In silico thrombectomy trials for acute ischemic stroke

Area of Science:

• Biomedical Engineering
• Computational Medicine
• Neurology

Background:

• In silico trials offer a method to accelerate medical device development and clinical trial optimization.
• Acute ischemic stroke treatment involves thrombectomy, a complex procedure with variable outcomes.
• Patient stratification is crucial for improving the efficacy of clinical trials.

Purpose of the Study:

• To demonstrate an in silico trial framework for thrombectomy in acute ischemic stroke.
• To compare treatment outcomes across different patient subpopulations and thrombectomy devices.
• To utilize a novel surrogate thrombectomy model for evaluating treatment success.

Main Methods:

• A device-specific surrogate thrombectomy model (logistic regression) was developed using finite-element model data.
• The surrogate model estimates the probability of successful recanalization.
• The model was integrated into an in silico trial framework and tested with three trial examples: validation, fibrin composition analysis, and device comparison.

Main Results:

• The surrogate model accurately reproduced recanalization rates from the MR CLEAN trial (p=0.6).
• Increased blood cell concentration in thrombi correlated with higher thrombectomy success, aligning with clinical data.
• A newer stent retriever showed improved recanalization in silico, though validation is pending.

Conclusions:

• In silico trials can guide medical device developers on device performance.
• This framework can identify optimal patient populations for clinical trials.
• The use of in silico trials has the potential to significantly reduce development time and costs.