Osseointegration of porous-coated implants during revision arthroplasty procedures is often impeded due to the presence of residual granuloma, particulate debris, and a sclerotic, dysvascular bone bed. We hypothesized that local infusion of recombinant fibroblast growth factor (FGF-2) would increase bone ingrowth in an in vivo model of tissue differentiation in the rabbit tibia in the presence of phagocytosable polyethylene particles. A drug test chamber (DTC) was implanted in the proximal medial tibial metaphysis of mature rabbits unilaterally. The chamber contained a 1× 1 × 5-mm tunnel for tissue ingrowth, and was connected to an osmotic diffusion pump. FGF-2 was infused at dosages of 0, 0.5, 5, 50, or 500 ng/day for a 3-week period, with subsequent harvesting of the ingrown tissue after each 3-week treatment. The effects of ultrahigh molecular weight polyethylene particles (0.5-μm diameter) on tissue ingrowth were determined by adding particles to the chamber at concentrations of 5.8 × 1011 (low dose) or 1.7 × 1012 (high dose) particles/mL, with and without infusion of 50 ng/day of FGF for 3 weeks. The tissue forming in the chamber was harvested after each treatment for histologic processing and morphometric analysis of bone ingrowth. Statistical analysis was performed using parametric tests (ANOVA), nonparametric tests (KruskalWallis test) and post hoc tests. In the absence of particles, infusion of 50 ng FGF-2 per day yielded the greatest amount of bone ingrowth. The high dose of particles suppressed bone ingrowth into the chamber, but the low dose particles did not (p = 0.0002, 95% confidence limits = 9.1918.80). Infusion of 50 ng FGF-2 per day significantly increased net bone formation in the presence of high-dose UHMWPE particles (p = 0.039, 95% confidence limits = 1.416.79). There was a trend for decreased numbers of vitronectin-receptor positive (osteoclast-like) cells with the addition of FGF-2, compared to particles alone (p = 0.08). Local delivery of FGF-2 may prove useful in mitigating the adverse effects of wear debris (e.g., in treating early osteolytic lesions), and facilitating osseointegration of revision total joint replacements in situations where the bone bed is suboptimal and residual particles and granulomatous tissue are present. © 2003 Wiley Periodicals, Inc. J Biomed Mater Res 65A: 454461, 2003 [ABSTRACT FROM AUTHOR]