MRI Can Predict Need for Surgery for Infants with Brachial Plexus Birth Injury
Andrea S Bauer, MD1, Peter Y. Shen, MD2, Ann E. Van Heest, MD3, M. Claire Manske, MD4, Martin J. Asis, MD5, Jennifer Chang, MD6 and Michelle A. James, MD7, 1Department of Orthopedic Surgery, Boston Children's Hospital, Harvard Medical School, Boston, MA, 2Kaiser Permanente, Santa Clara, CA, 3Department of Orthopaedic Surgery, University of Minnesota, Minneapolis, MN, 4Shriners Hospital for Children Northern California, Sacramento, CA, 5Gillette Children's Specialty Healthcare, St. Paul, MN, 6University of California Davis, Sacramento, CA, 7Department of Orthopaedic Surgery, Shriners Hospital for Children Northern California, Sacramento, CA
Introduction: Although magnetic resonance imaging (MRI) is an essential element of the diagnostic work-up for adults with brachial plexus injuries, it has not been routinely used for infants with brachial plexus birth injury (BPBI) as standard infant MRI protocols require sedation and don’t reliably establish whether they would benefit from surgery; instead, the decision to operate is based on trajectory of clinical recovery by age 5 months. The aim of this study was to develop a quick MRI protocol that can be performed without sedation or contrast and can identify infants who would benefit from surgery earlier than that decision could be made clinically.
Materials and Methods: This prospective multi-center study ((NAPTIME: Non-Anesthetized Plexus Technique for Infant MRI Evaluation) included infants with BPBI aged 28-120 days, from three tertiary care centers. Subjects had non-sedated, non-contrast rapid volumetric proton density MRIs on 3-Tesla scanners, with a total average scan time of 6.5 minutes. Neuroradiologists at each site calculated the NAPTIME nerve root injury score for subjects at their site (Figure 1); inter-rater reliability was performed on a subset of subjects. All subjects were evaluated with routine clinical exams up to 6 months of age, by which time the necessity of surgery was determined by the treating surgeon. Surgeons were blinded to MRI results. The ability of the NAPTIME score to discriminate the need for surgery was evaluated by estimating the area under the receiver-operating characteristic curve (AUC ROC), estimating sensitivity and specificity, across the range of NAPTIME scores.
Results: 102 families consented to enrollment. Of these, 65 infants successfully completed the NAPTIME MRI; 18 (28%) ultimately met clinical criteria for nerve surgery. The inter-rater reliability for the NAPTIME score was modest at 0.53 [0.29, 0.73]. The median NAPTIME score for subjects who met clinical criteria for nerve surgery was 16.2 (IQR 9.9, 18.9), while the median score for those who did not was 7.0 (IQR 5.0,10.5). The NAPTIME score predicted the need for surgery with an ROC AUC of 0.812 [0.688, 0.936]. A NAPTIME score of 13.25 offered specificity of 0.94 and sensitivity of 0.61 for need for nerve surgery, while a score of 10.5 gave specificity of 0.75 and sensitivity of 0.72.
Conclusion: Non-sedated, non-contrast MRI is a useful tool in determining severity of injury in BPBI. The NAPTIME score distinguishes which infants will meet criteria for reconstructive nerve surgery earlier than that decision can be made clinically.
Back to 2024 Abstracts