Gatifloxacine (GFLX)-containing poly(lactide-co-glycolide) (PLGA) was introduced to the pores and surfaces of porous β-tricalcium phosphate (βTCP) granules by melt compounding whereby no toxic solvent was used. The granular composite of GFLX-loaded PLGA and βTCP released GFLX for 42 days in Hanks' balanced solution and exhibited sufficient in vitro bactericidal activity against Streptococcus milleri and Bacteroides fragilis for at least 21 days. For in vivo evaluation, the granular composite was implanted in the dead space created by the debridement of osteomyelitis lesion induced by S. milleri and B. fragilis in rabbit mandible. After a 4-week implantation, the inflammation area within the debrided area was markedly reduced accompanied with osteoconduction and vascularization in half of the rabbits, and even disappeared in one of the six rabbits without any systemic administration of antibiotics. Outside the debrided area, inflammation and sequestrum were observed but the largest of such affected areas amounted to only 0.125 times of the originally infected and debrided area. These findings showed that the granular composite was effective for the local treatment of osteomyelitis as well as an osteoconductive scaffold which supported and encouraged vascularization.

Composites of gatifloxacin (GFLX)-loaded poly (lactic-co-glycolic) acid (PLGA) and β-tricalcium phosphate (βTCP) containing 0, 1, and 10 wt % GFLX (0, 1, and 10 wt % GFLX composites), and GFLX-loaded PLGA containing 1, 5, and 10 wt % GFLX (1, 5, and 10wt % GFLX-PLGA) as controls were fabricated and characterized in vitro and in vivo. On in vitro evaluation, the 10 wt % GFLX composite released GFLX over at least 28 days in Hanks' balanced solution and exhibited clinically sufficient bactericidal activities against Streptococcus milleri and Bacteroides fragilis from 1 h to 10 days. The 0, 1, and 10 wt % GFLX composites and 10 wt % GFLX-PLGA were implanted in bone defects created by debridement of osteomyelitis lesions induced by S. milleri and B. fragilis in the mandible of rabbits (n = 5). Four weeks after implantation of the 10 wt % GFLX composite, inflammation in the debrided area disappeared in all the rabbits, while inflammation remained in all the rabbits after implantation of the 0 wt % GFLX composite and 10 wt % GFLX-PLGA, and in three rabbits after implantation of the 1 wt % GFLX composite. Bone formation appears to be less intense for the 10 wt % GFLX composite than for the 1 wt % GFLX composite probably owing to the rapid degradation of the 10 wt % GFLX composite. These findings show that the GFLX composite is effective for the local treatment of osteomyelitis.