Should catheterization of patients with aortic stenosis include retrograde crossing of the aortic valve to assess the severity of aortic valve stenosis? What is the risk? Bernhardt and colleagues (page 32) present their provocative findings of cerebral embolism detected by magnetic resonance imaging (MRI) of aortic stenosis patients prospectively randomized to cardiac catheterization with or without retrograde catheterization of the stenotic aortic valve. The authors found a 22% incidence of acute cerebral events, 3% clinically apparent, in patients undergoing retrograde crossing of the aortic valve compared with a 0% incidence in patients without crossing of the aortic valve.
Although some of the details of catheterization were omitted, such as the catheter and wire used, as well as intensity of anticoagulation, it is clear that the valves were readily crossed (average fluoroscopy time about 3 minutes) despite the severity of aortic stenosis (average aortic valvular area 0.7 cm2). Interestingly, no correlation was found between risk factors usually thought to place the patient at increased risk (eg, patient’s age, aortic valve area, aortic atheroma, atrial fibrillation, left atrial and ventricular size or function, spontaneous echocardiographic contrast, carotid stenosis, hypertension) and the actual occurrence of cerebral emboli.1-3
The reported clinical rate of cerebrovascular accident is at the high end of what is generally expected; an earlier report yielded a clinical stroke rate of 0.9% in 435 patients undergoing retrograde catheterization with aortic stenosis.1 Considering the clinical rate of 3% in the current study, it is not surprising that the subclinical event rate was 22%. After all, clinically apparent strokes are just the “tip of the iceberg” of events.
Most cerebral embolic events go undetected as a result of small size or location in “silent areas” of the brain. The complete absence of events in the group without retrograde catheterization of the stenotic aortic valve also is not surprising. Transcranial Doppler echocardiography has demonstrated that microembolic signals are detectable routinely during cardiac catheterization and are influenced by the choice of catheter and guidewire, though most events typically are not clinically significant.4
Whereas direct measurement of the aortic valve area by echocardiography is often not possible because of valve calcification, Doppler and
2-dimensional echocardiography can provide an accurate assessment of the severity of aortic stenosis. Echocardiographic assessment of aortic stenosis is congruent with catheterization-assessed gradients and valve areas in 97% to 98% of patients, and both methods result in a similar decision of whether to operate.5,6 Many of the premier US hospitals already routinely rely on echocardiographic measurements instead of cardiac catheterization for the assessment of aortic stenosis. For example, the Mayo Clinic and Cleveland Clinic, respectively, reported invasive hemodynamic assessment of only 35% in 1992 (from 54% in 1990) and 30% in 1994 (from 64% in 1986) of patients with aortic stenosis.7,8
In a minority of cases, the Doppler pressure gradient can be misleadingly low and may necessitate catheterization to confirm the severity of aortic stenosis. Confirmation by catheterization is commonly needed when the Doppler beam cannot be aligned with the aortic jet. This may occur with coexisting aortic insufficiency and in severe left ventricular dysfunction when the gradient is low and significant discrepancies may be found between echocardiographic Doppler-predicted and catheterization-predicted valve area. Local practice also will dictate the accuracy of echocardiographic diagnosis and whether the cardiovascular surgeons are willing to operate solely on the basis of these findings.
What does the future hold? It is likely that the trend toward an increased reliance on noninvasive techniques for diagnosis will continue. MRI flow planimetry already has been demonstrated to yield similar results to catheterization and echocardiography in assessing aortic stenosis, and each year the ability of MRI to image the coronary arteries improves.9 It will not be surprising if, within a decade, catheterization laboratories are reserved primarily for interventions in which procedural risks are balanced by the benefits of the procedure. For diagnostic testing, a premium will be placed on safety, and noninvasive methods will dominate practice.