Correlation of magnetic resonance diffusion tensor imaging and clinical findings of cervical myelopathy

Despite significant advances in the development of diagnostic technology, the diagnosis of cervical myelopathy (CM) still remains based on the clinical findings, which do not provide the means for a sufficiently accurate diagnosis. Furthermore, conventional magnetic resonance imaging (MRI) using T1- and T2-weighted sequences lacks sensitivity to detect and characterize spinal cord lesions. Considering these uncertainties, several investigators have assessed the diagnostic value of diffusion tensor imaging (DTI), an advanced MRI technique that measures the diffusion of water molecules.

To determine the diagnostic value of DTI in CM in reliably characterizing spinal lesions and in associating them with the clinical findings.

Prospective cohort study.

Fifteen CM patients and five healthy volunteers without a history of neurological disorders or of symptoms as controls.

Symptoms and signs of CM were evaluated by the use of a modified Japanese Orthopedic Score and the other clinical findings. T2-weighed MRI was used to note the number of compressed levels. Diffusion tensor imaging results were measured according to two parameters, fractional anisotropy (FA) and apparent diffusion coefficient (ADC), at anterior, lateral, and posterior regions of interest (ROIs) in each of five cervical vertebrae, C3-C7.

On diagnosis of CM by clinical evaluation and findings from T2-weighted MRI, the 15 subjects were assigned to two subgroups based on complaints, symptoms, and signs. The nine subjects who had typical CM symptoms such as motor weakness, gait disturbance, clumsiness of the hands, and unilateral hypesthesia were assigned to the paralysis subgroup. The other six subjects, whose main symptom was pain and who had vague signs of upper motor neuron injury despite a definitive finding of CM by T2-weighted MRI, were assigned to the pain subgroup. Once assignments had been made, subjects underwent DTI done by the use of the same scanner as for T2-weighted MRI. Results of DTI for each subgroup and controls were averaged, and the mean was used for comparisons. Diffusion tensor imaging results from the paralysis subgroup were sorted into affected and unaffected sides according to the presence or the absence of symptoms.

The paralysis subgroup and the pain subgroup had similar findings from T2-weighted MRI on presentation. The paralysis subgroup had statistically significantly decreased FA values in the anterior and lateral ROIs on the affected side and in the anterior ROIs on the unaffected side, compared with controls. The paralysis subgroup also had statistically significantly increased ADC values in the anterior ROIs of the affected side, compared with controls. The pain subgroup showed significantly increased ADC values in anterior, lateral, and posterior ROIs.

Use of DTI to quantitatively compare compression in the cervical spinal cords of CM subjects and healthy controls explained individual differences in the clinical findings in the subjects. These findings even applied to CM subjects whose compressed spinal cords looked similar on conventional T2-weighted MRI. Therefore, DTI provided more accurate and reliable information than did conventional T2-weighted MRI about the relationship between spinal cord structure and clinical presentation of CM. Based on our DTI findings, we hypothesized that different clinical findings in CM are attributable to the stage of progression and the severity of pathologic change at presentation. We anticipate that the use of DTI to quantify the extent of myelopathological changes in CM could be more reliable than any other existing diagnostic tools and might provide invaluable information about selecting the optimal treatment for CM and predicting surgical outcomes and prognosis.