UKA v TKA
We have been able to identify 13 published clinical studies comparing the outcomes of total and unicompartmental knee replacement. Two studies are randomised controlled trials (RCTs), six are cohort studies, three are case-control studies and two are self-controlled studies of patients receiving TKA in one knee and OUKA in the other. In the following paragraphs, key results from these studies are summarised.
Ackroyd et al. (2002) performed the first RCT comparing UKA to TKA. Between 1989 and 1992, 102 knees suitable for UKA were randomised to receive either a St Georg Sled UKA or a Kinematic modular TKA. The early results demonstrated that the UKA group had fewer complications and more rapid rehabilitation than the TKA group. At five years, there was an equal number of failures in the two groups but the UKA group had more excellent results and a greater range of movement. At 15 years, 43 patients (45 knees) had died with their prosthetic knees intact (Ackroyd et al, 2002; Steele et al, 2006). The Bristol knee scores of the UKA group throughout the review period were better and, at 15 years, 15 (71.1%) of the surviving UKAs and 10 (53.1%) of the surviving TKAs had achieved an excellent score. The 15-year survival rate based on revision or failure for any reason was 90% for UKA and 79% for TKA. During the 15 years of the review, four UKAs and six TKAs failed.
Sun et al. (2012) randomised 56 patients to either an OUKA, or an AGC TKA (both Biomet, Warsaw, USA) and reported the results with a mean follow-up of 52 months. The UKA group had a shorter operative time, less blood loss, a lower transfusion requirement and fewer DVTs than TKA. KSS-Obj and mean range of movement was higher in the UKA group, but neither are described as being statistically significant. Seven patients in the UKA group required revision for tibial loosening (six patients) or subsidence. The authors attribute this to the learning-curve effect, as all were performed within the first two years of use of the OUKA. There were no revisions in the TKA group.
A case-control study, by Lombardi et al. (2009), compared early (mean 31 months) outcomes for TKA and UKA, with particular emphasis on speed of recovery. 103 consecutive UKA patients (115 knees) were matched to the same number of TKA patients (and knees) on the basis of age, gender, BMI and bilaterality. The number of revisions was similar as were the number of complications. Stiffness requiring manipulation under anaesthetic (MUA) was significantly more common in TKA (7/115 v 0/115, p=0.007). Patients with UKA had a higher mean haemoglobin at discharge (12.1 g/dL v 11.3 g/dL, p<0.001), shorter hospital stays (1.4 days v 2.2 days, p<0.001) and a better mean range of movement (77° of flexion v 67°, p<0.001). There was no statistically significant difference in KSS-Fcn or KSS-Obj.
Amin et al. (2006) matched 54 consecutive UKAs with 54 TKAs on age, gender, BMI, range of movement and pre-operative function. At a mean follow-up of 59 months, there was a superior range of movement, but inferior survival, in UKA compared to TKA. There were no differences in the outcome scores.
Sweeney et al. (2013) performed a retrospective cohort comparison of 317 OUKAs to 425 Advance TKAs (Wright Medical Technologies, Arlington, TN) from a single institution. The authors used hierarchical linear modelling to estimate the treatment effect of UKA and TKA at different time-points, allowing for clustering by surgeon and adjusting for age and gender. At six months, there was no significant difference between TKA and UKA.
Walton et al. (2006) retrospectively compared 183 UKAs with a non-matched group of 142 TKAs, reporting a mean OKS of 37.8 in the UKA group and 35.5 in the TKA group, a significant difference (p=0.04). Weale et al. (2001) retrospectively compared cohorts of patients undergoing OUKA (31 patients) and AGC TKA (130 patients), reporting inferior survival in the UKA group but with functional outcomes similar to TKA.
Three studies have been published comparing outcomes for UKA and TKA in the same patient. Costa et al. (2011) performed simultaneous bilateral knee replacement in 34 consecutive patients, performing UKA in one knee and TKA in the other. Five knees from the UKA group (all received the EIUS system (Stryker, Marwar, NJ)), were revised, four for peri-prosthetic fractures and one with unexplained pain. There was no difference between the groups on the basis of KSS-Fcn or KSS-Obj scores. Dalury et al. (2009) performed a retrospective review of all patients in a single centre who had received a TKA on one side and a UKA on the other. Range of motion was greater in the UKA group (123° v 120°), but there was no difference in functional scores between the groups. Of the 23 patients, 12 expressed a preference for their UKA knee and none expressed a preference for their TKA knee. A study by Laurencin et al. (1991) compared TKAs to UKAs in patients who had received both, reporting a superior range of motion in the UKA (123° v 110°). In this study, 44% of patients expressed a preference for their UKA, 12% for their TKA and 44% could perceive no difference between the function of their knees.
Rougraff et al. (1991) retrospectively compared 120 UKAs to 81 TKAs, finding superior survival in UKA (the end point being any reoperation) and significantly superior combined KSS scores in the UKA group.
In the USA, an independent telephone survey done at Washington University compared the level of satisfaction with various activities (Fig. 10.10(a)) and extent of residual symptoms (Fig. 10.10(b)) in a series of 353 mobile UKAs, 104 fixed bearing UKAs and 661 TKAs implanted in four centres (Berend et al, 2014). Overall, the results of mobile UKA were better than fixed UKA or TKA. The fixed bearing UKA did better than TKA on some questions and worse on others.
