Implant Prosthodontic Rehabilitation of Fibula Free-Flap Reconstructed Mandibles: A Memorial Sloan-Kettering Cancer Center Review of Prognostic Factors and Implant Outcomes

Purpose: This study aimed to estimate the cumulative survival rates (CSRs) of implants placed in reconstructed mandibles and to identify prognostic factors that may influence implant survival. Materials and Methods: The charts of 24 patients (10 male, 14 female) who had undergone mandibular resection and reconstruction with fibula free-flaps treated with implant-supported prostheses from April 1986 through December 2001 were reviewed. Information on demographics, surgical characteristics, treatment modalities, dentition, implant parameters, prostheses, and hyperbaric oxygen therapy (HBO) was gathered. Kaplan-Meier survival estimates were generated for the 100 implants that satisfied the inclusion criteria. Multivariate Cox proportional hazards regression models accounting for correlated implants within subjects were developed to identify prognostic factors for implant survival. Results: Ninteen implants had been placed in native mandible (3 in irradiated bone) and 81 in fibula bone flap. Six implants failed during the follow-up period (mean 51.7 months). The overall 5- and 10-year CSRs were 97.0% and 79.9%, respectively. In the univariate analysis, variables associated with implant survival were age, gender, chemotherapy, radiation therapy, HBO, irradiated bone, implant diameter, xerostomia, trismus, opposing dentition, and type of prosthesis. At 5 years, the CSR of implants in patients with HBO was 86.7%; HBO was statistically associated with an increased risk for implant failure (P = .005, hazard ratio = 19.79, 95% CI: 2.42 to 161.71). Discussion: The CSR was lower when implants were placed in a previously irradiated mandible. There is still a lack of reliable clinical evidence to support the effectiveness of HBO in these patients. Conclusions: A high survival rate was demonstrated for implants placed in fibula free-flap reconstructed mandibles. The finding that HBO was a risk factor can probably be attributed to the small sample size; further study is needed in this patient population. Keywords: fibula, implant survival, mandibular reconstruction