- Email: [email protected]
Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging
+Model
ARTICLE IN PRESS
DIII-847; No. of Pages 8
Diagnostic and Interventional Imaging (2016) xxx, xxx—xxx
ORIGINAL ARTICLE /Cardiac imaging
Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging F. Legou a, V. Tacher a, T. Damy b,c, V. Planté-Bordeneuve b,d, S. Rappeneau e, N. Benhaiem f, J. Rosso e, E. Itti e, A. Luciani a, H. Kobeiter a, A. Rahmouni a, J.-F. Deux a,b,∗ a
Radiology Department, Henri-Mondor Hospital, Paris Est Créteil University, Assistance publique—Hôpitaux de Paris, 94010 Créteil, France b Amyloid Network, Henri-Mondor Hospital, Paris Est Créteil University, 94010 Créteil, France c Cardiology Department, Henri-Mondor Hospital, Paris Est Créteil University, Assistance publique—Hôpitaux de Paris, 94010 Créteil, France d Neurology department, Henri-Mondor Hospital, Paris Est Créteil University, Assistance publique—Hôpitaux de Paris, 94010 Créteil, France e Nuclear Medecine department, Henri-Mondor Hospital, Paris Est Créteil University, Assistance publique—Hôpitaux de Paris, 94010 Créteil, France f Pathology department, Henri-Mondor Hospital, Paris Est Créteil University, Assistance publique—Hôpitaux de Paris, 94010 Créteil, France
KEYWORDS T2 ratio; Cardiac amyloidosis; Cardiac magnetic resonance imaging; Tissue characterization
Abstract Purpose: To detect if a difference of T2 ratio, defined as the signal intensity (SI) of the myocardium divided by the SI of the skeletal muscle on T2-weigthed cardiac magnetic resonance (CMR) imaging, exists between patients with systemic amyloidosis, by comparison to control subjects. To determine if a relationship exists between T2 ratio and the overall mortality. Materials and methods: CMR imaging examinations of 73 consecutive patients (48 men, 25 women; mean age, 63 years ±15 [SD]) with amyloidosis and suspicion of CA and 27 control subjects were retrospectively analyzed after institutional review board approval. Final diagnosis of CA was retained in case of histological confirmation of CA, typical pattern of CA on imaging and/or positivity of 99 Technetium-hydroxymethylene diphosphonate scintigraphy. Patients were divided in 2 groups according to the presence or the absence of CA. T2 ratios were calculated in patients with and those without CA and in control subjects with using analysis of variance. Prognostic value of T2 ratio was studied with a Kaplan—Meier curve.
∗ Corresponding author. Radiology department, Amyloidosis Network, Henri-Mondor hospital, 51, avenue Maréchal-de-Lattre-de-Tassigny, 94000 Créteil, France. E-mail address: [email protected] (J.-F. Deux).
http://dx.doi.org/10.1016/j.diii.2016.08.007 2211-5684/© 2016 Editions franc ¸aises de radiologie. Published by Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Legou F, et al. Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.007
+Model DIII-847; No. of Pages 8
ARTICLE IN PRESS
2
F. Legou et al. Results: Thirty-five patients (51%) had CA and 33 (49%) were free from CA. T2 ratio was lower in patients with CA (1.18 ± 0.29) than in patients without cardiac involvement (1.37 ± 0.35) (P = 0.03) and control subjects (1.45 ± 0.24) (P = 0.004). A T2 ratio of 1.36 was the best threshold value for predicting CA with a sensitivity of 63% and a specificity of 73%. Kaplan—Meier analysis showed a significant relationship between a shortened overall survival and a T2 ratio < 1.36. Conclusion: Patients with CA exhibit lower T2 ratio on CMR imaging by comparison with patients free of CA and control subjects. © 2016 Editions franc ¸aises de radiologie. Published by Elsevier Masson SAS. All rights reserved.
Amyloidosis is a group of uncommon systemic diseases characterized by the extracellular deposition of insoluble fibrillary amyloid proteins [1—3]. Many tissues and organs can be involved by amyloid proteins, such as liver, nerves, skin and heart. Myocardial involvement, also called cardiac amyloidosis (CA), is detected in 60% of patients with lightchain amyloidosis (AL) and between 5 to 25% of patients with transthyretin amyloidosis (ATTR), and is associated with a poor clinical prognosis [4,5]. Cardiac magnetic resonance imaging (MRI) is a useful examination to detect CA, usually with on late gadolinium enhancement images (LGE) that typically show a diffuse or circumferential subendocardial enhancement of the left ventricle, sometimes associated with enhancement of other cardiac chambers [6—9]. Other cardiac MRI signs were reported, such as a shortening the time of inversion and a suboptimal nulling pattern [6]. T2-weighted imaging can provide important clinical insight regarding the presence of myocardial edema in various cardiac diseases, such as recent myocardial infarction [10,11] or myocarditis [12—14]. In systemic amyloidosis, a reduction of T2 signal intensity has been reported in neck, spleen and the adrenal glands [15,16] and it has been hypothesized that  pleated-sheet ultrastructure protein may reduce water mobility and therefore T2 signal. A few studies, with conflicting results, have evaluated T2 imaging in CA [17,18] while it is likely that amyloid proteins deposit within the myocardium changes myocardial component and architecture of the heart and therefore influence T2 myocardial signal. In this study, we hypothesize that amyloid proteins deposit within myocardium induces a drop of myocardial signal on T2 imaging. The goal of this study was to detect if a difference of T2 ratio exist between patients with systemic amyloidosis, in comparison to control subjects on cardiac MRI. Our secondary endpoint was to determine if there was a relationship between the T2 ratio and the overall mortality.
Materials and methods Patients From September 2010 to September 2013, 73 consecutive patients with systemic histologically confirmed
amyloidosis and suspicion of CA, and referred to our institution for cardiac MRI, were considered for this retrospective study. There were 48 men and 25 women with a mean age of 63 years ±15 [SD]. Criteria for suspicion of CA were echocardiographic evidence of wall thickening (≥ 12 mm), increase of NT-proBNP level or cardiological symptoms (dyspnea with a mean New York Heart Association [NYHA] functional class > 2). All patients had histopathological confirmation of amyloidosis by tissue biopsies of the labial salivary gland (n = 39); nerve (n = 20); or gastrointestinal tissue (n = 14). Histopathological analysis was considered positive for amyloidosis in case of characteristic yellow-green birefringence under crossed polarizers on Congo red staining. Immunohistochemical stains were performed on tissue sections with the use of commercial antisera against kappa and lamba immunoglobulin light chains, transthyretin (TTR), lysozyme, apolipoprotein AI, and fibrinogen. In addition, 27 control subjects, included retrospectively, matching with patients for age and gender were also considered for this study. Baseline characteristics are reported in Table 1.
Cardiac MRI protocol All patients had cardiac MRI with a 1.5-T unit (Magnetom Avanto® , Siemens Medical Solutions, Erlangen, Germany) Table 1
Characteristics of the overall population. Patients with systemic amyloidosis (n = 73)
Control subjects (n = 27)
P
Men, n (%) Age (years)
48 (66) 63 ± 15 [28—88]
0.1 0.6
NT-proBNP level (pg/mL) ATTR (%) AL (%) Senile amyloidosis (%)
2306 ± 4496 [12—29,509] 48 (66) 19 (26) 6 (8)
15 (56) 58 ± 15 [30—89] NA NA NA NA
Data were presented as mean ± SD [range] or mean ± SD (%). AL: light-chain amyloidosis; ATTR: transthyretin amyloidosis; NA: non-assessable.
Please cite this article in press as: Legou F, et al. Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.007
+Model DIII-847; No. of Pages 8
ARTICLE IN PRESS
Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis equipped with a high-performance gradient sub-system (maximum amplitude, 40 mT/m; minimum rise, 200 s) and an 8-channel phased-array cardiac coil. Unenhanced cine steady state free precession (SSFP) sequences were performed for each patient. Contiguous short-axis sections encompassing the left ventricle were acquired from base to apex. The following parameters were used: TR/TE, 2.8/1.4 (apparent TR, 31.4 ms; 11 segments); flip angle, 82◦ ; matrix size, 192 × 192; FOV, 300 × 270 mm; slice thickness, 8 mm. Retrospective ECG gating was used with 25 phases per section. A short Tau inversion recovery (STIR) T2-weighted image was acquired in the middle short-axis section using the following parameters: the trigger pulse was activated every two heartbeats yielding a repetition time of 1500 to 2500 ms (depending on the heart rate), the echo time was 49 ms and the bandwidth was 235 Hz/pixel. The matrix size was 256 × 146 and the field of view 350 × 270 mm yielding a voxel size of 1.4 × 1.8 mm2 . Section thickness was 8 mm, turbo factor was 15 and TI was 150 ms. A surface coil intensity correction algorithm was used to compensate the myocardial intensity inhomogeneity. Late gadolinium enhancement (LGE) imaging was performed using a segmented three-dimensional (3D) inversion recovery (IR) gradient-echo T1-weighted technique 15 minutes after intravenous administration of a gadolinium chelate (gadoterate meglumine, Dotarem® , Guerbet, Roissy-Charles de Gaulle, France) at a dose of 0.2 mmol/kg of body weight. The sequence parameters were as follows: TR/TE, 3.9/1.4 ms; mean inversion recovery time, 220 ± 50 ms; flip angle, 10◦ ; matrix size, 192 × 192; FOV, 300 × 270 mm; number of sections, 12; slice thickness, 8 mm. Image acquisition lasted 12—20 seconds, depending on heart rate. TI scout parameters were as follows: TR/TE, 23.49/1.12 ms, flip angle: 30◦ , matrix size, 192 × 192 mm, FOV: 340 × 300, pixel size: 3.5 × 1.8 × 8 mm3 . Additional phase sensitive inversion recovery (PSIR) sequences were performed with the following parameters: TR/TE, 700/3.38 ms; Ti, 300 ms; flip angle, 25◦ ; matrix size, 256 × 256 mm; FOV, 340 × 300; pixel size, 1.8 × 1.3 × 8 mm3 . All control subjects had cardiac MRI with the same protocol. Indications of cardiac MRI were atypical chest pain, suspicion of left ventricular thickening on echocardiography and suspicion of arrhythmogenic right ventricular dysplasia (ARVD). All control subjects has a normal cardiac MRI examination defined as a normal left ventricular thickness (defined a wall thickness < 12 mm on end-diastolic short-axis cine images), a normal left ventricular ejection fraction (LVEF) (> 50%), normal left end-diastolic and end-systolic volumes, lack of segmental contraction impairment, no focal or diffuse myocardial edema on STIR T2 and no abnormal signal area detected on LGE sequences. All control subjects had also a normal right ventricular aspect and function.
3
readers were blinded to the clinical data. LGE and cine images were scored on a three-point scale: 0: no abnormal enhancement detected within cardiac chambers and normal left ventricular thickness, 1: doubtful enhancement within the left ventricle with normal left ventricular thickness, 2: diffuse or subendocardial enhancement of the left ventricle associated with left ventricular hypertrophy (≥ 12 mm on end-diastolic short-axis cine images).
Quantitative analysis LVEF, left ventricular mass and end-diastolic interventricular septal thickness were measured for all patients and control subjects on short-axis cine SSFP images using a dedicated workstation (Argus® ; Siemens Medical Solutions). Myocardial signal intensity (SI) was measured on the middle short-axis STIR T2 image (SIT2 myocardium ) from manual contouring of endocardial and epicardial left ventricular boundaries. All manual contouring of the myocardium were performed by the same operator (FL), in consensus with another author (JFD), with a thin boundary, and repeated 3 times per patients. The 3 values were averaged for each patient. SI of skeletal muscle (SIskeletal muscle T2 ) was measured on the same image using a region of interest (ROI) of at least 10 mm2 placed in a chest wall muscle. T2 ratio was calculated using the following formula: T2 ratio = ST2 myocardium /SIT2 skeletal muscle .
Scintigraphy evaluation and endomyocardial biopsy (EMB) Following cardiac MRI, 21 patients underwent a scintigraphy and 11 an endomyocardial biopsy (EMB). Six patients underwent both modalities. Mean delay between cardiac MR and scintigraphy and between cardiac MRI and EMB were 5 ± 3 and 10 ± 2 days, respectively. For scintigraphy, 99 Tc HMDP bone tracer (10 MBq/kg, Cis bio international, France) was injected intravenously according to the patient’s body weight. Images were acquired using a dual-head gamma camera system (Philips Precedence® SPECT/CT, Philips Healthcare, Amsterdam, The Netherlands). Bone phase whole-body scans were acquired 3 hours following administration of 99 Tc HMDP. Heart retention was visually scored in consensus by two experienced physicians (one nuclear medicine specialist with 20 years of experience [J.R.] and one radiologist with 2 years of experience [JF.D.]) blinded to all other patient data. This visual scoring was performed using a routine score system as described previously by Perrugini et al.: 0, no cardiac uptake and normal bone uptake; 1, slight cardiac uptake less marked than bone uptake; 2, moderate cardiac uptake with attenuated bone uptake; and 3, strong cardiac uptake with slight/absent bone uptake [19].
Diagnosis of CA Cardiac MRI analysis Qualitative analysis Two radiologists with 3 (F.L.) and 10 (J.-F.D.) years of experience in cardiac MRI, analyzed anonymously in consensus, LGE and cine SSFP images of all patients on a dedicated acquisition platform (Leonardo® ; Siemens Healthcare). The
The diagnosis of CA was retained in case of: • histological confirmation of CA on endomyocardial biopsy and/or; • significant heart retention of 99 Tc HMDP bone tracer on scintigraphy with a visual score equal ≥ 1 (Fig. 1) and/or; • typical pattern of CA on cardiac MRI; thickening of the left ventricle (end-diastolic thickness of left ventricular
Please cite this article in press as: Legou F, et al. Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.007
+Model DIII-847; No. of Pages 8
ARTICLE IN PRESS
4
F. Legou et al. investigate the correlation between T2 ratio and LGE score. Analysis of survival was performed with a Kaplan—Meier curve. Prism Version 5.0 GraphPad software was used for analysis (La Jolla, CA, USA).
Results Cardiac MRI Five patients were excluded due to poor image quality (3 had poor image quality on LGE images, 1 exhibited respiratory artifacts and one had a poor ECG gating). The remaining 68 patients and all the control subjects were analyzed. Thirty-one patients (46%) exhibited hypertrophy of the left ventricular wall associated with a significant enhancement of left ventricular myocardium on LGE images (scale 2) and were considered as having CA. The remaining 37 (54%) patients had none (n = 23; scale 0) or doubtful (n = 14; scale 1) enhancement and were considered free of CA. Myocardial mass, septal thickness and NT-proBNP level were significantly (P < 0.05) higher in patients ranked scale 2 (112 ± 29 g/m2 , 18 ± 4 mm and 4178 ± 5929 pg/mL, respectively) than in patients with scale 0—1 (77 ± 23 g/m2 , 10 ± 3 mm and 604 ± 1014 pg/mL, respectively). No control subject had positive LGE or left ventricular hypertrophy (mean septal thickness: 9 mm ± 2 [SD]).
Scintigraphy imaging and EMB
Figure 1. A 67-year-old man with cardiac amyloidosis. Wholebody 99 Tc HMDP image in anterior view shows a marked cardiac uptake (grade 3; arrow).
wall ≥ 12 mm) associated with a diffuse or circumferential enhancement of the left ventricle after gadolinium administration (grade 3 on the three-point scale cardiac MRI aforementioned) (Fig. 2).
Follow-up Survival during follow-up was performed by chart review and contact with the referring physician.
Statistical analysis Continuous data were expressed as mean ± standard deviation (SD) and ranges. The significance of differences between patients with CA, patients without CA and control subjects was analyzed with one way ANOVA test and Bonferroni correction. For subgroups analysis, a nonparametric Kruskal—Wallis test was used because the Kolmogorov—Smirnov test revealed a non-Gaussian distribution. T2 ratio values to predict CA were determined using a receiver-operating characteristic curve (ROC). Sensitivity, specificity were calculated using the thresholds previously defined. Differences were considered significant when P < 0.05. A Spearman correlation was conducted to
The 21 patients who had scintigraphy had a mean visual score of 1.5 ± 1.3 (0—3). Fourteen of these patients (67%) had a visual score ≥ 1 and were considered as having CA; all these patients were scale 2 on cardiac MRI. Cardiac involvement was confirmed histologically by EMB in 10 of 11 patients (91%) patients. Individual results of visual score of scintigraphy uptake, T2 ratio values and LGE score are reported for the 11 patients with EMB in the Table 2.
Diagnosis of CA Finally, taking into account cardiac MRI criteria, histological analysis and/or scintigraphy criteria, 35 patients (51%) were considered as having CA and 33 (49%) were considered to be free from cardiac disease. Among patients with CA, 20 patients (57%) were diagnosed with hereditary ATTR, 10 (29%) with AL and 5 (14%) with senile amyloidosis. Patients with CA exhibited significantly (P < 0.001) higher septal thickness (17 ± 5 mm), lower LVEF (52 ± 15%) and higher NT-proBNP level (3904 ± 5722 pg/mL) by comparison with patients without CA (10 ± 3 [SD] mm; 60 ± 9 [SD] % and 548 ± 987 [SD] pg/mL, respectively) (Table 3).
T2 ratio in patients with and without CA, and in control subjects T2 ratio was significantly lower in patients with CA (1.18 ± 0.29[SD]) than in patients without CA (1.37 ± 0.35 [SD]; P = 0.03) and control subjects (1.45 ± 0.24 [SD]; P = 0.004). T2 ratio was in the same range between patients without CA and control subjects (P = 0.9). All data are reported in Table 3. Examples of patients with and
Please cite this article in press as: Legou F, et al. Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.007
+Model DIII-847; No. of Pages 8
ARTICLE IN PRESS
Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis
5
Figure 2. A 72-year-old patient with cardiac amyloidosis (CA). The T2 ratio measured on short-axis STIR T2 image (a) was inferior to the 1.36 threshold, suggesting CA. LGE images acquired in the short-axis (b) and in 4 chamber planes (c) reveal left ventricular wall thickening and diffuse myocardial enhancement (arrows).
without CA and control patient are shown in Figs. 2—4. A weak negative correlation was detected between T2 ratio and cardiac MRI score (r = —0.38; P = 0.008). The ROC curves and the Youden index identified a threshold of 1.36 for the T2 ratio (AUC: 0.74; 95% CI [0.62—0.84]; P = 0.05) as the best thresholds for predicting CA on cardiac MRI. Sensitivity and specificity of 1.36 threshold to predict cardiac involvement on cardiac MRI were 63% and 73%, respectively.
Table 2 Values of visual score of scintigraphy, T2 ratio and LGE scale for the 11 patients with endomyocardial biopsy. All but one had histological evidence of CA on EMB.
Subgroup analysis The mean T2 ratio of the 31 patients with CA diagnosed on MR criteria exhibited a significantly (P < 0.05) lower T2 ratio (1.16 ± 0.30 [SD]) than patients with no CA (1.37 ± 0.34 [SD]) and than control subjects (1.45 ± 0.24 [SD]). Similarly, the mean T2 ratio of patients with CA diagnosed on histological analysis (1.22 ± 0.22 [SD]; n = 10) and patients with positive scintigraphy (1.26 ± 0.27 [SD]; n = 20) were significantly (P < 0.05) lower than this of the 27 control subjects (1.45 ± 0.24 [SD]). Regarding type of CA, we did not detect significant (P = 0.6) differences of T2 ratio between patients with AL (1.20 ± 0.31 [SD]; n = 10), ATTR (1.14 ± 0.29 [SD]; n = 20) and senile amyloidosis (1.32 ± 0.31 [SD]; n = 5).
Patients
EMB
Scintigraphy uptake
T2 ratio
LGE scale
Follow-up
1 2 3 4 5 6 7 8 9 10 11
+ + + + + + + + + + —
2 2 2 2 2 2 2 2 2 2 2
1.5 0.92 1.22 1.12 1.26 0.88 1.16 1.18 0.84 1.07 1.25
2 2 2 2 2 2 2 2 2 2 2
During the follow-up period of 18 ± 6 [SD] months (range: 5—30 months), 8 (23%) and 3 (9%) patients died respectively in the patients with and without CA. Among deceased patients with CA, 6 (75%) patients deceased from cardiac cause. None patient deceased from cardiac cause in patients without CA. The deceased patients had a lower T2 ratio (1.15 ± 0.33 [SD]) than those that did survive (1.33 ± 0.32 [SD]). Kaplan—Meier survival analyses revealed that patients who had a T2 ratio < 1.36 had poorer overall survival rate than patients who had a T2 ratio more than this threshold (P = 0.011) (Fig. 5).
EMB: endomyocardial biopsy; +: histological confirmation of cardiac amyloidosis on EMB. All but one patient had histological evidence of cardiac amyloidosis on EMB; LGE: late gadolinium enhancement.
Discussion In this study, we report that patients with CA exhibited a significant reduction of T2 ratio by comparison with patients
Please cite this article in press as: Legou F, et al. Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.007
+Model
ARTICLE IN PRESS
DIII-847; No. of Pages 8
6
F. Legou et al. Table 3 Septal thickness, left ventricular ejection fraction (LVEF), NT-proBNP level, T2 ratio and LGE score in three groups of patients. Patients with cardiac amyloidosis (n = 35) Septal thickness (mm) LVEF (%) NT-proBNP level (pg/mL) T2 ratio
17 52 3904 1.18
± ± ± ±
5 15 5722 0.29
Patients without cardiac amyloidosis (n = 33) 10 60 548 1.37
± ± ± ±
3 9 987 0.35
Control subjects (n = 27)
P
9±2 62 ± 7 NA 1.45 ± 0.24
0.02a 0.01a 0.05 < 0.04b
Data are expressed as mean ± SD except for LVEF (%). a Patients with CA vs. patients without CA. b Patients with CA vs. patients without CA (P = 0.038) and patient with CA vs. control subjects (P = 0.004).
Figure 3. A 45-year-old patient with systemic amyloidosis but without cardiac amyloidosis. The T2 ratio measured on short-axis STIR T2 image (a) is within the normal range (> 1.36). LGE images acquired in the short-axis (b) and in 4 chamber planes (c) show no abnormal enhancement within myocardium.
without CA and control subjects, and that T2 ratio could be used as a negative prognostic marker. Diagnosis of CA on cardiac MRI is usually suspected in case of left ventricular hypertrophy associated with a diffuse or subendocardial enhancement on LGE images. More recently, measurement of native myocardial T1 using mapping sequences has also been proposed as an additional marker to detect CA [20,21]. Myocardial T2 signal in CA remains controversial but still of interest because T2 sequences are widely available and do no require gadolinium administration. Therefore, suspecting CA on a simple measurement of T2 ratio could be useful from a clinical point of view. Wassmuth et al. have reported that myocardial T2 ratio was significantly reduced in patients with CA in comparison to control subjects [18]. By contrast, Sparrow et al.,
studying T2 value of myocardium with a mapping sequence, did not found significant difference between patients with CA and control subjects, and suggested that the calculation of myocardial T2 was not useful for the diagnosis [17]. In our study, we noticed a significant reduction of T2 ratio in the overall population of patients with CA but also in the subgroup of patients with histological confirmation of CA. These results trend to confirm the data of Wassmuth et al. [18]. In the present work, we explored mainly patients with trasnthyretin amyloidosis (60%) in contrast to the aforementioned study that included mostly patients with light-chain amyloidosis (94% of patients). Thus, it seems that the effect of the final aggregate of proteins on myocardial signal seemed to be present in both of these diseases despite different precursors of amyloid proteins between ATTR and
Please cite this article in press as: Legou F, et al. Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.007
+Model DIII-847; No. of Pages 8
ARTICLE IN PRESS
Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis
7
Figure 4. A 67-year-old control subject with cardiac disease. The T2 ratio measured on short-axis STIR T2 image (a) is in the normal range (> 1.36). LGE images acquired in the short-axis (b) and in 4 chamber planes (c) show no abnormal enhancement within myocardium.
Figure 5. Kaplan—Meier curve. Overall survival of patients with T2 ratio < 1.36 (dashed line) is significantly shorter than patients with CA and T2 ratio > 1.36 (solid line).
AL. Interestingly, the correlation between T2 ratio and level of gadolinium enhancement that we reported reinforce the relationship between amyloid proteins deposit and reduction of myocardial T2 signal intensity. To our knowledge, the explanation of this possible drop of myocardial signal has not been elucidated. One report have noticed a reduction of T2 signal intensity within extra CA lesion in one patient and suggested that local field heterogeneity could be involved [15]. On the other hand, it has been published that amyloid proteins contained an hydrophobic portion [22—24] and one can hypothesis that this portion may influence local myocardial water concentration and thus modify myocardial signal on T2 imaging.
Making the diagnosis of CA using only T2 ratio measurement could be attractive because no gadolinium chelate administration is needed. In addition, T2 ratio is not dependent of sequence type, manufacturer and magnetic field used, that facilitates its use in clinical practice. Nevertheless, it seems to be challenging because of a relatively low sensitivity (63%) of this measurement. Therefore, T2 ratio could be used as an additional diagnostic marker, especially in case of contraindication to gadolinium administration. Besides that measurement seems to be associated with shortened survival as we and previous authors [18] noticed and could be used to evaluate prognostic of the patients. Several limitations must be underlined in this study. First, only a minority of patients had histological confirmation of CA. Thus, it is difficult to affirm that all patients with positive MR imaging or scintigraphy criteria had cardiac involvement, even if a positive LGE is strongly associated with CA [25]. Second, we did not evaluate mapping sequences in this work that have been reported to be of interest to diagnose CA through T1 measurement [20,21]. Diagnosis confidence was reported to be higher with T2 mapping than with T2 weighed images [26,27] and a comparison between STIR T2 and T2 mapping sequences may have sense but was beyond the scope of this study. In conclusion, we report in a large cohort of patients with CA a decreased of T2 ratio in comparison to patients free of the disease and control subjects. This drop seemed to be associated with a poor prognostic on follow-up. Such measurement performed without the need of contrast medium
Please cite this article in press as: Legou F, et al. Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.007
+Model DIII-847; No. of Pages 8
ARTICLE IN PRESS
8
F. Legou et al.
injection, may be of interest in patients with renal failure or in whom cardiac MRI should be shortened.
Disclosure of interest The authors declare that they have no competing interest.
References [1] Benson MD. Pathogenesis of transthyretin amyloidosis. Amyloid 2012;19:14—5. [2] Falk RH, Comenzo RL, Skinner M. The systemic amyloidoses. N Engl J Med 1997;337:898—909. [3] Leung N, Nasr SH, Sethi S. How I treat amyloidosis: the importance of accurate diagnosis and amyloid typing. Blood 2012;120:3206—13. [4] Falk RH. Diagnosis and management of the cardiac amyloidoses. Circulation 2005;112:2047—60. [5] Roberts WC, Waller BF. Cardiac amyloidosis causing cardiac dysfunction: analysis of 54 necropsy patients. Am J Cardiol 1983;52:137—46. [6] Maceira AM, Joshi J, Prasad SK, et al. Cardiovascular magnetic resonance in cardiac amyloidosis. Circulation 2005;111:186—93. [7] Perugini E, Rapezzi C, Piva T, et al. Non-invasive evaluation of the myocardial substrate of cardiac amyloidosis by gadolinium cardiac magnetic resonance. Heart 2006;92:343—9. [8] Seeger A, Klumpp B, Kramer U, et al. MRI assessment of cardiac amyloidosis: experience of six cases with review of the current literature. Br J Radiol 2009;82:337—42. [9] Vogelsberg H, Mahrholdt H, Deluigi CC, et al. Cardiovascular magnetic resonance in clinically suspected cardiac amyloidosis: non-invasive imaging compared to endomyocardial biopsy. J Am Coll Cardiol 2008;51:1022—30. [10] Lonborg J, Vejlstrup N, Mathiasen AB, Thomsen C, Jensen JS, Engstrom T. Myocardial area at risk and salvage measured by T2-weighted cardiovascular magnetic resonance: reproducibility and comparison of two T2-weighted protocols. J Cardiovasc Magn Reson 2011;13:50. [11] Viallon M, Mewton N, Thuny F, et al. T2-weighted cardiac MR assessment of the myocardial area at risk and salvage area in acute reperfused myocardial infarction: comparison of stateof-the-art dark blood and bright blood T2-weighted sequences. J Magn Reson Imaging 2012;35:328—39. [12] Eitel I, Friedrich MG. T2-weighted cardiovascular magnetic resonance in acute cardiac disease. J Cardiovasc Magn Reson 2011;13:13.
[13] Luetkens JA, Doerner J, Thomas DK, et al. Acute myocarditis: multiparametric cardiac MR imaging. Radiology 2014;273: 383—92. [14] Stensaeth KH, Hoffmann P, Fossum E, Mangschau A, Sandvik L, Klow NE. Cardiac magnetic resonance visualizes acute and chronic myocardial injuries in myocarditis. Int J Cardiovasc Imaging 2012;28:327—35. [15] Gean-Marton AD, Kirsch CF, Vezina LG, et al. Focal amyloidosis on the head and neck: evaluation with CT and MR imaging. Radiology 1991;181:521—5. [16] Rafal RB, Jennis R, Kosovsky PA, Markisz JA. MRI of primary amyloidosis. Gastrointest Radiol 1990;15:199—201. [17] Sparrow P, Amirabadi A, Sussman MS, Paul N, Merchant N. Quantitative assessment of myocardial T2 relaxation times in cardiac amyloidosis. J Magn Reson Imaging 2009;30:942—6. [18] Wassmuth R, Abdel-Aty H, Bohl S, Schulz-Menger J. Prognostic impact of T2-weighted CMR imaging for cardiac amyloidosis. Eur Radiol 2011;21:1643—50. [19] Perugini E, Guidalotti PL, Salvi F, et al. Non-invasive etiologic diagnosis of cardiac amyloidosis using 99 mTc-3,3-diphosphono1,2-propanodicarboxylic acid scintigraphy. J Am Coll Cardiol 2005;46:1076—84. [20] Karamitsos TD, Piechnik SK, Banypersad SM, et al. Noncontrast T1 mapping for the diagnosis of cardiac amyloidosis. JACC Cardiovasc Imaging 2013;6:488—97. [21] Kramer CM, Chandrashekhar Y, Narula J. T1 mapping by CMR in cardiomyopathy: a non-invasive myocardial biopsy? JACC Cardiovasc Imaging 2013;6:532—4. [22] Bergstrom J, Gustavsson A, Hellman U, et al. Amyloid deposits in transthyretin-derived amyloidosis: cleaved transthyretin is associated with distinct amyloid morphology. J Pathol 2005;206:224—32. [23] Liu K, Cho HS, Hoyt DW, et al. Deuterium-proton exchange on the native wild-type transthyretin tetramer identifies the stable core of the individual subunits and indicates mobility at the subunit interface. J Mol Biol 2000;303:555—65. [24] Yang M, Lei M, Huo S. Why is Leu55—>Pro55 transthyretin variant the most amyloidogenic: insights from molecular dynamics simulations of transthyretin monomers. Protein Sci 2003;12:1222—31. [25] Syed IS, Glockner JF, Feng D, et al. Role of cardiac magnetic resonance imaging in the detection of cardiac amyloidosis. JACC Cardiovasc Imaging 2010;3:155—64. [26] Thavendiranathan P, Walls M, Giri S, et al. Improved detection of myocardial involvement in acute inflammatory cardiomyopathies using T2 mapping. Circ Cardiovasc Imaging 2012;5:102—10. [27] Montant P, Sigovan M, Revel D, Douek P. MR imaging assessment of myocardial edema with T2 mapping. Diagn Interv Imaging 2015;96:885—90.
Please cite this article in press as: Legou F, et al. Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.007
ARTICLE IN PRESS
DIII-847; No. of Pages 8
Diagnostic and Interventional Imaging (2016) xxx, xxx—xxx
ORIGINAL ARTICLE /Cardiac imaging
Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging F. Legou a, V. Tacher a, T. Damy b,c, V. Planté-Bordeneuve b,d, S. Rappeneau e, N. Benhaiem f, J. Rosso e, E. Itti e, A. Luciani a, H. Kobeiter a, A. Rahmouni a, J.-F. Deux a,b,∗ a
Radiology Department, Henri-Mondor Hospital, Paris Est Créteil University, Assistance publique—Hôpitaux de Paris, 94010 Créteil, France b Amyloid Network, Henri-Mondor Hospital, Paris Est Créteil University, 94010 Créteil, France c Cardiology Department, Henri-Mondor Hospital, Paris Est Créteil University, Assistance publique—Hôpitaux de Paris, 94010 Créteil, France d Neurology department, Henri-Mondor Hospital, Paris Est Créteil University, Assistance publique—Hôpitaux de Paris, 94010 Créteil, France e Nuclear Medecine department, Henri-Mondor Hospital, Paris Est Créteil University, Assistance publique—Hôpitaux de Paris, 94010 Créteil, France f Pathology department, Henri-Mondor Hospital, Paris Est Créteil University, Assistance publique—Hôpitaux de Paris, 94010 Créteil, France
KEYWORDS T2 ratio; Cardiac amyloidosis; Cardiac magnetic resonance imaging; Tissue characterization
Abstract Purpose: To detect if a difference of T2 ratio, defined as the signal intensity (SI) of the myocardium divided by the SI of the skeletal muscle on T2-weigthed cardiac magnetic resonance (CMR) imaging, exists between patients with systemic amyloidosis, by comparison to control subjects. To determine if a relationship exists between T2 ratio and the overall mortality. Materials and methods: CMR imaging examinations of 73 consecutive patients (48 men, 25 women; mean age, 63 years ±15 [SD]) with amyloidosis and suspicion of CA and 27 control subjects were retrospectively analyzed after institutional review board approval. Final diagnosis of CA was retained in case of histological confirmation of CA, typical pattern of CA on imaging and/or positivity of 99 Technetium-hydroxymethylene diphosphonate scintigraphy. Patients were divided in 2 groups according to the presence or the absence of CA. T2 ratios were calculated in patients with and those without CA and in control subjects with using analysis of variance. Prognostic value of T2 ratio was studied with a Kaplan—Meier curve.
∗ Corresponding author. Radiology department, Amyloidosis Network, Henri-Mondor hospital, 51, avenue Maréchal-de-Lattre-de-Tassigny, 94000 Créteil, France. E-mail address: [email protected] (J.-F. Deux).
http://dx.doi.org/10.1016/j.diii.2016.08.007 2211-5684/© 2016 Editions franc ¸aises de radiologie. Published by Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Legou F, et al. Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.007
+Model DIII-847; No. of Pages 8
ARTICLE IN PRESS
2
F. Legou et al. Results: Thirty-five patients (51%) had CA and 33 (49%) were free from CA. T2 ratio was lower in patients with CA (1.18 ± 0.29) than in patients without cardiac involvement (1.37 ± 0.35) (P = 0.03) and control subjects (1.45 ± 0.24) (P = 0.004). A T2 ratio of 1.36 was the best threshold value for predicting CA with a sensitivity of 63% and a specificity of 73%. Kaplan—Meier analysis showed a significant relationship between a shortened overall survival and a T2 ratio < 1.36. Conclusion: Patients with CA exhibit lower T2 ratio on CMR imaging by comparison with patients free of CA and control subjects. © 2016 Editions franc ¸aises de radiologie. Published by Elsevier Masson SAS. All rights reserved.
Amyloidosis is a group of uncommon systemic diseases characterized by the extracellular deposition of insoluble fibrillary amyloid proteins [1—3]. Many tissues and organs can be involved by amyloid proteins, such as liver, nerves, skin and heart. Myocardial involvement, also called cardiac amyloidosis (CA), is detected in 60% of patients with lightchain amyloidosis (AL) and between 5 to 25% of patients with transthyretin amyloidosis (ATTR), and is associated with a poor clinical prognosis [4,5]. Cardiac magnetic resonance imaging (MRI) is a useful examination to detect CA, usually with on late gadolinium enhancement images (LGE) that typically show a diffuse or circumferential subendocardial enhancement of the left ventricle, sometimes associated with enhancement of other cardiac chambers [6—9]. Other cardiac MRI signs were reported, such as a shortening the time of inversion and a suboptimal nulling pattern [6]. T2-weighted imaging can provide important clinical insight regarding the presence of myocardial edema in various cardiac diseases, such as recent myocardial infarction [10,11] or myocarditis [12—14]. In systemic amyloidosis, a reduction of T2 signal intensity has been reported in neck, spleen and the adrenal glands [15,16] and it has been hypothesized that  pleated-sheet ultrastructure protein may reduce water mobility and therefore T2 signal. A few studies, with conflicting results, have evaluated T2 imaging in CA [17,18] while it is likely that amyloid proteins deposit within the myocardium changes myocardial component and architecture of the heart and therefore influence T2 myocardial signal. In this study, we hypothesize that amyloid proteins deposit within myocardium induces a drop of myocardial signal on T2 imaging. The goal of this study was to detect if a difference of T2 ratio exist between patients with systemic amyloidosis, in comparison to control subjects on cardiac MRI. Our secondary endpoint was to determine if there was a relationship between the T2 ratio and the overall mortality.
Materials and methods Patients From September 2010 to September 2013, 73 consecutive patients with systemic histologically confirmed
amyloidosis and suspicion of CA, and referred to our institution for cardiac MRI, were considered for this retrospective study. There were 48 men and 25 women with a mean age of 63 years ±15 [SD]. Criteria for suspicion of CA were echocardiographic evidence of wall thickening (≥ 12 mm), increase of NT-proBNP level or cardiological symptoms (dyspnea with a mean New York Heart Association [NYHA] functional class > 2). All patients had histopathological confirmation of amyloidosis by tissue biopsies of the labial salivary gland (n = 39); nerve (n = 20); or gastrointestinal tissue (n = 14). Histopathological analysis was considered positive for amyloidosis in case of characteristic yellow-green birefringence under crossed polarizers on Congo red staining. Immunohistochemical stains were performed on tissue sections with the use of commercial antisera against kappa and lamba immunoglobulin light chains, transthyretin (TTR), lysozyme, apolipoprotein AI, and fibrinogen. In addition, 27 control subjects, included retrospectively, matching with patients for age and gender were also considered for this study. Baseline characteristics are reported in Table 1.
Cardiac MRI protocol All patients had cardiac MRI with a 1.5-T unit (Magnetom Avanto® , Siemens Medical Solutions, Erlangen, Germany) Table 1
Characteristics of the overall population. Patients with systemic amyloidosis (n = 73)
Control subjects (n = 27)
P
Men, n (%) Age (years)
48 (66) 63 ± 15 [28—88]
0.1 0.6
NT-proBNP level (pg/mL) ATTR (%) AL (%) Senile amyloidosis (%)
2306 ± 4496 [12—29,509] 48 (66) 19 (26) 6 (8)
15 (56) 58 ± 15 [30—89] NA NA NA NA
Data were presented as mean ± SD [range] or mean ± SD (%). AL: light-chain amyloidosis; ATTR: transthyretin amyloidosis; NA: non-assessable.
Please cite this article in press as: Legou F, et al. Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.007
+Model DIII-847; No. of Pages 8
ARTICLE IN PRESS
Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis equipped with a high-performance gradient sub-system (maximum amplitude, 40 mT/m; minimum rise, 200 s) and an 8-channel phased-array cardiac coil. Unenhanced cine steady state free precession (SSFP) sequences were performed for each patient. Contiguous short-axis sections encompassing the left ventricle were acquired from base to apex. The following parameters were used: TR/TE, 2.8/1.4 (apparent TR, 31.4 ms; 11 segments); flip angle, 82◦ ; matrix size, 192 × 192; FOV, 300 × 270 mm; slice thickness, 8 mm. Retrospective ECG gating was used with 25 phases per section. A short Tau inversion recovery (STIR) T2-weighted image was acquired in the middle short-axis section using the following parameters: the trigger pulse was activated every two heartbeats yielding a repetition time of 1500 to 2500 ms (depending on the heart rate), the echo time was 49 ms and the bandwidth was 235 Hz/pixel. The matrix size was 256 × 146 and the field of view 350 × 270 mm yielding a voxel size of 1.4 × 1.8 mm2 . Section thickness was 8 mm, turbo factor was 15 and TI was 150 ms. A surface coil intensity correction algorithm was used to compensate the myocardial intensity inhomogeneity. Late gadolinium enhancement (LGE) imaging was performed using a segmented three-dimensional (3D) inversion recovery (IR) gradient-echo T1-weighted technique 15 minutes after intravenous administration of a gadolinium chelate (gadoterate meglumine, Dotarem® , Guerbet, Roissy-Charles de Gaulle, France) at a dose of 0.2 mmol/kg of body weight. The sequence parameters were as follows: TR/TE, 3.9/1.4 ms; mean inversion recovery time, 220 ± 50 ms; flip angle, 10◦ ; matrix size, 192 × 192; FOV, 300 × 270 mm; number of sections, 12; slice thickness, 8 mm. Image acquisition lasted 12—20 seconds, depending on heart rate. TI scout parameters were as follows: TR/TE, 23.49/1.12 ms, flip angle: 30◦ , matrix size, 192 × 192 mm, FOV: 340 × 300, pixel size: 3.5 × 1.8 × 8 mm3 . Additional phase sensitive inversion recovery (PSIR) sequences were performed with the following parameters: TR/TE, 700/3.38 ms; Ti, 300 ms; flip angle, 25◦ ; matrix size, 256 × 256 mm; FOV, 340 × 300; pixel size, 1.8 × 1.3 × 8 mm3 . All control subjects had cardiac MRI with the same protocol. Indications of cardiac MRI were atypical chest pain, suspicion of left ventricular thickening on echocardiography and suspicion of arrhythmogenic right ventricular dysplasia (ARVD). All control subjects has a normal cardiac MRI examination defined as a normal left ventricular thickness (defined a wall thickness < 12 mm on end-diastolic short-axis cine images), a normal left ventricular ejection fraction (LVEF) (> 50%), normal left end-diastolic and end-systolic volumes, lack of segmental contraction impairment, no focal or diffuse myocardial edema on STIR T2 and no abnormal signal area detected on LGE sequences. All control subjects had also a normal right ventricular aspect and function.
3
readers were blinded to the clinical data. LGE and cine images were scored on a three-point scale: 0: no abnormal enhancement detected within cardiac chambers and normal left ventricular thickness, 1: doubtful enhancement within the left ventricle with normal left ventricular thickness, 2: diffuse or subendocardial enhancement of the left ventricle associated with left ventricular hypertrophy (≥ 12 mm on end-diastolic short-axis cine images).
Quantitative analysis LVEF, left ventricular mass and end-diastolic interventricular septal thickness were measured for all patients and control subjects on short-axis cine SSFP images using a dedicated workstation (Argus® ; Siemens Medical Solutions). Myocardial signal intensity (SI) was measured on the middle short-axis STIR T2 image (SIT2 myocardium ) from manual contouring of endocardial and epicardial left ventricular boundaries. All manual contouring of the myocardium were performed by the same operator (FL), in consensus with another author (JFD), with a thin boundary, and repeated 3 times per patients. The 3 values were averaged for each patient. SI of skeletal muscle (SIskeletal muscle T2 ) was measured on the same image using a region of interest (ROI) of at least 10 mm2 placed in a chest wall muscle. T2 ratio was calculated using the following formula: T2 ratio = ST2 myocardium /SIT2 skeletal muscle .
Scintigraphy evaluation and endomyocardial biopsy (EMB) Following cardiac MRI, 21 patients underwent a scintigraphy and 11 an endomyocardial biopsy (EMB). Six patients underwent both modalities. Mean delay between cardiac MR and scintigraphy and between cardiac MRI and EMB were 5 ± 3 and 10 ± 2 days, respectively. For scintigraphy, 99 Tc HMDP bone tracer (10 MBq/kg, Cis bio international, France) was injected intravenously according to the patient’s body weight. Images were acquired using a dual-head gamma camera system (Philips Precedence® SPECT/CT, Philips Healthcare, Amsterdam, The Netherlands). Bone phase whole-body scans were acquired 3 hours following administration of 99 Tc HMDP. Heart retention was visually scored in consensus by two experienced physicians (one nuclear medicine specialist with 20 years of experience [J.R.] and one radiologist with 2 years of experience [JF.D.]) blinded to all other patient data. This visual scoring was performed using a routine score system as described previously by Perrugini et al.: 0, no cardiac uptake and normal bone uptake; 1, slight cardiac uptake less marked than bone uptake; 2, moderate cardiac uptake with attenuated bone uptake; and 3, strong cardiac uptake with slight/absent bone uptake [19].
Diagnosis of CA Cardiac MRI analysis Qualitative analysis Two radiologists with 3 (F.L.) and 10 (J.-F.D.) years of experience in cardiac MRI, analyzed anonymously in consensus, LGE and cine SSFP images of all patients on a dedicated acquisition platform (Leonardo® ; Siemens Healthcare). The
The diagnosis of CA was retained in case of: • histological confirmation of CA on endomyocardial biopsy and/or; • significant heart retention of 99 Tc HMDP bone tracer on scintigraphy with a visual score equal ≥ 1 (Fig. 1) and/or; • typical pattern of CA on cardiac MRI; thickening of the left ventricle (end-diastolic thickness of left ventricular
Please cite this article in press as: Legou F, et al. Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.007
+Model DIII-847; No. of Pages 8
ARTICLE IN PRESS
4
F. Legou et al. investigate the correlation between T2 ratio and LGE score. Analysis of survival was performed with a Kaplan—Meier curve. Prism Version 5.0 GraphPad software was used for analysis (La Jolla, CA, USA).
Results Cardiac MRI Five patients were excluded due to poor image quality (3 had poor image quality on LGE images, 1 exhibited respiratory artifacts and one had a poor ECG gating). The remaining 68 patients and all the control subjects were analyzed. Thirty-one patients (46%) exhibited hypertrophy of the left ventricular wall associated with a significant enhancement of left ventricular myocardium on LGE images (scale 2) and were considered as having CA. The remaining 37 (54%) patients had none (n = 23; scale 0) or doubtful (n = 14; scale 1) enhancement and were considered free of CA. Myocardial mass, septal thickness and NT-proBNP level were significantly (P < 0.05) higher in patients ranked scale 2 (112 ± 29 g/m2 , 18 ± 4 mm and 4178 ± 5929 pg/mL, respectively) than in patients with scale 0—1 (77 ± 23 g/m2 , 10 ± 3 mm and 604 ± 1014 pg/mL, respectively). No control subject had positive LGE or left ventricular hypertrophy (mean septal thickness: 9 mm ± 2 [SD]).
Scintigraphy imaging and EMB
Figure 1. A 67-year-old man with cardiac amyloidosis. Wholebody 99 Tc HMDP image in anterior view shows a marked cardiac uptake (grade 3; arrow).
wall ≥ 12 mm) associated with a diffuse or circumferential enhancement of the left ventricle after gadolinium administration (grade 3 on the three-point scale cardiac MRI aforementioned) (Fig. 2).
Follow-up Survival during follow-up was performed by chart review and contact with the referring physician.
Statistical analysis Continuous data were expressed as mean ± standard deviation (SD) and ranges. The significance of differences between patients with CA, patients without CA and control subjects was analyzed with one way ANOVA test and Bonferroni correction. For subgroups analysis, a nonparametric Kruskal—Wallis test was used because the Kolmogorov—Smirnov test revealed a non-Gaussian distribution. T2 ratio values to predict CA were determined using a receiver-operating characteristic curve (ROC). Sensitivity, specificity were calculated using the thresholds previously defined. Differences were considered significant when P < 0.05. A Spearman correlation was conducted to
The 21 patients who had scintigraphy had a mean visual score of 1.5 ± 1.3 (0—3). Fourteen of these patients (67%) had a visual score ≥ 1 and were considered as having CA; all these patients were scale 2 on cardiac MRI. Cardiac involvement was confirmed histologically by EMB in 10 of 11 patients (91%) patients. Individual results of visual score of scintigraphy uptake, T2 ratio values and LGE score are reported for the 11 patients with EMB in the Table 2.
Diagnosis of CA Finally, taking into account cardiac MRI criteria, histological analysis and/or scintigraphy criteria, 35 patients (51%) were considered as having CA and 33 (49%) were considered to be free from cardiac disease. Among patients with CA, 20 patients (57%) were diagnosed with hereditary ATTR, 10 (29%) with AL and 5 (14%) with senile amyloidosis. Patients with CA exhibited significantly (P < 0.001) higher septal thickness (17 ± 5 mm), lower LVEF (52 ± 15%) and higher NT-proBNP level (3904 ± 5722 pg/mL) by comparison with patients without CA (10 ± 3 [SD] mm; 60 ± 9 [SD] % and 548 ± 987 [SD] pg/mL, respectively) (Table 3).
T2 ratio in patients with and without CA, and in control subjects T2 ratio was significantly lower in patients with CA (1.18 ± 0.29[SD]) than in patients without CA (1.37 ± 0.35 [SD]; P = 0.03) and control subjects (1.45 ± 0.24 [SD]; P = 0.004). T2 ratio was in the same range between patients without CA and control subjects (P = 0.9). All data are reported in Table 3. Examples of patients with and
Please cite this article in press as: Legou F, et al. Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.007
+Model DIII-847; No. of Pages 8
ARTICLE IN PRESS
Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis
5
Figure 2. A 72-year-old patient with cardiac amyloidosis (CA). The T2 ratio measured on short-axis STIR T2 image (a) was inferior to the 1.36 threshold, suggesting CA. LGE images acquired in the short-axis (b) and in 4 chamber planes (c) reveal left ventricular wall thickening and diffuse myocardial enhancement (arrows).
without CA and control patient are shown in Figs. 2—4. A weak negative correlation was detected between T2 ratio and cardiac MRI score (r = —0.38; P = 0.008). The ROC curves and the Youden index identified a threshold of 1.36 for the T2 ratio (AUC: 0.74; 95% CI [0.62—0.84]; P = 0.05) as the best thresholds for predicting CA on cardiac MRI. Sensitivity and specificity of 1.36 threshold to predict cardiac involvement on cardiac MRI were 63% and 73%, respectively.
Table 2 Values of visual score of scintigraphy, T2 ratio and LGE scale for the 11 patients with endomyocardial biopsy. All but one had histological evidence of CA on EMB.
Subgroup analysis The mean T2 ratio of the 31 patients with CA diagnosed on MR criteria exhibited a significantly (P < 0.05) lower T2 ratio (1.16 ± 0.30 [SD]) than patients with no CA (1.37 ± 0.34 [SD]) and than control subjects (1.45 ± 0.24 [SD]). Similarly, the mean T2 ratio of patients with CA diagnosed on histological analysis (1.22 ± 0.22 [SD]; n = 10) and patients with positive scintigraphy (1.26 ± 0.27 [SD]; n = 20) were significantly (P < 0.05) lower than this of the 27 control subjects (1.45 ± 0.24 [SD]). Regarding type of CA, we did not detect significant (P = 0.6) differences of T2 ratio between patients with AL (1.20 ± 0.31 [SD]; n = 10), ATTR (1.14 ± 0.29 [SD]; n = 20) and senile amyloidosis (1.32 ± 0.31 [SD]; n = 5).
Patients
EMB
Scintigraphy uptake
T2 ratio
LGE scale
Follow-up
1 2 3 4 5 6 7 8 9 10 11
+ + + + + + + + + + —
2 2 2 2 2 2 2 2 2 2 2
1.5 0.92 1.22 1.12 1.26 0.88 1.16 1.18 0.84 1.07 1.25
2 2 2 2 2 2 2 2 2 2 2
During the follow-up period of 18 ± 6 [SD] months (range: 5—30 months), 8 (23%) and 3 (9%) patients died respectively in the patients with and without CA. Among deceased patients with CA, 6 (75%) patients deceased from cardiac cause. None patient deceased from cardiac cause in patients without CA. The deceased patients had a lower T2 ratio (1.15 ± 0.33 [SD]) than those that did survive (1.33 ± 0.32 [SD]). Kaplan—Meier survival analyses revealed that patients who had a T2 ratio < 1.36 had poorer overall survival rate than patients who had a T2 ratio more than this threshold (P = 0.011) (Fig. 5).
EMB: endomyocardial biopsy; +: histological confirmation of cardiac amyloidosis on EMB. All but one patient had histological evidence of cardiac amyloidosis on EMB; LGE: late gadolinium enhancement.
Discussion In this study, we report that patients with CA exhibited a significant reduction of T2 ratio by comparison with patients
Please cite this article in press as: Legou F, et al. Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.007
+Model
ARTICLE IN PRESS
DIII-847; No. of Pages 8
6
F. Legou et al. Table 3 Septal thickness, left ventricular ejection fraction (LVEF), NT-proBNP level, T2 ratio and LGE score in three groups of patients. Patients with cardiac amyloidosis (n = 35) Septal thickness (mm) LVEF (%) NT-proBNP level (pg/mL) T2 ratio
17 52 3904 1.18
± ± ± ±
5 15 5722 0.29
Patients without cardiac amyloidosis (n = 33) 10 60 548 1.37
± ± ± ±
3 9 987 0.35
Control subjects (n = 27)
P
9±2 62 ± 7 NA 1.45 ± 0.24
0.02a 0.01a 0.05 < 0.04b
Data are expressed as mean ± SD except for LVEF (%). a Patients with CA vs. patients without CA. b Patients with CA vs. patients without CA (P = 0.038) and patient with CA vs. control subjects (P = 0.004).
Figure 3. A 45-year-old patient with systemic amyloidosis but without cardiac amyloidosis. The T2 ratio measured on short-axis STIR T2 image (a) is within the normal range (> 1.36). LGE images acquired in the short-axis (b) and in 4 chamber planes (c) show no abnormal enhancement within myocardium.
without CA and control subjects, and that T2 ratio could be used as a negative prognostic marker. Diagnosis of CA on cardiac MRI is usually suspected in case of left ventricular hypertrophy associated with a diffuse or subendocardial enhancement on LGE images. More recently, measurement of native myocardial T1 using mapping sequences has also been proposed as an additional marker to detect CA [20,21]. Myocardial T2 signal in CA remains controversial but still of interest because T2 sequences are widely available and do no require gadolinium administration. Therefore, suspecting CA on a simple measurement of T2 ratio could be useful from a clinical point of view. Wassmuth et al. have reported that myocardial T2 ratio was significantly reduced in patients with CA in comparison to control subjects [18]. By contrast, Sparrow et al.,
studying T2 value of myocardium with a mapping sequence, did not found significant difference between patients with CA and control subjects, and suggested that the calculation of myocardial T2 was not useful for the diagnosis [17]. In our study, we noticed a significant reduction of T2 ratio in the overall population of patients with CA but also in the subgroup of patients with histological confirmation of CA. These results trend to confirm the data of Wassmuth et al. [18]. In the present work, we explored mainly patients with trasnthyretin amyloidosis (60%) in contrast to the aforementioned study that included mostly patients with light-chain amyloidosis (94% of patients). Thus, it seems that the effect of the final aggregate of proteins on myocardial signal seemed to be present in both of these diseases despite different precursors of amyloid proteins between ATTR and
Please cite this article in press as: Legou F, et al. Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.007
+Model DIII-847; No. of Pages 8
ARTICLE IN PRESS
Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis
7
Figure 4. A 67-year-old control subject with cardiac disease. The T2 ratio measured on short-axis STIR T2 image (a) is in the normal range (> 1.36). LGE images acquired in the short-axis (b) and in 4 chamber planes (c) show no abnormal enhancement within myocardium.
Figure 5. Kaplan—Meier curve. Overall survival of patients with T2 ratio < 1.36 (dashed line) is significantly shorter than patients with CA and T2 ratio > 1.36 (solid line).
AL. Interestingly, the correlation between T2 ratio and level of gadolinium enhancement that we reported reinforce the relationship between amyloid proteins deposit and reduction of myocardial T2 signal intensity. To our knowledge, the explanation of this possible drop of myocardial signal has not been elucidated. One report have noticed a reduction of T2 signal intensity within extra CA lesion in one patient and suggested that local field heterogeneity could be involved [15]. On the other hand, it has been published that amyloid proteins contained an hydrophobic portion [22—24] and one can hypothesis that this portion may influence local myocardial water concentration and thus modify myocardial signal on T2 imaging.
Making the diagnosis of CA using only T2 ratio measurement could be attractive because no gadolinium chelate administration is needed. In addition, T2 ratio is not dependent of sequence type, manufacturer and magnetic field used, that facilitates its use in clinical practice. Nevertheless, it seems to be challenging because of a relatively low sensitivity (63%) of this measurement. Therefore, T2 ratio could be used as an additional diagnostic marker, especially in case of contraindication to gadolinium administration. Besides that measurement seems to be associated with shortened survival as we and previous authors [18] noticed and could be used to evaluate prognostic of the patients. Several limitations must be underlined in this study. First, only a minority of patients had histological confirmation of CA. Thus, it is difficult to affirm that all patients with positive MR imaging or scintigraphy criteria had cardiac involvement, even if a positive LGE is strongly associated with CA [25]. Second, we did not evaluate mapping sequences in this work that have been reported to be of interest to diagnose CA through T1 measurement [20,21]. Diagnosis confidence was reported to be higher with T2 mapping than with T2 weighed images [26,27] and a comparison between STIR T2 and T2 mapping sequences may have sense but was beyond the scope of this study. In conclusion, we report in a large cohort of patients with CA a decreased of T2 ratio in comparison to patients free of the disease and control subjects. This drop seemed to be associated with a poor prognostic on follow-up. Such measurement performed without the need of contrast medium
Please cite this article in press as: Legou F, et al. Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.007
+Model DIII-847; No. of Pages 8
ARTICLE IN PRESS
8
F. Legou et al.
injection, may be of interest in patients with renal failure or in whom cardiac MRI should be shortened.
Disclosure of interest The authors declare that they have no competing interest.
References [1] Benson MD. Pathogenesis of transthyretin amyloidosis. Amyloid 2012;19:14—5. [2] Falk RH, Comenzo RL, Skinner M. The systemic amyloidoses. N Engl J Med 1997;337:898—909. [3] Leung N, Nasr SH, Sethi S. How I treat amyloidosis: the importance of accurate diagnosis and amyloid typing. Blood 2012;120:3206—13. [4] Falk RH. Diagnosis and management of the cardiac amyloidoses. Circulation 2005;112:2047—60. [5] Roberts WC, Waller BF. Cardiac amyloidosis causing cardiac dysfunction: analysis of 54 necropsy patients. Am J Cardiol 1983;52:137—46. [6] Maceira AM, Joshi J, Prasad SK, et al. Cardiovascular magnetic resonance in cardiac amyloidosis. Circulation 2005;111:186—93. [7] Perugini E, Rapezzi C, Piva T, et al. Non-invasive evaluation of the myocardial substrate of cardiac amyloidosis by gadolinium cardiac magnetic resonance. Heart 2006;92:343—9. [8] Seeger A, Klumpp B, Kramer U, et al. MRI assessment of cardiac amyloidosis: experience of six cases with review of the current literature. Br J Radiol 2009;82:337—42. [9] Vogelsberg H, Mahrholdt H, Deluigi CC, et al. Cardiovascular magnetic resonance in clinically suspected cardiac amyloidosis: non-invasive imaging compared to endomyocardial biopsy. J Am Coll Cardiol 2008;51:1022—30. [10] Lonborg J, Vejlstrup N, Mathiasen AB, Thomsen C, Jensen JS, Engstrom T. Myocardial area at risk and salvage measured by T2-weighted cardiovascular magnetic resonance: reproducibility and comparison of two T2-weighted protocols. J Cardiovasc Magn Reson 2011;13:50. [11] Viallon M, Mewton N, Thuny F, et al. T2-weighted cardiac MR assessment of the myocardial area at risk and salvage area in acute reperfused myocardial infarction: comparison of stateof-the-art dark blood and bright blood T2-weighted sequences. J Magn Reson Imaging 2012;35:328—39. [12] Eitel I, Friedrich MG. T2-weighted cardiovascular magnetic resonance in acute cardiac disease. J Cardiovasc Magn Reson 2011;13:13.
[13] Luetkens JA, Doerner J, Thomas DK, et al. Acute myocarditis: multiparametric cardiac MR imaging. Radiology 2014;273: 383—92. [14] Stensaeth KH, Hoffmann P, Fossum E, Mangschau A, Sandvik L, Klow NE. Cardiac magnetic resonance visualizes acute and chronic myocardial injuries in myocarditis. Int J Cardiovasc Imaging 2012;28:327—35. [15] Gean-Marton AD, Kirsch CF, Vezina LG, et al. Focal amyloidosis on the head and neck: evaluation with CT and MR imaging. Radiology 1991;181:521—5. [16] Rafal RB, Jennis R, Kosovsky PA, Markisz JA. MRI of primary amyloidosis. Gastrointest Radiol 1990;15:199—201. [17] Sparrow P, Amirabadi A, Sussman MS, Paul N, Merchant N. Quantitative assessment of myocardial T2 relaxation times in cardiac amyloidosis. J Magn Reson Imaging 2009;30:942—6. [18] Wassmuth R, Abdel-Aty H, Bohl S, Schulz-Menger J. Prognostic impact of T2-weighted CMR imaging for cardiac amyloidosis. Eur Radiol 2011;21:1643—50. [19] Perugini E, Guidalotti PL, Salvi F, et al. Non-invasive etiologic diagnosis of cardiac amyloidosis using 99 mTc-3,3-diphosphono1,2-propanodicarboxylic acid scintigraphy. J Am Coll Cardiol 2005;46:1076—84. [20] Karamitsos TD, Piechnik SK, Banypersad SM, et al. Noncontrast T1 mapping for the diagnosis of cardiac amyloidosis. JACC Cardiovasc Imaging 2013;6:488—97. [21] Kramer CM, Chandrashekhar Y, Narula J. T1 mapping by CMR in cardiomyopathy: a non-invasive myocardial biopsy? JACC Cardiovasc Imaging 2013;6:532—4. [22] Bergstrom J, Gustavsson A, Hellman U, et al. Amyloid deposits in transthyretin-derived amyloidosis: cleaved transthyretin is associated with distinct amyloid morphology. J Pathol 2005;206:224—32. [23] Liu K, Cho HS, Hoyt DW, et al. Deuterium-proton exchange on the native wild-type transthyretin tetramer identifies the stable core of the individual subunits and indicates mobility at the subunit interface. J Mol Biol 2000;303:555—65. [24] Yang M, Lei M, Huo S. Why is Leu55—>Pro55 transthyretin variant the most amyloidogenic: insights from molecular dynamics simulations of transthyretin monomers. Protein Sci 2003;12:1222—31. [25] Syed IS, Glockner JF, Feng D, et al. Role of cardiac magnetic resonance imaging in the detection of cardiac amyloidosis. JACC Cardiovasc Imaging 2010;3:155—64. [26] Thavendiranathan P, Walls M, Giri S, et al. Improved detection of myocardial involvement in acute inflammatory cardiomyopathies using T2 mapping. Circ Cardiovasc Imaging 2012;5:102—10. [27] Montant P, Sigovan M, Revel D, Douek P. MR imaging assessment of myocardial edema with T2 mapping. Diagn Interv Imaging 2015;96:885—90.
Please cite this article in press as: Legou F, et al. Usefulness of T2 ratio in the diagnosis and prognosis of cardiac amyloidosis using cardiac MR imaging. Diagnostic and Interventional Imaging (2016), http://dx.doi.org/10.1016/j.diii.2016.08.007