|Home | About | Journals | Submit | Contact Us | Français|
Background. Cardiac magnetic resonance (CMR) imaging has evolved over the last decade into an indispensable diagnostic instrument. CMR imaging noninvasively provides structural, functional and morphological information with high spatial resolution and an unlimited field of view. Since October 2006 the VieCuri Medical Centre in Venlo has a CMR scanner at its disposal.
Objectives. The goal of this study was to analyse the impact of CMR imaging on diagnosis and treatment in daily practice in the setting of a medium-volume peripheral hospital.
Methods. All patients who underwent CMR imaging between October 2006 and November 2008 were included in this analysis. The medical history before and after the CMR scan, the application form for CMR imaging and the outcome of the scans were reviewed. CMR images, obtained using a 1.5-T magnetic resonance imaging system, were reviewed by a multidisciplinary team.
Results. In 235 patients CMR imaging demonstrated one or more abnormalities, whereas CMR imaging did not identify any abnormalities in 148 patients. CMR imaging confirmed an expected finding in 166 cases, identified an unexpected condition in 69 cases, ruled out an expected finding in 59 cases and ruled out a suspected condition in 89 cases. Due to better insight into diagnosis, CMR imaging resulted in a change of treatment in 166 of the total of 383 CMR scans (43%).
Conclusion. In a relevant number of cases CMR imaging leads to a change in the treatment of a patient, proving the value of CMR imaging as a diagnostic modality. Therefore, CMR imaging is an excellent opportunity for peripheral medical centres to improve efficiency and the standard of patient care. (Neth Heart J 2010;18:524–30.)
Cardiac magnetic resonance (CMR) imaging has evolved over the last decade into an indispensable diagnostic instrument. CMR imaging noninvasively provides structural, functional and morphological information with high spatial resolution and unlimited field of view.1,2 In the beginning the use of CMR imaging was only available in high-volume peripheral and academic medical centres. Bruder et al. (2009) demonstrated in the European Cardiovascular Magnetic Resonance (EuroCMR) Registry that CMR imaging is a safe procedure and that in 62% of the cases, its results have a strong impact on patient management. 1 Since October 2006, the VieCuri Medical Centre in Venlo has a CMR scanner at its disposal. The goal of this study was to analyse the impact of CMR imaging on diagnosis and treatment in daily practice in a non high-volume peripheral hospital.
All patients who underwent CMR imaging since its introduction in October 2006 until November 2008 were included in this analysis. To adequately determine the impact of CMR imaging on the treatment of each patient, their medical history before and after the CMR scan, the application form for the CMR scan and the outcome of the scans were reviewed. The medical history of each patient was reviewed using the electronic patient file. All CMR images were reviewed by a multidisciplinary team consisting of cardiologists and radiologists. Diagnosis was based on consensus within this expert team.
CMR images were obtained using a 1.5-T magnetic resonance imaging system (Magnetom Symphony 1.5T, Siemens, Erlangen, Germany). To assess cardiac morphology, function, flow, perfusion tissue injury and fibrosis, several CMR imaging sequences were used. All procedures were performed according to the standardised Society for Cardiovascular Magnetic Resonance (SCMR) recommended protocols.3
Transmural extent of infarction (TEI), as demonstrated with delayed enhancement (DE) imaging, has been shown to be a powerful predictor of contractile response after revascularisation and medical therapy.4 We defined myocardial tissue as viable and the myocardial infarction as subendocardial if the TEI was ≤50%.5 If the TEI of a myocardial scar was >50% it was defined as a transmural myocardial infarction without viability.
Hypertrophic cardiomyopathy was suspected when a typical asymmetrical pattern of hypertrophy was present. The suspicion of hypertrophic cardiomyopathy was confirmed when the hypertrophic areas showed scar on DE imaging in a typical, patchy, midmyocardial, non-typical coronary artery distribution pattern found at the junction of the interventricular septum and right ventricular free walls.6,7 Myocardial scar in hypertrophic regions is typically found in more than 80% of patients with hypertrophic cardiomyopathy,5 and is associated with an increased risk of sudden cardiac death.6,8
Due to exceptional right ventricle morphological and functional characterisation, CMR imaging is considered the noninvasive diagnostic modality of choice for arrhythmogenic right ventricular dysplasia/ cardiomyopathy (ARVD/C).2 Black blood T1- weighed turbo spin echo (TSE) imaging sequences demonstrated fat and fibrous collagen deposition of the right ventricle, typical for ARVD.2
Typical epicardial to midwall hyperenhancement in a non-typical coronary artery distribution pattern on DE imaging was the main diagnostic feature to suggest myocarditis.6 We did not perform CMR scans in the acute phase.
In the literature, clinical evidence of cardiac involvement in amyloidosis and sarcoidosis (Besnier-Boeck-Schumann disease) is found in approximately 50 and 6% of patients, respectively, rationalising CMR imaging in patients with these autoimmune diseases.6
In the inclusion period, 383 patients were referred for CMR imaging. The average age of the patients was 57 years (SD±15). Of these 383 patients, 152 (39.7%) were female and 231 (60.3%) were male.
The indications for CMR imaging are presented in table 1. In several cases there was more than one indication per patient.
CMR imaging had diagnostic image quality in 97% of cases (373 out of 383 patients), which is comparable with the results of the EuroCMR Registry (98%).1 In five cases the images were unusable due to movement artefacts. In four patients the CMR examination was performed to analyse the anatomical position of a coronary artery or a bypass graft. However, the coronary CT scan turned out to be a better modality to visualise these structures. In the remaining instance the wrong imaging sequence was used. This was corrected afterwards by performing a second CMR scan.
The results of the 373 usable CMR scans are listed in table 2. The results are sorted by frequency of presence.
The number of patients with a previous diagnosis and compared with diagnosis by CMR imaging are presented in table 3. In 235 of 383 patients (61%) one or more abnormalities were demonstrated using CMR imaging. In 166 of 235 patients (71%) the CMR scans demonstrated a condition that had been diagnosed previously. In 69 of these 235 patients (29%), the CMR scan demonstrated one or more conditions that had not been diagnosed previously.
In 148 of 383 patients (39%) the CMR scan did not demonstrate any abnormalities (table 3). In 89 of 148 patients (60%) no previously diagnosed structural abnormality had been demonstrated. Fifty-nine of these 148 patients (40%) had a previously diagnosed abnormality but the CMR scan could not confirm any abnormalities.
In 104 patients (27%) CMR imaging was performed to analyse a cardiomyopathy of unknown origin. This figure is comparable with the Euro-CMR Registry (31%).1
These patients had decreased left ventricular ejection fractions (LVEF) on echocardiography or left ventricular angiogram, without a definite diagnosis. Therefore a diagnostic CMR scan was performed (figure 1), resulting in an explanation for the decreased LVEF in 61 patients (59%). In 12 of these 61 patients a completely new diagnosis was identified.
In 103 of 383 patients the CMR imaging was performed to establish the presence of a myocardial scar in an infarct pattern and/or viability assessment. In 48 of 383 patients, 56 myocardial scars were identified with DE imaging. In 12 patients one scar was newly diagnosed, whereas in three patients two scars were newly diagnosed. Forty-four patients who were previously diagnosed with a myocardial infarction did not demonstrate any late enhancement.
Of the 56 identified myocardial scars, 33 transmural infarctions were seen, ten of which were newly diagnosed. In 23 patients a subendocardial infarction with viability was identified, eight of which were newly diagnosed.
In a total of 79 patients the indication to perform the CMR scan was to investigate the presence of a hypertrophic cardiomyopathy without left ventricular outflow tract obstruction (HCM), a hypertrophic cardiomyopathy with left ventricular outflow tract obstruction (HOCM), or left ventricular hypertrophy (LVH). These diagnoses were suspected based on ECG characteristics, Holter registrations of ventricular tachycardias and echocardiographic findings (figure 2).
LVH, HCM and HOCM were newly diagnosed using CMR imaging in 15, seven, and no patients, respectively.
We adapted the term ‘arrhythmogenic focus’ for substrate identification of ventricular tachyarrhythmia. In 70 patients, CMR imaging was performed to identify a possible substrate of possible or proven ventricular arrhythmias (VT) or collapse. In 53 patients, no explanatory abnormalities were discovered. However, in 17 patients (24%) a possible arrhythmogenic focus was identified (figure 3). A completely new diagnosis was found in 14 of these 17 patients. In one case ARVD was identified.
In 45 patients the indication for the CMR scan was to analyse valvular disorders. In 31 patients a valvular disorder was established. Interestingly, in the medical histories before the scan there were 99 patients with one or more valvular disorders; CMR imaging identified significant valvular disorders in only 44 patients.
In 42 patients the indication to perform the CMR scan was analysis of possible aortic abnormalities. In 26 patients an aortic abnormality was observed. In total 29 patients with aortic abnormalities were observed. In nine patients, the aortic abnormality had not been diagnosed before.
In 33 patients the indication to perform the diagnostic CMR scan was to evaluate whether the patient was suffering from myocarditis. In five of 33 patients myocarditis was identified. More importantly in the 340 patients in which CMR imaging was not performed to search for myocarditis, it was newly identified in 17 patients.
In ten patients with proven systemic sarcoidosis or amyloidosis CMR images demonstrated no cardiac involvement in both systemic autoimmune diseases.
In 23 patients the indication to perform the CMR scan was to evaluate pericardial disorders. In seven patients a pericardial abnormality was present on CMR images. In total, ten patients with pericardial abnormalities were identified using steady-state free-precession (SSFP) imaging sequences.
Ten patients were analysed for the presence of a tumour in or nearby the heart. In two patients a tumour was found. Additionally, in the CMR scans not performed to establish the presence of a tumour, one patient was found to have a tumour. In one case an apical thrombus was identified, in one case a lipoma was identified using fat suppression T2 sequences, and in one case a tumour with malignant features of 16 mm in diameter was identified, most likely a primary cardiac sarcoma.
Due to better insight into diagnosis, CMR imaging resulted in a change in treatment in 166 of the total of 383 CMR scans (43%). Based on the findings of the CMR scan, in 75 patients the medication was changed, 15 patients underwent a coronary angiogram, whereas 17 patients specifically did not receive a coronary angiogram, since CMR images showed transmural infarction.5 Seven patients underwent a SPECT scan to try to establish myocardial ischaemia, since in those patients a subendocardial myocardial infarction was demonstrated. Four patients received coronary interventions and 14 patients were judged eligible for surgery based on CMR imaging. Twenty patients received an internal cardiac defibrillator.
Of the 166 patients, 85 patients underwent further diagnostics or were referred for consults. Thirtynine patients were referred to the heart team to discuss further percutaneous or surgical intervention, 28 patients were referred to clinical genetics, and seven patients were referred to a different specialism. In four patients the CMR scan lead to an MRI for non-cardiac diagnostics, three patients underwent a CT scan, two patients underwent an echocardiogram and two patients underwent a positron emission tomography-CT scan.
The present study showed that it is very useful to be able to perform diagnostics with CMR imaging in daily clinical practice in a non high-volume centre. As illustrated, CMR imaging is used as an additional diagnostic tool. In 43% of patients CMR imaging had therapeutic consequences. When compared with the EuroCMR Registry (62%) this might seem a relatively small percentage.1 However, in contrast to the EuroCMR, in our study population 46 patients (12%) underwent CMR imaging as part of another study for unrecognised myocardial infarctions. Furthermore, it is important to take into account the learning curve of a new CMR centre and/or the implementation of the appropriateness criteria for CMR imaging into clinical practice.9
Our results also show that in 166 out of 383 patients CMR imaging did not establish any abnormalities, although in 59 patients a previous diagnosis had been established previously. CMR imaging proved useful in ruling out a certain diagnosis, which occurred in 89 patients. In our cases of cardiomyopathy of unknown cause, CMR imaging identified an explanatory cause in approximately six out of ten patients. In approximately one out of four cases myocarditis was diagnosed. When CMR imaging was performed to find an explanation for possible or proven VT or collapse, in one out of four patients a possible substrate for VT was identified.
When comparing the aforementioned rates of identifying explanatory causes, there is a big difference between CMR scans performed for analysis of cardiomyopathy ECI and scans performed to find an explanation for possible or proven VT or collapse (60 vs. 25%, respectively). Both indications for CMR imaging were given the highest score of appropriateness in the ACC/AHA/ACCF/ACR/ SCCT/SCMR/ASNC/NASCI/SCAI/SIR 2006 appropriateness criteria.9 Nevertheless, for costbenefit reasons, the necessity to perform CMR imaging should be critically judged in each case. Liberal use is not restricted by side effects since CMR imaging is a modality without harmful ionising radiation. However, the contrast agent for DE imaging, gadolinium, does have possible side effects.
The process of implementation of the CMR imaging modality in the regularly used diagnostic modalities required considerable efforts from cardiologists, radiologists and the supporting staff. Cardiologists involved with CMR imaging were trained to become acquainted with CMR imaging and referring cardiologists were taught in which cases CMR imaging was the correct diagnostic modality. Interpreting CMR images required expertise from a multidisciplinary team consisting of radiologists and cardiologists. Since both radiologists and cardiologists contributed to interpretation of the CMR images, reimbursement is justified for both.
For patients, physicians and the hospital itself, the availability of a CMR scanner definitely had a significant extra value. It is convenient for patients to have all their diagnostics done in one hospital. For the hospital, it is of great importance to limit the number of patients that have to be referred to other hospitals. Also, a hospital that is able to deliver a wide range of diagnostics has a good marketing position in comparison with other hospitals. Finally, for cardiologists it is convenient to deliver high quality care without being dependent on other hospitals. In sum, implementing CMR imaging in our peripheral medical centre has lead to a properly exploited diagnostic modality.
Until recently CMR scanning was limited to academic medical centres and high-volume peripheral centres. However, the implementation of CMR imaging in a peripheral medium-volume hospital, such as the VieCuri Medical Centre, has proven its value as a diagnostic modality. CMR imaging leads to a change in the treatment in a relevant number of patients. Therefore, CMR imaging is an excellent opportunity for peripheral medical centres to improve efficiency and the standard of patient care.
This is the CVOI e-learning article. The author has prepared 10 questions which are available through the website of our Cardiovascular Educational Institute (CVOI). Please follow the instructions below.
After finishing the questions you will be asked to fill in your name, hospital and e-mail address; then press the button “verzenden”.
When 6 out of the 10 questions are answered correctly, you acquire 1 accreditation point granted by the Quality Committee of our Netherlands Society of Cardiology (NVVC). The acquired accreditation point will be credited in your personal file in the GAIA system. You will also receive an enclosure by e-mail with all the correct answers.
Please note that the questions of every article are available on the CVOI website for just one month until the publication of the next e-learning article (on the 10th of each month).
Over a period of one year 10 e-learning articles will appear in 10 subsequent NHJ editions. In every NHJ edition (except the July/ August edition) a special e-learning article will be published recognisable by a special icon. On an annual basis you can collect 10 accreditation points. The accreditation points are credited in the GAIA system by the CVOI.
If you need additional information, please contact the CVOI by mail: email@example.com or by phone: 030-2345001.
We invite all readers of the NHJ to actively take part in this new, instructive and stimulating way of e-learning.
E.E. van der Wall
Netherlands Heart Journal