Article Text

Download PDFPDF

Non-invasive imaging
Clinical indications for cardiovascular magnetic resonance
  1. Moises Vasquez,
  2. Eike Nagel
  1. Institute for Experimental and Translational Cardiovascular Imaging, Klinikum der Johann Wolfgang Goethe-Universitat Frankfurt, Frankfurt am Main, Germany
  1. Correspondence to Professor Eike Nagel, Institute for Experimental and Translational Cardiovascular Imaging, Klinikum der Johann Wolfgang Goethe-Universitat Frankfurt, Frankfurt am Main 60590, Germany; eike.nagel{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Learning objectives

  • To familiarise the reader with the advantages and established indications for cardiovascular magnetic resonance (CMR) imaging modules common in clinical practice.

  • To recognise the role of CMR in the assessment of chest pain or heart failure, and in recognising cardiomyopathies.

  • To update the existing conceptions about the limitations and contraindications of CMR.


Cardiovascular magnetic resonance (CMR) is a highly versatile non-invasive and non-ionising multi-parametric imaging technique. Within the multimodality imaging setting of current clinical practice, CMR provides relative strengths in different aspects of the clinical work-up (table 1). It has become the reference standard for the evaluation of cardiac volumes and function. Its perfusion module has positioned itself as an accurate imaging stress test with indications for the proof of haemodynamically significant coronary artery disease (CAD) in all major international guidelines. Nonetheless, its unique advantage and distinctive feature resides in its tissue characterisation capabilities, including late gadolinium enhancement (LGE) for visualisation of regional replacement scar, T1 mapping for diffuse myocardial fibrosis and T2 mapping for myocardial oedema/inflammation. This in vivo depiction of myocardial tissue characteristics broadens its diagnostic power beyond the dichotomy of the presence or absence of a particular disease (eg, CAD) to include several options to explain a patient’s clinical picture (eg, myocarditis, Takotsubo, pericarditis, microvascular disease), allowing for guidance of an individualised and optimised therapy.1

Table 1

Strength of each cardiac imaging technique

European Society of Cardiology (ESC) guidelines generally consider CMR as the best alternative cardiac imaging modality for patients with non-diagnostic echocardiographic studies.2–5 This review demonstrates, however, many situations where it seems appropriate to consider CMR as a first-line tool to differentiate the various underlying aetiologies of an abnormal finding or presentation and guide individual therapy.

In past years the use of CMR was restricted to historically established indications such as the evaluation of complex congenital heart disease,6 7 diseases of the thoracic vessels8 and pericardium,9 and cardiac tumours.10 For more information on these historical indications, the reader is referred to the above citations. Today, technical improvements allow common clinical scenarios to be approached by CMR. As such, in the EuroCMR registry the most important indications were risk stratification in suspected CAD/ischaemia, investigation of myocarditis/cardiomyopathies, as well as assessment of viability (figure 1).11 These indications represent more than 80% of the cases and deal mainly with the evaluation of chest pain or heart failure (HF).11

Figure 1

Main indications for cardiac magnetic resonance examination (based on EuroCMR registry11). CAD, coronary artery disease.

Stable chest pain

The traditional approach to chest pain consists of ruling in or out CAD. While a thorough history and clinical examination is supposed to discriminate cardiac from non-cardiac causes, the high rate of negative invasive coronary angiography (ICA) as well as a systematic overestimation of prevalence in large clinical trials demonstrate that this is frequently insufficient.12–14 Despite these limitations, calculating a pretest probability (PTP) of suspected CAD based on the pain characteristics, patient’s age and gender is considered clinically appropriate and recommended in the ESC guidelines. The PTP then defines the diagnostic pathway to establish the diagnosis of CAD,15 and patients with intermediate to high PTP can be examined with perfusion CMR as a class IB recommendation.15 Since vasodilators do not induce true myocardial ischaemia, perfusion CMR does not pick up true ischaemia but demonstrates regional hypoperfusion (figure 2).

Figure 2

Patient with relevant coronary artery disease. Stress perfusion cardiovascular magnetic resonance shows hypoperfusion in 4/16 segments compromising more than 10% of the myocardium. A perfusion defect during the first pass of the contrast agent is evident in the lateral wall from basal to apical segments (A,D,G), which is not present during rest perfusion (C,F,I), and in the absence of an ischaemic scar on late gadolinium enhancement images (B,E,H). Coronary angiography confirmed a significant obstruction in the circumflex artery (J).

The most recent National Institute for Health and Care Excellence (NICE) recommendations on the investigation of stable chest pain has moved away from PTP and promotes CT coronary angiography (CTCA) as the first-line test. If CAD is found, the next step proposed by the guidelines is an assessment of myocardial ischaemia to identify patients most likely to benefit from ICA and revascularisation.16 The Scottish CT of the Heart (SCOT-HEART) trial, which evaluated the benefit of CTCA in patients presenting with stable chest pain, was paramount for the updated recommendations. Of all the patients who underwent CTCA in the study, an obstructed coronary artery was identified in only 25% of cases. Consequently, the remaining 75% of patients were left without an answer for their chest pain.17 18 A CMR examination can frequently provide answers in patients with hypoperfusion to guide them towards ICA as well as patients without hypoperfusion to explain the underlying cause of the chest pain.

The recent MR-INFORM trial compared perfusion CMR with the reference standard of ICA with fractional flow reserve (FFR) to guide management of stable angina patients. Perfusion CMR was associated with a lower incidence of coronary revascularisation than FFR and was non-inferior for major adverse cardiac events, demonstrating the value of perfusion CMR as a first-line test for guiding downstream management.19

Specifically, CMR provides the following differentiated answers in patients with stable chest pain:

  • Prognostically significant CAD (figure 2). An ischaemic area of >10% of myocardium is associated with a high risk of cardiovascular death and of myocardial infarction (>3% per year), prompting ICA and revascularisation (class IC indication).15 In the MR-INFORM trial an ischaemic burden of 6% was used to guide patients towards revascularisation. While there is good evidence to show that patients with no ischaemia should not be revascularised, there is much less evidence on the contrary (hence the level of evidence C). The ISCHEMIA trial will hopefully clarify whether patients with moderate to severe ischaemia benefit from revascularisation (

  • Prognostically non-significant CAD. In these patients a low-risk ischaemic finding (<5% of myocardium) is demonstrated as a correlate for symptoms rather than increased mortality.20 Patients should receive optimal medical therapy.15

  • Microvascular disease. Due to its high spatial and temporal resolution, CMR perfusion allows differentiation of microvascular disease based on the inflow pattern of the contrast agent. Diffuse circumferential homogenous subendocardial perfusion defects are suggestive of microvascular disease.21

  • No myocardial ischaemia. Importantly, rather than just excluding functionally significant CAD, CMR frequently provides guidance towards the underlying cause of the patient's symptoms (eg, left ventricular hypertrophy (LVH) or myocardial inflammation), allowing a more holistic approach to chest pain than just focusing on the coronary arteries (figure 3).

Figure 3

Patient evaluated with cardiovascular magnetic resonance for suspected coronary artery disease. The stress perfusion was normal but late gadolinium enhancement imaging revealed an alternative diagnosis demonstrating pericardial enhancement with mild pericardial effusion, suggesting the diagnosis of pericarditis.

Acute chest pain

Life-threatening conditions such as acute coronary syndromes (ACS), aortic dissection or pulmonary embolism can present themselves as acute chest pain. Clinical history, ECG and raised troponin levels often substantiate the diagnosis of ACS. In most cases an ICA will confirm the initial suspicion finding obstructive CAD, but approximately 10% of cases have non-significant CAD.22 While CMR can characterise the extent and location of the myocardial infarction, detect haemorrhage and microvascular obstruction and add prognostic information,23 this will rarely change management. As such, CMR is rarely indicated after ICA. In contrast, in patients with myocardial infarction with non-obstructive coronary arteries (MINOCA), CMR is the key diagnostic tool for reaching the correct diagnosis and guiding management.22 The main differential diagnoses are myocarditis, myocardial infarction (either with spontaneous thrombolysis or type II) and Takotsubo cardiomyopathy, which represent 75% of the cases, if CMR is performed within 2 weeks after symptom presentation.24 If CMR is performed early the diagnosis is reached in 84% of cases, whereas if it is performed later its diagnostic yield drops to 57%, emphasising the importance of timely CMR.24


Diagnosing myocardial inflammation is challenging due to a heterogeneity of clinical presentations ranging from dramatic (even fatal) clinical scenarios to completely asymptomatic courses.3 CMR is the best non-invasive imaging modality to diagnose myocardial inflammation. The diagnostic performance of the original Lake Louise Criteria (LLC) depends mainly on LGE imaging and a wide range of accuracies have been reported.25–28 The presence of a typical LGE pattern (non-ischaemic, epicardial location, predominantly in the lateral wall) aids in the diagnosis and has been shown to bear prognostic relevance and poor outcome (figure 4).29 The novel mapping techniques outperform LLC, native T1 and T2 being especially important in the acute setting, native T2 demonstrating chronic inflammation and LGE detecting irreversible damage. Mapping techniques should be used wherever sufficiently implemented.27 28 30–32

Figure 4

Patient with perimyocarditis. (A, B) There is evidence of epicardial lateral non-ischaemic late gadolinium enhancement (LGE) with mild pericardial effusion and pericardial enhancement on the LGE images. A movie can be found online supplementary file 1.

Heart failure (HF)

According to HF guidelines, CMR is recommended in non-diagnostic echocardiographic studies for the measurements of volume and ejection fraction (EF) (class IC indication).2 CMR should be considered to differentiate between ischaemic and non-ischaemic aetiologies as well as assessing the presence of specific cardiomyopathies in HF with reduced EF (HFrEF) (class IIaC) as well as HF with preserved EF (HFpEF) (class IC).2 While the guidelines focus on the optimal pathways to establish a diagnosis, there is less focus on the aetiological work-up which should be a core part of the clinical assessment. In patients with HF with either reduced or preserved EF, CMR provides a new diagnosis in approximately 30%.33 34 Given the high number of specific causes as well as frequent reclassification of the diagnosis in these patients following a CMR study, we strongly recommend performing a CMR study early in the process to avoid missing reversible causes.

Imaging of left and right ventricular structure and function and tissue characterisation with LGE and T1/T2 mapping facilitate diagnosis of the underlying aetiology. Frequently the primary diagnostic dilemma in patients with HF is whether an ischaemic origin is the cause. CMR may be used as a gatekeeper to ICA in patients presenting with HF of unknown aetiology.35 Subendocardial or transmural LGE attributable to a vascular territory is regarded as ischaemic. The transmurality of the hyperenhancement can predict the likelihood of functional recovery with segments demonstrating LGE of more than 50% transmurality regarded as non-viable (figure 5).36 LGE is also seen in various non-ischaemic diseases that cause HF. While a full discussion of other underlying causes is beyond this review, table 2 summarises the major CMR features. Most patients present themselves with HF symptoms (and, commonly, preserved EF), arrthyhmias or an incidental finding of LVH. Whenever LVH is unexplained or exaggerated, a CMR should be performed to differentiate the various hypertrophic phenocopies.

Figure 5

Ischaemic late gadolinium enhancement. (A) The transmurality of the infarction in the inferior and lateral wall is less than 50%, suggesting preserved viability. (B) The infarction in the septal and anterior wall compromises more than 75% of the septal thickness indicating a low likelihood of preserved viability.

Table 2

CMR characteristics in common infiltrative and hypertrophic cardiomyopathies

Assessment of specific cardiomyopathies

CMR is ideally suited for the detection and characterisation of specific cardiomyopathies. The tissue characterisation parameters with LGE and T1/T2 mapping are particularly important. The non-ischaemic patterns of LGE are:

  1. Intramyocardial (linear/striae: dilated cardiomyopathy, myocarditis, sarcoidosis; patchy: sarcoidosis, hypertrophic cardiomyopathy (HCM); or at the right ventricle insertion points: HCM, pulmonary hypertension).

  2. Subepicardial (Fabry disease, myocarditis, sarcoidosis, Chagas disease).

  3. Global subendocardial or transmural (amyloidosis, systemic sclerosis, post-transplant).37

While T1 mapping is particularly important in amyloidosis where values are starkly increased and Fabry disease where they are particularly decreased,38 it also adds up in the diagnosis of other cardiomyopathies where native T1 is mostly increased such as HCM, where native T1 values can help in the differential diagnosis with hypertensive cardiomyopathy39 or athlete’s heart.40

The diagnosis of HCM is based on the presence of a maximal left ventricle wall thickness of ≥15 mm in the absence of any systemic disease known to cause increased wall thickness.4 LGE is frequent among patients with HCM (approximately 40%),41 and its extent is associated with an increased risk of sudden cardiac death (SCD).41 42 LGE occupying ≥15% of left ventricle mass proved to be an independent predictor, associated with a twofold increase in SCD risk for relatively young asymptomatic patients with HCM without conventional risk factors (figure 6).41 The latest 2014 ESC guidelines on HCM do not yet support the use of LGE for SCD risk assessment.4 The Hypertrophic Cardiomyopathy Registry (HCMR), which has recruited 2750 patients, should clarify whether LGE predicts the risk of SCD independently of other risk factors (

Figure 6

(A,B) Patient with hypertrophic cardiomyopathy with asymmetric anterior and anteroseptal hypertrophy and an associated non-ischaemic diffuse scar in the area of maximal hypertrophy. A movie can be found online supplementary file 2.

Echocardiography has a limited role in determining the precise aetiology in patients presenting with LVH, such as HCM, amyloidosis, sarcoidosis or Fabry disease, as the echocardiographic appearance of the left ventricle may be non-specific. CMR is particularly useful in this patient group to guide management. Cardiac sarcoidosis is seen in approximately 25% of patients with sarcoidosis without known cardiac disease.37 The combination of increased native T1 and T2 values might help to recognise active disease in a similar fashion as in myocarditis.43 The LGE can adopt almost any pattern, but is commonly patchy, multifocal, with a predilection for the basal septum and right ventricle. Treatment with corticosteroids, immunosuppression, hydroxychloroquine or tumour necrosis factor-α inhibitors may cause regression of cardiac sequelae.43

Fabry disease is a rare X-linked lysosomal storage disorder which presents with low T1 values due to accumulation of fat (ie, glycosphingolipid within lysosomes). These patients most commonly have concentric LVH and LGE in the basal inferolateral wall. Fabry can mimic any form of HCM including asymmetric and apical forms.44 Treatment can be offered with intravenous enzymatic replacement therapy.

Cardiac amyloid presents with a hypertrophied ventricle and a restrictive filling profile on echocardiography. CMR has a characteristic pattern of global subendocardial LGE coupled with accelerated contrast washout. Native T1 (6SD above normal) and extracellular volume (frequently more than 50%) values are markedly increased.45 The two main forms of amyloidosis are the light chain amyloidosis called AL (A for amyloidosis and L for light chain) and transthyretin ATTR (TTR for transthyretin) based on the protein being deposited. In AL, chemotherapy (similar to multiple myeloma) and, in ATTR tafamidis, a novel therapy that binds to transthyretin preventing tetramer dissociation and amyloidogenesis may be indicated.46

Another specific cardiomyopathy amenable to therapy is haemosiderosis. It occurs due to increased catabolism of erythrocytes in patients having frequent blood transfusions (eg, due to thalassaemia major). Cardiac involvement develops later in comparison with other organs but limits the prognosis. Iron shortens all three CMR relaxation times (T1, T2 and T2*). T2* at 1.5 Tesla has been established as the reference standard for myocardial iron overload assessment. Due to greater sensitivity and reproducibility, T1 values might have an added diagnostic value. T2* at 3 Tesla is not recommended but may be replaced by T1 and T2 mapping.47 Chelation therapy has changed the prognosis of the disease and is recommended when T2* values at 1.5 Tesla are below 20 ms.48

Survivors of SCD

SCD is a dramatic clinical presentation, most frequently as a consequence of an ACS. However, in at least one-third of survivors this is not the case. When ICA is inconclusive (either no identifiable culprit lesion or unobstructed coronary arteries), the diagnosis is challenging and deciphering the pathogenesis is fundamental for management and prognosis. CMR can identify the aetiology in 50–69% of cases.49–51 underscoring the importance of including CMR as a routine clinical investigation in patients at risk for or after survived SCD.

Arrhythmogenic right ventricular cardiomyopathy (ARVC) is a frequent cause of SCD in the young. Although it predominantly affects the right ventricle, biventricular or predominant left ventricular involvement is increasingly recognised. CMR is considered the imaging modality of choice in evaluating patients with suspected ARVC. The CMR criteria are based on the identification of right ventricle involvement defined as right ventricular dilatation, right ventricular systolic dysfunction and regional wall motion abnormalities (eg, wall thinning and microaneurysms).52 It is important to stress, however, that no imaging modality in isolation can diagnose ARVC, and other criteria either on family history, ECG, monitoring or biopsy are required.52

Contraindications and limitations of CMR

The main limitation for a more frequent use of CMR remains the lack of local expertise and equipment even though this has improved, especially in the UK.53

Arrhythmia was previously considered a problem for reduced image quality. However, newer and faster imaging sequences as well as improved arrhythmia rejection techniques have minimised this problem.

Nephrogenic systemic fibrosis (NSF) is a disease associated with the use of certain gadolinium-based contrast agents (GBCAs, groups I and III). The withdrawal of market approval for linear contrast agents in Europe after the report of brain accumulation has taken these linear contrast agents out of current clinical practice.54 55 The use of low-risk contrast agents (group II) remains approved and reasonable, even in patients with severely reduced renal function (ie, eGFR <30 mL/min).55 An American College of Radiology statement (updated June 2018) “considers the risk of NSF among patients exposed to standard or lower than standard doses of group II GBCAs is sufficiently low or possibly non-existent such that assessment of renal function with a questionnaire or laboratory testing is optional prior to intravenous administration”.55 Likewise, the guidelines by the Canadian Association of Radiologists considered the position that GBCAs are absolutely contraindicated in severe kidney disease as outdated.56 However, GBCAs should only be administered if deemed necessary and in the lowest dose for diagnosis and local policies need to be taken into account.55 The contraindication for linear agents in severely reduced kidney function remains.

Many patients have implanted cardiac electronic devices. These can be divided into 'MR conditional' and 'legacy devices'. Neither are safe in an MRI environment unless recognised and specific protocols followed after generator and lead type identification. Legacy device scanning is now known to be safe in nearly every circumstance,57 58 but clinical roll-out remains limited globally, requiring strong cardiology/radiology cooperation and significant effort, especially for pacemaker-dependent or ICD patients. In larger devices artefacts are frequent, although a number of strategies mitigate this.59 Neither stents nor mechanical valves represent contraindications for CMR examinations.


CMR provides several imaging strategies that provide diagnostic and prognostic value often significantly beyond current diagnostic pathways. This review puts CMR imaging into the clinical perspective.

Key messages

  • Cardiac magnetic resonance (CMR) detects significant coronary artery disease and, depending on the pattern and extent of hypoperfusion, can also guide to a more optimised medical or interventional therapy.

  • Particularly in patients with “myocardial infarction and non-obstructive coronary arteries” (MINOCA), CMR provides a high diagnostic yield (eg, of myocarditis) if applied early.

  • Once heart failure has been suspected in patients with dyspnoea, echocardiography is used to define the heart failure syndrome based on ejection fraction and search for an underlying aetiology. If it is non-diagnostic or the aetiology remains unclear, CMR should be applied to evaluate the presence of a specific underlying cardiomyopathy, including patients with unexplained reduction of left ventricular function, heart failure with preserved ejection fraction, and incidental severe left ventricular hypertrophy.

  • Many patients with scenarios previously thought to be absolute contraindications for contrast CMR can now undergo scanning under appropriate protocols (eg, arrhythmias, severe renal impairment and legacy devices).

CME credits for Education in Heart

Education in Heart articles are accredited for CME by various providers. To answer the accompanying multiple choice questions (MCQs) and obtain your credits, click on the 'Take the Test' link on the online version of the article. The MCQs are hosted on BMJ Learning. All users must complete a one-time registration on BMJ Learning and subsequently log in on every visit using their username and password to access modules and their CME record. Accreditation is only valid for 2 years from the date of publication. Printable CME certificates are available to users that achieve the minimum pass mark.



  • MV and EN contributed equally.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Obtained.

  • Provenance and peer review Commissioned; externally peer reviewed.