Anteromedial Femoral Tunnel Placement Reduces Graft Failure Risk Without Affecting Rotational Stability in Anatomic Single‐Bundle Anterior Cruciate Ligament Reconstruction: A Systematic Review Authors: Haitham Kamel Haroun, Amr Mohamed Abdelhady, Mohamed Abdallah Obied, Mostafa Aly Elabd Published: 2026 ABSTRACT PurposeTo compare clinical outcomes and graft failure between anatomical femoral tunnel (FT) placement at the central and anteromedial bundle (AMB) positions.MethodsA systematic review was performed according to the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses guidelines. A search of PubMed, Cochrane, and Web of Science in November 2024 identified studies comparing AMB and central FT placement in anatomical single‐bundle anterior cruciate ligament reconstruction. The risk of bias was assessed using the Risk Of Bias in Non‐randomized Studies of Interventions tool. A qualitative synthesis was performed and presented graphically.ResultsSix studies with 841 patients (444 AMB FT, 397 central FT) were included. Weighted mean follow‐up was 34.8 and 40.7 months for the AMB and central groups, respectively. Overall risk of bias was serious in 4 studies and moderate in 2. AMB FT placement was associated with a significantly lower risk of graft failure (74%‐77% reduction) in 2 studies. However, concerns in risk of bias and consistency decreased the certainty in this finding. Evidence for patient‐reported outcomes and anterior stability was mixed and of uncertain clinical importance. Although some studies reported statistically significant benefits for AMB FT placement (differences of <7 points for patient‐reported outcomes and <1 mm for anterior stability), others found no significant difference. Rotational laxity showed a nonsignificant difference between groups in all studies.ConclusionsThe current evidence is limited by methodological flaws and risk of bias. It suggests that AM FT placement might be associated with a lower risk of graft failure compared with central placement. The AM position might achieve better patient‐reported outcomes and anterior stability, although the difference is not clinically significant. No evidence of a difference was found regarding rotational stability.Level of EvidenceLevel III, systematic review of Level II and III studies. Anterior cruciate ligament reconstruction (ACLR) techniques have continued to evolve over time. 1 From the early 1990s, orthopaedic surgeons performed arthroscopic ACLR using a transtibial (TT) technique. However, studies showed that the TT technique is unable to locate the femoral tunnel (FT) in the native anterior cruciate ligament (ACL) femoral footprint and results in unsatisfactory clinical outcomes. 2 , 3 Subsequently, anatomical ACLR using the tibial tunnel independent (TI) technique became the most used technique for FT placement inside the anatomical footprint. 1 However, single‐bundle ACLRs cannot replace the whole surface area of the native footprint despite anatomical placement of ACL graft tunnels. Because the femoral footprint is 3.5 times larger than the midsubstance of the ACL, a 9‐mm graft would cover around one‐third of the footprint surface area. 4 Hence, surgeons should decide in which portion of the footprint they would place the ACL graft tunnels during anatomical single‐bundle ACLR. 5 FT placement is the most critical step in ACLR. It determines the biomechanical properties of the ACL graft and has a significant effect on knee stability and clinical outcomes. 6 In addition, it determines the graft length change and tension pattern during knee range of motion, which, if excessive, could predispose to graft elongation and failure. 7 In the initial evolution of anatomical single‐bundle ACLR, surgeons advocated central FT placement to span both the anteromedial bundle (AMB) and posterolateral bundle, thus achieving part of rotational stability of the knee. In addition, a central FT position is intuitively the most “anatomical” position. 5 However, surgeons were always concerned by deviating from graft isometry and the liability of the ACL graft to elongate and disturb knee kinematics throughout the range of motion. 7 , 8 More recent anatomical studies showed the size mismatch between ACL midsubstance and its footprint. 9 In addition, a histological study differentiated ACL femoral insertion fibers into stout direct fibers just behind the lateral intercondylar ridge and membrane‐like indirect fibers more posterior (low). 10 Both these updated anatomical and histological studies led some surgeons to advocate anatomical eccentric anteromedial FT placement. 1 , 4 , 7 , 11 , 12 The rationale of this is to place the FT in the most isometric region of the femoral footprint. 4 , 7 The ideal target point for FT placement in anatomical single‐bundle ACLR is still unclear. 13 A 2021 systematic review analyzing trends among clinical studies of anatomical single‐bundle ACLR showed that, of studies reporting FT target points relative to the 2‐bundle anatomy, the target point was the central position and the AMB position in 78% and 18% of studies, respectively. 13 Several biomechanical studies have reported the superiority of FT placement in the central position. 14 , 15 , 16 , 17 , 18 In contrast, other biomechanical studies recommended placing the FT at the AMB position. 5 , 6 , 19 , 20 , 21 , 22 Both FT positions showed satisfactory and comparable graft healing on 1‐ and 1.5‐year follow‐up magnetic resonance imaging in anatomical single‐bundle ACLR. 23 , 24 However, several clinical studies showed conflicting results about graft failure rates after ACLR with FT placed at different anatomical positions. 1 , 11 , 25 , 26 The purpose of this study was to compare clinical outcomes and graft failure after anatomical single‐bundle ACLR with FT placed at AMB or central positions. The hypothesis was that anteromedial bundle (AMB) FT placement would be associated with a lower risk of graft failure and non‐inferior clinical outcomes compared with central FT placement. METHODS Search Methods We conducted an electronic literature search of the following databases from their inception date to November 2024: PubMed, Cochrane Central Register of Controlled Trials, and Web of Science. Searches were carried out in accordance with the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses 27 with a checklist provided in Supporting Information: Appendix 1. Prior to initiating the search, a review protocol was registered with the international prospective register of systematic reviews (PROSPERO) and published under registration number CRD42024518633. Our search term was as follows: (“Anterior Cruciate Ligament” OR ACL) AND (“femoral tunnel” OR “femoral aperture”). The full search strategy for each database is presented in Supporting Information: Appendix 2. The search was performed by 3 independent reviewers (M.A.E., A.M.A., H.K.H.). We also searched the references of the included studies for additional relevant studies. Criteria for Considering Studies for This Review The criteria for inclusion of the studies included clinical studies of anatomical single‐bundle ACLR directly comparing central and AMB FT positions with a minimum follow‐up period of 1 year. Studies should report clinical outcomes (functional outcome and/or knee stability) with or without graft failure. Abstracts, case series, and cadaveric studies were excluded. Only articles in English were included. Participants were deemed eligible if they were skeletally mature patients who underwent single‐bundle primary ACLR by TI technique with FT position within the footprint (anatomical ACLR). Patients with concomitant knee ligament surgeries were excluded. FT position was defined to be anatomical based on predefined criteria. These include 1 or both of 2 criteria. First, the authors had to state that they identified intraoperatively the native ACL insertion sites and/or bony landmarks and placed the FT in the native ACL insertion site. Second, authors would ascertain the achieved FT position by postoperative imaging. The intervention was classified as either central or AM FT position according to intraoperative guidance by remnants of native ACL bundles or bony landmarks and/or postoperative imaging ascertainment of FT placement within the reported positions for central or AM bundles of ACL insertion. 28 Data Collection and Analysis Study Selection Two independent investigators (M.A.E., A.M.A.) conducted selection of studies. An arbiter (the most senior, third author [H.K.H.]) solved any disagreement. Data Extraction and Management The 3 independent investigators extracted data from the included studies into Excel sheets. If there was any missing data in any study, correspondence with the article author was tried. The extracted data included study participant characteristics, surgical technique characteristics, and outcomes of interest including effectiveness and adverse outcomes. Effectiveness outcomes included patient‐reported outcomes (PROs) (O1) (Tegner activity score, Lysholm score, and International Knee Documentation Committee [IKDC] subjective score) and knee stability outcomes (O2) evaluated by either manual testing (O2a), instrumented laxity device (O2b), or stress radiography (O2c). Stability outcomes were extracted from the articles as follows: proportion of positive Lachman test results (G1 to 3), proportion of positive pivot‐shift test (G1 to 3) results, and anterior tibial translation either absolute or side‐to‐side difference. Adverse outcome included graft failure (O3). Assessment of Risk of Bias Each article was appraised by 2 independent reviewers (M.A.E., A.M.A.) separately. The arbiter solved any disagreement. We used the “Risk of Bias in Non‐Randomized Studies of Interventions assessment tool” 29 to assess the risk of bias in included studies. Statistical Analysis Dichotomous and continuous outcomes were synthesized using risk ratio and mean difference, respectively. Qualitative synthesis of the results from the included studies was provided and organized according to the outcome measurement tool and its endpoint. Data from included studies were not pooled to avoid incorrect reporting of results because of the lack of high‐level evidence. 30 Instead, these data were graphically represented in forest plots if the intervention effect was completely reported. The I 2 statistic was reported to explore heterogeneity among the studies. The I 2 value can range between 0% and 100%, with greater values indicative of more heterogeneity. If substantial statistical heterogeneity (I 2 > 75%) was detected, we explored the possible clinical and methodological reasons for this. The certainty of the body of evidence for each outcome was determined according to domains that could affect certainty. These analyses were conducted using RevMan software (version 5.4.1; Nordic Cochrane Centre, Copenhagen, Denmark). RESULTS Search Results The literature search extracted 3953 records. After we removed the duplicates, a total of 2999 records remained. After screening the titles and abstracts of these, 50 full texts remained. After a thorough reading, these studies were assessed for eligibility. Of these, 6 studies 23 , 24 , 31 , 32 , 33 , 34 were included. The remaining 44 studies were excluded for the following reasons: 29 studies were cadaveric. In 7 studies, FT was placed by TT technique. Two studies included nonanatomical FT position. 35 , 36 In 2 studies, grouping was performed by other surgical factors than FT position. 37 , 38 Two studies compared high and low FT positions. 39 , 40 Two studies compared anatomic FT positions not including AM and central positions 41 , 42 (Figure 1). PRISMA flow diagram of methodology used for inclusion and exclusion of studies. (AM, anteromedial; FT, femoral tunnel; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta‐Analyses; TT, transtibial.) Study Characteristics From the 6 studies included in our review, there were 5 observational studies and 1 interventional study. The observational studies included 4 retrospective cohort studies 31 , 32 , 33 , 34 and 1 prospective cohort study. 24 The interventional study was nonrandomized trial 23 registered in the Korean Clinical Research Information Service. Three studies included comparisons between 2 arms (AMB vs central FT positions). The other 3 studies were 3‐arm studies with 2 of them 23 , 32 including a third arm of FT posterolateral bundle position and the third 34 including a third arm of FT high central position. Participant Characteristics From a total of 841 patients, 444 patients underwent anatomical AM ACLR, and 397 patients underwent anatomical central ACLR. The AMB position group had a weighted mean age of 30.8 years, whereas the central position group had a weighted mean age of 29.5 years. The percentage of male patients in the AMB position and central position groups was 65.9% and 67.5%, respectively. In 1 study, 23 the ACL injury was partial in the AMB position group only. Time from injury to surgery for each group was reported in 1 study 24 with an average of 9.5 and 8.7 months in the AMB position and central position groups, respectively. Preinjury Tegner activity score was reported in 3 studies 23 , 24 , 33 where the weighted mean score was 4.6 in each of the 2 groups. Inclusion of meniscal injuries was reported in 5 studies. 24 , 31 , 32 , 33 , 34 Three of them 24 , 31 , 32 reported details of procedures performed for these meniscal injuries which were either partial meniscectomy or meniscal repair. Chondral lesions whether high grade (G3 or more) or requiring surgery were excluded in 3 studies. 23 , 24 , 33 One study 31 including chondral injuries reported details of procedures performed for these injuries which were chondroplasty, microfracture, or unspecified other surgeries. The concomitant procedures in both groups were comparable in the 3 studies that reported their details. 24 , 31 , 32 The AMB position group had a weighted mean follow‐up duration of 34.8 months, whereas the central position group had a weighted mean follow‐up duration of 40.7 months (Table 1). TABLE 1Characteristics of the Included Studies No. Study (1st Author, Year) Country Design (LoE) Comparison (Anatomic FT Position) Patients, n Men/Women, n Age, Year* Graft (Type/Source) FT Position, % (Depth*/Height*) as Measured by Quadrant Method on CTFollow‐Up, Months* 1Borton, 31 2018UKRetrospective cohort (III)AMB position166121/45 33 (24‐45) Hamstring autograftNA 42 (34‐48) Central position11382/31 27 (22‐38) 64 (59.5‐69) 2 Lee, 23 2017 (Groups) Republic of KoreaNonrandomized trial (II)AMB position2521/427.0 ± 13.0Soft tissue graft21.7/2023.5 ± 4.4Central position3323/1027.4 ± 9.525.9/33.825.2 ± 3.33 Seo, 32 2019 (Groups) Republic of KoreaRetrospective cohort (III)AMB position2520/531.7 ± 10.4Hamstring autograft24.8 ± 1.1/28.2 ± 1.430.8 ± 5.8Central position4637/933.9 ± 13.229.7 ± 1/35 ± 3.330.5 ± 7.0 4Tecame, 33 2022ItalyRetrospective cohort (III)AMB position12769/5829.4 ± 5.7Hamstring autograftNA 18.2 ± 2.4 Central position11771/4630.4 ± 4.8 22.8 ± 2.6 5Zhang, 24 2021ChinaProspective cohort (II)AMB position3419/1526.5 ± 6.7Hamstring autograft24.7/21.724Central position3521/1425.7 ± 4.829.7/35.1 6 Zhang, 34 2023 (Groups) ChinaRetrospective cohort (III)AMB position6743/2431.5 ± 11.5 Hamstring autograft NR60 ± 2.4Central position5334/1933.2 ± 9.360 ± 1.2 Note: Data in bold: significant difference between study groups.AMB, anteromedial bundle; CT, computed tomography; FT, femoral tunnel; LoE, level of evidence; NA, not applicable; NR, not reported. *Data are presented with standard deviations or ranges according to the data reported.© 2026 WILEY‐VCH GmbH Characteristics of the Included Studies Note: Data in bold: significant difference between study groups. AMB, anteromedial bundle; CT, computed tomography; FT, femoral tunnel; LoE, level of evidence; NA, not applicable; NR, not reported. Data are presented with standard deviations or ranges according to the data reported. Surgical Technique in Included Studies In 5 studies 23 , 24 , 32 , 33 , 34 reporting about surgeons, interventions in both groups were performed by the same surgeon in all of them. The procedure performed was standard single‐bundle ACLR in all studies except in 1 study 23 where remnant‐preserving ACL augmentation was performed in AMB FT position group. From these 5 studies with standard single‐bundle ACLR, the technique included a full‐tibial tunnel in all of them except in 1 study 31 with an all‐inside technique with “half‐tunnel” on both the femoral and tibial sides. The graft used was a soft tissue graft in all studies. In the 5 studies 24 , 31 , 32 , 33 , 34 reporting graft details, the graft was autologous hamstring in all of them. The diameter of the graft was reported in 4 studies 23 , 24 , 31 , 32 with averages ranging from 8 to 9 mm. All studies reported a femoral aperture localization strategy. The remnant ACL footprint was reported to be used in all studies except in 1 study 34 that used bony landmarks and fluoroscopy. In addition to the remnant femoral footprint, an offset guide 24 , 33 and bony landmarks 23 , 31 , 32 were used. FT was drilled by TI technique in all studies either by the anteromedial transportal technique in 4 studies 23 , 24 , 33 , 34 or the outside‐in technique in 2 studies. 31 , 32 The intra‐articular target point for the tibial tunnel had been aimed at the tibial footprint in all studies. In 5 of these studies, 24 , 31 , 32 , 33 , 34 it had been aimed at the center of the tibial footprint. Postoperative imaging ascertainment and quantification of anatomical placement of tibial tunnel within the footprint was performed in 1 study 32 where there was a nonsignificant difference in tibial tunnel position between the groups. Femoral fixation of ACL graft was suspensory fixation in all studies. Tibial fixation was aperture fixation by biodegradable interference screw in all studies except in 1 31 where it was suspensory fixation by adjustable loop. The interference screw was backed up by secondary fixation in all except in 1 study. 31 Patient assignment into one of the FT position groups was either performed intraoperatively intentionally during femoral aperture placement in 5 studies 23 , 24 , 31 , 33 , 34 or detected by postoperative imaging in 1 study. 32 From the 5 studies in which patients were grouped according to intraoperative placement, this FT placement was ascertained postoperatively by imaging in 3 studies. 23 , 24 , 34 See Table 1 for details of quantified FT position in included studies whenever reported. Quality Assessment of the Included Studies All studies reported about conflicts of interest (COI). Three studies 24 , 31 , 34 had potential conflicts of interest. One of them 31 was arbitrated to be of notable concern because there were financial interests related to one of the authors. The overall risk of bias judgement for the intervention effect for all outcomes scored serious risk in all studies except in 2 studies 32 , 34 where risk of bias judgement for the intervention effect scored moderate risk. Risk of bias about the methodological quality of the included studies is shown in Table 2. TABLE 2Risk of Bias Summary With Judgements of Each Domain for the Included StudiesStudy (1st Author, Year)1234567Overall Bias Bias Because of Confounding Bias in Selection of Participants Into the Study Bias in Classification of Interventions Bias Because of Deviations From Intended Interventions Bias Because of Missing Data Bias in Measurement of Outcomes Bias in Selection of Reported Result O1 O2 O3 O1 O 2a O 2b O 2c O3 O1 O 2a O 2b O 2c O3 Borton, 31 2018SLMMS*LL***LM S * * * S Lee, 23 2017SMLLSS*LM*L*L S S * S * Seo, 32 2019MLLLLLLLMLMLM M M M M M Tecame, 33 2022SMMLLLLLLL*LM S S S * S Zhang, 24 2020LMLLSSSLLL*LM S S S * S Zhang, 34 2023MLLLLLL*ML*MM * M M * M Note: Patient‐reported outcomes (O1); knee stability outcomes evaluated by manual testing (O2a), instrumented laxity device (O2b), stress radiography (O2c); graft failure (O3). Yellow for moderate risk, Red for serious risk.L, low risk; M, moderate risk; S, serious risk.*Outcome not assessed.© 2026 WILEY‐VCH GmbH Risk of Bias Summary With Judgements of Each Domain for the Included Studies Note: Patient‐reported outcomes (O1); knee stability outcomes evaluated by manual testing (O2a), instrumented laxity device (O2b), stress radiography (O2c); graft failure (O3). Yellow for moderate risk, Red for serious risk. L, low risk; M, moderate risk; S, serious risk. *Outcome not assessed. Qualitative and quantitative within‐studies comparison of participants and surgical technique characteristics was possible in most of the included studies. There was baseline equivalence of the 2 groups in all studies. However, there were concerns in some studies. The age was nonequivalent in 1 study 31 where the AM group patients were older. The degree of ACL injury was nonequivalent in 1 study 23 with less severe injury in AM position group. Follow‐up duration was nonequivalent in 2 studies 31 , 33 where the AM group had shorter follow‐up. Effects of Intervention Summary of finding table is provided in Table 3. Patient‐Reported Outcomes Tegner Activity Score at the Final Follow‐Up There were 4 studies 23 , 24 , 31 , 33 that provided data on Tegner activity score. There was a nonsignificant difference between the groups in all of them (632 patients). TABLE 3Summary of Findings Table Outcome Result Patient‐reported outcomesa‐ Tegner activity scoreNonsignificant difference between groups 632 (4 studies)b‐ Lysholm scoreInconsistent findings:721 (5 studies)2 studies (313 participants)Significantly better scores for AMB group∗ 3 studies (408 participants)Nonsignificant difference between groupsc‐ IKDC subjective scoreInconsistent findings:442 (4 studies)2 studies (313 participants)Significantly better scores for AMB group† 2 studies (129 participants)Nonsignificant difference between groupsKnee stability outcomesa‐ Positive Lachman testInconsistent findings:302 (2 studies)1 study (244 participants)Significantly less risk with AMB group1 study (58 participants)Nonsignificant difference between groupsb‐ Positive pivot‐shift testNonsignificant difference between groups562 (5 studies)c‐ Instrumented sagittal stability testingInconsistent findings:504 (4 studies)3 studies (384 participants)Significantly less translation with AMB group‡ 1 study (120 participants)Nonsignificant differenced‐ Sagittal stability testing by stress radiographyInconsistent findings:200 (2 studies)1 study (142 participants)Significantly less side‐to‐side difference with AMB group$ 1 study (58 participants)Nonsignificant difference Graft failure Graft failureInconsistent findings: 783 (5 studies)2 studies (513 participants)Significantly reduced risk with AMB position|| 2 studies (191 participants)Nonsignificant difference1 study (69 participants)No revisions in either group∗With a mean difference of 4.8 and 1.6 points.†With a mean difference of 6.7 and 1.8 points.‡With a mean differences of <1 mm; 0.5 and 0.74 mm.$With a mean difference of 0.6 mm.||With risk reduction by 74% to 77%.© 2026 WILEY‐VCH GmbH Summary of Findings Table With a mean difference of 4.8 and 1.6 points. With a mean difference of 6.7 and 1.8 points. With a mean differences of <1 mm; 0.5 and 0.74 mm. With a mean difference of 0.6 mm. With risk reduction by 74% to 77%. Lysholm Score at the Final Follow‐Up There were 5 studies 23 , 24 , 31 , 32 , 33 that provided data on Lysholm score. In 2 studies 24 , 33 (313 participants), there was significant better score in the AMB position group. However, the difference was 4.8 and 1.6 points. An inconsistent finding of nonsignificant difference between both groups was shown in the other 3 studies 23 , 31 , 32 (408 participants) whose data could not be graphically represented (Figure 2A). Forest plot for studies reporting on patient‐reported outcomes. (A) Lysholm score. (B) IKDC subjective score. (AM, anteromedial; CI, confidence interval; IKDC, International Knee Documentation Committee; IV, inverse variance method; SD, standard deviation.) International Knee Documentation Committee Subjective Score at the Final Follow‐Up There were 4 studies 23 , 24 , 32 , 33 that provided data on the International Knee Documentation Committee score. In 2 studies 24 , 33 (313 participants), there was significant better score in the AMB position group. However, the difference was 6.7 and 1.8 points. An inconsistent finding of nonsignificant difference between both groups was shown in the other 2 studies 23 , 32 (129 participants) whose data could not be graphically represented (Figure 2B). Knee Stability Outcomes Positive Lachman Test at the Final Follow‐Up There were 2 studies 23 , 33 that provided data on postoperative Lachman test. In 1 study 33 (244 participants), there was significantly less risk of positive Lachman test with the AMB position group. An inconsistent finding of a nonsignificant difference between both groups was shown in the other study 23 (58 participants) (Figure 3A). Forest plot for studies reporting on knee stability outcomes. (A) Positive Lachman test. (B) Positive pivot‐shift test. (C) Side‐to‐side difference in anterior tibial translation. (AM, anteromedial; CI, confidence interval; IV, inverse variance method; M‐H, Mantel‐Haenszel method; SD, standard deviation.) Positive Pivot‐Shift Test at the Final Follow‐Up There were 5 studies 23 , 24 , 32 , 33 , 34 that provided data on postoperative pivot‐shift test. In all of them (562 participants), there was a nonsignificant difference between both groups in the risk of positive pivot‐shift test (Figure 3B). Postoperative Instrumented Sagittal Stability Testing at the Final Follow‐Up There were 4 studies 24 , 32 , 34 that provided data on postoperative instrumented sagittal stability testing. In 2 studies 24 , 32 (140 participants), there was significantly less tibial translation with the AMB position group. However, the difference was less than 1 mm. A consistent finding was shown in a third study 33 (244 participants) whose data could not be graphically represented where the AM position group had significantly less anterior laxity and more anterior tibial translation reduction compared with the central position group with 0.5 and 0.74 mm mean difference, respectively. An inconsistent finding of nonsignificant difference between both groups was shown in a fourth study 34 (120 participants) (Figure 3C). Postoperative Sagittal Stability Testing By Stress Radiography at the Final Follow‐Up There were 2 studies that provided data that could not be graphically represented. In 1 study 32 (142 participants), the AM position group had significantly less side‐to‐side difference in AP tibial translation than the central position group with 0.6 mm mean difference. Inconsistent finding was shown in another study 23 (58 participants) where there was non‐significant difference between both groups in the risk of patients with more than 2 mm side‐to‐side difference in anterior tibial translation in both groups. Graft Failure Five studies 24 , 31 , 32 , 33 , 34 reported adverse outcomes of graft failure whether reported to be treated by revision ACLR or not. Graft failure was defined in 4 studies. 31 , 32 , 33 , 34 In 2 studies, 31 , 32 it was defined as graft rupture that happened after traumatic episodes. In the other 2 studies, 33 , 34 it was defined as remarkable sagittal and/or rotational instability on examination. In 1 of these 2 studies, 33 this could be replaced or associated with graft rupture on magnetic resonance imaging. For convenience, all these variations of report and definition of graft failure are referred to in this review as a common adverse outcome of “graft failure.” Five studies 31 , 32 , 33 , 34 provided data on graft failure risk. Data could be graphically represented in 4 studies. 31 , 32 , 33 , 34 In 2 studies 31 , 33 (513 participants), placement of FT at the AM position significantly reduced the risk of graft failure than placement at the central position in 2 studies by 74% and 77%. In the other 2 studies 32 , 34 (191 participants), there was a nonsignificant difference (Figure 4). In the fifth study 24 (69 participants) whose data could not be graphically represented, there were no revisions or complications from either 34 or 35 patients in AM or central position groups, respectively. Forest plot for studies reporting on graft failure. (AM, anteromedial; CI, confidence interval; M‐H, Mantel‐Haenszel method.) DISCUSSION The main finding of this review indicates that FT placement at the AMB position may reduce the risk of graft failure compared with central placement. Theoretically, this could be attributed to the less isometric profile of the central footprint ACLR, which may lead to higher graft strains. This finding is supported by 2 studies which reported a 74% and 77% reduction in failure risk with the AMB position. 31 , 33 However, the certainty of this finding is low owing to significant limitations: risk of intervention misclassification (lack of postoperative imaging confirmation) and potential confounding in favor of the AMB group (later surgery dates and shorter follow‐up) in those studies. 31 , 33 Furthermore, 3 other studies found no significant difference in graft failure risk between the 2 positions. 24 , 32 , 34 Additionally, the cause of graft failure was traumatic rupture in half of these studies. Regarding postoperative rotational stability, there was a nonsignificant difference in postoperative rotational stability, based on pivot‐shift testing, between the 2 FT positions across 5 studies. 23 , 24 , 31 , 32 , 33 , 34 However, the pivot‐shift test is a subjective measurement with low sensitivity that cannot detect small differences in rotational instability. For anterior knee stability, the review reveals that AM FT placement likely results in little to no difference in anterior stability compared with central placement, as measured by manual testing, instrumented laxity devices, or stress radiography. 24 , 32 , 33 The difference was statistically significant in 3 studies but nonsignificant in 2 other studies. 23 , 34 When the AM group showed better stability, the difference was clinically insignificant (less than 1 mm). 43 This pattern could be explained by the less isometric graft profile with central FT placement, which may lead to graft slackening during knee flexion. However, experimental and clinical studies suggest that this subclinical difference could lead to recurrent anterior laxity at longer‐term follow‐up. 7 , 44 , 45 For PROs, the review indicates that AM FT placement might result in little to no difference. Although the difference was statistically significant in 2 studies, it was nonsignificant in 3 studies. In addition, when the AM group showed better PROs, the difference was clinically insignificant, at less than 7 points. 46 Our review applied careful eligibility criteria to focus on the narrowly defined surgical technique of anatomical single‐bundle ACLR with the FT placed within the femoral footprint and drilled by the TI technique. The TI technique was selected because it is the most used technique for FT placement currently and is more associated with anatomical FT placement than the TT technique. 2 , 3 , 13 , 25 From all possible FT positions within the native ACL footprint, we selected the AMB and central FT positions for comparison. The reason is that, opposite to the posterolateral bundle position, 42 these are the 2 anatomical FT positions reported in the literature and used in clinical practice for single‐bundle ACLR. Our search also revealed studies comparing 2 other possible FT positions within the femoral footprint: the high and low (11 and 10 o'clock) FT positions. 39 , 40 We excluded these studies because the FT positions were differentiated only in the axial plane. However, the central FT position is not just lower but also shallower than the AMB position. The amount of evidence that had been included in this review was 6 studies involving a total of 841 patients, 444 patients underwent anatomical AM ACLR, and 397 patients underwent anatomical central ACLR. Although the major surgical factors (e.g., surgeon and graft type) were consistent across studies, substantial heterogeneity in the intervention effect on PROs and sagittal knee stability was observed. This could be explained by the risk of intervention misclassification in Tecame et al. study. In addition, there were participant differences in the severity of ACL injury and patients’ age in Lee et al. and Borton et al. studies, respectively. Also, the AM group in Tecame et al. had a later reconstruction and shorter follow‐up. These methodological and clinical heterogeneities could explain the inconsistent findings between Tecame et al., which showed better outcomes in the AM group, and other studies showing nonsignificant differences. To assess postoperative knee stability, we pooled the proportion of a positive Lachman and pivot‐shift test results, considering only grade 0 as normal to avoid controversy over classifying grade 1 as normal or abnormal. 43 Based on a comprehensive literature search, this review addresses a specific and previously underexamined aspect of ACLR. Although existing reviews have compared outcomes of different FT positions, they have done so within the context of comparing TT versus TI drilling techniques. Their findings cannot be directly compared with ours. The TT technique tends to place the FT in a high‐AM position, which may be outside the anatomical footprint. Consequently, they essentially compare nonanatomical versus anatomical FT placement. However, after correlating our findings with the most related literature, we noted the following: Moon et al. conducted a clinical study of anatomical ACLR comparing a nonisometric graft with a lower FT to a nearly isometric graft with a higher FT. 38 They show a result comparable to ours: a nonsignificant difference in PROs and pivot‐shift testing between both groups. Excellent clinical results and knee stability were also shown in a case series of anatomical ACLR with FT placed in the AMB position. 47 Some biomechanical studies showed comparable results to ours. Cross et al. compared anatomical AMB and central FT positions, finding no significant difference in time‐zero stability during pivot‐shift testing. 5 Similarly, Markolf et al. revealed that moving the FT from 11 o'clock to a more oblique location did not significantly affect knee kinematics during simulated pivot‐shift testing. 21 Our review finding of a possible reduced risk of graft failure with AM FT placement is consistent with 2 clinical reports. One study found that switching from an AMB to a central FT position increased graft failure rate in professional footballers. 25 The other, a large prospective cohort study with 6 to 15 years of follow‐up, reported a 3.5 times higher revision rate in the central FT group than in the AM FT group per 100 graft years. 25 Limitations The included studies had several limitations. First, although most of the included studies targeted the center of the tibial footprint, they did not ascertain the achieved tibial tunnel position except in 1 study. 32 However, tibial tunnel position could also affect ACL graft obliquity and ACLR outcomes. 31 Furthermore, all included studies exclusively used hamstring tendon autografts. Therefore, our findings may not be directly generalized to reconstructions using other grafts. Lastly, the weighted mean follow‐up duration was less than 3 years in both review groups, and a radiological evaluation for the development of osteoarthritis was lacking in all included studies. Although the review showed no difference in PROs between the anatomical FT positions, this does not mean that there is no difference. The detected subtle differences in knee kinematics between the 2 anatomical FT positions might be associated with the development or prevention of osteoarthritis in the long term. 48 Beyond the limitations inherent to the included studies, this review itself has methodological constraints that should be acknowledged. Quantitative synthesis of outcomes could not be performed because of the observational design of the included studies with substantial heterogeneity and risk of bias. 30 Furthermore, our literature search was limited to the English language which could introduce the risk of publication bias. CONCLUSIONS The current evidence is limited by methodological flaws and risk of bias. It suggests that AM FT placement might be associated with a lower risk of graft failure compared with central placement. The AM position might achieve better PROs and anterior stability, although the difference is not clinically significant. No evidence of a difference was found regarding rotational stability. SUPPORTING INFORMATION Additional supporting information can be found online in the Supporting Information section. DISCLOSURES The authors (H.K.H., A.M.A., M.A.O., M.A.E.) declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this article. Supporting information Supplementary Material