RADIOLOGICAL AND HISTOLOGICAL STUDY OF ASEPTIC LOOSENING USING A CEMENTED TIBIAL HEMIARTHROPLASTY IN THE RABBIT KNEE

Fourteen mature New Zealand white female rabbits had a unilateral cemented, stemmed, titanium, condylar-type tibial hemiarthroplasty, using an anteromedial arthrotomy of the right knee. The articular cartilage and minimal bone were resected. There were two prosthetic groups of seven animals each: a well-fixed, non-loose group and a loose group. In the non-loose group, the implant was inserted into the cement bed and axially compressed until the PMMA had cured. In the loose group, the same volume of cement was allowed to cure on the implant ex vivo; the prosthesis was then implanted to ensure that it was loose fitting. Radiographs were performed at zero and 3 months and graded for new lucent lines. Histological analysis was performed using undecalcified coronal sections, surface stained with toluidine blue with the prosthesis in situ, and the cement mantle preserved. Back-scattered electron microscopy was also performed. The mean cumulative grading of new lucent lines was 0.3 +/- 0.1 for the non-loose group and 2.2 +/- 0.4 for the loose group (P < 0.005). Non-loose prostheses were surrounded by a thin fibrous membrane or bone. Loose prostheses were surrounded by a thicker, fibrous tissue layer, containing histiocytes and giant cells which were more prevalent around cement particles, especially near the prosthetic tip. These findings parallel the histology found at cemented prosthetic interfaces in humans. The results of this study suggest that the fibrohistiocytic membrane commonly found around loose cemented implants may be the result of, rather than the cause of, the loosening process.