Impact of Admission Imaging Findings on Neurological Outcomes in Acute Cervical Traumatic Spinal Cord Injury

Variable and unpredictable spontaneous recovery can occur after acute cervical traumatic spinal cord injury (tSCI). Despite the critical clinical and interventional trial planning implications of this tSCI feature, baseline measures to predict neurologic recovery accurately are not well defined. In this study, we used data derived from 99 consecutive patients (78 male, 21 female) with acute cervical tSCIs to assess the sensitivity and specificity of various clinical and radiological factors in predicting recovery at one year after injury. Categorical magnetic resonance imaging parameters included maximum canal compromise (MCC), maximum spinal cord compression (MSCC), longitudinal length of intramedullary lesion (IML), Brain and Spinal Injury Center (BASIC) score, and a novel derived Combined Axial and Sagittal Score (CASS). Logistic regression analysis of the area under the receiver operating characteristic curve (AUC) was applied to assess the differential predictive value of individual imaging markers. Admission American Spinal Injury Association Impairment Scale (AIS) grade, presence of a spinal fracture, and central cord syndrome were predictive of AIS conversion at one year. Both BASIC and IML were stronger predictors of AIS conversion compared with MCC and MSCC (p = 0.0002 and p = 0.04). The BASIC score demonstrated the highest overall predictive value for AIS conversion at one year (AUC 0.94). We conclude that admission intrinsic cord signal findings are robust predictive surrogate markers of neurologic recovery after cervical tSCI. Direct comparison of imaging parameters in this cohort of patients indicates that the BASIC score is the single best acute predictor of the likelihood of AIS conversion.