The Arthroscopic Treatment of Glenohumeral Osteoarthritis

INTRODUCTION

Arthroscopic treatment of glenohumeral osteoarthritis offers certain advantages to the patient and surgeon. The foremost of these is the possibility of symptom relief with shoulder joint preservation. In many ways, prosthetic arthroplasty of the hip, knee, and shoulder has been so successful that one may wonder why joint-preserving surgery for shoulder arthritis would even be considered in the first place. This point of view neglects the many serious disadvantages of prosthetic shoulder arthroplasty. These disadvantages, especially critical to avoid in young patients, include limited prosthetic longevity, risk of reoperation or major complication (especially prosthetic joint infection), lifetime activity restrictions, and incomplete symptom relief.

On the other hand, arthroscopic shoulder joint preservation aims to provide symptom relief while minimizing complications, returning patients to activities without restrictions, and lowering costs. It should be noted that this chapter pertains only to all arthroscopic techniques for shoulder joint preservation. Specifically, we have not found a role in our practices for prosthetic humeral hemiarthroplasty with glenoid resurfacing, commonly confused with arthroscopic glenoid resurfacing (AGR). Hemiarthroplasty with glenoid resurfacing is joint sacrificing, has been found to be unpredictable,¹ and also does not offer the low complication rate seen with all AGR.

Arthroscopic joint preservation for shoulder arthritis also presents its own set of conundra. These may be grouped into three broad categories:

Patient Selection: Which patients benefit from jointpreserving techniques?
Indications: Which procedure(s) should be performed for a given patient?
Technical Issues: How can these operations be effectively and safely performed?

PATIENT SELECTION CONUNDRA

Patient selection likely is the most important factor for success in the arthroscopic treatment of glenohumeral arthritis. Patient and disease characteristics should be factored into the surgeon’s decision-making (Fig. 7-1). It goes without saying that chronologic age is relative, and the surgeon should take into account the relative physiologic age of the patient as much as possible. Additionally, as the average lifespan in developed countries continues to increase, “young” is getting older. Thus, the need for successful shoulder joint preserving techniques will only continue to increase.

The ideal patient for a prosthetic shoulder arthroplasty is elderly (e.g., ˜70 years of age and older) and lowdemand. In this demographic, if severe arthritis has failed nonoperative treatment, we offer the patient a prosthetic shoulder arthroplasty. On the other hand, milder arthritis in this age group can certainly be effectively treated with joint preservation, especially if stiffness is the primary complaint instead of pain.

Conversely, the ideal patient for a joint preserving procedure is young (fourth to fifth decade and younger) and
high-demand (e.g., weight lifters, laborers). In this age group, the rates of complications and revisions after prosthetic arthroplasty warrant consideration for a nonprosthetic surgical option. On the other hand, we consider joint preservation to be relatively contraindicated, even for young patients, when arthritis has resulted in severe glenoid or humeral bone loss.

FIGURE 7-1 Patient and disease characteristics are factored into the decision-making process for shoulder arthritis. Good candidates for prosthetic arthroplasty (yellow) are older, are lower-demand, and have more severe arthritis. Good candidates for joint preservation surgery are young and highdemand. When disease severity is high, our experience is that AGR has a better chance of relieving symptoms versus debridement alone for joint preservation.

Finally, the surgeon must account for and manage patient expectations. Some patients have negative perceptions of prosthetic joint replacement and may specifically seek an alternative to “joint replacement.” For patients in this category who might otherwise be “ideal” candidates for a prosthetic arthroplasty, we provide counseling about arthroscopic joint preservation. Given the low complication rate, we feel that it is reasonable to offer an arthroscopic option to highly motivated and well-informed patients who might not be “ideal” candidates for these operations.

INDICATION CONUNDRA

Once joint preservation has been decided upon, which arthroscopic procedure(s) should be performed? First, we make the preoperative decision for or against allograft glenoid resurfacing. Several authors have reported that the severity of clinical and radiographic arthritis is associated with inferior results with arthroscopic debridement procedures alone.² Therefore, we offer AGR when there are large osteophytes, no joint space remains on radiographs (typically axillary), and symptoms are reported as severe (e.g., ASES < 40). If nonoperative treatment has failed but symptoms are less severe (e.g., primarily stiffness) or for older patients, we will offer arthroscopic treatment without glenoid resurfacing. In this case, the primary operation involves an arthroscopic capsular release and manipulation under anesthesia with joint debridement.

Several additional arthroscopic procedures might be added to the primary procedure. These include biceps tenodesis versus tenotomy, subacromial decompression, and distal clavicle excision. We typically make a decision about the biceps intraoperatively. In the typical young and active patient, tenodesis is preferred over tenotomy, and we perform tenodesis for a SLAP lesion, biceps instability, or severe tendinopathy. We also usually make a decision about decompression after diagnostic arthroscopy of the subacromial space. Decision-making about the AC joint is discussed in detail in Chapter 6, but we usually rely on physical examination to make a preoperative decision about whether to perform distal clavicle excision.

Other arthroscopic procedures, specifically axillary neurolysis and inferior osteophyte debridement, have been proposed to be beneficial in treating shoulder arthritis with arthroscopy.² We have not found these to be necessary, and we consider the risk to the axillary nerve with these procedures to be too high to perform them routinely. Currently, we remove inferior osteophytes only as is necessary to perform the intended procedures, and this is uncommonly necessary. Even though we do not routinely debride inferior humeral osteophytes, we often debride and contour anterior and posterior humeral osteophytes that block internal and external rotation. Such debridement can restore glenohumeral congruency through a wider range of motion, thereby increasing the arc of rotation.

TECHNICAL CONUNDRA

The arthroscopic treatment of glenohumeral arthritis can be quite technically challenging since joint stiffness and osteophytes can restrict the surgeon’s ability to place instruments and implants into the joint. We have found several technical considerations to be of utmost importance in safely and efficiently performing these operations.

Setup

We consider the lateral decubitus position to be key when performing complex shoulder arthroscopy. The anesthesiologist will typically be much more comfortable with hypotensive anesthesia in this position, thus decreasing bleeding and allowing excellent visualization at moderate pump pressures (˜60 mm Hg). Lateral position also allows a second assistant, which we prefer over a static arm positioning device, to manipulate the arm during surgery. Specifically, the second assistant can adduct the arm over a bump to distract the joint space (Fig. 7-2).

Portals

Correct portal placement is critically important in performing shoulder arthroscopy in the arthritic shoulder. For all but the posterior portal, which is placed blindly, a spinal needle should be placed first prior to making the portal
incision to verify the correct angle of approach. Don’t remove the needle until the skin incision has been made and an instrument placed along the correct pathway. A detailed description of our portal locations can be found in the first volume in this series.³ In general, we find that our portals are placed much “lower” than is typically taught. For glenoid resurfacing, the surgeon must be comfortable in placing a posterolateral portal (Fig. 7-3) for posteroinferior glenoid anchor placement.

FIGURE 7-2 The working space in the glenohumeral joint during AGR can be greatly improved with the help of a skilled assistant manipulating the arm over a large bump.

Working with Allograft

Our preferred graft for glenoid resurfacing is ArthroFlex, (Arthrex, Naples, FL), which has been shown to have excellent mechanical properties and biocompatibility.⁴ We use the thickest graft currently available (3 mm) for glenoid resurfacing. We have found that a thicker graft has been associated with a lower revision rate in our practice of about 10% compared with the historical reported revision rate of about 20% for this operation.⁵,⁶

FIGURE 7-3 Clinical photograph showing a left shoulder during graft shuttling. At least four portals are required for AGR. A posterolateral portal (arrow) is often made percutaneously using a spinal needle for guidance and used for posterior anchor placement.

A thick graft poses some technical challenges, specifically in shuttling the graft and passing sutures. We have found that a thick graft is most easily shuttled through an anterosuperolateral portal without a cannula. The basic principles behind our standard technique are illustrated in Figure 7-4 and are expanded in great detail in the cases that follow.

Shuttling through alternative portals with or without a cannula (outer [blue] sheath of Gemini cannula or 10-mm PassPort cannula, Arthrex, Naples, FL) is also possible (“Bull’s-Eye” case).

We have found that suture passing is simplified by using the FastPass Scorpion suture passer (Arthrex, Naples, FL), taking care to change the Scorpion needle if it becomes dull or bent near its tip (typically after about six to eight suture passes through the graft).

History:

A 56-year-old male with pain and increasing stiffness, right shoulder, for the past 2 years. No specific injury, but he has been an avid weight lifter in the past.

Physical Exam:

Range of motion as follows: Elevation 120° (active and passive) External rotation 20° (active and passive) Internal rotation to the sacroiliac (SI) joint (active and passive)
Pain at the extremes of motion
Normal strength

Imaging:

X-rays show the mild-to-moderate degenerative arthritis. There is a small inferior humeral osteophyte. The glenohumeral joint space is approximately one-third the width of the normal joint space on the opposite side. There are no erosive changes of the glenoid.

FIGURE 7-4 Graft shuttling and initial fixation are technically challenging in AGR. Our standard technique is to place two to three inferior anchors through anterior and posterolateral portals. Next, one limb from each suture is retrieved out the anterosuperior portal, passed through the graft, and mulberry knots tied on the joint side of the graft (A). The other limb of each suture is then used to pull on the graft and shuttle it into the joint (B). These sutures can then be tied to gain initial fixation of the graft inferiorly (C).

Arthroscopic Findings:

Areas of grade III to IV chondromalacia on the glenoid and humeral head.
There is no posterior erosion of the glenoid.
Arthroscopic joint debridement and capsular release were performed (Fig. 7-5).
At 6 months post-op, the patient’s pain and range of motion were greatly improved (elevation 150°; external rotation 45°; internal rotation to the level of L2).

Pearls, Pitfalls, and Decision-Making:

This patient is in an age group that is not ideal for arthroplasty; his range of motion is markedly decreased; and he has no posterior erosion of the glenoid. This makes him an ideal candidate for arthroscopic capsular release. In patients with degenerative arthritis and motion loss, there are very high loads at the extremes
of motion due to eccentric edge loading at end-range motion. Arthroscopic capsular release provides a greater range of motion before edge loading occurs, thereby providing a greater pain-free arc of motion.

FIGURE 7-5 Right shoulder, anterosuperolateral viewing portal. Posterior capsular release is being performed with a pencil-tip electrode. H, humeral head; P, posterior capsule.

History:

A 63-year-old right hand-dominant man with shoulder pain for many years. He had noticed progressive increase in stiffness over the past few months.
He was unable to throw a ball with his grandsons, which was his primary complaint. The pain was intermittent and mild to moderate.
He stated that he could live with the pain if he could have better motion in the shoulder.

FIGURE 7-6 Preoperative x-rays show moderate arthritis (A) with a B2 glenoid (B) and a thick, type 2 acromion (C).

Physical Exam:

Active (AROM) and passive range of motion (PROM) of the right shoulder were 170° forward elevation, 40° external rotation with the arm at the side and abducted, internal rotation to the SI joint. Unable to reach across the body to the opposite shoulder.
5/5 strength in abduction, IR, ER with negative bear hug and belly press tests.

Imaging:

X-rays showed moderate glenohumeral arthritis with some preserved joint space on Grashey views but a biconcave glenoid with complete loss of joint space on the axillary (Fig. 7-6). The scapular Y view shows a thick, type 2 acromion.
MRI showed rotator cuff and long head of biceps tendinosis with glenohumeral arthritis and degenerative labral tearing (Fig. 7-7).

FIGURE 7-7 Coronal (A) and axial (B) MRI images confirm that glenohumeral arthritis is present without a full-thickness rotator cuff tear.

Arthroscopy Findings:

Low-grade partial articular tear of the supraspinatus, intra-articular rice bodies and synovitis, extrinsic impingement of the rotator cuff with subacromial bursitis. Extensive degenerative labral tearing.

FIGURE 7-8 The typical pattern of posteroinferior cartilage wear (A) as viewed from an anterosuperior portal in this right shoulder. After arthroscopic capsular release (B), the glenohumeral joint has much more working space. G, glenoid; H, humeral head.

Pearls, Pitfalls, and Decision-Making:

In this case, the critical decision-making point was in deciding which operation to perform. Even though this patient had moderate radiographic arthritis and a biconcave glenoid, his main symptom was stiffness. This type of patient will often greatly benefit from an arthroscopic capsular release and debridement and avoids the potential complications of a prosthetic arthroplasty.
Given the posterior subluxation and glenoid wear (Fig. 7-8), one may question the decision to perform
release of the capsule. Isn’t this overstretched and patulous? In fact, the posterior capsule is also tight and contracted, and it is just preferentially tight anteriorly. Otherwise, there would be no restriction to internal rotation.
Decision-making intraoperatively included management of the biceps/labral complex, rotator cuff, and subacromial space. Although we commonly perform biceps tenodesis for biceps/labral pathology, we do not think tenotomy or tenodesis should be done without a specific purpose—the “biceps killer” mentality. In this patient, the biceps had a firm attachment to the supraglenoid tubercle, had no instability in the groove, and was causing no discernable symptoms to the patient. Therefore, capsular releases were done preserving the medial sling and the biceps was retained (Fig. 7-9).
We treated the small supraspinatus PASTA with debridement to a stable margin. In addition to the significant subacromial bursitis, the PASTA provides additional justification for performing subacromial decompression. Of course, a 63-year-old rotator cuff has significant “intrinsic pathology.” But mechanical extrinsic compression from the CA arch, as in this case, often contributes to even partial articular rotator cuff tears through the development of tensile forces on the opposite side of the tissue from the compressive force7. Bursectomy, excision of the coracoacromial ligament, and anterior acromioplasty to a type 1 acromion provide excellent treatment for this constellation of intraoperative findings (Fig. 7-10).

FIGURE 7-9 A posterior capsular release (A) performed through a posterior working portal is critical to allow the patient to have improved internal rotation. Care must be taken around the 4 to 5 o’clock portion (right shoulder) of the release (B) so that the axillary nerve is not injured. G, glenoid; H, humeral head; A, anterior capsule; P, posterior capsule.

FIGURE 7-10 Extrinsic compression of the rotator cuff is relieved after converting the type 2 to a type 1 acromion using a bone block cutting technique (blue line). The anterior acromioplasty should be finished from a posterior working portal while viewing from a lateral portal. SS, supraspinatus muscle; SP, scapular spine.

History:

A 48-year-old right hand-dominant heavy laborer (self-employed carpenter) presented with bilateral shoulder osteoarthritis (Fig. 7-11), with left severely painful and the right with minimal symptoms.
He underwent AGR of his left shoulder with dermal allograft. At 4 years postoperatively, his VAS pain in the left shoulder was 1 (of 10), he rated the shoulder 80% normal, and his ASES score was 85. He had been able to return to full duties at work.
In the interim, the right shoulder became progressively painful. Radiographs were relatively unchanged, but he was unable to work or do leisure activities, such as hunting, because of pain. Nonoperative treatment failed.
He requested to have AGR of the right shoulder. The patient’s insurance company did not pay for the dermal allograft at the first operation, and he stated that he was more than willing to pay for the graft again because he had such a good result with the first surgery.

FIGURE 7-11 Preoperative Grashey (A, right; B, left) and axillary (C, right; D, left) x-rays demonstrate glenohumeral arthritis with large osteophytes and an early B2 glenoid on the left side.

Physical Exam:

Left shoulder (4 years post-op AGR) AROM was to forward elevation 135°, external rotation 20°, internal rotation to the SI joint.
Right shoulder AROM was to forward elevation 100° (Fig. 7-12), external rotation 10°, and internal rotation to the SI joint.
Both shoulders had 5/5 strength in abduction, internal and external rotation; negative bear hug and belly press tests; no AC joint tenderness.
The left shoulder had relatively smooth motion without pain or crepitus, but the right shoulder had painful crepitus with active or passive motion.

FIGURE 7-12 Right shoulder preoperative forward elevation.

Imaging:

Preoperative and postoperative radiographs for both shoulders are shown in Figures 7-11 and 7-13. Both had minimal joint space, large inferior osteophytes, although the left had static posterior subluxation and the right did not.

FIGURE 7-13 Postoperative Grashey (A, right; B, left) and axillary (C, right; D, left) x-rays of bilateral AGR demonstrate maintained joint space with better centering of the left shoulder on the axillary view.

Arthroscopy Findings: