ABC Models and the FDA
What follow is an excerpt from a presentation given at an international workshop on cardiovascular device standards. It is a summary comparison of the three types of models used in the device approval process: Animal, Bench and Computational. We are grateful to Tina Morrison, Deputy Director, Division of Applied Mechanics at FDA, friend and fellow ASTM and ASME committee member for the content in this article.
ANIMAL MODELS
Advantages
– Anatomical and physiological similarities may enable some short term safety evaluation. E.g., Deployment inaccuracies
– Importantly, they provide histological information that is not readily available in humans
Disadvantages
– Anatomical and physiological differences limit utility of some pre‐clinical testing
– Modes of deformation (e.g. SFA) are not the same
– Healing time course is accelerated compared to humans
– Inferences derived may be species‐dependent
BENCH‐TOP MODELS
Advantages
– Characterize a specific mechanical function of the device that may be difficult to examine in vivo or with a computer model
– Deform the device under 'worst‐case' conditions that may not be readily reproduced in the clinical setting
– Understand mechanical limitations
– Test the 'actual' manufactured device, with material inclusions, surface treatments, residual stresses, etc.
– Evaluate the initiation and nature of the fracture and wear using imaging techniques
Disadvantages
– Simple loading modes and amplitudes may not replicate the in vivo biomechanical environment
– No tissue or blood interaction with the device
– Limited or no incorporation of disease state (e.g., plaque or lesions)
– Device not responsive to biological reactions or vessel adaptations
– Limited knowledge gained if the device does not fail
– Test outcomes may not predict clinical performance
COMPUTATIONAL MODELS
Advantages
– Examine the structural performance of an entire device family or generations of devices under various loading conditions (whether they're physiologic or not)
– Evaluate the stress/strain distributions of a single or combined loading modes (e.g., pulsatile and bending)
– Investigate the 'hot spots' or potential failure zones on the device
– Investigate the durability of the device under a variety of loading scenarios (safety factors!)
Disadvantages
– Not all factors and scales can be added to the model (yet)
– Physiologically relevant boundary conditions and input data are not always known
– Device cannot respond to biological reactions and adaptations
– Cannot accurately model complex time‐dependent materials
– Lack robust validation for biomedical applications
– May be computationally expensive to examine complex loading and boundary conditions of an entire device family
Harness the Power of ABC Models
Many issues arise (e.g., ethical) if we were to evaluate all devices in randomized controlled, double‐blinded studies.
Therefore, the Division of Cardiovascular Devices at FDA encourages the use of the ABC Models (Animal, Bench, and Computer) to evaluate a device!
For most types of devices, they have learned much from the clinical studies to influence the nonclinical evaluation:
– Gain some safety information before the clinical evaluation;
– Therefore, single‐arm, relatively low number of patients is acceptable.
Ideally, a comprehensive evaluation plan should utilize the advantages of the ABC models to demonstrate safety and performance of the device.
The Bridge Between Bench‐Top to Bedside is Bi‐Directional at the FDA
– Clinical lessons greatly influence design and evaluation.
– The ABC models yield safety data (especially for new or modified devices) before the clinical evaluation.
– Understanding the Anatomical and Physiological / Pathophysiological variations are crucial for design and evaluation.
– Boundary conditions for the ABC models can be further advanced by clinical data and knowledge.
CONCLUSION
Animal, Bench and Computational models can improve medical device design and increase access to safe and effective therapies for patients.
Appropriately balanced and clinically relevant conditions and the development of new tests and methodologies can push the evaluation of devices forward.
Computational Modeling is a powerful tool for Regulatory Decision Making. Download our presentation given at an ASTM workshop on medical device standards about ABC Models: A Vision for Modeling and Simulation in Implantable Medical Device Design and Development: