Longitudinal outcomes following international multicentre experience with robotic aortic valve replacement^†

Abstract

Objectives

In an effort to maintain the technical aspects of traditional prosthetic surgical aortic valve replacement (AVR) while reducing invasiveness and facilitate options for concomitant operations, transaxillary lateral mini-thoracotomy endoscopic robotic-assisted aortic valve replacement (RAVR) has been introduced. The present data highlight the contemporary international collaborative experience.

METHODS

All consecutive patients undergoing standardized RAVR across 10 international sites (1/2020–7/2024) were evaluated using a central database with 1 year follow-up.

RESULTS

A total of 300 patients were analysed with a median predicted risk of 1.6% with aortic stenosis in 85.7%, nearly half with bicuspid valves. Biological prostheses were implanted in 220 (73.3%) with a median valve size 23 mm, 10% receiving aortic root enlargement, with 17% of all patients undergoing concomitant procedures. Median cross-clamp 120 min with no conversions to sternotomy. Median length of stay was 5 days, 4.3% with prolonged ventilation, 1.7% renal failure, 1.0% stroke and 8.3% required re-thoracotomy for evacuation of haemothorax. There were two 30-day operative mortalities (0.7%). The new permanent pacemaker rate for the full cohort was 2.6%. Of 163 patients with complete 1-year clinical and echocardiographic follow-up, mean aortic valve gradient was 10 mmHg and all but 2 patients (1.2%) had trace to no prosthetic or paravalvular insufficiency.

CONCLUSIONS

RAVR is safe and effective, providing the reproducible benefits of surgical AVR while affording a less invasive approach that permits the opportunity for concomitant procedures. For low and intermediate risk patients with aortic valve disease, RAVR is a potential reproducible alternative for patients and heart teams.

Open in new tabDownload slide

INTRODUCTION

In the current era of available minimally invasive transcatheter and surgical options for the initial management of symptomatic aortic valve disease, patients and providers continue to seek alternatives to a traditional sternotomy approach to surgical aortic valve replacement (SAVR). As short-term transcatheter aortic valve implantation (TAVI) non-inferiority trials facilitate incursions into lower surgical risk cohorts, debate continues regarding the optimal approach for these patients in the absence of longitudinal TAVI results compared to decades of SAVR outcomes [1–4].

Two common anterior chest wall options exist that facilitate minimally invasive aortic valve surgery, each with different advantages but also select opportunities. Upper partial sternotomy provides excellent visualization for a conventional SAVR with minimal need for cannulation alterations other than perhaps the optional adjunct of percutaneous femoral venous drainage. Right anterior thoracotomy (RAT) also permits SAVR via direct visualization or endoscopic video assistance [5–8]. While the RAT approach remains an excellent option for certain centres, including the ability to perform traditional sutured valves, it has yet to gain widespread use adoption to the common need for sutureless valves [7]. While both remain effective approaches with outcomes comparable to full sternotomy, both require anterior chest wall incisions involving either the sternum, or pectoralis musculature, and the RAT may or may not necessitate division of a rib or the right internal thoracic artery. Neither option readily permits the addition of concomitant intracardiac procedures (e.g. mitral, tricuspid, biatrial surgical ablation), although certain highly experienced centres may be able to navigate.

In an effort to maintain the technical aspects of traditional prosthetic SAVR, facilitate options for concomitant operations and to reduce invasiveness even further, transaxillary lateral mini-thoracotomy endoscopic robotic-assisted aortic valve replacement (RAVR) has been introduced as an alternative [9–11]. The objective of this review was to report the early longitudinal outcomes of RAVR at 10 international sites. We hypothesized that RAVR would provide a safe and effective alternative approach to traditional SAVR that may be safely adopted in a global multicentre experience through standardization of technique and outcome reporting.

METHODS

Procedure

The development, implementation and technical details of RAVR have been described previously [9–12]. To summarize, prior to induction of double-lumen endotracheal anaesthesia, patients receive upper extremity arterial monitoring and intra-thecal injection of 0.1 mg morphine sulfate. A 3 cm transaxillary right thoracotomy is performed in the 4th intercostal space at the level of the anterior axillary line. Cardiopulmonary bypass (CPB) is initiated via peripheral cannulation established through right common femoral artery and vein as well as right internal jugular vein as well as a 5 French distal perfusion catheter in the superficial femoral artery connected to the ipsilateral arterial cannula. An aortic root vent is then placed through the 3-cm working incision followed by a left ventricular vent through the right superior pulmonary vein via a separate chest wall stab incision. The DaVinci Xi robot (Intuitive Surgical, Sunnyvale, CA) is used with the camera port through the working incision (arm 2). Three additional ports include De-Bakey forceps (arm 1), long-tip grasping forceps (arm 3) and scissors/needle driver (arm 4) with patient positioning and port locations nearly identical to those used for robotic mitral valve (MV) surgery. A transthoracic aortic cross-clamp is then applied, and antegrade cardioplegic solution is delivered via the aortic root and/or directly via the coronary ostia in the setting of moderate or greater aortic insufficiency. Under full robotic assistance, a transverse aortotomy at or above the sinotubular junction is extended down to the midpoint of the noncoronary sinus to provide excellent visualization of the aortic valve. The robotic curved scissors and long-tip grasping forceps are utilized in all cases to facilitate the debridement of leaflets and all calcific debris with precise tableside aspiration assistance. Circumferential interrupted 2–0 braided polyester sutures are robotically placed from the ventricular side starting from the left-non commissure and proceeding circumferentially clockwise. Switching to a left-handed suture placement once approaching the right-non commissure facilitates tableside suture management and eliminates potential inadvertent instrument interference with the aortotomy. Once annular suture placement is complete, sizing is performed using conventional SAVR sizers. The sutures are passed through the sewing ring of the prosthesis and navigated through the working incision (Fig. 1). Suture fasters (Core-Knot; LSI Solutions, Victor, NY) facilitate securing the valve in place and the aortotomy is closed utilizing 4–0 polypropylene suture in 2 layers in a standard fashion. All patients receive atrial and ventricular pacing wires. The heart is then reanimated, the cross-clamp is released, the robot undocked, and the patient is weaned from CPB, decannulated and closed.

Patients

All consecutive adult patients (age >18) who underwent RAVR at robotic centres around the world performing 2 or more cases between January 2020 and July 2024 were included for analysis. A total of 300 consecutive patients underwent RAVR at 10 established robotic cardiac programmes, all following the identical approach. These included: Morgantown, USA: 217 (V. B. and L. W.), Barcelona, Spain: 22 (D. P.), Houston, USA: 14 (D. R.), Riyadh, Saudi Arabia: 13 (F. K.), Madison, USA: 11 (G.M.), Taipei, Taiwan: 9 (H.C.), Sao Paulo, Brazil: 6 (R. P.), Prague, Czech Republic: 3 (S.C. and J.V.), Sydney, Australia: 3 (T. Y.) and Boston, USA: 2 (S. M.).

To support data sharing, an international RAVR consortium was established, and data were contributed to a central database that included preoperative characteristics, operative details, postoperative events, as well as 30-day and 1 year echocardiographic and heart failure data. The deidentified multi-institutional collaborative database was created and housed at West Virginia University to prospectively follow all patients undergoing RAVR in strict accordance with institutional and international country-specific management of protected health information including the European General Data Protection Regulation. WVU Health Sciences Institutional Review Board approval was obtained with waiver of consent for analysis of de-identified data (Protocol #1709755537, approval 15 May 2022, reapproved 18 August 2023). Data underlying this article will be shared on reasonable request to the corresponding author in conjunction with journal data availability standards.

Categorical variables are presented as counts and percentages, while continuous variables are shown as mean ± standard deviation (SD) or median [25th, 75th percentiles] based on normality. Baseline characteristics, intra- and postoperative, 30-day, and 1-year outcomes were reported. All analyses were performed using SAS Version 9.4 (SAS Institute, Cary, NC).

RESULTS

Patient characteristics

A total of 300 consecutive patients undergoing RAVR at 10 global robotic cardiac programmes formed the study group. Patients had a median age of 67 years, were predominantly male (67.3%) and had a median Society of Thoracic Surgeons predicted risk of mortality of 1.6%. The majority had severe aortic stenosis (85.7%) with nearly half of patients having bicuspid valves and 44.7% with moderate or worse aortic regurgitation (Table 1).

Outcomes

Biological prostheses were implanted in 220 (73.3%) and mechanical valves in 80 (26.7%) (Table 2). The median valve size was 23 mm. A total of 30 patients (10%) underwent aortic root enlargement procedures to allow placement of a larger valve and 17% of all patients received other concomitant procedures including MV repair or MV replacement, left atrial appendage obliteration (LAAO) with or without biatrial cryothermic Cox Maze, and transaortic septal myectomy (Fig. 2). The median CPB time for all patients inclusive of concomitant procedures was 169 min with cross-clamp 120 min. There were no operative conversions to sternotomy. The postoperative mean aortic valve gradient was 9 mmHg, and no patient had more than trace-mild prosthetic or paravalvular regurgitation (Table 2).

The median length of stay was 5 days. Morbidity included 4.3% with prolonged ventilation, 1.7% renal failure and 1.0% stroke. There were no vascular complications. There were no 30-day valve-related reoperations, but 8.3% required re-thoracotomy for evacuation of haemothorax. One patient required temporary post-cardiotomy extracorporeal membrane oxygenation who was successfully discharged. There were two 30-day operative mortalities (0.7%). The new permanent pacemaker rate for the full cohort was 2.6% (7/300) and 2.3% (5/212) following isolated RAVR ± aortic root enlargement.

At 30-days postoperative follow-up, all patients were New York Heart Association (NYHA) class I-II. Of the 272 (90%) patients with evaluable transthoracic echocardiograms at 30-days, the mean aortic valve gradient was 10 mmHg and only 2 patients had more than trace prosthetic or paravalvular aortic regurgitation (Table 3). As of August 2024, 163 patients had completed 1-year clinical and echocardiographic follow-up with a mean aortic valve gradient of 10 mmHg, all with trace to no prosthetic or paravalvular insufficiency except 2 patients (1.2%) having 2+ aortic prosthetic insufficiency. To date, only 1 patient required valve reoperation at 2.5 years postoperatively for early symptomatic structural valve degeneration in a Trifecta bioprosthesis (Abbott, Santa Clara, CA).

DISCUSSION

The present study reports the safety and efficacy of the international RAVR experience with 3 principal findings. First, since the 1st case in 2020 at West Virginia University, RAVR has successfully propagated to 10 global centres as of August 2024 with numerous others in the process of commencing in late 2024 and early 2025. Second, the collaborative growth of RAVR through shared learning and procedural standardization has permitted reproducible and highly satisfactory early clinical outcomes. Finally, the transaxillary lateral approach to RAVR, essentially identical to that used for robotic mitral surgery, has permitted the successful performance of multiple concomitant procedures while providing longitudinal clinical and echocardiographic outcomes similar to those achieved by conventional mini-anterior or sternotomy SAVR. Given the international promulgation of TAVI into lower risk cohorts, minimally invasive options are increasingly sought by patients and providers and these data support RAVR as a potentially promising option for AVR.

Despite the exponential rise in TAVI, there exist numerous patient cohorts and pathoanatomies that currently remain best served by SAVR such as those of younger age, low risk or with bicuspid disease [9–14]. However, these patients are actively being explored for transcatheter alternatives. It is incumbent upon the surgical community to evolve in kind with reproducible least invasive surgical options that provide equal if not improved outcomes. To achieve this goal, several excellent minimally invasive approaches to SAVR with conventional biological or mechanical prostheses exist that include anterior chest minimally invasive options and, recently, transaxillary direct vision or endoscopic SAVR. Similar to RAVR in concept, the direct or endoscopic-assisted transaxillary approach enables traditional isolated SAVR implantation using shafted instruments with excellent results reported by the Dresden group, among others [15]. The RAVR platform affords a lateral transaxillary approach avoiding anterior skeletal or muscular disturbance while enabling the flexibility to add a multitude of concomitant procedures if determined to be in the best interest of the patient, including double or triple valve and/or biatrial Maze operations [5–9]. This added flexibility permits RAVR to be applied to an increased cohort of patients who may benefit from more than isolated SAVR, with 1-year outcomes that appear favourable and reproducible.

Need for permanent pacemaker implantation following aortic valve therapy, ranging between 11–29% following TAVI and 4–8% following SAVR, has been associated with reduced longitudinal outcomes including survival [16–18]. Following RAVR, despite a median valve size of 23 mm, the pacemaker implant rate for the total cohort was only 2.8% and 2.3% in isolated RAVR. A potential reason for this advantageous result may be the very clear visualization of the annular and subannular anatomy, including the membranous septum, and the ability to precisely place sutures to avoid impingement of related anatomy during prosthetic implantation. Similarly, this may further explain why the incidence of paravalvular leak remained consistently negligible across all RAVR sites.

As collective experience with the RAVR platform continues to evolve, further options have been possible. These have included aortic root enlargement, transaortic septal myectomy and aortic valve repair of primary aortic insufficiency due to degenerative aortic valve leaflet prolapse [19–21]. Traditional SAVR, particularly in low-risk patients, is associated with excellent long term durability and survival [3]. As a potentially safe minimally invasive alternative to traditional SAVR, the current report of the 1st 300 consecutive RAVR cases performed in an international multicentre experience provides initial evidence of its reproducibility and effectiveness at 30 days and 1 year.

This paper has several limitations. We acknowledge the potential limitation initial bias of case selection as programmes commenced RAVR; however, we felt this was appropriate and justified as RAVR was introduced as a novel procedure in each institution with a focus on patient and team safety. This noted, as RAVR becomes routine for centres, the heart team approach has made RAVR an all-comer 1st option for patients of low to intermediate surgical risk instead of TAVI in some centres. Further acknowledging that this report represents early multicentre results, care was taken to not directly compare RAVR outcomes with SAVR or TAVR or other minimally invasive approaches as the aim was merely to note observations of this experience in the context of others. Data reporting was voluntary for each centre including all consecutive patients undergoing RAVR with >95% complete data but cost data were not available. It deserves acknowledgement that while robotic use is rapidly growing, its full adoption remains slowed due to centre-specific and region-specific limitations of administrative and clinical support that we hope to be lessoned with recent global industry reinvestments in robotic cardiac surgery.

The present multicentre international experience highlights that in centres with established robotic experience, RAVR is reproducible and safe with excellent early results. As more centres initiate RAVR programmes, a dedication to reproducible quality patient care remains paramount as this therapy potentially becomes a more readily available option for heart team decision-making and for patients with symptomatic aortic valve disease.

FUNDING

This work was supported by National Insitute of Health National Heart Lung Blood Institute grant # 2UM1 HL088925 12 (to V.B. and J.H.M.).

Conflict of interest: Dr Serguei Melnitchouk discloses consulting fees from Medtronic. Dr Danny Ramzy discloses nonfinancial support from Edwards Lifesciences, Medtronic and Abbott. Dr Štěpán Černý discloses consulting fees from Intuitive Surgical. Dr Robert L. Smith discloses consulting fees from Edwards Lifesciences, Abbott, Medtronic and Artivion. No other authors have relevant disclosures.

DATA AVAILABILITY

The data underlying this article will be shared on reasonable request to the corresponding author.

Author contributions

Lawrence M. Wei: Conceptualization; Data curation; Investigation; Project administration; Writing—original draft; Writing—review & editing. Daniel Pereda: Data curation; Investigation; Writing—review & editing. Danny Ramzy: Data curation; Investigation; Writing—review & editing. Feras H. Khaliel: Data curation; Investigation; Writing—review & editing. Ghulam Murtaza: Data curation; Investigation; Writing—review & editing. J. Hunter Mehaffey: Data curation; Formal analysis; Investigation; Methodology; Validation; Writing—original draft; Writing—review & editing. Nai-Hsin Chi: Data curation; Investigation; Writing—review & editing. Robinson Poffo: Data curation; Investigation; Writing—review & editing. Štěpán Černý: Data curation; Investigation; Writing—review & editing. Jan Vojáček: Data curation; Investigation; Writing—review & editing. Tristan D. Yan: Data curation; Investigation; Writing—review & editing. Serguei Melnitchouk: Data curation; Investigation; Writing—review & editing. Alberto C. Weber: Data curation; Investigation; Writing—review & editing. Robert L. Smith: Data curation; Investigation; Writing—review & editing. Goya V. Raikar: Data curation; Investigation; Writing—review & editing. Ali Darehzereshki: Data curation; Investigation; Writing—review & editing. Arnar Geirsson: Investigation; Writing—review & editing. Arman Arghami: Investigation; Writing—review & editing. Jose L. Navia: Investigation; Writing—review & editing. Johannes Bonatti: Investigation; Writing—review & editing. Vinay Badhwar: Conceptualization; Investigation; Project administration; Supervision; Validation; Writing—original draft; Writing—review & editing

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Mateo Marin-Cuartas, Edoardo Zancanaro and the other anonymous reviewers for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• CPB

  Cardiopulmonary bypass

 
• LAAO

  Left atrial appendage obliteration

 
• NYHA

  New York Heart Association

 
• RAT

  Right anterior thoracotomy

 
• RAVR

  Robotic-assisted aortic valve replacement

 
• SAVR

  Surgical aortic valve replacement

 
• TAVI

  Transcatheter aortic valve implantation

Author notes