If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author: Robert A. Kaufmann, MD, Department of Orthopaedic Surgery, University of Pittsburgh, Kaufmann Medical Building, 3471 Fifth Avenue, Suite 1010, Pittsburgh, PA 15213.
The treatment of bidirectional ligament instability is proposed using a method that simultaneously tensions medial and lateral ligaments. Graft tension is maintained via plates that apply compression between the graft and bone.
Methods
We tested static varus and valgus elbow stability in six cadaver elbows with intact ligaments and capsules at five positions, and then created gross instability by dividing all soft tissue attachments. A ligament reconstruction was subsequently performed with and without nonabsorbable ligament augmentation. Elbow stability was measured and compared with the native state.
Results
The augmented and the nonaugmented ligament reconstructions provided stability to the lateral side with only 1.0 mm of increased deflection recorded for the augmented ligaments and 0.6 mm for the nonaugmented when compared with the native state. On the medial side, the deflection was greater after reconstruction compared with the native state with the augmented ligaments ranging between 1.0 and 1.8 mm and the nonaugmented ligament reconstruction ranging between 2.4 and 3.3 mm.
Conclusions
This novel ligament reconstruction maintained secure fixation between ligament and bone and allowed for maintenance of static stability at different degrees of elbow flexion.
Clinical Relevance
Restoring elbow stability using a method that minimizes ligament graft and which may not need to be removed could benefit management of bidirectionally unstable elbows, such as following interposition arthroplasty or substantial trauma.
Gross elbow instability may occur following elbow dislocations associated with severe soft tissue injuries and complex fractures. Elbow instability can also result from reconstructive procedures that require an extensive release of peri-articular soft tissues or interposition arthroplasty.
When managing gross elbow instability, the medial collateral and lateral ulnar collateral ligaments may need to be reconstructed. Various ligament reconstructions have been described and can broadly be divided into those that employ one long graft or two grafts.
When performing ligament reconstruction with a continuous loop, the graft must be pulled through bone tunnels. The tunnels cause friction that may make tensioning difficult. Portions of the ligament graft reside within the bones where they are not transmitting force. Asymmetric tension can lead to varus and valgus forces being imparted to the elbow.
In addition, when tightening the ligament reconstruction on one side, it may be difficult to maintain elbow reduction. Good medium-term results have been demonstrated for this one-graft continuous loop technique.
A two-graft approach requires placing two separate ligament reconstructions where symmetrical tensioning may be difficult.
Regardless of the type of ligament reconstruction chosen, the construct often needs to be protected with a nonhinged or hinged external fixator. These are temporary efforts to improve the healing environment and offload forces exerted during the reconstruction effort. Their drawback is that they are bulky and have a risk of pin tract infection.
Treatment of persistent instability after posterior fracture-dislocation of the elbow: restoring stability and mobility by internal fixation and hinged external fixation.
The ligament reconstruction may also be protected with an internal fixator, such as the IJS Elbow Stabilization System (Skeletal Dynamics), which uses pins to provide mechanical elbow stability without using a bulky external device. The goal of any mechanical fixator is to provide sufficient clinical stability to start early rehabilitation. However, both external and internal fixators require removal.
A system was designed to address some of the shortcomings of current methods for managing bidirectional elbow instability. The intent was to minimize the required amount of tendon graft by using a cylindrical ligament retention device that resides within bone and avoids graft traversing the width of the distal humerus (Fig. 1). The graft can be simultaneously tensioned and secured to the bone with compression generated by two plates that exhibit an aggressive tooth pattern (Fig. 2). The compressive force between the graft and the bone “locks in” 40 N of tensile force on the medial and lateral sides of the elbow by tightening two screws against the cross-locking plates. This force was chosen on the basis of a study demonstrating that medial collateral ligament tension of 20 N and 40 N reproduced native elbow kinematics.
Figure 1The cylindrical ligament retention device comes in three lengths to reside within the distal humerus. It can slide within the drill hole, allowing graft tension equalization.
Figure 2Two plates secure the graft limbs against bone after the ligaments are pretensioned for 3 minutes to minimize postimplantation stress relaxation. Ligament grafts may be augmented with an absorbable suture. The thumb arrow describes force being imparted to seat the olecronon. The four ligament arrows describe the application of 20N on each limb. The screwdriver arrow demonstrates tightening of the bolts to compress graft against bone.
During implantation, the grafts are placed along the axis of ulnohumeral rotation to minimize length changes during flexion and extension, which can result in graft elongation and failure.
Strength of ulnar fixation in ulnar collateral ligament reconstruction: A biomechanical comparison of traditional bone tunnels to the tension-slide technique.
This method for simultaneous reconstruction can be augmented with absorbable suture (polydioxanone [PDS] 1; Johnson & Johnson) to increase the tensile strength of this construct by transmitting forces through the suture and away from the graft. Each PDS 1 suture, when woven through the graft and sutured to itself on the dorsum of the ulna, should add additional strength in the acute setting and retain 60% of that strength at six weeks.
There was a risk that this method will not adequately restore static stability. Our goal was to measure the stability of the native elbow in abducted and adducted humeral positions and at different degrees of elbow flexion. We then sectioned all soft tissue structures and reconstructed the collateral ligaments and then compared these measurements. Furthermore, we aimed to identify whether ligament reconstruction augmentation with PDS 1 suture improves static construct stability.
Materials and Methods
Six whole cadaver arms were dissected. The number of specimens was determined after consideration of previous elbow ligament repair literature. Sample size calculations revealed that with an alpha of 0.05, a two-tailed analysis between two independent means would require a minimum of four trials in each group to reach a power of 0.8.
Biomechanical functional elbow restoration of acute ulnar collateral ligament tears: the role of internal bracing on gap formation and repair stabilization.
For each specimen, the forearm, hand, and wrist were left intact except to harvest the palmaris longus tendon and one flexor digitorum superficialis tendon of the middle or ring finger. The grafts were trimmed if they were greater than 3.5 mm in diameter and augmented with PDS 1. The anconeus and flexor carpi ulnaris muscles were elevated from the olecranon preserving the collateral ligaments while identifying the supinator crest and sublime tubercle. The supinator crest was used to direct the proper placement of a custom drill guide to create holes in the olecranon (Fig. 3). Once the holes were drilled, the ligament reconstruction plates were seated against the bone. The ligament reconstruction plates were held in place with nuts and bolts, the latter of which were also chosen as a reference against which measurements were taken. Placement of the plates away from the subcutaneous border of the ulna and under the flexor carpi ulnaris and extensor carpi ulnaris or anconeus muscles decreases the chance of skin irritation.
Figure 3Drill holes within the humerus and ulna are placed using anatomic landmarks. Static testing is performed using a custom-designed experimental setup that allows elbow flexion when a 13.3 N weight is applied to the wrist.
The humerus was stripped of all muscles, and the humeral head was removed. The points of isometry were the center of the capitellum laterally and the anterior inferior surface of the medial epicondyle. A K-wire was used to create a 2.0-mm indentation which was 7.0 mm proximal to the point of isometry on the medial and lateral epicondyle. These holes were marked to assist with measuring the distance (deflection) between these locations and the socket cap bolt head (Fig. 4). The proximal humerus was clamped into the experimental setup. The humerus was positioned in both 90° of abduction and 90° of adduction as a 13.3 N weight was applied to the wrist, and the deflection was measured at five positions of elbow motion (0°, 30°, 60°, 90°, and 120°).
Figure 4Caliper measurements of the distance between the two fixed points (epicondyle and socket cap bolt head of plate) that were taken at the five positions of elbow motion (0°, 30°, 60°, 90°, and 120°).
The elbow was then destabilized by reflecting all ligaments and capsule and prepared for ligament reconstruction. The centerline of rotation was identified using visible landmarks and two K-wires (one medial and one lateral) were placed along this axis. Another K-wire was placed 3 mm proximal and parallel to the centerline K-wires to account for the cannulated drill bit radius. These parallel K-wires were used to guide the cannulated drill in creating two holes through the medial and lateral epicondyle, which ensured that the graft would exit the distal humerus at the centerline of ulnohumeral rotation (Fig. 3). The augmented ligament was then passed through the eyelet of the cylindrical ligament retention device, and both were pushed into the hole that was created within the humerus (Fig. 1). The other ligament graft was passed through the other cylindrical ligament retention device eyelet, and the medial and lateral ligaments were simultaneously tensioned with approximately 40 N for 3 minutes, after which the plates were tightened against the bone (Fig. 2).
Interference screw fixation of soft tissue grafts in anterior cruciate ligament reconstruction: part 2: effect of preconditioning on graft tension during and after screw insertion.
This 40 N force on medial and lateral sides requires a combined 80 N force to be applied by the surgeon, which represents a substantial but submaximal one-handed pull, given that the maximum tension applied with a single hand pull by a sports medicine surgeon is 99 N.
The grafts are placed on the medial side so that they are collinear with the sublime tubercle, which is where the medial collateral ligament inserts. On the lateral side, the grafts run along the course of the lateral ulnar collateral ligament. The nuts and bolts were tightened with two-finger tightness, which allowed for the plates to contour to the irregular topography of the olecranon and compress the ligament graft into the bone. The transulnar bolts may be cut after nut tightening, so that bolt protrusion is minimized, but this was not done in this cadaver test. Before loading, grafts were marked with pen in two locations allowing for measurement of elongation and slippage through digital imaging, which was recorded before and after each trial.
After completion of the ligament reconstruction, the humeral shaft was secured to the test frame at 90° of abduction and then 90° of adduction so that a varus and valgus force could be exerted (Fig. 3). A clamp was used to apply a 13.3 N weight to the wrist while also allowing for maintenance of neutral forearm rotation without having to pin the forearm. After the arm was placed into the experimental setup, the distance between the points (deflection) was measured at five positions of elbow motion (0°, 30°, 60°, 90°, and 120°) (Figure 4, Figure 5). The suture augmentation was then cut in multiple locations to remove the PDS suture influence on the simultaneous ligament reconstruction, and again, the distance (deflection) between these points was measured at the aforementioned five positions of elbow motion.
Figure 5Deflection measurements in full extension and flexion of 4° of flexion were recorded in the native elbow and after the augmented and nonaugmented ligament reconstruction. MCL, medial collateral; LUCL, lateral ulnar collateral.
We compared the initial laxity (deflection) measurements of the native elbow at different positions on the medial and lateral sides with the augmented and nonaugmented ligament reconstruction measurements, which identified the impact of suture augmentation on ligament construct stiffness. We weighed the forearms and identified the centroid locations (the position where the forearm weight could be balanced). The distances from the centerline of ulnohumeral joint rotation to the forearm centroid and wrist (where the 13.3 N weight was applied) were measured, which allowed for calculation of the moment that was exerted on the elbow.
The descriptive analysis of the deflection results included the mean, SD, and change percentages compared with the control group. Our null hypothesis was that there was no significant statistical difference in lateral and medial deflection between control, PDS augmented reconstruction, and nonaugmented reconstruction groups. We used the analysis of variance test to compare the distances at each angle and reported the P value to assess statistical significance. A P value of less than .05 was considered statistically significant. We used Stata software for statistical analysis.
Results
The average weight of the six forearms was 23.5 N (21.3–26.1 N) and an average distance of 15 cm (13–17 cm) was measured from the approximate forearm centroid to the centerline of ulnohumeral rotation. This created an average moment of 3.5 Nm (3.1–3.7 Nm). When measured from the center of ulnohumeral rotation, the applied 13.3 N weight acted at an average distance of 24 cm (21–25 cm), which created an average moment of 3.2 Nm (3.0–3.5 Nm). The combined average moment acting on the elbow was 6.7 Nm (6.4–6.9 Nm).
No graft elongation or slippage was recorded through the measurement of sequential digital imaging. The deflection measurements from the control group and the augmented and nonaugmented ligament reconstruction groups for the medial side are shown in Table 1 and Figure 6 and for the lateral side are shown in Table 2 and Figure 7.
Table 1Medial Ulnar Collateral Ligament Reconstruction Distance Results in Centimeters
In comparison to control. Δ means difference or change.
P value
0
4.46
1.3
4.56
1.61
1
2.20
4.74
1.59
2.8
6.20
.95
30
4.73
1.46
4.91
1.72
1.8
3.80
5.06
1.68
3.3
6.90
.94
60
4.96
1.55
5.06
1.74
1.5
2.00
5.25
1.73
2.9
5.80
.95
90
4.88
1.41
5.04
1.68
1.6
3.20
5.12
1.67
2.4
4.90
.97
120
4.66
1.3
4.81
1.55
1.5
3.20
4.96
1.57
3
6.40
.94
PDS, polydioxanone.
Average distance measured between medial epicondyle and medial socket cap bolt head is listed with SD for each elbow position when static valgus stress was applied through weight of arm and 13.3 N weight at wrist. Percentage change relative to control is listed for augmented and nonaugmented ligament reconstruction.
∗ In comparison to control. Δ means difference or change.
Figure 6Graph showing the average distance measured between medial epicondyle and medial socket cap bolt head. Measurement range is shown as a black line. MUCL, medial collateral ligament; PDS, polydioxanone.
In comparison to control. Δ means difference or change.
P value
0
3.97
1.71
3.87
1.81
−0.1
−2.50
3.99
1.74
0.2
0.50
.99
30
4.37
1.72
4.34
1.88
−0.3
−0.70
4.36
1.88
−0.2
−0.20
.99
60
4.62
1.86
4.56
1.89
−0.6
−1.20
4.6
1.85
−0.2
−0.40
.99
90
4.58
1.91
4.59
1.81
−0.1
0.20
4.64
1.76
0.6
1.30
.99
120
4.61
1.69
4.47
1.63
−1.4
−3
4.5
1.66
−1.1
−2.30
.99
PDS, polydioxanone.
Average distance measured between lateral epicondyle and lateral socket cap bolt head is listed with SD for each elbow position when static varus stress was applied through weight of arm and 13.3 N weight at wrist. Percentage change relative to control is listed for augmented and nonaugmented ligament reconstruction.
∗ In comparison to control. Δ means difference or change.
Figure 7Graph showing the average distance measured between lateral epicondyle and lateral socket cap bolt head. Measurement range is shown as a black line. LUCL, lateral ulnar collateral; PDS, polydioxanone.
Lateral side deflection with the augmented ligament reconstruction was 1.0 mm more than the control in full extension and otherwise deflected less at all other positions of flexion. Lateral side deflection with the nonaugmented ligament reconstruction was 0.2 mm in full extension and 0.6 mm at 90° and otherwise deflected less than the control. The P value results from analysis of variance test did not show any statistically significant difference between control versus reconstruction with PDS (P = .99) or control versus reconstruction without PDS (P = .98) at each flexion angle.
Medial side deflection with the augmented ligament reconstruction ranged between 1.0 and 1.8 mm more than the control. Medial side deflection with the nonaugmented ligament reconstruction ranged between 2.4 and 3.3 mm. The P value results from the analysis of variance test did not show any statistically significant difference between control versus reconstruction with PDS (P = .94) or control versus reconstruction without PDS (P = .97) at each flexion angle.
Discussion
Medial and lateral elbow instability is a challenging problem. In this study, we recorded the native cadaveric stability at five different elbow positions of flexion in the abducted and adducted arm. The deflection of the elbow at these positions was then compared with measurements obtained after the elbow was destabilized and then reconstructed with a novel ligament reconstruction system. We used tendon grafts that were either unaltered or reinforced with PDS 1 suture. Valgus elbow instability is postulated when an opening that was greater than 2.0–3.0 mm occurred between the coronoid and the trochlea, and we considered this range to be our cutoff for maintaining medial and lateral stability.
The main finding of this study is that both the augmented and the nonaugmented ligament reconstruction provided substantial elbow stability. On the lateral side, only 1.0 mm of averaged increased deflection was measured, with most measurements demonstrating less laxity when compared with the control. The nonaugmented ligament reconstruction fared similarly to the augmented ligament on the lateral side.
On the medial side, the deflection after reconstruction was greater than on the lateral side, with the augmented reconstruction ranging between 1.0 and 1.8 mm more than the control and the nonaugmented ligament reconstruction ranging between 2.4 and 3.3 mm greater than the control. The nonaugmented ligament reconstruction reached 3.3 mm at 30° of flexion, which is greater than the suggested 3.0-mm cutoff for medial instability. Our data suggest that it may be advantageous to use PDS 1 augmentation, which has an ultimate tensile strength of 85 N, to increase the graft stiffness because augmented deviation was never greater than 1.8 mm.
Using a different type of suture with a higher tensile strength may have strengthened the biomechanical properties of this construct further. Still, we chose PDS 1 because it is the largest absorbable suture that is readily available and its absorption prevents the risk of stress shielding the tendon graft over time.
The deflection results of the control group cannot be employed to identify a position of greatest static laxity because the point of reference from which epicondyle measurements were taken was not at the point of isometry. We chose landmarks that were close to the point of isometry but would not be affected by drilling the distal humerus for placement of the graft.
The average moment that acted on the elbow during our testing was 6.7 Nm. We used the weight of the forearm acting at its centroid as well as a 13.3 N weight to exert substantial force and yet not cause catastrophic failure to our cadaver specimens. Our deflection measurements may have been increased by a greater moment. A static force analysis calculated an average medial or lateral force of 268 N, which was then split between two-graft limbs allowing each one to experience 134 N of force. Our grafts demonstrated no evidence of slipping or failure at this force, which compared favorably to a docking reconstruction that failed at 53.0 N as well as a biomechanical study that demonstrated 5-mm medial collateral reconstruction elongation failure at 102.7 N.
This ligament reconstruction used a palmaris or similar sized graft with a diameter of 3.5 mm and the stress (force/area) within the graft was then equal to 495 Bar. Realizing that the ultimate stress was 1,450 Bar and the expected graft stress was 495 Bar, the safety factor for the use of a palmaris graft in this experimental setup was 2.92.
Two similar studies identified catastrophic ligament reconstruction failure at 14.3 Nm and 13.3 Nm.
Applying the higher torque to the static analysis of this reconstruction, a static force of 572 N would be exerted on the medial and lateral sides, which would then divide between two-graft limbs that each experience 286 N. With an average palmaris longus diameter of 3.5 mm and a graft tension of 286 N for each limb, the stress (force/area) within the graft would be 1,057 Bar. Realizing that the ultimate stress is 1,450 Bar and the expected graft stress is 1,057 Bar, the safety factor for the use of a palmaris graft in this hypothetical setting would be 1.37. This ligament reconstruction should, in theory, be acceptable to prevent catastrophic failure with an applied torque of 14.3 Nm.
This study has multiple sources of error. Two limitations, which are common to many cadaveric studies, are that a small sample size (six cadavers) was used and the specimens were of advanced age (average age of 61 years). In all the specimens, ligament reconstruction integrity was maintained, and it is unlikely that additional cadaver specimens would have changed our findings.
Another potential shortcoming is that we altered the control deflection measurements by placing the medial and lateral plates before obtaining these values. The plate placement was performed so as not to violate the ligaments or capsule, and this method provided an immobile point of reference for our measurements. It is unlikely that destabilization occurred given that the control deflection measurements were small and the ligaments and capsule were preserved.
A limitation of our experimental setup is that it did not recreate dynamic stability through the actuation of humeral and forearm muscles. However, it has been shown that at 90° of abduction and adduction, dynamic forces are the smallest contributor to elbow stability, and the static stabilizers, such as those that we were testing, are the most important.
Another shortcoming is that we used flexor digitorum superficialis and palmaris tendons for the ligament reconstruction effort. We minimized differences by ensuring that the dimensions matched that of the palmaris longus graft for each specimen. The choice of graft may not matter as similar studies have demonstrated no significant difference in elongation, number of cycles to failure, or stiffness between reconstructions using a range of graft options.
As this experiment was designed as a reconstruction simulation to assess for static stability of the elbow, application of this data to the clinical setting remains hypothetical. It should be made clear that although the primary outcome of this study was to assess for graft catastrophic failure or elongation as measured through increased deflection of the elbow, it is still too early to understand whether this ligament reconstruction will be able to restore stability in the in vivo patient population that it is intended for. Additional study limitations are those common to many cadaveric studies, including the time-zero loading environment, testing at only 5° of elbow flexion, and the inability to test the effect of active muscle contraction resulting in dynamic joint compression.
Although consistent reference points were used for the measurement of graft elongation, there remains a potential for human error. Our experimental protocol minimized this by having two authors record the visual graft slipping, goniometer angle, and caliper deflection measurements.
One of the strengths of this study is that we used cadaver specimens with nonuniform bone dimensions and soft tissue integrity. Another strength is that all cadaver specimens underwent identical testing with tendon grafts of similar size. By excluding factors such as muscle and joint compression, the testing isolated the medial and lateral reconstructions.
This novel ligament reconstruction restored static elbow stability without evidence of graft failure when loaded with 6.7 Nm torque at different elbow positions. Absorbable PDS suture added stiffness to the construct. This technique allowed simultaneous tensioning of both tendon grafts, which may be beneficial when treating global elbow instability.
References
Zeiders G.J.
Patel M.K.
Management of unstable elbows following complex fracture-dislocations—the “terrible triad” injury.
Treatment of persistent instability after posterior fracture-dislocation of the elbow: restoring stability and mobility by internal fixation and hinged external fixation.
Strength of ulnar fixation in ulnar collateral ligament reconstruction: A biomechanical comparison of traditional bone tunnels to the tension-slide technique.
Biomechanical functional elbow restoration of acute ulnar collateral ligament tears: the role of internal bracing on gap formation and repair stabilization.
Interference screw fixation of soft tissue grafts in anterior cruciate ligament reconstruction: part 2: effect of preconditioning on graft tension during and after screw insertion.
Declaration of interests: R.A.K. owns Arrch Orthopedics that is designing this novel method of ligament reconstruction. No benefits in any form have been received or will be received related directly to this article.
00035-X&id=gr1.jpg){kind=link}
00035-X&id=gr2.jpg){kind=link}
00035-X&id=gr3.jpg){kind=link}
00035-X&id=gr4.jpg){kind=link}
00035-X&id=gr5.jpg){kind=link}
00035-X&id=gr6.jpg){kind=link}
00035-X&id=gr7.jpg){kind=link}