Aortic Valve Sclerosis

Aortic valve sclerosis is an important disease entity, although usually not because of hemodynamic considerations. Defined as calcification and thickening of the aortic valve without significant outflow obstruction (gradient <20 to 25 mm Hg), it is present in nearly 30% of the population older than age 65 and nearly 50% by age 85¹⁰ and is associated with a 50% increase in 5-year cardiovascular mortality rate.¹¹ Its incidence is as high as 15-fold greater than aortic valve stenosis; the two diagnoses can be differentiated from one another by hemodynamics and physical examination findings discussed subsequently. In keeping with the primary atherosclerotic nature of aortic sclerosis, it is associated with increasing age, male gender, hypertension, smoking, elevated low-density lipoprotein and lipoprotein(a) levels, and diabetes.¹⁰ The primary importance of aortic valve sclerosis is that it provides a window on the overall presence of vascular disease, in particular coronary artery disease.¹² Thus there should be a high index of suspicion of vascular disease in any patient with aortic valve sclerosis managed in the critical care setting.

Patients with aortic sclerosis do progress to aortic valve stenosis; a study of more than 2000 patients with valve thickening showed progression to severe AS in 2.5% over an average time interval of 7 years.¹³ The fact that so-called degenerative disease of the aortic valve is absent in nearly half of octogenarians is also important to consider because it implies that it is not only aging but other factors that result in leaflet thickening, calcification, and stenosis. In the Helsinki Aging Study from which data are reflected in Figure 32.1, additional analysis demonstrated that not only age but also hypertension and low body mass index independently predicted calcification of the aortic valve, and age and serum ionized calcium were independently associated with valvular stenosis.¹⁴ In general the process of sclerosis and then stenosis of the aortic valve appears, like atherosclerosis in general, to be an active inflammatory process, with deposition of lipoproteins, local inflammation with T lymphocyte and macrophage infiltration, fibroblast proliferation, and eventually osteoblast and bone formation.¹⁵ Similar to other vascular disease, endothelial disruption likely leads to the initial lipid deposition in the leaflet tissues. The areas of early focal plaque formation appear at the loci of greatest stress: on the aortic side of the leaflets at the flexion points. Because bicuspid valves have greater mechanical stress, the average age at presentation of patients with bicuspid aortic valve stenosis is significantly lower than in the tricuspid aortic valve stenosis that is typically seen in an elderly population.¹⁶

An important element in understanding calcific AS is lack of commissural fusion; unlike rheumatic mitral valve stenosis, in which fusion of the commissures is the primary cause of obstructive disease, lack of mobility of the aortic valve leaflets is the primary cause of obstruction in AS. Rheumatic AS is associated with commissural fusion but is now a rare finding, even in countries where rheumatic heart disease is prevalent, and is usually associated with mitral valve involvement and typically aortic insufficiency as well. More obscure causes, such as unicuspid and quadricuspid aortic valve disease, are uncommon, although presentation can be delayed to adulthood. In the case of unicuspid aortic valves there is a strong association with AS,¹⁷ whereas with quadricuspid aortic valves there is a high incidence of significant aortic insufficiency.¹⁸

Normal aortic valve area ranges from 3 to 4 cm². Significant resistance to outflow does not occur until the valve orifice is reduced more than 50%. Based on the simple hydraulic principle of Poiseuille’s law, a 50% reduction in valve diameter results in approximately a 16-fold increase in resistance. In practice, as the resistance to outflow rises, the left ventricle is subject to pressure overload, resulting in compensatory hypertrophy. This in turn normalizes wall stress because the latter is proportional to chamber diameter times pressure divided by wall thickness (the LaPlace principle). The degree of wall thickness is variable: In patients with inadequate hypertrophic response, wall stress is inordinately high and there is early dilation and dysfunction.¹⁹ In patients with a profound hypertrophic response disproportionate to valvular resistance, wall stress actually falls below normal and ejection fraction (EF) becomes supranormal,²⁰ a phenomenon that appears to be more common in women with AS. The EF for any degree of wall stress is predictable,¹⁹ and patients who have disproportionately poor ejection phase indices (afterload mismatch) typically manifest left ventricular (LV) dysfunction beyond the depression that would be associated with high wall stress (Fig. 32.3),²¹ a phenomenon described under low-gradient, low-output AS that follows.

LV hypertrophy, although it reduces wall stress, has several features that may be deleterious. With progressive hypertrophy, diastolic pressures may rise as LV compliance decreases, resulting in increased LV filling pressures. In addition, the myocardial supply-demand relationship is deleteriously affected by hypertrophy in this setting. With increased wall stress and increased wall mass, myocardial oxygen demand is increased; at the same time coronary flow reserve is decreased in AS,²² and in later stages of the disease diastolic perfusion pressure is lowered. Because of abnormal flow reserve, conventional testing for ischemia in the setting of significant aortic valve stenosis has inadequate specificity to identify the presence or absence of hemodynamically significant concomitant coronary artery disease.²³

Other features associated with AS that are important in the critical care setting include abnormal platelet function and decreased levels of von Willebrand factor,²⁴ with associated bleeding risk that improves with AVR. Another associated cause of bleeding is gastrointestinal angiodysplasia associated with aortic valve stenosis,²⁵ possibly exacerbated by the concomitant presence of von Willebrand syndrome.

The progression of AS is highly variable, but with the onset of moderately severe disease, the estimated mean decrease in valve area is on the order of 0.1 cm²/year.²⁶ Certain demographics and noninvasive findings appear to predict the rate of progression, including age older than 50 years and moderate to severe valve calcification.²⁷ The issue of progression is particularly important in patients who are asymptomatic of the disease because AVR before it is necessary is undesirable but must be weighed against the risk of sudden death. The latter has been studied in a number of longitudinal studies and is rare if patients are followed prospectively. The patients who are most likely to require early intervention are those with peak Doppler systolic velocities greater than 4 m/second or who have a rapid increase in serial transvalvular Doppler velocity measurements.^26,27 This threshold has been confirmed by other series.²⁸ Once patients become symptomatic, with a classic triad of heart failure, angina, or syncope, hemodynamic deterioration can be rapid with significant associated mortality risk.²⁹

Physical Examination

The hallmarks of the physical examination relate to the contour of the aortic pressure upstroke, with a low volume and delayed carotid upstroke (pulsus parvus et tardus), a late peaking systolic ejection murmur, and a diminished or absent aortic second sound. Figure 32.4 demonstrates the hemodynamics that manifest in an abnormal carotid pulse contour; unlike the brisk upstroke of LV pressure over time (dP/dt), the aortic pressure is slow to rise and of significantly lower volume. An ejection click may be present in young patients with congenital AS, but with increasing calcification of the valve, mobility decreases to a point at which an ejection click becomes unlikely. The aortic component of the second heart sound, reflecting deceleration and closure of the aortic valve, is diminished or absent in severe AS when the thickened aortic valve leaflets are poorly mobile, have little deceleration, and drift rather than snap shut. The murmur of AS is characteristically a crescendo-decrescendo murmur that is late peaking, consistent with the timing of the peak gradient as seen in Figure 32.4, and is typically best heard over the right upper sternum and clavicle. It reflects the high-velocity systolic jet directed into the ascending aorta. A second component, described by Gallavardin as a musical component, is best heard at the left lower sternal border. The latter confounds diagnosis because the murmur is suggestive of mitral regurgitation (MR), but in its true form it is purely secondary to aortic valve stenosis. Because handgrip raises resistance to LV ejection, the murmur should decrease if it is a true Gallavardin phenomenon, whereas when it is secondary to MR, it should increase.³⁰

Unfortunately, none of these findings has sufficiently high sensitivity or specificity for AS or for differentiating moderate from severe disease.³ Many patients with AS, particularly the elderly with stiff noncompliant vessels, will have a wide pulse pressure even with declining cardiac output and may have systolic hypertension. Hypotension may not be seen until late stages of the disease. The loudness of the murmur, which correlates to some degree with severity, is also not specific because body habitus and a variety of other factors affect the acoustics transmitted across the chest. Most importantly, in severe AS, as the cardiac output drops, the gradient generated decreases as well. In this setting the murmur may be quite soft, although sometimes still high pitched because of the high velocity across a tight constriction. A useful means of differentiating aortic sclerosis from stenosis is that in the former the systolic ejection murmur heard over the right sternum is typically midpeaking and mild to moderate in intensity, and it features a well-preserved aortic second heart sound. Although the physical findings described are variable, presence of the aortic second heart sound tends to exclude severe calcific AS.³¹

Noninvasive Evaluation

In contrast to the physical examination, echocardiographic techniques have progressively improved over the past several decades and availability in the critical care setting in industrialized nations is ubiquitous. The characteristic echocardiographic features of AS are decreased valve leaflet mobility, calcification in all except congenital AS in adolescents and young adults, and an augmented Doppler velocity that generally allows accurate estimation of the gradient. In congenital bicuspid AS in young adults, when commissural fusion is dominant, a characteristic doming pattern is seen, but this disappears with progressive valve calcification. The severity of calcification correlates with extent of obstruction by middle age, and typically Doppler signal velocity with peak and mean pressure gradient and valve area by continuity equation provide an accurate overall assessment. However, the gradient is highly dependent on flow across the valve, and in low-output states the gradient may result in underestimation of severity of disease; in high-output states such as in patients with augmented cardiac output caused by inotropic stimulation, endogenous high catecholamine states, and sepsis, the gradient may be disproportionately higher than the severity of stenosis would suggest. The latest ACC/AHA guidelines describe the threshold for severe AS as antegrade jet velocity greater than 4 m/second, gradient greater than 40 mm Hg, and aortic valve area less than 1 cm², with the caveat that the gradient and jet velocity depend on the overall transvalvular flow⁵ (see Table 32.1). Importantly, antegrade flow across the aortic valve does not equal cardiac output in patients with confounding conditions, including aortic insufficiency; thus, for example, a patient with mild to moderate AS and moderate aortic insufficiency may have a transvalvular gradient suggestive of severe AS. Finally, although most practitioners do not index the valve area, it is important to appreciate that in patients with large body surface areas, disease that might be considered moderate in smaller patients may be functionally severe.

Several caveats need to be considered before acceptance of noninvasive data in the critical care setting. Inadequate acoustic windows in some patients and difficulty in positioning patients on respirators and with multiple lines in place do limit echocardiographic imaging. Technical errors in recording and interpretation result in significant misdiagnosis: errors in recording angle, inadvertent imaging of MR jets instead of aortic outflow, assessing proximal velocity instead of the transvalvular signal, and selecting signals in the setting of arrhythmias that are not representative of mean heartbeats can all lead to skewed assessment of valve disease severity.³² Further, a number of other potential confounding variables can lead to overestimation or underestimation of the severity of AS,^6,33 and valve area calculations by the continuity equation in a low-flow setting may be inaccurate.³⁴ Because the continuity equation depends on measurement of the LV outflow track dimension, which is prone to inaccuracy, another measure of AS severity, the dimensionless index (the ratio of blood flow velocity across the LV outflow track to that across the aortic valve), has been widely adopted.³⁵ A number of other techniques for determining the severity of AS supplement the most commonly used noninvasive tools, including three-dimensional echo, computed tomography,³⁶ and magnetic resonance imaging (MRI).³⁷ Overall, there has been a major a shift in the gold standard from catheter-derived to noninvasive-derived determination of AS severity, and cardiac catheterization is no longer indicated for hemodynamic assessment of aortic valve disease severity when noninvasive findings are unequivocal.⁵

The echocardiogram provides important additional information including severity of LV hypertrophy, LV function and size, and concomitant disease of other heart valves. The presence of regional wall motion abnormalities, in the absence of a conduction disturbance, suggests concomitant coronary artery disease. The electrocardiogram (ECG) in AS is typically abnormal and frequently features LV hypertrophy, with ST-segment abnormalities in the lateral leads typically described as a strain pattern.³⁸ Although occasionally mistaken for anterior ischemia or infarction, with loss of R voltage across the precordium and occasionally with narrow QS complexes, the ECG is not specific for severity of AS. With increasing calcification of the perivalvular tissues, heart block is seen, typically late in the course of the disease.

Cardiac Catheterization

Cardiac catheterization is indicated in patients in whom the noninvasive data are equivocal, and coronary angiography is indicated in a range of settings identified in Box 32.1. The hemodynamics of AS are best recorded with a catheter or catheters placed simultaneously on either side of the aortic valve.^6,39 Unfortunately, most laboratories record femoral artery pressure in lieu of central aortic pressure or do a catheter pullback in lieu of simultaneous pressure tracings; both techniques can result in significant misinterpretation of the severity of aortic valve disease.⁴⁰ A variety of other errors in catheterization laboratory pressure measurements makes the catheter-derived valve area generally more variable than desirable in all except a few laboratories, including inherent errors in estimating rather than measuring oxygen consumption and in assuming that the Gorlin constant (originally established for a limited subset of patients⁴¹) and the valve area itself remain constant under varying loading conditions.³⁴ In the catheterization laboratory as well, transvalvular flow may be underestimated if there is concomitant aortic insufficiency, resulting in overestimation of the severity of aortic valve disease.

Medical Management

Medical management is aimed solely at patient stabilization because pharmacologic intervention has never been shown to prolong life and can achieve modest hemodynamic improvement at best. In the critical care setting the approach to the patient with AS normally parallels treatment of corresponding degrees of heart failure, albeit with caution, because the normal therapeutic approach to heart failure can be deleterious or fatal in the setting of severe AS. The traditional heart failure therapies, in particular, diuretics and vasodilators, need to be used with care because reduction in preload can cause hypotension, decreased cardiac output, and a downward spiral into refractory shock. In patients with small-volume hypertrophic left ventricles, cardiac output is particularly preload dependent.

Atrial arrhythmias, in particular, atrial fibrillation (AF), can lead to abrupt and severe decompensation because of loss of the atrial contraction component of LV filling. Cardioversion in the critical care setting is particularly helpful in severely decompensated patients, and maintenance of sinus rhythm in patients who are otherwise suitable (e.g., left atrium <6 cm, recent onset of AF, and absence of clot in the left atrium by transesophageal echocardiography) should be a priority.

Despite the caveats, vasodilator therapy, including nitroprusside, has been used in the critical care setting in patients with severe LV dysfunction and severe AS.⁴⁹ In a select group of patients not dependent on inotropes, cardiac output rose significantly and there was overall hemodynamic improvement. Similarly, intra-aortic balloon pumping has been used, although there are only isolated case reports of efficacy.^50,51 As would be expected, both nitroprusside and intra-aortic balloon pumping result in afterload reduction and some increase in transvalvular flow along with some increase in aortic valve gradient.

Hemodynamic monitoring is essential for the pharmacologic management of decompensated severe AS patients in the critical care setting. Because of the dependence on preload, a fine threshold exists between optimal filling pressures and pulmonary edema. Vasodilator therapy as described earlier results in peripheral vasodilation, but because of the fixed obstruction to outflow it may not provide sufficient increase in stroke volume and cardiac output to maintain systemic blood pressure. Inotropic agents are frequently required, and patients with inadequate contractile reserve may show limited improvement. Beta blockers are relatively dangerous, although tachycardia is hemodynamically unfavorable, even in sinus rhythm. Beta blockade, if employed, should be used with great caution. Beta blockers decrease contractility and, in addition, an increase in heart rate may be the only remaining compensatory mechanism for low stroke volume because of the fixed resistance to outflow. Patients with severe AS can enter a “death spiral” in which hemodynamic recovery becomes impossible.

Lipid-lowering therapy and converting enzyme inhibitors have been the subject of substantial investigation,⁵² although their role is in the chronic setting rather than during critical care. Two prospective randomized trials failed to show clear benefits of a statin^53,54 in slowing disease progression. Some preliminary evidence suggests potential benefits of angiotensin-converting enzyme (ACE) inhibitors in decreasing progression of aortic valve calcification,^55,56 but the available clinical data do not show slowing of AS progression.⁵⁷ Both lipid-lowering and converting enzyme inhibitors need to be tested in larger populations earlier in the course of aortic valve disease.

Percutaneous Interventions

Physical Examination

The hallmarks of acute severe AI include features consistent with congestive heart failure and low cardiac output including tachycardia, dyspnea, and signs of impaired cardiac output such as peripheral vasoconstriction. The wide pulse pressure that is a hallmark of AI may or may not be seen, in part because it is dependent on LV compliance. The classic AI murmur may not be heard because of the limited gradient between the aorta and left ventricle during diastole when LV diastolic pressures in some cases approach systemic diastolic pressures. Tachycardia and diminished effective forward flow in acute AI may offset the augmented systolic and wide pulse pressure seen in later stages of the disease, and these patients may have normal or occasionally diminished pulses, although more moderate acute AI may feature the more classic findings. Early closure of the mitral valve may result in a soft first heart sound, and distortions of leaflet anatomy may result in lack of a distinct aortic second heart sound. In some cases absence of a second heart sound in a patient presenting with cardiogenic shock may be the most prominent physical finding.⁷⁸

The dramatic physical findings in chronic AI are among the most familiar to physicians and trainees. A wide pulse pressure results primarily because of increased stroke volume, which augments systolic pressure,⁸⁵ and low aortic diastolic pressures because of volume runoff into the left ventricle. This in turn results in a large variety of associated eponymous physical findings including bounding carotid and peripheral pulses (Corrigan’s and water hammer pulses, respectively); Hill’s sign (dramatically higher systolic pressure in the legs than the arms because of the exaggerated effect of harmonics on systolic pressure waveforms, variously described as having a 20 or 40 mm Hg threshold); Duroziez’s sign (to and fro murmur over the femoral arteries); and Traube’s sign (“pistol shot” sound over the femoral artery during simultaneous compression). With LV dilation, the apical impulse is displaced laterally.

The intensity of the classic diastolic decrescendo murmur heard over the left midsternal border generally correlates well with the severity of AI,⁸⁶ although in later stages of decompensation (or in acute AI), when the gradient between the aorta and the left ventricle disappears in later stages of diastole, the murmur typically shortens. When the murmur is better heard over the right (rather than left) sternal border, the cause of the AI may be secondary to a dilated aortic root rather than a primary leaflet abnormality.⁸⁷ Two other murmurs are frequently heard with moderate to severe AI: a systolic ejection murmur consistent with increased transvalvular flow and on occasion an Austin-Flint murmur. The latter is the apical diastolic rumble occasionally confused with mitral stenosis and thought to originate from vibration of the anterior mitral leaflet caused by the diastolic regurgitant jet originating at the aortic valve; it has been described only in the setting of severe AI.⁸⁷ In rheumatic AI, if a diastolic rumble is heard, mitral stenosis should be excluded. An S₃ gallop is most likely to be present when severe AI occurs in a setting of depressed LV function⁸⁸ but may merely reflect substantial volume loading; the finding is not specific.

Noninvasive Evaluation

As with AS, noninvasive assessment has replaced cardiac catheterization for determining severity of AI. It provides both structural and physiologic information including visualization of the leaflets, annulus, and aortic root; semiquantitative assessment of the severity of AI; and characterization of LV size, hypertrophy, and function, as well as other structural heart disease. Transesophageal echocardiography has superior diagnostic sensitivity to transthoracic echo for certain parameters, in particular for detection of valvular vegetations.⁸⁹ A wide range of techniques for measurement of AI severity including width of the color Doppler jet compared with size of the LV outflow tract, width of the vena contracta, regurgitant volume, regurgitant fraction, and regurgitant orifice area are described in Chapter 8. The correlation of these parameters with severity of AI is listed in Table 32.1. Additional findings of importance are rate of jet velocity deceleration as diastasis is achieved between aortic and LV diastolic pressures and duration of diastolic flow reversal in the aorta.⁹⁰ In addition to echocardiography, cine MRI can assess severity of AI, size of chamber volumes, LV mass, wall thickness, and systolic function.⁹¹ LV size and function have also been followed serially by radionuclide ventriculography.⁹²

The size and function of the LV have been used to set thresholds at which AVR should be considered. The ACC/AHA guidelines⁵ and the European Society of Cardiology guidelines⁹³ have set slightly different values for class I and II recommendations. The echocardiogram-based recommendations, which vary with clinical settings, use thresholds that include LV end-diastolic dimension greater than 70 to 75 mm and end-systolic dimension greater than 50 to 55 mm, aortic root dimension greater than 50 to 55 mm (the lower threshold is for patients with bicuspid valves or Marfan syndrome),⁹³ and EF less than 50%. As the observational data have grown, these numbers have shifted slightly from the familiar “rule of 55s,” which recommended surgery for asymptomatic patients with EF less than 55% or LV end-systolic dimension greater than 55 mm. An important consideration is that the end-systolic and end-diastolic dimensions should be adjusted downward for patients with small body surface area. Keeping in mind that these threshold values can be confounded by comorbid conditions that affect chamber size and function including ischemic and other forms of cardiomyopathy, as well as multivalvular disease, is important.

The ECG was once an important tool in the serial assessment of patients with AI. Similar to AS, features of LV hypertrophy may be present. Progression of LV dilation and dysfunction were thought to correlate with development of an LV strain pattern, a phenomenon that has been confirmed.⁹⁴ As a result, the use of digitalis glycosides, which can cause a similar strain pattern, was discouraged in order to maintain relative specificity of the finding. Conduction disturbances remain an important feature, particularly heart block in the setting of endocarditis, discussed subsequently.

Cardiac Catheterization

As with the other valvular diseases discussed in this chapter, cardiac catheterization is indicated only for patients with inconclusive noninvasive findings or when noninvasive findings are discordant with the rest of the clinical picture. Typical findings include a wide aortic pulse pressure and an exaggerated rise in LV diastolic pressure, reflecting continuous retrograde filling during diastole (see Fig. 32.9). Cardiac catheterization in this setting normally includes aortic root angiography; although the latter requires a significant dye load (typically 30 mL/second for 2 seconds) to be diagnostically accurate. The hallmark of severe AI is greater opacification of the left ventricle than the aorta. The indications for coronary angiography in this setting are listed in Box 32.1.

Medical Management

In acute AI, medical therapy is targeted to stabilization pending AVR. Nitroprusside has been the standard for reducing peripheral vascular resistance and improving net forward flow across the aortic valve since the 1970s.⁹⁵ The use of arterial vasodilators is more complex when the patient is already hypotensive, and a combination of inotropes and afterload reduction should be considered. Although afterload reduction in this setting is highly beneficial, the otherwise optimal combination of afterload reduction and augmented pressure normally provided by intra-aortic balloon pumping in cardiogenic shock (see Chapters 7 and 30) is not a viable choice in AI. Counterpulsation during diastole increases the severity of AI with potentially catastrophic consequences; it represents an absolute contraindication to the intra-aortic balloon pump.

Besides vasodilator therapy, maneuvers to raise the heart rate may be helpful, especially in patients who are bradycardic or have a normal heart rate. Tachycardia disproportionately decreases the diastolic filling period, thereby decreasing the duration of regurgitation across the aortic valve. Both isoproterenol or electrical pacing to raise the heart rate have been used with therapeutic benefit.

In chronic AI, the evidence base is incomplete and somewhat controversial. The general consensus is that asymptomatic patients with mild to at most moderate AI with normal LV function who are not hypertensive do not require or benefit from vasodilator therapy.81,83 Only when patients are not surgical candidates is vasodilator use in chronic AI a class I indication.⁵ Class II indications are short-term treatment to optimize the patient’s hemodynamic status prior to AVR and, more controversially, long-term therapy for asymptomatic patients with severe AI but preserved LV function. Two important studies with divergent results have been performed: One, using nifedipine, appeared to demonstrate some slowing of progression to need for AVR compared with patients on digoxin.⁹⁶ The second, a placebo-controlled study, failed to demonstrate clear benefits of vasodilator therapy with either nifedipine or an ACE inhibitor,⁹⁷ but there are concerns regarding the size of the study, dose of drug used, and modest severity of disease at baseline. Oral vasodilators that have been described as having potential therapeutic benefit include several dihydropyridine calcium channel blockers (nifedipine and felodipine), as well as hydralazine. The appropriateness of vasodilator therapy regardless of severity of disease is clearer in patients with systemic hypertension, in whom reduction of afterload has dual benefits. The dose of drug should be sufficient to show at least some reduction in systolic pressure. There is no evidence base to recommend other forms of medical therapy, such as nitrates or drugs commonly used to treat congestive heart failure to prevent progression of disease in chronic asymptomatic AI. In the critical care setting, however, inotropes and vasodilators are appropriate if patients have congestive heart failure or a low output state, or both.

Aortic Valve Replacement

Two factors primarily determine when a patient should be referred for valve replacement: the presence of symptoms and findings of LV systolic dysfunction.⁵ Surgery results in superior long-term survival than medical therapy alone for symptomatic patients.^84,98 Similarly, patients with LV dysfunction have an inverse correlation between EF and mortality risk.⁹⁹ Three ACC/AHA class I indications for surgery, all for patients with severe AI, exist: (1) patients who have symptoms regardless of LV function; (2) patients with LV systolic dysfunction (EF less than or equal to 50%) regardless of symptoms; and (3) patients undergoing coronary bypass surgery or surgery on other heart valves or the aorta.

ACC/AHA class II indications use the thresholds defined by LV dimensions, again for patients with severe AI, using 55 mm as the end-systolic threshold and 75 mm as the end-diastolic threshold, but lowering those values by 5 mm when the rate of LV dilation is progressive or when patients have deteriorating exercise tolerance or an abnormal hemodynamic response to exercise. Patients with only moderate AI who are undergoing surgery on the aorta or coronary bypass are also considered to have class II indications.

The question of whether it is ever too late to perform AVR in patients with AI has been raised.⁸⁵ Because the disease involves high afterload and wall stress, there is postoperative improvement in EF¹⁰⁰ even in patients with severe baseline LV dysfunction. In general there is substantial reversal of the physiologic adaptations required by chronic severe AI in the postoperative state.¹⁰¹ Nevertheless, delay in referring patients for surgery is deleterious, with significant negative impact on outcomes¹⁰²; the prognosis is adversely affected both by the severity and duration of LV dysfunction.¹⁰³ This includes a fourfold increase in operative mortality rate for patients with EF less than 35% (14%) compared with those with EF greater than 50% (3.7%). Thus the trend has been for earlier referral of symptomatic patients and more aggressive noninvasive monitoring for early detection of deterioration in LV function or dimensions.

Special consideration should be given to patients with acute aortic valve endocarditis. Deteriorating hemodynamics, as well as extension of infection as manifest by progressive heart block and ring abscess formation on echo (Fig. 32.10), are among several factors that mandate early surgery¹⁰⁴; despite the semielective nature of the operation (with an operative mortality rate close to 10%) and the theoretical risk of an infected prosthesis, the long-term outcomes have been satisfactory, with 75% 10-year survival rate after hospital discharge.¹⁰⁵

Physical Examination

In acute MR the abrupt rise of left atrial pressure results from torrential retrograde flow, although the gradient between the left atrium and left ventricle may narrow substantially near the end of systole because of the substantial rise in left atrial pressure as shown in Figure 32.13. Thus the systolic murmur may decrease in intensity prior to the onset of the aortic second heart sound. A systolic thrill may be noted. The sudden and severe left atrial hypertension typically results in pulmonary edema with accompanying physical findings, as well as an increase in the pulmonic second sound and paradoxical splitting of the second heart sound. Because the adaptive mechanisms are not yet in place, features of LV volume overload are not seen, although a prominent precordial impulse, reflecting increased (albeit retrograde) ejection, may be noted. Unlike AI, the pulse pressure is not increased and may be reduced to reflect the lower forward stroke volume.

In chronic MR, precordial features of volume overload include a displaced point of maximal impulse and prominent precordial pulsation. An S₃ gallop is heard, especially with marked dilation of the left ventricle,⁸⁸ and a diastolic flow murmur representing increased transvalvular flow (rather than mitral stenosis) may be present. The first heart sound may be soft, reflecting lack of brisk apposition of the mitral valve leaflets. The murmur of MR commences immediately following the first heart sound, reflecting the observation that a significant amount of regurgitation takes place during what would otherwise be isovolumic systole (which normally commences immediately following mitral valve closure). The murmur of chronic MR is characteristically blowing; may be medium or high pitched; is best heard at the apex; radiates to the axilla and left scapula (when severe); and extends throughout systole, reflecting the presence of a relatively high gradient throughout LV ejection.

The location of radiation can provide a clue to the affected leaflet: anteriorly directed jets result from posterior leaflet dysfunction, and conversely anterior leaflet prolapse results in a murmur best heard toward the patient’s back. Because the Gallavardin phenomenon in AS mimics an MR murmur, it is important to differentiate these two when possible (AS and MR frequently occur concurrently, given the high LV systolic pressure with the former). The murmur of MR differs because it is holosystolic and increases rather than decreases with handgrip.

The intensity of the MR murmur reflects a number of phenomena relating to the size of the gradient, compliance of the left ventricle and left atrium, and contractility of the left ventricle but does not correlate well with extent of MR; thus, the murmur may be difficult to hear or inaudible, especially in the critical care setting.¹²²

Noninvasive Evaluation

Echocardiography provides visualization of the multiple structures responsible for causing MR including the valve leaflets, annulus, chordae tendineae papillary muscles, and left ventricle (see Fig. 32.11). Transesophageal echo provides superior imaging of the valve including assessment for vegetations in case endocarditis is suspected and better visualization of flail leaflets; it is particularly useful when the transthoracic views are incomplete or inadequate. In general, transesophageal echo is superior for fully assessing characteristics of the mitral regurgitant jet¹²³ and provides additional information in evaluating mitral valve morphology and motion. As with AI, Doppler methods allow assessment of the width of the regurgitant jet and size of the vena contracta, and semiquantitative measures of the regurgitant volume, regurgitant jet, and effective orifice area (see Table 32.1) are discussed in detail in Chapter 8. In patients presenting with acute heart failure and a hyperdynamic left ventricle, acute MR should always be considered.

Cardiac Catheterization

Although the indications for cardiac catheterization (see Box 32.1) are similar to those for AS and AI, an additional consideration is included as a class I indication: to assess pulmonary artery pressures at rest and if necessary with exercise to evaluate appropriateness for surgery (see valve replacement indications later). The size of the V wave is not an accurate indicator of MR severity and is not used as a decision-making criterion for surgery because it may be unremarkable even in severe MR when there is high left atrial compliance. In addition, the size of the V wave is typically recorded by wedge rather than direct left atrial pressure, which can result in significant underestimation of the severity of MR (Fig. 32.14). Right heart catheterization and oxygen saturation sampling allow differentiation between the two most common causes of abrupt mechanical deterioration in acute MI: severe MR and ventricular septal defect. Both conditions can feature a large V wave because of sudden volume loading of a noncompliant left atrium: secondary to retrograde flow across the mitral valve in the case of acute MR and antegrade flow of shunted blood from the LV across the pulmonary circulation in the case of a ventricular septal defect.

Medical Management

For acute, severe, symptomatic MR, the primary medical therapy consists of aggressive vasodilation with nitroprusside. It provides improved aortic flow, diminishes MR, and decreases left-sided filling pressures. In the setting of acute MR with shock, nitroprusside alone is usually considered undesirable because of poor end-organ perfusion if further hypotension is caused by the drug. Intra-aortic balloon pumping is highly effective in this setting¹²⁴ (see Chapter 7) because it lowers aortic impedance, improves cardiac output, and decreases regurgitant fraction at the same time as it restores blood pressure. A combination of an inotrope (typically dobutamine; a vasoconstrictor should be avoided) and nitroprusside can have similar salutary effects. When the acute MR is dynamic (i.e., occurs and resolves abruptly), the cause is frequently acute ischemia, and consideration should be given to evaluating the coronary circulation. In a small number of cases, acute coronary intervention has been documented to alleviate MR in the setting of acute ischemia.¹²⁵

Therapy for chronic MR is more controversial. Because afterload is typically not increased, the benefits of afterload reduction are without clear physiologic foundation. As such, ACE inhibitors have consistently been shown not to be effective for chronic MR.126–128 When concomitant systemic hypertension exists, physiologic benefits appear more convincing.¹²⁹ If patients have LV dysfunction superimposed on chronic MR, medical therapy with ACE inhibitors may be beneficial, and several studies have observed functional improvement with resychronization in this setting.¹³⁰ As with other valvular disorders, new-onset AF can result in abrupt hemodynamic deterioration and cardioversion should be considered with adjunctive pharmacotherapy to maintain sinus rhythm and rate control. Because of the sympathetic activation previously discussed, beta blockers appear to be beneficial^131,132 and may have additive benefit when used in combination with ACE inhibitors.¹³³ Both classes of drugs appear to be therapeutic when used after mitral valve repair.¹³⁴

Mitral Valve Surgery

Three types of surgeries are performed on the mitral valve: repair, replacement with preservation of the subvalvular apparatus and attachments, and replacement with transection of the chordae. Mitral valve repair has been consistently shown to improve outcomes compared with mitral valve replacement, both acutely and long term, with a fourfold lower operative mortality rate and a 30% improvement in 10-year survival rate.¹³⁵ With mitral valve replacement, patients who nevertheless have the chordal apparatus preserved have superior long-term outcomes compared with those who have the chords transected.¹³⁶ An additional benefit of repair is the avoidance of anticoagulation if patients are in sinus rhythm compared with replacement with a mechanical valve when anticoagulation is required.

In general, the timing of surgery for chronic MR should be prior to the onset of LV dysfunction. The most commonly used markers are symptoms and echocardiographic parameters including LVEF less than 60%.⁵ Patients who develop class III or IV symptoms, even transiently, have a nearly 10-fold increase in annual mortality rate,¹³⁷ with evidence that early surgery in the setting of severe MR improves overall survival. The 60% threshold for surgery has been shown to correlate with postoperative LV function¹³⁸ as well as survival.¹³⁹ More recently the effective regurgitant orifice has been shown to be strongly predictive of 5-year mortality risk, with patients having an orifice of 40 mm² or greater having an odds ratio of death greater than 5:1 compared with patients with orifice size of 20 mm² or less.¹⁴⁰ Preoperative end-systolic diameter of 40 mm or less has also been associated with superior outcomes. Patients with AF in general have a worse prognosis,¹⁴¹ although in large part because of embolic events rather than LV dysfunction, as do patients with pulmonary hypertension or right-sided heart failure, or both.¹⁴² Mitral valve surgery, even in the setting of depressed EF, may result in postoperative return to normal function.¹⁴³ Comparison of long-term outcomes between operated and medically treated patients with asymptomatic MR has invariably favored the former,¹⁴⁰ although much of the data depend on nonrandomized registry data. In general, the key to improved function and outcomes is preserving the integrity of the submitral apparatus,¹⁴⁴ and comparison of successful repair versus replacement has indicated that high-volume institutions (>140 mitral operations/year) have the best results with nearly twice the rate of successful repair as low-volume institutions.¹⁴⁵

A common teaching has been that repair or replacement of the mitral valve eliminates the “pop-off valve” represented by incompetent leaflets, resulting in dramatically increased afterload postoperatively. In turn, if significant preoperative LV dysfunction were present, the left ventricle would fail, with hemodynamic deterioration and subsequent death. In fact, careful investigation has demonstrated that postoperative deterioration of LV function was secondary to disruption of chordal integrity¹⁴⁶ and elimination of the favorable effects of the submitral apparatus on valvular-ventricular interaction rather than the elimination of a low-pressure decompression pathway. Afterload following mitral valve surgery with chordal transection does in fact increase, whereas patients subjected to mitral valve surgery with preservation of the chordal apparatus had decreased LV chamber volume and, in keeping with the LaPlace equation, reduced afterload.¹⁴⁷ Thus the recommendations regarding LVEF for surgery have changed dramatically, and successful repair of mitral valves in series of patients with EF less than 25% has been reported with operative risk of less than or equal to 5%.¹⁴⁸ The operative risk is in the 1% to 2% range for symptomatic patients with LV dysfunction but higher EF.¹⁴⁹ Similarly, the Society of Thoracic Surgeons database has reported mortality rates of approximately 5% with EF less than or equal to 30% but 3% with EF greater than 30%.¹⁵⁰ Propensity analysis of patients eligible for mitral valve repair who did or did not undergo the procedure has not shown a survival advantage in patients with mean EF less than 25%,¹³⁴ although a randomized trial has not been reported. Similarly, concomitant mitral valve surgery in patients undergoing coronary artery bypass graft (CABG) with moderate or greater MR did not clearly improve short-term survival.¹⁵¹ Combining mitral valve surgery with a mesh placed around the heart to help prevent dilation and thus lower wall stress may have additional benefit for patients with MR and LV dysfunction.¹⁴⁹ Only in the setting of an EF of less than 30% or end-systolic dimension greater than 55 mm, as well as anatomy or surgical experience that suggests chordal preservation is unlikely, do the current guidelines recommend medical therapy alone.

In symptomatic acute severe MR, mitral valve surgery is a class I indication regardless of level of LV function. Both with infective endocarditis,¹⁰⁴ as well as with papillary muscle rupture, surgery is mandated to be performed with as little delay as possible.¹⁵² However, the overall risk of mortality when MR accompanies acute myocardial infarction is substantially increased,¹⁵³ a phenomenon noted in patients undergoing emergent percutaneous intervention as well.¹⁵⁴ Chronic severe MR is much more complex, and the management algorithm is outlined in Figure 32.15.

Future Directions

A number of innovative technologies are being developed for percutaneous treatment of MR. The most extensively studied is based on a catheter-based edge-to-edge repair of the mitral leaflets designed to mimic a surgical procedure developed by Alfieri and colleagues.¹⁵⁵ The EVEREST (Endovascular Valve Edge-to-edge REpair STudy) II trial¹⁵⁶ compared 184 patients randomized to percutaneous versus 95 patients randomized to surgical mitral valve repair and had promising though somewhat nuanced results: although there was good safety profile and similar quality of life improvement, grade 2 or greater mitral insufficiency was present in nearly half the percutaneous repair patients at 1 year, compared with only 17% of the surgical repair patients. The surgical patients did have more adverse events, primarily blood transfusions. The surgical approach appeared clearly superior for patients with degenerative MR, with relative parity in outcomes in functional MR patients (those with structurally normal mitral valves, but typically dilatation of the mitral annulus). The Alfieri procedure is an uncommonly performed type of mitral valve repair, and as with most repair procedures typically involves placement of a prosthetic ring, an option not available with the percutaneous approach, and potentially a limitation of its long-term applicability as an isolated intervention.¹⁵⁷ In this context, it is worth noting that several percutaneous technologies are under development to constrict the coronary sinus that at least partially surrounds the mitral valve¹⁵⁸; they thereby attempt to mimic the mechanical effects of a surgically placed mitral ring. For now, percutaneous edge-to-edge repair has had fair acceptance outside the United States and is being performed primarily in the setting of functional MR. A number of other technologies are under development.

Physical Examination

The physical examination in HOCM features an abrupt early rise in systolic pressure followed by obstruction to LV outflow, resulting in a blunted pressure rise later in systole. This translates to the spike and dome pattern characteristic of HOCM on palpation of the carotid pulse. The typically hypertrophic septum (1.3 to 1.5 times the thickness of the posterior wall and typically 18 mm or greater in clinically significant HOCM) obstructs the outflow tract after initial ejection as LV volume decreases. Physiologic maneuvers or clinical states that lower LV volume abruptly such as standing, dehydration, or use of diuretics can substantially exacerbate the gradient and increase the intensity of the systolic ejection murmur characteristic of HOCM. Because the obstruction begins after the initial brisk ejection of blood from the ventricle, the onset of murmur is late after the first heart sound. It is best heard over the left base of the heart. In contrast to AS, the murmur typically does not radiate to the carotids, does not extend throughout systole, and decreases with squatting. The blunting of the carotid upstroke, so characteristic of AS, is not seen because of the vigorous early ejection.

Noninvasive Evaluation

Abnormalities of mitral valve function are common. Systolic anterior motion of the mitral valve is associated with further obstruction¹⁷⁰; in addition, MR is a typical secondary finding in HOCM. The ECG features LV hypertrophy, strain, and prominent Q waves in the inferior and precordial leads, with narrow QRS complexes being common. The echocardiogram is diagnostic for HOCM, providing both an anatomic basis and physiologic evidence of the degree of obstruction (Fig. 32.16).

Cardiac Catheterization

Cardiac catheterization in HOCM may reveal the standard triad of the Brockenbrough-Braunwald-Morrow sign in the postextrasystolic heartbeat: increase in the gradient between the left ventricle and the aorta, decrease in systemic blood pressure, and narrowing of the pulse pressure (Fig. 32.17). Other stimuli that increase outflow obstruction and increase gradient are Valsalva maneuver and amyl nitrite inhalation, in addition to postextrasystolic potentiation.

Medical Management

The management of critically ill patients with HOCM focuses on several core principles: optimizing LV volumes, specifically increasing preload while not decreasing afterload, decreasing contractility, and maintaining the patient in sinus rhythm. Thus the typical regimen of diuretics, vasodilators, and inotropes used in most patients presenting with congestive heart failure who do not have HOCM can severely exacerbate the outflow gradient and lead to further decompensation or death¹⁷¹ (as can intra-aortic balloon pumping); somewhat counterintuitively in this setting, negative inotropes, in particular beta blockers, combined with volume loading can stabilize patients.^172,173 Other negative inotropes including verapamil and disopyramide appear to have favorable hemodynamic effects and may represent an alternative to beta blockade.¹⁷⁴ In the patient in shock, if vasopressors are required, agents such as phenylephrine, which are primarily peripheral vasoconstrictors, should be used rather than inotropic agents. Verapamil has negative inotropic benefits and some utility as an antiarrhythmic agent, as well as diastolic relaxation properties, and has been used successfully in both chronic therapy and acutely decompensated patients with HOCM.¹⁷⁵ Verapamil does have some potentially undesirable vasodilator properties and can cause synergistic depression of conduction and contractility in combination with beta blockade; the effects of combined therapy, if used, should be monitored closely.¹⁷⁶

Patients with hypertrophic heart disease are highly dependent on the atrial contribution to cardiac output,¹⁷⁷ and patients with obstruction in particular depend on the additional LV filling from atrial contraction. AF is a cause of acute decompensation in HOCM patients, and cardioversion should be considered acutely. AF occurred in approximately one fourth of patients with HCM followed for 9 years and is predictive of heart failure–related death and stroke.¹⁷⁸ In terms of the latter, patients with HCM and AF appear particularly predisposed to thromboembolic events¹⁷⁹; anticoagulation has been shown to reduce the incidence significantly.¹⁸⁰ Left atrial size and function predict the onset of AF.¹⁸¹ The Maze procedure and catheter ablation of left atrial pathways have been used successfully in small series to prevent recurrent AF.^182,183 Amiodarone has been particularly effective for the maintenance of sinus rhythm in the population with AF and HCM¹⁸⁴ and may also have additional efficacy in preventing sudden death in patients with episodes of nonsustained ventricular tachycardia.^185,186 Ventricular fibrillation appears to be the primary cause of sudden death and can be prevented with implantable defibrillators for either primary or secondary prevention, with activation of the devices in 5% to 11% of patients annually depending on indications for placement¹⁸⁷; episodes of sustained ventricular tachycardia and a history of prior sudden death place patients in the highest risk category.

Percutaneous Intervention or Surgery

Once patients develop refractory symptoms despite pharmacologic management and have significant obstruction to LV outflow, surgical myomectomy or alcohol septal ablation must be considered. The choice is controversial and depends on LV and coronary anatomy, comorbid conditions, severity of coexistent mitral valve or other potentially surgical heart disease, and a variety of other clinical issues; equivalence of outcomes with the less invasive percutaneous approach has not been demonstrated.¹⁸⁸ Typical thresholds for intervention are a resting peak instantaneous gradient of 50 mm Hg or more, although depending on degree of symptoms and provokability, a variety of other thresholds have been applied.¹⁶⁶ Mitral valve replacement can frequently be avoided, especially in patients without intrinsic mitral disease, in whom MR is largely abolished by myomectomy and relief of the outflow gradient.¹⁸⁹ In patients with intrinsic mitral valve disease or for a variety of other anatomic considerations, mitral valve surgery including use of low-profile prostheses is used.¹⁶⁶ Dual-chamber pacing, once considered therapeutically effective,¹⁹⁰ with theoretical benefit from asynchronous septal contraction and secondary reduction in outflow tract gradient, has not been shown to have clear therapeutic benefit in randomized clinical trials, with the most likely functional benefits occurring in a limited subset of elderly patients.¹⁹¹

Physical Examination

The diagnosis of MS on physical examination is frequently missed, in part because it is uncommonly seen by most practitioners outside developing countries. Besides findings suggestive of congestive heart failure with a low output in the critical care setting, several findings correlate strongly with MS. A right ventricular lift may occur in patients with advanced and chronic pulmonary hypertension, and occasionally a palpable pulmonary second sound is felt. Both S₁ and P₂ are increased early in the course of the disease; with progressive deformity of the subvalvular apparatus, S₁ may become quite soft as mitral valve closure velocity decreases. The opening snap is preserved until the end stages of MS, and the severity of the disease correlates inversely with the length of time between S₂ and the opening snap. The diastolic rumble may not be heard if the patient is not placed in the lateral decubitus position and is difficult to hear in the critical care setting. Presystolic accentuation is the result of increased flow in late diastole caused by atrial contraction in patients in sinus rhythm. The diagnosis is often missed in both the outpatient and critical care settings,¹⁹⁵ and patients are sometimes treated for years for bronchitis, pneumonia, or asthma, when in fact the presentation is congestive heart failure with the underlying disease process being MS.

Noninvasive Evaluation

The echocardiogram is the primary tool for diagnosis of MS. The typical hockey stick appearance of the rheumatic anterior mitral leaflet is caused by partial restriction of mobility (Fig. 32.19). Other important characteristics of the mitral leaflets are degree of thickening, mobility, and calcification, all of which, along with extent of subvalvular disease, help to address suitability for balloon dilation or commissurotomy rather than valve replacement.¹⁹⁶ It is important to differentiate rheumatic MS from severe nonrheumatic MS, as shown in Figure 32.19. Transesophageal echo provides the requisite sensitivity to detect left atrial thrombus prior to balloon commissurotomy or cardioversion. Severity of mitral stenosis is graded by physiologic and anatomic criteria: The former depend on Doppler interrogation of transvalvular flow, with peak velocity used to estimate the gradient; rate of decompression of the left atrial pressure, or pressure half time, correlates inversely with valve area. The valve area is frequently directly measured using planimetry of a cross-sectional view. Because these methods have potential for error, a number of alternative methodologies have been proposed,^197–199 although there is no agreement on a gold standard.²⁰⁰ As with AS, a dramatic increase in gradient accompanies increased flow related to exercise or dobutamine infusion²⁰¹ (Figs. 32.20 and 32.21). Cardiac catheterization may be required if the noninvasive data are inconclusive; there is a tendency for the pulmonary wedge pressure/LV gradient to overestimate the severity of MS compared with left atrial/LV gradient measurement²⁰² (Fig. 32.22). Cardiac catheterization can also be used to assess pulmonary artery pressures and response to exercise if the clinical and noninvasive pictures are discordant.⁵

Medical Therapy

Medical therapy specifically aimed at improving hemodynamics in MS is basically twofold: slowing the heart rate and anticoagulation. In sinus rhythm, beta blockade is the preferred therapy; patients with significant bronchospasm or with coexistent, usually secondary, severe pulmonary hypertension with or without right-sided heart failure may not tolerate beta blockers. Diltiazem or low doses of beta blocker combined with diltiazem or verapamil are sometimes used in this setting. With combined therapy, care should be exercised that severe heart rate slowing or atrioventricular block does not occur, and verapamil may not be tolerated in right-sided heart failure. For the acutely decompensated patient with MS, diuretics, nitrates, inotropes (preferably agents that do not raise heart rate), and, to a lesser degree, arterial vasodilator therapy may all be beneficial.

Approximately 30% of patients with MS are thought to develop AF, with the risk increased by progressive increase in left atrial size. The risk of embolic events, especially stroke, is substantially higher than in AF patients without MS²⁰³ (up to 15% in unanticoagulated patients), possibly in part because of a hypercoagulopathy and increased platelet activation associated with MS,²⁰⁴ making the use of anticoagulation essential. It is not uncommon for the presenting finding to be an embolic event in a previously undiagnosed MS patient. Even in the setting of sinus rhythm,⁵ if spontaneous echo contrast is seen in the MS patient or left atrial size is greater than 55 mm, anticoagulation may decrease stroke risk²⁰⁴ and it should also be used in the sinus rhythm patient with prior embolic event or with left atrial thrombus detected by transesophageal echocardiography.

In the setting of new-onset AF and acute decompensation, cardioversion should be considered. Regardless of chronicity, heart rate slowing is essential; digoxin, beta blockers, diltiazem, or verapamil alone or in combination may be beneficial. Rate control only at rest is inadequate; many patients with low resting heart rates become tachycardic with minimal exercise. This is a vicious circle because the tachycardia may be a response to inadequate stroke volume; the tachycardia then lowers the diastolic filling period and further exacerbates the gradient. Control of tachycardia is important beyond its hemodynamic benefit; it also helps avoid tachycardia-induced cardiomyopathy.²⁰⁵

Percutaneous Intervention or Surgery

Percutaneous balloon mitral valvuloplasty is equivalent or superior to surgery in patients with favorable valve anatomy,206,207 characterized by mild to moderate impairment of leaflet mobility, valve thickness, valve calcification, and subvalvular disease, the components of the Wilkins-Weyman echo score.²⁰⁸ The scoring system has been shown to correlate with long-term outcomes.²⁰⁹ Surgery should be reserved for patients with unfavorable anatomy, more than mild MR, or persistent thrombus in the left atrium despite anticoagulation.⁵ The threshold for percutaneous or surgical intervention is a mitral valve area of 1.5 cm² or less. If patients are symptomatic and have favorable anatomy or if they are asymptomatic but have pulmonary hypertension (>50 mm Hg systolic at rest or >60 mm Hg with exercise), they have class I indications. Patients who have unfavorable anatomy and are functional class III or IV are considered to have a class II indication, especially if they are poor candidates for surgery. Asymptomatic patients with new-onset AF and favorable anatomy and symptomatic patients with mild MS (valve area >1.5 cm²) who have pulmonary artery pressure greater than 60 mm Hg, wedge pressure 25 mm Hg or greater, or gradient 15 mm Hg or greater during exercise are also considered class II candidates.⁵