Effect of human dermal allograft thickness on glenohumeral stability for superior capsular reconstruction in irreparable supraspinatus tears: A biomechanical analysis of the superior capsular reconstruction – A cadaveric study

Abstract

Purpose

The objective of this study was to compare the biomechanics of using a double layered human dermal allograft to a single layered human dermal allograft in superior capsular reconstruction.

Methods

Five cadaveric shoulders were tested. The superior translation of the humerus and the subacromial contact pressure were measured at 0°, 30° and 60° of glenohumeral abduction in the following six conditions: (1) intact rotator cuff, (2) irreparable supraspinatus tear, (3) superior capsular reconstruction using a double layered human dermal allograft with and (4) without posterior suturing, (5) superior capsular reconstruction using a single human dermal allograft with and (6) without posterior suturing.

Results

There was a significant increase in superior translation of the humerus and subacromial contact pressure when comparing torn supraspinatus to intact shoulder. All superior capsular reconstruction repairs lead to a reduction in superior translation and subacromial contact pressures compared to after the supraspinatus tear except for single layered superior capsular reconstruction repair without posterior suturing. There was no difference in superior translation and subacromial contact pressures comparing the intact shoulder to any of the superior capsular reconstruction constructs.

Conclusion

Superior capsular reconstruction using a single or double layered human dermal allograft improves superior translation after supraspinatus repair. There is some biomechanical benefit of a double layered human dermal allograft technique over a single layered graft technique in superior capsular reconstruction.

Level of Evidence

IV

Introduction

The rotator cuff has a role in maintaining stability of the glenohumeral joint through concavity compression, which counteracts the superior force vector of the deltoid during abduction and creates a fulcrum for concentric rotation of the humeral head.^1,2 Concavity compression particularly enhances stability in the mid-range of motion when the capsuloligamentous constraints are lax. ³ Disruption of concavity compression, due to a posterosuperior rotator cuff tear, can result in superior migration of the humeral head and subacromial impingement.^4,5

While most posterosuperior cuff tears are reparable, some large retracted chronic tears cannot be repaired, particularly if there is significant muscle atrophy and fatty infiltration.^6–9 For these irreparable cases, partial repairs, debridement, biceps tenotomy, tendon transfer and reverse shoulder arthroplasty have been utilized. ¹⁰ Bridging patch grafts, using human dermal allograft (HDA) patches, have also been used. Although some authors have reported good outcomes with this technique, others have reported high re-tear rates due to superior translation of the humerus from a lack of restoration of glenohumeral stability.^11–15

Superior capsular reconstruction (SCR) is a recently described surgical technique that addresses irreparable supraspinatus tears by reconstructing the superior capsule, using a graft spanning from the superior surface of the glenoid to the greater tuberosity. This creates a constraint to superior migration of the humeral head. ¹² In previous clinical and biomechanical studies, SCR completely restored native glenohumeral stability and function.^12,16

SCR was originally reported using fascia lata allograft (FLA). ¹⁶ Although previous studies have proven FLA to be effective, an identical technique using HDA was introduced to avoid donor site morbidity associated with harvesting the fascia lata autograft.^12,16–21

A previous biomechanical study by Mihata et al. ¹⁸ has shown that SCR with an 8 mm thick FL graft has greater stability than a 4 mm thick graft. However, no studies have investigated the effect of HDA graft thickness on shoulder stability. The aim of this study was to compare glenohumeral stability with a double layered HDA to a single layered HDA.

Materials and methods

Five fresh-frozen cadaveric shoulders were dissected for experimentation. These cadaver specimens were provided by the Sydney Clinical Skills and Simulation Centre, Royal North Shore Hospital for this biomechanical study. The mean age of donors was 68.2 (range 56–81) and the mean BMI was 22.4. Four of the specimens were from male donors and one was from a female donor. Only left shoulders were tested. The shoulders were thawed in room temperature before dissection. Each specimen had the skin, subcutaneous tissue and muscles removed but the shoulder capsule, tendinous insertions of the rotator cuff muscles, latissimus dorsi, pectoralis major and deltoid muscles were preserved. Each specimen was carefully examined for pre-existing rotator cuff tears. The scapulae were potted using Ardit Rapidset Mortar (Dunlop, ARDEX Australia Pty Ltd). The humeral shaft was transected at mid-humerus. An intramedullary threaded rod was placed inside the humerus and a nut was used to prevent subluxation of the humerus. No. 2 Force Fiber (Wright Medical Group N.V., Tennessee, USA) sutures were used to tag the tendinous insertions of each muscle in Krackow fashion to apply different loads (supraspinatus, 2; subscapularis, 2; infraspinatus, 1; teres minor, 1; deltoid, 3; pectoralis major, 2; latissimus dorsi, 2). ¹⁷ To obtain a consistent measure of the position of the humerus relative to the scapula, a screw was placed on the anterolateral aspect of the acromion and another was placed in the proximal bicipital groove. The scapula was secured in a custom made testing apparatus (Figure 1) and was placed at 20° of anterior tilt and 0° rotation. The humeral rod was also secured into the jig to allow controlled abduction of the humerus and axial rotation, and free elevation or depression of the humerus. Throughout the experiment, the position of the humerus was adjusted to ensure that it was positioned in the scapular plane. A digital inclinometer (Model DGTP, GemPro Ltd) was attached to the metal plate of the jig that stabilized the humeral rod to perform measurements at three different angles of humeral abduction (0°, 30° and 60° of glenohumeral abduction). ¹⁷ Each No. 2 FibreWire suture attached to the tendons was connected to fishing lines (Rovex 10X Formula Monofilament Leader Line 100m 40lb) to create muscle loads. The fishing lines were passed through a set of plates and then desired loads were applied over pulleys.

Figure 1.

Lateral view of the left shoulder mounted on the custom made testing apparatus at 0° of glenohumeral abduction.

Two different loading conditions were used: A balanced load was used for loading condition 1 (40 N on deltoid, 20 N on both pectoralis major and latissimus, 20 N on supraspinatus, 10 N on infraspinatus, 10 N on teres minor and 10 N on subscapularis); a superiorly directed load was applied for loading condition 2 (80 N on deltoid, 20 N on supraspinatus, 10 N on infraspinatus, 10 N on teres minor and 10 on subscapularis). These conditions were chosen based on previously published biomechanical studies on SCR repair with adjustments to the loading condition 1 to ensure that the humeral head was positioned in the centre of the glenoid. ¹⁷ Each shoulder was tested at 0°, 30° and 60° of glenohumeral abduction.

Superior translation of the humerus relative to the scapula was measured at each abduction angle with loading condition 1 and loading condition 2 using a MicroScribe. The distance between the screw on the acromion and the screw in the humerus was calculated along the z-axis (superior–inferior). The position resolution of the MicroScribe was 0.005″ (0.127 mm).

Subacromial peak contact pressure was measured with loading condition 2 using a Tekscan pressure measuring system using FlexiForce B201 (Tekscan, Inc., MA, USA) sensor with a maximum load of 667 N. The sensor was placed under the acromion and the pressure was measured at each abduction angle with a superiorly directed load.

Testing conditions

In this study, six conditions were tested (Figure 2): (1) an intact rotator cuff, (2) a complete supraspinatus tear, (3) after SCR using a double layered HDA (Graftjacket Maxforce Extreme 2.0 mm (86UM-4X07, Wright Medical Group N.V., Tennessee, USA)) with and (4) without posterior sutures, then (5) after SCR using a single layered HDA with and (6) without posterior sutures. For condition 1, only one specimen had a fully intact rotator cuff and this was used as control.

Figure 2.

Left showing a photograph of a left shoulder with the testing condition of SCR using HDA with posterior suturing. Right showing a close-up view.

After performing the initial test, the supraspinatus tendon was excised from its insertion and was separated from the infraspinatus and subscapularis tendons along the anterior and posterior borders of the supraspinatus tendon, leaving the remainder of the rotator cuff tendons intact. The tests were repeated on all specimens after inducing this irreparable supraspinatus tear (condition 2).

SCR was performed using a double layered HDA (Graftjacket Maxforce Extreme 2.0 mm (86UM-4X07, Wright Medical Group N.V., Tennessee, USA)). This was prepared according to the manufacturer's instructions and was then pre-stretched in all directions between artery forceps placed at each corner of the graft material. Two 3.5 mm Piton Anchors (Wright Medical B.V.) loaded with Forcefibre were placed in the superior glenoid spanning the width of the supraspinatus. The lateral repair was performed using an Arthrotunneler (Wright Medical B.V.) to create two transosseous tunnels between the medial and lateral margins of the supraspinatus footprint of the proximal humerus. The distances between the glenoid anchors and the medial tunnel holes were measured and the HDA patches were cut to fit the measurements with 10 mm of additional graft material laterally and 5 mm of additional graft material anteriorly, posteriorly and medially. The sutures associated with the anchors were placed through prepunched holes in the corners of the graft. The graft was then positioned by tightening and tying the horizontal mattress sutures. The lateral sutures were used to create a ‘suture bridge’ type repair utilizing the tunnels in the greater tuberosity. The posterior suture technique consisted of two margin convergence sutures passed from the posterior margin of the graft to the anterior edge of the infraspinatus. After the SCR using a double HDA with posterior suturing (condition 3) was tested on all specimens, the posterior sutures were released to test the SCR using a double HDA without posterior suturing (condition 4). A single layer of the HDA was carefully removed and posterior sutures were added to repeat the measurements on the SCR using a single HDA with posterior suturing (condition 5). Then, the posterior sutures were released again to test for the SCR using a single HDA without posterior suturing (condition 6).

Statistical analysis

All measurements were performed three times and the mean values of all the specimens were calculated. The inter-rater reliability (ICC value) of the testing system was assessed via a two-way random model.

Results were analysed using one-way ANOVA repeated measures with Bonferroni post-hoc correction with significance level set at 0.05. Statistical analysis was performed using Prism v7.04 (GraphPad Software, CA, USA), SPSS version 24 (IBM Corp.) and Microsoft Excel.

Results

Reliability of the test setup

The inter-rater reliability (ICC value) of the testing system was 0.835, which represents good reliability.

Effect of graft thickness

Superior translation

SCR using a double HDA with posterior suturing, double HDA without posterior suturing and single HDA with posterior suturing resulted in significantly less superior translation compared to the torn state. There was no difference in superior translation comparing the torn state and SCR using a single HDA without posterior suturing (Figure 3 and Table 1).

Figure 3.

Superior translation of the humerus for each testing condition at different abduction angles. D + : double layered graft with posterior suturing; D−: double layered graft without posterior suturing; HDA: human dermal allograft; S+: single layered graft with posterior suturing; S−: single layered graft without posterior suturing; SCR: superior capsular reconstruction (# statistically significant versus torn, + statistically significant versus S+, − statistically significant versus S−).

Table 1.

Superior translation of the humerus (mm).

AA	Torn	%	D+	%	D−	%	S+	%	S−	%
0°	8.2	357	2.1b,c,d	91	2.2b,c,d	96	4.1b	178	4.5	196
30°	8.7	311	3.0b,c,d	107	3.0b,c,d	107	5.0b	179	5.3	189
60°	6.6	254	2.7b,d	104	2.5b,d	96	3.7b	142	4.2	162

AA: abduction angle; D + : double layered graft with posterior suturing; D−: double layered graft without posterior suturing; S+: single layered graft with posterior suturing; S−: single layered graft without posterior suturing

Data are expressed as average. Per cent was calculated by dividing the mean value by the intact value at the corresponding angle (“b” statistically significant versus torn, “c” statistically significant versus S+, “d” statistically significant versus S−).

There was no difference in superior translation when comparing the intact state to any of the SCR constructs. When directly comparing the SCR constructs there was less superior translation with a double layered SCR with and without posterior sutures compared to single layered SCR with and without posterior sutures at 0° and 30° abduction and with single layered SCR without posterior sutures at 60° abduction (Figure 3 and Table 1).

Subacromial contact pressure

There was a significant reduction in subacromial contact pressure when comparing the torn state to all the SCR constructs. There was no difference in subacromial contact pressure when comparing the intact state to any of the SCR constructs (Table 2).

Table 2.

Subacromial peak contact pressure (MPa).

AA	Torn	%	D+	%	D−	%	S+	%	S−	%
0°	0.68	358	0.2b	105	0.22b	116	0.21b	111	0.33b	173
30°	1.24	413	0.35b,d	117	0.42b,d	140	0.47b	157	0.77b	257
60°	1.45	414	0.31b,d	89	0.36b,d	103	0.33b,d	194	0.68b	94

AA: abduction angle; D + : double layered graft with posterior suturing; D−: double layered graft without posterior suturing; S+: single layered graft with posterior suturing; S−: single layered graft without posterior suturing.

Data are expressed as average. Per cent was calculated by dividing the mean value by the intact value at the corresponding angle (“b” statistically significant versus torn, “c” statistically significant versus S+, “d” statistically significant versus S−).

When directly comparing the SCR constructs, there was less subacromial peak contact pressure with a double layered SCR with and without posterior sutures compared to single layered SCR without posterior sutures at 30° and 60° abduction, and with single layered SCR with posterior sutures compared to a single layered SCR without posterior sutures at 60° abduction.

Discussion

SCR using a HDA for the treatment of irreparable supraspinatus tears has been introduced as a potential alternative to SCR using FLA grafts.^17,20,21 SCR using a thicker 8 mm FLA grafts compared to 4 mm thick grafts has been previously reported to yield superior stability of the shoulder. ¹⁸ SCR using an 8 mm thick FLA graft showed improved superior stability compared to SCR using single layered HDA though subacromial contact pressures were restored to normal with both techniques. ¹⁷ Larger amounts of superior translation after SCR may have a negative impact on shoulder function after SCR and may also allow subacromial impingement which may lead to abrasion and failure of the graft over time.

The principal findings of this study are the use of a double layered HDA resulted in less superior translation compared to single layered HDA technique, particularly at lower abduction angles. The difference in subacromial contact pressure between a double layered HDA and a single layered HDA was less apparent though superiority with double layered HDA was noted in some testing conditions.

SCR using posterior sutures to the infraspinatus has previously been shown to reduce glenohumeral superior translation. ¹⁹ The addition of anterior sutures to subscapularis may reduce glenohumeral range of motion after SCR using FLA but this may not be the case after SCR using HDA. ¹⁷ The use of posterior and anterior sutures may also have a role in limiting graft elongation over time after SCR using HDA. ¹⁷ In this study, double layered graft with and without posterior sutures and single layered grafts with posterior sutures showed less glenohumeral translation and subacromial contact pressure compared to single layered grafts without posterior sutures in some testing conditions. Range of motion and graft elongation were not assessed in this study.

Due to the cadaveric nature of this study, there were several limitations. First, this study does not account for the biological healing potential. ¹² Furthermore, there were only two muscle loading conditions and they were static rather than dynamic. Due to the repair and testing sequence of each shoulder, the single layer graft may experience more stretching prior to testing. Additionally, only one specimen had an intact rotator cuff and this was used as control for all specimens. However, the main aim of this study was to examine the differences between the two repair groups (double layered versus single layered HDA) rather than to compare to the intact state.

Conclusion

The objective of this study was to compare the biomechanics of using a double layered HDA to a single layered HDA in the SCR. SCR using a single or a double layered HDA reduces superior translation of the humeral head after supraspinatus repair. Using a double layered HDA in the SCR provides greater stability than using a single layered HDA.