CLINICAL ORTHOPAEDICS AND RELATED RESEARCH
Number 392, pp. 68–74
Varus Tibial Joint Line Obliquity
A Potential Cause of Femoral Component Malrotation
Mark W. Pagnano, MD; and Arlen D. Hanssen, MD
In 60 consecutive total knee arthroplasties done
in 52 patients with primary osteoarthritis and
varus or neutral tibiofemoral alignment, the
posterior condylar angle was calculated intra-
operatively and averaged 3.98
(range, 0
–9
).
Eighteen knees had a posterior condylar angle
value less than 3
whereas 27 knees had a pos-
terior condylar angle value of 5
or greater. Fi-
nal rotational alignment of the femoral compo-
nent was set parallel to the transepicondylar
axis. Only one of these 60 knees required a lat-
eral retinacular release for proper patellar
tracking during the knee arthroplasty. When
compared with three previously defined angles
measured on the radiographs taken preopera-
tively, only the tibial plateau-tibial shaft angle
values were correlated significantly with the
value of the posterior condylar angle. As the tib-
ial varus joint line obliquity increased, there was
a distinct tendency for the transepicondylar axis
to be rotated more externally relative to the pos-
terior condylar axis. This variance suggests that
the use of the posterior condylar axis as a rota-
tional reference is inappropriate in many knees
with arthritis with varus or neutral tibiofemoral
alignment. In particular, varus tibial joint line
obliquity of more than 4
increases the likeli-
hood of femoral component malrotation when
the posterior femoral condyles are used to ref-
erence femoral component rotation.
From the Department of Orthopedic Surgery, Mayo
Clinic and Mayo Foundation, Rochester, MN.
Reprint requests to Mark W. Pagnano, MD, Mayo Clinic,
200 First Street Southwest, Rochester, MN 55905.
68
© 2001 Lippincott Williams & Wilkins, Inc.
Of the multiple variables that contribute to
proper patellar tracking after total knee arthro-
plasty, proper rotational alignment of the
femoral component has received considerable
attention.1–6,8–10,12–20 At least four different
methods have been proposed to determine
proper rotational alignment of the femoral com-
ponent.9,11,14,19 Those methods include the use
of the transepicondylar axis, 3
external rota-
tion based on the posterior condylar axis, the
anteroposterior (AP) trochlear line, and adjust-
ment of external rotation to create a symmetric
flexion gap after the completion of collateral
ligament balancing.
While doing total knee arthroplasty (TKA),
it is acknowledged that the posterior condy-
lar axis is an unreliable reference for rota-
tional orientation of the femoral component in
the valgus knee.2,9,12–14,20 It is recommended
that one of the other available methods be
used for femoral rotational alignment in these
valgus knees. In the varus knee, the posterior
condylar axis averages 3
internal rotation
when compared with the other measurement
methods.4,10 As a result, many knee instru-
mentation systems or prosthetic designs have
incorporated an average external rotational
correction of the posterior condylar axis to
balance flexion gap symmetry.11–14 With that
external rotation modification, the posterior
condylar surfaces of the femur (the posterior
condylar axis) still are used as the reference
Number 392
November, 2001
for rotation in many primary total knee
arthroplasties.
The posterior condylar angle is defined by
the angle formed by the intersection of the
transepicondylar axis and the posterior condy-
lar axis.9 In 88 knees with osteoarthritis and
varus or neutral deformity, the posterior
condylar angle averaged 3.3
(range, 0
–8
)
whereas the posterior condylar angle averaged
5.4
(range, 0
–10
) in 19 valgus knees.9 That
study revealed that some varus knees and
knees with a neutral deformity have a posterior
condylar angle that exceeds the average value
observed for the valgus knee. Furthermore, the
outlying values in the range of variance for the
varus knees and knees with a neutral deformity
almost equaled the outlying values observed in
the valgus knees.9
The current study was initiated to evaluate
the variability of the posterior condylar angle
in knees with osteoarthritis and neutral or varus
tibiofemoral alignment. Specifically, it was a
goal of this study to identify any characteristics
of those knees that deviated markedly from the
average 3
posterior condylar angle value. It
was hypothesized that knees could be identi-
fied preoperatively so that the surgeon could
predict when posterior condylar referencing
systems might be unreliable for proper rota-
tional alignment of the femoral component in
the varus knees or neutrally aligned knee with
osteoarthritis.
MATERIALS AND METHODS
The patients selected for inclusion in this study had
primary osteoarthritis of the knee with varus or
neutral tibiofemoral axial alignment. Varus align-
ment was defined as axial limb alignment less
than 0
and neutral alignment was defined as 0
to
7
. The authors excluded patients with rheuma-
toid arthritis and other inflammatory disorders such
as psoriatic arthritis or systemic lupus erythemato-
sus. Additional exclusion criteria included a history
or evidence of posttraumatic deformity (intraartic-
ular, femoral, or tibial), prior periarticular os-
teotomy (distal femoral, high tibial, or hip), septic
arthritis of the knee, prior ligamentous reconstruc-
tion, or any knee with valgus axial alignment. Pre-
Tibial Varus Deformity 69
vious arthroscopic debridement of the knee was not
a cause for exclusion.
Between February 1998 and February 1999, the
authors did 60 consecutive total knee arthroplas-
ties in patients who met the criteria for inclusion in
this study. The 60 total knee arthroplasties (53 pa-
tients) were done in 21 women and 32 men. In the
seven patients who had simultaneous bilateral to-
tal knee arthroplasties, there was one woman and
six men. The average age of the patients at the time
of knee arthroplasty was 68.2 years (range, 44–88
years). The right side was replaced in 26 knees (22
patients) and the left side was replaced in 34 (30
patients).
The total knee arthroplasties all were done with
a midline skin incision and a medial parapatellar
arthrotomy. Femoral and tibial sizing was done us-
ing design specific instrumentation provided by the
manufacturer. All of the knee arthroplasties were of
the cemented posterior-stabilized Sigma PFC de-
sign (Johnson and Johnson, Raynham, MA). The
tibial plateau was cut perpendicular to the tibial
shaft axis and with a neutral posterior slope. The
distal femur was resected using an intramedullary
guide to provide a flat distal femoral surface to al-
low intraoperative measurements.
Intraoperative measurements included the poste-
rior condylar axis, the surgical epicondylar axis, the
posterior condylar angle, the patellar height before
patellar osteotomy, and the height of the patella-
patellar component construct. The posterior condy-
lar axis was identified with the aid of a neutral pos-
terior condylar surface-referencing guide without
additional external rotation built into the guide. The
posterior condylar angle then was determined by
observing the difference between the surgical epi-
condylar axis and the posterior condylar axis as de-
scribed by Griffin et al9 (Fig 1). The final rotation of
the femoral component was set parallel to the surgi-
cal epicondylar axis. The tibial component was ro-
tated to align the midportion of the tibial component
with the anterior tibial crest. All patellas were resur-
faced with a cemented all-polyethylene domed
patellar component. The no-thumbs test was used to
assess patellar tracking. Satisfactory tracking in-
cluded a patella that was centralized on the trochlea
throughout a range of motion from 0
to 95
, with-
out a tendency for tilt or subluxation, and contact of
the patellar button with the medial and lateral fem-
oral condyles at 90
flexion. The tourniquet was de-
flated and patellar tracking was reassessed in any
patients in whom a lateral release was considered.

70 Pagnano and Hanssen
Preoperative radiographic analysis included mea-
surement of the tibiofemoral axis on a full-length
film obtained with the patients weightbearing with
their feet placed in a standardized position of neu-
tral rotation. Additional measurements included
calculation of three angles described by Cooke et
al.7 These angles include the femoral shaft-tibial
shaft angle, the femoral shaft-transcondylar angle,
and the tibial plateau-tibial shaft angle as shown in
Clinical Orthopaedics
and Related Research
Fig 1. The posterior angle is
formed by the intersection of
the posterior condylar axis (a
line referenced on the posterior
aspect of the femoral condyles)
and the surgical epicondylar
(transepicondylar) axis.
Figure 2. The values obtained for the posterior
condylar angles then were analyzed according to
the definitions of Cooke et al,7 which included sig-
nificance of an femoral shaft-transcondylar angle
greater than 9
, a tibial plateau-tibial shaft angle
more negative than 3.3
, and an femoral shaft-tibial
shaft angle less than 3.9
.7 Statistical analysis was
done using Student’s t test with the statistical sig-
nificance set at p  .05.
Fig 2. A schematic drawing
shows the three measurement
angles obtained on the preoper-
ative full length radiograph ob-
tained with the patient standing:
the femoral shaft-transcondylar
(FSXC) angle, the femoral shaft-
tibial shaft (FSTS) angle, and the
tibial plateau-tibial shaft (TPTS)
angle.
Number 392
November, 2001
RESULTS
In the overall group, the value of the posterior
condylar angle averaged 3.98
(range, 0
–9
).
Eighteen knees had a posterior condylar angle
value less than 3
whereas 27 knees had a pos-
terior condylar angle value of 5
or greater.
The femoral shaft-tibial shaft angle averaged
3.48
(range, 0
– 12
), the femoral shaft-
transcondylar angle averaged 6.28
(range,
3
–11
), and the tibial plateau-tibial shaft an-
gle averaged 4
(range, 0
– 10
). There
were no statistically significant differences in
any of the values when analyzed for gender,
side of surgery, or for the patient groups cate-
gorized as having varus or neutral axial align-
ment. When the values for the posterior
condylar angles were analyzed according to
the subgroups defined by the preoperative ra-
diographic measurements of Cooke et al,7 the
only angle of statistical significance was the
tibial plateau-tibial shaft angle (p  0.05).
The posterior condylar angle values also
were scatterplotted against the tibial plateau-
tibial shaft angle, femoral shaft-tibial shaft an-
gle, and femoral shaft-transcondylar angle val-
ues for each knee. The correlation between the
tibial plateau-tibial shaft angle and the poste-
rior condylar angle is readily apparent (Fig
3A). There was no significant correlation be-
tween the femoral shaft-transcondylar angle
and femoral shaft-tibial shaft angle and the
posterior condylar angle. (Fig 3B–C).
The patellar height before patellar os-
teotomy, measured intraoperatively, averaged
23.9 mm (range, 19–29 mm) in the overall
group whereas the patellar-patellar compo-
nent construct measurement averaged 23.6
mm (range, 19–28 mm). None of the patellar
heights were altered more than 2 mm and only
seven patellas were thicker than their pre-
osteotomy height. Only one (1.67%) of the 60
knees required a lateral retinacular release to
optimize patellar tracking during the total
knee arthroplasty. The values for the various
measurements in this patient were as follows:
tibial plateau-tibial shaft (3
); femoral shaft-
tibial shaft (1
); femoral shaft-transcondylar
Tibial Varus Deformity 71
(7
); posterior condylar angle (3
); patellar
height before patellar osteotomy (25 mm); and
patellar-patellar component construct height
(24 mm).
DISCUSSION
It is not known how much variance in the ro-
tational position of the femoral component in
total knee arthroplasty is acceptable. It is clear,
however, that the incidence of patellar mal-
tracking is affected by numerous surgical
technique variables and that optimal rotational
position of the femoral component is desir-
able.1–6,8–10,12–20 Although acknowledged to
be unreliable in the valgus knee, posterior
femoral condylar referencing guides still are
used routinely in many primary total knee
arthroplasties done in knees with varus or neu-
tral tibiofemoral axial alignment.
A recent study14 described the transepi-
condylar axis as the most consistent method of
measurement, of the four methods tested, to
recreate a balanced flexion space. Further-
more, in that study, 3
external rotation rela-
tive to the posterior condyles was the least
consistent method of measurement, especially
for valgus knees.14 The current study suggests
that posterior femoral condylar referencing
guides also are unreliable for many knees with
osteoarthritis and varus or neutral alignment.
In the current study, if femoral rotation had
been set using the posterior femoral condyles
with an empiric 3
external rotational adjust-
ment, 18 knee arthroplasties (30%) would
have been placed with excessive external rota-
tion whereas 27 knee arthroplasties (45%)
would have been implanted in 2
or more in-
ternal rotation.
Cooke et al7 described a specific deformity
in some knees with arthritis and varus align-
ment that have valgus angulation of the femo-
ral joint surface with proximal tibial varus (Fig
4). They indicated that recognition of this de-
formity predicts the need for specific operative
correction because patients with this entity
have a predilection toward internal rotation
malposition of the femoral component. Their
 Clinical Orthopaedics
72 Pagnano and Hanssen and Related Research

Fig 3A–C. (A) This scatterplot shows the

correlation between the posterior condy-
lar angle and the tibial plateau-tibial shaft
(TPTS) angle. (B) The lack of correlation
between the posterior condylar angle
and the femoral shaft-transcondylar
(FSXC) angle can be seen in this scat-
terplot. (C) The lack of correlation be-
tween the posterior condylar angle and
the femoral shaft-tibial shaft (FSTS) an-
C gle is shown.
Number 392
November, 2001
Fig 4. An anteroposterior radiograph shows a
knee with osteoarthritis in neutral alignment with
tibial varus joint line obliquity.
method of measurement defined four specific
measurements, three of which can be deter-
mined using standard radiographic techniques
and one measurement that requires a special-
ized radiographic technique (questor precision
radiography).7 In the current study, only those
measurements obtained by standard radi-
ographic techniques were used in an effort to
define the radiographic variables routinely ac-
cessible to the surgeon doing total knee arthro-
plasty.
Presumably, the reason for the propensity
for altered posterior condylar angles in patients
with increased tibial varus joint line obliquity
is that the distal femur has a corresponding
joint line obliquity, particularly posteriorly,
that represents a relative hyperplasia of the me-
Tibial Varus Deformity 73
dial femoral condyle. As with the valgus knee,
this femoral condyle hypoplasia places the
posterior condylar reference guides in exces-
sive internal rotation. Of the three measure-
ment angles of Cooke et al7 that were analyzed
in the current study, the femoral shaft-tibial
shaft and femoral shaft-transcondylar angles
were not correlated with variance of the poste-
rior condylar angle. It is unclear why the fem-
oral shaft-transcondylar angle, a measurement
of distal femoral joint line obliquity, did not
correlate with posterior condylar angle vari-
ance in the current study and this is in contrast
to the study of Cooke et al.7
As with Cooke et al,7 the tibial plateau-
tibial shaft angle did correlate positively with
the posterior condylar angle and in general, the
more negative the value for the tibial plateau-
tibial shaft angle, the more likely the value of
the posterior condylar angle was increased
(Fig 3A). This suggests that a simple preoper-
ative calculation of the tibial plateau-tibial
shaft angle on a radiograph obtained with the
patient standing will help alert the surgeon to
those knees in which the posterior condylar
axis may be an unreliable reference for rota-
tional position of the femoral component.
The transepicondylar axis was selected as
the rotational reference point for the actual
placement of the femoral components in the
current study because this is the preferred
method in primary and revision total knee
arthroplasty. Cooke et al7 also recommended
the use of the surgical epicondylar axis to de-
termine the correct femoral rotational position
in patients with tibial varus joint line obliquity.
The low rate of lateral retinacular release in the
current study, one of 60 knees, suggests that
this femoral rotational position combined with
the other surgical technique variables such as
tibial component rotation, limb alignment,
patellar thickness, and patellar component po-
sition facilitated proper patellar tracking.
In a study comparing the transepicondylar
axis, the posterior condylar axis, and the AP
line in normal knees and knees with os-
teoarthritis, it was reported that the AP axis
was rotated externally and was less reliable

74 Pagnano and Hanssen
than the posterior condylar axis for use in
alignment for total knee arthroplasty in knees
with medial femorotibial arthritis.12 However,
the transepicondylar axis measurement was
consistent in normal knees and knees with os-
teoarthritis.13
Although it is not possible, based on the
data in the current study, to advocate that the
transepicondylar axis is the best measurement
method, it does seem reasonable to suggest that
for many of these knees with varus or neutral
alignment the use of the posterior condylar axis
is inconsistent and potentially unreliable for
establishing femoral component rotation. In
particular, varus tibial joint line obliquity of
more than 4
should alert the surgeon to the
possibility of femoral component malposition
during total knee arthroplasty.
References
1. Akagi M, Matsusue Y, Mata T, et al: Effect of rota-
tional alignment on patellar tracking in total knee
arthroplasty. Clin Orthop 366:155–63, 1999.
2. Anouchi YS, Whiteside LA, Kaiser AD, et al: The
effects of axial rotational alignment of the femoral
component on knee stability and patellar tracking in
total knee arthroplasty demonstrated on autopsy
specimens. Clin Orthop 287:170–177, 1993.
3. Arima J, Whiteside LA, McCarthy DS, et al: Femoral
rotational alignment, based on the anteroposterior axis,
in total knee arthroplasty in a valgus knee: A technical
note. J Bone Joint Surg 77A:1331–1334, 1995.
4. Berger RA, Crossett LS, Jacobs JJ, et al: Malrotation
causing patellofemoral complications after total
knee arthroplasty. Clin Orthop 356:144–153, 1998.
5. Berger RA, Rubash HE, Seel MJ, et al: Determining
the rotational alignment of the femoral component in
total knee arthroplasty using the epicondylar axis.
Clin Orthop 286:40–47, 1993.
6. Churchill DL, Incavo SJ, Johnson CC, et al: The
transepicondylar axis approximates the optimal flex-
ion axis of the knee. Clin Orthop 356:111–118, 1998.
Clinical Orthopaedics
and Related Research
7. Cooke TD, Pichora D, Siu D, et al: Surgical impli-
cations of varus deformity of the knee with obliquity
of joint surfaces. J Bone Joint Surg 71B:560–565,
1989.
8. Eckhoff DG, Piatt BE, Gnadinger CA, et al: Assess-
ing rotational alignment in total knee arthroplasty.
Clin Orthop 318:176–181, 1995.
9. Griffin FM, Insall JN, Scuderi GR: The posterior
condylar angle in osteoarthritic knees. J Arthroplasty
13:812–815, 1998.
10. Griffin FM, Math K, Scuderi GR, et al: Anatomy of
the epicondyles of the distal femur: MRI analysis of
normal knees. J Arthroplasty 15:354–359, 2000.
11. Hungerford DS, Kenna RV: Preliminary experience
with a total knee prosthesis with porous coating used
without cement. Clin Orthop 176:95–107, 1983.
12. Nagamine R, Miura H, Inoue Y, et al: Reliability of
the anteroposterior axis and the posterior condylar
axis for determining rotational alignment of the fem-
oral component in total knee arthroplasty. J Orthop
Sci 3:194–198, 1998.
13. Nagamine R, White SE, McCarthy DS, et al: Effect
of rotational malposition of the femoral component
on knee stability kinematics after total knee arthro-
plasty. J Arthroplasty 10:265–270, 1995.
14. Olcott CW, Scott RD: A comparison of 4 intraoper-
ative methods to determine femoral component rota-
tion during total knee arthroplasty. J Arthroplasty
15:22–26, 2000.
15. Poilvache PL, Insall JN, Scuderi GR, et al: Rota-
tional landmarks and sizing of the distal femur in to-
tal knee arthroplasty. Clin Orthop 331:35–46, 1996.
16. Rhoads DD, Noble PC, Reuben JD, et al: The effect
of femoral component position on patellar tracking
after total knee arthroplasty. Clin Orthop 260:43–51,
1990.
17. Star MJ, Kaufman KR, Irby SE, et al: The effects of
patellar thickness on patellofemoral forces after
resurfacing. Clin Orthop 322:279–284, 1996.
18. Stiehl JB, Abbott BD: Morphology of the transepi-
condylar axis and its application in primary and re-
vision total knee arthroplasty. J Arthroplasty
10:785–789, 1995.
19. Stiehl JB, Cherveny PM: Femoral rotational align-
ment using the tibial shaft axis in total knee arthro-
plasty. Clin Orthop 331:47–55, 1996.
20. Whiteside LA, Arima J: The anteroposterior axis for
femoral rotational alignment in valgus total knee
arthroplasty. Clin Orthop 321:168–172, 1995.