Introduction: Cervical spine MRI with the neck in extension has been well described over the last 10 years, but its clinical value remains unknown.
Methods: We performed extension imaging in 60 patients in whom the initial neutral study showed borderline cord compression. Images were assessed using a previously validated grading system for cord compression. Multiple linear and area measurements were also obtained. Images were scored blindly and randomly. Inter- and intra-rater variability were determined in a subset of 20 cases. Independent clinical assessment utilised the Ranwat criteria.
Results: For most parameters inter/intra-observer variance of kappa/ICC > 0.6 was highly satisfactory. Standard MR was poor at discriminating between patients with and without myelopathy (ROC analysis, area under the curve (AUC), 0.52). This was considerably improved with extension imaging (AUC, 0.60), or by using the change in compression score between neutral and extension studies. Most measurements were not helpful; however, the ratio of cord area/CSF area at the level of maximum compression on extended images was the best discriminator (AUC, 0.71), as well as the presence of T2 change in cord substance (AUC, 0.68).
Conclusion: This is the first study to demonstrate added clinical value utilising extension MRI. In this cohort of difficult patients, when there was no T2 signal change in the cord, the presence of clinical myelopathy could only be predicted by utilising the data from extension imaging.