Author + information
- Vandana Sachdev, MD∗ (, )
- Hwaida Hannoush, MD,
- Stanislav Sidenko, BA, RCS,
- Shahryar G. Saba, MD,
- Pamela Sears-Rogan, MD,
- W. Patricia Bandettini, MD,
- Alessandra Brofferio, MD,
- Sujata M. Shanbhag, MD, MPH,
- Cynthia L. Brenneman, RN, RCS,
- Keith A. Horvath, MD,
- Myron A. Waclawiw, PhD and
- Andrew E. Arai, MD
- ↵∗Cardiology and Pulmonary Medicine Branch, NHLBI, National Institutes of Health, 10 Center Drive, CRC, 5NE-room 1436, Bethesda, MD 20892-1650
Echocardiography (echo) is the primary modality for assessment of mitral regurgitation (MR). American Society of Echocardiography guidelines recommend integrating numerous parameters to grade severity. Cardiac magnetic resonance (CMR) can also quantify MR severity. Uretsky et al. (1) reported that echo and CMR are discordant in grading MR severity and that one-third of patients classified as severe by echo are only mild by CMR.
We prospectively compared echo and CMR measurements of MR severity in 50 subjects. This study was approved by the National Heart, Lung, and Blood Institute Institutional Review Board (NCT01063322). Echo images were acquired using a Philips IE33 system (Andover, Massachusetts) and analyzed using Prosolv cardiovascular version 4.0 (Fujifilm, Indianapolis, Indianapolis). MR severity was determined using qualitative and quantitative parameters according to American Society of Echocardiography guidelines (2). Left ventricle (LV) volumes were obtained using the Simpson biplane method (3). Regurgitant volume (RV) was calculated by proximal isovelocity surface area (PISA) measurement. Regurgitant fraction was calculated as (RV/stroke volume [LV end-diastolic volume − LV end-systolic volume]). Ultrasonography contrast (Definity, Lantheus, Billerica, Massachusetts) was infused when needed. CMR images were acquired with 1.5-T or 3-T scanners (Avanto, Aera, or Skyra models, Siemens, Erlangen, Germany). Cine CMR images in multiple parallel short-axis views were used to quantify LV stroke volume. Velocity-encoded cine-phase contrast CMR was used to quantify aortic flow and mitral RV ([LV stroke volume − aortic stroke volume]).
Patients ranged from 36 to 82 years of age (mean 58 ± 10 years of age; 52% female). The cause of the MR was degenerative in 30 patients (60%), ischemic in 17 (34%), and other in 3 patients (6%). Echo and CMR were completed on the same day in all but 1 patient, who underwent both studies within 24 h. Blood pressure was not significantly different between studies, but heart rate was lower during the echo (64 ± 13 beats/min vs. 67 ± 12 beats/min, respectively; p = 0.03). LV ejection fraction was higher on echo (59 ± 10% vs. 56 ± 11%, respectively; p < 0.0001), and LV volumes were smaller (LV end-diastolic volume index 77 ± 20 ml/m2 vs. 107 ± 26 ml/m2, respectively; and LV end-systolic volume index 32 ± 13 ml/m2 vs. 48 ± 20 ml/m2, respectively; p < 0.0001). RV could be calculated by PISA measurements in 47 of 50 patients and were similar between echo and CMR (40 ± 31 ml vs. 41 ± 26 ml, respectively; p = ns), whereas regurgitant fraction was larger on echo (47 ± 28 ml vs. 36 ± 14 ml, respectively; p < 0.0001).
MR severity using PISA RV alone compared well with CMR in contingency table analysis (Figure 1A). There was complete agreement in 30 of 47 patients (weighted kappa = 0.65; SE = 0.07; 95% confidence interval [CI]: 0.50 to 0.79). A two-grade discordance was seen in only 4 of 47 patients. MR severity using PISA RV alone also correlated well with CMR RV (r = 0.79; 95% CI: 0.64 to 0.88; p < 0.0001), as did effective regurgitant orifice measurements (r = 0.81; 95% CI: 0.67 to 0.89; p < 0.0001). The bias between PISA RV and CMR RV measurements was only 0.6 ml, suggesting no clear overestimation by either method, although limits of agreement were wide (−44 to 43 ml).
The echo integrated assessment of MR did not agree as well with CMR. Complete agreement was seen in less than one-half of patients (weighted kappa 0.46; SE = 0.08; 95% CI: 0.29 to 0.62) (Figure 1B), and a 2-grade discordance was seen in 9 of 50 patients, a problem most common in patients with functional MR.
Our finding of a single echo-derived measurement of RV correlating well with CMR RV is in agreement with previous work (4). A unique observation from our study is that the integrated assessment by echo worsens the agreement with CMR. Because this is the currently recommended approach to estimating MR severity by echo, further confirmation of this is warranted.
Uretsky et al. (1) found that echo correlated weakly with CMR (r = 0.6), overestimated RV, and did not predict post-surgical LV remodeling as well as CMR. Our correlation between techniques was better and did not show a systematic overestimation by echo. Although many factors could explain intermodality differences, when echo and CMR are performed on the same day and compare similar measurements, there is reasonable agreement between techniques.
Please note: This research was supported by the Intramural Research Program of the National Heart, Lung, and Blood Institute, National Institutes of Health/Department of Health and Human Services. Dr. Arai has a research agreement with Siemens for MRI imaging. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. (New Techniques to Evaluate Mitral Regurgitation [Mitral Valve Regurgitation]; NCT01063322).
- Uretsky S.,
- Gillam L.,
- Lang R.,
- et al.
- Cawley P.J.,
- Hamilton-Craig C.,
- Owens D.S.,
- et al.