The principal finding of our study was that although ECS was not frequently performed after TAVI, it was associated with high mortality at 30 days and 1 year. One-third of patients had died at 1 year follow-up, even though procedural mortality was 0. In addition, the causes for mid-term mortality were not directly related to the initial TAVI-related complication or the ECS performed. Based on our findings, ECS after complicated TAVI should therefore not be omitted when deemed necessary.

These findings are in corroboration with those from previous studies. In the largest meta-analysis to date, including over 9000 patients from 46 studies, ECS after complicated TAVI was performed in 1.1% of cases—in two-thirds of patients, the femoral access route was used [5]. A more recent analysis of data from the EuRECS-TAVI, comprising 27,760 patients from 79 centres, showed an incidence of ECS after transfemoral TAVI of 0.76% [9]. Similar to our study results, embolisation and valve dislocation, guidewire perforation and annular rupture were the most common reasons for ECS in both studies [5, 9]. Patient profiles were also comparable to our study population, with a mean logistic EuroSCORE of ~20 (24 and 17, respectively) and mean age of ~82 years (81.3 and 82.4 years, respectively).

Even though TAVI is considered safe and effective in patients with a high, intermediate and low risk of operative mortality in multiple trials, in current clinical practice, those selected for TAVI are still usually elderly, frail patients with comorbidities who are deemed either inoperable or at high risk for SAVR [16]. The occurrence of a complication during TAVI for which sternotomy is deemed necessary therefore raises a potential ethical dilemma. Conducting ECS for complicated TAVI should be carefully considered in a multidisciplinary discussion with patients (if the clinical situation at the time permits this), cardiologists and surgeons, preferably prior to the procedure.

Our findings indicate that ECS after complicated TAVI can be performed safely, although it has a high but acceptable perioperative and 30-day mortality, taking into account the otherwise potentially lethal consequences. Nevertheless, the suspected nature of the complication should be taken into account as there are some studies describing a variety of outcomes depending on the type of complication (e.g. [15]). Event rates in our study were too low to make a distinction per type of complication, but the EuRECS-TAVI suggests that mortality of, for example, valve embolisation is ~22% and that of annular rupture is ~62% [9].

The decision to perform ECS after a complicated TAVI procedure should therefore also be driven by the nature of the complication itself. On the contrary, if the patient is not treated with ECS, the 30-day mortality may exceed the 18.8% found in this study. Furthermore, the 9 patients requiring ECS for valve embolisation had no procedural and 30-day mortality and therefore support the strategy to perform ECS despite comorbidities. In the TAVI population at our facility that did not undergo ECS during the same period, the cause of death in only 1 patient was valve prosthesis dislocation, thus supporting the necessity to perform TAVI in centres with cardiac surgical backup on site.

To our knowledge, the current study is the first to report complications other than mortality after ECS for complicated TAVI. In our cohort, 2/16 patients (12.5%) had to undergo re-exploration, which exceeds the average incidence for cardiac surgery in the Netherlands (~5%) [17]. However, this high re-exploration rate in our study did not come with an increased incidence of deep sternal wound infections, even though multiple procedures are a known risk factor for this complication [18]. The rate of permanent pacemaker implantation after ECS was also higher than the national average (~5%), which was most likely attributable to the preceding TAVI procedure as permanent pacemaker implantation after TAVI in general is estimated to be 14% [17,18,19].

This retrospective study had limited power to draw firm conclusions due to the low frequency of complications after ECS. Additionally, relevant parameters such as predictors of ECS and causality analysis could not be identified due to an insufficient number of events. Furthermore, the operator’s expertise and experience level may impact ECS after TAVI, although the EuRECS-TAVI found no difference in mortality between low- and high-volume centres [9]. Finally, data on long-term functional status and quality of life were not available.