Contents lists available at ScienceDirect

Clinical Biomechanics

journal homepage: www.elsevier.com/locate/clinbiomech

Review

Periprosthetic fracture fixation of the femur following total hip arthroplasty:A review of biomechanical testing – Part IIKatherine Wanga, Eustathios Kenanidisa,b, Mark Miodownika, Eleftherios Tsiridisb,Mehran Moazena,⁎

a Department of Mechanical Engineering, University College London, Torrington Place, London WC1E 7JE, UKbAcademic Orthopaedics Department, Papageorgiou General Hospital & CORE Lab at CIRI AUTH, Aristotle University Medical School, University Campus 54 124,Thessaloniki, Greece

A R T I C L E I N F O

Keywords:Periprosthetic femoral fractureBiomechanicsComputational modelFixation method

A B S T R A C T

Background: Periprosthetic femoral fracture is a severe complication of total hip arthroplasty. A previous reviewpublished in 2011 summarised the biomechanical studies regarding periprosthetic femoral fracture and itsfixation techniques. Since then, there have been several commercially available fracture plates designed spe-cifically for the treatment of these fractures. However, several clinical studies still report failure of fixationtreatments used for these fractures.Methods: The current literature on biomechanical models of periprosthetic femoral fracture fixation since 2010to present is reviewed. The methodologies involved in the experimental and computational studies of peri-prosthetic femoral fracture fixation are described and compared with particular focus on the recent develop-ments.Findings: Several issues raised in the previous review paper have been addressed by current studies; such asvalidating computational results with experimental data. Current experimental studies are more sophisticated indesign. Computational studies have been useful in studying fixation methods or conditions (such as bonehealing) that are difficult to study in vivo or in vitro. However, a few issues still remain and are highlighted.Interpretation: The increased use of computational studies in investigating periprosthetic femoral fracture fixa-tion techniques has proven valuable. Existing protocols for testing periprosthetic femoral fracture fixation needto be standardised in order to make more direct and conclusive comparisons between studies. A consensus on the‘optimum’ treatment method for periprosthetic femoral fracture fixation needs to be achieved.

1. Introduction

Periprosthetic femoral fractures (PFF) is a severe complication fol-lowing total hip arthroplasty (THA); the rate of intraoperative PFFranged from 0.1–27.8% and of postoperative from 0.07–18%. PFF aremore frequent in uncemented than cemented both in primary and re-vision THA (e.g. Biggi et al., 2010; Dubov et al., 2011; Fleischman andChen, 2015; Kenanidis et al., 2018). PFF account for approximately 6%of revision cases and are the third most common reason for revisionsurgery after aseptic loosing and infection (e.g. Lewallen and Berry,1998; Lindahl et al., 2006; Marsland and Mears, 2012). This number isexpected to rise substantially by 2030, with the increase in life ex-pectancy of the general population also leading to a rising incidence oftotal hip arthroplasties (THAs), with PFF also expected to rise pro-portionally (Della Valle et al., 2010).

PFF can occur intra-operatively or post-operatively, creating avariety of different fracture configurations at different locations; manyresearchers classify PFF based on fracture type, position on the femur,and bone quality. The Vancouver classification system is the mostwidely used and accepted classification system for PFF (Duncan andMasri, 1995; Learmonth, 2004; Moazen et al., 2011). Fractures classi-fied as Type A are fractures involving the trochanteric area. The ma-jority (approximately 75%, − Lochab et al., 2017; Lever et al., 2010) ofPFF, however, are Type B; located around and just distal to the tip ofthe stem, and are subdivided as B1 with the stem stable and good bonestock, B2 with the stem unstable and good bone stock, and B3 with stemunstable and significant bone loss. Type C are fractures located distal tothe stem (Capone et al., 2017; Leonidou et al., 2013; Tsiridis et al.,2009).

These fractures can be challenging to manage and treat, and are

https://doi.org/10.1016/j.clinbiomech.2018.12.001Received 31 August 2018; Accepted 4 December 2018

⁎ Corresponding author at: Department of Mechanical Engineering, University College London, Torrington Place, London, WC1E 7JE, UKE-mail address: M.Moazen@ucl.ac.uk (M. Moazen).

Clinical Biomechanics 61 (2019) 144–162

0268-0033/ © 2018 Elsevier Ltd. All rights reserved.

T

most commonly found in osteopenic elderly women, or in patients whohave experienced loosening of the femoral stem following low energytrauma (Kenanidis et al., 2018; Shah et al., 2011). Given the complexnature of PFF treatment, due to the combination of the fractured boneand existing prosthesis (Moazen et al., 2011), many factors are requiredto be taken into consideration in the treatment of PFF; e.g., sex, age,bone quality, fracture topography, previous hip revision procedures,implant stability, and types (e.g. cemented vs. uncemented stem - DellaValle et al., 2010). The Unified Classification System (UCS); a recentlyproposed treatment algorithm developed by Duncan and Haddad(Duncan and Haddad, 2014), outlines the principles of PFF treatment.Treatment for Type A fractures is dependent on two factors; fracturedisplacement and the importance of soft tissue attached. Non-displacedType A fractures are typically non-operative and treated conservatively.In cases of displacement of the greater trochanter, surgical treatmenttypically uses cerclage wires or hook cable plates for fixation. In cases ofthe lesser trochanter, if the fracture compromises the stability of theimplant, cerclage wiring and implant revision may be considered (Biggiet al., 2010; Schwarzkopf et al., 2013). Management of Type B fracturesis determined by subtype. B1 fractures can be treated by reduction andfixation using minimally invasive plate osteosynthesis (MIPO). In B2fractures, revision surgery with a longer stem is commonly used. B3fractures require more complex reconstruction or salvage procedures(megaprosthesis, allograft/stem composite). Type C fractures can betreated as a non-periprosthetic fracture. Specialized techniques can beused in some cases if hardware required for fixation will extend towardsthe implant, such as cerclages and unicortical screws (Capone et al.,2017; Duncan and Haddad, 2014).

While the Vancouver classification determine the treatment for PFF,many clinical cases still report failure of femoral fracture fixation due tomismanagement; the misclassification of B1 and B2 fractures is themain reason for the greater reported failure of B fractures (Kenanidiset al., 2018). For example, up to 20% of loose stems are missed onpreoperative radiologic evaluation; many surgeons also fail to ade-quately test stem stability in the operating room leading to in-appropriate selection of surgical methods for treatment (Fleischmanand Chen, 2015; Niikura et al., 2014). This suggests that protocol forclassifying PFF and subsequent fixation method is still insufficient. In-deed the reliability of any classification system depends on inter-ob-server and intra-observer consistency (Rayan et al., 2008). Optimalmanagement of PFF remains controversial and debated, given thatadequate fixation needs to be achieved without compromising thestability of the hip prosthesis. Although PFF is a rare complication,understanding risk factors and optimum treatment for fixation is still ofhigh importance, as one study documented a higher risk of death afterPFF compared with a similar population of patients undergoing un-complicated THA (Della Rocca et al., 2011; Lindahl et al., 2007).

Finite element (FE) analysis is a computational modelling techniquethat allows prediction of the mechanical behaviour of structures. Usedfor orthopaedic biomechanics since the early 1970's it has been in-creasingly utilized by a number of authors to study structural-me-chanical problems such as stress and strain analysis of bone, joints, andload-bearing implants (Huiskes and Chao, 1983; Kluess et al., 2010).Computer modelling allows a large number of scenarios to be testedwith little extra cost per test making it advantageous over traditionalexperimental studies. To optimise management of PFF fixation, therehave been a number of computational studies dedicated to simulatingtheir biomechanics.

In 2011, Moazen et al. summarised the biomechanical research in-vestigating PFF fixation following THA and its treatment methods.However, since then, there has been a large influx of biomechanical andcomputational studies carried out, and this is the basis of this paper.The aim of this paper was to provide an updated review of currentresearch relating to PFF following THA published since 2011; currently,available literature pertinent to the biomechanical analysis of PFFtreatment methods will be examined. Results of the experimental and

computational studies conducted from 2010 to present and their trendswere evaluated. Results from this review were critically compared toprevious studies, highlighting any evolutions in biomechanical analysisof treatment methods for PFF.

2. Methodology

Computerised scientific journal databases, i.e. Scopus, GoogleScholar, PubMed, and Web of Science were searched with the followingkeywords: Biomechanical testing, analysis, Finite element analysis,computational modelling, periprosthetic femoral fractures, and totalhip arthroplasty. All studies from the above-mentioned searches werethen reviewed; studies were included if they met the following criteria:(1) English Language; (2) Biomechanical or computational studies ofPFF after THA (3) femoral fractures. Additionally, all studies prior to2010 were excluded as they were reviewed previously (Moazen et al.,2011). In total 39 articles were retrieved, with 30 experimental studiesand 9 computational studies. In order to maintain linearity and con-tinuation, this paper will follow the same format as the previous review.

2.1. Experimental methods

A total of 30 experimental studies were reviewed. In many of thepresent experimental studies, the basic methodology described byMoazen et al. (2011) remained the same. The previous paper high-lighted three specific aspects in the experimental methodologies; typeof specimen, loading protocol, and methods of measurement. Meth-odologies in respect to those three aspects typically remained the same,and in-depth details of these can be referred back to the previous re-view. For most of the studies, mechanical performance is compared bystabilizing a periprosthetic fracture in both a cadaveric or syntheticfemur, and different loading protocols are applied to the construct (seeTable 1).

2.1.1. Specimen type and repeatabilityDespite basic methodology remaining the same, several noteworthy

factors have emerged from the reviewed studies; in particular, currentstudies using cadaveric femora use a higher number of specimenscompared to previous studies; where typical sample size ranged from 5to 16 cadaveric specimens, compared to a sample size range of 10 [5pairs – (Konstantinidis et al., 2010)] to 24 (Lehmann et al., 2010; Lenzet al., 2014) cadaveric specimens. One exception to this is Lenz et al.(2013) who used 45 cadaveric 70 mm segments of femora. In somestudies, authors used the same femur to test different fracture scenarios;Ebrahimi et al. (2012) utilized a single synthetic femur to test experi-mentally and computationally model and mimic the same femur whileintact, after injury, repair, and healing. While most studies used bonemineral density matched cadaveric femora, to ensure no lesions or pre-existing fracture, Lehmann et al. (2010) used an osteoporotic bonemodel, to represent the group with the highest incidence of PFF. Whilemost cadaveric bones used were fresh frozen, two studies used em-balmed femora (Demos et al., 2012; Konstantinidis et al., 2010).

2.1.2. Representation of loads and surrounding conditionsIn respect to loading modes and surrounding conditions, higher

loading modes have been used by several authors. In previous studies(Moazen et al., 2011), only 500 N could be seen used repeatedly fornon-destructive monotonic tests; in present studies, loads of 700 N(Choi et al., 2010; Graham et al., 2015) to 2500 N (Pletka et al., 2011)have been used. A loading mode not seen in previous papers is four-point bending (Lenz et al., 2016a, 2016b; Lever et al., 2010; Lochabet al., 2017) and in one case three-point bending (Choi et al., 2010);examples of these can be seen in Fig. 1. The basic experimental setupseen in most of the experimental studies can be referred back to theprevious review (Moazen et al., 2011). There is little consensus seen onloading protocols; loads to failure was also not consistent across the

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

145

Table1

Asu

mm

ary

ofth

esp

ecim

enpr

epar

atio

nan

dlo

adin

gpr

otoc

olin

labo

rato

ryst

udie

s.

Aut

hors

Spec

imen

num

ber

and

type

Pros

thes

isFr

actu

reLo

adin

gFe

mur

posi

tion

(Leh

man

net

al.,

2010

)24

Cada

veri

c(6

per

grou

p)b

Cem

ente

d,Ex

eter

,Str

yker

How

med

ica

Ost

eoni

csO

bliq

ue45

°ost

eoto

my,

leve

latt

ipof

hip

stem

(onl

yfo

rgr

oup

IV).

Four

-poi

ntbe

ndin

g–

load

appl

ied

at0.

1m

m/s

until

frac

ture

.H

oriz

onta

lpos

ition

.

(Lev

eret

al.,

2010

)12

mat

ched

pair

sCa

dave

ric

(5te

stm

odes

,15

test

case

s.)b

Cem

ente

d,(C

ompa

nyno

tm

entio

ned)

Obl

ique

45°o

steo

tom

yA

xial

com

pres

sion

–lo

adof

250

Nap

plie

d(t

wo

type

stes

ted

–ab

duct

ion,

and

forw

ard

flexi

on)

Tors

ion

–25

0N

appl

ied

toan

teri

oras

pect

offe

mor

alhe

adFo

ur-p

oint

bend

ing

(2ty

pes

test

ed-a

nter

o-po

ster

ior

and

med

io-la

tera

lfor

ces)

–25

0N

appl

ied

sym

met

rica

llyon

eith

ersi

deof

oste

otom

ysi

te.

Axi

al-2

0°of

abdu

ctio

n,an

d20

°for

war

dfle

xion

Tors

ion

and

4-po

intb

endi

ng–

Hor

izon

talo

rien

tatio

nto

sim

ulat

e90

°offl

exio

n(C

hoie

tal

.,20

10)

10Sy

nthe

ticCe

men

ted,

Zim

mer

,War

saw

,IN

Tran

sver

seos

teot

omy,

20m

mfr

actu

rega

pdi

stal

totip

ofst

em.

Sinu

soid

alax

iall

oadi

ngof

50–7

00N

for

100

cycl

esat

2H

z.

Thre

e-po

int

bend

ing

–Ve

rtic

alsi

nuso

idal

load

sof

50N

–50

0N

at2

Hz

for

300

cycl

es.

Tors

ion

–in

crea

sing

sinu

soid

alto

rsio

nalm

ovem

ents

3N

/m–

12N

/map

plie

dat

0.5

Hz

for

20cy

cles

.

All

test

sre

peat

edth

ree

times

for

each

cons

truc

tm

odel

.

25°o

fadd

uctio

n

(Kon

stan

tinid

iset

al.,

2010

)5

pair

sCa

dave

ricc

Cem

ente

d,Bi

cont

act,

Aes

cula

p,Tu

ttlin

gen,

Ger

man

yTr

ansv

erse

,10

mm

frac

ture

gap

dist

alto

stem

tip.

Axi

alan

dcy

clic

com

pres

sion

–10

00N

for

10,0

00cy

cles

,th

enpr

ogre

ssiv

ely

load

edto

failu

reat

100

N/2

000

cycl

es.

9°of

addu

ctio

n

(Ple

tka

etal

.,20

11)

9m

atch

edpa

irs

Cada

veri

cbCe

men

ted,

Ulti

ma,

DeP

uy,W

arsa

w,

INTr

ansv

erse

.10

mm

dist

alto

stem

tip.

Sinu

soid

alcy

clic

load

ing

–up

to10

,000

cycl

esfr

om0

to25

00N

axia

lfor

ce.

0-15

Nm

ofto

rsio

nat

rate

of1

Hz.

Vert

ical

orie

ntat

iona

(Sha

het

al.,

2011

)3

Synt

hetic

Cem

ente

d,Ki

ngPa

ckag

edM

ater

ials

Co.,

ON

,Can

ada

Tran

sver

seat

tipof

stem

–5

mm

frac

ture

gap

near

tipof

stem

Dis

plac

emen

tcon

trol

,max

imum

vert

ical

load

of10

00N

axia

lfor

ceat

rate

of5

mm

/min

appl

ied.

15°o

fadd

uctio

n

(Dem

oset

al.,

2012

)24

Cada

veri

c(6

per

grou

p)c

Cem

ente

d,10

0m

mst

raig

htm

etal

carr

iage

bolt

used

inst

ead

ofhi

pst

em.

Obl

ique

45°,

20m

mfr

actu

rega

pdi

stal

tohi

pst

emA

xial

com

pres

sion

tofa

ilure

at5

mm

/s.

Vert

ical

orie

ntat

ion,

fem

oral

shaf

tco

lline

arto

axis

oflo

adin

ga

(Len

zet

al.,

2012

a)12

Cada

veri

cbCe

men

ted,

Char

nley

hip

endo

pros

thes

is,D

ePuy

IN45

°-10

mm

dist

alto

tipof

pros

thes

is.

Cycl

icA

xial

bend

ing

at2

Hz

with

sync

hron

alsi

nuso

idal

axia

lloa

ding

at95

0N

for

10,0

00cy

cles

–A

xial

forc

esra

nged

from

50N

-100

0N

.Afte

r10

,000

cycl

es,i

ncre

ased

load

atra

teof

0.1

N/C

ycle

until

cata

stro

phic

failu

rest

artin

gfr

om10

00N

.

12°V

algu

s

(Len

zet

al.,

2012

b)8

Synt

hetic

(2gr

oups

)U

ncem

ente

d,M

athy

s,Be

ttla

ch,

Switz

erla

nd.

90°t

rans

vers

eos

teot

omy,

5m

mdi

stal

totip

ofst

em.

Cycl

icte

stin

gat

3H

zat

cons

tant

ampl

itude

of18

00N

for

first

5000

cycl

es.

Mon

oton

ical

lyin

crea

sing

sinu

soid

allo

adat

rate

of60

mN

/cy

cle

until

failu

rest

artin

gfr

om20

00N

20°V

algu

s

(Ebr

ahim

ieta

l.,20

12)

1Sy

nthe

ticCe

men

ted,

Exet

er,S

tryk

er,N

J,U

SATr

ansv

erse

5m

mga

p,23

3m

mfr

omto

pof

cem

ent

pott

ing

cube

.A

xial

load

,at

max

imum

of15

00N

,at

rate

of10

0N

/s15

°ofa

dduc

tion

(Len

zet

al.,

2013

)45

Cada

veri

c(s

egm

ents

-5pe

rgr

oup)

bN

opr

osth

esis

Non

e-70

mm

leng

thfr

agm

ents

cut

from

the

diap

hysi

sof

the

fem

urw

ere

used

Axi

allo

adto

failu

reat

rate

of50

N/s

Tors

iona

ltes

ting

atra

teif

2.5

Nm

/sun

tilco

nstr

uctf

ailu

re.

Axi

al–

Vert

ical

orie

ntat

ion

Tors

iona

l–H

oriz

onta

lor

ient

atio

n.(W

ähne

rtet

al.,

2014

)9

pair

s,m

atch

edCa

dave

ric.

(9pe

rgr

oup)

b

Unc

emen

ted,

Allo

clas

sic,

Zim

mer

,Sw

itzer

land

.Pr

oxim

alho

rizo

ntal

cut

and

45°d

ista

lcut

5m

mbe

low

stem

tip.

Cycl

icsi

nuso

idal

axia

lloa

ding

star

ting

at75

0N

,inc

reas

edat

0.1

N/c

ycle

at2

Hz

until

cons

truc

tfa

ilure

.Ve

rtic

alor

ient

atio

na

(Gie

sing

eret

al.,

2014

)17

Synt

hetic

(2gr

oups

,9

inN

CBgr

oup

and

8in

cont

rol)

Cem

ente

d,CP

T,Zi

mm

er,I

NO

steo

tom

y20

mm

dist

alto

tipof

stem

,6m

mga

pne

arst

emtip

.A

xial

load

of10

0-40

0N

and

tors

iona

lloa

dof

1-4

Nm

appl

ied

at1.

5H

zfo

r20

,000

cycl

es-o

steo

tom

yga

pth

enfil

led

with

cem

entt

osi

mul

ate

‘hea

led’

frac

ture

.The

nA

xial

load

of10

0-14

00N

and

1–10

.8N

mto

rsio

nall

oad

appl

ied

for

80,0

00cy

cles

7°Va

lgus

(Bra

ndet

al.,

2014

)8

Synt

hetic

Cem

ente

d,Ec

ofit,

Impl

antc

ast,

Buxt

ehud

e,G

erm

any.

15m

mbe

low

tipof

stem

Axi

allo

adto

failu

re–

cons

tant

incr

easi

nglo

adap

plie

dw

itha

star

ting

forc

eof

0N

6°Va

lgus

(Len

zet

al.,

2014

)24

mat

ched

,Cad

aver

icb

Cem

ente

d,Ch

arnl

ey,D

ePuy

,IN

Valg

us

(continuedon

nextpage

)

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

146

Table1

(continued)

Aut

hors

Spec

imen

num

ber

and

type

Pros

thes

isFr

actu

reLo

adin

gFe

mur

posi

tion

60°-

10m

mfr

omst

emtip

–D

ista

lpor

tion

offe

mur

and

plat

eem

bedd

edin

PMM

A.

Axi

albe

ndin

g–

50N

to20

0N

atra

teof

30N

/s.

Cycl

icte

stin

gat

rate

of2

Hz,

sync

hron

alax

iall

oadi

ngw

ithco

nsta

ntva

lley

load

of20

0N

.10

00N

peak

load

leve

linc

reas

edat

rate

of0.

1N

/cyc

leun

tilca

tast

roph

icfa

ilure

(Hoff

man

net

al.,

2014

)15

med

ium

Synt

hetic

(5fo

rea

chte

st)

Unc

emen

ted

VerS

ys,Z

imm

er,I

NO

bliq

ue45

°to

shaf

taxi

sat

the

leve

lofi

mpl

antt

ip.

Axi

alco

mpr

essi

on-l

oade

dto

500

Nat

20N

/s

Late

ralB

endi

ng–

load

edto

250

Nat

10N

/s

Tors

ion/

Sagi

ttal

bend

ing

–lo

aded

to20

0N

at10

N/s

Axi

alcy

clic

load

ing

–50

-500

Nlo

adap

plie

dat

3H

zfo

r10

,000

cycl

es.A

fter

cycl

iclo

adin

gfe

mur

ste

sted

agai

nfo

ral

lthr

eem

odal

ities

then

load

edto

failu

reor

100

mm

disp

lace

men

tin

tors

iona

l/sa

gitt

albe

ndin

g

10°a

dduc

tion

infr

onta

lpla

ne.

Vert

ical

lyin

sagi

ttal

plan

e.

(Sar

iyilm

azet

al.,

2014

)15

larg

e,le

ftSy

nthe

tic(5

for

each

test

)U

ncem

ente

d,Sy

nerg

y,Sm

ith&

Nep

hew

,TN

10m

mfr

actu

rega

pat

leve

lofp

rost

hesi

stip

–(t

rans

vers

e)Cy

clic

rota

tiona

lloa

ding

10re

peat

edcy

lindr

ical

twis

tsat

3H

zbe

twee

n0.

5an

d10

Nm

for

10,0

00cy

cles

Cycl

icax

iall

oadi

ng–

forc

eco

ntro

l-50

N-5

00N

for

1000

cycl

elo

adin

gs,w

ith10

repe

titio

nsat

a3-

Hz.

Axi

alFa

ilure

–di

spla

cem

ent

cont

rol–

forc

eap

plie

dw

ithsp

eed

of15

mm

/min

until

failu

re.

15°V

algu

sfo

rcy

clic

axia

llo

adin

g.

(Gri

ffith

set

al.,

2015

)12

larg

e,le

ft,sy

nthe

tic(6

for

each

test

)Ce

men

ted,

Exet

erfe

mor

alst

em45

°obl

ique

-25

mm

dist

alto

tipof

stem

,one

grou

pha

dm

idsh

aft

oste

otom

y(M

O)

(ana

tom

ical

lyre

duce

d)an

dth

eot

her

mid

shaf

tga

p(M

G)

(with

5m

mga

p)

Axi

alco

mpr

essi

on,d

ispl

acem

entc

ontr

ol–

prel

oade

d10

0N

to10

00N

,ver

tical

load

appl

ied

-500

Nfo

rM

O,2

50N

for

MG

.La

tera

lben

ding

–20

0N

vert

ical

load

at8

mm

/min

Tors

iona

lstiff

ness

–ve

rtic

allo

adof

200

Nat

8m

m/m

inA

xial

load

tofa

ilure

–pr

eloa

dof

100

Nat

load

rate

of8

mm

/min

tillc

atas

trop

hic

failu

re

Axi

al-2

5°ad

duct

ion

toco

rona

lpl

ane,

alig

ned

vert

ical

lyin

sagi

ttal

plan

e.La

teri

al–

hori

zont

alTo

rsio

nal-

Hor

izon

tal

(Gra

ham

etal

.,20

15)

5sy

nthe

ticCe

men

ted,

Exet

er,S

tryk

erSA

,Sw

itzer

land

.4

fixed

asif

anat

omic

ally

redu

ced.

1w

ith10

mm

gap

Axi

allo

ad–

disp

lace

men

tcon

trol

5m

m/m

in,m

ax50

0N

0°,1

0°,a

nd20

°add

uctio

nfo

rno

gap

mod

el.

10°f

orga

pm

odel

(Gw

inne

ret

al.,

2015

)20

larg

e,le

ft,sy

nthe

ticU

ncem

ente

d,A

llocl

assi

c,Zi

mm

er,

Switz

erla

nd.

Tran

sver

secu

tand

45°d

ista

lcut

atle

velo

fim

plan

ttip

.With

10m

mga

p.Cy

clic

sinu

soid

alax

iall

oadi

ngst

artin

gat

30N

.Inc

reas

edby

300

Nev

ery

1000

cycl

es.

Vert

ical

orie

ntat

iona .

(Lew

iset

al.,

2015

)30

Synt

hetic

Cem

ente

d,Zi

mm

er,W

arsa

w,I

NTr

ansv

erse

,25

mm

dist

alto

pros

thes

istip

.Dis

tal

part

offe

mur

not

used

tosi

mul

ate

segm

enta

lbon

elo

ss.

Tors

iona

lint

erna

lrot

atio

n.20

prec

ondi

tioni

ngcy

cles

at10

0N

/1H

z.Th

enlo

adin

gra

teof

8m

m/m

inun

tilfa

ilure

.A

xial

load

ing

–Ph

ase

I:lo

adof

4m

m/m

inun

til12

00N

.Ph

ase

II:4

mm

/min

until

failu

re/7

500

N.

Tors

iona

l-11

°ofp

rost

hesi

san

teve

rsio

n.A

xial

-13°

addu

ctio

n

(Fri

sch

etal

.,20

15)

24sy

nthe

ticU

ncem

ente

d,Zi

mm

er,W

arsa

w,I

NFe

mor

alne

ckos

teot

omy

10m

mpr

oxim

alto

less

ertr

ocha

nter

.Lo

ngitu

dina

lfra

ctur

eex

tend

ing

127

mm

dist

ally

Axi

allo

adof

50N

pre-

load

follo

wed

bylo

adin

gra

teof

0.8

mm

/min

and

term

inat

edaf

ter

disp

lace

men

tof2

0m

m.

Tors

ion

–ro

tatio

nald

ispl

acem

ents

appl

ied

atra

teof

2.4°

/s,

rota

ted

thro

ugh

40°u

ntil

failu

re.

25°a

dduc

tion,

0°an

teve

rsio

n

(Len

zet

al.,

2016

a)12

cada

veri

c,pa

ired

,(6

for

each

test

)bCe

men

ted,

Char

nley

,DeP

uy,I

N10

mm

dist

alto

tipof

pros

thes

is.–

orth

ogon

alto

shaf

tax

isof

fem

ur.

Axi

allo

adin

gan

ddi

spla

cem

enta

t10

Hz

4-po

int

bend

ing

and

tors

ion

test

edw

ithdi

spla

cem

ent

cont

rola

t0.

5m

m/m

in.U

pto

250

Nap

plie

d.

Cycl

icte

stin

gto

failu

rew

ithax

ialc

ompr

essi

onfr

om50

Nto

load

plat

eau

of20

0N

at30

N/s

,inc

reas

edpe

aklo

adat

500

Nat

0.1

N/c

ycle

.

Cycl

icte

stin

g-1

2°va

lgus

and

12°a

ntev

ersi

on

(Moa

zen

etal

.,20

16)

12la

rge,

left

synt

hetic

Cem

ente

d,Zi

mm

er,S

ulze

r,Sw

itzer

land

20m

mbe

low

tipof

stem

.A

xial

load

ing

–up

to70

0N

10°a

dduc

tion

(Gor

don

etal

.,20

16)

20sy

nthe

tic(5

for

each

test

)1.

Unc

emen

ted,

shor

tst

em(1

0),

Ana

Nov

aSo

litär

,Im

plan

Tec,

Aus

tria

140

mm

spir

alfr

actu

re(1

00m

mpr

oxim

alto

40m

mdi

stal

ofst

em)

Sinu

soid

alcy

clic

load

ing

-50

N-5

00N

at2

Hz

Axi

alSt

iffne

ss–

stro

keco

ntro

lled

0.02

mm

/sup

to50

0N

6°ad

duct

ion

(continuedon

nextpage

)

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

147

Table1

(continued)

Aut

hors

Spec

imen

num

ber

and

type

Pros

thes

isFr

actu

reLo

adin

gFe

mur

posi

tion

2.U

ncem

ente

d,lo

ngre

visi

onst

em(1

0),M

odul

arPl

us,S

mith

and

Nep

hew

,Aus

tria

Cycl

icsi

nuso

idal

fatig

uelo

adin

g-20

00N

max

load

,in

crea

sing

by15

0N

/500

cycl

esun

tilfa

ilure

(Len

zet

al.,

2016

b)12

cada

veri

c,pa

ired

,(6

for

each

grou

p)b

Cem

ente

d,Ch

arnl

ey,D

ePuy

,IN

Tran

sver

se,

10m

mdi

stal

totip

ofst

em,o

rtho

gona

lto

fem

ursh

aft

axis

Axi

albe

ndin

gto

200

Nat

30N

/sCy

clic

mec

hani

calt

estin

gat

2H

zw

ithsy

nchr

onic

axia

llo

adin

gin

crea

sed

at0.

1N

/cyc

lest

artin

gfr

om50

0N

until

cata

stro

phic

failu

re.

4-po

int

bend

ing

–25

0N

max

12°v

algu

san

d12

°ant

ever

sion

(Wal

cher

etal

.,20

16)

38sy

nthe

ticCe

men

ted,

Web

erst

anda

rdst

raig

htst

em,Z

imm

er.

Non

eto

sim

ulat

ehe

aled

peri

pros

thet

icfr

actu

resi

tuat

ion.

Axi

alco

mpr

essi

onat

500

Nov

er5

s.Th

en30

cycl

esfr

om40

0N

–150

0N

at0.

25H

zap

plie

d.To

rsio

nalt

estin

gto

0.6

Nm

over

5s,

then

30cy

cles

ofex

tern

alro

tatio

nfr

om0.

6N

m–5

0N

mat

0.25

Hz.

Load

tofa

ilure

atco

nsta

ntdi

spla

cem

entr

ate

of10

0m

m/

min

inax

iall

oadi

ng.

7°va

lgus

(Wäh

nert

etal

.,20

17)

10Sy

nthe

tic(5

for

each

grou

p)U

ncem

ente

d,A

llocl

assi

c,Zi

mm

erG

mbH

,Sw

itzer

land

.45

°dis

talc

utan

dho

rizo

ntal

cut5

mm

dist

alto

stem

tip.

Cycl

iclo

adin

gin

axia

lcom

pres

sion

at2

Hz

until

failu

re–

star

ting

atpe

aklo

adof

750

Nw

ithin

crem

ento

f0.1

N/

cycl

e.

Vert

ical

orie

ntat

iona

(Kon

stan

tinid

iset

al.,

2017

)20

cada

veri

cCe

men

ted,

Bico

ntac

t,A

escu

lap

AG

,G

erm

any.

Tran

sver

sebe

low

tipof

stem

.Fl

uctu

atin

gax

iall

oad

(sin

usoi

dalp

rofil

e,0.

5H

z,21

00N

)ap

plie

dto

pros

thet

icco

ne,r

epea

ted

for

20,0

00lo

adcy

cles

.St

anda

rdad

duct

ion

posi

tion

(Loc

hab

etal

.,20

17)

9pa

irs

ofca

dave

ricb

Cem

ente

d,D

ePuy

Sum

mit,

DeP

uySy

nthe

s,W

arsa

w,I

N.

45°o

bliq

ueos

teot

omy

25m

mdi

stal

totip

ofst

em.

5m

mfr

actu

rega

p4-

poin

tben

ding

–ra

teof

8m

m/m

inw

ithlo

adup

to25

0N

Tors

ion

and

Axi

alco

mpr

essi

on–

vert

ical

forc

eof

250

NA

xial

com

pres

sion

tofa

ilure

orup

tom

axim

umve

rtic

aldi

spla

cem

ent

of10

mm

.

20°a

bduc

tion

and

20°fl

exio

n

aA

utho

rdi

dn't

spec

ifyfe

mor

alpo

sitio

n,bu

tfr

omth

eim

ages

prov

ided

,we

belie

vest

anda

rdad

duct

ion

vert

ical

posi

tioni

ngw

asus

ed.

bFr

esh

froz

enca

dave

ric.

cFo

rmal

infix

edca

dave

ric

fem

ora.

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

148

studies, an issue that was raised previously. Boundary conditions,magnitudes, and direction of loads applied varied between authors,seen in Table 1.

The majority of studies reviewed here studied the biomechanicalperformance of typical variations of an Ogden construct; specificallyexamining the performance of the plate fixation and its fixation methodto the femur via screws, cables, wires or in some cases struts. However,several new trends and parameters may affect the outcome of thefixation method examined across the studies published that was notinvestigated previously; including fracture gap, type of plate used, and

screws and cement mantle integrity. These will be described below withan overview of the materials and methods, and updated parametersused in the studies.

2.2. Overview of recent developments

2.2.1. Fracture configurationMost studies simulated a Vancouver B1 type fracture in their stu-

dies. Introduction of an osteotomy to simulate PFF was most commonlygenerated using a saw; although fracture position and configuration

Fig. 1. Schematic diagram of different examples of loading methods used in tests.A) 4-point bending (Medio-lateral) (Lever et al., 2010).B) 3-point bending (Choi et al., 2010).C-D) The embedded femoral shaft bone was connected to the actuator of the testing machine via a xy-table. Setup for axial loading (C) and lateral torsional loading(D) shown. (Lenz et al., 2013).E) Test set up of specimen positioned in 12° valgus for cyclic testing. Distal part of femur is potted in PMMA cement (Lenz et al., 2012a).

Fig. 2. Schematic diagram of different fracture gapvariations used in experimental methods.A) Fracture Gap (Choi et al., 2010; Giesinger et al.,2014; Graham et al., 2015; Griffiths et al., 2015;Gwinner et al., 2015; Konstantinidis et al., 2010;Lochab et al., 2017; Sariyilmaz et al., 2014; Shahet al., 2011).B) No gap (Brand et al., 2014; Frisch et al., 2015;Griffiths et al., 2015; Hoffmann et al., 2014;Konstantinidis et al., 2017; Lehmann et al., 2010;Lenz et al., 2012a, 2012b, 2016a; Lever et al., 2010;Pletka et al., 2011).C) Fracture gap filled with cement (Giesinger et al.,2014).D) Fracture gap with a wedge-like cut (Gwinneret al., 2015; Wähnert et al., 2014, 2017).

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

149

varies between the studies (Table 1). In many studies, no fracture gapwas left after the osteotomy, in order to simulate a stable fracturepattern (Brand et al., 2014; Frisch et al., 2015; Griffiths et al., 2015;Hoffmann et al., 2014; Konstantinidis et al., 2017; Lehmann et al.,2010; Lenz et al., 2012a, 2012b, 2016a; Lever et al., 2010; Pletka et al.,2011). Other studies implemented a fracture gap (where the femur wasnot fixed as if anatomically reduced, and a gap was left between thefracture), typically below the tip of the hip stem prosthesis; fracture gapimplemented ranged from 5 mm to 20 mm (Choi et al., 2010; Giesingeret al., 2014; Graham et al., 2015; Griffiths et al., 2015; Gwinner et al.,2015; Konstantinidis et al., 2010; Lochab et al., 2017; Sariyilmaz et al.,2014; Shah et al., 2011). Fracture gaps were typically used to mimic afragmented fracture model (Sariyilmaz et al., 2014). Wähnert et al.(2014, 2017) and Gwinner et al. (2015) created a 45° and horizontal cutas the osteotomy gap, and a triangular wedge segment was removed.The fixed fracture with a gap between the proximal and distal frag-ments eliminates the compressive effect of the fragments, isolating theproximal fixation during testing and simulating a “worst-case” scenariowith a comminuted fracture with no cortical apposition (Demos et al.,2012). See Fig. 2 for examples of different fracture gap configurations.

A few studies investigated the effect of fracture gap and no fracturegap (Giesinger et al., 2014; Graham et al., 2015; Griffiths et al., 2015);Giesinger et al. (2014) filled the osteotomy gap with cement aftercreating a fracture to simulate ‘healed’ fracture situation. In two stu-dies, no fracture was created to simulate a healed periprosthetic frac-ture situation (Walcher et al., 2016) or a femur prior to fracture(Ebrahimi et al., 2012). Some studies did not use the distal part of thefemur distal to the osteotomy; the femur and plate construct was cutaccordingly (Brand et al., 2014; Lenz et al., 2012b, 2013, 2014; Lewiset al., 2015).

2.2.2. Plate typeWith the recent interest in advancing strategies for PFF treatment,

specialized plates have been developed for PFF, commercialized, andused in recent studies published; these include hook plates, lockingcompression plates (LCP), Variable Angle Locking plate (VA-LCP),locking attachment plates (LAP), Dall-Miles plates, cable-ready system,and non-contact bridging plate. Currently, the two main periprostheticsystems on the market and most notably studied are the LockingCompression Plate (LCP –Synthes, Solothurn, Switzerland) and Non-Contact Bridging Periprosthetic Proximal Femur Plate (NCB PP-Zimmer GmbH, Winterthur, Switzerland). Most researchers used thesesystems in their studies, and a few were interested in the direct com-parison of different construct systems (Konstantinidis et al., 2010; Leveret al., 2010; Lewis et al., 2015; Wähnert et al., 2014). Some authorsinvestigated the effect of strut allografts in place of a fracture plate or afracture plate used in conjunction with a strut (Choi et al., 2010; Lochabet al., 2017; Sariyilmaz et al., 2014). The biomechanical performance ofusing two fracture plates on a single fracture (Fig. 3) was also in-vestigated by several authors (Choi et al., 2010; Lenz et al., 2016a;Wähnert et al., 2017).

Several authors also studied use of bicortical screws for proximalplate fixation; Lochab et al., 2017; Griffiths et al., 2015; Gwinner et al.,2015; Hoffmann et al., 2014; Konstantinidis et al., 2010; Lenz et al.,2012b, 2014, 2016a, 2016b; Lewis et al., 2015; Wähnert et al., 2014,2017). One recent commercial development and a method used toachieve proximal bicortical fixation was the locking attachment plate(LAP); a clamp-on plate that is compatible and can be used in con-junction with a conventional locking compression plate (LCP) in thetreatment of PFF; the lateral arms allows for bicortical offset screwplacement laterally to the prosthesis stem (Synthes, Solothurn, Swit-zerland) (Lenz et al., 2016b). The design of the NCB PP plate (ZimmerGmbH, Winterthur, Switzerland) also allows for proximal bicorticalscrew fixation. Fig. 3 shows examples of typical variations of the PFFfixation construct used.

2.2.3. Screws and cement mantleA clinical concern regarding the way that a construct fixation is

applied is the potential breach of cement mantle integrity; in particular,cortical screw tips infringing the cement mantle and potentially leadingto substantial cement fracture and eventual hip implant loosening(Lever et al., 2010). Two authors (Kampshoff et al., 2010;Konstantinidis et al., 2017) studied the role of cement mantle integrityand screws in PFF. Konstantinidis et al. (2017) deliberately made amore brittle mantle by using hand-mixed rather than the advised va-cuum mixed cement, and Kampshoff et al. (2010) forgoed typical platefixation setup and investigated the effect of different screw implanta-tion techniques by directly drilling different screws in the cement.Brand et al. (2014) proposed and investigated a novel fixation method –intraprosthetic fixation; where screws that fixed the fracture plate tothe bone were also drilled and fixed to the cemented hip implant. An-other important factor to note is that the risk of fractures is higheraround the uncemented compared to the cemented implants(Fleischman and Chen, 2015). This is perhaps due to the higher inter-aoperative risk of fracture for uncemented implants (Wyatt, 2014). Tobest of our knowledge eight studies so far have investigated bio-mechanics of PFF fixation in uncemented hip implants (Frisch et al.,2015; Gordon et al., 2016; Gwinner et al., 2015; Hoffmann et al., 2014;Lenz et al., 2012b; Sariyilmaz et al., 2014; Wähnert et al., 2014, 2017).

2.3. Computational methods

A total of nine computational studies were reviewed in this paper,and the following section will examine the computational method used.Prior to 2010, there were only two computational studies investigatingthe biomechanics of PFF fixation. The previous review paper (Moazenet al., 2011) highlighted three main aspects in the computationalmethodologies; 1) representation of the femoral bone and fracture, 2)representation of the loads and surrounding conditions in silico, and 3)simulation predictions and accuracy. In-depth detail of these meth-odologies can be referred back to the previous paper. Here, develop-ments to these three aspects described above are discussed, with therepresentation of the femoral construct instead of the femoral bonebeing highlighted, as well as current trends.

2.3.1. Representation of the femoral construct and accuracyThe increase in present computational capabilities allow for more

geometrically accurate modelling of individual parts of the construct.Computational representation ranged from a simplified parametric FEmodel of a typical construct (Leonidou et al., 2015; Moazen et al., 2012)to more geometrically accurate 3D models. (Avval et al., 2016; Chenet al., 2012; Ebrahimi et al., 2012; Moazen et al., 2013, 2014; Shahet al., 2011; Wang et al., 2016). A clinical case was modelled using asimplified parametric FE model of the PFF fixation construct (Moazenet al., 2012). The bone, hip stem, and cement mantle were modelled asconcentric cylinders. A simplified representation of a fracture fixationplate was used, and screws were modelled as cylinders with no screwthread or head. The model was validated against a clinical case study,suggesting that simplified models are sufficient when modelling dif-ferent construct configurations. Older computational studies generatedlow resolution meshes [928–2184 elements (Mann et al., 1997; Mihalkoet al., 1992)] in comparison to current computational capabilities[61000–400,000 elements (Chen et al., 2012; Ebrahimi et al., 2012;Leonidou et al., 2015; Moazen et al., 2012; Wang et al., 2016)]. Allstudies used tetrahedral elements to mesh components.

2.3.2. Representation of the loads and surrounding conditionsIn almost all studies, FE models assumed the femur had linear,

isotropic, and elastic properties. Studies performed by several currentauthors showed that linear behaviour was a good approximation forfemurs when comparisons of FEA, synthetic femurs, and human cada-veric femurs were made (Dubov et al., 2011). However, in many

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

150

studies, the bone quality that was simulated experimentally and com-putationally were considered normal healthy bone stock, and not os-teoporotic bone seen in PFF patients. Although Dubov et al. (2011)noted that relative performance of constructs would likely remain thesame.

2.4. Overview of recent developments

2.4.1. Fixation methodsClassical computational studies of PFF fixation (Mann et al., 1997;

Mihalko et al., 1992) investigated the effects of different stem lengths astreatment methods, although Mihalko et al. (1992) also studied theeffect of plate fixation. Recent studies investigated a wider range ofdifferent fixation methods, and also the effect of fracture stability, bonequality, and fracture type. Fixation methods in present studies can bedivided into two categories. The first category considers the effect ofdifferent plate fixations (Avval et al., 2016; Moazen et al., 2012, 2014;Moazen et al., 2013; Wang et al., 2016), typically direct comparisonsbetween two plate types are made; such as rigid vs. flexible plating(Moazen et al., 2012), comparisons between the performance of stain-less steel (SS) vs. titanium (Ti) plate fixations and plate thickness(Moazen et al., 2012, 2013), double cable fixation vs. locking plate vs.multi-directional plate (Wang et al., 2016), double plating (Moazenet al., 2014), and lateral vs. anterior plating (Avval et al., 2016). Platefixation and long stem revision options under partial and full weightbearing conditions were also carried out by one group (Moazen et al.,2014). The second category considers the biomechanical performanceof different variations of a typical Ogden construct; typically this in-volves different configurations of cable, wires, or screws positions(Chen et al., 2012; Dubov et al., 2011). Four studies modelled un-cemented hip implants in their studies (Avval et al., 2016; Chen et al.,2012; Moazen et al., 2012; Wang et al., 2016)

2.4.2. Effect of fracture stability, bone quality, and fracture typeWhile the majority of computational studies focused on Vancouver

B1 type fractures; there were several authors did investigate treatmentmethods for different fracture types (Leonidou et al., 2015; Moazenet al., 2012, 2014), in one instance a Vancouver type C clinical case

comparing initially failed fixation vs a successful revision fixation wascarried out (Moazen et al., 2012). Femoral fracture stability and bonequality was also computationally modelled by several authors (Avvalet al., 2016; Ebrahimi et al., 2012; Leonidou et al., 2015; Moazen et al.,2013); Ebrahimi et al. (2012) investigated the stiffness and peak bonestress of the same femur after injury, repair, and healing with respect toits intact condition. Stable vs unstable fracture on plate fixation per-formance was also investigated (Moazen et al., 2013).

Avval et al. (2016) studied femoral density changes and bone re-modelling in the femur in response to a bone fracture plate and un-cemented hip stem implant using a validated mechano-biochemicalmodel. Bone was hypothesized as a thermodynamic system that ex-changes energy, matter, and entropy with its surroundings. The modelthey used assumed that the mechanisms of bone remodelling are exe-cuted by bone resorption and bone formation phases through fivebiochemical reactions (i.e. formation of multinucleated osteoclasts, oldbone decomposition, production of osteoblast activator, osteoid pro-duction, and calcification.)

One study, by Leonidou et al. (2015) modelled an osteoporotic bonesituation by developing three models with different canal thicknessratios (CTR) to represent poor, average, and best bone quality. Furtherthree models were developed with angle fractures varying from un-stable transverse (0°), and short oblique (146 °) to the stable long ob-lique configuration (76 °). Additional three models were developed withfracture at the tip of the stem, 4 mm, and 14 mm below the tip of thestem.

3. Results

Key results of the experimental and computational cases studied aresummarised in Table 2. Several studies using computational methodswere validated with experimental results (Dubov et al., 2011; Ebrahimiet al., 2012; Lenz et al., 2013; Moazen et al., 2013; Shah et al., 2011).The issue of lack of standardization between tests seen in past studiesstill exists, making it difficult to make direct comparisons. Most tests,like those seen in previous studies, show that increasing the overallrigidity of the construct increases the stability of the fracture. Rigiditywas measured by the overall stiffness of the instrumented femur or by

Fig. 3. Schematic diagram of different plate fixation methods onto a femoral construct with a hip stem.A-C: Schematic diagram of a typical Ogden construct (A) and a construct with an additional plate fixed with wires (B) or with screws (C). (Choi et al., 2010).D-E: Schematic diagram showing a construct with an additional LAP plate attached proximally to the plate (D), and with an additional LCP plate placed anteriorly(E). (Lenz et al., 2016a).

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

151

Table2

Asu

mm

ary

offix

atio

nm

etho

dan

dre

sults

ofth

ecu

rren

tla

bora

tory

and

com

puta

tiona

lstu

dies

inve

stig

atin

gbi

omec

hani

csof

the

peri

pros

thet

icfe

mor

alfr

actu

refix

atio

n.

Aut

hors

Test

case

Resu

lts

Expe

rim

enta

lstu

dies

Plat

ean

dst

rutfi

xatio

n

Late

ralp

late

fixat

ion

Stru

tfixa

tion

Prox

imal

Dis

tal

Posi

tion

Stru

tlen

gth

(mm

)Pr

oxim

alD

ista

l

Uni

cort

ical

Scre

wCa

ble/

wir

eBi

cort

ical

scre

wCa

ble/

wir

eCa

ble/

wir

e

(Leh

man

net

al.,

2010

)(a

)- 3 3

- - - -

- (b)

3(a

)(b

)3(

a)(b

)

- - - -

- - - -

- - - -

- - - -

Two

intr

amed

ulla

ryim

plan

tsin

fem

urw

ere

asso

ciat

edw

ithde

crea

sed

frac

ture

stre

ngth

betw

een

thes

eim

plan

ts.F

ract

ure

plat

ebe

twee

ntip

ofth

est

ems

lead

sto

good

stab

ility

rega

rdle

ssof

pres

ence

ofos

teot

omy

orre

trog

rade

naili

ng.

(Lev

eret

al.,

2010

)- - - 3 3 3

4C 4W

4C - - -

4 4 4 4 4 4

- - - - - -

- - - - - -

- - - - - -

- - - - - -

Scre

w-p

late

syst

emsp

rovi

ded

eith

ergr

eate

rore

qual

stiff

ness

com

pare

dto

cabl

e-pl

ates

inal

mos

tal

lcas

es.N

ost

atis

tical

diffe

renc

esbe

twee

nth

eth

ree

diffe

rent

plat

ing

syst

ems

used

toco

mpa

reca

ble

vssc

rew

fixat

ion

-Zi

mm

erCa

ble

read

ysy

stem

(Zim

mer

,IN,U

SA),

AO

cabl

e-pl

ate

syst

em(S

ynth

es,P

A,U

SA),

and

Dal

l-Mile

sca

ble

grip

syst

em(H

owm

edic

a,N

J,U

SA).

(Cho

iet

al.,

2010

)2 2 2/

2(c

)

2 - 2

4 4 4/4c

- Ant

-

- 188

-

- 2C -

- 2C -

Fixa

tion

usin

gdo

uble

plat

essh

owhi

ghes

tst

iffne

ss,h

owev

erre

sults

dem

onst

rate

dth

atus

eof

addi

tiona

lallo

graf

tst

rut

inco

njun

ctio

nw

itha

LCP

also

prov

ided

supe

rior

stiff

ness

com

pare

dto

sing

lelo

cked

plat

e(L

CP–S

ynth

es)

for

Vanc

ouve

rty

peB1

fem

oral

frac

ture

s.(K

onst

antin

idis

etal

.,20

10)

4(d

-5)

4BC

(e-B

C)- -

3 3- -

- -- -

- -Bi

cort

ical

scre

wpl

acem

ent

(NCB

plat

e;Zi

mm

er,I

N,U

SA)

show

edsu

peri

oran

dm

ore

stab

lean

chor

ing

com

pare

dto

unic

ortic

alsc

rew

fixat

ion

(LIS

Spl

ate;

Synt

hes,

Switz

erla

nd).

Mea

nfo

rce

resu

lting

insu

bseq

uent

mod

elfa

ilure

sim

ilar

inbo

thm

odel

s.Su

gges

ting

NCB

plat

ew

asno

tsup

erio

rto

the

LISS

plat

e;m

oreo

verN

CBsy

stem

show

edm

ater

ial

fatig

ueun

der

cycl

iclo

adin

g,su

gges

ting

incr

ease

dim

plan

tfa

ilure

rate

spa

rtic

ular

lyin

case

sof

dela

yed

bony

unio

n.(P

letk

aet

al.,

2011

)3 3

2 24 4

- -- -

- -- -

Type

ofpl

ate

and

wor

king

leng

thdi

dno

tsig

nific

antly

affec

tfai

lure

rate

,no

sign

ifica

ntdi

ffere

nces

was

foun

dbe

twee

nlo

ngan

dsh

ort

plat

esfo

rdi

spla

cem

ento

rro

tatio

nat

frac

ture

site

.Low

erbo

nem

iner

alde

nsity

sign

ifica

ntly

asso

ciat

edw

ithfa

ilure

.(S

hah

etal

.,20

11)

- 4 4

4C - 4C

4 4 4

- - -

- - -

- - -

- - -

Cabl

esab

sorb

edm

ajor

ityof

load

,fol

low

edby

plat

esan

dth

ensc

rew

s.O

ptim

alm

echa

nica

lsta

bilit

yca

nbe

achi

eved

usin

gca

bles

and

scre

ws,

then

scre

ws

–as

both

had

the

high

est

stiff

ness

es.I

fonl

yca

bles

are

used

clin

ical

ly,a

plat

ew

ithou

tpr

oxim

alho

les

reco

mm

ende

d.(D

emos

etal

.,20

12)

3(L

S)3

(LS)

3 -

- 3 3 3

4 4 4 4

- - - -

- - - -

- - - -

- - - -

Prox

imal

cabl

efix

atio

npr

ovid

essi

gnifi

cant

lyle

ssax

ials

tabi

lity

com

pare

dto

whe

nca

bles

and

scre

ws

wer

eus

ed.L

ocki

ngan

dno

n-lo

ckin

gsc

rew

cons

truc

tssh

owed

equi

vale

ntlo

ads

atfa

ilure

,and

supe

rior

inlo

adat

failu

reco

mpa

red

toca

bles

.(L

enz

etal

.,20

12a)

5 2+

3(e

-BC)

- -3 3

- -- -

- -- -

A-L

CP(p

roto

type

lock

ing

plat

e)w

ithpr

oxim

albi

cort

ical

and

unic

ortic

alsc

rew

fixat

ion

had

high

ernu

mbe

rof

cycl

esto

failu

reco

mpa

red

toco

nven

tiona

lLCP

usin

gpr

oxim

alun

icor

tical

scre

wfix

atio

n,an

dsh

owed

high

erco

nstr

ucts

tabi

lity

and

stre

ngth

.Bic

ortic

alsc

rew

posi

tioni

ngsh

owed

less

inte

rfra

gmen

tary

oste

otom

ym

ovem

ent,

sugg

estin

gim

prov

edos

teos

ynth

esis

inpe

ripr

osth

etic

frac

ture

s.(L

enz

etal

.,20

12b)

3 3(f

-2BC

)1

Ce-

- -- -

- -- -

- -LA

P-LC

Pco

nstr

uct

grou

pus

ing

addi

tiona

lpro

xim

albi

cort

ical

scre

wfix

atio

nha

dsi

gnifi

cant

lyhi

gher

stiff

ness

and

num

ber

ofcy

cles

tofa

ilure

com

pare

dto

cerc

lage

-LCP

cons

truc

t.U

seof

LAP

and

plac

ing

bico

rtic

allo

ckin

gsc

rew

sla

tera

llyat

pros

thes

isst

emca

nim

prov

est

abili

tyin

PFF

fixat

ion.

(continuedon

nextpage

)

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

152

Table2

(continued)

Aut

hors

Test

case

Resu

lts

Expe

rim

enta

lstu

dies

Plat

ean

dst

rutfi

xatio

n

Late

ralp

late

fixat

ion

Stru

tfixa

tion

Prox

imal

Dis

tal

Posi

tion

Stru

tlen

gth

(mm

)Pr

oxim

alD

ista

l

Uni

cort

ical

Scre

wCa

ble/

wir

eBi

cort

ical

scre

wCa

ble/

wir

eCa

ble/

wir

e

(Len

zet

al.,

2013

)(k

)- 1 -

1Ce

- -

- - 1

- - -

- - -

- - -

- - -

Both

scre

wfix

atio

nty

pes(

Uni

cort

ical

and

bico

rtic

al)s

how

edsi

gnifi

cant

lyhi

gher

ultim

ate

stre

ngth

and

stiff

ness

inax

ialc

ompr

essi

onan

dto

rsio

nco

mpa

red

toce

rcla

gefix

atio

n.Re

sults

ofm

echa

nica

ltes

tw

ere

visu

ally

confi

rmed

byFE

Afo

run

icor

tical

and

bico

rtic

alsc

rew

s.(W

ähne

rtet

al.,

2014

)2

(f–

2BC

)4

(g)

- -3 3

- -- -

- -- -

Both

fixat

ion

syst

ems

achi

eved

prox

imal

bico

rtic

alsc

rew

fixat

ion

arou

ndth

ehi

pst

em.L

AP-

LCP

cons

truc

tfo

und

less

stab

ledu

eto

less

rigi

dm

ain

plat

e.N

CBpl

ate

show

edsi

gnifi

cant

lyhi

gher

stiff

ness

and

cycl

esto

failu

re.

(Gie

sing

eret

al.,

2014

)4

–3

––

––

No

stat

istic

ally

sign

ifica

ntdi

ffere

nces

inax

ialn

orin

med

ial(

Varu

s)st

emm

igra

tion

com

pare

dto

aco

ntro

lgro

up.L

ocki

ngpl

ate

fixat

ion

ofa

PFF

with

stab

lece

men

ted

pros

thes

isdi

dno

tlea

dto

cem

ent

man

tlefa

ilure

.(B

rand

etal

.,20

14)

3 2(h

)- -

1 1- -

- -- -

- -In

trap

rost

hetic

fixat

ion

prov

ided

sign

ifica

ntly

high

erfa

ilure

load

sco

mpa

red

toun

icor

tical

lock

ed-s

crew

plat

ing.

Sign

ifica

ntin

crea

sein

prim

ary

stab

ility

with

out

wea

keni

ngth

eim

plan

t-cem

ent-f

emur

-mod

elth

atco

uld

lead

toea

rly

wei

ght-b

eari

ngpa

tient

mob

iliza

tion.

(Len

zet

al.,

2014

)3 - 4 3(

f-2

BC)

1Ce

4Ce

- -

2 2 2 2

- - - -

- - - -

- - - -

- - - -

Prox

imal

bico

rtic

alfix

atio

nus

ing

LAP-

LCP

cons

truc

tim

prov

esst

abili

tyof

prox

imal

plat

efix

atio

nin

Peri

pros

thet

icfr

actu

res.

Cerc

lage

cabl

e-s

crew

com

bina

tion

isva

luab

leal

tern

ativ

e,es

peci

ally

inos

teop

orot

icbo

ne.

Cerc

lage

ssh

ould

beus

edin

com

bina

tion

with

atle

ast

one

addi

tiona

lsc

rew

toac

hiev

est

able

fixat

ion.

(Hoff

man

net

al.,

2014

)6

(e-B

C)4 -

- 1W

3C

3 3 3

- - -

- - -

- - -

- - -

Prox

imal

bico

rtic

alsc

rew

plac

emen

tach

ieve

dm

axim

allo

adto

failu

rean

dm

axim

alto

rsio

nal/

sagi

ttal

bend

ing

stiff

ness

.Add

ition

ofun

icor

tical

scre

ws

incr

ease

dax

ials

tiffne

ssw

hen

cabl

efix

atio

nus

ed.L

ater

albe

ndin

gno

taff

ecte

dby

diffe

renc

esin

prox

imal

fixat

ion.

(Sar

iyilm

azet

al.,

2014

)2 2 2

2C - -

4 2 2

- Med

Ant

- 150

150

- 2C 2C

- 2C 2C

Med

ials

trut

allo

graf

tw

ithpl

ate

fixat

ion

show

edhi

ghes

tst

iffne

ssan

dfa

ilure

load

valu

esan

dle

astd

ispl

acem

enta

tfra

ctur

esi

te.S

ugge

stin

git

ism

echa

nica

llysu

peri

orm

etho

din

B1ty

pePF

Ffix

atio

ntr

eatm

entn

ear

tipof

THA

(Gri

ffith

set

al.,

2015

)5 2

(f–

4BC

)2C -

5 5- -

- -- -

- -LA

P-LC

Pco

nstr

uct

sign

ifica

ntly

stiff

erth

anca

ble

cons

truc

tund

erax

ial

load

with

bone

gap.

Offe

rsbe

tter

axia

lstiff

ness

com

pare

dto

cabl

eco

nstr

uct.

(Gra

ham

etal

.,20

15)

3 3 -

- 3C 3C

3 3 3

- - -

- - -

- - -

- - -

Uni

cort

ical

scre

ws

show

edst

iffes

tfo

rmof

fixat

ion

atal

lloa

ding

angl

es.

Sugg

ests

that

cabl

esre

sult

inth

epr

oxim

alsc

rew

sbe

ing

push

edin

toth

ebo

neas

itis

appl

ied,

caus

ing

scre

ws

tolo

osen

thei

rfix

atio

nto

bone

.Fr

actu

rega

pan

dno

gap

mod

elbe

have

diffe

rent

ly–

degr

eeof

frac

ture

redu

ctio

naff

ects

who

leco

nstr

ucts

tabi

lity

and

bend

ing

beha

viou

rof

fixat

ion.

(Gw

inne

ret

al.,

2015

)4 3

(e-B

C)- -

5 5- -

- -- -

- -Fa

ilure

mod

ein

unic

ortic

algr

oup

char

acte

rise

dby

scre

wpu

ll-ou

twith

noad

ditio

nalb

one

frac

ture

orfis

sure

.Pro

xim

albi

cort

ical

scre

wfix

atio

nsh

owed

nosc

rew

pull-

out,

and

had

high

ercy

cles

tofa

ilure

.Bic

ortic

algr

oup

also

show

edsi

gnifi

cant

supe

rior

ityof

scre

wpu

rcha

seco

mpa

red

toun

icor

tical

scre

ws.

How

ever

,mod

eof

failu

rere

sulte

din

seve

reco

mm

uted

frac

ture

patt

erns

com

pare

dto

the

unic

ortic

alsc

rew

s,w

hich

only

resu

lted

insc

rew

pull-

out.

(Lew

iset

al.,

2015

)- 4

LS4

LS

3Ce

- 2C

- - -

- - -

- - -

- - -

- - -

Prox

imal

bico

rtic

alsc

rew

fixat

ion

cons

truc

ts(L

AP

+SS

LCP,

Synt

hes,

PA,

USA

,and

TiN

CB,Z

imm

er,I

N,U

SA)

show

edhi

gher

max

imum

forc

esin

tors

iona

lloa

ding

com

pare

dto

cabl

e,un

icor

tical

lock

ing

scre

ws

(LS)

,and

(continuedon

nextpage

)

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

153

Table2

(continued)

Aut

hors

Test

case

Resu

lts

Expe

rim

enta

lstu

dies

Plat

ean

dst

rutfi

xatio

n

Late

ralp

late

fixat

ion

Stru

tfixa

tion

Prox

imal

Dis

tal

Posi

tion

Stru

tlen

gth

(mm

)Pr

oxim

alD

ista

l

Uni

cort

ical

Scre

wCa

ble/

wir

eBi

cort

ical

scre

wCa

ble/

wir

eCa

ble/

wir

e

4(2

f-4

BCea

ch)

(j)6

(e)

- -- -

- -- -

- -- -

cabl

e+

unic

ortic

alLS

cons

truc

ts.C

able

cons

truc

tssh

owed

the

low

est

max

imum

forc

es,i

nbo

thax

iala

ndto

rsio

nall

oadi

ng.

Bico

rtic

alTi

NCB

cons

truc

tsho

wed

high

erst

iffne

ssth

anth

ebi

cort

ical

SSLA

P-LC

Pco

nstr

uct

inax

iall

oadi

ng.

(Len

zet

al.,

2016

a)3

(f−

2BC

)/4(

i)3

(2f−

2BC

each

)(j)

- -

2/

2

2

- -

- -

- -

- -

LAP

-Dou

ble

LCP

plat

e(O

rtho

gona

l)co

nstr

uct

fixat

ion

show

edsi

gnifi

cant

lyhi

gher

stiff

ness

,cyc

les,

and

load

tofa

ilure

com

pare

dto

LAP

(x2)

-sin

gle

LCP

plat

eco

nstr

uct.

Add

ition

allo

ckin

gpl

ate

enha

nces

cons

truc

tsta

bilit

yan

din

crea

ses

cons

truc

tst

iffne

ssco

mpa

red

tosi

ngle

plat

efix

edw

ithtw

oLA

P.(L

enz

etal

.,20

16b)

2(G

T),3

3(f

–2

BC)

- -2 2

- -

- -

- -

- -

Hoo

kco

nstr

ucts

how

edsi

gnifi

cant

lylo

wer

cycl

esan

dlo

adto

failu

rean

dfix

atio

nst

reng

thco

mpa

red

toLA

P-LC

Pco

nstr

uct.

Plat

est

iffne

ssbe

twee

nth

etw

ogr

oups

wer

eco

mpa

rabl

ein

rang

e.U

seof

subt

roch

ante

rica

lbi

cort

ical

scre

wfix

atio

nis

aneff

ectiv

efix

atio

nm

etho

din

PPF

than

hook

plat

e,an

dis

less

influ

ence

dby

bone

stoc

kqu

ality

.Sug

gest

sth

atho

okpl

ate

isre

serv

edfo

rPPF

that

requ

ires

stab

iliza

tion

ofgr

eate

rtro

chan

tera

sit

ishi

ghly

BMD

depe

nden

t.(M

oaze

net

al.,

2016

)6

(e)

6(e

)- -

4 4- -

- -- -

- -Pr

oxim

albi

cort

ical

scre

wfix

atio

nus

ing

far

cort

ical

lock

ing

scre

ws

can

redu

ceov

eral

leffe

ctiv

est

iffne

ssof

lock

ing

plat

esan

din

crea

sefr

actu

rem

ovem

ent

whi

lem

aint

aini

ngov

eral

lstr

engt

hof

PFF

fixat

ion

cons

truc

tco

mpa

red

tobi

cort

ical

scre

wfix

atio

nus

ing

lock

ing

scre

ws.

Inun

stab

lefr

actu

res

alte

rnat

ive

fixat

ion

met

hods

such

aslo

ngst

emre

visi

onm

aybe

bett

er.

(Wäh

nert

etal

.,20

17)

2(f

-2BC

)2/

2(o)

- -3 3/

2(o)

- -- -

- -- -

Cons

truc

tstiff

ness

and

cycl

esto

failu

resi

gnifi

cant

lyhi

gher

indo

uble

-pl

ate

cons

truc

tco

mpa

red

toLC

P-LA

Pco

nstr

uct.

(Loc

hab

etal

.,20

17)

3 4(2

f)(l

)2C -

4,2C

(m)

4A

nt-

200

-2

Ce(n

)-

2C -LC

P-A

llogr

aft

cons

truc

tdem

onst

rate

dhi

gher

stiff

ness

valu

esin

com

pres

sive

abdu

ctio

n,to

rsio

n,an

dm

edia

l-lat

eral

four

-poi

ntbe

ndin

gco

mpa

red

toth

eLA

P-LC

Pco

nstr

uct.

No

diffe

renc

esid

entifi

edbe

twee

nth

etw

oco

nstr

ucts

inco

mpr

essi

vefle

xion

,ant

erio

r-po

ster

ior

bend

ing

orlo

adto

failu

rete

sts.

Long

stem

vssh

ort

stem

(Gor

don

etal

.,20

16)

Com

pari

son

of4

grou

ps,s

hort

stem

sve

rsus

long

stem

sfo

rth

eir

effec

tiven

ess,

and

lock

ing

plat

efix

atio

nve

rsus

cerc

lage

syst

em:

1–

Long

stem

/Cer

clag

e–(4

titan

ium

cerc

lage

band

san

d2

stab

ilize

rs)

2–

Long

stem

/Pla

te–(

NCB

,5pr

oxim

alun

icor

tical

scre

ws

and

4di

stal

bico

rtic

alsc

rew

s)3

–Sh

ort

stem

/Cer

clag

e–

(4tit

aniu

mce

rcla

geba

nds

and

2st

abili

zers

)4

–Sh

ort

stem

/Pla

te–(

NCB

,5pr

oxim

alun

icor

tical

scre

ws

and

4di

stal

bico

rtic

alsc

rew

s)

Resu

ltsin

dica

teth

atfo

rVan

couv

erB1

frac

ture

s,os

teos

ynth

esis

with

plat

efix

atio

nha

sno

biom

echa

nica

ladv

anta

ges

over

use

ofsi

mpl

ece

rcla

gesy

stem

–ce

rcla

geco

nstr

ucts

dem

onst

rate

dla

rger

stiff

ness

,lar

gers

tren

gth,

and

mor

ecy

cles

tofa

ilure

com

pare

dto

plat

eco

nstr

uct.

Revi

sion

with

alo

ngst

empr

ovid

essu

peri

orm

echa

nica

lsta

bilit

yre

gard

less

ofty

peof

oste

osyn

thes

isfix

atio

n,th

ussu

itabl

efo

rVa

ncou

ver

B1fr

actu

retr

eatm

ent.

Insh

orts

tem

incr

ease

dsu

bsid

ence

isse

enin

cerc

lage

syst

emco

mpa

red

topl

atin

g.

Plat

ean

dst

emdi

stan

ce(W

alch

eret

al.,

2016

)Bi

omec

hani

calp

erfo

rman

ceto

esta

blis

hsa

fedi

stan

ceof

plat

efr

omtip

offe

mor

alpr

osth

esis

.–A

mou

ntof

plat

eto

stem

over

lap

orw

heth

erth

ere

isa

safe

gap

betw

een

the

stem

and

the

plat

een

dto

redu

ceri

skof

futu

refr

actu

res.

All

NCB

dist

alpl

ates

wer

eat

tach

edto

the

fem

urat

ade

fined

dist

ance

from

the

stem

toth

epl

ate

atva

ryin

gga

psfr

om80

mm

gap

to60

mm

over

lap,

in20

mm

incr

emen

ts.

40m

mga

p–

40m

mov

erla

pco

nsid

ered

clos

egr

oup,

and

grea

ter

that

40m

mov

erla

por

dist

ance

cons

ider

edfa

rgr

oup.

Stra

inin

crea

sed

with

the

decr

ease

dov

erla

por

gap.

All

earl

yfa

ilure

soc

curr

edbe

twee

n20

mm

over

lap

and

gap.

Sign

ifica

ntly

less

stra

inin

the

far

grou

pin

both

axia

land

tors

iona

lloa

ding

.Su

gges

tsth

atre

sults

can

aid

orth

opae

dic

surg

eons

inpl

ate

posi

tioni

ngin

Vanc

ouve

rtyp

e-C

PFF

fixat

ion.

Redu

ctio

nin

post

-ope

rativ

eco

mpl

icat

ions

(continuedon

nextpage

)

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

154

Table2

(continued)

Aut

hors

Test

case

Resu

lts

Expe

rim

enta

lstu

dies

Plat

ean

dst

rutfi

xatio

n

Late

ralp

late

fixat

ion

Stru

tfixa

tion

Prox

imal

Dis

tal

Posi

tion

Stru

tlen

gth

(mm

)Pr

oxim

alD

ista

l

Uni

cort

ical

Scre

wCa

ble/

wir

eBi

cort

ical

scre

wCa

ble/

wir

eCa

ble/

wir

e

Dis

talp

late

swer

efix

edus

ing

2bi

cort

ical

scre

wsa

tthe

mos

tpro

xim

alsc

rew

hole

sand

2bi

cort

ical

scre

wsi

nth

efo

urth

and

fifth

hole

s.N

ofe

mor

alfr

actu

rew

asap

plie

dto

sim

ulat

esi

tuat

ion

ofhe

aled

peri

pros

thet

icfr

actu

rew

ithim

plan

tsst

illin

situ

.by

posi

tioni

ngth

epl

ate

ina

man

ner

that

may

redu

cest

ress

rise

rsth

atco

uld

lead

tofu

ture

frac

ture

s.

Cabl

es(F

risc

het

al.,

2015

)Bi

omec

hani

calr

espo

nse

ofce

rcla

gesy

stem

sin

fixat

ion

ofin

trao

pera

tive

PFF

ince

men

tless

THA

.Fou

rco

nstr

ucts

com

pare

d,1)

CoCr

(Cob

alt-C

hrom

e)ca

ble

2)H

ose

clam

p3)

Mon

ofila

men

tw

ire

4)Sy

nthe

ticca

ble.

No

plat

efix

atio

nus

ed.

Met

allic

cons

truc

tsw

ithpo

sitiv

elo

ckin

gsy

stem

perf

orm

edbe

st,

supp

ortin

ghi

ghes

tloa

dsw

ithm

inim

alim

plan

tsub

side

nce

(bot

hax

iala

ndan

gula

r)af

ter

load

ing.

CoCr

cabl

ean

dho

secl

amp

had

high

est

cons

truc

tst

iffne

ssan

dle

ast

redu

ctio

nin

stiff

ness

with

incr

ease

dlo

adin

g.

Scre

ws

(Kam

psho

ffet

al.,

2010

)Eff

ecto

fdiff

eren

tscr

ewim

plan

tatio

nte

chni

ques

onth

ein

tegr

ityof

loca

lcem

entm

antle

and

fixat

ion

stre

ngth

ofth

esc

rew

.Usi

ngdi

ffere

ntki

nds

oflo

ckin

gsc

rew

s.LC

S(S

ynth

es,O

berd

orf,

Switz

erla

nd)a

ndN

CB(Z

imm

er,W

arsa

w,I

N,U

SA),

with

orw

ithou

tafla

tten

edtip

,im

plan

ted

inun

icor

tical

and

bico

rtic

alco

nfigu

ratio

nsus

ing

diffe

rent

core

drill

size

s

No

unic

ortic

alsc

rew

indu

ced

crac

ks–

unic

ortic

alsc

rew

scan

beim

plan

ted

with

out

dam

agin

gce

men

tman

tle.S

crew

sw

ithsh

orte

ned

tip,s

mal

ler

flute

san

ddo

uble

thre

ads

wer

esi

gnifi

cant

lybe

tter

for

pull-

outr

esis

tanc

e.Bi

cort

ical

scre

ws

have

sign

ifica

ntly

high

erpu

ll-ou

tre

sist

ance

,but

incr

ease

risk

oflo

calc

emen

tman

tleda

mag

e.By

incr

easi

ngdr

illdi

amet

er,

onse

tof

crac

ksde

crea

sed,

but

sodo

espu

llou

tre

sist

ance

.(K

onst

antin

idis

etal

.,20

17)

Dam

age

anal

ysis

ofce

men

tm

antle

afte

rre

visi

onsc

rew

inse

rtio

n;In

fluen

ceof

thre

eva

riab

les

onth

ein

cide

nce

ofcr

ack

form

atio

nin

cem

entl

ayer

was

stud

ied;

scre

wty

pe.C

emen

tm

antle

thic

knes

s,an

dpo

sitio

nof

scre

wre

lativ

eto

cem

ent

man

tlean

dpr

osth

etic

stem

.

LCP

plat

esh

orte

ned

toha

vefo

urho

les

only

,and

appl

ied

late

rally

toth

efe

mur

atth

ele

velo

fthe

pros

thet

icst

em(p

roxi

mal

).Th

ree

type

sof

scre

wfix

atio

nm

etho

dsin

vest

igat

ed1)

four

bico

rtic

alno

n-lo

ckin

gsc

rew

s.2)

Four

unic

ortic

allo

ckin

gsc

rew

s.3)

Four

bico

rtic

allo

ckin

gsc

rew

s.

Crac

kfo

rmat

ion

foun

dto

beho

mog

eneo

usfo

rall

thre

esc

rew

type

s.Sc

rew

posi

tions

rela

tive

toce

men

tan

dpr

osth

etic

stem

was

divi

ded

into

four

cate

gori

es:1

)N

oco

ntac

tbe

twee

nsc

rew

and

cem

ent

man

tle.2

)Sc

rew

touc

hes

cem

entm

antle

and

ispa

rtia

llyw

ithin

it.3)

Scre

wis

entir

ely

with

inth

ece

men

tman

tle.4

)Scr

ewis

indi

rect

cont

actw

ithpe

ripr

osth

etic

stem

.

Sign

ifica

ntas

soci

atio

nbe

twee

nsc

rew

posi

tion

and

inci

denc

eof

crac

ksin

the

man

tle.P

roba

bilit

yof

dam

age

toth

ece

men

tman

tlein

crea

ses

sign

ifica

ntly

asth

edi

stan

ceto

the

impl

ante

dpr

osth

esis

decr

ease

s.Sc

rew

sin

dire

ctco

ntac

tw

ithth

epe

ripr

osth

etic

stem

show

edhi

ghin

cide

nce

ofcr

ack

dam

age

toth

ece

men

tman

tle.S

ugge

stst

hato

nly

the

posi

tion

ofth

esc

rew

sre

lativ

eto

the

cem

ent

man

tlean

d/or

peri

pros

thet

icst

emex

erts

sign

ifica

ntin

fluen

ceon

the

crac

kda

mag

eto

the

cem

ent

man

tle.

Com

puta

tiona

lstu

dies

(Dub

ovet

al.,

2011

)Bi

omec

hani

calp

erfo

rman

ceof

cabl

e-sc

rew

posi

tion

inre

pair

ing

peri

pros

thet

icfe

mur

frac

ture

sne

artip

ofa

tota

lhip

impl

ant

(Com

puta

tiona

lstu

dyof

the

expe

rim

enta

lstu

dyca

rrie

dou

tby

the

sam

egr

oup

Shah

etal

.,20

11).

Thre

edi

ffere

ntfix

atio

nm

etho

dsde

scri

bed

1)Co

nstr

uctA

-Cab

le-s

crew

pair

sin

posi

tions

1an

d2

prox

imal

tofr

actu

re2)

Cons

truc

tB-C

able

-scr

ewpa

irsi

npo

sitio

ns1

and

3.3)

Cons

truc

tC–

Cabl

e-sc

rew

pair

sin

posi

tions

1an

d4.

All

thre

eco

nstr

ucts

had

four

bico

rtic

alsc

rew

sdi

stal

ly.

Num

eric

alco

mpu

tatio

nat

1000

Nax

iall

oad

and

15°a

dduc

tion

ofth

efe

mur

show

edhi

gher

axia

lstiff

ness

and

high

ersu

rfac

est

ress

tran

sfer

toth

efe

mor

albo

nefo

rCon

stru

ctC

(cab

le-s

crew

pair

sin

extr

eme

posi

tions

).Su

gges

ting

inth

eca

seof

good

bone

stoc

k,op

timal

fixat

ion

can

beac

hiev

edby

Cons

truc

tC,

and

coul

dpo

tent

ially

redu

cebo

nere

-frac

ture

com

pare

dw

ithA

and

B–

asit

isex

pect

edth

atth

ehi

ghes

tst

iffne

ssm

ayac

hiev

eop

timal

mec

hani

cals

tabi

lity.

FEA

show

edex

celle

ntco

rrel

atio

nw

ithex

peri

men

talr

esul

ts.

(Moa

zen

etal

.,20

12)

Dev

elop

men

tand

anal

ysis

ofan

FEm

odel

ofa

Vanc

ouve

rty

peC

clin

ical

case

com

pari

ngm

echa

nica

leffe

cts

betw

een

two

impl

emen

ted

fixat

ion

met

hods

;whe

rein

the

initi

alfix

atio

nfa

iled

and

repl

aced

bya

seco

ndfix

atio

nth

atle

dto

heal

ing.

Initi

alpl

ate

fixat

ion

follo

win

gTH

Aus

eda

fem

oral

poly

axia

lpla

te(r

igid

fixat

ion

-PO

LYA

X,D

ePuy

,IN

,USA

),Re

frac

ture

revi

sion

used

aco

ndyl

arbl

ade

plat

e(fl

exib

lefix

atio

n-A

ngle

dBl

ade

Plat

e,Sy

nthe

s,PA

)

Rigi

dfr

actu

refix

atio

n(p

olya

xial

plat

e)w

ithsh

ortb

ridg

ing

leng

thin

the

case

ofPF

Fca

nsu

ppre

ssfr

actu

rem

ovem

entt

hatc

anpr

even

thea

ling

and

may

ultim

atel

yfa

il.In

cont

rast

use

ofa

flexi

ble

fixat

ion

non-

lock

ing

plat

ew

ithla

rger

brid

ging

leng

thpr

omot

edhe

alin

g.Ch

angi

ngbr

idgi

ngle

ngth

mad

ea

mor

esu

bsta

ntia

ldiff

eren

ceto

stiff

ness

and

frac

ture

mov

emen

tth

anot

her

para

met

ers.

Resu

ltssu

gges

ttha

taco

mpu

tatio

nala

ppro

ach

toco

mpa

rest

iffne

ssan

dfr

actu

rem

ovem

ent

ofdi

ffere

ntfix

atio

nco

nstr

ucts

can

help

dete

rmin

eop

timum

fixat

ion

met

hod

for

PFF.

(continuedon

nextpage

)

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

155

Table2

(continued)

Aut

hors

Test

case

Resu

lts

Expe

rim

enta

lstu

dies

Plat

ean

dst

rutfi

xatio

n

Late

ralp

late

fixat

ion

Stru

tfixa

tion

Prox

imal

Dis

tal

Posi

tion

Stru

tlen

gth

(mm

)Pr

oxim

alD

ista

l

Uni

cort

ical

Scre

wCa

ble/

wir

eBi

cort

ical

scre

wCa

ble/

wir

eCa

ble/

wir

e

(Che

net

al.,

2012

)Fi

nite

elem

ent

anal

ysis

perf

orm

edto

stud

yin

tern

albi

omec

hani

calf

orce

sdu

ring

fixat

ion

ofVa

ncou

ver

type

B1pe

ripr

osth

etic

frac

ture

with

anO

gden

cons

truc

tand

four

vari

atio

nsof

this

cons

truc

t.1)

A-P

roxi

mal

ly3

wir

es,d

ista

lly2

bico

rtic

alsc

rew

s.2)

B-Pr

oxim

ally

3w

ires

plus

2un

icor

tical

scre

ws,

dist

ally

2bi

cort

ical

scre

ws

3)C-

Prox

imal

ly3

wir

es,d

ista

lly2

bico

rtic

alsc

rew

spl

us3

wir

es4)

D-

Prox

imal

ly3

wir

espl

us2

unic

ortic

alsc

rew

s,di

stal

ly2

bico

rtic

alsc

rew

spl

us3

wir

es.

Resu

ltssh

owth

ator

igin

alba

sic

Ogd

enco

nstr

uct(

A)

fixat

ion

has

infe

rior

outc

ome

com

pare

dto

othe

rfix

atio

nm

etho

ds.A

dditi

onof

two

scre

ws

abov

eth

efr

actu

resi

te(C

onst

ruct

B)vi

sibl

yde

crea

sed

disp

lace

men

tan

dst

ress

.Add

ition

alw

ires

fixed

belo

wfr

actu

resi

tedo

notn

otic

eabl

yde

crea

seei

ther

von

Mis

esst

ress

orfr

actu

redi

spla

cem

ent(

Cons

truc

tC)

.Be

tter

fixat

ion

pow

eris

achi

eved

byus

ing

both

prox

imal

and

dist

alsc

rew

sin

trea

ting

Vanc

ouve

rB1

peri

pros

thet

icfr

actu

res

afte

rTH

A.

(Ebr

ahim

iet

al.,

2012

)Ex

peri

men

tala

ndco

mpu

tatio

nals

tudy

topr

edic

tove

rall

stiff

ness

and

peak

bone

stre

ssin

the

sam

efe

mur

afte

rinj

ury,

repa

ir,a

ndhe

alin

g,w

ithre

spec

tto

itsin

tact

cond

ition

.Fou

rsta

ges

wer

ede

scri

bed.

1)St

age

1–

inta

ctfe

mur

2)St

age

2–

mim

icke

dfe

mur

with

ahi

pst

em3)

Stag

e3

–m

imic

ked

5m

mfr

actu

rega

pre

pair

edw

ithpl

ate

and

scre

ws

4)St

age

4–

repr

esen

ted

com

plet

efr

actu

reun

ion.

FEm

odel

valid

ated

agai

nst

expe

rim

ents

and

re-a

naly

sed

usin

gcl

inic

al-le

velf

orce

of30

00N

Stag

e3

(im

med

iate

post

-sur

gica

lsce

nari

oof

peri

pros

thet

icfe

mor

alfr

actu

refix

atio

n)sh

owed

leas

tsta

ble

situ

atio

nco

mpa

red

tost

age

1,be

ing

the

mos

tvul

nera

ble

tore

-inju

ry;y

ield

ing

the

low

ests

tiffne

ssan

dhi

ghes

tbo

nest

ress

com

pare

dto

stag

e1

(Int

act

fem

ur).

Stag

e4

(hea

led

fem

ur)

show

edri

sein

stiff

ness

surp

assi

ngst

age

1an

dre

-dis

trib

utio

nof

stre

sses

back

tofe

mur

itsel

fcom

pare

dto

stag

e3.

Stud

yhi

ghlig

hts

the

pote

ntia

ladv

erse

effec

tsof

stre

ss-s

hiel

ding

and

high

stre

sses

thro

ugho

utth

esu

rgic

alpr

oces

san

dev

enaf

ter

frac

ture

heal

ing.

Sugg

ests

ast

iffne

ss-m

atch

ing

stra

tegy

infu

ture

desi

gnof

impl

ants

rela

tive

toth

ein

tact

fem

ur.

At1

500

N,F

Evs

Expe

rim

enta

lstr

ains

had

exce

llent

linea

rag

reem

ent.

(Moa

zen

etal

.,20

13)

FEm

odel

ofVa

ncou

vert

ype

B1PF

Ffix

atio

nw

ithin

ast

able

stem

with

good

bone

qual

ityde

velo

ped.

Effec

toff

ract

ure

stab

ility

onlo

ckin

gpl

ate

fixat

ion

perf

orm

ance

quan

tified

,and

com

pari

son

ofst

ainl

ess

stee

l(SS

)and

titan

ium

(Ti)

plat

ein

stab

lean

dun

stab

lefr

actu

reun

der

two

wei

ght-b

arin

gco

nditi

ons

-500

Nan

d23

00N

,ana

lysi

sca

rrie

dou

t.

Stre

ssan

dst

rain

onth

epl

ate

was

high

erin

the

unst

able

com

pare

dto

the

stab

lefix

atio

n.In

the

case

ofun

stab

lefr

actu

res,

itis

poss

ible

for

asi

ngle

lock

ing

plat

efix

atio

nto

prov

ide

the

requ

ired

mec

hani

cale

nvir

onm

entf

orca

llusf

orm

atio

nw

ithou

tsig

nific

antr

isk

ofpl

ate

frac

ture

,pro

vide

dpa

rtia

lw

eigh

tbe

arin

gis

follo

wed

.In

case

sw

here

part

ialw

eigh

tbea

ring

isun

likel

y,ad

ditio

nalb

iolo

gica

lfixa

tion

coul

dbe

cons

ider

ed.

(Moa

zen

etal

.,20

14)

Biom

echa

nica

lper

form

ance

ofsi

xdi

ffere

ntfix

atio

nm

etho

dsfo

rVan

couv

erB1

and

B2ty

pefr

actu

ress

tudi

ed.1

)Use

ofei

ght-h

ole

lock

ing

plat

e:fix

edla

tera

llyus

ing

3un

icor

tical

scre

wsp

roxi

mal

lyan

d4

bico

rtic

alsc

rew

sdi

stal

ly.2

)Ten

-hol

elo

ckin

gpl

ate:

4un

icor

tical

scre

ws

prox

imal

lyan

d4

bico

rtic

alsc

rew

sdis

tally

.3)D

oubl

elo

ckin

gpl

ates

:asw

ithm

etho

d1

plus

anad

ditio

nala

nter

iore

ight

-hol

elo

ckin

gpl

ate

fixed

usin

gth

ree

unic

ortic

alsc

rew

spro

xim

ally

and

thre

ebi

cort

ical

scre

wsd

ista

lly.4

)Rev

isio

nst

em(2

01m

m):

shor

tste

mus

edin

met

hod

1–3

repl

aced

bya

201

mm

long

stem

;the

cem

entm

ante

lwas

expa

nded

med

io-la

tera

llyto

fit.5

)Re

visi

onst

em(2

01m

m)a

ndei

ght-h

ole

plat

e:as

with

met

hod

4pl

usan

addi

tiona

leig

htho

lelo

ckin

gpl

ate

fixed

prox

imal

lyw

ithth

ree

unic

ortic

alsc

rew

san

ddi

stal

lyw

ithon

eun

icor

tical

and

two

bi-c

ortic

alsc

rew

s.6)

Revi

sion

stem

(241

mm

):as

with

met

hod

4ex

cept

stem

exte

nded

by40

mm

.1–3

repr

esen

tPFF

fixat

ion

met

hods

for

Vanc

ouve

rB1

frac

ture

s,an

d4–

6fo

rVa

ncou

ver

B2.

Indi

cate

that

intr

eatm

ent

ofB1

frac

ture

s,a

sing

lelo

ckin

gpl

ate

issu

ffici

entp

rovi

ded

part

ialw

eigh

tba

ring

isfo

llow

ed.

InB2

frac

ture

s,lo

ngst

emre

visi

onan

dby

pass

ing

frac

ture

gap

bytw

ofe

mor

aldi

amet

ers

are

reco

mm

ende

d.Lo

ngst

emre

visi

onco

uld

beco

nsid

ered

inal

lcom

min

uted

B1an

dB2

frac

ture

sw

hen

cons

ider

ing

risk

ofsi

ngle

plat

efr

actu

re.

(Leo

nido

uet

al.,

2015

)Co

mpa

ring

trea

tmen

tm

etho

dsfo

rdi

ffere

ntbo

nequ

ality

–th

ree

mod

els

with

diffe

rent

cana

lthi

ckne

ssra

tio(C

TR),

repr

esen

ting

poor

,av

erag

e,an

dgo

odbo

nequ

ality

.Fu

rthe

rth

ree

mod

els

wer

ede

velo

ped

with

angl

efr

actu

res

vary

ing

from

unst

able

tran

sver

se(0

°),s

hort

obliq

ue(1

46°)

,and

stab

lelo

ngob

lique

confi

gura

tion

(76°

).Co

mpa

riso

nsw

ere

also

mad

eon

thre

edi

ffere

ntm

odel

swith

the

frac

ture

atth

etip

ofth

est

em,4

mm

,and

14m

mbe

low

the

tipof

the

stem

.

Resu

ltssu

gges

ttha

tin

good

bone

qual

ityan

dac

ute

frac

ture

angl

es,s

ingl

elo

ckin

gpl

ate

fixat

ion

can

beco

nsid

ered

asan

appr

opri

ate

man

agem

ent

met

hod.

Conv

erse

lypo

orbo

nequ

ality

and

obtu

sefr

actu

rean

gles

alte

rnat

ive

met

hods

may

bere

quir

edas

fixat

ion

mig

htbe

unde

rhi

gher

risk

offa

ilure

.Sug

gest

sth

ator

thop

aedi

csu

rgeo

nsh

ould

take

into

cons

ider

atio

nth

ePF

Fto

pogr

aphy

and

bone

qual

ityan

dno

tent

irel

yre

lyon

Vanc

ouve

rcl

assi

ficat

ion

tofo

rmul

ate

atr

eatm

ent

plan

.(A

vval

etal

.,20

16)

Inve

stig

atio

nin

tofe

mor

alde

nsity

chan

gesi

nre

spon

seto

bone

frac

ture

plat

ean

dhi

pim

plan

t;lo

ng-te

rmbe

havi

ouro

fafe

mur

inre

spon

seto

thes

eim

plan

tsan

dfix

atio

nsw

ere

sim

ulat

ed.B

one

min

eral

dens

itych

ange

sev

alua

ted

for

late

ralp

latin

gan

dan

teri

orpl

atin

g(3

unic

ortic

alsc

rew

san

d5

bico

rtic

alsc

rew

sus

ed).

Resu

ltssh

owed

that

area

sdi

rect

lyun

der

the

plat

eex

peri

ence

dse

vere

bone

loss

(Up

to~

−70

%).

Som

ele

velo

fbon

efo

rmat

ion

(~+

110%

)w

asob

serv

edin

the

vici

nity

ofth

em

ostp

roxi

mal

and

dist

alsc

rew

hole

sin

both

late

rala

ndan

teri

orpl

ated

fem

urs.

Inre

spec

tto

bone

rem

odel

ling

resp

onse

,ant

erio

rpl

atin

gis

not

supe

rior

tola

tera

lpla

ting.

(continuedon

nextpage

)

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

156

Table2

(continued)

Aut

hors

Test

case

Resu

lts

Expe

rim

enta

lstu

dies

Plat

ean

dst

rutfi

xatio

n

Late

ralp

late

fixat

ion

Stru

tfixa

tion

Prox

imal

Dis

tal

Posi

tion

Stru

tlen

gth

(mm

)Pr

oxim

alD

ista

l

Uni

cort

ical

Scre

wCa

ble/

wir

eBi

cort

ical

scre

wCa

ble/

wir

eCa

ble/

wir

e

(Wan

get

al.,

2016

)(p

)Bi

omec

hani

calp

erfo

rman

ceof

thre

edi

ffere

ntPF

Ffix

atio

nm

etho

dsfo

rVan

couv

erty

peB1

frac

ture

sin

norm

alan

dos

teop

orot

icbo

new

asex

amin

edvi

aFE

mod

el.1

)D

oubl

eci

rcle

cabl

e2)

Trad

ition

allo

ckin

gtit

aniu

mpl

ate

(LCP

)3)

Mul

tidir

ectio

nall

ocki

ngpl

ate.

Stre

ssdi

stri

butio

n,st

iffne

ss,m

axim

umst

ress

and

rela

tive

disp

lace

men

twer

eco

mpa

red

unde

rsa

me

axia

land

tors

iona

lloa

ding

usin

gFE

A.

Mul

tidir

ectio

nall

ocki

ngpl

ate

syst

emsh

owed

high

erst

abili

tyan

dst

iffne

ss,m

ore

even

stre

ssdi

stri

butio

nun

der

the

sam

eax

iala

ndto

rsio

nal

load

ing

inbo

thno

rmal

and

oste

opor

otic

bone

than

doub

leci

rcle

-cab

lean

dtr

aditi

onal

lock

ing

titan

ium

plat

efo

rVa

ncou

ver

B1pe

ripr

osth

etic

frac

ture

s.

Key

toco

nten

t:C,

cabl

e;Ce

,cer

clag

e;W

,wir

e;BC

,Bic

ortic

alsc

rew

s;LS

,Loc

king

scre

ws;

GT,

grea

ter

troc

hant

er;A

nt,a

nter

ior;

Lat,

late

ral;

Med

,med

ial.

(a)

Onl

yhi

pst

emin

sert

ed.

(b)

Dis

talr

etro

grad

efe

mor

alna

ilus

ed,t

wo

lock

ing

scre

ws

used

prox

imal

ly,a

ndon

edi

stal

.(c

)A

dditi

onal

188

mm

ante

rior

lock

ing

plat

ew

asus

ed.

(d)

Inth

isin

stan

ce5

fixed

angl

ese

lf-cu

ttin

glo

ckin

gsc

rew

sus

edto

affix

plat

eto

late

ralf

emor

alco

ndoy

le.

(e)

Inth

isin

stan

ce,b

icor

tical

scre

ws

wer

eus

edfo

rpr

oxim

alfix

atio

nin

stea

dof

unic

ortic

alsc

rew

s.(f

)Lo

ckin

gat

tach

men

tpl

ate

(LA

P)w

asus

ed,a

dditi

onal

bico

rtic

alsc

rew

sus

edfo

rpr

oxim

alfix

atio

n.(g

)In

this

inst

ance

,pro

xim

alfix

atio

nin

volv

edtw

o4

mm

vari

able

angl

esc

rew

s(V

AS)

pass

ing

the

pros

thes

isan

dtw

o5

mm

VAS

pass

ing

the

stem

,to

achi

eve

bico

rtic

alfix

atio

n.(h

)In

this

inst

ance

,tw

oin

trap

rost

hetic

scre

ws

wer

eim

plan

ted

inpl

ace

ofth

ree

unic

ortic

alsc

rew

s.(i

)In

this

inst

ance

,ort

hogo

nalp

latin

gw

asim

plem

ente

d–

two

plat

esw

ere

used

,one

plac

edla

tera

lly(w

ithLA

P)an

don

ean

teri

or.

(j)In

this

inst

ance

two

LAP

was

used

.(k

)In

this

inst

ance

,LCP

leng

thw

asm

ade

shor

t,an

d70

mm

long

segm

ents

ofca

dave

ric

fem

oral

diap

hysi

sw

asus

ed.

(l)

Inth

isin

stan

ce,t

wo

LAP

wer

eap

plie

dus

ing

2tr

ansc

ortic

allo

ckin

gsc

rew

sin

the

oute

rm

ost

hole

s.(m

)In

this

inst

ance

,dis

talfi

xatio

nal

soin

clud

ed2

cabl

esin

conj

unct

ion

with

the

bico

rtic

allo

ckin

gsc

rew

s.(n

)Aut

hord

idn'

tspe

cify

num

bero

fcer

clag

eca

bles

with

crim

pssy

stem

was

used

,but

base

don

data

avai

labl

ean

dfr

omim

ages

we

thin

k2

cabl

esfo

rpro

xim

alan

d2

ford

ista

lfixa

tion

wer

eus

edin

allo

graf

tstr

utfix

atio

n.(o

)In

this

inst

ance

,an

addi

tiona

l5ho

le-L

CPpl

ate

was

fixed

ante

rior

lyto

the

fem

urin

conj

unct

ion

with

ala

tera

lLCP

plat

e.(p

)St

udy

didn

'tsp

ecify

spec

ific

bran

dof

plat

e,bu

twe

hypo

thes

ize

that

the

trad

ition

allo

ckin

gtit

aniu

mpl

ate

isth

eLC

Ppl

ate

byD

ePuy

Synt

hes

and

Mul

tidir

ectio

nall

ocki

ngpl

ate

isth

eN

CBpl

ate

byZi

mm

er.

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

157

motion across the fracture. However, the recent literature has indicatedthat biomechanically, better plate fixation is not dependent on the ri-gidity of a structure alone (Lujan et al., 2010; Moazen et al., 2012).

Results indicate that better plate fixation can be achieved by:

1) Fixation with screws, or screws with cables, in preference to cablesand wires. (Chen et al., 2012; Graham et al., 2015; Lenz et al., 2013;Lever et al., 2010; Shah et al., 2011; Wang et al., 2016)

2) Proximal fixation using bicortical screws instead of unicortical(Gwinner et al., 2015; Hoffmann et al., 2014; Konstantinidis et al.,2010; Lenz et al., 2014; Lewis et al., 2015); or addition of a LAP orLAP-like construct (Griffiths et al., 2015; Lenz et al., 2012b, 2016a)

3) Double plating (Use of additional plate in fixation) (Choi et al.,2010; Lenz et al., 2016a; Wähnert et al., 2017); or strut (Lochabet al., 2017; Sariyilmaz et al., 2014)

4) Intraprosthetic fixation (Brand et al., 2014)5) Use of long stem revision. (Gordon et al., 2016; Moazen et al., 2014)6) Larger bridging length (Moazen et al., 2012; Walcher et al., 2016)7) Application of far cortical locking technology (Moazen et al., 2016)8) Positioning of Screws or Cable-screws (Dubov et al., 2011;

Konstantinidis et al., 2017)

Many authors reported that in cases of good bone stock (typicallyVancouver B1 type fractures), fixation with plate and screws providedmost stability. Shah et al. (2011) showed that plate-screws with addi-tional proximal cable fixation were the best choice for healthy bone; incases of osteoporotic bone, a plate without proximal holes and proximalfixation with only cables was supported. A similar result to Shah et al.(2011) was reported by Demos et al. (2012). However, Graham et al.(2015); found that when unicortical screws are used in conjunctionwith cables, results in proximal screws being pushed into the bone as itis applied, causing screw loosening fixation to the bone. Furthermore,Gordon et al. (2016) showed that osteosynthesis using plate fixationoffered no biomechanical advantages over the use of a simple cerclagesystem. They suggested that revision with a longer stem would providesuperior mechanical stability regardless of the type of osteosynthesisfixation. A similar result could be seen in the computational study byMoazen et al. (2014) who also suggested long stem revision in both B1and B2 fractures when considering the risk of single plate fracture.However, Lewis et al. (2015) found that cable constructs failed in tor-sion by the femur rotating and loosening within the cables. The con-structs also had significantly less maximum force compared to all otherconstructs in both torsional and axial loading. They found that uni-cortical, and unicortical with cable specimens tended to fail by cata-strophic fracture of the femur due to cracks typically stemming frominsertion sites of the screws. Clinically, many studies have reported thatcerclage wiring alone has a high failure rate, and proximal unicorticalscrews in dynamic compression plates, while more stable than cerclagewiring along, are also inadequate (Schwarzkopf et al., 2013).

In regards to the likehood of cement mantle failure when usingscrews; Giesinger et al. (2014) found that plate fixation of PFF usingproximal screws with a stable cemented prosthesis didn't lead to cementmantle failure. In contradiction, Kampshoff et al. (2010) found that useof screws with shortened tip, smaller flutes and double threads, showedbetter pull out resistance, but increase the risk of cement mantle failure.Bicortical screws had significantly superior construct stability and pull-out resistance when compared to unicortical screws; however bicorticalscrews also increased the risk of local cement mantle failure. Ad-ditionally, Gwinner et al. (2015) also showed that the mode of failurewas more catastrophic in proximal bicortical screw fixation; with severecomminuted fracture patterns occurring, compared to screw pull-outwith less bone damage seen in the unicortical screws group.Konstantinidis et al. (2017) showed that the probability of cementmantle damage increases significantly the closer it is to the implantedprosthesis. Direct contact of screws with cement mantle resulted inhigher incidence of cement mantle crack damage. Lever et al. (2010)

also noted that in a clinical situation; cortical screw tips could nick thelateral surface of the femoral stem, resulting in metallic wear debrisforming during daily activities. Furthermore, some of the mechanicalstiffness measured may be due to screw impingement into cement; thusslightly overestimating stiffness levels that could be achieved in vivo.

Demos et al. (2012) found that there was no difference betweenlocking screws and non-locking screws. Many studies using bicorticalscrews or a LAP construct for proximal fixation showed higher rigiditycompared to unicortical screws. However, there were contradictions;Wähnert et al. (2014) found that use of LAP did not provide the moststability as it caused a less rigid plate. Moazen et al. (2016) found thatdistal far cortical locking screws can reduce the overall effective stiff-ness of locking plates and increase fracture movement. They also foundthat the overall strength of the PFF fixation construct was maintainedwhen compared to bicortical fixation with distal locking screws. How-ever, in unstable fractures, alternative fixation methods may be a bettertreatment option.

Fracture gap and bridging length were also found to influence thestability of a fixation construct; Graham et al. (2015) found that frac-ture gap model behaves differently to the no gap model and that thedegree of fracture reduction affects whole construct stability andbending behaviour of bone. Walcher et al. (2016) showed increasedstrain with decreased over-lap or gap of the plate to stem. An FE ana-lysis of a clinical case carried out by Moazen et al. (2012) suggested thatimplementing a fracture plate with a larger bridging length may pro-mote healing compared to a plate with shorter bridging length, dis-playing the importance of plate positioning in Vancouver type C PFFfixation.

4. Discussion

A total of 30 experimental and 9 computational studies publishedsince 2010 relating to PFF were reviewed in this paper. Several ad-vancements and differences were summarised compared to past studies;however, some issues still remain. Four main issues that were high-lighted in the previous review (Moazen et al., 2011), remain important;briefly, they are as follows;

1) Lack of standardization in methods used.2) Variation in the level of sophistication in both experimental and

computational models; in experimental studies, there is typically atrade-off between accuracy and consistency. In computational stu-dies, the balance is between realism and time for development andprocessing.

3) Biomechanical studies are primarily concentrated on Vancouvertype B1 fractures. With less focus on type A and C.

4) The relationship between results presented and the clinical situationneeds to be better defined. Two main issues that are clinically im-portant are, firstly the fracture heals, and secondly, the constructdoesn't fail.

Table 1 shows that there is still a lack of standardization for testingPFF. Current experimental studies still show a lack of consistency inboth testing procedures and measurements. This makes it difficult tomake direct and conclusive comparisons between findings. Biomecha-nical testing comparing the two main plates for PFF fixation (The LCPby DePuy Synthes, and NCP by Zimmer) typically use the same NCBplates but different DePuy Synthes plates, or plates of different lengths,making it difficult to make direct comparisons between the differentstudies and plates used (Konstantinidis et al., 2010; Lever et al., 2010;Lewis et al., 2015; Wähnert et al., 2014).

Modelling of the clinical problem is not easily done because eachPFFs case is different. The best approximation to the clinical challengein either experimental or computational studies is made by taking intoaccount all different parameters that affect the clinical result. Thusmodelling appropriate anatomic region and the stability of the fracture,

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

158

bone stock and the stability of the implant, and patients' characteristicsas demographics are the important basic requirements that we have toconsider when making the best experimental or computational study.Most of the biomechanical studies still concentrate on Vancouver typeB1 fractures, with no studies conducted on Vancouver type A; and onlyone experimental and one computational study (Walcher et al., 2016;Moazen et al., 2012;) on Vancouver type C fractures. This may be due tothe fact they are clinically less prevalent, and more easily treated(Brand et al., 2015; Capone et al., 2017; Fleischman and Chen, 2015;Lever et al., 2010). Vancouver type B2 and B3 fractures are morechallenging to conduct experimentally, with some studies using afracture gap to mimic an unstable fracture (Choi et al., 2010; Giesingeret al., 2014; Graham et al., 2015; Griffiths et al., 2015; Konstantinidiset al., 2010; Lochab et al., 2017; Sariyilmaz et al., 2014; Shah et al.,2011; Wähnert et al., 2014, 2017). However, it is important to note thatclinically, type B2 and B3 fractures are not only unstable fractures, butthe stem itself is unstable, meaning the stem has lost the connectionwith the surrounding bone and requires additional revision or treat-ment, typically with a longer stem (Schwarzkopf et al., 2013). In ad-dition, there are still several contradictions to which treatment methodis the ‘optimum’. The lack of standardization may be attributed to in-adequate understanding of treatment and differentiation betweenstable and unstable prosthesis; as failure to identify an unstable implantmay lead to treatment failure if osteosynthesis rather than revisionsurgery is performed (Schwarzkopf et al., 2013). Thus it is important toalso have biomechanical models that differentiate between stable andunstable prosthesis.

A distinct difference seen in present studies compared to older onesis the reduced use of struts and increased use of the LAP and doubleplating in the experimental studies. Of the 30 experimental studies,only three cases used struts in their biomechanical experiments. This isa stark contrast in comparison to the previous review, where of the 14experimental cases reviewed, eight studies used struts. This is in placeof the introduction and increase in testing the biomechanical perfor-mance of double plating and the use of a LAP or similar construct.Clinically, there is not much data regarding the use of the LAP, how-ever, there have been some reports of acceptable outcomes from usingan LAP to manage PFF with a well-fixed stem (Kim et al., 2017a, 2017b)or when stability of plate is insufficient (Kammerlander et al., 2013).Despite the significant decrease in the use of struts in biomechanicaltesting; clinically struts in conjunction with plate fixation are stillwidely used for PFF fixation treatment, with some studies showingpositive clinical outcomes (Barden et al., 2003; Khashan et al., 2013;Kim et al., 2017a, 2017b).

Another interesting and perhaps important development is the in-creased use of computational modelling in simulating PFF and itsfixation methods; possibly because researchers have realised the addedvalue of using this approach. The review of Moazen et al. (2011) re-ported only two computational studies; here nine cases were reviewed,ranging from simple models to more complex situations such as in-vestigating femoral density changes in response to bone fracture plateand hip implant (Avval et al., 2016), or modelling clinical cases(Moazen et al., 2012). While experimental studies remain the keycomponent of these biomechanical studies, there is no doubt in thevalue that computational studies bring to testing and evaluating effec-tive fixation methods in a greater range of fracture scenarios and morecomplex situations. Several computational studies were corroboratedagainst experimental results (Dubov et al., 2011; Ebrahimi et al., 2012;Lenz et al., 2013; Shah et al., 2011) demonstrating their validity.However, whether clinicians or researchers on the whole have con-fidence in the outcome of computational results over experimental isstill a matter of debate.

From a clinical point of view, the crucial outcome is that the frac-ture heals, return to pre-injury function, and the construct itself doesn'tfail. Much of the research has hence focused on construct stiffness; andthis is still the case in many of the present studies which highlight the

higher construct overall stiffness as the “better” fixation; this is despitestudies shown by several groups that locking plates (depending on howthey are applied) lead to overly rigid fixations that can supress callusformation (Lujan et al., 2010; Moazen et al., 2012). This can be partialsince we still do not know the overall stiffness of PFF fixations in situ,and that can be widely different to the way that they have been testedbiomechanically. An interesting development in response to this hasbeen the introduction of far cortical locking technology (Bottlang et al.,2009; Bottlang and Feist, 2011); commercially named MotionLoc, andcan be used in Zimmer NCB plates. The screws lock into the plate andbypasses the near cortex, reducing the effective stiffness of lockingplates compared to standard locking screws that are secured in bothnear and far cortices, limiting the rigidity of the fixation and supportingcallus formation. While there are some clinical data available that showsome positive results in the use of far cortical locking screws, particu-larly in distal periprosthetic femoral fractures (Bottlang et al., 2010;Ries et al., 2013; Wang et al., 2018); none, to the best of our knowledgehave reported any clinical data regarding PFF after THA specifically.

Thus more clinical data regarding the use of these new platingmethods or technologies needs to be reported to better translate andvalidate experimental and computational data. In this review, only onestudy (Moazen et al., 2016) focused on far cortical locking screws; againdemonstrating the importance of computational studies in testing morecomplex scenarios. There has been evidence of experimental andcomputational studies being translated into clinical practise; studies byGordon et al. (2016) and Moazen et al. (2014) advocated long stemrevision in cases of B1 and B2 fracture treatment; this aligns withclinical data of patients with failed B1 fracture osteosynthesis showedthat revision to a long stem provided good results (Cassidy et al., 2018;Randelli et al., 2018). Cassidy et al. (2018) suggested that revisionrather than repeat fixation, regardless of how well fixed the stem ap-pears would be optimum.

Present biomechanical studies used either cemented or uncementedhip stems; however, no studies made a no direct comparison betweenthe two and its effects on the biomechanical performance of the fixationconstruct. Thus it is difficult to say whether or not the literature for oneprosthesis implantation method can be applied to the other; conse-quently whether subsequent treatment methods derived from bio-mechanical studies where most studies used cemented prosthesis (22out of 30 experimental, and 4 out of 9 computational), can be used foruncemented and vice versa. Thus the relationship between cementedversus uncemented hip prostheses and its fixation methods needs fur-ther research in order to provide more clinically relevant data, this isparticularly paramount as the use of uncemented stems is increasing forTHA (Kim et al., 2015; Philippe et al., 2015).

It is also important to consider that clinically, there is differentbehaviour between cemented and uncemented THA. Failure is morelikely to occur in patients who underwent uncemented THA (Wyatt,2014). However, Wyatt (2014) noted that a 13-year long follow up ofTHA cases showed that there was no significant difference in revisionbetween implantation methods; suggesting the higher early revisionrate may be due to intraoperative events from an inexperienced surgicalteam. However, this contradicted the Swedish registry results, whichshow that uncemented stems are revised twice as frequently as ce-mented stems during the first five years, and that cemented stems wereten times less likely to require revision for periprosthetic fracture.

The Vancouver classification system for treating PFF was originallydeveloped for THAs with cemented femoral components (Duncan andMasri, 1995), and does not differentiate treatment between cementedand uncemented hip stems; thus raising the question of can directcomparisons for treatment of PFF to be made between cemented anduncemented prostheses. While the Vancouver classification system isreported to be reliable and valid, it is difficult to strictly apply rules fortreatment in some cases as there is no objective standard to assess thebone quality or prosthetic stability, and is an arguable drawback of theVancouver classification system (Park et al., 2011). Another caveat of

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

159

this system is that it cannot differentiate between stable and unstableprosthesis easily, which is one of the most crucial parts of treatment.Thus it would be useful if different types of PFF models that are noteasily recognised in the clinical setting could be simulated experimen-tally and computationally.

Another critical issue that needs to be discussed is the lack of os-teoporotic bone models; in most studies, the bone quality that was si-mulated experimentally and computationally could be considerednormal healthy bone stock, and not osteoporotic bone seen in patientswith high risk of PFF; with only 3 studies using osteoporotic bonemodels, two of which did compare bone quality (Lehmann et al., 2010;Leonidou et al., 2015; Wang et al., 2016). The same issues can be raisedas to whether or not results from current biomechanical studies can betranslated into clinical cases, and thus further studies using osteoporoticbone models is required.

While these issues still exist, it is important to recognize the im-proved strides made towards understanding the underlying issues ofPFF and its treatment methods. With the increased interest in PFF,many of the current studies show a higher level of sophistication intheir methods used. This is reflected in many of the studies showingmore consideration and highlighting parameters that may affect PFFthat were not previously tested in earlier studies (Moazen et al., 2011);such as fracture gap (simulating unstable fracture), cement mantle in-tegrity, bridging length, and plate type used. Comparison of bio-mechanical performance between constructs in different situations wasalso studied (e.g. before fracture, fracture with a plate, healed fracturegap - Giesinger et al., 2014; Graham et al., 2015; Griffiths et al., 2015).The interest in improving PFF fixation has also seen the development ofnew commercially available plates specifically designed for PFF. Sev-eral studies have made comparisons on the two major plates used forPFF; the LCP and NCL plate, as well the LAP (Griffiths et al., 2015; Lenzet al., 2012b, 2016b).

5. Conclusion

This review follows our earlier review of experimental and com-putational modelling of PFF fixation (Moazen et al., 2011). While therehave been improvements in the way biomechanical testing of PFFfixation is carried out, the lack of literature to address the situationsdescribed above hinders its translation into clinical practise. In parti-cular, the optimal treatment for Vancouver type B fractures remainscontroversial with experimental data not always reflecting actions oc-curring in situ. This is primarily due to available literature; whichmainly consists of small to medium-sized heterogeneous case studiesthat offer little comparative evidence (Fleischman and Chen, 2015).With the incidence of PFF expected to rise, a consensus on biomecha-nical testing methods, and subsequent optimum treatment methodsneed to be achieved. The effect of cemented versus uncemented pros-thesis on fixation methods needs further research, as well as the de-velopment of more osteoporotic bone models. An effective method canbe seen in using experimental methods in conjunction with computa-tional methods to help bridge this gap and develop more clinically re-levant models.

Conflict of interest

The authors confirm that there is no conflict of interest in thismanuscript.

Acknowledgements

This was supported by EPSRC Doctoral Training Partnership (DTP)Case Studentship (539270/173067). The authors declare that no com-peting interests exist.

References

Avval, P.T., Samiezadeh, S., Bougherara, H., 2016. Long-term response of femoral densityto hip implant and bone fracture plate: computational study using a mechano-bio-chemical model. Med. Eng. Phys. 38, 171–180. https://doi.org/10.1016/j.medengphy.2015.11.013.

Barden, B., Ding, Y., Fitzek, J.G., Löer, F., 2003. Strut allografts for failed treatment ofperiprosthetic femoral fractures: good outcome in 13 patients. Acta Orthop. Scand.74, 146–153. https://doi.org/10.1080/00016470310013860.

Biggi, F., Di Fabio, S., D'Antimo, C., Trevisani, S., 2010. Periprosthetic fractures of thefemur: the stability of the implant dictates the type of treatment. J. Orthop.Traumatol. 11, 1–5. https://doi.org/10.1007/s10195-010-0085-z.

Bottlang, M., Feist, F., 2011. Biomechanics of far cortical locking. J. Orthop. Trauma 25.https://doi.org/10.1097/BOT.0b013e318207885b.

Bottlang, M., Doornink, J., Fitzpatrick, D.D., Madey, S.M., 2009. Far cortical locking canreduce stiffness of locked plating constructs while retaining construct strength. J.Bone Joint Surg. Am. 91, 1985–1994.

Bottlang, M., Lesser, M., Koerber, J., Doornink, J., Von Rechenberg, B., Augat, P.,Fitzpatrick, D.C., Madey, S.M., Marsh, J.L., 2010. Far cortical locking can improvehealing of fractures stabilized with locking plates. J. Bone Joint Surg. Am. 92, 1652.https://doi.org/10.2106/JBJS.I.01111.

Brand, S., Klotz, J., Hassel, T., Petri, M., Ettinger, M., Bach, F.W., Krettek, C., Gösling, T.,2014. Intraprosthetic screw fixation increases primary fixation stability in peripros-thetic fractures of the femur-a biomechanical study. Med. Eng. Phys. 36, 239–243.https://doi.org/10.1016/j.medengphy.2013.07.016.

Brand, S., Ettinger, M., Omar, M., Hawi, N., Krettek, C., Petri, M., 2015. Concepts andpotential future developments for treatment of periprosthetic proximal femoralfractures. Open Orthop. J. 9, 405–411. https://doi.org/10.2174/1874325001509010405.

Capone, A., Congia, S., Civinini, R., Marongiu, G., 2017. Periprosthetic fractures: epide-miology and current treatment. Clin. Cases Miner. Bone Metab. 14, 189–196. https://doi.org/10.11138/ccmbm/2017.14.1.189.

Cassidy, J.T., Kenny, P., Keogh, P., 2018. Failed osteosynthesis of cemented B1 peri-prosthetic fractures. Injury 49, 1927–1930. https://doi.org/10.1016/j.injury.2018.07.030.

Chen, D.W., Lin, C.L., Hu, C.C., Wu, J.W., Lee, M.S., 2012. Finite element analysis ofdifferent repair methods of Vancouver B1 periprosthetic fractures after total hip ar-throplasty. Injury 43, 1061–1065. https://doi.org/10.1016/j.injury.2012.01.015.

Choi, J.K., Gardner, T.R., Yoon, E., Morrison, T.A., Macaulay, W.B., Geller, J.A., 2010.The effect of fixation technique on the stiffness of comminuted Vancouver B1 peri-prosthetic femur fractures. J. Arthroplast. 25, 124–128. https://doi.org/10.1016/j.arth.2010.04.009.

Della Rocca, G.J., Leung, K.S., Pape, H.C., 2011. Periprosthetic fractures: epidemiologyand future projections. J. Orthop. Trauma 25, 7–10. https://doi.org/10.1097/BOT.0b013e31821b8c28.

Della Valle, C.J., Haidukewych, G.J., Callaghan, J.J., 2010. Periprosthetic fractures of thehip and knee: a problem on the rise but better solutions. Instr. Course Lect. 59,563–575.

Demos, H.A., Briones, M.S., White, P.H., Hogan, K.A., Barfield, W.R., 2012. A bio-mechanical comparison of periprosthetic femoral fracture fixation in normal andosteoporotic cadaveric bone. J. Arthroplast. 27, 783–788. https://doi.org/10.1016/j.arth.2011.08.019.

Dubov, A., Kim, S.Y.R., Shah, S., Schemitsch, E.H., Zdero, R., Bougherara, H., 2011. Thebiomechanics of plate repair of periprosthetic femur fractures near the tip of a totalhip implant: the effect of cable-screw position. Proc. Inst. Mech. Eng. H J. Eng. Med.225, 857–865. https://doi.org/10.1177/0954411911410642.

Duncan, C.P., Haddad, F.S., 2014. The Unified Classification System (UCS): improving ourunderstanding of periprosthetic fractures. Bone Joint J. 96 (B), 713–716. https://doi.org/10.1302/0301-620X.96B6.

Duncan, C.P., Masri, B.A., 1995. Fractures of the femur after hip replacement. Instr.Course Lect. 44, 293–304. https://doi.org/10.1017/CBO9781107415324.004.

Ebrahimi, H., Rabinovich, M., Vuleta, V., Zalcman, D., Shah, S., Dubov, A., Roy, K.,Siddiqui, F.S., Schemitsch, E.H., Bougherara, H., Zdero, R., 2012. Biomechanicalproperties of an intact, injured, repaired, and healed femur: an experimental andcomputational study. J. Mech. Behav. Biomed. Mater. 16, 121–135. https://doi.org/10.1016/j.jmbbm.2012.09.005.

Fleischman, A.N., Chen, A.F., 2015. Periprosthetic fractures around the femoral stem:overcoming challenges and avoiding pitfalls. Ann. Transl. Med. 3, 234. https://doi.org/10.3978/j.issn.2305-5839.2015.09.32.

Frisch, N.B., Charters, M.A., Sikora-Klak, J., Banglmaier, R.F., Oravec, D.J., Silverton,C.D., 2015. Intraoperative periprosthetic femur fracture: a biomechanical analysis ofcerclage fixation. J. Arthroplast. 30, 1449–1457. https://doi.org/10.1016/j.arth.2015.02.026.

Giesinger, K., Ebneter, L., Day, R.E., Stoffel, K.K., Yates, P.J., Kuster, M.S., 2014. Can plateosteosynthesis of periprosthethic femoral fractures cause cement mantle failurearound a stable hip stem? A biomechanical analysis. J. Arthroplast. 29, 1308–1312.https://doi.org/10.1016/j.arth.2013.12.015.

Gordon, K., Winkler, M., Hofstädter, T., Dorn, U., Augat, P., 2016. Managing VancouverB1 fractures by cerclage system compared to locking plate fixation - a biomechanicalstudy. Injury 47, S51–S57. https://doi.org/10.1016/S0020-1383(16)47009-9.

Graham, S.M., Mak, J.H., Moazen, M., Leonidou, A., Jones, A.C., Wilcox, R.K., Tsiridis, E.,2015. Periprosthetic femoral fracture fixation: a biomechanical comparison betweenproximal locking screws and cables. J. Orthop. Sci. 20, 875–880. https://doi.org/10.1007/s00776-015-0735-3.

Griffiths, J.T., Taheri, A., Day, R.E., Yates, P.J., 2015. Better axial stiffness of a Bicortical

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

160

screw construct compared to a cable construct for comminuted Vancouver B1 prox-imal femoral fractures. J. Arthroplast. 30, 2333–2337. https://doi.org/10.1016/j.arth.2015.06.060.

Gwinner, C., Märdian, S., Dröge, T., Schulze, M., Raschke, M.J., Stange, R., 2015.Bicortical screw fixation provides superior biomechanical stability but devastatingfailure modes in periprosthetic femur fracture care using locking plates. Int. Orthop.39, 1749–1755. https://doi.org/10.1007/s00264-015-2787-6.

Hoffmann, M.F., Burgers, T.A., Mason, J.J., Williams, B.O., Sietsema, D.L., Jones, C.B.,2014. Biomechanical evaluation of fracture fixation constructs using a variable-anglelocked periprosthetic femur plate system. Injury 45, 1035–1041. https://doi.org/10.1016/j.injury.2014.02.038.

Huiskes, R., Chao, E.Y.S., 1983. A survey of finite element analysis in orthopedic bio-mechanics: the first decade. J. Biomech. 16, 385–409. https://doi.org/10.1016/0021-9290(83)90072-6.

Kammerlander, C., Kates, S.L., Wagner, M., Roth, T., Blauth, M., 2013. Minimally invasiveperiprosthetic plate osteosynthesis using the locking attachment plate. Oper. Orthop.Traumatol. 25, 398–408, 410. https://doi.org/10.1007/s00064-011-0091-1.

Kampshoff, J., Stoffel, K.K., Yates, P.J., Erhardt, J.B., Kuster, M.S., 2010. The treatment ofperiprosthetic fractures with locking plates: effect of drill and screw type on cementmantles: a biomechanical analysis. Arch. Orthop. Trauma Surg. 130, 627–632.https://doi.org/10.1007/s00402-009-0952-3.

Kenanidis, E., Tsiridis, E., Nečas, L., Rov\vnák, M., Buttaro, M., Scolaro, J.A.,Schwarzkopf, R., Statz, J.M., Ledford, C.K., Trousdale, R.T., 2018. Periprostheticfemoral fractures (PFFs). In: Tsiridis, E. (Ed.), The Adult Hip - Master Case Series andTechniques. Springer International Publishing, Cham, pp. 791–816. https://doi.org/10.1007/978-3-319-64177-5_39.

Khashan, M., Amar, E., Drexler, M., Chechik, O., Cohen, Z., Steinberg, E.L., 2013.Superior outcome of strut allograft-augmented plate fixation for the treatment ofperiprosthetic fractures around a stable femoral stem. Injury 44, 1556–1560. https://doi.org/10.1016/j.injury.2013.04.025.

Kim, M., Chung, Y., Lee, J., Park, J., 2015. Outcomes of surgical treatment of peripros-thetic femoral fractures in cementless hip arthroplasty. Hip Pelvis 27, 146–151.https://doi.org/10.5371/hp.2015.27.3.146.

Kim, M., Cho, J., Lee, Y., Shon, W., Park, J., Kim, J., Oh, J., 2017a. Locking attachmentplate fixation around a well-fixed stem in periprosthetic femoral shaft fractures. Arch.Orthop. Trauma Surg. 137, 1193–1200. https://doi.org/10.1007/s00402-017-2745-4.

Kim, Y., Mansukhani, S.A., Kim, J., Park, J., 2017b. Use of locking plate and strut onlayallografts for Periprosthetic fracture around well-fixed femoral components. J.Arthroplast. 32, 166–170. https://doi.org/10.1016/j.arth.2016.05.064.

Kluess, D., Wieding, J., Souffrant, R., Mittelmeier, W., Bader, R., 2010. Finite elementanalysis in orthopaedic biomechanics. Finite Elem. Anal. 151–171. https://doi.org/10.5772/52807.

Konstantinidis, L., Hauschild, O., Beckmann, N.A., Hirschmüller, A., Südkamp, N.P.,Helwig, P., 2010. Treatment of periprosthetic femoral fractures with two differentminimal invasive angle-stable plates: biomechanical comparison studies on cadavericbones. Injury 41, 1256–1261. https://doi.org/10.1016/j.injury.2010.05.007.

Konstantinidis, L., Schmidt, B., Bernstein, A., Hirschmüller, A., Schröter, S., Südkamp,N.P., Helwig, P., 2017. Plate fixation of periprosthetic femur fractures: what happensto the cement mantle? Proc. Inst. Mech. Eng. H J. Eng. Med. 231, 138–142. https://doi.org/10.1177/0954411916682769.

Learmonth, I.D., 2004. Aspects of current management the management of periprostheticfractures around the femoral stem. J. Bone Joint Surg. (Br.) 8686, 13–19. https://doi.org/10.1302/0301-620X.86B1.14864.

Lehmann, W., Rupprecht, M., Hellmers, N., Sellenschloh, K., Briem, D., Püschel, K.,Amling, M., Morlock, M., Rueger, J.M., 2010. Biomechanical evaluation of peri- andinterprosthetic fractures of the femur. J. Trauma 68, 1459–1463. https://doi.org/10.1097/TA.0b013e3181bb8d89.

Lenz, M., Gueorguiev, B., Joseph, S., Pol, Der, Van, B., Richards, R.G., Windolf, M.,Schwieger, K., De Boer, P., 2012a. Angulated locking plate in periprosthetic proximalfemur fractures: biomechanical testing of a new prototype plate. Arch. Orthop.Trauma Surg. 132, 1437–1444. https://doi.org/10.1007/s00402-012-1556-x.

Lenz, M., Windolf, M., Mückley, T., Hofmann, G.O., Wagner, M., Richards, R.G.,Schwieger, K., Gueorguiev, B., 2012b. The locking attachment plate for proximalfixation of periprosthetic femur fractures - a biomechanical comparison of twotechniques. Int. Orthop. 36, 1915–1921. https://doi.org/10.1007/s00264-012-1574-x.

Lenz, M., Perren, S.M., Gueorguiev, B., Höntzsch, D., Windolf, M., 2013. Mechanicalbehavior of fixation components for periprosthetic fracture surgery. Clin. Biomech.28, 988–993. https://doi.org/10.1016/j.clinbiomech.2013.09.005.

Lenz, M., Perren, S.M., Gueorguiev, B., Richards, R.G., Hofmann, G.O., FernandezDell'Oca, A., Höntzsch, D., Windolf, M., 2014. A biomechanical study on proximalplate fixation techniques in periprosthetic femur fractures. Injury 45, 71–75. https://doi.org/10.1016/j.injury.2013.10.027.

Lenz, M., Stoffel, K., Gueorguiev, B., Klos, K., Kielstein, H., Hofmann, G.O., 2016a.Enhancing fixation strength in periprosthetic femur fractures by orthogonal plating -a biomechanical study. J. Orthop. Res. 34, 591–596. https://doi.org/10.1002/jor.23065.

Lenz, M., Stoffel, K., Kielstein, H., Mayo, K., Hofmann, G.O., Gueorguiev, B., 2016b. Platefixation in periprosthetic femur fractures Vancouver type B1—trochanteric hookplate or subtrochanterical bicortical locking? Injury 47, 2800–2804. https://doi.org/10.1016/j.injury.2016.09.037.

Leonidou, A., Moazen, M., Skrzypiec, D.M., Graham, S.M., Pagkalos, J., Tsiridis, E., 2013.Evaluation of fracture topography and bone quality in periprosthetic femoral frac-tures: a preliminary radiographic study of consecutive clinical data. Injury 44,1799–1804. https://doi.org/10.1016/j.injury.2013.08.010.

Leonidou, A., Moazen, M., Lepetsos, P., Graham, S.M., Macheras, G.A., Tsiridis, E., 2015.The biomechanical effect of bone quality and fracture topography on locking platefixation in periprosthetic femoral fractures. Injury 46, 213–217. https://doi.org/10.1016/j.injury.2014.10.060.

Lever, J.O., Zdero, R., Waddell, J.P., 2010. The biomechanical analysis of three platingfixation systems for periprosthetic femoral fracture near the tip of a total hip ar-throplasty. Mech. Eng. Publ. Res. 1–8.

Lewallen, D.G., Berry, D.J., 1998. Periprosthetic fracture of the femur after total hip ar-throplasty: treatment and results to date. Instr. Course Lect. 47, 243–249.

Lewis, G.S., Caroom, C.T., Wee, H., Jurgensmeier, D., Rothermel, S.D., Bramer, M.A.,Reid, J.S., 2015. Tangential bicortical locked fixation improves stability in VancouverB1 periprosthetic femur fractures: a biomechanical study. J. Orthop. Trauma 29,e364–e370. https://doi.org/10.1097/BOT.0000000000000365.

Lindahl, H., Garellick, G., Regner, H., Herberts, P., Malchau, H., 2006. Three hundred andtwenty-one periprosthetic femoral fractures. J. Bone Joint Surg. Am. 88, 1215–1222.

Lindahl, H., Oden, A., Garellick, G., Malchau, H., 2007. The excess mortality due toperiprosthetic femur fracture. A study from the Swedish national hip arthroplastyregister. Bone 40, 1294–1298. https://doi.org/10.1016/j.bone.2007.01.003.

Lochab, J., Carrothers, A., Wong, E., McLachlin, S., Aldebeyan, W., Jenkinson, R., Whyne,C., Nousiainen, M.T., 2017. Do transcortical screws in a locking plate construct im-prove the stiffness in the fixation of Vancouver B1 periprosthetic femur fractures? Abiomechanical analysis of 2 different plating constructs. J. Orthop. Trauma 31,15–20. https://doi.org/10.1097/BOT.0000000000000704.

Lujan, T.J., Henderson, C.E., Madey, S.M., Fitzpatrick, D.C., Marsh, J.L., Bottlang, M.,2010. Locked plating of distal femur fractures leads to inconsistent and asymmetriccallus formation. J. Orthop. Trauma 24, 156–162. https://doi.org/10.1097/BOT.0b013e3181be6720.

Mann, K.A., Ayers, D.C., Damron, T.A., 1997. Effects of stem length on mechanics of thefemoral hip component after cemented revision. J. Orthop. Res. 15, 62–68. https://doi.org/10.1002/jor.1100150110.

Marsland, D., Mears, S.C., 2012. A review of periprosthetic femoral fractures associatedwith total hip arthroplasty. Geriatr. Orthop. Surg. Rehabil. 3, 107–120. https://doi.org/10.1177/2151458512462870.

Mihalko, W.M., Beaudoin, A.J., Cardea, J.A., Krause, W.R., 1992. Finite-element model-ling of femoral shaft fracture fixation techniques post total hip arthroplasty. J.Biomech. 25, 469–476. https://doi.org/10.1016/0021-9290(92)90087-H.

Moazen, M., Jones, A.C., Jin, Z., Wilcox, R.K., Tsiridis, E., 2011. Periprosthetic fracturefixation of the femur following total hip arthroplasty: a review of biomechanicaltesting. Clin. Biomech. 26, 13–22. https://doi.org/10.1016/j.clinbiomech.2010.09.002.

Moazen, M., Jones, A.C., Leonidou, A., Jin, Z., Wilcox, R.K., Tsiridis, E., 2012. Rigidversus flexible plate fixation for periprosthetic femoral fracture-computer modellingof a clinical case. Med. Eng. Phys. 34, 1041–1048. https://doi.org/10.1016/j.medengphy.2011.11.007.

Moazen, M., Mak, J.H., Etchels, L.W., Jin, Z., Wilcox, R.K., Jones, A.C., Tsiridis, E., 2013.The effect of fracture stability on the performance of locking plate fixation in peri-prosthetic femoral fractures. J. Arthroplast. 28, 1589–1595. https://doi.org/10.1016/j.arth.2013.03.022.

Moazen, M., Mak, J.H., Etchels, L.W., Jin, Z., Wilcox, R.K., Jones, A.C., Tsiridis, E., 2014.Periprosthetic femoral fracture — a biomechanical comparison between Vancouvertype B1 and B2 fixation methods. J. Arthroplast. 29, 495–500. https://doi.org/10.1016/j.arth.2013.08.010.

Moazen, M., Leonidou, A., Pagkalos, J., Marghoub, A., Fagan, M.J., Tsiridis, E., 2016.Application of far cortical locking technology in periprosthetic femoral fracturefixation: a biomechanical study. J. Arthroplast. 31, 1849–1856. https://doi.org/10.1016/j.arth.2016.02.013.

Niikura, T., Lee, S.Y., Sakai, Y., Nishida, K., Kuroda, R., Kurosaka, M., 2014. Treatmentresults of a periprosthetic femoral fracture case series: treatment method forVancouver type B2 fractures can be customized. Clin. Orthop. Surg. 6, 138–145.https://doi.org/10.4055/cios.2014.6.2.138.

Park, S.K., Kim, Y.G., Kim, S.Y., 2011. Treatment of periprosthetic femoral fractures in hiparthroplasty. Clin. Orthop. Surg. 3, 101. https://doi.org/10.4055/cios.2011.3.2.101.

Philippe, H., Nicolas, D., Jerome, D., Isaac, G., Alexandre, P., Jerome, A., FlouzatLachaniette, C.H., 2015. Long, titanium, cemented stems decreased late peripros-thetic fractures and revisions in patients with severe bone loss and previous revision.Int. Orthop. 39, 639–644. https://doi.org/10.1007/s00264-014-2528-2.

Pletka, J.D., Marsland, D., Belkoff, S.M., Mears, S.C., Kates, S.L., 2011. Biomechanicalcomparison of 2 different locking plate fixation methods in Vancouver B1 peripros-thetic femur fractures. Geriatr. Orthop. Surg. Rehabil. 2, 51–55. https://doi.org/10.1177/2151458510397609.

Randelli, F., Pace, F., Priano, D., Giai Via, A., Randelli, P., 2018. Re-fractures afterperiprosthtic femoral fracture: a difficult to treat growing evidence. Injury 49,S43–S47. https://doi.org/10.1016/j.injury.2018.09.045.

Rayan, F., Dodd, M., Haddad, F.S., 2008. European validation of the Vancouver classi-fication of periprosthetic proximal femoral fractures. J. Bone Joint Surg. (Br.) 90,1576–1579. https://doi.org/10.1302/0301-620X.90B12.20681.

Ries, Z., Hansen, K., Bottlang, M., Madey, S., Fitzpatrick, D., Marsh, J.L., 2013. Healingresults of periprosthetic distal femur fractures treated with far cortical lockingtechnology: a preliminary retrospective study. Iowa Orthop. J. 33, 7–11.

Sariyilmaz, K., Dikici, F., Dikmen, G., Bozdag, E., Sunbuloglu, E., Bekler, B., Yazicioglu,O., 2014. The effect of strut allograft and its position on Vancouver type B1 peri-prosthetic femoral fractures: a biomechanical study. J. Arthroplast. 29, 1485–1490.https://doi.org/10.1016/j.arth.2014.02.017.

Schwarzkopf, R., Oni, J.K., Marwin, S.E., 2013. Total hip arthroplasty Periprosthetic fe-moral fractures. A review of classification and current treatment. Bull. Hosp. JointDis. 71, 68–78.

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

161

Shah, S., Kim, S.Y.R., Dubov, A., Schemitsch, E.H., Bougherara, H., Zdero, R., 2011. Thebiomechanics of plate fixation of periprosthetic femoral fractures near the tip of atotal hip implant: cables, screws, or both? Proc. Inst. Mech. Eng. H J. Eng. Med. 225,845–856. https://doi.org/10.1177/0954411911413060.

Tsiridis, E., Pavlou, G., Venkatesh, R., Bobak, P., Gie, G., 2009. Periprosthetic femoralfractures around hip arthroplasty: current concepts in their management. Hip Int. 19,75–86.

Wähnert, D., Schröder, R., Schulze, M., Westerhoff, P., Raschke, M., Stange, R., 2014.Biomechanical comparison of two angular stable plate constructions for peripros-thetic femur fracture fixation. Int. Orthop. 38, 47–53. https://doi.org/10.1007/s00264-013-2113-0.

Wähnert, D., Grüneweller, N., Gehweiler, D., Brunn, B., Raschke, M.J., Stange, R., 2017.Double plating in Vancouver type B1 periprosthetic proximal femur fractures: abiomechanical study. J. Orthop. Res. 35, 234–239. https://doi.org/10.1002/jor.

23259.Walcher, M.G., Giesinger, K., du Sart, R., Day, R.E., Kuster, M.S., 2016. Plate positioning

in periprosthetic or interprosthetic femur fractures with stable implants—a bio-mechanical study. J. Arthroplast. 31, 2894–2899. https://doi.org/10.1016/j.arth.2016.05.060.

Wang, G., Wang, D., Mao, J., Lin, Y., Yin, Z., Wang, B., He, Y., Sun, S., 2016. Threedimensional finite-element analysis of treating Vancouver B1 periprosthetic femoralfractures with three kinds of internal fixation. Int. J. Clin. Exp. Med. 9, 10915–10922.

Wang, R., Zhang, H., Cui, H., Fan, Z., Xu, K., Liu, P., Ji, F., Tang, H., 2018. Clinical effectsand risk factors of far cortical locking system in the treatment of lower limb fractures.Injury. https://doi.org/10.1016/j.injury.2018.09.013.

Wyatt, M., 2014. Survival outcomes of cemented compared to uncemented stems inprimary total hip replacement. World J. Orthop. 5, 591. https://doi.org/10.5312/wjo.v5.i5.591.

K. Wang et al. Clinical Biomechanics 61 (2019) 144–162

162