Leveraging Computational Modeling for Medical Device Premarket Submissions: A Comprehensive Guide

Computational modeling and simulation (CM&S) has become an invaluable tool for medical device manufacturers in the development and regulatory approval process. This guide aims to help manufacturers effectively utilize CM&S to support premarket submissions to the FDA, ensuring compliance with regulatory expectations and maximizing the benefits of this powerful approach.

Why Computational Modeling Matters for Medical Device Submissions

1. Accelerates development: CM&S can reduce the need for physical prototyping and testing, speeding up the innovation process.
2. Reduces costs: By minimizing physical testing requirements, CM&S can significantly lower development expenses.
3. Enhances safety: Computational models allow for extensive virtual testing of device performance and safety across a wide range of scenarios.
4. Improves regulatory submissions: Well-executed CM&S can provide compelling evidence to support device safety and efficacy claims.
5. Facilitates personalized medicine: Patient-specific modeling enables the development of customized medical devices.

Examples of How CM&S Are Being Used at the FDA

Recently the FDA produced a report summarizing how they are evaluating and leverage CM&S for things like premarket and postmarket submissions (among other things) [1].

‍

In 2023, the FDA qualified use of the VICTRE tool [2] as part of the Medical Device Development Tool (MDDT) Program [3]. VICTRE is a comprehensive in-silico (computer-based) breast imaging pipeline developed to assess new medical devices, specifically digital mammography (DM) and digital breast tomosynthesis (DBT). The VICTRE pipeline is designed to replicate clinical trials for evaluating the performance of digital mammography and digital breast tomosynthesis devices. This tool provides a virtual environment to generate breast radiographic images, enabling the evaluation of these imaging devices without the need for actual clinical trials.

‍

The VICTRE tool is a significant advancement in the virtual assessment of breast imaging devices, providing a flexible and efficient alternative to traditional clinical trials.

‍

Step-by-Step Guide to Using Computational Modeling for Premarket Submissions

Step 1: Define the Question of Interest

‍

1. Clearly articulate the specific question, decision, or concern to be addressed using the computational model.
2. Ensure the question is focused and relevant to device safety or efficacy.
3. Avoid overly broad questions; instead, target specific aspects of device performance.

Example: "Is the device resistant to fatigue fracture under anticipated worst-case radial loading conditions?"

‍

Step 2: Establish the Context of Use (COU)

‍

1. Define how the model will be used to answer the question of interest.
2. Specify the type of modeling, key inputs and outputs, and how results will be interpreted.
3. Clarify if other information (e.g., bench testing, animal studies) will be used alongside model results.

Example: "Finite element analysis will be performed to identify worst-case device sizes for fatigue fracture. These devices will then be tested on the bench."

‍

Step 3: Assess Model Risk

‍

1. Evaluate the decision consequence: Consider potential patient harm if an incorrect decision is made based on model results.
2. Determine model influence: Assess how much the model contributes to addressing the question relative to other evidence.
3. Combine decision consequence and model influence to establish overall model risk (e.g., low, medium, high).

Step 4: Identify and Categorize Credibility Evidence

‍

Review the eight categories of credibility evidence outlined in the FDA guidance [4]:

• Code verification results
• Model calibration evidence
• Bench test validation results
• In vivo validation results
• Population-based validation results
• Emergent model behavior
• Model plausibility
• Calculation verification/UQ results using COU simulations

Select appropriate categories based on your model and COU. Plan to gather evidence from multiple categories when possible.

‍

Step 5: Define Credibility Factors and Set Goals

‍

1. For each category of evidence, define relevant credibility factors (e.g., software quality assurance, numerical code verification).
2. Create a gradation of activities for each factor, describing increasing levels of rigor.
3. Set credibility goals for each factor based on the assessed model risk.
4. Develop a plan to achieve these goals.

‍

Step 6: Perform Prospective Adequacy Assessment

‍

1. Evaluate if the planned credibility evidence will be sufficient to support using the model for the COU, given the risk assessment.
2. If inadequate, consider modifying the model, reducing model influence, modifying the COU, or revising the evidence generation plan.
3. Consider submitting a Q-submission to receive FDA feedback on your plan.

‍

Step 7: Generate Credibility Evidence

‍

1. Execute the planned studies and analyses to collect credibility evidence.
2. Ensure proper documentation of all methods, results, and analyses.

‍

Step 8: Conduct Post-Study Adequacy Assessment

‍

1. Re-evaluate achieved credibility levels for each factor.
2. Determine if credibility goals were met.
3. Assess if the totality of evidence supports using the model for the COU, given the risk assessment.
4. Consider proximity of model predictions to safety thresholds and uncertainties in predictions.

‍

Step 9: Prepare the Credibility Assessment Report

‍

1. Compile a comprehensive report following the recommended structure in the FDA guidance.
2. Include an executive summary, background, device description, model details, and full credibility assessment results.
3. Provide a clear rationale for the adequacy of the model credibility evidence.

Key Strategies for Success

1. Early engagement: Consider using the Q-submission process to get FDA feedback on your CM&S plan early in development.
2. Risk-based approach: Tailor the depth and breadth of credibility evidence to the assessed model risk.
3. Transparency: Clearly communicate model limitations and assumptions in your submission.
4. Validation emphasis: Prioritize validation activities that are most relevant to your COU.
5. Complementary evidence: Use CM&S to complement, not replace, other forms of evidence when appropriate.
6. Continuous improvement: Refine and update your model based on new data and validation results throughout the development process.

FDA Expectations:

The FDA has outlined their expectations in a guidance document. Refer to that for more details but we've provided an overview below:

‍

1. Credibility framework: Follow the generalized framework outlined in the guidance for assessing model credibility.
2. Risk assessment: Provide a clear rationale for the assessed model risk based on decision consequence and model influence.
3. Evidence categorization: Properly categorize and justify the types of credibility evidence provided.
4. Credibility factors: Define appropriate credibility factors and goals for each category of evidence.
5. Adequacy assessment: Conduct and document both prospective and post-study adequacy assessments.
6. Calculation verification: Include results demonstrating the numerical accuracy of COU simulations.
7. Validation relevance: Clearly explain the applicability of validation activities to the COU.
8. Uncertainty quantification: Provide uncertainty estimates for model predictions when possible.
9. Limitations disclosure: Explicitly state model limitations and justify why they don't reduce confidence in the COU.
10. Comprehensive reporting: Submit a well-organized Credibility Assessment Report following the recommended structure.

Conclusion

Leveraging computational modeling in medical device premarket submissions can significantly enhance the regulatory process when executed properly. By following this guide and adhering to FDA expectations, manufacturers can maximize the benefits of CM&S while ensuring regulatory compliance. Remember that CM&S is a powerful tool to support, not replace, other forms of evidence in demonstrating device safety and efficacy

References

[1] - https://www.fda.gov/media/163156/download

[2] - https://www.fda.gov/medical-devices/science-and-research-medical-devices/victre-silico-breast-imaging-pipeline

[3] - https://www.fda.gov/medical-devices/medical-device-development-tools-mddt

[4] - https://www.fda.gov/media/154985/download

‍

‍

Image Source: Created with assistance from ChatGPT, powered by OpenAI

‍

_{Disclaimer - https://www.cosmhq.com/disclaimer}