Aim of this study was to compare the postoperative range of motion of three types of total knee replacements.

They were 72 posterior cruciate ligament retaining knee prostheses (group I), 61 postero-stabilized (group II), 52 ultracongruent plates (group III). Inclusion criteria were primary arthritis with varus deformity inferior to 15 degrees (°), no previous surgery on the knee, body mass index inferior to 35, preoperative flexion superior to 110°. All prostheses were performed with the same ancillary with one unique surgeon (DH). Recovery and analgesia protocols were similar in the three groups. Mobility was measured using a goniometer.

Continuous data were tested for normal distribution using Kolmogorov-Smirnov test. Normally distributed data were analyzed with two tailed t-tests, whereas non-parametric data were analyzed with Mann-Whitney U test. Statistical significance was set at p < 0.05.

At 2 years follow-up, the group I demonstrated a mean flexion of 121.8° in preoperative period and 110.3° in postoperative period. They were respectively of 121.3 and 122.2° in the group II. Regarding group III, they were 121.6 °preoperatively and 118.4° postoperatively.

Results were significant (p< 0.05) between groups I and II, and groups I and III.

No statistic correlation was found between need of mobilisation under general anesthesia (p> 0.05), flexum (p> 0.05), knee score (p> 0.05), patient satisfaction depending on type of prostheses (p> 0.05).

Posterior cruciate ligament removal tends to offer a best postoperative flexion without significant influence on the knee score or patient satisfaction.

Competence of the extensor mechanism is the major determinant of functional outcome after resection of the proximal tibia and tumor prosthesis implantation. Restoration of a compromised active extension of the knee and an extension lag still remains a difficult challenge. Various techniques have been proposed in the past twenty years including direct attachment of the patellar ligament to the prosthesis, transposition of the medial gastrocnemius muscle possibly associated with other muscle flaps, transposition of the fibula and combination of these techniques. Transposition of the fibula was first reported by Kotz in 1983 but not sufficiently described, so that surgeons who want to plane and manage such a procedure can have some difficulties. We present our technique of fibula transposition and report the functional results about seven patients treated for high-grade sarcomas of the proximal tibia. Fibula transposition is carried out only if the entire fibula and its soft-tissues can be preserved. Resection of the tumor and reconstruction is carried out using the same anteromedial approach. After implantation of the prosthesis, the fibula and its muscles are mobilized anteriorly in a ‘baionnette’ shape obtained by performing a two-level osteotomy. The peroneal nerve and the anterior tibial vessels are previously identified and released to prevent tension on these structures during transposition. Care must be taken to preserve as much as possible the muscular insertions on the fibula so that probability of bone fusion increases. The biceps tendon and the lateral collateral ligament inserted in the fibular head are sutured to the patellar ligament. The knee is immobilized in a knee-ankle orthosis for 6 weeks. We have performed this technique in seven cases. A medial gastrocnemius muscle flap was associated in 3 cases to cover the prosthesis. Fusion was achieved in all cases. Full active extension was obtained in all cases with an extensor strength rated 5/5. All patients were ambulatory without external support at the last follow up.

Purpose: Patellar complications are among the most frequent after total knee arthroplasty. Encasing the patellar piece is one way of resisting the shear forces leading to loosening.

Material and Methods: We studied at more than five years the results obtained with a total knee prosthesis implanted with preservation of the posterior cruciate ligament (PCL). This prosthesis has an asymmetric encased patellar insert with a cemented central pivot. The instrumentation ensures patellar thickness. We reviewed 104 implants at more than five years. Six had been lost to follow-up. Ninety-eight implants were still in place.

Results: The following complications were observed: four fractures of the upper rim with little displacement (these fractures healed and pain regressed but the insert had moved); three vertical patellar fractures with little displacement (these fractures healed; two were symptomatic temporarily); one transverse fracture of the upper pole with displacement causing a defect in active extension; eight moderate asymptomatic impactions which were visible on the lateral x-ray (modified orientation of the insert with cement fracture). There was no significant difference for functional results (pain 40.9; movement 21.9; knee score 84.3) between patients with or without a patellar complication.

Discussion: Insertion of an asymmetric prosthesis increased the risk of an orientation error (two cases early in our experience). Encasing the patellar insert limits medialisation yet the centering was satisfactory (centred patella 95.2%, shift 3.6%, subluxation 1.2%). Encasing provides a peripheral wall protecting against transverse sheer forces. The lateral wall did not fracture, demonstrating its efficacy. The upper wall can fracture under the force of flexion without functional consequences. The other fractures, favoured by section of the lateral patellar wing (p< 0.05), were not treated. Moderate but certain impaction was noted in eight cases at a mean 3.5 years (1–6 years). It was due to failure of bony support under the effect of the compression forces applied on a small surface. The diameter of the encased patellar inserts was rarely more than 25 mm. Once the prosthesis is in place, the periphery of the patella is the only component articulating with the trochlea and its impaction does not cause further aggravation. This contact did not lead to pain in any patients.

Conclusions: Complications observed with encased patellar components differ from the better known apposed prostheses.

Purpose of the study: Fragile bone and weak soft tissues can create a serious challenge for arthroplasty of the rheumatoid arthritis shoulder. Patients seen late after rotator cuff tears become irreparable may also present a stiff shoulder, further complicating the procedure.

Material and methods: The purpose of this study was to assess outcome at more than five years in a prospective series of 12 patients with rheumatoid arthritis of the shoulder with an irreparable rotator cuff tear treated with a hemiarthroplasty with a mobile cup. The radiological and clinical results were compared with those obtained in a control series of ten bipolar humeral prostheses implanted for centered or excentered degenerative disease with irreparable cuff tears.

Results: The mean preoperative Constant score was 16.9 points: pain 2.5, activity 4.2, active mobility 9.5, strength 0.7. Active ROM was 63.8° for anterior elevation, 45° for abduction, and 12° for external rotation. At last follow-up, the mean postoperative Constant score was 39.4 points: pain 10.7, activity 10.8, active mobility 13.8, strength 4.1. Mean active anterior elevation was 83.7°, abduction 70.4°, and external rotation 29.1°. Outcome was not significantly different from the control group with degenerative joint disease (p< 0.05).

Discussion: The overall Constant score, especially the pain score, was significantly improved (p< 0.05). Improvement in joint motion was modest but comparable with other series in the literature and even better than with conventional hemiarthroplasty for the same indication. There were few complications, mainly superior subluxation favored by the preoperative infra-scapularis or infraspinatus tears. Glenoid wear was significant despite the dual mobility concept. There were no cases of loosening.

Conclusion: These results show that hemiarthroplasty with a mobile cut provides acceptable mid-term results for the advanced-stage rheumatoid shoulder with an irreparable rotator cuff tear. Results in this series were comparable with that in the control group of patients with degenerative joint disease. Consequently, the status of the rotator cuff appears to be more important that the inflammatory or degenerative etiology. Certain cuff tears involving the infrascapularis raise the risk of superoanterior instability and could be a limitation for this method. A more constrained prosthesis might be advisable.

Purpose of the study: Opinions vary concerning results after knee arthroplasty with preservation of the posterior cruciate ligament (PCL) in patients with rheumatoid disease. We report our findings in patients reviewed more than ten years after implantation in comparison with patients treated for osteoarthritis.

Material and methods: One surgeon implanted 43 knee arthroplasites (Kali) with preservation of the PCL (9 bilateral cases) in 31 women and 3 men, mean age 53 years (range 30–70 years). Outcome was assessed with the AKS clinical and radiological scores. Passive recur-vatum and posterior drawer at 90° flexion were measured radiographicaly at last follow-up. Outcome was compared with the results observed in a control group of 29 prostheses of the same type implanted for osteoarthritis (among a total of 203 implantations).

Results: There were no patients lost to follow-up: two patients were removed from the analysis due to infection on early wound necrosis and late metastatic infection. Eleven patients (16 prostheses) died before ten years follow-up; outcome was satisfactory for the prosthesis. Twenty-one patients (25 prostheses) were reviewed at more than ten years, mean follow-up 136 months. There was one case of supracondylar fracture which healed without sequela after osteosynthesis. The mean knee score was 34.3 preoperatively and 87.2 postoperatively with a mean function score improvement from 17 to 44 points. The pain score (47.3 points on average, was significantly improved while joint range of motion remained unchanged (117°). There were no worrisome lucent lines. Mean recurvatum measured radiographically was 6.9° (range 3–14°) and mean posterior drawer at 90° flexion was 4.2 mm. Outcome in the control group was the same excepting (p< 0.05) for lesser range of motion (109.7°) and better function score (62 points). Laxity (clinical and radiographic scores) were the same.

Discussion: The results obtained in patients with rheumatoid disease were satisfactory and the same as those obtained in patients with osteoarthritis and were comparable to those with prostheses sacrificing and replacing the PCL. There were no cases of prosthesis loosening. Complications were very limited and less frequent than among the entire population of 203 prostheses for degenerative disease.

Conclusion: Ligament alterations are not a contraindication for preservation of the PCL in patients with rheumatoid arthritis. Irrespective of the etiology, the main limitation on prosthesis longevity is polyethylene wear observed beyond ten years (ten changes of the plateau because of wear among 246 prostheses).

Purpose: At a time when total knee arthroplasty (TKA) with an ultra-congruent tibial plateau or a mobile plateau are advocated by many, the question is whether TKA with a fixed plateau preserving the posterior cruciate ligament (PCL) should be abandoned. We analysed the results of 500 Wallaby I TKA with preservation of the PCL and presenting asymmetrical and divergent femoral condyles with a fixed, also assymetrical tibial plateau, at mean follow-up of seven years (1–10).

Material and methods: This prospective multicentric study was performed by junior and senior surgeons. Mean patient age was 70.11 years and 91.4% of the patients had primary or secondary degenerative disease. Prior surgery had been performed in 130 knees, mainly for osteotomy (n=40, mostly tibial) and revision of uni-compartmental or total prostheses (n=18). The mean preoperative IKS knee score was 26.11 points, the function score was 29.54. Preoperative alignment was correct for 11.26% of the knees, 27.16% presented > 4° valgus and 61.56% > 3° varus. Nearly all tibial and patellar pieces were cemented. 5.8% of the femoral pieces were not cemented. A prosthesis was implanted on the patella in all knees except four. There were two early infections, one popliteal sciatic paralysis and twelve wound healing problems. General mobilisation under general anaesthesia was performed in 53 knees (10.6%).

Results: Twenty-two patients were not retained for analysis, fifteen lost to follow-up and seven deceased at one year. Among the 478 knees followed for one to ten years, there were six late infections (1.25%), one aseptic bipolar loosening (0.2%), 25 patellar fractures (5.23%) including three which required revision (two cerclage, one prosthesis removal), three periprosthetic femur fractures without effect on the clinical or radiographic outcome, one traumatic tear of the medial collateral ligament, and two secondary tears of the PCL without clinical consequences. There were no revisions for instability, generally considered the most frequent reason for TKA revision. The mean postoperative IKS score was 90.6 points and the function score was 59.7 points basically due to patient age and comorbidity. The eight-year survival (Kaplan-Meier method) was 98.2% (95%CI: 99.4–96.9) irrespective of the reason for revision and 99.2% (95%CI 100–98.4%) if the revision was performed for a mechanical problem. Radiographically, more than 70% of the knees were aligned (between 3° valgus and 2° varus) and more than 90% were between 5° valgus and 5° varus. Although it was difficult to measure radiographically polyethylene wear, there was only one case of > 2 mm wear with osteolysis found in 50 knees selected randomly among the knees with more than seven years follow-up.

Conclusion: This study, like others reported by authors preserving the PCL, show that preservation of the PCL limits the risk of instability, allowing excellent clinical and radiographic outcome without important polyethylene wear, opening perspectives for good long-term results.

Purpose: We report our results of revision procedures for severe acetabular loosening with saddle prostheses. The series concerned six patients, mean age 67 years age range 58–82 years, who initially presented rheumatoid disease (n=3), degenerative hip disease (n=3, one secondary to trauma), and radiation-induced hip disease (n=1). These patients had undergone one to three procedures for reconstruction with a bone graft and metal implants.

Material and methods: Failure was the indication for salvage with a saddle prosthesis due to massive loss of acetabular bone stock, complicated in five cases by an unstable hemipelvis due to transverse nonunion. An allograft was screw fixed in the iliac wing in two cases to stabilise the saddle prosthesis. Weight bearing was allowed early after surgery in all cases except two. The hip was immobilised in four cases for three to six weeks in a bermuda cast or with traction (n=4). Follow-up was 2 to 5 years.

Results: There was only one complication: stress fracture of the iliac bone at two years which did not heal. Bone graft with ostheosynthesis was proposed. Pain relief was nearly total in all cases. Four patients could walk without crutches and two walked with two crutches or an ambulator due to associated disease or alar fracture. The fixation was at 60 – 95° with abduction at 10–30°. All patients were satisfied and two felt the result was very superior to simple ablation of the prosthesis which they had experienced. Radiograpically, there was no evidence of ascension of the saddle but the risk remains at this short follow-up due to the rates described in the literature. One major migration was observed as was one stress fracture of a weak iliac bone which required removal of the implant.

Discussion: This difficult technique provides a better result and thus would be indicated for young patients able to sustain the intervention. It would be logical to associate a supporting bone graft on the iliac bone to prevent stress fracture and migration of the saddle. The saddle prosthesis appears to be the last resort after failure or insufficient results after prior reconstruction. The leading cause of failure is nonunion with hemipelvis mobility. This suggests the intervention should be considered before reaching this stage of well tolerated ace-tabular loosening.

Purpose: Patellar complications are among the most frequent after total knee arthroplasty. Encasing the patellar piece is one way of resisting the shear forces leading to loosening.

Material and methods: We studied at more than five years the results obtained with a total knee prosthesis implanted with preservation of the posterior cruciate ligament (PCL). This prosthesis has an asymmetric encased patellar insert with a cemented central pivot. The instrumentation ensures patellar thickness. We reviewed 104 implants at more than five years. Six had been lost to follow-up. Ninety-eight implants were still in place.

Results: The following complications were observed: four fractures of the upper rim with little displacement (these fractures healed and pain regressed but the insert had moved); three vertical patellar fractures with little displacement (these fractures healed; two were symptomatic temporarily); one transverse fracture of the upper pole with displacement causing a defect in active extension; eight moderate asymptomatic impactions which were visible on the lateral x-ray (modified orientation of the insert with cement fracture). There was no significant difference for functional results (pain 40.9; movement 21.9; knee score 84.3) between patients with or without a patellar complication.

Discussion: Insertion of an asymmetric prosthesis increased the risk of an orientation error (two cases early in our experience). Encasing the patellar insert limits medialisation yet the centering was satisfactory (centred patella 95.2%, shift 3.6%, subluxation 1.2%). Encasing provides a peripheral wall protecting against transverse sheer forces. The lateral wall did not fracture, demonstrating its efficacy. The upper wall can fracture under the force of flexion without functional consequences. The other fractures, favoured by section of the lateral patellar wing (p< 0.05), were not treated. Moderate but certain impaction was noted in eight cases at a mean 3.5 years (1–6 years). It was due to failure of bony support under the effect of the compression forces applied on a small surface. The diameter of the encased patellar inserts was rarely more than 25 mm. Once the prosthesis is in place, the periphery of the patella is the only component articulating with the trochlea and its impaction does not cause further aggravation. This contact did not lead to pain in any patients.

Conclusions: Complications observed with encased patellar components differ from the better known apposed prostheses.