Author + information
- Received October 30, 2017
- Revision received March 26, 2018
- Accepted March 26, 2018
- Published online June 18, 2018.
- Tushar Kotecha, MBChBa,b,c,∗,
- Ana Martinez-Naharro, MDa,b,∗,
- Thomas A. Treibel, MBBSb,d,
- Rohin Francis, MBBSa,b,
- Sabrina Nordin, MBBSb,d,
- Amna Abdel-Gadir, MBBSb,d,
- Daniel S. Knight, MDa,c,
- Giulia Zumbo, MDa,
- Stefania Rosmini, MDd,
- Viviana Maestrini, PhDd,e,
- Heerajnarain Bulluck, MBBSb,
- Roby D. Rakhit, MDb,c,
- Ashutosh D. Wechalekar, MDa,c,
- Janet Gilbertson, MSca,
- Mary N. Sheppard, MDf,
- Peter Kellman, PhDg,
- Julian D. Gillmore, MD, PhDa,c,
- James C. Moon, MDd,
- Philip N. Hawkins, PhDa,c and
- Marianna Fontana, PhDa,b,c,∗ (, )@UCL@RoyalFreeNHS
- aNational Amyloidosis Centre, University College London, Royal Free Hospital, London, United Kingdom
- bInstitute of Cardiovascular Science, University College London, London, United Kingdom
- cRoyal Free Hospital, London, United Kingdom
- dBarts Heart Centre, London, United Kingdom
- eDepartment of Cardiovascular, Respiratory, Nephrology, Anesthesiology & Geriatric Sciences, “Sapienza” University of Rome, Rome, Italy
- fMolecular and Clinical Sciences Research Institute, St. George's, University of London, London, United Kingdom
- gNational Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland
- ↵∗Address for correspondence:
Dr. Marianna Fontana, National Amyloidosis Centre, University College London, Royal Free Hospital, Rowland Hill Street, London NW3 2PF, United Kingdom.
Background Prognosis in light-chain (AL) and transthyretin (ATTR) amyloidosis is influenced by cardiac involvement. ATTR amyloidosis has better prognosis than AL amyloidosis despite more amyloid infiltration, suggesting additional mechanisms of damage in AL amyloidosis.
Objectives The aim of the study was to assess the presence and prognostic significance of myocardial edema in patients with amyloidosis.
Methods The study recruited 286 patients: 100 with systemic AL amyloidosis, 163 with cardiac ATTR amyloidosis, 12 with suspected cardiac ATTR amyloidosis (grade 1 on 99mTc-3,3-diphosphono-1,2-propanodicarboxylic acid), 11 asymptomatic individuals with amyloidogenic TTR gene mutations, and 30 healthy volunteers. All subjects underwent cardiovascular magnetic resonance with T1 and T2 mapping and 16 underwent endomyocardial biopsy.
Results Myocardial T2 was increased in amyloidosis with the degree of elevation being highest in untreated AL patients (untreated AL amyloidosis 56.6 ± 5.1 ms; treated AL amyloidosis 53.6 ± 3.9 ms; ATTR amyloidosis 54.2 ± 4.1 ms; each p < 0.01 compared with control subjects: 48.9 ± 2.0 ms). Left ventricular (LV) mass and extracellular volume fraction were higher in ATTR amyloidosis compared with AL amyloidosis while LV ejection fraction was lower (p < 0.001). Histological evidence of edema was present in 87.5% of biopsy samples ranging from 5% to 40% myocardial involvement. Using Cox regression models, myocardial T2 predicted death in AL amyloidosis (hazard ratio: 1.48; 95% confidence interval: 1.20 to 1.82) and remained significant after adjusting for extracellular volume fraction and N-terminal pro–B-type natriuretic peptide (hazard ratio: 1.32; 95% confidence interval: 1.05 to 1.67).
Conclusions Myocardial edema is present in cardiac amyloidosis by histology and cardiovascular magnetic resonance T2 mapping. T2 is higher in untreated AL amyloidosis compared with treated AL and ATTR amyloidosis, and is a predictor of prognosis in AL amyloidosis. This suggests mechanisms additional to amyloid infiltration contributing to mortality in amyloidosis.
Systemic amyloidosis is a fatal disease characterized by progressive deposition of abnormal, insoluble protein fibrils in the extracellular space, which disrupts normal tissue architecture and function (1). The associated progressive increase in ventricular mass and stiffness (2) are thought to explain much of the pathophysiology. Nearly all cardiac amyloidosis is monoclonal immunoglobulin light-chain (AL) (or primary systemic) type or transthyretin (ATTR) (formerly senile) type, with cardiac involvement being the major determinant of survival in both. While there are currently no proven disease modifying therapies available for ATTR amyloidosis, the prognosis of AL amyloidosis is improved with chemotherapy that suppresses aberrant light chain production from the underlying plasma cell clone in bone marrow (3). However, despite novel chemotherapy agents, 1-year mortality in established cardiac AL amyloidosis remains high, reflecting late diagnosis and need for better tolerated, less toxic, and more rapidly acting therapies (4).
Cardiovascular magnetic resonance (CMR) can be used to noninvasively measure cardiac amyloid burden. CMR can visualize, with late gadolinium enhancement (LGE), and quantitate, with T1 mapping, the continuum of cardiac amyloid deposition (5). Transmurality of LGE, elevation in native T1 and extracellular volume fraction (ECV) all correlate with amyloid burden and provide incremental information on outcome (5). Amyloid infiltration in cardiac ATTR amyloidosis is usually more severe than in AL in terms of ventricular wall thickness, left ventricular (LV) mass, transmurality of LGE, and ECV (6). However, severity of clinical heart failure and survival are generally worse in AL amyloidosis (median survival from presentation 6 months in AL and 6 years in ATTR) (2,7). This discordance remains poorly understood but has been ascribed to additional toxic effects of AL amyloid or faster rate of amyloid deposition in AL amyloidosis compared with ATTR amyloidosis, leading to increased myocardial damage (7,8).
T2 is a CMR biomarker that increases in myocardial edema, for example, in acute myocardial infarction (9) and myocarditis (10), but also in heart failure, where it appears to track chronic inflammation (11) and cardiac Fabry disease, where T2 may be elevated in LGE areas and tracks blood troponin release (12). While T2 ratio has been assessed in cardiac amyloidosis (13,14), T2 mapping has not previously been studied. Furthermore, no histological studies have demonstrated the presence of myocardial edema in cardiac amyloidosis.
The aim of this study was to assess the presence of myocardial edema using histology and T2 mapping in patients with amyloidosis and determine the prognostic significance in AL and ATTR amyloidosis subtypes.
Patients with amyloidosis were recruited and underwent comprehensive assessment at the National Amyloidosis Centre, Royal Free Hospital, London, United Kingdom, from 2011 to 2015. Patients were systematically followed up from date of CMR until March 22, 2017, the date of censoring. Accuracy of occurrence and date of death among deceased patients, and ongoing survival among those who were censored, was ensured on the basis of UK death certification data from the UK Office of National Statistics. A total of 286 patients were categorized into 3 groups.
AL amyloid patients
One hundred patients with biopsy proven systemic AL amyloidosis (61 men, mean age 63.6 ± 10.9 years) were recruited. Cardiac categorization was based on CMR findings (5). Categories were: 1) cardiac amyloidosis with transmural LGE (features on CMR consistent with cardiac amyloidosis and transmural LGE); 2) cardiac amyloidosis with subendocardial LGE (features on CMR consistent with cardiac amyloidosis and subendocardial LGE); and 3) no evidence of cardiac involvement (normal wall thickness on CMR with normal serum biomarkers and no LGE). Patients within groups 1 and 2 were considered to have definite cardiac involvement. Patients were also classified as treated or untreated. Untreated patients were defined as those that underwent CMR scanning before initiation of chemotherapy to treat AL amyloidosis or associated myeloma. Treated patients were those who had completed or were currently undergoing chemotherapy at the time of CMR scanning.
ATTR amyloidosis patients
A total of 163 consecutive patients (139 men, mean age 74.6 ± 8.1 years) with cardiac ATTR amyloidosis and 12 with possible cardiac ATTR amyloidosis (8 men, mean age 73.3 ± 14.4 years) were recruited. Cardiac ATTR amyloidosis was defined as the combination of heart failure symptoms with echocardiography consistent with or suggestive of cardiac amyloidosis, grade 2 or 3 cardiac uptake on the 99mTc-3,3-diphosphono-1,2-propanodicarboxylic acid (99mTc-DPD) scintigraphy in the absence of a monoclonal gammopathy or, in the presence of monoclonal gammopathy, a cardiac biopsy positive for TTR (15). Possible cardiac ATTR amyloidosis was defined by grade 1 cardiac uptake on 99mTc-DPD. All subjects underwent sequencing of exons 2, 3, and 4 of the TTR gene.
TTR gene mutation carriers
Eleven TTR gene mutations carriers were recruited (4 men, 36%; age 45.6 ± 7.5 years). These were defined as individuals with TTR gene mutation but no evidence of clinical disease, no cardiac uptake on 99mTc-DPD scintigraphy, and normal echocardiography, N-terminal pro–B-type natriuretic peptide (NT-proBNP) and troponin T.
Thirty healthy volunteers (21 men, mean age 54.4 ± 12.3 years) were recruited through advertising in the hospital, university and general practices. All had no history of cardiovascular disease, hypertension or diabetes mellitus. All had normal 12-lead electrocardiography and normal CMR scan.
All amyloidosis patients underwent transthoracic echocardiography, 12-lead electrocardiography, and 6-min walk test where health and patient choice permitted (e.g., not performed in the presence of severe arthritis, postural hypotension, or neuropathy), and provided blood samples for NT-proBNP. The ATTR group also underwent 99mTc-DPD scintigraphy. Nineteen patients underwent clinically indicated endomyocardial biopsy (e.g., if biopsies from other sites had failed to confirm the presence and type of amyloidosis).
The University College London/University College London Hospital Joint Committees on the Ethics of Human Research approved the study, and all participants provided written informed consent.
Patients with standard contraindications to CMR or glomerular filtration rate <30 ml/min/1.73 m2 were excluded.
All participants underwent CMR on a 1.5-T scanner (Avanto, Siemens Healthcare, Erlangen, Germany). A standard volume and LGE study was performed. The gadolinium-based contrast agent used was 0.1 mmol/kg gadoterate meglumine (Dotarem, Guerbet SA, Paris, France). The LGE sequence used was either standard fast low-angle single-shot inversion recovery (FLASH-IR) or true fast imaging with steady-state free precession (true-FISP) sequence with phase-sensitive inversion recovery (PSIR) or magnitude reconstruction. For native T1 mapping, basal and midventricular short axis, and 4-chamber long-axis were acquired using the shortened modified Look-Locker inversion recovery (ShMOLLI) sequence after regional shimming, as previously described (16). Post-contrast T1 mapping was performed using the same sequence and slice positions. For T2 mapping, a 4-chamber long-axis matching the T1 map was acquired using an investigational prototype (WIP 448B, Siemens Healthcare, Erlangen, Germany). This sequence (17) uses 3 single-shot T2-prepared steady state free precession (SSFP) readouts each separated with 3 heartbeats for T1 recovery. The echo times (TE) for the individual T2 preparations were 0, 25, and 45 ms. A monoexponential fit to a 2-parameter model, S=PDexp(-TE/T2), was used at each pixel to estimate proton density (PD) and T2.
CMR image analysis
All analysis was performed offline. LGE was graded as none, subendocardial, or transmural as previously described (5). T1 and T2 measurements were performed by drawing a region of interest in the basal to mid septum of the 4-chamber map. For ECV measurement, a single region of interest was drawn in each of the 4 required areas: myocardial T1 estimates (basal to midseptum in 4-chamber map) and blood T1 estimates (LV cavity blood pool in 4-chamber map, avoiding the papillary muscles) before and after contrast administration. Hematocrit was taken immediately before each CMR study. ECV was calculated as: myocardial ECV = (1 – hematocrit) × (ΔR1myocardium/ΔR1blood), where R1=1/T1 (18).
Biopsies were obtained from 19 patients. All procedures were performed percutaneously under fluoroscopic guidance and samples obtained from the right ventricle. Samples were stained with hematoxylin and eosin (HE) and interpreted by an experienced histopathologist (M.N.S.) blinded to the CMR results. Observations were graded for presence and extent of edema. To confirm presence of amyloid, histological analysis was performed by Congo red staining on 6-μm formalin-fixed and paraffin-embedded sections, and viewed under bright-field, cross-polarized, and fluorescent light. Three HE samples were unsuitable for analysis due to insufficient material so edema assessment was performed for 16 samples. Samples were also labeled with monospecific antibodies (CD68 for macrophages and CD45 for leukocytes) to assess for myocardial inflammation, which was defined as ≥14 infiltrating immune cells/mm2 (10,19,20). Six historical myocardial biopsy samples stained with HE from patients with confirmed acute myocarditis were also analyzed for the presence of edema and inflammation.
Statistical analysis was performed using SPSS Statistics version 24 (IBM Corporation, Armonk, New York). All continuous variables were normally distributed (Shapiro-Wilk), other than NT-proBNP and C-reactive protein, which were therefore log-transformed for bivariate testing; continuous variables are presented as mean ± SD with nontransformed NT-proBNP and C-reactive protein as median and interquartile range. Comparisons between groups were performed by 1-way analysis of variance, after checking for homogeneity of variance, with post hoc Bonferroni corrected pairwise comparisons. Linear association between T2 and other markers was assessed using Pearson’s correlation coefficient for continuous variables and Spearman’s correlation coefficient for ordinal variables. Cox proportional hazards regression analysis with univariable and multivariable modeling was used to assess predictors of mortality. Considering the low number of events, to avoid overfitting, separate multivariable models were performed, each with T2 and 2 other clinically relevant predictors. The log-rank test and Kaplan-Meier survival analysis were used to compare groups with elevated and normal T2.
We enrolled 286 patients. Baseline characteristics are shown in Table 1. The TTR mutations were: V122I (n = 32), T60A (n = 22), V30M (n = 10), S77Y (n = 2), and E54G, E89K, D38Y, G47V, E89L, I84S, I107F, and L12P in 1 case each. Of the AL patients, 54% were new referrals and had not commenced chemotherapy at the time of CMR scanning (untreated AL amyloidosis). The remainder were undergoing or had completed chemotherapy (treated AL amyloidosis). Of these treated patients, 24 (52%) had a full response or very good partial clonal response, 16 (35%) had a partial response, and 6 (2%) had no response to treatment. The majority of patients were New York Heart Association functional class I or II (84% AL amyloidosis group, 81% ATTR amyloidosis group). Overall, 46% of patients were in sinus rhythm, 40% in atrial fibrillation, and 3% in atrial flutter. All patients had a heart rate of <110 beats/min at the time of CMR scan. Median time from CMR to cardiac biopsy was 72 days.
Sixty-five (65%) AL amyloidosis and 163 (88%) ATTR amyloidosis patients had definite cardiac amyloidosis. Compared to patients with cardiac AL amyloidosis, patients with cardiac ATTR amyloidosis had higher LV mass index (137 ± 33 g/1.73 m2 vs. 101 ± 35 g/1.73 m2; p < 0.001), lower EF (54 ± 13% vs. 64 ± 13%; p < 0.001), and higher ECV (0.63 ± 0.10 vs. 0.51 ± 0.10; p < 0.001) (Figure 1). There were no significant differences in native T1 or serum NT-proBNP.
T2 was increased in amyloidosis (AL and ATTR) compared with healthy volunteers and highest in treatment naive AL amyloidosis patients (untreated AL amyloidosis 56.6 ± 5.1 ms; treated AL amyloidosis 53.6 ± 3.9 ms; ATTR amyloidosis 54.2 ± 4.1 ms; control subjects 48.9 ± 2.0 ms; p < 0.01 for all pairwise comparisons except treated AL amyloidosis vs. ATTR amyloidosis) (Figure 2A). These differences remained present after exclusion of patients with no cardiac amyloid and TTR mutation carriers (untreated AL amyloidosis 58.5 ± 5.5 ms; treated AL amyloidosis 55.1 ± 5.5 ms; ATTR amyloidosis 54.5 ± 4.1 ms; control subjects 48.9 ± 2.0 ms; p < 0.05 for all pairwise comparisons except AL amyloidosis treated vs. ATTR amyloidosis). ECV was highest in ATTR amyloidosis (untreated AL amyloidosis 0.48 ± 0.13; treated AL amyloidosis 0.42 ± 0.10; ATTR amyloidosis 0.60 ± 0.13, 0.28 ± 0.03; control subjects; p < 0.01 for all pairwise comparisons except untreated AL amyloidosis vs. treated AL amyloidosis) (Figure 2B). In the overall population, and separately in AL and ATTR, there was weak association between T2 and markers of disease severity including native T1, ECV, extent of LGE, LV mass index, and NT-proBNP, and no correlation with New York Heart Association functional class, heart rate, or C-reactive protein (Table 2).
At follow-up (mean 22.8 ± 14.7 months), 75 (26%) of 286 patients had died, 28 (28%) in the AL amyloidosis group and 47 (25%) in the ATTR amyloidosis group. In AL amyloidosis, survival curves indicate that at 18 months there was approximately 88% chance of survival if T2 was <55 ms compared with 67% if T2 was more than 55 ms (p = 0.01) (Figure 3), while in ATTR amyloidosis there was no relationship between T2 and prognosis (p = 0.126). Using Cox regression models, T2 predicted mortality in AL amyloidosis (hazard ratio [HR]: 1.48; 95% confidence interval [CI]: 1.20 to 1.82; p < 0.001) and remained significant after adjusting for ECV and NT-proBNP (HR: 1.32; 95% CI: 1.05 to 1.67; p < 0.05) (Table 3). ECV and NT-proBNP also remained independently predictive of mortality. After removal of patients with no cardiac involvement, the model remained predictive (HR: 1.37; 95% CI: 1.10 to 1.70; p < 0.05; HR after adjusting for ECV and NT-proBNP: 1.35; 95% CI: 1.04 to 1.74; p < 0.05). Alternative models were explored and T2 remained predictive of outcome after adjusting for New York Heart Association functional class and E/e′ (left ventricular early diastolic filling wave/lateral mitral annulus velocity) (HR: 1.34; 95% CI: 1.09 to 1.64; p < 0.01), LV ejection fraction and LV mass (HR: 1.44; 95% CI: 1.16 to 1.80; p = 0.001), and LV ejection fraction and E/e′ (HR: 1.41; 95% CI: 1.13 to 1.76; p < 0.01). For ATTR, T2 was not predictive of mortality (HR: 0.84; 95% CI: 0.68 to 1.04; p = 0.104).
Of the 16 biopsy samples analyzed, 14 (87.5%) had evidence of myocardial edema on HE, defined as expansion of the interstitial space between myocytes (21,22) in the absence of fibrosis or amyloid. Increased space between myocytes was also observed with Congo red staining under bright field and fluorescent light in patients who had evidence of edema on samples stained with HE. The extent of edema was assessed by visual analysis and ranged from 5% to 40%. All samples had evidence of amyloid infiltration on Congo red staining and demonstrated apple-green birefringence under cross-polarized light. There was no correlation between visually assessed percent edema on biopsy and myocardial T2 (r = –0.265; p = 0.321). Figures 4 and 5⇓⇓ show 2 examples of patients with and without myocardial edema and their respective T2 maps on CMR.
All samples had presence of macrophages (CD68+) and leukocytes (CD45+), but none of the samples reached diagnostic criteria for inflammatory infiltration (≥14 infiltrating immune cells/mm2) (19,20). It was observed that clustering of leukocytes and macrophages occurred around areas of amyloid infiltration (Figure 6). All of the myocarditis samples showed evidence of edema on HE, which appears similar to that seen in the amyloid samples. The myocarditis samples also showed extensive infiltration of immune cells, reaching diagnostic criteria for inflammation in all cases (Figure 7)
The presence and degree of cardiac involvement in amyloidosis is the major determinant of survival (23,24). In this study, we demonstrate evidence of myocardial edema in amyloidosis on histology and using CMR myocardial T2. Patients with untreated AL amyloidosis show the greatest increase in myocardial T2. Myocardial T2 is predictive of prognosis in AL amyloidosis even when adjusted for ECV and NT-proBNP, but not in ATTR.
CMR is now considered the imaging tool of choice for diagnosis of cardiac involvement in systemic amyloidosis. Recently, interest has emerged in measuring amyloid burden using CMR to improve the risk stratification models and provide insight into pathogenesis (25–27), but a discrepancy has emerged with the degree of amyloid infiltration typically being more severe in ATTR amyloidosis but survival being worse in AL amyloidosis (5,28). This paradox suggests that additional mechanisms beyond amyloid infiltration may contribute to the greater mortality in AL amyloidosis, such as light chain toxicity, previously demonstrated in vitro (8), or faster rate of amyloid deposition. In the present study, we confirm this discrepancy and, with the use of histology and T2 mapping, have demonstrated evidence of myocardial edema, a possible additional mechanism of myocardial damage in AL amyloidosis. The use of T1 mapping, LGE, ECV quantification, and T2 mapping allowed us to visualize and quantify both myocardial infiltration and edema, adding this latter dimension to tissue characterization that has traditionally focused solely on amyloid infiltration.
In the present study, we demonstrate T2 elevation in both types of amyloidosis. The degree of T2 elevation was highest in AL amyloidosis patients before initiation of chemotherapy, with treated AL patients having lower T2 levels, comparable to those found in ATTR amyloidosis. ECV and NT-proBNP are known to be independent predictors of outcome in both AL and ATTR amyloidosis (3,29). We demonstrate that additionally, T2 is also a predictor of mortality in AL amyloidosis after adjusting for ECV and NT-proBNP, supporting an independent role for myocardial edema in outcomes.
Our histological analysis is the first to show high prevalence of myocardial edema in patients with cardiac amyloidosis. Edema is difficult to assess histologically, being defined as the presence of increased space between cells (21,22) due to accumulation of serous fluid. Animal studies with heart water content measured at autopsy have shown that myocardial T2 is elevated in edematous tissue (30). Acute myocarditis is a condition known to be associated with myocardial edema and inflammation. When compared with myocarditis biopsy samples, our amyloid biopsy samples show similar cellular separation (consistent with edema) but a much lower level of immune cell infiltration. Edema can be secondary to an inflammatory process, systemic or localized, but can also be present without inflammation. In line with previous in vitro studies, we suggest that in cardiac amyloidosis edema is associated with light-chain or fibril toxicity or differing rates of amyloid deposition and not secondary to an inflammatory process (8,31). The absence of active inflammation is in keeping with previous work showing that there is rarely significant inflammation on histology or evidence of systemic inflammatory reaction to amyloid deposits (32). This is also consistent with the low C-reactive protein levels and minimal histological evidence of inflammation in our cohort. There was no correlation between percent edema on biopsy and CMR-measured T2, and there are a number of reasons for this. There is currently no validated method to objectively quantify edema on biopsy samples and therefore assessment in this study was by visual analysis. It is known that amyloid deposition is very patchy in affected organs (33). We believe that edema distribution too will be nonuniform and it is therefore not possible to accurately correlate extent of edema within a biopsy sample taken from a small area of myocardium with T2 value measured over a much larger area, and a different location. Future studies with postmortem whole heart assessment of water content would be required to confirm this hypothesis.
Both the elevated myocardial T2 and histology of biopsy samples suggest the presence of myocardial edema. We suggest that in cardiac amyloidosis edema is present but less prominent than infiltration, and this is consistent with the range of T2 values observed, which are lower than typically seen in myocarditis and acute myocardial infarction where edema is a much more prominent feature (9,10). These results suggest that cardiac involvement in AL amyloidosis is a spectrum characterized by variable degrees of amyloid infiltration and superimposed myocardial edema, with ECV and T2 defining separate processes that both contribute to risk (Central Illustration, Figure 1).
Previous work on myocardial T2 signal in cardiac amyloidosis has reported conflicting results. Two studies (n = 35 and n = 36, respectively) reported that myocardial T2 ratio (between myocardium and skeletal muscle) was reduced in patients with cardiac amyloidosis (13,14), while 1 study (n = 12) using T2 mapping reported no difference between patients with cardiac amyloidosis and control subjects (34). Several reasons could be responsible for these findings. First, amyloid deposition can be present in the skeletal muscle and skeletal muscle T2 may be activity dependent (35), introducing a possible confounder when the T2 ratio between myocardium and skeletal muscle is used. Secondly, these studies all had small patient numbers with no separate analysis of AL amyloidosis and ATTR amyloidosis patients. Third, the traditional dark blood turbo spin echo techniques as used in the first 2 of these studies may be prone to myocardial signal variation and signal loss (36).
T2 mapping measurements may also be prone to technical confounders. T2 measurement using multiple single-shot T2-prepared SSFP images (17) uses 3 recovery beats. The estimate of T2 is slightly confounded by tissue T1 for 2 reasons. First, there is T1-dependent regrowth of magnetization following T2-preparation because a linear phase encode order is employed, causing a slight T1-dependent bias (37). When we simulated this effect using our protocol, T2 falls <1 ms for a 100-ms elevation in T1. Second, there is incomplete magnetization using 3 recovery beats reducing measured T2 slightly as T1 increases. It has previously been demonstrated that age and heart rate have little effect on measured T2 in healthy volunteers (17). Within our cohort, there was no correlation of heart rate, age, or hematocrit with T2.
While T1 is elevated in cardiac amyloidosis, the correlation between native T1 and myocardial T2 is weak. Native T1 is a measure of myocardial relaxation influenced by the extracellular and intracellular compartments, both free water and water bound to large molecules such as collagen or amyloid (38), whereas T2 is more specific for free water. We propose that the measured T1 signal in cardiac amyloidosis is a composite of amyloid burden and myocardial edema whereas T2 is more specific for edema, with the overall population comprising a spectrum with varying degrees of amyloid infiltration and myocardial edema.
Our findings have potential implications for clinical management. Currently, patient stratification by CMR is mainly based on parameters linked to amyloid infiltration such as LGE pattern and ECV. A multiparametric mapping approach has potential to change this. T2 mapping is a noncontrast technique, which could also be applied to patients with renal failure, although further studies will need to validate the role of T2 mapping in this setting.
Further work is needed to confirm the hypothesis linking myocardial edema to light-chain or AL amyloidosis fibril toxicity or the typically rapid rate of amyloid deposition in AL amyloidosis. Our model suggests that if light-chain production is “switched off” using chemotherapy, T2 may be expected to fall, tracking reduction in free light chains and brain natriuretic peptide. This is supported by the fact that untreated AL patients demonstrated higher T2 levels than did those established on treatment. Prospective follow-up studies assessing changes within patients during chemotherapy will be able to more fully assess this hypothesis and explore the role of T2 mapping to monitor treatment response.
A limitation of this study is that we did not have information on the causes of death, as the National Amyloidosis Centre assesses patients from throughout the United Kingdom. As a result, when patients die locally, only notification of death rather than cause of death is received. This is a single-center study, with no paired scans performed before and after chemotherapy. Furthermore, the T2 values may not have the same reference range in other CMR scanners. Finally, troponin was measured only in a minority of patients and therefore not included in the final analysis.
Myocardial edema is present in cardiac amyloidosis as confirmed on histology and measured by CMR with T2 mapping. T2 mapping provides unique information in AL amyloidosis, with elevation associated with increased risk of death and remaining an independent predictor of prognosis after adjustment for known prognostic factors. These findings support the concept of AL amyloidosis not being a disease of pure infiltration, but one in which additional mechanisms contribute to mortality; the study highlights the potential role of CMR with multiparametric mapping for evaluating this patient population.
COMPETENCY IN MEDICAL KNOWLEDGE: Myocardial edema, a common histological finding in patients with cardiac amyloidosis, can be measured by CMR T2 mapping and may be a useful tool in risk stratification.
TRANSLATIONAL OUTLOOK: Longitudinal studies are needed to clarify the role of T2 mapping in monitoring the response to treatment in patients with cardiac AL amyloidosis.
The authors are grateful for the contributions of patients, and the administrative and clinical staff at the National Amyloidosis Centre and the Heart Hospital.
↵∗ Drs. Kotecha and Martinez-Naharro contributed equally to this work and are joint first authors.
This study was supported by the National Amyloidosis Centre, University College London. Dr. Gillmore has served on the advisory board for GlaxoSmithKline/Alnylam. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- 99mTc-3,3-diphosphono-1,2-propanodicarboxylic acid
- light chain
- confidence interval
- cardiovascular magnetic resonance
- extracellular volume fraction
- hematoxylin and eosin
- hazard ratio
- late-gadolinium enhancement
- left ventricular
- N-terminal pro–B-type natriuretic peptide
- steady-state free precession
- Received October 30, 2017.
- Revision received March 26, 2018.
- Accepted March 26, 2018.
- 2018 American College of Cardiology Foundation
- Falk R.H.
- Palladini G.,
- Dispenzieri A.,
- Gertz M.A.,
- et al.
- Fontana M.,
- Pica S.,
- Reant P.,
- et al.
- Fontana M.,
- Banypersad S.M.,
- Treibel T.A.,
- et al.
- Rapezzi C.,
- Merlini G.,
- Quarta C.C.,
- et al.
- Brenner D.A.,
- Jain M.,
- Pimentel D.R.,
- et al.
- Bulluck H.,
- White S.K.,
- Rosmini S.,
- et al.
- Lurz P.,
- Luecke C.,
- Eitel I.,
- et al.
- Bohnen S.,
- Radunski U.K.,
- Lund G.K.,
- et al.
- Nordin S.,
- Kozor R.,
- Bulluck H.,
- et al.
- Legou F.,
- Tacher V.,
- Damy T.,
- et al.
- Gillmore J.D.,
- Maurer M.S.,
- Falk R.H.,
- et al.
- White S.K.,
- Sado D.M.,
- Fontana M.,
- et al.
- Caforio A.L.,
- Pankuweit S.,
- Arbustini E.,
- et al.
- Leeson T.S.,
- Leeson R.,
- Paparo A.A.
- Ghugre N.R.,
- Pop M.,
- Thomas R.,
- et al.
- Dungu J.N.,
- Valencia O.,
- Pinney J.H.,
- et al.
- Vogelsberg H.,
- Mahrholdt H.,
- Deluigi C.C.,
- et al.
- Diomede L.,
- Rognoni P.,
- Lavatelli F.,
- et al.
- Pepys M.B.
- Akçakaya M.,
- Basha T.A.,
- Weingärtner S.,
- Roujol S.,
- Berg S.,
- Nezafat R.
- Maestrini V.,
- Treibel T.A.,
- White S.K.,
- Fontana M.,
- Moon J.C.