Purpose of the study: A multicentric study involving ten centres of shoulder surgery collected a retrospective series of 45 patients treated for infection of a total shoulder arthroplasty.

Material and methods: Mean age was 68 years (range 28–88) with 2/3 male. The initial indication was excentred joint degeneration (37%), centred degeneration (5%), necrosis (7%), fracture (34%), excentration without degeneration (12%), rheumatoid disease (5%). The implants were a reversed total prosthesis (59%), a hemiprosthesis (26%), a total anatomic prosthesis (15%). Nineteen patients (46%) had a ‘naive’ shoulder at implantation of the first prosthesis, including 70% without any specific medical condition: 30% fracture (hemiprosthesis) and 50% reversed total prosthesis.

Results: The infection developed immediately after implantation in 76% of the shoulders (abnormal pain, hematoma, early collection, inflammatory scar, early fistula). The preoperative Constant score (26 points) showed pain at 4, function at 8, mobility at 12, and force at 2. Osteolysis was noted in 30%. The infection lasted less than 3 months before revision in 27%, 3 to 12 months in 68%, and more than 12 months in 5%. Blood tests were normal in two patients. Staphylococcus epidermidis was isolated in 29%, Propioni bacterium acne in 31%, and these two germs in 5%, but no germ was identified in 24%. There was no loosening in 69% of the cases, making the revision invasive. A one-phase procedure was performed in 31% of patients and a two-phase procedure in 25%. Repeated wash-outs were performed in 25%. A reversed prosthesis was implanted at revision in 87% of the shoulders. These revisions produced a 22% complication rate and 27% were revised for repeated wash-outs, instability or definitive en bloc resection. An infectious disease specialist was called in for 75% of patients; a two-drug i.v. regimen for 10 days was instituted followed by oral antibiotics for 3 months. Outcome was assessed with more than 24 months follow-up for 26 patients. The Constant score at review was 47 points (pain 12, function 12, mobility 20, force 4).

Discussion: Seventy-seven percent of the patients were considered cured at more than 2 years. The two-phase surgery, undertaken when the infection had lasted 3 to 12 months (40%) or more than 12 months (60%), gave a Constant score at 51% and 100% cure. The one-phase option undertaken for infections lasting less than 3 months (40%) or 3 to 12 months (60%) gave a Constant score at 46% and 50% cure. Simple wash-out required 60% revision (repeated wash-out, or implant replacement) and 67% cure. In the event of complications after revision, the cure rate was 67%. In all, only 6% of the patients were dissatisfied.

Purpose: The goal of this retrospective study was to analyze the results of each kind of reoperation used in rotator cuff retear.

Materials and Methods: 38 rotator cuff retear were reoperated. 18 female and 20 male with a mean age was 57(range 40 to 73 years). Initial tear were repaired by open surgery with a procedure of trans osseous reinsertion in 26 cases. Arthro CT Scan, confirmed the recurrent tear and evaluated the size, number of tendon involved and the global fatty degeneration index (GFDI) of the muscle.The size of the retear was smaller in 9 patients, same in 5 and bigger in 23 with always a decreasing of the GFDI from 0.83 to 1.60. The mean time of reoperation was 29 months(range 1 to 120). We performed 14 trans osseous reinsertion, 6 latero lateral tendinous suture, 3 resection reinsertion, 5 deltoid flap, 3 latissimus dorsi flap, 4 reversed prosthesis, one arthrodesis, one subscapularis flap and one debridment for acute infection. A biceps tenodesis were performed in 20 cases.

Results: Results were evaluated with a mean follow-up of 33 months (range 5 to 145). The constant’s score improved from 46(22 to 68) preoperatively to 67(40–95) post operatively with a decreasing of the GFDI in 0.83 to 1.60. The best results were obtained with a trans osseous reinsertion (Constant’s score 70; gain 40%) and latero lateral tendinous suture(Constant’s score 78; gain 37%). However, the gain higher was obtained with the reversed prosthesis(Constant’s score 57;gain 88%).

Discussion and Conclusion: Arthro CT Scan is a good way to diagnose a recurrent rotator cuff tear and to choose the type of new operation when shoulder still painfull. The final result is higher if it is possible to do a transosseous reinsertion for a small retear with a good trophicity of the muscle (GFDI lower). However for pseudo paralytic shoulder, reversed prosthesis can improved the constant’s score from 34 to 64 with a restoration of a good elevation. Latissimus dorsi flap (deltoid and subscapularis flap) are usefull for painrelief but the improvement of motion is less.

Glenoid component loosening is a concern in long term outcome of total shoulder arthroplasty. Unfortunately revision of glenoid loosening remains very challenging regarding bone reconstruction and reimplantation of a glenoid component. A new design metal-back component (Arrow™) seems to get the procedure easier and lead to higly best clinical and radiological results than prior technics.

It is a retrospective study of 34 patients operated on for isolated aseptic glenoid component loosening between 1997 and 2007. Glenoid bone loss based on a new classification was rated in 5 degrees. All patients were reviewed at a mean follow up of 34 months (from 12 to 89). Pre and post operatively Constant score and SST have been evaluated. Fluoroscopic views and CT scan have been performed for radiologic assesment.

Mean delay for revision surgery was 58 months (from 28 to 134 months). 16 patients underwent only bone graft reconstruction. 18 patients underwent both bone reconstruction and glenoid implant reimplantation.

Autolog bone graft (iliac crest) was used in 19 cases, allograft bone in 5 cases, and both autolog and dried bone substitutes in 10 cases. The glenoid component was an uncemented metal-backed in 11 cases, and a polyethylene cemented in 2 cases. The new metal-backed bone ingrowth component is composed with a keel and a winglet securing bone graft fixation and implant stability. For the all cohort Constant score improves of 21 points (from 35 to 56). Gain of pain and active motion is statiscally significative (p< 0.005).Patients who underwent reimplantation of a new glenoid component had a higher clinical result (Constant score: 66) than those who underwent only a bone graft reconstruction (Constant score: 52) in term of pain and active forward elevation as well (p< 0.001). At maximum follow up (32 months) no radiolucent lines or loosening of the new glenoid component was noted. Patients who underwent only bone reconstruction showed a mild glenoid erosion due to subsidence and resorption of cancellous auto or allograft bone.

Revision of glenoid component depends on the glenoid bone stock specially about posterior defect. The use of cortical autolog bone graft and new metalback component allows better glenoid reconstruction, secure the stabiliy of the new gle and yields to goods clinical results restoring painless shoulder and function. Reimplantation of a glenoid component becomes an adequate procedure for total shoulder arthroplasty revision

Aims: The revision of hemi or total arthroplasty represents a difficult challenge for a shoulder surgeon. The purpose of this study was to report the results of a retrospective series of 30 reverse prosthesis.

Materials and Methods: 17 patients were female, 13 were male; the mean age was 68 (45–84). 14 patients had at least 2 surgeries before the arthroplasty. Of the failed arthroplasties, 26 were hemiarthroplasties and 4 were total shoulder arthroplasties. The indication for the revision of hemiarthroplasty were: 10 migration of the tuberosities, 7 secondary rotator cuff tears, 5 were stiff, 2 glenoiditis,, one humeral loosening and one sepsis. The indication for the revision of total arthroplasty were 3 loosenning of the glenoïd component and 2 secondary rotator cuff tear with a superior ascension of the humeral head. The rotator cuff was always irreparable. We used mainly a deltopectoral approach. During the removing of the humeral component and the cement, 6 fractures of the diaphysis occurred treated by cerclage. A bone graft was used at the level of the glenoïd (3) and of the humerus (4). We implanted 19 Delta and 11 Arrow reverse prosthesis. The patients were reviewed clinically using the Constant score with a mean follow up 3 years (range 1–5 years).

Results: The Constant score improved from 20 (6–39) pre operatively to 55 (44–87) post operatively. Shoulder flexion increased from 45 (20–100) pre operatively to 100 degrees (90–160) post operatively. Pain was improved from 3/15 to 13,66/15 post operatively. The worse results were obtained for the patients who were multioperated before the first arthroplasty. Four patients were reoperated: 2 for dislocations and two for material dysfunction(glenosphere). 6 patients were disappointed, 18 were satisfied and 6 patients were very satisfied subjectively with the procedure. No difference in term of clinical results between Delta and Arrow except no glenoid notch occurred with Arrow prosthesis.

Discusssion and conclusion: After a failure of an hemi or total arthroplasty, if the cuff is no functional or no reparable, reverse prosthesis is currently the best salvage procedure. Bone glenoid stock and deltoid muscle are the limited factors. Surgical tips and tricks should be known before to do this difficult operation. Others procedures can be discussed: reconstruction of the coracoacromial arch, gleno-humeral arthrodesis, deltoïd flap or latissimus dorsi flap and others semi-contrained prosthesis. Patients who had pseudo-paralytic painfull shoulder, without stiffness after an arthroplasty with less than 2 previous operations represent our best results.

Does the pre-operative range of active anterior elevation alter the outcome of reverse shoulder prosthesis? A review of 96 cases.

Aims: The aim of this retrospective study was to analyse the results of reverse shoulder prosthesis in massive, irreparable rotator cuff tears in terms of the preoperative active anterior elevation (AAE).

Materials and Methods: This was a retrospective study of 96 reverse shoulder prostheses in patients with a mean age of 74 years, and with an average follow-up of 30 months. We divided the patients into three groups for the purposes of the study. Group 1 had an AAE less than 60° (n=51); group 2 had an AAE between 60 and 120° (n=39); group 3 had an AAE above 120° (n=6). The majority had off-centre arthritis with a Fukuda Hamada classification of IV or V (n=77); only 19 fell within classes I, II or III. We excluded patients who had previously had a failed anterior cuff repair or arthroplasty and those with a diagnosis of rheumatoid arthritis.

Results: There was no significant difference in constant score between the three groups based on preoperative AAE: group 1: 63.50; group 2: 65.05; group 3: 65.16. Analysis of the 96 reverse prostheses in relation to the Fukuda Hamada classification also showed no difference: types I, II and III had a constant score of 64.37; types IV and V a score of 63.68. However, the improvement in AAE (I), rotation (RE1 and RI) and in power was significantly greater in shoulders with pseudoparalysis (p< 0.001): group 1: I = 71.43%; group 2: I = 33.48%; group 3: I = 3.03%. The degree of improvement of the pseudoparalytic shoulders (group 1) was influenced by the Fukuda Hamada classification (p< 0.01): 77.78% for stages I, II and III compared with 69% for stages IV and V.

Discusssion and conclusion: This retrospective study confirms that the reverse shoulder prosthesis is a beneficial treatment for massive, irreparable cuff tears in older patients with shoulder psuedoparalysis. Its use is debatable if the preoperative AAE is over 120°, in which case hemiarthroplasty may be a better option. The benefit of the reverse prosthesis is greatest in the shoulder with pseudoparalysis, no glenohumeral arthritis (Fukuda I, II or III), and no previous surgical intervention.

Purpose of the study: The aim of this biomechanical study was to assess the performance of the deltoid muscle in the absence of a rotator cuff using different models for shoulder prosthesis.

Material and methods: A computer model reproducing the three dimensions of the glenohumeral joint was use to analyze the force of the deltoid muscle during abduction movements in shoulders devoid of a rotator cuff. The three heads of the deltoid were analyzed in order to determine the most effective level of muscle tension. The lever arm of the deltoid was measured from 0–90° abduction. Using this 3D model, we simulated implantation of six different models of reversed prostheses in order to assess the biomechanical situation which would be the most favorable for the deltoid. Performance of the normal deltoid was compared with the performance of the deltoid after implantation of an anatomic prosthesis and after implantation of an reversed prosthesis. Several variables were studied: medial offset of the center of rotation, lateral offset of the humerus, lengthening of the deltoid muscle.

Results: Optimal deltoid performance (especially from 60–90° abduction) was observed if the center of rotation was offset medially and the humerus was offset laterally and lowered. A 10% increase in the length of the muscle fibres increased muscle performance 18%. Exaggerated lateral offest of the humerus increased deltoid performance between 30 and 60° abduction but lost its beneficial effect at 90° abduction. From 15° abduction, a scapular notch appeared when the humerus was off set medially. This could be avoided if the humerus was offset laterally with a less medial center of rotation. Beyond 150° abduction, superior impingement appeared irrespective of the type of prosthesis.

Discussion and conclusion: This biomechanical study proved the superiority of reversed prostheses compared with anatomic prostheses for massive rotator cuff tears. Medial offset of the center of rotation reduced shear forces on the glenoid. Lateral offset of the humerus increase via a pulley effect the lowering force of the deltoid. Lowering the humerus pulled on the muscle fibers of the deltoid and increased their performance. Dosing these three variables with an appropriate («ideal») design for the reversed prosthesis would optimize deltoid performance in patients with deficient rotator cuffs.

Purpose: Surgical prosthetic treatment of excentred degeneration of the shoulder joint has long been limited to simple humeral prostheses. Inverted total shoulder arthroplasty (ITSA) using the Grammont system appears to provide real improvement. What is the comparative outcome? Does the Grammont prosthesis give constant results? Is the simple humeral prosthesis still indicated for selected cases? In order to respond to these questions, we reviewed 84 patients who had undergone surgery between 1986 and 2000 for implantation of 52 simple humeral prostheses (mean follow-up 7.1 years) and 32 ITSA (mean follow-up 4.3 years).

Material and methods: Between 1986 and 1995, all patients with excentred degeneration of the shoulder joint were treated with a simple humeral prosthesis (36 cases). Between 1995 and 2000, either ITSA or simple humeral prosthesis (16 cases) was used. For the ITSA, indications were limited to excentred degeneration in patients aged over 70 years with a painful pseudoparalytic shoulder or contraindication for a simple prosthesis (alteration of the acromiocorocoid vault or the subscapular area). Simple prostheses were used either when an ITSA could not be implanted (patient too young, overly deteriorated glenoid, deltoid insufficiency) or when there was a preferential indication (excentred degeneration in young patients < 65 years, preservation of active motion, Favard type 1 or 2 glenoid…).

Results: For the simple humeral prosthesis series (1986–1995): the Constant score improved from 16.4 preoperatively to 489.6 postoperatively; radiographically there were one humeral loosening, three superior migrations, and three glenoid degenerations; three revision procedures were required. For the ITSA: the Constant score improved from 14.2 preoperatively to 61.6 postoperatively; radiographically there was one humeral loosening; 32% glenoid alterations in Nérot grades 3 and 4; one revision procedure was required. For the later simple humeral protheses (1995–2000): the Constant score improved from 18.2 preoperatively to 56.4 postoperatively (60.2 excluding cases with a simple prosthesis due to contraindication of an ITSA); radiologically there were one superior migration and two glenoid deteriorations.

Discussion: For a given indication, the results with the ITSA were better than with the simple prosthesis (p < 0.001). However, excluding salvage procedures where the ITSA could not be used, simple prostheses inserted for precise indications provided results nearly equivalent to the ITSA (p < 0.05) with less uncertainty concerning the future. For us, the inverted total shoulder prosthesis and the simple humeral prosthesis share indications for surgical repair of excentred degeneration of the shoulder joint.

Purpose: Revision of a humeral prosthesis or a total shoulder arthroplasty is a difficult therapeutic challenge. The purpose of this work was to report our experience in a retrospective series of 19 inverted Grammont prostheses implanted during revision procedures.

Material and methods: The series included 11 women and eight men, mean age 66.8 years (45–84). Nine patients had already undergone at least two procedures before the revision of their prosthesis. There were 19 revisions of humeral prostheses and four revision total shoulder arthroplasties. Causes leading to revision of the humeral prostheses were: infection one patient, secondary cuff tears six patients, anterosuperior dislocation of the prosthesis five patients, and stiff and painful shoulders three patients. For the total shoulder arthroplasties, three required revision for a loosened glenoid implant and two for secondary cuff tears with ascension of the humeral head. The same access was used as for the first procedure in most of the cases (deltopectoral or anterolateral approach). The rotator cuff was torn and retracted, or fibrous and non-functional. Ablation of the humeral implant and the cement led to fracture (error or infection) in six cases requiring cerclage for fixation. A complementary bone graft was needed in three cases at the glenoid level in three and at the humeral level in three others.

Results: The Constant score at mean follow-up of 36.4 months (12–60) was 55.07 (44–87) a clear improvement over the initial 13.87 (6–39) before revision. The score was less favourable for patients with several operations before revision. Mean anterior elevation was 29.33° (20–100) pre-operatievely and reached 118.67° (90–160) postoperatively. The pain score was 3/15 preoperatively and13.66/15 postoperatively.

Discussion: Other alternatives may be indicated in salvage situations with an unreparable cuff and a pseudo-paralytic shoulder: rectus or deltoid flaps, reconstruction of the acro-miocoracoid arch and glenohumeral arthrodesis.

Conclusion: The inverted Grammont prosthesis appears to be a useful salvage technique, particularly when the acro-miocoracoid arch is destroyed and the cuff is unreparable. The best chances of success are observed for revision of hemiarthroplasties of a pseudo-paralytic shoulder with complete passive mobility and an anterosuperior dislocation of the humeral prosthesis head.