Obesity and the absence of trochlear dysplasia increase the risk of revision in patellofemoral arthroplasty

THEKNE-02094; No of Pages 7 The Knee xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

The Knee

Obesity and the absence of trochlear dysplasia increase the risk of revision in patellofemoral arthroplasty Ming Han Lincoln Liow, Graham Seow-Hng Goh ⁎, Darren Keng-Jin Tay, Shi-Lu Chia, Ngai-Nung Lo, Seng-Jin Yeo Department of Orthopedic Surgery, Singapore General Hospital, 20 College Road, Academia, Level 4, Singapore 169865, Singapore

a r t i c l e

i n f o

Article history: Received 31 December 2014 Received in revised form 1 May 2015 Accepted 28 May 2015 Available online xxxx Keywords: Patellofemoral arthroplasty Obesity Trochlear dysplasia Outcomes Survivorship

a b s t r a c t Purpose: Proper indications and patient selection are paramount in Patellofemoral Arthroplasty (PFA). Although factors predicting outcomes in Total Knee Arthroplasty (TKA) have been studied, there are no such studies for modern PFA. This retrospective study reports the mid-term clinical and radiological outcomes and survivorship of PFA, investigating the risk factors associated with poorer outcomes and higher revision rates. Methods: Fifty-one patients (51 knees) with isolated patellofemoral arthritis underwent PFA using a second-generation implant. Mean follow-up was 4.1 years (range, 2.2–6.1). We stratified the cohort into Obese (body mass index (BMI) ≥ 30 kg/m2, n = 16), Overweight (BMI 25–29.9 kg/m2, n = 20) and Control (BMI 18.5–24.9 kg/m2, n = 15) groups. The same cohort was stratified based on the presence (n = 11) or absence (n = 40) of trochlear dysplasia (TD). Results: Mean Knee Society objective and function scores, Melbourne Knee score and Physical Component Score of Short-Form 36 improved significantly. Obesity was associated with no significant improvement in Melbourne Knee score and Knee Society function score (p N 0.05), a higher incidence of radiographic outliers and lower patient satisfaction. There was no significant difference in outcomes between the TD and Non-TD groups at two years (p N 0.05). Survivorship was 92.2%. Three revisions for progression of tibiofemoral osteoarthritis were noted in the Non-TD and obese or overweight groups. Conclusions: Patients with obesity and the absence of trochlear dysplasia are at a potentially higher risk of revision surgery to TKA and should be counseled that PFA may represent a “bridging” procedure. Level of Evidence: III © 2015 Elsevier B.V. All rights reserved.

1. Introduction Patellofemoral arthroplasty (PFA) may be indicated in carefully selected cases of advanced patellofemoral arthritis, especially in younger patients who have failed conservative management. Early implant designs had sharp, constraining trochlear grooves that were prone to complications such as maltracking and catching of the patella, thus leading to high revision rates [1–3]. However, modern PFA designs have improved on the problems that led to failure of the initial prostheses, making PFA a reasonable alternative to Total Knee Arthroplasty (TKA) in younger patients with isolated patellofemoral arthritis. In addition, PFA has proven benefits of greater bone conservation, shorter post-operative rehabilitation period and better knee kinematics in the sagittal plane due to preservation of the tibiofemoral articulation, menisci, and ligaments [4]. Current literature on second generation PFAs has shown good mid-term results, although varying survivorships of 80% to 95% have been reported [5–9]. This variation is presumably due to differences in patient selection, surgical practice and threshold for revision of poorly ⁎ Corresponding author. Tel.: +65 92386616; fax: +65 6224 9221. E-mail address: [email protected] (G.S.-H. Goh).

functioning implants. If predictive factors for revision surgery are identified and addressed, there is a potential to decrease the revision rate of PFA, allowing more patients to benefit from the advantages associated with PFA. Although the factors predicting outcomes in TKA have been extensively studied [10–14], there is currently no literature on the predictive factors for revision surgery in modern PFA. While the influence of body mass index (BMI) on outcomes of TKA remains controversial with some studies suggesting that it affects outcomes [11,12] and others reporting no effect [13,14], the extent to which obesity affects the outcomes and survivorship of PFA remains uncertain. In a study on the first-generation PFA, van Jonbergen et al. reported a calculated 10-year Kaplan–Meier survivorship of 84%, noting that obese patients with BMI greater than 30 kg/m2 were at increased risk for component failure [15]. As the pressures across the patellofemoral articulation are known to increase with a person's weight, obesity has been found to predispose a person to patellofemoral arthritis [15,16] and could lead to higher revision rates after primary PFA. Trochlear dysplasia is another common risk factor for isolated patellofemoral disease, with the degree of arthritis correlating significantly with the degree of dysplasia [17]. Nicol et al. reported the lowest rates of radiographic progression of tibiofemoral osteoarthritis [5] and Leadbetter et al. reported improved clinical outcomes [9] in patients

http://dx.doi.org/10.1016/j.knee.2015.05.009 0968-0160/© 2015 Elsevier B.V. All rights reserved.

Please cite this article as: Liow MHL, et al, Obesity and the absence of trochlear dysplasia increase the risk of revision in patellofemoral arthroplasty, Knee (2015), http://dx.doi.org/10.1016/j.knee.2015.05.009

2

M.H.L. Liow et al. / The Knee xxx (2015) xxx–xxx

with pre-existing trochlear dysplasia undergoing PFA. If the converse is true, the absence of trochlear dysplasia could also be another predictive factor of poor outcomes and revision surgery after primary PFA. Proper indications and patient selection are paramount to achieving good outcomes and low revision rates with PFA. The purpose of our study was to report the mid-term clinical and radiological outcomes, complication rates and survivorship of PFA using a modern implant, as well as to determine the risk factors associated with poorer clinical outcomes and higher revision rates, hypothesizing that obesity and the absence of trochlear dysplasia play a significant role.

2. Patients and methods Between 2008 and 2012, 51 consecutive patients were treated with PFA. All patients had isolated patellofemoral arthritis, with 40 patients having primary patellofemoral arthritis and 11 patients having trochlear dysplasia. Inclusion criteria that have been previously described were used in this study [18,19], including degenerative osteoarthritis confined to the patellofemoral joint, post-traumatic osteoarthritis, symptoms referable to patellofemoral joint degeneration that were unresponsive to non-operative intervention, patellofemoral malalignment or dysplasia, failed conservative procedures such as arthroscopic debridement, or extensor unloading procedures. Patients were questioned to determine if there was a familial history of early onset knee arthritis, as this might contribute to early progression of tibiofemoral osteoarthritis. Any patient with suspected tibiofemoral osteoarthritis, advanced chondromalacia or chondrocalcinosis, systemic inflammatory arthritis, complex regional pain syndrome, infection or who had a previous history of severe arthrofibrotic healing was excluded from receiving a PFA. The study was approved by our institutional review board. All surgeries were performed by three experienced arthroplasty surgeons using a single PFA prosthesis (SIGMA ® HP Partial Knee, DePuy, Warsaw, IN) at a single institution. The procedure was standardized utilizing an abbreviated medial parapatellar approach under tourniquet. Meticulous surgical technique

was adhered to for all knees. The femoral component was positioned in the most lateralized position on the anterior femur to maximize anterior coverage without medial or lateral overhang after referencing rotation parallel to the transepicondylar axis and perpendicular to the anteroposterior axis (Whiteside's line). The patella was prepared similarly to the principles of TKA utilizing a measured resection approach. A parallel resection at the patellar equator was made to avoid maltracking. Careful measurement of patella thickness in all quadrants confirmed an equal resection. Coverage was maximized with the patella button in a medialized position. The patella was then reconstituted to a normal thickness. Any sources of catching or snapping were addressed prior to implantation of the final prosthesis. Lateral releases were performed as needed for tracking purposes after the tourniquet was let down. Excess cement was removed to prevent third-body wear and formation of a destructive loose body when migrated into the tibiofemoral compartments. Drains were inserted at the end of the procedure at the discretion of the attending surgeon. Postoperatively, all patients received low molecular weight heparin and mechanical calf pumps for thromboprophylaxis. Rehabilitation was performed in accordance with our institution's integrated care pathway. Post-operatively, pain was managed in the inpatient setting with intramuscular pethidine and intravenous ondansetron as needed, and in the outpatient setting with oral analgesics such as paracetamol, gabapentin and arcoxia or naproxen with a proton pump inhibitor. The representative pre-operative and post-operative radiographs are shown in Figs. 1 to 4.

2.1. Outcome measures Two experienced independent physiotherapists performed the pre-operative and post-operative assessment of all patients. They were blind to the measurements of their colleagues. All the patients had preoperative range of motion, Melbourne Knee score [20], Knee Society scores, Short-Form 36 (SF-36) scores. The eight domains (Physical functioning, Social functioning, Role-Physical, Bodily Pain, Mental Health, Role-Emotional, Vitality, and General Health) of SF-36

Fig. 1. Pre-operative AP and lateral view.

Please cite this article as: Liow MHL, et al, Obesity and the absence of trochlear dysplasia increase the risk of revision in patellofemoral arthroplasty, Knee (2015), http://dx.doi.org/10.1016/j.knee.2015.05.009

M.H.L. Liow et al. / The Knee xxx (2015) xxx–xxx

3

Fig. 4. Post-operative skyline view.

Fig. 2. Pre-operative skyline view.

were transformed into two summary scores: the Physical Component Summary (PCS) and Mental Component Summary (MCS). The advantages of PCS and MCS are a smaller confidence interval and elimination of both floor and ceiling effect [21]. All scores were evaluated again at six months, one year, and two years post-operatively, together with an assessment of the patient's fulfillment of expectations and satisfaction rates with surgery. Expectation and satisfaction scores were rated out of a maximum of seven or six respectively, with higher scores indicating poorer results. We stratified expectation scores into: excellent, good, fair and poor. All radiographs were evaluated by the first author. The presence of trochlear dysplasia on pre-operative radiographs was assessed using the Dejour classification method [22]. Degenerative changes of the patellofemoral joint were assessed pre-operatively using the Iwano classification system [23]. Radiological outcomes were measured using the Insall–Salvati (Patella Alta N 1.5) and Caton–Deschamps (Patella

Alta N 1.3, Patella Baja b 0.6) ratios to assess for outliers. In addition, radiological assessment of the patellofemoral joint pre- and postoperatively were assessed with measurements of patella thickness, mediolateral translation and patella tilt using methods by Gomes et al. Data with regard to any further intervention was obtained from the hospital records. Complications were reviewed and survivorship was measured using revision to TKA as an endpoint. The 51 patients (44 females, seven males) had a mean age of 52.7 ± 7.5 years (Table 1). For the subgroup analysis, we stratified the cohort into three groups according to the World Health Organization (WHO) classification of obesity: Obese (BMI ≥ 30 kg/m2, n = 16), Overweight (BMI 25–29.9 kg/m2, n = 20) and Control (BMI 18.5–24.9 kg/m2, n = 15) groups. The mean BMI was 35.3 ± 4.2 kg/m2 for the Obese group, 27.9 ± 1.1 kg/m2 for the Overweight group and 22.9 ± 1.5 kg/m2 for the Control group. The mean BMI was significantly different (p ≤ 0.05) between each of the three groups. Among the total of 51 patients reviewed, 11 patients (mean age, 51.0 ± 8.7 years) had radiographic evidence of trochlear dysplasia (TD) and 40 patients (mean age, 53.1 ± 7.1 years) had no radiographic

Fig. 3. Post-operative AP and lateral view.

Please cite this article as: Liow MHL, et al, Obesity and the absence of trochlear dysplasia increase the risk of revision in patellofemoral arthroplasty, Knee (2015), http://dx.doi.org/10.1016/j.knee.2015.05.009

4

M.H.L. Liow et al. / The Knee xxx (2015) xxx–xxx

Table 1 Patient demographics (n = 51). Mean follow-up period in years (range) Mean age in years ± SD (range) Gender (%) Side (%) Body Mass Index in kg/m2 ± SD (range) Mean length of hospital stay in days (range)

Table 3 Subgroup analysis of improvement in clinical outcomes (only p-values shown) (n = 51). 4.1 (2.2 to 6.1) 52.7 ± 7.5 (39 to 72) 44 female (86%), seven male (14%) 19 Left (37%), 32 Right (63%) 28.7 ± 5.5 (20 to 43) 3.7 (2 to 8)

evidence of trochlear dysplasia (Non-TD) as described by Tecklenburg et al. [24]. They were grouped into the TD and Non-TD groups respectively. Of the 11 patients with trochlear dysplasia, four had type A dysplasia (36%), four had type B dysplasia (36%), none had type C dysplasia, and three had type D dysplasia (28%). All patients in the study had pre-operative radiographic evidence of patellofemoral arthritis as described by Iwano et al. [23], while no patients had significant tibiofemoral arthritis as defined by a Kellgren and Lawrence grade of three or four. All patients were followed up for a mean of 4.1 years (range, 2.2– 6.1 years). Data was prospectively collected pre-operatively, at six months, one year and two years post-operatively. 2.2. Statistical analysis All continuous data was expressed in terms of mean ± standard deviation of the mean. Repeated measures Analysis of Variance (ANOVA) and post hoc tests using the Bonferroni correction was used to determine any significant differences between the scores obtained at set time intervals before and after PFA. A one-way ANOVA was used to compare means between each subgroup. We defined statistical significance at thefive percent (p ≤ 0.05) level. Statistical analysis was carried out with SPSS software version 20.0 (SPSS Inc., Chicago, Illinois). 3. Results For analysis, 51 patients followed through all evaluations and no patients were lost to follow-up. An isolated PFA with a single implant was performed in all patients. No intraoperative complications such as intra-articular fracture, nerve or vessel damage were encountered. Significant improvements were noted in Melbourne Knee score (p b 0.001), Knee Society objective score (p b 0.001), Knee Society functional score (p = 0.002), and PCS (p = 0.001) pre-operatively to six months, one year and two years post-operatively. There was no significant difference in range of motion or MCS (Table 2). Of the 51 patients, intra-operative lateral retinacular release was performed (n = 17) for patients with patellar maltracking during implant trials and prior to cementation. The lateral release (n = 17) and non-lateral release groups (n = 34) both experienced a significant improvement in Knee Society objective and functional scores, Melbourne Knee scores and PCS at six months, one year and two years (p b 0.05). There were no significant differences in clinical outcomes and satisfaction rates at six months, one year and two years (p N 0.05). Lateral release did not affect the results in this study. Subgroup analysis of clinical outcomes in the Obese, Overweight and Control groups was performed demonstrating significant improvements (p ≤ 0.05) in mean Knee Society objective score and PCS pre-operatively to six months, one year and two years postoperatively (Table 3). Of the three groups, only the Control group experienced a significant improvement (p ≤ 0.05) in mean Knee Society function score. In addition, the Control group had a significantly greater knee flexion (p ≤ 0.05) six months, one year and two

Clinical outcomes Range of motion Flexion Extension Melbourne Knee Score Knee Society objective score Knee Society function score Short-Form 36 Physical Component Score Mental Component Score

Obese

Overweight

Control

TD

Non-TD

0.026 0.385 0.062 0.007 0.334

0.816 0.547 0.047 0.001 0.266

0.559 0.846 0.001 0.005 0.003

0.857 0.245 0.135 0.011 0.288

0.100 0.577 0.001 0.001 0.030

0.006 0.239

0.019 0.341

0.003 0.507

0.597 0.546

0.001 0.508

years post-operatively (Table 4). There was also a trend towards higher Knee Society function scores (p = 0.09) in the Control group two years post-operatively. Of note, the Obese group did not experience a significant improvement (p N 0.05) in Melbourne Knee score (Table 3). Only the Obese group experienced a significant improvement (p ≤ 0.05) in knee flexion, however, this was probably due to lower levels of preoperative flexion. There were no significant differences (p N 0.05) in the other clinical outcomes between the three groups. The mean Knee Society function score, Melbourne Knee score and PCS improved significantly (p ≤ 0.05) for only the Non-TD group when using repeated measures ANOVA statistical analysis (Table 3). However, when using a Student's t-test comparing preoperative and two years post-operative scores, significant differences were noted in Knee Society objective score, Knee Society function score and Melbourne Knee score for the Non-TD group, as well as for Knee Society function score and Melbourne Knee score for the TD group (Table 5). This discrepancy was likely due to the smaller sample size in the TD group. There was no significant difference (p N 0.05) in clinical outcomes between the TD and Non-TD groups two years post-operatively (Table 4). Looking at the overall cohort (n = 51), 73%, 77% and 76% had good or excellent fulfillment of expectations and 85%, 80% and 76% experienced good or excellent satisfaction with their surgeries six months, one year and two years post-operatively, respectively. Two years post-operatively, the proportion of patients that experienced good or excellent fulfillment of expectations in the Obese, Overweight and Control groups was 100%, 92% and 100% respectively; and the proportion of patients that experienced good or excellent satisfaction with their surgeries in the Obese, Overweight and Control groups was 75%, 83% and 89% respectively. At two years post-operatively, the proportion of patients that experienced good or excellent fulfillment of expectations in the TD and Non-TD groups was 100% and 96% respectively; and the proportion of patients that experienced good or excellent satisfaction with their surgeries in the TD and Non-TD groups was 86% and 82% respectively. The mean Insall–Salvati ratio increased significantly from 1.03 ± 0.190 to 1.12 ± 0.182° (p ≤ 0.05) and the mean Caton–Deschamps ratio increased significantly from 0.881 ± 0.197 to 0.940 ± 0.184° (p ≤ 0.05). No progressive radiolucent lines or other radiolucencies measuring greater than two millimeters in width were found around the implants in any knee. There were seven patients (13.7%) as radiological outliers in the Obese group, defined using the Insall–Salvati (Patella Alta N 1.5) and Caton–Deschamps (Patella Alta N 1.3, Patella Baja b 0.6) ratios. On the other hand, there were no radiological outliers in the Overweight or Control groups. The TD group had three patients (5.9%) as radiological outliers as compared to four patients (7.8%) in the Non-TD group. Mean pre-operative patella thickness (mm) was 26.96 ± 5.09 in the TD group and 25.40 in the Patellofemoral Osteoarthritis (PFOA) group. Mean post-operative patella thickness (mm) was 23.29 ± 1.72 in the TD group and 24.82 ± 2.81 in the PFOA group. 72.7% in the TD group and 81.1% in the PFOA group were symmetrically resurfaced (p = 0.675). Mean post-operative patella tilt was 9.8° ± 7.0° in the TD group and 11.1° ± 8.5° in the PFOA group (p = 0.639). Mean pre-operative medial-lateral translation (mm) was 14.44 ± 6.26 in the TD group and 9.19 ± 4.52 in the PFOA group. This was reduced post-operatively with a mean (mm) of 4.6 ± 7.0 in the TD group and 2.8 ± 3.9 in the PFOA group (p = 0.279). One superficial wound infection was encountered in a 42 year-old obese (BMI 31.9 kg/m2) female with no trochlear dysplasia on post-operative day eight.

Table 2 Improvement in clinical outcomes (n = 51). Clinical outcomes Range of motion Flexion (in degrees) Extension (in degrees) Melbourne Knee score Knee Society objective score Knee Society functional score Short Form-36 Physical Component Score Mental Component Score

Pre-op

Six months

One year

Two years

p-Value

126.6 ± 14.1 6.00 ± 6.9 12.6 ± 4.63 58.5 ± 19.9 65.9 ± 14.3

128.0 ± 13.9 4.60 ± 2.59 22.9 ± 6.98 89.1 ± 11.8 75.6 ± 19.7

131.6 ± 9.9 5.20 ± 4.26 24.7 ± 5.46 92.7 ± 8.30 81.6 ± 16.3

129.2 ± 12.1 3.30 ± 3.40 24.5 ± 5.87 89.8 ± 12.0 82.8 ± 19.1

0.430 0.405 b0.001 b0.001 0.002

26.8 ± 4.74 45.9 ± 13.0

41.6 ± 9.25 51.6 ± 15.8

45.3 ± 14.1 52.3 ± 12.6

45.4 ± 12.0 48.7 ± 15.6

0.001 0.119

Please cite this article as: Liow MHL, et al, Obesity and the absence of trochlear dysplasia increase the risk of revision in patellofemoral arthroplasty, Knee (2015), http://dx.doi.org/10.1016/j.knee.2015.05.009

M.H.L. Liow et al. / The Knee xxx (2015) xxx–xxx

5

Table 4 Comparison of clinical outcomes between subgroups at two years post-operatively (n = 51). Clinical outcomes Range of motion Flexion (degrees) Extension (degrees) Melbourne Knee Score Knee Society objective score Knee Society function score Short-Form 36 Physical Component Score Mental Component Score

Obese

Overweight

Control

p-Value

TD

Non-TD

p-Value

127.9 ± 5.2 2.63 ± 3.7 24.6 ± 4.8 95.3 ± 2.6 78.1 ± 19

129.8 ± 13 4.17 ± 4.7 22.0 ± 6.6 80.6 ± 17 68.8 ± 26

140.1 ± 10 3.67 ± 4.1 23.9 ± 8.6 85.6 ± 22 90.0 ± 13

0.050 0.732 0.758 0.186 0.093

135.4 ± 9.8 5.80 ± 0.8 22.9 ± 5.1 88.1 ± 12 72.9 ± 26

131.5 ± 12 2.46 ± 4.0 23.7 ± 7.4 85.5 ± 19 79.5 ± 21

0.447 0.080 0.792 0.739 0.501

44.2 ± 14 57.2 ± 12

43.3 ± 13 46.9 ± 14

50.3 ± 7.6 56.3 ± 7.2

0.313 0.090

44.3 ± 10 47.7 ± 15

46.6 ± 12 53.7 ± 11

0.660 0.254

At a mean of 4.1 years (range, 2.2–6.1 years), the overall survivorship, defined as need for conversion to a TKA, was 92.2%. A total of four PFAs were revised (Table 6). Three revisions were due to progression of osteoarthritis into the medial compartment. Of note, all three patients were in the Non-TD group — one was found to be obese while the other two were overweight. The fourth revision was due to patella maltracking, and the patient belonged in the Overweight and TD groups. Of note, there was no statistically significant difference (p N 0.05) in length of followup between patients with and without pre-operative trochlear dysplasia, as well as between the Obese, Overweight and Control groups. In addition, there was no significant correlation (p N 0.05) between pre-operative BMI and presence of trochlear dysplasia.

4. Discussion We report the outcomes of PFA performed for isolated patellofemoral arthritis using a modern implant at a single Institution. In our study, patients undergoing PFA for isolated patellofemoral arthritis had a significant improvement in Melbourne Knee score, Knee Society objective and functional scores and PCS. A higher pre-operative BMI was associated with no significant improvement in Melbourne Knee score and Knee Society function score. Patients in the Control group had a significantly greater flexion and there was also a trend towards higher Knee Society function scores at two years as compared to the Obese and Overweight groups. Higher pre-operative BMI was also associated with a higher incidence of radiographic outliers and lower patient satisfaction. Survivorship at a mean of 4.1 years was 92.2% with four revisions (7.8%). A substantial number of revisions were due to progression of tibiofemoral arthritis and were noted in the Non-TD and Obese or Overweight groups. There were several limitations to this study. This was a retrospective study although the patients were followed up prospectively. There was no matched cohort treated with TKA. In addition, survivorship data was obtained from hospital records but it is unknown whether patients had opted for revision surgery at another institution. The short duration of follow-up is another limitation. However, the authors will be following up on the clinical outcomes and survivorship of the patients enrolled in this study. As standing hip–knee–ankle radiographs were not available for review, we could not include Q-angle measurements in our study, and we acknowledge that coronal malalignment could have confounded the results in our study. Other potential confounders such as age, gender and pre-operative range of motion were not controlled for with use of multiple regression analysis, as the numbers in our study were few. Although criticism can be made regarding the small number of patients included in each group and the presumable lack of power, it is important to note that this is a relatively rare subgroup of patients with knee arthritis, and hence the same difficulty in obtaining numbers was encountered by similar studies [6,7]. Strengths of this study include the use of single

implant and surgery performed in a single institution, with standard postoperative rehabilitation protocols and radiological data. Clinically, patients in this study exhibited significant improvement in outcome measures of pain, function, satisfaction and fulfillment of expectations. Survivorship also mirrors TKA survivorship of 94% at a similar mean follow-up of 5.2 years [25]. Given these results, we suggest that PFA represents a viable alternative to TKA for patients with isolated patellofemoral arthritis, despite the current lack of literature directly comparing PFA and TKA. Clear contraindications for PFA have been reported, however, the ideal candidate for PFA remains highly controversial. There was no significant difference in pre-operative Knee Society function scores between the three groups, and yet the Control group showed a trend towards higher Knee Society function scores at two years postoperatively (p = 0.09). Lower patient satisfaction was also experienced in the Obese and Overweight groups. Most importantly, patients with higher pre-operative BMI experienced significantly less improvement in Melbourne Knee score and Knee Society function score. While it is possible that PFA patients with increased BMI exhibit lower postoperative functional improvement due to increased co-morbidities that may affect function, it remains uncertain if TKA would have provided these patients improved function when compared to PFA until further prospective studies have been conducted. Higher pre-operative BMI was also associated with a higher incidence of radiographic outliers. Obesity may have led to technical challenges during surgery, as increased subcutaneous tissue thickness in obese patients may have made exposure difficult, obscure bony landmarks, and obstruct accurate positioning of cutting guides [26], which may have resulted in an increased number of outliers and poorer outcomes. A number of studies suggest higher revision rates after TKA in obese patients [27], but there are few studies to date evaluating the influence of obesity on revision rates after PFA. With respect to survivorship, we found four revisions in the Obese and Overweight groups as compared to zero in the Control group. Three out of four revisions were due to progression of tibiofemoral arthritis to the medial compartment, a finding that resonates with the current literature as the main cause of failure in PFA [8]. Obesity has long been recognized as a risk factor for knee osteoarthritis due to excessive loading of weight-bearing joints [28], and our study demonstrates the association between obesity and higher revision rates due to progression of tibiofemoral arthritis after PFA. Van Jonbergen et al. shared similar results in a 13-year follow up study on a consecutive series of 185 Richards II prostheses [15]. In our cohort, isolated patellofemoral arthritis was seen at a slightly younger ages in patients who had trochlear dysplasia than in patients

Table 5 Comparison of the pre-operative and two years post-operative scores within the TD and Non-TD groups using the Student's t-test (n = 51). Clinical Outcomes

Melbourne Knee Score Knee Society objective score Knee Society function score

TD

Non-TD

Pre-op

2 years

p-Value

Pre-op

2 years

p-Value

13.8 ± 3.8 71.6 ± 13 66.8 ± 12

22.9 ± 4.8 88.1 ± 11 72.9 ± 24

0.038 0.001 0.157

14.0 ± 6.0 58.1 ± 19 67.1 ± 18

23.7 ± 7.2 86.1 ± 18 79.6 ± 20

0.001 0.001 0.013

Please cite this article as: Liow MHL, et al, Obesity and the absence of trochlear dysplasia increase the risk of revision in patellofemoral arthroplasty, Knee (2015), http://dx.doi.org/10.1016/j.knee.2015.05.009

6

M.H.L. Liow et al. / The Knee xxx (2015) xxx–xxx

Table 6 Patients who underwent revision surgery (n = 4). Case

Age at operation (yr)

Gender

BMI (kg/m2)

Presence of trochlear dysplasia

Reason

Revision surgery done

1 2 3 4

52 54 56 59

F M F F

33.6 27.9 27.5 27.2

No No No Yes

Progression of OA into medial compartment Progression of OA into medial compartment Progression of OA into medial compartment Patella maltracking with lateral subluxation

TKA TKA TKA TKA

who had patellofemoral arthritis of other causes, a finding previously shown in the literature [17]. A number of authors have reported better clinical outcomes in patients with arthritis secondary to trochlear dysplasia than in other groups. Argenson et al. found 73% good results in the dysplasia group and 54% good results in the group with primary osteoarthritis [16], while Leadbetter et al. found 83% good to very good results and overall higher patient reported outcome scores in those with arthritis secondary to trochlear dysplasia relative to those with other diagnoses [9]. In contrast, Van Jonbergen et al. recently found no differences in outcomes among the three different indications, namely trochlear dysplasia, post-traumatic patellofemoral arthritis or primary patellofemoral arthritis [15]. Concurring with the authors, we found no difference in clinical outcomes between the TD and Non-TD groups. However, we noted a significantly lower number of revisions due to tibiofemoral arthritis progression in the TD group as compared to the Non-TD group. In addition, we analyzed pre- and post-operative patellofemoral articulation radiological outcomes, which demonstrated restoration of patella thickness when compared to Dahm et al's study [29]. There was also restoration of patella tilt and medial-lateral subluxation values in the TD and PFOA groups, which were not statistically significant. However, we noted slightly elevated post-operative patella tilt and medial-lateral subluxation values in the TD group when compared to the PFOA group. We postulate that this could be explained by the use of an inlay implant which mimics the patient's natural anatomy, which may confer slightly reduced stability as the implant is unable to recreate the original trochlear anatomy. Although the same three patients who underwent revision in the Non-TD group were also overweight or obese, it is likely that the two variables are independent as there is no pathophysiological correlation between obesity and the absence of trochlear dysplasia. With regards to this, we found no significant correlation between pre-operative BMI and the presence or absence of trochlear dysplasia. Our study noted a higher revision rate in obese and overweight patients without trochlear dysplasia, which corroborates with a previous study by Nicol et al., who found that progression of tibiofemoral arthritis was much less likely in cases where the primary pathology of patellofemoral arthritis was trochlear dysplasia or patellafemoral malalignment rather than primary patellofemoral arthritis [5]. This finding was also supported by other authors who published case series in which the selection criteria primarily included patients with trochlear dysplasia or malalignment rather than primary patellofemoral arthritis [30,31]. It is thus reasonable to suggest that obese patients without trochlear dysplasia are at a relatively higher risk of revision as compared to non-obese patients with trochlear dysplasia, and the absence of trochlear dysplasia should be considered as a possible predictive factor of higher revision rates. One patient in the trochlear dysplasia group underwent revision, but this was done for persistent patellar maltracking post-operatively. 5. Conclusion In conclusion, our results demonstrate that in the appropriately selected patient with severe progressive patellofemoral osteoarthritis, PFA has been shown to be a successful procedure to improve pain and function, meeting patient's expectations and providing acceptable satisfaction scores. Our mid-term results further support PFA as a viable

option to delay TKA for the treatment of isolated patellofemoral arthritis in younger patients, because a TKA would invariably lead to revision surgery. However, careful patient selection remains paramount to the success of PFA. Patients with higher BMI may have less functional improvement following PFA and higher revision rates. Given that obesity is a potentially modifiable risk factor, pre-operative and postoperative management specifically targeting obese patients should be developed, although weight reduction will continue to be a challenge in these patients pre- and post-operatively. Additionally, isolated patellofemoral arthritis in the absence of trochlear dysplasia may be the harbinger of progressive age-related degeneration in the knee. Our study suggests that a combination of obesity and absence of trochlear dysplasia may be a strong predictive factor for revision to TKA. Until larger prospective studies with longer follow-up are conducted to delineate the role of obesity and the absence of trochlear dysplasia in the natural history of PFA, these select groups of patients should be counseled that PFA may represent a “bridging” procedure and that they are at a potential higher risk of revision surgery to TKA. Conflicts of interest None. Acknowledgments We would like to acknowledge our physiotherapist colleagues, Mr. William Yeo and Ms. Chong Hwei Chi, for assisting us in the preoperative and postoperative assessments of the patients. References [1] McKeever DC. Patellar prosthesis. J Bone Joint Surg Am 1955;37-A:1074–84. [2] Tauro B, Ackroyd C, Newman J, Shah NA. The Lubinus patellofemoral arthroplasty: a five-to ten year prospective study. J Bone Joint Surg Br 2001;83:696. [3] van Wagenberg JMF, Speigner B, Gosens T, de Waal Malefijt J. Midterm clinical results of the Autocentric II patellofemoral prosthesis. Int Orthop 2009;33:1603. [4] Hollinghurst D, Stoney J, Ward T, Pandit H, Beard D, Murray DW. In vivo sagittal plane kinematics of the Avon patellofemoral arthroplasty. J Arthroplasty 2007; 22(1):117. [5] Nicol SG, Loveridge JM, Weale AE, Ackroyd CE, Newman JH. Arthritis progression after patellofemoral joint replacement. Knee 2006:290–5. [6] Mont MA, Johnson AJ, Naziri Q, Kolisek FR, Leadbetter WB. Patellofemoral arthroplasty: 7-year mean follow-up. J Arthroplasty 2012;27:358. [7] Odumenya M, Costa ML, Parsons N, Achten J, Dhillon M, Krikler SJ. The Avon patellofemoral joint replacement: five-year results from an independent centre. J Bone Joint Surg Br 2010;92-B:56–60. [8] Ackroyd CE, Newman JH, Evans R, Eldridge JDJ, Joslin CC. The Avon patellofemoral arthroplasty: five year survivorship and functional results. J Bone Joint Surg Br 2007;89-B:310–5. [9] Leadbetter WB, Kolisek FR, Levitt RL, Brooker AF, Zietz P, Marker DR. Patellofemoral arthroplasty: a multi-centre study with minimum 2-year follow-up. Int Orthop 2009;33:1597. [10] Lingard EA, Katz JN, Wright EA, Sledge CB, Kinemax Outcomes Group. Predicting the outcome of total knee arthroplasty. J Bone Joint Surg Am 2004;86-A(10):2179–86. [11] Bozic KJ, Lau E, Ong K, Chan V, Kurtz S, Vail TP. Risk factors for early revision after primary TKA in Medicare patients. Clin Orthop Relat Res 2014;472:232–7. [12] Nunez M, Nunez E, del Val JL, Ortega R, Segur JM, Hernandez MV. Health-related quality of life in patients with osteoarthritis after total knee replacement: factors influencing outcomes at 36 months of follow-up. Osteoarthr Cartil 2007;15:1001–7. [13] Judge A, Arden NK, Cooper C, Kassim JM, Carr AJ, Field RE. Predictors of outcomes of total knee replacement surgery. Rheumatology 2012;51:1804–13. [14] Avyar V, Burnett R, Coutts FJ, van der Linden ML, Mercer TH. The influence of obesity on patient reported outcomes following total knee replacement. Arthritis 2012; 185208.

Please cite this article as: Liow MHL, et al, Obesity and the absence of trochlear dysplasia increase the risk of revision in patellofemoral arthroplasty, Knee (2015), http://dx.doi.org/10.1016/j.knee.2015.05.009

M.H.L. Liow et al. / The Knee xxx (2015) xxx–xxx [15] van Jonbergen HP, Werkman DM, Barnaart LF, van Kampen A. Longterm outcomes of patellofemoral arthroplasty. J Arthroplasty 2010;25:1066. [16] Argenson JN, Flecher X, Parratte S, Aubaniac JM. Patellofemoral arthroplasty: an update. Clin Orthop Relat Res 2005;440:50. [17] Dejour D, Allain J. Histoire naturelle de l'arthrose fe'moro-patellaire isole'e. Rev Chir Orthop 2004;90(Suppl. 5):1S69-129. [18] Leadbetter WB, Seyler TM, Ragland PS, Mont MA. Indications, contraindications, and pitfalls of patellofemoral arthroplasty. J Bone Joint Surg Am 2006;88(Suppl. 4): 122–37. [19] Ackroyd CE, Chir B. Development and early results of a new patellofemoral arthroplasty. Clin Orthop Relat Res 2005;436:7–13. [20] Feller JA, Knee RJ, Lang DM. Patellar resurfacing versus retention in total knee arthroplasty. J Bone Joint Surg Br 1996;78:226–8. [21] Ware Jr JE, Kosinski M, Bayliss MS, McHorney CA, Rogers WH, Raczek A. Comparison of methods for the scoring and statistical analysis of SF-36 health profile and summary measures: summary of results from the Medical Outcomes Study. Med Care 1995;33:AS264. [22] Dejour H, Walch G, Nove-Josserand L, et al. Factor of patellar instability: an anatomic radiographic study. Knee Surg Sports Traumatol Arthrosc 1994;2(1):19–26. [23] Iwano T, Kurosawa H, Tokuyama H, et al. Roentgenographic and clinical findings of patellofemoral osteoarthrosis. With special reference to its relationship to

[24] [25] [26] [27] [28] [29]

[30] [31]

7

femorotibial osteoarthrosis and etiologic factors. Clin Orthop Relat Res 1990;252: 190–7. Tecklenburg K, Dejour D, Hoser C, Fink C. Bony and cartilaginous anatomy of the patellofemoral joint. Knee Surg Sports Traumatol Arthrosc 2006;14:235–40. Parvizi J, Stuart MJ, Pagnano MW, Hanssen AD. Total knee arthroplasty in patients with isolated patellofemoral arthritis. Clin Orthop Relat Res 2001;392:147. Winiarsky R, Barth P, Lotke P. Total knee arthroplasty in morbidly obese patients. J Bone Joint Surg Am 1998;80(12):1770. Amin AK, Clayton RAE, Patton JT, et al. Total knee replacement in morbidly obese patients. J Bone Joint Surg Br 2006;88-B:1321–6. Felson DT, Anderson JJ, Naimark A, Walker AM, Meenan RF. Obesity and knee osteoarthritis: the Framingham Study. Ann Intern Med 1988;109(1):18–24. Dahm DL, Al-Rayashi W, Dajani K, Shah JP, Levy BA, Stuart MJ. Patellofemoral arthroplasty versus total knee arthroplasty in patients with isolated patellofemoral osteoarthritis. Am J Orthop (Belle Mead NJ) 2010 Oct;39(10):487–91. Argenson J, Guillaume JM, Aubaniac JM, et al. Is there a place for patellofemoral arthroplasty? Clin Orthop Relat Res 1995;321:162–7. Cartier P, Sanouiller JL, Khefacha A. Long-term results with the first patellofemoral prosthesis. Clin Orthop Relat Res 2005;436:47–54.

Please cite this article as: Liow MHL, et al, Obesity and the absence of trochlear dysplasia increase the risk of revision in patellofemoral arthroplasty, Knee (2015), http://dx.doi.org/10.1016/j.knee.2015.05.009