The effects of different types of bone cement intervertebral leakage on stress distribution in endplates of adjacent vertebrae: A finite element study

Abstract

Objective Finite element method was used to clarify the biomechanics effect of cement intervertebral leakage during vertebral augmentation. Present a novel classification of bone cement intervertebral leakage. Analyze the effect of stress changing of bone cement intervertebral leakage on adjacent endplate by finite element method. Methods Based on Churojana's classification method, we redefined diverse kinds of intervertebral leakage: as the Type I (intervertebral-extradiscal leakage), Type II (intradiscal leakage) and Type III (combined leakage). Type II was also been divided into IIa (anterior), IIb (central), IIc(posterior), IId (lateral) and IIe (cross-region) due to the location of the leaked bone cement.All the Type II cases were divided into 1 or 2 two subtypes according to whether the cement had reached the adjacent vertebral endplate. We established 3D reconstruction of volunteer thoracolumbar spine using Mimics 17.0 software, and using Geomagic 2015 to generate L1 vertebral compression fracture model. In the Ansys 17.0 software, we simulated the L1 bone cement leakage into the T12/L1 intervertebral space model. After validating the validity of the model, calculate the solution of the intact model, non-leakage model and various leakage models, the stress distribution of the caudal endplate of T12 was analyzed in neutral, flexion, extension, lateral bending and torsion. Results The maximum stress of inferior endplate of T12 vertebra of intact model is 11.476 MPa, 19.517 MPa, 16.879 MPa, 42.346 MPa, 43.033 MPa, 6.568 MPa, 6.568 MPa in neutral, flexion, extension, left bending, right bending, left rotation, right rotation respectively. For the non-leakage model, the maximal stress of adjacent vertebral endplate was 12.967 MPa (112.99%), 23.134 MPa (118.53%) and 20.403 MPa (120.88%) in neutral, flexion and extension compared to the intact model.No significant increasing can be found in other conditions. Compared to the non-leakage model, the stress of adjacent vertebral endplate is similar when type I leakage occurs. In type II leakage, the IIa1 was 28.506 MPa (123.40%) in the flexion; the IIa2 was 84.791 MPa (366.52%) in the flexion; the IIb2 was 14.138 MPa (122.82%) in the neutral and 27.313 MPa (118.06%) in the flexion; the IIc1 was 19.695 MPa (128.50%) in the extension; the IIc2 was 67.740 MPa (441.97%) in the extension, and the IId2(right) was 123.940 MPa (285.83%) in the right bending. In the left/right rotation motions, the stress values are small, ranging from 5.095-15.585 MPa. Conclusion After vertebral augmentation, the stress of adjacent vertebral endplate increased slightly. Type I leakage did not further increase the stress of adjacent vertebral endplates. Type II leakage increases the stress of adjacent endplates in the direction of leaked cement. Subtype 2 of Type II offer more stress than subtype 1. When the peripheral type of leakage (IIa, IIc and IId) occurred, if the spine flexes in the direction of leakage, then the stress increase of adjacent endplates will increase further. Key words: Cementoplasty; Spinal fractures; Biomechanics