1.1 Prosthetic instability

Prosthetic instability manifests itself as prosthetic dislocation where there is a gap between the humeral cup and glenoid sphere, with Wall et al. noting this in 7.5% of cases,¹⁰ Boileau et al. in 6.6%⁶ and Werner et al. in 8.6%.⁹ It is felt that prosthetic instability results as a consequence of humeral medialisation, which decreases the tension of the deltoid muscle. Deltoid muscle tension can be restored intra-operatively by increasing the offset on the humeral cup, however great care must be taken as over-tensioning can lead to acromial fractures.

1.2 Scapular notching

In all studies reviewed, scapular notching has been noted in more than 50% of cases.^8–11 The Grammont type RSA medialises the glenoid component thus lowering the humerus. This anatomical change can cause impingement of the medial aspect of the polyethylene humeral cup on the scapular neck inferiorly with the arm adducted. In addition to direct impingement, more severe cases of scapular notching may be caused by polyethylene particle wear causing osteolysis of the scapular neck (Fig. 1).

Close

Fig. 1 Illustrated standard reverse shoulder arthroplasty – inferior impingement with arm adducted.

Export to PPT

1.3 Deficient or absent external rotation

The Grammont type RSA has not demonstrated an improvement in active external rotation in the literature.^8–11 The limited lateral offset of the glenosphere limits the possibility of rotation of the humeral cup around it with the shoulder in adduction. In addition, the medialised centre of rotation decreases the amount of posterior deltoid muscle fibres that can be used to compensate for the absent external rotators.

1.4 Deficient or absent internal rotation

Just as the studies showed no significant increase in active external rotation, active internal rotation also shows no significant improvement.^8–11 Again this is explained by the medialised centre of rotation, which leads to limited excursion of the cup around the glenosphere in the horizontal plane and decreases the possibility of using the anterior deltoid to compensate for the absent internal rotators.

1.5 Loss of shoulder contour

Finally, the medialisation of the glenosphere and subsequent lowering of the humerus leads to a loss of the normal shoulder contour. This was raised as a concern in a study by Boileau et al.¹¹

To address these problems Boileau et al. adopted a novel approach to address the problematic issues encountered with the standard medialised RSA.¹² He advocated the use of an autologous bone graft harvested from the humeral head on a specifically designed baseplate to lateralise the prosthesis. They called it the bony increased-offset reversed shoulder arthroplasty (BIO-RSA) (Fig. 2).

Close

Fig. 2 Illustrated BIO reverse shoulder arthroplasty – no inferior impingement with arm adducted.

Export to PPT

3.1 Patients

Sixty consecutive patients underwent 65 procedures over a two-year period starting from September 2013 at a single centre. All procedures were completed by one consultant surgeon experienced in RSA surgery (MW). The indications for surgery were; cuff tear Arthropathy (44 patients), revision prosthesis (9 patients of which; 3 resurfacing hemiarthroplasty, 3 total shoulder replacements and 3 hemiarthroplasty), proximal humerus fracture (3 patients) and post-traumatic osteoarthritis (2 patients).

There were 40 men (62.7%), 3 of whom had bilateral surgery and 25 women (37.3%), 2 of whom had bilateral surgery. Their mean age at time of surgery was 74.1 years (range, 49 to 89) with an average follow up of 7.1 months (3.4–24 months).

Patients were allocated their procedure based on indication, surgeon preference and pre-existing bone quality of the glenoid.

3.2 Surgical technique

All patients were operated in beach chair position under general anaesthesia, regional anaesthesia or a combination of both. A standard deltopectoral approach was used and an Aequalis Reversed™ Shoulder Arthroplasty (Tornier Inc, Houston, Texas, USA) prosthesis was used in all cases.

The standard RSA was implanted using the manufacturers standard technique.¹⁴

The technique for implanting the BIO-RSA differs from the standard RSA. Following a standard deltopectoral approach and conservative capsulotomy, a 155° inclined humeral guide is placed on the apex of the humeral head and a retroversion rod is used to establish the desired retroversion (0° to 30° according to the trans-epicondylar axis). A 2.5 mm, threaded guidewire is then inserted into this axis and the humeral guide is removed.

The BIO-RSA, 29 mm graft reamer is guided along the threaded guidewire to create a flat surface of subchondral bone. A 8.3 mm cannulated drill is then advanced over the alignment pin to create a central hole of the graft before a bell saw is used to create a cylinder of cancellous bone 29 mm in diameter. A cutting guide is then used to harvest the desired thickness of bone graft (either 7 mm or 10 mm). The disc of pure cancellous bone is then extracted and inserted onto the BIO-RSA baseplate (Fig. 3).

Close

Fig. 3 Intra op images – bone graft harvesting from humeral head and application on glenoid.

Export to PPT

The glenoid is prepared using a 29 mm reamer over a guidewire (either 0° or 10° inferior tilt). It is advised to ream approximately 5 mm until cancellous and/or bleeding bone is reached to better facilitate bony integration of the BIO-RSA graft (Fig. 4).

Close

Fig. 4 Post op radiograph of BIO reverse shoulder arthroplasty.

Export to PPT

A central 8 mm hole is drilled over the glenoid guidewire, to accommodate the long post baseplate. Fixation is then obtained with impaction and two 4.5 mm convergent non-locking compression screws and two 4.5 mm divergent locking screws. The glenosphere is then fixed to the baseplate via a Morse taper and countersunk set screw.

3.3 Authors’ modifications to standard technique

3.3.1

3.3.1 Chamfering of the autologous bone graft

In cases of extreme ante- or retroversion of the glenoid, or excessive posterior wear, positioning of the guidewire as per the manufacturer's guidelines may result in excessive reaming into the scapular neck, or insufficient reaming of the rim of the glenoid. Using our technique, the guidewire is placed concentrically in the glenoid to allow optimum reaming. The ante- or retroversion is then corrected by chamfering the autologous graft to restore alignment before final implantation of the base plate (Fig. 5).

Close

Fig. 5 Modification of manufacturer’s technique – chamfering of autologous bone graft to accommodate ante- or retro-version.

Export to PPT

3.4 Rehabilitation

Patients were discharged 1–2 days following surgery with a Polysling for the first 2 weeks. The rehabilitation protocol was identical for both the BIO-RSA and the standard RSA.¹³ Following surgery, the patients were educated on sling management with immobilisation for two weeks. Patients were provided with an ice band and encouraged to continue application as required during the rehabilitation period. Commencement of the initial post-operative exercises was conditional upon the dissipation of the nerve block and included elbow, wrist and hand with shoulder pendular exercises and active assisted external rotation. The amount of external rotation permitted was assessed intra-operatively dependent on the quality of the Pectoralis Major and Subscapularis tendon repairs, typically between neutral and 30°.

Prior to discharge, an Occupational Therapy assessment was undertaken to assess the patients’ home environment and ability to complete activities of Daily Living (ADL).

At 2 weeks, the patients were advised to wean themselves from the sling, precautions being given to avoid weight-bearing through the operated arm, to restrict lifting to that of a cup/toothbrush and movements such as extension, internal rotation, and combined abduction with external rotation. Active assisted exercises within a 90° elevation and anterior to scapular plane, with initial recruitment for anterior deltoid and scapular disassociation work was commenced.

As patients gained pain-free range of movement and could perform static joint holds at a 90° flexion whilst supine, patients were progressed to active movements and advancement of the deltoid rehab protocol.

Between 6 and 12 weeks, patients progressed to active movements through range and a graduated deltoid training regime against gravity. An individually tailored regime was provided to each patient, focusing on movement quality in order to address scapular dyskinesis and poor movement patterns.

At 12 weeks, if patients could demonstrate pain-controlled functional range of movement with no signs of instability, specific strengthening goals were designed to achieve each patient's individual realistic goals or functional demands.

The timescales cited were for guidance only and were dependent on functional level pre-surgery, co-morbidity and status of the rotator cuff.

3.5 Follow up

All patients were retrospectively evaluated with a mean follow up period of 7.1 months (3.4–24 months). Their Oxford Shoulder Score (OSS) and range of movements were evaluated before and after surgery.¹⁴ Episodes of instability were noted along with other complications experienced during the follow up period.

Bone graft incorporation and formation of a scapular notch was assessed with regular radiography. Patients had a plain antero-posterior (AP) radiograph at day one, an AP and axillary at 2 weeks, with an AP, axial and lateral at 6 weeks and an AP and axial at 3 months. Finally a CT scan was advised at 3 and 6 months.

3.6 Statistical analysis

Descriptive statistics were reported for demographic data. Differences between groups were made using Student's t-test (paired) for reported outcomes and range of motion. The level of statistical significance was set at p < 0.05. Statistical analysis was performed with SPSS 22.0 software (SPSS Inc., Chicago, IL, USA).

4.1 Complications

There was one case of deep infection presenting with sero-sanguinous discharge from the superior part of the wound at 7 months. A wound swab grew Staphylococcus aureus and the patient had a revision of all mobile parts and a 3 month course of oral antibiotics. At 18 months her Oxford score was 39.

One patient had a post-operative haematoma secondary to low molecular weight heparin prophylaxis, this resolved with conservative management.