06475988

8/12/2019 06475988

http://slidepdf.com/reader/full/06475988 1/8

2214 IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 60, NO. 8, AUGUST 2013

The Biomechanical Effect of Notch Size, NotchLocation, and Femur Orientation

on Hip Resurfacing FailureZachary Morison, Michael Olsen, Gordon A. Higgins, Rad Zdero∗, and Emil H. Schemitsch

Abstract —For hip resurfacing, this is the first biomechanicalstudy to assess anterior and posterior femoral neck notching andfemur flexion and extension. Forty-seven artificial femurs wereimplanted with the Birmingham hip resurfacing (BHR) using arange of notch sizes (0, 2, and 5 mm), notch locations (superior,anterior, and posterior), and femur orientations (neutral stance,flexion, and extension). Implant preparation was done using im-ageless computer navigation, and mechanical tests measured stiff-ness and strength. For notch size and location, in neutral stancethe unnotched group had 1.9 times greater strength than the 5-mm

superior notch group (4539 N versus 2423 N, p = 0.047), and the5-mm anterior notch group had 1.6 times greater strength than the5-mm superior notch group, yielding a borderline statistical differ-ence (3988 N versus 2423 N, p = 0.056). For femur orientation, inthepresenceof a 5-mm anterior notch,femursin neutral stancehad2.2 times greater stiffness than femurs in 25◦ flexion (1542 N/mmversus 696 N/mm, p = 0.000). Similarly, in the presence of a 5-mmposteriornotch,femurs in neutral stance had2.8 times greater stiff-ness than femurs in 25◦ extension (1637 N/mm versus 575 N/mm, p = 0.000). No other statistical differences were noted. All femursfailed through the neck. The results have implications for BHRsurgical techniques and recommended patient activities.

Index Terms —Biomechanics, failure, femur, hip resurfacing,notch.

I. INTRODUCTION

H IP resurfacing arthroplasty is a bone-conserving alterna-

tive to total hip replacement in the treatment of end-stage

hip disease. Hip resurfacing presents several advantages to tradi-

tional total hip arthroplasty, especially for active young patients,

including preservation of proximal femoral bone stock, lowered

risk of dislocation, no violation of the intramedullary canal,

Manuscript receivedDecember 30,2012; revisedFebruary19, 2013; acceptedMarch 4, 2013. Date of publication March 7, 2013; date of current version July13, 2013. Asterisk indicates corresponding author.

Z. Morison and M. Olsen are with the Martin Orthopaedic BiomechanicsLaboratory, St. Michael’s Hospital, Toronto, ON M5B 1W8, Canada (e-mail:[email protected]; [email protected]).

G. A. Higgins is with the South Devon Foundation Hospital Trust,Torbay District General Hospital, Torquay, Devon TQ2 7AA, U.K. (e-mail:[email protected]).

∗R. Zdero is with the Martin Orthopaedic Biomechanics Laboratory, St.Michael’s Hospital, Toronto, ON M5B 1W8, Canada, and also with Departmentof Mechanical and Industrial Engineering, Ryerson University, Toronto, ON,Canada (e-mail: [email protected]).

E. H. Schemitsch is with the University of Toronto, Toronto, ON M5S 1A1,Canada (e-mail: [email protected]).

Digital Object Identifier 10.1109/TBME.2013.2251745

favorable wear characteristics, and excellent medium-term sur-

vival rates of about 95% [1]–[6].

Despite these potential benefits, there are still considerable

risks associated with hip resurfacing that may lead to a variety

of complications, such as adverse local tissue reaction [7]–[10],

necrosis of the femoral remnant [11], [12], unknown long-

term effects of increased serum metal ion levels [13], [14],

and femoral neck fracture [15], [16]. In particular, femoral neck

fracture appears to be the dominant clinical failure mechanism,accounting for almost 40% of hip resurfacing failures [17].

Femoral neck fracture is sometimes the result of surgical error

during femoral head preparation, which can introduce a notch in

the region of the femoral neck [15], [18]. The notch is thought

to be a stress riser that predisposes the femur-implant construct

to mechanical failure during postoperative physiological weight

bearing and other activities of daily living [19].

Femoral neck notching, however, has been the topic of very

few biomechanical studies on bone-conserving hip implants.

Davis et al. found that, compared to unnotched specimens, su-

perior neck notches of 2 and 5 mm weakened the proximal

femur by 24% (4034 N versus 5302 N) and 47% (2808 N ver-

sus 5302 N), respectively, during axial strength testing of theBirmingham hip resurfacing (BHR), while their correspond-

ing finite element analysis showed increased stresses for larger

notches [19]. Anglin et al. [20] discovered that femurs with

a hip resurfacing and a 2-mm superior neck notch achieved a

28% higher failure strength in axial compression (5250 N ver-

sus 4100 N) if the implant was placed in 10◦ valgus compared

to neutral alignment. Olsen et al. [21] showed that a 5-mm su-

perior neck notch statistically lowered the axial failure strength

of femurs implanted with a Birmingham mid-head resection by

19% (4060 N versus 5002 N) compared to control femurs with

no notch; however, a 2-mm superior notch did not reveal any

statistical effect. Although these researchers assessed the effectsof notches located on the superior aspect of the femoral neck,

there is no information in the literature investigating the influ-

ence of anteriorly or posteriorly located femoral neck notches

on hip resurfacing biomechanics. Moreover, these investigations

assessed femurs in the neutral single-leg stance phase of walk-

ing, but none to date have done so with the femur in flexion or

extension.

This study, therefore, measured the biomechanical stiffness

and strength of hip resurfaced femurs having anterior and poste-

rior neck notches of different sizes, while femurs were oriented

in neutral stance, flexion, and extension. This is the first inves-

tigation to do so.

0018-9294/$31.00 © 2013 IEEE

8/12/2019 06475988


MORISON et al.: BIOMECHANICAL EFFECT OF NOTCH SIZE, NOTCH LOCATION, AND FEMUR ORIENTATION 2215

Fig. 1. Typical femur-BHR construct with a superior neck notch. (a) Intactfemoral head, (b) superior notch created on prepared femoral head, (c) BHRinserted onto the prepared femoral head, and (d) radiograph of final femur-BHR construct. Notch sizes included 0 mm (i.e., no notch), 2 mm (anterior orposterior), and 5 mm (superior, anterior, or posterior).

Fig. 2. Mechanical test setup showing a proximal femur-BHR specimen inneutral stance, i.e., 7◦ adduction, 0◦ flexion, and 0◦ extension. Additional femurorientations used included 25◦ flexion (with 7◦ adduction) and 25◦ extension(with 7◦ adduction). The BHR head were free to rotate inside the smooth steelcup, whereas the distal condyles of the femur were rigidly secured using thesteel pins of a ring-like clamping system. Vertical force was applied onto the

BHR head in the downward direction.

II. METHODOLOGY

A. General Study Design

A series of artificial femurs received a BHR and were tested

over a range of neck notch sizes (0, 2, and 5 mm), neck notch lo-

cations (i.e., superior notch, anterior notch, and posterior notch)

(see Fig. 1), and simulated routine activities (i.e., walking,

climbing, and lunging) (see Fig. 2). Femurs were first non-

destructively tested under axial compression to obtain stiffness,

which was followed by destructive testing under axial com-

pression to obtain strength. Axial compression was chosen as

the loading mode, since it is commonly used for biomechani-

cal studies on femurs with a hip resurfacing [19]–[25] and be-

cause it is the dominant force experienced by the femoral neck

during walking [26]. Stiffness nondestructively assesses bone-

implant behavior at low loads, say, for activities that may not

cause bone-implant damage such as physician-prescribed pre-

cautionary“toe touch” weight-bearing immediately postsurgery.

Strength destructively assesses bone-implant behavior at high

loads, say, for activities that could cause bone-implant damage

such as full weight-bearing, walking, sports, etc.

B. Femur Properties

Forty-seven large left fourth-generation composite femurs

were obtained (Model #3406, Sawbones, Vashon, WA, USA).

The artificial cortical bone had a density = 1.64 g/cm3 , ulti-

mate tensile strength = 106 MPa, and modulus of elasticity =

16 GPa. The artificial cancellous bone had a “cellular” matrix

with a density = 0.32 g/cm3 , ultimate compressive strength =

5.4 MPa, and modulus of elasticity =

137 MPa. The femurshad an overall length of 485 mm from the top of the femoral

head to the distal end of the condyles, an intramedullary canal

diameter of 16 mm, a femoral head diameter of 52 mm, and a

native neck angle of 120◦ relative to the anatomical axis of the

femoral shaft.

C. Femur Randomization

Femurs were randomly divided into eight groups to be surgi-

cally prepared and implanted with a BHR (see Fig. 1). For simu-

lated neutral single-leg stance during walking (adduction = 7◦;

flexion = extension = 0◦), the femur-BHR test groups were no

notch, i.e., 0-mm notch (n = 6), 2-mm anterior notch (n = 6), 5-mm anterior notch (n = 6), 2-mm posterior notch (n = 6), 5-mm

posterior notch (n = 6), and 5-mm superior notch (n = 5). For

simulated climbing or lunging (adduction = 7◦; flexion = 25◦),

the femur-BHR test group had a 5-mm anterior notch (n = 6).

For simulated climbing or lunging (adduction = 7◦; exten-

sion = 25◦), the femur-BHR test group had a 5-mm posterior

notch (n = 6).

D. Femur Preparation

Femurs were prepared using imageless computer navigation

(VectorVision SR, BrainLAB, Feldkirchen, Germany) to ensure

accurate implant placement, as done previously by some of thecurrent authors [27]. Femurs were individually registered and

digitally mapped using the BrainLAB infrared camera and ar-

ray system in order to plan the alignment and entry point of the

guide wire. This wire provided the basis for the remainder of the

standard preparation of the femoral head, including postinser-

tion, central canal drilling, cylindrical reaming, planning, and

chamfering. To create a notch on the anterior or posterior neck,

implant position was planned in a posteriorly or anteriorly trans-

lated position, respectively. Notch sizes were verified by caliper

measurement. All implants were prepared in neutral alignment

with a 120◦ angle between the implant anchoring stem and the

femoral shaft. To ensure correct alignment, all stem shaft angles

8/12/2019 06475988



were verified to be within ±1◦ by plain digital radiographs. No

fractures were visible upon radiographic examination. Femurs

were finally implanted with the BHR (Smith and Nephew Inc.,

Memphis, TN, USA) following the surgical instructions of the

manufacturer. This was done by applying a layer of antibiotic

polymethyl-methacrylate cement (Stryker, Allendale, NJ, USA)

to the inside of the BHR shell, manually placing it overtop the

prepared femoral head, and then tapping its superior aspect with

a rubberhammer until it was seatedfirmly onto thefemoral head.

Femoral component implant size was determined to be 46 mm

by gaging the femoral neck using the standard BHR surgical

sizing instrument.

E. Mechanical Testing

All tests on intact femurs and femur-BHR constructs were

done with a mechanical testing system (Model #8874, Instron,

Norwood, MA, USA), which had a ±25 kN capacity, a 0.1 N

resolution, and a ±0.5% accuracy (see Fig. 2). Very similar test

regimes have been previously employed in biomechanical stud-

ies to determine stiffness and strength of hip resurfacing systems

under axial compression [19]–[25]. Proximally, femoral heads

were inserted into a smooth steel cup that was attached to the

load cell and linear displacement transducer of the mechanical

tester, which allowed simultaneous measurements of force and

displacement. Femoral heads were free to rotate inside the cup.

Distally, femoral condyles were fixed rigidly using threaded

bolts with sharp tips that were tightened through a steel ring-

like jig and into the superficial layer of the cortical bone. The

ring-like jig was mounted rigidly to the base of the mechanical

tester. A vertical load was then applied to produce axial com-

pression through the proximal steel cup to the femoral heads.

For stiffness tests, a 100 N preload was applied followed by theuse of displacement control to apply vertical loads to a max-

imum compression of 0.5 mm at a fixed rate of 10 mm/min.

Stiffness was defined as the average slope of the force-versus-

displacement curve from three trials on each specimen. All spec-

imens remained within the linear elastic range during stiffness

tests and did not incur any damage prior to the load-to-failure

testing, as indicated by the high values for the coefficient of

determination for the force-versus-displacement curves (R2 ≥

0.98). For strength tests, the same loading parameters were used

except that vertical load was applied until complete structural

collapse of the specimen was observed. Strength was defined as

the maximum force of the force-versus-displacement curve.

F. Statistical Analysis

Statistical analysis was performed using the software pack-

age SPSS 16 (SPSS Inc., Chicago, IL, USA). All data were

combined together into a single array for one-way analysis of

variance (ANOVA). Tamhane’s T2 analysis was used for all

post hoc analysis. A p-value less than 0.05 was considered

significant. There were 13 relevant pairwise comparisons ex-

tracted from the ANOVA analysis which examined the effect of

a change only in 1 parameter; other pairwise comparisons were

superfluous. Specifically, in neutral stance, the effect of notch

size was determined by comparing 0-mm (i.e., no notch) versus

Fig. 3. Mechanical test results for (a) stiffness and (b) strength. Statisticaldifferences ( p < 0.05) are indicated by symbols. Notch sizes were 0 mm (i.e.,no notch), 2 mm, and 5 mm. Notch locations were on anterior (i.e., ANT),

posterior (i.e., POS), and superior (i.e., SUP) sides of the neck. Femurs wereoriented in neutral single-leg stance, 25◦ flexion, and 25◦ extension.

2-mm versus 5-mm anterior notches, 0-mm versus 2-mm ver-

sus 5-mm posterior notches, and 0-mm versus 5-mm superior

notches. Similarly, in neutral stance, the effect of notch location

was determined by comparing 2-mm anterior versus 2-mm pos-

terior notches, and 5-mm anterior versus 5-mm posterior versus

5-mm superior notches. The effect of femur orientation was de-

termined by comparing 5-mm anterior and posterior notches at

neutral stance versus 25◦ flexion, and 5-mm anterior and poste-

rior notches at neutral stance versus 25◦ extension. A one-tailed

post hoc power analysis was finally performed to ensure thatthere were enough femurs per test group to detect all statistical

differences that were present, i.e., to avoid type II error. One un-

notched specimen failed prematurely during strength tests and,

thus, was only included in statistical analysis for stiffness, but

not for strength.

III. RESULTS

A. Effect of Notch Size During Neutral Stance

For stiffness, during neutral single-leg stance simulating

walking, there was no effect of notch size for any com-

parisons (0.865 < p < 1.000) [see Fig. 3(a)]. Similarly, for

8/12/2019 06475988



strength there was no effect of notch size for any comparisons

(0.518 < p < 1.000), except that the 0-mm, i.e., no notch, group

had about 1.9 times greater strength than the 5-mm superior

notch group (4539 N versus 2423 N, p = 0.047) [see Fig. 3(b)].

B. Effect of Notch Location During Neutral Stance

For stiffness, during neutral single-leg stance simulatingwalking, there was no effect of notch location for any compar-

isons (0.861 < p < 1.000) [see Fig. 3(a)]. Similarly, for strength

there was no influence of notch location for any comparisons

(0.108 < p < 1.000), except that the 5-mm anterior notch group

had about 1.6 times greater strength than the 5-mm superior

notch group, showing a borderline statistical difference (3988 N

versus 2423 N, p = 0.056) [see Fig. 3(b)].

C. Effect of Femur Orientation on 5-mm Notches

For a 5-mm anterior notch, femurs in neutral single-leg stance

had about 2.2times greater stiffness than femursin 25◦ of flexion

simulating climbing or lunging (1542 N/mm versus 696 N/mm, p = 0.000) [see Fig. 3(a)]. Similarly, in the presence of a

5-mm posterior notch, femurs in neutral single-leg stance had

about 2.8 times greater stiffness than femurs in 25◦ of extension

simulating climbing or lunging (1637 N/mm versus 575 N/mm,

p = 0.000) [see Fig. 3(a)]. However, for strength, the effect of

femur orientation was not apparent (0.724 < p < 0.999) [see

Fig. 3(b)].

D. Failure Modes

All notched femur-BHR specimens had fractures whichbegan

adjacent to the BHR rim near the notch and then propagated

vertically downward through the neck, ending at the top of thelesser trochanter (see Fig. 4). All unnotched specimens had

the same failure pattern except that the fracture started at the

superior edge of the BHR rim. All damage was through the

femoral neck, whereas none of the other regions of the femur or

the BHR implants themselves were damaged.

E. Post Hoc Power Analysis

For stiffness pairwise comparisons, analysis yielded average

powers of 29% (notch size), 33% (notch location), and 100%

(femur orientation). For strength pairwise comparisons, analy-

sis showed average powers of 47% (notch size), 67% (notch

location), and 55% (femur orientation).

IV. DISCUSSION

A. General Findings

This is the first study on hip resurfacing examining the biome-

chanical influence of anterior and posterior neck notching, as

well as femur flexion and extension. A 5-mm anterior notch

reduced stiffness during femur flexion, a 5-mm posterior notch

reduced stiffness during femur extension, and a 5-mm superior

notch reduced strength during femur neutral stance. Smaller

2-mm notches did not have an effect regardless of notch loca-

tion or femur orientation. All femurs failed through the neck.

Fig. 4. Typical failures occurred through the femoral neck for all femurshaving (a) no notch, (b) superior notch, (c) anterior notch, and (d) posteriornotch, regardless of notch size, notch location, or femur orientation.

Orthopaedic surgeons should endeavor to eliminate accidental

notching of the anterior, posterior, andsuperiorfemoral neck and

advise younger active patients to moderate activities involving

extreme flexion and extension.

B. Comparison to Prior Studies

There are no published biomechanical studies on hip resur-facing (or other bone-conserving hip implants) assessing the

influence of anterior or posterior femoral neck notches or femur

orientation. Moreover, there are few prior studies on the effect of

superior neck notching on hip resurfacing biomechanics, thus

making direct comparison to the current results challenging.

However, a few points of commonality with these prior studies

should be highlighted, as follows.

Davis et al. [19] evaluated 0-mm (i.e., unnotched), 2-mm,

and 5-mm superior neck notching on the properties of the BHR

using mechanical tests on artificial femurs plus finite element

modeling. Compared to the unnotched group, their 2- and 5-mm

superior notches lowered proximal femur axial strength, respec-

tively, by 24% (4034 N versus 5302 N) and 47% (2808 N versus5302 N). Their finite element analysis confirmed these data, by

showing increased stress in this region for larger neck notches.

Similarly, current tests demonstrated a 1.9-times difference in

axial strength of the 0-mm unnotched group (4539 N) versus the

5-mm notch group (2423 N) [see Fig. 3(b)]. Their specimens all

failed through the neck, similar to the current fracture modes.

The higher axial strengths for both notch sizes of the prior in-

vestigators may be due to their placement of the BHR stem in 5◦

valgus relative to the native femoral neck shaft angle, whereas

the current BHR stem was placed inline with the neck shaft.

Anglin and coworkers tested an unnamed hip resurfacing with

human cadaveric femurs to evaluate the influence of a 2-mm

8/12/2019 06475988



superior notch on implants placed neutrally inline with the na-

tive femoral neck versus a relative 10◦ valgus implant [20].

Their implants placed in 10◦ relative valgus achieved a 28%

higher axial strength compared to the neutral group (5250 N

versus 4100 N). This protective effect was only evident for fe-

murs with normal bone mineral density greater than 0.65 g/cm2 .

This was higher than the 2423 ± 424 N strength for current 5-

mm superior notch specimens in neutral stance [see Fig. 3(b)].

Their femurs all failed through the neck region, similar to the

present fracture modes. In contrast to the current experimen-

tal specimens and test regimes, their use of human femurs, an

unnamed hip resurfacing, 10◦ valgus implant placement, and

femurs potted just below the lesser trochanter would account

for any differences in measured strength values compared to

present data.

Olsen and colleagues evaluated 2- and 5-mm superior neck

notching in artificial femurs for the Birmingham mid-head Re-

section (BMHR), which is similar to hip resurfacing in that it

conserves proximal bone and does not violate the intramedullary

canal [21]. They found that a 5-mm superior neck notch sta-tistically lowered the axial failure strength of femurs by 19%

compared to unnotched control femurs (4060 N versus 5002 N),

whereas a 2-mm superior notch did not reveal any statistical

difference compared to the unnotched group (4367 N versus

5002 N). Additionally, femurs with a 5-mm notch with the

BMHR placed in 10◦ relative valgus had a protective effect

with a failure strength reaching 4469 N that was not statistically

different than the unnotched specimens. This was higher than

the 2423 ± 424 N strength for present 5-mm superior notch

femurs in neutral stance [see Fig. 3(b)]. Their specimens frac-

tured through the neck, similar to present failures. In contrast to

present tests, their use of the BMHR and a 10

◦

valgus implanta-tion accounts for some differences in strength levels compared

to current data.

C. Clinical and Biomechanical Implications

Current tests demonstrated that a 5-mm anterior or posterior

notch reduced stiffness during quasi-static loading regimes that

simulated climbing and lunging (i.e., 25◦ of flexion or exten-

sion). However, a 2- and5-mm anterior or posterior notch did not

mechanically compromise the femur’s structural integrity dur-

ing neutral single-leg stance simulating normal walking. Hip

resurfacing has been used especially for younger active patients

who may place their femurs in extreme flexion and extensionduring sports or other activities of daily living that require such

motions. Consequently, surgeons should be especially careful

to avoid surgical accidents that create anterior and posterior

notches for this patient group. Moreover, younger active pa-

tients should also be counseled to curtail any activities that

could involve extreme flexion or extension of their lower limbs.

Present findings also confirmed the importance of avoiding

notching of the superior femoral neck, since a 5-mm supe-

rior notch reduced failure strength during neutral single-leg

stance. This has also been established in the previous litera-

ture through finite element analysis, mechanical testing, and

clinical studies of hip resurfacing in the presence of superior

Fig. 5. 2-D forces and moments experienced during neutral single-leg stance

for (a) superior and (b) anterior (or posterior) neck notches. F = applied axialforce, M = resultant moment at the notch, V = resultant vertical shear forceat the notch, and N.A. = neutral axis.

neck notches [18]–[21]. Specifically, finite element analysis has

demonstrated that the stress riser created by the hip resurfacing

component is due to the disparity of the high stiffness cobalt

chrome prosthesis versus the low stiffness femoral neck [19].

In addition, finite element analysis has shown that a superior

femoral neck notch will further concentrate stress at the rim

of the femoral component, dramatically increasing the risk of

femoral neck fracture [19]. Mechanical testing of superior neck

notches can reduce femoral strength by 19–47% for hip resur-facing and related bone-conserving implants [19], [21]. Clinical

studies have reported that femoral neck fractures are accompa-

nied by superior neck notches in as many as 47% of cases [18].

Consequently, superior neck notching will continue to be a sub-

ject of concern and further investigation for improving the ser-

vice life of hip resurfacing implants.

Several trends observed presently can be explained using ba-

sic 2-D biomechanics. First, in neutral single-leg stance, su-

perior neck notches reduced stiffness and strength more than

anterior or posterior neck notches. The reason was that superior

notches were positioned along the outer edge of the femoral

neck cross section and, thus, experienced maximal stresses due

to the resultant moment M created by applied axial force F [see Fig. 5(a)], whereas anterior and posterior notches were po-

sitioned along the neck’s neutral axis which minimized stresses

caused by the resultant moment M [see Fig. 5(b)]. Second,

compared to neutral single-leg stance, flexion and extension

caused reduced stiffness and strength for anterior and posterior

neck notches. The reason was that flexion and extension caused

the anterior and posterior notches to rotate away from the neu-

tral neck axis closer to the point of axial force application F ,

thereby increasing their stress level created by the resultant mo-

ment M . It should be noted that, by definition, the magnitude

of the resultant vertical shear force V created by the applied

force F was identical for all notches regardless of notch size,

8/12/2019 06475988



notch location, or femur orientation. Since some of these trends

were not statistically significant due to small sample size, fu-

ture experiments with more specimens or computational studies

that evaluate stresses in more detail are needed to confirm these

remarks.

D. Addressing Possible Limitations

Artificial, rather than human, femurs were employed. The

architectural variation in cancellous human bone (which is

anisotropic) along applied stress lines does not exist in an ar-

tificial femur (which is isotropic) [28]. These artificial femurs

also had standardized geometry, thus not accounting for the

large variation in shape and size among human femurs which,

for example, can vary by a factor of 1.25 (overall length), 1.35

(midshaft diameter), 1.56 (neck length), and 1.12 (neck-to-shaft

angle) [29]. This may cause a different surface stress distribu-

tion in artificial versus human femurs. Even so, compared to

human cadaveric bones, artificial bones have advantages, such

as easy storage, low cost, availability, no biohazard, and reducedinterspecimen variability in properties [30], [31]. Artificial fe-

murs agree with healthy human femurs for the ratio of axial

versus torsional stiffness [29]. They also have similar results for

cortical and cancellous screw pullout stress relative to human fe-

murs [32], [33], and have shown clinical-type failure patterns at

the neck when tested intact or with a hip resurfacing [22], [34].

Moreover, the 3-D stress distribution in the proximal region

of artificial femurs undergoing axial compression is consistent

with data seen in clinical studies [35].

Muscle forces, especially abductor muscles, were not consid-

ered. Bitsakos et al. [36] showed that inclusion of muscle at-

tachments had a substantial bone-conserving effect during boneremodeling in various regions around a hip implant. However,

Stolk et al. [37] demonstrated that loading configurations which

include only hip joint contact force plus the abductor muscle can

sufficiently duplicate the mechanics of total hip replacements

during in vivo loading. Numerous previous studies have simpli-

fied things further by ignoring the abductor muscle for walking

simulations for traditional hip implants and hip resurfacing,

since the femur is primarily in a state of axial compression, as

done presently [19]–[25], [38]–[42]. If the abductor force was

simulated currently, a portion of the force would have mini-

mized the tensile stress along the superior aspect of the femoral

neck. This counteractive force would have affected the absolute

values of present results, but not the relative performance of thetest groups.

Only quasi-static axial compression was used. When im-

planted in vivo, however, hip resurfacings could potentially be

subjected to cyclic axial forces during activities of daily liv-

ing [43], [44]. Cyclic axial loading would also yield data about

potential fatigue failure of the hip resurfacing and/or stiffness

changes over time from micromotion at the implant-femur inter-

face [45]. However, the relative performance of the test groups

reported here would likely be similar in a “real world” clini-

cal scenario that involved cyclic forces for activities of daily

living and/or injury processes. It could also be suggested that

more realistic loading associated with clinical fracture risk for

a hip resurfacing is lateral impact to the trochanter due to a

fall, which causes femur neck fracture [46], [47]. Yet, axial

compression occurs during a common loading regime for the

lower limb, namely, weightbearing while walking [26], and has

been commonly used to assess fracture risk for hip resurfacing

previously [19]–[25].

Despite the present inspection of the prepared resurfaced fe-

murs visually and using computer navigation, it is possible that

the femoral neck may have been weakened during the BHR

impaction process using a rubber hammer, thereby contributing

to the femur’s weakening. Hip resurfacing impaction has been

suggested to create microfractures within the proximal femur

supporting the implant, thus inherently weakening the femur

surrounding the hip resurfacing. This phenomenon is challeng-

ing to detect, since previous investigators have noted the pres-

ence of microfractures only by histological analysis [48].

The volume of bone loss generated in each region of the

femoral head during accidental surgical notching should be

considered in the future for a proper engineering analysis of

hip resurfacing, since it alters the area moment of inertia of thefemoral head during loading. This may give new insights into

the mechanical stability and failure of hip resurfacings caused

by notches in different regions of the femoral head because,

for example, the bone loss created by a superior notch actually

extends (in tapered fashion) around to the anterior and poste-

rior surfaces of the femoral head. However, the current study

used only a simple linear designation of notching (i.e., 0, 2, and

5 mm), since it is the only standard that is universally used by

all orthopaedic surgeons and researchers [18]–[21].

Finite element modeling and analysis were not done to assess

stresses on the femur-BHR construct. Future researchers may

wish to do so to better understand the surface and internal stressdistributions following hip resurfacing. For example, some of

the current investigators have used computer modeling to com-

pute the stresses of femurs implanted with a BHR [19]. This

helped identify neck notches as areas of high stress, i.e., areas of

potential clinical failure, which was confirmed experimentally.

Since the current investigators performed destructive strength

tests to assess the failure load and mechanisms, computer mod-

eling was not needed to merely predict areas of potential failure.

Average post hoc powers ranged from 29% to 100%, indi-

cating that the study was underpowered for most comparisons,

i.e., inadequate number of specimens per test group. A good

study design is normally considered to have a minimum power

of 80%. However, to achieve 80% power, a much larger numberof samples would have been required, which would not have

been feasible either financially or practically.

V. CONCLUSION

This is the first biomechanical study on hip resurfacing ex-

amining anterior and posterior neck notching, as well as femur

flexion and extension. A 5-mm anterior or posterior neck notch

reduced stiffness, respectively, during flexion and extension,

i.e., climbing or lunging. A 5-mm superior notch reduced fe-

mur strength during neutral stance, i.e., walking. Consequently,

surgeons should attempt to eliminate accidental notching of

8/12/2019 06475988



the anterior, posterior, and superior neck during hip resurfac-

ing surgery, while younger active patients may be counseled to

conservatively engage in these activities.

ACKNOWLEDGMENT

The authors thank Smith and Nephew (Memphis, TN, USA)

for providing the BHR implants used in this investigation.

REFERENCES

[1] H. C. Amstutz, P. E. Beaule, F. J. Dorey, M. J. Le Duff, P. A. Campbell,and T. A. Gruen, “Metal-on-metal hybrid surface arthroplasty: Two to six-year follow-up study,” J. Bone Joint Surg. Amer., vol. 86, no. 1, pp. 28–39,2004.

[2] D. L. Back, R. Dalziel, D. Young, and A. Shimmin, “Early results of primary Birmingham hip resurfacings: An independent prospective studyof the first 230 hips,” J. Bone Joint Surg. Brit., vol. 87, no. 3, pp. 324–329,2005.

[3] R. B. C. Treacy, C. W. Mcbryde, and P. B. Pynsent, “Birmingham hipresurfacing arthroplasty: A minimum follow-up of five years,” J. Bone Joint Surg. Brit., vol. 87, no. 2, pp. 167–170, 2005.

[4] R. T. Steffen, H. Pandit, J. Palan, D. Beard, R. Gundle, P. Mclardy-Smith,D. W. Murray, and H. S. Gill, “The five-year results of the Birminghamhip resurfacing arthroplasty: An independent series,” J. Bone Joint Surg. Brit., vol. 90, no. 4, pp. 436–441, 2008.

[5] M. Khan, J. Kuiper, D. Edwards, E. Robinson, and J. Richardson, “Birm-ingham hip arthroplasty: Five to eight years of prospective multicenterresults,” J. Arthro., vol. 24, no. 7, pp. 1044–1050, 2009.

[6] L. Rahman,S. K. Muirhead-Allwood,and M. Alkinj, “What is themidtermsurvivorship and function after hip resurfacing?” Clin. Orthop. Relat. Res.2010, vol. 468, pp. 3221–3227, 2010.

[7] S. Glyn-Jones, H. Pandit, Y. M. Kwon, H. Doll, H. S. Gill, andD. W. Murray, “Risk factors for inflammatory pseudotumour formationfollowing hip resurfacing,” J. Bone Joint Surg. Brit., vol. 91, pp. 1566–1574, 2009.

[8] H. Pandit, S. Glyn-Jones, P. McLardy-Smith, R. Gundle, D. Whitwell,C. L. Gibbons, S. Ostlere, N. Athanasou, H. S. Gill, and D. W. Murray,

“Pseudotumours associated with metal-on-metal hip resurfacings,” J. Bone Joint. Surg. Brit., vol. 90, pp. 847–851, 2008.

[9] D. H. Nawabi, S. Gold, S. Lyman, K. Fields, D. E. Padgett, andH. G. Potter, “MRI predicts ALVAL and tissue damage in metal-on-metalhip arthroplasty,” Clin. Orthop. Relat. Res., 2013 Jan 26 [Epub ahead of print].

[10] T. P. Schmalzried, “The future of hip resurfacing,” Orthop. Clin. North Amer., vol. 42, no. 2, pp. 271–273, 2011.

[11] R. T. Steffen, N. A. Athanasou, H. S. Gill, and D. W. Murray, “Avascularnecrosis associated with fracture of the femoral neck after hip resurfacing:Histological assessment of femoral bone from retrieval specimens,” J. Bone Joint. Surg. Brit., vol. 92, pp. 787–793, 2010.

[12] J. Zustin, G. Sauter, M. M. Morlock, W. Ruther, and M. Amling, “Asso-ciation of osteonecrosis and failure of hip resurfacing arthroplasty,” Clin.Orthop. Relat. Res., vol. 468, pp. 756–761, 2010.

[13] A. J. Hart, S. Sabah, J. Henckel, A. Lewis, J. Cobb, B. Sampson,A. Mitchell, and J. A. Skinner, “The painful metal-on-metal hip resur-facing,” J. Bone Joint. Surg. Brit., vol. 91, pp. 738–744, 2009.

[14] S. J. MacDonald, “Metal-on-metal total hip arthroplasty: The concerns,”Clin. Orthop. Relat. Res., vol. 429, pp. 86–93, 2004.

[15] D. R. Marker, T. M. Seyler, R. H. Jinnah, R. E. Delanois, S. D. Ulrich,and M. A. Mont, “Femoral neck fractures after metal-on-metal total hipresurfacing: A prospective cohort study,” J. Arthro., vol. 22, no. 7 (Suppl3), pp. 66–71, 2007.

[16] R. T. Steffen, P. R. Foguet, S. J. Krikler, R. Gundle, D. J. Beard, andD. W. Murray, “Femoral neck fractures after hip resurfacing,” J. Arthro.,vol. 24, pp. 614–619, 2009.

[17] Australian Orthopaedic Association, National Joint Replacement Reg-istry, Annual Report 2009[Online]. Available:http://www.dmac.adelaide.edu.au/aoanjrr/documents/aoanjrrreport_2009.pdf.

[18] A. J. Shimmin and D. Back, “Femoral neck fractures following Birming-ham hip resurfacing: A national review of 50 cases,” J. Bone Joint. Surg. Brit., vol. 87, pp. 463–464, 2005.

[19] E. T. Davis, M. Olsen, R. Zdero, M. Papini, J. P. Waddell, andE. H. Schemitsch, “A biomechanicaland finiteelementanalysis of femoral

neck notching during hip resurfacing,” J. Biomech. Eng., vol. 131, no. 4,pp. 041002-1–041002-8, 2009.

[20] C. Anglin, B. A. Masri, J. Tonetti, A. J. Hodgson, and N. V. Greidanus,“Hip resurfacing femoral neck fracture influenced by valgus placement,”Clin. Orthop. Relat. Res., vol. 465, pp. 71–79, 2007.

[21] M. Olsen, P. M. Lewis, J. P. Waddell, and E. H. Schemitsch, “A biome-chanical investigation of implant alignment and femoral neck notchingwith the Birmingham mid-head resection,” J. Arthro., vol. 25, no. 6 (Suppl), pp. 112–117, 2010.

[22] E. T. Davis, M. Olsen, R. Zdero, J. P. Waddell, and E. H. Schemitsch,“Femoral neck fracture following hip resurfacing: the effect of alignmentof the femoral component,” J. Bone Joint. Surg. Brit., vol. 90, no. 11,pp. 1522–1527, 2008.

[23] E. T. Davis, M. Olsen, R. Zdero, G. M. Smith, J. P. Waddell, andE. H. Schemitsch, “Predictors of femoral neck fracture following hipresurfacing: A cadaveric study,” J. Arthro., vol. 28, no. 1, pp. 110–116,2013.

[24] M. Olsen, E. T. Davis, C. M. Whyne, R. Zdero, and E. H. Schemitsch,“The biomechanical consequence of insufficient femoral component lat-eralization and exposed cancellous bone in hip resurfacing arthroplasty,” J. Biomech. Eng., vol. 132, no. 8, pp. 081011-1–081011-7, 2010.

[25] M. Olsen, M. Sellan, R. Zdero, J. P. Waddell, and E. H. Schemitsch, “Abiomechanical comparison of epiphyseal versus metaphyseal fixed bone-conserving hip arthroplasty,” J. Bone Joint Surg. Amer., vol. 93 (Suppl 2),pp. 122–127, 2011.

[26] G. Bergmann, G. Deuretzbacher, M. Heller, F. Graichen, A. Rohlmann,J. Strauss, and G. N. Duda, “Hip contact forces and gait patterns fromroutine activities,” J. Biomech., vol. 34, no. 7, pp. 859–871, 2001.

[27] M. Olsen, E. T. Davis, J. P. Waddell, and E. H. Schemitsch, “Imagelesscomputer navigation for placement of the femoral component in resur-facing arthroplasty of the hip,” J. Bone Joint. Surg. Brit., vol. 91, no. 3,pp. 310–315, 2009.

[28] R. Huiskes and B. Van Rietbergen, “Biomechanics of bone,” in BasicOrthopaedic Biomechanics and Mechano-Biology, 3rd ed. V. C. Mow andR. Huiskes, Eds. Philadelphia, PA, USA: Williams & Wilkins, 2005,pp. 123–179.

[29] M. Papini, R. Zdero, E. H. Schemitsch, and P. Zalzal, “The biomechan-ics of human femurs in axial and torsional loading: Comparison of fi-nite element analysis, human cadaveric femurs, and synthetic femurs,” J. Biomech. Eng., vol. 129, no. 1, pp. 12–19, 2007.

[30] Sawbones (2013, Mar. 14). [Online]. Available: http://www.sawbones.com/catalog/pdf/us_catalog.pdf

[31] L. Cristofolini, M. Viceconti, A. Cappello, and A. Toni, “Mechanicalvalidation of whole bone composite femur models,” J. Biomech., vol. 29,no. 4, pp. 525–535, 1996.

[32] R. Zdero, K. Elfallah, M. Olsen, and E. H. Schemitsch, “Cortical screwpurchase in synthetic and human femurs,” J. Biomech. Eng., vol. 131,no. 9, pp. 094503-1–094503-7, 2009.

[33] R. Zdero, M. Olsen, H. Bougherara, and E. H. Schemitsch, “Cancellousbone screw purchase: A comparison of synthetic femurs, human femurs,and finite element analysis,” Proc. Inst. Mech. Eng. H, J. Eng. Med. ,vol. 222, no. 8, pp. 1175–1183, 2008.

[34] B. Nicayenzi, S. Shah, E. H. Schemitsch, H. Bougherara, and R. Zdero,“The biomechanical effect of changes in cancellous bone density on syn-thetic femur behaviour,” Proc. Inst. Mech. Eng. H, J. Eng. Med., vol. 225,no. 11, pp. 1050–1060, 2011.

[35] S. Shah, H. Bougherara, E. H. Schemitsch, and R. Zdero, “Biomechanicalstress maps of an artificial femur obtained using a new infrared thermog-

raphy technique validated by strain gages,” Med. Eng. Phys., vol. 34,pp. 1496–1502, 2012.[36] C. Bitsakos, J. Kerner, I. Fisher, and A. A. Amis, “The effect of muscle

loading on the simulation of bone remodeling in the proximal femur,” J. Biomech., vol. 38, pp. 133–139, 2005.

[37] J. Stolk, N. Verdonschot, and R. Huiskes, “Hip-joint and abductor-muscleforces adequately represent in vivo loading of a cemented total hip recon-struction,” J. Biomech., vol. 34, pp. 917–926, 2001.

[38] H. Ebrahimi, M. Rabinovich, V. Vuleta, D. Zalcman, S. Shah, A Dubov,K. Roy, F. S. Siddiqui, E. H. Schemitsch, H. Bougherara, and R. Zdero,“Biomechanical properties of an intact, injured, repaired, and healed fe-mur: An experimental and computational study,” J. Mech. Behav. Biomed. Mater., 2012, 2012 Dec;16:121-35. doi: 10.1016/j.jmbbm.2012.09.005.2012 Sept 20 [Epub ahead of print].

[39] S. Shah, S. Y. R. Kim, A. Dubov, E. H. Schemitsch, H. Bougherara, andR. Zdero, “The biomechanics of plate fixation of periprosthetic femoralfractures near the tip of a total hip implant: Cables, screws, or both?”

Proc. Inst. Mech. Eng. H, J. Eng. Med., vol. 225, no. 9, pp. 845–856,2011.

8/12/2019 06475988



[40] A. Dubov, S. Y. R. Kim, S. Shah, E. H. Schemitsch, R. Zdero, andH. Bougherara, “The biomechanics of plate repair of periprosthetic fe-mur fractures near the tip of a total hip implant: The effect of cable-screwposition,” Proc. Inst. Mech. Eng. H, J. Eng. Med., vol. 225, no. 9, pp. 857–865, 2011.

[41] J. P. Lever, R. Zdero, M. T. Nousiainen, J. P. Waddell, andE. H. Schemitsch, “The biomechanical analysis of three plating fixationsystems for periprosthetic femoral fracture near the tip of a total hiparthroplasty,” J. Orthop. Surg. Res., vol. 5:45, 2010.

[42] R. Zdero, R. Walker, J. P. Waddell, and E. H. Schemitsch, “Biomechanicalevaluation of periprosthetic femoral fracture fixation,” J. Bone Joint Surg. Amer., vol. 90, no. 5, pp. 1068–1077, 2008.

[43] G. Bergmann, F. Graichen, and A. Rohlmann, “Hip joint loading duringwalking and running, measure in two patients,” J. Biomech., vol. 26, no. 8,pp. 969–990, 1993.

[44] G. N. Duda, M. Heller, J. Albinger, O. Schulz, E. Schneider, and L. Claes,“Influence of muscle forces on femoral strain distribution,” J. Biomech.,vol. 31, pp. 841–846, 1998.

[45] M. Talbot, R. Zdero, and E. H. Schemitsch, “Cyclic loading of peripros-thetic fracture fixation constructs,” J. Trauma, vol. 64, no. 5, pp. 1308–1312, 2008.

[46] X. G. Cheng, G. Lowet, S. Boonen, P. H. Nicholson, P. Brys, J. Nijs,and J. Dequeker, “Assessment of the strength of proximal femur in vitro:Relationship to femoral bone mineral density and femoral geometry,” Bone, vol. 20, no. 3, pp. 213–218, 1997.

[47] A. M. Pankovich and J. A. Elstrom, “Intracapsular fractures of the prox-imal femur,” in Handbook of Fractures, J. A. Elstrom, W. W. Virkus,and A. M. Pankovich, Eds. Toronto, ON, Canada: McGraw-Hill, 2006,pp. 264–280.

[48] M. M.Morlock, N. Bishop, W. Ruther, G. Delling, andM. Hahn, “Biome-chanical, morphological, and histological analysis of early failures in hipresurfacing arthroplasty,” Proc. Inst. Mech. Eng. H, J. Eng. Med., vol. 220,no. 2, pp. 333–344, 2006.

Zachary Morison received the M.Sc. degree fromthe Institute of Medical Science, University of Toronto, Toronto, ON, Canada.

He is currently a Clinical Research Assistant atthe Martin Orthopaedic Biomechanics Laboratory,St. Michael’s Hospital, Toronto.

Michael Olsen received the Ph.D. degree in me-chanical engineering from the University of Toronto,Toronto, ON, Canada, where he is currently workingtoward the M.D. degree.

He is currently with the Martin OrthopaedicBiomechanics Laboratory, St. Michael’s Hospital,Toronto, where his research involves the biome-chanics of joint replacements and fracture fixationmethods.

Gordon A. Higgins received the B.Sc., M.B.Ch.B.,M.R.C.S., and F.R.C.S.(Tr & Orth) degrees fromManchester University, St. Andrew’s University, andthe Royal College of Surgeons, all in the U.K.

He is a Consultant Orthopaedic Surgeon with theSouth Devon Foundation Hospital Trust, Torbay Dis-trict General Hospital, Torquay, U.K. He is a spe-cialist in lower limb joint replacement surgery withan interest in total and resurfacing hip replacements,partial and total knee replacements, and arthroscopic(keyhole) surgery.

Rad Zdero received the Ph.D. degree from Queen’sUniversity, Kingston, Canada.

He is the Director of the Martin OrthopaedicBiomechanics Laboratory, St. Michael’s Hospital,Toronto, ON, Canada, and an Adjunct Professor inthe Department of Mechanical and Industrial Engi-neering, Ryerson University, Toronto. His researchinterests include testing orthopaedic trauma devices,characterizing human tissue, and developing newbiomaterials.

Emil H. Schemitsch received the M.D. degree from

the University of Toronto, Toronto, ON, Canada, andthe F.R.C.S.C. degree from the Royal College of Sur-geons of Canada.

He is an internationally recognized orthopaedicsurgeon interested in developing new devices andtechniques for joint replacement and fracture fixa-tion. He is the past President of the Canadian Or-thopaedic Association, the past Head of the Divisionof Orthopaedic Surgery at St. Michael’s Hospital,Toronto, ON, Canada, and is a Professor of Surgery

at the University of Toronto, Toronto.

Documents

06475988