See

discussions, stats, and author profiles for this publication at: http://www.researchgate.net/publication/233384453

The importance of lag screw position for the

stabilization of trochanteric fractures with a
sliding Hip screw: A subject-specific finite
element study.

ARTICLE in JOURNAL OF ORTHOPAEDIC RESEARCH · APRIL 2013

Impact Factor: 2.99 · DOI: 10.1002/jor.22266 · Source: PubMed

CITATIONS READS

15 44

3 AUTHORS:

Jérôme M Goffin Pankaj Pankaj

The University of Edinburgh The University of Edinburgh
9 PUBLICATIONS 315 CITATIONS 88 PUBLICATIONS 431 CITATIONS

SEE PROFILE SEE PROFILE

Hamish Simpson
The University of Edinburgh
238 PUBLICATIONS 5,344 CITATIONS

SEE PROFILE

Available from: Jérôme M Goffin

Retrieved on: 25 December 2015
The Importance of Lag Screw Position for the Stabilization of
Trochanteric Fractures with a Sliding Hip Screw: A Subject-Specific
Finite Element Study
Jérôme M. Goffin,1 Pankaj Pankaj,2 A. Hamish Simpson1
1
Department of Orthopaedic Surgery, The University of Edinburgh, The Royal Infirmary of Edinburgh at Little France, Old Dalkeith Road,
Edinburgh EH16 4SU, UK, 2School of Engineering, The University of Edinburgh, Edinburgh, UK
Received 3 July 2012; accepted 15 October 2012
Published online 8 November 2012 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/jor.22266

ABSTRACT: Using finite element analysis, we compared the biomechanical performance of a CT scan-based three-part trochanteric
fracture model (31-A2 in the AO classification) stabilized with a sliding hip screw for nine different positions of the lag screw (3
3
arrangement, from anterior to posterior and from inferior to superior). Our results showed that the volume of bone susceptible to
yielding in the head and neck region is the lowest for inferior positions and increases as the lag screw is moved superiorly. Overall, for
this specific subject, the models less likely to lead to cut-out are the ones corresponding to inferior middle and inferior posterior posi-
tions of the lag screw. In our study, the tip-apex distance (TAD) was anti-correlated with the risk of cut-out, as quantified by the
volume of bone susceptible to yielding, which suggests that a TAD >25 mm cannot be considered to be an accurate predictor of lag
screw cut-out. Further clinical studies investigating lag screw cut-out should attempt to find more reliable predictors of cut-out that
should better reflect the biomechanics and subject-specificity of the femoral head. ß 2012 Orthopaedic Research Society. Published by
Wiley Periodicals, Inc. J Orthop Res 31:596–600, 2013

Keywords: osteosynthesis; cut-out; yield strain; tip-apex distance

The latest Cochrane review comparing intramedullary
nails with extramedullary implants for the fixation of
extracapsular hip fractures concluded that the sliding
hip screw (SHS) should still be considered as the gold
standard device for the stabilization of such fractures.1
However, Born et al.2 reported that the cut-out rate
can be as high as 8% for hip screws. Furthermore, cut-
out of the lag screw from the femoral head is the most
common failure mode of fracture fixation with the SHS
since it accounts for 80% of implant failures. Recent
clinical studies reported data on cut-out rates and lag
screw position.3,4
However such studies suffer from the fact that the
absolute number of patients is not evenly distributed
across the different positions considered within the
femoral head, to the point that some positions are only
represented by a tiny percentage of the number of
patients included in those studies. Another drawback
is that the distribution of bone density across the fem-
oral head is patient-specific. Therefore, collecting data
on different lag screw positions taken from different
femoral heads quite likely requires a large number
of patients in any given position studied (within the
femoral head) to get statistically significant data from
which meaningful conclusions can be drawn.
To attempt to address these weaknesses, we consid-
ered a distinct approach with a different level of evi-
dence and decided to use the same femoral head for all
the lag screw positions evaluated to find the optimal
position for a given patient. The aim of this computa-
tional study was therefore to systematically assess
Conflict of interest: None.
Correspondence to: Jérôme M. Goffin (T: þ44-131-242-6465;
F: þ44 0131 242 6467; E-mail: j.goffin@ed.ac.uk)
ß 2012 Orthopaedic Research Society. Published by Wiley Periodicals, Inc.
how different positions of the lag screw could influence
the risk of cut-out.
MATERIALS AND METHODS
Finite element (FE) modeling was used to evaluate the im-
portance of an adequate positioning of the lag screw (sliding
hip screw) within the femoral head to minimize the risk of
cut-out of the lag screw from the head. We considered three-
part pertrochanteric fractures, classified as 31-A2 in the
Müller AO classification and featuring a lack of medial sup-
port at the level of the lesser trochanter. The angle of the
fracture line with the femoral shaft was assumed to be 438
(mean value of unpublished clinical data from our unit).
The intrusion distance of the medial fragment into the frac-
ture complex (Fig. 1) was assumed to be 30% (representative
value based on unpublished clinical data from our unit).
We used the CT scan of a cadaveric femur obtained from
PhysiomeSpace portal (https://www.physiomespace.com/).
Table 1 summarizes details of this bone that are relevant to
our study (the material properties of the head correspond to
those of our FE model, i.e., 75% density of the cadaveric
bone). This scan was converted to FE models with Simple-
ware software suite (Simpleware Ltd., Exeter, UK) and
imported as an orphan mesh (four-node linear tetrahedral
elements) into an FE solver (Abaqus, Simulia, Providence,
RI). Convergence tests were performed on all models to en-
sure a fine enough element discretization for strain analysis.
A linear relationship between CT Hounsfield units and ash
density was obtained from the calibration of the CT scanner
with a phantom; the following density–elasticity relationship
was chosen to convert apparent density into elastic modu-
3
lus5: EðMPaÞ ¼ 6; 950r1:49
app ðg=cm Þ. Ash density was assumed
6
to be 60% of apparent density. The assignment of inhomoge-
neous material properties to the FE mesh relied on algo-
rithms patented by Simpleware.7
The Omega3 Compression Hip Screw, a typical sliding
hip screw currently manufactured by Stryker Osteosynthesis
(Schoenkirchen, Germany), was used to stabilize our tro-
chanteric fracture model. It was considered to be made of

596 JOURNAL OF ORTHOPAEDIC RESEARCH APRIL 2013


Figure 1. Intrusion distance of the medial fragment into
the fracture complex is defined as the ratio of the length of the
medial wedge in the AP view to the length of the fracture line
and expressed as a percentage.
stainless steel characterized by a Young’s modulus of
195 GPa and a Poisson’s ratio of 0.3. A 1308 four-hole stan-
dard barrel plate was locked with Asnis III screws to the
distal femur.
Frictional contact interactions were assumed between the
different parts of the models. Friction coefficients were taken
from the literature: 0.46 for bone–bone interactions,8 0.42 for
bone–implant interactions,9 and 0.2 for implant–implant
interactions.10 The models were subjected to a load of
1,866 N corresponding to 300% body weight (BW), in accor-
dance with values reported for walking and stair climbing.11
Based on a study by Eberle et al.,12 the force vector was
pointing laterally in the coronal plane at an angle of 138 with
the axis of the femoral shaft. In the sagittal plane, it was
pointing posteriorly at an angle of 88 with the shaft axis. For
boundary conditions, the femoral head was constrained in
the plane orthogonal to the loading vector while the distal
end of the femur was constrained in all translational degrees
of freedom at a point located in the mid-coronal plane at
23 mm medially from the shaft axis. The femur could rotate
around this point about the frontal and sagittal axes.
To account for the clinical variability of lag screw position-
ing in the head, we divided the head into a 3
3 arrange-
ment, from anterior to posterior and from inferior to
Table 1. Donor Data
Side Sex Age at Death
Right Female 78 years
Average BMD of Average Elastic
Femoral Head Modulus of Head Head Diameter
107 mg/cm3 533 MPa 47 mm
LAG SCREW POSITION 597
superior, and systematically assessed the biomechanics of
our trochanteric fracture model by placing the lag screw in
each of these nine positions. Care was taken to place the tip
of the lag screw within 1 cm of the subchondral bone.13,14
We analyzed our FE models with 75% of the bone density of
the cadaveric bone taken from PhysiomeSpace in accordance
with a recent imaging study that showed that the average
bone mineral density (BMD) of femoral heads in hip fracture
patients was 75% of the BMD of an age-matched fracture-
free control group.15
To calculate the tip-apex distance corresponding to the
position of the lag screw, a best-fit sphere was calculated to
match the surface nodes of the head. The coordinates of the
apex were then obtained by taking into account the neck-
shaft angle and the femoral neck anteversion of the bone.
The AP and lateral views of the C-arm were then modeled by
rotating the axes of the system of coordinates by means of
3D rotation matrices. The distance between the apex and the
tip of the lag screw was then projected in both AP and lateral
views and subsequently added as suggested by Baumgaert-
ner et al.16
RESULTS
Contour plots showing minimum principal strains in
bone in a cross-section through the head and neck of
the femur are shown in Figure 2 for the nine lag screw
positions. This figure is based on the assumption that
failure of fixation by cut-out of the screw from the
head is likely to occur due to high compressive strains
in the whole head and neck region corresponding to
the proximal bone fragment. Areas characterized by
strains smaller (i.e., more compressive) than a cut-off
(yield strain) value of 0.9% minimum principal strain
were assigned a black color to emphasize the regions
where the bone was susceptible to yielding (undergo
irreversible deformations) in accordance with recently
published data.17,18
Figure 3 gives the tip-apex distances (TAD) corre-
sponding to the nine positions of the lag screw and the
volume of bone susceptible to yielding in the proximal
bone fragment (head and neck). To check whether a
correlation existed between the volume of bone suscep-
tible to yielding, which is related to the risk of cut-out,
and TAD, we computed Pearson’s correlation coeffi-
cient. We found a correlation coefficient of 0.75,
which indicates a significant (p-value ¼ 0.02) strong
anti-correlation between volume of bone susceptible to
yielding and TAD, that is, higher TADs correspond in
our study to smaller risks of cut-out.
Donor Height Body Mass
171 cm 64 kg
Neck Anteversion Neck-Shaft Angle
208 1358
JOURNAL OF ORTHOPAEDIC RESEARCH APRIL 2013
598 GOFFIN ET AL.
Figure 2. Minimum (compressive) principal strains in the head and neck region plotted in percent with a cut-off (yield strain) value
of 0.9%. Black regions have strains below 0.9% and are at higher risk of being involved in lag screw cut-out.
DISCUSSION
Most previously published biomechanical/computation-
al studies that focused on the position of the lag screw
did not systematically assess the range of positions
that are usually reported in large clinical studies,3 but
instead restricted their comparison to only two posi-
tions, that is, middle middle and inferior middle posi-
tions (SHS)19 or superior middle and inferior middle
(intramedullary nail).20 To our knowledge, our study is
the only exhaustive computational study in the sense
that we tried to sample the whole spectrum of lag
screw positions seen clinically.
The femoral neck anteversion characterizing the
bone used in this study is within the standard anatom-
ical range reported for female patients.21 Similarly,
the neck-shaft angle can be considered to be normal.22
The head diameter falls in the middle of the range
of head diameters reported by Asala et al.23 The
average BMD of the femoral head (107 mg/cm3, quan-
titative CT density) is slightly greater than the mean
(182 mg/cm3) minus two standard deviations (SD
JOURNAL OF ORTHOPAEDIC RESEARCH APRIL 2013
equal to 45 mg/cm3) obtained from the hip fracture
group in the study by Bousson et al.15 The average
elastic modulus of the head, obtained by converting
the average BMD using the density–elasticity relation-
ship, is within the range of values reported by
Sun et al.24 Our average modulus might be slightly
greater than the values published there because we
included, as did Bousson et al., the cortical shell in the
averaging.
We relied on the use of minimum principal strains
and a corresponding yield strain to evaluate the risk of
lag screw cut-out. Bone yielding and failure are indeed
ideally described by minimum (compressive) and maxi-
mum (tensile) principal strains. The stability of the lag
screw within the head and neck region depends on an
adequate anchorage of the lag screw in this region.
Therefore, it should be as important to avoid straining
the bone beyond its yield point in the region immedi-
ately superior to the thread of the lag screw as to avoid
doing so in the region surrounding the shaft of the
screw. Accordingly, we reported (Fig. 3) the volume of

Figure 3. Volume of cancellous bone in the head and neck
region with minimum principal strain below the compressive
yield strain value of 0.9%. It is a quantitative measure of the
risk of lag screw cut-out. Tip-apex distance corresponding to the
location of the lag screw tip.
bone susceptible to yielding in the whole head and
neck region. These volumetric results show that the
anchorage of the lag screw is more stable, that is,
the lag screw is less likely to be involved in cut-out,
for inferior positions and worsens as the lag screw is
moved superiorly. Generally, this seems to be related
to increasing regions of bone susceptible to yielding in
the femoral neck. When looking only at the contour
plots (Fig. 2), the two safest positions seem to be the
inferior middle and inferior posterior positions. The
contour plots do not show any bone susceptible to
yielding in the region immediately superior to the lag
screw thread in any of the models where the lag screw
is placed posteriorly.
It is a common belief to think that TAD is the
best predictor of lag screw cut-out. In our study, the
TADs corresponding to the six models with a middle or
superior position of the lag screw (group A, i.e., middle
anterior, middle middle, middle posterior, superior an-
terior, superior middle, and superior posterior) are all
smaller than the usual cut-off value of 25 mm. On the
other hand, the TADs for the models with a lag screw
positioned inferiorly (group B) are >25 mm. Based on
the current clinical knowledge related to TAD, one
would thus expect that the models in group A would
perform better than those in group B, in terms of
biomechanical outcomes. In fact, we show that the
opposite is true, that is, the risk of cut-out is smaller
for group B (Fig. 2) as quantified by the volume of
LAG SCREW POSITION 599
bone susceptible to yielding (Fig. 3). We therefore be-
lieve that our results can act as a counterexample to
the widely accepted statement ‘‘A tip-apex distance
>25 mm has been shown to be an accurate predictor of
lag screw cut-out.’’25 In logic, a single example that
shows the opposite of a statement is sufficient to prove
that this statement is false. Hence, based on our
results, we can state that a TAD >25 mm failed to be
an accurate predictor of lag screw cut-out for our sub-
ject-specific model.
Linear elasticity was used to model the material
properties in our FE models. For comparison purposes,
linear elasticity is indeed adequate to support our
results.26 Predictions obtained with a four-node tetra-
hedral mesh are similar to those produced by 10-node
tetrahedra if the mesh employed is sufficiently fine.27
We did not include muscle forces. Indeed, a recent
study8 on hip fractures concluded that additional mus-
cle loading, as suggested by Heller et al.,28 had no sig-
nificant effect on the results. There are discrepancies
in the values reported for the peak joint reaction forces
acting on the hip.11,29 In the light of these studies, we
believe that the use of a load corresponding to 300%
BW is adequate. Smaller loads would not affect the
general trends in our results.
Finally, in idealistic conditions, we could imagine
ourselves carrying out thorough surgical planning for
patients at risk of surgical complications (e.g., severe
osteoporosis or complex fracture patterns) whereby a
patient-specific FE model would be developed based on
a CT scan of the patient and different parameters,
such as type of implant and implant position, could
be systematically assessed so that the best evidence-
based treatment could be chosen and tailored to that
particular patient.
In conclusion, using FE modeling, we systematically
evaluated the risk of cut-out in a CT scan-based three-
part trochanteric fracture model by varying the posi-
tion of the lag screw to cover the whole spectrum of
positions seen clinically. Our models suggest that the
safest positions are the inferior middle and inferior
posterior ones. Furthermore, our study shows that
TAD is not an accurate predictor of lag screw cut-out.
We are therefore in dire need of more reliable predic-
tors of cut-out which could better reflect the inhomoge-
neous distribution of bone in the femoral head and
neck region.
ACKNOWLEDGMENTS
We thank Stryker Osteosynthesis and in particular Christian
Lutz, Geert von Oldenburg, and Claus Gerber, who have
generously provided us with 3D CAD models of the sliding
hip screw. The first author is grateful to Paul J. Jenkins
for helpful discussions and was supported by a Principal’s
Career Development Scholarship awarded by The University
of Edinburgh.
REFERENCES
1. Parker MJ, Handoll HH. 2010. Gamma and other cephalo-
condylic intramedullary nails versus extramedullary
JOURNAL OF ORTHOPAEDIC RESEARCH APRIL 2013
600 GOFFIN ET AL.
implants for extracapsular hip fractures in adults. Cochrane
Database Syst Rev CD000093.
2. Born CT, Karich B, Bauer C, et al. 2011. Hip screw migra-
tion testing: first results for hip screws and helical blades
utilizing a new oscillating test method. J Orthop Res 29:
760–766.
3. Hsueh KK, Fang CK, Chen CM, et al. 2010. Risk factors in
cutout of sliding hip screw in intertrochanteric fractures: an
evaluation of 937 patients. Int Orthop 34:1273–1276.
4. Pervez H, Parker MJ, Vowler S. 2004. Prediction of
fixation failure after sliding hip screw fixation. Injury 35:
994–998.
5. Schileo E, Taddei F, Malandrino A, et al. 2007. Subject-
specific finite element models can accurately predict strain
levels in long bones. J Biomech 40:2982–2989.
6. Schileo E, Dall’ara E, Taddei F, et al. 2008. An accurate
estimation of bone density improves the accuracy of subject-
specific finite element models. J Biomech 41:2483–2491.
7. Young PG, Beresford-West TBH, Coward SRL, et al. 2008.
An efficient approach to converting three-dimensional
image data into highly accurate computational models.
Philos Trans R Soc A 366:3155–3173.
8. Eberle S, Gerber C, von Oldenburg G, et al. 2010.
A biomechanical evaluation of orthopaedic implants for hip
fractures by finite element analysis and in-vitro tests. Proc
Inst Mech Eng H 224:1141–1152.
9. Hsu JT, Chang CH, Huang HL, et al. 2007. The number of
screws, bone quality, and friction coefficient affect acetabu-
lar cup stability. Med Eng Phys 29:1089–1095.
10. Sowmianarayanan S, Chandrasekaran A, Kumar RK. 2008.
Finite element analysis of a subtrochanteric fractured femur
with dynamic hip screw, dynamic condylar screw, and proxi-
mal femur nail implants—a comparative study. Proc Inst
Mech Eng H 222:117–127.
11. Heller MO, Bergmann G, Deuretzbacher G, et al. 2001.
Musculo-skeletal loading conditions at the hip during walk-
ing and stair climbing. J Biomech 34:883–893.
12. Eberle S, Gerber C, von Oldenburg G, et al. 2009. Type of
hip fracture determines load share in intramedullary osteo-
synthesis. Clin Orthop Relat Res 467:1972–1980.
13. Armstrong CG, Gardner DL. 1977. Thickness and distribu-
tion of human femoral head articular cartilage. Changes
with age. Ann Rheum Dis 36:407–412.
14. Rockwood CA, Green DP, Bucholz RW, et al. 2006. Rock-
wood and Green’s fractures in adults. Lippincott Williams &
Wilkins: Philadelphia, PA; London.
15. Bousson VD, Adams J, Engelke K, et al. 2011. In vivo
discrimination of hip fracture with quantitative computed
JOURNAL OF ORTHOPAEDIC RESEARCH APRIL 2013
tomography: results from the prospective European Femur
Fracture Study (EFFECT). J Bone Miner Res 26:881–893.
16. Baumgaertner MR, Curtin SL, Lindskog DM, et al. 1995.
The value of the tip-apex distance in predicting failure of
fixation of peritrochanteric fractures of the hip. J Bone Joint
Surg Am 77A:1058–1064.
17. Bevill G, Farhamand F, Keaveny TM. 2009. Heterogeneity
of yield strain in low-density versus high-density human tra-
becular bone. J Biomech 42:2165–2170.
18. Kopperdahl DL, Keaveny TM. 1998. Yield strain behavior of
trabecular bone. J Biomech 31:601–608.
19. Wu CC, Shih CH, Lee MY, Tai CL. 1996. Biomechanical
analysis of location of lag screw of a dynamic hip screw in
treatment of unstable intertrochanteric fracture. J Trauma
41:699–702.
20. Helwig P, Faust G, Hindenlang U, et al. 2009. Finite ele-
ment analysis of four different implants inserted in different
positions to stabilize an idealized trochanteric femoral frac-
ture. Injury 40:288–295.
21. Cibulka MT. 2004. Determination and significance of femo-
ral neck anteversion. Phys Ther 84:550–558.
22. Wright D, Whyne C, Hardisty M, et al. 2011. Functional and
anatomic orientation of the femoral head. Clin Orthop Relat
Res 469:2583–2589.
23. Asala SA. 2001. Sex determination from the head of the fe-
mur of South African whites and blacks. Forensic Sci Int
117:15–22.
24. Sun SS, Ma HL, Liu CL, et al. 2008. Difference in femoral
head and neck material properties between osteoarthritis
and osteoporosis. Clin Biomech (Bristol, Avon) 23:S39–S47.
25. Geller JA, Saifi C, Morrison TA, et al. 2010. Tip-apex dis-
tance of intramedullary devices as a predictor of cut-out fail-
ure in the treatment of peritrochanteric elderly hip
fractures. Int Orthop 34:719–722.
26. Schileo E, Taddei F, Cristofolini L, et al. 2008. Subject-
specific finite element models implementing a maximum
principal strain criterion are able to estimate failure risk
and fracture location on human femurs tested in vitro. J Bio-
mech 41:356–367.
27. Ramos A, Simoes JA. 2006. Tetrahedral versus hexahedral
finite elements in numerical modelling of the proximal
femur. Med Eng Phys 28:916–924.
28. Heller MO, Bergmann G, Kassi JP, et al. 2005. Determina-
tion of muscle loading at the hip joint for use in pre-clinical
testing. J Biomech 38:1155–1163.
29. Bergmann G, Deuretzbacher G, Heller M, et al. 2001. Hip
contact forces and gait patterns from routine activities.
J Biomech 34:859–871.