A roundtable discussion with members of the ASES

The number of shoulder arthroplasty procedures performed each year in the United States is increasing rapidly; estimates range from 55,000 to 80,000. Of these, about half are “reverse” shoulder arthroplasties (RSAs), in which the normal anatomy of the shoulder is switched. Rather than having the ball (top of the humerus) fit into the cup (glenoid socket) (Fig. 1), the glenoid is fitted with a metal ball, and the cup is fixed to the upper end of the humerus (Fig. 2).

This reverse approach was developed by the French surgeon Paul M. Grammont, MD, and approved in the United States by the U.S. Food and Drug Administration in 2004. Since then, it has become the procedure of choice when the patient has massive rotator cuff arthropathy and, as surgeons become more familiar with the approach and implant designs are refined, indications for its use are continuing to expand.

Recently, four members of the American Shoulder and Elbow Surgeons (ASES) conducted a roundtable discussion on RSA for AAOS Now. Joining moderator Richard J. Hawkins, MD, Steadman Hawkins Clinic of the Carolinas (Greenville, S.C.), were T. Bradley Edwards, MD, Fondren Orthopedic Group (Houston, Texas); Mark A. Frankle, MD, Florida Orthopaedic Institute (Temple Terrace, Fla.); and Joseph D. Zuckerman, MD, NYU Hospital for Joint Disease (New York City).

Dr. Hawkins: Let’s start by defining what an RSA is.

Dr. Edwards: A reverse shoulder involves a fixed fulcrum, so it basically puts some kind of a sphere or hemisphere variation on what is normally the socket side or the glenoid side, and on the humeral side, it puts a cup that allows the fulcrum to function and basically allows the deltoid to power the implant at least in elevation and abduction.

Dr. Hawkins: Why is this design effective in dealing with a patient who has a deficient rotator cuff?

Dr. Frankle: The normal shoulder has a balance of forces that keep the humeral head centered within the glenoid, regardless of arm position. When the rotator cuff is deficient or muscular insufficiency or imbalance exists, the humeral head may not stay centered in the socket and may dislocate superiorly. The joint becomes unstable and the fulcrum for normal elevation is lost. Reversing the articulation provides stability by neutralizing the forces that would normally be triggered by the rotator cuff.

Dr. Hawkins: Those principles were first described by Dr. Grammont, who suggested that the center of rotation of the glenoid sphere should be near the glenoid surface rather than the normal 15 degrees lateral. Certain implant designs enable us to get out to about 10 mm from the center of rotation. Now let’s look at the indications to do a RSA.

Dr. Zuckerman: The original indications were for a shoulder with glenohumeral arthritis and an incompetent rotator cuff. The primary indication was a massive rotator cuff tear or so-called rotator cuff arthropathy. The RSA offered a significant advantage over the standard anatomic shoulder arthroplasty. Hemiarthroplasty didn’t really offer significant improvements and put people in the limited goals category. So the RSA was a significant advancement in our treatment armamentarium.

Dr. Edwards: From that initial description of cuff tear arthropathy, the indications for RSA have expanded to include patients who have a dysfunctional cuff or a cuff that was not balanced enough to allow elevation. One example would be patients who had a conventional total shoulder arthroplasty (TSA) and now have a deficient, unbalanced rotator cuff. The patient can no longer lift the arm and an RSA would be used.

From there, indications expanded to patients without arthropathy but with large rotator cuff tears who cannot elevate the arm. These are the so-called “pseudoparalytic patients.” I think it’s important to note that these are not patients with just an isolated retracted supraspinatus tendon tear and full elevation or elevation beyond horizontal. These are patients who have a truly dysfunctional, unbalanced rotator cuff with no active elevation. In these patients, the RSA can restore the fulcrum and provide good range of motion.

RSA has also been used for a host of other arthropathies—such as rheumatoid arthritis—that are associated with dysfunctional cuffs. Finally, an RSA may be useful in trauma situations, such as a posttraumatic nonunion or malunion or an acute traumatic situation. In elderly patients who have a very limited chance that the tuberosities will heal around a hemiarthroplasty, the RSA is being used with great success.

Dr. Zuckerman: I think the indications are continuing to evolve. I will consider an RSA for an 85-year-old patient with glenohumeral arthritis and an intact rotator cuff with some degeneration but without a full-thickness defect. I now also use an RSA in patients with Parkinson’s disease, recognizing the variation in the severity of the condition. I think it’s been beneficial for those who have significant symptoms.

Dr. Hawkins: What about revising a failed hemiarthroplasty or a failed TSA to an RSA? Is that something that you see in your practice often?

Dr. Frankle: Well, I think this is an indication for an RSA, but I’ve learned that the mode of failure must be considered in planning a revision strategy. Patients with a failed hemiarthroplasty due to a four-part fracture resulting from tuberosities that retracted or never healed have humeral bone loss and scarring around the tuberosities. Often the glenoid is relatively pristine, but dealing with refractive scarring in the tuberosities is a challenge; if humeral bone loss is excessive, developing strategies to restore that can be important.

Patients with cuff tear arthropathy who were treated with hemiarthroplasty may experience progressive erosion against the glenoid or the acromion, resulting in increased pain. If glenoid bone loss or a thinned acromion exists, a surgeon must alter the technical methodology to fix those problems. I think it’s important to consider these aspects of the surgical and technical care to avoid complications and be prepared.

Dr. Zuckerman: Ten years ago, complex proximal humeral fractures were either fixed or underwent hemiarthroplasty. Now, new advancements in locking plate technology have reduced the role of hemiarthroplasty in the treatment of these fractures. Even so, probably a third or more of all complex humeral fractures are being treated with RSA as the primary treatment.

Dr. Hawkins: What is the technical difference between the surgical approach to an RSA compared to a conventional TSA?

Dr. Edwards: I use the same surgical approach for both—a deltopectoral approach. But the RSA gives the surgeon an option because the cuff is deficient. The surgeon can choose a superior approach with a small deltoid detachment or just a deltoid splitting, and several surgeons have good experience with that.

During the procedure, there is not a lot of difference in exposing the glenoid. The exposure sometimes can be easier in an RSA due to the difference in soft tissues. But other technical aspects of an RSA can be more difficult, such as ensuring stability and appropriate tension. Glenoid positioning is important to avoid some of the potential issues with scapular notching, depending on the implant system used.

The typical Grammont-style implant with a 155 degree neck inclination angle requires a medialized center of rotation and deltoid tension supplies a large part of the stability of the implant. Newer implants with a 145 degree neck inclination angle enable a different technique on the glenoid, somewhat more inferior. This means I can put less tension on the deltoid when reducing it.

Most important, however, is a full range of motion and no impingement anywhere that could lever the implant out. If it seems stable, I’m happy.

Dr. Hawkins: I think most shoulder surgeons use the deltopectoral approach, although some prefer the superior approach, which has advantages and disadvantages. I wonder if each of you might comment on repairing the subscapularis. In our opinion, repair of the subscap is unnecessary if a lateralized center of rotation is used.

Dr. Zuckerman: I prefer to repair the subscapularis if possible, but I also encourage an aggressive postoperative rehabilitation program, progressing to active range of motion at 2 weeks.

Dr. Frankle: I think that the subscapularis has an important role in internal rotation and provides horizontal forces that help stabilize the joint. I have even used an allograft in the repair.

Dr. Edwards: I always try to repair the subscapularis if possible, but I’ve also found that the prosthetic design may play a role in stability with regard to whether repair is important.

Dr. Hawkins: I think the difference may be related to center of rotation and lateralization. Dr. Zuckerman, do you have any concerns regarding complications in RSAs compared to TSAs?

Dr. Zuckerman: Complications can follow any surgical procedure, and I think our experience with the RSA over the last 10 years reflects a learning curve. My impression is that currently, the incidence of complications for RSAs is comparable to that after TSAs.

The risk of infection has been reduced significantly, as has the risk of instability. The most common long-term problem after anatomic shoulder replacement is rotator cuff failure, which tends not to be a problem after RSAs.

The issue of scapular notching—erosion of the inferior glenoid—appears to be implant- and technique-dependent. I think we now have enough implants choices and variations of surgical technique that even scapular notching and compromising the glenoid have been minimized.

Dr. Hawkins: In our series of more than 300 RSAs, we had a higher complication rate than with conventional TSA, although this has lessened with time and experience. Our main complications have been infection and dislocation.

Dr. Frankle: I think the main complication with RSAs is a postoperative periprosthetic fracture. These can be scapular or acromial, can occur many years afterward, and probably are related to a stress-induced problem.

In conventional TSA, my major concern is glenoid durability and loosening of the glenoid component, whether due to radiolucent lines underneath the glenoid cement interface or the glenoid implant interface.

Dr. Hawkins: Are complications treated differently in RSAs compared to TSAs? Let’s look at infections and dislocations.

Dr. Edwards: I treat them the same, depending on the timing of the infection and the patient. If the patient has an acute postoperative infection—whether conventional TSA or RSA—I will do a débridment change any changeable parts, and try to salvage the implant.

I’m more likely to treat a late or chronic infection with a staged replantation: Remove the implant, do serial débridement, treat the patient with an antibiotic spacer and a 6- to 8-week course of antibiotics, and insert another implant.

Dr. Zuckerman: Dislocations are less common now than they were initially. The first thing to determine is whether it is an acute or chronic dislocation because closed reduction may be used with an acute dislocation but not with a chronic situation.

The etiology of an unstable RSA must be carefully examined. It may be a soft tissue tension problem, a malposition problem, or an impingement problem involving heterotopic bone. In a fracture case, it’s possible that residual boney fragments are impinging the components and forcing them out. The treatment has to address the cause, so identifying the cause is important.

Dr. Hawkins: What can a patient expect in terms of range of motion and pain relief with a TSA versus an RSA?

Dr. Frankle: Patients with osteoarthritis (OA) who have TSAs will reliably have reduced pain and improved function from their preoperative status, and will clearly be satisfied.

Dr. Edwards: Patients who have RSAs will also have excellent pain relief and they should be able to obtain overhead function; depending on their preoperative motion, they will also experience a certain amount of improvement, between 140 degrees and 160 degrees of elevation.

But we have to emphasize what these patients can expect in external rotation. Some implants are designed to give better external rotation than others, but the primary predictor of active external rotation is the status of the posterior cuff.

Patients with some minor tearing in the infraspinatus or the teres minor can get a result that they will consider normal. Without any posterior cuff, a realistic expectation is that the shoulder is not going to move like a normal shoulder would.

Dr. Hawkins: What about functional outcomes after RSA? How soon can patients return to daily activities, sports, and so on?

Dr. Zuckerman: People need to use their upper extremity for daily activities. I caution them about lifting, particularly lifting with the arm away from the body, which really stresses the shoulder. The weight limit should be no more than 20 pounds. I will let patients play tennis to a limited extent. Golf is fine, although I think swings are inhibited with the implant.

Dr. Frankle: I try to individualize patient expectations and suggest that they think about prioritizing the activities they like the most. A lot of people in my area love to golf, so I let them go back to chipping and putting at around
6 weeks and the realistic expectation is to get back on the golf course by 3 months.

Dr. Hawkins: What about the patient who needs to use a walker or wheelchair? Is that patient a candidate for an RSA or a TSA?

Dr. Edwards: I think the RSA is a more reliable and predictable implant to use in that scenario. When I do a TSA and repair the subscapularis, I do not want that patient pushing or pulling up on the shoulder for a minimum of 2 to 3 months.

With an RSA, the risk of instability exists if the patient puts the arm in the wrong position when applying force on a walker. My experience has been that the patient can put some weight on it, particularly if the walker is fitted with a platform on the affected side.

Dr. Zuckerman: That’s an important consideration about patients who use a walker or cane. I think that use of a walker, cane, or some other device on the operative side is a contraindication for shoulder arthroplasty unless the patient has a significant amount of help and care afterward. That should be an important discussion.

Dr. Hawkins: Are there any other issues we should mention?

Dr. Frankle: The cost implications—RSA implants cost more. When we are faced with choices based on revenues, our responsibilities are more complicated. I’m not sure I have the answers, but I do think that the choice should be based on the implant that gives the most success. As we go forward, we need to track the cost in relation to other benefits for our patients.

Dr. Zuckerman: U.S. healthcare economics don’t take the long view when it comes to costs. Unfortunately, the cost of the RSA implant for a Medicare patient may equal or exceed the hospital’s reimbursement. In the long run, benefits or indirect costs may be lessened with an RSA, but that’s not how health care is handled. Frankly, I’ve been surprised that institutions have not pushed back about the exploding use of the RSA. We need studies to show the benefits of the RSA particularly when it can be used with a minimal complication rate for the patient’s benefit.

Dr. Hawkins: The RSA concept has been a great addition to our armamentarium and surgeons are getting better at selecting patients appropriately and improving outcomes. Understanding patient selection is important, and I hope this discussion will help readers in making those decisions.

Disclosures: Dr. Hawkins—Ossur, DJ Orthopedics, Wolters Kluwer Health – Lippincott Williams & Wilkins. Dr. Edwards—Tornier, Orthohelix, Shoulder Options, Kinamed, Journal of Shoulder and Elbow Surgery, Saunders/Mosby-Elsevier, Orthopedics. Dr. Frankle—DJ Orthopaedics, Tornier, Biomimetic. Dr. Zuckerman—Exatech, Inc., Hip Innovation Technology, Neostem, Orthonet, SLACK Incorporated, Wolters Kluwer Health – Lippincott Williams & Wilkins