Clinical Research in Cardiology (2021) 110:555–568

https://doi.org/10.1007/s00392-020-01771-1

ORIGINAL PAPER

Diagnostic value of cardiovascular magnetic resonance in comparison

to endomyocardial biopsy in cardiac amyloidosis: a multi-centre study
Grigorios Chatzantonis1 · Michael Bietenbeck1 · Ahmed Elsanhoury2 · Carsten Tschöpe2,3 · Burkert Pieske4 ·
Gloria Tauscher5 · Julia Vietheer6,7 · Zornitsa Shomanova1 · Heiko Mahrholdt5 · Andreas Rolf6,7 · Sebastian Kelle4 ·
Ali Yilmaz1

Received: 28 August 2020 / Accepted: 26 October 2020 / Published online: 10 November 2020
© The Author(s) 2020

Abstract
Background Cardiac amyloidosis (CA) is an infiltrative disease characterised by accumulation of amyloid deposits in the
extracellular space of the myocardium—comprising transthyretin (ATTR) and light chain (AL) amyloidosis as the most
frequent subtypes. Histopathological proof of amyloid deposits by endomyocardial biopsy (EMB) is the gold standard for
diagnosis of CA. Cardiovascular magnetic resonance (CMR) allows non-invasive workup of suspected CA. We conducted
a multi-centre study to assess the diagnostic value of CMR in comparison to EMB for the diagnosis of CA.
Methods We studied N = 160 patients characterised by symptoms of heart failure and presence of left ventricular (LV)
hypertrophy of unknown origin who presented to specialised cardiomyopathy centres in Germany and underwent further
diagnostic workup by both CMR and EMB. If CA was diagnosed, additional subtyping based on EMB specimens and
monoclonal protein studies in serum was performed. The CMR protocol comprised cine- and late-gadolinium-enhancement
(LGE)-imaging as well as native and post-contrast T1-mapping (in a subgroup)—allowing to measure extracellular volume
fraction (ECV) of the myocardium.
Results An EMB-based diagnosis of CA was made in N = 120 patients (CA group) whereas N = 40 patients demonstrated
other diagnoses (CONTROL group). In the CA group, N = 114 (95%) patients showed a characteristic pattern of LGE indica-
tive of CA. In the CONTROL group, only 1/40 (2%) patient showed a “false-positive” LGE pattern suggestive of CA. In
the CA group, there was no patient with elevated T1-/ECV-values without a characteristic pattern of LGE indicative of CA.
LGE-CMR showed a sensitivity of 95% and a specificity of 98% for the diagnosis of CA. The combination of a characteristic
LGE pattern indicating CA with unremarkable monoclonal protein studies resulted in the diagnosis of ATTR-CA (confirmed
by EMB) with a specificity of 98% [95%-confidence interval (CI) 92–100%] and a positive predictive value (PPV) of 99%
(95%-CI 92–100%), respectively. The EMB-associated risk of complications was 3.13% in this study—without any detri-
mental or persistent complications.
Conclusion Non-invasive CMR shows an excellent diagnostic accuracy and yield regarding CA. When combined with
monoclonal protein studies, CMR can differentiate ATTR from AL with high accuracy and predictive value. However,
invasive EMB remains a safe invasive gold-standard and allows to differentiate CA from other cardiomyopathies that can
also cause LV hypertrophy.

Keywords CMR · CA · EMB · Scintigraphy · Immunofixation

Abbreviations
Electronic supplementary material The online version of this AL Immunoglobulin light chain amyloidosis
article (https://doi.org/10.1007/s00392-020-01771-1) contains
ATTR Transthyretin amyloidosis
supplementary material, which is available to authorized users.
CA Cardiac amyloidosis
G. Chatzantonis and M. Bietenbeck contributed equally to this CAD Coronary artery disease
work. CIED Cardiovascular implantable electronic device
CKD Chronic kidney disease
* Ali Yilmaz
ali.yilmaz@ukmuenster.de CMR Cardiovascular magnetic resonance
ECG Electrocardiogram
Extended author information available on the last page of the article

13
Vol.:(0123456789)
556 Clinical Research in Cardiology (2021) 110:555–568

ECV Extracellular volume considered the gold standard for workup of cardiomyopa-
EMB Endomyocardial biopsy thies of unknown origin [9–11]. The need for invasive EMB
HCM Hypertrophic cardiomyopathy was recently questioned by Gillmore et al. in case of sus-
HFpEF Heart failure with preserved ejection fraction pected CA since the combined finding of a “positive” bone
IQR Interquartile range scintigraphy indicative of CA and the absence of monoclo-
LGE Late-gadolinium enhancement nal proteins resulted in a specificity as well as positive pre-
LV Left ventricle dictive value (PPV) of 100% for the diagnosis of cardiac
LV-EDV Left ventricular end-diastolic volume ATTR [12].
LV-EF Left ventricular ejection fraction In the present multi-centre study, we sought to assess
LVH Left ventricular hypertrophy the diagnostic value of CMR regarding the diagnosis of CA
MGUS Monoclonal gammopathy of undetermined in patients with unexplained left ventricular hypertrophy
significance (LVH)—in comparison to the current gold-standard EMB.
NPV Negative predictive value Similar to previous scintigraphic studies [12], we analysed
NSF Nephrogenic systemic fibrosis the specificity and PPV of CMR in patients with the com-
PPV Positive predictive value bined finding of a “positive” CMR study indicative of CA
QALE Query amyloid late enhancement and “negative” monoclonal protein studies.
ROC Receiver operating characteristic curve
ROI Region of interest
sFLC Serum-free light chain assay
sIFE Immunofixation electrophoresis in serum Methods
SPE Serum protein electrophoresis
TTE Transthoracic echocardiogram Study population

The present study group comprised N = 160 patients suf-

Introduction fering from symptoms of heart failure in the presence of
left ventricular (LV) hypertrophy (defined as maximal LV
Cardiac amyloidosis (CA) is caused by deposition of mis- wall thickness of ≥12 mm) that could not be explained by
folded amyloid fibrils in the extracellular space of the myo- abnormal loading conditions and who were referred to four
cardium resulting in a specific cardiomyopathy [1, 2]. The German specialised cardiomyopathy centres between 2016
two major forms of CA are transthyretin-related amyloidosis and 2019 for further diagnostic workup. Only patients who
(ATTR) and immunoglobulin light-chain amyloidosis (AL), underwent both non-invasive CMR and invasive EMB,
accounting for nearly 95% of cases [3]. ATTR comprises respectively, were included in this retrospective study. Note-
two subtypes: the acquired wild-type (wt) ATTR (wtATTR) worthy, patients with e.g. typical CMR findings of hyperten-
and the hereditary or mutant ATTR (mATTR) that is caused sive heart disease or hypertrophic cardiomyopathy (HCM)
by genetic mutations in the transthyretin (TTR) gene [1, 4]. who did not routinely undergo EMB were not included in
Cardiovascular magnetic resonance (CMR) has emerged this analysis. Moreover, the participating German centres
as a robust non-invasive imaging modality that offers com- share similar diagnostic algorithms including EMB workup
prehensive and detailed cardiac information regarding func- in every patient with suspected CA. If EMB workup resulted
tional and structural data [5]. In particular, late-gadolinium- in the diagnosis of CA, additional subtyping based on EMB-
enhancement (LGE)-imaging allows to robustly detect a specimens was performed to differentiate ATTR from AL
characteristic pattern of myocardial damage indicative of CA (and other forms). Furthermore, if CA was diagnosed based
[6]. Moreover, novel CMR techniques such as T1-mapping on CMR and/or EMB, additional monoclonal protein stud-
and measurement of the extracellular volume fraction (ECV) ies—comprising serum protein electrophoresis (sPE), serum
of the myocardium promise an improved assessment of even immunofixation electrophoresis (sIFE) and serum-free light
subtle structural changes in the myocardium. Some recent chain assay (sFLC)—were performed. In case of an EMB-
single-centre studies suggested a higher diagnostic yield and based diagnosis of ATTR, additional genetic testing of the
prognostic value of CMR based on T1-mapping and ECV transthyretin gene was ordered to differentiate between
measurement compared to conventional LGE-imaging [7, 8]. wild-type (wt)-ATTR and mutant (m)-ATTR. Patients with
However, multi-centre CMR data regarding the diagnostic an EMB-based diagnosis of CA (either ATTR or AL) were
yield of CMR parameters for the diagnosis of CA are still assigned to the CA group whereas patients with other car-
limited. diac diagnoses formed the CONTROL group (Fig. 1). Eth-
Despite considerable progress in non-invasive imaging ics approval was obtained by local authorities and written
modalities, invasive endomyocardial biopsy (EMB) is still informed consent was obtained by participants.

13
Clinical Research in Cardiology (2021) 110:555–568 557

Fig. 1 Cardiovascular magnetic resonance (CMR) images of a patient in the control group. Apart from LGE images, native T1-maps and
with a characteristic pattern of late-gadolinium-enhancement (LGE) ECV-maps are shown
indicative of cardiac amyloidosis (CA)—in comparison to findings

CMR acquisition centres and N = 93 out of 160 patients)—allowing to

measure extracellular volume fraction (ECV) of the myo-
CMR studies were performed on 1.5-T scanners (either cardium. A CMR-based diagnosis of CA was made if a
Ingenia or Achieva, Philips Healthcare, Best, the Neth- characteristic LGE-pattern indicative of CA was present.
erlands or Magnetom Aera, Siemens Medical Systems, T1-mapping images were obtained in three short-axis
Erlangen, Germany) in N = 151 patients and on a 3.0-T views prior to as well as 15–20 min after intravenous
scanner (Skyra, Siemens Medical Systems, Erlangen, contrast agent administration (Gadobutrol or Gd-DOTA
Germany) in N = 9 patients. The CMR protocol com- 0.15 mmol/kg) to determine native T1- and ECV-values.
prised cine- and LGE-imaging [magnitude only and Patients with chronic kidney disease (CKD) of stage 4 or
additional phase-sensitive inversion recovery (PSIR) in 5 [i.e. glomerular filtration rate (GFR) ≤30 ml/min] were
case of need] as well as native and post-contrast T1-map- not excluded by default. Furthermore, patients with car-
ping using a modified Look Locker inversion recovery diac devices were only excluded if they showed fractured,
(MOLLI) sequence (performed in 3 out of 4 participating epicardial or abandoned leads prior to the CMR study

13
558 Clinical Research in Cardiology (2021) 110:555–568

and/or poor image quality due to large device artefacts (Supplemental Figure 1). Continuous ECG, blood pressure
during the CMR scan. and pulse oximetry monitoring were performed throughout
the whole procedure. Fluoroscopy was used for guidance of
CMR data analysis the long-sheath and bioptome, and for targeting the region
of interest—which was defined in advance by non-invasive
CMR image analysis and interpretation were performed CMR. At least four EMB samples were collected from the
using commercially available software (cvi42—version RV and/or LV. Post-procedural echocardiography was per-
5.11.0, Circle Cardiovascular Imaging, Calgary, Alberta, formed to rule out or detect a pericardial effusion possibly
Canada). Analysis of ventricular volumes and function as caused by the biopsy procedure. Definition and assessment
well as LV mass was made by contouring the endocardium of biopsy complications was in accordance with previous
and epicardium in short-axis cine-images. First, LGE- studies [11].
images were assessed visually and the pattern, distribution
and extent of LGE was used to derive a CMR-based diag-
EMB workup (histology and immunohistochemistry)
nosis of the underlying cardiac disease—as illustrated in
detail elsewhere [18]. A CMR-based diagnosis of CA was
Histopathological and molecular pathological workup of
supposed if a characteristic LGE-pattern indicative of CA
biopsy samples were performed as described in detail previ-
(comprising all of the following criteria) was observed: (1)
ously [11, 16]. Myocardial inflammation indicative of myo-
subendocardial to transmural LGE pattern predominantly
carditis was defined on the basis of immunohistochemical
in the basal LV segments, (2) no LGE distribution correlat-
analyses based on hematoxylin/eosin and Masson trichrome
ing to the perfusion area of a coronary artery and suggest-
stainings, respectively. Furthermore, Masson trichrome and
ing an ischemic myocardial scar, (3) no sharp demarcation
hematoxylin/eosin staining as well as electron microscopy
and rather diffuse and extensive LGE pattern (Fig. 2a). In
allowed EMB-based diagnoses of (amongst others) hyper-
case of poor image quality of magnitude-only LGE images,
trophic cardiomyopathy (HCM), dilated cardiomyopathy
additional PSIR-LGE were assessed (Fig. 2b). LGE was
(DCM), cardiac sarcoidosis, transplant rejection, LV non-
described as non-characteristic, when the aforementioned
compaction cardiomyopathy (LVNC), mitochondrial cardio-
criteria were only partially fulfilled. In addition, the “Query
myopathy, toxic cardiomyopathy, glycogen-storage disease
Amyloid Late Enhancement” (QALE) score was reported as
and CA, respectively. Detection of amyloid was initially
described in more detail elsewhere [13]. T1- and ECV-maps
performed with Congo red staining of the formalin-fixed,
were assessed based on the consensus statement of SCMR.
paraffin-embedded myocardial tissue samples. Subsequently,
In those N = 9 patients who were studied at 3.0-T, only ECV
immunohistochemistry allowed to identify the type of pro-
values (but not absolute T1-values) were considered for fur-
tein subunit with the use of monospecific antibodies reac-
ther analyses. All CMR data analyses were performed offline
tive with the various types of amyloid. A negative result
by experienced readers.
for CA was presumed only after thorough investigation by
a specialised pathologist confirming negative histological
Monoclonal protein studies
and immunohistochemical examinations. Additional electron
microscopy studies were performed depending on the pre-
Patients with suspected CA were tested for the presence of
ceding clinical and imaging suspicion and the pathologist’s
monoclonal gammopathy by studying the presence of mono-
individual assessment.
clonal proteins in serum—suggestive of AL. First, sPE with
high-resolution agarose gel allowed the demarcation of an
M-gradient. If an M-protein was present, sIFE and sFLC Genetic testing
were performed to characterise its immunoglobulin chain
type (heavy vs. light). The presence of a monoclonal protein In those patients with EMB-proven ATTR amyloidosis,
was defined as an abnormal sFLC ratio (kappa-lambda ratio additional testing regarding the presence of a TTR gene
<0.26 or >1.65) or presence of a monoclonal band in sIFE. mutation was in general aimed at. However, due to medico-
legal reasons, genetic testing was only performed in N = 56
EMB procedure out of 160 patients. Informed written consent for genetic
testing in accordance with the Genetic Diagnostics Act
EMB specimens were obtained either from the RV septal (GenDG) was obtained. Thereafter, DNA extraction from
wall (via a femoral vein access site using a 7F long-sheath the patient’s blood with amplification by polymerase chain
with an angulated tip) and/or from the LV free wall (via reaction assay and sequencing of the coding region of the
a femoral artery access site using a 7F long-sheath with- TTR gene was performed. Mutations in the TTR gene asso-
out angulation following a retrograde approach) [14, 15] ciated with amyloid deposition are described elsewhere.

13
Clinical Research in Cardiology (2021) 110:555–568 559

Fig. 2 a Example of a patient with ATTR amyloidosis and a char- inversion recovery (PSIR) LGE-images. PSIR-LGE-images show an
acteristic late-gadolinium-enhancement (LGE)-pattern indicative improved image contrast and allow a better delineation of LGE in this
of cardiac amyloidosis. b Example of another patient with ATTR example
amyloidosis and both magnitude only and additional phase-contrast

13
560 Clinical Research in Cardiology (2021) 110:555–568

Statistical analysis in CONTROLs; p < 0.001] as expected due to the higher

prevalence of CA in elderly patients. Atrial fibrillation was
Statistical analysis was performed with SPSS (version 26.0, more common in CA patients (49 vs. 25%; p = 0.010).
IBM Corp., Armonk, NY). Normal distribution of numeric
variables was assessed with Shapiro–Wilk test. Continu- Endomyocardial biopsy (EMB) findings
ous variables are expressed as mean ± standard deviation
whereas skewed variables as median ± interquartile range. EMB was performed in all patients of the study group within
Categorical variables are expressed as frequency with per- a narrow time window after the CMR study [median time
centage. An independent two-sample T test was used for from CMR to EMB = 1 (0–3) months]. Regarding EMB-
comparison of normally distributed data and Mann–Whitney associated complications, there were four patients with myo-
U test for non-normally distributed variables. The Fisher’s cardial perforation resulting in pericardial effusion with need
exact test was used to compare non-continuous variables. for pericardiocentesis and one patient with a long-sheath
Receiver operating characteristic curves (ROC) were ana- induced non-sustained ventricular tachycardia. Importantly,
lysed to assess the specificity and sensitivity of different all complications were treated quickly and successfully, and
CMR measurements to identify patients with CA within the there were no detrimental or persistent complications at all.
whole study group. A p value <0.05 was considered statisti- The overall complication rate of the EMB procedure was
cally significant. Sensitivity and specificity of CMR in diag- 3.13%.
nosing CA as well as positive (PPV) and negative predic- The diagnosis of CA in EMB samples was made accord-
tive values (NPV) were calculated accordingly. Confidence ing to the aforementioned histopathological definitions in
intervals (CI) for sensitivity and specificity were calculated N = 120 patients. Amyloid-subtyping by targeted immuno-
with the exact Clopper–Pearson method; CI for the predic- histochemistry resulted in the diagnosis of cardiac ATTR
tive values was given as standard logit CI. amyloidosis in N = 92 (77%) patients and cardiac AL amy-
loidosis in N = 28 (23%) patients. In the remaining N = 40
CONTROL patients without cardiac amyloidosis, however,
Results with the presence of LV hypertrophy per definition, the fol-
lowing diagnoses were obtained: N = 10 HCM, N = 9 myo-
Study population carditis, N = 3 cardiac sarcoidosis, N = 3 transplant rejection,
N = 3 LVNC, N = 3 DCM, N = 2 mitochondrial cardiomyo-
The characteristics of the study population are summarised pathy, N = 2 toxic cardiomyopathy, N = 1 glycogen-storage
in Table 1. An EMB-based diagnosis of CA was made in disease, N = 1 cardiac involvement in filaminopathy and
N = 120 patients (CA group) whereas N = 40 patients dem- N = 3 non-ischemic cardiomyopathy of unknown aetiology.
onstrated other diagnoses (CONTROL group). Males and
females were equally distributed in the CA and CONTROL Genetic transthyretin testing results
group (p = 0.48). Median age differed significantly between
the two groups [75 (68–80) years in CA vs. 52 (46–61) years The majority of those N = 92 patients with biopsy-proven
ATTR amyloidosis underwent additional TTR gene test-
ing for potential pathogenic TTR gene mutations. Muta-
Table 1 Patient characteristics
tions in the TTR gene indicating the presence of mATTR
CA group CONTROL group P value were detected in N = 6 patients. The remaining patients with
N = 120 N = 40 biopsy-proven ATTR (94%) were classified as wtATTR.
Male, n (%) 100 (83) 31 (78) 0.48
Age, years 75 (68–80) 52 (46–61) <0.001 CMR findings
BMI, kg/m2 26 (±4) 26 (±6) 0.87
Hypertension, n (%) 87 (73) 22 (55) 0.05 The detailed anatomic, functional and structural CMR find-
Diabetes, n (%) 21 (18) 4 (10) 0.32 ings are listed in Table 2. Left ventricular ejection fraction
High cholesterol, n (%) 64 (53) 11 (28) 0.006 (LV-EF) was significantly reduced in the CONTROL group
Current smoker, n (%) 15 (13) 15 (38) 0.001 compared to the CA group [40 (32–51) % vs. 57 (46–62)
Coronary artery disease, 48 (40) 8 (20) 0.023 %; p < 0.001]. Accordingly, LV end diastolic volume (LV-
n (%) EDV) was markedly increased in the CONTROL group com-
Atrial fibrillation, n (%) 59 (49) 10 (25) 0.010 pared to the CA group (116 ± 41 ml/m2 vs. 77 ± 22 ml/m2;
Chronic kidney disease 11 (9) 0 (0) 0.07 p < 0.001). Due to our study inclusion criteria, LV hyper-
(GFR < 30 ml/min), trophy was present in all study patients—without a relevant
n (%)
difference in (global) LV mass between the CA and the
Bold P values indicate P < 0.05

13
Clinical Research in Cardiology (2021) 110:555–568 561

Table 2 CMR parameters CA group CONTROL group P value

N = 120 N = 40

Conventional CMR parameters

 LV-EF, % 57 (46–62) 40 (32–51) <0.001
 LV-EDV index, ml/m2 77 (±22) 116 (± 41) <0.001
 LV-ESV index, ml/m2 31 (23–42) 61 (44–99) <0.001
 LV mass index, g/m2 91 (71–105) 90 (68–110) 0.69
 LV mass/volume ratio, g/ml 1.17 (0.98–1.34) 0.83 (0.65–1.07) <0.001
 Max. LV wall thickness, mm 19 (17–22) 14 (13–18) <0.001
 RV-EF, % 49 (±11) 51 (± 10) 0.41
 RV-EDV index, ml/m2 82 (±22) 92 (± 28) 0.06
 RV-ESV index, ml/m2 39 (30–49) 44 (28–58) 0.34
LGE
 LGE extent, % 82 (52–100) 35 (20–59) <0.001
 LGE QALE score, n 12 (7–17) 4 (2–5) <0.001
 Characteristic LGE indicative of CA, n (%) 114 (95) 1 (2) <0.001
T1-mapping
 Native T1-mapping global, ms 1152 (1095–1245) 1043 (1036–1104) <0.001
 Native T1-mapping basal septal, ms 1139 (1086–1233) 1062 (1018–1137) 0.001
ECV
 ECV global, % 52 (±10) 34 (±10) <0.001
 ECV basal septal, % 53 (43–63) 31 (27–39) <0.001

Bold P values indicate P < 0.05

CONTROL group (p = 0.69). However, a significantly higher regarding the delineation of CA patients from the CON-
maximal LV wall thickness was observed in the CA group TROL group (Fig. 3). In particular, the parameter “charac-
compared to CONTROLs [19 [17–22] mm vs. 14 (13–18) teristic pattern of LGE indicative of CA” (assessed as char-
mm; p < 0.001]. acteristic for CA according to the aforementioned criteria)
Regarding myocardial structure analyses, a non-ischemic, showed an area under the curve (AUC) of 0.97 (95%-CI:
diffuse subendocardial to transmural pattern of LGE pre- 0.89–1.00; p < 0.001) whereas global ECV had an AUC of
dominantly present in the LV basal to midventricular seg- 0.87 (95%-CI: 0.77–0.98; p < 0.001) and global native T1 an
ments was detected in CA patients—with a substantially AUC of 0.76 (95%-CI: 0.60–0.92; p = 0.003).
higher myocardial LGE extent in comparison to the CON- LGE-imaging findings were further validated in compari-
TROL group [82 (52–100) % vs. 35 (20–59) %, p < 0.001]. son to biopsy findings. The respective LGE findings were
Accordingly, the assessment of the QALE score resulted in classified dichotomously into “characteristic LGE pattern
a significantly increased score in the CA group compared indicative of CA” vs. “LGE pattern NOT unequivocally
to the CONTROL group [12 (7–17) vs. 4 (2–5); p < 0.001]. indicative of CA”. There was only one patient with a “char-
Furthermore, both native T1- and ECV-values were signifi- acteristic LGE pattern indicative of CA” that could not be
cantly increased in the CA group compared to the CON- verified by EMB. Review of this patient’s case showed that
TROLs—not only in the basal septal wall but also in case of a RV biopsy was performed and that the bioptome was not
global LV assessment. In the CA group, there was no patient optimally positioned in the basal to mid part of the inter-
with elevated T1-/ECV-values without a characteristic pat- ventricular septum but rather in the apical part. Hence, a
tern of LGE indicating CA (in those patients with available potential “sampling error” due to suboptimal positioning of
T1-maps). Noteworthy, neither native T1- nor ECV-values the bioptome may have resulted in a “false-negative” EMB
were available in those six patients with EMB-based diag- result. Furthermore, there were 6 out of 120 (5%) patients
nosis of CA, but absence of a LGE-pattern indicative of CA. with EMB-proven CA who did not show a “characteristic
Subsequent ROC analyses showed excellent diagnostic LGE pattern indicative of CA” upon CMR. In comparison
yield for the parameters (a) predefined characteristic pattern to the gold standard EMB, the non-invasive CMR param-
of LGE indicative of CA, (b) global ECV—and a slightly eter “characteristic LGE pattern indicative of CA” showed a
reduced diagnostic value for (c) global native T1-mapping sensitivity of 95% and a specificity of 98% for the diagnosis

13
562 Clinical Research in Cardiology (2021) 110:555–568

Fig. 3 Receiver operating characteristic (ROC) curves illustrating the diagnostic yield of different CMR parameters regarding the diagnosis of
cardiac amyloidosis (CA)

Table 3 Sensitivity and specificity of CMR for the diagnosis of CA Combination of CMR and monoclonal protein
compared to EMB findings
Characteristic LGE pattern Sensitivity and
LGE indicative NOT indicative specificity (CI), % Presence of monoclonal proteins in the serum suggestive
of CA of CA of AL was found in all N = 28 patients with biopsy-proven
EMB- 114 (95) 6 (5) 95 (89–98) sensi- AL amyloidosis. In contrast, there were 6 out of 92 (7%)
proven tive patients with biopsy-proven ATTR amyloidosis who also
presence showed presence of serum monoclonal proteins—in the
of CA,
N = 120
absence of AL amyloidosis. Hence, monoclonal protein
EMB nega- 1 (2) 39 (98) 98 (87–100)
findings in these six patients with a median age of 73 years
tive for specific were assessed as monoclonal gammopathy of undeter-
CA, N = 40 mined significance (MGUS). The combined finding of
a “characteristic LGE pattern indicative of CA” (upon
CMR) AND negative monoclonal protein studies was 98%
of CA (Table 3). Accordingly, the positive predictive value specific for the diagnosis of cardiac ATTR amyloidosis.
(PPV) of LGE-CMR for the diagnosis of CA was 99% (95%- Moreover, such a combined finding showed a PPV of 99%
CI: 94–100%) whereas the respective negative predictive
value (NPV) was 87% (95%-CI: 75–93%).

Table 4 Combination of CMR results with monoclonal protein studies (MPS)

Characteristic LGE indicative of LGE pattern NOT indicative of Sensitivity and specificity PPV and NPV (CI), %
CA + negative MPS CA and/or positive MPS (CI), %

EMB-proven pres- 83 9 90 (82–95) sensitive NPV = 88 (80–93)

ence of ATTR-CA,
N = 92
EMB negative for 1 67 99 (92–100) specific PPV = 99 (92–100)
ATTR-CA, N = 68

13
Clinical Research in Cardiology (2021) 110:555–568 563

(95%-CI 92–100%) for the diagnosis of cardiac ATTR of CA, e.g. the “native T1 versus ECV paradox” in CA
amyloidosis when compared to biopsy findings (Table 4). that was recently addressed in detail elsewhere [21]. In a
recent meta-analysis, a total of 18 diagnostic (N = 2015)
and 13 prognostic CMR studies (N = 1483) using native
Discussion T1, ECV or LGE to diagnose and prognosticate CA were
included for analysis [22]. According to this meta-anal-
To the best of our knowledge, this is the largest multi- ysis, the parameter ECV showed a significantly higher
centre German study that evaluated the diagnostic value diagnostic odds ratio for CA than conventional LGE-
of CMR in comparison to EMB-proven CA. The results assessment. However, there was no significant difference
of our present study are based on real-world clinical data between LGE-assessment and native T1 for sensitivity,
from four German centres that are highly experienced in specificity and diagnostic odds ratio—regarding the diag-
both (a) conducting CMR studies (with a high volume of nosis of CA. Our present study does not allow to safely
>2.000 CMR studies per year and centre) and (b) perform- compare the diagnostic value of T1-mapping and/or ECV
ing EMBs (with a high volume of >100 EMB procedures measurement in comparison to LGE-imaging since map-
per year and centre) for workup of suspected cardiomyopa- ping was not performed in all patients. In this context, it
thies of unknown origin. This real-world German experi- needs to be considered that cut-off values for native T1
ence clearly illustrates that (a) non-invasive CMR allows or ECV derived from ROC analyses (in a specific group
to diagnose the presence of CA with a high sensitivity of of study patients) for ruling in or out the presence of CA
95% and an even higher specificity of 98% and that (b) the are—amongst others—determined by native T1- and
combined finding of a positive CMR study (indicative of ECV-values of the respective control group. A different
CA) and negative monoclonal protein studies is not only composition of the control group (e.g. higher percentage
highly specific for the diagnosis of cardiac ATTR amyloi- of HCM patients with extensive myocardial fibrosis vs.
dosis (specificity of 98%) but also highly trustable with a lower percentage of healthy controls without fibrosis) will
PPV of 99% (for the diagnosis of cardiac ATTR amyloi- result in different cut-off values. Hence, comparisons of
dosis)—proven by biopsy results. diagnostic CMR parameters based on ROC analyses (using
Already in 2008, the Stuttgart group assessed the diag- cut-off values) need to be considered carefully—with a
nostic value of LGE-CMR for the diagnosis of CA in a special attention to the control group of the underlying
rather small study group of 33 patients [6]: cardiac amy- study—and single-centre results cannot be transferred to
loidosis was detected by EMB in 15 out of 33 patients other centres by default.
who were studied more than a decade ago (on 1.5-T Mag- In the last years, innumerable studies addressing non-
netom Sonata at that time). Using the characteristic LGE- invasive diagnosis of CA were published by colleagues
pattern indicative of CA as a diagnostic criterion, they from the National Amyloidosis Centre in London/UK—
obtained a sensitivity of 80% and a specificity of 94% for deserving the designation the London/UK experience [7,
the diagnosis of CA—at that time. Obviously, both clini- 12, 17, 23]. Their study results moved the diagnostic field
cal experience in assessing LGE-CMR images as well as forward and substantially contributed to current recommen-
CMR-techniques have tremendously improved since that dations regarding the diagnostic approach in suspected CA
time allowing (amongst others) more experienced and [12, 24, 25]. Due to the efforts of these colleagues, bone
more accurate assessment of LGE-images today [17–19]. scintigraphy was established as a substantial non-invasive
Hence, the results of the present multi-centre study nicely method to diagnose CA [24–27]. However, this London/
illustrate this overall progress and clearly demonstrate that UK experience needs to be considered with some caution
both sensitivity and specificity of LGE-based diagnosis of and a non-reflected transfer of the UK-based results to other
CA have improved. EU countries should be avoided. As outlined by these col-
Obviously, today cardiac workup of patients with sus- leagues themselves, the patients presenting to this central
pected CA should be based on multi-parametric CMR UK centre are not “unselected”, but rather referred to this
including T1-mapping and ECV measurement—and not centre of experience with suspected CA. Consequently, the
limited to LGE-CMR—since several studies have shown cohort of patients studied at this centre—and in particular,
a superior diagnostic value of mapping-based approaches the “control group” used in their studies—does not reflect an
compared to conventional LGE-imaging [7, 8]. In princi- “unselected” cardiology population of patients. Obviously,
ple, native T1-mapping and ECV measurement are highly the same is true for our present study due to our methodolog-
suitable tools to detect (and quantify) even subtle amyloid ical approach in selecting patients. Importantly, the patients
deposits in the extracellular space of the myocardium [5, of this study were initially referred as “outpatients” to our
20]. However, there are still some puzzling data regard- specialised cardiomyopathy centres (mostly by resident
ing the comparison of native T1- and ECV-values in case cardiologists due to suspected cardiomyopathy for further

13
564 Clinical Research in Cardiology (2021) 110:555–568

diagnostic workup). Hence, we neither included, e.g. decom- of ATTR-CA by additional negative monoclonal protein
pensated inpatients presenting directly to our emergency or studies—similar to bone scintigraphy. Obviously, in case
intensive care units who also underwent CMR and/or biopsy of such clear CMR findings there is no need for additional
workup during their further hospitalisation, nor did we study diagnostic methods (such as bone scintigraphy)—and the
an unselected cardiology population. However, our control diagnostic algorithm suggested in some recommendation
group comprising only patients with LV hypertrophy should papers needs to be carefully revised. Based on the present
be more appropriate regarding validation of imaging meth- German experience, we suggest the diagnostic algorithm
ods for the diagnosis of CA. illustrated in Fig. 4.
To appropriately assess the value of the present CMR Finally, one may argue that additional EMBs are not
results in comparison to previous bone scintigraphy required in those patients with the combined finding of
results (e.g. the London/UK experience), a careful look a “positive” CMR study indicative of CA and “negative”
on the data of Gillmore et al. is required [12]: in a first monoclonal protein studies—similar to the scintigraphic
step, Gillmore et al. analysed the data of 374 patients approach of Gillmore et al. [12]. However, since our knowl-
with EMB and defined a “positive” bone scintigraphy edge on the underlying pathophysiology of amyloid depo-
indicative of CA as cardiac tracer uptake of either grade sition in the human myocardium is still very limited, and
1, 2 or 3 (vs. grade 0 as “normal” finding). This approach since CA is not only characterised by extracellular accu-
resulted in a sensitivity of 88% and a specificity of 87% mulation of amyloid fibrils but also by myocardial oedema,
for bone scintigraphy to detect CA (independent of sub- inflammation, and myocyte hypertrophy (with variable
type). In a second step, the authors focused on those degrees of each pathophysiology potentially resulting in
patients with ATTR-CA and received a very high sensitiv- differing effects on native T1, ECV and LGE-pattern), we
ity of >99% for bone scintigraphy to detect ATTR-CA— still need the histopathological information from EMBs in
but a rather low specificity of 68%. In a third step, the order to better understand our non-invasive imaging find-
authors changed their “positive” definition of bone scin- ings. This issue will be even far more important consider-
tigraphy and presumed that a cardiac tracer uptake of only ing upcoming specific therapies to treat ATTR-CA: we will
grade 2 or 3 was indicative of ATTR-CA (vs. grade 0 or 1 need the EMB information in order to better understand
defined as “normal/negative” findings). Such a different both (a) the therapeutic effect of the respective medication/
assumption resulted in a re-grouping of 42 (out of 374 therapy and (b) the change in non-invasive cardiac imag-
patients = 11%) from the positive group to the negative ing parameters that will be used for disease monitoring.
group, and in a sensitivity of 91% and a specificity of 87% Hence, considering the low risk of EMB complication even
for bone scintigraphy to detect ATTR-CA. In compari- in patients with CA—as confirmed in the present study, we
son, in our present CMR study, the presence of CA was suggest to continue to biopsy even patients with the com-
diagnosed based on LGE-CMR with a high sensitivity bined finding of a “positive” CMR study indicative of CA
of 95% and an even higher specificity of 98%—without and “negative” monoclonal protein studies. For the remain-
re-grouping of study patients by changing assumptions. ing patients, there is no doubt that EMB still represents the
Furthermore, Gillmore et al. analysed the specificity gold standard for workup of non-ischemic, unexplained
and PPV of a “positive” bone scintigraphy indicative of cardiomyopathy [9, 10, 16].
CA (defined as cardiac tracer uptake of either grade 2 Obviously, the EMB complication rate of 3.13% in the
or 3 vs. grade 0 or 1 as “normal/negative”) for ATTR- present study is somewhat higher compared to previous
CA when combined with the absence of monoclonal pro- studies suggesting a major complication rate of 0.64%
teins in N = 374 patients with EMB [12]: they obtained a for LV-EMB and of 0.82% for RV-EMB—proving that
specificity as well as PPV of 100% for the diagnosis of EMB is a safe procedure. Moreover, there are reports that
ATTR-CA. Based on this finding, they established their mentioned a myocardial perforation risk (= major com-
approach and diagnostic algorithm of bone scintigraphy- plication) in CA patients in up to 17.1% in case of RV-
based non-biopsy diagnosis of ATTR-CA that entered EMB and up to 6.6% in case of LV-EMB (Kristen et al.
some recommendation papers [24–26]—and is questioned Am J Hematology 2007;82:327–333)—and suggested a
by some interesting reports [28, 29]. In contrast, in our more “fragile” myocardium in case of CA with a higher
present study, the combined finding of a “positive” CMR risk of myocardial perforation in particular in case of
study indicative of CA and “negative” monoclonal protein RV-EMB. Obviously, the complication rate of EMB in
studies resulted in a similar specificity of 98% and PPV of the present study was tremendously lower as reported in
99% for the diagnosis of ATTR-CA (compared to biopsy the aforementioned study of Kristen et al. In the present
findings). Hence, our present results clearly suggest that study, myocardial perforation occurred in four patients all
non-invasive CMR allows both (a) to safely diagnose the belonging to the CA group—supporting the notion of a
presence of CA and (b) to further verify the presence “fragile” myocardium in case of advanced CA. However,

13
Clinical Research in Cardiology (2021) 110:555–568 565

Fig. 4 Schematic diagram representing the suggested diagnostic pathway for workup of cardiac amyloidosis (CA)

these complications occurred when a “stiff” bioptome Limitations

was used whereas no complications were observed when
using more smooth/elastic bioptomes. In our opinion, Similar to other previous studies, our study group does not
the risk of perforation in case of EMB is driven by (a) reflect an “unselected” cardiology population of patients
the properties of the bioptome, (b) the experience of the and comprised only those suffering from symptoms
interventionalist and (c) the composition of the myocar- of heart failure in the presence of left ventricular (LV)
dium—with a potentially more “fragile” myocardium in hypertrophy that could not be explained by abnormal load-
case of CA. Therefore, EMB procedures in patients with ing conditions. Obviously, we cannot exclude a potential
suspected CA are quite safe when they are performed by selection bias by selected patient referral to our special-
experienced interventionalists and preferably taken from ised cardiomyopathy centres. However, we believe that our
the LV using smooth bioptomes. approach is appropriate for evaluating the diagnostic yield
Finally, the issue of balancing diagnostic yield vs. proce- of CMR for the workup of CA (based on real-world clini-
dure-associated risk and costs is highly important regarding cal data), since our CONTROL group did not comprise
clinical decision-making in daily routine. However, a discus- healthy patients or patients without relevant structural
sion on this issue would go far beyond the scope of the pre- abnormalities—but rather severely diseased patients with
sent work and should be performed by expert panels based other cardiac diseases presenting to four German special-
on more comprehensive data regarding all of these issues. ised cardiomyopathy centres. Moreover, T1-mapping and
Nevertheless, we believe that the “diagnostic” data presented ECV data were not available in all patients—limiting the
in our manuscript will help such expert panels aiming at value of comparative analyses of LGE vs. native T1 and/
balancing diagnostic yield, procedure safety, method avail- or ECV. However, LGE-pattern per se showed a convinc-
ability, costs and practical consequences. ing diagnostic yield and T1-mapping-based approaches

13
566 Clinical Research in Cardiology (2021) 110:555–568

will obviously “strengthen”—but not “worsen”—this yield References

and particularly allow a monitoring of myocardial amyloid
load over time. 1. Connors LH, Sam F, Skinner M, Salinaro F, Sun F, Ruberg FL,
Berk JL, Seldin DC (2016) Heart failure resulting from age-related
cardiac amyloid disease associated with wild-type transthyretin: a
prospective. Observ Cohort Study Circ 133(3):282–290
Conclusion 2. Merlini G, Bellotti V (2003) Molecular mechanisms of amyloido-
sis. N Engl J Med 349(6):583–596
Non-invasive CMR shows an excellent diagnostic accuracy 3. Wechalekar AD, Gillmore JD, Hawkins PN (2016) Systemic amy-
loidosis. Lancet 387(10038):2641–2654
and yield regarding CA. When combined with monoclonal 4. Gillmore JD, Damy T, Fontana M, Hutchinson M, Lachmann HJ,
protein studies, CMR can differentiate ATTR from AL with Martinez-Naharro A, Quarta CC, Rezk T, Whelan CJ, Gonzalez-
high accuracy and predictive value. However, invasive EMB Lopez E, Lane T, Gilbertson JA, Rowczenio D, Petrie A, Hawkins
remains a safe invasive gold standard and allows to differ- PN (2018) A new staging system for cardiac transthyretin amyloi-
dosis. Eur Heart J 39(30):2799–2806
entiate CA from other cardiomyopathies that can also cause 5. Messroghli DR, Moon JC, Ferreira VM, Grosse-Wortmann L, He
LV hypertrophy. T, Kellman P, Mascherbauer J, Nezafat R, Salerno M, Schelbert
EB, Taylor AJ, Thompson R, Ugander M, van Heeswijk RB, Frie-
Author contributions GC and MB participated in the CMR exams, drich MG (2017) Clinical recommendations for cardiovascular
carried out the data and statistical analysis, and wrote some paragraphs magnetic resonance mapping of T1, T2, T2* and extracellular
of the draft version of the manuscript. AE, GT, JV and ZS participated volume: a consensus statement by the Society for Cardiovascular
in the CMR exams and in the analysis of the CMR data. CS, BP, HR, Magnetic Resonance (SCMR) endorsed by the European Asso-
HM, and SK participated in data collection and analyses, and criti- ciation for Cardiovascular Imaging (EACVI). J Cardiovasc Magn
cally reviewed the manuscript. AY supervised the study, wrote major Reson 19(1):75
paragraphs of the manuscript, critically reviewed the final manuscript 6. Vogelsberg H, Mahrholdt H, Deluigi CC, Yilmaz A, Kispert EM,
and drafted the manuscript. All authors read and approved the final Greulich S, Klingel K, Kandolf R, Sechtem U (2008) Cardiovas-
manuscript. cular magnetic resonance in clinically suspected cardiac amyloi-
dosis: noninvasive imaging compared to endomyocardial biopsy.
Funding Open Access funding enabled and organized by Projekt J Am Coll Cardiol 51(10):1022–1030
DEAL. 7. Fontana M, Banypersad SM, Treibel TA, Maestrini V, Sado DM,
White SK, Pica S, Castelletti S, Piechnik SK, Robson MD, Gil-
bertson JA, Rowczenio D, Hutt DF, Lachmann HJ, Wechalekar
Availability of data and materials The datasets used and/or analysed AD, Whelan CJ, Gillmore JD, Hawkins PN, Moon JC (2014)
during the current study are available from the corresponding author Native T1 mapping in transthyretin amyloidosis. JACC Cardiovasc
on reasonable request. Imaging 7(2):157–165
8. Martinez-Naharro A, Kotecha T, Norrington K, Boldrini M,
Compliance with ethical standards Rezk T, Quarta C, Treibel TA, Whelan CJ, Knight DS, Kellman
P, Ruberg FL, Gillmore JD, Moon JC, Hawkins PN, Fontana M
Conflict of interest None. (2019) Native T1 and extracellular volume in transthyretin amy-
loidosis. JACC Cardiovasc Imaging 12(5):810–819. https://doi.
Ethics approval and consent to participate The study protocol complies org/10.1016/j.jcmg.2018.02.006
with the Declaration of Helsinki. Ethics approval was obtained from 9. Ardehali H, Qasim A, Cappola T, Howard D, Hruban R, Hare JM,
local authorities and written informed consent was obtained from the Baughman KL, Kasper EK (2004) Endomyocardial biopsy plays
participants. a role in diagnosing patients with unexplained cardiomyopathy.
Am Heart J 147(5):919–923
Consent for publication N/A. 10. Cooper LT, Baughman KL, Feldman AM, Frustaci A, Jessup M,
Kuhl U, Levine GN, Narula J, Starling RC, Towbin J, Virmani R
(2007) The role of endomyocardial biopsy in the management of
Open Access This article is licensed under a Creative Commons Attri- cardiovascular disease: a scientific statement from the American
bution 4.0 International License, which permits use, sharing, adapta- Heart Association, the American College of Cardiology, and the
tion, distribution and reproduction in any medium or format, as long European Society of Cardiology Endorsed by the Heart Failure
as you give appropriate credit to the original author(s) and the source, Society of America and the Heart Failure Association of the Euro-
provide a link to the Creative Commons licence, and indicate if changes pean Society of Cardiology. Eur Heart J 28(24):3076–3093
were made. The images or other third party material in this article are 11. Yilmaz A, Kindermann I, Kindermann M, Mahfoud F, Ukena C,
included in the article’s Creative Commons licence, unless indicated Athanasiadis A, Hill S, Mahrholdt H, Voehringer M, Schieber M,
otherwise in a credit line to the material. If material is not included in Klingel K, Kandolf R, Bohm M, Sechtem U (2010) Comparative
the article’s Creative Commons licence and your intended use is not evaluation of left and right ventricular endomyocardial biopsy:
permitted by statutory regulation or exceeds the permitted use, you will differences in complication rate and diagnostic performance. Cir-
need to obtain permission directly from the copyright holder. To view a culation 122(9):900–909
copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. 12. Gillmore JD, Maurer MS, Falk RH, Merlini G, Damy T, Dis-
penzieri A, Wechalekar AD, Berk JL, Quarta CC, Grogan M,

13
Clinical Research in Cardiology (2021) 110:555–568 567

Lachmann HJ, Bokhari S, Castano A, Dorbala S, Johnson GB, autoimmune myocarditis in a patient with systemic lupus erythe-
Glaudemans AW, Rezk T, Fontana M, Palladini G, Milani P, matosus. Clin Res Cardiol 106(7):560–563
Guidalotti PL, Flatman K, Lane T, Vonberg FW, Whelan CJ, 21. Yilmaz A (2018) The "native T1 versus extracellular volume frac-
Moon JC, Ruberg FL, Miller EJ, Hutt DF, Hazenberg BP, Rapezzi tion paradox" in cardiac amyloidosis: answer to the million-dollar
C, Hawkins PN (2016) Nonbiopsy diagnosis of cardiac transthyre- question? JACC Cardiovasc Imaging
tin amyloidosis. Circulation 133(24):2404–2412 22. Pan JA, Kerwin MJ, Salerno M (2020) Native T1 mapping, extra-
13. Dungu JN, Valencia O, Pinney JH, Gibbs SD, Rowczenio D, Gil- cellular volume mapping, and late gadolinium enhancement in
bertson JA, Lachmann HJ, Wechalekar A, Gillmore JD, Whelan cardiac amyloidosis: a meta-analysis. JACC Cardiovasc Imaging
CJ, Hawkins PN, Anderson LJ (2014) CMR-based differentiation 13(6):1299–1310
of AL and ATTR cardiac amyloidosis. JACC Cardiovasc Imaging 23. Baggiano A, Boldrini M, Martinez-Naharro A, Kotecha T, Petrie
7(2):133–142 A, Rezk T, Gritti M, Quarta C, Knight DS, Wechalekar AD, Lach-
14. Tschope C, Kherad B, Schultheiss HP (2015) How to perform an mann HJ, Perlini S, Pontone G, Moon JC, Kellman P, Gillmore
endomyocardial biopsy? Turk Kardiyol Dern Ars 43(6):572–575 JD, Hawkins PN, Fontana M (2020) Noncontrast magnetic reso-
15. Yilmaz A, Klingel K, Kandolf R, Sechtem U (2012) Endomyocar- nance for the diagnosis of cardiac amyloidosis. JACC Cardiovasc
dial biopsy. The PCR-EAPCI textbook, 1st edn, Europa Edition. Imaging 13(1 Pt 1):69–80
http://www.pcronline.com/eurointervention/textbook/pcr-textb 24. Dorbala S, Vangala D, Semer J, Strader C, Bruyere JR Jr, Di Carli
ook/ MF, Moore SC, Falk RH (2014) Imaging cardiac amyloidosis: a
16. Baccouche H, Mahrholdt H, Meinhardt G, Merher R, Voehringer pilot study using (1)(8)F-florbetapir positron emission tomogra-
M, Hill S, Klingel K, Kandolf R, Sechtem U, Yilmaz A (2009) phy. Eur J Nucl Med Mol Imaging 41(9):1652–1662
Diagnostic synergy of non-invasive cardiovascular magnetic 25. Dorbala S, Vangala D, Bruyere J Jr, Quarta C, Kruger J, Padera
resonance and invasive endomyocardial biopsy in troponin- R, Foster C, Hanley M, Di Carli MF, Falk R (2014) Coronary
positive patients without coronary artery disease. Eur Heart J microvascular dysfunction is related to abnormalities in myocar-
30(23):2869–2879 dial structure and function in cardiac amyloidosis. JACC Heart
17. Fontana M, Pica S, Reant P, Abdel-Gadir A, Treibel TA, Ban- Fail 2(4):358–367
ypersad SM, Maestrini V, Barcella W, Rosmini S, Bulluck H, 26. Maurer MS, Bokhari S, Damy T, Dorbala S, Drachman BM,
Sayed RH, Patel K, Mamhood S, Bucciarelli-Ducci C, Whelan Fontana M, Grogan M, Kristen AV, Lousada I, Nativi-Nicolau
CJ, Herrey AS, Lachmann HJ, Wechalekar AD, Manisty CH, J, Cristina QC, Rapezzi C, Ruberg FL, Witteles R, Merlini G
Schelbert EB, Kellman P, Gillmore JD, Hawkins PN, Moon JC (2019) Expert consensus recommendations for the suspicion and
(2015) Prognostic value of late gadolinium enhancement cardio- diagnosis of transthyretin cardiac amyloidosis. Circ Heart Fail
vascular magnetic resonance in cardiac amyloidosis. Circulation 12(9):e006075
132(16):1570–1579 27. Yilmaz A, Bauersachs J, Kindermann I, Klingel K, Knebel F,
18. Francis R, Kellman P, Kotecha T, Baggiano A, Norrington K, Meder B, Morbach C, Nagel E, Schulze-Bahr E, Aus Dem Siepen
Martinez-Naharro A, Nordin S, Knight DS, Rakhit RD, Lockie T, F, Frey N (2019) Position statement of the German Society of
Hawkins PN, Moon JC, Hausenloy DJ, Xue H, Hansen MS, Fon- Cardiology (DGK) regarding “diagnosis and therapy of cardiac
tana M (2017) Prospective comparison of novel dark blood late amyloidosis.” Der Kardiologe 3:264–291
gadolinium enhancement with conventional bright blood imaging 28. Layoun ME, Desmarais J, Heitner SB, Masri A (2020) Hot hearts
for the detection of scar. J Cardiovasc Magn Reson 19(1):91 on bone scintigraphy are not all amyloidosis: hydroxychloroquine-
19. Kellman P, Xue H, Olivieri LJ, Cross RR, Grant EK, Fontana induced restrictive cardiomyopathy. Eur Heart J 41(25):2414
M, Ugander M, Moon JC, Hansen MS (2016) Dark blood late 29. Pilebro B, Suhr OB, Naslund U, Westermark P, Lindqvist P,
enhancement imaging. J Cardiovasc Magn Reson 18(1):77 Sundstrom T (2016) (99m)Tc-DPD uptake reflects amyloid fibril
20. Bietenbeck M, Florian A, Shomanova Z, Klingel K, Yilmaz A composition in hereditary transthyretin amyloidosis. Ups J Med
(2017) Novel CMR techniques enable detection of even mild Sci 121(1):17–24

Affiliations

Grigorios Chatzantonis1 · Michael Bietenbeck1 · Ahmed Elsanhoury2 · Carsten Tschöpe2,3 · Burkert Pieske4 ·

Gloria Tauscher5 · Julia Vietheer6,7 · Zornitsa Shomanova1 · Heiko Mahrholdt5 · Andreas Rolf6,7 · Sebastian Kelle4 ·
Ali Yilmaz1

Grigorios Chatzantonis Burkert Pieske

grigorios.chatzantonis@ukmuenster.de pieske@dhzb.de
Michael Bietenbeck Gloria Tauscher
michael.bietenbeck@ukmuenster.de gloria.tauscher@rbk.de
Ahmed Elsanhoury Julia Vietheer
ahmed.elsanhoury@charite.de j.vietheer@kerckhoff-klinik.de
Carsten Tschöpe Zornitsa Shomanova
carsten.tschoepe@charite.de zornitsa.shomanova@ukmuenster.de

13
568 Clinical Research in Cardiology (2021) 110:555–568

3
Heiko Mahrholdt DZHK (German Centre for Cardiovascular Research),
heiko.mahrholdt@rbk.de partner site Berlin, Charite,  Berlin, Germany
4
Andreas Rolf Department of Internal Medicine/Cardiology, Deutsches
a.rolf@kerckhoff-klinik.de Herzzentrum Berlin, Berlin, Germany
5
Sebastian Kelle Department of Cardiology, Robert-Bosch-Medical Centre,
kelle@dhzb.de Stuttgart, Germany
6
1 Department of Cardiology, Kerckhoff Hospital, University
Department of Cardiology I, University Hospital Münster,
Giessen, Bad Nauheim, Germany
Albert-Schweitzer-Campus 1, Building A1, 48149 Münster,
7
Germany DZHK (German Centre for Cardiovascular Research),
2 partner site Rhine Main, Frankfurt, Germany
Department of Cardiology, Centre for Regenerative
Therapies (BCRT), Campus Virchow and Berlin Institute
of Health (BIH), Berlin, Charite, Berlin, Germany

13