Latissimus Dorsi Tendon Transfer and Reverse Shoulder Arthroplasty: Restoring Mechanical Advantage by Transfer Posterior to the Long Head of the Triceps Tendon

Background: Patients undergoing reverse shoulder arthroplasty (RSA) for rotator cuff arthropathy may present with an external rotation (ER) lag due to posterior rotator cuff insufficiency. As a result, the addition of a latissimus dorsi (LD) tendon transfer in combination with RSA has become increasingly utilized. Initial descriptions of LD tendon transfer involved rerouting of the LD tendon posterior to the long head of the triceps tendon. However, more recent techniques involve rerouting the LD tendon anterior to the long head of the triceps tendon via the deltopectoral approach utilized for RSA. The purpose of this cadaveric study was to assess the biomechanical effect on ER force of LD tendon transfer anterior versus posterior to the long head of the triceps tendon in combination with RSA.

Methods: Eight fresh frozen cadaveric shoulders were utilized. A shoulder fellowship trained orthopedic surgeon performed RSA in each specimen. The rotator cuff tendon attachments were cut to simulate a massive rotator cuff tear. The LD tendon was transferred to the greater tuberosity first anterior to the long head of the triceps, and then posterior to the long head of the triceps. The specimens underwent biomechanical testing for each condition using an established cadaveric shoulder simulator with 6 degrees of freedom for glenohumeral joint motion.

Results: Both anterior and posterior LD tendon transfers successfully achieved ER of the humerus when force was applied to the LD tendon. Posterior LD tendon transfer resulted in less required force to achieve ER of the arm throughout range of motion compared to anterior LD tendon transfer (33.4N vs. 48.6N, p<0.001), which was maintained irrespective of glenohumeral abduction angle. Loading of the long head of the triceps tendon was associated with significantly increased force required for generation of ER only in the anterior LD tendon transfer condition (48.6N loaded vs. 39.4N unloaded; p<0.001).

Conclusion: LD tendon transfer posterior to the long head of the triceps resulted in more efficient humeral ER movements in the setting of RSA, irrespective of degree of shoulder abduction. Our results also demonstrate that long head of the triceps tension results in decreased efficiency of the LD tendon transfer when performed anterior to the long head of the triceps. While technically more difficult, transfer of the LD tendon posterior to the long head of the triceps provides a mechanical advantage that may improve clinical outcomes in patients with RSA lacking ER.