The purpose of this study was to characterise accurately, the extent and geometry, and produce representative rigid resin models of full thickness articular cartilage lesions of various types, shapes and sizes on the articular surface of pig patellae.

Ten adult pig patellae and three adult Ox patellae were obtained and cleared of adhering tissue. Full thickness lesions were induced from oval shaped to “U” shaped scarifications by careful use of a hand held bur, and the geometry noted by taking appropriate dimensions with a Vernier cailper in the horizontal and vertical planes, and plan view photographs. MRI images using fat-suppressed weighted 1.5 mm thick slices scans in the horizontal plane, were produced in DICOM format for conversion to SLE files used in the reconstruction in the computer. The patellae were then held in a stone plaster mix to produce a male mould of the articular surface. The computer images were generated and the physical dimensions taken with the Vernier calliper were recorded from the reconstructed image in the computer using graphics software. The computer data was used to produce a rigid full scale model of the articular surface in resin using laser stereolithography which is using in the rapid prototyping industry. The resin models were matched with the male plaster moulds to confirm an accurate match of the 3-dimensional shape of the computer generated in all the types of lesions we produced. It is proposed to use the rigid models to produce sterile templates that could be used by surgeons to fashion an area around a lesion using a suitable reamer/bur using a predetermined criteria of cartilage thickness (say 2 mm), and the same geometrical data would be used to produce a suitable semi rigid scaffold shaped to the lesion. Our study has shown that very accurate 3 dimensional data can be quickly processed from MRI images to produce, using current rapid prototyping techniques, templates and implants to fit lesions accurately in the patella. There is no reason why this technology could not be applied to any joint surface that can be accessed by MRI.

Bleeding from cancellous bone causes lamination within bone cement and at its prosthetic interfaces, and weakens the fixation of joint replacements. We examined the effects of anaesthesia and blood pressure on bleeding in human cancellous bone, and investigated the local response to freezing saline, 1:200,000 adrenaline and hydrogen peroxide. Spinal anaesthesia reduced cancellous bleeding by an average of 44%, local freezing saline by 24%. Saline at room temperature, adrenaline solution and hydrogen peroxide each reduced it by 14%. The effects of spinal anaesthesia and of freezing saline were additive: used together they reduced bleeding by 56%. The reduction of blood contamination of cement and its interfaces should contribute to better prosthetic fixation.

Since 1974, we have made a prospective study in Bristol of the results of unicompartmental knee replacement using the St Georg sledge prosthesis. A total of 115 knees in 100 patients have been followed up for 2 to 12 years (mean, 4 years 9 months). Results have assessed both by the Bristol knee score and by survivorship studies on the total series of 138 knees. Results were excellent or good in 86% and fair or poor in 14%. The survivorship study (based on a definition of failure which included significant pain or a dissatisfied patient or the need for revision) showed a cumulative success rate of 76.4% at six years, with no further failures after that time. Seven knees have been revised, in most cases for deterioration of the contralateral compartment. The operation is recommended as a satisfactory and durable form of treatment for osteoarthritis affecting a single tibiofemoral compartment.