Significance of site-specific radiation dose and technique for success of implant-based prosthetic rehabilitation in irradiated head and neck cancer patients—A cohort study

Abstract

Background: Radiotherapy aggravates implant-based prosthetic rehabilitation in patients with head and neck cancer. Purpose: To evaluate the impact of radiation dose at implant and parotid gland site for prosthetic rehabilitation. Material and methods: The retrospective study includes 121 irradiated head and neck cancer patients with 751 inserted implants. Radiation doses on implant bed and parotid gland site were recorded by 3-dimensional modulated radiation plans. Implant success was clinically and radiographically evaluated according to modified Albrektsson criteria and compared to treatment- and patient-specific data. Results: Implant overall survival after 5 years was 92.4% with an implant success rate of 74.9%. Main reasons for implant failure were marginal bone resorption (20.9%), implant not in situ or unloaded (9.6%) and peri-implantitis (7.5%). A mean radiation dose of 62.6 Gy was applied with a mean parotid dose of 35 Gy. Modulating radiation techniques went along with lower grades of xerostomia (p < 0.001). At implant site mean doses of 57.5, 42.0, and 32.3 Gy were recorded for oral, oropharyngeal, and hypopharyngeal/laryngeal carcinoma, respectively. Implant success inversely correlated to radiation dose at implant site. Strong predictors for implant failure in uni- and multivariate analysis were implant-specific dose >50 Gy (HR 7.9), parotid dose >30 Gy (HR 2.3), bone (HR 14.5) and soft tissue (HR 4.5) transplants, bad oral hygiene (HR 3.8), nonmodulated radiation treatment planning (HR 14.5), and nontelescopic prosthetics (HR 5.2). Conclusion: Radiotherapy impedes implant success in a dose-dependent manner at implant site. Modern radiation techniques effectively reduce xerostomia favoring implant-based prosthetic rehabilitation. Implantation in bone grafts is more critical and telescopic-retained overdentures should be preferred.