Total shoulder and hemiarthroplasty: glenoid failure, humeral osteolysis, revision surgery

The series was actually a combination of the individual experience of five surgeons, all of which used the Aequalis shoulder system.
   one performed uncemented hemiarthroplasty (31 cases),
   one performed cemented hemiarthroplasty (16 cases),
   one performed total shoulders with uncemented stem and cemented convex back glenoid (54 cases),
   one performed cemented stems and uncemented metal backed glenoids (34 cases), and

   one performed predominantly cemented stems with flat or convex backed glenoid components (183 out of his total of 260).

With respect to the glenoid components
   (1) over 2/3 rds of the glenoids had radiolucent lines on the immediate postoperative radiographs. Comment:We have found that postoperative radiolucent lines can be eliminated by the use of a sterile C02 jet that removes debris, clot, blood, and other fluids from the bone cement interface.
   (2) 39% of the glenoid components had loosened. Comment: this represents an annual loosening rate of over 4% per year. 

   (3) tilt of the glenoid component was found in 53.5%, subsidence in 38%, and loosening without tilt or subsidence of the component in 8.5%.
  (4) Twenty-nine (36.7%) of the seventy-nine metal-backed glenoid components were loose. All of the patients with loosening of the component had complete wear of the polyethylene insert.

Osteolysis of the greater tuberosity and/or the calcar was
   (1) found in over 40% of patients with a total shoulder replacement but in none of the shoulders with hemiarthroplasty
   (2) more frequent in shoulders with glenoid loosening than in shoulders without glenoid loosening. Of the 128 shoulders with a loose glenoid component, 121 (94.5%) had osteolysis of the greater tuberosity and/or calcar. These patients had significantly lower clinical outcome scores.
   (3) more frequent in shoulders with wear of the polyethylene insert. In patients treated with a metal-backed glenoid implant (polyethylene wear was present in fifty-three (67.1%) of seventy-nine shoulders). All shoulders with wear of the glenoid insert had osteolysis of the proximal humerus.
   (4) The observation that osteolysis only occurred in total shoulder replacements, supports the hypothesis that wear particles from polyethylene are likely to be responsible for osteolysis than are cement particles.

Stress-shielding was observed only with uncemented stems and was present in almost 2/3rds of uncemented humeral stems (this was the case for both total shoulders and hemiarthroplasties). Comment: With a press fit component the load is transferred from the component to the humerus only at certain locations, usually where the distal tip of the prosthesis is wedged into the firm diaphyseal bone. This unloads the proximal humeral metaphyseal bone and leads to bone loss through stress shielding. While this effect can be prevented by the use of cement, which enables the use of a less tight fitting diaphysis, cement complicates removal of the prosthesis should revision become necessary. The use of  impaction grafting obviates the problems of cement as well as the problem of stress shielding; this has been for many years our standard approach to fixation of the humeral implant.

Revisions
   (1) 53 revision procedures (13.4%) were performed.  Comment: this represents an annualized revision rate of 1.67% per year. Eight revisions were for soft tissue problems, 14 revisions were to a reverse total shoulder because of cuff failure and glenoid loosening, 21 other revisions involved removing a loosen glenoid component. One hemiarthroplasty was converted to a total shoulder, one periprosthetic fracture was fixed, and one humeral head implant was changed.
   (2) Revisions were performed in 1 of 66 hemiatrhoplasties (1.5%). Only 16 had moderate or severe glenoid erosion. 30 of 79 shoulders with metal backed glenoids (40%). 20 of 89 shoulders with a flat-backed all-polyethylene glenoid component (22.5%). 2 of 94 shoulders with a convex-backed glenoid component (2%). Comment: it is of interest that the revision rate for hemiarthroplasty was less than that for total shoulder.

Infection
   The authors note that a recently published study found that male sex, glenoid wear, humeral loosening, and humeral osteolysis are associated with a significantly increased likelihood for a positive culture of Propionibacterium acnes in revision shoulder arthroplasty. Comment: While clinically significant humeral loosening was nearly absent in this cohort and there was no evidence for low-grade infections in these patients, the authors’ approach to culturing for Propionibacterium is not stated and it is recognized that the presence of this organism is often overlooked unless specific culturing protocols are followed. It is of interest that the findings of humeral osteolysis and glenoid wear have been shown to be strongly associated with positive cultures for Propionibacterium.

Again, this is a most informative review of shoulder arthroplasties performed by expert surgeons and followed for a substantial period of time.