The Use of Hardened Bone Cement as an Impaction Grafting Extender

for Revision Hip Arthroplasty.

Mark Ruddy1, David P. FitzPatrick1, Kenneth T. Stanton1*

1School of Mechanical and Materials Engineering, University College Dublin, Belfield, Dublin 4,

Ireland

*Corresponding Author: kenneth.stanton@ucd.ie

Tel: +353 1 7161918

Fax: +353 1 7161942

Keywords: Morsellised bone, Impaction grafting, Total hip revision surgery, Bone cement

Abstract

Impaction bone grafting is a method of restoring bone stock to patients who have suffered significant

bone loss due to revision total hip surgery. The procedure requires morsellised cancellous bone

(MCB) to be impacted into the site of bone loss in order to stabilise the prosthesis with the aim of

long term resorption and reintegration of the impacted bone graft. Due to financial cost and the

potential to transmit disease, the use of supplementary material, known as an extender, is frequently

used to increase the graft material volume. This study investigates the use of Hydroset (Stryker Corp,

MA, USA), a hardened injectable bone cement (IBC) as an extender material and compares the

performance of the IBC in different weight percent inclusions to a commercially available bone graft

extender (GCP, BoneSave, Stryker Corp, MA, USA). The surgical impaction procedure was

standardised and samples were evaluated in terms of graft stiffness and height. It was observed that 30

wt.% IBC extended samples had significantly improved graft stiffness (p = 0.02) and no significant

different in height (p = 0.067) over a 100% MCB control sample. Cyclic loading, representative of

gait, found that the IBC subsided similarly to the commercial bone substitute in wt.% above 10%.

Shear testing of the impacted grafts showed no significant differences between GCP and IBC with

impaction forces determining the shear parameters of impacted grafts. The effects of the impaction

and cyclical loading procedures on extender particle sizes was assessed via particle size analysis. It

was found that the IBC extended samples demonstrated reduced friability, evident in the better

retention of particle size as a result of both impaction and gait representative loading compared to

that of the GCP samples. This indicates a potential reduction in issues arising from small particle

migration to joint surfaces. Scanning electron microscopy of the MCB particles with both GCP and

IBC as extenders showed retention of the porous trabecular structure post-testing which is essential

for revascularisation and bone growth into the graft.

1. Introduction

Removal of a worn or damaged primary hip replacement implants can necessitate extensive surgery

involving the extraction of any surrounding cement mantle and bone which is unsuitable for

promoting osteointegration of the new implant (Brewster et al., 1999). This process can lead to

significant loss of bone and several studies have shown that the outcome of revision surgery can be

dependent on the scale of this bone stock loss (Kavanagh et al., 1985). As a result, the restoration of

bone stock at the implant site is desirable; the current method of achieving this is impaction grafting

where morsellised bone is impacted into the bone cavity and this ensures a stable graft bed for the

new prosthesis (Slooff et al., 1984). The procedure for impaction bone grafting was first published in

2

1984 by Slooff et al.(Schreurs et al., 2004) and later adapted for use on the femur by Gie et al.(Gie et

al., 1993). It is a technique useful in cases where bone stock has been compromised to such an extent

that mechanical fixation of the implant is not possible.

The process of bone grafting has become extremely common with almost 1 million bone grafting

procedures being carried out in the United States each year; this figure is expected to grow by 13%

per year (Wise, 2000). Coupled with this is the increase in revision surgeries. The corresponding

increase in the need for impaction graft material has led to concerns over future availability (Galea et

al., 1998; Madhok et al.; Walschot et al., 2010). Allograft bone can also present a health hazard with

the transmission of disease from the donor to the recipient a possibility (Buck et al., 1989; Conrad et

al., 1995; Cook et al., 1995; Lord et al., 1988). The financial cost of allograft is also of concern

(Meding et al., 1997). In order to reduce costs and improve supply surgeons may frequently bulk-out

allograft material with a bone graft substitute, commonly known as an extender.

Extenders used in the literature for impaction grafting are limited. The majority of the studies

conducted have been on hydroxyapatite (Ca10(PO4)6(OH)2, HA) based materials. Granular calcium

phosphate bone substitutes (GCP), were the most widely used extenders in the literature that

investigated impactability and subsidence (Blom et al., 2005; Grimm et al., 2001; Harris, 1982; Van

Haaren et al., 2005; Verdonschot et al., 2001; Walschot et al., 2010). Additional materials such as

titanium and Bioglass™ have also been investigated as potential graft extenders (Brewster et al.,

1999; Walschot et al., 2010). The behaviour of Bioglass™ extended graft was only examined in shear

loading conditions (Brewster et al., 1999). Porous titanium was found to produce a stable elastic graft

however resorption and remodelling of titanium in vivo would not be possible.

Due to the limitations of phase pure HA biomaterials, a number of studies have been carried out to

examine the possibilities of composite materials. Most often, these will be mixtures of HA and β tri-

calcium phosphate (Ca3(PO4)2, βTCP) so as to give desired combinations of mechanical properties and

capabilities. Self-setting injectable bone cements (IBC) fall within this category of composite

materials.

IBC's are often supplied in two parts, a powder and a liquid, which must be mixed prior to insertion in

the body. These cements offer greater biocompatibilty over acrylic based cements and degrade more

quickly than phase pure HA. The mechanical properties of IBC's are comparable to healthy trabecular

bone and they can have good wet field characteristics (Clarkin et al., 2009). This is interesting for

impaction grafting as trabecular bone is the recommended source. These cements also display

excellent biocompatibility, bioactivity and osteoconductivity (Lewis, 2006). It is for these properties

that this study examines the use a commercial example HydroSet (Stryker Corp, MA, USA) in the

3

impaction grafting process. HydroSet is a two part cement supplied as a liquid (NA 3PO4, C6H9NO,

H2O) and a powder (CaHPO4, Ca4(PO4)2O, Na3C6H5O7) which are required to be mixed manually prior

to use.

A number of studies have stated that implant stability is highly dependent on the stiffness of the

impacted graft (Kärrholm et al., 1999). The method by which the impaction force is applied to the

graft material is divided into two schools of thought; those that slowly compact the graft (Giesen et

al., 1999; Ullmark and Nilsson, 1999) and those that apply a sudden impact force (Bavadekar et al.,

2001; Voor et al., 2004; Walschot et al., 2010). The present study utilises the sudden impact method

as the slow compaction of the graft suffers from issues in stress relaxation and is not as clinically

representative. However, the two methods do show commonalities, both test methods show that the

graft is viscoelastic and that the stiffness is variable over time. The correlation between the impaction

force applied, either gradual or sudden, and increasing stiffness has been found by both test methods.

Verdonschot et al. applied 98 compression cycles of 840 N to determine a MCB stiffness of 84.5 MPa

(Verdonschot et al., 2001). This is in comparison to the sudden impact method used by Bavadekar et

al. and Xu et al. who found mean maximum stiffnesses of 48 MPa (Bavadekar et al., 2001) and 68.06 

MPa (Xu et al., 2011) respectively. The higher stiffness of the gradual compressive impaction method

may be as a result of the higher compaction force applied. This force is above the threshold force

suggested to prevent femoral fracture during surgery (Flannery et al., 2010) and is therefore, not

representative of impaction forces that are commonly used during surgery.

The ability of an impacted graft to withstand the forces which will be experienced during the patient's

daily life is a major factor in the success of a procedure. An impacted graft which has not fully

stabilised the prosthesis can lead to movement of the prosthesis. This can result in pain for the patient

and accelerated wear of the prosthesis. Long-term stability is generally tested in terms of prosthesis

subsidence. Subsidence is the amount by which the graft moves or reduces in height. Subsidence is a

commonly used indicator for implant failure. Studies have found that the majority of the subsidence

occurs in the first 1000 cycles of testing (Blom et al., 2002; Van Haaren et al., 2005), which concurs

with clinical observations in which subsidence is evident in the first 6 months (Karrholm et al., 1994;

Onsten et al., 1995). A surgeon may evaluate an implant by viewing an X-ray scan and measuring the

amount of distance the implant has moved.

Previous studies have utilised sinusoidal loading profiles to examine the behaviour of extender

materials in reducing the subsidence of hip revision implants. Border et al. found that the introduction

of a TCP/HA extender reduced the subsidence of an artificial acetabular model compared to the pure

bone graft (Bolder et al., 2002a). Van Haaren et al. used a TCP/HA extender to a femoral model and

found that it also reduced subsidence in comparison to a pure bone graft (Van Haaren et al., 2005).

4

These studies tested the addition of extenders in models which have complex geometries and

interfaces between implant and the graft. As a result, there may be scope for other factors, such as

slippage between the implant and the graft which affects the outcome of the tests. In order to limit any

potential factors contributing to subsidence the present study resolved to test only the graft itself.

The use of sinusoidal loading profiles simplifies the feedback loop control required for monitoring of

long-term testing; however, a sinusoidal force pattern is not a perfect representation of the forces

experienced by the hip joint during gait. Large variation also exists in the literature studies on the

loads being applied to the graft material (Blom et al., 2002; Kligman et al., 2003; Malkani et al., 1996;

McNamara et al., 2012; Van Haaren et al., 2005). By calculating a load profile based on population

data and recorded forces generated during normal walking speed; a more relevant gait test method

may be applied to the graft materials.

The morphological form of an impacted graft material has led some researchers to equate it to the

study of particulate aggregates in other disciplines such as soil mechanics. Such particulate materials

have been found to be susceptible to shear failures which may lead to migration of the prosthesis. It

has been documented that the application of shear forces is the main mechanism of failure in

acetabular components in impacted grafts. The shear strength, τ f, of an impacted graft has been

described by the Mohr-Coloumb relationship:

τ f=c+σn tanϕ

where, σ n is the normal perpendicular stress applied on the shear failure plane, and c and ϕ, the

cohesion intercept and the friction angle, are constants for a given particulate material (Craig, 2004).

Previous studies have utilised this approach to investigate the shear properties of impacted graft

materials. Brennan et al. used shear testing to examine the effects of vibration during impaction on

pure allograft samples (Brennan et al., 2011). Dunlop et al. and Tanabe et al. examined the effects of

particle sizes and grading in separate studies; again this was conducted on pure allograft samples

(Dunlop et al., 2003; Tanabe et al., 1999). The effect of the impaction force applied and the manner in

which it was applied was investigated by Albert et al. (Albert et al., 2008).

These studies thereby examined the effects of various conditions on an impacted allograft sample. The

present study is concerned with the effects of extender materials on the shear properties of the

impacted graft. The addition of Bioglass™ by Brewster et al. offers some comparison in terms of the

effect the addition of an extender can have (Brewster et al., 1999); however this study did not vary the

Bioglass™ content by weight percentage but rather by particle grading. This makes it difficult to

5

determine the amount of Bioglass™ present within the samples for extender comparisons. Oakley et

al. tested the addition of porous and solid extender material in 50 and 100% mix ratios (Oakley and

Kuiper, 2006). However, the method of using unconfined compression tests to determine shear

cohesion offers the potential for highly variable results and does not represent the characteristics of

the graft under confined simultaneous normal and shear loading.

2. Materials and Methods

Bovine bone femoral heads were obtained from an abattoir and stored at -18 °C until required. Prior to

undergoing the morselisation process, the femoral heads were thawed at room temperature for two

hours. The femoral heads were sectioned into rough sections approximately 2 x2 x2cm using a

reciprocating saw. These sections were then morsellised using a high RPM domestic blender with a

power rating of 1000 W (Nutri Ninja Auto IQ, SharkNinja, USA) (Ruddy, 2016).

The manufacturer’s instructions for the use of HydroSet require the mixing of the supplied liquid and

powder phase for 45 seconds in the supplied mixing bowl followed by transfer to an injection syringe

and finally injection of the paste into the desired location. This process should take approximately 4

minutes and 30 seconds with set time ranging from 4 to 8 minutes. For the purpose of this study it was

not required to inject the IGC. The standard mixing phase was carried out and the mixture was left to

harden in the mixing bowl for approximately 12-15 hours at room temperature before testing. Once

set, a small dental press was used to granulate the cement mass. This consisted of a small stainless

steel container into which a serrated stainless steel lid is pressed, the serrated points on the lid result in

the cement mass breaking up into granules.

The granular bone substitute (BoneSave, Stryker Corp, MA, USA) in the current study was used as

supplied. It is a composite ceramic granule composed of 80% βTCP and 20% HA. Granules used in

this study are supplied with a mean diameter of 5 mm and have a stated macro-porostiy of 50%. The

GCP has been used in a clinical setting for a number of years and has been used as the gold standard

of extender materials in numerous studies (Arts et al., 2007; Blom et al., 2005; Insley and Streicher,

2004). In the present study it will be used as an additional performance comparison point.

Samples of morsellised bone, GCP and hardened IBC were separately laid out on sheets of black

paper and imaged using an Inspex HD 1080p Vesa camera (Ash Technologies Ltd., Ireland). The

resulting images were analysed using ImageJ software to calculate the particle size of the samples via

binary image thresholding (Abramoff et al., 2004).

6

X-Ray diffraction (XRD) analysis was carried out on samples of the morsellised bone, GCP and IBC

for composition analysis. XRD was performed on a Siemens D500 diffractometer (Munich, Germany)

with a Bragg-Brentano geometry (2θ/θ) and a Cu-Kα source was used to analysis samples. The X-Ray

tube potential and current were set to 40 kV and 30 mA, respectively. The test parameters used for

each test was an angular range of 10-90° 2θ, a step size of 0.04 and a step time of 1.5 seconds. When

testing samples 10 wt.% of silicon (Si, 99.5% pure, Johnson Matthey Alfa Products, Karlsruche,

Germany) was added to the samples as an internal reference to allow correction for non-linear peak

shift.

Scanning electron microscopy (SEM) of the morsellised bone was carried out with a Hitachi TM-

1000 SEM (Hitachi High Technologies Europe, Krefeld, Germany). In order to prevent damage to the

SEM as a result of outgassing, the marrow within the MCB particles was removed by soaking in

acetone for 24 hours at 37 °C followed by a further 24 hours drying at 37 °C.

The GCP and IBC extenders were also mechanically characterised. The Young’s Modulus was

determined by an unconfined compression test on a single granule of the extenders. This was

performed using a universal mechanical tester (Tinius Olsen, UK) at a test displacement velocity of

0.5 mm/min. The bulk density and apparent porosity were calculated according to ASTM C373-17

(ASTM-International, 2017). The hardness was determined using a Vickers micro-indenter.

The impaction process used was based on similar studies presented in the literature (Bavadekar et al.,

2001; Voor et al., 2004; Walschot et al., 2010). In order to examine the effects of the addition of

extender material on the graft properties, the extenders were added in 10, 30 and 50 wt. % to the

morsellised bone. The total weight for each sample was 5 g (Table 1). Individual samples of graft

material were placed in a constraining tube (diameter 20 mm) and subjected to an impaction force

applied by a drop weight striking a stainless steel rod telescoping inside the constraining tube. The

impacted graft was then placed in a compression testing machine equipped with a 1 kN load cell

(Tinius Olsen, UK) to measure the stiffness, or Young’s modulus. The stiffness was calculated as the

slope of the curve between 68 and 98% of the maximum load recorded. A limit of 80 N or 0.3 mm of

displacement was placed on the stiffness test to prevent additional compression than that provided by

the impact force. The cross-head displacement velocity was 0.5 mm/min. This measurement was

conducted on the grafts that were successively subjected to 5, 10, 15, 20, 25, 30, 40, 50, 60, 70, 80, 90

and 100 impacts.

7

Notation ExtenderExtender

Weight (g)MCB

Weight (g)

No. Impaction Samples

No. Cyclical Samples

MCB N/A N/A 5 5 5

10 GCP Granular Bone Substitute 0.5 4.5 5 5

30 GCP Granular Bone Substitute 1.5 3.5 5 5

50 GCP Granular Bone Substitute 2.5 2.5 5 5

10 IBC Injectable Bone Cement 0.5 4.5 5 5

30 IBC Injectable Bone Cement 1.5 3.5 5 5

50 IBC Injectable Bone Cement 2.5 2.5 5 5Table 1 Notation and graft mix ratios.

Cyclical compressive testing was performed to assess the graft materials ability to withstand long

term gait cycles. Samples were placed in the constraining tube, which was based on the same design

as the impactor rig but with added attachment points for securing in a dynamic test machine. Each 5 g

sample was impacted 30 times prior to cyclical loading. This was found to be an adequate number to

reach a stiffness or packing plateau and is also representative of the clinical surgical procedure. A

custom gait loading profile was then applied to the sample within the tube via an Instron servo-

hydraulic test machine (High Wycombe, UK). A reference patient weight of 117 kg, based on the

Centre for Disease Control and Prevention (CDC) statistics for the 90th percentile of the 60-65 age

range, was used as a worst case scenario to calculate the forces experienced (Fryar et al.). The

assumed weight and the data on the percentage body weight applied throughout the gait cycle allowed

the calculation of the loading profile. Each sample underwent 5000 cycles or until a total axial travel

of 8 mm was recorded. A limit of 8 mm was chosen to represent an implant failure due to findings in

the literature in assessing the long-term success of femoral prostheses (Loudon and Older, 1989;

Ostlere and Soin, 2003).

Samples which had undergone both the impaction and cyclical gait loading processes were gently

broken back up into individual granules by hand. The resulting granules of MCB and the extender

were re-imaged in order to examine the particle size of the materials post testing. Samples of post-test

MCB were treated in acetone to remove any remaining marrow and underwent SEM to analyse the

structure and any change in particle size.

8

A modified version of the test process outlined in ASTM D3080: Standard Test Method for Direct

Shear Test of Soils Under Consolidated Drained Conditions(ASTM-International, 2003) was used to

perform the shear testing on impacted graft samples. A small-scale direct shear box apparatus was

designed and built specifically. The test equipment consisted of a 20 mm internal diameter cylindrical

test cell, split horizontally, into which the graft material is placed. The bottom half of the test cell is

fixed and the upper half is free to move (Figure 1). An electrically driven stage, onto which a load cell

is mounted, is used to provide the force to shear the upper section of the test cell away from the lower

half. The resulting load cell force voltage is passed through a Wheatstone bridge sensor USB interface

to facilitate data logging. The displacement of the test cell is recorded via a linear displacement sensor

on the stage which is recorded via an analog to digital converter. All data signals are processed

through LabView 2014: Version 14.0 (National Instruments, Texas, USA).

Figure 1 Schema of shear box test.

After preparation of the graft samples, the shear box is filled uniformly in small horizontal layers.

Care was taken to ensure that no voids are present in the test material. The sample was impacted a

total of 30 times, as per the cyclical loading test, to simulate a surgically impacted sample. The

surface of the sample was checked to ensure it was level and any excess material was removed by a

sharp blade. The shear test cell was then secured in place and a normal stress was applied to the top of

the graft via a hanger system. The normal stresses applied to the samples were 171, 241 and 328 kPa.

These values were chosen as comparative values to those found in the literature (Albert et al., 2008;

Brewster et al., 1999; Dunlop et al., 2003; Tanabe et al., 1999). The test procedure was carried out at a

displacement velocity of 1 mm/min.

Due to the nature of the test procedure, the area (A c) of the test cell must be corrected for

displacement (IChemE, 1989);

9

Ac=D2

2(θ−cos (θ )sin (θ ) )

where;

θ=cos−1( δh

D )δh = horizontal displacement,

D = diameter of test cell.

Having calculated the corrected area, the shear stress (τ) and the vertical stress (σ v) can be calculated,

were T is the recorded shear force and N the applied normal force;

τ= TA c

σ v=NAc

Analysis of the shear test data was carried out using MatLab (Natick, MA, USA). In calculating and

plotting the mean stresses at each applied normal stress point, a linear regression model was produced

for each graft sample. By applying the Mohr-Coloumb law, the internal friction angle of the sample is

determined by the tan-1 of the slope of the linear regression model and the constant or the intercept

value corresponds to the interlocking or cohesion value of the graft.

Statistical analysis of the data was carried out using Minitab (Minitab Inc., PA, USA) with an

assumed alpha value of 0.05.

3. Results

Table 2 outlines the results of the mechanical characterisation of GCP and IBC granules. Figure 2

shows the XRD traces of the materials used in the study. It is evident that all the materials exhibit

similar chemical compositions. HA and β-TCP peaks a rough quantitative analysis of solubility can be

determined. It was found that BoneSave and HS had ratios of 0.41 and 1.98 respectively. These ratios

give an estimate of the relative solubility of the extender materials. The higher presence of β-TCP,

and this lower ratio, will increase the solubility of the extender in vivo.

Extende

r

Young’s Modulus

(MPa)

Porosity (%) Bulk Density

(g/cm3)

Hardness (Hv)

GCP 23 49.99 2.09 197.31

10

IBC 310 44 1.57 43.56

Table 2 Results of the mechanical characterisation of the extender materials.

Figure 2 XRD traces of the graft materials used,

Figure 3 shows the range of stiffness values achieved by impacted samples over the course of the

entire impaction procedure. The mean maximum stiffness achieved by the GCP extended samples of

48.30 ± 3.34 MPa was obtained by 10 GCP. This was a 3% increase on the mean stiffness, 46.76 ±

3.72 MPa, of MCB. The mean stiffness values of 33.90 ± 2.97 and 28.78 ± 2.65 MPa for 30 GCP and

50 GCP presented a decrease of 27.5% and 38.4% respectively over MCB. A correlation analysis of

these figures confirmed that there was a significant negative correlation between increasing GCP

content and stiffness (p = 0.01).

ANOVA analysis was carried out on all mix ratios to show the relationship with the 100% MCB

sample. It was found that GCP exhibited p-values of <0.001, <0.001 and 0.278 for 50 GCP, 30 GCP

and 10 GCP respectively. Assuming a significance value, p, of 0.05 this confirmed that there was a

significant difference between 50 GCP, 30 GCP and MCB. The lowest content GCP sample, 10 GCP,

was not found to be significantly different.

IBC was found to be comparable to a 100% MCB sample in two mix ratios tested. ANOVA analysis

found that 50 IBC and 10 IBC were not significantly different from MCB. 30 IBC was found to

11

significantly differ from the MCB stiffness (p = 0.02). However, unlike other extender mixes which

were found to be lower in stiffness, 30 IBC was found to have a significantly higher stiffness than

MCB. The mean maximum stiffness values for 10 IBC, 30 IBC and 50 IBC was found to be 47.55 ±

5.57, 51.67 ± 5.96 and 43.55 ± 3.49 MPa, respectively.

Figure 3 Comparison of the effects of increasing extender content on the stiffness of the imapcted

graft bed. Statistically significant differences between groups are illustrated on the Figure.

The impacted graft heights are given in Figure 4. Final graft heights of 11.96 ± 0.10, 10.46 ± 0.56 and

10.02 ± 0.72 mm were found for 10 GCP, 30 GCP and 50 GCP respectively. In comparison with the

minimum graft height of MCB of 11.74 ± 0.08 mm, ANOVA analysis found that 50 GCP and 30 GCP

were significantly different from MCB (p <0.0001 and 0.003, respectively). 10 GCP was found to

have no significant difference from MCB (p = 0.91).

12

Figure 4 Comparison of the effects of increasing extender content on the height of the imapcted graft

bed. Statistically significant differences between groups are illustrated on the Figure.

Similar to the stiffness results, it was found that two mix ratios of IBC had no significant difference

from the control MCB group. ANOVA analysis found that both 10 IBC and 30 IBC were not

significantly different from MCB; this was indicated by p-values of 0.961 and 0.067 respectively. 50

IBC was found to be significantly different, p <0.001, from the MCB sample group. The mean

minimum graft height for 10 IBC, 30 IBC and 50 IBC was found to be 11.43 ± 0.17, 11.07 ± 0.01 and

8.94 ± 0.94 mm.

As is the case in the literature, it was found that the introduction of ceramic-based extenders resulted

in a reduction in subsidence over the control bone graft (MCB). Final MCB subsidence of 1.36 mm

was recorded after 5000 cycles. GCP was found to decrease the subsidence across all wt.% (Figure 5).

IBC as an impaction graft extender showed similar results, following the trend of increasing extender

content to reduced subsidence (Figure 6). 10 IBC was found to perform similarly to MCB suggesting

that a 10% inclusion is not sufficient to influence the overall characteristics of the impacted graft.

Further addition of IBC showed a reduction in subsidence to levels comparable with GCP.

13

Figure 5 Subsidence of GCP extended samples due to the applied cyclical load over 5000 cycles.

Figure 6 Subsidence of IBC extended samples due to the applied cyclical load over 5000 cycles.

The shear parameters found via shear testing can be seen in Table 3. It was found that the shear

strength of extended graft samples to be higher than that of the 100% MCB graft sample group. This

would indicate that the addition of ceramic based materials to an impaction graft can increase the

shear strength of the graft. However, ANOVA analysis of these values found no significant difference

between all samples (p = 0.08). An angle of 35° was found for the internal friction angle for MCB.

14

The cohesive force for the samples was not seen to follow a clear trend. In order to examine these

parameters in more detail, analysis of co-variance (ANCOVA) was carried out on the linear

regression models fitted to samples. It was found that GCP and IBC showed no significant difference

in internal friction angle with p-values of 0.23 and 0.19 respectively, assuming a significance value of

0.05. The analysis of the cohesive forces found significant differences between MCB and GCP and

IBC with p-values of 0.0052 and 0.002 respectively.

Graft SampleCohesive Force

(kPa)Angle of Friction

(°)Mean Shear Strength

(328 kPa Normal Stress)

MCB 198 35 43710 GCP 119 45 451

30 GCP 32 52 467

50 GCP 127 46 462

10 IBC 93 37 448

30 IBC 127 48 491

50 IBC 100 52 507Table 3 Shear strength parameters for impacted grafts.

Analysis of the samples pre- and post-impaction showed how the overall mean particle size reduced in

all cases. Figure 7 illustrates the change in each material from the initial particle size pre-testing to the

post-impaction and cyclical gait loading tests. Initial particle sizes of 5.11 ± 2.01, 4.93 ± 0.88 and

2.55 ± 1.74 mm were found for MCB, GCP and IBC respectively.

Post-impaction GCP extender was found to have reduced to a mean size of 0.51 mm. Similarly IBC

was found to have reduced in mean particle size to 0.85 mm. The difference between these means was

found to be significant (p < 0.001). As per the extended samples, analysis of a pure MCB impacted

graft showed reduction in the particle size post impaction. This confirmed that the reduction in the

mixed samples was not solely as a result of the extender materials. The MCB particle diameter

reduced to a mean of 1.72 mm.

As was the case after the impaction testing, it was found that all samples decreased in mean particle

size after gait loading. MCB was found to have reduced by 58% to a mean size of 2.12 mm. GCP used

15

in an extended graft was found to have reduced from an initial mean size of 4.93 mm to 0.64 mm. As

with the analysis of impacted samples, it was observed that the GCP samples were found to be more

friable having broken down into a fine powder and none of the initial porous structure had been

retained. This is similar to the results found during the impaction stiffness testing, even with less

impacts being applied in the case of the gait loading.

The IBC sample was not seen to reduce in particle size as much as GCP, with a final particle size of

1.38 mm recorded, this was also found to be a statistically significant difference (p < 0.001). As with

the GCP extended samples, the IBC granules were evidently smaller after impaction; however, IBC

was observed to demonstrate greater cohesiveness with the MCB granules than GCP when handled

after testing.

SEM analysis was carried out to examine the pore structure of the morsellised bone used in testing. It

was found that for both post-impaction and gait loading, the morsellised bone retained the cancellous

structure when used with both GCP and IBC extenders. Figure 8 illustrates that there is no appreciable

difference in the pore structure between the control 100% MCB graft and MCB used in conjunction

with an extender material.

16

Figure 7 Particle size analyses. (A) MCB particle sizes pre- and post-testing. (B)

GCP particle sizes pre- and post-testing. (C) IBC particle sizes pre- and post-

testing.

17

Figure 8 Composite stitched SEM images of bone particles, used in conjunction with

different extenders, post impaction and post cyclical loading. (A-B) MCB. (C-D) GCP. (E-F)

IBC.

4. Discussion

A limited number of studies exist in the literature that investigate the effects of parameters such as

graft size, water and fat content, bone mill types, impaction grade and graft preparation on the

18

mechanical performance of impacted bone graft. However, it is widely accepted across the literature

that impacted graft stiffness is the primary determining factor of initial prosthesis stability. A wide

range of stiffness values have been reported for impacted MCB from 3-135 MPa (Bavadekar et al.,

2001; Cornu et al., 2009; Giesen et al., 1999; Verdonschot et al., 2001; Voor et al., 2000; Voor et al.,

2004). This range may be attributed to the variety of test conditions used across the literature sources.

In comparing to methods similar to those carried out in this study, a range of 40-70 MPa can be

considered for MCB (Bavadekar et al., 2001; Cornu et al., 2009). The stiffness values found for pure

MCB graft in the current study were within this range indicating a favourable comparison with prior

studies and alterative MCB preparation techniques.

It was found that 50 IBC and 10 IBC were not significantly different from the control MCB group. 30

IBC was found to be statistically different to the control MCB, but unlike other extenders that were

found to be significantly different, 30 IBC produced a higher stiffness. This result suggests that IBC is

a viable extender in terms of graft stiffness when used in the correct mix ratio. 10 GCP was found to

perform similarly to the control group which was not surprising as GCP has been utilised in impaction

grafting in both the acetabular cup and the femur with positive results (Arts et al., 2007; Blom et al.,

2005; Bolder et al., 2002b; Insley and Streicher, 2004).

A similar repeatable trend for the reduction in height of all samples was also seen. This was a process

with very little variability. This is a positive result for the surgeon as it was seen that after

approximately 30 impacts the height reduction reaches a plateau. This behaviour has also been

observed in the literature by Xu et al. and Cornu et al. (Cornu et al., 2009; Xu et al., 2011). This

indicates that subsequent impacts may not be necessary.

The levels of subsidence for samples consisting of purely MCB were found to be comparable to those

found in other studies (Blom et al., 2002; Van Haaren et al., 2005). Kligman et al. investigated the

differences in using cortical and cancellous bone in impaction grafting. It was found that cancellous

bone resulted in a mean subsidence of 1.32 ± 0.32 mm compared to 0.94 ± 0.26 mm for cortical bone

(Kligman et al., 2003). However, it is not recommended to use cortical bone due to the lack of

porosity and the higher elastic modulus which can affect the remodelling of the graft. The subsidence

of cortical bone found by Kligman et al. was still found to be higher than the rates recorded for many

of the extended grafts in this study which also offered more desirable characteristics.

Cyclical load testing suggested that the use of IBC as an extender material offered long term increases

in dimensional stability similar to GCP. This is in line with observations in the literature that the

introduction of ceramic based extenders results in a reduction in subsidence over control bone graft.

Blom et al. investigated the effects of the addition of 90 and 50 wt.% GCP to a allograft sample

19

(Blom et al., 2002). The highest subsidence recorded by Blom et al. for a 50 wt.% GCP mix was 1.04

mm and this compared favourably with the 0.219 mm measured for the equivalent IBC sample in this

study. The greater subsidence recorded by Blom et al. was judged to be as a result of the longer test

time of 20,000 cycles. A 50 wt.% GCP sample in this study was found to subside by 0.35 mm

suggesting greater stability was achievable from IBC.

Although the use of higher wt.% of IBC was shown to reduce the subsidence of the graft to very low

levels, approximately 0.5 mm, a word of caution must be noted. With the introduction of a material

with a much lower compressibility and less visco-elastic behaviour than that of bone, a greater

transmission of force to the femur may occur. This behaviour was seen to have resulted in an increase

in fractures to femurs during testing by Van Haaren et al. (Van Haaren et al., 2005).

It was observed that higher wt.% concentrations of GCP reduced the initial stiffness of the impacted

graft. This would suggest that a choice has to be made by the surgeon with regards to either reduced

long term subsidence or increased initial prosthesis stability. Increased stability can ensure that a

patient is back to full function more quickly after surgery which, in turn, can improve rehabilitation.

Due to the reduced stability of higher extender wt.% GCP, the patient may need to be more cautious

during rehabilitation which will slow healing and bone ingrowth. This was not as pronounced in the

IBC samples with 30 wt.% IBC showing the highest stiffness (51.67 MPa) and 50 wt.% IBC reducing

to 43.55 MPa compared to 28.78 MPa for 50 wt.% GCP.

The shear strengths of impacted grafts found in the literature varies widely. Under normal stresses of

350-370 kPa, shear strengths between 225 and 1700 kPa have been reported. In this study, for a

normal force of 328 kPa, a mean shear strength of 466 kPa across all the graft mixes was found.

Within the literature the definition of shear strength varies. A number of studies have not taken the

reducing cross sectional area into account as the test progresses and have calculated the shear strength

as the shearing force divided by the initial cross sectional area. The definition of shear stress at

different strains also gives rise to variability.

An increase in mean shear strength of 16% was recorded for IBC over MCB compared to 5% for

GCP. However, none of these increases were found to be statistically significant. The increases do

compare well to a reported increase of 7% in shear strength with the addition of Bioglass in a study by

Brewster et al. (Brewster et al., 1999). However, it was not stated if this increase was significant and

the sample size for each group included in the study was one. The mix ratio is also not stated.

It would have been expected that the introduction of ceramic granules into the impacted grafts would

increase the internal friction angle. This is due to the improved grading of the samples. A positive

20

correlation was calculated between increasing extender content and increasing friction angle.

However, these increases were, like shear strength not found to be statistically significant. This did

have some positive indications for the use of IBC as it was shown to be comparable to a widely

accepted extender with clinically proven results.

The use of an extender, such as IBC, with more angular characteristics than that of GCP should result

in greater particle interlocking/cohesion. This, in turn, should lead to an increase in shear strength as

evident by the Mohr-Coloumb relationship (↑c = ↑τ). However, such a relationship was not

observed in this study. This pointed to the impaction force dictating the shear parameters of an

impacted graft. Also, as MCB is a highly deformable material, the packing characteristic depends

more on the impaction force rather than granule characteristics.

The non-significance of the shear strength values recorded in this study indicate that once full

compaction of the graft has been achieved the shear properties are not as dependant on the granules

but rather the impaction force applied. Studies have shown that applying greater impaction energy or

impaction aids, such as vibration can increase the shear strength (Albert et al., 2008; Brennan et al.,

2011). The plasticity of the MCB may also reduce the effects of different particle shapes and sizes by

combining the graft into a single mass. This observation has also been noted by Albert et al. (Albert et

al., 2008). Testing graft samples which have not been fully compacted may reveal more granule

dependant shear properties.

Particle analysis of the graft mixture showed that the GCP extender graft had a large reduction in

particles and a significantly smaller mean particle size post-testing compared to that of the IBC. This

is of interest due to the as-supplied structure of the GCP exhibiting some of the features ideal for host

bone infiltration. In particular the 50% porosity and pore size. The results found in this study would

suggest that these features are of little relevance as none are retained after the impaction process,

therefore offering no appreciable porosity benefit over IBC. IBC also offers the benefit of retaining a

larger particle size which can reduce concerns over the possible migration of small GCP particles

from the graft site to other areas of the body where they may cause unwanted effects (Dunne et al.,

2015).

The distribution of the particle sizes, prior to impaction testing, offered by GCP and IBC was also of

interest. It was found that IBC offered a wider range of particles sizes; this is often referred to as

grading. It is well documented in the literature that for a granular material, the better graded a material

is the less void space present and the greater the density prior to impaction. This characteristic of IBC

as an extender may offer the potential to require less impaction energy to be imparted on it to reach a

21

compactness comparable to a similar wt.% GCP graft. As a result the energy transferred to the femur

is reduced thus reducing the potential of fracture occurring during surgery.

It is necessary for graft material to have a porous network to promote revascularisation and

remodelling. SEM imaging of the MCB particles used in both impaction and cyclical loading tests

revealed that for all extenders the trabecular structure was kept intact. This has shown that the use of

neither extender material affected the trabecular structure of the MCB. This offer benefits as

structurally damaged cancellous bone has been shown to have a much lower elastic modulus

(Walschot et al., 2010). This retention of the trabecular structure has been reported as a benefit of

larger particles (4-8 mm) as the internal porosity allows deformation and interlocking (Cornu et al.,

2009).

5. Conclusions

The use of GCP has been seen to be a completely viable option for an impaction graft extender based

on literature and clinical outcomes. However, issues surrounding the breakdown of GCP granules into

very small particles raises questions as to the claimed benefits of the initial porous structure. The very

small particles created during fracture under impact also raise concerns over the possible migration

from the graft site to other areas of the body where they may cause unwanted effects. The supplied

granule size distribution is also a limiting characteristic with studies showing increased mechanical

performance from more well graded distributions.

The injectable bone cement used in the current study is a readily available FDA approved material

which offers similar chemical attributes as natural cancellous bone and has been utilised successfully

as an injectable bone cement for a number of years. The potential of IBC to harden into a form with

both similar chemical and mechanical properties as cancellous bone has led to the investigation

performed in this study. IBC has been found to offer the potential to overcome a number of issues

with GCP. However, the setting reaction required to solidify the cement is an ongoing process and

there may be variations in the performance of HydroSet in the impaction process over time. The

environment in which the setting reaction takes place i.e. humidity and temperature, may also lead to

mechanical variation. This would require further study to evaluate the HydroSet granules at advancing

time points from the initiation of the setting process.

This study has shown that a well graded granular mix can be produced leading to increased initial

mechanical performance. IBC has also been found to withstand to forces experienced in the impaction

22

process better than GCP and as a result the fear of potential particle migration are reduced with the

use of IBC.

Mechanical testing both relevant to the surgical procedure and those demands put on the graft during

rehabilitation have shown positive signs for IBC. It has performed to a similar level as the equivalent

GCP extended sample and in some cases outperformed it. IBC showed increased impaction stiffness

over GCP with loading of 30 wt.% showing the best results. It was also found that in levels up to

50wt.% IBC outperformed GCP and resulted in less of a reduction in stiffness with increasing

percentage content. Cyclical loading based on force experienced during gait revealed that IBC

required content levels greater than 10% to reduce subsidence. This was in contrast to GCP which

immediately reduced subsidence at the smallest level of inclusion. Shear testing did not reveal

differences between GCP and IBC with impaction forces dictating the graft properties rather than the

extenders.

This study offers a starting point for the use of IBC or other related materials as extender materials.

Further testing would be required, in particular longer term loading to determine if the trends

observed in this study continued. A future unforeseen large subsidence would need to be guaranteed

against as such an occurrence would have significant consequences for a patient. An investigation into

the changes in IBC as a result of in-vivo degradation would also be essential. It would be important to

determine if the characteristics observed in this study are retained under more realistic in-vivo

conditions.

6. Acknowledgements

The authors wish to thank the Programme for Research in Third Level Institutions (PRTLI) for

funding of this project.

7. References

Abramoff, M.D., Magalhaes, P.J., Ram, S.J., 2004. Image processing with ImageJ. Biophotonics International 11, 36-42.

Albert, C., Masri, B., Duncan, C., Oxland, T., Fernlund, G., 2008. Impaction allografting—the effect of impaction force and alternative compaction methods on the mechanical characteristics of the graft. Journal of Biomedical Materials Research Part B: Applied Biomaterials 87, 395-405.

23

Arts, J.J.C., Walschot, L.H.B., Verdonschot, N., Schreurs, B.W., Buma, P., 2007. Biological activity of tri-calciumphosphate/hydroxyl-apatite granules mixed with impacted morsellized bone graft. A study in rabbits. Journal of Biomedical Materials Research Part B: Applied Biomaterials 81, 476-485.

ASTM-International, 2003. ASTM D3080, Standard Test Method for Direct Shear Test of Soils Under Consolidated Drained Conditions. ASTM International, West Conshohocken, PA.

ASTM-International, 2017. ASTM C373, Standard Test Methods for Determination of Water Absorption and Associated Properties by Vacuum Method for Pressed Ceramic Tiles and Glass Tiles and Boil Method for Extruded Ceramic Tiles and Non-tile Fired Ceramic Whiteware Products. ASTM International, West Conshohocken, PA.

Bavadekar, A., Cornu, O., Godts, B., Delloye, C., Tomme, J.V., Banse, X., 2001. Stiffness and compactness of morselized grafts during impaction: An in vitro study with human femoral heads. Acta Orthopaedica 72, 470-476.

Blom, A.W., Cunningham, J.L., Hughes, G., Lawes, T.J., Smith, N., Blunn, G., Learmonth, I.D., Goodship, A.E., 2005. The compatibility of ceramic bone graft substitutes as allograft extenders for use in impaction grafting of the femur. Journal of Bone & Joint Surgery, British Volume 87, 421-425.

Blom, A.W., Grimm, B., Miles, A.W., Cunningham, J.L., Learmonth, I.D., 2002. Subsidence in impaction grafting: the effect of adding a ceramic bone graft extender to bone. Proceedings of the Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine 216, 265-270.

Bolder, S., Verdonschot, N., Schreurs, B., Buma, P., 2002a. Acetabular defect reconstruction with impacted morsellized bone grafts or TCP/HA particles. A study on the mechanical stability of cemented cups in an artificial acetabulum model. Biomaterials 23, 659-666.

Bolder, S.B.T., Verdonschot, N., Schreurs, B.W., Buma, P., 2002b. Acetabular defect reconstruction with impacted morsellized bone grafts or TCP/HA particles. A study on the mechanical stability of cemented cups in an artificial acetabulum model. Biomaterials 23, 659-666.

Brennan, S.A., Brabazon, D., O’Byrne, J.M., 2011. Effect of vibration on the shear strength of impacted bone graft in revision hip surgery. Journal of Bone & Joint Surgery, British Volume 93, 755-759.

Brewster, N.T., Gillespie, W.J., Howie, C.R., Madabhushi, S.P.G., Usmani, A.S., Fairbairn, D.R., 1999. Mechanical considerations in impaction bone grafting. Journal Of Bone & Joint Surgery - British Volume 81, 118-124.

Buck, B.E., Malinin, T.I., Brown, M.D., 1989. Bone Transplantation and Human Immunodeficiency Virus: An Estimate of Risk of Acquired Immunodeficiency Syndrome (AIDS). Clinical Orthopaedics and Related Research 240, 129-136.

Clarkin, O.M., Boyd, D., Madigan, S., Towler, M.R., 2009. Comparison of an experimental bone cement with a commercial control, Hydroset™. Journal of Materials Science: Materials in Medicine 20, 1563-1570.

Conrad, E.U., Gretch, D.R., Obermeyer, K.R., Moogk, M.S., Sayers, M., Wilson, J.J., Strong, D.M., 1995. Transmission of the hepatitis-C virus by tissue transplantation. The Journal of Bone & Joint Surgery 77, 214-214.

Cook, S.D., Salkeld, S.L., Prewett, A.B., 1995. Simian immunodeficiency virus (human HIV-II) transmission in allograft bone procedures. Spine 20, 1338-1342.

24

Cornu, O., Schubert, T., Libouton, X., Manil, O., Godts, B., Van Tomme, J., Banse, X., Delloye, C., 2009. Particle size influence in an impaction bone grafting model. Comparison of fresh-frozen and freeze-dried allografts. Journal of biomechanics 42, 2238-2242.

Craig, R.F., 2004. Craig's Soil Mechanics. Taylor & Francis.

Dunlop, D.G., Brewster, N.T., Madabhushi, S.P.G., Usmani, A.S., Pankaj, P., Howie, C.R., 2003. Techniques to improve the shear strength of impacted bone graft the effect of particle size and washing of the graft. The Journal of Bone & Joint Surgery 85, 639-646.

Dunne, C.F., Gibbons, J., FitzPatrick, D.P., Mulhall, K.J., Stanton, K.T., 2015. On the fate of particles liberated from hydroxyapatite coatings in vivo. Irish Journal of Medical Science (1971-) 184, 125-133.

Flannery, O.M., Britton, J.R., O'Reilly, P., Mahony, N., Prendergast, P.J., Kenny, P.J., 2010. Threshold force required for femoral impaction grafting in revision hip surgery: A preliminary study in sow femurs. Acta Orthopaedica 81, 303-307.

Fryar, C.D., Gu, Q., Ogden, C.L., Anthropometric reference data for children and adults: United States, 2007–2010. National Center for Health Statistics.

Galea, G., Kopman, D., Graham, B.J.M., 1998. Supply and demand of bone allograft for revision hip surgery in Scotland. Journal of Bone & Joint Surgery, British Volume 80, 595-599.

Gie, G.A., Ling, R.S.M., Linder, L., 1993. Impacted cancellous allografts and cement for revision total hip arthroplasty. Journal Of Bone & Joint Surgery, British Volume 21.

Giesen, E.B.W., Lamerigts, N.M.P., Verdonschot, N., Buma, P., Schreurs, B.W., Huiskes, R., 1999. Mechanical characteristics of impacted morsellised bone grafts used in revision of total hip arthroplasty. Journal of Bone & Joint Surgery, British Volume 81, 1052-1057.

Grimm, B., Miles, A.W., Turner, I.G., 2001. Optimizing a hydroxyapatite/tricalcium-phosphate ceramic as a bone graft extender for impaction grafting. Journal of Materials Science: Materials in Medicine 12, 929-934.

Harris, W.H., 1982. Allografting in total hip arthroplasty: In adults with severe acetabular deficiency including a surgical technique for bolting the graft to the ilium. Clinical Othopaedics and Related Research, 150-164.

IChemE, 1989. Standard Shear Testing Technique For Particulate Solids Using The Jenike Shear Cell. Institution of Chemical Engineers, Warwickshire, UK.

Insley, G.M., Streicher, R.M., 2004. In Vitro and In Vivo Validation of a New Bonesubstitute for Loaded Orthopaedic Applications. Key Engineering Materials 254, 993-996.

Kärrholm, J., Hultmark, P., Carlsson, L., Malchau, H., 1999. Subsidence of a non-polished stem in revisions of the hip using impaction allograft evaluation with radiostereometry and dual energy x-ray absorptiometry. Journal of Bone & Joint Surgery, British Volume 81, 135-142.

Karrholm, J., Malchau, H., Snorrason, F., Herberts, P., 1994. Micromotion of femoral stems in total hip arthroplasty. A randomized study of cemented, hydroxyapatite-coated, and porous-coated stems with roentgen stereophotogrammetric analysis. The Journal of Bone & Joint Surgery 76, 1692-1705.

Kavanagh, F., Ilstrup, D.M., Fitzgerald, R.H., 1985. Revision total hip arthroplasty. The Journal of Bone & Joint Surgery 67, 517-526.

25

Kligman, M., Rotem, A., Roffman, M., 2003. Cancellous and cortical morselized allograft in revision total hip replacement: A biomechanical study of implant stability. Journal of Biomechanics 36, 797-802.

Lewis, G., 2006. Injectable bone cements for use in vertebroplasty and kyphoplasty: State-of-the-art review. Journal of Biomedical Materials Research Part B: Applied Biomaterials 76, 456-468.

Lord, C.F., Gebhardt, M.C., Tomford, W.W., Mankin, H.J., 1988. Infection in bone allografts. Incidence, nature, and treatment. The Journal of Bone & Joint Surgery 70, 369-376.

Loudon, Older, M., 1989. Subsidence of the femoral component related to long-term outcome of hip replacement. Journal of Bone & Joint Surgery, British Volume 71-B, 624-628.

Madhok, R., Lewallen, D.G., Wallrichs, S.L., Ilstrup, D.M., Kurland, R.L., Melton, L.J., Trends in the utilization of primary total hip arthroplasty, 1969 through 1990: A population-based study in Olmsted County, Minnesota, Mayo Clinic Proceedings, 1 ed, pp. 11-18.

Malkani, A.L., Voor, M.J., Fee, K.A., Bates, C.S., 1996. Femoral component revision using impacted morsellised cancellous graft a biomechanical study of implant stability. Journal of Bone & Joint Surgery, British Volume 78, 973-978.

McNamara, I., Rayment, A., Brooks, R., Best, S., Rushton, N., 2012. The effect of the addition of hydroxyapatite graft substitutes upon the hoop strain and subsequent subsidence of a femoral model during impaction bone grafting. Journal of the mechanical behavior of biomedical materials 5, 238-246.

Meding, J.B., Ritter, M.A., Keating, M.E., Faris, P.M., 1997. Impaction Bone-Grafting before Insertion of a Femoral Stem with Cement in Revision Total Hip Arthroplasty. A Minimum Two-Year Follow-up Study. The Journal of Bone & Joint Surgery 79, 1834-1841.

Oakley, J., Kuiper, J.H., 2006. Factors affecting the cohesion of impaction bone graft. Journal of Bone & Joint Surgery, British Volume 88, 828-831.

Onsten, I., Akesson, K., Besjakov, J., Obrant, K.J., 1995. Migration of the Charnley stem in rheumatoid arthritis and osteoarthritis. A roentgen stereophotogrammetric study. Journal of Bone & Joint Surgery, British Volume 77, 18-22.

Ostlere, S., Soin, S., 2003. Imaging of prosthetic joints. Imaging 15, 270-285.

Ruddy, M., 2016. Extender Materials for Use in Impaction Allografting for Revision Total Hip Arthroplasty. University College Dublin, pp. 86-87.

Schreurs, B.W., Bolder, S.B.T., Gardeniers, J.W.M., Verdonschot, N., Slooff, T., Veth, R.P.H., 2004. Acetabular revision with impacted morsellised cancellous bone grafting and a cemented cup a 15-20 year follow-up. Journal of Bone & Joint Surgery, British Volume 86, 492-497.

Slooff, T.J.J.H., Huiskes, R., van Horn, J., Lemmens, A.J., 1984. Bone grafting in total hip replacement for acetabular protrusion. Acta Orthopaedica 55, 593-596.

Tanabe, Y., Wakui, T., Kobayashi, A., Ohashi, H., Kadoya, Y., Yamano, Y., 1999. Determination of mechanical properties of impacted human morsellized cancellous allografts for revision joint arthroplasty. Journal of Materials Science: Materials in Medicine 10, 755-760.

Ullmark, G., Nilsson, O., 1999. Impacted corticocancellous allografts: recoil and strength. The Journal of Arthroplasty 14, 1019-1023.

26

Van Haaren, E.H., Smit, T.H., Phipps, K., Wuisman, P., Blunn, G., Heyligers, I.C., 2005. Tricalcium-phosphate and hydroxyapatite bone-graft extender for use in impaction grafting revision surgery an in vivo study on human femora. Journal of Bone & Joint Surgery, British Volume 87, 267-271.

Verdonschot, N., Van Hal, C.T.H., Schreurs, B.W., Buma, P., Huiskes, R., Slooff, T., 2001. Time-dependent mechanical properties of HA/TCP particles in relation to morsellized bone grafts for use in impaction grafting. Journal of Biomedical Materials Research 58, 599-604.

Voor, M.J., Nawab, A., Malkani, A.L., Ullrich, C.R., 2000. Mechanical properties of compacted morselized cancellous bone graft using one-dimensional consolidation testing. Journal of Biomechanics 33, 1683-1688.

Voor, M.J., White, J.E., Grieshaber, J.E., Malkani, A.L., Ullrich, C.R., 2004. Impacted morselized cancellous bone: Mechanical effects of defatting and augmentation with fine hydroxyapatite particles. Journal of biomechanics 37, 1233-1239.

Walschot, L.H.B., Schreurs, B.W., Buma, P., Verdonschot, N., 2010. Impactability and time-dependent mechanical properties of porous titanium particles for application in impaction grafting. Journal of Biomedical Materials Research Part B: Applied Biomaterials 95, 131-140.

Wise, D.L., 2000. Biomaterials and bioengineering handbook. Marcel Dekker New York, NY:.

Xu, Z.-J., Chen, L.-Y., Zhong, C., Tan, Y.-B., He, R.-X., 2011. Mechanical properties of 7-10 mm bone grafts and small slurry grafts in impaction bone grafting. Journal of Orthopaedic Research 29, 1491-1495.

27