KNEE
Effect of high tibial osteotomy on joint loading in symptomatic
patients with varus aligned knees: a study using SPECT/CT
Armin Mucha
•
Milos Dordevic
•
Anna Hirschmann
•
Helmut Rasch
•
Felix Amsler
•
Markus P. Arnold
•
Michael T. Hirschmann
Received: 13 April 2013 / Accepted: 2 May 2014 / Published online: 11 May 2014
Ó Springer-Verlag Berlin Heidelberg 2014
Abstract
Purpose The purpose was to prospectively evaluate the
outcome, in particular the SPECT/CT bone tracer uptake
(BTU) after high tibial osteotomy (HTO) due to symp-
tomatic varus malalignment. It was the hypothesis that the
BTU after HTO decreases in the medial compartment,
clinical outcome and the degree of correction correl ates
with BTU and asymptomatic patients after HTO reveals a
significantly decreased BTU in the medial subchondral
areas.
Methods Twenty-two consecutive patients with 23 knees
undergoing medial opening-wedge HTO for medial com-
partment overloading were assessed pre- and postopera-
tively (12 and/or 24 months) using Tc-99m-HDP-SPECT/
CT including our 4D-SPECT/CT protocol. BTU was
quantified and localized to specific biomechanically rele-
vant joint areas. Maximum absolute and relative values
(mean ± standard deviation, median and range) for each
area were recorded. Pre- and postoperative mechanical
alignment was measured. At 24 months after HTO, the
WOMAC score was used.
Results A significant decrease of BTU in the medial
subchondral zones after HTO was found (preoperatively
to 12 and 24 months postoperatively, p \ 0.01). BTU
normalized in all asymptomatic patients within
24 months. This decrease was partly seen in the lateral
compartments, but significantly higher in the medial
compartments (p \ 0.0001). A significant increase of the
BTU was noted in zones directly adjacent to the plate or
within the osteotomy zone (p \ 0.01). Decreased BTU
was observed in osteotomy zones at 24 months postop-
eratively following higher uptake values at 12 months
postoperatively. The average valgus correction of the
tibiofemoral angle was 5.9° ± 2.8°. Less stiffness corre-
lated significantly with a higher decrease in BTU
(p \ 0.05). Higher postoperative BTU significantly cor-
related with more pain (p \ 0.05). No statistical signifi-
cant associations between BTU and alignment correction
were found.
Conclusion In patients with medial compartment, over-
loading due to varus malalignment HTO led to a significant
decrease in BTU in the medial joint compartments.
SPECT/CT BTU patterns and intensity in these patients
pre- to 12 and 24 months postoperatively were seen. These
correlated significantly with pain and stiffness. Hence,
SPECT/CT coul d be used for assessment of adequate
correction and healing after HTO. SPECT/CT could be
further used to identify the optimal individualized correc-
tion for each patient and clinical scenario.
Clinical evidence Diagnostic prospective study, Level II.
Keywords Knee SPECT/CT High tibial osteotomy
Localization scheme Varus malalignment
Loading history
A. Mucha M. Dordevic M. P. Arnold
M. T. Hirschmann (&)
Department of Orthopaedic Surgery and Traumatology,
Kantonsspital Baselland, 4101 Bruderholz, Switzerland
e-mail: Michael.Hirschmann@ksbl.ch;
Michael.Hirschmann@unibas.ch
A. Hirschmann
Department of Radiology, University Hospital Basel, Basel,
Switzerland
H. Rasch
Institute for Radiology and Nuclear Medicine, Kantonsspital
Baselland, 4101 Bruderholz, Switzerland
F. Amsler
Amsler Consulting, Basel, Switzerland
123
Knee Surg Sports Traumatol Arthrosc (2015) 23:2315–2323
DOI 10.1007/s00167-014-3053-y
Introduction
Mechanical varus alignment of the knee is a common
deformity not always related to clini cal symptoms, but
known to be related to the development and progression of
knee osteoarthritis (OA) [4, 6, 9, 22, 32]. Based on com-
bined investigations of mechanical alignment and gait,
Johnson et al. [18] showed that the knee is predominantly
loaded in the medial compartment, even in a neutrally
aligned leg. In a knee with neutral mechanical alignment,
approximately 60–70 % of the loading forces affect the
medial tibiofemoral joint. From this fact, it is clear that
even a mild varus alignment could result in medial knee
pain and progressive medical compartment overloading. To
avoid medial compartment degeneration, it needs to be
identified at an early stage, and only then progression of
OA could be delayed or even prevented surgically. Jackson
and Waugh [17] and Coventry et al. [5] were the first to
propose high tibial osteotomy (HTO) as a viable, joint-
preserving treatment option for medial compartment OA in
varus deformed knees.
Current routine radiological imaging before HTO
includes anterior–posterior and lateral weight-bearing, long
leg radiographs [8] as well as a 45° posterior–anterior view
[28]. These radiographs are able to identify late OA
changes such as joint space narrowing, subchondral scle-
rosis and osteophyte formation, but fail to show over-
loading or early degenerative changes [37 ]. Postoperative
radiographs are also not capable of revealing the unloading
effect after HTO.
Recently, it was shown that combined single photon
emission-computerized tomography and conventional
computerized tomography (SPECT/CT) is able to visualize
mechanical overload ing of joint compartments. More spe-
cifically, SPECT/CT tracer uptake was correlated with
anatomical and mechanical alignment [15]. The intensity
of SPECT/CT tracer upta ke in the medial and lateral knee
compartment significantly correlated with varus or valgus
alignment of the knee [15]. A higher degree of radio-
graphically evident osteoarthritis was significantly related
to higher tracer uptake in the corresponding joint com-
partments [15]. It was concluded that SPECT/CT reflects
the specific loading pattern of the knee with regard to its
alignment [15].
The purpose of this prospective study was to evaluate
the clinical and radiological outcome after HTO due to
medial compartment overloading. In particular, the
SPECT/CT bone tracer uptake (BTU) pattern and intensity
were evaluated over the course of time. It was the
hypothesis that the BTU after HTO decreases and then
normalizes in the medial compartment, clinical outcome
and the degree of correction correlates with BTU and
asymptomatic patients after HTO reveals a significantly
decreased and normalized BTU in the medial subchondral
areas.
Materials and methods
Twenty-two consecutive patients (male/female 18:4, mean
age 47 ± 10) with twenty-three knees (male /female 19:4)
undergoing medial opening-wedge HTO for medial joint
compartment overloading or medial tibiofemoral osteoar-
thritis due to mechanical varus alignment were prospec-
tively included in this study. All subjects gave written
informed conse nt before inclusion into the study. Exclusion
criteria were posttraum atic osteoarthritis, open growth
plates, corticosteroid use within last 6 months, know n
history of avascular necrosis, osteochondrosis dissecans,
chondrocalcinosis, tumour, Paget’s disease, joint infection,
periarticular fracture, neuropathic arthropathy, reactive
arthritis, gout or a pregnancy. In a trial analysis prior to the
main study, it was observed that 5 of 27 patients showed
unaffected high BTU values after realignment surgery for
relevant medial zones due to meniscal chondrocalcinosis
and osteochondral lesions. These patients were excluded
for further analyses (Fig. 1).
Medial opening-wedge HTO was performed in every
patient according to published standard techniques by
experienced orthopaedic surgeons [19,
21, 26, 27, 33].
Fig. 1 Left SPECT/CT image of a 50-year-old male patient
12 months after HTO showing a consistently increased uptake on
the medial femoral condyle due to OCL which is only partially
affected by HTO. An increased tracer uptake can be seen in the
osteotomy gap indicating advancing osseous consolidation. Right
SPECT/CT image of a 50-year-old male patient 13 months after HTO
showing medial meniscal chondrocalcinosis with increased tracer
uptake
2316 Knee Surg Sports Traumatol Arthrosc (2015) 23:2315–2323
123
All procedures were performed in accordance with the
ethical standards of the responsible committee and with the
guidelines of the Helsinki Declaration of 1975, as revised
in 2008.
Anterior–posterior (AP) and lateral weight-bearing
radiographs as well as a 45° posterior–anterior view and
skyline views were routinely obtained before HTO,
4 months, 1 and 2 years postoperatively. Long leg weight-
bearing radiographs (pelvis to ankle) were performed pre-
operatively and 4 months postoperatively.
99mTc-HDP-SPECT/CT was performed preoperatively,
12 and/or 24 months after HTO using our previously
published 4D-SPECT/CT protocol [29 ]. All SPECT/CTs
were performed using a hybrid system (Symbia T16, Sie-
mens, Erlangen, Germany). Planar scintigraphic images
were taken in three phases, the perfusion phase (immedi-
ately after injection), the blood pool phase (2–5 min after
injection) and the delayed metabolic phase (at least 2 h
after injection). Two to three hours after injection of a
commercial 500 MBq 99 m-Tc-HDP (Malinckrodt, Wol-
lerau, Switzerland), a SPECT was performed with a matrix
size of 128 9 128, an angle step of 32° and a time per
frame of 25 s.
Bone tracer uptake was assessed in SPECT/CT images
using a customized software (IntroSPECT, OrthoImaging-
Solutions Ltd., London, UK), which is able to quantify
bone tracer uptake volumetrically in 3D and localize the
areas of increased or decreased BTU to specific biome-
chanical relevant joint areas [16]. For localization of the
BTU, a previously validated standardized localization
scheme was used (Figs. 2, 3), which was specifically tai-
lored for knees after HTO. In order to have comparable
SPECT volumes, we used an identical localiza tion scheme,
which was adjusted to both native and operated knees. The
scheme defines 9 femoral, 8 patellar and 13 tibial zones to
accurately map the examined tracer uptake volume. The
Fig. 2 The mapping scheme used for localization of areas of increased SPECT/CT tracer uptake values in knees after HTO (F femur, T tibia,
P patella, R reference zone, 1 lateral, 2 medial, 3 tibial intercondylar area, 4 distal plate area, s superior, i inferior, a anterior, p posterior)
Knee Surg Sports Traumatol Arthrosc (2015) 23:2315–2323 2317
123
femur (F) is divided into nine zones that include one ref-
erence zone in the shaft (R) and eight distal femoral zones.
Each distal femoral zone is represented with a number (1-
lateral, 2-medial) and 2 small letters (a-anterior, p-posterior
and i-inferior, s-su perior). The tibia (T) is divided into 13
zones that include each six proximal and distal tibial
regions and one shaft region. Each tibial zone is repre-
sented with a number (1-lateral plateau, 2-medial plateau,
3-tibial spine) and two small letters (a-anterior, p-posterior
and i-inferior, s-superior). In the tibia, the thickness of the
superior zones was defined as 10 mm to mimic the sub-
chondral bone plate as close as possible. Inferior tibial
zones include the osteotomy site and are defined by a
thickness of 50 mm. Zone 4 represents the distal part of the
plate including screws. The patella (P) is divided into four
zones (superomedial, superola teral, inferomedial and
inferolateral).
Maximum values (mean ± standard deviation, median
and range) for each area of the localization scheme were
recorded as well as normalized relative values for intensity
of bone tracer uptake calculated. For normalization, the
femoral shaft was used as reference region as previously
published [16].
Pre- and postoperative mechanical alignment was mea-
sured in degrees on SPECT/CT using a previously vali-
dated custom-made specialized software (Orthoexpert
v1.15, OrthoImagingSolutions Ltd., London, UK) [29].
The localization scheme showed near-perfect inter- and
intra-OR (intra-class correlation coefficient (ICC)[0.9) for
the measurement of tracer activity and localization in all
anatomical regions. For measurements of mechanical
alignment, there was a strong agreement betwee n the two
observers (an inter-OR of ICC = 0.99 and an intra-OR of
ICC = 0.98) [25].
Fig. 3 The mapping scheme used for localization of areas of increased SPECT/CT tracer uptake values in native knees ( F femur, T tibia,
P patella, R reference zone, 1 lateral, 2 medial, 3 tibial intercondylar area, 4 distal plate area, s superior, i inferior, a anterior, p posterior)
2318 Knee Surg Sports Traumatol Arthrosc (2015) 23:2315–2323
123
Clinical outcome was assessed at 24 months after
HTO using the German version of the WOMAC score
[34].
The study was approved by the local ethical committee.
Statistical analysis
Data were analysed using SPSS 17.0 (SPSS, Chicago, IL).
Means of two measurements of one observer were applied
for statistical analysis.
A nonparametric Spearman’s correlation was used to
compute associations between the WOMAC score and the
mechanical alignment with the tracer uptake in each rele-
vant area. Paired t test/Wilcoxon rank sum-test was used to
compare specific areas for pre- and postoperative SPECT/
CT tracer uptake. For all analyses, p \ 0.05 was consid-
ered statistically significant.
Results
A significant decrease of BTU in the medial subchondral
zones after HTO was found from preoperatively to 12 and
24 months follow-up (p \ 0.01). BTU normalized in all
asymptomatic patients within 24 months. The normalized
grading of BTU in SPECT/CT for each anatomical area of
the localization scheme is presented in Table 1 (values
represent difference between preoperative and postopera-
tive measurements).
A decrease of BTU was partly seen in the lateral com-
partments, but the decrease was significantly higher in the
deloaded medial tibial and femoral joint com partment
(p \ 0.0001, Fig. 4).
The achieved average valgus correction of the tibio-
femoral angle by HTO was 5.9° ± 2.8°. There were no
adverse events such as pseudo arthrosis, infection, loss of
correction or skin necrosis. The mean WOMAC score pain
(0–20) was 6.2 ± 5.6, WOMAC stiffness (0–8) was
2.8 ± 2.4, and the WOMAC daily activities (0–68) was
17.4 ± 16.4. The mean total score (0–96) was
25.4 ± 22.00 after HTO (Fig. 5).
Less stiffness with regard to the WOMAC score corre-
lated significantly with a higher decrease in SPECT/CT
BTU (p \ 0.05). Higher postoperative bone tracer uptake
significantly correlated with more pain (p \ 0.05).
A Spearman correlation analysis revealed no statistical
significant associations between SPECT/CT BTU and
alignment correction by HTO.
Discussion
The findings of this study have led to a better under-
standing of the in vivo loading of the different joint
Table 1 Mean, standard deviation, minimum and maximum of
grading for normalized values of SPECT/CT tracer uptake for
relevant areas of the localization scheme for the difference of
preoperative and postoperative measurements, including Wilcoxon’s
signed-rank test for the differences
N = 23 1saFe 2saFe 1spFe 2spFe 1iaFe 2iaFe 1ipFe 2ipFe 1sPat 2sPat 1iPat 2iPat
Difference pre- and postoperative measurements
Mean -0.29 -0.3 -0.5 -0.45 -0.35 -1.42 -0.28 -1.81 -0.78 -0.5 -0.55 -0.41
Median -0.32 -0.28 -0.3 -0.42 -0.37 -0.84 -0.25 -1.51 -0.6 -0.51 -0.53 -0.33
SD 0.46 0.5 0.82 0.91 0.62 1.37 0.65 1.55 1.26 0.93 0.71 0.59
Min -0.97 -1.52 -3.36 -3.01 -1.72 -4.39 -2.18 -5.7 -3.66 -2.52 -2.03 -1.76
Max 0.97 0.89 0.53 1.51 1.06 0.65 0.94 0.12 2.19 1.54 0.87 0.84
Wilcoxon test
Z -2.92 -2.95 -3.3 -2.68 -2.66 -4 -1.82 -4.17 -2.89
-2.28 -3.35 -3.15
p 0.0035 0.0032 0.001 0.0074 0.0078 0.0001 0.068 0 0.0039 0.0225 0.0008 0.0016
N = 23 1saTib 3saTib 2saTib 1spTib 3spTib 2spTib 1iaTib 3iaTib 2iaTib 1ipTib 3ipTib 2ipTib 4Tib
Difference pre- and postoperative measurements
Mean -0.2 -0.32 -1.02 0.16 0.18 -1.28 0.91 1.52 1.42 2.32 3.97 3.14 1.16
Median -0.14 -0.09 -0.78 -0.08 -0.04 -1.02 0.6 1.31 1.54 1.29 2.08 1.81 0.5
SD 0.5 0.69 1.36 0.87 0.88 1.75 1.84 2.2 2.02 3.58 4.55 4.1 1.88
Min -1.86 -2.43 -6.04 -1.17 -0.9 -8.11 -1.33 -1.26 -4.12 -0.53 0.38 -5.07 -0.29
Max 0.57 0.69 0.41 2.79 2.41 0.68 7.71 8.24 6.48 12.87 17.96 10.47 7.77
Wilcoxon test
Z -1.56 -2.28 -3.76 -0.18 -0.26 -3.98 -2.59 -2.98 -3.2 -3.67 -4.2 -3.41 -3.44
p 0.119 0.0225 0.0002 0.8552 0.7951 0.0001 0.0097 0.0029 0.0014 0.0002 0 0.0007 0.0006
Knee Surg Sports Traumatol Arthrosc (2015) 23:2315–2323 2319
123
