Can TAVI be performed without on-site cardiac surgery?

Abstract

Introduction

Aim of this analysis in to assess the prevalence and post-procedural outcomes of surgical bailout during transcatheter aortic valve implantation (TAVI).

Methods

Patients undergoing TAVI from September 2017 to March 2023 were enrolled from two high volume centers. All the procedures were performed with on-site cardiac surgery, but especially the scrubbed cardiac surgeon. The primary endpoint was in-hospital mortality of TAVI patients after emergent cardiac surgery (ECS). Secondary endpoints were intra-operative and 1-year mortality, and post-procedural complications such as acute kidney injury (AKI), stroke, myocardial infarction (MI), conduction abnormalities, need for inotropic support and intensive care unit (ICU) and in- hospital length of stay.

Results

A total of 1347 consecutive patients underwent transfemoral TAVI. Ten patients (0.74 %), representing the study population, reported intra-procedural complications requiring ECS: seven patients received a self-expandable prosthesis; three patients received a balloon expandable prosthesis. Indications for ECS included: type A dissection (n = 2), aortic annulus rupture (n = 1), left(n = 1) and right (n = 2) ventricle perforation, mitral valve apparatus damage (n = 2), prosthesis embolization (n = 2). Four patients required post-operative inotropic support. One case of minor stroke and one case of AKI (grade III) were reported. Three patients developed a post procedural left bundle branch block (LBBB). Median ICU and hospital length-of-stay were 4.5 (2–7.75) days and 14 (8–22) days, respectively. One case of in-hospital mortality was reported.

Conclusions

The on-site cardiac surgery, with the scrubbed heart surgeon, represents a life-saving resource for TAVI centers in case of ECS, and it is essential to achieve low-rate in-hospital mortality.

1. Introduction

Degenerative aortic stenosis (AS) is the most common valvular disease in developed countries, with a prevalence of 3 % among people over the age of 75 [1].

For decades, aortic valve replacement (AVR) has been recognized as the gold standard for the treatment of severe AS. Nowadays, transcatheter aortic valve implantation (TAVI), has become a well-established, safe and effective alternative to standard surgery in selected patients with severe AS.

Several trials reported that TAVI had non-inferior rates of death, strokes, and re-hospitalization compared to surgery at short and mid-term follow-up, even in patients at low and moderate risk [[2], [3], [4], [5], [6], [7]]. After more than two decades of worldwide experience, the incidence of severe complications requiring Emergency Cardiac Surgery (ECS) during TAVI has markedly dropped [8]. This, along with the dramatic increase in waiting list times, has open a heated debate about the feasibility of a TAVI program in centers without on-site cardiac surgery departments.

The aim of this study is to assess the prevalence of ECS and to investigate the short-and mid-term outcomes of surgical bailout during TAVI procedures in hospitals with the cardiac surgeon scrubbed supporting or performing the procedure.

2. Material and methods

2.1. Patient population and study design

From September 2017 to March 2023 data of consecutive patients undergoing TAVI for severe AS at two high-volume Cardiac Surgery Centers (AO Ordine Mauriziano, Turin Italy and AOU Città della Salute e della Scienza, Turin, Italy), were prospectively collected, and retrospectively analyzed.

Patients who underwent transfemoral (TF) TAVI and who experienced TAVI-related complications requiring ECS were enrolled in the study.

Procedural risk profile was calculated using EuroSCORE II [10].

ECS was defined, according to the Valve Academic Research Consortium-3 (VARC-3) consensus document, as any conversion to open sternotomy or thoracotomy during TAVI secondary to procedure-related complications [11]. Surgical procedures due to access-site complications were not included in this definition.

In hospital mortality consists of all-cause mortality occurred during the index procedure hospitalization.

Each patient has been candidated to TF TAVI after an accurate multidisciplinary evaluation based on clinical history, blood tests, electrocardiogram, transthoracic echocardiography, and computed tomography (CT). Particular relevance was given to pre-operative CT that is mandatory to select the best access for each patient in order to minimize access-related complications [12].

2.2. Operative technique [13,14]

All TAVI procedures were carried out under conscious sedation by a TF approach in a hybrid operating room with at least one senior cardiac surgeon (member of the Heart Team), one senior interventional cardiologist, one anesthetist, one perfusionist with the cardiopulmonary by-pass prepared and placed on stand-by and two nurses. No procedure was performed without at least one scrubbed cardiac surgeon who was involved as first or second operator.

Both new generation balloon-expandable (Sapien, Edwards Lifesciences, Irvine, CA, USA) and self-expandable prosthesis (Evolut [Medtronic, Minneapolis, MN, USA], Portico/Navitor [Abbott, Chicago, IL, USA ]), were implanted. If necessary, a valvuloplasty was performed before and/or after prosthesis implantation at the operator’s discretion.

In case of ECS local anesthesia was converted to general anesthesia. Surgery was performed via median sternotomy or right mini-thoracotomy, using cardiopulmonary by-pass and cardioplegic cardiac arrest, except for left/right ventricle perforation repairs that were performed off-pump.

Salvage of an embolized prosthesis into distal ascending aorta or aortic arch, as well as ascending aorta and hemiarch replacement for acute type A dissection, required circulatory arrest and deep hypothermia.

All reported complications occurred during TAVI procedure.

2.3. Outcomes and ethical considerations

Primary endpoint was in-hospital mortality. Peri-operative stroke, acute kidney injury (AKI), myocardial infarction (MI), need for inotropic support, conduction abnormalities (including new-onset atrial fibrillation, new-onset left bundle branch block and complete atrio-ventricular block requiring pace-maker implantation), intensive care unit (ICU) and in-hospital length of stay were defined as secondary endpoints.

Post-operative stroke was defined as clinical and radiological evidence of a new post-operative cerebrovascular event.

AKI’s definition was based on Kidney Disease: Improving Global Outcome (KDIGO) criteria [15].

Peri-operative myocardial infarction was recorded in the case of cTn-T value > 10 times the 99th percentile of the upper reference limit during the first 48 h with electrocardiographic (ECG) abnormalities and/or angiographic or imaging evidence of new myocardial ischemia/new loss of myocardial viability [16].

The degree of post-operative hemodynamic support was quantified using the Vasoactive Inotropic Score (VIS) [17].

All patients received an annual out clinic or telephone survey.

The follow up was completed on October 15, 2024.

The study was conducted in accordance with the ethical principles reported in the Declaration of Helsinki and the study design was approved by the local Comitato Etico Interaziendale A.O.U. Città della Salute e della Scienza di Torino – A.O. Ordine Mauriziano – A.S.L. Città di Torino” (protocol number 260-2022). Informed consent was obtained from all patients.

2.4. Statistical considerations

Data were prospectively collected into a dedicated database and retrospectively analyzed for a 6-year period. For continuous variables data were presented with media and standard deviation (SD) or median and interquartile range (IQR), for categorical variables data were represented with frequency and percentages.

All analyses were performed with SPSS 26.0 (IBM, Chicago, USA).

3. Results

During the study period 1347 patients underwent TAVI in the two indexed centers. Ten (0.74 %) patients required ECS for major complications.

In patient who underwent ECS, mean age was 81 ± 2.2 (range 78–84) and six patients were female. Two patients had undergone previous cardiac surgery. Three patients had advanced NYHA class at the time of the procedure. Pre-TAVI mean EuroSCORE II was 6.87 ± 6.3 % (range 1.69 – 24.75 %). Table 1.

Table 1.

Preoperative patients’ characteristics (n = 10).

Variables
Age, mean (SD), years	81 ± 2.2
Female, n (%)	6 (60)
BMI, mean (SD), kg/m2	25,93 ± 4.5
Hypertension, n(%)	8 (80)
Diabetes mellitus, n(%)	2 (20)
Dyslipidemia, n(%)	5 (50)
Smoke, n(%)	1 (10)
History of cerebrovascular event, n (%)	2 (20)
PAD, n(%)	4 (40)
COPD, n(%)	1 (10)
CKD III-IV, n(%)	2 (20)
Poor mobility, n(%)	3 (30)
History of heart failure, n(%)	1 (10)
Redo surgery, n(%)	2 (20)
EF, mean (SD)	56 ± 5.03
NYHA III-IV, n (%)	3 (30)
sPAP, mean (SD), mmHg	38.14 ± 8.45
Euroscore II, mean (SD)	6.87 ± 6.3

Table 1 SD: standard deviation, BMI: body max index, PAD: peripheral artery disease, COPD: chronic obstructive pulmonary disease, CKD: chronic kidney disease, EF: ejection fraction, NYHA: New York Heart Association, PAPs: systolic pulmonary artery pressure.

Detailed procedural data and indication for ECS are reported in Table 2. Seven patients received a self-expandable prosthesis, while three patients a balloon-expandable prosthesis. In one patient a valve pre-dilatation was performed, while in two cases a post-dilatation was required. One patient received concomitant PCI on the right coronary artery.

Table 2.

Indication for surgery, surgical and follow up data.

	Type of prosthesis	Size	Indication for surgery	Surgical approach	Surgical procedure	Outcome
Patient 1	Self-expandable	23	LV perforation	Sternotomy	Off pump sutured LV repair	Followed up for 59 months. No reportable event during FU period
Patient 2	Self-expandable	34	Aortic annulus rupture	Sternotomy	AVR + aortic annulus repair	Deceased after 32 months. No cardiac death.
Patient 3	Self-expandable	34	Aortic dissection	Sternotomy	AVR + ascending aorta replacement (HCA)	In hospital death
Patient 4	Self-expandable	27	Aortic dissection	Sternotomy	AVR + ascending aorta replacement (HCA)	Followed up for 24 months. No reportable event during FU period
Patient 5	Balloon-expandable	26	RV perforation	Sternotomy	Off pump suturless RV repair	Followed up for 18 months. 1 hospital readmission due to pneumonia
Patient 6	Balloon-expandable	23	RV perforation	Sternotomy	Off pump suturless RV repair	Followed up for 52 months. No reportable event during FU period
Patient 7	Self-expandable	29	Prosthesis embilization	Sternotomy	Embolized prosthesis removal	Followed up for 24 months. No reportable event during FU period
Patient 8	Self-expandable	26	Mitral valve apparatus demage	Sternotomy	AVR + MVR	Followed up for 20 months. No reportable event during FU period
Patient 9	Self-expandable	23	Prosthesis embilization	Sternotomy	AVR	Followed up for 15 months. No reportable event during FU period
Patient 10	Self-expandable	23	Mitral valve apparatus demage	Right minithoracotomy	Mitral valve repair	Followed up for 8 months. No reportable event during FU period

Table 2: LV: left ventricle, RV: right ventricle, AVR: aortic valve replacement, HCA: hypothermic circulatory arrest, MVR: mitral valve repair, FU: follow up.

Type A aortic dissection was reason for ECS in two patients, aortic annulus rupture in one patient, left ventricle (LV) perforation in one patient, right ventricle perforation in two patients, severe mitral regurgitation due to valve apparatus damage in two patients, and prosthesis embolization in two patients.

Surgery was performed through a full sternotomy in nine patients, whereas one patient received a right mini-thoracotomy. Off-pump repair of right or LV perforation was performed in three patients. AVR was completed in two patients; in one of these, aortic annulus repair was required. Hypothermic circulatory arrest was required in two cases of aortic arch and ascending aorta replacement and in one case of embolized prosthesis removal. One patient underwent mitral valve repair for damage to the mitral apparatus via right mini-thoracotomy while another patient required mitral and aortic replacement.

Post-operative outcomes are reported in Table 3. Four patients required post-operative inotropic support (VIS score 10.14 ± 7.02), one patient developed post-procedural AKI (grade III), while one patient had a post-operative minor stroke. New onset LBBB was reported in three patients.

Table 3.

Post-procedural outcomes.

Variables
Intra-operative mortality, n (%)	0
AKI, n (%)	1 (10 %)
Stroke, n (%)	1 (10 %)
New onset LBBB, n (%)	3 (30 %)
New onset AF, n (%)	0
PM implantation, n (%)	0
VIS score, mean (SD)	10.14 ± 7.02
MI, n (%)	0
ICU stay, days, median (IQR)	4,5 (2–7,75)
In-hospital stay, days, median (IQR)	14 (8–22)
In-hospital mortality, n (%)	1 (10 %)

Table 3. AKI: acute kidney injury, IQR: interquartile range,LBBB: left bundle branch block, AF: atrial fibrillation, PM: pace-maker, MI: myocardial infarction, ICU: intensive care unit.

No case of intra-operative mortality, new onset of atrial fibrillation (AF), PM implantation and post-procedural MI were reported.

Median ICU and hospital length-of-stay were 4.5 (2–7.75) and 14 (8–22) days, respectively.

In-hospital mortality occurred in 1 (10 %) case because of multiorgan failure.

Median follow-up period was 24 (16.5–42) months. One (11.1 %) out of nine discharged patients died during follow-up period due to lung malignancy. Only one (11.1 %) case of hospital re-admission, due to pneumonia, has been reported. Detailed follow-up data are reported in Table 2.

4. Discussion

The number of patients with AS needing TAVI is fast-growing, and it is expected to further increase 4- to 10-folds, leading directly to even longer waiting times, hospitalizations, and mortality of patients who could have gained advantage from TAVI [18,19]. This brings the attention on the chance to perform TAVI in hospital without on-site cardiac surgery department. However, despite advances in transcatheter procedures, TAVI still carries the risk for life-threatening complications that may require ECS procedures. In a snapshot of seven years’ experience in two high volume TAVI centers with on-site cardiac surgery and scrubbed cardiac surgeon in Italy, a low, but consistent rate of ECS has been reported (0.74 % of patients); the cardiac surgery team was able to address all the complications, from simple sternotomy in case of cardiac tamponade to complex cardiac surgery procedures for aortic dissection, and the overall in-hospital mortality was reported in 1 out of 10 patients.

Data on 47.546 patients from the Society of Thoracic Surgeons/American College of Cardiology (STS/ACC) Transcatheter Valve Therapy (TVT) registry show a 1.2 % rate of ECS during TAVI procedures. The most frequent causes of cardiac bailout were valve dislodgement (22 %), ventricular perforation (19.9 %), aortic annulus rupture (14.2 %), aortic dissection (8 %) and coronary occlusion (6 %); as expected, the need for ECS was associated with significantly higher in-hospital mortality when compared to patients who did not require surgical conversion (49.6 % vs 3.5 %, p < 0.0001) [20]. Interestingly, authors reported a reduction of surgical bailout incidence during the study period. The decreasing incidence of ECS was probably due to better patient selection, technological improvements in prosthesis and delivery systems and increased operator experience.

Similar data have been reported by the European Registry of Emergent Cardiac Surgery during TAVI (EurECS-TAVI) [21]. Among 27.760 patients who underwent TF-TAVI between 2013 and 2016, 2012 patients required ECS for TAVI complications (0.76 %), with a sensible decline in the reported rate, particularly from 1.07 % in 2013 to 0.73 % in 2016. LV perforation (28.3 %) and aortic annulus rupture (21.2 %) were the leading causes of surgical bailout. ECS was associated with 46 % in-hospital mortality and highest in case of ECS for annulus rupture (62.2 %), followed by coronary obstruction (54.5 %), aortic dissection (52 %), and LV perforation (50.8 %). One-year survival of the 114 discharged patients was 40.4 %.

It is important to note that LV perforation, which can be easily solved if a sternotomy in promptly performed by a scrubbed cardiac surgeon, is associated with a mortality of nearly 50 %.

Walther et al. analyzed intra- and post-procedural complications rate of 15.964 patients who underwent TAVI between 2011 and 2013 [22]. ECS was required in 201 patients (1.3 %). In-hospital mortality was significantly higher in patients experienced surgical bailout when compared to the overall population (42.3 % vs 5.2 %, p < 0.05). The proportion of patients undergoing ECS decreased from 1.6 % in 2011 to 1.1 % in 2013.

The reduction on ECS through several registries has increased the debate to perform TAVI in centers without on-site cardiac surgery. The first study to evaluate outcomes of patients who underwent TAVI in centers without on-site cardiac surgery was published by Eggebrecht and colleagues in 2014 [23]. Comparing 1254 patients who underwent TAVI at centers with on-site cardiac surgery to 178 patients treated at hospitals without on-site cardiac surgery they found no significant differences in terms of major post-procedural complications and 30-day mortality.

In an analysis of the Spanish TAVI registry Rao Gorrido et al. evaluated the safety and feasibility of TAVI at ten hospitals without on-site cardiac surgery but with a reference cardiac surgery center less than 90 km away [24]. Three-hundred-eighty-fours patients were enrolled and ECS was required in only one case (0.3 %) due to LV perforation. Other procedural complications such as cardiac tamponade, valve dislodgement and coronary obstruction were solved without surgical bailout.

In addition to these promising results and the decreasing of ECS incidence, the proponents of performing TAVI in centers without on-site cardiac surgery argue that the increasing of the number of centers authorized to perform TAVI could be helpful to face the increasing demand for TAVI and to reduce mortality during waiting list.

It is demonstrated that the decrease of TAVI complications requiring ECS is strictly related to the growing heart-team expertise [25,26]. Consequently, an expansion of TAVI to peripheral hospitals unable to ensure neither high procedural volumes, and at the beginning of the learning curve, nor the on-site cardiac surgery may lead to a worsening of TAVI outcomes. In a similar scenario the number of deaths due to the increasing number of procedural complications, together with the delay in their treatment, would exceed the mortality during the waiting list [27,28].

Wassef et al. demonstrated that a progressive improvement in clinical outcomes was associated with increasing TAVI experience [25]. Indeed, 30-day mortality was significantly higher in initial, early, and intermediate experience groups in comparison to very high experience groups.

Similarly, in their observational study including 60.538 TAVI procedures, Mao and colleagues found that high volume centers had a lower mortality. Interestingly, centers with high caseloads of both AVR and TAVI were likely to achieve the best outcomes [26].

The 2021 ESC/EACTS guidelines for the management of valvular heart disease recommended that TAVI must be performed in “Heart Valve Centers” that declare their expertise and outcome data, have interventional cardiology and cardiac surgery programs on-site and a structured collaborative Heart Team approach [9].

In 2021 the German Independent Institute for Quality and Efficiency in Health Care stated that in TAVI high volume centers, patients requiring ECS had a higher survival rate, a lower incidence of post-procedural complications and a lower re-hospitalization rate [29]. The importance of operator expertise is underlined by the fact that the latest European [9] and American [30] guidelines recommended performing TAVI in “Heart Valve Centers”.

In our analysis the incidence of TAVI complications requiring surgical bailout was similar to that reported in literature; however, post-procedural mortality was dramatically lower.

These results can be explained considering that all the procedures were performed in an hybrid room with a scrubbed cardiac surgeon, who is also an independent operator for TAVI and always perfectly know the clinical and anatomical case of the patient. The active involvement of the cardiac surgeon during TAVI procedure allows a prompt conversion to open cardiac surgery in case of operative complications requiring surgical bailout.

The main limitation of this study stems from its retrospective nature. An accurate assessment of the efficacy and feasibility of ECS requires further studies evaluating clinical outcomes of converted patients stratified by surgical risk. The second limitation lies in the absence of a control group.

5. Conclusion

The active role of the heart surgeon in TAVI procedures represents a life-saving resource for TAVI centers, and it is mandatory in order to guarantee a low rate of in hospital mortality in patients that experience severe complication during TAVI, especially in younger and low-risk patients.

Data availability statement

Data collected for the study will be made available by the corresponding author upon reasonable request after publication.

CRediT authorship contribution statement

V. Lodo: Writing – original draft, Data curation, Conceptualization. C. Barbero: Writing – review & editing, Writing – original draft, Data curation, Conceptualization. S. Salizzoni: Writing – review & editing, Validation, Supervision, Conceptualization. E. Zingarelli: Writing – original draft. M.La Torre: Validation, Conceptualization. Italiano G. Enrico: Writing – original draft. P. Centofanti: Validation, Conceptualization. M. Rinaldi: Visualization, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.