Original Article doi: 10.1111/joim.12324
Measurement of myocardial amyloid deposition in systemic amyloidosis: insights from cardiovascular magnetic resonance imaging A. Barison1,2, G. D. Aquaro1, N. R. Pugliese1, F. Cappelli3, S. Chiappino1, G. Vergaro1, G. Mirizzi1, G. Todiere1, C. Passino1,2, P. G. Masci1, F. Perfetto3 & M. Emdin1,2 From the1Fondazione Toscana Gabriele Monasterio; 2Scuola Superiore Sant’Anna, Pisa; and 3Regional Amyloid Centre, Azienda Ospedaliero Universitaria Careggi, Florence, Italy
Abstract. Barison A, Aquaro GD, Pugliese NR, Cappelli F, Chiappino S, Vergaro G, Mirizzi G, Todiere G, Passino C, Masci PG, Perfetto F, Emdin M (Fondazione Toscana Gabriele Monasterio, Pisa; Scuola Superiore Sant’Anna, Pisa; Azienda Ospedaliero Universitaria Careggi, Florence; Italy). Measurement of myocardial amyloid deposition in systemic amyloidosis: insights from cardiovascular magnetic resonance imaging. J Intern Med 2014; doi: 10.1111/joim.12324. Background. Cardiac involvement in systemic amyloidosis is caused by the extracellular deposition of misfolded proteins, mainly immunoglobulin light chains (AL) or transthyretin (ATTR), and may be detected by cardiovascular magnetic resonance (CMR). The aim of this study was to measure myocardial extracellular volume (ECV) in amyloid patients with a novel T1 mapping CMR technique and to determine the correlation between ECV and disease severity. Methods. Thirty-six patients with biopsy-proven systemic amyloidosis (mean age 70 9 years, 31 men, 30 with AL and six with ATTR amyloidosis) and seven patients with possible amyloidosis (mean age 64 10 years, six men) underwent comprehensive clinical and CMR assessment, with ECV estimation from pre- and postcontrast T1 mapping. Thirty healthy subjects (mean age 39 17 years, 21 men) served as the control group.
Introduction Systemic amyloidosis is an uncommon multisystem disease caused by the extracellular deposition of misfolded protein in various tissues and organs [1]. Cardiac involvement is a leading cause of
Results. Amyloid patients presented with left ventricular (LV) concentric hypertrophy with impaired biventricular systolic function. Cardiac ECV was higher in amyloid patients (definite amyloidosis, 0.43 0.12; possible amyloidosis, 0.34 0.11) than in control subjects (0.26 0.04, P < 0.05); even in amyloid patients without late gadolinium enhancement (0.35 0.10), ECV was significantly higher than in the control group (P < 0.01). A cutoff value of myocardial ECV >0.316, corresponding to the 95th percentile in normal subjects, showed a sensitivity of 79% and specificity of 97% for discriminating amyloid patients from control subjects (area under the curve of 0.884). Myocardial ECV was significantly correlated with LV ejection fraction (R2 = 0.16), LV mean wall thickness (R2 = 0.41), LV diastolic function (R2 = 0.21), right ventricular ejection fraction (R2 = 0.13), N-terminal fragment of the pro-brain natriuretic peptides (R2 = 0.23) and cardiac troponin (R2 = 0.33). Conclusion. Myocardial ECV was increased in amyloid patients and correlated with disease severity. Thus, measurement of myocardial ECV represents a potential noninvasive index of amyloid burden for use in early diagnosis and disease monitoring. Keywords: amyloidosis, biomarker, cardiac hypertrophy, cardiology, radiology.
morbidity and mortality, especially in primary light chain (AL) amyloidosis and in both wild-type and hereditary transthyretin (ATTR) amyloidosis. Endomyocardial biopsy is considered the gold standard for demonstrating cardiac amyloidosis (CA), but a combination of noncardiac biopsy with
ª 2014 The Association for the Publication of the Journal of Internal Medicine
1
A. Barison et al.
electrocardiography (ECG) [low-voltage QRS potentials], echocardiography [concentric left ventricular (LV) hypertrophy] and biomarkers [increased levels of N-terminal fragment of the pro-brain natriuretic peptide (NT-proBNP)] is commonly used in clinical practice to diagnose CA [2]. The role of novel techniques, such as scintigraphy and cardiac magnetic resonance (CMR), is under investigation [3, 4]. CMR provides functional and morphological information in patients with CA and significantly improves tissue characterization. The appearance of global, subendocardial-to-transmural late gadolinium enhancement (LGE) is highly characteristic of CA (despite suboptimal sensitivity as it is present in only two-thirds of patients with CA[3, 5, 6]) and correlates with prognosis [7]. Recently, equilibrium contrast CMR has demonstrated much higher extracellular myocardial volume in CA than in any other disease investigated with this technique [8], and accurate measurements of the expanded interstitium in amyloidosis might be useful for serial quantification of cardiac amyloid burden. The aim of this study was to investigate gadolinium contrast kinetics and interstitial expansion in patients with CA using a CMR T1 mapping technique and to assess the correlation between these results and clinical, biohumoral, morphological and conventional LGE features. For comparison, we included in the study not only healthy control subjects, but also a group of patients at risk of CA due to hypergammaglobulinaemia or TTR mutation without clinical evidence of amyloidosis.
Magnetic resonance assessment of cardiac amyloidosis
myocardial infarction, coronary revascularization, myocarditis, nonischaemic cardiomyopathy, congenital heart disease or primary moderate-tosevere valvular disease were excluded from the study. The study was approved by our institutional ethics committee, and written informed consent was obtained for each patient. According to current criteria [2], the diagnosis of systemic amyloidosis in the 36 patients was based on a positive biopsy in any organ (n = 30 AL and n = 6 ATTR amyloidosis), whilst the seven patients with possible systemic amyloidosis had a negative soft-tissue biopsy but possible cardiac involvement due to long-standing hypergammaglobulinaemia (n = 6) or TTR mutation (n = 1). Endomyocardial biopsy had been performed in only one patient with AL amyloidosis. Moreover, according to the same criteria [2], the heart was considered to be involved when the mean LV wall thickness (septum and posterior wall) was >12 mm at echocardiography, in the absence or presence of hypertension or other potential causes of LV hypertrophy (definite or probable CA, respectively). Troponin I was assessed using an immunochemiluminescence assay (ADVIA, Siemens, Erlangen, Germany). NTproBNP was also assessed using an immunochemiluminescence assay (Elecsys, Roche Diagnostics, Monza, Italy). Total immunoglobulin light chain concentrations were measured using agarose gel immunoelectrophoresis (SeaKem ME agarose, Cambrex Bio Science, Rockland, USA), and free light chain concentration was measured using immunonephelometry (BN II analyser, Dade Behring, Marburg, Germany).
Materials and methods Patient selection
CMR protocol
We prospectively recruited 36 patients with biopsyproven systemic amyloidosis (mean age 70 9 years, 31 men), seven patients with possible amyloidosis (mean age 64 10 years, six men) and 30 healthy control subjects (mean age 39 17 years, 21 men) (P < 0.001 for age difference between control group and both patient groups; nonsignificant difference for gender). Patients with any contraindication (metal devices, severe renal failure or claustrophobia) or unable to undergo CMR imaging (poor breath-holding capacity or arrhythmias with extremely variable RR intervals) were excluded from the study. All patients underwent clinical, echocardiographic, ECG and biohumoral evaluation, and a complete CMR scan. Patients with a previous history of
Study participants were examined using a 1.5-T CMR unit (CVi, GE Healthcare, Milwaukee, WI, USA) with dedicated cardiac software, phased-array surface receiver coil and vectocardiogram triggering. Ventricular function was assessed by breath-hold steady-state free-precession cine imaging in cardiac short axis and vertical and horizontal long axis. In cardiac short axis, the ventricles were completely encompassed by contiguous slices (8 mm thick); sequence parameters were as follows: field-of-view: 380–400 mm, repetition/echo time: 3.2/1.6 ms, flip angle: 60°, matrix: 224 9 192, phases: 30. For determination of the T1 value of the myocardium and blood cavity, a single mid-ventricular short-axis modified-cine inversion-recovery (MCine-IR) sequence was performed before and at fixed time
2
ª 2014 The Association for the Publication of the Journal of Internal Medicine Journal of Internal Medicine
A. Barison et al.
intervals (5, 10, 15 and 20 min) after a bolus of contrast agent (0.2 mmol kg 1 Gadodiamide, OMINISCAN, GE Healthcare, Waukesha, WI, USA). The MCine-IR sequence consisted of a nonselective adiabatic inversion pulse applied at the R-wave of the ECG and followed by a cine-segmented gradientecho acquisition extended to the following six RR intervals for precontrast imaging or four RR intervals for postcontrast imaging [9]; sequence parameters were as follows: field-of-view: 380–400 mm, repetition/echo time: 6/2.8 ms, flip angle: 8°, matrix: 224 9 192, phases: 40. LGE imaging was performed between 10 and 20 min after contrast agent administration using a segmented T1weighted gradient-echo inversion-recovery pulse sequence. In short-axis orientation, the LV was completely encompassed by contiguous slices (8 mm thick). Images were also acquired in vertical and horizontal long-axis views. Inversion time was individually adapted to suppress the signal of normal remote myocardium (220–300 ms). Sequence parameters were as follows: field-of-view: 380– 400 mm, slice thickness: 8 mm, repetition/echo time: 4.6/1.3 ms, flip angle: 20°, matrix: 256 9 192. CMR image analysis All CMR studies were analysed off-line using a workstation (Advantage Workstation, GE Healthcare) with dedicated software (MASS 6.1, Medis, Leiden, The Netherlands) by one experienced operator blinded to the clinical data. Using the stack of short-axis cine images, LV volumes, mass and global function were determined. LV volumes and mass were normalized to body surface area. LGE was assessed as (i) present or absent, (ii) focal (involving three or fewer adjacent LV segments) or diffuse (more than three LV segments), (iii) midwall, subendocardial, subepicardial or transmural and (iv) whether or not there was involvement of the basal, mid and/or apical segments of the LV. For T1 mapping analysis, a region of interest was drawn in the interventricular septum, in the LV blood pool and in the skeletal muscle (pectoralis or latissimus dorsi) for each frame of the T1 mapping series in a single mid-ventricular short-axis slice; when analysing the myocardium, care was taken to avoid the blood pool. An exponential fitting model was used to obtain the T1 from blood, myocardium and skeletal muscle with customwritten software (HIPPO-SW; Fondazione Toscana Gabriele Monasterio, Pisa, Italy). Of the 60 and 40 frames acquired for pre- and postcontrast analy-
Magnetic resonance assessment of cardiac amyloidosis
sis, respectively, only those with artefacts were excluded maintain the fitting error below 10% (usually
Measurement of myocardial amyloid deposition in systemic amyloidosis: insights from cardiovascular magnetic resonance imaging A. Barison1,2, G. D. Aquaro1, N. R. Pugliese1, F. Cappelli3, S. Chiappino1, G. Vergaro1, G. Mirizzi1, G. Todiere1, C. Passino1,2, P. G. Masci1, F. Perfetto3 & M. Emdin1,2 From the1Fondazione Toscana Gabriele Monasterio; 2Scuola Superiore Sant’Anna, Pisa; and 3Regional Amyloid Centre, Azienda Ospedaliero Universitaria Careggi, Florence, Italy
Abstract. Barison A, Aquaro GD, Pugliese NR, Cappelli F, Chiappino S, Vergaro G, Mirizzi G, Todiere G, Passino C, Masci PG, Perfetto F, Emdin M (Fondazione Toscana Gabriele Monasterio, Pisa; Scuola Superiore Sant’Anna, Pisa; Azienda Ospedaliero Universitaria Careggi, Florence; Italy). Measurement of myocardial amyloid deposition in systemic amyloidosis: insights from cardiovascular magnetic resonance imaging. J Intern Med 2014; doi: 10.1111/joim.12324. Background. Cardiac involvement in systemic amyloidosis is caused by the extracellular deposition of misfolded proteins, mainly immunoglobulin light chains (AL) or transthyretin (ATTR), and may be detected by cardiovascular magnetic resonance (CMR). The aim of this study was to measure myocardial extracellular volume (ECV) in amyloid patients with a novel T1 mapping CMR technique and to determine the correlation between ECV and disease severity. Methods. Thirty-six patients with biopsy-proven systemic amyloidosis (mean age 70 9 years, 31 men, 30 with AL and six with ATTR amyloidosis) and seven patients with possible amyloidosis (mean age 64 10 years, six men) underwent comprehensive clinical and CMR assessment, with ECV estimation from pre- and postcontrast T1 mapping. Thirty healthy subjects (mean age 39 17 years, 21 men) served as the control group.
Introduction Systemic amyloidosis is an uncommon multisystem disease caused by the extracellular deposition of misfolded protein in various tissues and organs [1]. Cardiac involvement is a leading cause of
Results. Amyloid patients presented with left ventricular (LV) concentric hypertrophy with impaired biventricular systolic function. Cardiac ECV was higher in amyloid patients (definite amyloidosis, 0.43 0.12; possible amyloidosis, 0.34 0.11) than in control subjects (0.26 0.04, P < 0.05); even in amyloid patients without late gadolinium enhancement (0.35 0.10), ECV was significantly higher than in the control group (P < 0.01). A cutoff value of myocardial ECV >0.316, corresponding to the 95th percentile in normal subjects, showed a sensitivity of 79% and specificity of 97% for discriminating amyloid patients from control subjects (area under the curve of 0.884). Myocardial ECV was significantly correlated with LV ejection fraction (R2 = 0.16), LV mean wall thickness (R2 = 0.41), LV diastolic function (R2 = 0.21), right ventricular ejection fraction (R2 = 0.13), N-terminal fragment of the pro-brain natriuretic peptides (R2 = 0.23) and cardiac troponin (R2 = 0.33). Conclusion. Myocardial ECV was increased in amyloid patients and correlated with disease severity. Thus, measurement of myocardial ECV represents a potential noninvasive index of amyloid burden for use in early diagnosis and disease monitoring. Keywords: amyloidosis, biomarker, cardiac hypertrophy, cardiology, radiology.
morbidity and mortality, especially in primary light chain (AL) amyloidosis and in both wild-type and hereditary transthyretin (ATTR) amyloidosis. Endomyocardial biopsy is considered the gold standard for demonstrating cardiac amyloidosis (CA), but a combination of noncardiac biopsy with
ª 2014 The Association for the Publication of the Journal of Internal Medicine
1
A. Barison et al.
electrocardiography (ECG) [low-voltage QRS potentials], echocardiography [concentric left ventricular (LV) hypertrophy] and biomarkers [increased levels of N-terminal fragment of the pro-brain natriuretic peptide (NT-proBNP)] is commonly used in clinical practice to diagnose CA [2]. The role of novel techniques, such as scintigraphy and cardiac magnetic resonance (CMR), is under investigation [3, 4]. CMR provides functional and morphological information in patients with CA and significantly improves tissue characterization. The appearance of global, subendocardial-to-transmural late gadolinium enhancement (LGE) is highly characteristic of CA (despite suboptimal sensitivity as it is present in only two-thirds of patients with CA[3, 5, 6]) and correlates with prognosis [7]. Recently, equilibrium contrast CMR has demonstrated much higher extracellular myocardial volume in CA than in any other disease investigated with this technique [8], and accurate measurements of the expanded interstitium in amyloidosis might be useful for serial quantification of cardiac amyloid burden. The aim of this study was to investigate gadolinium contrast kinetics and interstitial expansion in patients with CA using a CMR T1 mapping technique and to assess the correlation between these results and clinical, biohumoral, morphological and conventional LGE features. For comparison, we included in the study not only healthy control subjects, but also a group of patients at risk of CA due to hypergammaglobulinaemia or TTR mutation without clinical evidence of amyloidosis.
Magnetic resonance assessment of cardiac amyloidosis
myocardial infarction, coronary revascularization, myocarditis, nonischaemic cardiomyopathy, congenital heart disease or primary moderate-tosevere valvular disease were excluded from the study. The study was approved by our institutional ethics committee, and written informed consent was obtained for each patient. According to current criteria [2], the diagnosis of systemic amyloidosis in the 36 patients was based on a positive biopsy in any organ (n = 30 AL and n = 6 ATTR amyloidosis), whilst the seven patients with possible systemic amyloidosis had a negative soft-tissue biopsy but possible cardiac involvement due to long-standing hypergammaglobulinaemia (n = 6) or TTR mutation (n = 1). Endomyocardial biopsy had been performed in only one patient with AL amyloidosis. Moreover, according to the same criteria [2], the heart was considered to be involved when the mean LV wall thickness (septum and posterior wall) was >12 mm at echocardiography, in the absence or presence of hypertension or other potential causes of LV hypertrophy (definite or probable CA, respectively). Troponin I was assessed using an immunochemiluminescence assay (ADVIA, Siemens, Erlangen, Germany). NTproBNP was also assessed using an immunochemiluminescence assay (Elecsys, Roche Diagnostics, Monza, Italy). Total immunoglobulin light chain concentrations were measured using agarose gel immunoelectrophoresis (SeaKem ME agarose, Cambrex Bio Science, Rockland, USA), and free light chain concentration was measured using immunonephelometry (BN II analyser, Dade Behring, Marburg, Germany).
Materials and methods Patient selection
CMR protocol
We prospectively recruited 36 patients with biopsyproven systemic amyloidosis (mean age 70 9 years, 31 men), seven patients with possible amyloidosis (mean age 64 10 years, six men) and 30 healthy control subjects (mean age 39 17 years, 21 men) (P < 0.001 for age difference between control group and both patient groups; nonsignificant difference for gender). Patients with any contraindication (metal devices, severe renal failure or claustrophobia) or unable to undergo CMR imaging (poor breath-holding capacity or arrhythmias with extremely variable RR intervals) were excluded from the study. All patients underwent clinical, echocardiographic, ECG and biohumoral evaluation, and a complete CMR scan. Patients with a previous history of
Study participants were examined using a 1.5-T CMR unit (CVi, GE Healthcare, Milwaukee, WI, USA) with dedicated cardiac software, phased-array surface receiver coil and vectocardiogram triggering. Ventricular function was assessed by breath-hold steady-state free-precession cine imaging in cardiac short axis and vertical and horizontal long axis. In cardiac short axis, the ventricles were completely encompassed by contiguous slices (8 mm thick); sequence parameters were as follows: field-of-view: 380–400 mm, repetition/echo time: 3.2/1.6 ms, flip angle: 60°, matrix: 224 9 192, phases: 30. For determination of the T1 value of the myocardium and blood cavity, a single mid-ventricular short-axis modified-cine inversion-recovery (MCine-IR) sequence was performed before and at fixed time
2
ª 2014 The Association for the Publication of the Journal of Internal Medicine Journal of Internal Medicine
A. Barison et al.
intervals (5, 10, 15 and 20 min) after a bolus of contrast agent (0.2 mmol kg 1 Gadodiamide, OMINISCAN, GE Healthcare, Waukesha, WI, USA). The MCine-IR sequence consisted of a nonselective adiabatic inversion pulse applied at the R-wave of the ECG and followed by a cine-segmented gradientecho acquisition extended to the following six RR intervals for precontrast imaging or four RR intervals for postcontrast imaging [9]; sequence parameters were as follows: field-of-view: 380–400 mm, repetition/echo time: 6/2.8 ms, flip angle: 8°, matrix: 224 9 192, phases: 40. LGE imaging was performed between 10 and 20 min after contrast agent administration using a segmented T1weighted gradient-echo inversion-recovery pulse sequence. In short-axis orientation, the LV was completely encompassed by contiguous slices (8 mm thick). Images were also acquired in vertical and horizontal long-axis views. Inversion time was individually adapted to suppress the signal of normal remote myocardium (220–300 ms). Sequence parameters were as follows: field-of-view: 380– 400 mm, slice thickness: 8 mm, repetition/echo time: 4.6/1.3 ms, flip angle: 20°, matrix: 256 9 192. CMR image analysis All CMR studies were analysed off-line using a workstation (Advantage Workstation, GE Healthcare) with dedicated software (MASS 6.1, Medis, Leiden, The Netherlands) by one experienced operator blinded to the clinical data. Using the stack of short-axis cine images, LV volumes, mass and global function were determined. LV volumes and mass were normalized to body surface area. LGE was assessed as (i) present or absent, (ii) focal (involving three or fewer adjacent LV segments) or diffuse (more than three LV segments), (iii) midwall, subendocardial, subepicardial or transmural and (iv) whether or not there was involvement of the basal, mid and/or apical segments of the LV. For T1 mapping analysis, a region of interest was drawn in the interventricular septum, in the LV blood pool and in the skeletal muscle (pectoralis or latissimus dorsi) for each frame of the T1 mapping series in a single mid-ventricular short-axis slice; when analysing the myocardium, care was taken to avoid the blood pool. An exponential fitting model was used to obtain the T1 from blood, myocardium and skeletal muscle with customwritten software (HIPPO-SW; Fondazione Toscana Gabriele Monasterio, Pisa, Italy). Of the 60 and 40 frames acquired for pre- and postcontrast analy-
Magnetic resonance assessment of cardiac amyloidosis
sis, respectively, only those with artefacts were excluded maintain the fitting error below 10% (usually
