Arch Orthop Trauma Surg (2006) 126: 588–593

DOI 10.1007/s00402-006-0130-9

O R I GI N A L A R T IC L E

Suguru Ohsawa Æ Kazuya Hukuda Æ Yasuaki Inamori

Natsuo Yasui

High tibial osteotomy for osteoarthritis of the knee with varus deformity
utilizing the hemicallotasis method

Received: 22 March 2005 / Published online: 13 July 2006

Ó Springer-Verlag 2006

Abstract Introduction: A hemicallotasis method has been and was maintained at the final follow-up. The compli-
developed utilizing an external fixator as high tibial os- cations of this method were relatively few and consisted
teotomy (HTO), and satisfactory results of this method of pin-tract infection (8 knees), deep vein thrombosis (3
with the external fixator have been reported. This knees), and delayed union (2 knees). No peroneal nerve
external fixator has a universal joint that moves in all palsy or compartment syndrome was encountered. No
directions. We have recently designed a hemicallotasis knee was converted to total arthroplasty. However,
device for this operation. Methods: HTO for the knee administration of analgesics was necessary in ten knees
with varus deformity utilizing the hemicallotasis method at the final follow-up. Conclusion: The hemicallotasis
was performed on 44 knees. The patients had a mean age method easily determined the angle of correction even in
at operation of 65 years (range 49–82 years), a mean the knees with ligamentous laxity. Nevertheless, one of
follow-up period of 68 months (range 36–119 months), the major demerits of this method was a longer period of
and a mean preoperative knee score of 66 points (range application of the external fixator. The level of evidence
27–90 points). Results: The operated knees had a mean was level IV (case series).
knee score at the final follow-up of 86 points (range 51–
98 points), but the mean range of knee motion was not Keywords Osteoarthritis Æ Knee joint Æ
changed as follows. Before surgery, the mean flexion was Hemicallotasis Æ Delayed union Æ Femorotibial angle
129° (range 90–150°) and the mean extension was 5°
(range 30 to 0°), whereas at the final follow-up, the
corresponding values were 127° (range 85–150°) and 4° Introduction
(range 25 to 0°), respectively. Radiographically, the
femorotibial joint was classified as grade 2 in 9 knees, Joint-preserving operation for osteoarthritis of the knee
grade 3 in 21 knees, and grade 4 in 14 knees according to with varus deformity has been established for many
the classification of osteoarthritis (Kellgren and Lau- years [9]. Many authors [4, 7] have reported excellent
rence). The patellofemoral joint was also classified as results for high tibial osteotomy (HTO). However, the
grade 1 in 39 knees, grade 2 in 2 knees, and grade 3 in 3 conventional methods need fibular osteotomy or proxi-
knees. The mean femorotibial angle was 184° (4° varus) mal tibiofibular joint release in addition to the postop-
before surgery, 169° (11° valgus) after pin extraction, erative difficulty to change the angle of correction.
Moreover, the correction angle is difficult to be deter-
mined when the joint is unstable. Insall et al. [7] pro-
S. Ohsawa (&) posed that osteotomy was contraindicated in knees with
Department of Rehabilitation Medicine, Osaka Rosai Hospital,
1179-3, Nagasone-Cho, Sakai-Shi, 591-8025 Osaka, Japan more than 15° of varus deformity. The major disad-
E-mail: [email protected] vantage of the conventional osteotomy is no way to
Tel.: +81-722-523561 assess the mechanical axis postoperatively with full
Fax: +81-722-505492 weight bearing. To overcome these difficulties of the
K. Hukuda Æ Y. Inamori conventional methods, Turi et al. [23] developed a
Department of Orthopaedic Surgery, NTT West Osaka Hospital, hemicallotasis method utilizing an external fixator as
2-6-40, Karasugatuji, Tennohji-Ku, 543-0042 Osaka, Japan HTO for varus deformity of the knee. Other authors
have also reported good short-term results [5, 12], using
N. Yasui
Department of Orthopaedic Surgery, Tokushima University, the same external fixator. Magyer et al. [15] reported an
School of Medicine, 3-18-15, Kuramoto-Cho, adequate result using the Garches model (Orthofix Srl.
770-0042 Tokushima, Japan Bussolengo, Italy), which was applied to the anterior
 589

aspect of the tibia. The former device has a universal mechanical axis of the leg was on 30–40% lateral from
joint, which moves in all directions. The latter model is the midpoint of the joint [6, 21] opening the medial os-
applied to the anterior aspect of the tibia, which inter- teotomy site with the lateral cortex as a hinge. This
feres to check callus formation from an anteroposterior preoperative planning calculated the length of callotasis
(AP) view. We have recently designed a hemicallotasis and the period of elongation. Under an image intensifier
device for this operation, and herein we report the and without air tourniquet, the posterior pin of the
operative procedures and the results of our method. The proximal group of pins (‘‘the first pin’’) was initially
advantages of this device and hemicallotasis method are inserted parallel to the tibial plateau and the posterior
discussed. cortex of the plateau (Fig. 1c). The depth from the
plateau and from the posterior edge was approximately
7 mm each. Then, the pins of the distal part were in-
Materials and methods serted utilizing a template, and, two other pins of the
proximal group were inserted (Fig. 1). After all pins
From August 1991 to March 1998, we consecutively were set, the medial cortex of the tibia was osteotomized
operated on 55 knees with osteoarthritis of 40 patients with chisel at the middle of the tibial tuberosity leaving
using the hemicallotasis method. The indication of this the lateral cortex intact (Fig. 2). After checking that a
method was medial compartmental osteoarthritis of the mild valgus stress opened the osteotomy site under an
knee with a relatively good range of motion (ROM) image intensifier, the hemicallotasis device (Orthofix Srl.
(flexion: more than 100°). The mean age of the patients Bussolengo, Italy) was applied (Fig. 1). The patients
at operation was 65 years (range 49–82 years). Three were allowed to undergo full weight bearing postoper-
patients (4 knees) died of unrelated diseases. Seven pa- atively. After 2-week rest, hexagonal screws for the neck
tients (7 knees) withdrew from the follow-up. We and elongation (Fig. 1a, b) were loosened and hemical-
examined 44 knees of 30 patients for a mean follow-up lotasis was started at a speed of 0.25 mm over 4–6 h.
period of 68 months (range 36–119 months). Bilateral During elongation, weight bearing on the leg was not
involvement was shown in 25 patients and unilateral allowed. The longer the body of the external fixator, the
involvement in five patients. In bilateral involved pa- smaller the angle between the proximal pin and the axis
tients, three knees had been treated with conventional of the fixator (Fig. 2). When the mechanical axis of the
HTO and eight knees were treated conservatively. The leg reached 30–40% lateral from the midpoint of the
patients consisted of 23 women and 7 men. The mean joint in the AP radiographs taken in the standing posi-
body mass index (BMI) at surgery was 26 kg/m2 (range
19.5–35.1 kg/m2). Clinical assessment was performed
according to the Hospital for Special Surgery (HSS)
rating system as reported by Insall et al. [8]. Radio-
graphic assessment of osteoarthritis of the knee was
based on the classification of Kellgren and Lawrence
[10]. In brief, the categories are grade 0: normal; grade 1:
minute osteophyte, doubtfully significant; grade 2: defi-
nite osteophyte, unimpaired joint space; grade 3: mod-
erate diminution of the joint space; and grade 4: joint
space greatly impaired with sclerosis of the subchondral
bone. The extent of deformity was determined from
weight-bearing radiographs and was recorded as the
femorotibial angle (FTA) according to the method de-
scribed by Bauer et al. [1] as follows. A straight line was
drawn along the axis of the femoral shaft to intersect a
corresponding line drawn through the tibial shaft, and
the lateral angle between these two lines was measured.
The level of evidence was level IV (case series).

Operative procedures

We planned preoperatively according to the AP radio-

graphs of the weight-bearing leg. A proximal pin was
inserted parallel to the tibial plateau 7 mm below the
joint line. Three to four centimeters below the medial Fig. 1 The hemicallotasis device and pins are inserted into the
tibia. The device has two-axis free movement: one axis is neck
joint line, osteotomy was made on the medial side of the tilting (A screw for neck movement) and the other is elongation (B
tibia, and was extended upwards to the lateral cortex of screw for elongation). C A posterior pin of the proximal group of
the tibia, 1.5 cm below the lateral knee joint line. The pins, ‘‘the first pin’’ to be inserted
590

Results

Clinical results

The mean HSS knee score was 66 points (range 27–90

points) preoperatively, and 86 points (range 51–98
points) at the final follow-up. The knees of severe OA
had poorer results than that of moderate OA (grade 4,
79.6±13.2 points; grade 2, 3, 89.0±10.8 points,
P=0.0208). The mean preoperative ROM was 129°
(range 90–150°) in flexion and 5° (range 30 to 0°) in
extension. The mean ROM at the final follow-up was
127° (range 85–150°), and 4° (range 25 to 0°),
respectively. No knee was converted to total knee
arthroplasty (TKA) at the final follow-up. However, it
was necessary to administer nonsteroidal anti-inflam-
matory drugs for ten joints to suppress pain at the final
follow-up.
The postoperative complications consisted of pin-
tract infection (8 knees), deep vein thrombosis (3
knees), and delayed union (2 knees). One case needed
bone grafting to facilitate bone union. In another case,
joint fluid was excreted from ‘‘the first pin’’ that was
believed to invade the posterior side of the knee joint
cavity through the tibial plateau. The pin tract se-
Fig. 2 Case 22: a 49-year-old female suffering from osteoarthritis
of both knees. The change in the lower limb axis is shown.
creted joint fluid for 2 months, but it was cured
Radiographs in the standing position before surgery (a uneventfully. Pin-tract infection was treated by wash-
FTA=183°), immediately after the operation (b). Two or three ing with saline containing antibiotics, and cured
proximal and distal pins were inserted. Hemicallotasis was uneventfully. This complication usually occurred in the
completed (c FTA=169°). d After pin removal. Note that the ‘‘first pin’’. Neither peroneal nerve palsy nor com-
weight-bearing axis was moved to the lateral plateau. The fixator
was applied for 145 days partment syndrome was encountered. The proximal
and distal tibiofibular joints were symptom free during
and after elongation.
tion (Fig. 2c), the fixator was fixed and full weight
bearing was allowed. When the callus was mature, Radiographical results
loosening of the hexagonal screw started dynamization.
Full weight bearing was allowed to check the mechanical According to the classification of osteoarthritis [10], the
axis in the plain radiographs, which did not change after femorotibial joints were classified into grade 1 (0 knee),
1–2 weeks. The fixator was then taken off, and another 1 grade 2 (9 knees), grade 3 (21 knees), and grade 4 (14
or 2-week period elapsed before pin removal. After knees). The patellofemoral joints were also classified
checking for the absence of change in the mechanical into grade 1 (39 knees), grade 2 (2 knees), and grade 3 (3
axis, all of the inserted pins were removed (Fig. 2d). We knees). The mean FTA was 184° (range 172–197°) pre-
made radiograph examination of the leg about 10 times operatively, 169° (range 164–174°) at removal of the pins
before pin removal. and external fixator, 169° (range 164–174°) at 1 year
The appropriate statistical procedures were per- after surgery, and 169° (range 165–174°) at the final
formed by a computer using StatView 5.0 (SAS Institute follow-up, except for two patients with delayed union.
Inc., Cary, NC, USA). The statistical significance of The mean correction angle was 15.3° (range 3–31°). The
differences in the mean values was computed with the FTA of delayed union cases was 180 and 182° at the
unpaired Student t test and Mann–Whitney U test. The final follow-up.
level of statistical significance was set at a P value of less Fracture of the lateral cortex of the tibia was one of
than 0.05, two-tailed. All of the predictor variables were the serious problems in the hemicallotasis method, be-
used in the logistic regression model as dependent vari- cause this method requires an intact lateral cortex as a
ables. The model was simplified in a stepwise fashion by fulcrum (Fig. 2). When the lateral cortex was fractured,
removing the variables with P value of greater than 0.05. the proximal fragment was displaced and bone healing
Multiple regression analysis was used for studying the was delayed (Fig. 3). We performed ten osteotomies on
factors influencing the period of fixator application, and patients with fracture of the tibial lateral cortex and
this was simplified in the same way as described in the displacement of the proximal fragment. No fracture was
logistic regression model. identified in the other 34 knees. Logistic regression
 591

Fig. 4 Chronological changes of FTA in each knee. Pre preoper-

ative FTA, post at pin removal, numbers postoperative years

change of more than 5° was found in four legs. Three of

them had a fracture of the lateral cortex of the tibia, and
two of them were complicated with delayed union.

Period of fixator application

Fig. 3 Case 4: an 81-year-old male with osteoarthritis of both
knees. Lateral cortex fracture and proximal fragment transposition
occurred during hemicallotasis (a 2 weeks after starting the One of the problems of this method was long-term
elongation). Three months after completing the elongation with application of the fixator (128 days on average), which
insufficient FTA correction (b from 178 to 173°) forced the patients to abstain from bathing for a long
period and caused a higher risk of pin-tract infection.
Lower level osteotomy tended to fracture the lateral
analysis revealed the factors influencing the fracture and cortex, and transposition of the proximal fragment oc-
dislocation of the proximal fragment. A logistic model curred as mentioned above. The period of external fix-
for predicting the lateral cortex fracture was constructed ator application (mean ± SD) with transposition of the
by using all the clinical and radiological factors, as well proximal fragment (194±56 days) was significantly
as, the age, BMI, sex, body weight, FTA, and the level of (P<0.0001) longer than that without transposition
osteotomy, as described previously. This model was re- (108±34 days). Multiple regression analysis (R2=0.655)
fined by including only the fine predictors; namely, the revealed that the factors influencing the period of fixator
level of osteotomy (P=0.0020, 95% confidence interval; application were the final FTA (P=0.042, t=0.212) and
3.6–318), preoperative FTA (P=0.207), FTA at pin lateral cortex fracture (P<0.0001, t=5.15). The age
removal (P=0.741), and FTA at 1 year after surgery (P=0.563, t= 0.585), preoperative FTA (P=0.815,
(P=0.663). When transposition of the proximal frag- t=0.237), and FTA at 1 year after surgery (P=0.570,
ment occurred, the mean FTA at pin removal (169°) had t= 0.574) showed no significant correlations. We
a correction loss at 1 year after surgery (171°) and at the changed the operative procedure to use an image
final follow-up (171°). intensifier for determination of the proper osteotomy
level. Disease severity and the period of fixator appli-
Recurrence of deformity cation were also related to each other. The knees with
severe OA (grade 4) had longer period of application
A graphical assessment by sequential plotting of FTA of (166±67 days) than that of the knees with moderate OA
each knee revealed that recurrence of the varus defor- (grade 2, 3, 112±38 days, P=0.0016).
mity was minimal (Fig. 4). The mean FTA (± SD) was
184° (±5.4) preoperatively, 169° (±2.4) at pin removal,
169° (±3.1) at 1 year after surgery, and 170° (±3.8) at
the final follow-up. In our middle-term follow-up series, Discussion
FTA of most cases was unchanged as shown in the
graph (Fig. 4). The changes of FTA at immediate pin The advantages of hemicallotasis are postoperative
extraction and at the final follow-up were minimal. FTA correction of the osteotomy angle and determination of
592

the correction angle under the weight-bearing condition. The hemicallotasis method is an osteoplastic ap-
When the knee was unstable due to ligamentous and proach to the proximal tibia, which allows conversion to
capsular laxity or subluxation of the femorotibial joint, TKA more readily than the conventional HTO [17].
it was very difficult to determine the precise correction When the conventional osteotomy is applied in unilat-
angle preoperatively [7]. Some authors [2, 24] recom- eral involvement, the operated leg will be shorter than
mended the use of a jig to assist in the accuracy and the sound leg. Compression of the osteotomy site makes
reproducibility of the osteotomy cuts. Billings et al. [2] the tibia shorter than its original length, and conse-
recommended a newer jig and reported satisfactory quently the trouser leg length should be different. We
accuracy for the osteotomy cut. However, all these experienced a man who had undergone unilateral con-
methods could not determine the angle in the standing ventional HTO. He needed a shorter trouser leg on the
position, and planning of osteotomy could not account operated side to suit the operated leg length. The other
for the stability of the joints. Our method, however, side was subjected to the hemicallotasis method. He was
allowed easy determination of the angle postoperatively pleased with the absence of change in the leg length by
and in the standing position even with full weight our method, and without adjustment of the ipsilateral
bearing. trouser leg.
Another merit is low invasion to the knee joint. Our Insall et al. [8] reported that no favorable results were
method did not interfere with the knee joint, the observed in patients older than 60 years of age no matter
peroneal nerve or vessels, and there was no necessity what the degree of correction was. They recommended
of fibular osteotomy or release of the proximal tibi- TKA instead of HTO for more than 60-year-old pa-
ofibular joint. The extensor hallucis longus muscle is tients. However, Yasuda et al. [24] disagreed and re-
innervated by a motor branch of the deep peroneal ported no difference in the knee score between patients
nerve which mostly originates at 68–136 mm from the less than 70 years of age and those older than 70 years.
point of the fibular head and runs close to the fibular Our shortest time of fixator application was 75 days in a
periosteum [11]. This anatomy explained the frequency 70-year-old patient with a correction angle of 13°. Age
of the peroneal nerve damage by lateral side pin did not influence the lateral cortex fracture and trans-
insertion or by fibular bone cut. Charnley clamp, used position of the proximal fragment in the logistic
in the conventional method, needs trans-tibial pin regression analysis, or the period of external fixator
insertion. These pins caused weakness of the tibialis application in the multiple regression analysis.
anterior, the extensor hallucis longus, and the extensor Recently, medial opening wedge HTO has been
digitorum longus muscles, which were penetrated with developed [14, 19–21], which overcame the disadvan-
pins. tages of lateral closed wedge osteotomy in several points.
We did not encounter any compartment syndrome or There is no fibular cut and osteoplastic HTO, leading to
peroneal nerve palsy, because the pins were inserted only easy TKA revision. Nevertheless, opening of medial ti-
in the medial side of the tibia. Nakamura et al. [18] bia was made in one stage, i.e., a drastic change in the
reported the incongruence of the proximal tibiofibular osteotomy site occurred. Spahn [20] discussed medial
joint by our method. However, CT scanning of the opening wedge osteotomy using plate and screws when
proximal tibiofibular joint in several patients (data no the medial side was opened. Tibial head fracture was one
shown) did not show any incongruence of the joint, and of the serious complications of this method (18.2%). But
no patient complained of pain in the proximal or distal other series [14, 21] reported no complications as men-
tibiofibular joints after the operation through the final tioned above. On the contrary, our gradual opening
follow-up, either. The perioperative complications were resulted in no such fracture.
few. Only 8 of 220 pins were infected. Furthermore, However, this procedure had several disadvantages.
infection tended to occur in compromised hosts such as The worst problem was the pin insertion period
patients with diabetes mellitus, edematous legs caused (mean ± SD 128±53 days). This long period increased
by former tibial fractures, and with immunosuppressive the risk of pin-tract infection. Patients could not take a
drugs administration. Catagni reported that HTO using bath for a long time. Of the factors causing this pro-
Ilizarov method was complicated by wire infection in longed fixator application were the lateral tibial cortex
10% of cases. He considered it as a minor problem ra- fracture and transposition of the proximal fragment
ther than a complication [3]. which commonly resulted from the lower level of oste-
Rigid fixation by plating and screws for the osteoto- otomy cut as shown in Results. Our method of osteot-
my lowered the incidence of complications. Six compli- omy was changed from using an oscillating saw directly
cations in 69 knees (9%) occurred after releasing the to the tibia to using an osteotome under an image
proximal tibiofibular joint, leaving the fibula intact [2]. intensifier as described in Results. Furthermore, the line
Koshino et al. reported that in 299 HTOs using broad of osteotomy was made longer by cutting the tibia with
plate, three fractures, nine delayed unions, two infec- inclination toward the proximal lateral cortex through
tions, and three peroneal palsies occurred. They stated the middle of the tibial tuberosity in the AP view (Fig. 2)
that the difficulty in plating and screwing was realign- as described previously [22]. The period of fixator
ment of the fragments after fixation during the postop- application was shortened to 80–90 days (data not
erative course [13]. shown) by the aforementioned improvement. No lateral
 593

cortex fracture of the tibia was shown in a recent series. 5. Fowler JL, Gie GA, Maceachern AG (1991) Upper tibial val-
We excluded severely diseased knee (grade 4) for the gus osteotomy using a dynamic external fixator. J Bone Joint
Surg Br 73:690–691
method, where expected to have longer period of fixator 6. Fujisawa Y, Masuhara K, Shiomi S (1979) The effect of high
application and poorer results. On the contrary, the tibial osteotomy on osteoarthritis of the knee. An arthroscopic
plating and screw method required only 2 weeks of ab- study of 54 knee joints. Orthop Clin North Am 10:585–608
stain from bathing. This can be the most difficult point 7. Insall J, Shoji H, Mayer V (1974) High tibial osteotomy. A five-
year evaluation. J Bone Joint Surg Am 56:1397–1405
in our method. Recurrence of the deformity was rare in 8. Insall JN, Joseph DM, Msika C (1984) High tibial osteotomy
our cases as shown in Fig. 4. We believe that recurrence for varus gonarthrosis. A long-term follow-up study. J Bone
was caused by early pin extraction as mentioned previ- Joint Surg Am 66:1040–1048
ously [3] and delayed union due to fracture of the lateral 9. Jackson JP, Waugh W (1961) Tibial osteotomy for osteoar-
tibial cortex. thritis of the knee. J Bone Joint Surg Br 43:746–751
10. Kellgren JH, Lawrence JS (1957) Radiological assessment of
Our method had no effect on the patellofemoral osteo-arthrosis. Ann Rheum Dis 16:494–502
joint, such as Maquet procedure [16]. Our results 11. Kirgis A, Albrecht S (1992) Palsy of the deep peroneal nerve
showed that patellofemoral osteoarthritis diagnosed by after proximal tibial osteotomy. J Bone Joint Surg Am
plain radiography did not correlate with the patello- 74:1180–1185
12. Klinger HM, Lorenz F, Härer T (2001) Open wedge tibial os-
femoral pain during going up and down stairs. teotomy by hemicallotasis for medial compartment osteoar-
Furthermore, a patient showed improvement of the thritis. Arch Orthop Trauma Surg 121:245–247
patellofemoral joint by our method. These results 13. Koshino T, Morii T, Wada J, Saito H, Ozawa N, Noyori K
might explain the unnecessity for ventralization in the (1989) High tibial osteotomy with fixation by a blade plate for
medial compartment osteoarthritis of the knee. Orthop Clin
femorotibial medial compartment osteoarthritis North Am 20:227–243
involving the patellofemoral joint. 14. Lobenhoffer P, Agneskirchner JD (2003) Improvements in
Finally, our study had several limitations. Nonran- surgical technique of valgus high tibial osteotomy. Knee Surg
domized trial without control shows level IV of the level Sports Traumatol Arthrosc 11:132–138
of evidence. However, our patients were consecutively 15. Magyar G, Toksvig-Larsen S, Lindstrand A (1998) Open
wedge tibial ostoetomy by callus distraction in gonarthrosis.
operated and no other operation except TKA was ap- Acta Orthop Scand 69:147–151
plied in the period of the study. We believe the advan- 16. Maquet P (1963) Un traitement biomécanique de l’arthrose
tage of the procedure is obvious. fémoro-patellaire. L’avancement du tendon rotulien. Rev
In summary, Turi et al. [23] invented hemicallotasis Rheum 30:779–783
17. Mont MA, Alexander N, Krackow KA, Hungerford DS (1994)
as HTO for varusly deformed knee osteoarthritis. We Total knee arthroplasty after failed high tibial osteotomy.
developed a new external fixator for this procedure. The Orthop Clin North Am 25:515–525
result of this method was acceptable in the short-term 18. Nakamura E, Mizuta H, Kudo S, Takagi K, Sakamoto K
radiological and clinical assessment. The most difficult (2001) Open-wedge osteotomy with hemicallotasis. J Bone
Joint Surg Br 83:1111–1115
problem was the long-term application of the external 19. Siguier M, Brumpt B, Siguier T, Piriou Ph, Judet T (2001)
fixator. However, we believe that the merits of our new Ostéotomie tibiale de valgisation originale par ouverture in-
method outweigh the aforementioned demerits. terne sans perte de contact osseux. Technique et incidence sur la
rapidité de consolidation: à propos des 33 premiers cas. Rev
Chir Orthop 87:183–188
Acknowledgement We are grateful to Mr. Nobuo Shirahashi for his 20. Spahn G (2004) Complications in high tibial (medial opening
help in our statistical analysis. wedge) osteotomy. Arch Orthop Trauma Surg 124:649–653
21. Staubli AE, De Simoni C, Babst R, Lobenhoffer P (2003) To-
moFix: a new LCP-concept for open wedge osteotomy of the
References medical proximal tibia—early results in 92 cases. Injury
34(Suppl 2):B55–B62
22. Tsumaki N, Kakiuchi M, Sasaki J, Ochi T, Yoshikawa H (2004)
1. Bauer GCH, Insall J, Koshino T (1969) Tibial osteotomy in Low-intensity pulsed ultrasound accelerates maturation of
gonarthrosis (osteo-arthritis of the knee). J Bone Joint Surg Am callus in patients treated with opening-wedge high tibial oste-
51:1545–1563 otomy by hemicallotasis. J Bone Joint Surg Am 86:2399–2405
2. Billings A, Scott DF, Camargo MP, Hofmann AA (2000) High 23. Turi G, Cassini M, Tomasi PS, Armotti P, Lavini F (1987)
tibial osteotomy with a calibrated osteotomy guide, rigid L’osteotomia direzionale di ginocchio mediante la «emicallot-
internal fixation, and early motion. Long-term follow-up. J asi». Chir Org Mov LXXII:205–209
Bone Joint Surg Am 82:70–79 24. Yasuda K, Majima T, Tsuchida T, Kaneda K (1992) A
3. Catagni MA, Guerreschi F, Ahmad TS, Cattaneo R (1994) ten- to 15-year follow-up observation of high tibial osteo-
Treatment of genu varum in medial compartment osteoarthritis tomy in medial compartment osteoarthrosis. Clin Orthop
of the knee using the Ilizarov method. Orthop Clin North Am 282:186–195
25:509–514
4. Coventry MB (1973) Osteotomy about the knee for degenera-
tive and rheumatoid arthritis. Indications, operative technique,
and results. J Bone Joint Surg Am 55:23–48