Introduction

Since the introduction of the total condylar knee prosthesis in the 1970s, virtually all total knee arthroplasty (TKA) surgery is characterized by complete resection of the distal femoral and proximal tibial articular surfaces, and replacement with synthetic materials. Consequently, all abnormal cartilage and subchondral bone associated with osteoarthritis is removed from the primary weight bearing surfaces of the knee joint. Although there is little controversy about the routine resection of the femoral and tibial articular surfaces in the setting of multicompartmental osteoarthritis of the knee, optimal management of the patellar articular surface is less clear. Some knee surgeons advocate leaving the native patellar surface intact, others recommend routine resurfacing with a polyethylene component, while yet others recommend selective resurfacing based on one or more patient factors and/or intraoperative findings.¹

Available literature and quality of the evidence

A considerable body of higher‐quality evidence was identified to answer these questions. A search of Medline and EMBASE databases identified 11 English‐language meta‐analyses of randomized and/or quasi‐randomized controlled trials that address one or more of these questions,^2–12 with publication dates spanning from 2005 until 2018. These meta‐analyses aggregate data from between 8 and 28 individual level I and II studies published between 1982 and 2015, with a total of 53 studies appearing in one or more meta‐analyses.

Rationale and clinical comment

The primary goals of total knee replacement surgery are to improve patients’ quality of life on an elective basis, specifically in terms of reducing pain and functional limitations associated with degenerative disease of the Knee Consequently, any assessment of the relative outcomes of knee replacement surgery with or without patellar resurfacing should be made from the patients’ perspective, using validated patient‐reported outcome measures (PROMs). Patient‐reported outcome measures reflect the results of surgery that matter most to patients. They have a high importance when determining the effectiveness of TKA in treating patient symptoms, as well as when comparing different surgery performed using different techniques or implants. Although TKA can affect general health, joint/disease‐specific patient‐reported outcomes are more sensitive in detecting differences in outcomes of this procedure.

Available literature and quality of the evidence

Of the 11 meta‐analyses identified, four specifically assessed for differences in patient‐reported outcomes following TKA with or without patellar resurfacing.^2,7,9,12 These four meta‐analyses synthesized data from between 5 and 9 individual controlled trials, encompassing between 456 and 1102 individual TKA procedures. Three of these meta‐analyses assessed patient satisfaction,^7,9,12 while the fourth assessed patient‐reported general knee pain using a Visual Analog Scale (VAS).²

None of the meta‐analyses assessed outcomes in terms of validated joint‐specific outcome measures. Consequently, Medline, Embase, and the reference lists of all 11 meta‐analyses were searched to identify individual studies that assessed outcomes using joint‐specific PROMs, with three level I studies identified.^13–15 Given this available evidence, no lower‐quality studies were used to address this question.

Findings

Three meta‐analyses synthesized data from between five and nine studies to assess in the impact of patellar resurfacing on patient satisfaction with surgery,^7,9,12 with all three finding no significant difference between groups. Two meta‐analyses pooled data from five and nine trials respectively to determine aggregate satisfaction rates,^9,12 with both finding near‐identical proportions of satisfied patients in both groups, ranging from 90.0 to 92.1% among patients who underwent patellar resurfacing, and between 89.1 and 89.3% in patients who did not. The relative risk for dissatisfaction after TKA failed to reach significance in all three meta‐analyses, with a 95% confidence interval (CI) spanning one in all cases.

One network meta‐analysis with inclusion criteria limited to level I trials compared 10‐point VAS pain scores between three groups of patients across nine studies who underwent TKA: those with patellar resurfacing (n = 154), those with patellar denervation but no resurfacing (n = 135), and those with no denervation or resurfacing (n = 167).² Analysis of pooled results failed to identify any significant unadjusted mean difference (UMD) in pain scores associated with patellar resurfacing, irrespective of whether unresurfaced patellae were (UMD 0.11 points [−0.21 to 0.43]) or were not (UMD 0.11 [−0.21 to 0.44]) denervated.

Three individual trials were identified that reported joint‐specific patient reported outcomes, none of which identified any significant differences associated with patellar resurfacing.^13–15 Two trials reported components of the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) at follow‐up times between 1 and 10 years,^14,15 while one reported the Knee Osteoarthritis Outcome Score (KOOS) at three months to six years.¹³ All three studies were limited by small sample sizes (ranging from 16 to 54 per group) and notable loss to follow‐up. Differences in reporting (absolute vs change scores, time intervals) precluded pooling of WOMAC scores across the two studies. While a number of meta‐analyses reported outcomes measured using surgeon‐reported outcome measures such as the Knee Society Score (KSS) or Hospital for Special Surgery (HSS) knee score, nonpooled PROM data.

Overall, while a number of studies have reported differences in patient satisfaction associated with patellar resurfacing in TKA, the presently available evidence is limited by infrequent reporting of validated joint‐specific PROMs and variability in reporting. It is possible to say with moderate‐to‐high certainty that there is no clinically relevant difference in patient‐reported satisfaction with TKA associated with patellar resurfacing versus nonresurfacing. While the available evidence suggests no difference in joint‐specific PROMs associated with patellar resurfacing, certainty in this finding is low owing to marked limitations with the available evidence.

Resolution of clinical scenario

In terms of patient‐reported outcomes of TKA, the best available evidence suggests no difference in patient satisfaction with surgery irrespective of whether or not the patellar is resurfaced. However, insufficient evidence is available to guide decision‐making around patellar resurfacing based on potential differences in joint‐specific PROMs