Biomechanical effectiveness of tendon transfers to restore active internal rotation in shoulder with deficient subscapularis with and without reverse shoulder arthroplasty

Background Loss of active shoulder internal rotation can be very disabling. Several tendon transfers have been described for the management of an irreparable subscapularis (SSC) tear. The purpose of this study was to determine and compare the internal rotation moment arm (IRMA) of the sternal head of the pectoralis major (PM), latissimus dorsi (LD), and teres major (TM) when transferred to different insertion sites to restore shoulder internal rotation with and without reverse shoulder arthroplasty (RSA). Methods Six fresh-frozen right hemithoraces were prepared and evaluated using a custom tendon transfer model to determine the IRMA of different tendon transfers using the tendon and joint displacement method. Five tendon-transfer pairs were modeled using a single suture and tested before and after implantation of an RSA (Comprehensive; Zimmer-Biomet, Warsaw, IN, USA): PM to the insertion site of the SSC, LD to the anterior insertion site of the supraspinatus (SSP) tendon on the greater tuberosity, LD to SSC, TM to SSP, and TM to SSC. The SSC was not repaired at the end of the RSA procedure to simulate an SSC deficiency. The PM transfer was passed under the conjoined tendon when tested on the intact shoulder and above the conjoined tendon when tested with an RSA. Results Tendon transfers were shown to have a significant effect on IRMA. The effect of transferred tendons was significantly affected by the position of the humerus. With the humerus adducted, the IRMA of the TM-SSP (14.1 mm ± 3.1 mm) was significantly greater than the other transfers. With the humerus abducted to 90°, the IRMAs of the LD-SSP (30.0 mm ± 5.4 mm) and TM-SSP (28.4 mm ± 6.6 mm) were significantly greater than the IRMAs of other transfer options. The IRMA of the native shoulder differed significantly from that of the RSA state for all tendon transfers. With the humerus adducted to the side of the body, the IRMA of the RSA PM-SSC transfer was significantly greater than that without an RSA (19.0 mm ± 6.4 mm vs. 7.1 mm ± 0.9 mm), demonstrating increased efficiency for internal rotation in the RSA state. Conclusion Tendon transfers to restore shoulder internal rotation differ in effectiveness and may be affected by arm position and by implantation of a lateralized humerus/lateralized glenoid RSA. The LD potentially results in superior restoration of shoulder internal rotation in a native shoulder (given the risk of nerve compression with the TM transfer) compared with PM and should be considered as a potential tendon transfer to restore internal rotation in selected patients. In combination with a lateralized humerus/lateralized glenoid RSA, the fulcrum provided by the biomechanics of the semiconstrained implant allows the PM transfer to become a more efficient tendon transfer to restore active internal rotation.