Impact of cardiac magnetic resonance on the diagnosis and management of patients with cardiomyopathies
Categoría del artículo: Original Article
Publicado en línea: 05 sept 2024
Páginas: 169 - 178
DOI: https://doi.org/10.2478/rjc-2024-0021
Palabras clave
© 2024 Oana-Andreea Popa et al., published by Sciendo
This work is licensed under the Creative Commons Attribution 4.0 International License.
CMR outperforms other cardiac imaging techniques in identifying the type of cardiomyopathy and assessing cardiac function. [6] It can discriminate between ischemic and nonischemic cardiomyopathy by the type of myocardial scar. [7–10] CMR, particularly the LGE technique, plays a pivotal role in evaluating cardiomyopathy, and now, it is recommended by the current guidelines as the method of choice. [11,12]
However, errors in identifying cardiomyopathy etiology persist. If the underlying pathology of myocardial disease is unrecognized or misdiagnosed, patients remain at risk for inadequate treatment, high risk for sudden cardiac death, and no adequately implemented family screening for specific cardiomyopathies. [13–15]
We hypothesized that CMR could reveal previously undetected diseases in patients diagnosed/suspected of having cardiomyopathies by echocardiography or other cardiac imaging techniques. Furthermore, we hypothesized that CMR would enable diagnoses in a significant subset of patients suspected of having various forms of cardiomyopathy and/or can lead to a change in management.
In addition, we aimed to evaluate which CMR parameters could have a significant clinical impact (SCI).
This is a retrospective cohort of patients with suspected cardiomyopathies who underwent multiparametric CMR in two units: Emergency Clinical Hospital “Prof Dr. Agrippa Ionescu” and Emerald Medical Center, Bucharest, Romania, between January 2021 and March 2022. Demographic and clinical data were collected during the CMR assessment, including age, gender, coronary risk factors, history of myocardial infarction, New York Heart Association (NYHA) functional class, treatment, cardiac imaging history and indication. The study was conducted in accordance with the Declaration of Helsinki. All patients provided written informed consent.
The main objective of the study was to evaluate the impact of CMR on the diagnosis and management of patients with cardiomyopathies.
The secondary objective of the study was to identify predictors that influence the diagnosis and management of patients with cardiomyopathies.
The primary endpoint of the study was the significant clinical impact (SCI), a binary variable (yes/no), defined as change/no change in the diagnosis or change/no change in clinical management.
We conducted all CMR studies using a 1.5 Tesla scanner (Siemens, Erlangen Germany). The imaging protocol included breath-hold cine sequences in long-axis planes and sequential short-axis slices. Intravenous gadolinium-based contrast agents were administered to all patients, and we evaluated the presence of LGE. The LGE images were acquired 10–15 minutes after contrast agent injection. Our comprehensive imaging protocol followed current guidelines and recommendations, covering all clinically indicated sequences for patients with cardiomyopathies. [16,17] We determined left ventricular (LV) end-diastolic and end-systolic volumes using a software plug-in within the OsiriX program. Manual delineation of endocardial and epicardial contours occurred during end-diastolic and end-systolic phases (see
Volumetric assessment of left ventricular (LV) and right ventricular (RV) function in dedicated software. It involves contouring the LV endocardium (in red), epicardium (in green), and RV endocardium (in purple) in a short-axis stack at end-diastole (ED) and end-systole (ES).
Additionally, we assessed the anatomy and function of great vessels and valves. We also evaluated T1 and T2 mapping and STIR (Short Tau Inversion Recovery) images. The supervising physician, in collaboration with a cardiologist and a radiologist, determined the specific imaging protocol.
The final CMR diagnosis of the different cardiomyopathies was made according to current international guidelines/consensus documents. (see
CMR criteria for the majority of cardiomyopathy diagnoses
| Indication | Structural Features (cine-SSFP) | LGE | Native T1 | ECV | T2 |
|---|---|---|---|---|---|
-LV dilatation and disfunction - diffuse regional wall motion abnormalities - +/_RV dilatation and disfunction |
- typical mid-wall pattern | -slightly diffuse increased | -slightly increased | - slightly increased, especially in inflammatory etiologies | |
-no LV dilatation with LV disfunction -normal LV systolic function with + LGE |
-nonischemic pattern | -normal -slightly increased |
-normal -slightly increased |
-can be increased in inflammatory etiologies | |
-none -regional wall motion abnormalities -global LV dysfunction -pericardial effusion |
-patchy subepicardial pattern | -increased in the oedema zone | -increased in the oedema zone | -regional or global increased | |
-hypertrophy (≥15 mm LV wall thickness): symmetrical, asymmetric, apical - other anomalies: LV crypts, papillar hypertrophy, SAM |
-mid-wall, usually in the areas where the wall is the thickest -insertion points of the septum to the RV level |
-increased | -increased | -normal -slightly increased in the burnout type |
|
-RV dilatation, dysfunction, -akinetic, dyskinetic aneurysms at the RV level, often affecting the RV triangle: sub tricuspid area, RV ejection tract, and RV apex -LV systolic dysfunction, regional wall motion abnormality |
- transmural RV region(s) (inlet, outlet, and apex) - subepicardial or mid-wall at LV lateral/inferolateral, septum, or both -subepicardial circumferential infiltration in those with predominantly LV damage |
-decreased in fat infiltration -increased in scar region |
-non-specific | - can be increased at the level of LGE in acute phases | |
-concentric hypertrophy -IAS, LA wall thickening - LV dysfunction -pericardial effusion |
-diffuse subendocardial -nulling effect of the myocardium -dark blood aspect |
-diffuse increased | -very increased | -non-specific pattern - can be increased in those with AL form |
|
-global systolic dysfunction -regional or global wall motion abnormalities, -bi-ventricular dilatation or hypertrophy. -slight increase in left ventricular mass secondary to granulomatous expansion |
-mid-wall or sub-epicardial focal fibrosis, -transmural or subendocardial, but without a correlation to a coronary territory -papillary muscles, RV-free wall, and the atria can also be involved. |
-increased in the scar zone | - increased diffuse | -increased in areas with granulomas | |
-noncompacted-to-compacted myocardium ratio of 2.3:1 -trabeculated mass > 20% - LV disfunction/dilatation |
-non-specific pattern -most frecvent subepicardial |
-slightly increase | -increased diffuse | -normal | |
| -concentric important LV hypertrophy | -mid-wall inferolateral -mid-wall septal |
-decreased - pseudonormal |
-decreased | -slightly increased, basal anteroseptal | |
-normal -LV dysfunction -regional wall motion defects |
-focal fibrosis predominantly at the level of the lateral and inferolateral walls | -increased diffuse | -increased diffuse | -hypersignal at the LGE level |
DCM- dilated cardiomyopathy, NDLVC- non-dilated left vetricular cardiomyopathy, HCM- hypertrophic cardiomyopathy, ARVC- arrhythmogenic right ventricular cardiomyopathy, LV- left ventricle, RV- right ventricle, LA- left atrium, AS- interatrial septum, AL- light chain amyloidosis. According to [14,18,25].
We characterized the nonischemic scar as mid-myocardial or sub-epicardial focal fibrosis, while the ischemic scar refers to the typical subendocardial to transmural scar confined within a specific coronary vascular territory.
We considered a CMR exam to have a significant clinical impact when, following the examination:
A patient received a new diagnosis after the CMR exam (e.g., a patient referred for hypertrophic cardiomyopathy, but the CMR exam found changes susceptible to cardiac amyloidosis). (see There was a change in management (e.g., a patient known with LV non-compaction cardiomyopathy was found to have a LV thrombus, leading to the initiation of systemic anticoagulation). (see
Cardiac magnetic resonance (CMR) exam of a 65-year-old female addressed for hypertrophic cardiomyopathy assessment, the contrast-enhanced CMR identifies changes susceptible to cardiac amyloidosis. (A) The diastolic phase of a balanced steady-state free precession (bSSFP) cine sequence of the short axis shows predominant septal hypertrophy; the basal antero-septum is 18mm thick, while the other walls are 12mm. (B) LGE acquired in the short axis shows difficult nulling of the myocardium, with areas of transmural hyperenhancement without respecting a coronary artery distribution. (C) T1 mapping imaging identified elevated T1=1100 ms and very high extracellular volume (ECV=48%). (D) T2 mapping imaging revealed diffuse edema, T2 =52-58 ms.
Significant clinical impact definition. ICD- implantable cardioverter-defibrillator, EP-electrophysiology, PCI-percutaneous coronary intervention, CABG-coronary artery bypass grafting, CT- computer tomography
A patient with both a new diagnosis and a change in management was considered only once for the assessment of SCI.
We assessed data normality using the Kolmogorov-Smirnov test. Continuous variables were reported as either the mean (standard deviation [SD]) or the median (interquartile range [IQR]). LV end-systolic and end-diastolic volumes were adjusted for body surface area.
The R Foundation for Statistical Computing, R Core Team (2024) was used for statistical analysis. This included A: A language and environment for statistical computing (R Foundation for Statistical Computing, Vienna, Austria. URL
The significance α in the study was 0.05, so p values below 0.05 were considered statistically significant. A univariate binomial logistic regression (simple and multiple) was used, the dependent variable being the presence/absence of significant clinical impact, and the dependent variables were a series of clinical and imaging parameters measured in these patients.
Between January 2021 and March 2022, 594 patients performed CMR examinations in our institutions. Among these, 146 patients were diagnosed with ischemic cardiomyopathy, 84 had a final normal CMR exam, and 87 had other changes such as severe valvular heart disease, congenital heart disease, cardiac tumors, or pericardial effusion and were excluded. Five patients under 18 years of age were also excluded. The final group analyzed consisted of 272 patients with cardiomyopathy.
The mean age was 49 years (14), and 65% of participants were men. All patients underwent a comprehensive clinical cardiological examination and transthoracic echocardiography before being referred for CMR. In terms of major cardiovascular risk factors, 55% of patients had hypertension, 51% had hypercholesterolemia, 3% had diabetes, 15% were active smokers, and 4% had a family history of cardiomyopathy. Only 0.5% of patients had evidence of coronary artery disease (prior myocardial infarction, percutaneous coronary interventions, or significative coronary lesions at angiography); they were included in the cardiomyopathy group due to the predominant nonischemic etiology.
Baseline characteristics
| Age, years | 49 (±14) |
| Sex (male), % | 65% |
| Body surface area, m2 | 1.99 (±0.2) |
| Family history of cardiomyopathy | 4% |
| 21% | |
| 65% | |
| 13% | |
| 1% | |
| Hypertension, % | 55% |
| Diabetes mellitus, % | 3% |
| Hypercholesterolemia, % | 51% |
| Smoking % | 15% |
| Alcohol consumption, % | 4.5% |
| History of myocardial infarction, % | 0.5% |
| Aortocoronary bypass operation, % | 0% |
| PCI, % | 0.5% |
| CKD, % | 5% |
| Sinus Rhythm, % | 93% |
| Atrial Fibrillation, % | 7% |
| Complete LBBB, % | 12% |
| Frequent PVC, % | 7% |
| 0.5% | |
| Echocardiography | 100% |
| Coronary angiogram CT | 1% |
| Coronary angiography | 20% |
| SPECT | 1% |
| Treadmill test | 2% |
| DCM/ NDLVC | 45% |
| HCM/ LVH | 18% |
| Cardiac amyloidosis | 5.2% |
| ARVC | 4 % |
| Myocarditis | 3.6% |
| Cardiac sarcoidosis | 1% |
| LV non-compaction/hypertrabeculation | 2% |
| Anderson-Fabry disease | 2% |
| Tachycardia-induced cardiomyopathy | 1% |
| Chemotherapy induces cardiomyopathy | 0.5% |
| Neuromuscular cardiomyopathy | 0.5% |
| Arrhythmia/SND substrate | 2.5% |
| Ischemic cardiomyopathy | 1% |
| Other cardiomyopathy | 10% |
Regarding cardiac function, the average indexed LV end-diastolic volume was 113 (± 47), the average LV ejection fraction was 46% (±15%) and the average RV ejection fraction was 55% (±11). All patients had contrast-enhanced CMR;
LGE imaging detected scars in 177 patients (65%). Among these, 164 patients (92.6%) had non-ischaemic scars, while 13 patients (7.3%) had ischaemic scars and only one patient had a combination of ischaemic and non-ischaemic scars. Most patients had a single scar, 91 patients (51.4%), while 38 patients (21.5%) had 2 scars, 14 patients (7.9%) had 3 scars and 34 patients (19.2%) had more than 3 scars.
CMR parameters
| Cardiac dimensions and function | |
|---|---|
| LVEDV, ml | 224 (±100) |
| LVEDVi, ml/m2 | 113(± 47) |
| LVESV, ml | 129 (±92) |
| LVESVi, ml/m2 | 64 (±43) |
| LVSV | 92 (±58) |
| LVEF, % | 46 (±15) |
| LV mass i, g/m2 | 68(±27) |
| RVEDV, ml | 168 (±58) |
| RVEDVi, ml/m2 | 81 (±29) |
| RVESV, ml | 89 (±27) |
| RVESVi, ml/m2 | 43 (±26) |
| RVEF% | 55 (±11) |
| Oedema n, (%) | 15, (4,2%) |
| Native T1 (ms) | 1030(±50) |
| T2 (ms) | 49 (±19) |
| Scar present n, (%) | 177 (65%) |
| Non-ischaemic scar n, (%) | 164 (92%) |
| Ischaemic scar n, (%) | 13 (7.3%) |
| Number of scars (%)
· 1 scar · 2 scars · 3 scars · >3 scars |
· 91 (51.4%) · 38 (21.4%) · 14 (7.9%) · 34 (19.2%) |
The most common five final diagnoses of the CMR examinations were dilated cardiomyopathy (97 patients, 35.5%), non-dilatated left ventricular cardiomyopathy (56 patients, 20.58%), hypertrophic cardiomyopathy (29 patients, 10.66%) and cardiac amyloidosis (12 patients, 4.5%).
In 44% of cases, new diagnoses were discovered based on CMR findings. The most common new diagnoses included non-dilatated left ventricular cardiomyopathy (21% of new diagnoses), dilated cardiomyopathy (9.6% of new diagnoses), arrhythmogenic right ventricular cardiomyopathy (6.7% of new diagnoses), myocarditis (5.4% of new diagnoses), coronary artery disease (5.4% of new diagnoses), LV non-compaction cardiomyopathy (5% of new diagnoses), LV thrombus (2.4% of new diagnoses), amyloidosis (2.4% of new diagnoses) and sarcoidosis (1.8% of new diagnoses). Less frequent diagnoses included muscular dystrophy-related cardiomyopathy and chemotherapy-induced cardiomyopathy.
The change of diagnosis after CMR examination. DCM- dilated cardiomyopathy, NDLVC non-dilated left ventricular cardiomyopathy, HCM- hypertrophic cardiomyopathy, LVH- left ventricular hypertrophy, ARVC- arrhythmogenic right ventricular cardiomyopathy
(A) Contrast-enhanced cardiovascular magnetic resonance diastolic frame of cine image in 4 chamber view shows a noncompact aspect of the myocardium. (B) The early gadolinium enhancement image in the same four-chamber view highlights apical thrombi at the left ventricle (LV) and right ventricle (RV) levels, indicated by yellow arrows
Over 47% of cases resulted in changes in patient management due to CMR findings.
CMR findings directly influenced catheter-based procedures in 27% of patients: ICD/CRTD decision was impacted in 59 patients (22%), coronary angiography decision was changed in 9 patients (3%) and electrophysiology study was recommended in 5 patients (2%)
Change in medication: started a new medication (e.g., anticoagulation for the presence of LV thrombus) in 38 patients (14%), stopped cardiac medication in 2 patients (1%) (see Figure 2)
Hospital admission after a CMR exam was recommended in 2 patients (1%).
Other non-invasive investigations (scintigraphy, computer tomography scan, specific blood tests like genetic tests, etc.) were recommended in 10 patients (4%).
In approximately two-thirds of cases, we observed a notable clinical impact, 66% overall, which included a new diagnosis in 44% of cases and a change in management in 47% of cases. A total of 25.7% of patients had both a new diagnosis and a change in management.
In the univariable analysis, none of the demographic parameters showed a SCI (see
Demographic Parameters. Predictor of significant clinical impact - univariable analysis
| Predictor | N | Event N | OR (95% CI) |
p-value |
|---|---|---|---|---|
| 272 | 180 | 1.00 (0.99 to 1.02) | 0.733 | |
| 179 | 121 | 1.20 (0.71 to 2.03) | 0.492 | |
| 272 | 180 | 1.22 (0.53 to 2.96) | 0.643 | |
| 57 | 41 | 1.39 (0.74 to 2.71) | 0.312 | |
| 151 | 102 | 1.16 (0.70 to 1.93) | 0.559 | |
| 140 | 95 | 1.18 (0.71 to 1.96) | 0.517 | |
| 12 | 10 | 2.66 (0.68 to 17.6) | 0.212 | |
| 40 | 28 | 1.24 (0.61 to 2.64) | 0.568 | |
| 1 | 29 | 19 | — | |
| 2 | 176 | 110 | 0.88 (0.37 to 1.96) | 0.755 |
| 3 & 4 | 36 | 29 | 2.18 (0.72 to 6.97) | 0.175 |
When evaluating CMR parameters with significant clinical impact, we found that left ventricular ejection fraction (LVEF), indexed left ventricular end-systolic volume (LVESVi), indexed left ventricular stroke volume (LVSVi), right ventricular ejection fraction (RVEF), and, notably, the presence of LGE were significant (see
CMR parameters. Predictor of significant clinical impact-univariable analysis
| Predictor | N | Event N | OR (95% CI) |
p-value |
|---|---|---|---|---|
| 272 | 180 | 0.97 (0.95 to 0.99) | 0.002 | |
| 272 | 180 | 1.01 (1.00 to 1.01) | 0.067 | |
| 272 | 180 | 1.01 (1.00 to 1.02) | 0.012 | |
| 272 | 180 | 0.98 (0.96 to 1.00) | 0.026 | |
| 265 | 175 | 0.97 (0.95 to 0.99) | 0.002 | |
| 177 | 113 | 1.84 (1.11 to 3.07) | 0.019 | |
| 243 | 157 | 1.00 (1.00 to 1.01) | 0.507 | |
| 254 | 167 | 0.99 (0.93 to 1.01) | 0.387 |
Predictors that had an influence p-value <0.10 in the univariable analysis were entered into a multiple univariate binomial logistic regression. Then, using a backward selection algorithm, the variables with the highest associated p-values and/or with the highest VIF (variance inflation factor - a measure of multicollinearity) were removed from the model and the analysis shows that regardless of the presence of LGE, a 1-unit increase in LVESVi is associated with a 1% increase in the odds of SCI. Also, irrespective of the LVESVi value, the presence of LGE was associated with a significant increase in the odds of SCI (OR 1.72) (see
CMR parameters. Multiple univariate binominal logistic regression
| Predictor | N | SCI N | OR (95% CI) |
p-value | VIF |
|---|---|---|---|---|---|
| LVESVi | 272 | 180 | 1.01 (1.005 to 1.02) | 0.021 | 1.0 |
| LGE | 177 | 113 | 1.72 (1.03 to 2.89) | 0.038 |
LV-left ventricle, ESV-end-systolic volume, LGE- late gadolinium enhancement
Our study aimed to assess the significant clinical impact of contrast-enhanced CMR in the diagnosis and management of patients with cardiomyopathies. We observed a change between initial and post-CMR diagnosis in 44% of cases, with a significant clinical impact in 66% of patients. These findings are consistent with previously published data (see Table 7).
Comparison of CMR’s significant clinical impact in current studies
| Study | Study period | Study patient | n | Change in diagnostic (%) | Change in management (%) |
|---|---|---|---|---|---|
| Bruder O et al. [28] | 2007-2009 | euro CMR registry (German pilot) | 11, 040 | 16.4% | 61.8% |
| Bruder O et al. [20] | 2006-2012 | SCMR registry | 27,301 | 8.7 % | 61 % |
| Roifman et al. [19] | 2013-2019 | Patients with HF indications from the SCMR registry | 3,837 | 49% | - |
| Abbasi et al. [21] | 2013 | Patient with LVEF < 50%. | 150 | 52% | 65% |
| Lin et al. [3] | 2004-2017 | Patients Undergoing Cardiac Transplantation | 338 | 23 (7%) | |
| Kangala et al. [22] | 2017 | Patient with HFpEF | 154 | 27% | |
| Witek et al. [29] | 2008-2017 | Patient with HF of unknown etiology | 243 | 38. 7% | 16.9% |
| Onciul et al. [30] | 2018-2020 | Patient referred for stress CMR | 120 | 15.85% |
Rofman et al. used the Society for Cardiovascular Magnetic Resonance (SCMR) registry and found a 49% change in diagnosis in 3,837 patients scanned for heart failure. [19] From the European Cardiovascular Magnetic Resonance Registry (EuroCMR), Bruder et al. reported a 16% change in final diagnosis in more than 27,000 consecutive patients referred for CMR. [20] In addition, Abbasi et al. performed a review of CMR studies in 150 consecutive patients with LVEF ≤50%. They found that CMR had an impact on 65% of patients, leading to a new diagnosis in 30% of cases and influencing management decisions in 52% of cases. [21] Kangala et al. evaluated the impact of CMR on diagnosis and prognosis in a group of patients with heart failure with preserved ejection fraction (HFpEF). In 27% of patients, CMR revealed previously undetected pathologies that correlated with a worse prognosis. [22]
Numerous publications have demonstrated the ability of LGE to differentiate cardiac pathologies based on enhancement patterns. An LGE pattern extending from the subendocardium to the epicardium, corresponding to the territory of a specific coronary artery, is typical for ischemic cardiomyopathy, whereas the absence of such a pattern has a 94% specificity for non-ischemic cardiomyopathy. [23, 24] To pinpoint the specific etiology of non-ischemic cardiomyopathy using CMR, clinicians recognize specific LGE phenotypes based on their locations and patterns. For example, diffuse transmural LGE suggests cardiac amyloidosis, whereas multifocal epicardial LGE involving the right ventricular aspect of the basal interventricular septum may indicate cardiac sarcoidosis. In the absence of LGE, the differential diagnosis narrows to idiopathic dilated cardiomyopathy, familial cardiomyopathy, stress cardiomyopathy, peripartum cardiomyopathy, and toxic (alcoholic or anthracycline-related) cardiomyopathy. In addition, newer techniques such as T1 mapping can help to identify non-ischemic cardiomyopathy etiologies such as cardiac amyloidosis and Anderson-Fabry cardiomyopathy [14,25]. Given the crucial role of LGE in diagnostic discrimination, it would be reasonable to assume that its presence would be associated with a notable clinical impact, in line with our findings and those of Abbasi et al. [21].
In our study, native T1 time was not found to be a predictor of clinical impact, probably due to the limited number of patients with cardiac amyloidosis and Fabry disease.
Also, LVESVi and biventricular systolic function were found to be predictors of clinical impact in univariable analysis, with LVESVi identified as an independent predictor. These findings may be attributed to the fact that a large proportion of patients had dilated cardiomyopathy, and the most significant impact on patient management was observed in decisions related to ICD/CRTD implantation.
CMR has a crucial role in patients with cardiomyopathies as early intervention with appropriate drugs (e.g., immunosuppression in sarcoidosis, iron chelators in hemochromatosis), and ICD implantation may improve the prognosis associated with these diseases. [26,27]
Our findings support CMR as a valuable imaging tool to identify the underlying cause of myocardial disease. Based on our real-life population study, we recommend that CMR be considered in all patients with suspected cardiomyopathies, in addition to specific tests.
We did not perform a cost-effectiveness analysis of the integration of CMR into clinical routine for cardiomiopathies evaluation, but the German pilot data from the EuroCMR Registry showed that the use of CMR in clinical routine evaluation does not increase the cost of patient care but, on the contrary, reduces costs by 11-65%. [28]
However, larger multi-center studies evaluating outcomes and cost-effectiveness are needed.
The study has several limitations. Most importantly, it is a retrospective study. We did not have data on treatment and downstream clinical outcomes. We recognize that evaluating changes between pre-test indications and post-test diagnoses serves as an imperfect surrogate for assessing management changes. Despite this, CMR’s unique ability to detect new or alternate myocardial disease processes sets it apart from other imaging modalities. Other groups have recognized this, using similar surrogates to evaluate the clinical impact of CMR.
Cardiac magnetic resonance imaging has revolutionized the diagnostic approach to different cardiomyopathies by providing detailed information on myocardial structure, function, and tissue composition. This study highlights its important impact on clinical practice in patients with cardiomyopathies by establishing correct diagnosis and prognosis, improving patient outcomes, and guiding personalized treatment strategies.