Secondary Logo

Journal Logo

Intraoperative Evaluation of Mitral Stenosis by Transesophageal Echocardiography

Cherry, Anne D. MD; Maxwell, Cory D. MD; Nicoara, Alina MD

doi: 10.1213/ANE.0000000000001276
Cardiovascular Anesthesiology
Free
SDC

From the Department of Anesthesiology, Duke University School of Medicine, Durham, North Carolina.

Accepted for publication February 8, 2016.

Funding: Departmental.

The authors declare no conflicts of interest.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website.

Reprints will not be available from the authors.

Address correspondence to Anne D. Cherry, MD, Department of Anesthesiology, Duke University, DUMC Box 3094, Durham, NC 27710. Address e-mail to anne.cherry@duke.edu.

A 58-year-old woman with chronic renal failure, atrial fibrillation (AF), and mitral stenosis (MS) undergoing urgent bowel resection. Intraoperative transesophageal echocardiography (TEE) is requested to determine the significance of her MS and to guide hemodynamic management. What measurements or evidence would facilitate grading? What other consequences of MS should be evaluated by TEE?

Back to Top | Article Outline

PATHOPHYSIOLOGY

MS is defined as reduction in the mitral valve (MV) orifice area because of pathologic leaflet changes. Congenital MS is rare, resulting from abnormalities of the MV or subvalvular apparatus. In adults, MS is caused by either rheumatic or degenerative disease. Rheumatic MS is caused by autoimmune cross-reactivity between group A Β-hemolytic streptococcus antigens and endothelial tissue, resulting in inflammatory scarring of MV tissue. Degenerative MS causes annular calcification extending into the leaflets. Its incidence increases with age and in patients with renal failure, disorders of calcium metabolism, or after radiation.1 The distinct mechanisms of rheumatic and degenerative MS influence the emphasis of the TEE examination with respect to the structural differences, associated lesions, and reliability of echocardiographic modalities for severity grading.

Back to Top | Article Outline

QUALITATIVE ASSESSMENT (2D AND 3D)

The MV apparatus can be evaluated at the midesophageal level at different omniplane angles (4-chamber view at 0–10°, mitral commissural view at 50–70°, 2-chamber view at 80–100°, and long-axis view at 120–140°). The transgastric basal short-axis view shows commissural pathology, whereas the 2-chamber view displays the subvalvular apparatus. Three-dimensional (3D) imaging provides en face views of the MV from the left atrial (LA) or left ventricular (LV) perspective, demonstrating the relationship of the MV apparatus with adjacent structures.

Rheumatic MS manifests as commissural fusion, chordal shortening, leaflet thickening, and reduced leaflet mobility. The posterior leaflet is more often affected than the anterior.2 The disease involves predominantly the leaflet tips with relative sparing of the mitral annulus and leaflet bases. These result in the characteristic “hockey-stick” leaflet appearance (from the LA, a concave appearance of the leaflets in diastole) and minimal leaflet tip motion between opening and closing. The smallest MV area is at the leaflet tips (Fig. 1; Supplemental Digital Content 1, Video, http://links.lww.com/AA/B406). In degenerative MS, calcification of the MV annulus extends progressively into the leaflet tips (also posterior > anterior), resulting in reduced diastolic excursion with the smallest orifice area at the base of the MV leaflets (Fig. 2; Supplemental Digital Content 1, Video, http://links.lww.com/AA/B406).1 The subvalvular apparatus also may be calcified.

Figure 1.

Figure 1.

Figure 2.

Figure 2.

Table 1.

Table 1.

Several qualitative 2D scoring systems describe the MV morphology. The Wilkins score characterizes leaflet mobility, valve thickness, subvalvular thickening, and calcification. It was devised to determine suitability for balloon valvuloplasty and postprocedural mitral regurgitation (MR) risk3 but also communicates the varied locations and manifestations of MV pathology (Table 1). The Carpentier functional classification also describes mixed MV disease (stenosis and regurgitation) as type IIIa with leaflet restriction throughout the cardiac cycle. MS graded as severe by echocardiography is now further classified as symptomatic (stage D) or asymptomatic (stage C), to guide disease management.4

Back to Top | Article Outline

QUANTITATIVE ASSESSMENT

Table 2.

Table 2.

MS severity is determined using estimation of transmitral pressure gradient estimation by Doppler and measurements or calculations of mitral valve area (MVA) (Table 2). The normal MVA is 4 to 5 cm2. MVA of >1.5, 1.5 to 1, and <1 cm2 correspond to mild, moderate, and severe stenosis, respectively.

Back to Top | Article Outline

MVA by Planimetry

Planimetry by 2D imaging measures the MVA by tracing the MV orifice outline in mid-diastole in the transgastric basal short-axis view (Fig. 3A). It is a direct, anatomic measurement; therefore, it is less dependent on flow conditions5 and chamber compliance. Tachycardia may cause underestimation of MVA by planimetry because of incomplete opening during short cardiac cycles. Small adjustments in probe position and probe flexion or simultaneous orthogonal imaging may be needed to find the smallest orifice of the MV. In degenerative MS, calcific acoustic shadowing artifact may obscure the leaflets, making planimetry unreliable. In addition, the narrowest orifice may not be at the level of the leaflet tips.

Figure 3.

Figure 3.

Three-dimensional imaging may overcome these limitations by allowing simultaneous display of 3 orthogonal views and improve identification of the narrowest orifice (Fig. 3B).6 In severely distorted valves, even the flexibility of 3D imaging may not allow accurate planimetry. In 3D imaging, more so than in 2D, echodense objects can appear falsely large or thickened (blooming artifact), particularly with high gain settings. This would result in planimetered orifice area underestimation. Acquisition at low frame rate may prevent capture of the true mid-diastolic frame. Adding color flow Doppler (CFD) to 3D imaging may help locate the narrowest orifice area and allow functional orifice area measurement (Fig. 3C). The functional orifice area and the anatomic orifice area are not technically interchangeable. In stenotic valves, the functional orifice, which is measured by using CFD, occurs slightly distal to the anatomic orifice at the point where blood flow contracts and achieves maximal velocity. Therefore, the functional area is slightly smaller than the anatomic area. Although planimetry measures the anatomic MVA, planimetry by 3D with CFD and MVA calculations by pressure half-time (PHT), proximal isovelocity surface area (PISA) method, and continuity equation assess functional MVA. All methods correlate well with surgical findings.7 Three-dimensional with CFD also is limited by lower temporal resolution, CFD gain settings, and color artifacts.

Back to Top | Article Outline

Pressure Gradient

Transmitral pressure gradients are measured using continuous wave Doppler (CWD) of diastolic transmitral flow in any midesophageal view that allows optimal alignment of the Doppler beam and flow (within 15° of parallel). Visualization of flow direction using CFD may help ensure adequate alignment, especially in severe MV apparatus deformation where flow may be eccentric. Mean pressure gradients are used because peak velocities (and so also peak gradients) are more variable, being influenced by LA compliance and LV diastolic relaxation (Fig. 4A). Mean gradients of <5 mm Hg, 5 to 10 mm Hg, and >10 mm Hg correspond to mild, moderate, and severe stenosis, respectively. These cutoffs are based on heart rates of 60 to 80 bpm in sinus rhythm.8

Figure 4.

Figure 4.

Caution must be exercised in AF, tachycardia, or with high or low transmitral flow. AF results in beat-to-beat variation in flow. Therefore, mean pressure gradients should be averaged over 5 beats. Decreased diastolic filling time in tachycardia results in elevation of the LA pressure and the transmitral mean gradient. High transmitral flow because of hyperdynamic states (e.g., sepsis, hepatic failure) or severe MR will increase the mean gradient9 and cause overestimation of MS severity. Conversely, low cardiac output causes underestimation of MS severity. The mean pressure gradient also is sensitive to alterations in atrioventricular compliance, as seen in patients with diastolic dysfunction. Coexistent diastolic dysfunction, likely in older patients or those with ischemic heart disease, hypertension, LV hypertrophy, or aortic stenosis (AS) (factors also associated with degenerative MS), decreases the early diastolic transmitral gradient and underestimates MS severity.10

Back to Top | Article Outline

MVA Calculation

Three echocardiographic methods are used to estimate MVA: PHT, continuity method, and PISA. MVA can be derived invasively by the Gorlin formula, using the catheterization-derived cardiac output and transvalvular mean pressure gradient. Although all echocardiographic methods have shown good correlation with invasive MVA calculations and direct surgical measurements,7 each has inherent limitations in specific clinical situations.

Back to Top | Article Outline

MVA by PHT

The PHT represents the time elapsed as the peak diastolic transmitral pressure gradient decays to half its peak value. On the velocity scale, PHT is the time necessary for the peak transmitral velocity (Vmax) to decrease to Vmax/√2. PHT measurement is performed by tracing the transmitral flow slope from peak velocity to at least the point on the Doppler spectrum where velocity is two thirds of the peak velocity (Fig. 4B). Therefore, PHT measurement is inaccurate in situations of abbreviated diastolic filling such as sinus tachycardia or short cardiac cycles in AF. Also, fusion of the transmitral E and A waves in tachycardia prevents accurate measurement. If the mitral inflow Doppler tracing is bimodal and nonlinear, the deceleration slope should not be traced from the early part (representing initial, early diastolic decline in pressure gradient) but rather using extrapolation of the linear, midportion of the profile (Fig. 4C).11

MS results in slower decay of the pressure gradient between the LA and the LV and a prolonged PHT. PHT can be used in a simple, empirically derived formula to estimate the MVA: MVA = 220/PHT (milliseconds) (Table 2).12 The constant 220 is directly proportional to the product of the net compliance of the LA and LV and with the square root of the initial transmitral pressure gradient. Net chamber compliance and pressure gradient change in opposite directions, thus offsetting their effect on PHT. However, this assumption may not be valid with acute changes in chamber compliance.13

PHT is relatively unaffected by MR12 and rate-controlled AF.14 With AF, measurements should be made and averaged across 5 cardiac cycles, and tracing of mitral flow from short diastoles should be avoided.8 PHT measurements can be unreliable with tachycardia. With increasing heart rate, the PHT method overestimates MVA.9 Significant aortic regurgitation (AR) contributes to diastolic LV filling and causes significant shortening of PHT with MVA overestimation.15 This effect is highly dependent on chamber compliance and on the severity of MS, which may explain conflicting results in other studies regarding the impact of AR on PHT.13 Coexistent diastolic dysfunction, especially in patients with degenerative MS and concomitant AS, results in a shortened PHT and MVA overestimation8; PHT should not be used in these patients.

PHT is related to deceleration time (DT); PHT = 0.29 × DT, where DT is the time taken for the peak early diastolic velocity to decrease to zero. If the early diastolic transmitral flow velocity does not decrease to zero, the decay slope should be extrapolated to the zero baseline (Fig. 4B). MVA is calculated using MVA (cm2) = 759/DT (milliseconds).

Back to Top | Article Outline

MVA by Continuity Equation

The continuity method is based on the conservation of mass. Transmitral stroke volume (SV) is equal to the SV through the left or right ventricular (RV) outflow tracts (LVOT or RVOT). For example, MV SV = LVOT SV; therefore, MVA = (LVOT CSA × LVOT VTI)/(MV VTI), where VTI = velocity time integral and CSA = cross-sectional area. LVOT or RVOT VTI is measured using pulse wave Doppler with the sample volume placed approximately 5 mm proximal to the aortic or pulmonary valves, respectively. The LVOT and RVOT diameter also should be measured at this proximal point. Diameter measurements are made carefully because these values are squared and can exaggerate error. MV VTI should be measured using CWD to capture the velocity profile where the orifice is smallest, with beam alignment optimization as performed for pressure gradient measurement.

The continuity equation is less dependent on cardiac output than the pressure gradient. A major limitation of the continuity equation is that it cannot be used if there are significant intracardiac shunts or regurgitant jets at measurement points or anywhere between them. AR or pulmonary regurgitation results in overestimation of MVA, whereas MR results in underestimation of MVA. Also, because measurements are made in several different imaging planes and cardiac cycles, irregular rhythms and variable loading conditions during acquisition invalidate the assumption of equivalent beat-to-beat flow.

Back to Top | Article Outline

MVA by PISA

The PISA method depends on flow acceleration; as flow accelerates toward a narrow orifice, it forms hemispheric “shells” of increasing velocity and decreasing radius. Each hemisphere is considered an isovelocity surface. Calculation of MVA by PISA is an application of the principle of continuity of flow; peak flow rate across the area of any hemispheric shell must equal peak flow rate through the MV orifice. Because the MV orifice is the narrowest point in the path of blood flow, it corresponds to the location of maximal diastolic flow velocity (Vmax) as measured by CWD through the MV (Fig. 5A); applying these variables to the continuity equation yields the MVA: MVA = (2 π R2 × Va)/Vmax, where R = radius of the hemispheric shell, Va = velocity at the level of the shell or aliasing velocity (Fig. 5B; Supplemental Digital Content 2, Video, http://links.lww.com/AA/B407). This formula assumes flow acceleration toward a flat, narrow orifice. In MS, however, the MV leaflets form a funnel at an angle α, which constrains the isovelocity shell to an arc of α degrees. Therefore, the MVA formula should account for the angle correction: MVA = (2 π R2 × Va)/Vmax × α/180.

Figure 5.

Figure 5.

The PISA method is not confounded by chronic or acute changes in LA pressure and LV compliance.16 Being independent of reference valve flow, PISA is not affected by significant MR or AR. Loading conditions also are less likely to influence the calculations because the acquisition of all formula components can be accomplished in a couple of beats. There are, however, several limitations to this method. Measurement of the α angle is not automated and is most often estimated (introducing error). Accurate MVA calculation also is dependent on accurate measurement of the isovelocity surface radius because squaring this value exaggerates error. The PISA changes shape as the Va changes. PISA is flatter at high velocities and more upright at lower velocities. A hemispheric shape is usually achieved with Va set between 30 and 40 cm/s. The PISA model also is based on the assumption of hemispheric flow convergence toward the stenotic MV orifice and may not be true with an elliptical orifice such as in rheumatic MS. Some of these limitations may be overcome by the measurement of the PISA from a 3D data set with CFD, but this requires postprocessing software.

Back to Top | Article Outline

TEE EVALUATION FOR CONSEQUENCES OF MS

TEE for MS should include evaluation of other structural and functional changes associated with impaired transmitral flow (Table 3). Patients with LA enlargement may have or develop AF. Hemodynamically significant stenosis and sluggish LA blood flow predispose to thrombus formation. TEE in these patients may reveal spontaneous echo contrast (SEC) in the LA. Positive and negative predictive values for thrombus with SEC are 44% to 48% and 82% to 93%, respectively,17,18 so SEC should prompt thorough evaluation of LA appendage and chamber walls for thrombus.

Table 3.

Table 3.

Patients with MS also may have LV dysfunction. LV contractility is difficult to assess because ejection can be reduced by impaired LV preload and elevated afterload (low cardiac output–induced vasoconstriction). Reduced LV contractility has been reported in patients with aggressive inflammation secondary to rheumatic MS.19 Both diastolic dysfunction and MS can result in increased LA pressure and pulmonary congestion. A very short isovolumic relaxation time (<70 milliseconds) and systolic blunting of pulse wave Doppler pulmonary venous flow (systolic velocity < diastolic velocity) are both correlated with increased left atrial pressure (LAP). Unfortunately, the relative contribution of MS versus diastolic dysfunction to increased LAP is difficult to distinguish by TEE but is important to consider for 2 reasons: (1) diastolic dysfunction also impairs LV filling and may lower the mean transmitral gradient, causing underestimation of MS severity and (2) severe diastolic dysfunction may not improve after MS intervention and may impact long-term patient outcome. Unfortunately, MS confounds most echocardiographic modalities of diastolic assessment (pulmonary venous Doppler, mitral inflow Doppler, and annular tissue Doppler). Because patients with isolated MS normally have low to normal LV end-diastolic pressure versus diastolic dysfunction, in which LV end-diastolic pressure is high, direct LV pressure measurements during cardiac catheterization can help distinguish between the 2.

Other valves may be affected in rheumatic or degenerative MS. MR may be present with either etiology.20,21 Infective endocarditis, AR, and tricuspid regurgitation (TR) are associated with rheumatic MS,22 whereas calcific AS is more common with degenerative disease. In combined MS and AS, pressure gradients may lead to inaccurate severity grading because MS may falsely lower gradients by imposing a low-flow state across the aortic valve.

Finally, increased LAP because of MS may precipitate pulmonary congestion and result in vascular remodeling, RV impairment, RV dilation, and tricuspid annular dilation leading to functional TR. Pulmonary artery systolic pressure (PASP) can be estimated using CWD of a TR jet to estimate the systolic peak gradient between the RV and the right atrium. RV systolic pressure (mm Hg) is calculated as the sum of right atrial pressure and the peak gradient associated with the TR jet. In the absence of significant RVOT or pulmonary valve gradients, RV systolic pressure approximates PASP. PASP <30 mm Hg correlates with mild MS and >50 mm Hg correlates with severe. However, a wide range of PASP can be present for a given valve area.8 TR repair at the time of surgery for rheumatic MS deserves special consideration. The higher burden of pulmonary hypertension and subtle rheumatic involvement of the tricuspid valve apparatus make these patients more prone to progression of TR after MV surgery.23 The incidence of echocardiographically significant late TR after MV replacement has been found to be as high as 68% in patients with rheumatic MS.24

MS poses significant challenges during surgery, depending on the etiology of the disease. Significant mitral annulus calcification seen in patients with degenerative MS may contribute to paravalvular leaks or cardiac rupture at the atrioventricular groove or LV free wall after MV replacement. In patients with rheumatic MS, valve repair with complex techniques such as leaflet enlargement and commissurotomy is possible, although more controversial because of the progressive nature of the rheumatic disease.25 In summary, echocardiographic grading of disease severity is important for management decision making; the strengths and limitations of each quantitative method, along with supportive evidence, should be carefully considered.

Back to Top | Article Outline

Teaching Points

  • The smallest MVA is at the leaflet base in degenerative MS and at the leaflet tips in rheumatic MS. Two-dimensional or 3D planimetry provides direct measurement. CFD-facilitated planimetry measures the functional MVA.
  • MVA is calculated by PHT (MVA = 220/PHT). The PHT method is limited by LV hypertrophy, significant AS, severe AR, and tachycardia.
  • MVA can be calculated by PISA and by continuity equation. The PISA method is limited mostly by geometric assumptions and potential technical errors. The continuity equation method is limited by variations in beat-to-beat filling, as well as the presence of intracardiac shunts or regurgitant jets at the measurement points or anywhere between them.
  • The mean pressure gradient facilitates MS grading. It must be averaged over 5 beats in AF and does not correlate well with severity in the setting of low or high transmitral flows or in the presence of tachycardia.
Back to Top | Article Outline

DISCLOSURES

Name: Anne D. Cherry, MD.

Contribution: This author helped write the manuscript.

Attestation: Anne D. Cherry approved the final manuscript.

Name: Cory D. Maxwell, MD.

Contribution: This author helped write the manuscript.

Attestation: Cory D. Maxwell approved the final manuscript.

Name: Alina Nicoara, MD.

Contribution: This author helped write the manuscript.

Attestation: Alina Nicoara approved the final manuscript.

This manuscript was handled by: Martin London, MD.

Back to Top | Article Outline

REFERENCES

1. Payvandi LA, Rigolin VHCalcific mitral stenosis.Cardiol Clin201331193202
2. Krapf L, Dreyfus J, Cueff C, Lepage L, Brochet E, Vahanian A, Messika-Zeitoun DAnatomical features of rheumatic and non-rheumatic mitral stenosis: potential additional value of three-dimensional echocardiography.Arch Cardiovasc Dis20131061115
3. Wilkins GT, Weyman AE, Abascal VM, Block PC, Palacios IFPercutaneous balloon dilatation of the mitral valve: an analysis of echocardiographic variables related to outcome and the mechanism of dilatation.Br Heart J198860299308
4. Skubas NJ, Shernan SK, Bollen BAn overview of the American College of Cardiology/American Heart Association 2014 Valve Heart Disease Practice Guidelines: what is its relevance for the anesthesiologist and perioperative medicine physician?Anesth Analg201512111328
5. Mohan JC, Patel AR, Passey R, Gupta D, Kumar M, Arora R, Pandian NGIs the mitral valve area flow-dependent in mitral stenosis? A dobutamine stress echocardiographic study.J Am Coll Cardiol200240180915
6. Schlosshan D, Aggarwal G, Mathur G, Allan R, Cranney GReal-time 3D transesophageal echocardiography for the evaluation of rheumatic mitral stenosis.JACC Cardiovasc Imaging201145808
7. Faletra F, Pezzano A Jr, Fusco R, Mantero A, Corno R, Crivellaro W, De Chiara F, Vitali E, Gordini V, Magnani P, Pezzano A Sr.Measurement of mitral valve area in mitral stenosis: four echocardiographic methods compared with direct measurement of anatomic orifices.J Am Coll Cardiol19962811907
8. Baumgartner H, Hung J, Bermejo J, Chambers JB, Evangelista A, Griffin BP, Iung B, Otto CM, Pellikka PA, Quiñones MEAE/ASEEchocardiographic assessment of valve stenosis: EAE/ASE recommendations for clinical practice.Eur J Echocardiogr200910125
9. Voelker W, Regele B, Dittmann H, Mauser M, Ickrath O, Schmid KM, Karsch KREffect of heart rate on transmitral flow velocity profile and Doppler measurements of mitral valve area in patients with mitral stenosis.Eur Heart J1992131529
10. Zaid RR, Barker CM, Little SH, Nagueh SFPre- and post-operative diastolic dysfunction in patients with valvular heart disease: diagnosis and therapeutic implications.J Am Coll Cardiol201362192230
11. Gonzalez MA, Child JS, Krivokapich JComparison of two-dimensional and Doppler echocardiography and intracardiac hemodynamics for quantification of mitral stenosis.Am J Cardiol19876032732
12. Hatle L, Angelsen B, Tromsdal ANoninvasive assessment of atrioventricular pressure half-time by Doppler ultrasound.Circulation1979601096104
13. Thomas JD, Wilkins GT, Choong CY, Abascal VM, Palacios IF, Block PC, Weyman AEInaccuracy of mitral pressure half-time immediately after percutaneous mitral valvotomy. Dependence on transmitral gradient and left atrial and ventricular compliance.Circulation19887898093
14. Bryg RJ, Williams GA, Labovitz AJ, Aker U, Kennedy HLEffect of atrial fibrillation and mitral regurgitation on calculated mitral valve area in mitral stenosis.Am J Cardiol1986576348
15. Flachskampf FA, Weyman AE, Gillam L, Liu CM, Abascal VM, Thomas JDAortic regurgitation shortens Doppler pressure half-time in mitral stenosis: clinical evidence, in vitro simulation and theoretic analysis.J Am Coll Cardiol199016396404
16. Salem Omar AM, Tanaka H, AbdelDayem TK, Sadek AS, Raslaan H, Al-Sherbiny A, Yamawaki K, Ryo K, Fukuda Y, Norisada K, Tatsumi K, Onishi T, Matsumoto K, Kawai H, Hirata KComparison of mitral valve area by pressure half-time and proximal isovelocity surface area method in patients with mitral stenosis: effect of net atrioventricular compliance.Eur J Echocardiogr20111228390
17. Black IW, Hopkins AP, Lee LC, Walsh WFLeft atrial spontaneous echo contrast: a clinical and echocardiographic analysis.J Am Coll Cardiol199118398404
18. Daniel WG, Nellessen U, Schröder E, Nonnast-Daniel B, Bednarski P, Nikutta P, Lichtlen PRLeft atrial spontaneous echo contrast in mitral valve disease: an indicator for an increased thromboembolic risk.J Am Coll Cardiol198811120411
19. Carabello BAModern management of mitral stenosis.Circulation20051124327
20. Akram MR, Chan T, McAuliffe S, Chenzbraun ANon-rheumatic annular mitral stenosis: prevalence and characteristics.Eur J Echocardiogr2009101035
21. Carpentier AF, Pellerin M, Fuzellier JF, Relland JYExtensive calcification of the mitral valve anulus: pathology and surgical management.J Thorac Cardiovasc Surg199611171829
22. Marijon E, Mirabel M, Celermajer DS, Jouven XRheumatic heart disease.Lancet201237995364
23. Shiran A, Sagie ATricuspid regurgitation in mitral valve disease incidence, prognostic implications, mechanism, and management.J Am Coll Cardiol2009534018
24. Porter A, Shapira Y, Wurzel M, Sulkes J, Vaturi M, Adler Y, Sahar G, Sagie ATricuspid regurgitation late after mitral valve replacement: clinical and echocardiographic evaluation.J Heart Valve Dis199985762
25. Zakkar M, Amirak E, Chan KM, Punjabi PPRheumatic mitral valve disease: current surgical status.Prog Cardiovasc Dis20095147881

Supplemental Digital Content

Back to Top | Article Outline
Copyright © 2016 International Anesthesia Research Society