Bone graft healing in reconstruction of maxillary atrophy.

PURPOSE: Evaluate correlations between volume change for iliac crest bone grafts in maxillary reconstruction (graft volume change [GVC]) and bone mineral density (BMD), bone volume fraction (BVF), hematologic bone metabolic factors (I), and identify indicators of implant failure (II). MATERIAL AND METHODS: Forty-six consecutive patients had their edentulous atrophic maxilla reconstructed with free autogenous bone grafts from anterior iliac crest. Endosteal implants were placed 6 months after graft healing. Computer tomography was performed after 3 weeks and 6 months after grafting. Bone biopsies were taken from the internal table of donor site for calculation (BVF), and blood samples were collected. Implant stability was measured at placement with resonance frequency analysis and expressed as implant stability quotient (ISQ). Implant failure was registered. RESULTS: GVC in onlay bone graft was 37%. The BVF in iliac crest biopsies was 32%. Serum-IGFBP3 differed with 79% of the samples over normal range. Fifteen patients had one or more implant failures prior to loading (early failures). Forty-two patients were followed for a minimum of 3 years after implant loading and, in addition, 6/42 patients had one or more implants removed during the follow-up (late failures). GVC correlated to decreased BMD of lumbar vertebrae L2-L4 (Kruskal-Wallis test, p=.017). No correlation was found between GVC and hematologic factors (Pearson correlation test) or between GVC and BVF (Kruskal-Wallis test). No correlation was found between ISQ and GVC (Pearson correlation test, p=.865). The association between implant failures and the described factors were evaluated, and no significant correlations were found (unconditional logistic regression). CONCLUSION: Onlay bone grafts decrease 37% during initial healing period, which correlate to BMD of lumbar vertebrae L2-L4. No other evaluated parameters could explain GVC. The evaluated factors could not explain implant failure.