Purpose of the study: Glenohumeral dysplasia is an uncommon cause of degenerative disease of the shoulder joint. In this context, arthroplasty is a therapeutic challenge due to the bony deformations.

Material and methods: Between 1998 and 2004, simple humeral prostheses were implanted in eight shoulders (seven patients, two men and five women, mean age 49.5 years). There was no procedure on the glenoid cavity. A Neer II was used for four shoulders (two dysplasic cases with short 63 mm stems) and a Neer III for four shoulders.

Results: At mean follow-up of 4.5 years (maximum 7 yers) the Neer outcome was satisfactory for five patients and non-satisfactory for two. Five of the seven patients were satisfied with their operation. The Constant scores improved: from 3.5 to 11.8 for pain, 9.8 to 16.6 for activity, and 13.8 to 24.4 for active mobility. Active anterior elevation was 114° on average, external rotation 25°, and internal rotation at level L3. The overall constant score was 52.8 points with a weighted score of 43%. Radiographically, there were no lucent lines around the humeral implant. Anterior dislocation occurred in one shoulder six months after the initial operation. Capsuloligament revision was performed but the implant was left in place. For one other shoulder, secondary rotator cuff tears limited the function outcome, but the prosthesis was not revised.

Discussion and conclusion: The results were average, but did allow our patients to resume nearly normal activity without pain. Looking at the failures in this small series suggests that the status of the rotator cuff is the main prognostic factor. Neither glenoid deformation nor the lack of replacement appeared to have an effect on the final outcome. Deformation of the proximal end of the humerus may require use of a shorter stem which should be available at the time of the operation.

Purpose of the study: The proximal humerus is a common localization for solitary endchondroma. Levy (Clin Orthop2004, 431) emphasized the frequency of associated muscle and tendon disease. Treatment is generally curettage-autograft filling. Use of calcium phosphate bone substitute has been validated (A. Uchida et al. J Bone Joint Surg (Br) 90, F. Gouin Rev Chir Orthop 95, R. Mirzayan J Bone Joint Surg (Am) 2001). This retrospective analysis was conducted to determine the signs and symptoms and report the results of surgical treatment obtained in a consecutive series of 15 patients with metaphyseal enchondroma treated in the same unit.

Material and methods: This series included twelve women and three men, mean age 48.2 years (range 38–73). All complained of pain. Two also had signs of calcification and six presented a cuff tendinopathy. Eight had had one or more joint injections. On average, the enchondromas measured 3.1 cm on the ap view and 3.6 cm on the lateral view. Magnetic resonance imaging (MRI) demonstrated the presence of a subacromial effusion in 13/16 shoulders, supraspinatus tendinopathy in six, calcifications in three, and acromioclavicular arthropathy in three. Curettage was followed by filling with biphased tricalcium phosphate (SBM, Lourdes) associated in nine shoulders with acromioplasty-bursectomy and in two with resection of a calcification.

Results: There were no postoperative complications. Mean follow-up was six months. All patients recovered joint motion, seven were pain free, six complained of pain at exercise and two had episodic pain. There were no local signs of substitute intolerance. Follow-up was greater than one year in 12 patients and greater than two years in eight: seven shoulders were pain free, three presented pain at exercise, and two required analgesic drugs. Radiographically, the limit between the bone substitute and the cancellous bone was imprecise; the bone substitute could not be readily visualized in four shoulders, had faded out in three, and was visible in five.

Discussion: The association of enchondroma and a rotator cuff pathology is common suggesting the tumor could affect disease expression. Imaging provides strong arguments favoring a benign disease. Use of bone substitute for filling is reliable and avoids the need for an iliac graft.

Conclusion: A fortuitously discovered or painful enchondroma of the humerus should be treated by curettage-filling with bone substitute as soon as the nature of the tumor has been clearly identified and/or strong uptake on scintigraphy visualized. This is a supplementary operative argument suggesting an associated cuff pathology.

Purpose: Non-traumatic osteonecrosis of the humeral head is an uncommon condition. Arthroplasty remains the treatment of choice when the head has lost its spherical shape. The purpose of this work was to assess clinical and radiological results in a monocentric consecutive series of 19 shoulders.

Material and methods: Twelve men and five women, mean age 56 years were treated for humeral head osteonecrosis with a shoulder prosthesis. Seven patients had quadripolar necrosis: two had four prostheses (2 shoulder + 2 hip), three had three prostheses (1 shoulder + 2 hip), and two had two prostheses (1 shoulder + 1 hip). Osteonecrosis was idiopathic in six cases, related to corticosteroid treatment in ten, radiotherapy in two, and Gaucher disease in one. The Arlet and Ficat classification showed stage II=3, stage IV=13 and stage V=3. There were rotator cuff tears in two patients. Fourteen simple humeral prostheses and five total shoulder prostheses were implanted.

Results: At mean seven years follow-up (2–12), the Neer classification showed excellent outcome in seven shoulders, satisfactory outcome in nine, and unsatisfactory outcome in three. Eighty-nine percent of the patients were satisfied. The Constant pain score improved from 1.5 points to 11.5 points, the activity score from 7.2 to 14.6, and motion from 15 to 27. Active anterior flexion was 120°, external rotation 34°, internal rotation at L3. The rough Constant score was 58 and the weighted score 78%. There were no lucent lines around the humeral implants, but a complete line was present around two glenoid implants which were radiographically loosened. These two shoulders presented glenoiditis with glenoid wear. No surgical revisions have been performed to date.

Discussion: Shoulder prosthesis has given satisfactory results for osteonecrosis of the humeral head with resolution of pain in 80% of patients. Shoulder motion remains limited in relation with the often significant preoperative stiffness. Results have been better in patients with less advanced disease preoperatively and preserved motion. In our series, poor results were found in patients whose osteonecrosis was radiation-related.

Purpose: The results of coracoid blocks for the treatment of anterior shoulder instability are satisfactory (SOFCOT 1991 and 1999). Little work has been devoted to complications of this operation. We report our results after surgical revision in a consecutive series of seventeen patients.

Material and methods: Fourteen men and three women, mean age 34 years, required a second interention after anterior coracoid block: for recurent instability in ten cases and for painful stiff shoulder in seven. Radiographic assessment demonstrated a conflict between the block and/or the screw with the humeral head in 13 cases and signs of joint degeneration in three. The block was malpositioned in eight shoulders. The stabilisation procedure (Bankart ± capsuloplasty) was performed for these unstable shoulders with debridement and removal of the screw and or the block for painful stiff shoulders. The subscapular tendon was normal preoperatively in two shoulders, fibrous or thin in eleven and torn in one. The time from the first operation to revision was eleven years on the average.

Results: At mean follow-up of 21 months, the patients were assessed with the Duplay score. Results were good or excellent for eleven patients (70% for stabilised unstable shoulders and 57% for debrided painful stiff shoulders), fair for four and poor for two. Clinical assessment of the subscapular demonstrated a deficiency in ten shoulders. The force of internal rotation of the operated shoulder was 3.3 kg less than for the controlateral shoulder. Computed tomography demonstrated significant fatty degeneration of the subscapular in four patients. Glenohumeral joint degeneration was observed in nine shoulders. The most significant prognostic factor for final outcome was the number of prior interventions (p< 0.01).

Discussion: The result of revision surgery after coarcoid block depends on the clinical presentation. While the results for painful stiff shoulders remained very limited due to frequent intra-articular lesions, adjunction of a capsuloplasty with or without reinsertion of the glenoid rim yielded satsifactory results in more than two-thrids of the unstable shoulders. Involvement of the subscapular muscle appears to be related to multiple interventions in these shoulders, as well as to the deleterious effect of the block (Picard 1998, Glasson 1999) and continues to be the crucial prognostic factor for final outcome.

Purpose: Reaming is classically contraindicated for open leg fractures. For certain authors, reaming can favour bone healing without increasing the risk of infection (Court-Brown JBJS 90B and 91B, Wiss Coor 95). The aim of this retrospective analysis of patients treated in a single centre was to validate these notions and determine the role of locked centromedullary nailing (LN) with reaming for the treatment of open leg fractures.

Material and methods: Between 1989 and June 2000, 141 open leg fractures were fixed with locked centromedullary nailing without reaming in 103 men and 38 women, mean age 34 years, who were mainly accident victims (2-wheel vehicles 43%, 4-wheel vehicles 22%). Multiple trauma was present in 18.7% of the cases and multiple fractures in 28%. Skin wounds were (Gustilo classification): type I 81 (57%), type II 38 (27%), type IIIA 14, and type IIIB 8. There was a simple fracture in 50% of the cases, a wedge fracture in 32%, and comminution in 18% with bifocal fractures in 10 cases. Osteosynthesis was performed within a mean 5.5 hours (2–18) and deferred in six cases. The Grosse and Kempf nail was used in all cases with reaming (man 11). Static locking was used in 88% of the cases. type I, II and IIA skin wounds were sutured after debridement. Three aponeurotomies were performed as preventive measures. Type IIIB wounds were treated by early plasty. A brief antibiotic prophylaxis was given in all cases.

Results: There was one aggravation of the comminution, two dismantelings subsequent to unauthorised weight-bearing, three compartment syndromes and one lateral sciatic popliteal paralysis. Two patients died from their multiple injuries. Four patients developed infection: two healed without removing the nail, one was amputated (free flap failure). One patient consulted another unit. Ten patients who were not residents of our area were lost to follow-up. Dynamisation was performed in 31 patients (25.6%) at a mean 4.4 months. Four patients with delayed healing cured after a new nailing with secondary reaming. Delay to bone healing was related to the type of fracture (p < 0.01): 4.2 months for type A (AO classification), 5.2 months for type B and 5.9 months for type C. Bone healing was correlated with Gustilo type (p < 0.05): 4.5 months for types I, 4.6 months for type II, 5.8 months for types III. Six patients developed nonunion: four were revised with success after a new nailing and secondary reaming (two lost to follow-up). Delayed healing and non-union were related to type of fracture (A = 3.8%, B = 15.6%, C = 18%) and soft tissue damage (Gustilo I: 4.1%; II: 10.7%; III: 15.8%).

Discussion conclusion: Locked centromedullary nailing with reaming is appropriate when the skin wound is minimal; dynamisation and/or replacement of the nail with secondary reaming should be discussed early in case of delayed healing.

Purpose: We report a series of 16 GUEPAR total elbow prostheses implanted in a single centre.

Material and methods: Between 1988 and 1996, sixteen GUEPAR prostheses were implanted in 13 patients (three bilateral implantations). There were 11 women and two men, mean age 61 years (51–81). Twelve patients (14 elbows) had rheumatoid polyarthritis and one patient (two elbows) had post-traumatic degenerative disease. The V transtricepital approach was used in 15 cases and the Bryan-Morrey approach in one. Postoperatively, the elbows were immobilised at 45° flexion for the normal period (18 days) followed by active mobilisation. Results were analysed with the Mayo Clinic score. The radiographs were examined in search for lucent lines and signs of loosening and prosthetic instability.

Results: At a mean follow-up of four years (2–12), the mean Mayo Clinic score had improved from 33 to 75 points (45–100). Eleven elbows were pain free at last follow-up. Extension and flexion progressed 22° giving a postoperative amplitude of 34° to 129°. Pronation supination progressed by 15° giving a 154° rotation amplitude. The function score improved from 6 to 18 points. Seven of the sixteen elbows achieved normal function. Outcome was excellent for seven elbows, good for one, fair for three and poor for five. In two elbows, instability required changing the ulnar implant. There were four implant loosenings that required revision at 24, 36 ,36 and 48 months after the initial implantation. The radiographic analysis demonstrated a complete lucent line around the humeral and ulnar implant in one case, around the ulnar implant in one case. There were two peroperative fractures of the humerus and on postoperative fracture due to a fall. Ulnar paresthesia was observed in two cases requiring secondary neurolysis in one. There were no infections or secondary injury to the triceps.

Discussion: The GUEPAR prosthesis is a non-constrained prosthesis essentially indicated for rheumatoid polyarthritis. If the intrinsic stability is lost, the implant is contraindicated if there is loss of bone stock or if the instability is major. In selected cases, a generally painless elbow with recovery of the functional amplitude can be achieved with this prosthesis. Nevertheless, the presence of four early loosenings in our series as well as two instabilities suggest this implant should be abandoned in favour of a semi-constrained implant.

Purpose: We report our experience with 23 Coonrad-Morrey total elbow prostheses.

Material and methods: Between July 1997 and February 2001, we implanted 34 Coonrad-Morrey total elbow pros-theses in 33 patients. Twenty-three patients (23 implants) were reviewed at a mean 24 months follow-up, maximum 40 months. There were three men and 20 women, mean age 62 years (42–69). Twelve patients had rheumatoid polyarthritis, the principal indication. There were also four recent fractures of the distal humerus, two nonunions, and one patient with post-traumatic osteoarthritis. One patient had sequelar osteoarthritis since childhood. Finally three revisions were performed for loosening of a GUEPAR prosthesis in two cases and a GSBIII prosthesis in one. Results were assessed with the Mayo Clinic score. We searched for lucent lines around the implants, polyethylene wear, and incorporation of the bone graft behind the anterior wing of the implant on plain radiographs.

Results: At last follow-up, the mean Mayo Clinic score had improved from 25 to 89 points (70–100). Before surgery, 17 patients had severe pain. At last follow-up, eight patients had occasional pain. Extension was improved by 10°, flexion by 27° giving a postoperative amplitude of 29° to 132°. Prona-tion supination progressed by 37° giving a rotation amplitude of 127°. The function score improved from 4 to 21 points. Sixteen of the 23 patients had normal elbow function. Outcome was excellent in 13 patients, good in eight, and fair in two. There were no lucent lines visible on the radiographs. There was no sign of polyethylene wear. The bon graft was incorporated behind the implant in 20 cases and was not visible in three. Complications included one peroperative fracture, one cutaneous dehiscence, one post-operative fracture of the olecranon due to a fall, and persistent ulnar paresthesia in four patients requiring secondary neurolysis in one.

Discussion, conclusion: The Coonrad-Morrey semi-constrained prosthesis provides a response to a large range of situations. The dominant indication is rheumatoid polyarthritis, but trauma patients can benefit from this reliable therapeutic solution giving a satisfactory rate of success. A satisfactory functional amplitude is generally achieved with this implant and the elbow is generally pain free.

Purpose of the study: A retrospective series of 40 patients who underwent simultaneous intramedullary nailings for bilateral femoral shaft fractures was analyzed. The aim of our study was to verify that simultaneous nailing without reaming does not increased risk of fat embolism and to assess clinical and radiological outcome.

Material and method: This series included 27 men and 13 women, mean age 27.8 years, who underwent first intention intramedullary nailing between 1986 and February 1999. Thirty-two patients had multiple fractures. Mean ISS was 23 (range 9 to 59). Among the 80 femoral shaft fractures, 15 were open fractures, 3 were associated with sciatic paralysis, and 4 were complicated by an interruption of the femoral vessels. The AO classification was: type A = 44; type B = 25; type C = 11. Mean delay to simultaneous centromedullary nailing was 3.8 days: surgery was performed on the day of arrival for 25 patients. General anesthesia was used in all cases with respiratory assistance (FIO2 = 50 to 100 p. 100). Mean nail diameter was 11.6 (range 10–14). Gurd criteria and PaO2 were followed to assess pulmonary function. Clinical and radiological outcome was assessed using the modified Thorensen criteria.

Results: Preoperatively, PaO2 was < 87 mmHg in 8 patients. Four of these patients showed a discrete drop off and three improved well above the normal level. Only one patient experienced an important decrease but did not develop respiratory distress. Among the 32 patients with a normal level preoperatively, PaO2 remained in the normal range in 18, fell to a limit level but below 87 mmHg in 4, and showed a substantial drop off of 46 to 172 mmHg in 10. Two of these 10 patients developed respiratory distress due to fat embolism which was fatal in one case. One other patient died in the immediate postoperative period of an undetermined cause. All of the other patients recovered normal gas levels within a few hours or days. There were four cases of phlebitis, including one with pulmonary embolism, one case of respiratory distress by pulmonary superinfection, and one case of septicemia. Both femoral fracture sites became infected in one patient. Malunion occurred in two cases. Two vascular repairs of the femoropopliteal axis were unsuccessful, leading to above knee amputations. Thirty-four patients have been examined after a minimal 12 months follow-up (mean 30 months). Outcome was excellent for 48 femurs, good for 10 and fair for 10.

Discussion: This continuous series of simultaneous bilateral femoral shaft intramedullary nailings appears to be the only such report to date. The clinical and radiological outcomes were comparable with those achieved in one-side femoral fractures. The risk of fat embolism is inevitable after long bone fractures. Many factors favoring the risk are recognized, the most important being delay to fixation. Reaming creates excessive pressure in the medullary canal and could thus contribute to the risk. The presence of an associated chest trauma is not a formal contraindication if effective hematosis is preserved as evidenced by the blood gases.

Conclusion: Simultaneous nailing of bilateral femoral shaft fractures can be performed if blood gases remain acceptable and minimal reaming is used.