UW Shoulder and Elbow Academy: Reverse total shoulder in revision of failed anatomic arthroplasty

The Effect of Glenoid Bone Loss and Rotator Cuff Status in Failed Anatomic Shoulder

Arthroplasty after Revision to Reverse Shoulder Arthroplasty

These authors evaluated outcomes and risk for re-revision in patients with a failed anatomic total shoulder arthroplasty (TSA) revised to reverse shoulder arthroplasty (RSA) based on rotator cuff deficiency and glenoid bone loss.

Fourteen patients were treated with retention of humeral stem and conversion to reverse shoulder arthroplasty using a convertible platform. Six of the revisions were performed as a staged procedure: two were due to infection, three were due to the severity of glenoid bone loss and one was due to intra-operative myocardial infarction. The five staged revisions for infection or glenoid bone loss had loose glenoids on preoperative radiographs.

Two year outcomes were available in 123 cases. The overall mean preoperative SST score was 2 with no significant difference in preoperative SST scores if the glenoid was fixed or loose (p=0.893) or if the rotator cuff was intact or deficient (p=0.627). The overall mean postoperative SST score was 5 with no significant difference in postoperative SST scores if the glenoid was fixed or loose (p=0.876) or if the rotator cuff was intact or deficient (p=0.234).

The overall re-revision rate was 11.4% with a mean time to re-revision of 22 months (range 0-89 months). The most common reason for subsequent revision was glenoid-sided failure (11/18). There were three RSAs revised for humeral loosening, three with subsequent infections, and one revised for instability. Of the 11 RSAs with glenoid sided failure, all had loose glenoids (100%) on preoperative radiographs at time of the initial revision. This is compared to 53.8% (n=85) cases with loose glenoids on preoperative radiographs in the group that were not subsequently revised (n=83) or those revised for reasons other than glenoid-sided failure (n=2, humeral loosening and infection, respectively). Rates of rotator cuff deficiency were similar between the revision groups.

The odds ratio was 1.8 for subsequent re-revision with glenoid loosening compared to those without loose glenoids on preoperative radiographs.

Comment: The success of anatomic and reverse total shoulder arthroplasties depends in large part on the preservation of glenoid bone stock. This is why the desire to change glenoid version needs to be curbed by the need to preserve bone. It is also why surgeons should carefully consider implants that would compromise glenoid bone stock should they need to be removed.

It is of note that in the hands of one of the most experienced shoulder surgeons, more that one in ten of these revisions required re-revision. It is likely that the failure rate would be higher for many of the rest of us.

For this reason, we consider revising failed anatomic total shoulders with glenoid deficiency to a hemiarthroplasty as shown in the three cases below; each of these patients had pain relief and satisfactory function two years after the revision.