- Email: [email protected]
Importance of the posterior bundle of the medial ulnar collateral ligament
ARTICLE IN PRESS J Shoulder Elbow Surg (2016) ■■, ■■–■■
www.elsevier.com/locate/ymse
ORIGINAL ARTICLE
Importance of the posterior bundle of the medial ulnar collateral ligament Dave R. Shukla, MDa,*, Elan Golan, MDb, Philip Nasser, MSMEa, Maya Culbertson, MFSb, Michael Hausman, MDa a
Leni & Peter May Department of Orthopaedic Surgery, Icahn School of Medicine at Mount Sinai, New York, NY, USA Maimonides Department of Orthopaedic Surgery, Maimonides Medical Center, Brooklyn, NY, USA
b
Background: There has been a renewed interest in the pathomechanics of elbow dislocation, with recent literature having suggested that the medial ulnar collateral ligament is more often disrupted in dislocations than the lateral ligamentous complex. The purpose of this serial sectioning study was to determine the influence of the posterior bundle of the medial ulnar collateral ligament (pMUCL) as a stabilizer against elbow dislocation. Methods: An elbow dislocation was simulated in 5 cadaveric elbows by mechanically applying an external rotation moment and valgus force. Medial ulnohumeral joint gapping was measured at 30°, 60°, and 90° of flexion in an intact elbow after sectioning of the medial collateral ligament’s anterior bundle (aMUCL) and then after sectioning of the pMUCL as well. Results: After sectioning of the aMUCL, the pMUCL was able to stabilize the joint against dislocation. After aMUCL sectioning, the proximal joint space significantly increased by 4.2 ± 0.6 mm at 30° of flexion and 2.6 ± 0.3 mm at 60° of flexion, although it did not dislocate. The gapping increase of 0.9 ± 0.6 at 90° of flexion did not reach significance. After sectioning of the pMUCL (after having already sectioned the aMUCL), all of the specimens frankly dislocated at all flexion angles. Conclusions: An intact pMUCL can prevent elbow dislocation and limited joint subluxation to within 6.6 mm. Our findings indicate that repair or reconstruction may be warranted in certain circumstances (ie, residual instability after operative management of a terrible triad injury or after aMUCL reconstruction). Level of evidence: Basic Science Study; Biomechanics © 2016 Journal of Shoulder and Elbow Surgery Board of Trustees. All rights reserved. Keywords: Elbow instability; elbow dislocation; medial collateral ligament posterior bundle; posterolateral rotatory instability; elbow trauma; recurrent elbow instability
This study was conducted at the Icahn School of Medicine at Mount Sinai, New York, NY, USA. *Reprint requests: Dave R. Shukla, MD, Leni & Peter May Department of Orthopaedic Surgery, Icahn School of Medicine at Mount Sinai, 5 East 98th Street, New York, NY 10029, USA. E-mail address: [email protected] (D.R. Shukla).
An elbow dislocation can result from complete disruption of the stabilizing soft tissue structures, particularly in cases in which the osseous stabilizers remain uninjured. Classic teaching has indicated that disruption of the soft tissues originates from the lateral side, progressing anteriorly and then medially,14-16 described as the circle of Horii. However, there is a lack of consensus agreement on the origin of the
1058-2746/$ - see front matter © 2016 Journal of Shoulder and Elbow Surgery Board of Trustees. All rights reserved. http://dx.doi.org/10.1016/j.jse.2016.04.008
ARTICLE IN PRESS 2 disruptive forces as other reports have discussed a possible medially originating mechanism.8,18,19,21 Of the medial collateral ligamentous complex, the anterior bundle of the medial ulnar collateral ligament (aMUCL) is the primary restraint against valgus instability.7 The posterior bundle of the medial ulnar collateral ligament (pMUCL) has been described as a minor secondary constraint.3 Several biomechanical studies have discussed the role of the pMUCL in conferring stability within the context of evaluating other ligaments.2-4,7,12 Whereas the pMUCL’s influence on stabilizing against valgus forces and rotation moments has been studied, the importance of the posterior bundle as a secondary stabilizer against dislocation, which involves several simultaneous forces, has not been quantified. The purpose of this serial sectioning study was to determine the influence of the pMUCL as a stabilizer against elbow dislocation and medial joint instability and to introduce a novel method of quantifying elbow stability than has been traditionally used.
Methods Specimen preparation Five right-sided fresh frozen cadaveric elbows, which had been transected at the mid humerus and the mid forearm, were obtained from a cadaveric donor program. All of the specimens were from female donors, with an average age of 77 (range, 74-79) years. Each specimen was fully thawed at room temperature before dissection.
Specimen preparation All of the skin and subcutaneous tissues were dissected to the layer of the investing fascia. All of the muscles were sharply excised, except for the common extensors. The joint capsule and medial and lateral ligamentous joint complexes were left intact. The specimens were secured in aluminum pots using polymethyl methacrylate to allow attachment to the fixture within the testing machine. A 10 × 20mm anterior capsular window was created to assist with proper joint alignment during potting and to ensure congruous articular contact during testing. The humeral shaft was potted into one aluminum pot, and the radius and ulna were separately potted into another aluminum pot. After potting of the radius and ulna, an osteotomy of the radial shaft was performed, always distal to the bicipital tuberosity. This was performed so that a fixed radius would not limit motion and alter the kinematics through pivoting of the elbow about a constrained, rigid radiocapitellar articulation during the simulated dislocation event. During potting, care was taken to ensure that the natural carrying angle of each elbow was taken into account and that the articular surface remained parallel to the base of the pot (ie, horizontal).
D.R. Shukla et al. The humeral pot was mounted on a custom jig that allowed adjustment of the elbow flexion angle, in increments of 15° from 0° to 90° (Fig. 1). The jig was mounted onto a custom-made, lowfriction X-Y stage, which itself was mounted into a 6-axis load cell.
Mechanical testing A dislocation event that originated with disruption of the medialsided soft tissue stabilizers was mechanically simulated using a servohydraulic materials testing system (MTS) machine (MTS Systems, Eden Prairie, MN, USA), and joint gapping was measured using a 3-dimensional (3D) motion tracking camera system (Vicon, Denver, CO, USA). The MTS machine was equipped with a low-friction X-Y stage, which was mounted on a 6-axis load cell (ATI Industrial Automation, Apex, NC, USA). The MTS machine was equipped with an actuator that allowed the application of an axial load, a torsional moment (ie, rotation), and bending in the X and Y planes. The forearm pot was mounted into the clamp that was attached to this actuator (ie, top of the machine) and secured with a collar clamp. The humeral pot was secured to the jig at the bottom of the machine (Fig. 1). Once the specimen was secured within the clamps and before the application of any loads, it was ensured that the joint was congruous and seated in its natural position. This was accomplished by direct visualization of the anterior ulnohumeral articulation and by performing minor adjustments in rotation of both the humeral and forearm pots, such that the load cells indicated that the forces across the joint were as close to 0 as possible (always <0.05 N · m of torque). After this step, the specimen was subjected to the cycle of loading. After application of an axial load of 25 N that was applied to simulate a joint compressive force, the machine placed the elbow into 5° of valgus by bending the forearm pot in the X-plane. An external rotation moment of 2.5 N · m was simultaneously applied. The motion of the ulna was similar to that which would occur in a posterolateral rotatory subluxation event. This meant that the forearm pot was externally rotated until the load cell detected a torsional force of 2.5 N · m. Once 2.5 N · m of torque was reached, the machine stopped torque application and then internally rotated the forearm
Mounting of the specimen The forearm (ie, radius and ulna pot) was secured into the multiaxis hydraulic actuator, which applied the axial load, valgus force, and external rotation moment.
Figure 1 Schematic of the testing apparatus showing the direction of axial load (25 N), rotation of the X-axis motor causing a 5° valgus angulation, and external rotation of the actuator.
ARTICLE IN PRESS Posterior ulnar collateral ligament prevents instability
3
Figure 2 (A) Schematic of the medial ulnohumeral articulation. The circles with letters represent the placement of the markers. The distance between markers A and C was termed the proximal joint, and the distance between markers B and C was termed the distal joint. (B and C) The anterior bundle (B) and the posterior bundle (C) of the medial collateral ligament (MCL) were sharply incised at their respective ligamentous midsubstances, along the joint line. back to its original position, while the valgus bending returned from its position of 5° to 0°.
3D motion capture Gapping at the medial ulnohumeral joint was measured using the Vicon 3D motion capture system. The system used in this study relied on optical–passive motion capture, which used retroreflective markers that were tracked by 4 infrared cameras. The cameras were placed in the appropriate positions around the specimen and calibrated before each testing session in accordance with the manufacturer’s recommendations. The accuracy of the system’s measurement was validated before specimen testing. This was accomplished through translation of the markers by a known value that was then correlated to the value observed by the tracking system. The system’s error was found to be within 0.05 mm. The reflective markers were 6.4 mm in diameter. The Vicon system tracked the center points of the markers throughout their course. The markers were mounted on truncated 0.045-inch K-wires, which were then placed in the subchondral bone in the following locations: marker A was placed at the most proximal insertion point of the pMUCL on the olecranon (Fig. 2, A); marker B was placed just posterior to the insertion of the aMUCL on the humerus; marker C was placed just posterior to the aMUCL, just proximal to the articular surface; and marker X was placed on the medial epicondyle. Marker X was not used in the measurements but was used, as recommended by the manufacturer, to increase the accuracy of the tracking system. The distance between markers A and C was termed proximal joint gapping; the distance between markers B and C was termed distal joint gapping.
scribed by Floris et al,5 confirmation of the anterior bundle’s fibers was facilitated by identification of a macroscopically visible ridge. After sectioning of the aMUCL, the elbow was again tested at each flexion angle. After testing of the “injured” (ie, cut aMUCL) elbow at each of the 3 flexion angles, the pMUCL was transected (Fig. 2, C), and the elbow was again tested at each flexion angle.
Data analysis All data were reported as the mean ± standard error. The data were modeled with the use of pair-wise comparisons using a Student t-test, with a significance level of P ≤ .05.
Results After sectioning of the aMUCL, the ulnohumeral articulation became unstable, but the pMUCL was able to prevent frank dislocation. The proximal joint space increased by 4.2 ± 0.6 mm (P = .002) at 30° of flexion, 2.6 ± 0.3 mm (P = .005) at 60° of flexion, and 0.9 ± 0.6 (P = .9) at 90° of flexion (Fig. 3). The distal joint space increased by 3.8 ± 0.6 mm (P = .004) at 30°, 3.3 ± 0.8 mm (P = .02) at 60°, and 1.9 ± 1.1 mm (P = .18) at 90° (Fig. 4). After sectioning of the pMUCL (after having already sectioned the aMUCL), all of the specimens frankly dislocated at all flexion angles and failed to reach the threshold torque limit that had been set at 2.5 N · m, beyond which the machine stopped applying any additional torque.
Surgical conditions
Discussion First, the intact, uninjured specimen was tested at 30°, 60°, and 90° of flexion. To change the elbow flexion angle, the inferior (humeral pot) collar clamp was loosened and the jig was adjusted. Care was taken to ensure that the ulnohumeral articulation remained well seated before retightening of the collar clamp in preparation for the next testing cycle. After testing of the elbow at 90° of flexion (ie, completion of the “intact” cycle), the aMUCL was sharply incised along the medial ulnohumeral joint at the midportion of the ligament (Fig. 2, B). Care was taken to ensure that only the aMUCL was cut. As de-
This study demonstrated that under a simultaneously applied axial valgus load and external rotation moment, with the aMUCL sectioned and with the radiocapitellar center fixed, preventing radiocapitellar subluxation or lateral ligamentous complex injury, an intact pMUCL can prevent posterolateral dislocation and limit the extent of valgus laxity and ulnohumeral subluxation.
ARTICLE IN PRESS 4 Our findings emphasize the importance of the pMUCL in conferring some stability to an injured elbow, particularly at 30° and 60° of ulnohumeral joint flexion. After the aMUCL was incised, the maximal amount of proximal and distal joint gapping that occurred in any specimen was 6.6 mm and 5.5 mm, respectively. Both of these maximal values occurred at 30° of elbow flexion. We did not find that there was a significant increase in joint gapping after the aMUCL injury at 90°. This was likely due to the inherent osseous stability of the joint at 90° of flexion, given that the bone constraints were not injured in this study (ie, coronoid and ulnohumeral articulation).
Figure 3 There was an increase in proximal ulnohumeral joint gapping after injuring the anterior bundle of the medial collateral ligament (aMUCL) at both 30° (P = .002) and 60° (P = .005) of elbow flexion. Data are means ± standard error of the mean. The single (*) and double (**) asterisks denote significant differences between columns.
Figure 4 There was an increase in distal ulnohumeral joint gapping after injuring the anterior bundle of the medial collateral ligament (aMUCL) at both 30° (P = .004) and 60° (P = .02) of elbow flexion. Data are means ± standard error of the mean. The single (*) and double (**) asterisks denote significant differences between columns.
D.R. Shukla et al. These data prove that the pMUCL does function in more than just a minor capacity in its contribution to elbow stability, and the findings are important as they contribute to the presently minimal knowledge of the pMUCL. Whereas the role of the aMUCL as the primary restraint against valgus instability has been studied and established,3,7,13 there are relatively fewer studies that have focused on the biomechanical importance of the pMUCL.2-5,8,16 Furthermore, although the MUCL has been examined within the context of other studies, there are even fewer biomechanical studies that have focused specifically on the posterior bundle of the MUCL.17 The present data carry several clinical implications, particularly relating to operative management of elbow fracturedislocations (ie, terrible triad injuries). For terrible triad injuries, an accepted surgical tactic is to address the medial collateral ligament only if instability persists after management of the coronoid, radial head, and lateral collateral ligamentous complex. It has been recommended that one need not address the MUCL if the elbow is congruous from 30° to full flexion in at least 1 position of forearm rotation9; other authors have reported that addressing the MUCL is not usually necessary.6,10 However, this study demonstrates that the pMUCL is an important secondary stabilizer against external rotation and valgus instability. After repair or reconstruction of the injured structures following a fracture-dislocation (ie, coronoid, radial head, and lateral ulnar collateral ligament), it may be prudent to assess the integrity of the pMUCL and to repair it along with the aMUCL; this may protect the injured structures and further minimize the risk of recurrent instability or fixation failure. The findings of this study are also important given the recent data suggesting that a medially originating dislocation mechanism is more common than previously thought. It has been classically taught that most elbow fracturedislocations occur secondary to a posterolateral rotatory force. Internal rotation of the body relative to a fixed forearm11,14-16,22 causes disruption of the lateral stabilizers, progressing anteriorly and then medially. However, there are recent compelling studies that support a medial-sided initiation of injury. Schreiber et al demonstrated, through both a review of magnetic resonance imaging19 and by review of 62 in vivo elbow dislocations captured on YouTube,20 that most elbow dislocations occur secondary to a hyperphysiologic valgus moment on an extended elbow, noting that this would require a requisite MUCL disruption, followed by soft tissue disruption from the medial to lateral side. Rhyou and Kim, through evaluation of magnetic resonance imaging, observed a high proportion of MUCL distraction injuries along with valgus bone contusion that affected the radial head and capitellum, which supported their belief that dislocations may originate on the ulnar side.18 Our data demonstrate that the pMUCL confers secondary stability against dislocation secondary to this external rotation and valgus mechanism. Several of our observations are consistent with findings from other studies. We noted the greatest increase in both proximal and distal joint gapping at 30° of flexion. Morrey et al
ARTICLE IN PRESS Posterior ulnar collateral ligament prevents instability demonstrated that the ulnar collateral ligament was the primary restraint against valgus loads at 30° of flexion,13 whereas Callaway et al reported that specifically the posterior bundle was a secondary restraint at 30° of flexion.3 Of the cited studies that have examined the ulnar collateral ligament, only the data reported by Pollock et al17 have focused specifically on the pMUCL, rather than within the context of a study examining the entire ulnar collateral ligament. That study, which would most closely approximate this study, examined the change in varus and valgus angulation with passive elbow flexion and different forearm rotations after sectioning of the pMUCL. This study differs primarily in 2 ways. First, in this study, the aMUCL was sectioned, followed by pMUCL sectioning. In the study by Pollock et al,17 the pMUCL was sectioned, but the aMUCL was never sectioned. Second, this study differs in that 3 forces were simultaneously, not sequentially, applied to simulate a dislocation event. Therefore, a direct comparison of our data to their findings would not be appropriate. The methods of this study differ from several of the existing biomechanical studies that have examined medialsided elbow instability. Other investigators have used MTS-generated torque-rotation curves,3 bone markers with camera-capture,1 potentiometers with strain gauges,4,5,21 and electromagnetic tracking systems.2 Whereas an MTS machine was also used in this study to effect the simulated dislocation event, elbow instability was directly measured using 3D precision markers within an error of 0.05 mm in this study. We believe this to be a reliable and accurate system for measuring joint instability. There are several limitations to this study. First, we were unable to incorporate the radiocapitellar articulation in an unhindered manner that would appropriately simulate normal physiology, given the restrictions to our testing method. However, care was taken to leave the lateral soft tissues uninterrupted, including the common extensor mass and lateral collateral ligament complex. The lateral ulnar collateral ligament remained intact, as care was always taken to preserve both its distal humeral origin and insertion on the crista supinatoris. Another limitation is that the common flexor mass, which is an important secondary stabilizer, was not activated. Our method of assessment of the joint space gapping would not have been possible with an intact common flexor mass. A third limitation is that there was some inherent constraint within the system, as the humerus and forearm were potted in polymethyl methacrylate. The proximal radial osteotomy was performed in an effort to minimize the constraint involving the radiocapitellar articulation. It is possible that if the radius osteotomy were not performed and the radius was potted in varying degrees of rotation (ie, possibly full pronation, neutral, and full supination), different results may have been noted. However this would have tripled the number of specimens required and would have resulted in a significant increase in the resources required for this study. This, combined with the complexity of achieving the simultaneous application of forces, may have resulted in our being
5 unable to replicate a physiologic dislocation with complete accuracy. However, 3 pilot specimens were used before any actual testing in an attempt to eliminate as many sources of error as possible, and we believe our results to be valid. A final limitation is that an anterior capsular window was created to visualize the ulnohumeral articulation and to ensure congruency before testing. It is possible that the crossing capsular fibers may have conferred some stability to the elbow, particularly at low flexion angles. However, we do not think that an intact anterior capsule would have reduced the valgus instability to a degree such that a significant difference between surgical conditions would not have been observed. The findings of this study demonstrate that the posterior bundle of the MUCL is an important secondary stabilizer in the context of posterolateral rotatory instability. It is unlikely that the pMUCL is ever injured in isolation. Rather, it usually is attenuated or ruptured in conjunction with other stabilizing structures, such as the lateral collateral ligament complex and aMUCL in a posterolateral dislocation, or along with a radial head fracture and coronoid fracture in a terrible triad fracture-dislocation. During operative management of such injuries, it is important that these other stabilizers be repaired or reconstructed if indicated. The aMUCL remains the primary ligament that prevents valgus instability. After injury to the aMUCL, some stability will be lost and subluxation will occur, most notably between 30° and 60° of flexion. However, an intact pMUCL can prevent frank dislocation. In circumstances that necessitate exposure or operative management of the medial ligamentous complex (ie, fracturedislocation), we recommend surgically addressing the pMUCL by performing a repair, at a minimum. If a surgeon is performing a procedure with the intent of restoring stability to an injured joint, it is beneficial to repair all of the structures that might improve the stability of that joint to provide the greatest chance of success. In addition, in the event that the repair or reconstruction of an aMUCL fails, which is uncommon but does occur, the pMUCL can limit severe instability. As we did not evaluate the most optimal method of surgically managing pMUCL (ie, repair vs. reconstruction), we did not include a detailed discussion of reconstruction as that would not fall within the scope of this study. However, that is currently the subject of biomechanical evaluation and is currently under way at our institution. Repair of the native pMUCL can be performed using any suture configuration that approximates the ligamentous tissue if there is a midsubstance injury. If the pMUCL has avulsed from either the humerus or ulna, the surgeon can employ whichever method of repair that he or she is most comfortable with (eg, bone tunnels, suture anchor). As regards reconstruction, work at our institution is currently under way to determine the optimal method, and we cannot at present recommend 1 method in particular. At present, the senior author (M.H.) has routinely surgically addressed the pMUCL in circumstances of residual medial instability after operative fixation of terrible triad fracture-dislocations, with promising results (unpublished data).
ARTICLE IN PRESS 6
D.R. Shukla et al.
Further study is required to determine the optimal manner in which the pMUCL should be addressed.
7. 8.
Conclusion 9.
The pMUCL can confer some stability to the aMUCLdeficient elbow in a simulated ulnohumeral dislocation model. In the setting of an acute injury or residual instability after surgical management of an elbow dislocation, repair or reconstruction of the pMUCL may be warranted.
10.
11.
Disclaimer
12.
The authors, their immediate families, and any research foundation with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.
13.
14. 15. 16.
References 1. Armstrong AD, Dunning CE, Ferreira LM, Faber KJ, Johnson JA, King GJ. A biomechanical comparison of four reconstruction techniques for the medial collateral ligament–deficient elbow. J Shoulder Elbow Surg 2005;14:207-15. http://dx.doi.org/10.1016/j.jse.2004.06.006 2. Armstrong AD, Ferreira LM, Dunning CE, Johnson JA, King GJ. The medial collateral ligament of the elbow is not isometric: an in vitro biomechanical study. Am J Sports Med 2004;32:85-90. http://dx.doi.org/10.1177/0363546503258886 3. Callaway GH, Field LD, Deng XH, Torzilli PA, O’Brien SJ, Altchek DW, et al. Biomechanical evaluation of the medial collateral ligament of the elbow. J Bone Joint Surg Am 1997;79:1223-31. 4. Eygendaal D, Olsen BS, Jensen SL, Seki A, Sojbjerg JO. Kinematics of partial and total ruptures of the medial collateral ligament of the elbow. J Shoulder Elbow Surg 1999;8:612-6. 5. Floris S, Olsen BS, Dalstra M, Sojbjerg JO, Sneppen O. The medial collateral ligament of the elbow joint: anatomy and kinematics. J Shoulder Elbow Surg 1998;7:345-51. 6. Forthman C, Henket M, Ring DC. Elbow dislocation with intra-articular fracture: the results of operative treatment without repair of the medial
17.
18.
19.
20.
21.
22.
collateral ligament. J Hand Surg Am 2007;32:1200-9. http://dx .doi.org/10.1016/j.jhsa.2007.06.019 Hotchkiss RN, Weiland AJ. Valgus stability of the elbow. J Orthop Res 1987;5:372-7. Josefsson PO, Johnell O, Wendeberg B. Ligamentous injuries in dislocations of the elbow joint. Clin Orthop Relat Res 1987;221-5. Mathew PK, Athwal GS, King GJ. Terrible triad injury of the elbow: current concepts. J Am Acad Orthop Surg 2009;17:137-51. McKee MD, Pugh DM, Wild LM, Schemitsch EH, King GJ. Standard surgical protocol to treat elbow dislocations with radial head and coronoid fractures. Surgical technique. J Bone Joint Surg Am 2005;87(Suppl 1):22-32. http://dx.doi.org/10.2106/JBJS.D.02933 McKee MD, Schemitsch EH, Sala MJ, O’Driscoll SW. The pathoanatomy of lateral ligamentous disruption in complex elbow instability. J Shoulder Elbow Surg 2003;12:391-6. http://dx.doi .org/10.1016/S1058-2746(03)00027-2 Morrey BF, An KN. Functional anatomy of the ligaments of the elbow. Clin Orthop Relat Res 1985;84-90. Morrey BF, Tanaka S, An KN. Valgus stability of the elbow. A definition of primary and secondary constraints. Clin Orthop Relat Res 1991;18795. O’Driscoll SW, Bell DF, Morrey BF. Posterolateral rotatory instability of the elbow. J Bone Joint Surg Am 1991;73:440-6. O’Driscoll SW, Morrey BF, Korinek S, An KN. Elbow subluxation and dislocation. A spectrum of instability. Clin Orthop Relat Res 1992;186-97. Osborne G, Cotterill P. Recurrent dislocation of the elbow. J Bone Joint Surg Br 1966;48:340-6. Pollock JW, Brownhill J, Ferreira LM, McDonald CP, Johnson JA, King GJ. Effect of the posterior bundle of the medial collateral ligament on elbow stability. J Hand Surg Am 2009;34:116-23. http://dx.doi.org/10 .1016/j.jhsa.2008.09.016 Rhyou IH, Kim YS. New mechanism of the posterior elbow dislocation. Knee Surg Sports Traumatol Arthrosc 2012;20:2535-41. http://dx.doi.org/10.1007/s00167-011-1872-7 Schreiber JJ, Potter HG, Warren RF, Hotchkiss RN, Daluiski A. Magnetic resonance imaging findings in acute elbow dislocation: insight into mechanism. J Hand Surg Am 2014;39:199-205. http://dx.doi.org/10 .1016/j.jhsa.2013.11.031 Schreiber JJ, Warren RF, Hotchkiss RN, Daluiski A. An online video investigation into the mechanism of elbow dislocation. J Hand Surg Am 2013;38:488-94. http://dx.doi.org/10.1016/j.jhsa.2012.12.017 Sojbjerg JO, Helmig P, Kjaersgaard-Andersen P. Dislocation of the elbow: an experimental study of the ligamentous injuries. Orthopedics 1989;12:461-3. Wake H, Hashizume H, Nishida K, Inoue H, Nagayama N. Biomechanical analysis of the mechanism of elbow fracture-dislocations by compression force. J Orthop Sci 2004;9:44-50. http://dx.doi.org/10 .1007/s00776-003-0735-6
www.elsevier.com/locate/ymse
ORIGINAL ARTICLE
Importance of the posterior bundle of the medial ulnar collateral ligament Dave R. Shukla, MDa,*, Elan Golan, MDb, Philip Nasser, MSMEa, Maya Culbertson, MFSb, Michael Hausman, MDa a
Leni & Peter May Department of Orthopaedic Surgery, Icahn School of Medicine at Mount Sinai, New York, NY, USA Maimonides Department of Orthopaedic Surgery, Maimonides Medical Center, Brooklyn, NY, USA
b
Background: There has been a renewed interest in the pathomechanics of elbow dislocation, with recent literature having suggested that the medial ulnar collateral ligament is more often disrupted in dislocations than the lateral ligamentous complex. The purpose of this serial sectioning study was to determine the influence of the posterior bundle of the medial ulnar collateral ligament (pMUCL) as a stabilizer against elbow dislocation. Methods: An elbow dislocation was simulated in 5 cadaveric elbows by mechanically applying an external rotation moment and valgus force. Medial ulnohumeral joint gapping was measured at 30°, 60°, and 90° of flexion in an intact elbow after sectioning of the medial collateral ligament’s anterior bundle (aMUCL) and then after sectioning of the pMUCL as well. Results: After sectioning of the aMUCL, the pMUCL was able to stabilize the joint against dislocation. After aMUCL sectioning, the proximal joint space significantly increased by 4.2 ± 0.6 mm at 30° of flexion and 2.6 ± 0.3 mm at 60° of flexion, although it did not dislocate. The gapping increase of 0.9 ± 0.6 at 90° of flexion did not reach significance. After sectioning of the pMUCL (after having already sectioned the aMUCL), all of the specimens frankly dislocated at all flexion angles. Conclusions: An intact pMUCL can prevent elbow dislocation and limited joint subluxation to within 6.6 mm. Our findings indicate that repair or reconstruction may be warranted in certain circumstances (ie, residual instability after operative management of a terrible triad injury or after aMUCL reconstruction). Level of evidence: Basic Science Study; Biomechanics © 2016 Journal of Shoulder and Elbow Surgery Board of Trustees. All rights reserved. Keywords: Elbow instability; elbow dislocation; medial collateral ligament posterior bundle; posterolateral rotatory instability; elbow trauma; recurrent elbow instability
This study was conducted at the Icahn School of Medicine at Mount Sinai, New York, NY, USA. *Reprint requests: Dave R. Shukla, MD, Leni & Peter May Department of Orthopaedic Surgery, Icahn School of Medicine at Mount Sinai, 5 East 98th Street, New York, NY 10029, USA. E-mail address: [email protected] (D.R. Shukla).
An elbow dislocation can result from complete disruption of the stabilizing soft tissue structures, particularly in cases in which the osseous stabilizers remain uninjured. Classic teaching has indicated that disruption of the soft tissues originates from the lateral side, progressing anteriorly and then medially,14-16 described as the circle of Horii. However, there is a lack of consensus agreement on the origin of the
1058-2746/$ - see front matter © 2016 Journal of Shoulder and Elbow Surgery Board of Trustees. All rights reserved. http://dx.doi.org/10.1016/j.jse.2016.04.008
ARTICLE IN PRESS 2 disruptive forces as other reports have discussed a possible medially originating mechanism.8,18,19,21 Of the medial collateral ligamentous complex, the anterior bundle of the medial ulnar collateral ligament (aMUCL) is the primary restraint against valgus instability.7 The posterior bundle of the medial ulnar collateral ligament (pMUCL) has been described as a minor secondary constraint.3 Several biomechanical studies have discussed the role of the pMUCL in conferring stability within the context of evaluating other ligaments.2-4,7,12 Whereas the pMUCL’s influence on stabilizing against valgus forces and rotation moments has been studied, the importance of the posterior bundle as a secondary stabilizer against dislocation, which involves several simultaneous forces, has not been quantified. The purpose of this serial sectioning study was to determine the influence of the pMUCL as a stabilizer against elbow dislocation and medial joint instability and to introduce a novel method of quantifying elbow stability than has been traditionally used.
Methods Specimen preparation Five right-sided fresh frozen cadaveric elbows, which had been transected at the mid humerus and the mid forearm, were obtained from a cadaveric donor program. All of the specimens were from female donors, with an average age of 77 (range, 74-79) years. Each specimen was fully thawed at room temperature before dissection.
Specimen preparation All of the skin and subcutaneous tissues were dissected to the layer of the investing fascia. All of the muscles were sharply excised, except for the common extensors. The joint capsule and medial and lateral ligamentous joint complexes were left intact. The specimens were secured in aluminum pots using polymethyl methacrylate to allow attachment to the fixture within the testing machine. A 10 × 20mm anterior capsular window was created to assist with proper joint alignment during potting and to ensure congruous articular contact during testing. The humeral shaft was potted into one aluminum pot, and the radius and ulna were separately potted into another aluminum pot. After potting of the radius and ulna, an osteotomy of the radial shaft was performed, always distal to the bicipital tuberosity. This was performed so that a fixed radius would not limit motion and alter the kinematics through pivoting of the elbow about a constrained, rigid radiocapitellar articulation during the simulated dislocation event. During potting, care was taken to ensure that the natural carrying angle of each elbow was taken into account and that the articular surface remained parallel to the base of the pot (ie, horizontal).
D.R. Shukla et al. The humeral pot was mounted on a custom jig that allowed adjustment of the elbow flexion angle, in increments of 15° from 0° to 90° (Fig. 1). The jig was mounted onto a custom-made, lowfriction X-Y stage, which itself was mounted into a 6-axis load cell.
Mechanical testing A dislocation event that originated with disruption of the medialsided soft tissue stabilizers was mechanically simulated using a servohydraulic materials testing system (MTS) machine (MTS Systems, Eden Prairie, MN, USA), and joint gapping was measured using a 3-dimensional (3D) motion tracking camera system (Vicon, Denver, CO, USA). The MTS machine was equipped with a low-friction X-Y stage, which was mounted on a 6-axis load cell (ATI Industrial Automation, Apex, NC, USA). The MTS machine was equipped with an actuator that allowed the application of an axial load, a torsional moment (ie, rotation), and bending in the X and Y planes. The forearm pot was mounted into the clamp that was attached to this actuator (ie, top of the machine) and secured with a collar clamp. The humeral pot was secured to the jig at the bottom of the machine (Fig. 1). Once the specimen was secured within the clamps and before the application of any loads, it was ensured that the joint was congruous and seated in its natural position. This was accomplished by direct visualization of the anterior ulnohumeral articulation and by performing minor adjustments in rotation of both the humeral and forearm pots, such that the load cells indicated that the forces across the joint were as close to 0 as possible (always <0.05 N · m of torque). After this step, the specimen was subjected to the cycle of loading. After application of an axial load of 25 N that was applied to simulate a joint compressive force, the machine placed the elbow into 5° of valgus by bending the forearm pot in the X-plane. An external rotation moment of 2.5 N · m was simultaneously applied. The motion of the ulna was similar to that which would occur in a posterolateral rotatory subluxation event. This meant that the forearm pot was externally rotated until the load cell detected a torsional force of 2.5 N · m. Once 2.5 N · m of torque was reached, the machine stopped torque application and then internally rotated the forearm
Mounting of the specimen The forearm (ie, radius and ulna pot) was secured into the multiaxis hydraulic actuator, which applied the axial load, valgus force, and external rotation moment.
Figure 1 Schematic of the testing apparatus showing the direction of axial load (25 N), rotation of the X-axis motor causing a 5° valgus angulation, and external rotation of the actuator.
ARTICLE IN PRESS Posterior ulnar collateral ligament prevents instability
3
Figure 2 (A) Schematic of the medial ulnohumeral articulation. The circles with letters represent the placement of the markers. The distance between markers A and C was termed the proximal joint, and the distance between markers B and C was termed the distal joint. (B and C) The anterior bundle (B) and the posterior bundle (C) of the medial collateral ligament (MCL) were sharply incised at their respective ligamentous midsubstances, along the joint line. back to its original position, while the valgus bending returned from its position of 5° to 0°.
3D motion capture Gapping at the medial ulnohumeral joint was measured using the Vicon 3D motion capture system. The system used in this study relied on optical–passive motion capture, which used retroreflective markers that were tracked by 4 infrared cameras. The cameras were placed in the appropriate positions around the specimen and calibrated before each testing session in accordance with the manufacturer’s recommendations. The accuracy of the system’s measurement was validated before specimen testing. This was accomplished through translation of the markers by a known value that was then correlated to the value observed by the tracking system. The system’s error was found to be within 0.05 mm. The reflective markers were 6.4 mm in diameter. The Vicon system tracked the center points of the markers throughout their course. The markers were mounted on truncated 0.045-inch K-wires, which were then placed in the subchondral bone in the following locations: marker A was placed at the most proximal insertion point of the pMUCL on the olecranon (Fig. 2, A); marker B was placed just posterior to the insertion of the aMUCL on the humerus; marker C was placed just posterior to the aMUCL, just proximal to the articular surface; and marker X was placed on the medial epicondyle. Marker X was not used in the measurements but was used, as recommended by the manufacturer, to increase the accuracy of the tracking system. The distance between markers A and C was termed proximal joint gapping; the distance between markers B and C was termed distal joint gapping.
scribed by Floris et al,5 confirmation of the anterior bundle’s fibers was facilitated by identification of a macroscopically visible ridge. After sectioning of the aMUCL, the elbow was again tested at each flexion angle. After testing of the “injured” (ie, cut aMUCL) elbow at each of the 3 flexion angles, the pMUCL was transected (Fig. 2, C), and the elbow was again tested at each flexion angle.
Data analysis All data were reported as the mean ± standard error. The data were modeled with the use of pair-wise comparisons using a Student t-test, with a significance level of P ≤ .05.
Results After sectioning of the aMUCL, the ulnohumeral articulation became unstable, but the pMUCL was able to prevent frank dislocation. The proximal joint space increased by 4.2 ± 0.6 mm (P = .002) at 30° of flexion, 2.6 ± 0.3 mm (P = .005) at 60° of flexion, and 0.9 ± 0.6 (P = .9) at 90° of flexion (Fig. 3). The distal joint space increased by 3.8 ± 0.6 mm (P = .004) at 30°, 3.3 ± 0.8 mm (P = .02) at 60°, and 1.9 ± 1.1 mm (P = .18) at 90° (Fig. 4). After sectioning of the pMUCL (after having already sectioned the aMUCL), all of the specimens frankly dislocated at all flexion angles and failed to reach the threshold torque limit that had been set at 2.5 N · m, beyond which the machine stopped applying any additional torque.
Surgical conditions
Discussion First, the intact, uninjured specimen was tested at 30°, 60°, and 90° of flexion. To change the elbow flexion angle, the inferior (humeral pot) collar clamp was loosened and the jig was adjusted. Care was taken to ensure that the ulnohumeral articulation remained well seated before retightening of the collar clamp in preparation for the next testing cycle. After testing of the elbow at 90° of flexion (ie, completion of the “intact” cycle), the aMUCL was sharply incised along the medial ulnohumeral joint at the midportion of the ligament (Fig. 2, B). Care was taken to ensure that only the aMUCL was cut. As de-
This study demonstrated that under a simultaneously applied axial valgus load and external rotation moment, with the aMUCL sectioned and with the radiocapitellar center fixed, preventing radiocapitellar subluxation or lateral ligamentous complex injury, an intact pMUCL can prevent posterolateral dislocation and limit the extent of valgus laxity and ulnohumeral subluxation.
ARTICLE IN PRESS 4 Our findings emphasize the importance of the pMUCL in conferring some stability to an injured elbow, particularly at 30° and 60° of ulnohumeral joint flexion. After the aMUCL was incised, the maximal amount of proximal and distal joint gapping that occurred in any specimen was 6.6 mm and 5.5 mm, respectively. Both of these maximal values occurred at 30° of elbow flexion. We did not find that there was a significant increase in joint gapping after the aMUCL injury at 90°. This was likely due to the inherent osseous stability of the joint at 90° of flexion, given that the bone constraints were not injured in this study (ie, coronoid and ulnohumeral articulation).
Figure 3 There was an increase in proximal ulnohumeral joint gapping after injuring the anterior bundle of the medial collateral ligament (aMUCL) at both 30° (P = .002) and 60° (P = .005) of elbow flexion. Data are means ± standard error of the mean. The single (*) and double (**) asterisks denote significant differences between columns.
Figure 4 There was an increase in distal ulnohumeral joint gapping after injuring the anterior bundle of the medial collateral ligament (aMUCL) at both 30° (P = .004) and 60° (P = .02) of elbow flexion. Data are means ± standard error of the mean. The single (*) and double (**) asterisks denote significant differences between columns.
D.R. Shukla et al. These data prove that the pMUCL does function in more than just a minor capacity in its contribution to elbow stability, and the findings are important as they contribute to the presently minimal knowledge of the pMUCL. Whereas the role of the aMUCL as the primary restraint against valgus instability has been studied and established,3,7,13 there are relatively fewer studies that have focused on the biomechanical importance of the pMUCL.2-5,8,16 Furthermore, although the MUCL has been examined within the context of other studies, there are even fewer biomechanical studies that have focused specifically on the posterior bundle of the MUCL.17 The present data carry several clinical implications, particularly relating to operative management of elbow fracturedislocations (ie, terrible triad injuries). For terrible triad injuries, an accepted surgical tactic is to address the medial collateral ligament only if instability persists after management of the coronoid, radial head, and lateral collateral ligamentous complex. It has been recommended that one need not address the MUCL if the elbow is congruous from 30° to full flexion in at least 1 position of forearm rotation9; other authors have reported that addressing the MUCL is not usually necessary.6,10 However, this study demonstrates that the pMUCL is an important secondary stabilizer against external rotation and valgus instability. After repair or reconstruction of the injured structures following a fracture-dislocation (ie, coronoid, radial head, and lateral ulnar collateral ligament), it may be prudent to assess the integrity of the pMUCL and to repair it along with the aMUCL; this may protect the injured structures and further minimize the risk of recurrent instability or fixation failure. The findings of this study are also important given the recent data suggesting that a medially originating dislocation mechanism is more common than previously thought. It has been classically taught that most elbow fracturedislocations occur secondary to a posterolateral rotatory force. Internal rotation of the body relative to a fixed forearm11,14-16,22 causes disruption of the lateral stabilizers, progressing anteriorly and then medially. However, there are recent compelling studies that support a medial-sided initiation of injury. Schreiber et al demonstrated, through both a review of magnetic resonance imaging19 and by review of 62 in vivo elbow dislocations captured on YouTube,20 that most elbow dislocations occur secondary to a hyperphysiologic valgus moment on an extended elbow, noting that this would require a requisite MUCL disruption, followed by soft tissue disruption from the medial to lateral side. Rhyou and Kim, through evaluation of magnetic resonance imaging, observed a high proportion of MUCL distraction injuries along with valgus bone contusion that affected the radial head and capitellum, which supported their belief that dislocations may originate on the ulnar side.18 Our data demonstrate that the pMUCL confers secondary stability against dislocation secondary to this external rotation and valgus mechanism. Several of our observations are consistent with findings from other studies. We noted the greatest increase in both proximal and distal joint gapping at 30° of flexion. Morrey et al
ARTICLE IN PRESS Posterior ulnar collateral ligament prevents instability demonstrated that the ulnar collateral ligament was the primary restraint against valgus loads at 30° of flexion,13 whereas Callaway et al reported that specifically the posterior bundle was a secondary restraint at 30° of flexion.3 Of the cited studies that have examined the ulnar collateral ligament, only the data reported by Pollock et al17 have focused specifically on the pMUCL, rather than within the context of a study examining the entire ulnar collateral ligament. That study, which would most closely approximate this study, examined the change in varus and valgus angulation with passive elbow flexion and different forearm rotations after sectioning of the pMUCL. This study differs primarily in 2 ways. First, in this study, the aMUCL was sectioned, followed by pMUCL sectioning. In the study by Pollock et al,17 the pMUCL was sectioned, but the aMUCL was never sectioned. Second, this study differs in that 3 forces were simultaneously, not sequentially, applied to simulate a dislocation event. Therefore, a direct comparison of our data to their findings would not be appropriate. The methods of this study differ from several of the existing biomechanical studies that have examined medialsided elbow instability. Other investigators have used MTS-generated torque-rotation curves,3 bone markers with camera-capture,1 potentiometers with strain gauges,4,5,21 and electromagnetic tracking systems.2 Whereas an MTS machine was also used in this study to effect the simulated dislocation event, elbow instability was directly measured using 3D precision markers within an error of 0.05 mm in this study. We believe this to be a reliable and accurate system for measuring joint instability. There are several limitations to this study. First, we were unable to incorporate the radiocapitellar articulation in an unhindered manner that would appropriately simulate normal physiology, given the restrictions to our testing method. However, care was taken to leave the lateral soft tissues uninterrupted, including the common extensor mass and lateral collateral ligament complex. The lateral ulnar collateral ligament remained intact, as care was always taken to preserve both its distal humeral origin and insertion on the crista supinatoris. Another limitation is that the common flexor mass, which is an important secondary stabilizer, was not activated. Our method of assessment of the joint space gapping would not have been possible with an intact common flexor mass. A third limitation is that there was some inherent constraint within the system, as the humerus and forearm were potted in polymethyl methacrylate. The proximal radial osteotomy was performed in an effort to minimize the constraint involving the radiocapitellar articulation. It is possible that if the radius osteotomy were not performed and the radius was potted in varying degrees of rotation (ie, possibly full pronation, neutral, and full supination), different results may have been noted. However this would have tripled the number of specimens required and would have resulted in a significant increase in the resources required for this study. This, combined with the complexity of achieving the simultaneous application of forces, may have resulted in our being
5 unable to replicate a physiologic dislocation with complete accuracy. However, 3 pilot specimens were used before any actual testing in an attempt to eliminate as many sources of error as possible, and we believe our results to be valid. A final limitation is that an anterior capsular window was created to visualize the ulnohumeral articulation and to ensure congruency before testing. It is possible that the crossing capsular fibers may have conferred some stability to the elbow, particularly at low flexion angles. However, we do not think that an intact anterior capsule would have reduced the valgus instability to a degree such that a significant difference between surgical conditions would not have been observed. The findings of this study demonstrate that the posterior bundle of the MUCL is an important secondary stabilizer in the context of posterolateral rotatory instability. It is unlikely that the pMUCL is ever injured in isolation. Rather, it usually is attenuated or ruptured in conjunction with other stabilizing structures, such as the lateral collateral ligament complex and aMUCL in a posterolateral dislocation, or along with a radial head fracture and coronoid fracture in a terrible triad fracture-dislocation. During operative management of such injuries, it is important that these other stabilizers be repaired or reconstructed if indicated. The aMUCL remains the primary ligament that prevents valgus instability. After injury to the aMUCL, some stability will be lost and subluxation will occur, most notably between 30° and 60° of flexion. However, an intact pMUCL can prevent frank dislocation. In circumstances that necessitate exposure or operative management of the medial ligamentous complex (ie, fracturedislocation), we recommend surgically addressing the pMUCL by performing a repair, at a minimum. If a surgeon is performing a procedure with the intent of restoring stability to an injured joint, it is beneficial to repair all of the structures that might improve the stability of that joint to provide the greatest chance of success. In addition, in the event that the repair or reconstruction of an aMUCL fails, which is uncommon but does occur, the pMUCL can limit severe instability. As we did not evaluate the most optimal method of surgically managing pMUCL (ie, repair vs. reconstruction), we did not include a detailed discussion of reconstruction as that would not fall within the scope of this study. However, that is currently the subject of biomechanical evaluation and is currently under way at our institution. Repair of the native pMUCL can be performed using any suture configuration that approximates the ligamentous tissue if there is a midsubstance injury. If the pMUCL has avulsed from either the humerus or ulna, the surgeon can employ whichever method of repair that he or she is most comfortable with (eg, bone tunnels, suture anchor). As regards reconstruction, work at our institution is currently under way to determine the optimal method, and we cannot at present recommend 1 method in particular. At present, the senior author (M.H.) has routinely surgically addressed the pMUCL in circumstances of residual medial instability after operative fixation of terrible triad fracture-dislocations, with promising results (unpublished data).
ARTICLE IN PRESS 6
D.R. Shukla et al.
Further study is required to determine the optimal manner in which the pMUCL should be addressed.
7. 8.
Conclusion 9.
The pMUCL can confer some stability to the aMUCLdeficient elbow in a simulated ulnohumeral dislocation model. In the setting of an acute injury or residual instability after surgical management of an elbow dislocation, repair or reconstruction of the pMUCL may be warranted.
10.
11.
Disclaimer
12.
The authors, their immediate families, and any research foundation with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.
13.
14. 15. 16.
References 1. Armstrong AD, Dunning CE, Ferreira LM, Faber KJ, Johnson JA, King GJ. A biomechanical comparison of four reconstruction techniques for the medial collateral ligament–deficient elbow. J Shoulder Elbow Surg 2005;14:207-15. http://dx.doi.org/10.1016/j.jse.2004.06.006 2. Armstrong AD, Ferreira LM, Dunning CE, Johnson JA, King GJ. The medial collateral ligament of the elbow is not isometric: an in vitro biomechanical study. Am J Sports Med 2004;32:85-90. http://dx.doi.org/10.1177/0363546503258886 3. Callaway GH, Field LD, Deng XH, Torzilli PA, O’Brien SJ, Altchek DW, et al. Biomechanical evaluation of the medial collateral ligament of the elbow. J Bone Joint Surg Am 1997;79:1223-31. 4. Eygendaal D, Olsen BS, Jensen SL, Seki A, Sojbjerg JO. Kinematics of partial and total ruptures of the medial collateral ligament of the elbow. J Shoulder Elbow Surg 1999;8:612-6. 5. Floris S, Olsen BS, Dalstra M, Sojbjerg JO, Sneppen O. The medial collateral ligament of the elbow joint: anatomy and kinematics. J Shoulder Elbow Surg 1998;7:345-51. 6. Forthman C, Henket M, Ring DC. Elbow dislocation with intra-articular fracture: the results of operative treatment without repair of the medial
17.
18.
19.
20.
21.
22.
collateral ligament. J Hand Surg Am 2007;32:1200-9. http://dx .doi.org/10.1016/j.jhsa.2007.06.019 Hotchkiss RN, Weiland AJ. Valgus stability of the elbow. J Orthop Res 1987;5:372-7. Josefsson PO, Johnell O, Wendeberg B. Ligamentous injuries in dislocations of the elbow joint. Clin Orthop Relat Res 1987;221-5. Mathew PK, Athwal GS, King GJ. Terrible triad injury of the elbow: current concepts. J Am Acad Orthop Surg 2009;17:137-51. McKee MD, Pugh DM, Wild LM, Schemitsch EH, King GJ. Standard surgical protocol to treat elbow dislocations with radial head and coronoid fractures. Surgical technique. J Bone Joint Surg Am 2005;87(Suppl 1):22-32. http://dx.doi.org/10.2106/JBJS.D.02933 McKee MD, Schemitsch EH, Sala MJ, O’Driscoll SW. The pathoanatomy of lateral ligamentous disruption in complex elbow instability. J Shoulder Elbow Surg 2003;12:391-6. http://dx.doi .org/10.1016/S1058-2746(03)00027-2 Morrey BF, An KN. Functional anatomy of the ligaments of the elbow. Clin Orthop Relat Res 1985;84-90. Morrey BF, Tanaka S, An KN. Valgus stability of the elbow. A definition of primary and secondary constraints. Clin Orthop Relat Res 1991;18795. O’Driscoll SW, Bell DF, Morrey BF. Posterolateral rotatory instability of the elbow. J Bone Joint Surg Am 1991;73:440-6. O’Driscoll SW, Morrey BF, Korinek S, An KN. Elbow subluxation and dislocation. A spectrum of instability. Clin Orthop Relat Res 1992;186-97. Osborne G, Cotterill P. Recurrent dislocation of the elbow. J Bone Joint Surg Br 1966;48:340-6. Pollock JW, Brownhill J, Ferreira LM, McDonald CP, Johnson JA, King GJ. Effect of the posterior bundle of the medial collateral ligament on elbow stability. J Hand Surg Am 2009;34:116-23. http://dx.doi.org/10 .1016/j.jhsa.2008.09.016 Rhyou IH, Kim YS. New mechanism of the posterior elbow dislocation. Knee Surg Sports Traumatol Arthrosc 2012;20:2535-41. http://dx.doi.org/10.1007/s00167-011-1872-7 Schreiber JJ, Potter HG, Warren RF, Hotchkiss RN, Daluiski A. Magnetic resonance imaging findings in acute elbow dislocation: insight into mechanism. J Hand Surg Am 2014;39:199-205. http://dx.doi.org/10 .1016/j.jhsa.2013.11.031 Schreiber JJ, Warren RF, Hotchkiss RN, Daluiski A. An online video investigation into the mechanism of elbow dislocation. J Hand Surg Am 2013;38:488-94. http://dx.doi.org/10.1016/j.jhsa.2012.12.017 Sojbjerg JO, Helmig P, Kjaersgaard-Andersen P. Dislocation of the elbow: an experimental study of the ligamentous injuries. Orthopedics 1989;12:461-3. Wake H, Hashizume H, Nishida K, Inoue H, Nagayama N. Biomechanical analysis of the mechanism of elbow fracture-dislocations by compression force. J Orthop Sci 2004;9:44-50. http://dx.doi.org/10 .1007/s00776-003-0735-6