Thrombotic vs Bleeding Risks Post-TAVI: Insights From GALILEO

PARIS, France—In patients who undergo a successful TAVI procedure, subsequent thromboembolic events are less frequent but carry a higher risk of death compared with bleeding, according to new subanalysis of the GALILEO trial.

While the PCI patient population has been thoroughly studied in terms of the timing and effect of postdischarge ischemic and bleeding complications, this kind of research has not yet been done on those who undergo TAVI, said Gennaro Giustino, MD (Mount Sinai Hospital, New York, NY), who presented the findings during EuroPCR 2022. Additionally, “long-term predictors of survival after successful uncomplicated TAVR procedure remain unclear. Age is usually seen as the major predictor of survival,” he told TCTMD.

For the study, Giustino and colleagues used data from the GALILEO trial (mean age 80 years; mean STS score 4) to look at the timing of subsequent thromboembolic and bleeding events and their effects on mortality. Primary results from that trial have previously shown that rivaroxaban-treated TAVI patients had increased risks of all-cause mortality, thromboembolic events, and bleeding compared with those on antiplatelet therapy.

Overall mortality from the trial was 9.2% over 2 years—a 5.5% rate of CV death—with a median time to death of 248 days. Older age over 85 years (adjusted HR 2.04; 95% CI 1.38-3.03) and male sex (adjusted HR 2.26; 95% CI 1.48-3.48) were the strongest baseline predictors of mortality, followed by hemoglobin less than 10 g/dL, chronic obstructive pulmonary disease (COPD), PAD, eGFR < 45 mL/min/1.73m², and NYHA class III or IV.

The median time to first thromboembolic event was 151 days, with stroke being the most common type of event followed by MI. Bleeding events happened sooner, at a median of 66 days post-TAVI, with BARC types 2 and 1 being most frequent. However, the time to death was shorter following a thromboembolic event than after a BARC 2 or 3 bleeding event (median 36 vs 178 days), even after excluding thromboembolic events that resulted in death within 7 days.

All categories of BARC bleeding except type 1 were associated with mortality, but the relationship between thromboembolic events and death was substantially greater (adjusted HR 8.41; 95% CI 5.10-13.87)—almost double the risk compared to BARC 3 bleeding (adjusted HR 4.34; 95% CI 2.31-8.15), for example. Notably, age (adjusted HR 1.04; 95% CI 1.01-1.08), male sex (adjusted HR 2.23; 95% CI 1.44-3.46), and COPD (adjusted HR 2.11; 95% CI 1.30-3.41) were each associated with mortality but to a lesser degree.

“Balancing the risk of thromboembolic events versus bleeding complications even after a successful TAVR remains challenging and requires further evaluation to further understand the risk of balancing thrombosis and bleeding in these patients,” Giustino concluded. “Our routine aggressive antithrombotic strategies are not being shown to be beneficial versus a less aggressive strategy in a number of trials, so we need to better understand what is actually happening.”

Because the GALILEO trial included all-comers, Giustino said the findings can be generalized to any patient undergoing TAVI at any surgical risk without contraindications to anticoagulation. That said, because of the mean patient age of 80 years, he continued, “obviously [in] younger patients with lower surgical risk, the incidence of these events is going be lower. Therefore, the risk/benefit profile of an antithrombotic therapy would be different from one in an intermediate to high risk population. This needs to be put into perspective.”

As for why minor bleeding would have a prognostic effect, Giustino said this is “similar to what has been also shown in the PCI literature. These mild bleeds are probably a marker of the frailty or the sickness and the comorbidity burden of the patients.”

Following the presentation, session co-moderator Franz-Josef Neumann, MD (University Heart Center Freiburg-Bad Krozingen, Germany), said what jumps out to him is the big difference in risks posed by the different types of events, as compared with the post-PCI population. “In the PCI setting, the intake of thromboembolic events and the bleeding events are almost equal, but in your data set the impact of the thromboembolic events [was] . . . two times higher than the impact of bleeding events,” he noted.

“This is the major finding of the analysis,” Giustino confirmed. “It's different from the PCI setting where there are multiple studies showing that the prognostic effect of an MI is probably similar to the prognostic effect of a major bleed. In this setting, we need to remember that most of the thrombotic events are strokes, which are inherently associated with higher morbidity and mortality.”

Additionally, several trials evaluating antithrombotic strategies after TAVI, including GALILEO, have now shown that more potent antithrombotic therapies do not reduce thromboembolic complications but rather increase the risk for bleeding.

“However, this analysis emphasizes that thromboembolic events after a successful TAVR are still pretty frequent and are associated with significant impact on morbidity and mortality,” Giustino said. “So therefore, it's true that probably routine aggressive antithrombotic therapy is not warranted in most people that require TAVR, but this needs to be evaluated with further studies to understand the mechanisms of thromboembolic events longitudinally after a valve replacement.”

Specifically, he said he’d like to see more work looking at the relationship between anticoagulation and subclinical leaflet thrombosis and the clinical effect. “The other thing is understanding if it does have an impact on TAVR durability,” Giustino said. “What I would like to see is retesting antithrombotic therapies in lower-risk patients that may have a better risk/benefit profile to tolerate a more aggressive antithrombotic approach, which does not need to be indefinite. It can be also for a short period of time. . . . This is similar to what surgeons have been doing for many years.”

Lastly, Giustino said he would like to see “a better characterization of what are the mechanisms of these events after TAVR. . . . For example, understanding the incidence of subclinical arrhythmias or what are the mechanisms of myocardial infarctions in these patients.”