The Knee 25 (2018) 167–176

Contents lists available at ScienceDirect

The Knee

Outcomes of total knee arthroplasty in degenerative

osteoarthritic knee with genu recurvatum
Seung-Suk Seo a, Chang-Wan Kim b, Chang-Rack Lee b,⁎, Jin-Hyuk Seo c, Do-Hun Kim c, Ok-Gul Kim c
a
Department of Orthopedic Surgery, Haeundae Bumin Hospital, 584 Haeundae-ro, Haeundae-gu, Busan, Republic of Korea
b
Department of Orthopedic Surgery, Inje University Busan Paik Hospital, 75 Bokji-ro, Busanjin-gu, Busan, Republic of Korea
c
Department of Orthopedic Surgery, Bumin Hospital, 59 Mandeok-daero, Buk-gu, Busan, Republic of Korea

a r t i c l e i n f o a b s t r a c t

Article history: Background: This study aimed to assess the incidence of genu recurvatum without neuromuscular
Received 30 June 2017
disorders in knees that underwent navigation-assisted total knee arthroplasty (TKA), to evaluate
Received in revised form 17 October 2017
short-term radiologic and clinical results of navigation-assisted TKA in genu recurvatum, and to
Accepted 31 October 2017
evaluate differences in results according to the degree of pre-operative hyperextension and type
of implant and insert.
Keywords:
Methods: This study retrospectively reviewed 510 knees that underwent navigation-assisted TKA
Genu recurvatum
from January 2005 to December 2011. The incidence of knees that showed hyperextension
Navigation
Outcome
of ≥5° (genu recurvatum) on navigation, and the accompanying alignment were evaluated. It
Total knee arthroplasty assessed radiologic, intraoperative, and clinical results in recurvatum and control groups by
using propensity score matching.
Results: A total of 465 knees underwent navigation-assisted TKA for degenerative osteoarthritis.
Genu recurvatum was observed in 55 knees (11.8%). Of these, 41 knees (74.5%) had degree of
hyperextension between five degrees and 10°, and 47 (85.4%) had varus alignment. The thick-
ness of the resected distal femur in the recurvatum group (7.6 ± 1.6 mm) was less than that
in the control group (8.4 ± 1.4 mm, P = 0.001). The thickness of the insert in the recurvatum
group (12.5 ± 2.3 mm) was greater than in the control group (10.8 ± 1.5 mm, P b 0.001). The
sagittal alignment at the final follow-up was 1.3 ± 3.4° in the control group and −0.1 ± 0.7° in
the recurvatum group (P = 0.003). Subgroup analyses in the recurvatum group showed no sig-
nificant difference in sagittal alignment and patient-related outcomes by degree of pre-operative
hyperextension and implant/insert type (P N 0.05 for all parameters).
Conclusions: Genu recurvatum was not uncommon among patients undergoing primary TKA.
This review obtained satisfactory short-term clinical and radiologic results, with a smaller distal
femoral resection and thicker insert.
© 2017 Elsevier B.V. All rights reserved.

1. Introduction

Genu recurvatum is related to neuromuscular disorders such as poliomyelitis [1–3]. In cases where limbs have been affected
by a neuromuscular disorder, total knee arthroplasty (TKA) can result in pain relief. However, concerns have been raised about
recurrence of knee instability and functional deterioration in patients with poor pre-operative quadriceps function. Hence, TKA
is contraindicated in knees with severe muscle weakness. It has been reported that in knees without neuromuscular disorder, a
recurvatum deformity can develop in conditions such as genu valgum and rheumatoid arthritis [4,5]. Genu recurvatum has been reported
as being uncommon among patients who have undergone TKA. Some studies (before the early 2000s) have reported a b1% incidence of

⁎ Corresponding author.
E-mail address: leechangrack@inje.ac.kr (C.-R. Lee).

https://doi.org/10.1016/j.knee.2017.10.008
0968-0160/© 2017 Elsevier B.V. All rights reserved.
168 S.-S. Seo et al. / The Knee 25 (2018) 167–176

genu recurvatum in knees that required TKA [6–8]. However, these studies were limited by lower diagnostic accuracy, as they either used
a goniometer to identify genu recurvatum or did not clearly describe the hyperextension measurement method. In 2012, Mullaji et al.
objectively confirmed genu recurvatum using navigation and reported an incidence rate of 3.9% [9]. It is interesting that the incidence
increased when genu recurvatum was checked by navigation. The difference in reported genu recurvatum incidence rates between
previous literature and Mullaji et al. may be attributable to anesthesia, diagnosis method, and racial differences [9]. A few studies have
investigated the incidence of genu recurvatum so far, and more research on this topic is required.
Satisfactory clinical and radiologic results have been reported for TKA on knees with genu recurvatum [6–9]. However, existing
literature is limited by small numbers of cases or a follow-up period b2 years. Moreover, when evaluating TKA results in knees
with genu recurvatum, factors likely to affect the results should be considered such as the degree of pre-operative hyperextension
and type of implant and insert. However, it is believed that there is no available literature on these factors.
Considering the lack of literature related to genu recurvatum in TKA and the methodological problems of previous studies,
there is a need for more research on genu recurvatum in knees undergoing TKA. Therefore, the current authors developed the
following specific research questions: (1) Is the incidence of genu recurvatum in knees having undergone TKA as rare as reported
in previous studies? (2) Are short-term radiologic and clinical results good? (3) Do clinical results differ according to the degree of
pre-operative hyperextension and implant and insert type?
The current study aimed to (1) assess the incidence of genu recurvatum without neuromuscular disorders in knees that
underwent navigation-assisted TKA, (2) evaluate short-term radiologic and clinical results of navigation-assisted TKA in genu
recurvatum, and (3) evaluate differences in results by the degree of pre-operative hyperextension and type of implant and insert.
It was assumed that the incidence of genu recurvatum would be higher in knees undergoing TKA than those previously reported,
and that TKA results would be good, regardless of the degree of pre-operative hyperextension and implant or insert type.

2. Materials and methods

This study was approved by the Institutional Review Board. From January 2005 to December 2011; 612 primary TKAs were
implemented by one surgeon at the current institution. Among them, 510 cases of TKA were implemented by using navigation.
Inclusion criteria for the genu recurvatum group in the current study were (1) degenerative osteoarthritis and (2) hyperextension
≥5° under spinal or general anesthesia on navigation. Patients excluded were those with a follow-up period b 2 years; with history
of fracture or ligament injury in the ipsilateral knee; and with inflammatory arthritis, neuropathic arthropathy, or neuromuscular
disorders. The study attempted to compare the results between the recurvatum group and the control group. To minimize selec-
tion bias, the control group was selected from patients who had received TKA due to degenerative osteoarthritis during the same
period without pre-operative genu recurvatum ≥5°, according to the propensity score matching (1:2, nearest neighbor matching
without replacement). Matching variables included age, sex, body mass index, and pre-operative femorotibial angle. Navigation-
assisted TKA was performed in almost all cases during the study period.
The navigation used in the included cases was the OrthoPilot computer-aided surgery system (B. Braun, Aesculap, Tuttlingen,
Germany), with OrthoPilot Tibia First with Soft Tissue Management TKA v4.0, v4.2, and v4.4 software (the different versions of the
software used corresponded to software updates). The minimal detectable angular and length change of the navigation system
were one degree and one millimeter, respectively. The surgical technique for navigation-assisted TKA is described in detail
elsewhere [10,11]. In brief, patients were placed in a supine position on the surgery table under general or spinal anesthesia.
The knee was exposed by medial parapatellar arthrotomy. The navigation tracker was located on the distal femur approximately
six centimeters above the articular surface and on the proximal tibia six centimeters below the articular surface. After registering
the knee, hip, and ankle joints in the navigation system, anatomical landmarks were registered, including the intercondylar notch
and medial and lateral epicondyles. Next, the alignment of the lower extremity was evaluated on coronal and sagittal planes. Bone
resection was performed on the proximal tibia, distal femur, anterior and posterior femur (in that order). The proximal tibia was
resected to be 0° to the mechanical axis of the tibia on the coronal plane, with a three-degree posterior slope on the sagittal plane.
Flexion and extension gaps were checked after resection of the proximal tibia. Bone resection of the distal femur was targeted so
that the distal transverse plane cut was 0° to the mechanical axis of the femur. In cases of a large extension gap, resection of the
distal femur was reduced by one to two millimeters [12]. Soft tissue release was performed with the knee in full extension such
that the difference in the medial and lateral gaps was b 3 mm. Next, the medial and lateral gaps were measured with the knee at
90° flexion. The medial and lateral flexion gaps were evaluated using navigation when the femoral component was rotated within
the range of external rotation from 0 to five degrees, with the knee at 90° flexion. Rotation of the femoral component was deter-
mined such that the difference in the medial and lateral flexion gaps was b3 mm.
After anterior and posterior bone resection of the distal femur, a trial implant was inserted and the alignment and soft tissue
balance checked. A tension device was used to evaluate the gap size and stability during surgery. Both medial and lateral gaps
were measured using navigation after applying tension, and only cases with a difference of b 3 mm were accepted. The thickness
of the insert was chosen such that on navigation, when the knee was extended after inserting a trial insert, the sagittal alignment
remained within the flexion contracture from 0 to three degrees. A thicker insert was chosen if genu recurvatum was confirmed,
and a thinner insert was chosen when flexion contracture was N3°. Implantation was executed in one stage in the tibia, femur,
and patella. Coronal and sagittal alignment was re-measured after implantation. The type of implant used depended on the
following. Initially, the procedure was performed without sacrificing the posterior cruciate ligament (PCL) (i.e., the posterior
cruciate-retaining (CR) type). The PCL was sacrificed if it was over-released during soft tissue balancing or if there was a mis-
match in the flexion-extension gap after soft tissue release (i.e., the posterior cruciate-stabilizing (PS) type). The implant
 S.-S. Seo et al. / The Knee 25 (2018) 167–176 169

types used included the CR-type E-motion (B-Braun, Aesculap, Tuttlingen, Germany) and CR-type and PS-type Triathlon
(Stryker, Mahwah, NJ, USA). The distal thickness of the femoral component of each implant was nine and eight millimeters,
respectively.
The mechanical femorotibial angle, joint line height, and patella height were used for radiologic assessment. The mechanical
femorotibial angle was defined as the angle formed by the mechanical axes of the femur and tibia on a standing whole-leg
anteroposterior radiograph. Varus alignment was expressed as a positive number and valgus alignment as a negative number.
Joint line height was defined as the distance between the line connecting the most distal ends of the medial and lateral femoral
condyles, and a line parallel to the fibular head apex (Figure 1A and B) [13]. Change in joint line height was defined as the
pre-operative joint line height minus the postoperative joint line height. A positive number represented a decrease and a negative
number an increase of the joint line height. Insall-Salvati index was used to evaluate patellar height.
Radiologic parameters were measured twice by two orthopedic surgeons at a one-week interval. Measurement of radiologic
parameters was performed using Picture Archiving Communication System software (PACS; Infinitt, Seoul, Korea) with a minimal
detectable angular change of 0.1°. The intra-class correlation coefficient (ICC) was used for the test–test reliability of the radiologic
parameter measurements. The ICC for intra-observer reliability was 0.87–0.91 for all measurements and the inter-observer reli-
ability was 0.77–0.85.
Clinical assessment was based on the degree of hyperextension at full extension of the knee and patient-reported outcome
measures (PROMs). The degree of hyperextension measured by physical examination in the outpatient clinic may be inaccurate
for reasons such as measurement error and each patient's pain perception. Therefore, the current study measured the degree
of hyperextension under spinal or general anesthesia using navigation, and only included cases that showed hyperextension
≥5° on navigation in the recurvatum group. The value that was found on navigation, when the surgeon lifted the patient's heel
after locating the patella at the center, was used for the degree of hyperextension (Figure 2A and B). Hyperextension ≥10° was
defined as severe and hyperextension between five and 10° was defined as mild. However, as the degree of hyperextension
could not be measured under anesthesia using navigation before surgery or during the postoperative follow-up period, hyperex-
tension was measured at these points by physical examination using long-arm goniometer. After locating the patella at the center,
light pressure was put on the knee to achieve full extension. Then, the degree of hyperextension was checked by lifting the heel,
and the measurement recorded. The degree of hyperextension or flexion contracture measured by navigation or goniometer

Figure 1. Measurement of joint line height in preoperative (A) and postoperative (B) X-rays.
170 S.-S. Seo et al. / The Knee 25 (2018) 167–176

Figure 2. Evaluation of sagittal alignment using navigation: (A) position for measuring sagittal alignment under anesthesia and (B) value of sagittal alignment that
appears on navigation.

was defined as sagittal alignment, with hyperextension expressed as a negative number and flexion contracture as a positive
number. PROMs included the American Knee Society knee and function scores and the Western Ontario and McMaster Universities
Osteoarthritis Index score [14,15]. The thicknesses of the medial and lateral condyles of the resected distal femur and
proximal tibia were also measured. The thickness of the thickest part of the resected bone was measured in mm using flat-
tipped calipers.
SPSS Version 23.0 (IBM Corp., Armonk, NY, USA) was used for statistical analyses. The significance level was set at P b 0.05.
Radiologic and clinical results were compared between the control and recurvatum groups. Matching variables were compared
between the recurvatum and control groups both before and after the matching, and appropriateness of the matching was
checked using standardized mean difference. Independent t-test was used for the comparison of continuous variables and Fisher's
exact test was used for comparison of categorical variables.
Subgroup analyses were performed in the recurvatum group, with patients classified according to the degree of pre-operative
hyperextension on navigation (five to 10° vs. ≥ 10°), type of implant (PS vs. CR prosthesis), and type of polyethylene insert
(mobile vs. fixed bearing). The clinical results (sagittal alignment, PROMs) of each subgroup were compared (Mann–Whitney
U test).
The thickness of the resected bone was measured in the femoral and tibial medial and lateral condyles. However, the thickest
part of the resected bone could be the medial or lateral condyle, depending on factors such as alignment, location of arthritic
compartment (medial or lateral), extent of articular cartilage wear, and femoral bowing. The thickest part of the resected bone
was compared, regardless of anatomical location (Mann–Whitney U test).
 S.-S. Seo et al. / The Knee 25 (2018) 167–176 171

Table 1
Overall demographic data.

Characteristic Control group Recurvatum group P Standardized mean difference

a
Before PS matching
Sex, male:female 42:368 1:54 0.045 −0.625
Age, years 70.3 ± 6.5 67.8 ± 5.6 0.007 −0.442
(41 to 93) (58 to 78)
BMI, kg/m2 24.7 ± 3.4 24.1 ± 3.4 0.227 −0.176
(15.1 to 33.7) (17.0 to 33.0)
Pre-operative mFTAb 9.7 ± 6.0 8.0 ± 7.4 0.056 −0.230
(−5.5 to 25.7) (−10.3 to 25.5)

After PS matching
Sex, male:female 2:108 1:54 1.0 0.000
Age, years 67.7 ± 6.9 67.8 ± 5.6 0.933 0.016
(55 to 93) (58 to 78)
BMI, kg/m2 24.4 ± 3.5 24.1 ± 3.4 0.674 −0.071
(15.1 to 33.0) (17.0 to 33.0)
b
Pre-operative mFTA 8.0 ± 5.4 8.0 ± 7.4 0.987 0.002
(−5.5 to 25.4) (−10.3 to 25.5)

Data are presented as means ± standard deviation (range) unless otherwise indicated.
BMI, body mass index; PS, propensity score; mFTA, mechanical femorotibial angle.
a
Propensity score matching.
b
Mechanical femorotibial angle (measured on the standing weight-bearing whole leg anteroposterior radiographs).

3. Results

Of 510 cases that received navigation-assisted TKA, 465 knees underwent surgery for degenerative osteoarthritis. Among the
research subjects, 19 cases with a follow-up period b2 years, 13 cases with inflammatory arthritis, eight cases with osteonecrosis,
and five cases with a history of previous surgeries due to fracture or ligament injury were excluded from the sample. Hyperex-
tension ≥5° on navigation was observed in 55 of 465 knees (11.8%). Patient demographic data are summarized in Table 1. All
patients in the recurvatum group complained of osteoarthritis-related symptoms pre-operatively such as knee joint pain
and swelling, decreased range of motion and mechanical symptoms, and no patient complained of genu recurvatum-related
symptoms (back knee gait).
A total of 110 TKAs without pre-operative genu recurvatum were matched with 55 cases of TKAs with genu recurvatum and
constituted the control group. Before the matching, a significant difference of sex and age was observed between the control
group and recurvatum groups (P = 0.045 and P = 0.007, respectively). After the matching, all matching variables including
sex and age were successfully matched. The absolute value of the standardized mean difference of each matching variable was
b0.1.
Average follow-up period of the control group and recurvatum groups was 32.8 months ±9.6 months (24–72 months) and
33.3 ± 10.5 months (24–60 months), respectively. The implants used in this study were Triathlon and E-motion. In the
recurvatum group, Triathlon (Stryker, Mahwah, NJ, USA) was used in 23 cases and E-motion (B-Braun, Aesculap, Tuttlingen,
Germany) in 32 cases. For the Triathlon implants, a CR design was used in six cases and a PS design in 17 cases. A fixed-type
insert was used in all Triathlon cases. For the E-motion implants, a CR design and a mobile-type insert were used in all cases.
Table 2 describes the distribution of mechanical femorotibial angle and sagittal alignment (degree of hyperextension) in knees
with genu recurvatum. In total, 41 of the 55 knees (74.5%) showed hyperextension between five and 10° and 47 knees (85.4%)
showed varus alignment.
Table 3 summarizes the radiologic and intraoperative results of the control and recurvatum groups. There were no significant
differences in radiologic results between the two groups, with the exception of change in the joint line height before and after
surgery, which was 0.6 ± 3.8 mm in the control group and 2.6 ± 2.7 mm in the recurvatum group (P = 0.001). A comparison
of intraoperative results showed that the pre-operative sagittal alignment in the control group was 3.5 ± 4.4° and that in the
recurvatum group −8.3 ± 3.8° (P b 0.001). Postoperative sagittal alignment was 0.3 ± 1.9° in the control group and −0.1 ± 2.6°
in the recurvatum group (P = 0.332). Comparisons of resected bone thickness showed that the average thickness of the resected

Table 2
Distribution of knees with genu recurvatum on navigation according to the mechanical femorotibial angle and degree of hyperextension.

Hyperextensiona Mechanical femorotibial angle

Valgus Neutral (−3° to 3°) Varus

≤5° to b10° 3 2 36 41 (74.5%)

≥10° 2 1 11 14 (25.5%)
Total 5 (9.1%) 3 (5.5%) 47 (85.4%) 55 (100%)
a
Sagittal alignment measured on navigation (degree of hyperextension).
172 S.-S. Seo et al. / The Knee 25 (2018) 167–176

Table 3
Comparison of radiologic and intraoperative results between control group and recurvatum group.

Control group Recurvatum group P

(n = 110) (n = 55)

Radiologic
Coronal alignmenta (mFTA), ° Pre-operative 8.0 ± 5.4 (−5.5 to 25.4) 8.0 ± 7.4 (−10.3 to 25.5) 0.056
Postoperative 1.1 ± 2.4 (−4.0 to 8.4) 1.1 ± 3.0 (−5.8 to 6.7) 0.956
b
Patellar height Pre-operative 1.1 ± 0.2 (0.7 to 2.0) 1.0 ± 0.1 (0.8 to 1.3) 0.169
Postoperative 1.0 ± 0.3 (0.6 to 2.2) 1.1 ± 0.1 (0.8 to 1.4) 0.104
Δ Joint line height, mm 0.6 ± 3.8 (−9.0 to 11.0) 2.6 ± 2.7 (−4.0 to 7.0) 0.001

Intraoperative (navigation)
Coronal alignment (mFTA), ° Pre-operative 5.9 ± 5.0 (−4 to 22) 5.0 ± 5.1 (−9 to 22) 0.252
Postoperative 0.5 ± 1.5 (−2 to 4) 0.1 ± 1.2 (−3 to 4) 0.107
Sagittal alignmentc Pre-operative 3.5 ± 4.4 (−4 to 16) −8.3 ± 3.8 (−26 to −5) b0.001
Postoperative 0.3 ± 1.9 (−3 to 5) −0.1 to 2.6 (−5 to 5) 0.332
Thickness of resected bone, mmd Femur 8.4 ± 1.4 (3.5 to 12.5) 7.6 ± 1.6 (5.0 to 12.0) 0.001
Tibia 7.7 ± 1.3 (3.5 to 10.5) 7.3 ± 1.5 (3.0 to 10.0) 0.103
Thickness of insert, mm 10.8 ± 1.5 (9.0 to 14.0) 12.5 ± 2.3 (9.0 to 16.0) b0.001
Implant E-motion CR: 71 E-motion CR: 32
Triathlon CR: 16 Triathlon CR: 6
Triathlon PS: 23 Triathlon PS: 17

Data are presented as means ± standard deviation (range).

mFTA, mechanical femorotibial angle (varus, positive number; valgus, negative number); CR, posterior cruciate-retaining type; PS, posterior cruciate-stabilizing type.
a
Alignment measured on the standing weight-bearing whole leg anteroposterior radiographs.
b
Insall-Salvati index.
c
Alignment at full extension on navigation (hyperextension, negative number; flexion contracture, positive number).
d
Thickness of the thickest part of the medial and lateral condyle of the resected proximal tibia or distal femur.

femur in the recurvatum group (7.6 ± 1.6 mm) was statistically significantly smaller than that in the control group (8.4 ± 1.4 mm)
(P b 0.001). The thickness of the insert in the recurvatum group (12.5 ± 2.3 mm) was greater than that in the control group
(10.8 ± 1.5 mm, P b 0.001).
Comparison of clinical results between the control and recurvatum groups showed that the sagittal alignment at full knee
extension at final follow-up was 1.3 ± 3.4° in the control group and −0.1 ± 0.7° in the recurvatum group. The difference was
significant (P = 0.003) (Table 4). However, there were no significant differences in PROMs between the control and recurvatum
groups (Figure 3). In subgroup analysis by degree of pre-operative hyperextension as well as implant and insert type, there were
no significant differences in PROMs between the severe hyperextension group (≥10°) and the mild hyperextension group (five to
10°), between the CR TKA group and the PS TKA group, and between the fixed-bearing group and mobile-bearing group (P N 0.05,
all; Figure 4A–C). Patients in the recurvatum group showed no complications after TKA, such as recurrence of recurvatum, insta-
bility, loosening, or osteolysis.

4. Discussion

The principle finding of this study was that the incidence genu recurvatum (hyperextension ≥5°) was not uncommon when
evaluated using navigation. Moreover, approximately 85% of the genu recurvatum cases showed varus alignment. Although the
recurvatum group had a larger change of joint line height after TKA than the control group, there were no significant differences
in clinical results at the minimum two-year follow-up. The degree of pre-operative hyperextension and type of implant and insert
did not affect clinical outcomes.
Genu recurvatum can develop in conditions other than neuromuscular disorders, including genu valgum and rheumatoid
arthritis [1–3,5,7]. The incidence of genu recurvatum was previously reported as b1% [6,7]. However, among knees that received

Table 4
Comparison of clinical results between control group and recurvatum group.

Hyperextensionb Sagittal alignmenta, °

Pre-operative Final follow-up

Control group (n = 110) 7.1 ± 7.8 (−3 to 30) 1.3 ± 3.4 (0 to 15)
Recurvatum group (n = 55) −3.2 ± 5.2 (−22 to 6) −0.1 ± 0.7 (−5 to 0)
Mild (≤5° to b10°) (n = 41) −1.7 ± 3.3 (−9 to 6) −0.1 ± 0.8 (−5 to 0)
Severe (≥10°) (n = 14) −7.6 ± 7.0 (−22 to 0) 0.0 ± 0.0 (0 to 0)
Pc b0.001 0.003
a
Sagittal alignment measured on physical examination (hyperextension, negative number; flexion contracture, positive number).
b
Sagittal alignment measured on navigation.
c
Control vs. recurvatum group.
 S.-S. Seo et al. / The Knee 25 (2018) 167–176
Figure 3. Comparison of clinical results between the recurvatum group and the control group (patient-reported outcomes). KSKS, Knee Society knee score; KSFS, Knee Society function score; WOMAC, Western Ontario and
McMaster Universities Osteoarthritis Index score.

173
 174
S.-S. Seo et al. / The Knee 25 (2018) 167–176
Figure 4. Subgroup analysis in the recurvatum group by degree of preoperative hyperextension and type of implant and insert (A) Comparison between mild hyperextension group and severe hyperextension group (B)
Comparison between posterior cruciate-substituting design and posterior cruciate-retaining design (C) Comparison between mobile bearing and fixed bearing. KSKS, Knee Society knee score; KSFS, Knee Society function
score; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index score.
 S.-S. Seo et al. / The Knee 25 (2018) 167–176 175

navigation-assisted TKA for degenerative osteoarthritis at the current institution, the incidence of cases with hyperextension ≥5°
on navigation was 11.8%. This result is inconsistent with previously reported results. It is difficult to accurately evaluate the cause
for differences in the rate of incidence, due to the heterogeneity of each study. However, the marked differences between the in-
cidence rate in the current study and those previously reported may partially be attributed to anesthesia, diagnosis method, and
racial differences. Although Mullaji et al. studied Asian subjects and evaluated genu recurvatum by using navigation, as in the cur-
rent study, the incidence of genu recurvatum was 3.9% in their study [9]. Although this number was higher than those in other
previous studies, it considerably differed from the current results. The difference in the incidence of genu recurvatum between
Mullaji et al. and the current study could be attributed to the difference of definition of genu recurvatum [9]. Mullaji et al. defined
genu recurvatum as a case with hyperextension N 5°, whereas the current study included hyperextension ≥5° in the recurvatum
group [9]. There were several other potential reasons such as minute differences of the method of approach, difference of navi-
gation tool, and racial difference.
However, even with differences of genu recurvatum diagnosis methods or racial differences, the incidence of genu recurvatum
in the current study was markedly different from other studies. Therefore, it is unlikely that the incidence of genu recurvatum in
osteoarthritic knees is as low as previously reported (b 1%). Moreover, some studies reported a relationship between genu
recurvatum and genu valgum [4,5]. However, interestingly, 85.4% of the cases with genu recurvatum in the current study had
varus alignment. It is thought that the incidence rate of genu recurvatum and the accompanying alignment should be discussed
more. Noyes et al. classified knees with varus alignment into primary-, double-, and triple-varus knee syndromes, with consider-
ation of tibiofemoral osseous and geometric alignment, abnormal knee motion limits, abnormal knee joint positions, and specific
ligament defects [16]. Triple-varus knee syndrome has three characteristics: varus tibiofemoral osseous alignment, marked sepa-
ration of the lateral tibiofemoral compartment, and increased external tibial rotation and hyperextension with an abnormal varus
recurvatum position. It has also been reported that chronic excessive tensile forces in the posterolateral ligament structures can
cause varus recurvatum [17]. The proportion of people with varus alignment is relatively high among Asians, and coronal bowing
of femur or tibia is reported to be prevalent in Asian populations [18–20]. Furthermore, as progress to severe varus osteoarthritic
knee requiring TKA takes a long time, the incidence of knees with varus recurvatum is unlikely to be as low as previously report-
ed. Further studies on the incidence of genu recurvatum using a larger number of cases and including people from diverse races
are required.
In the current comparison of radiologic and intraoperative results between the control and recurvatum groups, significant dif-
ferences in the change in the joint line height and thickness of resected distal femur were found. This may be attributed to the
fact that the senior author attempted to obtain joint stability with lesser resection of distal femur when implementing TKA on
patients with genu recurvatum. A change in the joint line can alter knee joint biomechanics, and may also affect clinical results.
However, PROMs obtained at the final follow-up did not show significant differences between the control and recurvatum groups.
Although the average joint line height changed by 2.6 mm, it was thought that the size of the change was not sufficiently large to
affect clinical results. Unlike the PROM findings, a difference in sagittal alignment in full knee extension was observed between
the control and recurvatum groups at the final follow-up. Intraoperative results showed no significant difference in postoperative
sagittal alignment measured by navigation after implantation between the two groups. However, there was a significant
difference in sagittal alignment measured by physical examination at the final follow-up. The control group had an average flexion
contracture of approximately one degree, whereas all cases were capable of full extension (0°) in the recurvatum group. As the
difference was only one degree, it is difficult to determine a clinical meaning. Nevertheless, the fact that all cases in the
recurvatum group showed no flexion contracture deserves attention. The average follow-up period in the current study was
relatively short (around 33 months); therefore, mid- or long-term follow-up will be required in future studies to more accurately
evaluate clinical results.
In general, even with severe varus or valgus deformity, clinical results after TKA have been reported to be satisfactory, regard-
less of the type of implant and insert [21–24]. However, no previous analysis has been conducted on these factors in patients
with genu recurvatum who underwent TKA. The current study performed subgroup analyses to determine if the degree of
pre-operative hyperextension and type of implant and insert affected the clinical results. However, no significant differences in
clinical results were observed. If joint stability can be obtained without a large change of joint line height when performing
TKA, it is likely that clinical results will be satisfactory, regardless of the degree of pre-operative hyperextension and type of
implant and insert.
Lastly, the proportion of patients complaining of genu recurvatum-related symptoms in the group of patients assigned to the
recurvatum group was estimated, and interestingly, it was found that no patient complained of genu recurvatum-related
symptoms. All patients complained of osteoarthritis-related symptoms such as pain, and this observation can be attributed to
the restriction of study subjects to those with severe osteoarthritic knees that required TKA. Furthermore, as patients with
neuromuscular disease or rheumatoid arthritis were excluded from this study, the results of the present study may differ from
those of previous studies that included patients with genu recurvatum.
This study had several limitations. First, it was a retrospective study with a short follow-up period. Second, as the number of
cases included in this study was small, statistical power could not be determined. Third, among the patients that were included in
this study, one case was male and the rest were female. Further studies will be required to clarify whether the incidence of genu
recurvatum or clinical results differ by gender. Fourth, this study used one navigation system; the incidence of genu recurvatum
or intraoperative results may differ if a more sensitive navigation system is used. Finally, although sagittal alignment at full knee
extension was measured by a single orthopedic surgeon, measurements in the outpatient clinic (using a goniometer) had a high
possibility of measurement error.
176 S.-S. Seo et al. / The Knee 25 (2018) 167–176

5. Conclusions

Genu recurvatum was not uncommon among patients who underwent primary TKA. Satisfactory short-term clinical and radio-
logic results were obtained using a smaller distal femoral resection and thicker insert.

Conflict of interest

The authors declare that they have no competing interests.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

References

[1] Giori NJ, Lewallen DG. Total knee arthroplasty in limbs affected by poliomyelitis. J Bone Joint Surg Am 2002;84-A:1157–61.
[2] Patterson BM, Insall JN. Surgical management of gonarthrosis in patients with poliomyelitis. J Arthroplasty 1992;7(Suppl):419–26.
[3] Tigani D, Fosco M, Amendola L, Boriani L. Total knee arthroplasty in patients with poliomyelitis. Knee 2009;16:501–6.
[4] Scott WN, Insall JN. Knee Society (U.S.). Insall & Scott surgery of the knee. Philadelphia, PA Rosemont, IL: Elsevier/Churchill Livingstone in association with the
Knee Society; 2012.
[5] Tew M, Forster IW. Effect of knee replacement on flexion deformity. J Bone Joint Surg Br 1987;69:395–9.
[6] Krackow KA, Weiss AP. Recurvatum deformity complicating performance of total knee arthroplasty. A brief note. J Bone Joint Surg Am 1990;72:268–71.
[7] Meding JB, Keating EM, Ritter MA, Faris PM, Berend ME. Total knee replacement in patients with genu recurvatum. Clin Orthop Relat Res 2001:244–9.
[8] Meding JB, Keating EM, Ritter MA, Faris PM, Berend ME. Genu recurvatum in total knee replacement. Clin Orthop Relat Res 2003:64–7.
[9] Mullaji A, Lingaraju AP, Shetty GM. Computer-assisted total knee replacement in patients with arthritis and a recurvatum deformity. J Bone Joint Surg Br 2012;94:
642–7.
[10] Bae DK, Song SJ. Computer assisted navigation in knee arthroplasty. Clin Orthop Surg 2011;3:259–67.
[11] Rhee SJ, Seo CH, Suh JT. Navigation-assisted total knee arthroplasty for patients with extra-articular deformity. Knee Surg Relat Res 2013;25:194–201.
[12] Whiteside LA, Mihalko WM. Surgical procedure for flexion contracture and recurvatum in total knee arthroplasty. Clin Orthop Relat Res 2002:189–95.
[13] Lee KJ, Moon JY, Song EK, Lim HA, Seon JK. Minimum two-year results of revision total knee arthroplasty following infectious or non-infectious causes. Knee Surg
Relat Res 2012;24:227–34.
[14] Insall JN, Dorr LD, Scott RD, Scott WN. Rationale of the knee society clinical rating system. Clin Orthop Relat Res 1989:13–4.
[15] Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Validation study of WOMAC: a health status instrument for measuring clinically important patient
relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee. J Rheumatol 1988;15:1833–40.
[16] Noyes FR, Barber-Westin SD, Hewett TE. High tibial osteotomy and ligament reconstruction for varus angulated anterior cruciate ligament-deficient knees. Am J
Sports Med 2000;28:282–96.
[17] Hughston JC, Jacobson KE. Chronic posterolateral rotatory instability of the knee. J Bone Joint Surg Am 1985;67:351–9.
[18] Lasam MP, Lee KJ, Chang CB, Kang YG, Kim TK. Femoral lateral bowing and varus condylar orientation are prevalent and affect axial alignment of TKA in Koreans.
Clin Orthop Relat Res 2013;471:1472–83.
[19] Song MH, Yoo SH, Kang SW, Kim YJ, Park GT, Pyeun YS. Coronal alignment of the lower limb and the incidence of constitutional varus knee in Korean females.
Knee Surg Relat Res 2015;27:49–55.
[20] Yau WP, Chiu KY, Tang WM, Ng TP. Coronal bowing of the femur and tibia in Chinese: its incidence and effects on total knee arthroplasty planning. J Orthop Surg
(Hong Kong) 2007;15:32–6.
[21] 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-A:2671–6.
[22] Lee BS, Lee SJ, Kim JM, Lee DH, Cha EJ, Bin SI. No impact of severe varus deformity on clinical outcome after posterior stabilized total knee arthroplasty. Knee Surg
Sports Traumatol Arthrosc 2011;19:960–6.
[23] Miyasaka KC, Ranawat CS, Mullaji A. 10- to 20-year follow-up of total knee arthroplasty for valgus deformities. Clin Orthop Relat Res 1997:29–37.
[24] Ritter MA, Faris GW, Faris PM, Davis KE. Total knee arthroplasty in patients with angular varus or valgus deformities of N or = 20 degrees. J Arthroplasty 2004;19:
862–6.