The Superior Capsular Reconstruction

Lessons Learned and Future Directions

John M. Tokish, MD; Justin L. Makovicka, MD


J Am Acad Orthop Surg. 2020;28(13):528-537. 

In This Article

Superior Capsular Reconstruction Controversies and Considerations

Autograft Versus Allograft

SCR using dermal allograft has shown early encouraging results but has not reached the level of success seen when using fascia lata autograft. Biomechanical cadaveric research has compared fascia lata allografts with human dermal allografts in SCR (Table 2).[25] Shoulders with simulated irreparable rotator cuff tears were tested in three conditions: (1) SCR using a fascia lata autograft with anterior and posterior suturing, (2) SCR using human dermal allograft with anterior and posterior suturing, and (3) SCR using human dermal allograft with posterior suturing only. Both the superior glenohumeral joint force and subacromial contact pressure were completely restored in all three reconstruction scenarios, but although superior translation was completely restored with the fascia lata graft, it was only partially restored with the dermal allograft. Also, of note, despite partial restoration of the superior translation with the dermal allograft, it still remained greater than that of the intact condition. The ability of the dermal allograft to restore joint force and subacromial contact pressure suggests that it might decrease symptoms seen with subacromial impingement, but the ability of SCR to restore shoulder function may be dependent on its ability to depress the humeral head, in which the dermal allograft was only partially able to do. In addition, the dermal allografts were found to elongate 15% during the testing, while the length of the fascia lata grafts was unchanged.[25] However, it should be noted that a limitation of biomechanical cadaveric studies are that only time-zero effects are measured. Thus, the effect of biologic healing and graft incorporation on results cannot be determined, although dermal allograft and fascia lata grafts likely have different capacities to incorporate.

Graft Thickness

The difference in graft thickness may play a major role in the outcomes seen after SCR. The native shoulder capsule has been found to be about 4.4 to 9.1 mm thick at the attachment of the greater tuberosity, with the fascia lata autografts used for SCR in Mihata's work falling within this range (6 to 8 mm thick).[14,26] By contrast, the maximal thickness of commercially available dermal allografts are 3 to 4 mm.[16] Mihata et al[26] compared the subacromial peak contact pressure and glenohumeral superior translation with both a 4- and 8-mm fascia lata graft (Table 2). Subacromial peak contact pressure was decreased after SCR using both a 4-mm (113% decrease at 30° glenohumeral abduction) and 8-mm graft (158% decrease at 30° glenohumeral abduction).[26] By contrast, only the 8-mm graft was found to decrease superior translation after SCR (130% decrease at 30° glenohumeral abduction), and they concluded that an 8-mm graft had greater stability than a 4-mm graft.

The effect of graft thickness on outcomes is even more evident when stratifying clinical results by the thickness of the graft used. The dermal allografts used in Denard et al.'s reconstructions ranged in size from 1 to 3 mm in thickness. The overall rate of success reported in their reconstructions was 67.8%, while only 40% in those with 1-mm thick grafts.[16] When reconstructions done with 1-mm grafts were excluded, the success rate rose to 75.5% in Hamada grade 1 or 2 shoulders. They also found graft thickness to play a role in the rate of graft healing, as a greater percentage of those who healed on postoperative MRI were 3 mm thick. The healed group also demonstrated decreased postoperative pain and higher ASES scores. However, long-term follow-up on graft healing is needed because thicker grafts may not completely incorporate leading to late failure of the graft after initial healing. Hirahara et al's[27] reconstructions were also improved with thicker grafts. Mean ASES scores increased to 90 (86.46 in all), and mean VAS pain scores dropped to 0 (0.38 in all) when only including reconstructions in which a graft >3 mm thick was used.

Graft Tensioning Position

It has been demonstrated that a proper functioning superior capsule after reconstruction is critical for joint stability. The size of the superior capsule in the native shoulder changes depending on abduction and rotation of the shoulder, with the distance between the superior glenoid and greater tuberosity decreasing with increasing amounts of shoulder abduction.[26,28–31] Stiffness of the shoulder is seen with tight, thick capsules, while laxity is seen with elongation of the capsule, thus an optimal position for reconstruction improves force couples and provides for a well-functioning shoulder.[26] Although SCR in original studies was done with the arm at 45° of shoulder abduction, Mihata et al[26] determined through biomechanical analysis that SCR restored superior stability of the shoulder at time 0, when the graft is attached between 15° to 45° of shoulder abduction (10° to 30° glenohumeral abduction) (Table 2). After simulation of an irreparable rotator cuff tear, subacromial peak contact pressures and superior translation were markedly decreased after SCR with the graft tensioned at both 15° and 45° of shoulder abduction.

Incorporation With Native Rotator Cuff

In an irreparable rotator cuff tear, the superior capsule has a defect leading to transverse discontinuity of shoulder capsule.[13,15] Mihata et al evaluated the biomechanical role of side-to-side suturing of the graft during SCR to residual rotator cuff tissue in restoring shoulder stability after irreparable cuff tears (Table 2). Superior translation was markedly decreased when the SCR was repaired to the remaining posterior and anterior cuff but was not decreased in SCR alone.

In a second biomechanical study, Mihata et al demonstrated that anterior and posterior side-to-side suturing of a fascial lata allograft during SCR led to a decrease in total glenohumeral range of motion (ROM) (Table 2).[25] This, along with their previous work that noted no change in outcomes with the addition of anterior side-to-side sutures, indicates that anterior side-to-side sutures may not be necessary nor desired when using a fascia lata graft for SCR. By contrast, anterior and posterior side-to side suturing when using a dermal allograft for SCR did not result in decreased total ROM, while dermal allograft SCR with only posterior side-to-side suturing actually increased glenohumeral ROM relative to an intact shoulder. Owing to the increased dermal allograft elasticity, proper graft tensioning can be achieved by securing the graft both anteriorly and posteriorly to remaining cuff tissue, thus limiting elongation in the medial-lateral direction and improving overall stability.[27] The unlikeliness of shoulder stiffness, along with the potential advantage of limiting medial-lateral graft elongation, led Mihata et al[25] to recommend both anterior and posterior side-to-side suturing when using a dermal allograft for SCR. Hirahara et al reported that their patients who had anterior marginal convergence done, in addition to posterior marginal convergence, during SCR had a notable increase in ASES scores compared with those that only had posterior.[27]


Rotator cuff tears are often accompanied by bony impingement of the undersurface of the acromion.[32,33] Although acromioplasty may eliminate this, it also has the potential to increase superior instability due to the loss of the coracoacromial arch and ligament. In a cadaveric analysis, Mihata et al[34] compared the effects of SCR with and without acromioplasty on shoulder biomechanics (Table 2). SCR without acromioplasty markedly shifted the humeral head inferiorly, decreased superior translation, and decreased subacromial peak contact pressure. Adding acromioplasty to SCR decreased subacromial contact area, but did not change humeral head position, superior translation, or subacromial peak contact pressure compared with SCR alone. It was concluded that acromioplasty decreased the subacromial contact area without increasing subacromial contact pressure and thus may safely be done with SCR to decrease the risk of postoperative abrasion and tearing of the graft beneath the acromion.[32]