and Methods for the Biomechanical Assessment of Osseointegrated Transfemoral Prostheses: a Systematic Review

The amputee population according to the World-Health-Organization is about 40 million. However, there is a high abandon rate of socket prostheses for the lower limb (25%-57%). The direct connection between the external prosthesis and the patient's bone makes osseointegrated prostheses for transfemoral amputees advantageous (e.g., improvement of the motor control) compared to socket prostheses, which are currently the gold standard. However, similarly to other uncemented prostheses, the osseointegrated ones are at risk of aseptic loosening and adverse bone remodelling caused by stress-shielding. The preclinical assessment of these prostheses has already been evaluated using different methods which did not provide unanimous and comparable evidence. To compare data from different investigations, a clear and detailed overview of the methods used to assess the performance is necessary. In this review 17 studies investigating the primary stability, stress shielding and stress concentration of osseointegrated transfemoral prostheses are examined. Primary stability consists in the biomechanical stability upon implant insertion. Primary stability is assessed measuring extraction force (either with a pull-out or a push-out test) and micromotion at the interface between the implant and the host bone with LVDT ( test) or numerical models. Stress-shielding causes adaptive changes in the bone density around metal implants, and thus in the bone strength and stiffness. Stress-shielding is assessed with strain gauges or numerical models measuring the load transfer and the strain distribution on the surface of the femur, and between the implant and the bone respectively. Stress concentration can lead to the formation of cracks inside the bone, resulting in fractures. The stress concentration is assessed measuring the load transfer and the strain energy density at the interface between the implant and the bone, using numerical models. As a result, a global view and consensus about the methods are missing from all these tests. Indeed, different setup and loading scenario were used in the test, while different model parameters (e.g., bone properties) were used in the numerical models. Once the preclinical assessment method is established, it would be important to define thresholds and acceptance criteria for each of the possible failure scenarios investigated.