• Aortic Stenosis

Some years ago, a pleasant 76-year-old patient presented to our Valve Clinic for a second opinion on her aortic stenosis (AS). She was reassured by her primary cardiologist that her valve ‘was not quite there yet for surgery’, but continued to be troubled by mild dyspnoea. She just wanted to make sure nothing was missed. Her home echocardiogram showed a valve area of 1.0 cm², with a peak velocity of 3.8 m/s and a mean gradient of 36 mm Hg; the official interpretation on the home report was that of moderate AS.

Ever since echocardiography was shown to reliably estimate aortic valve gradients and valve area,1 ,2 it has de facto replaced catheterisation for assessment of AS severity. Indeed, both European and American professional societies endorse echocardiography as initial approach, with additional studies (CT, invasive haemodynamic assessment) reserved for those patients in whom disease severity is indeterminate, or when clinical and echocardiographic findings are discordant.3-5 However, the traditional cut-off values used for the definition of severe AS (aortic valve area (AVA) <1.0 cm, mean gradient >40 mm Hg and peak velocity >4 m/s) have been criticised as being intrinsically inconsistent, as a gradient over 40 mm Hg is more likely to be associated with AVA <0.8 cm.6 7

In their Heart manuscript, Mehrotra and colleagues evaluate the outcomes of patients with severe AS stratified by AVA, and compare them with outcomes of moderate AS.8 The major finding is that about two-thirds of patients with AVA 0.8–1.0 cm2 have normal flow, low gradient, and that outcomes in this subgroup are similar to those of patients with moderate AS in the short term (1.5 years); patients with high gradients or low-flow, low-gradient AS do as poorly as patients with AVA <0.8 cm².

There are a few points that merit discussion. First, decision-making in AS is among the most complex in valvular heart disease. Assessment of AS severity is only the starting point and a complicated one at that (figure 1). The anatomical defect posed by the restricted valve opening has variable haemodynamic consequences depending on flow conditions. Indeed, a low stroke volume (with or without reduced ejection fraction) will result in low gradients/velocities despite a severely narrowed valve; an increase in stroke volume (with dobutamine administration for low ejection fraction, nitroprusside for severely increased vascular resistance, exercise or postectopic beat) uniformly yields higher gradients in true severe AS. As such, traditional measures of AS severity (AVA, velocity, gradient) must be assessed in the context of flow rather than in isolation.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

The complexity of decision-making in AS. Decision-making in AS is among the most complex in valvular heart disease, with severity of disease, presence of symptoms, comorbid conditions and risk/benefit of aortic valve replacement all playing interconnected roles in patient outcome. AS, aortic stenosis; AVA, aortic valve area; AVR, aortic valve replacement; LV, left ventricle.

Once presence of severe AS has been established, the second challenge is to assess existence of symptoms and understand to what extent they are linked to AS. The classical triad of dyspnoea, angina and syncope can be present in nearly 40% of patients with mild AS,9 and therefore can be attributed to comorbid conditions (coronary artery disease, hypertension, diastolic dysfunction and so on) rather than valvular disease. Conversely, 15%–65% of patients with asymptomatic severe AS who undergo stress testing have abnormal findings,10 suggesting that many patients under-report or do not recognise symptoms in AS. Obviously, aortic valve replacement (AVR) will not help symptoms in patients with mild AS, whereas it may be life-saving in ‘asymptomatic’ severe AS with abnormal stress findings.

Of course, patient outcome is what matters most, and that is where our science is the weakest. Most of the studies on outcomes in AS (including the current one by Mehrotra and colleagues) include AVR among the endpoints. This is understandable (cardiac surgery is indeed a substantial event), yet biased as the decision to proceed with AVR is subjective. Indeed, the presence of a perceived risk factor (such as AVA <1.0 cm2 in the present study) is more likely to result in closer follow-up and referral for AVR. Using mortality alone would be preferable in outcome studies, but would require a substantially larger number of patients.

To return to our patient, was AS moderate because the mean gradient was <40 mm Hg and the valve area did not cross below 1.0 cm²? We felt that disease severity was advanced enough to justify AVR, and that brings home the last point. To be systematic in our thinking, we create rigorous patterns and cut-off values for all diseases. While that is essential for practising medicine, we must resist our inclination to become fascinated with numbers. Diseases, including AS, are a continuum rather than a step-like process (figure 2). Nothing magical happens to the individual patient with AS when the gradient changes from 39 to 41 mm Hg, with diabetes when glucose swings from 125 to 127 mg/dL, or with obesity when the body mass index increases from 29.9 to 30.1 kg/m2. As beautifully shown in the present study, our category bins may contain a mixture of diseases, and the clinicians must use their skills to understand what is going on with the patient sitting in front of them. This is why practising medicine remains in part a form of art — and why we love it.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

The continuum of disease versus the fascination with numbers. Disease progression is a continuous process. We must not let our fascination with cut-off numbers get us into thinking in a stepwise mode. AS, aortic stenosis.