These authors evaluated the minimum two year outcomes, revisions, and complications between a first-generation cemented modular humeral implant (Group 1, N=400) and a second-generation non-modular, primarily uncemented humeral implant (Group 2, N=231) in reverse total shoulder arthroplasty (RSA) with 135-degree neck shaft angle and varying degrees of metallic glenosphere offsets. Impaction autografting was performed on all press fit cases.
The average age of the study population was 70.6 years (range 22-91); for female patients 71.6 years (range 22-91); and for male patients 69.1 years (range 24-87). The majority of patients were diagnosed with cuff tear arthropathy (35%), followed by massive cuff tear without osteoarthritis (22%), osteoarthritis (17%), massive cuff tear with osteoarthritis (7%), malunion/nonunion (5%), inflammatory arthritis (4%), acute fracture (6%), infection (1%) and other (avascular necrosis. chronic dislocation, post-traumatic osteoarthritis, instability, congenital brachial plexus palsy; 4%).
All patients received a subscapularis repair, even if not fully reparable (e.g. loss of tendon). The postoperative protocol emphasized a physician-directed home therapy program and included the use of a shoulder immobilizer for six weeks with gentle pendulum exercises. This will then progress to a light sling and active-assisted range of motion in the supine position. As tolerated by the patient, they are then allowed to progress with a focus on strengthening and stretching exercises. At no point in time are patients prescribed formal physical therapy.
Both groups of patients had similar improvement in clinical outcomes.
The incidence of humeral loosening for the cemented group was 3.6%, whereas in the uncemented group it was 0.4% (p= 0.01).
A total of 28 shoulders treated with cementing technique (4.0%) and 6 patients treated by press-fit technique (1.5%) were revised (p= 0.028).
The rate of postoperative acromial fractures within the first year was 3.4% in the cemented group and 1.8% in the uncemented group (p= 0.177).
A total of 23 shoulders (10 shoulders were revised prior to 2-year follow-up) treated with the modular implant (23/607; 3.8%) and 11 shoulders (6 shoulders were revised prior to 2-year follow-up) treated with the monolithic implant (11/478; 2.3%) were revised (p=0.219).
In Group 1, the causes for revision included recurrent instability (n=7), infection (n=7), humeral loosening (n=4), periprosthetic fracture (n=2), glenosphere dissociation (n=1), baseplate failure (n=1), and failure at the modular junction (n=1).
In Group 2, the causes for revision included periprosthetic fracture (n=3), baseplate failure (n=3), recurrent instability (n=2), humeral loosening (n=1), infection (n=1), and glenosphere dissociation.
The average time to revision in the modular group was 43 months (range: 1 to 161 months) and in the monolithic group was 30 months (range: 1 to 87 months; p=0.318).
A total of 28 shoulders treated with cementation technique (4%) and 6 patients treated by press-fit (1.5%) were revised (p= 0.028).
In the cemented group, the causes for revision included recurrent instability (n=7), infection (n=8), humeral loosening (n=5), periprosthetic fracture (n=5), glenosphere dissociation (n=1), baseplate failure (n=1), and failure at the modular junction (n=1).
In the press-fit group, the causes for revision include baseplate failure (n=3), instability (n=2), and glenosphere dissociation (n=1). The average time to revision in the cemented group was 44 months (range, 1 to 161 months) and in the uncemented group was 13.5 months (range, 1-38 months; p= 0.005).
Comment: This article represents the good outcomes and low complication rate from the practice of a surgeon with a very large experience in reverse total shoulder arthroplasty. It is informative to see the surgical and rehabilitation protocols resulting from this large experience.
Our practice is similar to that represented by Group 2 - using a non-modular humeral implant fixed with impaction autografting. Our technique is detailed in this link.
===
