Anterolateral rotatory instability after rupture of the anterior cruciate ligament (ACL) is assessed clinically mainly with the pivot-shift test: when a combined internal-rotation and valgus load is applied in early flexion, the lateral tibial plateau subluxates anteriorly relative to the lateral femoral condyle and then reduces abruptly as flexion deepens under tension in the iliotibial band. Besides the centrally placed intra-articular ACL, the anterolateral side of the knee contains, from superficial to deep, the superficial iliotibial band (ITB) and its iliopatellar band extension; the proximal and distal Kaplan fiber bundles and the condylar strap that connect the deep surface of the ITB to the posterolateral distal femur; the capsulo-osseous layer (COL) of the ITB; and, at the capsular level, the anterolateral ligament (ALL) and the anterolateral capsule. Together these structures form the anterolateral complex (ALC), which lies lateral to the axis of tibial internal rotation and provides extra-articular restraint against internal rotation.
There has been clear disagreement in earlier work on how the anterolateral complex should be defined anatomically and on which of its parts contribute most to mechanical restraint. Some early anatomic and biomechanical reports treated the ALL as the principal ligament controlling anterolateral rotatory instability, whereas a recent systematic review concluded that cadaveric biomechanical experiments show only a limited role for the ALL complex in restraining anterior tibial translation and internal rotation. Taking together the available gross anatomy, histologic sections, sequential soft-tissue sectioning on robotic and optical kinematic setups, and acute clinical imaging studies, the evidence more consistently supports the view that, when the superficial ITB and its distal femoral attachments remain intact, the ITB and its deep femoral connections (the Kaplan fiber system and the capsulo-osseous layer) carry the main extra-articular anti–internal-rotation role at larger flexion angles; the anterolateral ligament and the anterolateral capsule provide a conditional, synergistic secondary restraint, and the size of their contribution measured in vitro is jointly influenced by whether the superficial ITB has already been freed, whether the underlying capsule is opened with the ALL, the order of sectioning, and the outcome measure (the share of resistance under predetermined trajectory replay, or the increment in laxity angle under a fixed load).
1. Anatomic Layers of the Anterolateral Complex and Disagreements on Nomenclature
The 2017 International Anterolateral Complex Consensus Group Meeting (Getgood et al., 2019; the meeting received educational support from Smith & Nephew, and the authors declared no conflicts of interest) defined the terminology for this region: the anterolateral complex comprises the superficial and deep ITB, the periosteal–fascial fibers by which the ITB attaches to the distal femur (the Kaplan fiber system), the capsulo-osseous layer, and the anterolateral ligament lying within Seebacher layer 3 (the superficial layer of the lateral capsule, anterior to the lateral collateral ligament)[6]. Equating the ALL with the entire anterolateral complex, or assigning the capsulo-osseous layer and the capsule inconsistently, is an important source of the differences among published reports.
Table 1. Anatomic attachments of the principal structures of the anterolateral complex and disagreements on nomenclature
| Structural layer and name | Principal attachments and relations | Anatomic recognition and naming disagreements |
|---|---|---|
| Superficial ITB and iliopatellar band | The main distal insertion spreads widely onto Gerdy's tubercle; the curved anterior fibers (iliopatellar band) insert on the lateral patella and the lateral border of the patellar tendon[2,4] | The densest, most superficial fascial layer of the anterolateral knee; middle-layer oblique fibers cross those of the superficial layer[2] |
| Proximal and distal Kaplan fiber bundles and condylar strap | The proximal and distal bundles attach to the posterolateral bony ridge 53.6 mm and 31.4 mm proximal to the lateral femoral epicondyle, respectively, and the superior lateral genicular artery runs immediately distal to the distal bundle[4]; the condylar strap lies distal to that artery and connects the region proximal to the lateral epicondyle / adjacent to the lateral gastrocnemius tendon with the deep surface of the ITB 41–58 mm proximal to Gerdy's tubercle[5] | Anchors the distal ITB to the posterolateral distal femur; the condylar strap was identified in every specimen approached from behind (between biceps femoris and ITB) (11/11) and could not be accurately recognized when the ITB was reflected from the front, because that approach cuts the connection (0/5)[5] |
| Capsulo-osseous layer of the ITB (COL) | Arises from the area about 19.1 mm proximal to the lateral epicondyle and from fascia adjacent to the lateral head of gastrocnemius, is reinforced posteriorly by the aponeurosis of the short head of biceps femoris, arches over the superficial surface of the lateral capsule, and inserts distally on the lateral tibial tubercle behind Gerdy's tubercle[2,4] | Herbst et al. (2017) and Godin et al. (2017) assign it to the deep ITB and separate it from the capsule[2,4]; Daggett et al. (2018) treat the COL and the ALL as the same structure and report that separation from the underlying capsule is inconsistent[3] |
| Anterolateral ligament and anterolateral capsule (ALL & anterolateral capsule) | Dissection of embalmed specimens (Claes et al., 2013) identified an origin from a prominence on the lateral epicondyle slightly anterior to the origin of the lateral collateral ligament (LCL)[1]; fresh-frozen specimens and the international consensus more often place it posterior and proximal to the epicondyle or intermingled with the LCL origin[6]. Midway it attaches to the body of the lateral meniscus, with the inferior lateral genicular vessels between the ALL and the lateral meniscus, and distally it inserts near the midpoint of the line joining Gerdy's tubercle and the tip of the fibular head[1,6] | The international consensus and Herbst et al. (2017) place it in the lateral capsule (Seebacher layer 3); Herbst et al. identified a mid-third capsular ligament in 7/20 specimens[2,6]; Kittl et al. (2016) separately defined extracapsular fibers superficial to the LCL as the ALL[9] |
1. Deep Femoral Attachments of the Iliotibial Band
The superficial ITB has a broad bony insertion on Gerdy's tubercle. Its deep surface is anchored to the femur by fiber structures that extend from the distal femoral metaphysis to the region of the lateral epicondyle, so that it can restrain the joint across flexion–extension and rotation[2,4]. Godin et al. (2017; some authors disclosed consulting or research support from Arthrex, Smith & Nephew and Ossur) confirmed, in 10 fresh-frozen knee specimens, that discrete proximal and distal Kaplan fiber bundles were present in every specimen, attaching to the posterolateral bony ridge 53.6 mm and 31.4 mm proximal to the lateral epicondyle, respectively, with the superior lateral genicular artery constantly running immediately distal to the distal bundle[4]. In a separate uniaxial tensile failure test on 10 fresh-frozen male knees, the structural ultimate failure load of the distal Kaplan bundle was 170.2 N (stiffness 36.3 N/mm) and that of the proximal bundle was 71.3 N (stiffness 22.6 N/mm)[4]. These values are the structural mechanical properties of isolated single-bundle specimens under uniaxial tension; they cannot be equated directly with the functional load sharing of an intact living knee under rotational load.
Distal to the distal Kaplan bundle, fibrous connections between the deep surface of the ITB and the region of the lateral epicondyle remain. Herbst et al. (2017), in 20 fresh-frozen knee specimens (mean age 47.6 years), found typical Kaplan fibers in all 20, and in 16 (80%) there were also two thinner accessory condylar insertions anteroproximal to the lateral epicondyle[2]. Landreau et al. (2019; some authors disclosed royalty, consulting or research support relationships with Smith & Nephew, Medacta and Arthrex) dissected 16 unpaired fresh-frozen knees by an anterior approach (5 knees) or a posterior approach (11 knees): with the posterior approach, a flat condylar strap (mean thickness 0.88 ± 0.41 mm) was separated distal to the distal Kaplan bundle and the superior lateral genicular artery in all 11 specimens, connecting the region of the lateral epicondyle with the deep surface of the ITB and becoming qualitatively taut on clamp palpation during passive internal rotation at 30°–60° of flexion; with the anterior approach, the condylar strap could not be accurately identified in any of the 5 specimens, because reflecting the ITB cuts that deep connection[5].
Immediately proximal to the lateral epicondyle, the ITB gives off the capsulo-osseous layer (COL) into the deep plane. Godin et al. (2017) measured the center of the femoral attachment of the COL 19.1 mm proximal to the epicondyle and the center of the tibial attachment 19.5 mm from the center of Gerdy's tubercle[4]; Herbst et al. (2017), using the tip of Gerdy's tubercle and the lateral articular cartilage margin as references, measured the COL insertion 23.4 ± 1.6 mm posterior to Gerdy's tubercle and 6.6 ± 0.9 mm distal to the lateral cartilage margin[2]. The two studies used different bony reference points, but both confirm that the COL continues from above the lateral epicondyle to the lateral tibial tubercle behind Gerdy's tubercle and, together with the deep Kaplan fibers and the condylar strap, forms the anchoring system of the ITB on the posterolateral distal femur[2,4].
2. Localization of the Anterolateral Ligament and Interstudy Anatomic Differences
Claes et al. (2013) dissected 41 unpaired elderly embalmed knees (mean age 79 years) and isolated an ALL in 40 (97%), describing an origin from a prominence on the lateral epicondyle slightly anterior to the LCL origin, an oblique anteroinferior course with attachment to the middle third of the body of the lateral meniscus, the inferior lateral genicular vessels running between the ALL and the outer margin of the lateral meniscus, and a distal insertion near the midpoint of the line joining Gerdy's tubercle (21.6 ± 4.0 mm posterior to it) and the tip of the fibular head (23.2 ± 5.7 mm anterior to it), 6.5 ± 1.4 mm distal to the lateral cartilage margin; their schematic placed this structure within capsular layer 3[1]. Subsequent fresh-frozen dissections and the international consensus (Getgood et al., 2019) note that the femoral attachment of the ALL more often lies posterior and proximal to the lateral epicondyle and intermingles with fibers of the LCL origin, and locate it as a capsular thickening within the superficial layer of the lateral capsule (Seebacher layer 3), anterior to the LCL[6].
Comparing Herbst et al. (2017) with Daggett et al. (2018; some authors were Arthrex consultants or received educational support from several companies) shows that the anatomic differences mainly concern specimen condition, the method of exposure and the scope of the structural definition:
- Herbst et al. (2017), in 20 fresh-frozen knees without an applied external internal-rotation moment, separated the layers from superficial to deep and distinguished the capsulo-osseous layer of the ITB from the underlying anterolateral capsule; the authors found no independent superficial ligament that simultaneously matched all of the origin and insertion descriptions of Claes et al., but observed in 35% (7/20) of specimens a mid-third capsular ligament formed by fusion of the superficial and deep layers of the capsule anterior to the LCL[2].
- Daggett et al. (2018) included the paired knees of 12 fresh-frozen cadavers (24 knees, median age 76.5 years); 12 left knees underwent layer-by-layer gross dissection, and in 12 right knees the anterolateral tissues were resected en bloc and a single horizontal H&E histologic section was cut at the joint-line plane[3]. On gross dissection the authors assigned the ALL and the deep capsulo-osseous layer of the ITB to the same structure (layer 4), and on pages 4–5 of the text they note that this layer cannot consistently be separated from the underlying capsule; on the single-plane joint-line sections of the right knees, 8/12 specimens simultaneously showed four dense, regularly organized connective-tissue structures—the ITB, ALL/COL, lateral meniscus and LCL—while the remaining 4/12 sections contained only some of these structures (2 with meniscus alone, 2 with three structures), reflecting the fact that a single horizontal plane did not capture every structure at the same height in every specimen, not that the ligaments were anatomically absent[3].
In addition, Claes et al. (2013) divided the ITB 6 cm proximal to the epicondyle in embalmed specimens, reflected it distally and applied an internal-rotation moment at 60° of flexion to bring tensed fiber bundles into relief[1], whereas layer-by-layer dissection of fresh-frozen specimens (Herbst et al., 2017) or a posterior approach (Landreau et al., 2019) distinguished superficial from deep structures while preserving fascial continuity[2,5]. Together with differing assignments of the capsulo-osseous layer (to the deep ITB or into the ALL), this means that the tissue range denoted by "ALL" in the literature is not entirely the same. At the tibial insertion, the international consensus (Getgood et al., 2019) also states clearly that the site of a Segond avulsion fracture receives the combined attachments of the capsulo-osseous layer of the ITB, the anterolateral ligament/capsule and the anterior aponeurotic expansion of the short head of biceps femoris, and is not the exclusive insertion of any single ligament[6].
2. Mechanical Division of Labor by Flexion Angle and Experimental Conditions
Biomechanical experiments that assess the function of the anterolateral structures use mainly two methods. In one, a six-degree-of-freedom robot records a baseline motion trajectory, the target structures are sectioned in sequence, the trajectory is replayed, and the share of resistance or of restraining torque carried by each structure under that predetermined trajectory is calculated from the drop in load (for example the robotic series of Parsons et al., 2015; Kittl et al., 2016; and Willinger et al., 2023)[7,9,11]. In the other, structures are sectioned in sequence under a fixed applied load, and the increment in tibial internal-rotation angle or anterior translation after each cut is measured directly (for example the optical kinematic series of Geeslin et al., 2018, and Willinger et al., 2023)[10,11]. The two measures have different physical meanings and different denominators and cannot be compared interchangeably.
Table 2. Design and principal results of four cadaveric sequential soft-tissue sectioning biomechanical experiments
| Study and sample | Measurement method and loading | Handling of the superficial ITB and sectioning design | Principal anti–internal-rotation / simulated pivot-shift results |
|---|---|---|---|
| Parsons et al. (2015) [7,8]: 12 knees dissected, 11 tested | Robotic predetermined trajectory replay, share of resistance; 134 N anterior drawer, 5 N·m isolated internal rotation (0°–90°) | Before testing the superficial ITB was already freed and reflected to Gerdy's tubercle; medial parapatellar incision left open; block-randomized sectioning of ACL, LCL and ALL (sectioning the ALL also opened part of the underlying lateral capsule) | Under anterior drawer, ALL + part of capsule shared 1%–4% of resistance (ACL 49%–60%); under 5 N·m internal rotation the share rose from 5% at 0° to 30% at 35°, 37% at 45°, 44% at 60° and 45% at 75° |
| Kittl et al. (2016) [9]: 16 knees (8 ACL-intact, 8 ACL-deficient) | Robotic predetermined trajectory replay, share of resistance; 90 N anteroposterior translation, 5 N·m isolated internal rotation (0°–90°), simulated pivot shift (4 N·m internal rotation + 8 N·m valgus, 15°–45°) | Superficial ITB retained (no muscle load applied); fixed superficial-to-deep sequence: sITT (superficial ITB, biceps femoris complex and lateral retinaculum), dcITT (deep ITB + COL + Kaplan fibers), extracapsular ALL superficial to the LCL, anterolateral capsule Cap | Under 5 N·m internal rotation: in ACL-intact knees at 30°, dcITT 25.6% ± 9.0%, sITT 18.1% ± 9.1% (whole ITB 44% ± 10%, rising to 76% ± 11% at 60%), ALL 11.0% ± 9.7%; in ACL-deficient knees at 30°–90°, whole ITB 58%–80%, ALL only 2.3%–4.9% |
| Geeslin et al. (2018) [10]: 20 knees enrolled, 9 pairs / 18 knees analyzed | Robot under fixed load, laxity increments; 5 N·m internal rotation (0°–90°), simulated pivot shift (5 N·m internal rotation + 10 N·m valgus at 15° and 30°), 88 N anterior drawer | Superficial ITB insertion retained and the incision sutured; ACL cut first, then paired knees randomized to ALL first or proximal/distal Kaplan bundles first | In ACL-deficient knees at 60°, 75° and 90°, the internal-rotation increments after cutting the Kaplan bundles (+1.6°, +2.1°, +2.2°) were significantly larger than after cutting the ALL (+0.3°, +0.4°, +0.5°); under simulated pivot shift at 15° and 30°, the additional internal-rotation increment from cutting either structure alone was only 0.4°–0.5° |
| Willinger et al. (2023) [11]: 22 knees enrolled (robotic 10, kinematic analysis 10) | Robotic cohort (10 knees), share of resistance; optical kinematic cohort (another 10 knees), 5 N·m internal rotation and simulated pivot shift (5 N·m internal rotation + 8 N·m valgus) laxity increments | Superficial ITB retained; compared ACL, Kaplan fibers (KFs), combined anterolateral capsule and ALL (C/ALL), and lateral meniscus posterior root (LMPR); kinematic cohort used 3 sectioning orders (3–4 knees per subgroup) | Robotic cohort under 5 N·m internal rotation: at 90°, KFs shared 44% ± 23% of resistance, C/ALL 14% ± 13%; at 60° and 90°, KFs significantly higher than C/ALL; in the kinematic cohort, internal-rotation and simulated pivot-shift laxity increments were larger when KFs or C/ALL were cut later rather than first |
1. Force Sharing Between the Iliotibial Band and the ALL–Capsule Under Predetermined Trajectory Replay
As shown in Table 2, Parsons et al. (2015), in 11 fresh-frozen knees, found that under a 134 N anterior drawer the ALL sectioning step shared only 1%–4% of anterior resistance from 0° to 90°, whereas under 5 N·m isolated internal rotation its anti–internal-rotation contribution at 35° (30%), 45° (37%), 60° (44%) and 75° (45%) of flexion exceeded that of the ACL at the same angles (original Table 4, p. 673)[7].
Calculation error in the identification rate reported by Parsons et al. (2015) (p. 672): The first sentence of the Results on page 672 of that paper states that the ALL was present in all but one specimen ("83% identified")[7]. Against the Methods on page 670, the Discussion on page 673, Table 2 on page 671 and the legends to Figs. 3–4, 12 knees were dissected in the whole series and only 1 was excluded because no clear ALL was seen, so that 11 knees completed mechanical testing (11/12 is about 91.7%, whereas 83.3% corresponds to 10/12)[7]. This chapter does not use the erroneous 83%; nothing in the original text or in the subsequent correspondence shows that the percentage miscalculation altered the mechanical data from the 11 tested knees.
To understand the 30%–45% share of internal-rotation resistance measured by Parsons et al. (2015), two specific experimental conditions must be kept in view. Page 670 of the original and the authors' subsequent formal reply (Parsons et al., 2015, p. NP22)[7,8] state clearly that before testing the superficial ITB had already been freed from the proximal specimen and stripped to its distal insertion on Gerdy's tubercle, excluding any restraining effect of the superficial ITB on translation and internal rotation, and that the medial parapatellar incision was left open; at the same time, page 671 of the original and point 3 on page NP22 of the reply state clearly that when the ALL was cut at the level of the tibial plateau, part of the underlying lateral capsule was opened with it, so that the values measured at that step are in fact the combined contribution of "ALL plus part of the underlying lateral capsule," not that of an isolated ALL (a contemporaneous letter raised methodological points about interaction among capsular parts and about normalization of resultant force vectors; the authors' formal reply explained that net force vectors were subtracted and then renormalized, and the letter alone cannot be used to discard the whole data set)[7,8].
Kittl et al. (2016), in 16 knees with the superficial ITB retained (no additional muscle tension applied), sectioned in a fixed superficial-to-deep order sITT, dcITT, the extracapsular ALL superficial to the LCL, and the whole anterolateral capsule Cap, and measured the share of resistance by replaying the original intact trajectory (original Tables 1–3, pp. 350–351)[9]. Under 5 N·m isolated internal rotation, whether the ACL was intact or deficient, the whole ITB system (sITT + dcITT) carried the principal internal-rotation resistance at 30°–90° of flexion (44%–76% in ACL-intact knees, 58%–80% in ACL-deficient knees), whereas the extracapsular ALL accounted for 11.0% ± 9.7% at 30° in ACL-intact knees and only 4.4%, 4.9%, 2.3% and 3.3% at 0°, 30°, 60° and 90° in ACL-deficient knees (Cap did not exceed 4.1% at any angle); under simulated pivot shift in ACL-deficient knees (15°, 30°, 45°), the deep dcITT alone carried 27.2%–35.4% of the anti–internal-rotation torque (at 45° the whole ITB together accounted for 72% ± 14%), whereas the extracapsular ALL accounted for 6.9%, 5.8% and 2.9% at 15°, 30° and 45°, respectively[9].
The numerical differences between Parsons et al. and Kittl et al. are jointly influenced by several design factors: besides whether the superficial ITB was freed beforehand, the two studies differ in how they define the ALL and the capsule, in the tissue included in each sectioning step (step 1 of Kittl et al., sITT, includes not only the superficial ITB but also the tendons and aponeurotic arms of the long and short heads of biceps femoris and the lateral patellofemoral and patellomeniscal retinacula; see original Table 1, p. 350), in the order of sectioning and in the denominator used to calculate resistance percentages[7–9]. The contribution of 52.2%–55.8% for sITT at 60°–90° of flexion in Kittl et al. therefore represents the overall effect of that composite superficial-tissue sectioning step in the predetermined sequence, and cannot be attributed to the isolated superficial ITB alone; the shares of resistance listed above for the deep dcITT under isolated internal rotation and simulated pivot shift, by contrast, directly support an important antirotatory role for the deep ITB–Kaplan fiber–capsulo-osseous layer unit[9].
2. Laxity Increments and Synergistic Restraint Under Randomized and Multi-Sequence Sectioning
Geeslin et al. (2018; internal funding from the Steadman Philippon Research Institute, consumables provided by Arthrex, some authors disclosing consulting or funding relationships with Arthrex, Smith & Nephew, Ossur and ConMed) enrolled 10 pairs (20 knees) of fresh-frozen male paired knees, retained the superficial ITB insertion and sutured the incision after each step; the ACL was cut first, then paired knees were randomly assigned to "ALL first, then proximal/distal Kaplan fiber bundles" or the reverse order (Results–Specimen Preparation, p. 1355, record that the fourth pair was excluded after preliminary analysis because the kinematic change after cutting the Kaplan fibers exceeded the mean by 2 standard deviations, leaving 9 pairs / 18 knees)[10]. Under 5 N·m isolated internal rotation in ACL-deficient knees, cutting the ALL significantly increased internal rotation at every tested angle from 15° to 90° except 0°, and cutting the Kaplan fiber bundles significantly increased internal rotation at 15° and at 60°–90°; at 60°, 75° and 90° of flexion, the internal-rotation angle increments after cutting the Kaplan fiber bundles (+1.6°, +2.1°, +2.2°) were significantly larger than after cutting the ALL (+0.3°, +0.4°, +0.5°), whereas at 0°–45° the difference between the two increments was not statistically significant[10]. Under simulated pivot shift at 15° and 30° of flexion (5 N·m internal rotation + 10 N·m valgus), adding a cut of the ALL alone or of the Kaplan fiber bundles alone to the ACL-deficient state increased the internal-rotation angle by only an additional 0.4°–0.5° (anterior translation increments 0.3–0.6 mm, of which the 0.6 mm translation increment after cutting the Kaplan fibers at 30° was not statistically significant), whereas cutting both produced additive rotational and translational laxity[10]. This shows that under simulated pivot shift at low flexion angles, the absolute angular increments from isolated sectioning of a single secondary anterolateral structure are small, and the single-structure angular differences from in vitro experiments should not be taken directly as criteria for clinical surgical indications.
Willinger et al. (2023; the investigators' institution received funding from Smith & Nephew) enrolled 22 unpaired fresh-frozen knees and divided them into two independent 10-knee analysis cohorts: 10 knees on a robot for share of resistance under predetermined trajectory replay, and another 12 knees (10 after exclusion of 2 for technical failure) on an optical kinematic testing frame for free-range-of-motion laxity increments; the study cut the anterolateral capsule and the ALL together as a single complex (C/ALL)[11]. In the robotic cohort, relative to the sum of the structures measured in that experiment, the ACL carried 88% (at 90°) to 94% (at 0°) of anterior-drawer resistance; under 5 N·m internal rotation, the Kaplan fibers (KFs) carried 44% ± 23% of internal-rotation resistance at 90° of flexion and C/ALL carried 14% ± 13%, and at 60° and 90° of flexion the anti–internal-rotation contribution of the KFs was significantly higher than that of C/ALL[11]. The lateral meniscus posterior root (LMPR) mainly restrained valgus and posterior translation in ACL-intact knees, and cutting it did not produce significant anterolateral rotatory instability; in ACL-deficient knees, cutting the LMPR produced a small increase in anterolateral rotatory laxity[11]. In the optical kinematic cohort, because each of the three secondary-structure sectioning-order subgroups contained only 3–4 specimens, single-step sectioning increments were reported descriptively without hypothesis testing (original p. 1139): the descriptive results showed that when KFs or C/ALL were cut as a later step, the internal-rotation and simulated pivot-shift laxity increments they produced were larger than when they were cut first, supporting the view that KFs and C/ALL share and mutually compensate for the anterolateral anti–internal-rotation load[11].
3. Scope of the Recent Systematic Review (Hohmann et al., 2025) and Verification Against the Primary Data
Hohmann et al. (2025) published a systematic review of 22 cadaveric biomechanical studies (284 specimens). Although the title mentions ACL-deficient and ACL-reconstructed knees, the Abstract and the conclusions in the body state that, in zero-time-point biomechanical testing of cadaveric specimens, no conclusive evidence was found that the ALL plays a significant role in restraining anterior tibial translation or internal rotation in ACL-intact or ACL-reconstructed knees[12]. When citing this review, its scope and statistical handling must be stated clearly: the inclusion criteria on page 4883 excluded literature that studied the ITB, the Kaplan fibers or lateral extra-articular tenodesis (LET), so the review does not assess the function of the ITB–Kaplan fiber system; at the same time, pages 4884 and 4896 state clearly that pooled summary effect sizes were not calculated for the load–displacement kinematic data at each flexion angle, and that the forest plots simply list individual studies (most comparisons containing only 2–7 small-sample studies)[12].
Checking the core summary table and Discussion of Hohmann et al. (2025) back against the primary experimental papers also reveals several original-text errors that affect quantitative interpretation:
Data transcription, group assignment and citation errors in the systematic review by Hohmann et al. (2025) (pp. 4886, 4897–4898, 4899.e1):
- Errors in sample size, flexion angle and column assignment for Parsons et al. (2015) (Table 2, p. 4886): Table 2 lists the specimen number for Parsons et al.[7] as 41 (the primary study in fact dissected 12 knees and mechanically tested 11), misprints the 37% internal-rotation contribution at 45° of flexion as "4°: 37%" (original Table 4 is 45°: 37%), and places that group's ALL resistance percentages, which were measured by replaying an intact baseline trajectory, in the "ACL Deficient" column of Table 2, although the original experiment did not measure those percentages in a separate ACL-deficient cohort[7,12].
- Misattribution of resistance shares among different anatomic structures from Kittl et al. (2016) (Table 2, p. 4886): Table 2 lists the ALL internal-rotation contribution of Kittl et al.[9] in ACL-intact knees at 30° of flexion as "30°: 26%". Against Table 2 on page 350 of Kittl et al., the 25.6% ± 9.0% (about 26%) at 30° of flexion is in fact the contribution of the deep ITB and capsulo-osseous layer (dcITT); under the same conditions the extracapsular ALL contributed 11.0% ± 9.7% (capsule Cap 3.1% ± 2.7%)[9,12].
- Confusion of testing angles and loading protocols (Discussion p. 4897 and Appendix p. 4899.e1): The Discussion and Appendix misstate the internal-rotation contribution at 35° of flexion from Table 4 of Parsons et al. (30%) as 30° of flexion, and mismatch the isolated-internal-rotation ALL contributions measured by Kittl et al. in ACL-deficient knees (2%–5%; original Table 2 actually measured at 0°, 30°, 60° and 90° of flexion, corresponding to 4.4%, 4.9%, 2.3% and 3.3%) onto the flexion angles of the simulated pivot shift (15°–45°, original Table 3)[7,9,12]. At the same time, the Discussion on page 4897 attributes the differences between the two studies to "Parsons et al. using isolated internal rotation and Kittl et al. using combined loads," omitting the fact that Kittl et al. likewise completed 5 N·m isolated internal-rotation testing and overlooking the differences between the two studies in whether the superficial ITB was stripped and in the scope of structural sectioning[7,9,12].
- Mismatched primary-study citations (Table 2, p. 4886, and References, p. 4898): In Table 2 the reference superscript for Kittl et al. (2016) is 7, but reference 7 on page 4898 is in fact the 2018 review by Kittl et al. in Clin Sports Med, not the 2016 Am J Sports Med primary experimental paper[12].
This chapter therefore does not use the erroneous quantitative transcriptions or protocol attributions for those two experiments in Table 2 and the Discussion of Hohmann et al. (2025); the relevant numbers are taken from the primary experimental papers. The review's methodological discussion of the small samples, older specimen ages, high heterogeneity of testing protocols and limited ability of in vitro results to reflect living clinical conditions in zero-time-point cadaveric studies is retained[12].
3. Clinical Interpretation of the Pivot Shift and Acute Imaging
The clinical pivot-shift test applies a combined internal-rotation and valgus load in roughly the 15°–45° flexion interval and reflects the combined restraining state of the intra-articular ACL, the superficial and deep ITB–Kaplan fiber–capsulo-osseous layer unit, the anterolateral ligament and capsule, and the lateral meniscus, among other structures[6,11]. Because the superficial and deep anterolateral structures act synergistically and can compensate for one another in restraining internal rotation, and because soft-tissue injury may undergo repair and remodeling as time after injury passes, an abnormal signal in a single anterolateral structure on preoperative MRI does not stand in a simple one-to-one relationship with the grade of the pivot-shift test under anesthesia.
Tokura et al. (2024) retrospectively assessed 105 patients with primary ACL injury, of whom 98 had both the ALL and the Kaplan fibers identifiable on preoperative MRI: no injury in 43 (43.9%), isolated ALL injury in 32 (32.7%), isolated Kaplan fiber injury in 12 (12.2%), and combined ALL and Kaplan fiber injury in 11 (11.2%)[13]. Neither the preoperative manual pivot-shift IKDC grade under anesthesia (P = .247) nor tibial reduction acceleration measured by an electromagnetic system (group means 1.4, 1.4, 1.4 and 1.7 m/s², P = .856) differed statistically significantly among the four groups[13].
Wording error in the formula for reduction acceleration in Tokura et al. (2024) (p. 41): Section 2.4 on page 41 of that paper describes the tibial reduction acceleration measured by the electromagnetic system as the "second derivative of the velocity of the anteroposterior translation with respect to time"[13]. In physics the first derivative of velocity with respect to time (or the second derivative of displacement with respect to time) is acceleration (unit m/s²); the second derivative of velocity with respect to time is in fact jerk (unit m/s³), which contradicts the unit (m/s²) reported in Table 2 on page 42 of the same paper[13]. This is a methodological wording error and cannot on that account be taken to show that the acceleration values actually output by the device software were miscalculated; the manual pivot-shift grades under anesthesia are unaffected by the wording error. This chapter retains the report that the manual and electromagnetic measurements showed no significant between-group differences, but the acceleration calculation cannot be reproduced from the wording of that sentence as written.
The absence of between-group pivot-shift differences in that study is constrained by two features of its design: MRI for the whole series was completed at a mean of about 6.0–8.8 days after injury, whereas the pivot-shift test under anesthesia was performed preoperatively, that is at a mean of 70.0–82.0 days after injury[13], so that the two examinations were separated by about two and a half months (although the study did not obtain a preoperative repeat MRI or measure tissue stiffness directly, subacute soft-tissue repair during the waiting period may weaken the association between acute imaging injury and the preoperative pivot shift); in addition, the combined ALL and Kaplan fiber injury group contained only 11 patients, a small sample[13].
Cui et al. (2024) included 540 patients with acute ACL tear who completed MRI and a manual pivot-shift test under anesthesia within 14 days of injury (mean about 7.3 days) (476 adults, 64 adolescents), and reported that in adults partial or complete ALL rupture (grades 2–3) was associated with a high-grade pivot shift (grades 2–3) (OR = 2.271, 95% CI 1.499–3.442), whereas in adolescents Kaplan fiber injury (grades 1–3) was associated with a high-grade pivot shift (OR = 19.98, 95% CI 2.367–168.669)[14]. Checking the formally published pages of that paper, however, Tables 1–3 and the Discussion contain several numerical contradictions and unstable regression estimates:
Within-table count contradictions, quasi-complete separation with unstable regression, and conflicting statements in the text of Cui et al. (2024) (pp. 3–6):
- Contradictions in basic demographic and meniscal injury counts in Tables 1–2 (pp. 4, Tables 1–2): In Table 1 the adolescent group (n = 64) has an age range of 11–17 years, yet the mean and SD are given as 16.1 ± 7.8 years (an SD of 7.8 years is impossible within a range of only 6 years); Table 1 lists 362 men and 112 women in the adult group, summing to 474 rather than the 476 in the table header (the whole series 400 + 138 = 538 rather than 540); lateral meniscus anterior horn (LMA) injury is listed as 0 (0.0%) in adults and 3 (4.7%) in adolescents in Table 1, yet the total column of Table 1 gives 18 (3.3%), and Table 2 reports 4 (2.0%) LMA injuries in the adult high-grade pivot-shift subgroup (n = 201), contradicting the 0 in adults in Table 1[14].
- Quasi-complete separation in the adolescent subgroup regression, anomalous output for a zero-variance variable, and conflicting statements of significance in the text (Tables 1–2 on p. 4 and Table 3 and Discussion on p. 5): From Table 1 (64 adolescents in total) and Table 2 (36 adolescents with a high-grade pivot shift, leaving 28 with a low-grade pivot shift), all 18 adolescents with Kaplan fiber partial/complete rupture (KF disruption, grades 2–3) were in the high-grade pivot-shift group (univariable 2×2 cell frequencies 18, 0, 18, 28), and all 3 with LMA injury were likewise in the high-grade group, producing quasi-complete separation and sparse cells; the multivariable logistic regression for adolescent KF disruption in Table 3 on page 5 is unstable, giving an anomalous extreme OR = 1.64×10⁹ (95% CI lower bound 0, upper bound missing, P = .998), and LMA injury likewise gives an anomalous extreme OR = 9.26×10¹⁰ (P = .999); at the same time, the zero-variance variable "adolescent medial meniscus anterior horn (MMA) injury," with counts of 0 in both Tables 1 and 2, nonetheless yields a finite estimate OR = 0.147 (95% CI 0.008–2.553, P = .188) in Table 3. In addition, the Discussion on page 5 first states that the association with "KF injury" in adolescents was no longer significant in multivariable analysis, then in the next sentence states that adolescent "KF injury" increased the risk of a high-grade pivot shift nearly 20-fold, whereas Table 3 in fact shows P = .006 for KF injury (OR = 19.98) and P = .998 for KF disruption, so that the terminology and the statements of significance in the text contradict each other[14].
- Inconsistent definition of posterior lateral meniscus injury (Methods p. 3 and Limitations p. 6): The Methods on page 3 state that posterior lateral meniscus injury (LMP) includes longitudinal tears and posterior root tears, whereas the Limitations on page 6 state that the combined analysis excluded root tears ("except for root tear"), an inconsistency[14].
Given the within-table data contradictions and the unstable adolescent regression model, this chapter does not use the adolescent subgroup OR = 19.98 of Cui et al. (2024) as an independent effect size, as evidence that age determines which anterolateral structure is the dominant stabilizer, or as a basis for choosing a surgical procedure, nor does it read their posterior meniscus data as a clear lateral meniscus posterior root effect. The acute-phase frequency distributions presented in Tables 1–2 (for example that all 18 adolescents with Kaplan fiber grade 2–3 rupture within 14 days of injury were in the high-grade pivot-shift subgroup) and the association reported in adults (OR = 2.271) are retained only as acute cross-sectional observational clues; until the original count contradictions are clarified, those numbers cannot serve as stable clinical prediction rules or surgical decision criteria, and an observation window of 14 days after injury, or comparison across studies, does not itself prove that tissue has undergone no healing or that any given examination window has a proven diagnostic advantage.
When preoperative MRI of patients with ACL injury is read clinically, signal abnormality in the anterolateral structures should be assessed together with the interval from injury to imaging, concomitant meniscal injury and the actual findings of the pivot-shift test under anesthesia; imaging abnormality of a single anterolateral structure, or the small angular increments that follow soft-tissue sectioning in vitro, should not be equated directly with the clinical grade of rotatory instability.
References
[1] Claes S, Vereecke E, Maes M, Victor J, Verdonk P, Bellemans J. Anatomy of the anterolateral ligament of the knee. J Anat. 2013;223(4):321-328. PMID: 23906341. DOI: 10.1111/joa.12087
[2] Herbst E, Albers M, Burnham JM, Fu FH, Musahl V. The Anterolateral Complex of the Knee. Orthop J Sports Med. 2017;5(10):2325967117730805. PMID: 29051903. DOI: 10.1177/2325967117730805
[3] Daggett M, Stephenson C, Dobson J, et al. Anatomic and Histological Study of the Anterolateral Aspect of the Knee: A SANTI Group Investigation. Orthop J Sports Med. 2018;6(10):2325967118799970. PMID: 30345320. DOI: 10.1177/2325967118799970
[4] Godin JA, Chahla J, Moatshe G, et al. A Comprehensive Reanalysis of the Distal Iliotibial Band: Quantitative Anatomy, Radiographic Markers, and Biomechanical Properties. Am J Sports Med. 2017;45(11):2595-2603. PMID: 28609131. DOI: 10.1177/0363546517707961
[5] Landreau P, Catteeuw A, Hamie F, Saithna A, Sonnery-Cottet B, Smigielski R. Anatomic Study and Reanalysis of the Nomenclature of the Anterolateral Complex of the Knee Focusing on the Distal Iliotibial Band: Identification and Description of the Condylar Strap. Orthop J Sports Med. 2019;7(1):2325967118818064. PMID: 30729141. DOI: 10.1177/2325967118818064
[6] Getgood A, Brown C, Lording T, et al. The anterolateral complex of the knee: results from the International ALC Consensus Group Meeting. Knee Surg Sports Traumatol Arthrosc. 2019;27(1):166-176. PMID: 30046994. DOI: 10.1007/s00167-018-5072-6
[7] Parsons EM, Gee AO, Spiekerman C, Cavanagh PR. The biomechanical function of the anterolateral ligament of the knee. Am J Sports Med. 2015;43(3):669-674. PMID: 25556221. DOI: 10.1177/0363546514562751
[8] Parsons EM, Gee AO, Spiekerman C, Cavanagh PR. The Biomechanical Function of the Anterolateral Ligament of the Knee: Response. Am J Sports Med. 2015;43(8):NP22. PMID: 26232458. DOI: 10.1177/0363546515597218
[9] Kittl C, El-Daou H, Athwal KK, et al. The Role of the Anterolateral Structures and the ACL in Controlling Laxity of the Intact and ACL-Deficient Knee. Am J Sports Med. 2016;44(2):345-354. PMID: 26657572. DOI: 10.1177/0363546515614312
[10] Geeslin AG, Chahla J, Moatshe G, et al. Anterolateral Knee Extra-articular Stabilizers: A Robotic Sectioning Study of the Anterolateral Ligament and Distal Iliotibial Band Kaplan Fibers. Am J Sports Med. 2018;46(6):1352-1361. PMID: 29558208. DOI: 10.1177/0363546518759053
[11] Willinger L, Athwal KK, Holthof S, Imhoff AB, Williams A, Amis AA. Role of the Anterior Cruciate Ligament, Anterolateral Complex, and Lateral Meniscus Posterior Root in Anterolateral Rotatory Knee Instability: A Biomechanical Study. Am J Sports Med. 2023;51(5):1136-1145. PMID: 36917838. DOI: 10.1177/03635465231161071
[12] Hohmann E, Keough N, Molepo M, Arciero R, Imhoff A. The Anterolateral Ligament Complex Has Limited Impact on Anterior Tibial Translation or Internal Rotational Stability in Anterior Cruciate Ligament-Deficient and Anterior Cruciate Ligament-Reconstructed Knees: A Systematic Review of Biomechanical Cadaver Studies. Arthroscopy. 2025;41(11):4882-4899.e15. PMID: 40306468. DOI: 10.1016/j.arthro.2025.04.043
[13] Tokura T, Nagai K, Hoshino Y, et al. Injuries to both anterolateral ligament and Kaplan fiber of the iliotibial band do not increase preoperative pivot-shift phenomenon in ACL injury. Asia Pac J Sports Med Arthrosc Rehabil Technol. 2024;36:40-44. PMID: 38595931. DOI: 10.1016/j.asmart.2024.03.004
[14] Cui J, Cheng S, Luo Y, et al. Influence of Injury to the Kaplan Fibers and Anterolateral Ligament on Pivot-Shift Phenomenon in Adult and Adolescent Patients With Anterior Cruciate Ligament Injury: A Retrospective Cohort Study. Orthop J Sports Med. 2024;12(11):23259671241288250. PMID: 39512602. DOI: 10.1177/23259671241288250