A controlled trial of rivaroxaban after transcatheter aortic-valve replacement (GALILEO)
Selected in The New England Journal of Medicine by D. Giacoppo , E. Bahena-López
The GALILEO trial was a randomized, open-label, event-driven, international, multicentre trial comparing two different preventive strategies for thromboembolic complications in patients who underwent successful TAVI and had not established indication for oral anticoagulation.
References
Authors
Dangas GD, Tijssen JGP, Wöhrle J, Søndergaard L, Gilard M, Möllmann H, Makkar RR, Herrmann HC, Giustino G, Baldus S, De Backer O, Guimarães AHC, Gullestad L, Kini A, von Lewinski D, Mack M, Moreno R, Schäfer U, Seeger J, Tchétché D, Thomitzek K, Valgimigli M, Vranckx P, Welsh RC, Wildgoose P, Volkl AA, Zazula A, van Amsterdam RGM, Mehran R, Windecker S; GALILEO Investigators
Reference
N Engl J Med. 2019 Nov 16 [Epub ahead of print]
Published
November 2019
Link
Read the abstractReviewers
Our Comment
This joint review is part of the PCRonline GLOBAL Journal Club Initiative by selected members of the EAPCI/PCR Journal Club and EAPCI young ambassadors, and is based on the underlying idea of “Bringing peers together, exchanging ideas, towards a common standard of care”.
Why this study – the rationale/objective?
Hypoattenuated leaflet thickening, a thin layer of thrombus covering the aortic side, and reduced leaflet motion have been reported by multi-detector computed tomography after transcatheter aortic valve implantation (TAVI) in variable proportions of patients. Although clinical implications of such phenomenon are unclear, recent observational studies have shown a possible causal relationship between subclinical leaflet thrombosis and thromboembolic events.
Current guidelines recommend dual antiplatelet therapy after TAVI, though this advice remains largely empirical since based on observational reports and expert consensus. In other clinical settings new-generation oral anticoagulants have shown some advantages compared with traditional therapy. Whether oral anticoagulation can reduce the risk of subclinical leaflet thrombosis and thromboembolic events after TAVI is unknown.
How was it executed – the methodology?
The Global Study Comparing a Rivaroxaban-Based Antithrombotic Strategy to an Antiplatelet-Based Strategy after Transcatheter Aortic Valve Replacement to Optimize Clinical Outcomes (GALILEO) trial was a randomized, open-label, event-driven, international, multicentre trial comparing two different preventive strategies for thromboembolic complications in patients who underwent successful TAVI and had not established indication for oral anticoagulation. Patients were randomly assigned to rivaroxaban at a dose of 10 mg once daily plus aspirin at a dose of 75-100 mg once daily for 3 months, followed by rivaroxaban monotherapy, or to clopidogrel at a dose of 75 mg once daily plus aspirin at a dose of 75-100 mg once daily for 3 months, followed by aspirin monotherapy. Patients assigned to the antiplatelet therapy-based strategy who were not taking clopidogrel received a single loading dose of ≥300 mg.
The primary efficacy endpoint was the composite of death from any cause or thromboembolic events, including any stroke, myocardial infarction, symptomatic valve thrombosis, systemic embolism, deep-vein thrombosis, or pulmonary embolism. The primary safety outcome was the composite of life-threatening, disabling, or major bleeding events. Secondary endpoints included composite and individual outcomes.
All the endpoints were adjudicated by an independent clinical-events committee. Data analyses were conducted externally and an independent data and safety monitoring board provided oversight by periodically reviewing all reported serious adverse events.
What is the main result?
The trial was early terminated for safety concerns in August 2018, after reaching the 42% of total planned events. A total of 1644 patients underwent randomization at median time of 2 [0-8] days after successful TAVI; 826 patients were allocated in the rivaroxaban plus aspirin group and 818 in the clopidogrel plus aspirin group. Follow-up was complete in 96.8% of patients, vital status was available in 98.0% of patients; and median trial duration was 17 [13-21] months.
In the intention-to-treat analysis, the primary endpoint occurred more frequently in the rivaroxaban-based group than in the clopidogrel-based group (105 vs. 78 events, 12.7% vs. 9.5%; hazard ratio 1.35; 95% confidence interval, 1.01 to 1.81). Results were consistent across prespecified subgroups. All-cause death was significantly higher in the rivaroxaban-based group than in the clopidogrel-based group (64 vs. 38 events, 7.7% vs. 4.6%; hazard ratio 1.69, 95% confidence interval, 1.13 to 2.53). The effect was mainly driven by the non-cardiovascular cause component (29 vs. 11, 3.5% vs. 1.3%; hazard ratio 2.67, 95% confidence interval 1.33 to 5.35). The incidence of stroke was similar between groups (30 vs. 25 events, 3.5% vs. 3.1%; hazard ratio 1.20, 95% confidence interval 0.71 to 2.05), with no differences in the ischemic and haemorrhagic components. The incidences of other thromboembolic endpoints, including symptomatic valve thrombosis, pulmonary embolism, deep vein thrombosis, systemic embolism, and myocardial infarction did not significantly differ between groups.
In the intention-to-treat analysis, the incidence of the composite of life-threatening, disabling, or major bleeding was not significantly different between strategies (46 vs. 31 events, 5.6% vs. 3.8%; hazard ratio 1.50; 95% confidence interval, 0.95 to 2.37), though there was a numerical trend towards an increased risk in the rivaroxaban-based group. The analysis of the individual components of the endpoints revealed no difference in life-threatening and disabling bleeding events between groups, while the incidence of major bleeding according to the Valve Academic Research Consortium in the rivaroxaban-based group was double as compared with the antiplatelet-based group (30 vs. 15 events, 3.6% vs. 1.8%; hazard ratio 2.02; 95% confidence interval, 1.09 to 3.76). Results remained consistent by using other definitions of bleeding.
Critical reading and the relevance for clinical practice
The GALILEO trial indicates that, compared with double antiplatelet therapy for 3 months followed by aspirin monotherapy, the addition of rivaroxaban to aspirin for 3 months followed by rivaroxaban monotherapy not only is not able to reduce the incidence of stroke and other thromboembolic events but is also associated with, harm mainly due to increased mortality and major bleeding.
Although previous reports have revealed substantial proportions of thrombus among the components of debris found in cerebral embolic protection systems after TAVI, in the GALILEO trial the similar incidences of stroke and systemic thromboembolism between strategies likely indicate the presence of alternative and partially unknown mechanisms leading to these complications. In addition, although the results of the GALILEO-4D substudy showed reduced incidences of significant leaflet thickening and motion impairment at 3-month follow-up in patients assigned to the rivaroxaban-based strategy, in the GALILEO trial these advantages did not translate to any apparent clinical benefit in terms of valve function at 3-year follow-up. Indeed, the incidences of symptomatic valve thrombosis were very low and not statistically significant. These findings should promote new investigations to define with higher accuracy the causal relationship between subclinical leaflet thrombosis and accelerated valve dysfunction.
The increased mortality observed in patients assigned to the rivaroxaban-based strategy needs to be considered as related to the higher incidence of major bleeding in the same group. Indeed, the absence of significant differences in terms of life-threatening and disabling bleeding events between groups does not exclude a detrimental effect of major bleeding over time.
A tailored application of a rivaroxaban-based strategy after TAVI according to the individual bleeding propensity finds weak support according to available data. Subgroup analyses from the GALILEO trial might help to identify the patient that could take advantage of oral anticoagulation added to antiplatelet therapy. However, the search for an effective preventive strategy for thromboembolic complications after TAVI should also account for the accumulated evidence with surgical aortic valve replacement that have already shown a more favourable safety profile of biological prostheses compared with mechanical prostheses as a result of differences in anticoagulation requirements.
No comments yet!