There is a strong association of changes in device and procedural factors with changes in long-term mortality, likely driven by their association with short-term complications of transcatheter aortic valve replacement (TAVR), a new registry analysis shows.

Presenting at the Transcatheter Cardiovascular Therapeutics (TCT) 2022 conference in Boston, Suzanne V. Arnold, MD, MHA, from the University of Missouri-Kansas City, explained why patient outcomes have improved since the commercial approval of TAVR over a decade ago.

“There have been a lot of moving parts during that time period,” she said. “So patient risk has decreased as TAVR was introduced to both intermediate- and then low-risk patients.

“But also, the devices and periprocedural care have also markedly changed over that time period.”

Study essentials

The study objective was to dissect the contribution of changes in patient risk and procedures to improvements in TAVR outcomes over time.

Here, 161,196 patients, who underwent TAVR at 596 U.S. hospitals from 2011-2018, were included.

As part of the Society of Thoracic Surgeons/American College of Cardiology Transcatheter Valve Therapy (STS/ACC TVT) Registry, one of the primary outcomes of the study was 30-day death rates.

The study team also looked at 30-day composite adverse events, which included death, stroke, stage 3 acute kidney injury (AKI), major or life-threatening bleeding, moderate or severe paravalvular leak (PVL), and 1-year death rates.

The study identified a number of mediator clusters – groups of different patient procedural factors that are adjusted for as whole collections of effects to understand potential improvement mediators over time.

The mediator clusters included demographics, non-cardiovascular comorbidities, cardiovascular comorbidities, device factors and non-device-related procedural factors.

Main findings

Discussing the results of outcomes over time, the study team noted a 30-day death rate of 6.7% in 2011-2012 decreasing to 2.4% in 2018.

Similarly, the 1-year death rate was 19.9% in 2011-2012 and decreased to 10.1% in 2018.

Improvements in 30-day complications were also discussed, specifically the 30-day composite outcome (25.3% in 2011-2012 vs. 10.5% in 2018), 30-day stroke (2.4% in 2014 vs. 1.5% in 2018) and 30-day bleed (12.2% in 2011-2012 vs. 5.4% in 2018).

Other complications looked at in the study included 30-day AKI stage 3 (2.4% in 2011-2012 vs. 0.8% in 2018) and 30-day moderate/severe paravalvular leak (PVL; 7.7% in 2011-2012 vs. 1.5% in 2018).

“The composite was primarily driven by improvement in reductions in 30 days major or life-threatening bleeding, and reduction in PVL,” Arnold commented.

30-day death significance

Delving deeper into the significance of the 30-day death result, Arnold explained that the odds ratio (OR) of 0.82 (95% confidence interval [CI], 0.80-0.84] – a value less than 1 – indicated that death was less likely for later procedures.

For the 30-day composite adverse events, adjusted for non-device procedural factors, an OR of 0.96; (95% CI, 0.95-0.97) was determined.

“The greatest movement in that shift in odds ratio towards 1 is with adjustment for device and procedural factors,” Arnold pointed out.

“This indicates that the majority of improvement that we have been seeing in short-term mortality over time is driven by these procedural factors when we look at their big composite adverse events.

“We see a similar pattern in that device and procedural factors play a significant role, but there remains some unexplained association of time with outcomes,” she added.

“And so we have attributed this as to a learning curve or refinement and operator skill. So it is improvements that cannot be explained by other patient and procedural factors.”

The pattern was slightly different for 1-year death, with an OR of 0.98 (95% CI, 0.97-0.99), adjusted for non-device procedural factors.

“As we would expect, differences and changes in patient risks had shown an association with the improvement in outcomes,” Arnold said. “But devices, nonprocedural and non-device procedural factors also played a significant role.”

Concluding thoughts

Arnold concluded her presentation by stating that although U.S. TAVR patients have become younger and healthier over time, it appeared that changes in devices and procedural care were primarily driving the observed reduction in short-term outcomes.

This was in contrast to changes in patient comorbidities that were associated with reductions in long-term mortality. Learning curve or within an operator skill was also associated with reductions in complications.

“We believe our findings emphasize the importance of device iterations and non-device procedural factors in improving short-term mortality and complications with TAVR,” she concluded.

“This may have important implications for future device innovation, particularly as we move to the treatment of other forms of valvular heart disease.”

Image Credit: Jason Wermers/CRTonline.org