5.6 Elbow arthroscopy

Burman reported a cadaveric study of elbow arthroscopy as early as 1932, but it was not until recent times that the technique became popular for treating elbow disorders. Advances in instrumentation as well as arthroscopic skills are the most probable reasons for the increase in use of the arthroscope in and around the elbow joint.

Elbow arthroscopy remains a technically challenging procedure. The reported complication rate is ten times higher than that of shoulder or knee surgery and careful attention to detail is required to help avoid iatrogenic neurovascular complications. The purpose of this chapter is to explore the current indications and contraindications for the procedure, give a description of preoperative planning, patient positioning, and the portals available, and review the results reported in the literature.

The original indications for elbow arthroscopy were for diagnosis and for removal of loose bodies. As surgeons became more proficient in this procedure, however, the range of pathology dealt with by arthroscopy has expanded. An example of its application is the osteoarthritic elbow where arthroscopy can be used after non-operative treatment has failed, as an intermediate step before arthroplasty. Indication is closely allied to experience and the indications have been related to five stages of surgical experience by Savoie as below:

Debridement of the arthritic elbow and removal of loose bodies remain the most common indications for elbow arthroscopy, but simultaneous debridement of impinging osteophytes (Figure 5.6.1) and release of capsular contracture can also be performed. In fact, a three-dimensional computed tomography scan prior to the operation in order to identify all potential osteophytes needing debridement has been recommended.

Debridement of ECRB in the treatment of lateral epicondylitis (tennis elbow) can be effectively performed arthroscopically.

Instability, osteochondritis dissecans, and fractures are more recent additions to the indications of elbow arthroscopy.

If debridement/synovectomy is entertained in the deformed rheumatoid elbow, this should be performed by an experienced arthroscopist as the risk of neurovascular injury is particularly high.

With regards to the paediatric population, Micheli et al. (2001) have reported a series of 47 patients aged 3.5–17 years who underwent elbow arthroscopy for osteochondritis dissecans, arthrofibrosis and joint contracture, synovitis, acute trauma, and posterior olecranon impingement syndrome. Using a modified Andrews elbow scoring system (MAESS), 85% of these patients had a good to excellent result with 90% returning to sports free of any limitations. They concluded that elbow arthroscopy was both safe and effective in the paediatric and adolescent population as long as there was a careful selection process and the arthroscopist was experienced.

Disordered anatomy caused by trauma, arthritis (including rheumatoid disease), previous surgery (especially ulnar nerve transposition), or an overlying cellulitis may be regarded as contraindications. Congenital conditions may be a relative contraindication. Biceps tendon injury and isolated radial tunnel syndrome are also relative contraindications.

On the day of the operation the patient is seen on the ward and placed in the position for surgery. This is because the skin around the elbow is very mobile and will alter in position with changes in patient position. The medial and lateral epicondyles, radial head, subcutaneous olecranon tip, ulnar nerve, and area of interest are then marked with a permanent marker. Whilst identifying these structures the authors go through the procedure and the structures at risk, to reinforce informed consent.

General anaesthesia allows full muscle relaxation and enables the patient to be positioned as required without discomfort (especially if the patient is positioned in the lateral decubitus or prone position). We routinely use additional regional blocks for postoperative pain relief although these do not allow for immediate postoperative exclusion of nerve injury.

Original description: Andrews and Carson (1985).

Positioning: the patient is positioned supine at the edge of the operating table. The affected limb is suspended by overhead traction of 5–10lb (2.25–4.5kg) (e.g. ‘Chinese finger traps’). The shoulder is abducted to 90 degrees and the elbow flexed to 90 degrees.

Advantages: since the elbow is held in an anatomic position, orientation of the structures is made easier for the arthroscopist. Excellent visualization of the anterior joint can be achieved. This is particularly useful for fixation of anterior radial head or coronoid fractures. Importantly, it also allows for maximal airway exposure if there are any anaesthetic concerns.

Disadvantages: it is more difficult to access the posterior compartment. In addition, an assistant is required to prevent the arm from swinging when instrumented. This can be partly overcome by modifying the position—having the shoulder flexed to 90 degrees thereby suspending the forearm over the chest. A secondary advantage of this modification is that the anterior neurovascular structures are pulled away from the working area by the force of gravity.

Original description: Poehling et al. (1989).

Positioning: the patient is positioned prone with the shoulder abducted 90 degrees. The upper arm is placed in a holder allowing 90 degrees of elbow flexion.

Advantages: gravity pulls the neurovascular structures further anteriorly, away from the working area. Full extension to near full flexion of the arm can be produced. The entire elbow is readily accessible.

Disadvantages: patient needs intubation for airway protection.

Original description: O’Driscoll and Morrey (1992).

Positioning: the patient is placed in the lateral position with the affected side up. The upper arm rests on a padded support. The table is tilted 20 degrees towards the operating surgeon to avoid potential antecubital fossa compression and minimize patient roll-back (Figures 5.6.2).

Advantages: the position allows excellent access to the anterior and posterior aspects of the elbow without some of the disadvantages of the positions mentioned earlier, particularly regarding anaesthesia implications.

Disadvantages: access to the medial side is limited but can be improved by positioning the elbow higher than the shoulder.

Of all the joints, arthroscopy of the elbow arguably carries the greatest risk of iatrogenic neurovascular damage. It is therefore essential to have a thorough knowledge of the anatomy of the available portals (Box 5.6.2) and their precise locations as margins for error are very small. The distance to the nerves from each portal has been elegantly documented by a number of cadaveric studies.

We regard it as essential to distend the elbow prior to making the entry portal. This is achieved by using 20–30mL of arthroscopic fluid (Figure 5.6.3A). Capsular distention pushes the neurovascular structures further away from the initial portal site and allows for easier entry into the joint. The site of injection is the centre of a triangle formed by the lateral epicondyle, subcutaneous olecranon tip, and radial head. This site is known as the ‘soft spot’ or infracondylar recess.

In establishing a portal, care must also be taken to minimize the risk of injury to the superficial cutaneous sensory nerves. Only the skin is incised, with a small blade and then a haemostat is used for blunt dissection in the pericapsular tissue. A blunt arthroscopic trochar and cannula are next inserted aiming towards the subcutaneous olecranon tip. Saline issuing from the cannula confirms intra-articular placement.

The authors begin by establishing the anterolateral portal followed by the anteromedial portal using a within-out technique. However, some prefer to begin with the anteromedial portal.

Steinmann (2003) has given an excellent account of the described portals in his review of elbow arthroscopy (Figures 5.6.3B,C).

Due to an unacceptably high risk of injury to the ulnar nerve, there are no posteromedial portals.

Position: 3cm distal, 1cm anterior to lateral epicondyle. The trochar is aimed towards the centre of the elbow joint and passes through ECRB and supinator. Clearly the precise measurement will depend on body habitas and will need to be modified appropriately.

Structures at risk: radial nerve (1.4mm away on average) and posterior antebrachial nerve (7.6mm away on average).

Function: passing all instruments through this port minimizes the risk of injury to the radial nerve. Used to scope the anterior trochlea, coronoid fossa, coronoid process, proximal radioulnar joint, and radiocapitellar joint.

Position: 2cm proximal, 1cm anterior to lateral epicondyle.

Structures at risk: radial nerve (9.9mm) and posterior antebrachial nerve (6.1mm).

Function: preferred by some surgeons to the anterolateral portal due to its greater distance from the radial nerve. However, scoping the coronoid fossa or medial trochlea may be more difficult.

Position: the centre of a triangle formed by the lateral condyle, subcutaneous olecranon tip, and radial head (the soft spot).

Structures at risk: lateral antebrachial nerve (between 6–16mm in 90 degrees elbow flexion and mid pronation).

Function: scoping the intra-articular olecranon tip and fossa, posterior trochlea, olecranon articular surface, posterior capitellum, radial head, and radioulna articulation. May be of particular use in osteochondritis dissecans of the capitellum and for the removal of loose bodies.

Position: 2cm distal, 2cm anterior to medial epicondyle. The trochar is aimed at the centre of the elbow joint and passes through the brachialis and common flexor origin.

Structures at risk: medial antebrachial cutaneous nerve (1mm away). Median nerve 7–14mm away with the elbow flexed at 90 degrees. The ulnar nerve is, on average, 20mm away in 90 degrees of flexion but it should be palpated prior to making the portal, checking for any possible subluxation from the cubital tunnel. Needless to say, particular caution is due in the case of a prior ulnar nerve transposition—in which case the surgeon would be advised to avoid this portal.

Function: good visualization of the anterolateral capsule, radiocapitellar, and ulnohumeral joints. Useful as an accessory portal when working in the medial gutter.

Position: 2cm proximal to the medial epicondyle, immediately anterior to the intermuscular septum. The trochar is aimed towards the radial head gliding across the anterior distal humerus.

Structures at risk: medial antebrachial cutaneous nerve (2.3mm average), ulnar nerve (12mm average), and median nerve (12.4mm average). The same considerations regarding the ulnar nerve apply with this portal as described earlier.

Function: good visualization of the radial head, lateral capsule, and coronoid (Figure 5.6.4) but poor visualization of the radiocapitellar joint and radial fossa. Useful as a retractor portal.

Position: lateral joint line level with the tip of the olecranon with the elbow in 90 degrees elbow flexion. Trochar aimed at centre of olecranon fossa.

Structures at risk: average distances at 90 degrees of flexion—medial brachial cutaneous nerve 22mm, posterior antebrachial cutaneous nerve 22mm, and ulnar nerve 26mm. the ulnar nerve is palpated before portal placement.

Function: excellent visualization of the posterior elbow (Figure 5.6.5) including the olecranon fossa and tip, medial and lateral gutters, and the posterior radiocapitellar joint where loose bodies are frequently missed.

Position: 3cm proximal to the tip of the olecranon. Sometimes a scalpel may be needed to penetrate the elbow joint due to the bulk of triceps. Trocar aimed at centre of olecranon fossa.

Structures at risk: average distances at 90 degrees flexion:—medial brachial cutaneous nerve 15mm, posterior antebrachial cutaneous nerve 30mm, and ulnar nerve 20mm. The ulnar nerve is palpated before portal placement.

Function: debridement of the posterior elbow joint. As an accessory portal to the posterolateral portal.

Position: 2cm proximal to direct posterior portal.

Function: a Howarth elevator can be passed to elevate the joint capsule allowing better visualization of the olecranon fossa.

The author (S.M.H.) also uses an accessory lateral portal positioned on the radiocapitellar joint line directly between the anterolateral and the direct lateral portals. It has proven very valuable for debridement of osteochondritic lesions and removal of posteriorly placed loose bodies.

Adams et al. (2008) have recently published their results of arthroscopic osteophyte resection and capsulectomy. Forty-one patients with primary osteoarthritis in 42 elbows showed significant improvements in mean flexion, extension, supination, and Mayo Elbow Performance Index scores with 81% good to excellent results. Pain decreased significantly. Complications were rare and included heterotopic ossification and ulnar dysesthesias.

Cohen et al. (2000) have compared the results of arthroscopic debridement with an open Outerbridge–Kashiwagi (OK) procedure (Figure 5.6.6). Overall they found no difference between the two procedures although arthroscopic intervention provided better pain relief, whereas open debridement resulted in an improved flexion range.

Savoie et al. (1999) have reported a series of 24 patients with osteoarthritis treated with arthroscopic debridement, partial resection of the coronoid and olecranon processes, and fenestration of the olecranon fossa. Eighteen of the 24 patients had arthroscopic radial head excision. They report significant pain reduction and an improvement of average arc of motion of 81 degrees. The paper records a complication rate comparable to open ulnohumeral arthroplasty.

Kelly et al. (2007) have reported a series of 24 patients with elbow joint osteoarthritis treated with arthroscopic ulnohumeral arthroplasty, leaving the radial head intact. Debridement of the radial head, anterior and posterior osteophytes, and capsular release alone was performed. Their results showed 14 patients with an excellent result, seven good, three fair, and one poor. The average flexion–extension arc improved by 21 degrees. Twelve patients had no limitations in their daily activities, and 12 experienced occasional problems. No surgical complications were reported. They concluded that resecting an arthritic radial head was not essential.

The rheumatoid elbow presents a particular challenge to the elbow arthroscopist mainly caused by deformity, obscured vision, and synovial hypertrophy (Figure 5.6.7).

Lee et al. (1997) have reported a series of 14 elbows in 11 patients with good 3–4-year results (93% excellent/good) but with subsequent rapid deterioration.

Horiuchi et al. (2002) have reported results of arthroscopic synovectomy in 29 elbows (27 patients). They graded preoperative radiographs from 1–4 using the system of Larsen et al. (1977) forming three groups: 1/2, 3, and 4. They found only elbows with Larsen grade-1 or 2 arthritis had a favourable long-term result with regard to total function. The postoperative results were unsatisfactory for Larsen grade-4 elbows.

Nerve injury is the predominant concern when performing arthroscopic release of arthrofibrosis. Phillips et al. (1998) reported a series of 25 patients with elbow arthrofibrosis treated with arthroscopic debridement. At 18 months, all patients demonstrated increased motion and decreased pain, although one patient had a reoperation due to continued symptoms. They found that patients with post-traumatic arthritis tended to have greater contractures preoperatively, compared to those patients with degenerative arthritis, but had greater improvement postoperatively. This series reported no perioperative or postoperative complications and concluded that arthroscopic release and debridement of arthrofibrotic elbow joints achieved equal improvement to that of open techniques, with less morbidity and earlier rehabilitation.

Jones et al. (1993) have reported a series of 12 patients with good results, although one patient had residual posterior interosseous nerve palsy.

Menth-Chiari et al. (2001) have reported a series of 12 patients who underwent arthroscopic excision of the radial head either due to post-traumatic arthritis after fracture of the radial head, or due to rheumatoid arthritis. All except one patient had significant pain relief and complete relief of mechanical symptoms. There were no infections or neurovascular injuries reported.

Owens et al. (2001) have reported a series of 16 patients with recalcitrant tennis elbow who underwent arthroscopic surgical debridement of ECRB. There was only 75% follow-up data due to patients’ military reassignment, but the group found that all patients had an improvement with an average return to unrestricted work at 6 days.

Peart et al. (2004) report a series of 87 patients, 54 treated with an open and 33 with an arthroscopic procedure. No significant difference in outcome was noted although an arthroscopic release led to an earlier return to work and required less postoperative therapy.

A systematic review of the literature found similar results in endoscopic, percutaneous, and open procedures.

Rahusen et al. (2006) reported a series of 15 patients with osteoarthritis dissecans treated with arthroscopic debridement (Figure 5.6.8). Their results showed no significant improvement in range of motion, but use of the MAESS showed a significant improvement from 65.5% (poor) to 90.8% (excellent). There was a reduction in pain level and the average return to work occurred 3 months postoperatively.

The role of drilling or microfracture of the capitellum remains unproven.

There are a number of reports of novel use of the arthroscope in managing conditions of the elbow such as osteoid osteoma and treatment of posterolateral elbow impingement in professional boxers^. The treatment of throwers’ elbow/plica has also seen an increasing role of the arthroscope, and the author (S.M.H.) reports arthroscopic removal of a silastic radial head replacement. These are indications of the developing nature of the technique.

Kelly et al. (2001) have reported the complications of elbow arthroscopy in 473 patients. Four (0.8%) had a complication classified as major (septic arthritis). Minor complications occurred in 11%, which included prolonged drainage from, or superficial infection at, a portal site after 33 procedures, 12 transient nerve palsies (five ulnar, five superficial radial, one each of posterior interosseous, medial antebrachial cutaneous, and one anterior interosseous palsy). Rheumatoid arthritis followed by contracture was the most significant risk factor for temporary nerve palsy.

This study found no permanent nerve injuries. It also reports the findings of The Arthroscopy Association of North America who conducted two separate surveys. Of the 1648 elbow arthroscopies in both surveys, only one nerve injury was documented.

The authors of this comprehensive study of complications suggest that the use of retractors was the single most important technical step in preventing serious nerve injuries especially when performing synovectomy or capsulectomy. The second factor was arthroscopic/open identification of the nerve when performing a capsulectomy around nerves.

With regards to prolonged portal drainage, Kelly and colleagues found that all were from lateral portals and recommended the use of a locked horizontal mattress suture to minimize the risk of this complication.

As with many areas of arthroscopic surgery, the technique of elbow arthroscopy, its indications, and its success has evolved considerably in recent years. However, it is a technically demanding and a dangerous procedure for which a detailed working knowledge of the surrounding neurovascular anatomy is essential. It remains the preserve of the experienced arthroscopist and we strongly concur with Savoie (2007) that graduated progression where indication is titrated against experience remains the mandatory approach.

The authors are very grateful to Alun Jones and Andy Biggs at Robert Jones and Agnes Hunt Hospital, Oswestry, for their help in producing the illustrations.