Orthopaedics & Traumatology: Surgery & Research (2010) 96, 856—860

ORIGINAL ARTICLE

Post-traumatic knee osteoarthritis treated by

osteotomy only
S. Lustig a, F. Khiami b, P. Boyer c, Y. Catonne b, G. Deschamps d, P. Massin c,∗ ,
The French Hip and Knee Societye

a
Albert-Trillat Surgical Center, North Lyon Teaching Hospital Center, Lyon 1 Claude-Bernard University, 8, rue des Margnolles,
69300 Caluire-et-Cuire, France
b
Orthopaedic Surgery Department, La Pitié-Salpêtrière Hospital, Paris 6, Pierre et Maire Curie University, 47—83, boulevard de
l’Hôpital, 75651 Paris cedex 13, France
c
Orthopaedic Surgery Department, Bichat Hospital, Paris Diderot University, 46, rue Henri-Huchard, 75877 Paris cedex 18, France
d
Medical and Surgical Orthopaedics Center, 2, rue du Pressoir, 71640 Dracy le Fort, France
e
56, rue Boissonade, 75014 Paris, France

Accepted: 3 June 2010

KEYWORDS
Summary
Osteotomy;
Background: Osteotomies to address lower extremity post-traumatic deformities are more
Post-traumatic knee
complex than standard osteotomies performed for congenital deformities, standard osteotomies
arthritis;
and their outcomes are not well known. We performed a multicentric retrospective study
Knee
of these cases. We hypothesized that osteotomy without total knee replacement to correct
fracture malunion deformities can provide long-term relief from athritic pain.
Patients and methods: Twenty-eight patients, mean age 46.4 years old, underwent, between
2000 and 2008, osteotomy for post traumatic osteoarthritis which had resulted in intraar-
ticular malunion in six patients and extraarticular malunion in 22 cases. The initial trauma
had occurred a mean 17.3 years before. There were 11 valgus and 17 varus deformities.
Two osteotomies were performed in the callus to correct intraarticular malunion. The other
osteotomies were performed outside the callus: in 25 cases to correct coronal plane deformities
(nine tibial, 11 femoral and five tibial and femoral), including nine cases with associated dero-
tation. Osteotomies were performed on the distal femoral metaphysis and the proximal tibia.
There was also one case of supramalleolar derotation osteotomy of the tibia. All 28 patients
were contacted again for a consultation. There was a postoperative clinical and radiographic
follow-up of at least 2 years for all patients despite four lost to follow-up patients. There were
18 patients with Ahlback grade 2 arthritis, nine grade 3 and one grade 4.
Results: Two patients with an intraarticular malunion finally underwent revision surgery
to receive total knee replacement because of persistent pain. These patients had
grade 3 and 4 arthritis respectively and undercorrection persisted in the coronal plane.

∗ Corresponding author. Tel.: +33 1 40 25 75 03.

E-mail address: [email protected] (P. Massin).

1877-0568/$ – see front matter © 2010 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.otsr.2010.06.012
Osteotomy for post-traumatic knee arthritis 857

Four patients underwent repeated surgery for stiffness, early infection treated with debride-
ment and antibiotics as well as femoral pseudarthrosis (two cases). After a mean follow-up of
3.8 years, the pain score had improved significantly with more marked improvement in extraar-
ticular malunions (P = 0.03). Functional improvement was moderate (equivalent in patients with
Ahlback grades 2 and 3 arthritis) and articular range of motion did not change. Osteotomy cor-
rected valgus and varus deformities with a mean realignment effect of 9◦ and 10◦ respectively.
Discussion: Osteotomy should correct the three components of the traumatic deformity at the
distal femoral metaphysic level to allow mechanical axis and rotation anomalies correction, and
at the proximal tibia level for realignment purposes. Supramalleolar tibial osteotomy should be
performed for tibial derotation. Pain relief with osteotomy had better outcomes when dealing
with extraarticular malunions. In unicompartmental grade 2 and 3 arthritis, the indications
can be fairly broad in young patients. Besides providing temporary relief, osteotomy facilitates
future total knee replacement surgery in these cases.
Level of evidence: Level 4; non controlled, retrospective study.
© 2010 Elsevier Masson SAS. All rights reserved.

Introduction tis operated on between January 2000 and January 2008

responding to these inclusion criteria were included in
The development of arthritis at a distance from an intra- the study. Twenty-eight cases were identified and patients
or periarticular knee fracture is frequent [1]. Rasmussen were called for a follow-up visit (19 men and nine women,
reported a 21% incidence of post-traumatic arthritis a mean mean age: 46 years old) (range 19—70) (Table 1). Malunions
7.3 years after surgery in a series of 192 patients treated for included 13 tibial malunions, 12 femoral malunions, and
tibial plateau fractures [2]. There is the same risk in frac- three combined femoral and tibial malunions. Six malunions
tures of the distal femur, which are rarer, especially if they were intraarticular and 22 were extraarticular (Table 2).
are intraarticular [3]. Most resulted in coronal/AP plane deformities, 11 valgus and
These fractures often occur in active young patients, 17 varus. Although all of the files did not include a CT scan
in whom it is preferable to delay total knee replacement examination, a rotational abnormality of more than 10◦ was
surgery. The presence of material in the knee, prior surgery clinically identified and confirmed on X-ray in 10 cases of
or the presence of post-traumatic stiffness complicates extraarticular malunion (Table 3).
planning. To our knowledge only two studies in the litera- Analysis of demographic data according to the site of the
ture have reported the results of a series of osteotomies for malunion (Table 1) shows that patients with intraarticular
post-traumatic knee arthritis. In the series by Demsar [4], malunions were significantly younger (P = 0.05). A mean of
14 patients were followed for between 2 and 7 years with two surgical interventions had been performed on the knee
poor results in 43% of cases. More recently Narashima et al. before osteotomy (range 1—9) and the mean delay between
[5] published a series of seven femoral osteotomies with the initial trauma and osteotomy was 17.3 ± 9.1 years
intramedullary nail fixation for post-traumatic knee arthri- (2—41). This interval was significantly shorter for intraar-
tis with a mean follow-up of 5 years. Union was obtained in ticular malunions (4 ± 2 years compared to 21.2 ± 11 years
a mean 28 months. Results were satisfactory except for one for extraarticular malunions; P < 0.01).
failure requiring total knee replacement revision surgery. Initial treatment for femoral fractures included 11 inter-
Our study, performed during the French Hip and Knee nal fixations (six plates and five endomedullary nailings) and
Society (Société française de la hanche et du genou) sym- four cases of orthopedic conservative treatment with long-
posium (Paris 2009) reporting the results of a retrospective term traction, and for tibial fractures, 11 internal fixations
multicentric study including 28 cases of post-traumatic knee (nine plates and two endomedullary nailings) and five ortho-
arthritis treated by osteotomy with a mean postoperative pedic conservative treatment (immobilization in a cast).
follow-up of 44 months, should provide a significant contri- At osteotomy all cases of knee arthritis were unicompart-
bution to this topic. The goal was to identify the indications mental. There were 18 cases of Ahlblack grade 2 arthritis
for this type of surgery according to the type of malunion [6]), nine grade 3 and one grade 4. Angular deviations
and the extent of existing arthritis. The hypothesis was that in absolute values were similar for patients with grade 2
an osteotomy to correct a malunion deformity could provide (9 ± 6◦ ) and grade 3 (8 ± 5◦ ) arthritis. The osteotomy was
long-term pain relief from arthritis. performed on the callus in two cases and included elevation
of an intraarticular malunion. The other osteotomies were
performed outside the callus: in 25 cases for coronal plane
Patients and methods correction (nine tibial, 11 femoral and five combination
tibial and femoral) (Fig. 1) including nine cases with simulta-
Eight French hospital centers participated in collecting this neous correction of a rotational abnormality. All osteotomies
retrospective series of post-traumatic knee arthritis on a were performed on the metaphysis, distal for the femur, and
femoral and/or tibial malunion, treated by femoral and/or proximal for the tibia. Osteotomies of the distal femoral
tibial osteotomy. All cases of post-traumatic knee arthri- metaphysis corrected coronal plane and rotational defor-
858 S. Lustig et al.

Table 1 The series.

IA M DM D Total series

n 6 9 5 8 28
Age (years) 38.4 47.7 47.2 51.2 46.4
Sex-ratio (men [M]/women [W]) 6/0 5/4 3/2 6/2 19/9
Interval between the initial trauma and the osteotomy (years) 4 17 21.4 25.8 17.3
No. of prior surgical procedures 3 1 1 3 2
IA: intraarticular; M: metaphyseal; DM: diaphyseal-metaphyseal; D: diaphyseal.

Table 2 Location of malunion.

IA M DM D Total series

n 6 9 5 8 28
Location
Tibia 5 3 2 3 13
Femur 1 6 1 4 12
Femur + Tibia 2 1 3
IA: intraarticular; M: metaphyseal; DM: Diaphyseal- metaphy-
seal; D: Diaphyseal

mities. On the other hand, only coronal plane corrections

could be obtained with osteotomies of the proximal tib-
ial metaphysis. In one last case, correction of a rotational
malalignment alone was performed by supramalleolar tibial
osteotomy [7].
Twenty-four patients were seen a mean 3.8 years after
surgery (2—8 years), the four remaining patients were con-
sidered to be lost to follow-up. However, there was a
minimum follow-up of 2 years in all 28 cases with a stan-
dard clinical and radiographic evaluation including knee Figure 1 Sixty-five-year-old patient presenting with medial
X-ray, standing view and a patello femoral view. The clin- tibio femoral osteoarthritis on a post-traumatic varus knee,
ical scores were based on a simplified scale derived from 31 years after a fracture of the femoral diaphysis treated sur-
the International Knee Society (IKS), which is better adapted gically with nail fixation. a :the varus deformity is the result of
to retrospective studies [8]. The function score was based an extra-articular diaphyseal malunion of the femur as well as
on two parameters worth 50 points each (walking perimeter, a congenital tibial deformity. The HKA angle is 165◦ compared
and going up and down stairs), added to a negative 20 points to 175◦ on the controlateral side. c: arthritis is stage 3; b: a
if the person required support walking (canes, crutches). A double femoral and tibial osteotomy was performed to correct
score of 90 or more was a sign of a very good result, a score both deformities and obtain normal correction (postoperative
of 70—89 a good result and a score of 60—69 an average HKA 181◦ ). c: at 2 years of follow-up the patient is pain-free
result. Any lower scores were considered poor results. Pain and satisfied; arthritis seems stable.
was evaluated separately on a 50-point scale.

Statistical analysis Results

The Mann-Whitney test for unpaired groups was used to com- Early revision surgery was necessary in two patients who
pare quantitative results of the two groups of patients. A underwent total knee replacement, and thus osteotomy was
P-value of less than 0.05 was considered to be significant. considered to be a failure. Malunion was intraarticular in
both patients, aged 43 and 58 years old. The first presented
Table 3 Spatial direction of malunions.
with significant residual varus and grade 4 arthritis follow-
ing a medial split-depression fracture of the tibial plateau.
IA M DM D Total Series A valgus osteotomy of the tibial metaphysis was unsuccess-
ful and did not correct the angular deformity (obtained HKA
Spatial location
correction 175◦ ) so total knee replacement with a rotating
Frontal/Coronal 6 9 4 7 26
hinge constraint had to be performed 8 months after the
Sagittal 6 1 1 8
osteotomy. The second patient presented with residual val-
Rotational 3 4 5 12
gus in a split fracture of the lateral tibial plateau, which had
IA: intraarticular; M: Metaphyseal; DM: diaphyseal- been insufficiently elevated during initial surgery (Fig. 2).
metaphyseeal; D: Diaphyseal. There was Ahlback grade 2 lateral arthritis. An intraarticular
Osteotomy for post-traumatic knee arthritis 859

Figure 2 a, b, c: a 56-year-old patient presenting with a lateral fracture of the tibial plateau treated by osteosynthesis, which
was insufficiently elevated, resulting in valgus knee; d: an osteotomy was performed on the callus, with elevation of the lateral
tibial plateau and plate osteosynthesis; e: 15 months after the osteotomy, because of incapacitating and persistent pain and slight
persistent valgus, total knee replacement was performed.

osteotomy was attempted to elevate the entire tuberosity, necessity of total knee replacement by several years. This
resulting in good joint congruency but slight persistent val- improvement was even more marked in extraarticular malu-
gus (HKA angle 193◦ ). There was significant residual pain, nions, or if the knee joint had never been treated surgically,
requiring total knee replacement 1 year later. but was indirectly affected by a coronal deformity or rota-
Four additional complications required specific surgical tional malalignment. At final follow-up in this series the good
treatment. Arthroscopic arthrolysis was performed for post- results in pain could be attributed to the trophic effect of
operative stiffness. Early infection was successfully treated osteotomies and different results depending on the initial
by surgical draining and appropriate antibiotic treatment. stage of osteoarthritis could only be obtained in a long-term
Two femoral pseudarthroses required surgical revision with study.
decortication and grafts resulting in union in both cases. The two failures in this study were intraarticular malu-
Osteotomy improved pain in most cases. The mean pain nions with late stage cartilaginous lesions (grades 3 and 4).
score improved significantly in the entire group (13.7 pre- The intraarticular osteotomy performed in these cases was a
operative versus 34.2 at the final follow-up; P = 0.03). Pain difficult and invasive procedure, even if Marti et al. [9] and
improved more in extraarticular malunions (22-point gain Kerkhoffs et al. [10] have reported good long-term results
versus 10-point gain for intraarticular malunions, P = 0.04). in their studies. The decision must be made between this
Mean function also improved significantly from 75.4 pre- indication and a metaphyseal osteotomy of the tibia for
operatively to 81.1 at the final follow-up, for a mean extraarticular realignment which is less invasive and which
improvement of 5.7 points whatever the type of malunion gave longer lasting results in the other cases in our series,
(P = 0.17). The function and pain scores were not signifi- even if pain relief was less satisfactory.
cantly different at the final follow-up between patients with In fact it is difficult to draw conclusions about the failures
Ahlback grade 2 and 3 arthritis (follow-up of 3 ± 2 years for in this study. In the first case there was undercorrection.
grade 2 with a function score of 85 ± 33 and a pain score In the other case, although the osteotomy in the callus
of 34 ± 17, and follow-up of 3 ± 2 years for grade 3 with a seems to have restored joint congruency, it only partially
function score of 88 ± 8 and a pain score of 37 ± 12, respec- corrected tibiofemoral alignment. This is why it may be
tively). Results were excellent in 18 cases, good in five, necessary to associate osteotomy of the callus with extraar-
average in two and poor in three. There was no change ticular osteotomy to obtain durable results in particular for
in preoperative range of motion with a mean preoperative malunion in fractures of the tibial spine or tibial tuber-
active extension decreasing from 3◦ to 2◦ at the final follow- cle of the tibial plateau which associate epiphyseal joint
up, and a mean flexion of 124◦ which remained unchanged. malunion with a metaphyseal deformity. Indeed, besides
After osteotomy the mean preoperative HKA angle of restoring joint congruency, the priority should be to restore
189◦ (180—196◦ ) in 11 valgus malunion deformities had been tibiofemoral alignment by respecting, in our opinion, the
corrected to 181◦ postoperatively (177—187◦ ), for a mean rule of obtaining slight overcorrection in valgus osteotomies
angular correction of 8◦ . After osteotomy the mean preop- and normal correction in varus osteotomies. According to
erative HKA angle of 171◦ (158—179◦ ) in 17 varus malunions recommendations, three of the deformities in our series
had been corrected to 179◦ postoperatively (168—183◦ ), or were markedly undercorrected (two with varus at 168◦ and
a mean angular correction of 7◦ . 175◦ and the other with 187◦ valgus), four other patients who
had initial varus deformities were slightly undercorrected
Discussion (final HAK 178◦ ) as well as two patients with initial valgus
deformities (final HKA, 182 and 183◦ ). While the degree of
valgus in our study was similar to that found in primary knee
When osteotomy was performed on unicompartmental grade
osteoarthritis (9◦ in our series versus 11.6◦ for Backstein
2 or 3 arthritis it significantly relieved pain, delaying the
860 S. Lustig et al.

et al. [11] in a series of 40 femoral varus osteotomies), the Acknowledgements

degree of varus was greater (a mean 9◦ in our series versus
1.3◦ for Takeuchi et al. [12] in a series of 57 valgus tib- The authors would like to thank Professors Argenson,
ial osteotomies). Combined tibial and femoral osteotomies Catonné, Huten, Judet, Massin, Neyret and Piriou, and the
were necessary when deformities included both bones, and Doctors Bonnin and Deschamps who provided the files for
in particular in case of a traumatic deformity on one bone this multicentric study.
associated with a marked congenital deformity on the other
(Fig. 1).
Correction of the femur is generally performed by open
References
osteotomy on the distal metaphysis allowing correction of
frontal as well as rotational malalignments. A series of 14 [1] Bedi A, Haidukewych GJ. Management of the posttraumatic
arthritic knee. J Am Acad Orthop Surg 2009;17:88—101.
cases of femoral derotation with static interlocking nail-
[2] Rasmussen PS. Tibial condylar fractures as a cause of degener-
ing were reported by Stahl et al. [13]. The authors used ative arthritis. Acta Orthop Scand 1972;43:566—75.
an endomedullary saw and obtained union in all cases with [3] Kettelkamp DB, Hillberry BM, Murrish DE, Heck DA. Degenera-
derotation abnormalities of up to 63◦ , emphasizing the tive arthritis of the knee secondary to fracture malunion. Clin
interest of this closed procedure. When osteotomy is per- Orthop 1988:159—69.
formed on the diaphyseal femur, nail fixation seems to [4] Demsar A. Osteotomy of the tibia in the prevention and
provide better results, even if union takes longer. [5]. In therapy of post-traumatic knee arthrosis. Acta Chir Iugosl
our series, although the open osteotomies were performed 1977;24(Suppl. 2):219—23.
on the femoral metaphysis they were complicated by pseu- [5] Narasimha Murty A, El Zebdeh MY, Ireland J. Femoral shaft
darthrosis in two cases, when fixation with simple staples osteotomy for malunited femoral fractures presenting with
varus OA of the knee. Knee 2004;11:445—9.
was used. Osteotomy of the proximal tibial metaphysis
[6] Ahlback S. Osteoarthrosis of the knee. A radiographic investi-
cannot correct frontal plane deformities, while rotational gation. Acta Radiol Diagn (Stockh) 1968;(Suppl. 277):7—72.
malalignment should be corrected distally in the supra- [7] Judet H, Judet J. La réorientation de l’articulation tibio-
malleolar region [7]. Indeed, high tibial derotations are tarsienne. Chirurgie 1971;97:638—45.
known to be difficult, with a high rate of complications [8] Massin P, Lautridou C, Cappelli M, et al. Total knee arthro-
including fibular paralyses or compartment syndromes [14]. plasty with limitations of flexion. Orthop Traumatol Surg Res
The total rate of complications in this series was sim- 2009;95:1—6.
ilar to that in ‘‘conventional’’ series of tibial (15% for [9] Marti RK, Kerkhoffs GM, Rademakers MV. Correction of lateral
Insall [15]) or femoral (16,3% for Cameron et al. [16]) tibial plateau depression and valgus malunion of the proximal
osteotomies. tibia. Oper Orthop Traumatol 2007;19:101—13.
[10] Kerkhoffs GM, Rademakers MV, Altena M, Marti RK. Com-
In conclusion, after planning based on a three-
bined intra-articular and varus opening wedge osteotomy
dimensional preoperative evaluation of the malunion, an for lateral depression and valgus malunion of the proximal
osteotomy of the femoral and/or tibial metaphysis that part of the tibia. Surgical technique. J Bone Joint Surg Am
corrects all aspects of the initial deformity can provide 2009;91(Suppl. 2 Pt 1):101—15.
good intermediate term results. In our experience an [11] Backstein D, Morag G, Hanna S, Safir O, Gross A. Long-term
extraarticular osteotomy is preferable, even in intraartic- follow-up of distal femoral varus osteotomy of the knee. J
ular malunions if there are no major problems of joint Arthroplasty 2007;22:2—6.
congruency. Osteotomy of the metaphysis should be fixed [12] Takeuchi R, Ishikawa H, Aratake M, et al. Medial opening
with a plate and screws, or a blade plate, which maintains wedge high tibial osteotomy with early full weight bearing.
good contact between the fragments. The best indica- Arthroscopy 2009;25:46—53.
[13] Stahl JP, Alt V, Kraus R, et al. Derotation of post-traumatic
tion is moderate unicompartmental knee arthritis, but in
femoral deformities by closed intramedullary sawing. Injury
young subjects lasting results can be obtained in later 2006;37:145—51.
stages even in the presence of complete joint space [14] Krengel WFr, Staheli LT. Tibial rotational osteotomy for idio-
narrowing. In case of failure, correction of the initial pathic torsion. A comparison of the proximal and distal
deformity facilitates the total knee replacement proce- osteotomy levels. Clin Orthop 1992;283:285—9.
dure. [15] Insall JN. Osteotomy. In: W R, Insall JN, Scorr WN, et al.,
editors. Surgery of the knee. 2nd ed New York Churchill Liv-
ingstone; 1993. p. 635—75.
Conflict of interest statement [16] Cameron HU, Botsford DJ, Park YS. Prognostic factors in the
outcome of supracondylar femoral osteotomy for lateral com-
No conflicts of interest for any of the authors. partment osteoarthritis of the knee. Can J Surg 1997;40:114—8.