Introduction

Recurrent patellar dislocation has an annual incidence rate ranging from 5.8 to 77.8 per 100,000, with the highest incidence rate being in young and active people [1,2,3]. Failure to treat patellar dislocation can lead to patellar instability, persistent knee pain, and patellofemoral osteoarthritis eventually. Hence, appropriate treatment is needed.

Regarding patellar dislocation, the medial patellofemoral ligament (MPFL) plays a critical function in the patellofemoral joint as a primary stabilizer. Treating a patellar dislocation is challenging for orthopedic surgeons due to the complex procedures required and possible unsatisfactory results such as frequent recurrence. Although medial soft-tissue realignment surgery is the conventional treatment to medialize the patella, these procedures do not reconstruct or repair the MPFL. A rather high recurrent instability rate of 27% has been reported after medial capsule reefing [4,5,6,7].

Recent studies have indicated that MPFL reconstruction is associated with favorable clinical outcomes [8,9,10,11]. Bitar et al. [12] reported that treatment with MPFL reconstruction using a graft produced good results, based on the analyses of postoperative recurrences and the better final clinical score results. Despite previous results, when surgeons perform MPFL reconstruction using a graft, there is debate regarding graft choice, particularly on whether an autograft or an allograft should be used.

Previous studies have reported clinical outcomes of MPFL reconstruction using an autograft such as a semitendinosus, a patellar tendon, or a gracilis tendon [12,13,14,15,16]. Mikashima et al. [17] have suggested that autografts are better than allografts because they can achieve better results using an autogenous tendon without anything surpassing it in terms of autologous histocompatibility. Conversely, Hohn et al. [18] suggested that the use of an allograft can preserve autogenous tissue and may be preferable in patients with connective tissue disorder or ligamentous laxity. They found that MPFL reconstruction using allograft tissue resulted in a low risk of recurrent instability, perhaps comparable to what has been published by others who have used autograft tissue. In the same vein, some authors have reported that allograft tissues have some advantages over autografts in terms of donor-site morbidity, including loss of strength, faster recovery, decreased surgical time, and use in patients with connective tissue disorder [19,20,21]. Despite several graft-fixation methods having been used for different types of graft, no consensus has been reached about the ideal kind of graft.

To clarify these discrepancies and establish evidence for selecting graft materials for MPFL reconstruction, the purpose of this study is to review the use of an allograft or autograft in MPFL reconstruction. We hypothesized that both autograft and allograft materials would yield favorable results for MPFL reconstruction.

Materials and methods

Literature search

We used multiple comprehensive databases to find studies that reported clinical outcomes of MPFL reconstruction using an autograft or an allograft for patellar dislocation. This study adhered to the Cochrane Review Methods. Reporting was conducted in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Statement. To identify relevant studies, controlled vocabulary and free-text words described in Additional file 1 were used to search MEDLINE, EMBASE, the Cochrane Central Register of Controlled Trials, Web of Science, and SCOPUS databases between January 2000 and September 2017. Due to the recent development of surgical techniques and equipment, past research results that are too old may have a heterogeneous effect on recent research results. Thus, only studies after the year 2000 were included and analyzed. All relevant studies were identified regardless of language, publication type (article, poster, conference article, instructional course lectures, etc.), publication journal, or publication year. This search was updated in September 2017, including reference lists of studies and any review articles identified. Reference lists of the investigated studies were scrutinized to identify any possible additional publications not found through electronic or manual searches. In cases of two or more studies by the same author, we determined whether patients had been “duplicated.” If duplicated, only the latest study was included.

Eligibility criteria

Studies were included in our investigation according to the following eligibility criteria: (1) subjects were humans who had received MPFL reconstruction using an autograft or an allograft, (2) studies that evaluated clinical outcomes of MPFL reconstruction, and (3) researchers conducted level-I, -II, -III, or -IV evidence studies. Studies were excluded if they did not evaluate the effect of surgical technique, focused on revision surgery, included patellar dislocation after total knee arthroplasty, had subjects with congenital disease, or connective tissue disorders, only reported non-clinical outcome measures or intra-operative measures, consisted of level-V evidence (case report, technical note, and letters to editor), were review articles, animal studies, or in vitro studies. Detailed criteria are summarized in Table 1.

Table 1 Inclusion and exclusion criteria
Full size table

Data collection and analysis

Two authors independently assessed the titles or abstracts of studies identified with the search strategy. Subsequently, a full paper review was conducted for the final inclusion. Uncertainty regarding the study inclusion was resolved through discussion and consensus. Data were extracted by authors using predefined forms. They were then checked for accuracy. We extracted data of study characteristics and patient demographics (Table 2). Clinical outcomes, such as the Kujala score (mean and standard deviation (SD) of preoperative and postoperative score), Lysholm score, Tegner score, redislocation rates (at final follow-up), instability episodes, subjective results, reoperation rates, range of motion (ROM), and perioperative complications, are revealed in Table 3.

Table 2 Characteristics of the included studies on medial patellofemoral ligament reconstruction for patellar dislocation using an autograft versus an allograft
Full size table
Table 3 Clinical outcomes of the included studies on medial patellofemoral ligament reconstruction for patellar dislocation using autograft versus allograft
Full size table

Assessment of methodological quality

Two investigators independently assessed the methodological quality of each study using the Coleman methodology score [36]. Each study was assessed using 10 methodological criteria, resulting in a final score ranging from 0 to 100. A perfect score of 100 indicated a study design that avoided the influence of chance, various biases, and confounding factors. Each author scored the methodological quality of each study twice, with a 10-day interval between assessments. Any disagreement between authors was resolved through discussion or review by a third investigator.

Results

Study identification

A total of 2151 relevant articles were initially identified. Of these, 432 were duplicates or published before the year 2000 in these databases. After screening the remaining 1719 articles using titles and abstracts, all but 34 were excluded because they were not relevant to the purpose of the present study. A full-text review of these 34 articles resulted in the exclusion of 12 articles because they did not meet the inclusion criteria. The remaining 22 clinical studies were included for data extraction and systematic review (Fig. 1) [9, 11,12,13,14,15,16,17, 22,23,24,25,26,27,28,29,30,31,32].

Fig. 1
figure 1

Flow diagram of the Preferred Reporting Items of Systematic Reviews and Meta-Analyses (PRISMA)

Full size image

Quality of included studies

The mean modified Coleman methodology score of these included studies was 78.1 ± 8.2 (range, 66 to 100). The results of the mean Coleman methodology score for each criterion are shown in Table 4.

Table 4 Overall Coleman methodology score for each criterion
Full size table

Data abstraction (qualitative analysis)

Medial patellofemoral ligament reconstruction using an autograft

Kujala scores

Among 21 studies on MPFL reconstruction with autograft, 20 studies [9, 12,13,14,15,16,17, 22, 24,25,26,27,28,29,30,31,32,33,34,35] evaluated the Kujala score as a primary clinical outcome. Five randomized controlled trials (RCTs) [12, 24,25,26, 35] and 15 retrospective studies [9, 13,14,15,16, 22, 27,28,29,30,31,32,33,34] reported the Kujala score in MPFL reconstruction with an autograft, consisting of a total of 698 subjects. The reported range of postoperative mean Kujala score was from 80.5 to 96.0 points. There were significant differences between preoperative and postoperative Kujala scores in all 20 studies. Regarding surgical techniques, Wang et al. [33] found that double-bundle (DB) MPFL reconstruction showed better outcomes compared to single-bundle (SB) MPFL reconstruction. Kang et al. [24] reported that a Y-shape graft technique had favorable outcomes compared to a C-shape graft technique. Conversely, Niu et al. [26] and Zhao et al. [35] reported that MPFL reconstruction had significantly favorable Kujala scores compared to medial soft-tissue realignment surgery. However, Astur et al. [15] reported that there were no statistically significant differences in Kujala score between the endobutton and anchor fixation groups. Han et al. [16] reported that the results of the Kujala score were not associated with the presence of cartilage lesion, or sex.

Patellar instability (redislocation or subluxation)

Of 21 studies (714 subjects) on MPFL reconstruction with an autograft, only three studies [29, 33, 35] reported patellar redislocation after surgery. Redislocation occurred in 10 (1.4%) patients. Wang et al. [33] reported that patellar redislocation occurred more frequently in SB MPFL reconstruction compared to that in DB MPFL reconstruction. Although patellar redislocation did not occur, six studies [11, 13, 17, 22, 25, 30] reported that the persistent apprehension sign remained in their patients (10 patients, 1.4%).

Subjective results

Various clinical evaluation tools were used to investigate the subjective results after MPFL reconstruction using an autograft. For patients who underwent surgery, the percentage of good or excellent satisfaction ranged from 71.4 to 100.0% [9, 11, 12, 24, 25, 27, 28, 30, 31, 33]. Ma et al. [25] found that there were no significant differences in subjective questionnaire scores between medial retinaculum plasty and MPFL reconstruction with autograft groups. In terms of graft type, Kang et al. [24] reported a good or excellent rate of 97.5% in the Y-shape graft group compared to 83.3% in the C-shape graft group with significant difference.

Perioperative complications

Among 12 studies [9, 11, 15,16,17, 25,26,27, 29, 31,32,33] that dealt with perioperative complications, three [26, 27, 32] reported no perioperative complications after MPFL reconstruction with an autograft. Furthermore, six studies [9, 15, 16, 25, 29, 31] reported postoperative arthrofibrosis or limitations in the ROM. Flexion deficit was particularly prominent after the surgery. However, extension deficit was not found. Mikashima et al. [17] reported that there were two cases of patellar fracture in patients using an autograft. There were no infections or vascular problems such as deep vein thrombosis. However, one study [33] reported two cases of superficial wound infection.

Medial patellofemoral ligament reconstruction using allograft

clinical evaluation scales

MPFL reconstruction using allografts was also subjected to qualitative analysis. To evaluate clinical outcomes after MPFL reconstruction using allografts, only one study [23] was included. Using clinical knee evaluation scales, such as the KOOS (Knee Injury and Osteoarthritis Outcome Score), Lysholm, Tegner, and VR-12 (Veterans RAND 12-Item Health Survey), Dragoo et al. [23] have investigated whether MPFL repair is superior to MPFL reconstruction using a semitendinosus allograft. They found that there were no statistically significant differences in clinical outcomes between the two techniques. Thus, they concluded that MPFL repair or reconstruction with an allograft might lead to clinically acceptable results at 2-year follow-up.

Perioperative complications

One study reported perioperative complications after MPFL reconstruction with an allograft. Dragoo et al. [23] reported that, despite one report of postoperative recurrent dislocation in their MPFL repair cohort with a recurrence rate of 4%, there were no recurrent dislocations in any patients initially treated with MPFL reconstruction. Furthermore, there were no other surgical complications, including stiffness, infections, painful metalwork, or wound problems at the final follow-up.

Discussion

In the present study, we assessed evidence from clinical studies that evaluated treatment outcomes after MPFL reconstruction using autograft or allograft materials. Although direct comparative studies were unavailable, the Kujala score and subjective results from the majority of studies indicated that an autograft for MPFL reconstruction yielded satisfactory clinical outcomes after MPFL reconstruction. However, no new outcome has been drawn from the use of allografts. The present study showed low rates of occurrence of perioperative complications in both groups. Furthermore, the rate of postoperative patellar instability was low at about 2.8%, and this value is similar to the pooled estimated value of postoperative redislocation rate observed in a previous review [37]. The results of the present systematic review partly supported our hypothesis that either autograft or allograft materials would yield favorable results for MPFL reconstruction. However, due to insufficient data description, direct comparison between both groups was not performed; thus, which technique yields better improvements in clinical outcome for MPFL reconstruction remains inconclusive.

Although many studies have investigated graft materials after anterior cruciate ligament (ACL) or posterior cruciate ligament (PCL) reconstruction, direct comparisons of clinical outcomes after MPFL reconstruction with autograft versus allograft are rarely reported. Only one study performed a direct comparison of an autograft versus an allograft for MPFL reconstruction [38]. However, that study was not included in the present review because it did not satisfy our inclusion due to the short-term follow-up period. In that study, Calvo Rodriguez et al. [38] reported that one patient received revision surgery due to poor positioning of the anchors. Furthermore, one patient had a non-displaced patellar fracture related to the bone tunnel and another patient had a flexion deficit. These three patients had received an allograft for MPFL reconstruction. Although three cases of perioperative complications occurred in their subjects, recurrent dislocations or graft-related complications were not observed. Ultimately, there were no significant differences in clinical outcomes between the two groups. Unlike that study, the present study did not conduct a direct comparison for MPFL reconstruction using autograft versus allograft. However, according to Sillanpaa et al.’s classification [39], almost all studies reporting the Kujala score were classified in the “good” category (85–94 points) for both groups. The results of the present study are similar to those of Calvo Rodriguez et al. Both studies revealed that MPFL reconstruction using both grafts had a favorable clinical outcome. To strengthen the evidence of these results, prospective (high-quality large-scale) comparative studies with similar clinical conditions are encouraged.

There is critical debate regarding the various surgical procedures concomitantly performed with MPFL reconstruction considering numerous predisposing factors, such as trochlear dysplasia, patellar height, graft types, rotational abnormalities of the tibia and femur, and the anterior tibial tuberosity to trochlear groove (TT-TG) distance [40]. To evaluate one independent factor, removing all confounding factors is ideal to reduce the risk of bias. For this reason, some authors have intentionally removed these confounding variables from consideration by narrowing their inclusion criteria [40]. However, strict control of all confounding factors affecting clinical outcomes is limited in practice. This concept is associated with “effectiveness” (heterogenous, more practical, “real-world”) studies in normal clinical conditions likely encountered in a real clinical trial [41]. Hence, the findings of the present study should be interpreted with great caution because the data involved were extracted from somewhat heterogenous studies. Besides, concomitant surgeries, such as lateral retinacular release and tibial tuberosity transfer, might increase surgery-related complications. Similarly, Buckens et al. [42] have considered that the heterogeneity of their series, with different concomitant procedures, might underestimate the real success of MPFL reconstruction. As such, our results imply that isolating MPFL reconstruction using autografts or allografts might produce more satisfactory results. If the authors want to focus on the “efficacy” (homogenous subjects, interventions, comparators, and outcome measures), future investigations should aim to establish more uniform criteria for selecting patients to undergo this procedure.

Based on the Coleman scales to assess the methodological quality, almost all the criteria in each study revealed a higher score. However, major sections of methodological deficiencies remained, including study size and type of the study. Theoretically, large-scale prospective studies would provide the rigorous control of potentially confounding factors. Thus, the present study critically appraised and synthesized the available evidence on this topic to provide a conclusion to a debatable issue. Further prospective studies are needed in the future to address methodological limitations. Screening and data extraction of the present study were carried out by two independent reviewers. This is one strength of our study. Although several recent systematic reviews have focused on ACL or PCL reconstruction with either an autograft or an allograft, less is known regarding autograft versus allograft for MPFL reconstruction. This study provides valuable evidence in support of MPFL reconstruction using an autograft or an allograft.

Despite its strengths, our study has some limitations. First, a relatively small number of prospective studies were included on each topic in our systematic review. There are few previously published original prospective studies with low risk of bias on this topic which is an absolute limitation. A review that is based on low-quality studies can affect conclusions. Second, in addition to demographic factors such as sex, age, and weight, technical factors regarding surgical methods also need to be controlled, including the transpatellar tunnel technique or non-transpatellar tunnel technique, various graft types, and fixation methods because they might affect the results following MPFL reconstruction. Third, we did not fully consider concomitant procedures that could affected clinical outcomes, such as tibial tuberosity transfer, lateral retinacular lengthening, or trochleoplasty. In other words, the methodologies of the studies included here are different from each other; they have heterogeneity. Due to such heterogeneity and the absence of direct comparative studies, we could not compare these two graft materials using statistical methods or conclude which graft material was better.

Conclusions

Although many studies showed favorable clinical results for MPFL reconstruction using an autograft, the clinical results of MPFL reconstruction using an allograft have not yet been sufficient to achieve a meaningful clinical result due to low evidence. Direct comparisons were not conducted because there were very few studies on allografts; thus, further research in this area should be performed in the future.