Early to midterm radiographic loosening of the press fit radial head prosthesis

Background Radiographic loosening is frequently seen around the radial head (RH) implant. The hypothesis of this study was that radiographic loosening will be more frequent in patients in which the RH prosthesis was implanted due to elbow trauma leading to instability that required lateral collateral ligament repair (LCL). Materials and methods A retrospective review of the patients who had RH implantation between 2012 and 2019 was performed. Evaluation included evidence of radiographic loosening, stress shielding, formation of heterotopic ossification, and rate of removal of the implant. Range of motion of the elbow at the latest follow up was also recorded. Results At a mean follow up of 18 months (range 1.4 – 80) eight out of 25 patients had radiographic loosening around the implant. The radial head implant was removed in 8 patients (in 3 due to painful radiographic loosening, in 4 due pain without radiographic loosening and in 1 due to infection). Radiographic loosening around the RH implant had no association with LCL repair (p=0.18) or future removal of implant (p=0.18) or the diagnosis of Monteggia lesion (p=0.68). In addition, removal of the RH implant had no association with prior LCL repair (p=0.60) or the diagnosis of Monteggia lesion (p=0.15). Stress shielding was seen in 5 patients and was of no clinical significance. Heterotopic ossification was seen in 12 patients and was classified as Class I in 3, IIA in 3, IIC in 6, according to the Hastings Classification. The average flexion-extension arc was 23° to 130°, and average pronation-supination was 76° to 69°. Conclusion One third of the patients had radiographic loosening around the RH implant at a mean follow up of 18 months. Pain with or without radiographic loosening were the main reasons for removal of the implant. No associations were found between the development of radiographic loosening and LCL repair at the time of RH replacement. Limitations of this study are: (a) the retrospective design (b). the small sample size and the possibility of a type II statistical error.