gms | German Medical Science

Objective: Diffusion tensor magnetic resonance imaging (DTI) is a frequently used technique for the preoperative work-up of intracerebral tumors. A shift of fiber tracts can be made visible and might influence surgical tactics in order to reduce morbidity. In contrast, DTI of the spinal cord is technically challenging due to low signal intensity-to-noise ratio of the small volume of cord tissue, and by artefacts related to an inhomogeneous magnetic field. Recent studies in cervical spinal cord injuries found that in contrast to conventional MR imaging DTI allows the quantification of spinal cord injury which correlates with the extend of the injury. Here, we report our first experience of preoperative DTI in cervical intramedullary pathologies.

Methods: Nine patients with cervical intramedullary pathologies were measured preoperatively in a 1.5 tesla MRT. Additional to conventional MR-imaging diffusion tensor imaging (DTI) sequences were performed, prolonging the average scan time approximately 5 to 7 minutes. From the DTI sequences fiber tracts were visualized and the ascertainable fractional anisotropy (FA) as specific value for the fiber integrity was quantified within the posterior funiculus. Furthermore the FA was measured in areas of the intramedullary edema, within the lesion, and in spinal cord tissue unaffected by the intramedullary pathology.

Results: We found the FA inside a cervical intramedullary glioblastoma and a pilocytic astrocytoma were significantly reduced about one standard deviation compared to those areas unaffected by the intramedullary tumor. In contrast the FA inside areas of intramedullary edema tended to display increased values enabling discrimination between edema and tumor tissue. Intramedullary cavernomas appear to be a subgroup with poor differentiation between lesion and surrounding tissue due to artefacts. Furthermore, there seems to be the possibility to discriminate different degrees of cervical myelopathy. Whether this discrimination allows a prediction of the postsurgical outcome has to be evaluated in further studies.

Conclusions: Identifying the FA in our patient collective under standard measuring conditions allowed a confident differentiation between tumor and edema of the spinal cord. Further technical improvements such as pulse-triggered imaging have the potential to increase the sensitivity of DTI. Our experience suggests that spinal DTI might become a supplementation to standard contrast MRI in patients with intramedullary pathologies.