The new england journal of medicine perspective Heart Failure in Aortic Stenosis — Improving Diagnosis and Treatment Michael R. Zile, M.D., and William H. Gaasch, M.D. The development of heart failure in patients with tional to the stroke volume divided by the square aortic stenosis is associated with a high mortality root of the pressure gradient. Therefore, if the stroke rate — unless aortic-valve replacement is per- volume declines, as it does in some patients with formed. There is an especially high risk of death aortic stenosis in whom heart failure has developed, among patients with aortic stenosis and a decreased there is a proportional decline in the pressure gra- ejection fraction. Before surgery is performed in dient. Under these low-flow conditions, the cal- such patients, initial management must include an culated effective aortic-valve area may indicate the evaluation of the severity of the stenotic lesion and presence of severe aortic stenosis, despite a low the functional state of the left ventricle; in addition, transvalvular pressure gradient. A mean pressure the heart failure must be treated and the patient’s gradient that is less than 30 mm Hg in a patient with condition stabilized. It is possible to pursue both what appears to be severe aortic stenosis (an aortic- of these goals simultaneously with echocardio- valve area of <1 cm2) indicates what is referred to as graphic techniques or cardiac catheterization tech- “low-gradient aortic stenosis.” niques, together with selected pharmacologic in- An example of a low pressure gradient in a pa- terventions. Proper evaluation and treatment require an un- derstanding of the pathophysiology of heart failure ABC in patients with aortic stenosis. Hypertrophic re- Aortic modeling provides a compensatory mechanism by pressure which the left ventricle can generate increased sys- tolic pressures while maintaining normal systolic Left ventricular wall stress (afterload) and a normal ejection frac- pressure tion. If, however, the hypertrophic remodeling is Base Line DB NP DB NP DB NP inadequate, systolic wall stress will be increased. There is an inverse relation between wall stress and Stroke volume 0 or 0 or the ejection fraction: the presence of afterload ex- Pressure gradient 0 0 0 cess results in a decline in the ejection fraction. Aor- Aortic-valve area 0 0 0 0 tic-valve replacement can increase the ejection frac- Diagnosis True aortic Relative aortic Severe left tion by correcting the afterload excess created by a stenosis stenosis ventricular dysfunction truly stenotic valve. Benefit from aortic- Yes Possible No A second mechanism that can produce a de- valve replacement pressed ejection fraction in patients with aortic ste- nosis is a decline in the intrinsic contractility of the Hemodynamic Response to Dobutamine (DB) and Nitroprusside (NP) in Pa- myocardium. Aortic-valve replacement may have lit- tients with Aortic Stenosis and a Low Transvalvular Pressure Gradient. tle or no effect on the ejection fraction if decreased The base-line panel shows an example representing a typical patient with low- contractility coexists with “relative” aortic stenosis gradient aortic stenosis. The hemodynamic response to treatment with dobu- tamine and nitroprusside helps to clarify whether such a patient has true aortic (or “pseudo” aortic stenosis) or if there is a primary stenosis (Panel A), relative aortic stenosis (Panel B), or severe left ventricular cardiomyopathy (see Figure). dysfunction (Panel C). A zero denotes no change, a single arrow an increase in In true severe aortic stenosis, the aortic-valve the value for the particular variable, and two arrows a large increase in the value. area is constant (or nearly constant) and is propor- n engl j med 348;18 www.nejm.org may 1, 2003 1735 Downloaded from www.nejm.org by GEOFFREY K. LIGHTHALL MD on May 06, 2003. For personal use only. No other uses without permission. All rights reserved. The new england journal of medicine tient with true severe aortic stenosis is shown in A third group of patients has little or no response Panel A of the Figure. The low pressure gradient is to dobutamine or nitroprusside (Panel C). These a consequence of a low stroke volume, which re- patients have severe and generally irreversible left sults in large part from increased systolic wall stress ventricular dysfunction and a poor prognosis. Al- (afterload excess). The intravenous administration though these patients have aortic-valve disease, the of dobutamine or nitroprusside produces an in- valve disease itself does not contribute substantial- crease in the stroke volume and an increase in the ly to the left ventricular dysfunction. Medical therapy systolic pressure gradient across the aortic valve, is directed at the cardiomyopathic failing ventricle. but there is no change in the calculated aortic-valve Once heart failure develops in patients with true area. The absence of a change in the valve area un- aortic stenosis, aortic-valve replacement should be der different hemodynamic conditions indicates the performed as soon as the patient is medically sta- presence of true severe aortic stenosis. Patients with ble. Medical treatment includes the judicious use true stenosis have a salutary response to aortic-valve of diuretics and may, in selected cases, require the replacement. use of a positive inotropic agent or a vasodilator. However, not all patients with a low pressure Vasodilators such as nitroprusside have generally gradient and a calculated aortic-valve area of less been thought to be contraindicated in patients with than 1 cm2 in fact have true severe aortic stenosis. aortic stenosis because of the danger of hypoten- Some of these patients have aortic-valve disease sion. In this issue of the Journal, Khot et al. (pages without severe stenosis (relative aortic stenosis). In 1756–1763) report their experience with the thera- these patients, the decreased ejection fraction and peutic use of nitroprusside in patients with true low pressure gradient are caused primarily by de- aortic stenosis and a decreased ejection fraction. creased contractility. It is difficult to differentiate They reasoned that aortic valvular obstruction and patients with true aortic stenosis from those with increased total vascular resistance act in concert to relative aortic stenosis, because low-flow states lim- load the left ventricle and that such a doubly loaded it the accuracy of the valve-area calculation. It is un- ventricle might benefit from a reduction in systemic der these circumstances that selected pharmaco- resistance. They found a substantial increase in car- logic interventions can be helpful. diac output (with both high-gradient and low-gradi- An example of a low pressure gradient in a pa- ent aortic stenosis), and they concluded that nitro- tient with relative aortic stenosis is shown in Panel prusside was a safe and effective therapeutic bridge B of the Figure. Such patients have a variable aortic- to aortic-valve replacement. However, the patients valve area that is dependent on the stroke volume. they studied were highly selected, were invasively The administration of dobutamine or nitroprusside monitored, and were being treated in an intensive produces an increase in flow across the aortic valve care unit. There are substantial risks associated with that is well out of proportion to the increase in the this approach that should limit its wide application. transvalvular pressure gradient; during infusion of Whether treatment with positive inotropic agents nitroprusside, the gradient may even decrease. As a would produce a similar effect with fewer risks re- result, the calculated aortic-valve area increases by mains to be determined. at least one third, or 0.3 cm2, and may exceed 1 cm2. Patients with aortic stenosis, a depressed ejec- Such a patient has relative aortic stenosis, and the tion fraction, and a low transvalvular pressure gra- value of aortic-valve replacement in this condition dient continue to pose a management challenge. Ju- remains in doubt. It is not known whether medical dicious use of pharmacologic interventions in an treatment significantly alters the outcome, nor is it invasive setting has the potential to improve diag- clear that the response to dobutamine or nitroprus- nosis and treatment. side is predictive of the subsequent response to medical therapy. Given the rapid advances in surgery From the Cardiology Division, Department of Medicine, Gazes and valve design, some patients with relative aor- Cardiac Research Institute, Medical University of South Carolina, and the Ralph H. Johnson Department of Veterans Affairs Medical tic stenosis may benefit from aortic-valve replace- Center — both in Charleston, S.C. (M.R.Z.); and the Department of ment. In general, however, although optimal treat- Cardiovascular Medicine, Lahey Clinic, Burlington, Mass. (W.H.G.). ment remains uncertain, treatment of relative aortic stenosis should be focused on medical therapy. 1736 n engl j med 348;18 www.nejm.org may 1, 2003 Downloaded from www.nejm.org by GEOFFREY K. LIGHTHALL MD on May 06, 2003. For personal use only. No other uses without permission. All rights reserved..