Log in to MyACC
Outcomes of SAVR After Failed TAVR
Quick Takes
- Paravalvular leak, failed repair, and structural prosthetic deterioration are the most common indications for SAVR after previous TAVR.
- Patients who undergo SAVR after TAVR have worse than expected mortality and morbidity.
Study Questions:
What are the outcomes after surgical aortic valve replacement (SAVR) performed for transcatheter aortic valve replacement (TAVR) failure?
Methods:
This is a retrospective review of patients who underwent SAVR between July 2011 and March 2015 after previous TAVR in the Society of Thoracic Surgeons (STS) Adult Cardiac Surgery Database (ACSD). Patients who had undergone initial TAVR before 2006 or >100 months prior to SAVR were excluded. For 18 patients with missing 30-day mortality data, status was imputed to be alive. Outcomes evaluated included operative mortality, stroke, prolonged ventilation, new renal failure, new-onset atrial fibrillation, new pacemaker placement, and discharge location. STS Predicted Risk of Mortality (PROM) was calculated for SAVR as a “re-operation” (whether redo sternotomy was performed or not), and patients were categorized into Low Risk (STS PROM <4%), Intermediate Risk (STS PROM 4-8%), and High Risk (STS PROM >8%) groups. Observed/Expected (O/E) mortality ratios were calculated and compared using a Wald test.
Results:
Of 233 patients who were identified as having undergone previous TAVR or having had a TAVR valve explanted at the time of surgery, 123 were included in the final cohort. Median age was 77 years, and 38% were female. The more common indications for re-operation were paravalvular leak in 19 patients (15.5%), structural prosthetic deterioration in 14 (11.4%), failed repair in 13 (10.6%), sizing or position issues in 13 (10.6%), and prosthetic valve endocarditis in 12 (9.8%). Indication was Other in 26 (21.1%) or Unknown/Missing in 22 (17.9%). Twenty-one patients were categorized as Low Risk, 30 as Intermediate Risk, and 72 as High Risk for mortality from re-operation.
Operative mortality overall was 17.1% (n = 21), and stratified by risk group was 14.3% in Low Risk patients, 10.0% in Intermediate Risk patients, and 20.8% in High Risk patients. Median cardiopulmonary bypass time was 146 minutes (min) (interquartile range [IQR], 117-198 min), median cross-clamp time was 102 min (IQR, 74-132 min), and median operative time was 321 min (IQR, 253-412 min). Re-operation after SAVR during the index hospitalization (e.g., re-exploration for bleeding) occurred in 17 patients (13.8%). Stroke occurred in four patients (3.3%). Other postoperative morbidities occurred frequently in this cohort (prolonged ventilation in 40.7%, new renal failure in 10.4%, new pacemaker placement in 14.6%). O/E ratios were higher than expected for all risk groups but only significant for the Low Risk group: Low Risk O/E 5.48 (95% confidence interval [CI], 1.17-13.93), Intermediate Risk O/E 1.66 (95% CI, 0.35-4.40), and High Risk O/E 1.16 (95% CI, 0.68-1.79).
Conclusions:
Patients who undergo SAVR after TAVR have longer operative times and experience operative mortality and morbidity that is worse than expected.
Perspective:
Despite multiple limitations of the study, including small sample size, selection bias in terms of surgeon determination as to operative candidacy after failed TAVR, exclusion of patients who might have undergone more complex re-operations after TAVR (e.g., multivalvular surgery, concomitant coronary revascularization, or aortic procedure), missing 30-day status for 15% of the cohort, and an STS Risk Calculator that has not been designed to take into account prior TAVR as a risk variable, this study addresses an important question with the largest series of patients to date and provides an initial “signal” and glimpse into future concerns regarding performing TAVR as the initial procedure in low- and intermediate-risk patients with severe symptomatic aortic stenosis. If mortality and morbidity are higher than expected, perhaps initial SAVR in low- and intermediate-risk patients, followed by valve-in-valve TAVR for the second procedure, will have better outcomes than initial TAVR followed by SAVR. Future studies with larger cohorts will be needed to flesh this out.
Clinical Topics: Arrhythmias and Clinical EP, Cardiac Surgery, Geriatric Cardiology, Invasive Cardiovascular Angiography and Intervention, Valvular Heart Disease, Atrial Fibrillation/Supraventricular Arrhythmias, Aortic Surgery, Cardiac Surgery and Arrhythmias, Cardiac Surgery and VHD, Interventions and Structural Heart Disease
Keywords: Aortic Valve Stenosis, Atrial Fibrillation, Pacemaker, Artificial, Cardiac Surgical Procedures, Cardiopulmonary Bypass, Geriatrics, Heart Valve Diseases, Heart Valve Prosthesis, Renal Insufficiency, Reoperation, Risk Factors, Sternotomy, Transcatheter Aortic Valve Replacement
< Back to Listings
