Discussion
The present study describes the HT treatment decision algorithm for a defined cohort of consecutives patients with severe AS from a tertiary referral hospital and two additional satellite hospitals. A local consensus document was used to define the patients who needed to be discussed in the HT. After HT discussion, more than 50% of the patients were allocated to the TAVR group, and approximately 20% to each of the other treatment groups. Baseline characteristics that determined the allocation group were mainly older age, concomitant MR, surgical risk scores, renal function and frailty. Patients without intervention had a 1-year mortality rate which was three times higher than either of the intervention groups, mainly driven by cardiovascular death. Readmission rates at 1-year was close to 50% in both the TAVR and SAVR groups.
The HT concept stems from 2 randomized controlled trials comparing surgical and percutaneous strategies in coronary artery disease and AS (10,11). The purpose of the HT in these trials was to choose suitable candidates for both interventions. The function of contemporary HT discussions should be to apply the clinical acumen of the HT members to the selection of patients for medical, transcatheter and surgical treatment, as basing this decision purely on risk scoring systems may not properly reflect specific high-risk characteristics in some patients (12). As such, the HT approach provides a patient-specific decision based on the overall patient profile.
The role of the HT has become more prominent in recent years since the introduction of TAVR as an alternative treatment for severe AS, particularly in elderly patients with multiple comorbidities. The treatment of severe AS is a perpetually evolving area, with current evidence attesting to its value in lower risk patients (13,14), making continuous evaluation of HT decisions in this changing clinical environment of paramount importance. In our center, the HT format was designed by cardiac surgeons, and interventional, clinical and imaging cardiologists who take part in weekly meetings. Referral physicians also participated via video-link. The routine schedule of a weekly meeting dedicated entirely to patients with AS, allowed all resources to be focused on these patients to optimize the HT decision making process. The number of candidates referred to the HT increased over the years, suggesting a greater penetrance of the HT concept and, probably, greater value being placed in HT decisions by the referring physicians.
A critical aspect of the HT is to determine which patients would not benefit from an invasive approach and avoid futility (15). Reasons to avoid an invasive treatment were heterogeneous, but in general invasive treatment was avoided in patients considered to have a high-risk of mortality. The most common reasons to avoid invasive treatment were severe comorbidities or acute and critical illness. Despite careful decision making to avoid futile invasive procedures, a relatively high percentage of patients who underwent an invasive treatment died (~15%) or were readmitted (~40%) within one year, suggesting that the patient selection process could still be improved. Specially, patients referred to the TAVR group were very old and had several significant comorbidities, which conferred a high risk of dying from non-cardiovascular causes (more than half of the patients in this cohort). Continuous evaluation of the HT decisions, with a special focus on these high-risk patients, should be implemented to identify patients at higher risk of mortality and attempt to diminish future futile interventions. In our cohort HT decisions could be reconsidered if changes in the clinical situation arose and approximately 10% of the patients were subsequently changed from the initial allocated group.
Several factors (older age, significant MR, frailty, eGFR and surgical risk scores) defined by the CART analysis impacted the clinical decision making in accordance to others (6). Significant MR was identified as an important co-morbidity which increased the likelihood of referring the patient for surgery, particularly in those without frailty. However, the concomitant role of significant MR in patients with AS is still unresolved (16). Previous reports showed a negative impact of untreated baseline ≥ grade III MR in both TAVR (17,18) and SAVR populations (19,20). On the other hand, a double valve intervention is associated to higher perioperative mortality than isolated SAVR (21). Among those with significant MR, frailty was a determinant factor to choose TAVR over SAVR. Frailty has been described as a strong predictor of peri-procedural complications and mid-term outcomes after cardiac surgery (22,23). However, in the setting of TAVR, frailty may not be significantly related to peri-procedural mortality or morbidity; although, it appears to have impact on mid-term outcomes (24). This may be due to the less physiologically stressful nature of TAVR compared to SAVR (25).
In our study, chronic kidney disease (CKD), a frequent comorbidity in patients with AS(26), increased the chance of referring the patient for TAVR over SAVR. A previous study in patients with CKD has shown better short-term outcomes in TAVR compared to SAVR (27). Despite the well-recognized limitation of surgical risk scores in predicting outcomes in TAVR patients, systematic calculation of theses scores provided useful and objective information for the HT discussion. Current European and American guidelines support their use for surgical risk stratification based on the inclusion criteria of randomized trials (2,3,28). Moreover, surgical scores still provide a general risk prediction in individualized patient and may predict a futile intervention (29).
The purpose of the study was not to compare different therapeutic options, due to the non-randomized nature of the interventions and the selection bias by the HT. However, general results were in accordance with previous registries and randomized trials with high-risk patients(11,28). While acute kidney injury, new onset atrial fibrillation and significant bleeding were higher in the SAVR group, vascular complications were more frequent in TAVR patients. Stroke and in-hospital mortality were similar. While residual aortic regurgitation tended to be higher in TAVR patients, valve hemodynamics overall were better in this group. Also, both invasive groups showed similar results in terms of long-term mortality and rehospitalization rates.
This study has the inherent limitations of any observational study without an external adjudication event committee. However, it demonstrates the practical real-world issues related to the current management of AS patients within a HT format. In this study we present the results of a single HT with a limited sample size. Additionally, not all patients with AS were evaluated by the HT. Although, wide-ranging criteria were prospectively set to define those that should be referred to the HT, a number of patients were treated without HT discussion and were not included in our study. Patients in the MT arm were not specifically followed-up and outcomes were identified in a retrospective fashion. This study spans the time period before the publication of trials on low risk patients, and the results cannot be extrapolated to current practice relating to patients in lower risk groups.