Femoral rotation unpredictably affects radiographic anatomical lateral distal femoral angle measurements


Miles, J.E.

Veterinary and Comparative Orthopaedics and Traumatology 29(2): 156-159

2016


To describe the effects of internal and external femoral rotation on radiographic measurements of the anatomical lateral distal femoral angle (a-LDFA) using two methods for defining the anatomical proximal femoral axis (a-PFA). Digital radiographs were obtained of 14 right femora at five degree intervals from 10° external rotation to 10° internal rotation. Using freely available software, a-LDFA measurements were made using two different a-PFA by a single observer on one occasion. Mean a-LDFA was significantly greater at 10° external rotation than at any other rotation. The response of individual femora to rotation was unpredictable, although fairly stable within ±5° of zero rotation. Mean a-LDFA for the two a-PFA methods differed by 1.5°, but were otherwise similarly affected by femoral rotation. If zero femoral elevation can be achieved for radiography, a-LDFA measurements do not vary much with mild femoral rotation (±5°). Outside of this range, a-LDFA varies unpredictably with femoral rotation.

Brief
Communication
156
Femoral
rotation
unpredictably
affects
radiographic
anatomical
lateral
distal
femoral
angle
measurements
James
E.
Miles
Department
of
Veterinary
Clinical
and
Animal
Sciences,
Faculty
of
Health
and
Medical
Sciences,
University
of
Copenhagen,
Frederiksberg
C,
Denmark
Keywords
Femur,
varus,
radiography,
canine
Objective:
To
describe
the
effects
of
internal
and
external
femoral
rotation
on
radio-
graphic
measurements
of
the
anatomical
lat-
eral
distal
femoral
angle
(a-LDFA)
using
two
methods
for
defining
the
anatomical
proxi-
mal
femoral
axis
(a-PFA).
Methods:
Digital
radiographs
were
obtained
of
14
right
femora
at
five
degree
intervals
from
10°
external
rotation
to
10°
internal
ro-
tation.
Using
freely
available
software,
a-LDFA
measurements
were
made
using
two
different
a-PFA
by
a
single
observer
on
one
occasion.
Correspondence
to:
James
E.
Miles
Department
of
Veterinary
Clinical
and
Animal
Sciences
Faculty
of
Health
and
Medical
Sciences
University
of
Copenhagen
DyrImgevej
16
1870
Frederiksberg
C
Denmark
Phone:
+45
2057
0840
Fax:
+45
3533
2929
E-mail:
jamiasund.ku.dk
Introduction
Measurement
of
the
anatomical
lateral
dis-
tal
femoral
angle
(a-LDFA)
may
be
relevant
for
planning
the
surgical
management
of
medial
patellar
luxation
whenever
femoral
malalignment
is
considered
to
be
causative
(1).
Femoral
rotation
can
affect
radio-
graphic
a-LDFA
measurement,
with
exter-
nal
rotation
increasing
measured
values
Results:
Mean
a-LDFA
was
significantly
greater
at
10°
external
rotation
than
at
any
other
rotation.
The
response
of
individual
fe-
mora
to
rotation
was
unpredictable,
al-
though
fairly
stable
within
±5°
of
zero
ro-
tation.
Mean
a-LDFA
for
the
two
a-PFA
meth-
ods
differed
by
15°,
but
were
otherwise
similarly
affected
by
femoral
rotation.
Clinical
significance:
If
zero
femoral
elev-
ation
can
be
achieved
for
radiography,
a-LDFA
measurements
do
not
vary
much
with
mild
femoral
rotation
(±5°).
Outside
of
this
range,
a-LDFA
varies
unpredictably
with
femoral
rotation.
Vet
Comp
Orthop
Traumatol
2016;
29:156-159
http://dx.doi.org/10.3415NC0T-15-06-0107
Received:
June
19,
2015
Accepted:
October
18,
2015
Epub
ahead
of
print:
January
20,
2016
and
internal
rotation
decreasing
measured
values
(1).
Measurement
of
a-LDFA
requires
identification
of
an
anatomical
proximal
femoral
axis
(a-PFA)
and
a
distal
trans-
condylar
axis.
Several
methods
for
identify-
ing
a-PFA
have
been
described,
all
of
which
produce
slightly
different
values
for
a-LDFA,
but
which
largely
respond
simi-
larly
to
femoral
elevation
(2).
The
impact
of
femoral
rotation
on
a-LDFA
has
been
evaluated
experimentally
using
surface
rendered
computed
tomo-
graphic
(CT)
reconstructions
(3).
A
near-
linear
response
was
observed,
with
a-LDFA
values
decreasing
by
approximately
one
de-
gree
for
every
two
degrees
of
rotation
over
a
range
of
20°
external
rotation
to
20°
inter-
nal
rotation
(3).
The
reconstruction
criteria
used
for
this
study
presented
the
femur
in
a
similar
orientation
to
approximately
24°
of
elevation,
although
it
is
generally
recom-
mended
that
femoral
elevation
be
zero
de-
grees
for
radiographic
investigations
of
a-LDFA
(4).
This
study
investigated
the
effect
of
li-
mited
femoral
rotation
at
the
accepted
op-
timal
femoral
elevation
on
radiographic
measurements
of
a-LDFA.
Two
methods
of
identifying
the
a-PFA
were
used
and
com-
pared,
to
identify
possible
advantages
with
respect
to
femoral
rotation.
Material
and
methods
Fourteen
right
canine
femoral
specimens
were
obtained
from
the
University
of
Co-
penhagen
(Anatomy
Section,
Department
of
Veterinary
Clinical
and
Animal
Sciences,
Faculty
of
Health
and
Medical
Sciences).
The
intercondylar
width,
de-
fined
as
the
distance
between
the
most
caudal
aspects
of
the
condyles,
was
measured
for
later
calculation
of
rotation
accuracy.
Femora
were
positioned
at
zero
elev-
ation,
defined
as
horizontal
alignment
of
the
lesser
trochanter
and
supracondylar
eminence,
with
the
aid
of
foam
wedges
as
previously
described
(1,
2).
Schattauer
2016
Vet
Comp
Orthop
Traumatol
2/2016
113
112
t
t
I
D
-
Method
TO
0—Method
SY
CT
method
Ext10°
Ext5°
Optimal
Int5°
Int10°
Femoral
positioning
110
-
100
LJ-
90
so
Ext10°
Ext5°
Optimal
Int5°
Int10°
Femoral
positioning
110
100
0
90
so
J.
E.
Miles:
Femoral
rotation
affects
radiographic
anatomical
lateral
distal
femoral
angle
Figure
1
Methods
for
determin-
ing
the
anatomical
lateral
distal
femoral
angle
(a-LDFA).
A)
Method
TO
defines
the
anatomical
proximal
femoral
axis
(a-PFA)
as
a
line
passing
through
the
centres
of
circles
centred
on
the
diaphy-
sis
at
one-third
and
one-half
of
the
femo-
ral
length.
The
angle
between
a-PFA
and
the
distal
transcondy-
lar
axis
laterally
is
the
a-LDFA.
B)
Method
SY
defines
the
a-PFA
as
a
line
passing
through
the
centres
of
circles
centred
on
the
proxi-
mal
femoral
metaphy-
sis
and
the
diaphysis
at
one-half
of
the
fe-
moral
length.
The
angle
between
a-PFA
and
the
distal
trans-
condylar
axis
laterally
is
the
a-LDFA.
Initial
digital
radiographs
were
obtained
with
the
caudal
aspects
of
the
femoral
con-
dyles
placed
in
contact
with
the
radio-
graphic
cassette:
this
position
was
defined
as
zero
rotation.
Subsequently,
radiographs
were
obtained
with
internal
and
external
rotations
of
five
degrees
and
10°,
achieved
by
the
use
of
thin
plastic
wedges
used
for
flooring
installation.
Wedge
thicknesses
of
2
mm,
3
mm
and
4
mm
were
used
accord-
ing
to
the
measured
intercondylar
width.
Radiographs
were
saved
as
DICOM
files.
Files
were
anonymized
using
commer-
cial
softwarea
and
then
randomized
and
a
Sante
DICOM
Editor:
Santesoft,
Athens,
Greece
b
Ant
Renamer:
Developed
by
Antoine
Potten,
avail-
able
at:
http://www.antp.be/
c
Synedra
View
Personal:
Synedra
information
tech-
nologies
GmbH,
Innsbruck,
Austria
read
on
one
occasion
using
freely
available
softwareb
,
c.
Femoral
varus
was
measured
using
two
previously
described
methods
to
define
the
anatomical
proximal
femoral
axis
(2).
The
SymAx
method
(SY)
utilized
the
centres
of
circles
centred
on
the
proxi-
mal
femoral
metaphysis
and
the
femoral
diaphysis
at
one-half
the
femoral
length,
and
the
Tomlinson
method
(TO)
utilized
the
centres
of
circles
centred
on
the
femoral
diaphysis
at
one-third
and
one-half
the
fe-
moral
length
from
proximal
to
distal
(10.Figure
1).
Data
were
analysed
using
a
statistical
software
packaged.
A
two-way
repeated
measures
analysis
of
variance
with
pairwise
d
IBM
SPSS
Statistics
for
Wmdows
20,
Release
2011:
IBM
Corp.,
Armonk,
New
York,
USA
Figure
2
The
effect
of
femoral
rotation
on
the
radiographic
anatomical
lateral
distal
femoral
angle
(a-LDFA).
A)
Mean
a-LDFA
values
for
method
TO
(squares
and
dashed
lines)
and
method
SY
(circles
and
solid
lines),
using
the
axis
definitions
shown
in
10•Figure
1
at
the
five
studied
femoral
rotations.
Error
bars
represent
one
stan-
dard
deviation.
The
grey
dotted
line
(CT
method)
represents
mean
values
for
a-LDFA
for
six
femora
derived
from
a
study
of
the
effect
of
rotation
on
CT
measurements
(3).
B)
The
underlying
changes
in
a-LDFA
for
14
femora
using
method
SY
are
shown
for
the
five
studied
femoral
rotations,
de-
monstrating
the
inconsistent
response
in
a-LDFA
to
femoral
rotation.
Int.
=
internal,
Ext.
=
external
comparisons
was
used
to
compare
the
ef-
fect
of
rotation
for
each
varus
measure-
ment
method,
using
Bonferroni
correction
to
compensate
for
multiple
comparisons.
The
threshold
for
statistical
significance
was
set
at
five
percent.
Mean
calculated
internal
and
external
ro-
tation
angles
of
five
degrees
(range:
4.6°
-
Results
Vet
Comp
Orthop
Traumatol
2/2016
Schattauer
2016
J.
E.
Miles:
Femoral
rotation
affects
radiographic
anatomical
lateral
distal
femoral
angle
158
6.0°)
and
9.7°
(range:
9.0°
-
10.8°)
were
ob-
tained,
based
on
the
intercondylar
dimen-
sions
of
the
femora
in
this
study.
Mean
fe-
moral
length
was
19.5
cm
(±1.5
cm).
Mauchley's
test
indicated
that
the
assump-
tion
of
sphericity
for
the
effect
of
rotation
had
been
violated
(x
2
(9)
=
63.56,
p
<0.001),
and
the
degrees
of
freedom
were
corrected
using
Greenhouse-Geisser
esti-
mates
of
sphericity
(e
=
0.43).
Both
femoral
rotation
and
choice
of
measurement
method
significantly
(p
<0.001)
affected
measured
a-LDFA.
Pairwise
comparisons
showed
that
measured
a-LDFA
at
10°
ex-
ternal
rotation
differed
significantly
(cor-
rected
p
<0.001)
from
values
at
five
degrees
of
external
or
internal
rotation,
and
zero
rotation,
but
not
10°
internal
rotation.
Mean
a-LDFA
values
yielded
by
method
SY
were
1.5°
(95%
confidence
interval:
1.1°
-
1.8°)
lower
than
using
method
TO.
Graphical
analysis
showed
that
the
changes
due
to
rotation
were
not
consistent
for
all
femora,
with
the
mean
a-LDFA
val-
ues
concealing
both
positive
and
negative
trends
in
measured
a-LDFA
(
Figure
2).
Discussion
The
key
fmding
of
this
study
was
the
un-
predictable
response
in
radiographic
measured
varus
to
induced
femoral
ro-
tation.
Although
femoral
rotation
during
radi-
ography
is
widely
considered
to
affect
sub-
sequent
measurements
of
a-LDFA,
the
only
pre-existing
investigation
of
this
phenom-
enon
is
a
study
of
CT
reconstructions
which
were
positioned
to
optimize
the
view
of
the
weight-bearing
surfaces
of
the
femo-
ral
condyles
(3).
A
study
of
the
effect
of
fe-
moral
elevation
on
radiographic
measure-
ments
of
a-LDFA
noted
the
confounding
effect
of
the
distal
aspects
of
the
trochlear
ridges
on
accurate
identification
of
the
fe-
moral
condylar
outlines
at
low
femoral
elevations
(5).
The
data
presented
here
sug-
gests
that
this
confounding
effect
leads
to
unexpected
and
unpredictable
variations
in
measured
a-LDFA
at
zero
elevation,
which
is
in
stark
contrast
to
the
linear
effect
seen
at
moderate
femoral
elevation
using
CT
(3).
While
rotation
from
five
degrees
exter-
nal
to
five
degrees
internal
rotation
pro-
duces
an
approximately
eight
degree
de-
crease
in
measured
a-LDFA
with
elevated
CT
reconstructions,
the
effects
of
rotation
appear
minimal
on
radiographs
taken
at
zero
elevation
(3).
It
is
likely
that
the
changes
in
articular
surface
orientation
seen
with
elevated
CT
reconstructions
also
occurred
in
the
radio-
graphs
in
this
study,
but
were
masked
at
low
rotations
by
superimposition
of
the
distal
trochlear
ridges.
This
presents
a
co-
nundrum
for
the
clinician.
If
the
radio-
graph
is
taken
at
zero
(optimal)
elevation,
then
rotation
under
±5°
appears
to
have
little
consequence
for
subsequent
measure-
ments
of
a-LDFA.
However,
if
the
femur
is
elevated
during
radiography,
then
even
li-
mited
rotation
could
have
clinically
impor-
tant
effects
on
a-LDFA
measurements.
Various
methods
to
increase
the
chances
of
obtaining
a
zero
elevation
image,
such
as
elevation
of
the
dorsum
of
the
patient,
con-
trolling
femoral
positioning
with
fluor-
oscopy
prior
to
radiography
and
the
use
of
horizontal
beam
radiography
have
been
described
(1,
4).
However,
in
most
in-
stances,
the
clinician
remains
uncertain
as
to
the
degree
of
elevation
actually
present
and
therefore
to
the
importance
of
any
fe-
moral
rotation.
The
differences
in
measured
a-LDFA
angle
between
methods
have
been
pre-
viously
described
and
no
advantage
or
dis-
advantage
could
be
ascribed
to
either
method
in
relation
to
the
effects
of
femoral
rotation
on
their
use
(2).
None
of
the
femora
used
in
this
study
originated
from
chondrodystrophic
dogs,
and
generalization
of
the
results
presented
here
to
other
femoral
shapes,
particularly
those
with
pronounced
procurvatum,
should
be
made
with
caution.
Rotation
in
this
study
was
restricted
to
±10°,
as
clinical
experience
suggested
that
beyond
these
limits
radiographs
would
be
dearly
unacceptable
for
measurement
pur-
poses,
but
that
clinicians
might
well
be
tempted
to
use
radiographs
with
a
more
re-
stricted
amount
of
rotation.
Clinicians
should
be
wary
of
the
inter-
action
of
femoral
elevation
and
rotation.
If
optimal
(zero)
elevation
can
be
assured,
the
effects
of
mild
(±5°)
femoral
elevation
on
a-LDFA
appear
to
be
minimal.
If
elevation
is
present,
then
clinically
important
changes
in
measured
a-LDFA
may
be
seen.
Acknowledgements
The
authors
thank
Professor
Thomas
Eriksen
for
the
critical
reading
of
the
manuscript,
and
the
Radiology
section
at
the
Department
of
Veterinary
Clinical
and
Animal
Sciences
for
their
assistance
with
image
collection.
Institutional
funding
was
used
for
this
project.
Preliminary
results
were
presented
as
a
clinical
research
ab-
stract
and
poster
at
the
British
Small
Ani-
mal
Veterinary
Association
Congress,
9th-12th
April,
2015.
Conflict
of
interest
There
are
no
conflicts
of
interest
to
declare
for
this
paper.
References
1.
Swiderski
JK,
Radecki
SV,
Park
RD,
et
aL
Compari-
son
of
radiographic
and
anatomic
femoral
varus
angle
measurements
in
normal
dogs.
Vet
Surg
2008;
37:43-48.
2.
Miles
JE,
Mortensen
M,
Svalastoga
EL,
et
al.
A
comparison
of
anatomical
lateral
distal
femoral
angles
obtained
with
four
femoral
axis
methods
in
canine
femora.
Vet
Comp
Orthop
Traumatol
2015;
28:
193-198.
3.
Oxley
B,
Gemmill
TJ,
Pink
J,
et
al.
Precision
of
a
novel
computed
tomographic
method
for
quantifi-
cation
of
femoral
varus
in
dogs
and
an
assessment
of
the
effect
of
femoral
malpositioning.
Vet
Surg
2013;
42:
751-758.
4.
Dudley
RM,
Kowaleski
MP,
Drost
WT,
et
aL
Radiographic
and
computed
tomographic
deter-
mination
of
femoral
varus
and
torsion
in
the
dog.
Vet
Radiol
Ultrasound
2006;
47:
546-552.
5.
Jackson
GM,
Wendelburg
KL.
Evaluation
of
the
ef-
fect
of
distal
femoral
elevation
on
radiographic
measurement
of
the
anatomic
lateral
distal
femoral
angle.
Vet
Surg
2012;
41:
994-1001.
Schattauer
2016
Vet
Comp
Orthop
Traumatol
2/2016