The Worst Redo Option After TAVR?

For the rising number of patients needing aortic reintervention, transcatheter aortic valve replacement (TAVR) explantation carried the highest mortality risk, two registry analyses affirmed.

In Society of Thoracic Surgeons (STS) data, the number of TAVR patients getting a subsequent surgical valve (SAVR) rose significantly over the decade from 2011 to 2021, as did the number of patients with valve-in-valve TAVR followed by SAVR, whereas the number of SAVR patients getting a second SAVR procedure remained statistically stable.

The highest operative mortality rate occurred in the TAVR-SAVR group, which was significant compared with SAVR-SAVR (17% vs 9%) in both a risk-adjusted model (OR 1.53, P=0.004) and a propensity score matched cohort (OR 1.74, P=0.020).

Surgical explant of valve-in-valve TAVR carried a 12% operative mortality rate that was not significantly greater than that for SAVR-SAVR, reported Shinichi Fukuhara, MD, of the University of Michigan in Ann Arbor, and colleagues in JACC: Cardiovascular Interventions.

Separately, in an analysis of the EXPLANTORREDO-TAVR international registry, redo aortic valve replacement also rose substantially from just three in 2009 to 35 redo TAVRs and 38 TAVR explants in 2021.

Again, TAVR followed by SAVR had higher mortality -- both at 30 days (13.6% vs 3.4%, P<0.001) and at 1 year (32.4% vs 15.4%, P=0.001), albeit driven by the early difference according to a landmark analysis, reported Gilbert Tang, MD, MSc, MBA, of the Mount Sinai Health System in New York City, and colleagues.

An editorial accompanying the two studies acknowledged how little evidence clinicians have had to rely on as the longer life expectancy of today's TAVR patients and expansion of TAVR-in-SAVR procedures has brought treatment of bioprosthetic failure into focus.

Giuseppe Tarantini, MD, PhD, and Tommaso Fabris, MD, both of the University of Padua Medical School, Italy, agreed with Fukuhara and colleagues' call for "[c]areful assessment of concurrent cardiac lesions and future TAVR repeatability during the initial valve selection process."

That's going to require developing more accurate prediction models for life expectancy so that patients who will outlive the durability of a TAVR prosthesis can be determined prospectively, the editorialists noted. Research outside cardiology has suggested that only about 16% of life expectancy predictions were accurate to within a year and still only accurate in about two-thirds of cases when predicting whether a patient would live 10 years or longer.

Tarantini and Fabris offered an algorithm for choice on first procedures:

• Low surgical risk patients under age 75 should get SAVR with a mechanical or biological TAVR-friendly approach (no externally mounted leaflets or sutureless and stentless valves that increase risk of coronary ostial impairment on TAVR) or TAVR as a second choice if redo-TAVR is feasible.
• Intermediate or high risk younger patients and those over age 75 of any risk level should get TAVR as the first choice and SAVR as the second choice for those 75- to 80-year-olds at low risk if redo TAVR is not feasible (such as low coronary takeoff and small aortic root dimensions).
• When TAVR is the first intervention, "redo TAVR with low implantation of a short-frame balloon-expandable valve in a tall-frame supra-annular THV [transcatheter heart valve] results in lower risk for coronary obstruction without significant hemodynamic impairment despite leaflet overhang, particularly if the mode of failure of the first THV was regurgitation."

The STS database study included 1,126 patients who got SAVR after a prior TAVR, 674 who got SAVR after a prior SAVR and TAVR, and 29,306 who had repeat SAVR. The study period covered 2011 through 2021. Patients were excluded if they had a mechanical or other prior non-bioprosthetic valve, endovascular procedures, emergent TAVR valve explantation, unplanned SAVR, prior ventricular assist device or heart transplant, prior root replacement, or non-aortic valve pathology such as trauma.

The regression analysis included about half of patients with available STS predicted mortality risk scores; the propensity matched analysis included 433 pairs of patients.

The international EXPLANTORREDO-TAVR database included 181 TAVR explantation patients and 215 who had a redo-TAVR for transcatheter heart valve failure during a separate admission from the initial TAVR. The database included procedures at 29 centers from May 2009 through February 2022.

Limitations of both studies included lack of data on mechanism of structural valve deterioration (stenosis vs regurgitation), as well as their retrospective design and potential for time selection and learning-curve biases. Moreover, the editorialists noted, "the nonrandomized decision to perform SAVR vs TAVR reintervention represents a potential selection bias."

While the STS database was large, it lacked granularity, Tarantini and Fabris argued, adding that "the limited information related to the type of TAVR explanted limits the evaluation on the most appropriate sequence of redo interventions."

While both studies added an important piece of the puzzle, it will be "important to fill the missing piece of information about the mid- and long-term performance of redo TAVR compared with TAVR-in-SAVR," the editorialists noted. "If these 2 reinterventions have comparable outcomes, TAVR repeatability might be as important as THV durability."