Purpose of the study: A stiff shoulder in internal rotation is a classical complication of obstetrical injury to the brachial plexus. The condition generally associates glenohumeral dysplasia. In 2003, Pearl proposed arthroscopic release of the shoulder, with or without latissimus dorsi transfer as an alternative to an open procedure. We report a series of 13 patients who underwent this procedure.

Material and method: From 2004 to 2007, 13 children aged 1 to 11 years (range 3.5) underwent surgery. The procedure was an arthroscopic tenotomy of the intra-articular portion of the subscapularis associated with release of the anterior capsuloligament structures without tendon transfer. A thoracobrachial resin cast maintained the shoulder in maximal external rotation for six weeks. The functional outcome was assessed on the basis of the passive and active range of motion and the Mallet score. Pre and postoperative MRI was used to assess glenohumeral dysplasia and check its correction.

Results: Mean follow-up was 23 months (range 5–40). There were no complications. Mean immediate postoperative gain was 53 (range 30–70) for external rotation (RE1). On average, passive RE1 improved from −5.4 to +57.7. Eight patients (61%) recovered active external rotation (57.5 on average). Mean active abduction improved from 45.8 to 56.5. Active internal rotation declined in 38% of the shoulders. The Mallet score improved in 69% of the shoulders. The best results were observed for children aged less than 4 years. There was one failure, related to major dyplasia which could not be reduced with an open procedure. Correction of the glenohumeral dysplasia was noted on 7 of the 9 MRI performed and the retroversion angle of the glenoid improved from −28.2 to −25.6, on average.

Discussion: Open techniques are aggressive and only variably effective. A majority of our patients (8/13) achieved improved passive and active RE1 solely with the arthroscopic procedure. Correction of the glenohumeral dysplasia with growth appears to correspond to the clinical improvement but further follow-up is needed.

Conclusion: Our results are comparable with those reported by Pearl. This is a minimally invasive easy-to-perform technique. For children aged less than 4 years, systematic tendon transfer does not appear to be necessary.

Purpose of the study: Arthroscopic repair of rotator cuff tears is a painful technique. We conducted a randomized controlled trial to analyze prospectively the level of postoperative pain after this procedure and to assess the efficacy of three anti-pain techniques.

Material and method: This prospective randomized study included 140 patients attending two centres specialized in shoulder surgery who underwent arthroscopic rotator cuff repair from January to November 2008. Preoperatively, patients were assigned to three anti-pain techniques: interscalenic catheter (KTIS) (n=44 patients), subacromial catheter (KTSA) (n=64 patients), unique subacromial injection of 2% ropivacaine (INJ) (n=32 patients). The procedure was performed under general anaesthesia in all cases. Criteria used to evaluate the efficacy of the anti-pain techniques were: pain as assessed by a visual analogue scale (VAS) the day before operation to the 30th postoperative day; daily consumption and total dose of opiates; subjective satisfaction with pain management assessed on day 30.

Results: The pain profile was the same as demonstrated last year with a painful displacement on day 1 and progressive decline in pain through day 30 when the pain level was lower than preoperatively. Two risk factors were noted: isolated distal tears of the supraspinatus and concomitant acromioplasty. On day 0, the VAS was ≤ 1.7 in the KTIS group versus 3.6 and 3.5 respectively in the KTSA and INJ groups (p< 0.05). On day 1, the respective values were 3.1, 3.5 and 3.8 (NS). On day 2 and beyond, there was no difference between the three groups. Overall morphine consumption on day 0 was 10.8 mg in the INJ group versus 6.9 and 2.1 mg in the KTSA and KTIS groups, with a significant difference between the INJ and the KTIS groups. There was no difference beyond day 1. Patient satisfaction with pain management during the first 30 postoperative days reached 94.4, 82.9 and 84% respectively for INJ, KTIS and KTSA, with no statistically significant difference between the groups.

Conclusion: The interscalenic catheter technique is more effective than the two other techniques. The subacromial catheter is not better than a single injection of local anaesthetic and is associated with greater risk, suggesting its use should be revisited.

Purpose of the study: Several series have been reported on arthroscopic treatment of anterior instability. Few authors have focused on patient outcome after recurrent instability following arthroscopic stabilisation. Did these patients undergo revision surgery? What proportion? What were the results of secondary surgical coracoids block?

Material and methods: This was a retrospective analysis of 53 failures after arthroscopic stabilisation collected among a cohort of 182 patients who underwent surgery in our institution between 1988 and 2006. At last follow-up, shoulder function was noted using the Walch-Duplay score. The degree of joint degenerative disease was noted on the radiographs using the Samilson classification.

Results: Mean time to recurrence after arthroscopic stabilisation was 21 months (range 3–114). Patients were reviewed at mean 68 months. Twenty-four patients (45%° had not had revision surgery: 17 (32%) had declined a new operation and 7 (13%) had a unique episode of instability. Twenty-nine patients (55%) had revision surgery: 27 underwent an open procedure in our institution for a coracoids block. The revision was performed in another institution for two patients who were excluded from the analysis. Mean time between the two operations was 29 months. At last follow-up, 89% of the reoperated patients were satisfied. The mean Walch-Duplay score was 83.6/100 (activity=18.5; stability=15.9; pain=23.9; mobility=24.2). The Duplay score was 100 for 48% of the reoperated patients; 41% had persistent apprehension. Three patients (11%) developed recurrent dislocation at a mean 23 months (19–29). Among the 53 patients included in the study, 26% had moderate osteoarthritic lesions (Samilson 1 or 2). The reoperated patients were free of such lesions. Hyperlaxity, age, and sport practiced did not have any impact on surgical revision.

Discussion: In this overall series of 53 patients, 20 (37%) retained an unstable shoulder. Among them, 17 had declined new surgery. Eleven percent of the reoperated patients developed subsequent recurrence. This rate is higher than after first-intention blocks. Published series of arthroscopic revisions reported a higher recurrence rate (Kim, Arthroscopy 2002: 21 % recurrence; Neri, JSES 2007: 27 % recurrence).

Conclusion: The Latarjet block is the treatment of choice after failure of arthroscopic stabilisation, despite a high recurrence rate.

Purpose of the study: There are several goals for the treatment of septic nonunion of the leg: control the infection, achieve healing, preserve function. The purpose of this work was to report the results obtained with a two-phase technique using a cement spacer.

Material and method: From 1994 to 2007, 27 patients were treated for a septic nonunion of the tibia (19 proven, 8 suspected). There were 22 women and 5 men, mean age 39 years (range 16–66). The first phase of the surgical technique involved “cancerological” cleaning and insertion of an antibiotic cement spacer. Osteosynthesis was performed if necessary. Antibiotics were adapted to sample results. The second phase involved an autologous bone graft with osteosynthesis after biological markers had returned to normal and an antibiotic window. Antibiotics were then discontinued if samples were negative. Patients were reviewed with physical examination, radiology, and laboratory tests at one year.

Results: Mean follow-up was 4 years (range 1–11). At the first phase, mean bone defect after cleaning was 5 cm (range 3–8); osteosynthesis procedures were required for 22 patients (81.5%), mainly with plate fixation. Mean time to the second phase was 4 months (range 1.5–22). At the second phase, bone loss was filled with isolated bone fragments (44%) or associated with a tricortical graft (52%) or a plate nail combination (37%). Bacteriological samples were negative for 25 patients after the second phase. Six patients required surgical revision for recurrent aseptic non-union (22%).

Discussion: All patients healed at mean one year with a tolerable misalignment in 37%. At last follow-up there were no cases of infection. Nineteen patients had residual stiffness of the ankle or knee but 80% had resumed their sports activities and 85% their occupational activities.

Conclusion: A two-phase surgical treatment of septic non-union of the leg is effective. We were able o achieve cure of the infection in all patients with per primam healing in 78% in addition to an acceptable functional outcome. The spacer offers the advantage of preparing a bed for the graft and preserving autonomy between the two phases.

Purpose of the study: The PFM-R (Zimmer) prosthesis is a straight modular stem made of sanded titanium designed to favour spontaneous bone reconstruction. The purpose of this work was to analyse the clinical and radiographic results of this implant.

Material and methods: This was a consecutive prospective series of 154 patients who underwent surgery from 1998 to 2007 (15 first-intention prostheses and 139 revisions for severe loosening [Paprosky stages 3 and 4]). Revision included a clinical evaluation (PMA score) and radiographic assessment (migration, bone regeneration, stress shielding, osteointegration, Le Béguec score) as well as a survival analysis.

Results: At mean 4.6 years follow-up (1–10 years), three patients had died, two were lost to follow-up and 18 implants were removed, 11 for infection, six for migration and one for defective technique. The PMA score improved from 8 (0–16) to 15.8 (5–18). Mean pivot impaction was 4 mm (0–50), statistically dependent on initial bone stock, form of the isthma, the corticomedullary index in the implantation zone, length of anchor, and time to weight bearing, but not femorotomy nor zone of primary stability. Bone stock was good in 73% at poor in 27% (15 stress shielding, nine infectious osteolysis, 16 absence of bone regeneration). Stress shielding was related to length and diameter of the implanted pivot (p< 0.05). Bone regrowth was statistically dependent on the number of prior operations, type of stem explanted (cemented), initial bone stock, form of the isthma and quality of the surgical reconstruction. The implant was osteointegrated in 128 cases (86%). The analysis of the prosthetic anchoring showed that primary stability was mainly diaphyseal (90%) then secondarily global (83%). Osteointegration depended statistically on the number of prior operations, initial bone stock, form of the isthma and bone regrowth, but not femorotomy. The overall Le Béquec score reached 14.7 (2–20) at last follow-up.

Discussion: The PFM-R enabled bone regeneration and osteointegration in the majority of patients. The quality of the femoral reconstruction around the implant appears to be fundamental. Massive stems should be avoided as they lead to stress shielding. The limits for use of this implant are osteopenia and absence of an isthma.

Introduction: Treatment of acetabular defects can be difficult, especially in case of roof destruction. Since 9 years, we use a variant of Paprosky’s technique which consists in rebuilding the roof by structural allograft and acetabular reinforcement ring. The purpose of this study is to present this technique and the follow up results.

Patients: This retrospective study concerns 21 patients (23 hips) with severe acetabular bone loss (8 cases of stage 2 and 15 cases of stage 3 of Paprosky): 4 septical and 19 aseptical loosening. Between 1998 and 2005, all patients were operated with the same surgical technique using an allogeneic structural allograft (femoral head or distal femur) and an acetabular reinforcement ring (20 of KERBOULL, 3 of GANZ) associated with a cemented PE cup.

Method: Review included a clinical and X-ray evaluation (analysis of the refocusing of the hip, the positioning and the stability of implants and the graft incorporation).

Results: Mean duration of follow-up is 3,5 years [1–8,3]. Preoperative PMA score rised from 6,6 [0–12] to 15,8 [12–18] in postoperative. There was no peroperative complication. After surgery, 2 cases of early hip dislocation required PE block; 2 cases of sepsis were treated, one by washing and one by a surgical revision. In 60% of cases, immediate total weight bearing was allowed.

The immediate postoperative X-rays showed that the rotation center of the hip was 5,2 mm [0–10] far from the ideal rotation center (26% of cases: 0 mm) and the PE cup was implanted with a lateral inclination of 42,5° [30–55]. In postoperative X-ray follow up, one case of acetabular aseptic loosening was found which didn’t need hip revision. In all other cases no modification of implants position neither of hip rotation center was noted. In 79% of cases, we had total graft incorporation; in 17% of cases, an non evolutive radiolucent area between graft and bone and in 4% of cases (loosening) a graft migration.

Conclusion: The use of a structural allograft combined with acetabular reinforcement ring allows hip reconstruction in severe acetabular bone loss with good medium term results.

Purpose of the study: Orthopedic treatment is generally proposed for minimally displaced fractures of the scapula. Surgery is indicated in the event of medialization or disorientation of the glenoid cavity. The purpose of this study was to identify rules for management of problematic cases.

Material and methods: This retrospective study included 27 patients, 22 men and five women, mean age 46.7 years (range 22–83 years). The fracture was limited to the scapula in 17 and was part of thoracic syndrome in ten. Injury to the plexus was noted in three patients. Associated lesions were noted in 55% of patients. Orthopedic treatment was proposed for 15 patients (group 1) and surgical treatment for 12 (group 2): neck osteosynthesis (n=4), clavicle fixation (n=5), combined osteosynthesis (n=3).

Results: Mean follow-up was 47.3 months; 21 patients were reviewed (four lost to follow-up, two deaths). Per primam healing was achieved for all fractures. There were no complications related to the surgical procedure in group 2 and no secondary displacement in either group. Mean time to resumed occupational activity was four months (range 0.2–25 months): 2.2 months (1–5 months) in group 1 and 5.5 months (0.2–25 months) in group 2. The age and gender weighted Constant score for the overall series was 96.2% (range 80–100%) with 81% excellent and very good functional outcomes. In group 1, the Constant score was 95.4% (81–100%) with 70% excellent and very good functional outcomes. It was 97% (80–100%) in group 2 with 91% excellent results. At last follow-up, there was only one case of Samilson grade 2 osteoarthritis.

Discussion and conclusion: The functional outcome observed after treatment of scapular fractures is satisfactory. Orthopedic treatment should be reserved for non-displaced fractures. If there is significant medial offset of the glenoid cavity or instability, osteosynthesis of the scapular neck is indicated, possibly with reinforcement by fixation of the clavicle. Isolated osteosynthesis of the clavicle stabilizes the scapular ring but does not reduce the scapular fracture.

Purpose of the study: Resection of the lateral quarter of the clavicle is an effective treatment for painful acromio-clavicular arthropathy. An open procedure can create a disgraceful or painful scar or be associated with secondary instability and muscle weakness. Arthroscopic resection would avoid these complications.

Material and methods: Between 1992 and 2002, 35 patients were treated arthroscopically for isolated painful acromioclavicular arthropathy after failure of medical treatment. Twenty-seven patients (mean age 44 years) were reviewed clinically (Constant score, subjective outcome, stability) and radiographically (quality of resection, coraco-clavicular space, ossification) at more than two years follow-up. Acromioplasty was performed in all cases to improve exposure in 14 or because of an aggressive acromion (n=7). A scarf orthesis was worn to prevent pain. Self-controlled rehabilitation exercises were proposed.

Results: There were no postoperative complications excepting one case of retractile capsulitis in the context of an occupational accident. Excluding this case, mean sick-leave was five weeks (range 3–20). At mean follow-up of seven years, all scars were pain free and minimally visible. There was no problem with frontal or sagittal instability. The Constant score improved 24 points on average compared with the preoperative score with a significant gain for pain (+9 points). Two patients were disappointed: one retained a sequellar capsulitis and one persistent pain after insufficient resection. On average, the resection measured 10 mm (range 6–20 mm). Three patients had an insufficient posterosuperior resection. The coracoclavicular interval remained unchanged in all cases and four patients presented secondary ossification of the resection zone. Two patients were sensitive to acromioclavicular palpation with a positive cross arm test (one had an insufficient resection and the other ossifications at last follow-up).

Discussion and conclusion: Arthroscopic acromio-clavicular resection reduced operative morbidity. The operation does not destabilize the joint. The total joint surface area can be resected to avoid a residual pos-terosuperior impingement which would be a source of persistent pain. The presence of secondary ossifications in the zone of resection has led us to propose NSAID treatment although the prophylactic effect remains to be demonstrated.

Purpose of the study: The revision modular femoral prosthesis (PFM-R) (Centerpulse) uses a cone-shaped straight femoral stem with winglets for press-fit revision fixation in the femoral shaft. The risk of secondary migration further into the shaft is well known. The purpose of this paper was to assess the degree of PFM-R stem migration and to search for predictive factors in order to better ascertain the limits of this type of revision anchorage in the femoral shaft.

Material and methods: Fifty-three files (48 patients) were reviewed retrospectively at minimum none months follow-up. Mean patient age was 56.6 years. Revision was required for loosening (SOFCOT classification): grade 1 (n=5), grade 2 (n=8), grade 3 (n=17), grade 4 (n=10). For 13 hips, the implant to be replaced was not cemented (monobloc stem, infection), or a first-intention PFM-R.

Results: Mean migration was 4.1 mm (0–17 mm), less than 5 mm in 73.5% of cases. There was no correlation between migration and SOFCOT grade. Conversely, there was a significant difference in migration between the grade 1 and grade 3B and between Paprosky grade 1 and 4 (p=0.05). The degree of migration was correlated with the length of the persistent isthma (p< 0.0001), with the morphology of the isthma [conic/inverted conic (p< 0.01), conic/cylindric (p< 0.051)], with the cortical index (p=0.06), the length of the anchor wings (p=0.051), but not with the length of the femorotomy bridge. The length of the wing anchorage was correlated with the length of the persistent isthma (p=0.002) and with the morphology of the isthma [(conic/inverted conic p=0.02), cylindric/inverted conic (p=0.02)], but did not increase significantly with bridging length. There was a trend towards migration in osteoporotic bone (p=0.07).

Discussion and conclusion: Use of a straight stem for anchorage in the femur is associated with secondary migration which depends on the quality of the bone in the anchorage zone and the extent of the press-fit. The persistence of a cone-shaped or cylindrical isthma measuring greater than 4 cm associated with a cortical index greater than 45% and a long wing anchorage can provide excellent primary stability. It would be useless to use long stems since they do not increase the quality of anchorage. The absence of an isthma and the presence of osteoporosis are limitations for this concept.

Purpose: Preoperative morphology must be carefully assessed for proper surgical planning for patients with arthroplasty with massive rotator cuff tears, but many morphological aspects are poorly understood. The purpose of this study was to assess the technical implications of this situation.

Material and methods: We included patients with arthropathy with massive rotator cuff tears who had a complete clinical and radiographic preoperative work-up. We analysed the morphological aspects of the acromion, the humerus and the glenoid cavity.

Results: One hundred forty-two patients (148 shoulders) were included. The acromion presented a fracture or lysis in 13 shoulders and was thinned or had an imprint in 37. It was normal in 70 and hypertrophic in 16. The humerus showed signs of necrosis in 31 shoulders, with a washed out trochiter in 7. Glenoid wear was classed in four stages: E0 or normal glenoid (n=51), E1 or centred wear (n=32), E2 or biconcave aspect (n=46), and E3 or major wear with superior concavity (n=13). Inverted prostheses were implanted in 80 shoulders and non-constrained prostheses in 68. The non-constrained prostheses exhibited progressive ascension of the humeral head in 63% with wear of the glenoid vault. Clinical deterioration led to revision in two patients. The non-constrained prostheses inserted in patients with an E2 glenoid had a significantly lower Constant score (p< 0.05) than the others. A notch appeared in the scapular column in 65%; of the constrained prostheses. This notch was favoured significantly in glenoids classed E2 or E3 preoperatively. The preoperative aspect of the humerus did not appear to affect clinical and radiographic outcome.

Discussion: A thin or lysed acromion associated with an E2 glenoid constitutes a poor indication for non-constrained prosthesis. In this situation, an inverted prosthesis should be used taking care to avoid orienting the glenosphere upwardly, a technically difficult task. Good indications for non-contrained prostheses should probably be limited to shoulders with a normal or thickened acromion and and E1 glenoid.

Purpose: The purpose of this study was to analyse changes in foot dynamics secondary to tibiotalar arthrodesis and examine the adaptation mechanisms induced by wearing shoes.

Material and methods: A 3D gait analysis was performed (Vicon 370) in ten patients with a tibiotalar arthrodesis fixed in a neutral position and in ten matched controls. Recordings were made in three conditions: walking barefooted, walking at a comfortable self-chosen speed wearing shoes, and walking at maximal speed wearing shoes. We measured tibial inclination in the sagittal plane, knee flexion, and the tibia-forefoot angle produced at heel lift-off. The distance of the ground reaction force (GRF) from the heel was measured during the weight-bearing phase and at lift-off. Statistical comparisons were made with the contralateral side and the control group.

Results: Heel lift-off came significantly earlier on the arthrodesis side compared with the contralateral side and with the control group. At heel lift-off, the knee was in complete extension in all three groups. The GRF moved forward more rapidly but remained more posterior on the arthodesis side at heel lift-off in comparison with the contralateral side and the control group. Wearing shoes enabled later heel lift-off on the arthrodesis side and increased tibial inclination at lift-off as well as decreased speed of the anterior displacement of the GRF. The GRF however remained more posterior than in the control group. At maximal walking speed, heelk lift-off came earlier on the arthrodesis side and at the same time a lesser anterior tibial inclination and a more posterior position of the GRF.

Discussion: Early heel lift-off on the arthrodesis side allows anterior inclination of the tibia to continue and to increase stride length. Heel lift-off however occurs when the GRF has not yet advanced to the metatarsophalangeal position, thus increasing stress on the rear and mid foot. Wearing shoes improves the kinematic parameters and decreases stress on the joints below the arthrodesis. Increased walking speed aggravates the perturbed foot dynamics when walking.

Purpose: Management of septic nonunion of long bones is a difficult challenge requiring a multidisciplinary approach. The purpose of this study was to report our results with a two-stage technique using a spacer (Masquelet technique).

Material and methods: Between June 1997 and July 2001, eleven patients were treated for septic nonunion (n=7) or suspected septic nonunion (n= =4). There were seven men and four women: mean age 38 years (26–51). Nonunion involved the humerus (n=1), the femur (n=1), and the tibia (n=9). The same surgical technique was used in all cases: “carcinologic” debridement with gap filling using antibiotic cement and osteosynthesis when necessary, followed by a second procedure two months later to remove the spacer and insert an autologous bone graft when laboratory results had returned to normal.

Results: Mean follow-up was three years (1–5). All patients achieved per primam bone healing within 4.5 months (3–6) despite a mean bone gap of 55 mm (15–100) after avivement. Intraoperative samples taken during the second procedure were negative and there was no recurrent infection or need for revision.

Discussion: This two-phase technique has provided encouraging results in terms of “infectious cure” and bone healing. A standardised approach to the treatment of septic nonunion of long bones as used in our centre should provide data validating this technique.

Purpose: The purpose of this study was to assess long-term functional outcome after arthroscopic acromioplasty for full-thickness rotator cuff tears, to evaluate the efficacy of complementary procedures (biceps tenotomy, extended acromioclavicular resection), and to examine the course of anatomic lesions.

Material and methods: From 1988 to 1994, 141 full-thickness rotator cuff tears were treated by arthroscopic acromioplasty. Ninety-eight patients, mean age 60 years, were reviewed clinically, radiographically, and sonographically at more than eight years. The mean preoperative Constant score was 48.5 points. The tear involved the supraspinatus in 18 cases, the supraspinatus and infraspinatus in 40, the supraspinatus and the subscapularis in ten, and all three tendons in 20. Coronal extension showed a distal tear in four, an intermediate tear in 52, and a retracted stump in 32. Systemic acromioplasty was associated in 36 patients with tenotomy of the long head of the brachial biceps and in 44 with acromioclavicular extension.

Results: At mean follow-up of 10.7 years (8–13.5), the mean Constant score was 60 points. The clinical outcome was excellent or good in 39.7%, fair in 45.5%, and poor in 14.8. 62.5% of the patients were satisfied or very satisfied. The height of the subacromial space was 5.19 mm at last follow-up compared with 4.3 mm preoperatively. The antalgesic effect of biceps tenotomy was significant when the initial acromiohumeral space was less than 5 mm. This procedure did not produce any significant change in the subacromial height or development of osteoarthritis at last follow-up. Sonography showed stability of the size of the tear in 83.8% of the cases. The results were significantly less satisfactory when the initial tear involved the subscapularis or all three tendons.

Conclusion: The clinical results of arthroscopic acropmioplasty for full-thickness tears show long-term stability. Biceps tenotomy improves the antalgesic effect significantly, particularly if the subachromial space measures less than 5 mm preoperatively, without causing significant radiological degradation. Extension to the acromioclavicular level should be systematic in patients with acromioclavicular pain preoperatively and/or radiological anomalies.

Purpose: In 1988, the conclusions of the SOFCOT symposium appear to condemn use of cemented implants for revision of loosened femoral elements. Eleven years later, at the 1999 Symposium, Vidalain demonstrated that cementing remains a reasonable alternative. The purpose of this retrospective analysis was to estimate long-term results of revisions using a long cemented stem without bone stock reconstruction.

Material and methods: From 1987 to 1995, 135 patients (135 hips) underwent revision surgery with this technique. We retained for analysis 95 hips (15 lost to follow-up 25 deceased patients) in 66 women and 29 men, mean age 70 years at revision (42–86). The preoperative Postel Merle d’Aubigné (PMA) function score was 8/18. Femoral implants in place had been cemented in 80% of the hips. Using the SOFCOT criteria, 65% of the loosenings were stage II, 29.5% stage III. All were aseptic loosenings and the same revision technique was used in all cases: removal-replacement of both prosthetic elements, use of a long cemented stem without associated grafting.

Results: Mean follow-up was eight years (60–157). Thirty nine percent of the patients had early postoperative complications. The mean function score at last follow-up was 14.8/18 with 62.4% of the outcomes considered good or very good. Age, restitution of the rotation centre, quality of cementing, and stem/femur fit influenced the result significantly. Radiographic analysis showed a progressive lucent line in 32% of the hips, only 36% of the femoral stems were totally free of lucent lines. Cumulative actuarial survival was 87% at 14 years and fell to 65.5% considering certain radiographic loosening as the endpoint.

Conclusion: Use of a long cemented femoral stem for revision total hip arthroplasty provides acceptable long-term functional outcome. Results are significantly affected by imperfect technique. This type of implant, which destroys any hop for restoration of bone stock, should be reserved of elderly subject where a more “ambitious” procedure were be too risky.

Purpose: The purpose of this study was to analyse outcome of shoulder prostheses after radiotherapy, to define a specific clinicoradiological entity, and evaluate incidence of complications.

Material and methods: Fourteen shoulder prostheses were implanted in 13 women who had been treated for breast cancer with complementary radiotherapy and one man treated for Hodgkin’s lymphoma. The time from radiothearpy to implantation was 16 years, seven months. Two forms were identified on the preoperative x-rays: seven cases with typical avascular osteonecrosis according to the Arlet and Ficat classification, and seven cases with a radiographic presentation of arthritis or degenerative disease. Humeral prostheses were used in five cases and a total shoulder arthroplasty in nine.

Results: Four implants had to be removed, three for sepsis, and five patients required revision surgery. The mean postoperative Constant score for the ten prostheses still in place was 53.1 points with a mean elevation of 111° at three years seven months follow-up. The gain in pain score was 8.5 points with a mean result of 10.9 points. The results were different depending on the initial radiological form, with less favourable outcome observed in typical osteonecrosis.

Discussion: This study demonstrated a particular radio-clinical entity independent of classical osteonecrosis of the humeral head. The surgical procedure was more difficult and the outcome was less satisfactory than in the classical forms with a high rate of complications.

Purpose: The purpose of this work was to compare outcome after hemiarthroplasty and after inverted arthroplasty for the treatment of excentred degeneration of the shoulder joint.

Material and methods: This multicentric study inclued 136 patients, 110 women and 26 men, mean age 72.4 years (range 55 – 86 years) with 142 shoulder prostheses. Among the 142 shoulders operated on, 62 underwent Aequalis hemiarthroplasty and 80 inverted arthroplasty. The preoperative handicap was more severe in the hemiarthroplasty group (Constant score 26) than in the inverted arthroplasty group (Constant score 23).

Results: Clinical and radiological follow-up data were collected at a mean follow-up of 44 months for the Aequalis prostheses and at 45 months for the inverted prostheses. There were seven failures requiring five revision procedures: three in the hemiarthroplasty group and four in the inverted arthroplasty group.

The differences between the hemi- and inverted arthroplasty groups concerned the Constant score, which was significantly better in the inverted prostheses (65.5) than in the hemiarthroplasties (46.1), for all subscores. Active elevation was 138° for the inverted prostheses and 97° for the Aequalis prostheses (p < 0.01). Mean external rotation in position 1 was 22° for the Aequalis prostheses and 11° for the inverted prostheses (p < 0.01) with no difference in elevation rotation. These results remained equivalent and significant over time, even beyond five years.

Radiographically, there was one case with an anomalous humeral component (impaction) in the inverted group. For the inverted prostheses, there were three migrations that have not been revised to date and three partial screw loosenings. The main problem was the development of notches in the scapular column observed in 50% of the cases including 20% which reached the lower screw. For the Aequalis prostheses, the main problem was deterioration of the acromial vault observed in 50% of the cases leading to altered function.

Discussion: Inverted prostheses provide clearly better mid-term results, even after five years. With hemiarthroplasty there is a long-term risk of vault wear. The long-term course of the inverted prostheses is worrisome due to the notches in the scapular column and the stress on the fixation. These prostheses should be used preferably for old persons with a deteriorated vault.

Purpose: Does the postoperative abduction brace facilitate healing after rotator cuff repair? We have answered this question empirically using, since 1995, a premodelled abduction brace for four weeks for all patients, together with early passive rehabilitation. The purpose of this work was to assess the efficacy of this brace and compare outcome with that in patients treated in 1994–1995 with and without the brace.

Material and methods: This retrospective analysis was conducted in 72 patients, mean age 58 years. Contant functional score and imaging of the tendon healing (arthroscan 56%, MRI 44%) were used as assessment criteria. Four homogeneous groups were identified: Group 1 (40 patients, mean age 55 years) was composed of patients with a distal tear of the supraspinatus; twenty patients in this group were treated with a “elbow-to-body” sling (group 1a) and twenty others with the abduction brace (group 1b); Group 2 (32 patients, mean age 60 years) was composed of patients with an intermediary tear of the supraspinatus, partially extending anteriorly or posteriorly, fifteen patients in this group were treated with a “elbow-to-body” sling (group 2a) and seventeen others with the abduction brace (group 2b).

Results: Mean follow-up was 49 months. The weighted Constant score improved from 57.3% to 89.1%. Imaging demonstrated recurrent tears in 25 patients (35%). In group 1 (distal tear of the supraspinatus), the weighted Constant score at last follow-up was 89% (93% in group 1a and 86% in group 1b). There were eight recurrent tears (20%), (25% in group 1a and 15% in group 1b). Use of the abduction brace (group 1a) had no effect. In group 2 (intermediat tear of the supraspinatus), weighted Constant score was 88% (82% in group 2a and 94% in group 2b). There were seventeen recurrent tears (53%) (71% in group 2a and 41% in group 2b). At last follow-up, there was no significant difference for functional score or tear recurrence.

Discussion and conclusion: These findings suggest the postoperative abduction brace is not beneficial after repair of non-retracted distal cuff tears. On the contrary, for intermediate tears, it allows a clear improvement in the final Constant score and a lower rate of recurrent tears. This study provides information useful for choosing the postoperative management of patients undergoing repair of rotator cuff tears.

Purpose: Infection is a serious complication of total knee arthroplasty. Surgical strategies based on removal-reinsertion of the prosthesis in two times with antibiotic therapy has proven its efficacy. The use of a spacer between the two operations has been proposed to facilitate reimplantation. Since 1993, we have used this two-time procedure with an articulated spacer in an attempt to optimise functional outcome yet maintain anti-infection efficacy. We report our experience.

Material and methods: This retrospective series included 28 patients, 21 women (75%) and seven men (25%) who underwent surgery between December 1993 and February 2000.Mean age of the patients at revision was 67 years (18–83). Medical and surgical risk factors for infection were present in 64% and 54% of the cases respectively. Delay between prosthesis surgery and onset of the first signs of infection was 29 months (four days–222 months). A single-germ infection was involved in 18 cases (64%) and a multiple-germ infection in nine (36%). The infection was acute in 32% of the cases and chronic in 68%. There were eight fistulae (28%). Bacteriology reported staphylococcal infection in 25 cases (including 13 S. epidermidis), streptococcal infections in five, anaerobic germs in seven (corynebacterium in five) and Gram-negative germ (pseudomonas) in one. Delay between diagnosis of infection and insertion of the articulated spacer was 11 months (four days–62 months). The first operation consisted in removal of the prosthesis, wide excision of the synovial and infected tissues and insertion of the two articulated pieces, modelled with antibiotic-impregnated cement. Weight-bearing was authorised with crutches and an articulated brace. Rehabilitation exercises were performed to maintain joint amplitude. The prosthesis was reimplanted three months later (1.5–7 months). All prostheses were reimplanted with cement: two prostheses with posterior preservation, 20 posterior stabilised prostheses, and six hinge prostheses. The patients were given antibiotics for eleven months (1–25 months). The IKS score was used to assess functional outcome. Cure of infection was assessed on clinical, biological and radiographic findings.

Results: All patients were seen at a mean follow-up of 35 months (8–78). Follow-up was greater than 24 months in 68% of the patients. We had three cases (11%) of recurrent infection: one acute infection and two septic loosenings. At reimplantation, we had complications in seven patients (25%) ten of whom required revision surgery, six for mechanical complications (three dislocations, three aseptic loosenings). Mean IKS score was 136 points (50–190) with 79 points (30–100) for the knee and 67 points (20–90) for function. Mean flexion amplitude was 94° (45–115°).

Discussion, conclusion: With this method, joint mobility can be maintained between the two operations, greatly improving patient comfort. The mid-term results in terms of infection cure have been satisfactory (89% cure). Nevertheless, the final functional result can be disappointing, due to the persistence of pain (low-grade infection, difficult implant fixation…). The removal-reinsertion strategy using a single operation would in our opinion still have its indications.

Purpose: Arthroscopic arthroplasty for painful shoulder is not always successful. Repeated arthroscopy is sometimes discussed for patients with recurrent pain. The purpose of this work was to determine the technique and identify indications.

Material and methods: We conducted a retrospective analysis of 24 patients (13 men, 11 women), mean age 52 years, who underwent repeated arthroscopy after failure within 31 months of arthroscopic acromioplasty with no other intervention. Three groups were formed: group 1 (7 patients): subacromial impingement due to tendinitis with intact supraspinatus; group 2 (11 patients): subacromial impingement secondary to cuff tear; group 3 (6 patients): calcified tendinopathy. In group 1, the acromion was type 1 in three cases, type 2 in three and type 3 in one; repeated arthroscopy included complementary acromioplasty. In group 2, all patients had complementary acromioplasty; six of them with biceps tenotomy. In group 3, the remaining calcification was removed in all patients and complementary acromioplasty in three.

Results: Mean follow-up was 21 months. In group 1, outcome was satisfactory in three patients (43%), irrespective of the acromial morphology. In group 2, outcome was satisfactory in six patients (55%), including five with acromioplasty with tenotomy and only one with complementary acromioplasty alone. In group 3, outcome was satisfactory in five patients (83%); one failure was attributed to remaining calcification; acromioplasty did not influence outcome.

Discussion, conclusion: The acromion should not be considered as the principal cause of failure after first acromioplasty. In this series, only 33% of the patients who had complementary acromioplasty had a good outcome at last follow-up. Biceps tenotomy in patients with rotator cuff tears and removal of the calcium deposit in patients with calcified tenopathies should be considered first.

Purpose: Arthroscopic treatment of calcified tendinopathy is classically performed in two times: exploration of the glenohumeral joint followed by subacromial arthroscopy to evacuate the calcification. In our experience, glenohumeral arthroscopy has only exceptionally provided a diagnostic element. In addition, the principal complication of this procedure is retractile capsulitis which may be a secondary effect of glenohumeral aggression. We conducted a retrospective analysis to assess the effect of systematic exploration of the glenohumeral joint.

Material and methods: Two homogeneous groups of patients were identified. Group 1 included 32 patients who had had glenohumeral arthroscopy then resection of the calcifications using a bursoscope. Group 2 included 32 patients whose treatment was limited to subacromial arthroscopy for resection. The preoperative Constant score (52 in group 1 and 54 in group 2), disease duration (34 and 40 months respectively), and localisation of the calcification were comparable for the two groups. Acromioplasty was not performed in these patients. All were reviewed at minimal follow-up of 6 months for assessment of the Constant score and a radiography study.

Results: At last follow-up the mean Constant score was 70 in group 1; calcifications had disappeared in 84% of the cases and delay to recovery (total pain relief and return to work) was 11 months. There were 4 cases of postoperative capsulitis (12.5%). The mean Constant score was 79 in group 2; calcifications had disappeared in 78% of the cases and delay to recovery was 6.5 weeks (p = 0.0001). There was one case of retractile capsulitis (3%). In group 1, glenohu-meral arthroscopy did not lead to the discovery of specific elements except in two cases where it identified partial tear of the deep aspect of the supraspinatus. Acromioplasty was never performed.

Discussion-Conclusion: Systematic glenohumeral arthros-copy is not warranted in patients undergoing treatment for calcified tendinopathy. The fact that glenohumeral exploration did not disclose any particular element and had no effect on healing and capsulitis rates favours the use of a subacromial approach alone.