Distinguishing Cardiac Amyloidosis and Hypertrophic Cardiomyopathy by Thickness and Myocardial Deformation of the Right Ventricle

Objectives To compare right ventricular thickness (RVT) and deformation of cardiac amyloidosis (CA) and hypertrophic cardiomyopathy (HCM) patients. Methods Sixty CA (mean age 58 ± 10 years; 33 males (55%)) and sixty HCM patients (mean age 55 ± 14 years; 27 males (45%)) were retrospectively enrolled. RVT, global radical peak strain (GRPS), global longitudinal peak strain (GLPS), and global circumferential peak stain (GCPS) were analyzed. To determine the cutoff values of the RVT and RV strain parameters for distinguishing CA from HCM, the areas under the receiver operating characteristic curve (AUCs) were analyzed. Results RVT of CA patients was significantly thicker than that of HCM patients (7.8 ± 2.1 vs 5.9 ± 1.3, p < 0.001). Moreover, significantly decreased RV-GRPS (12.1 ± 6.9 vs 23.5 ± 12.1, p < 0.001), RV-GCPS (−3.4 ± 2.2 vs −5.6 ± 3.5, p < 0.001), and RV-GLPS (−4.6 ± 2.3 vs −11.1 ± 4.9, p < 0.001) were observed in CA patients compared with HCM patients. RVT and RV strain demonstrate comparable diagnostic accuracy in differentiating CA from HCM. In particular, RV-GLPS combined with RVT showed the best performance for discriminating CA from HCM (AUC = 0.92, 95% CI: 0.85 to 0.96, p = 0.0001). Conclusions Right ventricular myocardial thickness and deformation of CA patients was more severe than HCM patients. RV-GLPS combined with RVT presents an excellent diagnostic performance in distinguishing CA and HCM.


Introduction
Cardiac amyloidosis (CA) is defined by the presence of extracellular amyloid deposition within the myocardium of the whole heart, leading to biventricular wall thickening with impaired relaxation and the loss of ventricular elasticity [1]. Due to the ventricular hypertrophy caused by amyloid deposition, CA has often been misdiagnosed as hypertrophic cardiomyopathy (HCM), which has main macroscopic characteristics of myocardial wall thickening and myocyte hypertrophy [2,3]. Clinically, the differentiation of CA from HCM is extremely important owing to the diverse therapeutic options and difference in long-term prognosis.
As an important differential diagnostic index, the structure and function of the left ventricle (LV) has been identified and shown to discriminate between CA and HCM to a certain extant [4][5][6]. ough right ventricle thickness (RVT) is extensively involved in CA but less in HCM [7], the differences in RVT and RV deformation were underestimated. Due to the high spatial resolution and difficult acoustic windows, cardiovascular magnetic resonance (CMR) imaging is now considered the gold standard technique for RV morphological study. Moreover, CMR tissue tracking (TT) technique, which could measure cardiac muscle motion and both LV and RV deformation, has emerged as more sensitive indicators than the ejection fraction (EF) [8][9][10][11][12]. us, the aims of this study were as follows: (1) to assess and compare RV thickness (RVT) and RV deformation parameters derived from the CMR-TT technique between CA and HCM patients and (2) to further identify the most valuable RV parameters for differentiating CA from HCM.

Study Population.
is study was approved by our institutional review board. We retrospectively studied 63 patients with CA (mean age 58 ± 11 years; range 25-81 years, 35 males [56%]) from 2015 to 2019. All patients were diagnosed with light chain amyloidosis. A diagnosis of light chain amyloidosis was made based on a biopsy of subcutaneous fat or an involved organ with the demonstration of typical Congo red birefringence, the detection of a monoclonal protein in the serum or urine and/or a monoclonal population of plasma cells in the bone marrow [13]. Furthermore, the diagnosis criteria of CA were based on a consensus opinion from the 10th International Symposium on Amyloid and Amyloidosis: LV wall thickness >12 mm without another known cause, as shown by echocardiography or cardiovascular magnetic resonance imaging [14]. Other than the bone marrow, the other tissue specimens for biopsy were obtained from the kidney (n � 13, 22%), liver (n � 1, 2%), and fat (n � 3, 5%). e exclusion criteria included (1) congenital heart disease (n � 0); (2) coronary artery disease (n � 1); (3) severe arrhythmia (n � 1); and (4) poor quality CMR images (n � 1). Finally, 60 CA patients (mean age 58 ± 10 years; range 25-81 years, 33 males [55%]) were enrolled.
We further included 60 patients with HCM (mean age: 56 years, range: 18-83 years; 27 males [45%]). e diagnostic criteria for subjects with HCM were based on the guidelines from the European Society of Cardiology [15]. All HCM patients enrolled in our study had nonobstructive hypertrophic cardiomyopathy, which can be divided into the following categories according to the segments of hypertrophic myocardium: (1) interventricular septal hypertrophic (n � 39), (2) anterolateral wall hypertrophic (n � 11), (3) posterior wall hypertrophic (n � 4), and (4) diffuse LV hypertrophic (n � 6). All of the patients were matched to CA patients in terms of maximum thickness of the LV segments. A total of 30 agematched healthy volunteers (mean age: 56 years, range: 24-80 years; 17 males [57%]) were enrolled as normal controls. e inclusion criteria for normal controls included no hypertension (blood pressure <140/90 mmHg), normal 12-lead electrocardiography (ECG), and no history or symptoms of cardiovascular disease or diabetes. All patients and normal controls underwent CMR imaging for morphologic and deformation analysis.

CMR Imaging Protocol.
CMR was performed using a 3.0-T whole-body scanner (Magnetom Tim Trio; Siemens Healthineers, Erlangen, Germany) with an 18-element body phased-array coil and ECG-triggering device during breath holding. Steady-state free precession sequences were performed to acquire consecutive short-axis cines covering the LV from the mitral valve level to the apex in 8 slices (TR/TE 37.66 ms/1.2 ms, flip angle 39°, FOV 280 mm × 373 mm, matrix size 146 mm × 280 mm, slice thickness 8 mm), while two-chamber long-axis and four-chamber cine series were acquired using the same sequences.

Imaging Analysis.
All image analyses were performed using commercially available software (cvi42; Circle Cardiovascular Imaging, Inc. Calgary, Canada). To measure cardiac function, endocardial and epicardial traces were performed manually in serial short-axis slices at the enddiastolic and end-systolic phases. Global LV/RV systolic function, including LV/RV end-diastolic volume (EDV), end-systolic volume (ESV), and LV/RV ejection fraction, were computed. LV and RV myocardial strain analysis was performed by loading the long-axis four-chamber and shortaxis slices into the tissue tracking module (Figures 1(a) and 1(b)).
e RVT was determined three times of the midventricular, and the average thickness was calculated (Figures 1(c) and 1(d)). e global feature tracking parameters were acquired automatically, including global radical peak strain (GRPS), global longitudinal peak strain (GLPS), and global circumferential peak stain (GCPS). Peak systolic strain rate (PSSR, maximum strain rate in absolute value over all phases starting from diastole until the next systole) and peak diastolic strain rate (PDSR, maximum strain rate in absolute value over all phases starting from systole until the next diastole) were also analyzed. Positive and negative symbols represent different directions of motion. As previously described [16], a normal RVT value was defined as ≤7 mm. According to the criteria, we divided CA and HCM patients into subgroups according to patients with RV hypertrophy (RVT >7 mm) or without RV hypertrophy (RVT ≤7 mm).

Biomarkers.
In CA and HCM patients, the blood was collected within two weeks of the CMR examination to measure myoglobin, creatine kinase isoenzyme, N-terminal pro-brain natriuretic peptide (NT-proBNP), cardiac troponin T, triglyceride (TG), and cholesterol (CHOL) levels.

Reproducibility Analysis.
To verify the reproducibility and reliability of RV functional assessments by CMR in our cohort, intra-and interobserver variability were calculated for the RV strain parameter measurements in 15 randomly selected CA patients and 15 randomly selected HCM patients. For intraobserver variability, the parameters were measured twice by the same observer with a minimum interval of two weeks between serial assessments. Interobserver variability was assessed using measurements obtained by a second independent observer blinded to all other analyses. ese values are presented as mean ± standard deviation (SD).

Statistics Analysis.
All statistical analyses were performed using a commercially available software package (SPSS for Windows, version 25.0, SPSS Inc., Chicago, IL; GraphPad, version 7.00, GraphPad Software, Inc., La Jolla, CA, USA; and MedCalc, version 9.3.0.0., MedCalc Software, Mariakerke, Belgium). All data were evaluated for normal distributions using the Kolmogorov-Smirnov test. Homogeneity of variance was evaluated using Levene's test. e results are expressed as mean ± standard deviation (SD) or median (interquartile range [IQR], 25%-75%). An independent samples t-test and Mann-Whitney U test were used to evaluate the baseline and strain parameters of the subgroups. To determine the cutoff values of the different RV parameters for diagnosing CA among patients with increased wall thickness, receiver operating characteristic curves were constructed, and the Youden index was used. Multiple receiver operating characteristic curves were compared based on the methodology described by Delong et al. to determine the differential diagnostic capacity [17]. Inter-and intraobserver variability for the RV strain parameters were assessed in 30 patients with CA (n � 15) and HCM (n � 15) using the Bland-Altman method, the results of which were presented as percentage mean bias ± SD and 95% confidence interval (CI). A two-sided p value < 0.05 was considered statistically significant for all tests. Table 1  TG, and CHOL were found between the CA and HCM groups.

Differences in Diagnostic Performance for Detecting CA or HCM.
e diagnostic performances of the RV parameters are displayed in Table 5. After calculating the sensitivity, specificity, areas under the curve (AUCs), and cutoff values, we found that the RV parameters, including the RVEF, RVT, and RV strain parameters, had AUCs of 0.68-0.92 for distinguishing CA from HCM. Figure 2 shows the comparison of the AUCs of RV parameters for detecting CA. Overall, the RV parameters, especially RV-GLPS combined with RVT, showed the largest AUCs (AUC 0.92, 95% CI 0.85-0.96, p � 0.0001) and balanced high sensitivity (sensitivity 81.4%, 95% CI 69.1-90.3) and specificity (specificity 89.8%, 95% CI 79.2-96.1). In contrast, traditional RV parameters such as RVEF showed a low diagnostic efficiency (AUC 0.68, 95% CI: 0.59 to 0.76, p � 0.0003). e receiver operating characteristic curves of different RV deformation indices for distinguishing between CA and HCM are shown in Figure 3. Figure 4, the average% of the difference of all measurement results is below 2.4, and most points fall within the 95% consistency limit. Among them, all points of intraobserver of RV-GCPS, RV-GRPS, and interobserver of RV-GLPS fall within the 95% consistency limit. 1 (3.3%) point of intraobserver of RV-GLPS did not fall within the 95% consistency limit. 2 (6.7%) point of interobserver of RV-GRPS and RV-GCPS did not fall within the 95% consistency limit.

Discussion
Since the prognosis and therapeutic options greatly differ between diseases, the differentiation between CA and HCM has always been a difficult problem in clinical practice. In the present study, we find that the degree of RVT was more severe in CA patients and the RV deformation derived from the CMR-TT technique showed a more significant decline in CA patients than in HCM patients. We further demonstrated        Table 4. e cross-   : Intra-and interobserver consistency analyses for right ventricular (RV) global radial peak strain (GRPS), circumferential peak strain (GCPS), and longitudinal peak strain (GLPS). e agreements of the RV strain parameters are shown as the percentage mean bias ± SD and 95% limits. e percentage mean bias of the intraobserver agreement was between −1.8%∼2.1% and that of the interobserver agreement was −2.2%∼2.4%. measurement of extracellular volume fraction also valuable tool for evaluation of CA [18]. However, different machines and renal failure make it difficult to diagnose CA. Bellavia et al. determined that RVT was more severe in advanced CA patients, and Doppler myocardial imaging measures of the RV can identify early impairment of cardiac function or stratify risk of death in CA patients [19]. Besides Doppler myocardial imaging, noncontrast CMR-TT based on routine cine images also could help evaluate the severity of myocardial deformation from various diseases with high spatial resolution [11,[20][21][22][23]. A recent literature also proves that CMR-TT is a reliable method for distinguishing between CA and HCM without administration of gadolinium-contrast [24]. eir study showed no significant differences between AUCs for the LGE pattern (0.994), LV GRPS (0.898), and GCPS (0.880) (all p > 0.109). e difference is that they compare the strain of LV; however, we compared the strain of RV. e study of Reddy et al. [25] proved that CMR can be a potent tool for accurate functional assessment of strain and strain rates involving both LV and RV for CA patients, including GRPS (39.7 ± 3.5 vs 13.6 ± 5.1), GCPS (−18.2 ± 1.5 vs −8.1 ± 1.7), and GLPS (−15.3 ± 0.9 vs −6.6 ± 1.4) of the LV and GRPS (−12.1 ± 1.6 vs −6.9 ± 1.2) and GLPS (−15.7 ± 2.1 vs −8.5 ± 1.4) strain of the RV. Nevertheless, their sample size is only 5, and normal people were used as control. Our results first revealed the difference of right ventricular thickness and strain between CA and HCM with reasonable cases. e application of tissue tracking has led to a deeper understanding of dysfunction processes in CA, both in the LV and RV.

Limitations
ere are some limitations in our study. First, a classification of HCM was not performed, which may cause some deviations when these results are applied for all HCM patients. But thinking of obvious differences in morphology between CA and HCM patients with LV outflow tract obstruction or apical hypertrophy, patients with interventricular septum hypertrophy or LV free wall hypertrophy were particularly difficult to diagnose with CA. erefore, only patients with interventricular septum hypertrophy or LV free wall hypertrophy were enrolled in our study. Second, CMR sequences that evaluated the changes in histology, such as the late gadolinium delayed enhancement, mapping, or extracellular volume sequences, were not included in the present study because it is difficult to identify in the RV to some extent.

Conclusions
As a promising method, CMR tissue tracking can be used for the quantitative analysis and early detection of subclinical RV deformation. Right ventricular myocardial thickness and deformation of CA patients were more severe than HCM patients. Most importantly, RV-GLPS combined with RVT presented the largest AUC and balanced a high sensitivity and specificity to differentiate between CA and HCM.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.