Article Text

Download PDFPDF

Original research article
Echo-Doppler determinants of outcomes in patients with unoperated significant mitral regurgitation in current era
  1. Asim M Rafique1,
  2. Parham Zarrini2,
  3. Nirmal Singh3,
  4. Roy Beigel4,
  5. Rigved Tadwalkar1,
  6. Meshe Chonde5,
  7. Leandro Slipczuk2,
  8. Bojan Cercek2,
  9. Saibal Kar2 and
  10. Robert J Siegel2
  1. 1UCLA Medical Center, Los Angeles, California, USA
  2. 2Cedars-Sinai Heart Institute, Los Angeles, California, USA
  3. 3University of Arizona, Tucson, Arizona, USA
  4. 4Sheba—Tel Hashomer Hospital, Ramat Gan, Israel
  5. 5UPMC Presbyterian, Pittsburgh, Pennsylvania, USA
  1. Correspondence to Dr Robert J Siegel; siegel{at}

Statistics from

Key questions

What is already known about this subject?

  • Patients with uncorrected severe symptomatic mitral regurgitation (MR) have a significantly increased risk of morbidity and mortality. Several clinical and echocardiographic prognosticators of outcomes in patients with unoperated severe MR have been described in the past, but limited data exist in the current era.

What does this study add?

  • This study provides an updated analysis of echo-Doppler parameters in determining outcomes in patients with severe MR who are not candidates for surgery or percutaneous repair in the context of current American College of Cardiology (ACC)/American Heart Association (AHA) guidelines.

How might this impact on clinical practice?

  • This study will help clinicians to better risk stratify patients with severe MR and highlights that the current ACC/AHA guidelines provide limited incorporation of echo-Doppler parameters in the preoperative risk stratification of patients with severe MR.


Patients with uncorrected severe symptomatic mitral regurgitation (MR) have a significantly increased risk of morbidity and mortality.1–4 However, in the EuroHeart survey, approximately one-half of patients with severe symptomatic MR did not undergo surgery, most frequently due to comorbidities.5 However, there are no data from the current era, which addresses the natural history of patients with unoperated severe MR to determine the prognostic value of various echo-Doppler and clinical variables. The objective of our study was to identify clinical and echo-Doppler predictors of 5-year outcomes in patients with significant symptomatic MR who did not undergo surgical intervention in the current era.


In this observational study, we retrospectively evaluated the medical records of consecutive patients with significant MR (≥3+) from December 2005 through December 2008, who were considered high risk and did not undergo surgery during a 5-year follow-up and were not deemed to be candidates for the MitraClip procedure.6 ,7 We excluded patients with concomitant moderate or greater aortic stenosis or aortic regurgitation, and patients with prior percutaneous or surgical repair or replacement of the mitral and/or aortic valves.6 ,7 Hospital medical records and the social security death index were used to obtain and verify the 5-year mortality data. Patients were stratified into two groups as survivors versus non-survivors at 5 years of follow-up. The Institutional Review Board approved the study, and the waiver for patient consent was granted because of the retrospective nature of the study.

Two independent reviewers (RJS and AMR) evaluated the echocardiograms. Significant MR was defined as per the ACC/AHA guidelines, using a combination of qualitative and quantitative parameters including MR colour jet area to left atrial area ratio (JA/LAA) ≥40%, Doppler vena contracta width (VC) ≥0.7 cm, mitral regurgitant volume (Rvol) ≥60 mL/beat, regurgitant fraction ≥50%, effective regurgitant orifice area (EROA) ≥40 mm2, density and width of MR continuous wave (CW) jet, pulmonary vein flow (PVF) with systolic reversal, and enlarged left atrium (LA) and left ventricle (LV).6–8

Transmitral early (E) peak velocity, its deceleration time (DT) and transmitral atrial (A) wave velocity were obtained with a pulsed wave sample at the tip of the mitral valve.9 Early diastolic velocity of the lateral aspect of the mitral annulus (E) was measured by Doppler tissue imaging (TDI). PVF was obtained on an apical four-chamber view.9 ,10

The maximum velocity of the tricuspid regurgitation (TR) jet was measured using CW Doppler as per American Society of Echocardiography (ASE) recommendations.8 ,11 Right ventricle (RV) systolic pressure was estimated based on the modified Bernoulli equation and assumed to be equal to the pulmonary artery systolic pressure (PASP) in the absence of RV outflow obstruction. PASP was calculated by adding the tricuspid valve pressure gradient to the estimated right atrial pressure (RAP).11 RAP was considered as: 3 mm Hg for inferior vena cava (IVC) diameter <1.7 cm and ≥50% collapse with inspiration; 7 mm Hg for IVC diameter ≥1.7 cm and ≥50% decrease in the diameter with inspiration; 12 mm Hg if <50% collapse on inspiration; ≥15 mm Hg if the IVC was dilated without any collapse.8 ,11

Statistical analysis

Statistical analysis was performed with the statistical software program IBM SPSS V.21.0 (IBM Inc). Continuous data were presented as mean±SD. Categorical data were presented as an absolute number or percentages. Between-groups comparisons of baseline data were performed using the independent-samples t-test. All categorical variables were compared between the two groups using the Pearson χ2 test. Clinical and statistical variables were entered into Cox regression models to evaluate the independent predictors of 5-year survival. Kaplan-Meier analysis was used to estimate survival. The log-rank test was used to compare survival across two groups. A p value <0.05 was considered statistically significant. Given the retrospective and observational nature of the study, the sample size was not calculated.


Medical records of 83 patients with significant MR and available 5-year follow-up data were reviewed. Eight patients were excluded due to: prior mitral valve repair (MVR; n=2), MitraClip treatment (n=1), or MVR/mitral valve replacement (MVRe; n=3) during the follow-up period, or absence of a reliable TR signal on echocardiography (n=2). We included 75 patients (45 male) with a mean age of 72±11 years. The mean mitral EROA was 54±30 mm2, the mitral Rvol was 80±37 mL/beat, the VC was 0.6±0.2 cm and the mean mitral JA/LAA was 42.3±12 mm Hg. TDI of the lateral mitral annulus was available for 64 (85%) patients, and PVF was interpretable in 62 (83%) patients. The reasons for non-operability are shown in table 1.

Table 1

Reasons for not performing mitral valve surgery

Overall, 29 (39%) patients died during the 5 years of follow-up. As shown in table 2, non-survivors (n=29) compared with survivors (n=46) were older (77±9.8 vs 68±14 years, p=0.006); had more coronary artery disease (61% vs 35%, p=0.031), diabetes (21% vs 6.5%, p=0.057), pulmonary hypertension (71% vs 37%, p=0.004), atrioventricular block (14% vs 2.3%, p=0.073), higher New York Heart Association (NYHA) class (2.7±0.8 vs 2.3±0.8, p=0.037), higher brain natriuretic peptide (BNP) level (1157±717 vs 427±502, p=0.024, n=18) and a trend for a higher incidence of syncope (7.1% vs 0%, p=0.066). Overall, there was no significant difference in the prevalence of symptoms between the two study groups (93.1% vs 84.8%, p=0.281). Other comorbidities and medical treatment were comparable between the two groups.

Table 2

Comparison of baseline clinical variables between non-survivors and survivors

The use of β-blockers, ACE inhibitors or angiotensin receptor blockers, or spironolactone was significantly higher in non-survivors compared with survivors (87% vs 61%, p=0.033). The use of these medications tended to be more frequent in non-survivors compared with survivors both in functional MR (FMR; 100% vs 78%, p=0.134) and degenerative MR (DMR; 79% vs 56%, p=0.147). Patients on these medications tended to have worse: NYHA class (2.6±0.7 vs 2.1±0.9, p=0.044), left ventricular ejection fraction (LVEF; 58±17% vs 65±12%, p=0.182), forward stroke volumes (47±19 vs 71±42, p=0.051), PASP (43±16 vs 36±16, p=0.176), RAP (9±6 vs 8±6, p=0.782), BNP (896±1232 vs 491±195, p=0.561) and creatinine (2.1±2.9 vs 1.0±0.4, p=0.289).

As shown in table 3, non-survivors compared with survivors had: similar LV systolic (36.4±13.1 vs 34.3±9.2, p=0.448) and LV diastolic (55.5±9.7 vs 53.6±6.8, p=0.369) dimensions, with a higher prevalence of an LVEF<50% (20.7% vs 4.3%, p=0.026), lower LV stroke volume (49.5±21.8 vs 66.7±47.8, p=0.038), higher mitral E velocity (116.1±28.9 vs 99.5±30.7, p=0.024), higher E/E (12.7±4.6 vs 9.8±4, p=0.008), higher prevalence of E/E≥15 (32% vs 10.5%, p=0.034), higher PASP (52.6±18.7 vs 36.7±14, p<0.001), similar E/A≥2 (18.2% vs 26.3%, p=0.473), similar DT<140 ms (21.4% vs 10.9%, p=0.216), and more TR and RV dysfunction (26% vs 4%, p=0.035).

Table 3

Comparison of echocardiographic variables between non-survivors and survivors

Non-survivors compared with survivors had: severe MR (72% vs 70%, p=0.791), FMR (41% vs 22%, p=0.069), eccentric MR jet (64% vs 78%, p=0.242), mitral annular calcification (70% vs 36%, p=0.021), JA/LAA (72.4% vs 45.7%, p 0.023), VC≥0.7 (34.5% vs 22.2%, p=0.246), Rvol≥60 mL (58.6% vs 76.1%, p=0.11), EROA≥40 mm2 (51.7% vs 76.1%, p=0.029) and mean EROA (47±23 vs 59±32, p=0.074; table 4). Of the 22 patients with FMR, the 5-year mortality was 54% compared with 32% of 53 patients with DMR (p=0.032).

Table 4

Comparison of echocardiographic variables related to MV anatomy and function

Figure 1 shows the areas under the curve (AUC) from receiver-operator characteristic curve for the various parameters. Lateral E/E≥8 was 92% sensitive and 39% specific; lateral E/E≥15 was 32% sensitive and 89% specific; PASP≥50 mm Hg was 59% sensitive and 87% specific; PASP≥60 mm Hg was 28% sensitive and 89% specific; RAP≥10 mm Hg was 36% sensitive and 81% specific (AUC=0.62, 95% CI 0.48 to 0.76, p=0.098); mitral E velocity ≥120 cm/s was 43% sensitive and 80% specific (AUC 0.66, 95% CI 0.53 to 0.78, p=0.027); and mitral JA/LAA ≥40% was 66% sensitive and 60% specific (AUC 0.71, 95% CI 0.60 to 0.83, p <0.01) to predict 5-year mortality.

Figure 1

(A) ROC for the prediction of 5-year mortality in unoperated significant MR. AUCs and corresponding p values are shown for lateral annulus E/E (AUC=0.69, 95% CI 0.56 to 0.83, p=0.01). (B) ROC for the prediction of 5-year mortality in unoperated significant MR. AUCs and corresponding p values are shown for PASP (AUC=0.76, 95% CI 0.64 to 0.88, p<0.001). AUC, area under curve; PASP, pulmonary artery systolic pressure; ROC, receiver-operator characteristic curve.

Using Kaplan-Meier curves, the 5-year all-cause mortality was 67% in patients with E/E≥15 compared with 33% in patients with E/E<15 (p=0.008, figure 2A). Using Kaplan-Meier curves, the 5-year mortality was 74% for patients with a PASP≥50 mm Hg compared with 23% for patients with PASP<50 mm Hg (p<0.001, figure 2B).

Figure 2

(A): Five-year mortality in unoperated patients with severe MR with E/E≥15 was significantly higher compared with patients with E/E<15 (67% vs 33%, p=0.008). (B): Five-year mortality in unoperated patients with severe MR with PASP≥50 was significantly higher compared with patients with PASP<50 (76% vs 23%, p<0.001). DMR, degenerative MR; EROA, effective regurgitant orifice area; FMR, functional MR; MR, mitral regurgitation; PASP, pulmonary artery systolic pressure.

Patients with PASP≥50 mm Hg and E/E≥15 (n=8) had a 5-year mortality of 75% while patients with PASP<50 mm Hg and E/E<15 (n=38) had a lower 5-year mortality of 21%. Only seven patients had LVEF<50%, so further stratification was not possible. In the subgroup of patients with LVEF≥50%, patients with PASP≥50 mm Hg had a significantly worse survival than patients with PASP<50 mm Hg (22% vs 67%, p<0.001).

In the multivariate Cox regression analysis presented in table 5, the only significant predictor predictive of 5-year mortality was a PASP≥50 mm Hg (HR 2.2, 95% CI 1.4 to 3.3, p<0.001). Age, gender, NYHA class ≥3, LVEF<50%, E/E≥15, EROA≥40 and MR Rvol≥60, aetiology were not significant independent predictors in the multivariate model. Using the above parameters as continuous variables in the linear regression model, the significant predictors included PASP (p=0.001) and E/E (p=0.011).

Table 5

Cox regression model using clinical and echocardiographic variables


Patients with moderate to severe or severe MR who do not undergo surgical repair or replacement because of comorbidities have a high 5-year mortality. In the multivariate analysis, significant predictors of 5-year mortality were PASP and E/E. We did not find EROA, Rvol and VC to be predictive of mortality in patients with significant MR.

The likely mechanisms of increased mortality in patients with elevated and persistent PASP most likely include more advanced LV and LA remodelling in response to chronic volume overload from severe MR resulting in irreversible changes in the myocardium due to fibrosis and the permanent dysfunction of endothelium in the pulmonary vasculature. Elevated filling parameters in the setting of pulmonary hypertension therefore represent a more advanced disease stage where early intervention is warranted. Early optimisation of medical therapy for congestive heart failure (CHF) is prudent to improve pulmonary hypertension (PH) and hence outcomes in patients with MR by preventing late LV irreversible remodelling.12 The role of specific therapy for PH should be evaluated in patients with MR with an elevated transpulmonary gradient. MVR/MVRe is now an alternative treatment for patients with severe symptomatic MR who are not operative candidates.13–15 It is important to carefully evaluate the patients undergoing MVR/MVRe or MitraClip as increasingly more complex and sick patients are being referred for these procedures.16–18 Swaans et al19 showed that high surgical-risk patients with severe symptomatic MR treated with transcatheter MV repair show similar survival rates comparable to patients undergoing surgery, with both groups showing a survival benefit compared with conservative treatment. Appropriate risk stratification and proper selection of these high-risk patients and optimisation of filling pressures may improve outcomes of these patients with poor surgical options and with an increased mortality with medical therapy alone.


Patients with unoperated significant MR have a high mortality rate. Elevated filling pressures as determined by PASP and mitral E/E were most significant predictors of 5-year survival in patients with unoperated significant MR. Optimisation of filling pressure preprocedure or earlier intervention preceding development of irreversible changes may further improve outcomes in patients undergoing evaluation for MVR/MVRe or MitraClip.


This is a retrospective study; however, few studies have evaluated the prognosis of untreated MR in the modern era as most patients with severe MR undergo surgery or percutaneous intervention. This study is unique in providing an updated evaluation of echo-Doppler prognosticators in patients with severe MR. PASP was evaluated using echocardiography, which reflects the filling pressure under true resting conditions, avoiding the impact of analgesia or sedation and intravenous fluids on cardiac haemodynamics.


View Abstract

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.