Introduction

Anterior cruciate ligament (ACL) rupture is one of the most common knee injuries affecting more than 200,000 patients every year [1, 2]. Arthroscopic ACL reconstruction is recommended to restore knee stability in young patients, particularly those hoping to return to pivoting sports, and more than 120,000 procedures are performed every year in United States [3, 4].

National web-based registries revealed that the most commonly used technique to perform ACL reconstruction is the single-bundle [5,6,7]. However, different strategies for ACL femoral tunnel drilling have been described, including the arthroscopic transtibial (TT) technique, anteromedial (AM) portal technique, outside-in, and retrograde drilling [8].

The discrepancy between the tunnel position and the native ACL insertion site with the TT technique has been criticized by some authors as potentially leading to a “non-anatomical” ACL reconstruction and positioning of the graft in a more vertical orientation [9].

The popularity of tibial-independent femoral tunnel techniques has increased over the years to improve the accuracy of femoral tunnel placement achieving a more anatomic ACL position [10].

Despite the growing interest in the AM portal drilling technique, some studies reported an increased risk of revision compared to the traditional TT technique [11, 12].

Many clinical trials have evaluated the drilling of femoral tunnels through the AM and TT portals during ACL reconstruction, but there is no high-quality and updated, evidence-based agreement on which approach results in the best outcomes and the literature is often inconclusive [13,14,15].

The purpose of this study is to systematically review and meta-analyze randomized controlled trials reporting the comparative clinical and functional outcomes, postoperative complications, and radiological outcomes of single-bundle anterior cruciate ligament reconstruction (ACLR) performed using the TT approach or AM technique.

Materials and methods

Criteria for considering studies for this review

Prospective randomized studies reporting comparative outcomes of arthroscopic ACL reconstruction performed with TT and AM transportal femoral drilling were considered eligible for inclusion.

Participants were patients with a diagnosis of ACL rupture with an indication of ACL reconstruction.

Non-randomized trials, cohort studies, retrospective studies, case reports, technical notes, editorial commentaries, ex vivo, biomechanical, pre-clinical, and clinical studies without adequate quantitative or qualitative data were excluded.

The primary endpoint of this research was to systematically review and meta-analyze the clinical outcome in patients who were randomized for the TT approach or AM technique in ACL reconstruction.

The primary outcome measure was considered the IKDC objective score.

Secondary endpoints were to assess and compare the objective anterior laxity, the pivot-shift test, the subjective IKDC score, the Lysholm score, the Tegner scale, the failure rate and the femoral tunnel length.

Search methods for identification of studies

A systematic review of the literature has been performed, following the Cochrane Handbook of Systematic Reviews of Interventions [16] and Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) [17] for study selection (Fig. 1).

Fig. 1
figure 1

The PRISMA flowchart for study selection

Full size image

A systematic search from January 1st, 1990, to January 1st, 2023, was performed in the following databases: the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE/PubMed, Embase, Scopus, the Science Citation Index Expanded from Web of Science, ScienceDirect, CINAHL, and LILACS. The research was conducted using the following keywords alone and in all the various combinations: “ACL” AND “anterior cruciate ligament” AND “reconstruction” AND “transtibial” OR “conventional” AND “anteromedial” OR “transportal.”

The selection process was based on the participants, intervention, control, outcome, and study design (PICOS).

Two reviewers (AMM, PG) independently screened each title and abstract collected from the primary electronic search. In case of a relevant title and abstract, the full-text version was obtained.

All references of each study were screened to find any additional relevant paper potentially missed with the first review process. The two reviewers independently followed the same checklist to screen all studies and evaluate the eligibility criteria. Disagreements were resolved by consensus agreement with a third reviewer (TD).

Data collection and analysis

The level of evidence of included studies was evaluated through the adjusted Oxford Centre For Evidence-Based Medicine 2011 Levels of Evidence [18]. The quality of the studies was defined using the grading of recommendations assessment, development, and evaluation (GRADE) system [19], rating the quality of evidence in systematic reviews. After the evidence is collected and summarized, the GRADE system provides explicit criteria for rating the quality of evidence that include study design, risk of bias, imprecision, inconsistency, indirectness, and magnitude of effect.

The risk of bias was assessed with the revised tool to assess the risk of bias in randomized trials (RoB 2) and are reported in Fig. 2 [20].

Fig. 2
figure 2

Risk of bias assessment with RoB 2 tool

Full size image

Stepwise analysis of study design, aim of the study, level of evidence, journal, year of publication, country, number of procedures included in the study, graft type, femoral and tibial fixation, mean age, and follow-up were independently conducted by each reviewer. Discrepancies in data extraction were discussed and resolved by a consensus meeting between the authors.

All studies were assessed for the primary and secondary outcome measures.

The analysis was conducted separately for patients with an AM or transportal ACL reconstruction (study group) and patients with a TT or conventional technique (comparator).

Data were extracted and recorded for a stepwise analysis. Basic information about each study including population features, country, number of patients, mean age at surgery, study design, level of evidence, and mean follow-up was extracted and is summarized in Table 1.

Table 1 General characteristics of the included studies
Full size table

Specific features of measured outcomes were accurately assessed, and data are summarized in Table 2.

Table 2 Details of methods and outcomes of included studies
Full size table

Statistical analysis

All analyses were completed with Review Manager 5.4.1 software (Cochrane Collaboration, Oxford, UK), and a p value funnel plot was used to analyze the existence of publication bias for the primary outcome measure (Fig. 3).

Fig. 3
figure 3

Funnel plot for the primary outcome measure (IKDC objective score) reporting the risk of publication bias of the assessed studies

Full size image

For each included study, mean differences (MD) and 95% CI were calculated for continuous outcomes, while risk ratio (RR) and 95% CI were calculated for dichotomous outcomes.

Statistical heterogeneity among the studies was assessed using the χ2 test and I2. A fixed-effect model was applied when I2 < 40%, and a random-effect model when I2 ≥ 40%.

A p value of less than 0.05 was considered statistically significant.

Results

Description of the studies

A total of 1500 records were identified from the electronic database research. After initial screening, 87 studies were retrieved for a full assessment. A total of 13 prospective randomized studies [21,22,23,24,25,26,27,28,29,30,31,32,33] involving 989 patients who underwent arthroscopic single bundle ACL reconstruction (486 in the TT group and 503 in the AM group) were finally included in this systematic review. The summary of the selection process is reported in Fig. 1.

The analysis of the risk of bias revealed that 4 studies [22, 23, 27, 31] had some concerns and 9 studies [21, 24,25,26, 28,29,30, 32, 33] had a low risk of bias (Fig. 2).

The analysis of the quality of the studies according to the GRADE system revealed that 5 studies [21, 29, 30, 32, 33] had high quality, 2 studies [23, 28] had moderate quality, and the remaining 6 studies [22, 24,25,26,27, 31] had low quality.

The details of GRADE assessment are reported in electronic supplemental materials (ESM).

Results of meta-analysis

Functional outcome scores

IKDC objective

Among the 13 investigations, only 7 studies [21, 22, 24,25,26,27, 30] reported qualitative data on objective IKDC score.

These studies involved 601 patients including 297 TT and 304 AM ACL reconstructions.

Statistical heterogeneity was χ2 = 4.65; I2 = 0%; p = 0.59, and a fixed-effect model was used for analysis.

Although all studies reported no significant differences in IKDC objective scores between AM and TT groups, the meta-analysis of pooled data showed a significant difference in IKDC objective scores at the final follow-up assessment (RR = 0.80; 95% CI 0.70–0.91; p < 0.001) (Fig. 4).

Fig. 4
figure 4

Forest plot for IKDC objective score

Full size image

Precisely, 63.8% of patients in AM group versus 50.2% of patients in the TT group had IKDC grade A (p < 0.001).

IKDC subjective

Among the 13 investigations, only 5 studies [21, 25, 26, 29, 33] reported quantitative data on subjective IKDC score and 2 studies [27, 31] did not provide standard deviations.

These studies involved 410 patients including 196 TT and 214 AM ACL reconstructions.

Statistical heterogeneity was χ2 = 1.09; I2 = 0%; p = 0.90, and a fixed-effect model was used for analysis.

All studies reported no significant differences in IKDC subjective scores between AM and TT groups and the meta-analysis confirmed this finding (MD = − 0.11; 95% CI − 0.31 to 0.08; p = 0.26) (Fig. 5).

Fig. 5
figure 5

Forest plot for IKDC subjective score

Full size image

Lysholm score

Among the 13 investigations, only 7 studies [21,22,23, 25, 26, 29, 33] reported quantitative data on Lysholm score and 2 studies [28, 31] did not provide standard deviations.

These studies involved 640 patients including 315 TT and 325 AM ACL reconstructions.

Statistical heterogeneity was χ2 = 4.91; I2 = 0%; p = 0.56, and a fixed-effect model was used for analysis.

Only 2 studies [22, 23] reported better Lysholm scores in AM ACL reconstruction, and the meta-analysis confirmed this significant finding (MD = − 0.73; 95% CI − 1.20 to − 0.26; p = 0.002) (Fig. 6).

Fig. 6
figure 6

Forest plot for Lysholm score

Full size image

Tegner activity scale

Among the 13 investigations, only 5 studies [22, 25, 26, 29, 31] reported quantitative data on the Tegner activity scale.

These studies involved 355 patients including 177 TT and 178 AM ACL reconstructions.

Statistical heterogeneity was χ2 = 11.45; I2 = 65%; p = 0.002, and a random-effect model was used for analysis.

There were no differences in the post-operative Tegner activity scale in TT and AM groups (MD = − 0.27; 95% CI − 0.66 to 0.13; p = 0.18) (Fig. 7).

Fig. 7
figure 7

Forest plot for Tegner activity scale

Full size image

Objective Knee laxity tests

Anterior tibial translation

Among the 13 investigations, 9 studies [21,22,23,24, 26, 29,30,31, 33] investigated instrumental anterior tibial translation.

These studies involved 780 patients including 385 TT and 395 AM ACL reconstructions.

Statistical heterogeneity was χ2 = 12.20; I2 = 34%; p = 0.14, and a fixed-effect model was used for analysis.

Meta-analysis revealed that patients of AM group had significantly less anterior tibial translation (MD = 0.36; 95% CI 0.21 to 0.52; p < 0.001) (Fig. 8).

Fig.8
figure 8

Forest plot for anterior tibial translation

Full size image

Pivot-shift test

Among the 13 investigations, 10 studies [21, 22, 24,25,26,27, 29,30,31, 33] investigated manual pivot-shift test.

These studies involved 801 patients including 394 TT and 407 AM ACL reconstructions.

Statistical heterogeneity was χ2 = 4.80; I2 = 0%; p = 0.85, and a fixed-effect model was used for analysis.

Meta-analysis revealed that patients of AM group had significantly lower grade pivot-shift test than the TT group (OR = 0.46; 95% CI 0.32–0.65; p < 0.001) (Fig. 9).

Fig. 9
figure 9

Forest plot for pivot–shift test

Full size image

Graft failure

Among the 13 investigations, 8 studies [21, 22, 24, 26, 27, 29, 30, 33] investigated graft failure.

These studies involved 771 patients including 397 TT and 374 AM ACL reconstructions.

Statistical heterogeneity was χ2 = 2.68; I2 = 0%; p = 0.85, and a fixed-effect model was used for analysis.

Meta-analysis revealed that patients of AM group had lower but not significant failure rate when compared to the TT group (RR = 0.54; 95% CI 0.28–1.06; p = 0.07) (Fig. 10). Precisely, the graft failure rate was 2.8% for the AM group and 5.3% for the TT group.

Fig. 10
figure 10

Forest plot for graft failure

Full size image

Publication bias

A funnel plot was performed with the IKDC objective score as the indicator. A total of 7 studies [21, 22, 24,25,26,27, 30] were included in the analysis. The 7 points in the funnel plot suggest a lower impact of publication bias on the results (Fig. 3). According to Cook’s distances, none of the studies could be considered to be overly influential. Neither the rank correlation nor the regression test indicated any funnel plot asymmetry (p = 0.88 and p = 0.77, respectively).

Radiological outcomes

Radiological outcomes were reported in 10 studies [22, 24, 25, 27,28,29,30,31,32,33] and were assessed with different imaging techniques. Four studies [22, 27, 30, 31] evaluated femoral tunnel position with X-rays, 4 studies [25, 28, 29, 32] performed a 3D-CT scan, and 3 studies [24, 31, 33] investigated post-operative ACL positioning with MRI. There was a high heterogeneity of assessed outcomes and meta-analysis was not possible, except for femoral tunnel length. A descriptive analysis of radiological outcomes is summarized in Table 3.

Table 3 Radiological outcomes
Full size table

Femoral tunnel length

Only 4 studies [25, 29, 32, 33] reported detailed quantitative data with standard deviations of femoral tunnel length. One study [27] reported only mean values without standard deviations.

These studies involved 218 patients including 107 TT and 111 AM ACL reconstructions.

Statistical heterogeneity was χ2 = 46.62; I2 = 98%; p < 0.001, and a random-effect model was used for analysis.

Four studies [27, 29, 32, 33] reported a significantly longer femoral tunnel in the TT group than in AM group and 1 study [25] did not report a significant difference in tunnel length between the two techniques.

Meta-analysis of pooled data showed that patients who underwent the TT technique had significantly longer femoral tunnel (MD: 8.11; 95% CI 1.22 to 15.01; p = 0.02) (Fig. 11).

Fig. 11
figure 11

Forest plot for femoral tunnel length

Full size image

The femoral tunnel length was not associated with better clinical outcomes or reduction of failure rate in any included research article.

Discussion

The main finding of this study was the significant difference in terms of objective IKDC, Lysholm score, TAS, anterior tibial translation and pivot-shift test in favor of AM technique in comparison with TT technique. No statistically significant difference in terms of subjective IKDC score and graft failure was reported.

Chen et al. [34], in a meta-analysis considering 5 RCT studies on 479 patients, evaluated the clinical outcome and complications rate between AM and TT techniques in single-bundle ACL reconstruction. The authors reported higher functional outcomes in patients undergoing AM technique, as well as a lower incidence of pathological side-to-side difference (SSD), pivot-shift and Lachman test in the AM group in comparison with the TT group.

In a recent meta-analysis including 10 RCTs and 8 Prospective Comparative Trials, Mao et al. [35] reported the clinical and radiological outcomes of both the TT and the AM techniques in single bundle ACL reconstruction. The study considered a population of 53,888 ACL reconstructed patients, 729 of them extracted from RCT studies and 53,159 from prospective controlled trials, including national registry data. Higher values of passive range of motion were reported in favor of the AM technique, despite no difference in subjective outcomes and postoperative activity level. Furthermore, an increased SSD was recorded in the TT group notwithstanding a comparable risk of ACL failure between AM and TT patients.

These findings are consistent with the current literature, reporting a more anatomic tunnel placement associated with the AM technique leading to a biomechanical behavior closer to the native ACL, decreasing the risk of anterior translation and rotational instability [36, 37].

Some authors, in early studies reporting the results of the AM technique, rose concerns about an increased risk of graft failure in comparison with the traditional TT technique [11, 12]. The data of the current study evidenced no difference between AM and TT techniques in terms of graft failure. However, analysis of radiological outcomes revealed that AM group had a significantly shorter femoral tunnel leading to less graft-to-bone interface. As the femoral tunnel length was not related to significant changes of clinical outcome and re-rupture, the increased failure risk reported in the early studies [11, 12] could be also explained by the physiological learning curve related to the introduction of a new surgical technique, leading to an excessive posterior placement of the ACL femoral tunnel. Consistent with this hypothesis, Clatworthy et al. [38] reported a normalization of the ACL failure risk, comparable to TT, after slightly anteriorizing the femoral tunnel, as proposed by the anatomical study of Kawaguchi et al. [39] concerning the biomechanical role of the fan-shape-like fibers of the ACL.

Some authors proposed a modified TT technique [25] in order to overcome the limitations of the traditional TT, reporting a tunnel placement almost comparable to the independent tunnel technique. In a systematic review and meta-analysis, Li et al. [40] compared the clinical outcome between conventional TT, modified TT and AM techniques. The authors considered 10 RCTs and 16 retrospective/prospective controlled trials, including 2202 patients. No difference in clinical outcomes between modified TT and AM technique was reported, nevertheless, AM resulted superior to conventional TT in terms of functional outcomes, post-operative pivot-shift test and SSD. Due to limited sample size, the author was not able to report statistically significant data regarding the comparison between modified and conventional TT techniques. In comparison with the AM technique, the conventional TT could lead to a more anterior and superior graft femoral tunnel placement, different from the original ACL footprint, therefore determining a residual instability leading to lower clinical outcomes.

The results of the current meta-analysis seem to confirm this hypothesis, demonstrating higher results in terms of both objective clinical outcomes and knee stability in AM group, despite no difference in terms of subjective outcomes and graft failure risk.

To the best of our knowledge, this is the larger meta-analysis including only RCTs available in the literature. In comparison with the meta-analysis published by Chen et al. [34], including 5 RCTs and 479 patients, 916 patients from 13 RCTs were considered in the current paper. In the recent meta-analysis by Mao et al. [35], a large cohort was considered, including 729 patients from RCT studies and 53,159 patients from prospective studies. The inclusion of registries data [41, 42] exponentially increased the study population; nevertheless, these kinds of data could be affected by some bias such as confounders variables due to missing data regarding patient selection, no quality control on collected data and possible interference by medical industries [42]. In the current meta-analysis, only RCTs were considered and all the studies except one provided a Level of Evidence I.

Considering the current meta-analysis, the following limitations should be considered: The included trials have short to mid-term follow-up leading to a possible underestimation of graft failure. Moreover, the overall population is relatively limited mainly due to the small cohort of patients included in single RCTs and the assessed radiological outcomes are highly heterogeneous precluding a meta-analysis of outcomes (except for femoral tunnel length).

Furthermore, several confounding variables as anterolateral complex injuries or low-grade posterolateral corner distractions, and combined procedures such as anterolateral reconstruction or lateral extra-articular tenodesis that significantly influence the anterior and rotational knee stability and the clinical outcomes have not been assessed and properly considered.

Finally, the majority of studies did not specify the difference between the conventional and the modified TT precluding a direct comparison of these techniques and relative sub-groups.

Conclusion

ACL reconstruction through AM portal technique provides better objective IKDC, Lysholm score, TAS, anterior tibial translation and pivot-shift test at mid-term follow-up.

No statistically significant difference in terms of subjective IKDC score and graft failure was reported.