By Michael H. Crawford, MD, Editor

SYNOPSIS: The authors of a consortium observational study of patients with severe aortic stenosis undergoing valve replacement found that a pre-procedure cardiac MRI that detects myocardial fibrosis is predictive of post-procedure mortality and may be useful for making decisions about valve replacement in asymptomatic patients.

SOURCES: Musa TA, Treibel TA, Vassiliou VS, et al. Myocardial scar and mortality severe aortic stenosis. Circulation 2018;138:1935-1947.

Cavalcante JL, Sorajja P. Not too little and not too late. Circulation 2018;138:1948-1950.

Myocardial scar tissue has been identified as a critical factor in the transition between compensated and decompensated aortic stenosis (AS). Scar tissue can be detected by cardiac MRI (CMR) using late gadolinium enhancement (LGE). A United Kingdom consortium evaluated the utility of LGE-CMR for identifying long-term mortality in patients undergoing aortic valve replacement. The secondary endpoint was cardiovascular (CV) death. All 674 patients at six U.K. centers with severe AS who underwent a CMR and aortic valve replacement were included. Surgical aortic valve replacement (SAVR) was performed on 399 patients, and transcatheter AVR (TAVR) was performed on 275 patients. Patients with any LGE detected were more likely to be male, to have experienced a myocardial infarction (MI), and to exhibit more left ventricular (LV) dysfunction. After a median 3.6 years of follow-up, 22% of patients had died; almost half of these deaths were CV-related.

The authors evaluated 52 clinical parameters potentially predictive of mortality. LGE-CMR revealed that scar tissue was present in 51% of all study patients. In about one-third of these patients, the scar tissue was in an MI pattern. The factors independently associated with mortality were age (hazard ratio [HR], 1.5; 95% confidence interval [CI], 1.11-2.04; P < 0.01), Society of Thoracic Surgeons risk score (HR, 1.12; 95% CI, 1.03-1.22; P < 0.01), and scar presence (HR, 2.39; 95% CI, 1.4-4.1; P = 0.001). Also, the presence of scar tissue compared to no scar tissue independently predicted mortality (26% vs. 13%; P = 0.001) and CV mortality (15% vs. 5%; P = 0.002) regardless of the valve replacement procedure. Every 1% increase in scar burden was associated with an 11% higher mortality and an 8% higher CV mortality. The authors concluded that myocardial scar detected by LGE-CMR was independently associated with two-fold higher mortality in patients with severe AS after valve replacement regardless of the replacement technique.


Current intervention decision-making in AS patients is mainly concerned with symptoms, the severity of AS, and left ventricular (LV) systolic function. Since AS is largely a disease of older individuals, symptoms often are hard to define. Cardiologists assess AS severity mainly using echocardiography, which is complex and subject to potential errors. Reduced LV function probably is a late manifestation of AS, which increases the risk of replacement procedures.

Aortic valve replacement mortality is in the range of 1-2% by SAVR or TAVR. Thus, newer, more sensitive markers of when to replace the valve are needed. The results of previous observational studies have suggested that LV hypertrophy, abnormal exercise tests, nonsustained ventricular tachycardia on ECG monitoring, degree of valve calcification on imaging, or biomarkers such as BNP or troponin may identify higher-risk patients before classic indicators appear. However, none of these have reached wide-spread use or a class I recommendation in guidelines. Part of the reason for this slow uptake of new indicators for valve replacement is the lack of prospective randomized, controlled trials. The studies that have been conducted so far involved measuring a marker preoperatively and then seeing whether it predicts mortality after valve replacement. This study of LGE-CMR is of the same design, but it is compelling because the presence of myocardial fibrosis cannot be good.

Should clinicians start using LGE-CMR in decision-making for intervention in AS? Several considerations temper enthusiasm for adopting LGE-CMR for this purpose. First, high-quality CMR is not readily available outside specialized centers and is costly. Second, scarring is common; it was detected in half the patients. Also, in two-thirds of the patients with fibrosis detected, it was noninfarct-related. Many cardiac conditions can lead to patchy myocardial fibrosis, which makes relating it to AS potentially problematic. In fact, some have advocated using LGE positivity as a futility marker in AS. Perhaps its presence means a cardiologist waited too long to start treatment. Such advocates may have a point as almost one-quarter of the patients in this study were dead after 3.6 years of median follow-up. There are other limitations to this study. Selection biases existed; patients with pacemakers or implantable cardioverter-defibrillators and advanced renal disease were excluded due to the requirements for CMR. The authors did not analyze invasive hemodynamics or biomarkers, and there was no follow-up CMR. Also, myocardial fibrosis can occur due to aging alone. Some studies have shown that valve replacement is associated with new scar formation. In addition, investigators lumped SAVR and TAVR together; during the study period, TAVR techniques likely improved. Interestingly, mortality was higher among TAVR patients at all time points, but these data were not adjusted for age.

Currently, the use of any parameters outside of what is recommended in the guidelines is experimental. Fortunately, there are two randomized trials underway that may help inform the use of other markers for decision-making in AS in the future. One is a prospective evaluation of using newer indicators vs. the guideline recommendations (EVOLVED-AS). The other concerns the use of TAVR in asymptomatic patients with severe AS (EARLY-TAVR). Although the creators of EARLY-TAVR are not using CMR for decision-making, the trial will be important for establishing the validity of replacing the aortic valve in asymptomatic patients with severe AS vs. watchful waiting for guideline indications.