INTRODUCTION

Ischemic mitral regurgitation (IMR) is mitral regurgitation (MR) due to complication of coronary disease and not fortuitously associated with it. Two types can be individualized. Acute MR secondary to ruptured papillary muscle is a rare but often fatal complication of myocardial infarction (MI) and will not be discuss here. We focus on functional MR, much more common, caused by ischemic myocardial alterations which occurs without any intrinsic valve disease.

QUANTITATIVE EVALUATION OF THE SEVERITY OF MITRAL REGURGITATION

IMR was often underrated because of poor sensibility of auscultation. Indeed, in contrast to organic mitral regurgitation in which there is a good correlation between the intensity of the systolic murmur and the degree of regurgitation, IMR is often silent. Echocardiography is now the method of choice for MR evaluation. Color flow imaging allows an easy visualization of the regurgitation and frequency of IMR is estimated between 15 and 20% after MI. However, if color imaging is very sensitive, it has major limitations for the assessment of severity of MR.1,2 We have pre-viously demonstrated that color imaging overestimates the severity of the regurgitation. In 170 patients (58 with functional MR and 112 with organic MR) we prospectively evaluated the severity of MR by the regurgitant jet area obtained by color imaging and the regurgitant volume (RVol)/regurgitant fraction measured by quantitative Doppler and quantitative two-dimensional echocardiography. Quantitative Doppler is based on measurement of mitral and aortic stroke volumes and quantitative two-dimensional echocardiography on left ventricular end-diastolic and end-systolic volumes (calculated by the biplane method of disks). Patients with functional MR had larger jet area (10.6 ± 5.3 vs 8.2 ± 5.3 cm²; P=.0006) but smaller RVol (28 ± 14 vs 55 ± 46 ml; P=.004). Thus, same jet areas correspond to smaller RVol in functional MR than in organic MR. In addition to the 2 above mentioned methods, the Proximal Isovelocity Surface Area (PISA) provides a third quantitative method to evaluate the severity of the regurgitation. The PISA method is based on analysis of the flow convergence proximal to the regurgitant orifice.3 This method is simple, fast reproducible and has been proven reliable by multiple investigators.3,4 The three methods allow evaluating the RVol, which indicates the volume overload induced by MR, and the effective regurgitant orifice (ERO), which measures the severity of anatomic lesions. An example of calculations of RVol and ERO using the PISA method is presented in Figure 1. ERO is the ratio of regurgitant flow (2 x Π xradius of the flow convergence2 xaliasing velocity) and peak regurgitant velocity. The RVol is calculated as follow: RVol=ERO x MR TVI. These quantitative methods provide the tools to reliably evaluate the prognosis and mechanisms of IMR.

Fig. 1. Quantification of MR by the PISA method. ERO is the ratio of the regurgitant flow (2 x Π xradius of the flow convergence2 xaliasing velocity) and the peak regurgitant velocity. RVol is calculated as follow: RVol=ERO xMR TVI. LV indicates left ventricle; LA, left atrium.

PROGNOSIS OF ISCHEMIC MITRAL REGURGITATION

Pioneering series have underscored the prognosis signification of IMR.5-7 In the acute phase of MI, IMR was frequent and carried an adverse prognosis but the acute MI by itself could have influenced the results. The SAVE study (Survival And Ventricular Enlargement) showed that mild MR was associated with high mortality. However, severity of MR was graded by angiography, the study excluded grade 3 or 4 MR and inclusion was limited to the 16 days after MI. Therefore the prognosis implication of large range of chronic IMR remained uncertain.

In a previous study, we have evaluated the consequences of IMR on hemodynamic (assessed by systolic pulmonary artery pressure (PAP) measured by Doppler).8 Although ejection fraction and end-systolic volume were significantly correlated to systolic PAP, correlations were extremely weak (Figure 2A). This result corroborates to the well-known poor correlation between ejection fraction and functional status. In contrast, systolic PAP was more strongly correlated to ERO (Figure 2B) and diastolic function (evaluated by the deceleration time). By multivariate analysis, these 2 variables were the strongest predictors of systolic PAP in association with age.

Fig. 2. Determinants of systolic pulmonary artery pressure (PAP) (measured by Doppler echocardiography) in functional mitral regurgitation. Correlation between systolic PAP and ejection fraction (EF) (A, left graph), end-systolic volume index (ESVI) (A, right graph), and ERO (B) measured by quantitative Doppler and quantitative two-dimensional echocardiography. (Taken from Enriquez-Sarano M8.)

Recently, we have demonstrated that, in the chronic phase after MI, IMR presence was associated with excess mortality independently of all baseline characteristics (age, gender, NYHA functional class, severity of the coronary disease...) and degree of ventricular dysfunction. At 5 years, survival of patients without mitral regurgitation was 61% but was only 38% in patients with MR (P<.0001) (Figure 3A). Even when the potential artificial ejection fraction increase due to MR was taken into account, IMR presence remained an independent predictor of mortality. Moreover, the mortality risk was directly related to the severity of the regurgitation. Patients with Rvo≥30 mL/beat demonstrated higher mortality than those with RVol<30 mL/beat (Figure 3B). Similarly, patients with ERO≥20 mm² displayed a higher mortality than those with ERO<20 mm². Of note, ERO was a stronger prognostic indicator than RVol. Therefore, IMR presence and severity are significant predictors of mortality and these results suggest that patients with Rvol≥30 mL/beat or ERO≥20 mm² should be more actively managed. This study also underlined that criteria for severe MR should be different for organic and ischemic MR. Thus, severe organic MR is defined by Rvol≥60 mL/beat and ERO≥40 mm²,9 whereas values of 30 mL/beat and 20 mm² should be considered for ischemic MR.

Fig. 3. Long term survival after diagnosis according to the presence of IMR (A) and the degree of MR graded by the RVol) (B).

MECHANISMS OF ISCHEMIC MITRAL REGURGITATION

IMR is not the result of organic mitral lesions but of incomplete closure of normal leaflets. Several mechanisms have been proposed to explain the occurrence of IMR. It is now clear that isolated ischemia of papillary muscle is insufficient to produce MR.10,11 Kaul and al demonstrated, in dogs, that selective ischemia of papillary muscle was ineffective to produce MR in contrast to global left ventricular ischemia. Conversely, left ventricular remodeling is of major importance. Several experimental studies or short series have suggested that more than left ventricular volume or ejection fraction, left ventricular shape was associated to IMR.12-15 Our group has recently demonstrated influence of local left ventricular remodeling on IMR. In 138 patients we prospectively evaluated: a) global left ventricular remodeling (volumes, stress and sphericity) and function; b) mitral valve deformation, and c) severity of the regurgitation. By multivariate analysis, main determinants of ERO were loss of annular contraction and mitral valve deformation or tenting (both P<.0001) (Figure 4A and 4B). This deformation is characterized by insufficient systolic leaflet body displacement toward the annulus, with coaptation limited to leaflet tips resulting in MR (Figure 5A to 5C). In turn, mitral tenting was directly determined by apical and posterior papillary muscle displacement. Global left ventricular size, sphericity, stress and systolic function had no or minimal additional independent association with the degree of IMR and tenting. Thus, in contrast to inferior MI, despite significant left ventricular enlargement, no MR is often observed in anterior MI because the relations between papillary muscles and mitral valve are not affected. This study also underlined the lack of influence of annular dilation on IMR. Ratio of leaflets to annular surface area is >2,16 and a considerable annular dilatation would be required to result in inadequate mitral coaptation. This was also nicely confirmed in a recent study.17 The authors compared annular size and degree of MR in 25 patients with lone atrial fibrillation and 24 patients with cardiomyopathy (ischemic or idiopathic). Although patients with atrial fibrillation and cardiomyopathy had similar annular size, significant MR only occurred in patients with cardiomyopathy. Therefore isolated annular dilatation does not cause significant MR. Animal experiments using three-dimensional echocardiography have also confirmed that IMR is related to changes of the geometry of the mitral valve apparatus and not to global left ventricular remodeling.18,19

Fig. 4. Determinants of ERO in IMR. Correlation between ERO and valvular tenting area (A) and annular contraction (B).

Fig. 5. Schematic representation of IMR mechanisms. Left: myocardial infarction without local left ventricular remodeling: no MR. Middle: local left ventricular remodeling (apical and posterior displacement of papillary muscle) leading to mitral valve tenting and MR. Right: increase of MR by additional loss of annular contraction.

CONCLUSION

IMR is a frequent complication of coronary disease. Its presence and degree have major prognostic implications and underscore the importance of its detection and quantification. A regurgitant volume ≥30 mL/beat and an effective regurgitant orifice ≥20 mm² defined a subgroup at high risk of mortality, which should be actively managed. Degree of IMR is associated to the loss of annular contraction and the mitral valve deformation or tenting, which is determined by the degree of local remodeling (apical and posterior displacement of papillary muscle) independently of global left ventricular remodeling and function. To date mitral ring annuloplasty is applied to correct IMR, most of the time in association with coronary artery bypass. However, this technique is only directed at reducing the annular size alone and may be ineffective to prevent long term recurrence of MR and to improve survival. Surgical corrections aiming at specifically correcting local alterations of left ventricular remodeling to minimize tenting have been proposed20-23 and should be further investigated.

Based on a presentation at the X International Symposium on Ischemic Heart Disease, Bilbao, Spain 22-24 May 2002.

Dr. Messika-Zeitoun was supported by a grant from the Federation Française de Cardiologie.

Correspondence: Dr. M. Enriquez-Sarano.

Division of Cardiovascular Diseases, Mayo Clinic,

200 First St. SW, MN 55905 Rochester, MN, USA.