Abstract
Introduction
The management of early-onset pediatric deformity has been revolutionized by the development of adjustable growing rods. These rods contain a rare-earth magnet that rotates when triggered by an externally applied magnetic field resulting in rod elongation. This patient group includes those with underlying neuromuscular conditions who occasionally require further MRI of the neuraxis during development and, given the magnetic properties of the rod, they are conventionally contraindicated for this imaging modality once implanted. We have investigated the behavior of the MAGEC (Ellipse Technology) rod to determine whether MRI adversely affects the elongation properties of the rod or indeed directly causes any detrimental consequences, including rod elongation, shortening, heating, or significant artifact to outline implications for this patient group.
Materials and Methods
We designed an in-vitro experiment using two MAGEC rods secured in an MRI compatible restraint system to investigate MRI behavior of magnetic growing rods during scanning. We used a 1.5T Philips MRI and a restraint system designed to secure the rods within this environment. Repeated gradient echo multiscan sequence MRI of the neuraxis was performed to evaluate whether the rods elongated, contracted, or rotated during scanning, while we used MRI compatible heat-sensors attached to the outer housing of the rods to assess temperature changes. We also assessed the ability of the rods to lengthen after completion of the MRI and a phantom model to evaluate the amount of artifact induced by the rare-earth magnet.
Results
During repeated MRI of the neuraxis, the MAGEC rod mechanism was not triggered; however, the rods did rotate to bring the internal rare-earth magnet in line with the surrounding electromagnetic field, while only small forces were generated causing rod movement within the restraint system. The ability of the rods to lengthen after MRI was not enhanced or impaired and no significant heating effect on the external rod housing was demonstrated. The assessment of imaging with a phantom model did reveal a significant degree of artifact up to 30 cm because of the presence of the magnet within the electromagnetic field and, therefore, only MRI of the head and cervical spine could be interpreted with normal placement of MAGEC rods within the pediatric patient.
Conclusion
This study demonstrated that there are no detrimental effects of MRI on the MAGEC rod. In addition, while the degree of the artifact demonstrated prohibits imaging of the thoracolumbar spine, the MRI images of the head and cervical spine remain unimpaired. Finally, we did not demonstrate any significant force acting on the rod by the electromagnetic field that would cause implant dislodgement or loosening during MRI.