While Rare, Surgical Removal of TAVR Carries Risks: Global EXPLANT-TAVR Registry

Surgical removal of TAVR valves for a variety of reasons is associated with low procedural mortality but fairly high rates of death at 1 year, even when performed at expert centers, according to data from the EXPLANT-TAVR multicenter global registry.

The collaboration’s efforts to provide a look at international explant rates and outcomes has implications for future studies, said Gilbert H.L. Tang, MD (Mount Sinai Health System, New York, NY), during a presentation last week at TCT Connect 2020.

“The findings may guide decision-making in terms of initial TAVR versus SAVR as the first intervention, in terms of patient lifetime management strategy,” he observed.

At 1 year, the mortality rate was 28.6%. Although the researchers did not perform a multivariable analysis due to a low rate of events overall, univariate analysis showed the greatest predictors of all-cause mortality at 30 days and 1 year to be diabetes, cerebrovascular disease, and in-hospital life-threatening bleeding. The most frequent in-hospital complication was new pacemaker at 13.9%, followed by atrial fibrillation (10.1%) and acute renal failure (8.1%). In-hospital mortality was 10.8%. Mean length of hospital stay was 16.1 days. The stroke rate at 1 year was 10.1%.

Commenting following the presentation, Charles J. Davidson, MD (Northwestern Memorial Hospital, Chicago, IL), pointed out that even at 30 days, the mortality rate was almost 11%. Tang said much of that likely is due to in-hospital mortality, which included multiorgan failure and sepsis.

“The 30-day mortality is pretty reflective of data we’ve seen that were recently published,” he added. “I would also like to caution that this mortality rate is based on expert centers that participate in our registry.”

While explantation cases are infrequent and data on patients undergoing the procedure are limited, injury to native aortic anatomy following removal of the valve, as well as increased risk of death, are growing concerns, Tang added. He noted that in a recent published case series of over 1,400 patients, the explant rate was low at 1%, but the in-hospital mortality rate was high at 11.8%. Similarly, a report from the Society of Thoracic Surgeons (STS) database showed an operative mortality rate of 17.1%, and worse outcomes compared with those who underwent SAVR as the initial strategy. Just this month, Tang and colleagues published a look at TAVR explantation rates in the Centers for Medicare and Medicaid Services (CMS) database from 2012 through 2017, showing 1-year mortality approaching 23% among a cohort of over 132,000 patients.

Weighing Risk and Contemporary Relevance

The purpose of EXPLANT-TAVR was to create an international registry to provide more clarity on explantation rates beyond single-center experiences and established national databases. The study included 188 patients treated at 36 centers from January 2010 to June 2020. Participating centers were located in the United States, Canada, Austria, Germany, France, Italy, Spain, and Switzerland.

The mean STS risk at the time of the initial TAVR was 4.8% and 28% of patients were deemed to be at low surgical risk. The median time to surgical explantation was 12.9 months. Among the reasons for explantation were endocarditis (41.4%), structural valve degeneration (23.6%), paravalvular leak (15.9%), prosthesis-patient mismatch (7.3%), and valve migration (3.2%). Slightly less than half of all cases were urgent/emergent. Approximately one in five patients had unfavorable anatomy for redo TAVR. Balloon-expandable devices accounted for slightly more of the procedures than self-expanding/mechanically-expanding devices combined, at 53.4%.

Only 14.4% of patients required root replacement, but nearly half had other concomitant cardiac procedures during the explantation, including mitral repair or replacement, tricuspid repair, and CABG. Procedural mortality was low at 0.9% (two patients).

Davidson zeroed in on the lack of comparison data on redo procedures to help understand the risks and benefits.

“It's difficult because we don't have a surgical comparator group, and it's almost impossible to have that because the threshold for a second operation is much higher than the threshold for a first operation . . . so that unfortunately, does color TAVR in somewhat of a negative light,” he said. Davidson asked if valve-in-valve was considered in any cases, such as paravalvular leaks that may have been amenable to percutaneous repair.

Tang said that redo procedures are an area where evidence is lacking and that the registry investigators are still going through the data to look at the issue. He also cautioned that the long length of the study means contemporary data are mixed with data that are close to a decade old.

“So, I think we may have to break down the more-contemporary data set in terms of how many patients are not feasible for a redo TAVR,” Tang said. Davidson added that a contemporary look should also figure in other factors such as newer valve-stent platforms and declining rates of paravalvular leak, structural valve deterioration, and patient-prosthesis mismatch, when extrapolating risk of explantation to younger, lower-risk populations.