Animal models of osteomyelitis. Knowledge, hypothesis, and speculation.

Each animal model has provided insights. Particularly important was the considerable resistance of bone to infection without manipulation (no morrhuate, fracture, rod, wax, or prosthesis). Such perturbations allow bone infection with much smaller inocula. Typical inocula decreases are 1000 to 10,000 fold. Staphylococci may have a selective advantage in bone because of specialized or tropic binding, perhaps to cartilage or collagen. Osteoclast-induced resorption of hydroxyapatite might explain the distribution of some osteomyelitis. Increased osteoclast activity could link the susceptible metaphyseal regions, the repetitively traumatized diabetic foot, a history of blunt bone trauma, fracture, and perhaps even nearby soft tissue infection. Diagnosis remains difficult; gallium-67 and indium111 labeled WBC probably deserve additional investigation. Therapeutic failures in the rabbit and rat models mirror clinical experience. Clindamycin, rifampin, and quinolones are promising. Neither systemic nor local antimicrobial prophylaxis is well studied yet.