Diffusion Tensor Imaging and Tractography ofTraumatic Brachial Plexus Palsies: PreliminaryExperience

PURPOSE In traumatic injuries of the brachial plexus the preoperative assessment of nerve root avulsions is essential for surgical planning. In the diagnostic workup MR myelography has almost completely replaced CT myelography as the first choice examination. To our knowledge there are no reports on the application of diffusion tensor Imaging to the study of traumatic brachial plexus palsies. Evaluation of the clinical feasibility of diffusion tensor tractography of the brachial plexus in a selected population of patients with traumatic brachial plexus injuries. MATERIALS & METHODS Ten patients (8 males, 2 females, mean age 27.4 years, range 16-52 years) with traumatic brachial plexus palsies (6 left, 4 right) have been examined with 1.5 T scanner (Siemens, Avanto) with the following imaging protocol: a) 3D MR myelography (3DMRM) with multiplanar reconstructions along the course of cervical nerve roots (3D T2 fast spin echo sequences, 0.7 mm isotropic voxel), b) 3D T2 STIR SPACE sequences (TR 2000, TE 149, TI 160, 1 mm isotropic voxel), C) DTI (64 contiguous axial sections from C3 to T2, TR 9200 TE 90, 12 directions, 2.2 mm isotropic voxel, 5 runs, peripheral triggering, FOV 250 mm). Diffusion tensor imaging data were preprocessed coregistering every single direction of each run with the first b0 and corrected for eddy currents with FSL (FMRIB Software Library; Oxford, UK) FDT tool. Tractography of the brachial plexus and cervical spinal cord was obtained placing multiple 3D ROIs along the extraforaminal course of C5-T1 nerve roots on MPR reformatted b0 coronal images. Nerve root avulsions were defined on MR tractography on the basis of a loss of continuity with the spinal cord fibers. Brachial plexus tractography diagnostic accuracy was assessed comparing it with 3DMRM findings on a per root basis on 100 nerve roots. RESULTS Three-dimensional MRM detected 18 avulsed nerve roots and 32 intact roots on the injured side. Brachial plexus tractography had 94% sensitivity, 87% specificity, 60% PPV and 98% NPV compared to 3DMRM, with 1 false negative and 12 false positives. The relatively elevated false positive rate is mainly related to the evaluation of C8-T1 nerve roots, which are displayed in a suboptimal way, due to their anatomical course adjacent to bone. CONCLUSION Diffusion tensor tractography is a cutting-edge technique for studying peripheral nerves, which cannot be performed as readily and reliably in the spinal cord and brachial plexus for technical and anatomical reasons such as pulsation artifacts, motion of organs in the thorax and large number of longitudinal fibers. Compared to 3DMRM brachial plexus tractography is characterized by an elevated diagnostic accuracy in the assessment of nerve root avulsions, which can be displayed easily as loss of continuity with spinal cord fibers. Although DTI of the brachial plexus is still an experimental technique which needs further testing in order to produce meaningful clinical results, our preliminary experience suggests its promising role in the assessment of nerve continuity in traumatic injuries of the brachial plexus.