Abstract

Autologous osteochondral grafting has demonstrated positive outcomes for treating articular cartilage defects by replacing the damaged region with a cylindrical graft consisting of bone with a layer of cartilage, taken from a non-loadbearing region of the knee. Despite positive clinical use, factors that cause graft subsidence or poor integration are relatively unknown. The aim of this study was to develop finite element (FE) models of osteochondral grafts within a tibiofemoral joint and to investigate parameters affecting osteochondral graft stability.

Initial experimental tests on cadaveric femurs were performed to calibrate the bone properties and graft-bone frictional forces for use in corresponding FE models, generated from µCT scan data. The effects of cartilage defects and osteochondral graft repair were measured by examining contact pressure changes using in vitro tests on a single cadaveric human tibiofemoral joint. Six defects were created in the femoral condyles which were subsequently treated with osteochondral autografts or metal pins. Matching µCT scan-based FE models were created, and the contact patches were compared. Sensitivity to graft bone properties was investigated.

The bone material properties and graft-bone frictional forces were successfully calibrated from the initial tests with good resulting levels of agreement (CCC=0.87). The tibiofemoral joint experiment provided a range of cases to model. These cases were well captured experimentally and represented accurately in the FE models. Graft properties relative to host bone had large effects on immediate graft stability despite limited changes to resultant cartilage contact pressure.

Model confidence was built through extensive validation and sensitivity testing, and demonstrated that specimen-specific properties were required to accurately represent graft behaviour. The results indicate that graft bone properties affect the immediate stability, which is important for the selection of allografts and design of future synthetic grafts.

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Osteoarthritis (OA) affects over 8.75 million people in the UK creating the need for early stage interventions. Osteochondral (OC) grafting has been used to repair full thickness lesions but the efficacy of this therapy is questionable.

As a first step in developing a testing framework able to predict the potential and suitability of OC grafts for repair, here, we present experimental data to be used in informing boundary conditions, input parameters and testing sequences for developing and verifying an FE model of the interaction of OC grafts and surrounding host tissue.

Ten OC cylindrical grafts (height: 10mm; diameter: n=5–6.5mm; n=5–8.5mm) were harvested from adult porcine femurs (60–70kg). Unconfined compression tests were conducted using an Instron3365 with a 500N load cell and a BioBath filled with PBS at 37ºC. The OC grafts (prior to separation of cartilage and bone) and cartilage underwent four 5% strain (of cartilage layer) steps with displacement rate of 0.005mm/sec, each followed by a 45-minute relaxation. Final strain was 20%. Bone underwent a single displacement of 20% strain of bone at same displacement rate.

Young's moduli ranged from 6.2–42.0MPa, 0.7–3.9MPa, 46.8–123.7MPa for OC graft, cartilage and bone, respectively. The coefficient of variance between OC Grafts, cartilage and bone was 70.6%, 71.8%, and 25.2%, respectively.

Dispersion between samples was high. This may be due to intrinsic tissue variability but also due to the testing protocol: for cartilage in particular, the load was at the low end of the load cell capacity and the sample aspect ratio was poor for compressive testing. This work provides insight in understanding the effect of individual patient's and/or individual grafts used during osteochondral grafting. The results compel the need to accurately model these tissues when developing specimen-specific FE models for OC grafting.

Osteochondral (OC) grafting is one available method currently used to repair full thickness cartilage lesions with good results clinically when grafting occurs in patients with specific positive prognostic factors. However, there is poor understanding of the effect of individual patient and surgical factors. With limited tissue availability, development of Finite Element (FE) models taking into account these variations is essential. The aim of this study was to evaluate the effect of altering the material properties of OC grafts and their host environment through computer simulation.

A generic FE model (ABAQUS CAE 2017) of a push-out test was developed as a press-fit bone cylinder (graft) sliding inside a bone ring (host tissue). Press-fit fixation was simulated using an interference fit. Overlap between host and graft (0.01mm–0.05mm) and coefficient of friction (0.3–0.7) were varied sequentially. Bone Young's moduli (YM) were varied individually between graft and host within the range of otherwise derived tissue moduli (46MPa, 82MPa, 123MPa).

Increasing both overlap and frictional coefficient increased peak dislodging force independently (overlap: 490% & frictional coefficient: 176% across range tested). Increasing bone modulus also increased dislodging force, with host bone modulus (107%, 128%, and 140% increase across range, when Graft YM = 123MPa, 82 MPa, and 46MPa, respectively) having a greater influence than graft modulus (28%, 19% and 10% increase across range, when Host YM = 123 MPa, 82MPa and 46MPa, respectively).

As anticipated increasing overlap and friction caused an increase in force necessary to dislodge the graft. Importantly, differentially changing the graft and host material properties changed the dislodging force indicating that difference between graft and host may be an important factor in the success or failure clinically of osteochondral grafting.

Osteoarthritis is a debilitating disease affecting over 1.7 million people in the UK annually. Total ankle replacements are an increasingly sought option for repairing a late stage arthritic ankle, but result in the removal of significant portions of bone regardless of tissue quality. Hence, the mapping of bone quality would allow the use of targeted treatments at earlier stages of the disease. This study aims to develop characterisation methodologies using porcine tissue to investigate the mechanical properties of subchondral bone in the ankle.

N=11 talar bone plugs (6mm diameter) were extracted from porcine ankles and embedded into Delrin endcaps using a thin layer of PMMA cement. These were scanned under micro-CT (16 microns) and subjected to quasi-static uniaxial compression to determine apparent stiffness for each specimen. Specimen-specific continuum FE models were developed, with material properties derived from the greyscale value of the underlying image. A python-based least squares regression (Opti4Abq, N=6) was used to minimise the difference between experimental and model stiffness values, to determine the coefficient linking greyscale and mechanical properties. Apparent stiffness, elastic modulus and compressive strength were compared to BV/TV, which was derived using BoneJ (a bone image plugin for the NIH ImageJ).

The results show positive correlations between BV/TV and compressive strength, stiffness and Young's modulus. Average BV/TV across all samples was 0.45. Average experimental and computational stiffness were 986N/mm and 891 N/mm respectively. A 21.8% RMS error was found using the validation set (N=5), which was of similar order to the calibration set. Some specimens saw issues with misalignment of the specimen faces and the loading platens, likely causing overestimation of mechanical properties.

This study has developed methods that can be translated for use with human ankle bone and will lead to the development of an accurate means of mapping arthritic bone in the ankle.

Intervertebral disc (IVD) degeneration is one of the major causes of back pain. A number of emerging treatments for the condition have failed during clinical trial due to the lack of robust biomechanical testing during product development. The aim of this work was to develop improved in-vitro testing methods to enable new therapeutic approaches to be examined pre-clinically. It forms part of a wider programme of research to develop a minimally invasive nucleus augmentation procedure using self-assembling hydrogels.

Previous static testing on extracted IVDs have shown large inter-specimen variation in the measured stiffness when specimen hydration and fluid flow were not well controlled. In this work, a method of normalising the hydration state of IVDs prior-to and during compressive testing was developed.

Excised adult bovine IVDs underwent water-pik treatment and a 24-hour agitated bath in monosodium citrate solution to maximise fluid mobility. Specimens were submerged in a saline bath and held under constant pressure for 24 hours, after which the rate of change of displacement was low. Specimens were then cyclically loaded, from which the normalised specimen stiffness was determined. A degenerate disc model was developed with the use of enzymatic degeneration, allowing specimens to be tested sequentially in a healthy, degenerate, and then treated state. Self-assembling peptide-GAG hydrogels were tested using the developed method and the effect of treatment on stiffness and disc height were assessed.

Compared to previous static tests, the improved method reduced the variation in the normalised specimen stiffness. In addition, statistically significant differences were seen before and after enzymatic degradation to simulate degeneration, thus providing controls against which to evaluate treatments. The augmentation of the nucleus with the hydrogel intervention reduces the stiffness of the degenerate disc towards that of the healthy disc. This method is now being used to further investigate nucleus augmentation devices.

Acetabular tissue damage is implicated in osteoarthritis (OA) and investigation of in situ acetabular soft tissues behaviour will improve understanding of tissue properties and interconnections. The study aim was to visualise acetabular soft tissues under load and to quantify displacements using computed tomography (CT) scans (XtremeCT, Scano Medical).

A CT scan (resolution 82 μm) was performed on the disarticulated, unloaded porcine acetabulum. The femoral head was soaked in Sodium Iodide (NaI) solution and cling film wrapped to prevent transfer to the acetabular side. The joint was realigned, compressed using cable ties and re-scanned. The two images were down-sampled to 0.3 mm. Acetabular bone and soft tissues were segmented. Bony features were used to register the two background images, using Simpleware ScanIP 7.0 (Synopsys), to the same position and orientation (volume difference < 5%). Acetabular soft tissues displacements were measured by tracking the same points at the tissue edges on the two acetabular masks, along with difference in bone position as an additional error assessment.

The use of radiopaque solution provided a clear contrast allowing separation of the femoral and acetabular soft tissues in the loaded image. The image registration process resulted in a difference in bone position in the areas of interest equivalent to image resolution (0.3 mm, a mean of 3 repeats by same user). A labral tip displacement of 1.7 mm and a cartilage thickness change from 1.5 mm unloaded to 0.9 mm loaded, were recorded.

The combination of contrast enhancement, registration and focused local measurement was precise enough to reduce bone alignment error to that of image resolution and reveal local soft tissue displacements. These measurement methods can be used to develop models of soft tissues properties and behaviour, and therapy for hip tissue damage at early stage may be reviewed and optimised.

Ligaments and tendons are connective tissues with a highly hierarchical structure, from collagen fibres, to fibrils and fascicules. Their intricate structural arrangement produces an anisotropic non-linear elastic mechanical behaviour and a complex damage pattern before failure. Recent constitutive models have been developed with all parameters describing the structure of the tissue, with the advantage that they can in theory be measured on the tissue rather than being phenomenologically-derived. This is an ideal framework to model damage as its onset and propagation can be associated to changes in the structure directly.

In this preliminary study, the possibility to identify damage mechanisms in the tissue structure using in silico models was analysed for both the anterior cruciate ligament, with fascicules forming a helix with its longitudinal axis, and the patellar tendon, with fascicules co-aligned with its longitudinal axis. Tissues of interest were modelled as cylinders submitted to uniaxial tension. Damage was modelled as either a reduction of collagen volume fraction with increased strain, assuming the number of collagen fibres sustaining load decreases as fibres fail, or a reduction of the modulus of the fibres, assuming pre-failure damage of the fibres. Each damage mechanism was associated with a damage variable with different fibre stretch threshold for damage initiation and assuming linear variation of damage until an arbitrary failure point.

The apparent behaviour of the modelled tissues was significantly different as damage thresholds, damage mechanisms, type of fascicules were varied.

This preliminary work showed that using a structural constitutive model to describe occurrence and propagation of structural damage in an in silico model of hierarchical connective tissues is a framework that can clearly differentiate at a macroscopic level between different values of damage threshold and different damage mechanisms for tissue with co-aligned or helical fascicules.

The clinical uptake of minimally invasive interventions for intervertebral disc, such as nucleus augmentation, is currently hampered by the lack of robust pre-clinical testing methods that can take into account the large variation in the mechanical behaviour of the tissues. Using computational modelling to develop new interventions could be a way to test scenarios accounting for variation. However, such models need to have been validated for relevant mechanical function, e.g. compressive, torsional or flexional stiffness, and local disc deformations.

The aim of this work was to use a novel in-vitro imaging method to assess the validity of computational models of the disc that employed different degrees of sophistication in the anatomical representation of the nucleus.

Bovine caudal bone-disc-bone entities (N=6) were dissected and tested in uniaxial compression in a custom-made rig. Forty glass markers were placed on the surface of each disc. The specimens were scanned both with MRI and micro-CT before and during loading. Specimen-specific computational models were built from CT images to replicate the compression test. The anatomy of the nucleus was represented in three ways: assuming a standard diameter ratio, assuming a cylindrical shape with its volume matching that measured from MRI, and deriving the shape directly from MRI. The three types of models were calibrated for force-displacement. The radial displacement of the glass markers were then compared with their experimental displacement derived from CT images.

For a similar accuracy in modelling overall force-displacement, the mean error on the surface displacement was 35% for standard ratio nucleus, 38% for image-based cylindrical nucleus, and 32% for MRI-based nucleus geometry.

This work shows that, as long as consistency is kept to develop and calibrate image-based computational models, the complexity of the nucleus geometry does not influence the ability of a model to predict surface displacement in the intervertebral disc.

Introduction

Geometric variations of the hip joint can give rise to repeated abnormal contact between the femur and acetabular rim, resulting in cartilage and labrum damage. Population-based geometric parameterisation can facilitate the flexible and automated in silico generation of a range of clinically relevant hip geometries, allowing the position and size of cams to be defined precisely in three dimensions. This is advantageous compared to alpha angles, which are unreliable for stratifying populations by cam type. Alpha angles provide an indication of cam size in a single two-dimensional view, and high alpha angles have been observed in asymptomatic individuals.

Parametric geometries can be developed into finite element models to assess the potential effects of morphological variations in bone on soft tissue strains. The aim of this study was to demonstrate the capabilities of our parameterisation research tool by assessing impingement severity resulting from a range of parametrically varied femoral and acetabular geometries.