Deep Learning in Prediction of Late Major Bleeding After Transcatheter Aortic Valve Replacement

Abstract Background About 40% of patients undergoing transcatheter aortic valve replacement (TAVR) have a history of atrial fibrillation (AF) and an additional 10% develop AF after TAVR. However, there is paucity of data regarding the optimal antithrombotic regimen following TAVR in patients with a clinical indication for oral anticoagulants (OAC). Purpose To compare the prognostic impact of OAC plus at least one antiplatelet agent (APT) versus OAC therapy alone in patients undergoing TAVR. Methods We systematically searched the literature for studies evaluating the comparative efficacy and safety of OAC + APT versus OAC alone in TAVR. Random-effect meta-analysis was performed comparing clinical outcomes between the two groups. All-cause mortality and cardiovascular mortality were the efficacy outcomes. Stroke and major bleeding, defined as Bleeding Academic Research Consortium bleeding types 3 to 5, constituted the safety outcome. Results Overall, 398 titles and abstracts were identified through database searching. Four observational studies were selected, for a total of 1929 patients. After a median follow-up of 18.5 months (IQR 11.3–29.3), OAC + APT increased major bleeding events compared to OAC alone (OR=1.79; 95% CI 1.21–2.66; P=0.004) with no difference in stroke (OR 01.02; 95% CI 0.52–2.01; P=0.95), all-cause mortality (OR=1.07; 95% CI 0.78–1.47; P=0.66) and cardiovascular mortality (OR=1.08; 95% CI 0.79–1.47; P=0.62). Conclusion A combination strategy of OAC + APT provides increased risk of bleeding compared to OAC therapy alone in patients undergoing TAVR with similar outcomes in terms of stroke, all-cause mortality and cardiovascular mortality; therefore, when feasible, it should be advised not to add APT on top of OAC therapy in patients without other clinical indications for APT treatment. Funding Acknowledgement Type of funding source: None

Download Full-text

Deep learning-based prediction of early cerebrovascular events after transcatheter aortic valve replacement

Scientific Reports ◽

10.1038/s41598-021-98265-5 ◽

2021 ◽

Vol 11 (1) ◽

Author(s):

Taishi Okuno ◽

Pavel Overtchouk ◽

Masahiko Asami ◽

Daijiro Tomii ◽

Stefan Stortecky ◽

...

Keyword(s):

Deep Learning ◽

Aortic Valve ◽

Aortic Valve Replacement ◽

Predictive Model ◽

Valve Replacement ◽

Transcatheter Aortic Valve Replacement ◽

Feature Reduction ◽

Risk Of Death ◽

Transcatheter Aortic Valve ◽

Cerebrovascular Events

AbstractCerebrovascular events (CVE) are among the most feared complications of transcatheter aortic valve replacement (TAVR). CVE appear difficult to predict due to their multifactorial origin incompletely explained by clinical predictors. We aimed to build a deep learning-based predictive tool for TAVR-related CVE. Integrated clinical and imaging characteristics from consecutive patients enrolled into a prospective TAVR registry were analysed. CVE comprised any strokes and transient ischemic attacks. Predictive variables were selected by recursive feature reduction to train an autoencoder predictive model. Area under the curve (AUC) represented the model’s performance to predict 30-day CVE. Among 2279 patients included between 2007 and 2019, both clinical and imaging data were available in 1492 patients. Median age was 83 years and STS score was 4.6%. Acute (< 24 h) and subacute (day 2–30) CVE occurred in 19 (1.3%) and 36 (2.4%) patients, respectively. The occurrence of CVE was associated with an increased risk of death (HR [95% CI] 2.62 [1.82–3.78]). The constructed predictive model uses less than 107 clinical and imaging variables and has an AUC of 0.79 (0.65–0.93). TAVR-related CVE can be predicted using a deep learning-based predictive algorithm. The model is implemented online for broad usage.

Download Full-text

Deep Learning-based Prediction of Early Cerebrovascular Events after Transcatheter Aortic Valve Replacement

10.1101/2021.09.10.21263380 ◽

2021 ◽

Author(s):

Taishi Okuno ◽

Pavel Overtchouk ◽

Masahiko Asami ◽

Daijiro Tomii ◽

Stefan Stortecky ◽

...

Keyword(s):

Deep Learning ◽

Aortic Valve ◽

Aortic Valve Replacement ◽

Predictive Model ◽

Valve Replacement ◽

Transcatheter Aortic Valve Replacement ◽

Feature Reduction ◽

Risk Of Death ◽

Transcatheter Aortic Valve ◽

Cerebrovascular Events

Background: Cerebrovascular events (CVE) are one of the most feared complications of transcatheter aortic valve replacement (TAVR). CVE appear difficult to predict due to their multifactorial origin incompletely explained by clinical predictors. We aimed to build a deep learning-based predictive tool for TAVR-related CVE. Methods: Integrated clinical and imaging characteristics from consecutive patients enrolled into a prospective TAVR registry were analysed. CVE comprised any strokes and transient ischemic attacks. Predictive variables were selected by recursive feature reduction to train an autoencoder predictive model. Area under the curve (AUC) represented the model`s performance to predict 30-day CVE. Results: Among 2,279 patients included between 2007 and 2019, both clinical and imaging data were available in 1,492 patients. Median age was 83 years and STS score was 4.6%. Acute (<24 hours) and subacute (day 2-30) CVE occurred in 19 (1.3%) and 36 (2.4%) patients, respectively. The occurrence of CVE was associated with an increased risk of death (HR [95%CI]: 2.62 [1.82-3.78]). The constructed predictive model uses less than 107 clinical and imaging variables and has an AUC of 0.79 (0.65-0.93). Conclusions: TAVR-related CVE can be estimated using a deep learning-based predictive algorithm. The model was implemented online for broad usage. (https://www.welcome.alviss.ai/#/cvecalculator).

Download Full-text