The indications for UKA have evolved and have had an impact on the failure rate of the procedure. The classic indications, as noted by Kozen and Scott [37], include patients with degenerative arthritis in one compartment, age greater than 60 years old, weight less than 82 kg, low-impact work/lifestyle, minimal rest pain, minimum flexion of 90° with less than 5° of flexion contracture, angular deformity less than 10° of varus or 15° of valgus, intact anterior cruciate ligament, and intact opposite compartment. Using these criteria, the use of UKA has been reported to vary from 6 to 30% of patients undergoing knee arthroplasty [12, 38, 39]. Some authors have advocated use of UKA in the younger, more active patient as the first in a series of arthroplasties because of the perceived ease of revisability [40, 41]. While the incidence of osteoarthritis has remained constant, the use of UKA has increased in the early twenty-first century due in part to the popularity of minimally invasive surgery (MIS). Over the last decade, the use of UKA has declined as reflected in the AOANJRR data, but it remains a viable option for OA of one compartment of the knee [28].

Similar to any joint reconstructive procedure, there are patient-related factors, surgical technique factors, and implant-related factors that can contribute to failure.

The ideal diagnosis for unicompartmental arthroplasty is osteoarthritis or osteonecrosis without metaphyseal involvement involving either the medial or lateral compartment of the knee. Patients with inflammatory arthritis or chondrocalcinosis should be avoided. Some authors have noted increased failure rates in obese patient [9, 26]. Younger-aged patients (40–60 years old) were not associated with a higher revision rate [40]. However, the recent AOANJRR noted a significantly higher revision rate in patients less than 65 years old [28].

Surgical technique has evolved over the last two decades with use of intramedullary/extramedullary instruments, computer-assisted techniques, and robotic guidance. Some argue that UKA is technically more demanding than TKA, with a larger learning curve. If technical errors do occur, UKA is less forgiving than TKA. The experience of the surgeon and/or center has been associated with the rate of failure for this procedure [30]. In one study, a specialty center had a lower failure rate versus results from a multicenter group with less experienced surgeons. Seven of eight revisions in this series occurred in the first ten procedures at each hospital [41]. Review of data from the Swedish Knee Registry revealed the risk of revision for failed UKA to be 1.63 times greater for less experienced surgeons versus a more experienced group. In the United States, 70% of TKAs are performed by surgeons who perform 30 or fewer a year. If the indications for UKA are 10–20% of patients considered for arthroplasty, then the question of the minimum number of procedures to maintain proficiency is warranted.

Most authors have advocated slight undercorrection of the deformity in UKA to avoid overload of the unresurfaced opposite compartment. On the tibial side, avoidance of varus placement of the tibial component is important to avoid increase stress on the cancellous bone. The importance of implant-to-implant alignment and proper soft tissue tensioning have also been advocated. Bone cuts are conservative, but the surgeon must avoid overstuffing the compartment with implant. This leads to overcorrection and subluxation of the implants and joints. Extensive soft tissue releases are not necessary in patients undergoing UKA, as deformity is not typically significant. Fixation with cement has led to better short- and long-term results in UKA, versus use of cementless implants, and appears to be the most appropriate fixation at this time.

First-generation implants had all-polyethylene tibial components . Several authors cited thin polyethylene—less than 6 mm in thickness—as a risk factor for failure in these first-generation implants. Second-generation implants with metal backing had overall thinner polyethylene, particularly at the periphery, which led to an increase of polyethylene wear as a failure mode [9]. White et al. reported that the wear pattern of varus knees with early disease is anterior and peripheral. Retrieval of these second-generation implants revealed a similar pattern of wear [42]. Thus, the greatest stresses were placed on the thinnest polyethylene (Fig. 21.4). Polyethylene sterilized with gamma radiation in air and a long shelf life led to early catastrophic failure in a series of UKAs reported by McGovern et al. At a mean of 18 months after index UKA, 49% of the implants were either revised or scheduled for revision secondary to polyethylene wear [43].

Current UKA implants are resurfacing in nature . Fixed-bearing implants attempt to strike a balance between optimizing contact area and limiting constraint between the implants. Modular systems avoid thinner polyethylene at the locking mechanism and use polyethylene that avoids oxidative degradation. Mobile-bearing implants have a more constraining surface geometry, but fixation is not compromised due to the mobile bearing. Current designs of fixed and mobile-bearing implants make polyethylene wear in the first decade of use unusual.

Mechanisms of failure have varied over the last three decades and from various studies. The most common causes are component loosening, progressive arthritis, polyethylene wear, and mechanical problems.

The initial evaluation of a patient with a painful UKA is similar to that of a patient with a painful TKA, and the approach outlined in Chap. 3 is used. As previously noted, revision for pain without a clear-cut etiology of the pain is only rarely successful. The surgeon must ask: “What has failed?”

Failure of polyethylene in a modular implant can be associated with an intact femoral component and tibial base plate, loosening of one or both implants, and associated osteolysis. Failure of fixation may occur with one or both implants and may be associated with some degree of bone loss. Progression of disease most likely will involve the opposite compartment, but occasionally the patellofemoral joint. This is confirmed with weight-bearing radiographs, as well as a sunrise view of the patella .