Session: 07-07-02: Computational Modeling in Biomedical Applications II

Paper Number: 172805

Biomechanical Evaluation of Bone Graft and Implant Material Choice in a Pelvic Reconstruction Using Finite Element Analysis 

INTRODUCTION

A hemipelvectomy is an invasive surgical procedure in which portion of the pelvis is resected to remove a tumor. This procedure leads to a gap in the pelvic ring, causing a loss of stability for patients. To restore the structural stability of a resected pelvis, surgeons oftentimes perform pelvic reconstruction surgeries. There is a large variability across reconstruction surgeries and hence there is no gold standard technique. This has led to a lack of understanding of the biomechanics of a reconstructed pelvis. In this study, we use finite element modeling to shed light on the biomechanics of a reconstructed pelvis using three different kinds of bone grafts: tibia, femur and fibula. We further extend the study to evaluate structural performance of the different implant materials used to affix the graft.

METHODS

Computed Tomography (CT) imaging data was obtained for a patient that underwent an external hemipelvectomy with sacrectomy from the MD Anderson Cancer Center. The pelvic ring was reconstructed using an ischial-sparing, pedicled tibia and fillet of thigh flap technique. Healthy and reconstructed 3D models were prepared using pre- and post-operative clinical images. Two additional models were created digitally by modifying the tibia model to include a femur and a fibula graft instead. Based on the imaging, cortical and cancellous regions were assigned to the bones with linear elastic material properties obtained from literature. Each model has between 300,000 and 400,000 second order tetrahedral elements. The models were tested for a sitting scenario by fixing the bottom and applying a 500N downward force to simulate the weight.

For the implant analysis, idealized implant and bone geometries were created for the tibia-ischium region. An implant material analysis was conducted using titanium, stainless steel, magnesium, polyether ether ketone (PEEK) and carbon fiber PEEK (CF-PEEK). To simulate the tightening of the screws, a pretension of 750N was applied to each screw. This was followed by a 500N vertical load on the top surface of the tibia, representing the body weight, similar to the graft analysis. Both studies utilized ABAQUS’s static analysis solver.

RESULTS

In each reconstruction, the bone graft shoulders majority of the stresses. The femoral graft has the lowest von-mises stress development, followed by tibia and then the fibula. This result was also validated using a 2D truss model, demonstrating that the area of graft is the most important parameter in determining stress. Given that the femur has the largest area, followed closely by the tibia, their results are comparable, while the fibula has a much higher stress due to its small area.

Each graft was also evaluated for yield. This result highlighted that only the uneven region of the tibia which is in contact with the spine yields. This area is limited in size and is not expected to compromise the overall structural integrity. Since the femur and fibula cases were digital reconstructions, none of them have any uneven surfaces in contact with other bones. Overall, none of the cases are expected to yield.

For the implant analysis, the von-mises stresses in the screws were compared with the yield stress of the material. All metal implants remained well below their yield threshold while minimizing the separation between the joined bones, with stainless steel demonstrating the best performance. PEEK and CF-PEEK implants had higher screw stresses with values crossing the yield threshold, making them unsuitable for this reconstruction.

CONCLUSION

Surgical choices made during pelvic reconstructions can influence structural performance. Using finite element analysis, the femur and stainless steel were identified as the optimal bone graft and implant material, respectively. This methodology can be extended to additional reconstructions and may assist surgeons in making informed decisions during surgical planning for pelvic reconstructions.

Presenting Author: Ritika Menghani Rice University

Presenting Author Biography: Ritika Menghani is a postdoctoral researcher at Rice University focusing on projects that use computational modeling for biomedical applications.

Authors:

Ritika Menghani Rice University
Karthik Tappa MD Anderson Cancer Center
Peiyan Li Rice University
Katelyn Kevorkian MD Anderson Cancer Center
Alexander Mericli MD Anderson Cancer Center
Valerae Lewis MD Anderson Cancer Center
Justin Bird MD Anderson Cancer Center
Raudel Avila Rice University