Echocardiography remains the first-line imaging modality for the diagnosis of hypertrophic cardiomyopathy (HCM), but cardiac magnetic resonance (CMR) is increasingly used to assess phenotypic manifestations of the disease due to its high spatial resolution, absence of acoustic window limitations and ability to characterize tissue, in what has been termed a form of in-vivo histological assessment.1 It is particularly valuable in HCM, giving reliable information on parameters that have prognostic significance for which echocardiography can be inadequate, such as thickness of ventricular wall segments and the presence of apical aneurysms, as well as the degree of mitral regurgitation and abnormalities of the subvalvular apparatus.

One CMR technique that has aroused great interest is late gadolinium enhancement (LGE). This has been validated for the detection and quantification of myocardial necrosis,2 but also provides important information for the characterization of diseases of the myocardium, particularly in terms of patterns and distribution.3

The value of LGE in HCM has been the subject of intense debate. The myocardial fibrosis that is the hallmark of HCM is histopathologically diffuse in both replacement and interstitial forms, and is assumed to be a substrate for tachyarrhythmias and sudden death. While the latter is uncommon in the general HCM population, the subgroups most at risk have yet to be fully identified4; the currently proposed risk markers have low accuracy in predicting which patients are at risk.4,5

LGE on CMR, which essentially shows the focal distribution of fibrosis, and can present in a wide variety of patterns, is common in HCM patients (56–67% in large series6–8), and is also found in some asymptomatic mutation carriers. LGE is not currently considered a risk marker for sudden death, but might be considered a potential risk modifier, particularly when seen in extensive areas of the myocardium.4 What evidence is available on the clinical significance of LGE in HCM?

Firstly, it is known to be a valid indicator of fibrosis. LGE was initially validated in a post-mortem study of a patient with end-stage HCM, in which histopathological study revealed extensive fibrosis; the contrast agent was distributed in areas of fibrosis and increased collagen and there was a clear correlation between CMR and histological findings. However, this case – with severe ventricular dysfunction – is not representative of most HCM patients in clinical practice.9 Characterization of the histological mechanisms underlying LGE in patients with preserved ventricular function was hampered by the lack of an animal model of HCM. Recently, Moravsky et al.,10 analyzing myectomy specimens from patients with obstructive HCM, showed that LGE reliably reflects total myocardial fibrosis when appropriate analytical techniques are used, but they were unable to distinguish replacement from interstitial fibrosis, in which there is expansion of the extracellular matrix, which some studies suggest has different implications for the genesis of arrhythmias. Characterization of myocardial fibrosis in HCM by LGE is thus not easy, not only because of the difficulty in differentiating the two types of fibrosis, as described above, but also because areas of focal contrast deposition can be found adjacent to areas of diffuse distribution, which are beyond the spatial resolution of CMR.

Secondly, in terms of clinical significance, early studies showed that LGE is linked to the electrophysiological substrate, and is associated with ventricular arrhythmias on Holter ECG monitoring.11 However, its relationship to prognosis is still a controversial subject. The presence of LGE was associated with greater risk of death in two large studies of asymptomatic or mildly symptomatic patients: Bruder et al. found an independent risk of sudden death that was higher in patients with LGE than with traditional risk factors,6 while according to O’Hanlon et al., LGE was related only to a composite endpoint of adverse cardiac outcomes, but the risk was independent and increased with the extent of LGE.7 The high prevalence of LGE in HCM, in contrast to the relative rarity of sudden death in the general HCM population, has intensified the debate, and some argue that it is an insufficiently accurate marker to indicate the need for prophylactic implantation of a cardioverter-defibrillator.

In a study of 1300 HCM patients, Chan et al.12 found no association between the presence of LGE and risk of progressive heart failure symptoms/cardiovascular death and sudden death. However, a significant linear relation was evident between the extent of LGE and these endpoints, and it was independently associated with an increased risk of sudden death, even after controlling for traditional sudden death risk factors. LGE in ≥15% of the myocardium conferred a more than three-fold higher risk of sudden death. These findings imply that arrhythmic risk in HCM could be reclassified according to the extent of myocardial fibrosis.

It should be noted that other studies suggest that the extent of LGE is not related directly to the degree of ventricular dysfunction, and that patients with low-normal ejection fraction may have a similar degree of LGE to those with end-stage HCM and severe dysfunction.13

Despite the limitations of the study by Caetano et al. in this issue of the Journal,14 including the small population size, the fact that it was a convenience sample and hence subject to bias arising from referral for CMR, the older age of the patients compared to other series (and therefore greater likelihood of comorbidities), and their relative heterogeneity, the presence of LGE was associated with hemodynamically and morphologically more severe or advanced forms of the disease, as well as with traditional risk factors. This is in agreement with current thinking, based on large populations, that focal fibrosis identified by LGE on CMR has clinical significance.

New CMR techniques that enable mapping and quantification of diffuse fibrosis15 may add prognostic information, as well as widening understanding of the underlying mechanisms and improving individual clinical decisions.

Such a prospect suggests new research pathways for this poorly understood entity. It is to be hoped that innovative answers will be found to the challenging question of the relation between clinical phenotype and outcome in HCM, in areas ranging from genetics to cell biology and environmental interactions