Predicting Healthy C5 Spinal Nerve Stumps Eligible for Grafting with MRI, Tinel Test, and Rhomboid Electromyography: A Retrospective Study of 295 Consecutive Brachial Plexus Surgeries

Background MRI, Tinel test, and rhomboid electromyography (EMG) can be used to predict whether C5 spinal nerve stumps are healthy and eligible for grafting in acute adult brachial plexus injuries, but their comparative diagnostic efficacies have not been evaluated. Purpose To compare diagnostic performances of MRI, Tinel test, and rhomboid EMG in predicting healthy C5 spinal nerve stumps that are eligible for grafting. Materials and Methods This retrospective study included consecutive adult patients with acute brachial plexus injury who underwent microreconstructions between January 2008 and December 2018. Healthy C5 spinal nerve stumps eligible for grafting with preceding normal intradural nerve rootlets were diagnosed by an experienced neuroradiologist using an avulsion severity score system based on fast imaging employing steady-state acquisition or FIESTA neurography, which separates intradural nerve rootlets into normal, partial mild, severe, and complete avulsions. Tinel test and rhomboid EMG results were assessed by an experienced nerve surgeon and neurologist, respectively. The accuracy, sensitivity, and specificity of the three tests were compared using microdissection as the reference standard. P < .05 was considered to indicate a significant difference. Results A total of 251 patients (mean age, 31 years ± 13 [standard deviation]; 217 men) with brachial plexus injuries who had ­undergone MRI (n = 251), Tinel test (n = 235), rhomboid EMG (n = 181) and MRI, Tinel test, and EMG (n = 172) were ­included. Accuracy, sensitivity, and specificity, respectively, in predicting healthy C5 spinal nerve stumps eligible for grafting were 93% (233 of 251), 84% (58 of 69), and 96% (175 of 182) for MRI; 59% (139 of 235), 56% (37 of 66), and 60% (102 of 169) for Tinel test; and 39% (71 of 181), 85% (43 of 50), and 21% (28 of 131) for rhomboid EMG. MRI (area under the receiver operating characteristic curve [AUC], 0.90; P < .001) ­outperformed MRI and Tinel test (AUC, 0.74), Tinel test (AUC, 0.59), and rhomboid EMG (AUC, 0.53). Conclusion MRI performed best in the prediction of healthy graftable C5 spinal nerve stumps in acute adult brachial plexus injuries. © RSNA, 2021 Online supplemental material is available for this article.