Session Item

Osteoradionecrosis (ORN) of the mandible is a severe late complication of external beam radiotherapy (EBRT) for oral cavity cancer (OCC) that is difficult to manage and can have a significant impact on quality of life. Several risk factors for the development of ORN for head and neck cancers have been identified, however, knowledge of risk factors for ORN after postoperative EBRT (PORT) for OCC is limited. The goal of this study was to describe the incidence and determine risk factors of mandibular ORN in patients treated with PORT for OCC.

All OCC patients (N=227) treated with PORT at the Erasmus Medical Center between 2010 and 2018, with a minimum of one year disease free follow-up, were included in a retrospective cohort. The median age was 66 (range 24-91), 58.6% was male, and 48.9% of the primary surgeries involved a marginal or segmental mandible resection. Frequently prescribed dose schedules were 33x2Gy (49.3%) and 30x2Gy (27.8%). Follow-up was censored at the first of the following events: end of follow-up (standard follow-up 5 years), death, disease recurrence or additional head and neck RT. Dose-volume data were extracted from treatment plans. Cumulative incidence rates of mandibular ORN were computed using the Kaplan Meier method. Risk factors for the development of mandibular ORN were evaluated with Cox regression models (uni- and  multivariable).

We observed 41 cases of ORN of the mandible (crude incidence 18.1%, 39 within 5 years), with 13 mandibular fractures (31.7%) and 15 patients with orocutaneous fistulas (36.6%). 92.7% of patients were symptomatic or required treatment (CTCAE grade ≥2). The Notani score (based on panoramic radiograph), was available for 35 patients (87.8%), with N=18 Grade 3, N=10 Grade 2 and N=7 Grade 1. The estimated cumulative incidence was 8.4% (SE 1.8) at 1 year, 15.9% (SE 2.5) at 3 years, and 19.8% (SE 3.0) at 5 years (Figure 1). Univariable analysis (Table 1) showed that being an active smoker at diagnosis, N-stage, any mandible resection as primary surgery, fibula reconstruction of the mandible and tumor location at the floor of mouth, were significantly associated with increased ORN risk. Looking at dosimetric factors, we found that the Dmean of the mandible was higher in patients with ORN (mean 41.1 Gy) than in patients without ORN (35.6 Gy). This was similar for the DMax (69.4 Gy vs 65.7 Gy) and V60 (37.9% vs 22.9%). Multivariable analysis (HR, 95% CI) showed that smoking at diagnosis (2.17, 1.12-4.22) and V60 (1.03, 1.01-1.04) remained significant risk factors. 

Patients treated with PORT for OCC are at high risk for ORN, with a 5-year cumulative incidence of 19.8%. Smoking at diagnosis significantly increases the risk, with a 2.18 times higher chance of developing ORN. We also found a strong relation with mandibular RT dose. We found that even small changes in treatment planning can decrease the risk of ORN, as a 1% increase of V60 leads to a patient being 3% more likely to develop ORN.