Back to 2026 Abstracts
Magnetic Resonance Neurography as the Diagnostic Anchor in Multimodal Prediction of C5 Graftability: Evidence from 402 Adult Brachial Plexus Injuries
Ying - Hsuan Lee, MD
1, Johnny Chuieng-Yi Lu, MD, MSCI
2, Tommy Nai-Jen Chang, MD
3; David Chwei-Chin Chuang, MD
3(1)Chang Gung Memorial Hospital, Taoyuan City, Taiwan, (2)Plastic and Reconstructive Surgery, Chang Gung Memorial Hospital, Taoyuan, Taiwan, (3)Chang Gung Memorial Hospital, Taoyuan, Taiwan
Introduction: Accurate preoperative assessment of C5 nerve root graftability is crucial for optimal surgical planning in brachial plexus injuries. The distinction between ruptured and avulsed spinal nerves determines reconstructive options, yet no single diagnostic test has proven universally reliable. This study evaluates the diagnostic accuracy of MRN (magnetic resonance neurography) compared to clinical examination and electrodiagnostic studies in predicting C5 graftability.
Materials & Methods: A retrospective analysis was conducted on 402 adult patients with traumatic brachial plexus injuries who underwent nerve reconstruction surgery at tertiary medical center between September 2008 and November 2024. Patients were categorized into panplexus (C5-T1, n=166) and non-panplexus (n=236) injury patterns. Diagnostic modalities included physical examination (root avulsion pain, Tinel sign, muscle power assessment), electromyography (EMG) of C5 paraspinal and rhomboid muscles, and high-resolution MRN with avulsion severity scoring. Intraoperative surgical inspection of C5 appearance served as the reference standard.
Results: Among 402 patients, 207 (51.5%) had C5 avulsion and 195 (48.5%) had graftable C5 roots. MRN demonstrated the highest predictive value with an odds ratio of 9.171 (95% CI: 5.795-14.514, p<0.001). In multivariate analysis, MRN achieved an area under the curve of 0.801 when combined with Tinel sign and rhomboid muscle power assessment. Diagnostic accuracy varied significantly between injury patterns: For panplexus injuries, a parsimonious model combining MRI and Tinel sign yielded an AUC of 0.828, while in non-panplexus injuries, the optimal model included Tinel sign, levator scapulae muscle power, and MRN (AUC 0.766). Tinel sign was the only physical examination finding that remained statistically significant across all patient subgroups (p<0.001). EMG of the C5 paraspinal and rhomboid muscles showed limited diagnostic value in determining C5 graftability.
Conclusions: MRN serves as the primary diagnostic tool for C5 nerve root graftability assessment with superior accuracy. Optimal prediction requires an injury-specific approach, as diagnostic performance varies between injury patterns. Combining MRN with selective clinical findings (Tinel sign and muscle power assessments) creates synergistic effects that enhance decision-making beyond individual tool performance. This multimodal, MRN-centered strategy enables precise surgical planning and improved brachial plexus reconstruction outcomes
Back to 2026 Abstracts
