Effect of cerebral embolic protection devices on TAVI outcomes: a network meta-analysis

Background Cerebral embolic protection devices (CEPD) have been evaluated in multiple randomized trials during transcatheter aortic valve implantation (TAVI). Methods A systematic literature search including PubMed, Scopus and Cochrane databases was conducted to identify randomized controlled studies (RCTs) comparing CEPDs in patients undergoing TAVI. A random-effects network meta-analysis (NMA) was performed within a frequentist framework. The primary outcome was periprocedural stroke up to 30 days. Secondary outcomes included all-cause mortality, disabling stroke, non-disabling stroke, volume and number of ischemic lesions, bleeding, vascular complications and acute kidney injury (AKI). Network diagrams were constructed for each outcome to depict direct comparisons among interventions in the included RCTs. Treatment ranking probabilities for the different CEPDs were estimated using P-scores, which reflect the relative performance of each intervention. Results A total of 9 RCT studies encompassing 23,072 patients (CEPD n=5,813, control n=5,638) were included in the NMA. No cerebral embolic protection device significantly reduced mortality, stroke, acute kidney injury, bleeding, or vascular complications compared with no-CEPD group, except ischemic cerebral lesions. Claret Montage dual filter system significantly reduced both total cerebral lesion volume (mean difference [MD] -370.40 [-500.48; -240.32]) and the number of lesions (MD -4.30, 95% CI -8.31; –0.29) versus no-CEPD group. According to p-scores, the Sentinel device ranked highest among all evaluated treatments for stroke-related outcomes, including overall stroke, disabling and non-disabling stroke, as well as for bleeding and vascular complications. However, Sentinel was not associated with a statistically significant reduction in the risk of stroke (RR 0.88, 95% CI 0.69–1.12) or mortality (RR 0.86, 95% CI 0.53–1.39). Conclusions CEPDs during TAVI were not associated with a significant reduction in periprocedural stroke. The relative efficacy of alternative devices remains uncertain due to the limited data. Future trials should focus on better risk stratification to identify TAVI patients with a higher baseline risk of stroke, as well as on the development of next-generation devices.