Better outcomes after minimally invasive surgeries compared to the standard invasive medial parapatellar approach for total knee arthroplasty: a meta-analysis

Purpose Minimally invasive surgery (MIS) for total knee arthroplasty (TKA) is often marketed as being able to speed up healing times over standard invasive surgery (SIS) through the medial parapatellar approach. The advantages of these minimally invasive approaches, however, are not yet definitively established. A meta-analysis of studies comparing peri-operative and post-operative differences and long-term complications of MIS versus SIS for TKA was conducted. Methods This meta-analysis was conducted following the PRISMA guidelines. The Pubmed, Google Scholar, Scopus, and Embase databases were accessed in September 2020. All clinical trials comparing minimally-invasive versus standard approaches for TKA were considered. Only studies reporting quantitative data under the outcomes of interest were included. Methodological quality assessment was performed using the PEDro appraisal score. Results This meta-analysis covers a total of 38 studies (3296 procedures), with a mean 21.3 ± 24.3 months of follow-up. The MIS group had shorter hospitalization times, lower values of total estimated blood loss, quicker times of straight-leg raise, greater values for range of motion, higher scores on the Knee Society Clinical Rating System (KSS) and its related Function Subscale (KSFS). Pain scores, anterior knee pain and revision rate were similar between MIS and SIS. SIS allowed a quicker surgical duration. Conclusion The present meta-analysis encourages the use of minimally invasive techniques for total knee arthroplasty. However, MIS TKA is technically demanding and requires a long learning curve. Level of evidence III, meta-analysis of clinical trials.


Introduction
The most common exposure for total knee arthroplasty (TKA) is the medial parapatellar arthrotomy (MPP) using standard incision surgery (SIS) [22]. Despite the satisfying surgical outcomes, this approach has been criticized as it produces extensive damage to the knee extensor muscle mechanism, and it may negatively affect the patellar blood supply [45]. Thus, a less invasive MPP which allowed to spare the quadriceps (quadriceps-sparing approach = QS) [30] and the limited-or mini-medial parapatellar approach (MMPP) [50]. The midvastus and subvastus [59] approaches became respectively the mini-midvastus (MMV) [15], and the mini-subvastus (MSV) [17]. MIS for TKA uses a surgical incision shorter than 14 cm, thus offering an attractive alternative for both surgeons and patients. Despite the large number of published studies comparing the MIS and 1 3 standard approaches, there is still lack of consensus concerning the best approach for TKA. Under these premises, a meta-analysis comparing MIS versus the traditional MPP SIS approach for TKA was conducted, investigating outcomes and long-term complications between the two approaches. The goal of the present study is to update current evidence and offer new insights concerning the surgical exposure to the orthopaedic surgeons.
It was hypothesised that MIS for TKA may achieve superior surgical outcomes than the MPP SIS approach.

Search strategy
This meta-analysis was performed according to the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA guidelines) [39]. The search parameters were defined as follows: (P) Population: patients requiring TKA; (I) Intervention: SIS TKA though the MPP approach; (C) Comparison: MIS TKA; (O) Outcomes: peri-operative data, functional scores, complications.
The search was performed in September 2020. The databases accessed were Pubmed, Google Scholar, Scopus, and Embase, without any limitation on time of publication. The following keywords were used: total knee arthroplasty, total knee replacement, prosthesis, combined with minimallyinvasive, medial parapatellar, mini-medial parapatellar, minivastus, subvastus, quadriceps-sparing, and further combined with anterior knee pain, revision, range of motion, scores, blood loss, surgical duration, outcomes. Two independent authors (**; **) performed the database search. If title and related abstract matched the topic, the full-text article was accessed. The bibliographies for each article of interest were screened by hand. Disagreements between the authors were debated and solved.

Eligibility criteria
Two independent authors (**; **) screened articles for inclusion. All clinical trials comparing minimally-invasive approaches for TKA to the standard approach were considered for inclusion. According to the authors' capabilities, articles in English, French, Spanish, Italian, and German were included. Only clinical trials with evidence levels I to III were considered according to the Oxford Centre of Evidenced-Based Medicine [23]. Every type of TKA (cruciate or bi-cruciate retaining, posterior stabilized) was considered eligible. No distinction was made between different MIS approaches. Studies taking advantage of a navigation system were also included. Case series, reviews and metaanalyses, editorials and expert opinions were excluded. Biomechanical, in vitro, animal and cadaveric studies were also excluded. Only studies reporting quantitative data under the outcomes of interest were included.

Outcomes of interest
Two independent authors (**, **) screened the included studies and extracted the following generalities: year, type of study, number of knees, duration of follow-up (in months), surgical approach, percentage of osteoarthritic and female patients, mean age and body mass index (BMI) (kg/m 2 ). For each approach, the following peri-operative endpoints were collected: duration of surgery and hospitalization, total estimated blood loss (intra-operative and post-operative). Functional scores included range of motion (ROM), knee flexion, time of straight-leg raise (SLR), the Knee Society Clinical Rating System (KSS) and its related Function Subscale (KSFS) [42] and the visual analogic scale for pain (VAS). Procedure-related complications, anterior knee pain and need for revision were also retrieved.

Methodological quality assessment
For methodological quality assessment, the PEDro scale was applied. This scale is a validated system for evaluating the quality of clinical trials [38]. Two authors (**; **) who already had extensive experience with this score independently evaluated each article. The PEDro scale evaluates studies based on the criteria: clearly eligibility criteria, allocation, baseline comparability, blinding, follow-up, analyses, point estimates and variability. A final mean value > 6 is considered to indicate good methodological quality.

Statistical analysis
The statistical analysis was performed by the main author (**). For the assessment of baseline comparability, the IBM SPSS Software was used. The unpaired t-test was performed, with values of P > 0.5 considered satisfactory. Statistical analyses were performed using Review Manager Software 5.3 (the Nordic Cochrane Collaboration, Copenhagen). For continuous variables, the inverse variance method with mean difference (MD) effect measure was adopted, while for 1 3 binary data, the Mantel-Haenszel method with Odds Ratio (OR). The confidence interval was set to 95% in all comparisons. A fixed effect was set as default for every comparison. Heterogeneity was assessed through the Chi-square (χ 2 ) and Higgins Tests (I 2 ). If χ 2 > 0.5, the I 2 test was evaluated. I 2 test values of 25, 50 and 75% detected respectively low, moderate and high levels of heterogeneity. If high heterogeneity was detected, a random effect model was used. The forest and funnel plot were performed to establish a visual representation of the effect measure and risk of publication bias, respectively.

Literature search
The initial literature search resulted in 2218 articles, of which 592 were duplicates. 1201 did not match the eligibility criteria and a further 371 did not report quantitative data under the outcomes of interest. Another 16 articles were excluded because of uncertain results or untrustworthy data origin. In the end, this left 38 articles for inclusion: 22 RCTs and 16 n-RCTs. Figure 1 shows the flow-chart of the literature search.

Methodological quality assessment
The mean PEDro score for included studies showed high quality at 7.3/10. The main limitation reflected in this score results from the impossibility of blinding the surgeons. This score evidenced that in 27% (10/38) of studies the follow-up was too short and 24% (9/38) did not provide any randomization of the samples. The PEDro score assigned to each study is shown in Table 1.

Risk of publication bias
To evaluate the risk of publication bias, the funnel plot of the most reported outcome (surgical duration) was performed. The plot (Fig. 2) shows a moderate symmetrical distribution of the referral points. There is adequate distribution with respect to the no-effect line. Consequently, this metaanalysis shows a moderate risk of publication bias for.

Patient demographic
In the present study, data from 3281 patients were collected. The mean follow-up time was 21.3 ± 24.3 months. 1697 patients had undergone TKA using a MIS approach, and 1584 patients using the SIS approach. There was baseline comparability between age, BMI, gender and diagnosis (P > 0.5). Study generalities and patient demographic are shown in Table 2, while Table 3 shows in detail the baseline characteristic of the two cohorts.

Outcomes of interest
The traditional SIS approach allows a shorter surgical duration (MD − 15.51; CI 9.79-21.23; P < 0.0001, Fig. 3). The MIS group was associated with a shorter hospitalization length ( The visual analogic scale, the rate of anterior knee pain and revisions (Fig. 6) were similar between the two cohorts. Table 4 shows the main results of the meta-analyses, while Table 5 displayed the complications.

Discussion
The main finding of the present meta-analysis is that MIS approaches may offer clinical and functional benefits over the conventional MPP SIS approach for TKA. Peri-operatively, these include a significantly shorter length of hospitalization and time to straight-leg raise, along with a reduction in total estimated blood loss. Overall, patients undergoing MIS TKA achieve greater ROM, flexion and KSS and KSFS scores, and the surgical procedure is shorter. Concerning complications, the two approaches yielded similar results.
MIS TKA procedures have been introduced to minimize quadriceps disruption, resulting in better quadriceps strength [19,41,44]. Furthermore, the shorter incision and limited knee arthrotomy, and avoidance of patellar eversion and dislocation and hyperflexion of the tibiofemoral joint, produce less damage to the muscles, collateral ligament, and posterior capsule. All these features may result in faster recovery [12,16]. However, given to the difficulty in execution, the longer learning curve and the need for special instruments, MIS TKA has not become very popular [2,36]. The instrumentation for MIS TKA necessitates special retractors and jigs (e.g., the sided cutting tools). These instruments require 1 3 adequate technical training. The revised sided cutting tool allow the coronal and sagittal bony cuts in one step, while, with the traditional instrumentation, two bone cuts are necessary. To assist the surgeon, the use of mobile windows can facilitate exposure of knee surfaces, and adequately trained assistants have also been recommended [26,33,52]. In addition to the new surgical instrumentation, new implants designed specifically for MIS TKA (e.g., the uncemented tibial plateau with smaller keel) have been introduced into the market, along with specific recoated stems and modular implants. These implants are designed specifically for those situations with reduced visibility of the surgical field.  [45]. Moreover, for MIS TKA, navigation systems, along with patient-specific instrumentation, recently gained popularity [21,37,61]. However, despite noticeable improvements, there are still controversial, and the reliability and feasibility, cost-effectiveness and clinical advantages of these new tools and new implants is uncertain. This has discouraged many surgeons from performing minimally-invasive TKAs, and the MPP remains the most common approach for TKA.  [5] Y N N Y N N N Y Y Y Y 6 Boerger et al. 2005 [6] Y Huang et al. 2015 [24] Y N N Y N N N Y Y Y Y 6 Jung et al. 2009 [25] Y [53] Y N N Y N N N N Y Y Y 5 Unnanuntana et al. 2012 [54] Y   The reduction in hospitalization time for MIS patients offers great potentials for cost-savings. Notwithstanding, MIS approaches require a long learning curve for the whole surgical team [1,31]. Reduced visibility of anatomical landmarks, the number of surgical steps, and the need for different equipment play a role in extending the duration of surgery. Once surgeons are more familiar with the less invasive procedures, operating times do decrease [11,26], and eventually no difference in surgical time are found after surgeons received adequate training.
The endpoint total estimated blood loss was evaluated under a random effect method, given the high grade of heterogeneity. This can be explained by the different protocols of tourniquet, drainages and antifibrinolytic agents used in the various studies. Thus, even though this endpoint resulted statistically significant in favour of the MIS group, this result must be interpreted with caution. The time to straight-leg raise is used to assess functional recovery of the quadriceps muscle after a TKA; the statistically significant reduction in time detected among the MIS group is noteworthy. The faster restoration of function of the extensor muscle mechanism may arise from to the limited knee arthrotomy and smaller incision in MIS TKA, together with the avoidance of patellar eversion. Reduced damage to soft tissues may also explain the statistically significant improvement of the analysed scores. The visual analogic scale for pain, even if not statistically significant, was remarkably lower in the MIS group. Similar consideration can be inferred also to the KSS and KSFS, which resulted statistically significant better outcomes in favour of the MIS group. A statistically significant improvement of joint motion was observed (ROM and flexion). Some studies found that MIS TKA resulted in an improvement of ROM and flexion in the early post-operative period, which disappeared after one week and three months [3,32]. During TKA performed by SIS approach, the quadriceps tendons and muscles are incised and re-sutured, resulting in scar tissues and fibrosis, which can explain the reduced joint motion. However, evidences are lacking, and future studies should investigate and compare the trend of favourable joint motion in MIS over the time. This study encourages orthopaedic surgeons to consider MIS TKAs approaches, notwithstanding the difficulties that arise from the longer learning curve. In light of the present results, further studies should investigate the best approach for MIS TKA and establish with greater stringency what the correct indications for MIS TKA are.

Limitations
Given of the high overall heterogeneity, all comparisons were analysed under a random effect method. Moreover, the funnel plot detected a moderate risk of publication bias. Articles were compared regardless to the type of pre-and peri-operative protocols. The patient anatomical characteristics, time of tourniquet, use of antifibrinolytic agents and antibiotics administration, type of technique, type of implants, use of drainages and post-operative rehabilitation and antithrombotic protocols were source of major differences which could not be adjusted statistically. Limited follow-up times represent another important limitation of this study, resulting in unreliable data concerning long-term complications and implant survivorship. A lack of distinction between MIS approaches (QS, MMV, MSV, MMPP), a reflection of the available data, represent an important limitation of this study. This was necessary to improve the amount of pooling data for inclusion. Further studies are required to investigate the pros and cons of the various approaches separately when enough suitable studies will have been published. Another important limitation of this study is the number of analysed endpoints, which was limited by insufficient data in the literature which would allow for analysis of further endpoints. Computer-assisted TKA were not considered, and this may represent another limitation. Moreover, limitations in reported durations of followup prevents a more robust analysis of long-term benefits and risks. Given these limitations, results from the present study must be interpret with caution. Strong points of the present work, on the other hand, are represented by the widespread nature of the literature search, along with the strict eligibility criteria, its methodological quality assessment, and good baseline comparability. This is crucial to provide more reliable and homogeneous results, leading to greater strength of scientific evidence.

Conclusion
MIS approaches may offer clinical and functional benefits over conventional SISI MPP approach for TKA. Perioperatively, MIS patients experience lower total estimated blood loss and a reduction in hospitalization time. Postoperatively, MIS patients demonstrate improved joint function and other outcome scores during follow-up. Minimally invasive approaches for TKA involve a technically more complicated execution which requires a long learning curve for the whole surgical team. These results must be interpreted within the limitations of the present study.