Keywords
arthroplasty, replacement, knee - osteoarthritis, knee - patella
Introduction
Total knee arthroplasty (TKA) is a surgical procedure with proven clinical efficacy[1]
[2]
[3]
[4] resulting in a substantial functional and quality of life improvement in patients
with gonarthrosis.[5]
[6]
The first modern total knee prosthesis (TKP) was developed by a Canadian orthopedist,
Frank Guston, in 1960; subsequently, it was improved by John Insall in the 1970s,
and the patellar component was introduced by Townley and Insall in 1972.[7] The introduction of a patellar component in TKA reduced the occurrence of anterior
knee pain, but resulted in new complications, such as component failure, instability,
patellar fracture, extensor tendon rupture, patellar avascular necrosis, and other
soft tissue injuries.[8] Although such complications were attributed to a poor surgical technique and inadequate
implant positioning, the fear of sustaining them hindered the routine adoption of
patellar replacement.[8]
Over the years, surgical techniques and implants have improved, significantly reducing
the occurrence of complications in patellar arthroplasties, and dividing surgeons
regarding their adopted routine procedure.[9] Some authors suggest routine patellar replacement since it improves pain, functional
scores, and patient satisfaction, with a lower rate of reintervention due to pain
persistence.[10] Others defend patellar retention as routine due to the risk of complications and
the potential advantage of patellar bone stock maintenance with similar satisfaction
and functional rates.[11]
[12]
[13] Additionally, some still recommend patellar replacement in selected cases, believing
that its routine performance is not supported by the literature.[14]
The present study aims to functionally evaluate subjects submitted to primary TKA
with patellar joint resurfacing and to compare them with patients in whom the patellar
joint surface was retained.
Materials and Methods
A total of 191 patients who underwent primary TKA from January 2012 to December 2014
for primary gonarthrosis were initially selected from our database to participate
in the study. All of them had at least 5 years of follow-up, received the same implant
(Advance Medial-Pivot Knee System, Microport Orthopaedics, Arlington, TN, USA) and
were operated on by two experienced surgeons from an orthopedic reference hospital.
After medical records analysis, 33 patients were excluded; 7 were deceased, 2 were
unable to participate due to comorbidities not related to TKA, and 24 who were lost
to follow-up. No patient met the exclusion criteria regarding complications with major
functional impairment. Finally, 158 patients were included in the study. Four patients
underwent bilateral TKA in different years: two subjects were submitted to a patellar
arthroplasty in one knee, while the other patella was preserved; one patient had both
patellae retained; and one patient underwent a bilateral arthroplasty. In these cases,
both knees were assessed separately. The final sample consisted of 162 knees, including
81 knees with patellar resurfacing and 81 knees with patellar joint surface retention.
Knees were divided into two groups: those with patellar resurfacing during TKA and
those with patellar joint surface retention. Patellar resurfacing was selectively
indicated in case of moderate to severe patellofemoral arthrosis detected at a gross
assessment during surgery. Only patellae with articular surfaces in good conditions
were retained.
All patients were contacted by telephone by one of the authors (de Campos Júnior L.
R.) and answered the Lequesne[15] and the Western Ontario and McMaster Universities (WOMAC) and its subdivisions questionnaires.[16] During contact, subjects were invited to participate in the study and, if they agreed,
the call was resumed.
Patients who agreed to participate in the study signed an Informed Consent Form (ICF).
The study was previously approved by the Research Ethics Committee of the institution.
Statistical Analysis
Quantitative variables were described as mean and standard deviation or median and
interquartile range values. Categorical variables were described as absolute and relative
frequencies. Mean values were compared using the t-Student test. In case of asymmetry,
the Mann-Whitney test was applied. The Pearson chi-squared test was used to compare
proportions. The level of significance was set at 5% (p < 0.05) and the analyzes were performed with IBM SPSS Statistics for Windows, version
25.0 (IBM Corp., Armonk, NY, USA).
Results
A total of 46 patients were male and 112 were female; the study groups were homogeneous
both in gender distribution and in number of knees ([Table 1]). The ages of the subjects at the time of surgery, questionnaires application, and
the interval between them were also homogeneous ([Table 2]).
Table 1
|
Male
|
Female
|
Total
|
|
Patellar retention
Patellar resurfacing
|
23
24
|
58
57
|
81
81
|
|
47
|
115
|
162
|
Table 2
|
Variables
|
With no patellar resurfacing[*]
|
With patellar resurfacing[*]
|
p-value[**]
|
|
Age at surgery (years old)
|
71.9 ± 9.5
|
71.3 ± 6.3
|
0.614
|
|
Age at data collection (years old)
|
77.9 ± 9.8
|
77.2 ± 6.4
|
0.595
|
|
Follow-up (years)
|
5.9 ± 0.7
|
5.9 ± 0.8
|
0.918
|
The median Lequesne score was 3.5 points for the group with patellar retention and
2.5 points for the group with patellar resurfacing; this difference was not sufficient
to generate statistical significance ([Table 3]).
Table 3
|
Scores
|
With no patellar resurfacing[*]
|
With patellar resurfacing[*]
|
p-value[**]
|
|
median (P25-P75)
|
median (P25-P75)
|
|
Lequesne
WOMAC
|
3.5 (0.5–7)
|
2.5 (0.5–7)
|
0.585
|
|
Pain
|
0 (0–2)
|
0 (0–1)
|
0.036[**]
|
|
Rigidity
|
0 (0–0)
|
0 (0–0)
|
0.796
|
|
Physical Activity
|
7 (1–12.5)
|
5 (0–9)
|
0.190
|
|
Total
|
8 (2–15)
|
5 (0.5–11.5)
|
0.169
|
In the WOMAC score, in all three evaluated areas (pain, stiffness, and difficulty
to perform daily/physical activities) and the global score, the only significant difference
was observed in pain, with a higher score in the group with patellar retention ([Table 3] and [Figure 1]). There was no significant difference in joint stiffness scores between groups ([Table 3]). Although there was no significant difference regarding the difficulty in daily/physical
activities score (p = 0.190) and the global score (p = 0.169), median, 25th percentile and 75th percentile values were slightly higher among patients with patellar retention ([Table 3] and [Figures 2] and [3]). Since scores were expressed as natural numbers, some median values were equal
to zero because most subjects had the same score in a given subdomain.
Fig. 1 Western Ontario and McMaster Universities (WOMAC) score for pain.
Fig. 2 Western Ontario and McMaster Universities (WOMAC) score for difficulty in daily living
activities.
Fig. 3 Global Western Ontario and McMaster Universities (WOMAC) score.
Three patients presented complications during the study. Two subjects had surgical
wound dehiscence, including one submitted to patellar resurfacing and another with
patellar retention. Both underwent surgical debridement and antibiotic therapy with
good outcomes. The third patient had an early periprosthetic infection and underwent
a new surgical procedure for polyethylene change, washing and debridement, in addition
to antibiotic therapy. One patient excluded due to loss at follow-up had a late knee
infection and underwent a limb amputation.
Discussion
Over the years, several studies have evaluated the difference in outcomes between
the performance or not of patellar arthroplasty. Recently, Ha et al.[17] performed the first prospective randomized study in which patients had one knee
submitted to TKA with patellar resurfacing and the other knee had the patella retained
at the same surgical time and using the same implant model. After 5 years of follow-up,
60 patients (120 knees) were reevaluated. The authors found out that knees submitted
to patellar resurfacing had less anterior pain (p < 0.001) and a lower incidence of patellar crepitation (p < 0.001). Although both
knees showed functional improvement, it was significantly higher in those submitted
to patellar resurfacing (p < 0.001). Neither knee presented complications nor required surgical revision. In
a satisfaction assessment, 47% of the patients preferred the knee submitted to patellar
resurfacing, while 46% were indifferent.[17]
Migliorini et al.[18] performed a meta-analysis of 31 articles, totaling 4,132 knees, while Longo et al.[19] carried out another meta-analysis with 35 articles and a total of 5,535 knees. Both
found less anterior knee pain in patients who underwent patellar resurfacing (p = 0.02 and p = 0.00001, respectively); this same group also presented lower revision rates (p <0.0001 and p = 0.00001, respectively). Only the first set of authors observed a statistically
significant functional difference, with better function in patients undergoing patellar
resurfacing (p = 0.009).
Better pain outcomes in patients undergoing patellar resurfacing was a common finding
between our study and the literature; we believe that this is a well-defined outcome.
Although we did not show any significant difference regarding the global WOMAC score
and its subdomain for difficulty in daily living/physical activities, higher median
values, relatively higher 25th and 75th percentiles and a considerably low p-value in the patellar retention group lead us to believe that the difference could
become significant if the study population was larger. This hypothesis seems plausible,
given the significance evidenced by Migliorini et al.[18] and Ha et al.,[17] and the consistent findings of Longo et al.,[19] who believe there is a trend for better function after patellar resurfacing. Further
studies are certainly required for functional comparison between these groups of patients
to generate solid evidence.
The three studies showed lower revision rates in the patellar resurfacing group.[17]
[18]
[19] Migliorini et al.[18] believe that this is related to individual dissatisfaction with pain persistence
after patellar retention, leading surgeons to reoperate and perform a resurfacing
procedure.[18] However, we tried to evaluate only the functional outcomes from two scenarios in
ideal situations, excluding patients with complications leading to great functional
impairment. We believe that these complications and their analysis would affect group
outcomes in a nonrepresentative way. We emphasize that despite this definition, only
one patient who was excluded due to loss at follow-up had a serious complication which
was not caused by patellar resurfacing. We believe that outcomes related to the incidence
of revision must not be considered because there is a greater tendency to indicate
revision in symptomatic patients with retained patellae. There are some questionable
points in the aforementioned meta-analyzes: lack of description of the surgeons' skills
and of the implant model in most studies, heterogeneity among studies,[19] use of different implants and inclusion of studies with 2 years of follow-up.[18]
In a cost-effectiveness study, Zmistowski et al.[20] reviewed 14 prospective randomized studies investigating different outcomes from
surgeons who chose selective or nonselective patellar arthroplasty. After nonselective
patellar arthroplasty, the persistence of anterior knee pain was 20.9% in the patellar
retention group and 13.2% in the patellar resurfacing group (p < 0.001), with reoperation rates for patellar conditions of 3.7 and 1.6% (p < 0.01), respectively. In studies excluding patellae with arthrosis, the incidence
of anterior pain was equivalent between groups, that is, 3.1% in the patellar retention
group and 3.2% in the patellar resurfacing group (p = 0.97), while the rate of reoperation due to patellar pain persistence dropped to
1.2 and 0% (p = 0.06), respectively. After assessing the outcomes, complications and related costs,
the study concluded that the routine performance of patella resurfacing does not have
the best cost-benefit relationship and that patellar retention is better because it
avoids revisions due to persistent pain and minimizes risks inherent to patellar arthroplasty.[20]
Since our service performs selective patellar arthroplasty procedures, we believe
that, hypothetically, we exposed the patellar retention group to the best possible
perspective in terms of clinical outcomes – after all, by preserving only patellae
in good macroscopic conditions, we can state that the higher WOMAC scores for pain
in the patellar retention group did not result from moderate or advanced arthrosis.
However, even with selective arthroplasty, the difference between groups was significant,
in contrast to Zmistowski et al.[20] We believe that the selection may also have been a reason for not identifying a
significant difference between groups regarding global WOMAC and difficulty in daily
living activities scores due to the probable reduction of the real difference that
could have occurred in the absence of selection.
We recognize limitations in our study, such as sample size, lack of assessment of
quality of life and mental health scores, and the application of questionnaires over
the telephone. It is worth mentioning that all patients received the same prosthesis
model and that the procedures were performed by the same group of experienced surgeons.
Since we used scores validated for the Brazilian population,[15]
[16] the clinical and functional evaluation were based only on subjective criteria reported
by patients, preventing objective data appraisal by an evaluator. We believe that
this results in reliable outcomes – as Epstein et al.[21] states, symptoms referred by the patient are always the most important data. Instruments
based on patients' report can provide data that are not achieved by physiological
assessments and that may have greater reproducibility on quality of life than clinical,
biochemical, and physiological indices.[21]
Conclusion
The present study revealed no significant difference between the group that underwent
patellar resurfacing and the group with patellar retention in terms of Lequesne score,
global WOMAC score and WOMAC scores for difficulty in daily living/physical activity
and stiffness scores. There was a significant difference only in WOMAC score for pain,
with a better outcome in the group undergoing patellar resurfacing even in the presence
of selective arthroplasty.
