A paper from Kadum and colleagues (Int Orthop 2016) demonstrated 70 TESS arthroplasties and they reviewed, particularly the radiographs, center of rotation, humeral head height, neck shaft angle. They concluded that this is a challenging operation. It is important to determine the correct level of the humeral head cuts to avoid varus or valgus inclination; and to obtain the correct head size.
Another multi-centered study paper from Giles Walch, Pascal Boileau and others, (Collin, et al., Int Orthop. 2017)—47 stemless arthroplasties—2 revisions. But note that 17 of 47 or 36% had superior and lateral radiolucent lines. The authors concluded that radiographic lucent lines require continued observation and remain a concern in this population.
Another paper out of Germany (Beck et al., Int Orthop 2018) 51 stemless arthroplasties; 31 follow-ups (61%), revision rate 9.7%. Note that 29 of the 51 or 39% were lost to follow-up.
Another paper out of Vienna (Heuberer, et al., BMC Musculoskelet Disord. 2018) compared 73 stemless and 110 stemmed arthroplasties. They reviewed radiographs for humeral radiolucency, osteolysis or stress shielding. There were no clinical differences between the groups. But radiologic changes were noted in 37%, both superiorly and at the calcar.
Habermeyer and colleagues have published extensively in this area (J Shoulder Elbow Surg 2015). They first reviewed 78 stemless Eclipse shoulder reconstructions with a follow-up of 72 months. Complication rate 13%; revision rate 9%. The bone density decreased close to 35% over the course of this time. Two years later he published with co-authors and a multi-centered group (Habermeyer, et al., J Shoulder Elbow Surg 2017). A fewer number of stemless Eclipse—49 with 88% follow-up at 9 years. This group began in 2005 and yet they had a revision rate of 0% and one radiolucency. It’s not clear how many of the patients from the first group were in the second group.
So, why do I prefer short-stemmed implants?
There’s no difference in clinical outcomes between stemless and stemmed designs. Stemless implants are designed for problems that are rarely clinically relevant. Stemless implants have high numbers of radiolucencies in the literature, yet no long-term data. All stemless designs are not created equal. There are differences with respect to insertion technique, the presence of a collar, the type of surface coating and the bone contact surface area. And the question is will there be higher failure rates and less successful long-term outcomes as a result of these differences.
For now, stemless implants are a triumph of technology over reason.
Moderator Thornhill: Let me ask you a question—and this is for my own edification. Let’s define stemless. I think what you’re talking about is something that removes the humeral head and has some fixation in the metaphysis, but nothing in the diaphysis. Is that correct?
Dr. Romeo: Yes, that’s correct. What we’re calling stemless is where the head is removed and there is minimal preparation of the metaphyseal bone; no preparation of the diaphyseal bone. The stemless devices are really limited by predicate. And Frank mentioned 55mm…the predicate on U.S. market is you have to have a 7cm stem on the lateral aspect. So, if you try to make it shorter than that, you have to prove to the FDA that it adds value. And then the long stems are anything beyond that.
Moderator Thornhill: So, it’s important to remember that you’re talking about a little short something on the metaphyseal side, but not just an onlay surface.
Frank, you’re saying that a short-stemmed implant is more predictable. The data are better. Why not do what Tony does?
Dr. Cordasco: I think the humeral side of a total shoulder replacement has not generally been a problem in the last 40-some-odd years. We’ve, again, begun to shorten the stem over the course of time because it’s pretty clear that it’s not necessary to have a very long stem. Even in periprosthetic fractures some people are talking about not necessarily bypassing that with a longer stemmed implant. So, my concern about this particular device is that there are multiple devices…we’re not sure which one is the best. One of the goals is to prevent stress shielding and yet many of the outcomes in the European literature demonstrate just that in the proximal humerus. The big concern is we try to fix a small problem to develop an implant that will provide longer term…much more significant I should say…problems going forward. In my view, the devil’s in the details.
