Commentary & Perspective on
"Shoulder Arthroplasty for Arthritis After Instability Surgery"
by John W. Sperling, MD, et al.

Commentary & Perspective by
Louis U. Bigliani, MD*,
New York-Presbyterian Hospital, New York, NY

First and foremost, this study clearly demonstrates that the results of shoulder arthroplasty following instability surgery are not as good as those in patients with primary osteoarthritis. Arthroplasty after instability surgery is more difficult due to soft-tissue contractures, posterior subluxation, and loss of glenoid bone. Sperling et al have shown that the results of shoulder arthroplasty in this patient population are not as consistent or as predictable in regard to pain, range of motion, and component survival. At our institution, a similar group of patients who had shoulder arthroplasty after instability surgery have had results that confirm those of this study.

Of concern is the fact that these arthroplasties were performed in a relatively young group of patients. Also, these were active, predominantly male, patients who would probably be involved in a high level of physical activity and, possibly, heavy labor. Therefore, the total success rate of only 45% (fourteen of thirty-one patients had excellent or satisfactory results), is worrisome. Furthermore, eleven of the thirty-one patients (35%) in this series underwent revision surgery, and ten (32%) had persistent pain. These results highlight the fact that successful treatment of this patient population is extremely difficult and challenging. In this group of younger individuals, a perfect arthroplasty to achieve the longest possible survival of the implant would be ideal. It is apparent that soft-tissue balancing and proper implant selection are crucial in patients who have had previous instability surgery, as many had postoperative instability and component failure.

It is important to emphasize that instability surgery does adversely affect the subscapularis muscle. In this series, there were a variety of different surgical procedures that had been performed for the treatment of shoulder instability. In our experience, the most difficult shoulder arthroplasties involve patients who have had a previous Bristow procedure or a Putti-Platt procedure. In these cases, the subscapularis is shortened or severely contracted, making it very difficult to achieve adequate length of the subscapularis tendon to allow attachment to the lesser tuberosity and provide adequate external rotation. Although it was not mentioned by Sperling et al. in this article, it would be helpful to know how many patients in their series had subscapularis insufficiency, which may have contributed to the failures.

A constricted and shortened subscapularis also leads to reduction in external rotation, which, in turn, leads to posterior subluxation and bone loss in the posterior part of the glenoid. Thus, a vicious cycle occurs, which makes for a much more difficult surgical reconstruction, and hence, less predictably excellent results. In this series, it was noted that the patients had an average external rotation of only 4° preoperatively, which is much less than the external rotation in the average patient with osteoarthritis. Also, several patients had severe glenoid erosion and bone loss, which makes positioning of the glenoid component extremely difficult.

The rate of revision surgery in this group is high (35%), much higher than that in patients who have undergone arthroplasty for the treatment of primary osteoarthritis. This high revision rate is due to inadequate soft tissue, as well as inadequate bone stock, in patients who have had instability surgery. Therefore, these patients must be evaluated preoperatively with use of an axillary radiograph and either a CAT (computerized axial tomography) scan or MRI (magnetic resonance imaging). These studies enable a more adequate evaluation of the glenoid vault in terms of its depth and erosion. If surgery is attempted, techniques to lengthen the subscapularis adequately must be considered, as this will definitely be needed to achieve soft-tissue balancing.

There were differences in the results between the two treatment groups in this series. The rate of satisfactory results in the hemiarthroplasty group was superior to that in the total shoulder arthroplasty group—60% and 38%, respectively. However, it appears that some of the failures in the total shoulder cohort were attributable to failure of an ingrowth glenoid component. It is important also to emphasize that 30% of the patients in the hemiarthroplasty group had revision to a total shoulder arthroplasty because of pain due to glenoid arthritis. In such a small group of patients, this rate is too high. Also, there was an unacceptably high rate of revision of the glenoid components.

We prefer to use a glenoid replacement because it centers the humeral head and allows for better soft-tissue balancing. We have observed high rates of subluxation, glenoid arthritis, and revision in patients who have had a hemiarthroplasty for the treatment of glenohumeral arthritis after instability surgery. In this study by Sperling et al., 50% of the patients in the hemiarthroplasty group had glenohumeral subluxation.

There was also a high rate of humeral periprosthetic radiolucency in this study. Many of the humeral prostheses were not cemented but were press-fit, and some had a bone-ingrowth surface. It appears that cemented fixation should be considered for this group of patients in whom soft-tissue balancing cannot be ideal.

The main message of this article by Sperling et al. is that in this subgroup of patients with glenohumeral arthritis—namely, those who have had previous instability surgeries, surgical reconstruction is extremely difficult, which leads to a higher failure rate and early revision surgery to achieve an excellent range of motion and pain relief.

*The author did not receive grants or outside funding in support of the research or preparation of this manuscript. He did not receive payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the author is affiliated or associated.