LITERATURE REVIEW J Neurosurg 130:1333–1345, 2019 Timing of surgery in traumatic brachial plexus injury: a systematic review Enrico Martin, BS,1,2 Joeky T. Senders, BS,1,2 Aislyn C. DiRisio, BS,2 Timothy R. Smith, MD, PhD, MPH,2 and Marike L. D. Broekman, MD, PhD, JD1,2 1Department of Neurosurgery, University Medical Center Utrecht, The Netherlands; and 2Computational Neuroscience Outcomes Center, Department of Neurosurgery, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts OBJECTIVE Ideal timeframes for operating on traumatic stretch and blunt brachial plexus injuries remain a topic of de- bate. Whereas on the one hand spontaneous recovery might occur, on the other hand, long delays are believed to result in poorer functional outcomes. The goal of this review is to assess the optimal timeframe for surgical intervention for traumatic brachial plexus injuries. METHODS A systematic search was performed in January 2017 in PubMed and Embase databases according to the PRISMA guidelines. Search terms related to “brachial plexus injury” and “timing” were used. Obstetric plexus palsies were excluded. Qualitative synthesis was performed on all studies. Timing of operation and motor outcome were collect- ed from individual patient data. Patients were categorized into 5 delay groups (0–3, 3–6, 6–9, 9–12, and > 12 months). Median delays were calculated for Medical Research Council (MRC) muscle grade ≥ 3 and ≥ 4 recoveries. RESULTS Forty-three studies were included after full-text screening. Most articles showed significantly better motor outcome with delays to surgery less than 6 months, with some studies specifying even shorter delays. Pain and quality of life scores were also significantly better with shorter delays. Nerve reconstructions performed after long time intervals, even more than 12 months, can still be useful. All papers reporting individual-level patient data described a combined total of 569 patients; 65.5% of all patients underwent operations within 6 months and 27.4% within 3 months. The highest percentage of ≥ MRC grade 3 (89.7%) was observed in the group operated on within 3 months. These percentages de- creased with longer delays, with only 35.7% ≥ MRC grade 3 with delays > 12 months. A median delay of 4 months (IQR 3–6 months) was observed for a recovery of ≥ MRC grade 3, compared with a median delay of 7 months (IQR 5–11 months) for ≤ MRC grade 3 recovery. CONCLUSIONS The results of this systematic review show that in stretch and blunt injury of the brachial plexus, the optimal time to surgery is shorter than 6 months. In general, a 3-month delay appears to be appropriate because while recovery is better in those operated on earlier, this must be considered given the potential for spontaneous recovery. https://thejns.org/doi/abs/10.3171/2018.1.JNS172068 KEYWORDS brachial plexus injury; surgery; timing; outcome; systematic review; trauma; peripheral nerve RAUMATIC brachial plexus injury is a rare yet dev- injury, which makes almost every case unique. There is astating event that is most commonly noted in an overall consensus that in brachial plexus injury, elbow young, active adult males involved in traffic ac- flexion is the first goal of repair, followed by shoulder sta- Tcidents.45,54,59 Although the injury itself is not fatal, life- bility.7,9,14,18,33,40,55,71,87 Many other aspects of treating these long disability usually follows and can be difficult to re- patients, including the ideal time for operating on stretch verse.45,54,59 Most of the knowledge about the treatment of and blunt injuries, remain a topic of discussion among pe- traumatic brachial plexus injury comes from single-center ripheral nerve surgeons.7 observational retrospective studies. This is partially due to While some authors advocate for very early repair of the large heterogeneity in presentation of brachial plexus traumatic brachial plexus injury,13,38,41,51,88 others suggest ABBREVIATIONS AFRS = average final result of surgery; DASH = Disability of the Arm, Shoulder and Hand questionnaire; IQR = interquartile range; MRC = Medical Research Council; PRISMA = Preferred Reporting Items for Systematic Reviews and Meta-Analysis; SF-36 = 36-Item Short-Form Health Survey; VAS = visual analog scale. ACCOMPANYING EDITORIAL See pp 1330–1332. DOI: 10.3171/2018.2.JNS1881. SUBMITTED August 20, 2017. ACCEPTED January 10, 2018. INCLUDE WHEN CITING Published online June 1, 2018; DOI: 10.3171/2018.1.JNS172068. ©AANS 2019, except where prohibited by US copyright law J Neurosurg Volume 130 • April 2019 1333 Unauthenticated | Downloaded 10/06/21 09:48 AM UTC E. Martin et al. that long delays can still result in good functional recov- studies included are summarized using box and bar plots. eries.50,58,70 Many groups recommend waiting at least 3 Further subgroup analysis was performed based on surgi- months before surgery1,14,18,19,24,44,46,59,67,72 because sponta- cal delay. Level of injury groups were made, distinguish- neous recovery might occur.21 Many groups also discour- ing C5–6, C5–7, C5–T1, and infraclavicular lesions. Box age delays longer than 6 months8,18,24,33,44,53,55,67,91 because plots were made using the statistical program R (version long denervation times can decrease muscle strength. This 3.3.2, R Core Team, 2016). results from a combination of three processes: a reduced regenerative capacity in chronically axotomized proximal Results nerve stumps, a decreased capacity of distal nerve stumps After removal of duplicates, a total of 1161 citations to support regenerating axons, and an inability of atrophied 29,30,34 were identified in the PubMed and Embase databases. muscle to recover from chronic denervation. Timing One hundred ninety-four potentially relevant articles were is essential because nerve axons regenerate at a speed of selected through title/abstract screening, of which 43 stud- 81 only 1–2.5 mm per day, and denervation times include ies were selected for qualitative synthesis after full-text both the delay in surgery and also the time before a nerve screening (Fig. 1). reaches its target. The purpose of this study is to review how the length of Study Characteristics delay to surgery affects outcomes. The maximum length The majority of studies included in our study were ret- of delay at which surgeons should still be able to perform rospective observational studies, and only 7 of the includ- successful nerve repairs is also reviewed. ed studies were prospective cohort studies23,50,63,77,90,94,95 (Table 1). The studies included a total of 2204 patients, and Methods among studies that reported sex, 89.15% of patients were Literature Search male. The median age of patients in the included studies A systematic search was performed in both PubMed was 28 years, with an interquartile range (IQR) of 26–32.6 and Embase databases according to the PRISMA (Pre- years. Surgeries performed included nerve transfers using ferred Reporting Items for Systematic Reviews and Meta- donor nerves or recipient nerves, nerve grafts, simple neu- Analysis) guidelines, in order to identify all potentially rolysis, or a combination of these procedures. The median relevant articles as of January 2017. The search string was surgical delay was 6 months (IQR 5–7.65 months), with a built with the help of a professional librarian using search range of 0–240 months. Follow-up times ranged from 6 terms related to “brachial plexus injury” and “timing.” The months to more than 38 years, with a median of 3.45 years exact search syntaxes for PubMed and Embase are shown (IQR 2–5.15 years). in the Appendix. Studies were included that looked at tim- ing of operation in traumatic brachial plexus injury and Outcomes in Motor Function Assessed by MRC Grade either showed clear conclusions on timing of operation or All studies that assessed Medical Research Council included both timing and postoperative outcome in tables. (MRC) muscle grade outcomes found a significant20,23, Exclusion criteria included lack of full text, obstetric bra- 31,41,49,57,68,78,83 or nonsignificant4,19,61,64,67,91,93 improved recov- chial plexus surgery, irrelevant data, case series with fewer ery after early operation versus late operation (Table 2). than 10 patients, review articles, patients with secondary Twelve studies dichotomized both surgical delay and MRC operations, overlapping data, and languages other than grade outcomes,4,23, 31,43,49,50,57,67,68, 78, 91,93 whereas 3 groups English, Dutch, French, and German. The initial review dichotomized surgical delay20,41,83 or MRC grade outcome was conducted by two independent authors (E.M. and only.19,61,64 Several studies investigated if operations per- J.T.S.). Disagreements were solved through discussion, in formed even earlier than 6 months are beneficial for motor which one other author was involved (M.L.D.B.). outcome; cutoff points were 2 months,41 3 months,4,57,78,91 and 4 months.49,83 All studies showed better results for op- Data Extraction and Synthesis erations performed even earlier than 6 months after injury. 49 4 41 Data extracted from each study included year of pub- However, only Liu et al., Altaf et al., and Jivan et al. lication, study type, number of patients, range and me- were able to show statistical significance. Whereas Samii et al.67 showed no useful recovery in dian age of patients, gender percentages, levels of bra- 4 chial plexus lesion included, cause of injury, preoperative operations performed after 12 months and Altaf et al. even reported no useful recoveries after 5 months’ delay, assessment, range and median delay of surgery, type of 50 43 surgery performed, outcome of timing, outcome measure, Liverneaux et al.