Risk factor assessment for the development of osteoradionecrosis.

PURPOSE Osteoradionecrosis (ORN) of the jaws has been extensively studied. However, controversy still exists regarding its etiology, risk factors, and the underlying mechanism of disease. The purpose of this study was to identify and evaluate significant risk factors for the development of ORN. PATIENTS AND METHODS This was a retrospective cohort study of 82 Massachusetts General Hospital patients radiated for head and neck cancer between 1984 and 2005. Patient records were reviewed to collect demographic information, medical and dental history (including dental intervention or trauma), tumor specific data, treatment details, and follow-up. Biologic variables (ie, age and gender) of potential significance were also evaluated. The major outcome variable was the development of ORN or lack of development of ORN. The time from radiation to ORN, or for non-ORN patients, time to last follow-up visit, was computed. Univariate analyses identified candidate variables associated with ORN (P < .15). Cox proportional hazards regression was used to evaluate these candidate variables as well as biologically relevant variables. Significant prognostic factors for the development of ORN (P < .05) were identified. RESULTS Multivariate regression identified the following variables as significantly associated with decreased ORN risk: higher body mass index (P = .02) and use of steroids (P = .02). Radiation dose greater than 66 Gray (P = .03) was associated with an increased ORN risk. CONCLUSIONS Optimization of nutritional status, use of steroids, and limitation of total radiation dose may minimize the risk of ORN.