Fig. 31.1
The Mallet score
The components of the Mallet score should recorded separately because some of the components can be normal in the presence of severe contractures. Changes in the individual components of the score are therefore considered a more accurate than the total score [7]). Bae et al. [23] investigated the reliability of the Mallet score. Eighty children with OBPIs were examined by two trained examiners on two occasions. They found good inter-observer and intra-observer reliability.
How to Prevent Glenohumeral Dysplasia?
Studies identified and included in this chapter were of levels III to V.
During the first 24 months of life there is a role for nerve surgery with the primary aim of reinnervating muscles. At the time of nerve surgery concomitant shoulder surgery should be performed if the patient has a structural abnormality [24]. After 24 months of life the primary aim of surgery should be to improve any glenohumeral abnormality.
Nonsurgical Treatment
All patients with an OBPI who have a shoulder that is congruently reduced should undergo physiotherapy to prevent contractures and maintain a functional arm. This requires an experienced therapist and parents who are able to engage with therapy at home. The aim of therapy is to encourage the child to use the affected arm without stressing them. Therapy should focus on external rotation with the shoulder in an adducted position. One technique that is used is to lay the child on the normal side and prompt them to play with a toy with the affected arm. Another way to prevent contractures is to encourage parents to position the arm in an abducted and externally rotated position on a pillow after the child has fallen asleep. Surgery should be considered if despite physiotherapy the patient develops a subluxed shoulder, a dislocated shoulder or has a progressive loss of external rotation [25].
Subscapularis Release and Tendon Transfers
When nonsurgical treatment has failed and an internal rotation contracture with a congruent joint is present surgical lengthening of subscapularis with or without a tendon transfer to rebalance the external rotators should be considered.
Some surgeons advocate performing a tendon transfer at the time of a subscapularis release. Others advocate a tendon transfer if active external rotation has not been achieved by 2 years of age. A latissimus dorsi transfer to the greater tuberosity is the most commonly used tendon transfer [15, 25, 26]. In 1978 Hoffer et al. [25] wrote one of the earliest descriptions of tendon transfers for OBPIs. They described an open anterior release through a deltopectoral approach combined with a posterior incision to transfer latissimus dorsi to the greater tuberosity. They reported successful surgery on 11 children aged between 2 and 7 years.
Following Hoffer’s study others have replicated there results. Thatte et al. [15] reported a series of 150 patients out of 305 patients treated for an OBPI. The surgery differed from that described by Hoffer in that the 150 patients all underwent a transfer of both latissimus dorsi and teres major. Patients were followed up for a mean of 4 years (range 2.5–8 years). They reported that all patients had an improvement in shoulder function and 35 (23 %) made a complete recovery.
Aydin et al. [26] reported a similar series of 46 patients who had a latissimus dorsi transfer and were followed up for a mean of 41 months. They analysed the results of those with severe contractures (<90° abduction) and moderate contractures (>90° abduction) separately. In the group with severe contractures they found a significant increase in abduction from 62.5° to 131.4° and external rotation from 21.4° to 82.6°. In the group with moderate contractures they found a significant increase in abduction from 99.4° to 140° and external rotation from 33.2° to 82.7°. There was no significant difference between the two groups. These study findings have been reproduced by others and demonstrate that even in the presence of severe contractures a latissimus dorsi transfer can rebalance the shoulder muscles sufficiently to achieve improved external rotation and abduction [27].
Most studies report the short to medium term results of tendon transfers [5]. The longest follow up study we identified was by Pagnotta et al. [28] who followed up 203 patients who had undergone a latissimus dorsi transfer at 1, 3, 6, 10 and 15 years. They found that following a latissimus dorsi transfer abduction began to deteriorate at 6 years, however, external rotation was preserved. More long term studies are required to establish if there is deterioration in function over time and to what extent.
Tendon transfers of the lower trapezius to infraspinatus have been performed. This is usually as a revision procedure when latissimus dorsi and teres major have already been utilised. Using trapezius has the theoretical advantage of utilising a muscle that is not innervated by the brachial plexus. Trapezius also normally acts in phase with the posterior cuff. Bertelli et al. [29] reported the outcome in seven children at 2 and 4 years post surgery. They found the mean increase in external rotation to be 47° at 2 years and 54.3° at 4 years. We could not identify any other papers investigating trapezius transfers however it is a viable option and is worthy of further study.
Several studies have shown that in the short to medium term a latissimus dorsi transfer and anterior release can improve shoulder function.
How to Treat Established Glenohumeral Dysplasia?
When the glenohumeral joint is subluxed or dislocated but the articular surfaces are preserved joint reduction by soft tissue releases and osteotomies can be performed. A step wise approach to achieve this has been described by Di Mascio et al. [30]. They advocate that as little surgery as is necessary to achieve a congruent stable reduction throughout a functional range of motion should be performed. Step one is an open reduction; step two is a lengthening of subscapularis and an anterior capsule release; step three is an internal rotation humeral osteotomy and step four is a glenoplasty. They used this technique to treat 29 patients with a mean age of 5 years and reported the results at a mean follow up of 34 months. They reported a mean increase in abduction of 24° and external rotation of 54°.
Step One – Open Reduction
In Di Mascio’s series they used an open reduction through a deltopectoral approach. This enables good access to the anterior structures and also allows access to the humeral neck for an osteotomy if required. None of the 29 patient in Di Mascio’s series was treated by open reduction alone. We could not identify any case series reporting the outcome of open reduction without a muscle lengthening procedure in our literature search. An open reduction alone is not therefore considered a standard treatment.
Step Two – Lengthening of Subscapularis and Anterior Capsule Release
The coracoid is often elongated and is a physical barrier to reduction. In this situation a coracoid osteotomy is necessary to reduce the shoulder. This should be done before any other releases because a coracoid osteotomy releases its soft tissue attachments and improves access to other structures. If a coracoid osteotomy is not being performed or if despite a coracoid osteotomy a concentric reduction is not possible a subscapularis and pectoralis major lengthening should be performed. If this does not achieve a concentric reduction throughout the range of motion then the SGHL, MGHL, IGHL, CHL and rotator interval should all be fully released [30]. Brachialis and teres major may also be contracted and require releasing.
Most surgeons advocate that the releases are performed as an open operation, however, an arthroscopic release of the subscapularis and capsule alone may also be considered. This has the advantage of enabling a formal assessment of the glenoid which is difficult to image in infants. Pedowitz et al. [31] reported a series of 22 children undergoing an arthroscopic anterior capsule and subscapularis release to reduce subluxed shoulders. Post operatively they all were placed in spica casts. They found that they were able to reduce the shoulders successfully.
Kozin et al. [32] performed arthroscopic releases in 29 children; 16 of which also underwent tendon transfers. They followed these patients up at one year and found both an improvement in clinical results and a maintained reduction of the humeral head.
Pearl et al. reported similar results in their series of 33 children. The younger cohort of 19 children with a mean age of 1.5 years underwent an anterior release alone. The older cohort of 14 children with a mean age of 6.7 years underwent an anterior release and latissimus dorsi transfer. They found a significant improvement in external rotation in both groups. 18 of these patients’ had a preoperative MRI demonstrating a pseudoglenoid and this had improved in 12 of the 15 children who had post operative MRI scans.
Not all series have shown good results using soft tissue surgery alone to treat established glenohumeral dysplasia. Hultgren et al. [33] followed up 118 patients for a minimum of 7 years. All patients had undergone soft tissue procedures alone to reduce subluxed glenohumeral joints. They reported a high failure rate with 66 (56 %) patients developing a subsequent subluxation and 23 (19 %) of these requiring a humeral osteotomy.
Step Three – Internal Rotation Humeral Osteotomy
Di Mascio et al. [30] described performing a derotational humeral osteotomy to restore a more normal humeral neck version when the retroversion is greater than that that which will remodel. Our literature search identified two papers reporting the outcomes of internal rotation humeral osteotomies. Sibinski et al. [34, 35] performed 10 internal rotation osteotomies on 25 patients undergoing a latissimus dorsi transfer. At a mean of 3.8 years there was an improvement in Mallet scores and internal rotation but other movements did not improve significantly.