Low-gradient, low ejection fraction aortic stenosis (AS) represents about 5?10% of all cases of severe AS and is the most challenging subgroup of patients to manage.1 2 The term low-gradient, low ejection fraction AS is usually applied to patients with a mean gradient<30 mm Hg (or 40 mm Hg), an aortic valve area (AVA) <1 cm2, and an ejection fraction (EF) <35% (or 40%).1?7 Low EF in such patients may be caused by severe low-flow AS with inadequate compensatory left ventricular hypertrophy, called afterload mismatch, but also by another myocardial disease (such as extensive fibrosis, associated cardiomyopathy or myocardial infarction (MI)),5 in which case, AS is not the primary problem. The essential difficulty for clinicians is to distinguish true severe low-flow AS, responsible for low EF, from pseudo-severe AS comprising mild-to-moderate AS associated with another cause of left ventricular dysfunction (LVD). A very low gradient may be seen in true severe low-flow AS, while the decreased AVA seen in pseudo-severe AS reflects poor opening of the aortic valve directly related to low transvalvular flow. For the clinician, the two main questions in low-gradient, low ejection fraction AS are: How severe is the AS? Which patients can benefit from surgery?

   Determining the true severity of low-gradient, low ejection fraction AS may be a challenging problem. Physical examination is often misleading with a soft systolic ejection murmur despite severe AS. Echocardiography is the key examination for visualising the aortic valve, and measuring the maximum aortic velocity, pressure gradient, valve resistance, AVA and EF. As initially reported by deFilippi et al,6 dobutamine echocardiography is extremely helpful to potentially distinguish severe from non-severe AS. Dobutamine may increase the stroke volume (SV) and consequently the aortic transvalvular flow. Contractile reserve (CR) during dobutamine echocardiography is currently defined by an increase of SV >20%.2 7 In pseudo-severe AS, an increase of SV results in an increase of AVA to reach a non-severe range (>1.2 cm2), whereas the gradient (<30 mm Hg) is not significantly altered owing to the larger AVA.3 The additional value of a new index derived from dobutamine echocardiography, the projected effective AVA at a flow rate of 250 ml/s, has been proposed to distinguish patients with true severe low-gradient, low ejection fraction AS from patients with pseudo-severe disease.8

   In this issue of Heart, Burwash et al, using dynamic positron emission tomography, demonstrate that patients with true severe low-gradient, low ejection fraction AS have a marked tendency to higher resting myocardial blood flow and more severe impairment of myocardial flow reserve than patients with pseudo-severe AS (see page 1627).9 Although these preliminary data are promising, larger studies are necessary to confirm the value of these parameters to distinguish these two entities and to determine their possible role and feasibility in routine clinical practice. The same group reported that B-type natriuretic peptide is significantly lower, but with a considerable overlap of values, in pseudo-severe low-gradient, low ejection fraction AS than in true severe AS.10

   In true severe low-gradient, low ejection fraction AS, the dobutamine-induced increase of SV leads to a marked increase of gradient with only a slight change in AVA.3 6 7 These patients are considered to have severe low-gradient, low ejection fraction AS with CR. In contrast, in some patients, dobutamine infusion does not induce any change in SV and consequently no change in gradient or AVA.3 6 7 These patients are considered to have low-gradient, low ejection fraction AS without CR, but the severity of AS remains undetermined. In these patients, severe aortic valve calcification on echocardiography, fluoroscopy or CT is an indicator of considerable AS.

   In selected patients with low-gradient, low ejection fraction AS, aortic valve replacement (AVR) is associated with a dramatic improvement of long-term survival compared with medical management.2 7 11 12 Moreover, among patients who survive AVR, an improvement in functional status is usually seen and the EF increases by at least 10% in more than 80% of cases.4 Perioperative mortality in low-gradient, low ejection fraction AS ranges from 8 to 21% in recent publications.1 3 7 11 12In a multicentre European study, we reported a significant decrease of operative mortality, from 20% in the 1990s to 10% over the past 5 years.12 Perioperative mortality and lower long-term postoperative survival are mainly related to older age, presence of comorbidities, very low pressure gradient (<20 mm Hg), severe associated coronary artery disease or history of MI, and absence of CR with dobutamine.1 12 A high operative mortality of about 30% is reported in patients with low-gradient AS without CR.2 4 7 12 Conversely, those patients with CR had an operative mortality of only 5?8%. Patients with severe AS and CR therefore have an acceptable operative risk and AVR may improve long-term survival and functional status in most cases.2 7

   In the recent European guidelines, surgery is advised in this group of patients with CR.2 Because of the very high operative risk in the absence of CR, many clinicians consider that the absence of CR constitutes a contraindication to AVR. However, in these patients, the prognosis with medical management is extremely poor and there is a trend, in a small patients series, towards better survival after AVR.7 Moreover, in those patients who survive the perioperative period, functional status improved in more than 90% of cases and LVEF improved by >10% in more than 60% of cases.4 Perhaps the main determinant for low preoperative EF in these patients without CR is an afterload mismatch that cannot be corrected by inotropic stimulation (dobutamine). Therefore, the absence of CR with dobutamine in low-gradient, low ejection fraction AS does not systematically indicate irreversible LVD. According to the European guidelines,2 "surgery can be performed in these patients but decision-making should take into account clinical condition, in particular the presence of comorbidity, degree of valve calcification, extent of coronary artery disease and feasibility of revascularisation". Thus the decision for AVR surgery should be made case by case and the absence of CR should not be considered to be an absolute contraindication to AVR. In our opinion, surgery could be considered in patients with calcified aortic valve without CR in the absence of scarring due to extensive MI and excessive comorbidities, and in the presence of a basal mean gradient >20 mm Hg. A heart transplant should also be considered in eligible patients. As the currently available data are unable to identify clearly the subset of patient who will have a better outcome with surgery, further studies must be conducted in large series of patients with low-gradient, low ejection fraction AS without CR. In the near future, percutaneous AVR could be a valuable alternative for surgical high-risk patients with low-gradient, low ejection fraction AS.Patients with pseudo-severe AS should be carefully followed up to detect any change in the severity of AS and of LVD. Medical treatment, which may induce inverse remodelling, must be systematically optimised. In patients with increased QRS duration and ventricular asynchrony, cardiac resynchronisation should be discussed. AVR is classically not recommended in reports of pseudo-severe AS. For example, patients with low-gradient, low ejection fraction AS with an AVA on dobutamine echocardiography clearly greater than 1.3 cm2 and with another cause of LVD, such as a large scar of MI, are not suitable for AVR. However, because the increased afterload due to the "moderate" AS may be not well tolerated in the presence of a severe associated LVD, surgery, in our opinion, should be carefully discussed in some patients with an AVA of about 1.2?1.3 cm2 on dobutamine. Unfortunately, the currently available data on pseudo-severe AS are too limited to allow any definitive conclusions about the management of these patients, and further multicentre studies including larger number of cases are mandatory in this heterogeneous subset of patients.