This study found that very good results can be achieved when the FLO is used to treat lateral compartment failure after medial OUKR and appears to achieve clinical outcomes that are as good, if not better, than if a TKR was used. In this elderly patient cohort (mean age 74 at the time of FLO) there were no early complications, no re-operations related to the FLO and at 5 years there was a mean OKS of 40. Evidence from the literature would suggest that conversion to bi-UKR is advantageous over TKR, due to faster recovery, fewer complications, and better function, in appropriate circumstances [9, 12, 21]. The main concern with a staged bi-UKR is that similar to UKR, the procedure will have a higher failure rate than TKR in the long term [16]. However, as the progression of arthritis in the contralateral compartment can no longer occur, the reoperation rate of bi-UKR may well be comparable to that of TKR.

Categorical distribution of pre-operative Oxford Knee Score (OKS) and 1-year post-operative OKS according to Kalairajah et al. [10]

1-year post-operative Oxford Knee Score (OKS) versus patient age at the time of lateral UKR (p = 0.02). Simple linear regression with 95% confidence bands shown

There was one re-operation 14 years after the initial medial OUKR, for a bearing exchange following dislocation of the medial mobile bearing. It is now recommended that the medial bearing thickness should be assessed on a well-aligned pre-operative AP radiograph, before the FLO. If the bearing appears very thin then, during the operation it should be replaced, and any associated impingement addressed. As increased wear is usually caused by impingement and impingement is the main cause of dislocation, this approach should prevent dislocation in the future.

At one and two years postoperatively, the mean OKS was 42, with 87% of the patients achieving an excellent or good OKS. These results are similar to those achieved following primary UKR [18]. A negative correlation between age and OKS at one year post-operatively (p = 0.02) is to be expected as increasing age is associated with both decreasing activity and more major medical conditions [4]. There was an increase in mean TAS post-operatively, but this increase is not significant, as expected in an elderly population with low baseline activity. No correlations were found between clinical outcome scores and the severity of the damage to the lateral femoral or tibial condyle, gender, BMI, or time between medial and lateral UKR operations. This suggests that when considering whether to perform staged bi-UKR these factors need not be taken into consideration. However, the appropriate indications for the procedure need to be satisfied.

The literature would suggest that the early post-operative OKS of 42 for these patients, with lateral osteoarthritis following medial UKR, treated with lateral UKR are appreciably better than scores of patients, with lateral osteoarthritis following medial UKR, treated with TKR. For example, following conversion of UKR to TKR, Pearse et al. reported a mean OKS of 30 and Jonas et al. reported a mean OKS of 32 [9, 21]. However, the outcome of revisions of UKR to TKR is influenced by the reason for revision. As a conversion of a medial UKR with lateral osteoarthritis to TKR is usually a simple primary TKR the results may be better than those for conversions of UKR to TKR for other reasons. Kerens et al. reported a mean OKS at 1 year of 38 in a cohort that of patients mainly revised for lateral osteoarthritis, which is still not as good as the results of revision with a FLO. [12]. When comparing the difference in outcome of a medial UKR treated with lateral UKR or TKR, the most similar reports are those comparing bi-UKR and TKR. These studies have shown better results for bi-UKR in multiple scores including Western Ontario and McMaster Universities (WOMAC), Knee Society (KSS), OKS and EQ-5D scores [11, 14]. In addition, bi-UKR have improved biomechanical and functional results measured both in-vitro, using cadaveric knees, and in-vivo, using gate analysis [5]. These functional improvements were attributed to the bone and ACL-preserving nature of bi-UKR when compared to TKR.

In this study, the FLO were all implanted without medical complications, re-admissions, or early mortality which is to be expected considering the minimally invasive nature of the procedure. These results are supported by previous studies, showing a shorter recovery time for staged bi-UKR than TKR [8]. Furthermore, as the procedure is a minimally invasive UKR, the risk of medical complications such as stroke, myocardial infarction, thromboembolism, deep infection and early mortality is about half that of TKR [13]. This is particularly important as most patients undergoing these procedures are elderly. From the patient’s perspective, by the time they develop lateral osteoarthritis they have had a well-functioning medial UKR for many years. As a result, they tend not want their medial UKR removed and prefer a lateral UKR. Unlike national registries, they do not consider their medial UKR a failure and are pleased to have the opportunity to have another UKR.

The main limitation of the study is that the sample size is small, making it difficult to extrapolate conclusions to a larger population. However, lateral osteoarthritis after UKR, despite being the most common cause for revision, is rare so it is difficult to do a large study. Another limitation is that there was no matched group of patients with lateral osteoarthritis following medial UKR treated with TKR, so to compare outcomes we were reliant on the literature.