As there have been no standard methods to determine the thickness
of resection for the proximal tibia pre-operatively, these are determined
with intra-operative findings, which primarily depend on the surgeon’s
experiencesThe purpose of this study was to clarify the correlation between
the change of limb alignment and the change of joint line height
at the medial proximal tibia and distal femurThe hypothesis was that the change of limb alignment correlated
with the change of joint line heightThere was a significant correlation between the correction of
FTA (δFTA) and the joint line elevation of the tibia (JLET)This observation can make it possible to know the requirement
of elevation of joint line to obtain their desired correction of
limb alignment and to predict the requirement of bone resection
at proximal tibia pre-operativelyStrength - With this study’s results, the surgeon can pre-operatively
know the requirement of the elevation of the joint line in order
to obtain the desired correction of limb alignmentLimitation - the amount of the ligamentous balance was not quantified
Introduction
Unicompartmental knee arthroplasty (UKA) is an effective surgical
procedure for relieving pain and restoring function in patients
with localised osteoarthritis (OA) of the knee.[1-4] Accuracy of implant positioning and
restoring the proper limb alignment are required for the long-term survival
of the implants.[5] A
neutral correction or slight undercorrection in the coronal alignment
is advocated in the literature.[6] Hernigou
et al[7] showed
that severe undercorrection of the deformity was a cause of increased
polyethylene wear. On the other hand, OA in the contralateral compartment
can occur after surgery where there has been overcorrection of the
pre-existing varus deformity.[8]Some studies have reported factors that can affect post-operative
limb alignment.[9-11] Kim et al[12] reported that limb
alignment was affected by the thickness of the bearing, rather than
alignment of the femoral and tibial implants. While the limb alignment
is determined by the position of the implant in total knee arthroplasty
(TKA), it is corrected by the restoration of the joint line at the affected
compartment in UKA.[13,14] The correction
of limb alignment is mainly formed by the elevation (restoration) of
the joint line at the tibia and femur, especially for medial OA.The change of the joint line at the tibial side is affected by
the difference between the depth of resection at the proximal tibia
and the thickness of the bearing insert.[10] Similarly, at the femoral side,
the change of the joint line is affected by the depth of resection
at the distal femur. If the pre-operative varus deformity is very
mild and no correction is required, the proximal tibia and distal
femur should be resected for the same thickness of the thinnest component.
In contrast, if there is a pre-operative varus deformity due to
wear of cartilage and subchondral bone at the medial tibia and femur,
the joint line should be restored by elevation of the medial joint
line. In this case, the bearing insert should be thicker than the
resected bone. As there have been no standard methods to determine
the thickness of resection of the proximal tibia pre-operatively,
resection thickness is determined intra-operatively based on the
surgeon’s experience. Information about the relationship between
the joint line elevation (JLE) and the change of limb alignment
would be useful for pre-operative planning. However, to our knowledge,
no studies about the relationship between reconstruction of the
joint line and correction of limb alignment have been reported.The purpose of this study was to clarify the correlation between
the change of limb alignment and the change of joint line height
at the medial proximal tibia and distal femur in UKA. The clinical
relevance of this study is that based on this information, the surgeon
can pre-operatively predict the requirement of bone resection at
proximal tibia to obtain their desired correction of limb alignment.
In this study, the pre- and post-operative radiographs of 42 patients
who underwent UKA for medial compartment were retrospectively reviewed,
and the changes of medial joint line and the femorotibial angle
(FTA) were measured. The hypothesis was that the change of limb
alignment would correlate with the change of joint line height.
Materials and Methods
A consecutive series of 42 medial UKAs between December 2008
and June 2013 at our institution were reviewed retrospectively.
There were nine males and 33 females with a mean age of 71.7 years
(59 to 86). The inclusion criteria included patients with OA or
osteonecrosis at the medial compartment of the knee, age > 60 years,
no flexion contracture more than 15° or no limitation of flexion
< 100°, and no dysfunction of cruciate ligaments and collateral
ligaments. Furthermore, a long-leg anteroposterior (AP) radiograph
was taken while a 15 kg valgus stress was being applied to the knee.
If the mechanical axis of the leg was passed at the point from medial
40% to 60% of the knee joint surface on the valgus stress long-leg
radiograph, the case was included.[15] Exclusion criteria were inflammatory
joint diseases or severe osteoporosis, previous high tibial osteotomy
and septic lesion, severe obesity > 35 of body mass index, and severe
bone defect > 10 mm. There was no case of arthroplasty for lateral
compartment OA. This study protocol was reviewed and approved by
the institutional review board at our university. All patients had
given their informed consent in advance, from which their radiographs
would be used for retrospective studies without their specific permission.
Surgical procedures
All procedures were undertaken using minimally invasive techniques
with a quadriceps-sparing approach without reversing or subluxation
of the patella. At first, the proximal tibia was prepared using
an extramedullary system. The resection angle was adjusted to be
perpendicular to the mechanical axis of the tibia in the coronal
plane and to be parallel to the posterior slope of the original
medial tibial plateau in the sagittal plane. The rotational alignment
of the tibia was adjusted to be parallel to the medial wall of the
intercondylar notch in 90° of flexion.[16] The depth of resection for the proximal
tibia was determined according to an intra-operative joint laxity
test in order that a 9 mm extension gap would be obtained after
the resection of the proximal tibia. After the resection of the
proximal tibia, the cutting alignment and the gaps in both of extension
and flexion were assessed. If either gap was < 9 mm, an additional
1 mm to 2 mm cut or adjustment of the posterior slope was performed. Ligamentous
releases were limited to the deep layer of the medial collateral
ligament on the tibial side. After the resection of the proximal
tibia, a spacer block with optimal thickness to obtain a neutral
limb alignment was
inserted to fill the extension gap, and the distal femur was resected
in parallel to the tibial cut surface in order to accommodate the
thickness of the femoral implant. Similarly, the posterior femur
was also resected in parallel to the tibial cut surface in flexion
to accommodate the thickness of the implant. With trial components
in place, the movement and laxity of the knee were tested. The thickness
of the polyethylene insert was chosen to leave a 2 mm laxity between
the components in both extension and flexion. All implants used
were the Zimmer Unicompartmental (Zimmer Inc., Warsaw, Indiana),
which were metal-backed cemented tibial components. Following surgery, patients
began routine physiotherapy with weight bearing as tolerated.
Radiographic evaluation
The patients were assessed radiographically both pre- and post-operatively
with standing AP radiographs. The radiographs were taken on the
tangential direction to the tibial baseplate or medial joint surface
of the proximal tibia. The FTA was measured and the correction of
FTA (the difference between the pre- and post-operative radiograph; δFTA) was
calculated for each subject. Next, the depth of resection at the
proximal tibia was measured using the following method (Fig. 1):
both the anatomical axis of the tibia and a line perpendicular to
the anatomical axis from the lowest point of the joint line (termed
‘line A’) were drawn. The distance from the tip of anatomical axis
to the intersection between the anatomical axis and line A was measured
on the pre-operative AP film (‘distance A’). Similarly, both the
anatomical axis and a line perpendicular to the anatomical axis
from the centre of the bottom line of the tibial baseplate (‘line
B’) were drawn. The distance from the most proximal point of anatomical
axis to the intersection between the anatomical axis and line B
was measured on the post-operative AP film (termed ‘distance B’). The difference between distance A and
distance B was defined as the depth of resection of the proximal
tibia. Finally, a line parallel to the tibial base plate passing
through the lower point of femoral implant (‘line C’) was drawn. The
distance from line C to a line of the tibial baseplate was measured
(‘distance C’).Pre-operative anteroposterior
radiographs showing a) the distance between the adductor tubercle
as the distal point on the medial condylar slope of the femur and
the joint line (line D) and b) the depth of resection was defined
as differences between distance A and distance B. Line C was the
thickness of the tibial insert used to correct magnification.The details of the thickness of the tibial insert were obtained
from the operative records of each patient. The joint line elevation
of the tibia (JLET) was defined as the difference between the depth
of resection and the thickness of the tibial insert. At the femoral
side, the change of the joint height of the femur was measured based
on a method previously described.[17] The perpendicular distance between
the adductor tubercle as the distal point on the medial condylar
slope of the femur and the joint line was measured (‘distance D’). The
joint line down the femur (JLDF) was defined as the difference between
pre- and post-operative distance D.The magnification of each radiograph was calculated using the
thickness of the tibial insert measured in the post-operative film,
and the thickness obtained from the operative records. The JLET
and JLDF were adjusted by the amount of image magnification.
Statistical analysis
The correlation between δFTA and the JLET was assessed using
a simple linear regression model. JMP software version 9.0.2 (SAS,
Cray, North Carolina) was used to analyse the data with the level
of significance set at p < 0.05. A power analysis was performed
on correlations (r = 0.5, significant level = 0.05, and power =
0.80) to indicate whether sample sizes of 28.2 on correlations could
address the questions. The δFTA and JLET were conducted by one observer
(UK) and were repeated in a blinded manner during the course of
two sessions. Intraobserver reliabilities, evaluated with the use
of the intraclass correlation coefficient, were 0.821 and 0.881
for δFTA and JLET, respectively. Two observers (UK and KM) independently measured 20 randomly selected
radiographs. The mean δFTA measured (by UK) was 5.5 (standard deviation
(sd) 2.3) and that of KM was 5.5 (sd 2.0). Similarly,
the mean JLET measured by UK was 4.2 (sd 1.3) and that
of KM was 4.2 (sd 1.2). Inter-observer reliabilities, evaluated
with the use of the interclass correlation coefficient, were 0.912
and 0.806 for δFTA and JLET, respectively.
Results
The mean pre- and post-operative FTA were 180.5° (sd 2.4, 172.2°
to 184.8°) and 175.0° (sd 2.0, 168.5° to 178.9°), respectively.
The mean δFTA was 5.5 (sd 2.0, 2.3° to 10.1°).The bearing sizes used in this study were 8 mm to 12 mm (mean
bearing thickness 9.1 mm). The mean depth of resection at the proximal
tibia was 4.7 mm (sd 1.5, 2.0 to 9.3). The mean JLET was
4.4 mm (sd 1.4, 2.1 to 7.8). There was a significant correlation
between δFTA and the JLET (correlation coefficient 0.494; p = 0.0009). JLET
was calculated using the following equation: JLET = 0.4 δFTA + 2.4
(Fig. 2).Graph showing the correlation between
joint line elevation of the tibia (JLET) and the correction of the
femorotibial angle (δFTA). The slope was 0.4, and the intercept
was 2.4. JLET was calculated using the following equation: JLE =
0.4 δFTA + 2.4.The mean JLDF was 2.7 mm (sd 2.1, -2.0 to 6.7). There
was no correlation between δFTA and the JLDF (correlation coefficient
-0.158, p > 0.05; Fig. 3).Graph showing the correlation between
the joint line down the femur (JLDF) and the correction of the femorotibial
angle (δFTA). There was no correlation.The mean change of joint space (defined as JLET + JLDF) was 7.1
mm (sd 2.4, 2.0 to 11.5). There was no correlation with
the δFTA (correlation coefficient 0.156, p > 0.05; Fig. 4).Graph showing the correlation between
the mean change of the joint space (defined as joint line elevation
of the tibia (JLET) + joint line down the femur (JLDF)) and the
correction of the femorotibial angle (δFTA). There was no correlation.
Discussion
The most important finding of this study was that the change of limb alignment demonstrated a correlation with
the JLET following medial UKA. With this result in mind, the surgeon
can pre-operatively calculate the required joint line elevation
in order to obtain their desired correction of limb alignment.The desired correction angle in the coronal alignment can be
estimated from δFTA, which is calculated by subtracting the FTA
on a pre-operative valgus stress radiograph from that on a standing
AP radiograph.[15,18] Tashiro et al[19] suggested that
the mean post-operative limb alignment demonstrated a significantly
strong correlation with the values on the pre-operative valgus stress
radiographs, while the correlation between both the pre- and the
post-operative standing alignment, was moderate. A neutral, or slight
under correction is recommended as an ideal post-operative coronal
alignment.[6]The surgeon can predict the requirement of bone resection at
the proximal tibia from the desired correction angle in order to
obtain the ideal post-operative alignment. This information is useful
when undertaking pre-operative planning of medial UKA. The equation
of the regression line was as follows: JLET = 0.4 δFTA + 2.4 (Fig.
2). The reason for the equation intercept may be the pre-operative
thickness of cartilage. With this equation, it is suggested that,
for example, it is necessary to elevate the joint line 4.4 mm from
the bony surface of the tibia if the surgeon plans to correct the
FTA by 5° valgus. In this case, the surgeon should resect 4 mm from
the proximal tibia to use an 8 mm bearing insert. However, the condition
of the remaining cartilage varies among the cases, and this may
have caused the variations of the results of this study. Therefore,
the surgeon should determine the amount of resection of the proximal
tibia in consideration of the thickness of the remaining cartilage.According to a report from the Australian Orthopaedic Association
National Joint Replacement Registry,[20] the cumulative percentage revision
of UKA at eight years was 13.2% compared with 5.0% for TKA. Another
registry reported that the primary reasons for revision arthroplasty
were progression of OA in other compartments (48%).[21] Furthermore, a
study presented lateral OA as the main reason for UKA revision.[22] Squire et al[23] reported that
62 of 136 knees showed progression of OA in the contralateral compartment
and seven knees underwent revision for this reason. The risk factors
for UKA revision were younger patient age, increased patient weight,
use of a thinner tibial component, increased tibial component posterior
slope, and lesser or over correction of pre-operative varus limb
alignment.[24]Post-operative alignment is important in order to prevent the
progression of degenerative arthritis and implant-related complications.
Several studies have reported the effect of post-operative alignment
of the knee on the long-term outcome of UKA.[6,7,25,26] For UKA, neutral correction or slight
under correction has been recommended as over correction may increase the
risk of degeneration in the opposite compartment. Significant under
correction, on the other hand, may accelerate polyethylene wear
and recurrence of deformity.[7,27]In contrast to the TKA, the alignment correction of the leg in
UKA depends not on the alignment of the component itself, but on
the elevation (restoration) of the joint line on the affected side.
In this study, there was a significant correlation between δFTA
and the JLET, while the change of the joint line of the femur had no
significant correlation with δFTA. Because the distal femur is usually
resected by the same thickness of the femoral implant, the change
of the joint line of the femur was relatively small in our case
series. At the tibial side, the elevation of the joint line is determined
by the thickness of the implant relative to the bone resection.
If the bone resection was 2 mm with the use of 8 mm thickness of
the implant, the joint lift length would be 6 mm. If the amount
of resection was too short to obtain the optimal joint lift with
use of the thinnest component, there would be a risk of over correction
of the limb alignment, or the tibia would need to be re-cut. If
the amount of resection was too large, there would be a risk of
fracture of the tibia or subsidence of the implant.[28-31] Therefore, determining the thickness of
bone resection at the proximal tibia is one of the critical points
in the operative procedure of UKA. There has not been a standard
method to determine or predict the optimal thickness of bone resection
pre-operatively, and thus, it primarily depends on the surgeon’s
experience.There are some limitations in our study. First, we have not quantified
the amount of the ligamentous balance in each case. However, the
same surgical criteria for soft-tissue balancing were carried out
throughout this studied group. Ligamentous releases were limited
to the deep layer of the medial–collateral ligament on the tibial
side. In addition, the spacer-guided system is reported to approximate
valgus laxity.[32] Furthermore,
the purpose of this study was to clarify the relationship between
JLE and the change of limb alignment, which is not related to the
ligament balance. Second, the depth of tibial resection was evaluated
radiographically. The radiograph may cause some variations as a
result of technical inconsistencies. However, we chose the radiograph
that was taken at a consistent direction and the rate of magnification
was adjusted accordingly. Furthermore, radiographic evaluation was
performed using standard weight-bearing radiographs. For analysis
of the alignment of the lower limb, long-leg standing radiographs
are required. However, the purpose of this study was to clarify
the relationship between JLE and the change of limb alignment, therefore,
the use of a standard weight-bearing radiograph for pre- and post-operative
radiographic evaluation was reasonable for the purpose of this study.
Also, Skyttä et al[33] reported
that the standard AP knee radiograph appeared to be a valid alternative
to the hip-to-ankle radiograph for determining the coronal planes
of the knee. Finally, we did not assess the clinical outcome with
post-operative alignment. However, the assessment of clinical outcome
was not the main purpose of this study, and would be a focus of further
research.In conclusion, this study indicates that there is a significant
correlation between joint line elevation and the change of limb
alignment in UKA. The JLET was calculated using the following equation:
JLET = 0.4 δFTA + 2.4. This observation can make it possible to
predict the required amount of bone resection of the proximal tibia
pre-operatively, in order to achieve JLET, and the desired correction
of limb alignment.
Authors: Thomas J Aleto; Michael E Berend; Merrill A Ritter; Philip M Faris; R Michael Meneghini Journal: J Arthroplasty Date: 2008-02 Impact factor: 4.757
Authors: F Iacono; G F Raspugli; D Bruni; G Filardo; S Zaffagnini; W F Luetzow; M Lo Presti; I Akkawi; G M Marcheggiani Muccioli; M Marcacci Journal: Knee Surg Sports Traumatol Arthrosc Date: 2013-12-22 Impact factor: 4.342
Authors: Patrick Weber; Christian Schröder; Rüdiger Paul Laubender; Andrea Baur-Melnyk; Christoph von Schulze Pellengahr; Volkmar Jansson; Peter E Müller Journal: Knee Surg Sports Traumatol Arthrosc Date: 2013-07-24 Impact factor: 4.342
Authors: F Zambianchi; V Digennaro; A Giorgini; G Grandi; F Fiacchi; R Mugnai; F Catani Journal: Knee Surg Sports Traumatol Arthrosc Date: 2014-03-30 Impact factor: 4.342
Authors: Richard A Berger; R Michael Meneghini; Joshua J Jacobs; Mitchell B Sheinkop; Craig J Della Valle; Aaron G Rosenberg; Jorge O Galante Journal: J Bone Joint Surg Am Date: 2005-05 Impact factor: 5.284