Dr. Lederman: I appreciate the opportunity to debate Butch, who I have the utmost respect for. Butch probably has a breadth of experience and a breadth of knowledge beyond his years, and he’s always probably the worst person you want to debate.
To achieve an anatomic total shoulder replacement one of the most important factors is anatomic humeral reconstruction. The indications for the stemless and the resurfacings are quite similar-to avoid the diaphysis and to be able to put the humeral head independent of the diaphyseal axis and perhaps, easier revisions which may be one of the main indications.
As Butch mentioned, the ability to recreate the center of rotation of the humeral head was challenging with the first and second generation implants. Resurfacing arthroplasty was introduced in 1986 so it’s been around for quite some time to try to deal with this. With improvements in stem designs with third and fourth generation implants, the ability to recreate the normal anatomy has gotten much better.
Does the stem really matter? We have a lot of options.
What matters is the ability to reproducibly get the anatomy right, to get glenoid access and to get the ability to revise without challenges. The advantage of a surface replacement is that it is a relatively simple technique. We can maintain the normal anatomy, there is a large surface area for ingrowth, and it may be cheaper by the nature of its one-piece design.
So, what does literature tell us?
An original study by Levy and Copeland (JBJS-Br, 2001): 285 prostheses, 7-year follow up, 8% revision rate primarily due to glenoid-based problems and they were implanted both as a total and as a hemi.
Looking at patients in the 50-year-old range—Dave Bailie, et al. (JBJS-A, 2008) did 36 patients of hemiarthroplasties with 3 biologic resurfacings, 35 of the 36 were satisfied at a follow up of a little over 3 years. Levy, et al. (JSES, 2015) looked at a mean follow up 14.5 years with 54 shoulders and 49 patients. There were 2 revisions for infections, 7 revised to a stem, but none of these were due to humeral component complications. It had an overall revision free survivorship at an average of 14.5 years of 81%.
Looking at the Australian Registry in 2016, resurfacing and stemless represent only 12% with the stemmed implant being the primary go-to implant—at least in Australia. As we look at the revision rates at 7 years, the revision rates are similar but slightly higher revisions for resurfacing over stemmed implants (12.6 v. 10.3, respectively). The main reason for revision is only 18% when it’s related to the implant. The others are related to issues such as glenoid-based problems or rotator cuff insufficiency.
Butch is right, the ability to get these resurfacing implants right is difficult. Proper sizing and positioning is a challenge and the same challenges exist for the stemless arthroplasty. They are mostly placed in varus when they are missed. They can be oversized or undersized and getting it right can be challenging. Multiple authors have pointed this out and a lot of the surgeons performed better at anatomically reproducing the premorbid humeral head anatomy with a stemmed arthroplasty compared to a resurfacing arthroplasty.
Stemless devices are built to have robust metaphyseal fixation. Is there potentially a problem in removing these? Nowadays short stem devices can be easily removed preserving the bone envelope of the proximal humerus.
There are many options for anatomic humeral replacement. The resurfacing implants are bone preserving, there is minimal concern for stress shielding and loosening over the long term, they have stood the test of time in use for 30 years and they are revisable.
Stemmed devices may be more predictable for anatomic restoration, and stemless may offer improved glenoid exposure as a primary advantage.
