Allografts as alternative to autografts in primary posterior cruciate ligament reconstruction: a systematic review and meta-analysis

Purpose Following posterior cruciate ligament (PCL) rupture, autografts and allografts are routinely used for its reconstruction. This study investigated the efficacy and safety of allografts for primary PCL reconstruction, comparing them to autografts in terms of patient-reported outcome measures (PROMs), functional tests, and complications. Methods This study followed the PRISMA guidelines. PubMed, Web of Science, Google Scholar, Embase, and Scopus were accessed in October 2022. All the clinical studies investigating the outcomes of primary PCL reconstruction using allografts, or comparing the outcomes of allografts versus autografts, were accessed. The outcomes of interests were: instrumental laxity, range of motion (ROM), Telos stress radiography, drawer test, International Knee Documentation Committee (IKDC), Tegner Activity Scale, and the Lysholm Knee Scoring Scale. Data on complications were also recorded. Results A total of 445 patients were included. The mean follow-up was 45.2 ± 23.8 months. The mean age of the patients was 30.6 ± 2.2 years. The time span between the injury and surgical intervention was 12.9 ± 10 months. Overall, 28% (125 of 445 patients) were women. Good baseline comparability was found between the two cohorts. No difference was found in terms of Lysholm Score, ROM, Tegner Scale, IKDC, arthrometer laxity, drawer test, and Telos stress radiography. No difference was found in the rates of anterior knee pain and revision. Conclusion Allografts can be considered a suitable alternative to autografts for PCL reconstruction. Level of evidence III.


Introduction
The posterior cruciate ligament (PCL) provides primary stabilization for posterior tibial translation over the femur [7, 36,46,58]. The PCL is typically injured by a high-energy trauma directed posteriorly to the proximal tibia with the flexed knee, an event which typically occurs in traffic accidents or contact sports practice [37,55]. Sport-specific PCL injuries incidence ranges from 1 to 4% [3]. The PCL has better healing capability and greater strength than the anterior cruciate ligament [17,60]. Indeed, PCL tears can be managed conservatively when up to 10 mm of posterior knee translation is present [1]. However, surgical PCL reconstruction may be necessary in patients with greater laxity associated with instability [38,54]. Both autografts and allografts have been employed for PCL reconstruction [4, 46,53]. In 2018, Belk et al. [4] performed a systematic review including five studies [2,27,31,52] comparing allografts versus autografts for PCL reconstruction. Since then, further available studies have not yet been considered for systematic reviews [13,21,33,47,48,56,57,59,60].
The present study investigated the role of allografts for primary PCL reconstruction, comparing them to autografts in terms of patient-reported outcome measures (PROMs) and functional tests. It was hypothesized that allografts and autografts achieve similar outcome following reconstruction of the PCL.

Search strategy
This systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses: the PRISMA checklist [40]. The PICOT algorithm was primarily developed:

Data source and extraction
Two authors (F.M. and A.P.) independently performed the literature search in October 2022. PubMed and Web of Science were accessed. Subsequently, Google Scholar, Embase, and Scopus were accessed to identify further articles. The following keywords were used in combination: knee, posterior cruciate ligament, PCL, tears, injury, surgery, reconstruction, allograft, autograft, instability, BPTB, Achilles, hamstring, tibialis anterior, quadriceps, laxity, patient-reported outcome measures, PROMs, laxity, stability, complication, anterior knee pain, failure, revision. The same authors independently performed the initial screening of the resulting titles. If title and abstract matched, the article full-text was accessed. The bibliographies of the full-text articles were also screened. Disagreements were solved by a third author (N.M.).

Eligibility criteria
All the clinical studies investigating the outcomes of primary PCL reconstruction using allografts, or comparing the outcomes of allografts versus autografts, were accessed. Given the authors' language capabilities, articles in English, German, Italian, French, and Spanish were included. Studies of level I-IV of evidence, according to Oxford Centre of Evidence-Based Medicine [23], were considered. Editorials, cohort studies, reviews, technical notes, narrative reviews, expert opinion, and letters were excluded. Animal, biomechanics, and cadaveric studies were also excluded. Articles combining PCL reconstruction with ACL reconstruction were excluded. Articles reporting data on revision settings were not considered. Only articles reporting quantitative data under the outcomes of interest were considered for inclusion.

Data extraction
Two authors (F.M. and A.P.) independently performed data extraction. Study generalities (author, year, journal, type of study) and patient baseline demographic information (number of samples with related gender and mean age, time span between the injury and the index surgery, length of the follow-up) were collected. The following data were retrieved at last follow-up for both grafts: (1) functional tests: instrumental laxity, range of motion (ROM), Telos stress radiography, drawer test; (2) PROMs: International Knee Documentation Committee (IKDC), Tegner Activity Scale, Lysholm Knee Scoring Scale; (3) complications. The instrumental laxity was evaluated using the arthrometers KT-1000 or KT-2000 (MEDmetric Corp, San Diego, California). Both these devices applied a posterior translation force of 134 N on the tibial plateau over the femur condyles.

Methodology quality assessment
For the methodological quality assessment, the Coleman Methodology Score (CMS) was used. The CMS is widely used to evaluate the methodological quality of scientific articles for systematic reviews and meta-analyses [12]. This score analyses the included papers under several items, including study size, length of the follow-up, surgical approach, type of study, and the description of diagnosis, surgical technique, and rehabilitation. Additionally, outcome criteria assessment, procedures for assessing outcomes, and the subject selection process are also evaluated. The CMS rates articles with values comprised between 0 (poor) and 100 (excellent). A mean value greater than 60 points is considered satisfactory.

Statistical analysis
All statistical analyses were performed by the main author (F.M.). All the studies reporting the outcomes of PCL reconstruction using allografts were compared to autografts. Data from autografts were obtained from the included studies which compared between the two grafts. Continuous variables were analysed using the mean difference (MD) effect measure. The Student t test was performed, with values of P < 0.05 considered statistically significant. The odds ratio (OR) effect measure was used to investigate the rate of complications, with values of the 2 test < 0.05 considered statistically significant. Studies that directly compared allografts versus autografts which also reported measure of data dispersion (standard deviation or confidence interval) were included in the meta-analysis. The meta-analyses were performed using the Review Manager Software 5.3 (The Nordic Cochrane Collaboration, Copenhagen). The inverse variance was adopted for continuous variables, with MD effect measure. Dichotomous data were evaluated through a Mantel-Haenszel analysis, with OR effect measure. The comparisons were performed with a fixed model effect as set up. Heterogeneity was assessed through the Higgins-I 2 test. If I 2 test > 50%, high heterogeneity was detected. In cases of heterogeneity, a random model effect was used. The confidence intervals (CI) were set at 95% in all comparisons. The overall effect was considered statistically significant if P < 0.05. The funnel plot of the most commonly reported Station, USA) to assess funnel plot asymmetry, with values of P < 0.05 indicating statistically significant asymmetry.

Search result
The literature search identified 122 articles comparing allografts versus autografts for primary posterior cruciate ligament reconstruction. Of these, 30 duplicates were excluded. An additional 67 articles were excluded because of the following reasons: revision setting (n = 4), type of study (N = 50), combined with ACL reconstruction (N = 7), combined allografts and allografts (N = 6). An additional 15 articles were excluded because they lacked quantitative data under the outcomes of interest. This left 10 comparative studies for inclusion: 7 retrospective studies, and three prospective studies (Fig. 1).

Methodological quality assessment
The CMS identified some limitations and strengths in the present study. The study size and the length of the follow-up were adequate in most of studies. Surgical approach, diagnosis, and rehabilitation protocols were generally well described. Outcome measures and timing of assessment were often well defined, providing moderate reliability. The procedures for assessing outcomes and subject selection were often biased and not satisfactorily described. Concluding, the CMS for the articles was 66 points, attesting the acceptable quality of the methodological quality assessment.

Risk of publication bias
The funnel plot of the most reported outcome was performed (Fig. 2). The graph evidenced minimal asymmetry in the referral point disposition. However, the Egger's test did not evidence any statistically significant asymmetry (P = n.s.). In conclusion, the risk of publication bias was low to moderate.

Patient demographics
A total of 445 patients were included. The mean followup was 45.2 ± 23.8 months. The mean age of the patients was 30.6 ± 2.2 years. The time span between the injury and surgery was 12.9 ± 10 months. Overall, 28% (125 of 445 patients) were women. Good baseline comparability was found between the two cohorts in terms of mean age, length of the follow-up, time span between the injury and the surgical intervention. The demographics of the included studies is shown in Table 1.

Outcomes of interest
No difference was found between allografts and autografts in Lysholm Score, ROM, Tegner Activity Scale, IKDC, arthrometer laxity, drawer test, and Telos stress radiography. These results are shown in greater detail in Table 2.

Complications
No difference was found in the rate of anterior knee pain and revision. These results are shown in greater detail in Table 3.
No difference was found in Lysholm Score, ROM mean, Tegner Activity Scale, and arthrometer laxity. The forest plots of the comparisons included in the meta-analysis are shown in Fig. 3.

Discussion
The main finding of the present study is that allografts demonstrated similar outcome to autografts for PCL reconstruction.  3.5 ± 0.8 3.5 ± 1.0 0.0 n.s Telos stress radiography (mm) 9.1 ± 6.7 2.9 ± 0.1 6.2 n.s . In a biomechanical study, Achilles allografts showed more similar biomechanical characteristics of a native PCL compared to quadriceps allografts [19].
The findings of the present study must be interpreted with some limitations. The relatively small number of studies available for inclusion represents an important limitation. Further, 76% (10 of 13 included studies) were retrospective, increasing the risk of selection bias. Moreover, the analyses were conducted irrespective of the source and the strands of the grafts. The current literature lacks prospective analyses with blinding or sample randomization. The eligibility criteria and the procedure protocols were often biased and small between-study differences were evidenced. Allografts must be increased for revision PCL reconstruction, but there is no evidence yet to recommend allografts compared to autograft for primary PCL reconstruction.

Conclusion
Allografts are an effective and safe alternative to autografts in PCL reconstruction.