Steroid-induced versus cryoinsult-induced femoral head osteonecrosis: Statistical measurement of histologic abnormality focalization

Systematic studies of treatment options for femoral head osteonecrosis have been hindered by the lack of an animal model that mimics the human event of femoral head collapse. The (bipedal) emu model of osteonecrosis has proven to progress to collapse following an external insult. Since osteonecrosis clinically has multiple etiologies, various alternatives are seemingly appropriate to initiate lesions for research purposes. This study compared osteonecrosis initiated by a local cryoinsult, to that initiated by a large systemic dose of steroid (methylprednisolone acetate), and by a combination of steroid and cryoinsult. One year post-insult, the animals were sacrificed and their femoral heads harvested. A custom-written computer program was used to expedite analysis of the global distribution of histologically apparent osteonecrosis, as determined by the Ficat scale. Following whole-head analysis, grade distributions were calculated, and a measure of spatial clustering of histologic abnormality was calculated using Moran's I statistic. Normal femoral heads showed very small amounts of histological abnormality, and no significant spatial clustering. Femoral heads receiving isolated cryoinsult had a greater incidence of high-grade abnormality, as well as statistically significant spatial clustering of that abnormality. Steroid-only animals had a higher percentage of diffuse, low-grade abnormality. The group receiving both steroid and cryoinsult showed the highest overall abnormality, with statistically significant clustering of that high-grade abnormality. Cryoinsult, both in isolation and in combination with a steroid regimen, appeared necessary to mimic the clinically typical focal lesions of osteonecrosis.