Chapter 8: Arthroscopic Subacromial Decompression

Impingement

Our understanding of the impingement syn­drome has greatly improved over the past few years and there is no doubt that this will continue for the next decade. It is becoming apparent that impingement is just the final common pathway for a number of different pathologies. Perhaps the best way to look at impingement is anatomically and functionally.

The subacromial space is not a true space but a potential space. Thus it has a roof and a floor, but no real walls. This potential space is lined by the subacromial bursa. The subacro­mial space is roofed by the acromion, the acromioclavicular joint, the extreme distal por­tion of the clavicle, the coracoacromial ligament and, anteriorly, the coracoid process itself. The floor of the space is the rotator cuff, the coracohumeral ligament and the long head of biceps. Once distended, the potential space becomes a space: the anterior wall consists of the coracoid process, the lateral and posterior walls are formed by the deltoid and the medial wall is made of bands from the coracoacromial ligament to the acromioclavicular joint and the musculotendinous junction of supraspinatus (Figure 8.1).

Figure 8.1 Anatomy of the subacromial space.

The subacromial space can therefore be compromised either by the roof coming down, or the floor coming up. The roof can only come down by some part of it being thicker anatomically, but the floor can come up both anatomically (by a thickening of the floor), or functionally (by upward subluxation of the glenohumeral joint, which can be physiological or pathological).

A 'lower' roof

The acromion

The shape of the anterior acromion varies from person to person, and may be flat or hooked. The Bigliani classification recognizes three types of shape, varying from flat (type 1) to fully hooked (type 3). The best way to appreciate the acromial shape is to take a subacromial arch radiograph. The relevance of the Bigliani type 3 acromion is that the hook will have to be surgically excised.

The acromioclavicular joint

Figures vary but, on average, 30 per cent of patients with impingement will have degenerate changes visible in the acromioclavicular joint on an anteroposterior radiograph of the shoul­der. A proportion of these will have inferior osteophytes (Figure 8.2) projecting down into the subacromial space. The capsule of the acromioclavicular joint may also be thickened. Obviously, to remove the osteophytes, the capsule has to be resected first.

Figure 8.2 Inferior osteophytes projecting down from the acromioclavicular joint.

The coracoacromial ligament

We have already shown that the coracoacro­mial ligament inserts underneath the acromion and not onto the front of the acromion, so any thickening of this ligament at its insertion point will project into the subacromial space.

The coracoid process

Gerber et al[1] have described a variety of anterior impingement caused by the proximity of the coracoid to the front of the shoulder, which is relieved by the operation of coraco-plasty.

A 'higher' floor (functional)

Physiological

Matthews and Fadale[2] have shown how the subacromial space is diminished with abduc­tion of the arm. Sigholm et al,[3] using a microca-pillary infusion technique, have measured the pressure in the subacromial space. At rest, the pressure was 8mmHg, rising to 32mmHg with the arm abducted to 45 degrees and to 56 mmHg with the arm abducted to 45 degrees while holding a 1 kg weight. This 'physiological impingement' has relevance both to work-related and sport-related impingement.

Pathological

Upward subluxation of the glenohumeral joint may be caused by instability, or by an abnor­mality of muscular or nervous control of the rotator cuff. In particular, rotator cuff tears can lead to upward subluxation (Figure 8.3). Weakness of the rotator cuff, secondary to suprasca­pular nerve entrapment, may cause upward subluxation.

Figure 8.3 Upward subluxation of the humeral head following rotator cuff tear.

Obviously an understanding of the aetiology of impingement is vital to the correct manage­ment. For instance, if a particular patient has impingement secondary to instability, then the capsule should be repaired. Subacromial decompression will only make this patient worse. A patient with anterior impingement between the cuff and the coracoid needs a coracoplasty and will not get better with a subacromial decompression. A patient with work-related physiological impingement may need only to change his posture, for example, to working with the elbow at the side instead of abducted. Patient selection is the key to suc­cess with subacromial decompression.4

Natural history of impingement

Neer[5] classified the impingement syndrome into three stages according to changes in the rotator cuff, caused by repeated abrasion against the roof of the subacromial space. Stage 1 causes oedema and haemorrhage in the cuff, occurs in the younger patient, is often work- or sport-related and tends to settle with conservative measures. Stage 2 causes fibro­sis and 'tendinitis' or partial thickness tears, occurs in an older age group and is more resistant to therapy. Stage 3 implies full thickness rotator cuff tears, rupture of the biceps tendon and bone changes.

Chard et al[6] reviewed 137 patients with impingement treated conservatively. At a mean of over 18 months, 35 still had active tendinitis, a further 40 still had residual pain and 8 had developed pain due to other causes. Only 54 resolved and these were distinguished by early presentation and a history of overuse unrelated to occupation. In 29 patients, function was still impaired and 2 lost their jobs. Chard and colleagues concluded that rotator cuff tendinitis is not an early self-limiting condition, that a sizeable proportion did not resolve with conser­vative management and that improvements in management were needed.

Indications for Surgery

For open or closed surgery, the diagnosis of impingement must be watertight, subluxation must have been excluded, the patient must be over 35 years, must have failed to respond to conservative management, and must have had the symptoms in excess of 6 months, with no improvement.

Diagnosis

A scheme for diagnosis is given in the algorithm shown in Figure 8.4. The patient will typically be a man of over 40 years who complains of true shoulder pain, which is exacerbated in the mid-range of elevation, a mid-range painful arc. The onset is usually insidious. If there is a history of trauma, then cuff tear and dislocation or sub­luxation should be excluded. If the painful arc is work- or sport-related then it may be phy­siological and a change in posture may be of benefit during working or sporting activities. Unfortunately, most patients will have worked this out for themselves before attending a surgeon and have not achieved success.

Figure 8.4 Diagnostic algorithm for impingement.

During the examination, particular attention should be paid to the tests for dislocation, the apprehension tests, sulcus sign and any joint laxity. If there is pain on cross-body adduction, attention should be paid to the acromioclavicu­lar joint and the coracoid. If subcoracoid impingement is suspected, a CT scan should be performed. If there is excessive muscle wasting or weakness, the surgeon should suspect a cuff tear or neurological cause.

Three plain radiographs should be taken, an anteroposterior view of the shoulder, an axial and a subacromial arch view. If there are bone changes on the anteroposterior view (cysts or sclerosis of the greater tuberosity, sclerosis of the undersurface of the acromion, or a decrease in the acromiohumeral interval below 7 mm) then the patient has stage 3 impinge­ment or cuff tear and is a poor candidate for arthroscopic treatment. Degenerative change in the acromioclavicular joint with inferior spikes means that the outer 1 cm of the clavicle may need to be excised, and this is best done with open surgery. If the arch view shows a Bigliani type 3 acromion, then a bony procedure must be performed.

The 'impingement test' should now be per­formed, 2 ml 1 per cent lignocaine (US: lido-caine) being injected just under the anterior acromion. The patient is re-examined at 5 minutes and, for a diagnosis of impingement to hold good, the patient's painful arc should have improved dramatically.

If the patient has had no previous conserva­tive treatment and the time course is under 6 months' duration, then a single steroid injection of 40 mg depomedrone, or equivalent dosage of triamcinalone, may be injected under the anterior acromion, the patient being re­examined in 6 weeks. If the patient fails to respond to conservative measures, then surgery can be considered.

ASD: Ellman Method

Equipment

Many surgeons perform ASD with powered shaver systems alone. However, the soft tissue covering under the acromion (the insertion of the coracoacromial ligament) is very resistant to a powered resector, and virtually immune to a burr, as it is firm and rubbery. This means that electrosurgical apparatus is really necessary and cuts surgical time considerably.

Method

The patient is placed in the lateral position, prepared, draped and a shoulder arthroscopy is performed. The arthroscope is then placed in the subacromial bursa (see Chapter 5). Ellman[9] suggests that the posterior portal is used for an inflow and that the arthroscope is inserted 1 cm below and 1 cm anterior to the posterior angle of the acromion (Figure 8.5).

Figure 8.5 Portals for arthroscopic subacromial decompression.

Two 18-gauge spinal needles are now inserted to outline the anterior border of the acromion (Figure 8.6). The medial needle is placed just in front of the acromioclavicular joint and the lateral needle in front of the anterolate­ral angle of the acromion.

Figure 8.6 Needles are inserted to mark the edges of the coracoacromial ligament.

The soft tissue must now be removed from the undersurface of the acromion, and this means that the coracoacromial ligament must be divided. Ogilvie Harris (personal communi­cation, 1988) has examined the blood supply of the coracoacromial ligament in cadavers. He has shown that the largest vessels are at the coracoid end. At the midway point there are usually three vessels in the ligament, but only 1 in 20 had a blood vessel within 1 cm of the insertion of the ligament. The entry point for the instruments is 3-4 cm from the acromion, in direct line with the two pins (Figure 8.7). The coracoacromial ligament is divided from the medial marker pin to the lateral marker pin, using the electrosurgical apparatus (Figure 8.3). For the electrosurgical apparatus to func­tion, the saline must be flushed out of the bursa and sterile distilled water for irrigation must be instilled.

Figure 8.7 The entry point for the instruments is 3-4 cm from the acromion in direct line with the two pins.

Once the coracoacromial ligament has been divided, the area of resection should be cir­cumscribed, using the electrosurgical appar­atus (Figure 8.9). The tissue within this circle is then morsellized with the cutting diathermy so that it hangs down in strips.

Figure 8.8 The coracoacromial ligament is divided with the electrosurgical apparatus.

Figure 8.9 The area for resection is circumscribed by the electrosurgical apparatus.

The electrosurgical apparatus is now exchanged for the powered full radius resector and all the morsellized tissue resected so that the undersurface of the bone can be seen (Figure 8.10). The full radius resector is now exchanged for an acromionizer burr and the acromioplasty started. Ellman recommends that the burring starts from the anteromedial corner of the acromion, with a deepening hole 3 mm deep which is then extended as a trough from front to back for a distance of 2.5 cm.

Figure 8.10 The full radius resector is used to remove the soft tissue under the acromion.

Figure 8.11 The acromionizer burr is then used to resect the undersurface of the acromion.

Further troughs are now made across the width of the acromion (Figure 8.11). Finally the whole area is polished by circular passes of the burr and the process repeated until the required depth is achieved.

Osteophyte under the acromioclavicular joint should be resected and this requires removal of the inferior capsule before the bone can be burred away.

As in any surgical procedure, the key is visualization, and this is the problem with ASD. Firstly, the surgeon is working in a restricted space and, secondly, the tissues tend to bleed.

Thirdly, the soft tissue does not peel away cleanly and the surgeon finds himself perform­ing an arthroscopic procedure amongst dense seaweed in extremely murky water. Access can be improved by traction of the limb in a caudad manner with a shoulder holder, but the arm must be kept at the side and not abducted as this restricts vision. High pressure irrigation must be used. This does not necessarily mean an infusion pump, although many surgeons find this helps, but there must be adequate flow, under adequate pressure, which means a high head of pressure, a high drip stand and wide-bore inflow tubing. As already stated, an elec-trosurgical apparatus cuts down bleeding and operation time considerably.

Postoperative management

The shoulder swells to an alarming extent following this procedure but this tends to reduce within hours. A sling should not be worn. The patient should perform pendulum exercises every morning and every evening for 1 minute. Strenuous overhead activities, such as tennis and throwing, should be avoided for at least 4-6 weeks. At 6 weeks, strengthening exercises are begun.

Results

Ellman described his 2-5 year results at the American Academy of Orthopaedic Surgeons' meeting in 1989.[7] He started ASD in 1985 and performed 10 cases in that year. By 1987 he had performed 50 cases and by 1989 217 cases. He reported on the 87 cases with a follow-up of 2-5 years. Of these, 79 per cent had stage 2 impingement and 21 per cent had stage 3 - that is, a full thickness cuff tear. The average follow-up was 34 months and the average age 49 years. The patients had received on average four steroid injections prior to surgery and nearly all had experienced night pain. The patients were assessed using UCLA scores and 85 per cent were satisfac­tory, 76 per cent returning to sport. Of the patients with stage 2 impingement, 89 per cent were satisfactory, but those with cuff tears had worse results, only 65 per cent being satisfac­tory. The group with cuff tears had an average age of 74 years and Ellman believes that open treatment is preferable to ASD if a cuff tear is present.

Gartsman also presented his results at this meeting.[10] He showed that results in stage 2 impingement were good, 88 per cent being satisfied, but this dropped to 82.5 per cent satisfaction in the face of a partial thickness tear, and fell even further to only 45 per cent satisfaction in the presence of a full thickness tear.

Warren et al11 presented results in 40 patients undergoing ASD and debridement of the acromioclavicular joint under scalene block, performed between 1984 and 1986. Of these, 20 had stage 2 impingement and 10 stage 3. The average age was 43 years. All had a glenohumeral arthroscopy and 31 abnormali­ties were found, 11 had labral pathology, 10 had full thickness tears, three had partial thickness tears and seven had tendinitis of the long head of the biceps. Overall, 80 per cent were excellent or good as rated by the Hospital for Special Surgery (HSS) scoring system, 14 per cent were fair and 16 per cent poor. The average time to recovery was 3.8 months, most were in work at 10 days and 77 per cent returned to their premorbidity sporting level. Five of the six who failed to return to sport had anteroinferior labral tears (Bankart lesions). Warren stresses that impingement secondary to instability should not be treated by ASD and that patients with moderate size tears should have these repaired at open surgery.

Open or Closed Surgery?

Before electing for closed surgery, both surgeon and patient must be convinced of its superiority over open surgery. Presently this is debatable.

The method of open surgery has been described by Neer:[5] 2 cm of anterior deltoid are released from the anterior acromion and the coracoacromial   ligament   and  anteroinferior acromion are excised. Any osteophytes under the acromioclavicular joint are excised and the rotator cuff is checked for evidence of tears. The deltoid is then carefully reattached and the skin closed. Hawkins[4] reviewed 108 patients who underwent open acromioplasty, and who had no cuff tear. All patients had shoulder pain for over 1 year before operation, despite conservative treatment. Open acromioplasty was successful in 87 per cent of patients. The operation was less successful in women, those with limited preoperative movement, those who were involved in work compensation, and those whose pain had started after direct trauma. In most surgeons' hands, closed acro­mioplasty is not as successful as this, although the recent results of Ellman and Kay[7] at 85 per cent satisfactory for stage 2 impingement and of Alcheck et al[8] at 82 per cent excellent and good are comparable.

There is no doubt that arthroscopic subacro­mial decompression (ASD) is both more difficult and generally takes longer than open proce­dure. Although ASD has these drawbacks, and tends not to be as successful as open acromio­plasty, it does have advantages. These are that it is a day-case procedure, for some reason it is less painful, and the patient has faster rehabilitation. There is no doubt that arthrosco­pic surgery has a good image amongst the population at large and, to an extent, the patient now demands minimally invasive surgery if possible.