CHAPTER 42 Complications of Elbow Arthroscopy

INTRODUCTION

The potential for complications with elbow arthroscopy is as well known as that of any orthopedic technique. This may explain in part why elbow arthroscopy is still not widely performed by most orthopedic surgeons. Other reasons include (1) relative infrequency of elbow involvement and (2) difficulty of effectively and accurately performing arthroscopy in a very congruent joint, especially when the pathology alters the joint capacity.

INFORMATION AND CLINICAL EXPERIENCE

Although suffering from a deserved reputation for significant potential complications, there is still relatively little information in the orthopedic literature regarding the frequency of complications from elbow arthroscopy. With the exception of a Mayo report, what data exist pertain to a single or limited case reports often associated with an anatomic study.^1,^3,^6,¹⁹^–²²

We will consider this complication in the following order: (1) anatomy; (2) pathology being treated; (3) procedures available to the arthroscopist; (4) incidence of reported complications; (5) Mayo’s experience; and (6) recommendations.

The elbow is one of the most congruous joints in the body; hence, the ability to manipulate the joint to separate the articular surfaces and allow better visualization is extremely limited, and multiple portals may be needed.^2,^9,¹⁸ Furthermore, the capacity of the joint is limited in the normal situ-ation and is even further curtailed by most pathology. O’Driscoll and colleagues²⁴ demonstrated an average normal capacity of approximately 30 mL. Post-traumatic and degenerative processes result in contracture of the joint, often allowing less than 10 mL of intra-articular distension. The soft tissue envelope of the elbow is extremely thin, the capsule being separated from the skin by a thin layer of subcutaneous tissue in some locations. Thus, the tendency for the portals to “seal” is limited. This feature predisposes to chronic drainage and the possibility of infection.¹⁵

Neurovascular Structures

Without question, the greatest concern regarding elbow anatomy is the proximity of the radial and ulnar nerves that cross the joint in proximity to the capsule. The relationship of the radial, median, and ulnar nerves to the capsule in both the distended and the nondistended positions has been studied extensively.^1,^3,^6,¹⁹^–²² ^,30 ^,31 ^,33 Furthermore, the vulnerability of cutaneous nerves had also been studied in relationship to the arthroscopic portal sites. These data are summarized in Table 42-1. It is particularly important to note that distension of the joint does alter the relative location of cutaneous nerves as well as the radial and median nerves referable to the portal site. However, a distended joint in no way protects either of these nerves from an intra-articular procedure.²⁰ As a matter of fact, the distended capsule may theoretically render these nerves more, rather than less, vulnerable. The most vulnerable nerve anatomically is the posterior interosseous nerve.¹⁰ This nerve may typically be 5 to 10 mm from an anterolateral portal. However, there is significant variation, and in some instances. the nerve can be as close as 2 to 3 mm to the capsule because it lies over the radial neck. Similarly, the median nerve demonstrates a variation of approximately 5 mm between the distended and the nondistended capsule referable to the anteromedial portal. However, once again, the distended capsule approximates the nerve—it does not separate the nerve. Although it is clearly the most protected, median nerve injury has been reported.¹¹ Of great concern is that of a concurrent injury to the brachial artery or vein. Finally, the ulnar nerve actually rests on the medial capsule.³⁰ The greatest risk consists of procedures performed in the posteromedial corner of the elbow.¹² However, injury from portal placement has also been reported.⁷

TABLE 42-1 Nerve Distances to Capsule and Portals

In general, these nerves are not vulnerable to portal insertion if the portal sites are accurately defined and the joint is distended. The risk of nerve injury is as follows: least risk, the median nerve; moderate risk, the ulnar nerve; greatest risk, the radial nerve.

PATHOLOGY

The nature of the pathology influences potential risks of complications ²² (Table 42-2).

TABLE 42-2 Authors’ View of the Relative Risk of Complication by Diagnosis^*

Rheumatoid Arthritis

The most common reason for arthroscopy of an elbow involved with rheumatoid arthritis is synovectomy. The capsule is extremely thin in these patients, and therefore, the nerves are at significant risk of injury with this procedure. One patient developed a temporary ulnar nerve paresthesia owing to the instability associated with rheumatoid synovitis and the vulnerability of the nerve referable to a thin capsule covered by proliferative synovium.

Loose Body

The removal of loose bodies is still probably the best and most common indication for elbow arthroscopy in both the degenerative and the post-traumatic elbow.^4,^23,^25,²⁷ The complications are uncommon unless the portal site strays from the recommended positions or débridement is required. Both generally place the radial nerve at risk.

Degenerative Arthritis

Primary degenerative arthritis is a good indication for elbow arthroscopy.^4,¹⁴ The se-lection of a procedure may include removal of loose bodies, débridement of the coronoid and olecranon osteophytes, and anterior and posterior capsular release. The anterior capsular release places the median and, particularly, the radial nerve at risk. The posterior release may place the ulnar nerve at risk. In addition, the multiple portal sites necessary for effectively carrying out this procedure may further add an element of vulnerability to these nerves. In addition to the degenerative as well as the post-traumatic elbow, the capsular capacity is very limited,²⁴ making even a diagnostic procedure something of a risk particularly for articular scuffing. As confidence increases and a greater amount of bone is removed, the potential for ectopic bone has also been reported⁸ (Fig. 42-1).