Abstract
Background:
For anterolateral rotatory instability as a result of secondary soft tissue injuries in anterior cruciate ligament (ACL)-deficient knees, there is increasing interest in secondary stabilizers to prevent internal rotation (IR) of the tibia.
Purpose:
To determine which secondary stabilizer is more important in anterolateral rotatory instability in ACL-deficient knees.
Study Design:
Controlled laboratory study.
Methods:
The lower extremities of 10 fresh-frozen cadavers (20 extremities) without anterior-posterior or rotational instability were included. Matched-pair randomization was performed, with each side per specimen assigned to 1 of 2 groups. In group 1, the ACL was sectioned, followed by the anterolateral ligament (ALL); in group 2, the ACL was sectioned, followed by sequential sectioning of the posterolateral meniscocapsular complex (PLMCC) and posteromedial meniscocapsular complex (PMMCC). The primary outcome was the change in relative tibial IR during a simulated pivot-shift test with 5 N·m of IR torque and 8.9 N of valgus force. The secondary outcomes were the International Knee Documentation Committee grade in the pivot-shift test and the incidence of the grade 3 pivot shift.
Results:
In group 1, compared with baseline, the change in relative tibial IR at 0° of knee flexion was 1.4° (95% CI, –0.1° to 2.9°; P = .052) after ALL release. In group 2, it was 2.5° (95% CI, 0.4° to 4.8°; P = .007) after PLMCC release and 4.1° (95% CI, 0.5° to 7.8°; P = .017) after combined PLMCC and PMMCC release. Combined PLMCC and PMMCC release resulted in greater change of tibial IR with statistical significance at 0°, 15°, and 30° of knee flexion (P = .008, .057, and .004, respectively) compared with ALL release. The incidence of grade 3 pivot shifts was 10% in group 1 and 90% in group 2.
Conclusion:
Posterior meniscocapsular laxity caused an increase in relative tibial IR as much as ALL injury in ACL-deficient knees in our simulated laboratory test, and greater anterolateral rotatory instability occurred with posterior meniscocapsular injury compared with ALL injury.
Clinical Relevance:
Repair of the injured posterior meniscocapsular complex may be an important treatment option for reducing anterolateral rotatory instability in the ACL-deficient knee.
Keywords
Anterolateral rotatory instability is commonly observed in anterior cruciate ligament (ACL)-deficient knees, and the pathologic motion is the combined result of tibial internal rotation (IR) and anterior tibial translation. Residual rotational instability had been shown to occur in more than 20% of patients after ACL reconstruction. 40 In previous studies, a residual anterolateral rotatory instability was associated with persistent knee discomfort and the progression of osteoarthritis. 19,22 The cause of anterolateral rotatory instability is still unknown; it is thought be associated with multiple factors such as meniscus, meniscotibial ligament, collateral ligament, joint capsule, anterolateral complex, and bony morphology of the femoral condyle and tibial plateau. 41 Recently, secondary soft tissue injury in ACL-deficient knees is emphasized; thus, there is increasing interest in secondary stabilizers to prevent tibial IR. 38
The anterolateral ligament (ALL) is the most highlighted secondary stabilizer to prevent tibial IR. 5,11,34 Recent clinical studies also showed an improvement in anterolateral rotatory instability with ALL reconstruction. 23 However, several anatomic studies have reported various results for the anatomic structure of the ALL, and there is controversy regarding whether the ALL is a clearly distinguished functional anatomic structure. 5,9,15,42 Other biomechanical studies reported that the ALL plays a minor role in controlling tibial IR in ACL-deficient knees. 21,32,35 A recent clinical study reported that an ALL injury detected by magnetic resonance imaging did not affect rotational instability. 27 Based on the various evidence mentioned above, reasonable doubts have been raised as to whether the ALL is a contributing factor to the anterolateral rotatory instability.
In another view, a few authors advocated that ramp lesions and associated posterior meniscocapsular laxity (Figure 1), which are commonly observed associated injuries in the ACL-deficient knee, were associated with anterolateral rotational instability ‡‡ ; the ramp lesion and posterior meniscocapsular laxity are also attracting attention as secondary stabilizers to prevent rotational instability.

Diagnostic arthroscopic findings of associated ramp lesions (arrows) in an ACL-deficient knee with knee flexed at 90°. (A) Notch view, with 70° optic, rotated at 6 o’clock. In acute ACL injury, posteromedial ramp lesion could be observed between posteromedial meniscotibial ligament and medial meniscus. (B) Notch view, with 70° optic, rotated at 4 o’clock. In acute ACL injury, posterolateral ramp lesion could be observed between posterolateral meniscotibial ligament and lateral meniscus. (C) Notch view, with 70° optic, rotated at 7 o’clock. In the chronic ACL-deficient knee, the posteromedial meniscotibial ligament is attached to the medial meniscus, but the ligament is shown stretched and the space between medial meniscus and posteromedial joint capsule is widened. ACL, anterior cruciate ligament.
If considering the relative tibial movement according to the position of the femur in anterolateral rotatory instability, 38 ramp lesions and posterior meniscocapsular laxity might also be the cause of anterolateral rotatory instability.
The purpose of the study was to determine which secondary stabilizer injury plays a more important role in the anterolateral rotatory instability in ACL-deficient knee through a simulated cadaveric experiment. The hypothesis was that simulated ramp lesions combined with posterior meniscocapsular laxity will produce a comparable degree of anterolateral rotatory instability to an ALL injury.
Methods
Trial Design
This cadaveric experiment was conducted between March 1 and December 31, 2019. Institutional review board approval was not required because deidentified cadaveric specimens are exempt from review at our institution. Full-limb cadaveric specimens including the pelvis were used for this simulated laboratory study (Figure 2). Subjective physical examination was performed with the anterior drawer test and pivot-shift test by a single board-certified orthopaedic surgeon (D.K.L.) with experience in sports medicine and knee arthroscopy, and specimens without anterior and posterior as well as rotational instability were included. A total of 14 fresh-frozen cadavers (28 extremities) with no formaldehyde preservation were included in the study. Four specimens (8 extremities) were used to perform the pilot test and to establish the experimental protocol, and the remaining 10 specimens (20 extremities) were included in the testing (mean age, 80.2 years; range, 70-94 years; 4 male and 6 female cadavers; mean height, 160.8 cm; range, 152-172 cm). All specimens were stored at room temperature for 12 hours before the procedures.

Flowchart of the study. ACL, anterior cruciate ligament; ALL, anterolateral ligament; AP, anterior and posterior; exam, examination; IR, internal rotation; IKDC, International Knee Documentation Committee; PLMCC, posterolateral meniscocapsular complex; PMMCC, posteromedial meniscocapsular complex.
Repetitive passive cyclic flexion/extension movement and 5 N·m of internal and external rotational torque were applied to all specimens to reduce the potential effect of joint stiffness and rigidity. 8 To control for bony morphological differences among individual specimens, 1 extremity of each cadaver was assigned randomly to group 1, and the opposite extremity was assigned to group 2. In group 1, the ACL was sectioned followed by the ALL. In group 2, the ACL was sectioned followed by sectioning of the posterolateral meniscocapsular complex (PLMCC), followed by sectioning of the posteromedial meniscocapsular complex (PMMCC). In group 2, the order of sectioning was determined based on the assumption that PLMCC would be expected to occur in the setting of acute ACL injury, while PMMCC is a consequence of ongoing ACL insufficiency.
Measurement of Tibial IR
To simulate the clinical situation in which the actual pivot shift occurred when performing a physical examination, the instrument (Figure 3) was manufactured to apply constant rotational torque without cutting the lower extremity. Therefore, the anatomic structures of the muscles, tendons, and ligaments could be preserved. The device was also designed to allow the knee flexion to change at 15° intervals. The round rotating plate was connected to the cast boot. Therefore, the rotation of the ankle was restricted when applying the constant IR torque.

Instrument used to apply constant rotational torque without cutting the lower extremity, preserving the muscles, tendons, and ligaments. Based on previous literature, 5 N·m of IR torque and 8.9 N of valgus force were applied for the experiment. The round rotating plate was connected to the cast boot, and the rotation of the ankle was restricted when applying the constant IR torque. IR, internal rotation.
Experiment Protocol
The experimental protocol was set after reviewing the existing research method. 2,8 The specific experiment protocol was as follows. Constant rotational torque and valgus force were applied for the experiment. Based on previous literature, the appropriate force value of valgus and IR torque were determined. 2,8 The radius of the round rotating plate was 4.5 cm, the IR torque value was 5 N·m by loading 25 pounds (11.3 kg), and 2 pounds (0.9 kg) of weight were applied to create a valgus force of 8.9 N (Figure 3 and Appendix Figure A1). An electronic inclinometer was used for measuring tibial IR. To measure the relative bony tibial IR compared with the femur, the femoral rotation was manually fixed to 0° in the inclinometer with an external fixator. Each measured value was measured twice, and the average value of the 2 measurements was used for statistical analysis. Intraobserver reliability was evaluated to confirm that the manufactured instrument had appropriate accuracy to measure the change in rotation.
Anatomic Dissection Protocol of Each Structure
All surgical procedures were performed by a single board-certified orthopaedic surgeon (D.K.L.). The biomechanical study was performed by 3 medical personnel: 2 board-certified orthopaedic surgeons (D.K.L. and J.H.K.) and 1 medical student (J.T.). The average time of testing for 1 specimen was approximately 8 to 9 hours. To confirm the hypothesis of the study, we tried to reimplement an anatomic model similar to the clinical situation, in which posterior meniscocapsular laxity is produced by repetitive pathologic motion in chronic ACL insufficiency. However, we thought that it was difficult to reproduce sufficient meniscocapsular laxity of the clinical situation with simple cutting between the periphery of the posterior meniscus and capsule. Therefore, additional posterolateral and posteromedial capsular release was performed at the same time. The anatomic dissection protocol is described below.
Step 1: ACL Release
A medial parapatellar approach was performed to expose the ACL. The ACL was cut completely at the tibial attachment with a scalpel (Figure 4A). After cutting the ACL, the joint capsule was closed meticulously.

Images of anterior dissection during experimental protocol steps 1 and 2. (A) ACL cut at the tibial attachment. (B) ALL cut at the femoral insertion site. ACL, anterior cruciate ligament; ALL, anterolateral ligament; Ant, anterior; Lat, lateral; Med, medial; Post, posterior.
Step 2: ALL Release
The Gerdy tubercle and lateral epicondyle were palpated before the incision. A longitudinal incision was made connecting the 2 anatomic landmarks, and the iliotibial band was split to expose the ALL. During the procedure, care was taken not to damage the Kaplan fibers. Based on previous studies, the anatomic position of the ALL was confirmed before the procedure. 11,34 Since it is difficult to confirm the exact anatomic location of the ALL due to the capsular structure, the tight part of the capsule was determined when IR torque was applied first (Supplemental Video 1). 7 Considering the possibility that the ALL could be incompletely cut, the tight part was detached from the lateral epicondyle in a single large flap 1 cm in width. At the same time, the location of the lateral collateral ligament was checked while taking care not to damage the lateral collateral ligament. The flap was dissected distally to the level at which the lateral meniscus was exposed (Figure 4B). After complete dissection of the ALL, the iliotibial band was meticulously repaired.
Step 3: PLMCC Release
The posterior capsule was exposed through the posterior approach between both gastrocnemius muscles. The superior border of the lateral meniscus posterior horn was confirmed using the spinal needle to determine the position of transverse incision. Then, a 4-cm transverse incision was made using a scalpel along the superior border of the lateral meniscus posterior horn. A 2-cm vertical incision was added in the proximal direction at the medial incision starting point. Therefore, iatrogenic posterolateral meniscocapsular laxity was reproduced, and the posterolateral meniscotibial ligament was exposed. Complete cutting was performed with an additional 4-cm vertical incision between the posterolateral meniscotibial ligament and lateral meniscus to simulate a posterolateral ramp lesion (Figure 5, A and B). During the procedure, the location of the popliteus and posterior cruciate ligament was checked, and care was taken not to damage these structures.
Step 4: PMMCC Release
Posteromedial meniscocapsular laxity and posteromedial ramp lesion were simulated in the same manner as above (Figure 5C). During the procedure, the locations of the posterior oblique ligament and posterior cruciate ligament were checked, and care was taken not to damage these structures.

Images of posterior dissection during experimental protocol steps 3 and 4. (A) Posterior capsule was exposed through the posterior approach between both gastrocnemius muscles. (B) L-shaped incision (arrows) made at the posterolateral capsule to simulate posterolateral meniscocapsular laxity. Additional complete cutting was performed between posterolateral meniscotibial ligament and lateral meniscus to simulate a posterolateral ramp lesion. (C) L-shaped incision (arrows) made at the posteromedial capsule to simulate posteromedial meniscocapsular laxity. Additional complete cutting was performed between posteromedial meniscotibial ligament and medial meniscus to simulate a posteromedial ramp lesion. Lat, lateral; Med, medial; PL, posterolateral; PM, posteromedial.
Outcome Measures
The primary outcome was the change in relative tibial IR when 5 N·m of IR torque was applied with 8.9 N of valgus force. The secondary outcomes were the International Knee Documentation Committee (IKDC) grade in the pivot-shift test and the incidence of grade 3 pivot shift. The primary and secondary outcomes were measured in each step, including premanipulation status. Tibial IR was measured at 0°, 15°, and 30° of knee flexion in each step. After measuring tibial IR, the instrument was removed, and 2 orthopaedic surgeons (D.K.L. and J.H.K.) independently evaluated the IKDC grade in the pivot-shift test, recorded as 0 (normal), 1 (glide), 2 (clunk), or 3 (locked subluxation). 41 Interobserver reliability of the IKDC grading was evaluated.
Statistical Analysis
The sample size was determined based on a previous similar cadaveric study in which rotational instability was measured. 36 Intraobserver reliability of the tibial IR measurements and interobserver reliability of the IKDC evaluation were calculated using the intraclass correlation coefficient (ICC).
The change in relative tibial IR was determined as the difference from the premanipulation baseline value. The Wilcoxon signed-rank test was performed to confirm which anatomic structure most affected the change in relative tibial IR. The Mann-Whitney test was performed to compare the change in relative tibial IR between the groups. A P value of <.05 was considered to be statistically significant, and P values of ≥.05 and <.10 were considered to be borderline statistically significant. The measured values were analyzed using SPSS (Version 23.0; IBM Corp).
Results
The ICC values for intraobserver reliability of the tibial IR measurements and for interobserver reliability of the IKDC grading were 0.999 and 0.922, respectively, indicating excellent reliability for both measures.
Change in Relative Tibial IR
The change in relative tibial IR at different knee flexion angles for both experimental groups is shown in Figure 6. After ALL release in group 1, the change in tibial IR from ACL release alone was 1.0° (95% CI, –0.9° to 1.5°; P = .052), 1.4 (95% CI, –0.3° to 3.7°; P = .024), and 0.7 (95% CI, –0.7° to 3.1°; P = .053) at 0°, 15°, and 30° of knee flexion, respectively.

Change in relative tibial IR at different knee flexion angles. The relative tibial IR increased after each anatomic structure was released (ACL, ALL, PLMCC, and PMMCC). Compared with ALL release, greater change in tibial IR was seen after combined PLMCC and PMMCC release at 0°, 15°, and 30° of knee flexion (P = .008, .057, and .004). P values in red indicate statistically significant differences (P < .05) ACL, anterior cruciate ligament; ALL, anterolateral ligament; IR, internal rotation; PLMCC, posterolateral meniscocapsular complex; PLMCS, posterolateral meniscocapsular complex; PMMCC, posteromedial meniscocapsular complex; PMMCS, posteromedial meniscocapsular complex.
After PLMCC release in group 2, the change of tibial IR from ACL release alone was 2.0° (95% CI, –0.6° to 4.1°; P = .007), 2.2 (95% CI, 0.5° to 5.1°; P = .005), and 1.7 (95% CI, 0.1° to 3.2°; P = .005) at 0°, 15°, and 30° of knee flexion, respectively. After PMMCC release in group 2, the change of tibial IR from PLMCC release was 1.6 (95% CI, –0.7° to 4.8°; P = .017), 1.0 (95% CI, –0.9° to 2.9°; P = .074), and 0.9 (95% CI, –0.9° to 4.4°; P = .066) at 0°, 15°, and 30° of knee flexion, respectively. The total change of tibial IR from ACL release to combined PLMCC and PMMCC release was 4.1 (95% CI, 0.5° to 7.8°; P = .017), 3.6 (95% CI, 1.6° to 8.1°; P = .074), and 3.1 (95% CI, 1.2° to 6.5°; P = .066) at 0°, 15°, and 30° of knee flexion, respectively. Compared with ALL release, greater change in tibial IR was seen after combined PLMCC and PMMCC release at 0°, 15°, and 30° of knee flexion (P = .008, .057, and .004, respectively).
Change in Pivot-Shift Grade and Incidence of Grade 3 Pivot Shift
The IKDC grade in the pivot-shift test increased gradually after the release of each anatomic structure (ACL, ALL, PLMCC, and PMMCC). In group 1, 10% of the extremities (1/10) showed grade 3 pivot shift. In group 2, 90% of the extremities (9/10) showed a grade 3 pivot shift. The specific changes of the pivot shift after releasing each anatomic structure are shown in Table 1 and Supplemental Video 2.
IKDC Grade Change in the Pivot-Shift Test a
a Data are reported as No. of specimens (%). ACL, anterior cruciate ligament; ALL, anterolateral ligament; IKDC, International Knee Documentation Committee; PLMCC, posterolateral meniscocapsular complex; PMMCC, posteromedial meniscocapsular complex.
Discussion
The principal finding of this study was that posterior meniscocapsular laxity has an impact on the change in relative tibial IR comparable with that of an ALL injury. Grade 3 pivot shift also showed higher incidence after combined PLMCC and PMMCC release. The results suggest that not only ALL injury but also posterior meniscocapsular laxity contribute considerably to anterolateral rotatory instability in the ACL-deficient knee. In our laboratory simulation, ALL injury also affected tibial IR in early knee flexion and the IKDC grade in the pivot-shift test. However, the change in relative tibial IR at full extension was statistically greater in the posterior meniscocapsular release group than in the ALL release group in our study. Considering that tibial subluxation occurs with the knee nearly fully extended in a pivot shift, the role of an ALL injury as the secondary stabilizer to prevent tibial IR was not as great as that of posterior meniscocapsular laxity.
Similar opinions have also been reported in other studies. 17,21,25,30,32,43 In a study that detected the presence of ALL injuries with magnetic resonance imaging evaluations, the degree of the anterolateral rotatory instability was not correlated with the presence of ALL injuries. 27 In addition, several authors have pointed out that the effect on anterolateral rotatory instability might not be significant because the ALL acts mainly during knee flexion. 21,25,43 Nevertheless, recent clinical studies have shown that combined ACL and ALL reconstruction improves residual anterolateral rotational instability. 23
The pivot-shift mechanism can be explained by 2 factors. The first is posterolateral ramp lesion and posterolateral meniscocapsular laxity. In acute ACL injury, posterolateral subluxation of the femur (relative tibial IR) can cause additional injuries to the PLMCC, with lateral compartment bony contusion. 31 PLMCC acts as a secondary stabilizer that resists posterolateral subluxation of the femur in the ACL-deficient knee. 20,25 In the case of substantial damage to the PLMCC or remaining laxity without adequate treatment, posterolateral subluxation of the femur is increased. Therefore, relative tibial IR is increased, as shown in our study results. Recently, several studies have also shown that anterolateral rotatory instability is associated with PLMCC injury. 4,6,16,18,29 Some studies reported that anterolateral rotatory instability could be improved by meniscocapsular repair. 16,18,20 In another recent study, it was reported that the presence of a lateral notch sign greater than 2 mm, suggesting severe bone contusion and high-energy trauma, is associated with a high-grade pivot shift. 26 Considering the greater bone contusion associated with more soft tissue damage at the PLMCC, the result also indirectly suggests that PLMCC injury might affect anterolateral rotatory instability.
The second factor is the anteromedial shifting of the rotational axis caused by ACL injury and posteromedial ramp lesion with posteromedial meniscocapsular laxity. In states of ACL insufficiency, the axis of the pivot is shifted in the anteromedial direction, 30,37 and the PMMCC acts as a secondary stabilizer for resisting anterior translation of the tibia and rotational instability of the femur. 8 Cam-like action of the femur and stress concentration on the medial meniscus posterior horn induces laxity of the PMMCC by repeated tibial anterior translation and pivot-shift motion if the ACL is not treated in the early phase. 37 Furthermore, the laxity of the PMMCC induces anterior translation of the tibia, which causes the rotational axis to be displaced further anteriorly. 1,8,10,14,36,39 Therefore, relative tibial IR is increased, as shown in our study results. Other biomechanical studies showed similar results, 8,10,36 and several previous clinical studies reported that PMMCC laxity is associated with tibial anterior instability. 1,14,39
The characteristics of anterolateral rotatory instability of the knee are similar to those of shoulder anterior instability. Anterior dislocation of the shoulder injures the anterior capsule, labrum, and glenoid and creates a Bankart lesion. Similarly, posterolateral subluxation of the femur results in posterolateral ramp lesion. In addition, the lateral notch sign of the bone contusion and deformation could occur in the knee, similar to the Hill-Sachs lesion in the shoulder. Interestingly, damaged anterior capsulolabral complex repair, which is called Bankart repair, has been shown to have successful surgical outcomes in the shoulder joint. However, in rotational instability of the knee, ALL reconstruction became the preferred treatment. Recently, with growing interest in anterolateral rotatory instability, surgical indications for ALL reconstruction during ACL reconstruction have tended to increase, even in patients with a primary ACL injury, in an effort to diminish the remaining rotational instability and reduce ACL graft failure. We agree that ALL reconstruction is an effective and powerful surgical option to improve rotational instability, 23 but it might be concerning to perform ALL reconstruction in the primary ACL-deficient knee because treatment of ramp lesions could be neglected during ALL reconstruction in the limited surgical time. Several recent studies have also emphasized that the repair of ramp lesions during ACL reconstruction is the crucial factor in improving rotational stability. 8,16,18,20,28,36
Limitations
The limitations of this study were as follows. First, there is a possibility of β errors occurring due to small sample size; some of the values in our results may not be statistically significant. Second, the mean age of the cadavers in the experiment was old, and the degree of osteoarthritis or meniscal injury was not evaluated. Hence, the present results might be different from results in a clinical situation with younger patients. However, we tried to increase the reliability in limited research circumstances by excluding cadavers with anterior and posterior as well as rotational instability. Third, there might be a problem with the validity of the measuring method, as pivot shifts were not evaluated with existing validated methods (navigation system, electromagnetic sensor, or accelerometer, and so on). Therefore, anterior translation was not assessed in a simulated laboratory test. Fourth, the cutting sequence used in our study may have limitations. In a clinical setting, medial meniscus ramp lesions may occur independently without lateral meniscus ramp lesions. Although the order of cutting had to be established for the purposes of the study, it is possible that it may not be representative of the clinical situation. Another limitation is that this is a time-zero study that aimed to simulate cumulative capsular injury, which may not occur as a single injury through repeated pivot-shift injuries. The mechanical properties could not be reimplemented with a single cut at the meniscocapsular junctional area. Therefore, capsular sectioning was performed to simulate a condition characterized by chronic stretching. The results could be limited in a real clinical situation because the analysis has time-zero mechanical properties and improper cadaver storage or handling of biological tissues, including multiple cycles of freezing and refreezing, can adversely degrade the mechanical properties of tissues for biomechanical investigation.
Conclusion
Posterior meniscocapsular laxity caused as great an increase in relative tibial IR as ALL injury in the ACL-deficient knee in our simulated laboratory test; greater anterolateral rotatory instability occurred with posterior meniscocapsular injury compared with ALL injury. Repair of the injured posterior meniscocapsular complex may be an important treatment option for reducing anterolateral rotatory instability in the ACL-deficient knee.
A video supplement for this article is available at https://journals.sagepub.com/doi/full/10.1177/23259671231188712#supplementary-materials.
Footnotes
Acknowledgment
The authors thank the Department of Anatomy, Catholic Institute for Applied Anatomy, College of Medicine, The Catholic University of Korea, for providing the cadaveric specimens and general support. D.K.L. appreciates all coauthors for their efforts in conducting this study.
Notes
Final revision submitted March 15, 2023; accepted April 24, 2023.
The authors have declared that there are no conflicts of interest in the authorship and publication of this contribution. AOSSM checks author disclosures against the Open Payments Database (OPD). AOSSM has not conducted an independent investigation on the OPD and disclaims any liability or responsibility relating thereto.
Ethical approval was not sought for the present study.
Appendix
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
