265 / 460
D.L. Dahm
264
Patient demographics were collected for both
PFA and TKA groups. Preoperative and
postoperative
Knee Society Clinical Rating
System
(KSS) scores were calculated from
standardizedquestionnairesgivenprospectively
to all patients as part of surgery and at regular
intervals thereafter. Preoperative Tegner
activity scores and
University of California,
Los Angeles
(UCLA) activity scores were
obtained retrospectively for patient records.
Postoperative Tegner and UCLA scores that
were not obtainable through chart reviews were
obtained by telephone interview. Preoperative
and postoperative radiographs were reviewed.
Trochlear dysplasia was assessed using the
Dejour classification method [13]. Patellar
position was determined using the Insall Salvati
index [14].
Statistics
Compared comparisons between the PSA and
TKA groups were performed using Wilcoxon
signed-rank tests. Multivariate regression was
used to analyze the outcome effects of any
independent variable showing its significant
difference between the groups. All regression
models were analyzed for power and fit with
significance set at .05. The statistical analysis
was performed using JMP6 statistical discovery
software (SAS, Inc., Cary, NC).
Results
Between January 2003 and December 2005,
3500 patients underwent a knee arthroplasty
procedure at our institution. We identified 205
knees (5.8%) coded in our database for patello
femoral arthritis. After detailed radiographic
review, 45 knees (1.3%) with isolated patello
femoral arthritis were identified. There were 23
knees (0.7%) that underwent PFA and 22 knees
(0.6%) that underwent TKA.
Mean follow-up was 29 months (range, 24 to
49 months) in the PFA group and 27 months
(range, 24 to 33 months) in the TKA group.
There were no statistically significant diffe
rences between the groups in regards to gender,
race, body mass index, average number of prior
knee surgeries, smoking status, or employment
status. Mean age at the time of surgery was 60
years (range, 39-81 years) in the PFA group
and 69 years (range, 44-83 years) in the TKA
group (p=0.01). There were no statistically
significant differences between the PFA and
TKA groups in relation to mean preoperative
Kellgren and Lawrence score, mean pre
operative Iwano score or the presence/absence
of trochlear dysplasia. Mean preoperative KSS
scores, KSS function scores, Tegner scores,
and UCLAscores were not statistically different
between groups. Likewise, mean preoperative
range of motion was similar between groups.
Mean postoperative Knee Society clinical
rating systems scores were 89 and 90 in the
PFA and TKA cohorts respectively. Mean
UCLA scores were 6.6 in the PFA group and
4.2 in the TKA group (p<0.0001). Mean blood
loss (p=0.03) and hospital stay (p=0.001) were
significantly lower among PFA patients. Linear
regression analysis showed that blood loss,
hospital stay, and functional outcomes were not
affected by age as an independent variable. No
significant complications occurred in the PFA
group. There was one deep vein thrombosis in
the TKA group; additionally, one patient in the
TKA group required a manipulation under
anesthesia. At last follow-up, no patient in
either group had required revision knee
arthroplasty.
Discussion
Thesurgicaltreatmentofadvancedsymptomatic
patellofemoral arthritis remains somewhat
controversial. Satisfactory results have been
reported for both PFA and TKA in this setting
[6-10, 15]. In the present study we retrospec
tively compared the clinical and functional
outcomes of patients who underwent either
PFA or TKA for treatment of isolated patello
femoral arthritis. Demographically the two
cohorts were remarkably similar. Radiographs
were reviewed carefully to ensure that only
patients treated for isolated patellofemoral
arthritis were included. Although the TKA