We here provide a structured review and meta-analysis on eBAV and eTAVI in CS including the totality of existing evidence available. The main findings of our study are that both eBAV and eTAVI in this setting are associated with high rates of mortality.
Clinical challenges
Treatment of CS due to AS remains a challenging scenario and its optimal clinical management still needs to be determined and, up to date, no randomised trial investigated the optimal interventional therapy of this cohort. Current guidelines suggest BAV to be considered as a bridge to SAVR or to transcatheter intervention30 in haemodynamically unstable patients and in those with AS who require urgent high-risk non-cardiac surgery (recommendation level IIb, evidence C).7 However, clear guidance concerning the interventional treatment of CS is still missing and its optimal timing still needs to be determined. This comprehensive review provides the largest analysis performed so far and valuable insights to guide clinical decision-making in CS patients.
Additionally to the need for clear and evidence-based guidelines, the number of patients presenting with CS due to severe AS has been rising over the last years.27 29
Mortality and complications in eBAV and eTAVI
In general, mortality rates in patients presenting in CS due to decompensated AS are high. A study comparing elective and urgent TAVI outcomes showed that patients undergoing urgent TAVI experienced a 4-fold increase in 30-day mortality rates (17.5% for urgent TAVI compared with 4% for elective TAVI, p=0.001), along with an increased incidence of cardiovascular mortality within 30 days (25.3% in non-elective vs 15.1% in elective, p=0.043), life-threatening bleeding (11.5% in non-elective vs 4.1% in elective, p=0.018) and vascular complications (11.5% in non-elective vs 4.6% in elective, p=0.031).31 Given the high-risk setting in decompensated AS patients, our meta-analysis points towards a favourable pooled estimated 30-day and 1-year mortality rate in the eTAVI group. A recent meta-analysis investigating 36 886 patients undergoing eTAVR found that while urgent TAVR was associated with higher mortality and readmission rates compared with elective TAVR, it demonstrated a reduced mortality risk and comparable safety profile relative to urgent SAVR or BAV, supporting the feasibility of eTAVI in CS settings.32 This result is in line with the only previous smaller meta-analysis in this field.33 Wernly et al reported a 30-day mortality rate for eBAV (total sample size 238) of 46.2% (CI 30.3% to 62.5%; I²=74%) and for eTAVR (total sample size 73) of 22.6% (95% CI 12.0% - 35.2%; I² =26%)33 also pointing towards eTAVI for the preferred emergency procedure. While this prior analysis only incorporated 8 studies including 311 patients, our meta-analysis is now reporting outcomes for eBAV and eTAVI for a total of 2811 patients from 17 studies.
Furthermore, a recent US study evaluated readmission rates in patients who received TAVI or BAV as an urgent procedure. In the urgent TAVI group, the rate of all-cause readmissions within 30 days was notably reduced (15.4% vs 22.5%, with an adjusted HR (aHR) of 0.92 (0.90 - 0.95), p<0.001) when contrasted with the urgent BAV group. This trend persisted for readmissions at 90 days, where the aHR was 0.75 (p=0.005). Readmissions at 30 days due to cardiovascular reasons and congestive heart failure were also decreased in the urgent TAVI group (aHR of 0.93, p<0.001 and aHR of 0.98, p=0.040, respectively) compared with those in the BAV group. Moreover, the urgent TAVI group experienced a significant reduction in 90-day readmissions for cardiovascular reasons.34 While this study has not been included in our analysis as the urgent procedure did not meet the criteria for CS, these findings are in line with our results regarding the favouring trend towards eTAVI and our pooled estimated rates for adverse events are only slightly higher. The 30-day mortality rates observed in this study reflect the outcomes of a critically ill population with inherently high mortality. While these rates provide valuable real-world insights into the outcomes of eTAVI and eBAV, they should not be interpreted as evidence of causality, as the lack of a non-interventional arm precludes definitive conclusions about the direct impact of these procedures on survival. Investigating outcomes in a non-interventional cohort would be both ethically and practically challenging, as withholding potentially life-saving interventions in this setting is not feasible.
Even when evaluating long-term outcomes at 5 and at 7 years after TAVI, as reported by Ichibori et al, patients identified with high or prohibitive surgical risk exhibited encouraging long-term survival outcomes post-TAVI procedure, with rates of 58.8% at 5 years and 45.3% at 7 years.35 Also interestingly, mortality rates in early BAV studies from 199114 remain consistent decades after,20 indicating that mortality is rather phenotype-associated and procedure-associated than related to gained expertise or improved technique of the method over the years. Nevertheless, none of the studies included in the analysis and cited so far were randomised. Therefore, a selection bias definitely needs to be acknowledged, particularly in those countries and healthcare systems with limited quotas of TAVI available.
In addition, eBAV can cause severe acute regurgitation, which can worsen the already compromised haemodynamics of CS. Nevertheless, our results show a rather low pooled estimated rate of severe aortic regurgitation for eBAV (6%) but a relatively high rate for eTAVI (4%).
Of note, in the largest prospective non-randomised trial on eTAVI in CS patients,27 authors emphasised that despite the expected higher mortality rate in CS patients compared with non-CS patients, mortality rate was associated with the degree of CS.27 It seems that mortality in the CS group was not predominantly influenced by complications arising from the procedure but is instead determined by the severity of CS.27 Given the critical nature of CS, the timing of intervention is particularly important, as any delay could lead to irreversible organ damage. Further research should focus on defining the optimal time point for intervention, as well as exploring patient selection criteria and long-term outcomes. While there are no longitudinal studies investigating the optimal time point for interventional treatment in CS, one study showed that within an acute heart failure cohort, a trend for reduced all-cause mortality at 2 years was noted for patients undergoing TAVI within 60 hours of admission in comparison to those treated later.36 However, the 30-day all-cause mortality rates were similar between both groups.36 Further studies should focus on the right time point of intervention in CS patients with structural investigation of CS severity with right heart catheterisation.37
Patient selection for eTAVI and eBAV might additionally impact patient outcomes. eTAVI is generally performed on patients who are less critically ill, with better preserved left ventricular function and lower rates of coronary artery disease. In contrast, eBAV is often chosen for patients with more severe left ventricular dysfunction and a higher prevalence of coronary artery disease, conditions that are associated with increased mortality.38 Although eTAVI has shown favourable outcomes, these differences in patient selection suggest that eBAV may still be the preferred intervention for patients with higher risk profiles or those in environments where TAVR is not available.38
Moreover, previous studies often employed older-generation devices that had larger delivery profiles and a higher risk of paravalvular leak. The use of newer-generation devices, which have smaller delivery systems and improved sealing mechanisms, may reduce these risks and improve outcomes in this patient population.
Ethical and practical considerations for treatment decision-making
Notably, studies in the meta-analysis that reported a cause-specific mortality stated that mortality was mostly not due to complications such as bleeding, stroke or myocardial infarction but due to multiorgan failure or septic conditions,25 resulting in pooled estimated rates for procedural complications being considerably lower than pooled estimated mortality rates in-hospital and after 30 days. The findings suggest that concurrent conditions, including chronic renal failure, severe three-vessel disease and frailty, could significantly influence the rate of early mortality following procedures such as eTAVI or eBAV. In order to take these considerations into account as well as patient will and ethical aspects, a brief emergency heart team meeting might be beneficial to improve outcomes and avoid futility.39 Such a team discussion should also include consideration of precipitating factors or triggers (eg, acute coronary syndrome, atrial fibrillation, infections) that should be addressed before proceeding with eTAVI.39
Limitations
Despite two studies,19 24 all other studies included in the meta-analysis were single-arm studies, making a direct group comparison unfeasible. This reliance on single-arm studies limits our ability to draw definitive conclusions, as it restricts the analysis to a pooled estimated rate for each endpoint. Thus, our findings should be considered hypothesis-generating only. Furthermore, secondary endpoint data were not available in all studies and, most importantly, were not assessed at the same time and using the same standards in all studies. Some studies assessed secondary endpoint data at 30 days while others reported those data in hospital. In addition, we noted a relevant heterogeneity of the reported secondary endpoints, which did not always comply with the VARC-2/3 criteria (predominantly in the older studies, table 2).40 This heterogeneity, combined with the lack of standardisation, particularly for the secondary endpoints, introduces variability that can obscure the true effects and reduce the reliability of our findings. Therefore, these data should be interpreted with caution, and the generalisability of the results may be limited. In addition, in 23% of the eBAV patients, the procedure was used as a bridging procedure for a valve replacement with either TAVI or SAVR, adding a further bias in our analysis (table 1). However, a clear indication of the time point of the second intervention is not available, as only median and ranges are provided.
Moreover, the studies included in our meta-analysis were conducted over a wide time range, contributing to heterogeneity and potentially impacting the reliability of pooled estimates.
The studies incorporated into our analysis cover an extensive historical period, reflecting the comprehensive nature of our research. However, our three oldest studies from the 1990s had a limited weight in our pooled rates, mainly due to their small sample sizes. Despite this, the evolution in clinical practice and patient management over time introduces another layer of variability, which may impact the study’s conclusions. Of note, our leave-one-out sensitivity analyses did not detect any differences (online supplemental figure 1).
In addition, CS is a condition with a wide range of severity, which is not easy to harmonise and compare. The studies included in our meta-analysis are characterised by relevant heterogeneity and did not report outcomes according to the SCAI classification. A further relevant limitation is the lack of information concerning the strategy used to achieve valve sizing in eTAVI. In elective cases, CT scan is the gold standard to evaluate access site and annulus size. However, anecdotal cases of urgent TAVI with sizing using transoesophageal echocardiography or BAV have been reported.41 The absence of high-resolution anatomical data from CT scans could have influenced acute outcomes and prognosis, further complicating the interpretation of our results.