Optimal Tension in Reverse Total Shoulder Replacement: A Tool to Help Determine?

A blog commentary on the article “Clinical significance of intraoperative glenohumeral joint load evaluation using a novel humeral sensor in navigated reverse total shoulder arthroplasty” by Wang, et al in Seminars in Arthroplasty: JSES

Introduction

Reverse shoulder replacement (rTSA) was originally developed for the arthritic shoulder with an irreparable rotator cuff tear to allow function in the absence of the rotator cuff. One of the primary principles is lengthening of the deltoid for more effective function. The optimal amount of deltoid lengthening is yet to be determined. rTSA has been referred to as a “Goldilocks” operation. If the reverse shoulder replacement is put in too loose, there’s inadequate tension on the deltoid, which may lead to weakness and significant risk of instability. Over tensioning the rTSA may be a risk factor for acromial or scapular spine stress fracture, postoperative pain possibly due to tension in the conjoined tendon, or tension related neuropathies. So, not too tight, not too loose – just right. Currently, determination of the optimal tension is very subjective and changes with experience. In the article “Clinical significance of intraoperative glenohumeral joint load evaluation using a novel humeral sensor in navigated reverse total shoulder arthroplasty,” the authors describe the use of an intraoperative load measuring trial implant to measure intraoperative joint load in various positions of the arm.

Methodology

The authors performed a pilot study to evaluate the safety and utility of an intraoperative load sensor. Their secondary goal was to correlate these loads to outcomes and complications. All 15 surgical procedures were performed by a single surgeon with a standardized technique. Surgeries were planned and performed using ExactechGPS navigation. For consistency, the same glenosphere size was chosen for all patients (38mm), and the subscapularis tendon was not repaired at the conclusion of the procedure. Trialing was carried out with the trial liners that incorporated a load sensor. The device, Orthosensor, is a load sensor with three sensors covered in plastic that convert digital voltages to measurements in pounds of force. Measurements were recorded via Bluetooth onto a tablet in the operating room. Once the surgeon was satisfied with passive range of motion and stability, measurements were recorded in four positions: neutral (N), with the arm in 0⁰ of abduction and flexion and neutral rotation; across chest (AC) with thumb tip of operative arm placed to contralateral acromioclavicular joint; behind back (BB) with thumb tip over most superior point of ipsilateral iliac crest; and overhead (OH), with arm in 90⁰ of forward flexion, 60⁰ of abduction, and 60⁰ of external rotation. The surgeon was blinded to the recordings.

Results

Patients were evaluated at 3- and 12-months post-surgery, and radiographs were performed at 6 weeks and 12 months. All joint load measurements and wireless recordings were made without complications; at 12 months there were no stress fractures or episodes of instability. A wide variation in joint loads was measured. At neutral, the mean load was 6.1 lbF. In the other three positions (the functional positions), the mean loads were in the 30-40 lbF range. The load in BB position significantly correlated with both N and AC positions, but no OH. At 3 months, no clinical correlations were identified. At 12 months, the ASES score was significantly associated with load in the BB position. In addition, the ASES pain subscore was significantly correlated with load in the BB and AC positions. Despite these correlations, motion behind the back was still rated as difficult in many patients.

Discussion

A primary weakness of the study is its small numbers (due to it being a pilot study). This limits the ability to find thresholds – perhaps loads that are too tight at which point complications like stress fractures occur or outcomes decrease, or conversely loads that are too low, leading to instability or weakness. Another limitation is that recordings were performed with muscle relaxation passively. This may not correlate to usage of the arm actively. Reproducibility of the arm positions is another possible limitation, along with single measurements. However, a prior cadaveric study showed high repeatability of measurement. Repair of the subscapularis, when performed, may further complicate the understanding of this. Finally, this study did not evaluate range of motion for potential correlation, though this introduces more variability.

Conclusion

In my opinion, the authors have performed an excellent pilot study to help in the objective determination of optimal tension in rTSA. Technological advances are helping us better understand factors associated with outcomes and complications as well as helping with optimal surgical techniques. As we begin to combine these technologies, such as preoperative planning, intraoperative navigation and load-sensors, augmented and even customized implants, postoperative outcome and complication databases, and machine learning and artificial intelligence, we can see a potential future where we as surgeons can optimize a surgery for a particular patient, based on their diagnosis, preoperative function, imaging, and postoperative goals, while minimizing complications.