What keeps a shoulder stable – Is there an ideal method for anterior stabilisation?

The gleno-humeral joint is by far the most mobile in the human body and allows us to place the hand in almost every desired position within a large sphere. The extreme range of motion, however, comes with a cost, since this joint is also the one most often afflicted with instability.

Mobility is provided by the unconstrained, shallow, ball-and-socket construction, a loose joint capsule and a few ligaments that are not tight in the resting position. The anterior ligaments are tight only in the outer range of external rotation and abduction. The abducted and externally rotated position is also the most vulnerable position for anterior dislocation of the joint if an addition of an external force (such as fall on an out-stretched arm) cannot be counteracted by muscles and ligaments. The position would also be in the wind-up and cocking phases of throwing and similar activities that can provoke symptoms of instability because of the large muscular forces and angular velocity that can be produced.

To allow extreme mobility, the joint has an inherent laxity but for normal function, control and stability is essential which is provided by the surrounding musculature.^1,2 Seventeen muscles, including the rotator cuff and deltoid, attach to the scapula and are involved in control of scapular motion and thereby the positioning of the glenoid, the fundament for arm function. In addition, the pectoralis major and latissimus dorsi significantly contribute to arm motion and strength, and synchronisation of all muscle activities is a prerequisite. Smooth sequential muscle activation is regulated by proprioception mediated by neurological sensory organs in ligaments, joint capsule, skin, tendons and muscles. Also, visual input and sensory afferents from the rest of the body contribute to function. It is well described that proprioception is disturbed in patients with anterior shoulder instability,^3–6 assumedly as a result of a dislocation and soft tissue lesion but a bilateral proprioceptive dysfunction also in the uninjured shoulder has been described and consequently a pre-injury deficit cannot be excluded.

Anterior shoulder instability is a common and usually easy diagnosis. With recurrent traumatic dislocations or subluxations, the management is often surgical with mostly a reliable and rewarding outcome. The literature is replete with case series describing good results after surgical stabilisation but with careful reading, it also becomes evident that in a varying but often substantial proportion of patients, surgery can fail with persistent instability.^7–12 The solution to this has often been a search for a new method that would universally solve the instability problem. It appears that with every generation of shoulder surgeons, a new method emerges that the advocates hope will take care of all cases of anterior instability.

There are numerous techniques for surgical stabilisation and with the advent of arthroscopic techniques, the so-called arthroscopic Bankart repair became the gold standard despite the fact that the arthroscopic method was associated with almost exactly twice the recurrence rate as with open Bankart repair (according to a high-quality study on the subject by Mohtadi et al. ¹³ ). The arthroscopic technique aims to repair the Bankart lesion with avulsion of the inferior gleno-humeral ligament, and adjacent labrum to the glenoid, as did the original Bankart procedure, ¹⁴ but an open traditional operation also includes ligament fixation with osteo-suture, suture of the capsule and the partially released subscapularis tendon. These parts of the procedure may be of significant relevance for the result, but their relative contributions have often been overlooked. The potential impact can be illustrated by methods like the Putti-Platt, Magnusson-Stack and Neer’s capsular shift which are all soft tissue procedures aimed at capsular imbrication or tightening of the subscapularis tendon.^14–17 The reported results of these operations have described similar or even better success rates in preventing re-dislocations than the more recently popular methods. However, loss of external rotation and osteoarthritis in long term have been a concern leading to these procedures being abandoned.

Arthroscopic stabilisation has dominated the last two decades but due to reported high recurrence rates after this procedure, up to over 25% in some series,^7–12 the search for a new procedure has continued. More recently, the Latarjet operation has regained popularity following the reports by Bradley and Walch 2002 ¹⁸ and Bouju et al. 2014 ¹⁹ and Metais et al. in 2016. ²⁰ This method and the quite similar Bristow procedure have been reported in several case series with good outcomes including low rates of recurrent instability, at least in study format and in the hands of high-volume surgeons.^21–25 The technique has been considered difficult with a long learning curve and complication rates have been a concern.^26,27 So far there is no solid scientific support for superiority of this procedure over others.^10,28–31

Arthroscopic Latarjet is also described. ³² Recently, this technique has developed further and is acknowledged by arthroscopic equilibrists and many surgeons seem attracted by the idea and the medical industry has provided devices for the procedure.^33–36

It is well known that young age at the first dislocation is associated with a higher risk of recurrence. ³⁷ The reason for this is largely unknown but it has been speculated that inherent laxity could be a factor. Conversely, the risk of persistent instability decreases with age even though a Bankart lesion is almost always present. This seems to demonstrate that there are other associated contributing factors. Bankart fracture of the anteroinferior glenoid rim has been reported as such a potential risk factor, but scientific evidence is lacking, and non-operative management of even large fragments has been reportedly successful. ³⁸ A benefit with the Latarjet method is that it includes transposition of the coracoid tip that can fill bony defects. On the other hand, patients with only small fracture fragments may sometimes also experience chronic instability; the size of a fracture fragment is apparently not a reliable discriminating factor for decision of surgical or non-operative treatment. Large, engaging Hill–Sachs impression fracture is also an associated lesion that may increase the risk of instability symptoms and may need attention with for example, remplissage or bone graft.

Several studies, completed or ongoing, have compared open stabilisation according to Latarjet or the very similar Bristow procedure, with the arthroscopic Bankart operation.^39–42 Comparing open and arthroscopic procedures with different methods for stabilisation may seem inappropriate since the open procedures previously seem to have proven more reliable in terms of preventing recurrent dislocations, perhaps thanks to the ‘side effects’ of an open operation. Few studies have compared open and arthroscopic methods using the same principle for stabilisation. In systematic reviews, Hurley et al. ³⁰ found six studies and Malahias et al. ⁴³ five studies comparing open and arthroscopic Latarjet but these studies were of low or moderate quality. It may however also be inappropriate to compare methods rather than patients if the aim is to provide a successful treatment for as many patients as possible. It is likely that the patho-mechanisms, and thereby the resulting soft tissue injury, differs between patients and it is probable that not one single procedure will reliably treat every patient with anterior shoulder instability. If all patients with anterior shoulder instability are treated with one or the other method, there will always be some that do not fit into the algorithm. This can be exemplified and illustrated by the so-called HAGL (humeral avulsion of gleno-humeral ligament) lesion^44,45 in which the main part of the injury is located on the humeral side instead of the glenoid, and ALPSA (anterior labral periosteal sleeve avulsion) ⁴⁶ and SLAP (superior labral tear from anterior to posterior), ⁴⁷ examples of soft tissue lesions described associated with anterior dislocations. There will be individuals with specific demands and there will be those who have different injuries or combinations of lesions that are not addressed by the respective technique. Different methods may have an indication in accordance with the pathologic morphology.

Furthermore, De Palma described different directions of anterior dislocations, the subcoracoid being the most frequent, but also subclavicular, subglenoid and intrathoracic dislocations may occur, and it appears obvious that the different directions of dislocation will be encompassed by different soft tissue injuries. ⁴⁸ In most anterior dislocations, a Bankart lesion, avulsion of the antero-inferior labrum and inferior gleno-humeral ligament will be present. A dislocated humeral head is however much larger than the space allowed by the Bankart lesion and stretching of the capsule and subscapularis tendon most probably also occurs. Moseley and Overgaard ⁴⁹ investigated the condition of the subscapularis tendon after anterior dislocations and found it stretched and described partial lesions in most of the examined patients, a finding later underlined by DePalma et al. ⁵⁰ and Gamulin et al. ⁵¹ The exact contribution of these lesions in recurrent instability is unknown but appears as a factor of importance. In most of the open procedures for stabilisation, release and repair of the subscapularis tendon will be a substantial part of the operation and the currently popular Latarjet method creates a sling mechanism with the tendon that is also probably an important feature.

Regardless of surgical technique it appears that there is a tendency to convert a neuromuscular problem into a mechanical, two-dimensional one. Under normal circumstances, shoulder stability is functional, dynamic, active and three-dimensional, provided by delicately synchronised musculature. Many surgical methods aim at passive stabilisation by creating an anterior restraint against translation of the humeral head. This view might infer a risk of stiffness and in long-term create osteoarthritis by disturbing the normal biomechanics in which sliding is an integral part.

In summary, an anterior shoulder dislocation will cause several patterns of patho-morphology and there is not a single procedure that can address all lesions. Therefore, it appears that many previous and currently ongoing studies comparing two treatment methods will fail to demonstrate a clinically meaningful difference unless relevant associated lesions are identified, acknowledged and treated. Application of different methods addressing relevant lesions and adjusted to each patient´s age, level of activity and other individual factors is likely necessary to improve outcomes of surgical management of anterior shoulder instability.^52–54 Perhaps Bankart lesions and partial subscapularis tears should be repaired to improve proprioception as well as stability, perhaps the Latarjet method should primarily be used in patients with anteroinferior bony defects and some of the open old soft tissue procedures can perhaps still be indicated for patients with incompetent capsule or for revisions. Future research needs to consider the potential complexity of these factors and injuries causing anterior shoulder instability.