The clinical outcome of total knee arthroplasty is compromised by a previously implanted medial unicondylar knee arthroplasty

Objective To investigate the clinical outcome of patients that underwent conversion of a medial unicondylar knee arthroplasty (UKA) to a total knee arthroplasty (TKA) and to compare that outcome to patients that underwent primary TKA. It was hypothesized that those groups would significantly differ in terms of knee score outcome and implant survival. Methods A retrospective-comparative study was conducted utilizing data from the Federal state’s arthroplasty registry. Included were patients from our department that undergone a conversion of a medial UKA to a TKA (UKA-TKA group). The Western Ontario and MacMaster Universities Osteoarthritis Index (WOMAC) from preoperative and 1-year postoperative was used. Moreover, the implant survival was analyzed. Results In the UKA-TKA group, there were 51 cases (age 67 ± 10, 74% women), and in the TKA group, there were 2247 cases (age 69 ± 9, 66% women). The one-year postoperative WOMAC total score was 33 in the UKA-TKA group und 21 in the TKA group (p < 0.001). Similarly, the WOMAC pain, WOMAC stiffness, and WOMAC function scores were significantly worse in the UKA-TKA. After 5 years, the survival rates were 82% and 95% (p = 0.001). The 10-years prosthesis survival was 74% and 91% in the UKA-TKA and TKA groups, respectively (p < 0.001). Conclusions Based on our findings it is concluded that patients who received a TKA after UKA have inferior results than those that directly receive a TKA. This is true for both patient-reported knee outcome and prosthesis survival. Converting UKA to TKA should not be seen as an easy operation, but should rather be done by surgeons with considerable experience in both primary and revision knee arthroplasty.


Introduction
It was reported that early failure modes for medial unicondylar knee arthroplasty (UKA) were mainly bearing dislocation, fracture, and infection, whereas the major late failure mode was progression of osteoarthritis [1], disregarding the type of UKA: cemented, uncemented or hybrid [2]. Regardless of the reasons for revision, when a UKA has to be revised to a total knee arthroplasty (TKA), it is a common discussion in the orthopedic community whether the outcome of total knee arthroplasty (TKA) is compromised by a previously implanted medial UKA.
A comprehensive literature analysis was carried out with this regard. Overall 29 original articles dealing with the subject of converting UKA to TKA were identified [3,4,34] (Table 1).
Interestingly, the majority of those studies only reported non-controlled case series of patients who's UKA were converted to TKA. Only 6 studies compared patients with UKA converted to TKA (UKA-TKA) with patients that received primary TKA [4,13,22,23,27,30]. Regarding implant survival (or cumulative revision rate) and patientreported outcome (PRO) like knee scores as strongest possible outcome parameters, it seems that neither of those studies applied both. Moreover, some of those six publications suffered from rather low sample sizes (e.g., 21 vs. 28 [13]) and others followed their patients only over relatively short periods (e.g., 4 years [2]). Overall, only one previous research group analyzed the cumulative revision rate of UKA-TKA compared to TKA [27]. In summary, the majority of publications were case series of UKA to TKA conversions (Level of Evidence 4). Those studies that compared UKA-TKA to TKA (Level of Evidence 3) suffered from the above-mentioned limitations.
Consequently, it was the aim of the current study to investigate the clinical outcome of patients that underwent conversion of a medial UKA to a TKA (UKA-TKA group). And to compare that outcome to patients that underwent primary TKA (TKA group). It was hypothesized that those groups would significantly differ in terms of knee score outcome (H1) and implant survival (H2).

Methods
The study design was retrospective-comparative. Data from the arthroplasty registry was utilized after approval by the ethics committee of the Medical University (approval No. AN2016-0207). Patients who previously underwent either primary TKA (TKA group) or conversion of a medial UKA to a TKA (UKA-TKA group) at our department were considered. Cases were excluded in the case of incomplete WOMAC data. Among those patients in the registry databank with primary TKA, those with revision implants (very complex primaries) or tumor prosthesis were excluded, therefore leaving only cruciate retaining (CR) and posterior stabilized (PS) implants for the TKA group. For patient-reported outcome measurement, the Western Ontario and MacMaster Universities Osteoarthritis Index (WOMAC) score [5] was available from the arthroplasty registry. It had been applied in the German language version [32] the day before surgery and again postoperatively 1 year after surgery.
The WOMAC questionnaire collects data on pain, stiffness, and physical function. Every item was completed on an 11-point scale and converted for analysis purposes to a scale from 0 to 100%, 0 denoting the best and 100% the worst response. The score for each of the three main dimensions is defined as the sum of all item scores divided by the number of items. The total score was defined as the sum of pain, stiffness, and function scores divided by three. Prosthesis survival data was also taken from the arthroplasty registry data bank.
For statistical analysis, Stata 13 (StataCorp. 2013. Stata Statistical Software: Release 13. College Station, TX: StataCorp LP) was used. Means and standard deviations were calculated as descriptives. Independent T-Tests were applied to test for differences in WOMAC scores between the groups. Statistical tests were always performed twotailed. Alpha was defined as 0.05. We estimated cumulative revision-free survival from date of surgery until date of revision, date of death or end of follow-up, whichever occurred first, by applying the Kaplan-Meier method. Differences in survival curves were tested using the generalized Fleming-Harrington test of equality, with parameters q and p chosen at p = 0.0, q = 0.03.

Results
In the UKA-TKA group, there were 51 cases (age 67 ± 10, 74% women), and in the TKA group, there were 2247 cases (age 69 ± 9, 66% women). For the UKA-TKA group, the reasons for revision are provided in Table 2.
In 22 cases, a cruciate retaining design was used (Stryker, Scorpio, and Triathlon CR). In 27 cases, a posterior stabilized design was used (Stryker, Scorpio, and Triathlon PS). One case needed a semi-constrained (Triathlon TS) and one case a constrained implant design (Link, Rotating Hinge, Endo Model). In the primary TKA group, the implants used were Scorpio CR, Scorpio PS, Triathlon CR, and Triathlon PS. The 1-year postoperative WOMAC total score was 33 in the UKA-TKA group and 21 in the TKA group (p < 0.001). Similarly, the WOMAC pain, WOMAC stiffness, and WOMAC function scores were significantly worse in the UKA-TKA group one year postoperatively (0.001 < p < 0.007, H1, Table 3).

Discussion
Regarding the hypotheses of the study, the most important findings were that (a) UKA-TKA provided significantly worse WOMAC scores than TKA and (b) that UKA-TKA led to significantly inferior implant survival than TKA.  Regarding the Oxford knee score 6 months postoperatively, UKA-TKA patients of Pearse et al. showed a mean score of 30 compared to a mean score of 37 in the TKA group (p < 0.001). The results from that study are in good agreement with the findings of the current study. Except for the fact that the UKA-TKA survival rate of the current study was even lower (10-years survival of 74%). To our best knowledge, the study of Pearse was the only previous study that also analyzed both patient-reported outcome and 10-years implant survival like it was done in our study.
Regarding the other five studies which previously compared UKA-TKA to TKA, none of them investigated survival rates [4,13,22,23,30]. Miller et al. retrospectively investigated 35 UKA-TKA and 100 TKA and analyzed Knee Society Score and complication rates [22]. The authors reported inferior Knee Society Score among UKA-TKA patients and also a higher complications rate. Also Jarvenpaa et al. conducted a retrospective study on 21 UKA-TKA and 28 TKA patients and collected data on complications, range of motion, WOMAC, VAS pain, and walking distance [13].
For the UKA-TKA patients, the authors reported significantly worse WOMAC pain and WOMAC stiffness, what is in good agreement with the current study.
Becker et al. retrospectively analyzed 28 UKA-TKA and 28 TKA patients [4]. The authors used the Knee Society Score and the WOMAC as patient-reported outcome and additionally also analyzed the range of motion, the insert thickness, and also radiographic parameters. Becker et al. reported significantly thicker inserts, lower range of motion, and worse knee scores among the UKA-TKA patients. The latter fact again being congruent with our findings.
The only conflicting findings come from O'Donnell et al. [23]. The authors compared 55 UKA-TKA with 55 TKA and analyzed the Knee Society Score, the range of motion, and radiographic outcome parameters. In contrast to the findings of the current study, O'Donnell et al. did not identify significant differences between the two groups and hence concluded that the clinical outcome of UKA-TKA is similar to that of TKA. Sarraf et al. reported on a large population of 374 UKA-TKA and 251,803 TKA but only investigated the height of the insert and the constraint needed [30]. The authors reported significantly higher insert thicknesses in the UKA-TKA group (10 mm vs. 12.8 mm) and also a more frequent demand of implant constraint in the UKA-TKA group (4.19% vs. 2.15%). Due to the absolute incongruence in types of outcome parameters used, that study cannot be compared with the current study. In synopsis of those four previous studies [4,13,22,23] that analyzed only patientreported outcome (but not survival) in both UKA-TKA and TKA patients, three found worse scores among UKA-TKA patients and one did not. Those findings from previous research are now supported by the findings from the current study.
Regarding reasons for converting a UKA to a TKA (types of UKA failure), the current study found pain, loosening, and progression of osteoarthritis as the major reasons. This is in perfect agreement with the 2021 reports from the arthroplasty registries of the Great Britain and Australia. The following limitations shall be acknowledged.
First, it was a retrospective study with the typical weaknesses associated with such studies: selection bias, information bias, inability to investigate parameters other than those previously collected during clinical routine, reliance on data collected by others etc.
Second, although previously suggested [33], we did not succeed in collecting physical activity data and healthrelated quality of life data in conjunction with the kneespecific WOMAC data. Third, the types of implants in the UKA-TKA group were heterogenous, both the UKA being explanted and the TKA implants used for revision. However, the latter limitation is true for all previous studies that compared UKA-TKA to TKA [4,13,22,23,27,30]. Another limitation of the study is that only those cases from the arthroplasty registry could be included who were previously operated at our institution. The same type of investigation with all cases from the registry would have been more powerful. Furthermore, it is also regarded as limitation that we cannot explain the causality of the fact that UKA-TKA had inferior outcome than TKA. It may be speculated whether this is due to the fact that for a second time a soft tissue approach has to be performed.
It may also be discussed whether the bone loss or medial instability which has to be handled during many UKA-TKA procedures is to blame for the impaired outcome. It should also be acknowledged that we only investigated TKA performed with off the shelf implants, although others suggested that better results may be achieved with more personalized solutions [35]. Finally, the lack of a power analysis is acknowledged.
It is regarded as strengths of the current study that it is the second study so far that investigated both patient-reported outcome (WOMAC) and implant survival over a period of 10 years.
The study findings are regarded as of high clinical relevance. First, the procedure of converting UKA to TKA should not be seen as an easy operation, but should rather be done by surgeons with considerable experience in both primary and revision knee arthroplasty. Second, the inferior outcome of a later conversion of UKA to a TKA should be discussed with a patient already when opting for a UKA.

Conclusions
Based on our findings it is concluded that patients who received a TKA after UKA have inferior results than those that directly receive a TKA. This is true for both patientreported knee outcome and prosthesis survival. Converting UKA to TKA should not be seen as an easy operation, but should rather be done by surgeons with considerable experience in both primary and revision knee arthroplasty.
Funding Open access funding provided by University of Innsbruck and Medical University of Innsbruck. No industry funding declared.

Conflict of interest No conflicts of interest reported.
Ethical approval Ethical approval was granted by local Ethics.committee Informed consent Not applicable, as registry data was used.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.