TAVI In Heart Failure, How Much Risk is Acceptable?

This editorial refers to 'Heart failure in severe aortic valve stenosis: prognostic impact of left ventricular ejection fraction and mean gradient on outcome after transcatheter aortic valve implantation', by M. Gotzmann et al., published in this issue on pages 1155–1162. Transcatheter valve implantations like TAVI and MitraClip implantations are becoming more important treatment options in patients with heart failure who cannot undergo an operation or for whom the operations risk for conventional surgical therapies is considered too high.1,2 The PARTNER trial provided evidence that transcatheter aortic valve implantations (TAVI) give a significant survival benefit in patients who are not surgical candidates2 and give comparable results in patients with high operative risk, compared with surgical aortic valve replacement (SAVR).3 In recent years the number of TAVI procedures has increased at a high rate. At the moment over 50,000 TAVI procedures have been performed. Due to the success of TAVI, there tends to be a shift towards performing TAVI in lower-risk categories, despite the fact that there is a complete lack of evidence about these groups. Although TAVI gives a significant survival benefit in the previously-described groups, the mortality rates at 1 year are still high.2–4 It is therefore important carefully to select patients who will benefit from the procedure, and even more important to identify those patients who will not benefit from TAVI. In this issue of the European Journal of Heart Failure, Gotzmann et al.5 present the results of a TAVI population stratified by left ventricular ejection fraction (LVEF) and mean gradient over the aortic valve, based on a proposal from the Quebec group.6 Originally this division was made by echocardiography. In patients with an aortic valve area index (AVAi) beneath 0.6cm2, flow was measured by indexed stroke volume (SVI). Gradients were measured by Bernoulli's equation. In the article by Gotzmann only mean gradient and LVEF are used to identify these subgroups. A normal LVEF with a small AVAi, indicates low flow, but no SVI are reported. The group with low flow/low gradients and a preserved LVEF is often referred to as paradoxical AOS.6 The paradox is in the fact that with a severe AOS and a normal LVEF, a normal flow is expected, which is contradictory to what is found in these patients. The diagnosis is difficult and physicians often do not recognize this entity of severe AOS. Subsequently, these patients are not referred for AVR and therefore do not receive optimal treatment. Another difficulty in the diagnosis of paradoxical AOS is that the diagnosis is susceptible to measurement error. Errors in measurement of the LVOT diameter and gradients are common and have to be excluded. Due to these facts some physicians even doubt if the diagnosis actually exists. The study of Hachicha6 reports that the patients with paradoxical AOS had smaller LV cavities and a greater degree of concentric remodelling. The lower values for SVI and transvalvular flow observed in this group are presumably caused by the association of both a lower LVEF and a smaller ventricle, compared with the group with normal flow and high gradient. Although the LVEF in paradoxical AOS is above 50%, there are more subtle signs of LV dysfunction (fractional shortening, concentric hypertrophy) reflecting a group with more advanced disease. Unfortunately important data about flow (SVI) and LV-function (fractional shortening) are not reported in the article of Gotzmann. The LV dimensions mentioned in the article are more or less the same between the groups with a normal LVEF. The LV-mass index differs, indicating that the normal LVEF/low gradient group has more hypertrophic ventricles. This is in line with the idea of the paradoxical AOS as stated in the article of Hachicha et al.6 The spontaneous prognosis of the low flow/preserved LVEF group is worse than in the group with high gradient a preserved LVEF.7 However, after undergoing SAVR, a comparable excellent result has been found in the groups with preserved LVEF.6 It is therefore surprising that Gotzmann et al. find an increased mortality in the group with low gradient and normal LVEF after TAVI. When comparing the baseline characteristics of the patient groups in this article, it becomes apparent that the low gradient with preserved LVEF group has more co-morbidities compared with the high gradient/preserved LVEF group (Renal function, previous CABG, previous MI, CAD), resulting in higher Euroscores. So, more likely this difference in the baseline characteristics is the cause of the higher mortality found in this group. Furthermore, it has to be stated that with these small numbers no firm conclusions can be drawn. Besides mean gradient over the aortic valve, Gotzmann et al. stratify their patients with severe AOS by LVEF. There is conflicting evidence whether LVEF is a predictor of mortality in TAVI patients. In SAVR outcome is strongly dependent on preoperative LV-function.8–10 In TAVI, however, a few smaller studies suggest that patients with a reduced LVEF have a comparable mortality compared with groups with a preserved LVEF.11 These findings are in contrast to results reported in some larger TAVI registries, where it was found that a decreased LVEF is an independent predictor of survival.12,13 An important factor is that patients with a reduced LVEF tend to have more and more advanced co-morbidities, which is also illustrated in the article of Gotzmann et al., Table 1. The higher burden of illnesses is reflected in higher operative risk scores and eventually a higher mortality. The very high mortality in the low LVEF/low gradient group makes it clear that there is need for a more stringent selection. The question arises if we have to decline these patients for TAVI. Gotzmann states that despite a high mortality many of these patients show significant improvement of LV-function. These findings are in accordance with the findings of others.11,14 Rejecting this group of patients for TAVI means withholding a reasonable amount of people an important survival benefit and improvement of quality of life. To come to a better selection, the focus must lie on the risk factors of this group of patients. Possibly cardiac contractile reserve at dobutamine stress echocardiography (DSE) can play a role in differentiating which patients will benefit from TAVI in the low flow/low EF group. There is some evidence for the use of DSE in SAVR.15,16 Also, BNP/NTproBNP levels can play a role in this decision.17 On the other hand, we have to keep an eye on the risk factors that do not depend on the LV-function. Currently STS score and Euroscore are most often used to estimate risk in TAVI patients. Although these scores have proven to be of value, they are not designed to estimate the risk in a TAVI population and have a low discriminatory ability. Several pre-procedural risk factors for TAVI have already been indentified: liver cirrhosis, COPD, pulmonary hypertension (sPAP.60 mmHg), LVEF, diabetes mellitus, renal insufficiency, NYHA class III/IV, mitral valve regurgitation (3 + or 4 +).4,12,13 Recently, focus tends to shift to frailty indices. However, none of these risk factors individually are powerful enough to predict risk of patients undergoing TAVI. Eventually a TAVI specific risk score is the best alternative to predict risk in TAVI patients. To reach this goal, further research is needed. The difficulty for physicians in selecting patients for TAVI is that the patients, who have the highest risk, are most in need for treatment. These patients have the worst prognosis and quality of life. It is often stated that at a certain point, patients who have too many co-morbidities or who are too frail will not benefit from TAVI any longer. In this so-called 'Cohort C' population (referring to the Partner trial), TAVI must be avoided. The idea makes sense but the real world is not that strict. There is no 'point of no return'; more likely it is a continuum. If the estimated risk of the procedure increases, the mortality rates increase subsequently. The question to be answered is: where do we draw the line? Further research on risk factors and long-term results is needed to answer this question more adequately and to give us a firm base in decision-making. Finally, the physicians must decide if the benefits for the individual patient outweigh the risk of the procedure, keeping in mind the recently-reported overuse of TAVI.18 This decision has to be made by the multidisciplinary heart team, which contains, at the least, thoracic surgeons, echo-cardiologists and interventional cardiologists.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.