WASHINGTON – A decision-assistance tool will soon be available that is designed to help intermediate-risk patients with severe aortic stenosis and their physicians better compare each patient’s expected outcome from surgical or transcatheter valve replacement based on each patient’s individual clinical and demographic features.
The decision tool will be available as both a web-based calculator and a downloadable app. It is derived from the outcomes of 4,732 patients who underwent surgical aortic valve replacement (SAVR) during 2011-2013 and who were included in the registry maintained by the Society of Thoracic Surgeons (STS), as well as an equal number of closely matched patients who underwent transcatheter aortic valve replacement (TAVR) during 2014-2015 and entered in the Transcatheter Valve (TVT) registry run by STS and the American College of Cardiology. Tool development also used longer-term outcomes data collected through Medicare.
Mitchel L. Zoler/Frontline Medical News
Dr. Howard C. Hermann
After receiving patient-specific data, the decision tool estimates the patient’s short-term and 1-year predicted risks for death and stroke and likelihood of being discharged home, as well as the predicted number of days the patient would remain alive and out of the hospital during the first postprocedural year, J. Matthew Brennan, MD, said at the annual meeting of the American College of Cardiology.
Cardiologists and cardiac surgeons “are desperately looking for something like this” because, currently, the only option is to estimate a patient’s risk after SAVR or TAVR using tools developed only from patients who underwent one of these procedures. The new tool gives clinicians and patients a way to compare the two options for an individual patient in a way that “minimizes the biases,” said Dr. Brennan, an interventional cardiologist at Duke University in Durham, N.C., and principal investigator for development of the website and the tool.
Mitchel L. Zoler/Frontline Medical News
Dr. J. Matthew Brennan
Dr. Brennan and his associates developed the decision assistance tool with funding from the Patient-Center Outcomes Research Institute, and it will be available online at no cost at valveadvice.org. The website was already up and running when Dr. Brennan announced the tool during the meeting, and it currently has patient-centered information about aortic valve disease and the options available for treating it. He expects the decision tool to be posted on the site by April or May.
The data he and his associates used to create the decision tool came from a total of more than 197,000 SAVR patients entered into the STS SAVR database during 2011-2013 and more than 25,000 TAVR patients enrolled in the TVT registry during 2014-2015. They used propensity score matching to identify 4,732 matched patients from each group. The patients averaged 81-82 years old, nearly half were women, and their average STS risk score was 5.5-5.8, which meant that patients fell in an intermediate-risk range by this criterion. Just over three-quarters of the TAVR patients had their procedure done via a transfemoral route.
The analysis showed that, overall, 1-year mortality and stroke rates following each type of procedure were not significantly different, and several subgroup analyses failed to identify any type of patient who fell outside this overall pattern. The TAVR patients had a stroke rate that continued to rise during 12-month follow-up, compared with a much flatter pattern among the SAVR patients. While this did not result in an excess stroke rate, the pattern over time suggested that TAVR patients may not have received optimal anticoagulant treatment during the year following their procedure, Dr. Brennan said. The number of months that patients were alive and not hospitalized was also very similar in the SAVR and TAVR groups.
While the 1-year outcomes were very similar, the periprocedural outcomes showed several statistically significant differences. In-hospital mortality was significantly higher in the SAVR patients at 5%, compared with 3% in the TAVR patients. The SAVR patients also were significantly more likely to develop a need for dialysis and a need for red cells and to have a doubled duration in their ICU stay and in their postprocedural length of hospitalization, compared with TAVR patients. On the other hand, TAVR patients were significantly more likely to need a new pacemaker while hospitalized and had a 10 times higher rate of major vascular complications. The stroke rates were very similar in both arms, Dr. Brennan reported.
According to Dr. Brennan, the most striking difference in hospital outcomes was the discharge destination for patients: 70% of TAVR patients went home after their procedural hospitalization, compared with 41% of the SAVR patients. Discharges home following periprocedural hospitalization were “substantially higher” with TAVR, he said.
Another notable feature of these data was how they contrasted with the 1-year outcomes reported from the German Aortic Valve Registry at the American Heart Association Scientific Sessions in November 2016. In the German registry, 1-year mortality after propensity-score adjustment was about 11% among SAVR patients and about 14% among TAVR patients treated with a transfemoral approach, a statistically significant difference, reported Nicolas Werner, MD, of the Ludwigshafen (Germany) Clinic.
“I think the reason we see a difference in the German data is they weren’t able to remove from their analysis the really high-risk patients” who preferentially underwent TAVR, suggested Dr. Brennan. “We had the ability to match patients who had equipoise for undergoing SAVR or TAVR. That’s why our results are more consistent with the findings from the TAVR clinical trials.”
“One of the most important findings from [Dr. Brennan’s] study is [that] it makes the German Registry results look like the outliers rather than the results from the TAVR clinical trials,” commented Howard C. Hermann, MD, professor of cardiovascular disease and director of the cardiac catheterization laboratories at the University of Pennsylvania in Philadelphia.
The data analysis and development of the valveadvice.org website and decision tool received no commercial support. Dr. Brennan had no disclosures. Dr. Hermann has received honoraria and research support from Edwards Pharmaceuticals, research support from Medtronic, and honoraria and research support from several other companies.
On Twitter @mitchelzoler