2020 ACC/AHA Guideline for the Management of Patients With Valvular Heart Disease: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines

1. Recurrent rheumatic fever is associated with a worsening of rheumatic heart disease. However, infection with group A streptococcus does not have to be symptomatic to trigger a recurrence, and rheumatic fever can recur even when the symptomatic infection is treated. Prevention of recurrent rheumatic fever requires long-term antimicrobial prophylaxis rather than recognition and treatment of acute episodes of group A streptococcus pharyngitis. The recommended treatment regimens and duration of secondary prophylaxis are shown in Tables 6 and 7.

The risk of developing IE is highest in patients with a prosthetic valve, prior IE, or congenital heart disease with residual flow disturbances.³ IE has been reported to occur after transcatheter aortic valve implantation (TAVI) at rates equal to or exceeding those associated with surgical aortic valve replacement (SAVR) and is associated with a high 1-year mortality rate of 75%.^23,24 IE may also occur after valve repair with prosthetic material, which results in high in-hospital and 1-year mortality rates, even with surgical intervention.^25,26 IE appears to be more common in heart transplant recipients than in the general population, according to limited data.³ The risk of IE is highest in the first 6 months after transplantation because of endothelial disruption, high-intensity immunosuppressive therapy, frequent central venous catheter access, and frequent endomyocardial biopsies.³ Persons at risk of IE can reduce potential sources of bacterial seeding by maintaining optimal oral health through regular professional dental care and the use of appropriate dental products, such as manual, powered, and ultrasonic toothbrushes; dental floss; and other plaque-removal devices.

Transient bacteremia is commonly seen in routine activities such as brushing teeth and flossing (20% to 68%), using toothpicks (20% to 40%), and simply chewing food (7% to 51%). The incidence of IE after most procedures is low, with no controlled data supporting the benefit of antibiotic prophylaxis. Indiscriminate use of antibiotics can be associated with the development of resistant organisms, Clostridium difficile colitis, unnecessary expense, and drug toxicity. The rate of transient bacteremia during or immediately after endoscopy is 2% to 5%, and the organisms typically identified are unlikely to cause IE.^11,27,28 The rate of bacteremia does not increase with biopsy, polypectomy, or sphincterotomy. Some gastrointestinal procedures are associated with rates of bacteremia higher than that for simple endoscopy; these procedures include esophageal dilation (as high as 45%), sclerotherapy (31%), and endoscopic retrograde cholangiopancreatography (6% to 18%).²⁹ However, no studies have shown reduced rates of IE with antibiotic prophylaxis. Surgery, instrumentation, or diagnostic procedures that involve the genitourinary tract may cause bacteremia. In the absence of infection, the rate of bacteremia after urinary tract procedures is low. In patients with bacteriuria, antimicrobial therapy before elective procedures, including lithotripsy, typically is provided.³⁰

The 4 large RCTs^8–14 comparing NOACs with warfarin included small numbers of patients with VHD, prior valve repair, and bioprosthetic valves (excluding moderate to severe rheumatic MS and mechanical heart valves). In addition to the subsequent meta-analyses,^1,15–17 examinations of insurance claims data and large registries¹⁸ have consistently confirmed no signal for a differential effect between NOAC and VKA therapy.^19,20 More consistently observed is a net clinical benefit, with fewer events in patients using NOACs than in patients on VKA therapy. Validation of the CHA₂DS₂-VASc risk schema in patients with VHD (excluding moderate to severe rheumatic MS and mechanical heart valves) has been performed in large registries,² confirming the applicability of this score. Bioprosthetic valves do not appear to be independent predictors of thromboembolic events in patients with AF.¹⁹

The coexistence of AF and rheumatic MS is common and confers a substantial risk of thromboembolic events. These patients have been specifically excluded from NOAC trials, yet a single registry study and a US claims database analysis do suggest that NOACs may be potentially preferable.^21,22 These findings need further validation, and currently the use of NOACs cannot be supported over VKA (target international normalized ratio [INR] of 2.5).

Postoperative AF after VHD intervention is associated with increased stroke and mortality rates^3,4 irrespective of the CHA₂DS₂-VASc score. Anticoagulation in this setting may reduce these endpoints. There are conflicting data about the safety and efficacy of NOAC therapy in patients early after implantation of a bioprosthesis.^5,6,23 Until more data are available, the writing committee favors using VKA for patients with AF in the first 3 months after surgical or transcatheter bioprosthetic valve implantation to prevent thromboembolic events. The optimal duration of anticoagulation is not well defined. Repeat evaluation is encouraged in all patients to detect arrhythmia recurrence in the context of their CHA₂DS₂-VASc scores.

The phase II study comparing dabigatran to warfarin (RE-ALIGN [Randomized, Phase II Study to Evaluate the Safety and Pharmacokinetics of Oral Dabigatran Etexilate in Patients after Heart Valve Replacement]) was halted prematurely because of excess stroke and bleeding in the dabigatran group. Until there is an explanation of why these adverse events occurred, there is insufficient evidence to support the use of NOACs for patients with mechanical heart valves.⁷

In patients who have had a valve intervention, most cardiologists continue to see patients for a clinical history and physical examination at annual intervals, or more frequently if needed for symptoms or concurrent conditions. A baseline TTE study is recommended after all valve interventions, including replacement with a prosthetic valve (see Section 11.1). This baseline postprocedural study ideally is performed 1 to 3 months after intervention to ensure loading conditions have returned to normal, but in some cases it may need to be done during the index hospitalization for the patient’s convenience. The timing of subsequent periodic imaging after valve intervention is based on the type of valve prosthesis or repair, length of time after valve intervention, residual valve dysfunction, ventricular size and systolic function, and any concurrent cardiac conditions (Table 12). TTE is the standard approach for periodic imaging, supplemented by TEE when prosthetic mitral valve dysfunction is a concern (see Section 11.1). Additional imaging with CT, fluoroscopy CMR, or PET is reserved for patients for whom there is concern about valve dysfunction (see Section 11.1) or endocarditis (see Section 12.1).^1,2

Hypertension is common in patients with AS, may be a risk factor for AS, and adds to the total pressure overload on the LV in combination with valve obstruction. Concern that antihypertensive medications might result in a decrease in cardiac output has not been corroborated in studies of medical therapy, including 2 small RCTs, likely because AS does not result in “fixed” valve obstruction until late in the disease process. In 1616 patients with asymptomatic AS in the SEAS (Simvastatin Ezetimibe in Aortic Stenosis) study, hypertension (n=1340) was associated with a 56% higher rate of ischemic cardiovascular events and a 2-fold higher mortality rate (both P<0.01) than those seen in normotensive patients with AS, although no impact on progression of valve stenosis leading to symptoms requiring AVR was seen. Medical therapy for hypertension follows standard guidelines, starting at a low dose and gradually titrating upward as needed to achieve blood pressure control. There are no studies addressing specific antihypertensive medications in patients with AS, but diuretics may reduce stroke volume, particularly if the LV chamber is small at baseline. In theory, ACE inhibitors may be advantageous because of the potential beneficial effects on LV fibrosis, in addition to control of hypertension. Consideration should be given to a higher target blood pressure for patients with AS than is recommended for the general population, but this is an underexplored area, and further data are needed before a different target blood pressure can be recommended for patients with AS.^1–3,9–13

Concurrent CAD is common in patients with AS, and all patients should be screened and treated for hypercholesterolemia, with GDMT used for primary and secondary prevention of CAD. In RCTs of statin therapy for mild to moderate AS, although aortic valve event rates were not reduced, the rate of ischemic events was reduced by about 20% in the statin therapy group even though these patients did not meet standard criteria for statin therapy.^4–6,14,15

In patients undergoing TAVI, observational and registry data show that those who were treated with renin–angiotensin system blocker therapy after the procedure had a lower 1-year mortality rate than those not treated with renin–angiotensin system blocker therapy, with a relative risk reduction of about 20% to 50% and an absolute risk reduction between 2.4% and 5.0%. When stratified by LVEF, having a prescription for a renin–angiotensin system inhibitor, versus no prescription, was associated with a lower 1-year mortality rate among patients with preserved LVEF but not among those with reduced LVEF.^7,8,16,17

Despite experimental models and retrospective clinical studies suggesting that lipid-lowering therapy with a statin might prevent disease progression of calcific AS, 3 large well-designed RCTs failed to show a benefit, either in terms of changes in hemodynamic severity or in clinical outcomes, in patients with mild to moderate valve obstruction. Thus, at the time of publication, there are no data to support the use of statins for prevention of progression of AS.^7,8,16,17

In symptomatic patients with severe high-gradient AS (Stage D1), ample evidence demonstrates the beneficial effects of AVR on survival, symptoms, and LV systolic function.^35,41–46 The most common initial symptom of AS is exertional dyspnea or decreased exercise tolerance. Clinical vigilance is needed to recognize these early symptoms and proceed promptly to AVR. More severe “classical” symptoms of AS, including HF, syncope, or angina, can be avoided by appropriate treatment at the onset of even mild symptoms. Outcomes after surgical or transcatheter AVR are excellent in patients who do not have a high procedural risk.^41,43–45 Surgical series demonstrate improved symptoms after AVR, and most patients have an improvement in exercise tolerance, as documented in studies with pre- and post-AVR exercise stress testing.^41,43–46 Historical observation studies on outcomes in symptomatic patients with severe AS have been confirmed in RCTs comparing TAVI with palliative care in patients with a prohibitive surgical risk. The choice of surgical versus transcatheter AVR for patients with an indication for AVR is discussed in Section 3.2.4.^{1–3,5,6,12–16,35,42,47–55}

In asymptomatic patients with severe AS and normal LV systolic function, the survival rate during the asymptomatic phase is similar to that of age-matched controls, with a low risk of sudden death (<1% per year) when patients are followed prospectively and when patients promptly report symptom onset. However, in patients with a low LVEF and severe AS, survival is better in those who undergo AVR than in those treated medically. The depressed LVEF in many patients is caused by excessive afterload (afterload mismatch), and LV function improves after AVR in such patients. If LV dysfunction is not caused by afterload mismatch, survival is still improved, likely because of the reduced afterload with AVR, but improvement in LV function and resolution of symptoms might not be complete after AVR.^{17,23,24,56–62}

Prospective clinical studies demonstrate that disease progression occurs in nearly all patients with severe asymptomatic AS. Symptom onset within 2 to 5 years is likely when aortic velocity is ≥4.0 m/s or mean pressure gradient is ≥40 mm Hg. The additive risk of AVR at the time of other cardiac surgery is less than the risk of reoperation within 5 years.^{12–16,63–65}

Mean pressure gradient is a strong predictor of outcome after AVR, with better outcomes seen in patients with higher gradients. Outcomes are poor with severe low-gradient AS but are still better with AVR than with medical therapy in those with a low LVEF, particularly when contractile reserve is present. The document “Echocardiographic Assessment of Valve Stenosis: EAE/ASE Recommendations for Clinical Practice” defines severe AS on dobutamine stress testing as a maximum velocity >4.0 m/s with a valve area ≤1.0 cm² at any point during the test protocol, with a maximum dobutamine dose of 20 mcg/kg per minute.⁶⁶ The recommendation for AVR in these patients is based on outcome data in several prospective nonrandomized studies. LVEF typically increases by 10 LVEF units and may return to normal if afterload mismatch was the cause of LV systolic dysfunction. If dobutamine stress testing indicates moderate, not severe AS, GDMT for HF can be continued without AVR. Patients without contractile reserve may also benefit from AVR, but decisions in these high-risk patients must be individualized because outcomes are poor with either surgical or medical therapy. The role of TAVI in these patients is currently under investigation.^{17,22–24,59,60,67}

A subset of patients with severe AS presents with symptoms and with a low velocity, low gradient, and low stroke volume index, despite a normal LVEF. Low-flow, low-gradient severe AS with preserved LVEF should be considered in patients with a severely calcified aortic valve, an aortic velocity <4.0 m/s (mean pressure gradient <40 mm Hg), and a valve area ≤1.0 cm² when stroke volume index is <35 mL/m². Typically, the LV is small, with thick walls, diastolic dysfunction, and a normal LVEF (≥50%). The first diagnostic step is to ensure that data were recorded and measured correctly. If hypertension is present, blood pressure is controlled before reevaluation of AS severity. Next, valve area is indexed to body size because an apparent small valve area may be only moderate AS in a small patient; an aortic valve area index ≤0.6 cm²/m² suggests severe AS. Transaortic stroke volume is calculated by Doppler or 2D imaging. Measurement of a CT calcium score often is helpful. Evaluation for other potential causes of symptoms ensures that symptoms are most likely attributable to valve obstruction. Although the survival rate after TAVI is lower in patients with low-flow severe AS than in patients with normal-flow severe AS, AVR appears beneficial, with an increase in stroke volume and improved survival as compared with medical therapy.^{18,25–27,54,68–76}

Exercise testing may be helpful in clarifying symptom status in patients with severe AS. When symptoms are provoked by exercise testing, the patient is considered symptomatic and meets a COR 1 recommendation for AVR; symptoms are symptoms, whether reported spontaneously by the patient or provoked on exercise testing. The rate of symptom onset within 1 to 2 years is high (about 60% to 80%) in patients without overt symptoms who demonstrate 1) a fall of ≥10 mm Hg in systolic blood pressure from baseline to peak exercise or 2) a significant decrease in exercise tolerance as compared with age and sex normal standards. Management of patients with a lack of appropriate rise in BP with exercise is less clear. Decisions about elective AVR in these patients include consideration of surgical risk, patient preferences, and clinical factors, such as age and comorbid conditions.^{13,28,77–82}

In patients with very severe AS and an aortic velocity ≥5.0 m/s or mean pressure gradient ≥60 mm Hg, the rate of symptom onset is approximately 50% at 2 years. On multivariable analysis of a large cohort of adults with asymptomatic AS (>500 patients), an aortic velocity ≥5 m/s was associated with a >6-fold increased risk of cardiovascular mortality (hazard ratio [HR]: 6.31; 95% CI: 2.61–15.9).³³ A randomized trial of SAVR versus continued surveillance showed a significant survival benefit to early surgery in patients with aortic velocity ≥4.5 m/s.³¹ In patients very severe asymptomatic AS and low surgical risk, a decision to proceed with AVR or continue watchful waiting takes into account patient age, avoidance of patient–prosthesis mismatch, anticoagulation issues, and patient preferences.^31–33,39

An elevated serum BNP level is a marker of subclinical HF and LV decompensation. In a cohort of 387 asymptomatic adults with severe AS, elevated BNP levels were associated with an increased 5-year risk of AS-related events, with a hazard ratio for a BNP level >300 pg/mL (3 times normal) of 7.38 (CI: 3.21 to 16.9).³² Serum BNP levels also are predictive of symptom onset during follow-up and persistent symptoms after AVR.³⁶

Hemodynamic progression eventually leading to symptom onset occurs in nearly all asymptomatic patients with AS once the aortic velocity reaches ≥2 m/s. Although the average rate of hemodynamic progression for calcific stenosis of a trileaflet valve is an increase in aortic velocity of about 0.3 m/s per year, an increase in mean gradient of 7 to 8 mm Hg per year, and a decrease in valve area of 0.15 cm² per year, there is marked variability between patients in disease progression. Predictors of rapid disease progression include older age, more severe valve calcification, and a faster rate of hemodynamic progression on serial studies. In patients with an aortic velocity >4 m/s in addition to predictors of rapid disease progression, symptom onset is likely in the near future, so there is less benefit to waiting for symptom onset. Thus, elective AVR may be considered if the surgical risk is low and after consideration of other clinical factors and patient preferences.

In adults with initially asymptomatic severe AS, the rate of sudden death is low (<1% per year). However, an aortic velocity ≥5 m/s or an LVEF <60% each is associated with higher all-cause and cardiovascular mortality rates in the absence of AVR.³¹ A multivariate analysis of predictors of death in a large cohort (>500 patients) showed a >4-fold higher risk of cardiovascular death for those with an LVEF <60% than for those with a higher LVEF (HR: 4.47; 95% CI: 2.06 to 9.70).³³ A progressive decrease in LVEF is most likely in those with an LVEF <60% before AS becomes severe.^8,9,11 Evaluation for other causes of a decline in LVEF is appropriate, particularly when AS is not yet severe, but a progressive decline in LV systolic function is of concern and should prompt more frequent evaluation; and consideration of AVR when repeat studies show a progressive decline in LVEF without other cause with a lack of response to medical therapy. The presence of at least 3 serial imaging studies showing a consistent decline in LVEF ensures that the changes seen are not simply attributable to recording, measurement, or physiological variability.^8–11

Hemodynamic progression eventually leading to symptom onset occurs in nearly all asymptomatic patients with AS. The survival rate during the asymptomatic phase is similar to age-matched controls, with a low risk of sudden death (<1% per year) when patients are followed prospectively and when patients promptly report symptom onset. The rate of symptom onset is strongly dependent on the severity of AS, with an event-free survival rate of about 75% to 80% at 2 years in those with a jet velocity <3.0 m/s, compared with only 30% to 50% in those with a jet velocity ≥4.0 m/s. Patients with asymptomatic AS require periodic monitoring for development of symptoms and progressive disease (Section 3.1). In patients with moderate calcific AS undergoing cardiac surgery for other indications, the risk of progressive VHD is balanced against the risk of repeat surgery or TAVI (Sections 4.3.3 and 10). This decision must be individualized on the basis of the specific operative risk in each patient, clinical factors such as age and comorbid conditions, valve durability, and patient preferences.^{13,49,62–64}

1. Although qualitative measures of AR severity are adequate in many situations, when AR is significant (Stages B and C), quantitative measures of regurgitant volume and effective regurgitant orifice (ERO) area¹ are better predictors of clinical outcome.^2,3 Measures of LV systolic function (LVEF or fractional shortening) and LV end-systolic dimension (LVESD) or LV end-systolic volume are predictive of the development of HF symptoms or death in initially asymptomatic patients (Stages B and C1) and are significant determinants of survival and functional results after surgery in asymptomatic and symptomatic patients (Stages C2 and D).^2–18,26 Symptomatic patients (Stage D) with normal LVEF have a significantly better long-term postoperative survival rate than those with depressed systolic function.

2. Auscultation has high specificity for detecting AR but low sensitivity and diagnostic accuracy.²⁷ TTE can identify AR in patients who have been deemed to be at risk on the basis of the presence of known aortic dilation or a condition associated with abnormal aortic valve function, such as a BAV.

3. TTE and CMR are useful for evaluating patients in whom there is discordance between clinical assessment and severity of AR by TTE or when TTE images are suboptimal. CMR imaging provides accurate and reproducible measures of regurgitant volume and regurgitant fraction in patients with AR, as well as assessment of aortic morphology, LV volume, and LV systolic function. Cardiac catheterization with LV and aortic angiography, as well as quantitation of regurgitation severity, is another option.^{20–25,28–30}

Severe AR is associated with a wide pulse pressure, such that systolic blood pressure is higher than in patients without AR even when systemic vascular resistance is normal. Transaortic stroke volume increases further with medications that lower heart rate, such as beta blockers, which may result in a paradoxical apparent increase in blood pressure. Vasodilating drugs, such as ACE inhibitors or ARBs, do not affect heart rate and thus may reduce systolic blood pressure without a substantial reduction in diastolic blood pressure in patients with chronic AR.^1,2,5–8

In symptomatic patients who are candidates for surgery, medical therapy is not a substitute for AVR. However, medical therapy is helpful for alleviating symptoms in patients who are considered to be at very high surgical risk because of concomitant comorbid medical conditions.^5,9

Symptoms are an important indication for AVR in patients with chronic severe AR, and the most important aspect of the clinical evaluation is taking a careful, detailed history to elicit symptoms or diminution of exercise capacity. Patients with chronic severe AR who develop symptoms have a high risk of death if AVR is not performed,⁵² and survival and functional status after AVR are related to the severity of preoperative symptoms, assessed either subjectively or objectively with exercise testing.^1–4 Even among symptomatic patients with a severe reduction in LVEF (<35%), AVR results in improved survival rate.^5–7

LV systolic function is an important determinant of survival and functional status after AVR.^{3–5,8,9,12,53–61} Outcomes are optimal when surgery is performed before LVEF decreases below 55%.^16,25,26 In asymptomatic patients with LV systolic dysfunction, postoperative outcomes are better if AVR is performed before onset of symptoms.⁵³

Patients with chronic severe AR may be referred for other types of cardiac surgery, such as CABG, mitral valve surgery, or surgery for correction of dilation of the aortic root or ascending aorta. In these patients, AVR will prevent both the hemodynamic consequences of persistent AR during the perioperative period and the possible need for a second cardiac operation in the near future. Patients undergoing surgical repair or replacement of the aortic root or ascending aorta may be candidates for aortic valve–sparing procedures.^33–39

LVESD in patients with chronic AR reflects both the severity of the LV volume overload and the degree of LV systolic shortening.^54,62 An elevated LVESD often reflects LV systolic dysfunction with a depressed LVEF. If LVEF is normal, an increased LVESD indicates a significant degree of LV remodeling and is associated with subsequent development of symptoms and/or LV systolic dysfunction and an increased mortality rate after AVR.^17,20,21 Most studies have used unadjusted LVESD, but indexing for body size is important, particularly in women or small patients.^13,19,52 Recent data indicate that the LVESD index threshold for optimal postoperative survival may be even smaller than 25 mm/m²,^14–16 but more outcome data, and ideally an RCT, of earlier intervention are needed. LV volumes may be a more sensitive predictor of cardiac events than LVESD index in asymptomatic patients,¹⁸ but more data are needed to determine the threshold values of LV systolic volume that best predict postoperative outcomes.

In patients with moderate AR who are undergoing other forms of cardiac surgery, such as CABG, mitral valve surgery, or replacement of the ascending aorta, the decision to intervene on the aortic valve concurrently includes consideration of several factors, including aortic valve anatomy, aortic root size and shape, regurgitant severity, other comorbidities, and patients’ preferences and values. Patients undergoing surgical repair or replacement of the aortic root or ascending aorta may be candidates for a valve-sparing procedure.^33–39

LVEDD, a marker of the severity of LV volume overload in patients with chronic AR, is significantly associated with clinical outcomes in asymptomatic patients, and progressive increases in LVEDD are associated with subsequent need for surgery.^{16,17,20,25–28} In asymptomatic patients, it is important to ensure that apparent changes in LV size or LVEF are not due simply to measurement or physiological variability. In addition, confirmation of severe regurgitation by quantitative measures of AR severity with TTE, TEE, or, when needed, CMR provides confidence that AR is the cause of LV dilation or decrease in LVEF. When there is an apparent significant fall in EF or increase in LV size, repeat imaging typically is performed at 3- to 6-month intervals unless there is clinical deterioration.

TAVI for isolated chronic AR is challenging because of dilation of the aortic annulus and aortic root and, in many patients, lack of sufficient leaflet calcification. Risks of TAVI for treatment of AR include transcatheter valve migration and significant paravalvular leak.^29–32 TAVI is rarely feasible, and then only in carefully selected patients with severe AR and HF who have a prohibitive surgical risk and in whom valvular calcification and annular size are appropriate for a transcatheter approach.

Patients with rheumatic MS with AF and prior embolic events are at high risk of arterial embolization when AF or an LA thrombus is present. Treatment with VKA anticoagulation will decrease the incidence of these events.^3–7,16 It is controversial whether long-term anticoagulation should be given to patients with rheumatic MS in normal sinus rhythm on the basis of LA enlargement or spontaneous contrast on TEE.^17,18 Patients with very large left atria have more spontaneous echocardiographic contrast and lower LA appendage Doppler velocities,¹⁹ which have been associated with a higher rate of embolic events, but no data directly link large left atria to embolic events. Non–vitamin K oral anticoagulation has not been studied in patients with rheumatic MS, and these patients were excluded from the randomized AF trials. In addition to the much higher risk of embolization with rheumatic valve disease as compared with other causes of valve disease, there is concern that rheumatic disease also affects the atrial muscle, resulting in an increased risk of blood flow stasis and thrombosis in the body of the LA, as well as the LA appendage.¹ Further studies are required to confirm these findings.²

Patients with rheumatic MS are prone to developing atrial arrhythmias—specifically AF. Significant detrimental hemodynamic consequences may be associated with the acute development of AF, primarily from the rapid ventricular response, which shortens the diastolic filling period and increases LA pressure.¹⁶ The treatment of acute AF consists of anticoagulation and control of the heart rate response with negative dromotropic agents. If the rate cannot be adequately controlled with medications, cardioversion may be necessary to improve hemodynamics. In the stable patient, the decision for rate control versus rhythm control is dependent on multiple factors, including the duration of AF, hemodynamic response to AF, LA size, prior episodes of AF, and history of embolic events. It is more difficult to achieve rhythm control in patients with rheumatic MS because the rheumatic process itself may lead to progressive fibrosis and enlargement of the atria, fibrosis of the internodal and interatrial tracts, and damage to the sinoatrial node.

The use of negative dromotropic agents for the treatment of symptoms in patients with rheumatic MS in normal sinus rhythm has been controversial. Although a reduction in heart rate and prolongation of the diastolic filling period will decrease the transmitral mean gradient, studies have shown that treatment with beta blockade may not improve or may even decrease exercise tolerance, most likely because of the limitation of the cardiac output attributable to a limited stroke volume and chronotropic incompetence.^9–11 Nonetheless, there are now several randomized trials that have examined beta blockers and ivabradine in patients with rheumatic MS and have shown that either drug can increase exercise duration and improve symptoms,^2–15 To explain these differences, the earlier trials that found no benefit of beta blockers were performed in older patients with underlying chronotropic incompetence, whereas the randomized trials showing benefit were performed primarily in younger patients with higher resting and exercise-induced heart rates. Thus, the use of beta blockers or ivabradine to improve symptoms may be effective only in patients who do not have underlying chronotropic incompetence. When medical therapy is considered for relief of symptoms in patients with rheumatic MS, it must be remembered that intervention with PMBC relieves symptoms in those patients with an appropriate valve morphology.

Randomized trials have established the safety and efficacy of PMBC as compared with surgical closed or open commissurotomy in patients with a favorable valve morphology with less than 2+ MR in the absence of LA thrombus.^6,8–12 PMBC is performed by advancing one or more balloon catheters across the mitral valve and inflating them, thereby splitting the commissures. Favorable valve morphology consists of mobile and relatively thin valve leaflets, which are free of calcium, in the absence of significant subvalvular fusion.^18,19,22,23 An anatomic mitral morphology score can be used to determine suitability for PMBC and to evaluate the appearance of the commissures and degree of calcification.^1,24,25 Clinical factors, such as age, NYHA class, and presence or absence of AF, are also predictive of outcome. Older patients with lower gradients (<10 mm Hg) will not have as good an outcome as patients with higher gradients, probably because of other concomitant problems that cause symptoms, such as LV diastolic dysfunction and LA noncompliance, measured by net atrial–ventricular compliance.^26–30 PMBC should be performed only by experienced operators, with immediate availability of surgical backup for potential complications. Long-term follow-up has shown 70% to 80% of patients with an initial good result after PMBC to be free of recurrent symptoms at 10 years, and 30% to 40% are free of recurrent symptoms at 20 years.^1–7

Mitral valve surgery is an established therapy for rheumatic MS, with the preferred approach being commissurotomy (either closed, where the valve is opened blindly through the LA or LV, or open, which allows more extensive surgery under direct visualization) when anatomy is favorable.^31–36 However, in the presence of severe valvular thickening and subvalvular fibrosis with leaflet tethering, mitral valve replacement may be the best option. In addition to those who have suboptimal valve anatomy (or failed PMBC), patients with moderate or severe TR may also have a better outcome with a surgical approach that includes tricuspid valve repair.³⁷ Patients undergoing surgical commissurotomy at centers with a high level of expertise may have better long-term outcomes than those undergoing PMBC.^6,7 Because the natural history of rheumatic MS is one of slow progression over decades, surgery should be delayed until the patient has severe limiting symptoms (NYHA class III or IV), particularly if mitral valve repair is contemplated.

Although most patients with rheumatic MS who are asymptomatic will do well for years without intervention, an elevation of pulmonary artery pressure is an indication that there is progressive elevation of LA pressure affecting the pulmonary circulation. An elevated pulmonary pressure can be assessed by Doppler echocardiography. Although there may be a decrease in pulmonary pressure after relief of the rheumatic MS,³⁸ some patients will have developed intrinsic pulmonary vascular disease, as evidenced by a poorer long-term survival rate in patients who have pulmonary hypertension before intervention.^14,39 An elevated pulmonary arterial resistance before intervention is associated with RV dysfunction and TR after the procedure.^40–42 Thus, PMBC may prevent the adverse consequences of irreversible pulmonary hypertension if it can be performed with a high success rate and low risk in patients who are developing pulmonary hypertension. Correction of the MR before irreversible changes occur can be curative. Thus, in chronic primary MR, MR is the disease.

The new onset of AF may be an indication for proceeding with PMBC in the asymptomatic patient with a favorable valve morphology for several reasons. First, AF may be the equivalent of symptom onset, signifying that rheumatic MS is resulting in progressive LA damage. Second, AF increases the risk of thromboembolic events in patients with rheumatic MS. In addition, a shortened diastolic filling interval with AF and a rapid ventricular response further increase LV pressure. Finally, the presence of AF is associated with worse outcomes in patients with rheumatic MS and with suboptimal results after PMBC.⁴³ In theory, lowering a high LA pressure after PMBC might be beneficial in restoring normal sinus rhythm. Although there is no randomized trial to prove the effectiveness of intervening early, there is a documented improvement in P-wave dispersion after PMBC, which may affect the ability to restore normal sinus rhythm.¹⁵

Some patients have symptoms from rheumatic MS even with a mitral valve area >1.5 cm² and a resting mean transmitral gradient <10 mm Hg. This may be related to the variability and reliability of measuring a mitral valve area by either planimetering a short-axis image of the mitral valve or using a diastolic half-time for indirect calculation of the mitral valve area. There are also patients who have a relatively low gradient at rest who generate a much higher gradient with exercise, with symptoms developing from the higher LA pressure. Thus, in these patients in whom there is a discrepancy between the clinical symptoms and the resting hemodynamics, exercise testing with measurement of the mean transmitral gradient or the direct pulmonary artery wedge, or both, is useful.^44–48 Patients who increase their gradients to >15 mm Hg with exercise have been shown to improve symptomatically after PMBC.¹⁶

Both anatomic valve morphology and the presence of commissural calcification predict successful PMBC. However, in all such series, this predictive ability is not absolute, with 42% of patients with an anatomic valve Wilkins morphology score >8^22,23,31,32 having an optimal outcome (25% increase in mitral valve area to >1.5 cm²) and 38% of patients with commissural calcium having event-free survival at 1.8 years.^18,19,22,23 Accordingly, severely symptomatic patients who are poor surgical candidates may benefit from PMBC even with suboptimal valve anatomy.¹⁷ Patients who refuse surgery may also be offered PMBC after discussion about the potential complications associated with this procedure when it is performed in patients with suboptimal valve anatomy.

Patients with MR and LV dysfunction experience myocardial damage and HF. With onset of LV systolic dysfunction, surgery is usually indicated. However, in those patients in whom surgery (or transcatheter repair) is not performed or will be delayed, medical therapy for systolic dysfunction may be helpful to treat the LV dysfunction alone. Although data specific to patients with MR with LV dysfunction are sparse, treatment of such patients would consist of the standard regimen for HF, including beta-adrenergic blockade, ACE inhibitors or ARBs, and possibly aldosterone antagonists.^1–3 Perhaps the best data exist for the use of beta blockers,¹ which reverse LV dysfunction in experimental MR.² Patients who are receiving beta blockers may have better surgical outcomes and delayed onset of LV dysfunction as compared with those not taking these medications.³ ACE inhibition has not been effective in experimental MR with LV dysfunction. Because aldosterone antagonism is thought to work in part by inhibiting fibrosis, its role in MR, where little fibrosis occurs, is unclear.

Because vasodilator therapy appears to be effective in acute severe symptomatic MR, it seems reasonable to attempt afterload reduction in chronic asymptomatic MR with normal LV function in an effort to forestall the need for surgery. However, the results from the limited number of trials addressing this therapy have been disappointing, demonstrating little or no clinically important benefit.^4–8 Conversely, because vasodilators decrease LV size and mitral closing force, they may increase mitral valve prolapse, worsening rather than decreasing severity of MR.⁶ The foregoing does not apply to patients with concomitant hypertension. Hypertension must be treated because of the well-known morbidity and mortality associated with that condition and because increased LV systolic pressure by itself increases the systolic transmitral gradient and worsens severity of MR.

Primary MR is a mechanical problem of the leaflet coaptation that has only a mechanical solution—that of mitral valve mechanical intervention. The onset of symptoms that results from severe MR worsens prognosis even when LV function appears to be normal,^1,2 and the negative prognosis extends even to mild symptoms.² Thus, the onset of symptoms is an indication for prompt mitral valve surgery.

The goal of therapy in MR is to correct it before the onset of LV systolic dysfunction and its subsequent adverse effect on patient outcomes. The ideal time for mitral valve surgery is when the patient’s LV approaches but has not yet reached the parameters that indicate systolic dysfunction (LVEF ≤60% or LVESD ≥40 mm).^3–7,16 Because symptoms do not always coincide with LV dysfunction, imaging surveillance is used to plan surgery before severe dysfunction has occurred. If moderate LV dysfunction is already present, prognosis is worse after mitral valve operation.^{5–7,9,10,16} Thus, further delay (although symptoms are absent) will lead to greater LV dysfunction and a still worse prognosis. Because the loading conditions in MR allow continued late ejection into a lower-impedance LA, a higher cutoff for “normal” LVEF is used in MR than in other types of heart disease. Although it is clearly inadvisable to allow patients’ LV function to deteriorate beyond the benchmarks of an LVEF ≤60% or LVESD ≥40 mm, some recovery of LV function can still occur even if these thresholds have been crossed.^5,32

Repair success increases with surgical volume and expertise, which is a principle guiding surgical referral.^21,27 However, mitral valve replacement is preferable to a poor repair. The results of a minimally invasive approach may be similar to those of a full median sternotomy if the minimally invasive operation is performed by highly experienced surgeons.^33–38 When leaflet dysfunction is limited so that only annuloplasty and repair of the posterior leaflet are necessary, an operative mortality rate of <1%, long-term survival rate equivalent to that of the age-matched general population, approximately 95% freedom from reoperation, and >80% freedom from recurrent moderate or severe MR at 15 to 20 years after operation are expected.^23,24,39,40

The onset of symptoms, LV dysfunction, or pulmonary hypertension worsens the prognosis for MR. Careful surveillance may result in timing of valve surgery before these negative sequelae occur. However, an attractive alternative strategy for treating severe chronic primary MR is to perform early mitral repair before these triggers are reached. Early mitral repair avoids the need for intensive surveillance and also obviates the possibility that patients might become lost to follow-up or delay seeing their clinician until advanced LV dysfunction has already ensued.^4,13,16,22 For the early mitral repair strategy to be effective, a durable repair must be provided. An unwanted valve replacement and its attendant risks, or a failed repair necessitating reoperation, could be a complication of this approach. Thus, there must be a high degree of certainty that a durable repair can be performed. This certainty comes from the track record of the surgical team in operating on the specific type of lesion under consideration. Thus, asymptomatic patients should be treated in a Comprehensive Valve Center.^{21,24,27–29} In excellent hands, patients with severe MR from flail leaflets who undergo early operation as opposed to watchful waiting have a lower risk of developing HF and lower mortality rates.^4,13,15

MR may lead to progressively more severe MR as the initial level of MR causes LV dilation, which increases stress on the mitral apparatus, causing further damage to the valve apparatus, more severe MR, and further LV dilation—thus initiating a perpetual cycle of ever-increasing LV volumes and MR. Longstanding volume overload leads to irreversible LV dysfunction and a poorer prognosis. Patients with severe MR who develop an LVEF <60% or LVESD ≥40 mm have already developed LV systolic dysfunction.^5,6 One study has suggested that for LV function and size to return to normal after mitral valve repair, the LVEF should be >64% and LVESD <37 mm.¹⁶ Thus, when longitudinal follow-up demonstrates a progressive decrease of LVEF toward 60% or a progressive increase in LVESD approaching 40 mm, it is reasonable to consider intervention.⁴¹ In severe MR, TTE is recommended at 3- to 6-month intervals or more frequently as the ventricle enlarges.

Mitral transcatheter edge-to-edge repair (TEER) with the anterior and posterior leaflets clipped together at ≥1 locations is safe and effective in treating severely symptomatic patients with primary MR who are at high or prohibitive surgical risk.^17,18,42 Studies of TEER with a mitral valve clip have demonstrated improved symptoms and a reduction in MR by 2 to 3 grades, leading to reverse remodeling of the LV. Superior outcomes were shown with surgery versus TEER, and thus it is only the patients who are at high or prohibitive risk for surgery for whom TEER is performed.

Rheumatic mitral valve disease is less suitable for mitral repair compared with complex degenerative disease. Durability of the repair is limited by thickened or calcified leaflets, extensive subvalvular disease with chordal fusion and shortening, and progression of rheumatic disease. Freedom from reoperation at 20 years, even in experienced hands, is in the 50% to 60% range. In a large series from Korea, repair was accomplished in 22% of patients operated on for rheumatic disease.¹⁹ One-third of these patients who underwent repair had significant stenosis or regurgitation at 10 years. Repair of rheumatic mitral valve disease should be limited to patients with less advanced disease in whom a durable repair can be accomplished or to patients in whom a mechanical prosthesis cannot be used because of anticoagulation management concerns.⁴³

Mitral valve repair is the procedure of choice for isolated severe primary MR limited to less than one-half of the posterior leaflet, and mitral valve replacement is inappropriate unless mitral valve repair has been attempted and was unsuccessful.^{11–14,21,22} Surgical repair of primary MR has been remarkably successful. Repair of isolated degenerative mitral disease, when leaflet dysfunction is sufficiently limited that only annuloplasty and repair of the posterior leaflet are necessary, has led to outcomes distinctly superior to those with biological or mechanical mitral valve replacement.^11–14 Repair is associated with an operative mortality rate of <1%, long-term survival rate equivalent to that of age-matched general population, approximately 95% freedom from reoperation, and >80% freedom from recurrent moderate or severe (≥3) MR at 15 to 20 years after surgery.^15,39 As much as one-half of the posterior leaflet may be excised, plicated, or resuspended. Posterior leaflet repair has become sufficiently standardized in this situation so that repair, rather than mitral valve replacement, is the standard of care. Execution of this procedure with a success rate ≥95% should be the expectation of every cardiac surgeon who performs mitral valve procedures.

TTE is essential in patients with MR to identify patients with primary MR and those with secondary forms of MR. In general, in patients with LV systolic dysfunction and symptoms of HF, the presence of chronic secondary MR of any severity is associated with a worse prognosis than that seen in the absence of MR. Most patients with secondary MR have global LV dysfunction, but in some patients, a limited but strategically placed wall motion abnormality may also cause chronic secondary MR. An initial TTE helps establish the cause of chronic secondary MR and also serves as a baseline for future comparisons. In patients with secondary MR, severe MR is defined as an ERO ≥40 mm², but outcome studies have shown poor prognosis in those with moderate MR (ERO ≥20 mm²).^1,2

Prognosis is poor for both ischemic and nonischemic MR, but ischemic MR lends itself to the possibility of revascularization and potential improvement in LV function if CAD has led to large areas of hibernating viable myocardium. Long-term results of the STICH (Surgical Treatment for Ischemic Heart Failure) trial demonstrated an improved 10-year survival rate in patients with ischemic cardiomyopathy and LVEF <35% who underwent CABG plus GDMT as compared with those randomized to GDMT alone. CT angiography is usually adequate to rule out significant CAD and thus rule out ischemic MR. If CAD is detected and noninvasive testing demonstrates areas of viability, coronary arteriography is pursued to better define the anatomy for potential revascularization.^14,15 Although the presence of myocardial viability did not determine the effect of revascularization on survival in the STICH trial, there is a subset of patients with viable myocardium in whom the ischemic MR will respond to revascularization.^16–18

Clinical trials have identified anatomic considerations, detectable by TEE, that can identify patients with secondary MR who have a valve morphology amenable to TEER. In the COAPT (Cardiovascular Outcomes Assessment of the MitraClip Percutaneous Therapy for Heart Failure Patients with Functional Mitral Regurgitation) trial of patients with secondary MR and HF, exclusion criteria included vertical coaptation length <2 mm in valves with leaflet tethering, evidence of calcification in the grasping area of the A2 or P2 scallops, presence of a significant cleft of A2 or P2 scallops, and lack of both primary and secondary chordal support. These are similar criteria to the earlier EVEREST trial. TEE is standard preprocedural imaging to determine suitability for TEER.^3–8

During mitral TEER, TEE assists in guiding positioning of the clip(s), assessing success of the procedure, determining whether more than a single clip is necessary to reduce MR, and assuring that the clip(s) has not created MS.^4,7,9–13

Chronic secondary MR usually develops as a result of LV systolic dysfunction. Thus, standard GDMT for HF forms the mainstay of therapy. Diuretics, beta blockers, ACE inhibitors or ARBs, and aldosterone antagonists help improve symptoms and/or prolong life in patients with HF in general and probably do so even when HF is complicated by chronic secondary MR. GDMT can reduce LV volumes (reverse remodeling) in many patients, which reduces severity of secondary MR.^1–11

Secondary MR is often responsive to GDMT (including coronary revascularization or cardiac resynchronization therapy in appropriate patients). Optimization of GDMT should be under the supervision of a cardiologist expert in the treatment of patients with HF to achieve optimal results and to determine with the MDT when symptoms are truly refractory to GDMT before decisions are made for surgical or transcatheter treatment.^9,12

The COAPT trial of transcatheter treatment of secondary MR demonstrated improvement in survival, hospitalization, symptoms, and quality of life in patients with persistent symptoms despite optimization of GDMT who were randomized to TEER, as compared with those randomized to continued GDMT. In contrast, MITRA-FR (Multicentre Randomized Study of Percutaneous Mitral Valve Repair MitraClip Device in Patients With Severe Secondary Mitral Regurgitation) enrolled patients with greater degrees of LV enlargement and less severe MR (mean ERO area 0.31 cm² versus 0.41 cm²) and reported no benefit of TEER in reducing the composite endpoint of death or hospitalization as compared with medical therapy. In addition, the inclusion criterion in MITRA-FR of an LVESD up to 70 mm represents extreme dilation; in contrast, in the COAPT trial, the mean LVESD was smaller (52±9 mm), and even the LVEDD rarely exceeded 70 mm (mean 62±7 mm). Thus, the enrollment criteria in COAPT trial (LVEF between 20% and 50%, LVESD ≤70 mm, pulmonary artery systolic pressure ≤70 mm Hg, and persistent symptoms [NYHA class II, III, or IV] while on optimal GDMT) are the current standard selection criteria for TEER for secondary MR. Observational studies have suggested that a greater reduction in MR severity with TEER is associated with greater LV and LA reverse remodeling.^1–8,48,49 The exact anatomy and mechanism of MR also needs to be taken into consideration when determining candidacy for transcatheter repair.

There is no proof that surgical correction of chronic secondary MR is effective in prolonging life, but observational studies and a substudy of the randomized STICH trial suggest that it is wise to address the mitral valve during CABG for severe CAD when secondary MR is severe. Although it may be hoped that the revascularization will recruit hibernating myocardium and reduce chronic secondary MR, this has not been demonstrated, and failing to correct chronic severe secondary MR may leave the patient with severe residual MR. The risks and benefits of additional surgical interventions should be weighed in patients with LV systolic dysfunction.^9–13 For patients with secondary MR undergoing operation for other valve disease, see Section 10.2 (Timing of Intervention for Mixed Valve Disease).

MR may develop in patients with preserved LV systolic function who have progressive LA dilation, leading to enlargement of the mitral annulus and malcoaptation of the leaflets.^51,52 This may arise in conditions such as HF with preserved LVEF, restrictive cardiomyopathy, and nonobstructive hypertrophic cardiomyopathy. These patients often have associated AF, which may contribute to the progression of LA and annular dilation, thus increasing the severity of MR,^18,53 and successful ablation of AF may reduce or eliminate MR.⁵³ Isolated annular dilation accounts for <20% of patients referred for surgery of severe MR in the STS database, but it is also the etiology with the highest mitral valve repair rates (85%).^16,17 The limited data addressing mitral valve repair in patients with annular dilation related to AF indicate low operative risk.^18–20

There is limited evidence that mitral valve surgery improves survival in symptomatic patients with secondary MR. In addition, surgery may improve symptoms and quality of life in these patients whose symptoms persist despite GDMT. Small RCTs demonstrate that mitral valve surgery reduces chamber size and improves peak oxygen consumption in chronic severe secondary MR. Ischemic or dilated cardiomyopathy presents different challenges for mitral repair. Regurgitation is caused by annular dilation, as well as by apical and lateral displacement of the papillary muscles. New techniques have facilitated mitral repair in this situation, but durability of the repair is dependent primarily on regression or progression of ventricular dilation. If the heart continues to dilate, long-term durability of the repair is moot; the survival of the patient is limited.^9,12,21–43

5. In an RCT of mitral valve repair versus mitral valve replacement in patients with severe ischemic MR, there was no difference between repair and mitral valve replacement in survival rate or LV remodeling at 2 years. However, the rate of recurrence of moderate or severe MR over 2 years was higher in the repair group than in the replacement group, leading to a higher incidence of HF and repeat hospitalization. The lack of apparent benefit of valve repair over valve replacement in secondary MR versus primary MR, with less durable repairs in secondary MR, highlights that primary and secondary MR are 2 different diseases.^{9,12,21–32,44–47}

TTE can distinguish primary TR (abnormal valve leaflets) from secondary TR (normal valve leaflets), define any associated left-sided valvular or myocardial disease, and provide an estimate of pulmonary artery systolic pressure.^11–15 Characterization of the severity of TR relies on an integrative assessment of multiple parameters, as recommended by the American Society of Echocardiography and European Association of Echocardiography,^1,2 but many limitations remain. In patients with TR undergoing left-sided valve surgery, an annular diastolic diameter >40 mm (or >21 mm/m²) indicates an increased risk of persistent or progressive TR after isolated mitral valve surgery.¹³ Pulmonary artery systolic pressure is estimated from maximal TR velocity. Assessment of RV systolic function is challenged by geometric and image acquisition constraints, as well as by variability in RV loading condition.^16,17 Normal RV systolic function is defined by several parameters, including tricuspid annular plane systolic excursion >16 mm, tricuspid valve systolic annular velocity >10.0 cm/s, and RV end-systolic area <20.0 cm² or fractional area change >35%. Other imaging modalities, such as 3D TEE, magnetic resonance imaging, and CT scan, may provide more accurate information on the status of the RV.

When physical examination and TTE data on estimated pulmonary artery systolic pressure are either discordant or inadequate, invasive measurement of pulmonary artery pressures and pulmonary vascular resistance can be helpful to guide clinical decision-making in individual patients.^3–5 A weak TR signal or the presence of severe TR may result in underestimation of pulmonary systolic pressure; direct invasive measurement can resolve this uncertainty. Data from invasive measurement are essential for patients in whom the cause of pulmonary hypertension is uncertain or when assessment of pulmonary vascular reactivity after vasodilator challenge is needed. Direct measurements of right atrial pressure may also be useful for clinical decision-making. Right ventriculography may further aid in the evaluation of the severity of TR and the status of the RV. Thermodilution cardiac output measurements may be inaccurate with severe TR, and thus a Fick cardiac output should be measured to apply to the calculation of pulmonary resistance.

Patients with severe TR usually present with signs or symptoms of right-sided HF, including peripheral edema and ascites. Low-salt diet and support stockings may be helpful. Diuretics can be used to decrease volume overload in these patients. Loop diuretics are typically provided and may relieve systemic congestion, but their use can be limited by worsening low-flow syndrome. Aldosterone antagonists may be of additive benefit, especially in the setting of hepatic congestion, which may promote secondary hyperaldosteronism.

Medical therapies for management of severe TR (Stages C and D) are limited. Attention should be focused on the underlying etiologies in patients with secondary TR. Reduction of pulmonary artery pressures and pulmonary vascular resistance with specific pulmonary vasomodulators may be helpful to reduce RV afterload and secondary TR in selected patients with pulmonary hypertension.^1,2 GDMT is effective for secondary TR attributable to HF with reduced LVEF. Restoration of normal sinus rhythm may be effective for secondary TR attributable to annular dilation associated with AF.^3,4

Severe TR of either a primary or secondary etiology may not improve predictably after treatment of the left-sided valve lesion and reduction of RV afterload; as such, severe TR should be addressed as part of the index procedure.^1,2,28–31 Reoperation for severe, isolated TR after left-sided valve surgery is associated with a perioperative mortality rate of 10% to 25%.^1,29 Tricuspid valve repair does not add appreciably to the risks of surgery.^1,2,28–31 There has been a significant increase in the number of tricuspid valve repairs performed for this indication over the past decade. Tricuspid valve repair is preferable to replacement, but replacement may be necessary if there is marked dilation of the annulus or intrinsic disease of the tricuspid leaflets.^28,31 Observational data have shown a lower operative risk with tricuspid valve repair than with replacement, but this may be related to patient selection, given that the latter would be inserted in patients with a severely dilated annulus and abnormal leaflets to prevent recurrent or residual regurgitation. The risks and benefits of tricuspid valve operation should be carefully considered in the presence of severe RV systolic dysfunction or irreversible pulmonary hypertension because of the possibility of RV failure after operation.

Left uncorrected at the time of left-sided valve surgery, mild or moderate degrees of secondary TR may progress over time in approximately 25% of patients and result in reduced long-term functional outcome and survival.³² Risk factors for persistence or progression of TR include tricuspid annulus dilation (>40 mm diameter or 21 mm/m² diameter indexed to body surface area on preoperative TTE measured at end diastole; >70 mm diameter on direct intraoperative measurement of the intercomissural distance), degree of RV dysfunction or remodeling, leaflet tethering height, pulmonary artery hypertension, AF, and intra-annular RV pacemaker or implantable cardioverter-defibrillator leads.^{3–10,33–36} Several observational studies and one prospective RCT have demonstrated the benefit of tricuspid repair at the time of mitral valve surgery for progressive TR (Stage B) with tricuspid annulus dilation on echocardiographic and functional parameters, although data on outcomes such as survival and major adverse events are lacking.^{3–10,33–35} Because the severity of TR may be dynamic, dependent on the preload and pulmonary pressures, a past history of signs or symptoms of right-sided HF indicates the propensity to develop more severe TR and should be considered an indication for concomitant tricuspid valve repair.

In patients with symptomatic severe primary TR, reduction or elimination of the regurgitant volume load by tricuspid valve surgery can alleviate systemic venous and hepatic congestion and decrease reliance on diuretics.^11,12,20 Patients with severe congestive hepatopathy may also benefit from surgery to prevent irreversible cirrhosis of the liver. Quality and duration of long-term survival are related to residual RV function. In patients with severe symptomatic primary TR from either device leads or endomyocardial biopsy, TR develops rapidly, and surgery can be done before the onset of RV dysfunction.^11,37 Correction of symptomatic severe primary TR (Stage D) in patients without left-sided valve disease would preferentially be performed before the onset of significant RV dysfunction or end-organ damage. Randomized studies of early intervention are lacking, and the benefit might be limited by the risk of intervention, suboptimal reduction in TR severity, or suboptimal durability of currently available approaches to tricuspid valve repair and replacement.

There is now recognition that TR can develop in association with AF and annular dilation (a form of secondary TR).^23–25 Notably, AF-related TR appears to represent a fundamentally different pathophysiology from other forms of secondary TR, with greater basal dilation and annular enlargement, as compared with the RV elongation with leaflet tethering seen in patients who have secondary TR caused by pulmonary hypertension or myocardial disease.²⁴ These patients with AF-related TR have rapid progression of TR severity and right-sided chamber dilation. In appropriately selected symptomatic patients with AF-related severe TR, quality of life and symptoms can be improved by surgical intervention for TR. In patients undergoing intervention, overall outcomes are better in those without severe RV dysfunction or end-organ damage. Newer surgical techniques and a better selection process resulted in an acceptable operative mortality rate (<4% to 5%) for isolated TR in selected patients.^{2,11,12,15–19,22,38}

The optimal timing of tricuspid valve surgery for asymptomatic or minimally symptomatic patients with severe primary TR has not been established. Extrapolation from limited experiences reported for patients with stable carcinoid heart disease and patients with a flail tricuspid leaflet, as well as application of the management principles adopted for patients with severe MR, suggest that serial assessments of RV size and function might trigger consideration of corrective surgery in selected patients with severe primary TR when a pattern of continued deterioration can be established and the surgical risk is considered acceptable.^13,14 In otherwise healthy patients without other comorbidities, such as patients with severe TR attributable to trauma, the surgical risk associated with tricuspid valve operation is low (<1% to 2% operative mortality rate) in the absence of RV dysfunction or pulmonary hypertension.

Isolated tricuspid valve surgery for severe TR historically has been performed relatively late in the natural history of the disease, when patients have become symptomatic with signs of right-sided HF. Unadjusted mortality rates for isolated tricuspid valve surgery have therefore exceeded those reported for isolated aortic or mitral valve surgery, and this trend has been even more pronounced for reoperative tricuspid surgery late after left-sided valve surgery.^1,2,39 This high reoperative mortality rate is likely related to the advanced nature of RV failure encountered at the time of the second procedure, residual pulmonary hypertension, LV dysfunction, and other valve abnormalities. The hazards imposed by reoperation have influenced decision-making for initial repair of functional TR at the time of left-sided valve surgery in an attempt to prevent the development of severe TR later after the left-sided valve surgery. However, if there is no significant pulmonary hypertension or severe RV systolic dysfunction, operation for severe symptomatic isolated TR years after surgery for left-sided disease may improve symptoms of right-sided HF, if done before the onset of severe RV dysfunction or end-organ damage with either hepatic or renal dysfunction.^11,18

Currently, isolated moderate AS or moderate AR is placed in Stage B, progressive disease for which no therapeutic action is indicated. However, some patients with moderate mixed disease develop symptoms that stem from their valve disease. Formerly, the argument was raised that if there were no AR, the aortic jet velocity and gradient would be correspondingly lower and would not meet the definitions for severe AS, and obviously there would be only moderate AR by definition. Therefore, no action was recommended. However, recent data suggest that the natural history of moderate mixed disease behaves similarly to that of pure severe AS^1,2 and that moderate mixed disease has a mortality risk similar to that of pure severe AS. Thus, valve replacement is warranted for the symptomatic patient if the patient’s data fulfill any of the criteria for severe AS. The decision about whether to proceed with TAVI versus SAVR is discussed in Section 3.2.4.2.

For patients with mixed moderate AS/AR who have developed LV dysfunction, as evidenced by an LVEF of <50%, and who have no other reason for LV dysfunction, valve disease is presumed to be the cause. In such patients, SAVR is indicated.^1,2

TTE after valve implantation or repair provides an assessment of the procedural results and serves as a baseline against which comparison can be made for any change. TTE provides accurate measurements of transvalvular velocities and pressure gradients, as well as detection and quantitation of transvalvular and paravalvular leak.^1–4 Normal transvalvular velocities and gradients vary across different types and sizes of prosthetic valves but are also affected by patient-specific factors, including body size and cardiac output. The postoperative study, recorded when the patient is asymptomatic and in a stable hemodynamic state, provides Doppler flow data for a specific valve in an individual patient. In addition, TTE provides assessment of other valve disease(s), pulmonary artery pressure, atrial size, LV and RV size and function, and pericardial disease.

Bioprosthetic or repaired native valve dysfunction typically presents with the insidious onset of HF symptoms or a change in the auscultatory findings. More abrupt and severe symptoms may occur with infective endocarditis or rupture of a valve cusp. Patients with mechanical valve dysfunction may present with HF, shock, thromboembolic events, hemolysis, or a change in auscultatory findings. Presentation may often be acute or subacute because of thrombus formation and more abrupt impairment of leaflet opening or closure. Attention should be directed to the trend in recent INR determinations. Prosthesis–patient mismatch and functional stenosis of a repaired native valve are also to be considered in the evaluation of patients with HF symptoms. Repeat noninvasive assessment begins with transthoracic echocardiography, comparison with the index postoperative study when available, and the use of other modalities as dictated by the clinical context and preliminary findings.

TTE is the preferred approach for initial assessment of suspected prosthetic valve dysfunction because it allows for measurement of transvalvular velocity, gradient, and valve area. TTE also allows quantitation of LV volumes and LVEF, an estimate of pulmonary artery systolic pressure, and evaluation of right heart function. However, the LA side of a prosthetic mitral valve is obscured by acoustic shadowing from the TTE approach, resulting in reduced sensitivity for detection of prosthetic MR and prosthetic mitral valve thrombus, pannus, or vegetation. TEE provides superior imaging of the LA side of the mitral prosthesis and is accurate for diagnosis of prosthetic mitral valve dysfunction.^5,6 Both TTE and TEE are also needed for patients with prosthetic aortic valves in whom the posterior aspect of the valve is shadowed on the TTE approach and the anterior aspect of the valve is shadowed on the TEE approach.^7,8 TEE has superior sensitivity for the detection of vegetations and abscess formation in patients with suspected prosthetic valve (or annuloplasty ring) endocarditis. With mechanical valve obstruction, fluoroscopy or CT imaging can also be helpful for detection of reduced motion caused by pannus ingrowth or thrombus.

Studies based on TTE follow-up estimate that approximately 30% of patients with a surgical aortic valve bioprosthesis develop evidence of valve dysfunction over the 10 years after implantation (defined as an increase in mean gradient of ≥10 mm Hg or a worsening of transprosthetic regurgitation from mild to moderate or from moderate to severe).⁹ The incidence of clinically important structural valve deterioration increases markedly more than 10 years after surgery, such that routine annual TTE studies thereafter are reasonable.^10,11 Risk factors associated with accelerated (<5 years) valve deterioration include young age (<60 years) at implantation, smoking, diabetes mellitus, chronic kidney disease, initial mean gradient ≥15 mm Hg, and valve type.^9,12 The selective adoption of an earlier, annual TTE screening program may be considered for at-risk patients on an individual basis, as up to 13% of patients with a surgical aortic valve develop hemodynamic valve dysfunction at a median of 6.7 to 9.9 years after implantation.¹² Patients typically remain asymptomatic until valve dysfunction is severe enough to result in adverse hemodynamic consequences or AF. Depending on the valve type and mechanism of regurgitation, some patients with asymptomatic, significant prosthetic valve regurgitation may require reintervention. For example, if prosthetic regurgitation is attributable to a bioprosthetic leaflet tear, more severe acute regurgitation may occur suddenly and cause clinical decompensation. With prosthetic valve stenosis, TTE diagnosis while the patient is asymptomatic alerts the clinician to the need for more frequent follow-up. A standardized definition and grading system for structural valve deterioration for surgical and transcatheter aortic valves have been proposed.¹³ In patients with mechanical valve prostheses, routine annual TTE evaluation is not needed if the postoperative baseline study is normal and no clinical change is apparent. Many of these patients require TTE studies for other indications, however, such as for the assessment of LV function, pulmonary artery pressure, or other cardiac or valve disease.

Durability data for bioprosthetic TAVI valves are less robust than the data for surgically implanted bioprosthetic valves. To date, the intermediate-term durability of TAVI valves has compared favorably with that of SAVR valves, as reported in randomized trials and registries.^14–21 For the most part, these data reflect observations made in older patients and may not be applicable to younger populations (eg, <70 years). TAVI-based protocols typically include routine TTE before discharge and at 30 days and 1 year, in part because of reporting requirements. In the absence of clinical change, routine annual TTE studies are reasonable as experience continues to accumulate.

The choice of valve prosthesis in each patient is based on consideration of several factors, including valve durability, expected hemodynamics for valve type and size, surgical or interventional risk, the potential need for long-term anticoagulation, and patient values and preferences. The trade-off between the risk of reintervention for bioprosthetic valve deterioration and the risk of long-term anticoagulation should be discussed. Some patients prefer to avoid repeat surgery and are willing to accept the risks and inconvenience of lifelong anticoagulant therapy. Other patients are unwilling to consider long-term anticoagulation because of the inconvenience of monitoring, the attendant dietary and medication interactions, and the need to restrict participation in some types of physical activity. The incidence of structural deterioration of a bioprosthesis is greater in younger patients, but the risk of bleeding from anticoagulation is higher in older patients. In patients with shortened longevity or multiple comorbidities, a bioprosthesis might be more appropriate. In women who desire subsequent pregnancy, the issue of anticoagulation during pregnancy is an additional consideration (see pregnancy-related issues in Section 13.5).^20,21

Anticoagulant therapy with VKA is necessary in all patients with a mechanical valve to prevent valve thrombosis and thromboembolic events. If anticoagulation is contraindicated or if the patient refuses VKA therapy, an alternative valve choice is appropriate. Newer anticoagulant agents have not been shown to be safe or effective in patients with mechanical heart valves.

Patients <50 years of age at the time of AVR incur a higher and earlier risk of bioprosthetic valve deterioration.^{4,10,11,22–24} Overall, the predicted 15-year risk of needing reoperation because of structural deterioration is 22% for patients 50 years of age, 30% for patients 40 years of age, and 50% for patients 20 years of age, although it is recognized that all bioprostheses are not alike in terms of durability.¹¹ Anticoagulation with a VKA can be accomplished with acceptable risk in most patients <50 years of age, particularly in compliant patients with appropriate monitoring of INR levels. Thus, the balance between valve durability and risk of bleeding and thromboembolic events favors the choice of a mechanical valve in patients <50 years of age, unless anticoagulation is not desired, cannot be monitored, or is contraindicated.

Uncertainty and debate continue about which type of AVR is appropriate for patients 50 to 65 years of age. Newer surgical bioprosthetic valves may show greater freedom from structural deterioration, specifically in the older individual, although a high late mortality rate in these studies may preclude recognition of valve dysfunction.^11,15–19 The risks of bleeding and thromboembolism with mechanical prostheses are low, especially in compliant patients with appropriate INR monitoring. Several studies have shown a survival advantage with a mechanical prosthesis in this age group. Alternatively, large retrospective observational studies have shown similar long-term survival rates in patients 50 to 69 years of age undergoing mechanical versus bioprosthetic valve replacement.^22–24 In general, patients with mechanical valves experience a higher risk of bleeding because of anticoagulation, whereas individuals who receive bioprosthetic valves experience a higher rate of reoperation attributable to structural deterioration of the prosthesis, as well as perhaps a decrease in survival rate.^6,25–27 Several other factors should be considered in the choice of type of valve prosthesis (see Section 11.1). Ultimately, the choice of mechanical versus bioprosthetic valve replacement for all patients, but especially for those between 50 and 65 years of age, should be made in a shared decision-making process that must account for the trade-offs between durability (and the need for reintervention), bleeding, and thromboembolism.¹

In patients >65 years of age at the time of bioprosthetic AVR, the likelihood of primary structural deterioration at 15 to 20 years is only about 10%.^28–31 In addition, older patients are at higher risk of bleeding complications related to VKA therapy and more often require interruption of VKA therapy for noncardiac surgical and interventional procedures. It is reasonable to use a bioprosthetic valve in patients >65 years of age to avoid the risks of anticoagulation because the durability of the valve exceeds the expected years of life.

In general, patients with mechanical valve replacement experience a higher risk of bleeding because of anticoagulation, whereas individuals who receive a bioprosthetic valve replacement incur a higher risk of repeat intervention attributable to structural valve deterioration. In patients <65 years of age, observational data suggest better long-term outcomes with a mechanical mitral valve replacement, even when the risks and inconvenience of long-term VKA anticoagulation are considered. In a propensity-matched analysis from New York’s Statewide Planning and Research Cooperative System (SPARCS), although there was no survival difference for patients 50 to 69 years of age undergoing mechanical versus bioprosthetic mitral valve replacement,⁷ the rates of reoperation were lower (HR: 0.59) with a mechanical valve, though stroke risk (HR: 1.62) was higher. In the 2017 report from the California Office of Statewide Health Planning and Development,¹ for patients who underwent mitral-valve replacement and were 40 to 69 years of age, receipt of a biological prosthesis was associated with a mortality rate significantly higher than that seen with receipt of a mechanical prosthesis.¹ The choice of a mechanical mitral valve in patients <65 years of age who are good candidates for anticoagulation should account for these observational, nonrandomized data and abide by the principles of shared decision-making.^1,7,10

Hazards associated with anticoagulation increase with age, and rates of structural valve deterioration decline significantly. In patients >65 years of age, the ratio of valve durability to life expectancy supports the use of a bioprosthetic mitral valve replacement, which allows avoidance of the risks of long-term VKA anticoagulation in these older patients. In 1 observational study, the expected durability of a bioprosthetic mitral valve replacement was 11.4 years in patients <60 years of age, 16.6 years in those 60 to 70 years of age, and 19.4 years in those >70 years of age.¹¹ In the 2017 report from the California Office of Statewide Health Planning and Development,¹ overall survival rates were similar for patients 70 to 79 years of age who underwent mechanical versus bioprosthetic mitral valve replacement, and bleeding risk was lower with a bioprosthetic valve.¹

Replacement of the aortic valve with a pulmonary autograft (the Ross procedure) is a complex operation involving replacement of the aortic valve by the patient’s own pulmonic valve, along with placement of a pulmonic valve homograft. The Ross procedure allows the patient to avoid a prosthetic heart valve and the risks of anticoagulation, and it provides excellent valve hemodynamics. However, both the pulmonic homograft in the pulmonic position and the pulmonary autograft (the neoaortic valve) are at risk of valve degeneration. The failure of the Ross procedure is most often attributable to regurgitation of the neoaortic valve in the second decade after the operation. In addition, at least half of pulmonic homograft valves require reintervention within 10 to 20 years. Calcification of the homograft and adhesions between the homograft and neoaorta may increase the difficulty of reoperation. The Ross procedure typically is reserved for younger patients with appropriate anatomy and tissue characteristics in whom anticoagulation is either contraindicated or undesirable, and it is performed only at Comprehensive Valve Centers by surgeons experienced in this procedure.^12–14,32

Mechanical prosthetic valve thrombosis is diagnosed by an abnormally elevated velocity or gradient across the prosthesis, with either limited leaflet motion or attached mobile densities consistent with thrombus, or both. Prosthetic valve obstruction is usually defined as a mean transvalvular gradient increase >50% (or an increase >10 mm Hg across an aortic prosthesis) compared with baseline, after exclusion of other causes, such as a high-output state. When mechanical valve thrombosis is suspected, imaging and Doppler data from TTE provide information on valve function, estimated pulmonary pressures, and LV size and systolic function.⁹ Leaflet motion should be visualized with CT or TEE (particularly for a mitral prosthesis) or fluoroscopy (for an aortic prosthesis).^6,7,10–12 Prolonged periods of observation under fluoroscopy or TEE may be required to diagnose intermittent obstruction. The presence and quantification of thrombus and pannus should be evaluated by either TEE or CT.^6,7,10–12

The decision between surgery and systemic fibrinolysis for symptomatic left-sided mechanical valve thrombosis should be individualized (Table 23) after review by the heart valve team, while engaging the patient in a process of shared decision-making and accounting for local experience and expertise. The overall 30-day mortality rate with surgery is 10% to 15%, with a lower mortality rate of <5% in patients with NYHA class I or II symptoms.^2,3,7 Recent studies using an echocardiogram-guided, slow-infusion, low-dose fibrinolytic protocol have shown hemodynamic success rates >90%, with embolic event rates <2% and major bleeding rates <2%.¹³ Systemic fibrinolysis is therefore an acceptable alternative to reoperation in patients at high or prohibitive surgical risk and in patients who have a small thrombus burden, mild HF symptoms (NYHA class I or II), and low bleeding risk. Absence of surgical expertise should be considered in the clinical decision-making process as a factor that favors thrombolytics, whereas recurrent valve thrombosis favors a surgical approach (Table 23).

Bioprosthetic valve thrombosis appears to be more common with transcatheter than with surgical bioprosthetic valves. Leaflet thrombosis often is suspected on the basis of an increased transvalvular velocity on routine echocardiographic monitoring and can be confirmed by the finding of hypoattenuation of the valve leaflets on CMR imaging. When there is clinical evidence of stenosis, 3D TEE or 4D CT imaging may be useful to detect a layer of valve thrombus, which may respond to treatment with oral anticoagulation.

Multiple small nonrandomized studies support the use of VKAs to treat patients with clinical and subclinical bioprosthetic valve thrombosis after both SAVR and TAVI.^1,3,4,6,7 VKA anticoagulation can result in a significant reduction of transvalvular gradient, improved leaflet motion, and clinical improvement.^4,6,8

TTE and TEE assessment can appropriately detect and quantify prosthetic valve stenosis.¹ TTE within 3 months after valve implantation is useful to provide baseline data on valve hemodynamics and ventricular function. In some patients, the orifice area of the implanted prosthesis may be inadequate to meet the cardiac output demands of the patient, even when the prosthetic valve itself is functioning normally. This circumstance, termed patient–prosthesis mismatch, is associated with a high transvalvular gradient, persistent LV hypertrophy, and an increased rate of cardiac events after valve replacement.^10,11 Diagnosis in the setting of bileaflet mechanical valves is complicated by nonlaminar patterns of blood flow, for which significant pressure recovery may be present; thus, a high velocity in the central narrow slit-like orifice may not correlate with prosthetic valve stenosis or patient–prosthesis mismatch. Prosthetic valve stenosis is distinguished from patient–prosthesis mismatch by comparison with the early postoperative baseline study and by visualization of the appearance and motion of the valve leaflets. Prosthetic valve stenosis is characterized by a clinical course of progressive increase in transvalvular velocity and pressure gradient in conjunction with abnormal thickened/calcified leaflets (for bioprosthetic valves) or evidence of pannus formation (with mechanical valves).

The motion of the leaflets of a mechanical valve is best evaluated radiographically with fluoroscopy or cine-CT imaging because strong reflections from the mechanical valve obscure motion on echocardiographic imaging in most patients, particularly as assessed on images from the transthoracic approach. With fluoroscopic imaging, the angle of imaging must be adjusted to demonstrate leaflet motion from a side view, permitting measurement of the angles of opening and closure that can be compared with expected values for that valve type. Cine-CT images are obtained at a high frame rate focused on the prosthetic valve in a 3D acquisition. Compared with fluoroscopy, cine-CT 3D images are less operator dependent for measuring opening and closing angles. In addition, cine-CT allows detection of pannus or thrombus on or adjacent to the valve, which is not possible with fluoroscopy. When excessive gradients are present with normal leaflet motion and no thrombus, either patient–prosthesis mismatch or pannus formation is present (or both).

Stenosis of a bioprosthesis may occur because of progressive structural valve degeneration or pannus formation. However, stenosis can also occur because of a thin layer of thrombus on the valve cusps, which is reversible with oral anticoagulation therapy. Bioprosthetic valves are less thrombogenic than their mechanical counterparts. However, the diagnosis of subclinical bioprosthetic valve thrombosis has increased with the use of CT imaging and as a function of the increased numbers of implanted bioprosthetic valves, including TAVI.^3–7 When there is clinical evidence of bioprosthetic valve stenosis, 3D TEE or 4D CT imaging may be useful to detect a layer of valve thrombus as the cause of the obstruction.

Reoperative surgery for prosthetic valve stenosis is associated with acceptable mortality and morbidity rates in the current era, but the risks are typically higher than those estimated at the time of initial surgery because of older patient age, clinical status at the time of intervention, and reoperative status.^2,16 The decision to proceed with surgical versus transcatheter intervention is based on available expertise, individual patient and valve characteristics, and shared decision-making.

Catheter-based therapy with transcatheter ViV has emerged as an acceptable alternative to reoperative surgery for the treatment of high– and prohibitive–surgical risk patients with bioprosthetic AS.^4,5 Although coronary artery obstruction is more common with aortic ViV procedures than with TAVI for native AS, rates of paravalvular leak and permanent pacemaker implantation are lower with aortic ViV procedures than with TAVI for native AS. Annulus rupture has not been reported. Transcatheter ViV also has been successfully performed for failed surgical bioprostheses in the mitral, pulmonic, and tricuspid positions, although LV outflow obstruction may occur after mitral ViV implantation.

A subset of patients presents with stenosis of a bioprosthetic valve attributable to leaflet thrombosis that results in decreased mobility of the leaflets. Leaflet thrombosis can occur from 1 month to years after implantation. If the patient is stable and has no contraindication to long-term anticoagulation, a trial of oral anticoagulation with VKA may result in resolution of the thrombus and improvement in hemodynamics.^6–13 However, these patients are at increased risk of recurrent thrombosis (if the anticoagulation is stopped) and early structural deterioration, and thus they require close follow-up.¹⁷

Echocardiography is the primary imaging modality to assess the location and quantify the severity of prosthetic valve regurgitation, often requiring both TTE and TEE approaches.^1,3 Although TTE provides superior assessment of transvalvular gradients, chamber sizes, and function, TEE is better suited to identify the cause and location of regurgitation and is essential for prosthetic mitral valve because of acoustic shadowing on TTE. Even with prosthetic aortic valves, acoustic shadowing may affect detection of a paravalvular leak by either TTE or TEE, with TTE being suboptimal to assess posterior paravalvular leak and TEE suboptimal to assess anterior defects.³ The Valve Academic Research Consortium (VARC) has suggested an approach for assessment of paravalvular leak severity and constructed a 5-class grading scheme.⁷

3D echocardiography plays a significant role in determining the precise location and size of the paravalvular leak in patients undergoing intervention. For a successful transcatheter paravalvular leak closure, adequate paravalvular leak assessment includes 1) precise location of the defect(s), 2) precise dimensions, 3) orientation of the defect in relation to the sewing ring and prosthetic valve occluders or leaflets, and 4) location and orientation of the subvalvular structures. Real-time 3D TEE allows optimal visualization of the defects and direct guidance for catheter movement and positioning of the implanted device(s) during the transcatheter closure procedure.^4,5 3D TEE also allows assessment of residual regurgitation after device placement. Limitations of 3D TEE include artifacts of ultrasound imaging (eg, dropout, acoustic shadowing, reverberation artifacts) and reduced temporal and spatial resolution.⁸ Transcatheter closure using intracardiac echocardiography guidance is possible and alleviates the need for conscious sedation or anesthesia but allows only 2D and color Doppler imaging.⁹

Surgery is a viable therapeutic option in many patients with symptomatic paravalvular leak and is associated with reasonable outcomes.¹ The risks associated with surgical intervention depend on the procedure required, be it suture repair or repeat AVR. Although surgical reoperation is associated with acceptable mortality and morbidity rates in the current era, it still carries a higher risk than the initial surgery. Kaneko and colleagues examined a cohort of 3380 patients from the STS database (2011–2013) who underwent elective isolated reoperative AVR, and they demonstrated a higher (but acceptable) operative mortality rate than that seen with initial AVR (4.6% versus 2.2%; P<0.0001) and relatively low complication rates.¹³ This was true even among octogenarians who underwent reoperative AVR.³ In a cohort of 136 consecutive patients who underwent surgical correction for a non–endocarditis-related aortic or mitral paravalvular leak (1986–2001), surgical correction of the paravalvular leak was associated with acceptable operative mortality (6.6%) and morbidity rates.¹ More recently, Shah and colleagues reported an operative mortality rate of 3% among 495 patients undergoing surgery for paravalvular leak,¹⁴ with higher risk associated with mitral than with aortic valve procedures (odds ratio: 1.66; 95% CI: 1.25–2.20). These findings are consistent with the findings of Bouhout and colleagues,¹⁵ who reported operative mortality rates of 8% among mitral valve, 3% among aortic valve, and 14% among double-valve patients in a total cohort of 190 patients undergoing surgery indicated for paravalvular leak. Estimates of operative risk for individual patients can be calculated by using the STS risk calculator (http://riskcalc.sts.org/stswebriskcalc/).¹⁶

Prosthetic valve deterioration can result in regurgitation attributable to leaflet calcification and noncoaptation or leaflet degeneration with a tear or perforation. Even in asymptomatic patients with severe prosthetic regurgitation, valve replacement is reasonable because of the risk of sudden clinical deterioration if further leaflet tearing occurs. IE should be excluded or concurrently treated. If a “watchful waiting” approach is taken in asymptomatic patients with severe prosthetic valve regurgitation, referral to a Comprehensive Heart Valve Center is prudent.

In some patients, operative risk is high, or surgery is not feasible. Nonrandomized studies have demonstrated clinical success with percutaneous paravalvular leak closure performed by expert operators under the supervision of an MDT at a Comprehensive Valve Center. Procedural success rates for percutaneous paravalvular leak closure, typically defined by no more than mild residual regurgitation and the absence of death and major complications, are highly variable. In a large single-center cohort, percutaneous repair of 141 paravalvular defects was attempted in 115 patients, with an achieved overall success rate of 77% and a 30-day complication rate of 8.7%.⁶ In another study of 126 patients who underwent percutaneous paravalvular leak repair, Sorajja and colleagues reported a 3-year survival rate of 64.3%.⁵ The degree of residual regurgitation affects symptom improvement and survival. In a cohort of 231 consecutive patients (2006–2017) who underwent percutaneous mitral paravalvular leak closure, the reduction of paravalvular leak to mild or less was achieved in 70% of patients.^7,8 Those patients with mild or less residual paravalvular leak had a survival rate at 3 years of 61%, compared with a rate of 47% in patients with greater degrees of residual paravalvular leak (P=0.002).^7,8 Notably, treatment of HF symptoms with paravalvular leak closure is more successful than is treatment of hemolysis. IE should be excluded before attempted paravalvular leak repair.

The Valve-In-Valve International Data registry examined outcomes of transcatheter ViV procedures in 459 patients, of whom about 30% had isolated regurgitation and 30% had mixed lesions.¹⁰ Within 1 month after the ViV procedure, 7.6% of patients died, 1.7% had a major stroke, and 93% of survivors experienced good functional status. The 1-year survival rate was 83.2%.¹⁰ Several systematic reviews have compared outcomes of transcatheter ViV with those of reoperative SAVR. In 1 report, ViV had similar hemodynamic outcomes to repeat surgery and lower stroke risk and bleeding risk than repeat surgery.¹¹ A meta-analysis of 498 patients demonstrated no significant differences in early and mid-term all-cause mortality rates with ViV or reoperation.¹⁷ In another meta-analysis of 342 patients, reoperative AVR was compared with transcatheter ViV for failed degenerated aortic bioprosthesis and the group undergoing reoperative AVR had a lower all cause mortality with superior hemodynamic outcomes.¹⁸ Thus, although transcatheter ViV appears to be a safe and feasible alternative to repeat SAVR in patients who are inoperable or at high surgical risk, repeat SAVR should remain the standard of care, particularly in low-risk patients.¹⁸

Beta-blocker medications are used to control heart rate or to treat arrhythmias. However, maternal use of beta blockers has been associated with a newborn birth weight approximately 100 g lower than that of newborns whose mothers did not take beta blockers.⁶ The use of beta blockers with beta-1 selectivity avoids the beta-2 effects on uterine relaxation. The incidence of fetal growth retardation is lower with metoprolol treatment than with atenolol treatment in pregnancy.^1–6

Diuretic medications can alleviate the effects of volume overload in pregnant women with VHD and HF symptoms (Stage D). However, reduction of volume overload must be balanced against the reduction in placental blood flow associated with diuretic medications.^7,8 Additionally, data from ROPAC suggested that maternal diuretic use was associated with rates of low birth weight and fetal mortality that were higher than for women not taking any medications. In part, this association was attributable to the severity of the underlying HF requiring treatment.^7,8

ACE inhibitors and ARBs are strongly associated with fetal malformations when used by women during pregnancy.^9–11

A preconception TTE is used to assess valve function, ventricular function, and pulmonary artery systolic pressure. Preconception TTE can help identify women with valve dysfunction who may benefit from valve intervention before conception. Results can facilitate patient counseling about specific risks of pregnancy.

The management of prosthetic heart valves during pregnancy is substantially different from the management of prosthetic heart valves in a nonpregnant patient. There is a much higher risk of mechanical valve thrombosis during pregnancy because of the hypercoagulable state. Choosing the appropriate anticoagulation strategy to balance risks to the mother and fetus requires a team familiar with management of prosthetic heart valves in pregnancy to provide comprehensive counseling. The management of anticoagulation also requires specialized expertise, and frequent titration of VKA or heparin doses is needed.^2,10 Transvalvular gradients increase during pregnancy because of increased heart rate, plasma volume, and stroke volume.¹ In the event that valve intervention is required during pregnancy, the comprehensive Heart Valve Team and maternal-fetal medicine team is required to optimize maternal and fetal outcomes.

Each anticoagulation strategy has relative advantages and disadvantages in terms of maternal and fetal safety, but there is no anticoagulation strategy that is consistently safe for the mother and fetus. The maternal mortality rate is >1%, and serious maternal and fetal complications are common, even with modern mechanical heart valves and careful management.^4–6 After counseling, some women with mechanical heart valves may choose not to become pregnant, whereas others may wish to proceed with pregnancy, so comprehensive and candid counseling about the risks of pregnancy with a mechanical heart valve is important.

Pregnancy is a time of increased risk of mechanical valve thrombosis, so there should be high suspicion of valve thrombosis in women with an embolic event, clinical deterioration, symptoms of HF, or a pronounced increase in valve gradients or valve regurgitation during pregnancy. TEE is useful to visualize leaflet motion and thrombus burden. Fluoroscopy and gated cardiac CT are also useful in evaluating patients with suspected valve thrombosis.^7,11,12

The risk of catastrophic valve failure or stroke is prohibitively high for women with mechanical heart valves who cannot take dose-adjusted and frequently monitored anticoagulation throughout pregnancy.^1–10

Much of the maternal morbidity and mortality during pregnancy occurs in women who are not receiving appropriate doses of anticoagulation because of improper administration, improper monitoring, or medication nonadherence. Women who are not able to receive therapeutic anticoagulation or do not have access to frequent monitoring and dose adjustment are at prohibitive risk for pregnancy.^13–15

No anticoagulation strategy is optimally safe for both the mother and the fetus. Warfarin is safest for the mother but crosses the placenta and can cause fetal intracranial hemorrhage; fetal loss; and teratogenicity, particularly at doses >5 mg/d and when given during the first trimester, keeping in mind that the warfarin dose needed to maintain a therapeutic INR may change during pregnancy. Neither UFH nor LMWH crosses the placenta, but each is associated with higher rates of maternal complications than are seen with warfarin.^{3,4,6,16,18,20,23,25,35,36} Depending on a woman’s values and priorities, she may choose an anticoagulation strategy that minimizes maternal risk, minimizes fetal risk, or attempts to achieve a balance between maternal and fetal risk. Physicians should not assume a woman’s values or preferences, nor should physicians supplant their own preferences for those of the patient. Counseling and shared decision-making allows for a woman and her physician to choose the best anticoagulation to achieve the woman’s goals.

Warfarin crosses the placental barrier and results in anticoagulation of the fetus, as well as the mother. There is a higher risk of fetal intracranial hemorrhage if the mother is fully anticoagulated with warfarin during vaginal delivery. Women taking warfarin can minimize the risk by switching to a heparin preparation before planned delivery.^{5,8,13,17–20}

Although LMWH does not result in an anticoagulated fetus, the risk of maternal hemorrhage is high if delivery occurs while the mother is on LMWH. Therefore, it is recommended that the mother be hospitalized before planned delivery, with discontinuation of long-acting anticoagulation and initiation of intravenous continuous infusion of UFH to keep aPTT >2 times control levels.^19,20

Intravenous heparin should be stopped long enough before delivery to reduce risk of maternal bleeding and allow safe placement of epidural anesthesia (typically at least 6 hours). Exact timing should be coordinated with the obstetrics and anesthesia teams.^19,20

Because warfarin results in an anticoagulated fetus, there is a high risk of fetal intracranial hemorrhage if vaginal delivery is attempted in a woman who is anticoagulated with warfarin. If a woman goes into labor while on warfarin, appropriate reversal of anticoagulation followed by cesarean section reduces the risk of fetal intracranial hemorrhage.^3,22,23

The teratogenic effects of warfarin are dose dependent. The rate of warfarin embryopathy is reduced (<3%) but not eliminated if the daily dose of warfarin is ≤5 mg/d.^{3,6.16,18,23,25,36,37} In most women, the effective dose of warfarin typically does not vary significantly during pregnancy.³⁸ For women who require a dose ≤5 mg/d, continuation of low-dose warfarin throughout pregnancy poses the lowest combined risk to mother and fetus.^3,6,16 The number of reported pregnancies on low-dose warfarin is relatively small, and not all publications have found improved fetal outcomes on low-dose warfarin,¹¹ so caution should be exercised until more data are available.

If warfarin is taken in doses >5 mg/d during the first trimester of pregnancy, there is a >30% risk of fetal loss or embryopathy. For women who require >5 mg/d to maintain a therapeutic INR, replacing warfarin with dose-adjusted LWMH during the first trimester reduces fetal loss.^{3,6,10,15,16,23,39,40} While the patient is taking LMWH, anti-Xa levels should be monitored at least weekly and the dose adjusted accordingly. Fixed dosing is never appropriate, because it is associated with high maternal morbidity and mortality.⁴¹ After the first trimester, the fetal toxicity of warfarin is substantially lower, so switching back to warfarin for the second and third trimesters results in a reasonable balance between maternal safety and fetal safety.

In regions where LMWH is unavailable or cost-prohibitive, or if anti-Xa levels cannot be monitored, continuous infusion of UFH can be used as an alternative to LMWH during the first trimester for women who require a warfarin dose of >5 mg/d.³⁷ If UFH is used during the first trimester, the dose should be adjusted to maintain an aPTT 2 times control. There are several disadvantages to UFH compared with LMWH: Women are at greater risk for line infections, osteoporosis, and heparin-induced thrombocytopenia, so UFH should be reserved for situations where dose-adjusted LMWH is not feasible.¹⁰ Intermittent subcutaneous injection of UFH is not an acceptable alternative because it is associated with prohibitive rates of valve thrombosis.⁴² UFH is associated with very high rates of valve thrombosis, stroke, and death in pregnant women with mechanical heart valves during the second and third trimesters.^3,4,43

In carefully selected women with thrombosis of a mechanical heart valve during pregnancy, low-dose, slow-infusion, tissue-type plasminogen activator (tPA) can be an alternative to surgical valve replacement. Women who are optimal candidates are hemodynamically stable and have obstructive prosthetic valve thrombosis, valve thrombosis with embolic complications, or nonobstructive valve thrombosis with a thrombus >10 mm.⁴⁴ Given the high rates of fetal loss with cardiac surgery during pregnancy, thrombolysis is an attractive alternative for appropriately selected, hemodynamically stable women with mechanical valve thrombosis.

Although the teratogenicity of warfarin is highest during the first trimester, there is still a risk of pregnancy loss or fetal hemorrhage when warfarin is taken during the second and third trimesters. Therefore, after appropriate counseling, some women may choose to avoid warfarin entirely throughout pregnancy. For these women, dose-adjusted LMWH throughout pregnancy is the safest alterative. LMWH throughout pregnancy is associated with a higher rate of thrombotic complications than warfarin. However, many of the thrombotic events occur when LMWH is administered improperly or monitored erratically or if patients are nonadherent.^8,14,15,39 When administered and monitored meticulously, LMWH can be safe.¹² Effective dose monitoring includes weekly measurements of anti–factor Xa levels, with additional monitoring after dose adjustment.¹³ Measurement of trough levels to maintain a trough Xa level >0.6 IU/mL may help women maintain therapeutic anticoagulation while on LMWH.^12,13

When warfarin is taken at a dose ≤5 mg/d, the risk of warfarin embryopathy is reduced but not entirely eliminated. Some women, after discussion with their physicians, may choose to substitute LMWH for low-dose warfarin during the first trimester to eliminate the risk of warfarin embryopathy. This choice improves fetal outcomes but at the cost of increased maternal thrombotic complications.^1–3,12

Low-dose aspirin is regarded as safe during pregnancy and can be continued in women with mechanical heart valves if needed for other indications. There may be noncardiac indications for aspirin in pregnant women, such as prevention of preeclampsia.²⁸

Studies using subcutaneous LMWH at a fixed dose without monitoring of anti-Xa levels in pregnant patients with mechanical prostheses found a high risk of valve thrombosis and maternal death. In pregnant women treated with dose-adjusted LMWH, the dose of LMWH required to maintain an adequate anti-Xa level 4 to 6 hours after administration increases throughout pregnancy.^9,15,39,41

A randomized clinical trial of dabigatran in nonpregnant patients with mechanical heart valves showed an increased rate of thromboembolic and bleeding complications with dabigatran compared with warfarin.⁴⁵ The safety and effectiveness of anti-Xa direct oral anticoagulants has not been established in patients with mechanical heart valves. Additionally, the safety of anti-Xa direct oral anticoagulants in pregnancy is unknown.^30–32

Anti-Xa direct oral anticoagulants have not been shown to be safe in patients with mechanical heart valves, so they should not be used in pregnancy.

The prevalence of CAD in patients with severe AS ranges between 15% and 80%.⁵ and varies depending on the definition of CAD used and the populations examined.⁶ The impact of CAD on outcomes is controversial,^7,8 although one report singled out severe CAD (defined by a SYNTAX score >22) and incomplete revascularization as the only independent predictors of death after TAVI.⁹ Assessment of the coronary anatomy is important in patients with severe AS to rule out obstructive CAD. Invasive coronary angiography is commonly performed. In patients with a low pretest probability of CAD, contrast-enhanced coronary CT angiography¹⁰ has an excellent negative predictive value.^11,12 In patients with normal renal function, an option is to combine contrast-enhanced coronary CT angiography with CT assessment of the peripheral circulation and heart structure as an initial imaging test, reserving coronary angiography for the event that the contrast-enhanced coronary CT angiography is nondiagnostic or significant CAD is found. Invasive functional assessment of coronary lesions in TAVI candidates by using fractional flow reserve or instantaneous wave-free ratio is safe and feasible.^13–15 Instantaneous wave-free ratio may be particularly attractive because it does not require the administration of a vasodilator and is less influenced by the effect of the stenotic aortic valve, although randomized clinical trials validating the utility of both are ongoing.

There are no RCTs to inform clinical practice on the benefits and timing of PCI in patients undergoing TAVI. The decision to perform PCI is therefore driven by myriad clinical factors (eg, presence of angina or ischemia, ability to take dual-antiplatelet therapy before TAVI) and anatomic factors (eg, lesion location and complexity, technical feasibility) and should be individualized. Overall, nonrandomized studies suggest that PCI before TAVI is safe and feasible,¹ even in patients with left main disease.² Conceptually, pre-TAVI PCI also allows a safer procedure and circumvents future post-TAVI PCI, which can be occasionally challenging. Staged PCI before TAVI is a common strategy in clinical practice and is associated with lower contrast volume and renal failure than is the strategy of TAVI with concomitant PCI,¹ although the timing of pre-TAVI PCI remains controversial.

Multiple RCTs have been conducted to define the optimal management of CAD in patients without VHD based on the SYNTAX score to define those least amenable to percutaneous treatment. Subsets of patients shown to have superior freedom from major adverse cardiac events include those with complex left main disease and those with a SYNTAX score >33.³ Accordingly, a surgical approach is reasonable in this subset of patients. Among SAVR patients, revascularization for those with significant CAD (>50% stenosis) has been shown to impact late risk of mortality favorably.⁴

CAD is frequently present among patients with VHD^1–4 and may contribute to angina pectoris among those with aortic valve disease.^3,4 Knowledge of coronary anatomy contributes to risk stratification, in addition to directing concomitant coronary revascularization. There is a very low prevalence of CAD among men <40 years of age and premenopausal women with no atherosclerotic risk factors^2,5–7 or history of mediastinal radiation.⁸

Functional MR occurs in patients with structurally normal valve leaflets and chordae because of LV dysfunction, including regional wall motion abnormalities or global dilation with displacement of the papillary muscles, leaflet tethering, annular dilation, and decreased closing forces from reduced contractility.^9–11 Because this LV dysfunction may be attributable to CAD and accompanying myocardial ischemia, the assessment of coronary anatomy status is necessary to complete the diagnosis and allow evaluation of revascularization options.

Contrast-enhanced coronary CT angiography is an alternative to coronary angiography among selected patients who are at low to intermediate pretest probability of CAD before valve surgery.¹² This does not include patients who have active symptoms of angina, those with documented ischemia, or those with a prior history of CAD, all of whom should have selective coronary angiography. Recent studies, most often in the setting of a pre-TAVI evaluation, have demonstrated diagnostic sensitivity of >90%, specificity of 60% to 90%,^13–15,21 and accuracy of >90%.²¹ Contrast-enhanced coronary CT angiography may be safer than coronary angiography in selected patient populations, such as those with IE and vegetations on the aortic valve. However, a positive contrast-enhanced coronary CT angiogram, defined as the presence of epicardial CAD, requires confirmation with invasive coronary angiography to establish the need for and extent of CABG. The risk of radiation exposure and renal failure because of the contrast injection should be taken into consideration.

The presence of uncorrected CAD has been shown to negatively impact both perioperative^22,23 and late outcomes of surgery for VHD.¹⁹. Accordingly, concomitant CABG has been favored. These studies of concomitant CABG at the time of valve surgery have demonstrated little or no adverse impact on the acute perioperative mortality rate, despite increased cross-clamp and cardiopulmonary bypass times. Moreover, combined CABG and valve surgery reduces the rate of perioperative myocardial infarction, and incomplete revascularization is associated with greater postoperative LV systolic dysfunction and a reduced survival rate after surgery as compared with patients who receive complete revascularization. For more than a decade, improved myocardial preservation techniques have been associated with reduced overall operative mortality rates, and it has become standard practice to bypass all significant coronary artery stenoses, when possible, in patients undergoing valve surgery. In patients with a significant stenosis of the left anterior descending artery, a left internal thoracic artery graft should be used if possible. Hybrid PCI followed by surgical valve repair or replacement has been reported favorably but is restricted to patients at high risk with a combined surgical approach.²⁰