Imaging biomarkers in aortic valve disease: it is time to shift the focus to the myocardium

This editorial refers to ‘Postoperative myocardial fibrosis assessment in aortic valvular heart diseases—a cardiovascular magnetic resonance study’, by L.T. Pires et al., https://doi.org/10.1093/ehjci/jead041. Significant aortic stenosis (AS) or aortic regurgitation (AR) is associated with adverse remodelling of the left ventricle (LV) that is accompanied by interstitial and/or replacement fibrosis. Interstitial fibrosis refers to the maladaptive increased production of extracellular matrix as a response of the myofibroblast to pressure and/or volume load while replacement fibrosis occurs due to irreversible myocyte death.1 Cardiac magnetic resonance imaging (MRI) with parametric mapping [T1 mapping, extracellular volume fraction (ECV), and indexed extracellular volume (iECV)] and late gadolinium enhancement (LGE) imaging is the non-invasive modality of choice for the assessment and quantification of interstitial and replacement fibrosis, respectively. This is clinically important as the degree of fibrosis is associated with clinical outcomes, including the need for urgent aortic valve replacement and mortality.2–4 Thus, quantifying the degree of fibrosis is a potential parameter to guide early intervention. However, whether intervention for AS or AR impacts the progression of fibrosis is not well known. Pires et al. report on their study that quantified the degree of myocardial fibrosis as evaluated by MRI in a prospective cohort of 99 patients with severe aortic valve disease (32% AR, 67% AS) and evaluated the impact of aortic valve replacement (AVR) on fibrosis.5 In summary, they showed that AR was associated with more diffuse fibrosis than AS as assessed by ECV and iECV; surgery was associated with reverse remodelling as assessed by LV mass and volume, decrease in the total amount of interstitial fibrosis as assessed by iECV, but no change in replacement fibrosis as assessed by LGE in both groups. The authors should be congratulated on this important study with potential clinical implications. It underscores the importance of quantifying myocardial fibrosis as a potential clinical parameter to guide the timing of aortic valve intervention. Aortic stenosis or AR is not just a disease of the valve. Rather, the accompanying pressure and volume overload lead to a secondary disease of the myocardium for which irreversible damage may occur if intervention is delayed. Unfortunately, intervention is often delayed until symptom onset which tends to be subjective at the early stage of the disease and therefore gives room for irreversible myocardial damage to occur. There is therefore a need for imaging biomarkers to serially monitor the health status of the myocardium at the tissue level and guide the timing of intervention. The iECV, calculated by multiplying the global ECV by the indexed LV end-diastolic myocardial volume and expressed in mL/m2, provides an assessment of the total extracellular space. It is worth noting that the authors use end-diastolic myocardial volume rather than end-diastolic ventricular cavity volume. This can also be calculated as iECV = ECV × (LV mass index/1.05) where 1.05 g/mL is the specific gravity of the LV myocardium.6 Indexed extracellular volume can be serially monitored to assess the growth of interstitial fibrosis, a marker of myocardial stress. As shown by the authors, the degree of interstitial fibrosis is inversely associated with LV ejection fraction. It is reasonable to assume that the development of fibrosis precedes and predict the decrease in LV ejection fraction and, therefore, may represent an important biomarker for subclinical LV dysfunction. However, it is important to note that when comparing the degree of interstitial fibrosis between the two types of aortic valve disease (stenosis vs. regurgitation) or serially over time, the iECV rather than ECV should be used. As shown by the author, the ECV (a more common parameter in clinical practice) which is often reported as the percentage of total myocardial volume is normalized to the myocardium and therefore may not reflect the absolute change in the degree of interstitial fibrosis when there is a concomitant change in cellular mass. For example, aortic valve surgery was associated with a decrease in the amount of interstitial fibrosis (as assessed by iECV) in both groups but no change in ECV in AR and paradoxical increase in ECV in AS (Figure 1). Hence, using ECV alone might have led to a wrong conclusion about the impact of surgery. More importantly, the development of replacement fibrosis on LGE indicates the onset of myocardial damage which cannot be reversed by aortic valve intervention and should, therefore, necessitate prompt intervention.7 The combination of iECV and LGE imaging has been proposed as potential risk stratification tool in AS patients.8 Illustration of the shortcoming of using ECV for serial assessment in two hypothetical patients with AR and AS. Both patients have same amount of interstitial fibrosis (iECV) before surgery and experience similar decrease (i.e. 5 mL/m2 less) in the amount of interstitial fibrosis after surgery. However, % ECV, which is normalized to the total myocardial mass volume, is stable in the AR patient and higher in the AS patient after surgery. AR, aortic regurgitation; AS, aortic stenosis; iCV, indexed total cellular volume; iECV, indexed extracellular volume; ECV, extracellular volume. The total end-diastolic myocardial volume can be calculated as the sum of the iECV and iCV. Using the formula in the in the text, the LV mass index (LVMi) can also be calculated. The LVMi decreased from 105 to 80 and 126 to 80 g/m2 in the AR and AS patients, respectively. In a prior multicentre cardiac MRI study, Everett et al.7 demonstrate in a cohort of 61 patients with asymptomatic AS that both interstitial fibrosis (assessed by iECV) and replacement fibrosis (assessed by LGE) progress rapidly over time in the absence of intervention. Somewhat similar findings had also been demonstrated in a cohort of patients with isolated chronic AR.3 Among patients with AS or AR, AVR is associated with reverse LV remodelling, including LV mass, and mechanics as assessed by echocardiogram.9 The postoperative structural change at the tissue level (as assessed by MRI) following AVR has been previously reported for AS patients by Treibel et al.10 Their findings have now been reproduced by Pires et al. In addition, the study by Pires et al. adds to the literature by showing that, similar to AS patients, interstitial fibrosis is reversible in AR after AVR but replacement fibrosis does not. As expected, ECV remain stable after AVR in AR patients due to balance regression unlike in AS patients in which there is disproportionate cellular regression and, therefore, post-AVR ECV is higher. There is however need for additional areas of research to move the field forward. Although it is reasonable to assume that fibrosis precedes LV dysfunction in AS or AR patients, this has not been serially documented in a longitudinal study. When this is proven to be true, it will provide credence for serial evaluation of fibrosis for early detection of myocardial dysfunction similar to the use of strain in echocardiography. Also, the optimal threshold for intervention based on the degree of interstitial and/or replacement fibrosis is not known. There is need for studies evaluating the optimal discriminatory threshold of iECV and %LGE above which the rate of adverse events significantly increased. A prior study had demonstrated a stepwise increase in the risk of death based on the presence of iECV and LGE: low risk with normal myocardium (defined as normal iECV and no LGE), intermediate risk with elevated iECV, and highest risk with the presence of LGE. However, this can be improved further using a combination of the quantitated values of these parameters to guide intervention. Lastly, sex difference in LV remodelling has been well documented in patients with AS and AR.11–13 However, there is a paucity of MRI data on sex differences in structural changes at tissue level in these patients. Such information will provide more clue to the mechanism of sex differences in LV remodelling as well as underscore the need for a sex-specific approach in the use of these MRI parameters to guide the timing of intervention. None declared.