This study reviewed the long-term outcomes of total knee arthroplasty (TKA) in patients with severe valgus knee deformity (variant-III). 32 patients (37 knees) underwent TKA with an average follow up of 10 years. The mean preoperative valgus alignment of 33 degrees was corrected to nearly neutral alignment postoperatively. Clinical and functional outcomes significantly improved based on HSS knee scores and range of motion. No revisions were required. Complications included 3 transient peroneal nerve palsies and 2 DVTs, but no infections or loosening. TKA can successfully treat severe valgus deformity with proper soft tissue balancing and implant selection.

1. Chin Med J 2014;127 (6)1062
DOI: 10.3760/cma.j.issn.0366-6999.20132488
Department of Orthopaedics, Beijing Jishuitan Hospital, Beijing
100035, China (Zhou XH, Zhang L and Zhou YX)
Department of Orthopeadic, Xinqiao Hospital, Third Military
Medical University, Chongqing 400037, China (Wang M and Liu C)
Correspondence to: Dr. Wang Min, Department of Orthopeadic,
Xinqiao Hospital, Third Military Medical University, No. 2 Xinqiao
Street, Chongqing 400037, China (Tel: 86-23-68774308. Email:
52solar@sina.cn)
Original article
Total knee arthroplasty for severe valgus knee deformity
Zhou Xinhua,Wang Min, Liu Chao, Zhang Liang and Zhou Yixin
Keywords: total knee arthroplasty; variant-III valgus knees
Background Primary total knee arthroplasty (TKA) in severe valgus knees may prove challenging, and choice of implant
depends on the severity of the valgus deformity and the extent of soft-tissue release. The purpose of this study was to
review 8 to 11 years (mean, 10 years) follow-up results of primary TKA for varient-III valgus knee deformity with use of
different type implants.
Methods Between January 2002 and January 2005, 20 women and 12 men, aged 47 to 63 (mean, 57.19±6.08) years old,
with varient-III valgus knees underwent primary TKA. Of the 32 patients, 37 knees had varient-III deformities. Pie crusting
was carefully performed with small, multiple inside-out incisions, bone resection balanced the knee in lieu of soft tissue
releases that were not used in the series. Cruciate-retaining knees (Gemini MKII, Link Company, Germany) were used in
13 knees, Genesis II (Simth & Nephew Company, USA) in 14 knees, and hinged knee (Endo-Model Company, Germany)
in 10 knees. In five patients with bilateral variant-III TKAs, three patients underwent 1-stage bilateral procedures, and two
underwent 2-stage procedures. All implants were cemented and the patella was not resurfaced. The Hospital for Special
Surgery (HSS) knee score was assessed. Patients were followed up from 8 to 11 years.
Results The mean HSS knee score were improved from 50.33±11.60 to 90.06±3.07 (P <0.001). The mean tibiofemoral
alignment were improved from valgus 32.72º±9.68º pre-operation to 4.89º±0.90º post-operation (P <0.001). The mean
range of motion were improved from 93.72º±23.69º pre-operation to 116.61±16.29º post-operation (P <0.001). No
patients underwent revision. One patient underwent open reduction and internal fixation using femoral condylar plates
for supracondylar femoral fractures secondary to a fall at three years. Three patients developed transient peroneal nerve
palsies, which resolved within nine months. Two patients developed symptomatic deep vein thrombosis that was managed
with rivaroxaban and thrombo-embolic deterrent stockings. There was no incidence of pulmonary embolism. Postoperative
patient satisfaction was 80.7±10.4 points in the groups. Prosthetic survival rate was 100% at mean 10 years post-
operative.
Conclusions Not only hinged implants can be successfully used in variant-III valgus knees. As our results show, if proper
ligament balancing techniques are used and proper ligament balance is attained, the knee may not require the use of a
more constrained components. Our results also present alternative implant choices for severe knee deformities.
Chin Med J 2014;127 (6): 1062-1066
Severe valgus knees are defined as those with a valgus
alignment of >20° on standing anteroposterior
radiographs1,2
and classified as variant-III,3
which account
for 5% of all valgus deformities. Variant-III deformities
are distinguished by attenuation of the medial capsular
ligament complex, marked lateral soft-tissue contracture,
and marked osseous deficiency of both the lateral femoral
condyle and the lateral tibial plateau. Soft-tissue balance
is difficult to achieve and a constrained type of implant is
often required. The purpose of this report was to review the
long-term results of knee arthroplasty achieved in knees
with a preoperative severe valgus orientation.
METHODS
Between January 2002 and January 2005, 20 women and
12 men, aged 47 to 63 years old (mean, 57.19±6.08),
with severe valgus knees, underwent primary total knee
arthroplasty (TKA) by a senior surgeon for rheumatoid
arthritis (n=10) or primary osteoarthritis (n=22).
Demographic data on the patients are summarized as
height (160.13±9.54) cm, weight (65.06±12.65) kg, and
body mass index (BMI) (25.15±2.93) kg/m2
. With regard
to these patients, 37 knees had severe valgus deformities
defined as type-III.4
The preoperative weight bearing valgus
(tibiofemoral alignment) was between 21° and 52° (mean,
32.72°±9.68°) (Figures 1A, 2A and 3A). There was no
history of proximal tibial or distal femoral osteotomy. Any
patient who had neurological disorders, diabetes mellitus,
severe osteoarthritis of the hip joints or muscular imbalance
in the lower extremities was excluded from the study. Each
patient was subjected to spinal epidural anesthesia. The
knee was exposed in flexion using a midline longitudinal
skin incision and a lateral parapatellar arthrotomy that
permited direct exposure of the contracted soft tissue in the
lateral aspect of the kneel. In this orientation an extensive
lateral release is much easier to perform. A tourniquet was

2. Chinese Medical Journal 2014;127 (6) 1063
used. In five patients with bilateral variant-III TKAs, three
patients underwent 1-stage bilateral procedures, and two
underwent 2-stage bilateral procedures.
The tibia was then prepared using an extra-medullary
jig at first, the femur was prepared using intramedullary
instruments. The distal femur was resected in 6º of valgus.
The rotational alignment of the femur was referenced to
the Whiteside line and transepicondylar axis. In some
cases of valgus deformity, the lateral epicondyle is not
prominent, and defining a distinct lateral point for the axis
may be imprecise. With these axises in view, the amount of
posterior femoral condylar deformity is estimated. A spacer
block was then used to check the flexion and extension
gaps. The selective release of the tight lateral structures
technique (Pie crusting) was carefully performed with
small, multiple inside-out incisions. The release always
started from the posterolateral capsule at the level of the tibial cut and then moved anteriorly to involve the lateral
Figure 3. A 78-year-old male patient suffering right severe knee
deformity. A: AP radiograph of right knee. B, C: AP and lateral
radiograph of right knee postoperation.
Figure 1. A 67-year-old female patient suffering from OA with
left severe valgus deformity. A: Left knee clinical photograph
preoperation. B: Clinical photograph postoperation. C, D:
Radiograph of left knee pre-operation. E, F: Radiograph after
TKA with a CR prosthesis (Link company, Germany).
Figure 2. A 64-year-old female patient with severe left knee
valgus deformity. A: Left knee clinical photograph pre-
operation. B: Pre-operative full leg standing X-ray shows valgus
knee deformity. C, D: Radiograph after TKA with use of hinge
knee prosthesis (Link company, Germany). E: Left knee clinical
photograph post-operation. F: Radiograph of left knee shows
dislocation of patellar.

3. Chin Med J 2014;127 (6)1064
collateral ligament and the iliotibial band. The popliteus
tendon was released in nine knees, and gap changes
were tested by spacer block after each puncture until gap
symmetry was achieved. All 37 severe knees obtained
symmetrical flexion and extension gaps under fractional
lengthening by means of gradual release of the lateral
collateral ligament (LCL). Based on the competency of the
medial collateral ligament, advancement of the ligament
was not needed in any of the knees. Lateral release must
be performed in a graduated, organized stepwise fashion to
provide adequate stability without over releasing. Finally,
we assessed gap balancing by placing the trial components
and applying varus and valgus stress to the knee again until
gap symmetry was seen. Patellar maltracking was judged
by a no-thumb technique.
Valgus was predominantly of femoral origin, bone deficits
in the lateral femoral condyle (15 knees) or lateral the
tibial plateau (three knees) may require bone-grafting as
estimated, and the remaining defects in six lateral femoral
condyles were filled with autograft bone taken from other
cuts during the procedure. Bone resection balance of the
knee in lieu of soft tissue releases were not used in the
series, and a tibial tubercle osteotomy is not suggested here
yet.
In the 37 knees with a variant-III deformity, cruciate-
retaining knees (Gemini MK II, Link Company, Germany.
Figure 1E and 1F) were used in 13 knees, Posterior
Sacrified implant Genesis II (Simth & Nephew Company,
USA, Figure 3B and 3C) in 14 knees, and hinged knee
(Endo-Model Company, Germany. Figure 2C and 2D) in 10
knees. All implants were cemented and the patella was not
resurfaced. Postoperatively the knees were placed in 10º of
flexion for 3–4 days to prevent stretching of the peroneal
nerve, active and passive range-of-motion exercises (within
a range of 10º to 70º) were allowed. The mean duration
of hospital stay was 14 days with a range of 12–15 days.
Patients were followed up by the same surgeon at 3, 6, and
12 months, and then yearly.
The Hospital for Special Surgery (HSS) knee score was
assessed. Limb alignment, patellar position, and evidence
of loosening or osteolysis were assessed using standing
anteroposterior, lateral, and skyline radiographs according
to the Knee Society roentgenographic evaluation system.
Mediolateral stability was assessed by a varus-valgus
stress in full extension. Anteroposterior stability was
assessed by an anteroposterior stress test in 30º and 90º of
flexion. The patient’s active range of motion was measured
using a goniometer. Patient satisfaction was assessed by
the questioner as 100 points at full mark, with the best
condition equaling 100 points.
Statistical analysis was performed using SPSS 13.0 (SPSS,
Inc., Chicago, IL), the paired t-test or independent-samples
t-test was employed for comparison of preoperative and
postoperative results. Results were considered statistically
significant at P <0.05.
RESULTS
Routine postoperative care after all total knee arthroplasties
included early mobilization, walking with full weight-
bearing, and quadriceps-strengthening and range-of-motion
exercises beginning after 2–3 days postoperatively under a
physiotherapist’s supervision.
All patients were followed up for 8 to 11 years (mean of
10 years), all knees had good patellar position and were
clinically stable in both mediolateral and anteroposterior
planes. The mean HSS knee score was improved from
50.33±11.60 (range, 30º–70º) preoperation to 90.06±3.07
(range, 83º–95º) postoperation. There was significant
improvement in the postoperative knee score (P <0.001).
The mean tibiofemoral alignment was improved from
valgus 32.72º±9.68º (range, 21º–52º) preoperation to
4.89º±0.90º (range, 3º–9º) postoperation (Figures 1B, 1E,
1F, 2C-2E, 3B and 3C). There were significant differences
between the tibiofemoral alignments at preoperation and
those at postoperation (P <0.001). The mean range of
motion improved from 93.72º±23.69º (range, 47º–127º)
preoperation to 116.61º±16.36º (range, 81º–136º)
postoperation. There were significant differences between
the preoperation and postoperation range of motion in terms
of arc of range of motion (P <0.001, Table 1). No patient
underwent revision. One patient underwent open reduction
and internal fixation using femoral condylar plates for
supracondylar femoral fractures secondary to a fall at three
years. Three patients developed transient peroneal nerve
palsies, which resolved within nine months. Two patients
developed symptomatic deep vein thrombosis; there was no
incidence of pulmonary embolism.
Four patients had moderate knee pain (4–5 by VAS
evaluation), three of them complained of posterior knee
pain at passive maximum flexion and one had severe pain
for the first seven months after surgery. Only one patient
walked with a cane; a bilateral TKRs case. The other
patients could walk without canes and could go up or down
stairs slowly. Twenty-four patients could participate in
recreational sports such as fishing and gardening.
On the skyline radiographs a normal patellar position was
observed in 36 knees and abnormal patellar tilt greater than
4° was observed in only one knee. Radiographic evaluation
showed no femoral and tibial osteolysis adjacent to
components in any patients at the time of the latest follow-
up. No tibial or femoral component had radiographic
evidence of loosening, and no subsidence was detected.
No late-onset instability was displayed in the follow-up.
Postoperative patient satisfaction was 80.7±10.4 points in
the groups. Prosthetic survival rate was 100% at a mean of
Table 1. Range of motion (degree)
Time Extension Flexion Arc
Preoperation –0.78±2.49 92.94±23.43 93.72±23.69
Postoperation –0.33±1.49 116.28±16.29*
116.61±16.36*
*
A significant difference was observed between the preoperation and postoperation
in terms of flexion and arc of range of motion (P <0.001).

4. Chinese Medical Journal 2014;127 (6) 1065
10 years postoperatively
DISCUSSION
To our knowledge, the mean preoperative tibiofemoral
valgus angle (32.72°±9.68°) in this study is the highest
reported in the literature. All total knee prosthesis achieved
satisfactory clinical results in variant-III valgus knees
arthroplasty in our study and no revisions were needed for
any reason during a mean follow-up of 10 years.
Adequate soft-tissue balance in variant-III valgus knees
during TKA is really challenging. Any tight lateral
structures were released by means of pie crusting with a
low complication rate with this practice. Some authors
advocate reconstruction of the medial collateral ligament
complex. In this study we did not have medial advancement
of the medial collateral ligament as a means to limit the
degree of soft-tissue release. This avoided drawbacks such
as non-union at the medial collateral ligament advancement
site, increased surgery time, and delayed mobilization,5
it also avoided late-onset instability in 100% of these
variant-III valgus knees. In our study implants in variant-
III knees would not have led to late-onset instability after
nearly 10 years. Good stability and deformity correction
were reported in all knees, and there was no incidence of
flexion instability, mediolateral instability, and there was no
patellar dislocation. A tibial tubercle osteotomy is suggested
in cases where the Q angle was >20º, but it has inherent
risks of non-union and subsequent extensor mechanism
problems, it was not used in our cases. Augmentation of the
deficient condyle has been suggested by some authors. In
our study, augmentation was applied in 18 cases, although
Koskinen et al6
reported that augmentation of the deficient
condyle with bone graft or metal blocks was not necessary.
In the last 25 years, a number of different surgical
techniques and implants have been proposed in TKRs,
but the variant-III valgus knee still presents a formidable
reconstructive challenge. It is unknown whether and how
implant variables influence outcomes in the variant-III knee
deformities, and there is also no consensus on the degree of
implant constraint that should be used in these cases. Our
results present an alternative implant choice for the variant-
III valgus knee deformities in younger patients.
Choosing the constraint level according to balancing
difficulties appears to be more prudent. Girard et al7
published results that each increase of 1º in laxity in valgus
augmented the risk of inserting a constrained prosthesis
by 1.9 folds (P=0.000 3). Adolph et al3
recommended
that more severe deformity may require more aggressive
releases and a posterior stabilized constrained, or hinged
total knee arthroplasty; hinged implants are recommended
in elderly patients with severe irreducible valgus.8
But
authors have raised concerns that these high-constraint
prosthesis and hinge prostheses represent a higher risk of
loosening and exposure to technical difficulties in case
of revision.9
Girard et al7
emphasized that high-constraint
prostheses have preferentially been used in elderly and less
active patients, as their survival at five years is less than
70%, despite the introduction of newly-conceived, modern
implants. Insall and Easley10
suggested that a posterior
cruciate ligament-substituting prosthesis should be selected
for elderly patients, and many authors have advocated the
use of PCL-substituting implant designs, avoiding concerns
with PCL balancing and dealing with a potentially abnormal
native ligament.11
Conversely, others have published results
that supported the use of PCL-retaining designs with
various soft tissue balancing techniques.12,13
McAuley et al4
concluded cruciate-retaining implants could be used in a
wide range of patients with valgus osteoarthritis. Posterior
cruciate-retaining prostheses have significantly improved
the survival rate in comparison with posterior cruciate-
stabilizing prostheses at fifteen years.14
As our results show, if proper ligament balancing techniques
are used and proper ligament balance is attained, the knee
may not require the use of a more constrained component.
Aglietti et al15
confirms that selective multiple-puncture
release of the lateral soft tissues of the knee joint corrects
moderate to severe valgus deformities independently from
the implant. We have preferred to use the cruciate-retaining
model in valgus patients in order to save bone stock for
future possible needs, especially among younger patients.
We have achieved excellent results in varus or valgus
deformity TKAs with the implants. The choice of whether
to use a PCL-retaining design or posterior-stabilized design
for variant-III TKA is based on limited data, and few trials
have been published presenting results of this prosthesis in
variant-III valgus knees.
In a study using a cruciate-retaining implant, no
significant difference was reported in terms of knee score,
alignment, or rate of surgical revision in patients with
severe (≥20º) varus or valgus deformity and matched
controls.16
Revisions were due to surgical errors and
might have been avoided using proper surgical technique
rather than implants. Cruciate-retaining implants can be
successfully used in valgus knees and implant survival
can be improved if at least one of the lateral-stabilizing
structures is preserved.4,9
Aglietti et al15
reported a series
of patients where a rotating and anteroposterior gliding
mobile bearing, cruciate-retaining knee was implanted, and
no hinged knees implants were used, although 13% of the
patient’s preoperative weight bearing valgus mechanical
axis was greater than 20°.
There are some limitations regarding the findings of our
current clinical study due to the absence of a large series
and long term follow-up, and more extensive functional
analyses have to be employed to finally evaluate the
theoretical advantage of implants. A constrained implant
is also an important alternative selection for a successful
TKA for variant-III valgus knees, especial for elderly
patients over 75 years. If proper soft-tissue balance cannot
be achieved or there is no functional medial collateral
ligament present in elderly patients, more constrained
implants should be used. The results present an alternative

5. Chin Med J 2014;127 (6)1066
implant choice for the variant-III valgus knee deformities
for patients less than 65 years old.
REFERENCES
1. Beverland D. Management of the severe varus and valgus knee
using the low contact stress rotating platform. Orthopedics 2006;
29 (Suppl): S60-S63.
2. Ranawat AS, Ranawat CS, Elkus M, Rasquinha VJ, Rossi R,
Babhulkar S. Total knee arthroplasty for severe valgus deformity.
J Bone Joint Surg 2005; 87: 271-283.
3. Adolph V, Lombardi Jr, Kathleen L. An algorithmic approach
to total knee arthroplasty in the valgus knee. J Bone Joint Surg
2004; 86 (Suppl 2): 62-71.
4. McAuley JP, Collier MB, Hamilton WG, Tabaraee E, Engh GA.
Posterior cruciate-retaining total knee arthroplasty for valgus
osteoarthritis. Clin Orthop Relat Res 2008; 466: 2644-2649.
5. Elkus M, Ranawat CS, Rasquinha VJ, Babhulkar S, Rossi R,
Ranawat AS. Total knee arthroplasty for severe valgus deformity.
Five to fourteen-year follow-up. J Bone Joint Surg Am 2004; 86:
2671-2676.
6. Koskinen E, Ville R, Pekka P, Harilainen A, Sandelin J, Tallroth
K, et al. Results of total knee replacement with a cruciate-
retaining model for severe valgus deformity: a study of 48
patients followed for an average of 9 years. Knee 2011; 18: 145-
150.
7. Girard J, Amzallag M, Pasquier G, Mullieza A, Brosseta T,
Gougeonc F, et al. Total knee arthroplasty in valgus knees:
Predictive preoperative parameters influencing a constrained
design selection. Orthop Traumatol Surg Res 2009; 95: 260-266.
8. Williot A, Rosset P, Favard L, Brilhault J, Burdin P. Total knee
arthroplasty in valgus knee. Orthop Traumatol Surg Res 2010;
96 (Suppl): S37-S42.
9. Pour AE, Parvizi J, Slenker N, Purtill JJ, Sharkey PF. Rotating
hinged total knee replacement: use with caution. J Bone Joint
Surg (Am) 2007; 89: 1735-1741.
10. Insall JN, Easley ME. Surgical techniques and instrumentation
in total knee arthroplasty. In: Surgery of the Knee, 3rd ed. Insall
JN, Scott WN, eds. Churchill Livingstone: New York; 2001:
1717-1738.
11. Anderson JA, Baldini A, MacDonald JH, Pellici PM, Sculco TP.
Primary constrained condylar knee arthroplasty without stem
extensions for the valgus knee. Clin Orthop Relat Res 2006;
442: 199-203.
12. Politi J, Scott R. Balancing severe valgus deformity in total
knee arthroplasty using a lateral cruciform retinacular release. J
Arthroplasty 2004; 19: 553-557.
13. Engh GA. The difficult knee: severe varus and valgus. Clin
Orthop Relat Res 2003; 416: 58-63.
14. Abdel MP, Morrey ME, Jensen MR, Morrey BF. Increased long-
term survival of posterior cruciate-retaining versus posterior
cruciate-stabilizing total knee replacements. J Bone Joint Surg
Am 2011; 93: 2072-2078.
15. Aglietti P, Lup D, Cuomo P, Baldini A, De Luca L. Total knee
arthroplasty using a pie-crusting technique for valgus deformity.
Clin Orthop Relat Res 2007; 464: 73-77.
16. Ritter MA, Faris GW, Faris PM, Davis KE. Total knee
arthroplasty in patients with angular varus or valgus deformities
of > or = 20 degrees. J Arthroplasty 2004; 19: 862-866.
(Received September 25, 2013)
Edited by Wang De