Salvage Cardiac Surgery Following Failed Transcatheter Intervention – Do Current STS Prediction Models Appropriately Estimate Mortality?

J. Malas¹, Q. Chen², D. Tam², A. Peiris³, T. Gunn⁴, D. Emerson⁵, W. Cheng⁴, M. Bowdish⁶, J. Chikwe^7
1Department of Cardiac Surgery, Smidt Heart Institute, Cedars-Sinai Medical Center, Los Angeles, California ²Cedars Sinai Medical Center, Los Angeles, California ³Cedars-Sinai Medical Center, Smidt Heart Institute, Los Angeles, California ⁴Cedars-Sinai Medical Center, Los Angeles, California ⁵Cedars Sinai, Los Angeles, California ⁶Cedars Sinai Medical Center, La Canada, California ⁷Cedars-Sinai, Los Angeles, California

Purpose: The requirement for transcatheter valve and coronary programs to be supported by on-site cardiac surgery is increasingly debated. Limited data describe incidence and outcomes of emergency salvage cardiac surgery following transcatheter therapy, and Society of Thoracic Surgeons (STS) risk models may not accurately predict mortality in this challenging population.

Methods: Consecutive patients undergoing salvage cardiac surgery for complications of transcatheter valve, coronary, or ablation procedures at a high-volume program between January 2008 to June 2023 were identified from a prospective database. Salvage surgery was defined as emergency operation within 24 hours of attempted transcatheter procedure. The primary outcome was survival to hospital discharge, stratified by index transcatheter procedure, use of extracorporeal membrane oxygenation (ECMO) and compared to the preoperative STS predicted risk of mortality (PROM).

Results: Of 58,009 patients undergoing transcatheter therapy (24,197 percutaneous coronary intervention (PCI), 41.7%, and 7,857 transcatheter valve therapy, 13.5%), 36 (0.06%) patients underwent salvage cardiac surgery (Figure 1). Of these, 17 (47.2%) were deemed high/prohibitive surgical risk candidates prior to their index procedure. The most common interventional procedures leading to salvage cardiac surgery were transcatheter aortic valve replacement (TAVR) (n=13) and percutaneous coronary intervention (PCI) (n=12) (Table 1). The most common indication for salvage cardiac surgery following TAVR was aortic root rupture (46.2%, n=6), followed by left ventricular apical rupture (15.4%, n=2) and device malposition/migration (15.4%, n=2). Salvage operation following PCI was most commonly secondary to coronary perforation (25%, n=3), coronary dissection (16.7%, n=2), and coronary occlusion (16.7%, n=2). ECMO was required in 8 patients (22.2%) undergoing salvage procedures. Overall survival to hospital discharge was 66.7% (24/36). In-hospital mortality was 25% (3/12) in those undergoing salvage after PCI, compared to the median STS PROM of 11.7%. In-hospital mortality was 54% (7/13) in those undergoing salvage after TAVR, compared to the median STS PROM of 11.8%. In-hospital mortality was 75% (6/8) in patients requiring ECMO support for salvage cardiac surgery.

Conclusion: Salvage surgery was required in fewer than 1/1000 transcatheter procedures at a high-volume center, which may not represent experience in lower volume programs. Salvage surgery was associated with excess mortality particularly among TAVR patients with aortic root rupture. Current STS prediction models may underestimate the risk in this population.

Identify the source of the funding for this research project: No funding