Aortic stenosis (AS) is now very frequently encountered, to the extent that aortic valve replacement (AVR) has become the second most frequently performed cardiac surgical procedure. Concomitantly, the most frequently encountered etiology and physiopathology of this entity have changed considerably. Indeed, 30 years ago, patients with AS were typically younger individuals with either rheumatic heart valve disease or congenitally bicuspid aortic valves. The reduction in aortic valve area (AVA) was usually fixed and due to a fusion of the valvular commissures, and the primary disease was deemed to be limited to the valve, without any particular involvement of the left ventricle, the ascending aorta, or the rest of the vascular tree. In contrast, AS is at present most frequently encountered in older individuals, and the most prevalent etiology is a degenerative process that is more or less akin to a variant form of atherosclerosis. The result of this process is valve leaflet thickening with decreased mobility but typically without commissural fusion. Hence, the reduction in AVA is not fixed and will vary in relation with prevailing hemodynamic conditions (i.e., the valve will tend to open more in response to an increase in cardiac output). In addition, these individuals frequently have other manifestations of atherosclerosis, such as decreased systemic arterial compliance, which may contribute to an additional increase in left ventricular (LV) hemodynamic load, and/or coronary artery disease, resulting in a decrease in LV function beyond that which would have been expected if the disease had been purely limited to the valve. Indeed, a reduction in systemic arterial compliance entails an increase of both the systolic and pulse pressure (e.g., for a similar stroke volume, the blood pressure can be 160/80 mm Hg instead of 120/80 mm Hg), in which case the LV systolic pressure given a peak-to-peak valvular gradient of 50 mm Hg could be ±210 mm Hg instead of ±170 mm Hg. Likewise, the coexistence of myocardial ischemia due to coronary artery disease may entail a decrease in ejection fraction beyond that expected for a given degree of AS severity. Hence, the clinical presentation of the disease is often much more complex and multifaceted than previously believed, and clinicians are often faced with new diagnostic and therapeutic challenges requiring new solutions. Such challenges and solutions are elegantly illustrated by two reports in the current issue of JASE .

In the first study, Lee et al. examined the relation between valvuloarterial impedance ( Z _va ) and longitudinal myocardial strain in patients with paradoxical low-flow (PLF) AS. The latter is a relatively new entity, and Z _va is a relatively new parameter designed to measure the global LV hemodynamic load. Hence, the results of the study confirm that the adverse outcomes in patients with AS are more closely related to Z _va than to parameters such as AVA and gradients. As mentioned, patients with degenerative AS often have a concomitant decrease in systemic arterial compliance resulting in systolic hypertension, and their ventricles are therefore submitted to a double increase in load (i.e., both valvular and vascular). From the standpoint of pathophysiology, it should therefore be of no surprise that decreases in LV function and adverse clinical outcomes have been found to be more closely related to a parameter such as Z _va , representing the total increase in hemodynamic load, rather than to AVA or gradients, which reflect only the increase in valvular load due to valve stenosis. In this context, it may seem somewhat intriguing that the patients with PLF AS reported by Lee et al. had marked increases in Z _va compared with the patients with normal-flow AS but were nonetheless normotensive. Unfortunately, values for systemic arterial compliance or vascular resistance are not reported for these patients, but the most likely explanation for this observation is that their peripheral blood pressures were pseudonormalized because of their low-output state. Indeed, systemic arterial compliance and vascular resistance are invariably increased in patients with Z _va levels of this magnitude, and pseudonormalization of peripheral blood pressures has been observed in up to 30% of patients with PLF AS and/or Z _va > 4.5 mm Hg · mL ⁻¹ · m ² . In contrast to previous studies, the patients with PLF AS reported by Lee et al. also had, on average, relatively high gradients (i.e., >40 mm Hg) despite a substantial decrease in cardiac output; a similar finding was observed in a minority (26%) of our own patients with PLF AS and was associated with more severe AS (average indexed AVA, 0.3 cm ² /m ² ) compared with the patients with PLF AS with lower gradients. The fact that the patients with PLF AS in Lee et al. ’s had on average normal rather than decreased values for LV dimensions is also atypical. Indeed, patients with PLF AS and high gradients usually have markedly decreased LV dimensions (the average LV end-diastolic dimension in our series was 43 ± 5 mm compared with 50 ± 4 mm in this series) and their low-output state is deemed to be due to restrictive physiology in relation to marked concentric hypertrophy and a small LV cavity. This observation, plus the fact that these patients also had increased LV systolic volumes, a relative decrease in ejection fraction, and a marked increase in relative wall thickness, may in fact suggest that these patients are at the upper end of the spectrum of severity and that their ventricles might be about to dilate progressively and on the brink of developing a frank decrease in LV ejection fraction. Notwithstanding these considerations, the bulk of these results constitute additional evidence that patients with PLF AS, whether associated or not with low gradients, are definitely at a more advanced stage of their disease in relation to a more pronounced increase in global LV hemodynamic load. Hence, in our series and the present series, average Z _va values were 5.3 and 5.6 mm Hg · mL ⁻¹ · m ² , respectively, in patients with PLF AS compared with 4.1 and 3.62 mm Hg · mL ⁻¹ · m ² in patients with normal-flow AS.

The use of LV longitudinal shortening as a surrogate of disease severity and increase in LV hemodynamic load is also worth emphasizing from the standpoints of both pathophysiology and diagnostic accuracy. Following a 1979 study of LV geometry and radial, circumferential, and longitudinal function in patients with various forms of valvular disease, Dumesnil et al. concluded, “To our knowledge, the fact that ventricular longitudinal axis shortening may be selectively decreased in patients with AS, as we found in this study, has never been previously observed and there is no clear explanation for this phenomenon. Since subendocardial myocardial fibers are mostly oriented longitudinally, one might hypothesize that the decrease in longitudinal axis shortening is a consequence of the subendocardial abnormalities that have been observed in patients with aortic valve disease. Whether measurements of ventricular longitudinal axis shortening in these patients may become useful in estimating the extent of such abnormalities as well as the severity of the disease remains to be determined.” The present study, along with other previous studies, further documents this hypothesis by showing that the decrease in global longitudinal strain is largely determined by the extent of the increase in global LV hemodynamic load and extent of LV concentric remodeling. Future studies will be necessary to determine if this consequence is predominantly related to an increase in subendocardial wall stress or if it is not also due to structural damage in relation to myocardial fibrosis. Moreover, and as emphasized by the investigators, this finding is also important from a diagnostic standpoint. Indeed, the main pitfall associated with the echocardiographic diagnosis of PLF AS is an error in the calculation of the stroke volume, because this measurement is included in the calculation of most parameters (e.g., AVA, systemic arterial compliance, and Z _va ) used to make the diagnosis. Moreover, it is derived from two separate measurements (i.e., LV outflow tract diameter and LV outflow tract time-velocity integral), each with potential for error. Hence, longitudinal myocardial strain by speckle tracking represents an independent measurement that can be performed both routinely and reproducibly, and as implied by the results of the present study, it should improve our diagnostic accuracy in providing independent corroboration that PLF AS is indeed present.

The study by Iwahashi et al. in this issue goes one step further in showing that, in the final analysis, prognosis is determined primarily by the overall effect of disease(s) on the myocardium. Indeed, these results add to a mounting body of evidence showing that B-type natriuretic peptide (BNP) level is a more robust predictor of outcome than any other parameter reflecting stenosis severity or increase in hemodynamic load. In this context, it should be emphasized that the pathophysiology of adverse outcomes in AS is primarily related to an imbalance between the global increase in LV overload, independent of whether it be of valvular and/or vascular origin, and LV reserve both at rest and during exercise. Hence, in patients with isolated AS, the extent of myocardial abnormalities is essentially related to the severity of the disease, which can be described in relatively simple terms such as AVA, peak velocities, and gradients. However, as emphasized above, elderly patients with degenerative AS often have other vascular abnormalities, such as a decrease in systemic arterial compliance with concomitant systolic hypertension and associated coronary artery disease with related LV dysfunction. Hence, BNP measurements bear an analogy with valvuloarterial impedance; both of these variables correlate better with myocardial abnormalities and outcomes than the usual parameters of AS severity, because they are measures of the overall impact of disease(s) on the left ventricle. Conceptually, BNP may also have an added advantage over Z _va in that it is probably more sensitive to assess the additional impact of conditions such as coronary artery disease and cardiomyopathy on the myocardium.

Unfortunately, Iwahashi et al. do not report Z _va or systemic vascular resistance in their study. The fact that BNP correlated better with LV mass index than any other index of AS severity, however, tends to further corroborate the concept that myocardial damage and clinical outcomes are determined primarily by the total burden of disease(s) on the ventricle rather than by AS severity alone. Moreover, the observation that elevated BNP and adverse clinical outcomes were present in many patients with preserved ejection fractions is consistent with the conclusions of Lee et al. to the effect that myocardial damage is often present in patients with severe AS and preserved ejection fraction. With regard to the clinical use of BNP, there are two caveats worth mentioning. First, the threshold values for adverse events appear to vary considerably from one study to the other. Hence, unless the value are unequivocally elevated (e.g., >500–600 pg/mL), an individual result should not be interpreted in isolation but rather in light of other clinical variables; also, it may serve as a benchmark for further monitoring the patient’s evolution. Second, as observed in this study, the adverse outcomes associated with elevated BNP appear to occur even in patients who receive appropriate treatment (i.e., AVR). This finding is somewhat disturbing in that it may suggest that irreversible myocardial damage has already occurred and that optimal results with regard to LV function and symptomatic status might have been better achieved had the operation been performed earlier in the course of the disease (i.e., before BNP levels had started to increase). A most interesting observation in this regard is also that of Bergler-Klein et al. in patients with severe AS, low gradients, and reduced ejection fractions, in whom BNP levels >550 pg/mL were the best predictor of mortality regardless of treatment (medical vs surgical) or whether the patients had myocardial contractile reserve. Hence, it appeared difficult from these data to determine which patients would have benefited from AVR and whether the assessment of myocardial contractile reserve had any relevance with regard to clinical decision making. Indeed, it might well be that patients with very high BNP levels may in fact be beyond treatment by classical AVR and should perhaps be considered for transcatheter AVR or even heart transplantation rather than surgical AVR. Notwithstanding these considerations, the study by Iwahashi et al. is further evidence that BNP is a very robust predictor of severity and outcomes in AS. From a practical standpoint, we therefore believe it should be routinely measured and become an integral part of the clinical decision-making process in patients with AS. Furthermore, it might also become a very interesting tool for follow-up from the standpoint of cost and benefit. Indeed, the continued observation of low and stable levels of BNP in an asymptomatic patient might preclude the repeated and unnecessary use of more expensive investigations.

In summary, these two interesting studies further confirm that although historically, AS severity has been expressed in terms of gradients and valve areas, it seems clear that these measures alone provide only a partial picture of the severity of AS and its effect on the heart and on outcomes. Hence, newer measures such as valvuloarterial impedance, LV longitudinal shortening, and BNP, which provide a broader view of the effect of net overload (both valvular and vascular) on the myocardium, have additive value and ought to be considered more routinely in evaluating patients with AS.