Purpose: Development of knee arthroplasty has led surgeons to reconsider their tibial osteotomy techniques. Closed wedge subtraction which can lead to a deformed epiphyseal callus and an oblique joint space have been progressively replaced by more anatomic open wedge techniques. The bone gap created by open osteotomy is general filled with an iliac autograft, implying morbidity of the harvesting site, or inert material. For the last eleven years, we have used dehydrated corticocancellous grafts which maintain the collagenous framework and mechanical properties as well as the ingrowth properties close to autografts without the inconvenience of bone harvesting. After making the horizontal osteotomy cut above the tubercle, tricortical allografts are fashioned at the operative site and implanted without impaction. Fixation is achieved by staples or plate screwing. Mechanical stability is always achieved within the usual time allowing weight bearing at 40 days.

Material and methods: Twenty patients who underwent this procedure more than nine years earlier were reviewed for this study.

Results: Two patients died. One underwent a second procedure for a tri-compartmental prosthesis seven years after osteotomy. There were no local or general complications at short- or long-term. Radiological ingrowth was observed 12 to 24 months postoperatively allowing removal of the osteosynthesis material and bone biopsies which proved bone ingrowth. Control gonometry at 45 days and one year validated the reliability of the technique and achievement of the desired biomechanical objective, confirming the absence of lost correction.

Discussion: This simple and reliable technique has several advantages: simple less aggressive surgery, immediate mechanical stability, absence of material-related complications, constant radiographic ingrowth proven histologically.

Conclusion: The quality of the long-term results make this technique a reasonable alternative for iliac autografting. It should be preferred over non-inhabitable inert material and over heterografting which has its own complications (mechanical instability, pseudoseptic lysis).

Purpose: Among 115 cemented patellofemoral Bechtol-Blazina prostheses implanted between 1974 and 1991, we reviewed 59 knees in 41 women and nine men (nine bilateral implants). Material and methods: Mean patient age was 60 years and mean follow-up was ten years (6–16). The main indication was primary degenerative disease (82%) followed by chondrocalcinosis (6%), chondromalacia (7%) and secondary degeneration (5%). A previous procedure had been performed on 18 knees (14 patellofemoral and four femorotibial). Associated procedures were tubercle translocation (n=15), tibial osteotomy (n=2), and tension of the medial wing (n=1). The International Knee Society score was noted.

Results: According to the IKS, outcome was good or excellent in 91% of the knees and insufficient in 9% (knee score and function score). Mean ten-year survival determined with the actuarial method was 84.4%. It fell from 91.1% to 75.5% between the ninth and eleventh year due to femorotibial deterioration. There were no cases of phlebitis, infection or prosthetic loosening. Causes of failure were predominantly femorotibial (eight degradations) rather than femoropatellar (two snaps and three painful syndromes involving the lateral parapatellar area). Thirteen reoperations were necessary: tri-compartmental arthroplasty (n=8), modification of the trochlear orientation (n=1), release of the lateral patellar wing (n=3), demaquetisation (n=1). Total or nearly total pain relief was achieved for 80% of the knees. The mean flexion angle was 123° and stair climbing was considered normal in 91% of the patients. All knees were stable. Radiographically, there were six cases of polyethylene remodelling and one case of wear associated with major femorotibial deterioration. This absence of significant wear was, in our opinion, essentially due to the absence of a metal-backed patellar insert.

Discussion: The following indications can be retained on the basis of this series: the ideal indication is isolated patellofemoral osteoarthritis with dysplasia, relative indications are advanced-stage patellofemoral osteoarthritis with axial deviation having no clinical or radiographic (tangent view) impact. In this case, the patellofemoral implant allows postponing insertion of a tricompartmental prosthesis in middle-aged patients. Contraindications are patellar chondromalacia, chondrocalcinosis, low patella, and presence of associated femorotibial degradation.

Conclusion: The long-term clinical and radiographic results after patellofemoral arthroplasty make this procedure the treatment of choice for isolated and advanced-stage patellofemoral osteoarthritis on a normally axed limb when a conservative procedure cannot be performed or has failed. It provides effective pain relief and good knee stability. The operation is difficult and must use congruent implants to avoid the problem of prosthetic instability and patellar snaps.

Purpose: We analysed complications observed after 1771 implantations of single-compartment knee prostheses performed since 1974.

Material and methods: Sterilised all polyethylene Marmor implants were used from 1974 to 1984. Metal-backed tibial implants with or without cement and condyle resurfacing were added after 1984.

Results: Mean survival of the 207 Marmor cemented prostheses implanted from 1974 to 1984 was 93% at 12 years. There were nine significant complications: seven loosenings, three with infection, and two deteriorations of the other compartment. There were five minor complications: two instabilities, one patellar and one anterior laxity; two stiff joints; one meniscal lesion in the other compartment and one hamstring tendinitis. A non-adjustable cobalt-chromium metal backed insert was added to the polyethylene lining in 1984. For the 790 implants using this prosthesis, 48 had to be revised (6%) four to eight years after implantation for polyethylene wear due to an insufficient thickness for size 7.5 mm and gamma sterilisation.

These observations led us to use, since 1991, two new polyethylene inserts sterilised with ethylene oxide: an adjustable titanium metal-back support with cement in 329 cases and without cement in 171 with hydroxyapatite surfacing, and a modified Marmor with complete thickness (min 9 mm) peripheral cortical support used in 274 cases. Fourteen revisions were required for: screwing problems in three cases early in our experience, three loosenings, three inappropriate indications, two patellar problems, one deterioration of the other compartment, one knee instability on an oblique plateau and unexplained residual knee pain. There was no evidence of significant polyethylene wear, even for the oldest cases in our series.

Conclusion: This long-term analysis of single-compartment knee prostheses has shown that at the tibial level the essential elements are the minimal thickness of the polyethylene, the type of sterilisation, and the type of metal back. Resurfaced condyles have demonstrated their superiority over those requiring a resection since only one femoral loosening was observed. Revision for total knee arthroplasty is not different from first intention replacements.