Skip to main content

Echo vs. Cardiac MRI for Grading Mitral Regurgitation Severity

A comparison of echocardiographic-measured severity of mitral regurgitation with that of cardiac MRI-determined severity demonstrated poor agreement in the diagnosis of severe mitral regurgitation. Only cardiac MRI was predictive of left ventricular reverse remodeling after mitral valve surgery.

By Michael H. Crawford, MD, Editor

SYNOPSIS: A comparison of echocardiographic-measured severity of mitral regurgitation with that of cardiac MRI-determined severity demonstrated poor agreement in the diagnosis of severe mitral regurgitation. Only cardiac MRI was predictive of left ventricular reverse remodeling after mitral valve surgery.

SOURCE: Uretsky S, Animashaun IB, Sakul S, et al. American Society of Echocardiography algorithm for degenerative mitral regurgitation: Comparison with CMR. JACC Cardiovasc Imaging 2022;15:747-760.

Accurately identifying patients with severe mitral regurgitation (MR) is important clinically. The American Society of Echocardiography (ASE) has published guidelines describing the integration of various measures of MR severity to arrive at a final determination. However, this approach has not been validated against an independent comparator.

In a multicenter study, researchers enrolled 197 patients with MR who were imaged by echo and cardiac MRI. They excluded 27 patients with secondary MR, eight with atrial fibrillation, and 10 with incomplete data, leaving a study population of 152 (mean age = 62 years; 59% men with primary MR). The echoes were graded using the six key parameters recommended by the ASE guidelines.1 Four of the six parameters were required to diagnose definitely mild or severe MR. Cardiac MRI phase-contrast images were used to calculate aortic and pulmonary flow values, and the difference represented the MR volume flow. For comparison to echo, cardiac MRI MR volume was categorized as mild (< 30 mL), moderate II (30 mL-44 mL), moderate III (45 mL-59 mL), and severe (≥ 60 mL). Regurgitant fraction by cardiac MRI was divided into mild (< 30%), moderate II (30%-39%), moderate III (40%-49%), and severe (≥ 50%). The median time between the two studies was six days (range was 0-21 days). Average heart rate and blood pressure were not different between the two studies.

In patients with ASE definitely severe MR, only 52% showed cardiac MRI severe MR, and 10% showed mild MR by cardiac MRI. Of those with definitely mild MR by ASE, 100% showed mild MR by cardiac MRI regurgitant volume. By cardiac MRI regurgitant fraction, only 31% of those with definitely severe MR by ASE showed severe MR, and 7% demonstrated mild MR. Among those with definitely mild MR by ASE, only 60% were mild by cardiac MRI; the rest were moderate II. Also, in 42 patients who underwent mitral valve interventions (93% surgical, 7% clip), severe MR by cardiac MRI was an independent predictor of post-intervention reductions in left ventricular end-diastolic volume (i.e., reverse remodeling) at six months, whereas ASE severe MR was not. The authors concluded agreement between ASE-defined MR severity and cardiac MRI was suboptimal. Because cardiac MRI severity was predictive of left ventricular reverse remodeling after corrective interventions, cardiac MRI should be part of surgical decision-making in patients with degenerative MR.


Anyone who has tried to quantitate MR by echo should not be surprised by these findings. However, the magnitude of the discrepancies is sobering. Of those with definitely severe MR by echo criteria, less than half showed severe MR by cardiac MRI. This means echo parameters are markedly overestimating the severity of MR. Also, in patients who underwent surgical correction of presumably severe MR, only cardiac MRI severe MR was predictive of reverse left ventricular remodeling at six months. This suggests patients were operated on for lesser degrees of MR where positive left ventricular postoperative changes would not be expected. Of course, it does not mean negative left ventricular changes were not prevented by “earlier” surgery. The authors did not explore that concept here. Surprisingly, the echo quantitation of MR agreed more closely with cardiac MRI in eccentric jets compared to central jets of MR. Since eccentric jets tend to underestimate the severity of MR, perhaps this acted to correct the usual overestimation of severity by echo. Interestingly, central jets were better at estimating mild MR.

When individual echo parameters were considered, proximal isovelocity surface area (PISA) radius and vena contracta width were 85% sensitive for identifying cardiac MRI severe MR. Systolic flow reversal in a pulmonary vein was 75% sensitive, but the much-heralded flail leaflet was only 65% sensitive for detecting cardiac MRI severe MR. All the other echo measures showed sensitivities ≤ 50%. On the other hand, specificity was relatively good by all the echo measures. Also, if vena contracta width was ≤ 0.3 cm and PISA radius was ≤ 0.3 cm, MR almost always was mild by cardiac MRI. In addition, if the regurgitant jet was not holosystolic, no one exhibited severe MR by cardiac MRI.

There were several limitations to this study. First, although theoretically attractive, there is no definitive proof cardiac MRI is the gold standard for determining MR severity. In fact, cardiac MRI regurgitant volume and regurgitant fraction do not always agree on the severity of MR. I prefer regurgitant volume because it is not a ratio as regurgitant fraction is. With ratios, one never knows which variable is dominating the resultant value. Second, these investigators did not include Doppler echo regurgitant volume and fraction or effective regurgitant orifice area because they are not widely used and depend on other variables that may be inaccurate. Third, 3D echo and transesophageal echo were not included, yet many patients considered for mitral valve interventions undergo these procedures. Fourth, stress echo results were not pursued, although the response to exercise often is useful in gauging the severity of MR. Whether any of this information would have enhanced echo accuracy is unclear.

Since cardiac MRI is not widely available and may not be reimbursed, cardiologists must do the best job we can with the recommended echo measures and advanced echo techniques available while keeping in mind echo may overestimate the severity of MR.


  1. Zoghbi WA, Adams D, Bonow RO, et al. Recommendations for noninvasive evaluation of native valvular regurgitation. A report from the American Society of Echocardiography developed in collaboration with the Society for Cardiovascular Magnetic Resonance. J Am Soc Echocardiography 2017;30:303-371.