Keywords
unicompartmental knee arthroplasty - unicompartmental knee replacement - osteoarthritis
- knee - epidemiology
Introduction
Unicompartmental knee arthroplasty (UKA) was first introduced in the 1970s for patients
with severe tibiofemoral unicompartmental osteoarthritis.[1]
[2]
It is a cost-effective treatment in patients with osteoarthritis of the medial or
lateral tibiofemoral compartment, consistently achieving a high rate of patient-reported
outcomes in the good/excellent category.[3] In addition, a higher rate of return to sport than that of total knee arthroplasty
(TKA) is reported.[4] The United Kingdom arthroplasty register shows that between 8% and 15% of the patients
that need knee arthroplasty can be successfully treated with UKA.[5]
Our institution is a university health centre that belongs to the Chilean private
care network. Despite this, around half of our patients are covered by public health
insurance (the Chilean National Health Found, Fondo Nacional de Salud, FONASA, in
Spanish). In Chile, UKA has not become a widely-used technique such as TKA. Unfortunately,
no objective data is available, as the national coding system does not differentiate
between UKA and TKA. However, our centre is one with the highest volume of UKA procedures
in Chile. Different reasons are given for this lower amount of UKA procedures performed,
such as: the lower level of access to the learning the technique, its slower learning
curve, and its greater difficulty compared to TKA.[6] However, UKA has shown good results in international series both in young and geriatric
patients,[7]
[8] so it is of great interest to further spread this surgical procedure so that it
is incorporated into the arsenal of treatment for patients with knee osteoarthritis
in our country.[9]
The purpose of the present study is to describe the patient-reported functional outcomes
of a historical cohort of patients undergoing UKA in our centre and compare the results
with those of international reports. We hypothesized that the outcomes of our cohort
are comparable to those of international series regardless of follow-up, age, and
the side of the prosthesis.
Methods
A historical cohort study was designed and carried out at Hospital Clínico Universidad
de Chile. Patients who undergo knee arthroplasty at our centre are registered according
to the Chilean national code system under number 2104153. All patients that underwent
UKA between 2003 and 2019 were included.
The selection criteria to submit patients to UKA are pain located in the compromised
compartment, body mass index under 34 kg/m2, and a reducible varus or valgus alignment. Patients with diffuse knee pain, moderate
or severe irreducible coronal malalignment, or chondropathy grade ≥ 3 or Kellgren
y Lawrence ≥ III in the contralateral compartment are not candidates for UKA. If the
patient does not complain of anterior knee pain, patellofemoral chondropathy is not
an exclusion criterion.[10] Also, UKA is contraindicated for patients with neuromuscular impairment or knee
instability – such as anterior cruciate ligament tear. Meanwhile, all patients who
underwent TKA or patellofemoral arthroplasty were excluded. [Figure 1] shows the flow chart of the selection of patients.
Fig. 1 Flow chart of the selection of patients.
All clinical records of patients who underwent knee arthroplasty between January 2003
and May 2019 were revised. The data extracted included the birth date, date of surgery,
gender, type of UKA, and side in which it was performed. Also, successive medical
controls were reviewed for the following complications: reoperations, superficial
infection, deep infection, arthrofibrosis, need for mobilization under anesthesia,
tibial subsidence, and referral to the chronic pain unit. Before surgery, all patients
undergo a weight-bearing anteroposterior (AP) radiograph, a lateral knee radiograph,
a Schuss radiograph, an AP full-length lower limb radiograph, and a magnetic resonance
imaging (MRI) scan ([figure 2]).
Fig. 2 Imaging studies routinely performed for the selection of patients. weight-bearing
anteroposterior (AP) magnetic resonance imaging scan (A), showing severe osteoarthritis of the lateral compartment (B), AP full-length lower limb radiograph (C), Schuzz view (D). The postoperative AP knee radiograph is also shown (E).
To avoid selection biases, the postoperative radiograph was revised in all cases to
confirm that the patient had undergone UKA. Furthermore, the UKA procedure was corroborated
with the patients who were successfully contacted for the phone interview. The clinical
records of the patients were extensively revised to reduce the risk of bias inherent
to a retrospective study.
All patients underwent surgery by the anterior knee approach. A medial parapatellar
arthrotomy was performed in cases of medial UKA; meanwhile, a lateral parapatellar
arthrotomy was performed in cases of lateral UKA. Only one fixed bearing model of
prosthesis was used: ZUK (Zimmer-Biomet, Warsaw, IN, US) and Journey (Smith & Nephew,
London, UK). First, the tibial cut is performed with the aim of presering as much
of the tibia as possible, and to reproduce the native slope. Then, the femoral cut
is performed to restore the joint line and to select the appropriate size of the femoral
component ([figure 3]).
Fig. 3 Lateral UKA was performed through lateral parapatellar arthrotomy. The first tibial
cut (A), femoral cut (B), and the UKA (C) are shown.
In June 2020, an independent evaluator (DP) contacted all patients by phone. A validated
transcultural adaptation of the Western Ontario and McMaster Universities Arthritis
Index (WOMAC) was applied by phone.[11]
[12] The WOMAC is an instrument to assess the functional capacity of patients with knee
and hip osteoarthritis that was further validated to evaluate knee arthroplasty.[13] It is considered one of the best self-reported instruments in osteoarthritis patients
due to its psychometric properties.[14] It consists of 3 dimensions: 5 items assessing pain (score: 0 to 20), 2 items for
stiffness (score: 0 to 8), and 17 items for physical function in daily activities
(score: 0 to 68). The score is higher as the patients feel unwell.[12] The scale has been validated to be self-administrated and administrated by phone.[15] Although it is not the primary function of the scale, the scores of the 3 dimensions
can be added and reported in standardized way from 0 to 100.[14]
Our local ethics committee approved the present study. All patients successfully contacted
by phone gave verbal consent to participate in the survey using the telephone form
approved by the ethics committee.
Data were summarized as medians, ranges, and interquartile ranges (IQRs, 25th to 75th percentiles). The proportion of patients successfully contacted by phone was compared
to the total cohort in terms of age (patients older than 70 years of age), gender,
the percentage of medial UKAs, the median follow-up, and the proportion of patients
with more than 5 years of follow-up. The WOMAC score was compared regarding the type
of UKA (medial or lateral), age (older or younger than 70 years of age) and follow-up
(longer or shorter than 5 years) using the non-parametric median test.
A significance of 5% was adopted, and a 95% confidence interval was estimated. The
data was processed using the Stata (StataCorp LLC, College Station, TX, US) software,
version 15.
Results
A total of 78 patients, comprising 94 UKA procedures, were included; 16 patients underwent
bilateral UKA (20.5%). The median age was 64 years (range: 43 to 85 years; IQR: 58
to 71 years), and 60 patients (77%) were women. Of the total amount of UKA procedures,
72 were medial (76.6%), and 22 (23.4%) were lateral ([Table1]). The first UKA was carried out in 2003, a year in which only 1 (1.1%) procedure
was performed. In 2009 and 2013, only 1 (1.1%) UKA was performed as well. The mode
was found in 2016, in which a total of 14 (14.9%) UKAs were performed.
Table 1
|
Total
|
Follow-up: June 30, 2020
|
|
Number of patients
|
78
|
60 (76.9%)
|
|
Number of UKAs
|
94
|
72 (76.7%)
|
|
Median age*
|
64 (43 to 85)
|
62 (43 to 85)
|
|
Older than 70 years of age*
|
25 (26.6%)
|
15 (25.0%)
|
|
Men*
|
19 (20.2%)
|
11 (21.6%)
|
|
Medial UKA**
|
72 (76.6%)
|
42 (70.0%)
|
|
Lateral UKA
|
22 (23.4%)
|
20 (30%)
|
|
Median follow-up**
|
5.0 (1.1 to 16.7)
|
4.7 (1.1 to 16.1)
|
|
Follow-up of 5 years or longer**
|
47 (50%)
|
35 (48.6%)
|
|
Revision to TKA**
|
1 (1.1%)
|
0
|
|
Tibial subsidence
|
1 (1.1%)
|
0
|
|
Needed second surgery**
|
2 (2.1%)
|
2 (2.8%)
|
|
Superficial infection**
|
2 (2.1%)
|
2 (2.8%)
|
|
Deep infection**
|
0
|
0
|
In total, 3 (3.3%) patients required another surgery after the UKA: 1 patient required
revision to TKA, 1 patient underwent lateral meniscectomy, and the last patient needed
a pes anserine release. Revision TKA for intractable diffuse knee pain was performed
after three years in another institution. Among the surgical complications, 2 (2.2%)
patients had a superficial infection, and none of the patients had deep infection,
required mobilization under anaesthesia, or needed a referral to the chronic pain
unit. In total, 1 (1.1%) patient aged 86 years old suffered tibial subsidence that
did not require revision. After three months of conservative treatment, the bone was
healed, and the patient reached full range of motion ([figure 4]). He died eight years after surgery due to cardiovascular disease.
Fig. 4 Tibial Subsidence. Preoperative radiograph (A), immediate AP radiograph (B). After three weeks, an inclination is observed on the AP knee radiograph (C). Bone healing can be observed after three months, without evidence of further subsidence
or loosening (D). Knee range of motion at four years of follow-up is shown (E,F).
A total of 60 patients (76.9%), corresponding to 72 UKA procedures, were successfully
contacted by phone for the final follow-up. At the last follow-up, 3 (3.3%) patients
were dead, and 15 (19.2%) could not be located. The epidemiological characteristics
are shown in [table 1]. The median follow-up was of 4.4 years (range: 1.1 to 16.2 years; IQR: 2.5 to 10.1
years). The median pain score on the WOMAC was 1 (range: 0 to 12; IQR: 0 to 3), the
stiffness score was 0 (range: 0 to 4; IQR: 0 to 0), and the physical function score
was 2 (range: 0 to 29; IQR: 0 to 10). The median total WOMAC score was 4 (range: 0
to 44; IQR: 0 to 13) ([figure 5]).
Fig. 5 The WOMAC index and its subdimension distribution.
Fig. 6 Distribution of WOMAC scores among patients younger or older than 70 years of age
at the time of the surgery.
The patients submitted to lateral UKA (n = 20, 32%) achieved better total scores on
the WOMAC than those submitted to medial UKA (n = 42, 68%). There were only statistically
significant differences regarding the total WOMAC score (median test; p = 0.04329) and the WOMAC pain score (median test; p = 0.0160) ([table 2]).
Table 2
|
WOMAC dimension
|
Medial UKA
|
Lateral UKA
|
p-value*
|
|
Pain
|
1 (0 to 12)
|
0 (0 to 4)
|
0.0160
|
|
Stiffness
|
0 (0 to4)
|
0 (0 to 2)
|
0.0640
|
|
Physical function
|
5 (0 to 29)
|
0 (0 to 17)
|
0.0566
|
|
Total
|
6 (0 to 44)
|
0 (0 to 21)
|
0.0432
|
No differences in WOMAC was found between older than 70 years and younger patients
([Figure 6]). Patients with a follow-up longer than 5 years had a median total WOMAC score of
8 (range: 0 to 26; IQR: 1 to 13). Meanwhile, patients with follow-up shorter than
5 years have a median total WOMAC score of 1 (range: 0 to 44; IQR: 0 to 9). This difference
did not achieve statistical significance (median test; p = 0.1329) ([figure 7]).
Fig. 7 Distribution of WOMAC scores among patients with follow-up longer or shorter than
five years.
Discussion
The present study shows that patients undergoing UKA had a high proportion of good/excellent
results on the WOMAC index. No differences were found regarding the WOMAC score in
terms of age (older or younger than 70 years) and follow-up time (longer or shorter
than 5 years). Only 1 (1.1%) revision to TKA was found in the present study, probably
because of the number of patients and the broad follow-up time. The median survival
time reported[16] using revision to TKA as failure is of 94.4% at 5 years, and of 89.1% at 10 years.
The functional outcomes of the present historic cohort were excellent. Two significant
issues are essential to have good results in UKA: patient selection and volume of
procedures.[10] Patient selection is crucial to obtain excellent outcomes, especially in centres
with a low volume of surgeries performed.[17] A novel preoperative scoring system to predict good results was developed by Antoniadis
et al.,[18] which could be helpful to identify the best candidates for UKA.
The coronal alignment of the lower limb has been a hot topic regarding UKA indications.[19] In our approach, the physical exam is mandatory to establish if the malalignment
is reducible or fixed. Fixed malalignment between 5° and 10° in the mechanical alignment
is considered a relative indication. Particular attention should be paid when an intraarticular
malalignment is found, as in UKA procedures, it has consistently shown that good satisfaction
is achieved when the articular height and alignment are restored.[20]
[21] Meanwhile, more than 10° is a contraindication for UKA. Stress radiographs have
been proposed to address the reductibility of the coronal alignment; nevertheless,
not consistent findings have been reported.[22]
[23]
[24]
The volume of surgeries performed is crucial for good outcomes, and the present series
shows that our proportion of patients undergoing knee arthroplasty is just above the
one reported in the United Kingdom registry.[5] Moreover, twelve procedures are reported[25] to be the threshold to achieve a lower rate of aseptic failure, so our average of
six UKAs per year is insufficient. Hence, we think that the good outcomes of the present
study are associated to the fact that the first author (CI) was involved in every
case in the first 14 years revised in the present cohort.
Patients with severe femorotibial unicompartmental knee osteoarthritis have three
surgical options: TKA, UKA, and osteotomies. These three procedures have been used
in different types of patients, but some indications may overlap.[27] Cost-effectiveness studies[3] show that UKA is more efficient in patients older than 60 years of age; meanwhile,
osteotomies are more useful in patients younger than 60 years. Moreover, studies[27]
[28] in patients older than 75 years of age show that UKA had similar functional outcomes
but a lower complication rate than that of TKA. In our cohort, there were no statisticcally
significant differences regarding the WOMAC score between patients older than 70 years
of age and younger patients.
A significant concern in knee arthroplasty design is the ability to reproduce the
natural biomechanics. A finite-element study[29] showed that UKA is more accurate in reproducing the native biomechanics of the knee
than TKA, even if the lower limb alignment is in a mild varus or valgus. These findings
are explained as UKA preserves the cruciate ligaments, being more anatomic than TKA.
This is essential for current patients, who are more active, demand better results,
and expect a better quality of life after joint replacement than decades ago.[30] Many of them even aspire to return to sports, which is consistently more likely
in patients undergoing UKA than TKA.[4]
The rate of lateral UKA has been reported[31] to be 5% to 10% of that of the rate of medial UKA, reaching better functional outcomes
but a lower survival rate. Nevertheless, a recent systematic review[32] showed that the survival rates at 5, 20, and 25 years between medial and lateral
PUR are similar. In the present cohort, there is a significant tendency for better
results in lateral UKA, but this should be interpreted with caution. Patients who
require medial or lateral UKA have a different knee problem, so the trend found shows
that lateral UKA is as good a treatment as medial UKA for unicompartmental osteoarthritis.
These groups of patients had different knee problems, making it difficult to compare.
There is a cultural appreciation that performing a lateral UKA is more complex; however,
this is related to the fact that the volume of lateral surgeries performed is lower
than that of medial surgeries. Nevertheless, many reports[33]
[34] show that good outcomes can be achieved with the lateral procedure, as reported
in the present study.
Robotic UKA has been proposed to increase implant orientation and size, functional
outcomes and survival rates.[35]
[36] Nevertheless, a recent metanalysis[37] shows no difference in the median survival time between robotic-assisted and conventional
surgeries. Moreover, the functional outcomes achieved have not been significantly
better in robotic-assisted UKA either. Nevertheless, this could be explained by the
ceiling effect of the functional-outcome scales, because robotic-assisted UKA has
shown substantially more accuracy in implant orientation.[38]
[39] Robotic-assisted surgery is probably the future in UKA. However, for now, a better
understanding of the appropriate implant orientations, trained surgeons, and better
patient access is needed to improve the results.
A limitation of the present study is that this is not a captive cohort. Therefore,
patients could have been revised to TKA at another centre. But we achieved a successful
follow up of 76.9% of the patients, reaching a good proportion of the historical cohort.
Also, the only revision to TKA reported was performed at another centre, which notified
the surgeon (CI) when the revision surgery was done, but this case could not be contacted
in 2020.
Conclusion
For patients with femorotibial unicompartmental knee osteoarthritis, UKA is an effective
and reproducible treatment. Age seems not to affect the functional results, and UKA
is an effective treatment in patients older than 70 years of age. These results should
encourage knee surgeons to learn this technique and those responsible for public health
policies to consider UKA for patients with knee osteoarthritis.
