Backgrounds

Knee osteoarthritis (KOA) is a prevalent and chronic joint disease, most commonly observed in middle-aged and elderly people [1]. For end-stage unicompartmental osteoarthritis of the knee, uncompartmental knee arthroplasty (UKA) is a successful surgical treatment [2,3,4,5,6]. The benefits of UKA over total knee arthroplasty (TKA) include improved knee range of motion, rapid functional recovery, more natural steps, improved kinematics, and fewer complications [7,8,9]. At present, two methods of fixation are employed for UKA-inserted prostheses: cementless and cemented. The two types of prostheses differ in terms of complications, postoperative efficiency, operating procedures, and prosthesis design [10]. For an extended period, the predominant fixation method for knee arthroplasty has been cemented Oxford UKA [11]. However, cemented prostheses present challenges in the form of wear particles and bone cement residue, which can potentially lead to prosthesis loosening and failure [12, 13]. As a consequence of advances in material science, cementless prostheses are now frequently employed for hip replacements with highly favourable outcomes. Furthermore, they have begun to be used with considerable success for UKA of the knee joint in recent years [14, 15]. In light of the aforementioned considerations, it is evident that there is currently no definitive answer as to which prosthesis is best suited for UKA surgery. Despite the initial fixation in the case of cemented Oxford UKA being accurate the operation is slightly complicated and requires a high level of surgical skill. Problems such as inadequate cement insertion, excessive thickness of cement preventing good seating of the prosthesis, and incomplete cement removal resulting in debris formation can lead to intra-articular impingement, wear and tear, loosening, and unexplained pain, ultimately resulting in the failure of the prosthesis. Cementless prostheses are relatively straightforward to implant; however, if not performed correctly, there is a the risk of periprosthetic fractures or early prosthesis loosening. In contrast, the cementless Oxford UKA is more effective at integrating with human bone tissue, enabling the bone to grow into the prosthesis’ porous titanium and hydroxyapatite coatings, which stabilises it in place. This circumvents the issues associated with cemented fixation and extends the prosthesis’ life span [16, 17]. Furthermore, enhanced bone development into the prosthesis and bone may be achievable due to the absence of shear between the prosthesis and the bone [18]. Nevertheless, a significant debate persists regarding the optimal choice between cementless Oxford UKA and cemented Oxford UKA. According to Campi et al., cementless prosthetic fixation represents a secure and advantageous alternative to bone cement. Furthermore, the findings revealed that non-cemented prostheses demonstrate comparable clinical outcomes, failure rates, and reoperation rates to cemented prostheses. However, they exhibited a lower incidence of periprosthetic radiolucent lines and shorter surgery times [19]. Nevertheless, some scholars proposed an alternative viewpoint, with studies such as Bruni et al. [20]. Consequently, they have recommended the use of standard bone-cemented prostheses. Therefore, in order to establish an evidence-based rationale for UKA surgery, we sought to evaluate the efficacy and prognosis of the two Oxford UKA prostheses through a meta-analysis.

Materials and methods

Inclusion and exclusion criteria

Inclusion criteria: (1) All patients presenting with osteoarthritis of the medial compartment of the knee and an intact anterior cruciate ligament were included in the study; (2) The interventions were cementless UKA and cemented UKA; (3) The outcome indicators included: surgery time, OKS scores, KSS clinical scores, KSS functional scores, the presence of periprosthetic translucency lines, the rates of revision and reoperation, the five-year prosthesis survival rates, the complication rate, etc. (4) In the event that multiple reports included the same patients, the literature with the longer follow-up period was selected for inclusion.

Exclusion criteria: (1) reviews, case reports, and meta-analysis types of literature; (2) patients who had undergone high tibial osteotomy or UKA revision surgery; (3) duplicated published literature; (4) literature with inconsistent outcome indicators; (5) literature with the unavailability of the full text; and (6) animal, cadaver, and basic studies. Outcome metrics: operative time, OKS score, KSS Clinical Score, KSS Functional Score, the presence of periprosthetic radiolucent lines, the revision rate, the reoperation rate, the five-year prosthesis survival rate, and the complication rate.

Search strategy

A comprehensive search of the literature was conducted using the following databases: Pubmed, The Cochrane Library, China National Knowledge Infrastructure (CNKI), Embase, Web of Science, MEDLINE, and other relevant sources.

The search terms included “unicompartmental knee arthroplasty”, “unicompartmental knee replacement”, “UKA”, “UKR”, “cemented”, “cementless”, “uncemented”, and “knee osteoarthritis”. The search strategy employed was as follows: ([“cemented” OR “cementless” OR “uncemented”]) AND ([“UKA” OR “UKR” OR “unicompartmental knee arthroplasty” OR “unicompartmental knee replacement”]) AND ([“osteoarthritis, knee”]).

Data extractions

The data were extracted from the original literature by two independent researchers. The data was extracted by a systematic review of the titles and abstracts of the literature, with the full text of any relevant articles read in full if necessary. The extracted data included the following information: (1) authors; (2) publication date; (3) location of the study; (4) sample size; (5) mean age; (6) gender; and (7) outcome indicators, namely surgery time, OKS scores, KSS clinical scores, KSS functional scores, periprosthetic radiolucent lines, revision rates, reoperation rates, prosthesis survival, and complication rates. In the event of a discrepancy in the processing of the data, the decision was referred to a third investigator (the senior chief medical officer).

Evaluation of the quality of literature

The quality of the included literature was evaluated by two independent researchers. In the event of a discrepancy, the decision was referred to a third researcher (a senior chief medical officer) for resolution. The quality of the included randomised controlled trials was evaluated using the Cochrane Risk of Bias (ROB) assessment tool. The evaluation was conducted in accordance with seven main criteria: randomized sequence generation, allocation concealment, blinding of patients and physicians, blinding of outcome assessors, incomplete outcome data, selective reporting of study results, and other biases. The quality of the studies was also evaluated using the Newcastle-Ottawa scale, with a total score of 9. The studies were evaluated in three areas: study subject selection, between-group comparability, and exposure factor measurement. A maximum of two points could be awarded for between-group comparability, and a maximum of one point could be awarded for the remaining 7 entries, with scores ranging from zero to nine. The scores were divided into three levels, with scores < 5 being low-quality studies that could not be included in the analysis, 5–7 being moderate-quality studies, and ≥ 8 being high-quality studies.

Statistical analysis

A meta-analysis of the statistical data was conducted using RevMan 5.4 software. The odds ratios (ORs) and their 95% confidence intervals (CIs) were employed to calculate dichotomous variables, whereas the mean differences (MDs) and their 95% confidence intervals were used to calculate continuous variables. The observed differences were deemed statistically significant at a probability level of P < 0.05. The results of the individual studies were tested for heterogeneity using I2. When I2 ≤ 50%, indicating less heterogeneity between groups, a fixed-effect(FE) meta-analysis was performed, and when I2 ≥ 50%, indicating higher heterogeneity between groups, random-effects(RE) meta-analysis was conducted. Subgroup analysis and sensitivity analysis were employed to explore the source of heterogeneity.

Results

Literature search and selection processes

A total of 1,419 pieces of literature were obtained through the search. Duplicates were excluded: 108. Following a review of the titles and abstracts, 1293 papers were excluded as they did not fit the topic of the article; Eighteen papers were identified as meeting the topic criteria; Six papers were excluded after a comprehensive review of the article content and the outcome indicators; The data from 2558 patients across 12 papers were included in the study (Fig. 1).

Fig. 1
figure 1

Flow diagram of study identifcation and selection

Full size image

Basic characteristics of the included studies

Prior to surgery or treatment, there was no statistical difference in the sample size, mean age, or sex ratio between the two groups before surgery or treatment (Table 1). All included literature met the predefined inclusion criteria.

Table 1 Characteristics of included literature studies
Full size table

Inclusion of study quality evaluation results

The study included a total of 12 pieces of literature, comprising 5 randomized controlled trials, 3 prospective studies, and 4 retrospective studies. The results of the NOS scores indicated that there were 5 high-quality studies, 7 medium-quality studies, and no low-quality studies (Table 1). The literature quality assessment graph is shown in Figs. 2 and 3.

Fig. 2
figure 2

Risk of bias assessment of included studies

Full size image
Fig. 3
figure 3

Risk of bias assessment of included studies

Full size image

Meta-analysis results

Surgery time

Four studies were conducted to compare the surgical time for UKA using two types of prostheses (Fig. 4), with data from 516 patients included in the analysis. There was no heterogeneity between the studies (I2 = 0%, P = 0.99), so a fixed effects model was used. The results demonstrated that the operative time was significantly higher with cemented Oxford UKA than with cementless Oxford UKA [MD = 9.91, 95% CI (7.64, 12.17), P < 0.00001].

Fig. 4
figure 4

Forest plot of surgery time

Full size image

OKS scores

We conducted a comparative analysis of the baseline OKS scores of the two prostheses (Fig. 5), which revealed no statistically significant difference between the baseline scores for the two prosthesis types. A total of nine studies compared the OKS scores for UKA using two prosthesis types (Fig. 6), encompassing data from 1,281 patients. Of these, 658 were in the cemented group and 623 in the cementless group. There was no heterogeneity between the studies (I2 = 0%, P = 0.82), so a fixed effects model was used. The results demonstrated that in terms of postoperative OKS scores, cemented Oxford UKA exhibited significantly lower outcomes than cementless Oxford UKA [MD=-1.58, 95% CI (-2.30, -0.86),P < 0.0001].

Fig. 5
figure 5

Forest plot of baseline scores for OKS scores

Full size image
Fig. 6
figure 6

Forest plot of oxford knee score

Full size image

KSS clinical scores

A comparison of the baseline scores of the KSS clinical scores for the two types of prostheses (Fig. 7), revealed no significant difference between the baseline scores of the two types of prostheses. A total of five studies compared the KSS clinical scores of UKA using two types of prostheses (Fig. 8), which included data from 922 patients. Of these, 479 were in the cemented group and 443 in the cementless group. The heterogeneity between studies was low (I2 = 27%, P = 0.24), thus a fixed effects model was employed. The results demonstrated that, with regard to postoperative KSS clinical scores, cemented Oxford UKA was significantly inferior to cementless Oxford UKA [MD=-1.81, 95% CI (-3.26, -0.37), P = 0.01].

Fig. 7
figure 7

Forest plot of baseline scores for KSS scores

Full size image
Fig. 8
figure 8

Forest plot of knee society knee score

Full size image

KSS function scores

A comparison of the baseline scores of the KSS function scores of the two prostheses (Fig. 9), revealed no significant difference between the baseline scores of the two types of prostheses. Three studies compared the KSS functional scores of UKA using two types of prostheses (Fig. 10), comprising data from 796 patients. A total of 415 patients were included in the cemented group and 381 patients were included in the cementless group. No heterogeneity was observed between studies (I2 = 0%, P = 0.82), thus a fixed effects model was employed. The findings indicated that, in regard to postoperative KSS functional scores, cemented Oxford UKA exhibited a significantly lower outcome than cementless Oxford UKA [MD=-1.72, 95% CI (-3.27, -0.17), P = 0.03].

Fig. 9
figure 9

Forest plot of baseline scores for KSS scores(function score)

Full size image
Fig. 10
figure 10

Forest plot of knee society knee score(function score)

Full size image

Periprosthetic radiolucent lines

A total of five studies were conducted to compare the incidence of postoperative periprosthetic radiolucent lines in UKA using two types of prostheses (Fig. 11). The studies included data from 297 patients. Of the total number of cases, 149 were in the cemented group, with a total of 54 instances of postoperative periprosthetic radiolucent lines. In the cementless group, comprising 148 cases, 22 cases exhibited postoperative periprosthetic radiolucent lines. A high degree of heterogeneity was observed between studies (I2 = 71%, P = 0.008), necessitating the use of a random effects model. With regard to the occurrence of postoperative periprosthetic radiolucent lines, the results showed that cemented Oxford UKA was significantly more prevalent than cementless Oxford UKA [OR = 3.62, 95% CI (1.08, 12.13), P = 0.04].

Fig. 11
figure 11

Forest plot of radiolucent lines

Full size image

Revision rates

A total of five studies were conducted to compare the postoperative revision rates of UKA using two types of prostheses (Fig. 12). The studies included data from 2106 patients. Of these, 1,012 were in the cemented group, with a total of 56 revisions performed. The cementless group comprised 1094 cases, with 28 cases undergoing revision. No heterogeneity was observed no heterogeneity between studies (I2 = 0%, P = 0.76), and thus a fixed-effects model was employed. With regard to the postoperative revision rate, the findings indicated that cemented Oxford UKA exhibited significantly higher incidence than cementless Oxford UKA [OR = 2.22, 95% CI (1.40, 3.53), P = 0.0007].

Fig. 12
figure 12

Forest plot of revision rates

Full size image

Reoperations

A total of four studies were conducted to compare the postoperative reoperation rates for UKA using two types of prostheses (Fig. 13), which included data from 1818 patients. Of these, 869 were in the cemented group, with a total of 33 instances of postoperative reoperation. In the cementless group, 949 cases were observed, with 30 instances of reoperation occurring post-operatively. The heterogeneity between studies was low (I2 = 14%, P = 0.32), and thus a fixed-effects model was employed. With regard to the postoperative reoperation rate, the findings indicated no statistically significant difference between cemented Oxford UKA and cementless Oxford UKA [OR = 1.21, 95% CI (0.73,1.99), P = 0.46].

Fig. 13
figure 13

Forest plot of reoperations

Full size image

Five-year prosthesis survival rate

A total of four studies were conducted to compare the postoperative five-year prosthetic survival rates for UKA using two types of prostheses (Fig. 14). The studies included data from 325 patients. Of these, 159 were in the cemented group, with a five-year prosthetic survival rate of 92.4%. The cementless group comprised 153 cases with a five-year prosthesis survival rate of 94.8%. The heterogeneity was observed between studies (I2 = 0%, P = 0.56), and thus a fixed-effects model was employed. With regard to five-year postoperative prosthetic survival, the findings indicated no statistically significant difference between cemented Oxford UKA and cementless Oxford UKA [OR = 0.76, 95% CI (0.30,1.90), P = 0.56].

Fig. 14
figure 14

Forest plot of prosthetic survival rates

Full size image

Complication rates

A total of six studies were conducted to compare the postoperative complication rates of UKA using two types of prostheses (Fig. 15), which included data from 1,098 patients. Of these, 559 were in the cemented group, with a complication rate of 8.2%. The cementless group comprised 539 cases, with a complication rate of 6.1%. No heterogeneity was observed between studies (I2 = 0%, P = 0.75), and thus a fixed-effects model was employed. With regard to postoperative complications, the findings indicated no statistically significant difference between cemented Oxford UKA and cementless Oxford UKA [OR = 1.36, 95% CI (0.85,2.16), P = 0.20].

Fig. 15
figure 15

Forest plot of complications

Full size image

Publication bias analysis and sensitivity analysis

The RevMan 5.4 software was employed to analyze publication bias for nine outcome indicators, including surgery time, OKS score, KSS Clinical Score, KSS Functional Score, periprosthetic radiolucent line, revision rate, reoperation rate, five-year prosthesis survival rate, and complication rate. The resulting funnel plot exhibited basic symmetry (Fig. 16), indicating the absence of significant publication bias. Significant heterogeneity was observed in the studies that examined the inclusion of periprosthetic radiolucent lines. Consequently, we conducted a sensitivity analysis, whereby studies were excluded one by one. However, this did not result in a reduction of the observed In contrast, no significant heterogeneity was identified in the other groups of studies. Furthermore, the heterogeneity remained unchanged after performing the sensitivity analysis. The combined effect size remained statistically significant, and the forest plot structure did not change significantly. This suggests that the results of the meta-analysis performed in this study were stable.

Fig. 16
figure 16

Publishing bias funnel plots

Full size image

Discussion

UKA represents an efficacious treatment for end-stage knee osteoarthritis, exhibiting a lower propensity for injury and a reduced incidence of postoperative complications relative to TKA. The number of UKA has witnessed a notable surge in recent years. The initial iteration of Oxford UKA was predominantly reliant on cemented fixed prostheses. However, in recent years, cementless fixed prostheses have demonstrated promising outcomes in UKA. It is therefore necessary to analyze the different prosthesis fixation methods in Oxford UKA. The primary objective of this study was to assess the efficacy and prognosis of the two fixation modalities by comparing the surgical time, OKS score, KSS clinical score, KSS functional score, presence of periprosthetic radiolucent lines, revision rate, reoperation rate, five-year prosthesis survival rate, and complication rate. The findings provide a framework for clinicians to select the most appropriate prosthesis for clinical treatment.

The findings of this study indicate that, with regard to surgical time, the cemented Oxford UKA prosthesis required a significantly longer surgical procedure than the cementless Oxford UKA group. This is primarily attributable to the exclusion of cement fixation process was omitted in the cementless Oxford UKA group, which rendered the procedure more straightforward. Furthermore, the use of bone cement was identified as a potential risk factor for deep vein thrombosis following artificial arthroplasty, with an increased risk of 8.89 times that observed in the cementless group. The cementless Oxford UKA group may have a potential advantage in terms of blood clot formation due to the shorter surgery times and absence of cementation risks. With regard to the knee OKS score, KSS clinical score, and KSS functional score, cementless Oxford UKA exhibited significantly higher values than cemented Oxford UKA. This may be attributed to the cement interface, where dislodged cement particles have the potential to cause joint swelling, pain, and synovitis. Furthermore, the presence of osteitis and large cement particles necessitates arthroscopic removal. Additionally, the presence of excess cement may irritate the soft tissues or affect the tension of the soft tissues, resulting in pain or limitation of joint movement. cause pain or limitation of joint movement [31, 32]. The cemented Oxford UKA group exhibited a higher incidencez of postoperative periprosthetic radiolucency lines and postoperative revision rates than the cementless Oxford UKA group. The present study revealed a high degree of heterogeneity in the periprosthetic radiolucency linesz. It is postulated that this may be attributed to the inclusion of both partial and complete periprosthetic radiolucency lines, that were taken as the final outcome in some of the studies. The majority of partial periprosthetic radiolucency lines are located within 2 mm and are, therefore, susceptible to observational bias. However, it is important to differentiate them from physiologic radiolucency lines. Complete periprosthetic radiolucency represents an early indication of aseptic loosening and is diagnosed when the patient also presents with knee pain. In a study by Pandit, it was demonstrated that some radiolucent lines around the cementless prosthesis disappear within a year post-surgery. This indicates that the bone growth that occurs fills the void, suggesting that with the cementless Oxford UKA, even if there is a slight lack of fixation at the time of fitting, the bone growth at a later stage can compensate for minor surgical technical deficiencies [22]. It was observed that once the cementless Oxford UKA had effective bone ingrowth, effective fixation could be ensured in the following years. However, it is important to note that radiolucent line around the cementless Oxford UKA prosthesis needs to be continuously observed. This may explain the significantly lower incidence of periprosthetic radiolucent lines and the lower revision rate observed for cementless Oxford UKA compared to cemented Oxford UKA. With regard to the five-year survival of the prosthesis, no significant difference was observed between the two fixation modalities. It would be beneficial for future studies to focus on more long-term prosthesis survival over a longer period of time. With regard to postoperative reoperation and complication rates, no significant difference was observed between the two fixation modalities. It is noteworthy noting that this study is not the first meta-analysis to have been conducted comparing the outcomes of cemented Oxford UKA with those of cementless Oxford UKA in the treatment of medial knee osteoarthritis. In a study by Ma [10], no significant difference was observed in postoperative knee OKS scores or KSS scores between cemented Oxford UKA and cementless Oxford UKA. This is contrary the findings of our study, which demonstrated that the postoperative knee OKS score and postoperative KSS scores of the cementless Oxford UKA group were significantly higher than those of the cemented Oxford UKA group. As in our study, the surgical time for cementless Oxford UKA was significantly shorter than that for cemented Oxford UKA, while cementless Oxford UKA exhibited a lower revision rate and fewer periprosthetic radiolucent lines.

This systematic review and meta-analysis presents a comprehensive compilation of clinically controlled studies on cementless Oxford UKA and cemented UKA to date. The findings indicate that the cementless Oxford UKA is associated with a shorter surgical time, superior postoperative knee OKS scores, KSS clinical scores, and KSS functional scores, a reduced incidence of periprosthetic radiolucent lines, and a lower revision rate compared to the cemented Oxford UKA. It is therefore evident that the prospect of cementless Oxford UKA is even more promising. It should be noted that the present study is not without limitations. For instance, the number of studies included in the analysis was relatively low for some of the outcome indicators, such as the KSS functional score, the reoperation rate, and the five-year prosthesis survival rate. Furthermore, the current study was limited to short- and medium-term postoperative follow-up, and most of the included studies were prospective and retrospective, which may affect the quality of the research. Also, the sample sizes of some of the included studies were small, which may affect the statistical efficacy. Therefore, future studies need more high-quality RCT studies with large samples, high quality, and long-term follow-up to validate the conclusions.

Conclusion

In comparison to cemented Oxford UKA, cementless Oxford UKA demonstrated a reduction in surgical time, an improvement in knee OKS functional score, KSS clinical score, and KSS functional score, as well as a reduction in the incidence of periprosthetic radiolucent lines, and the revision rate. However, further research is required to validate these conclusions with the inclusion of additional high-quality RCT studies, particularly those with long-term follow-up.