidosis.
