OBJECTIVE Nerve reconstruction after 6 months of denervation time in brachial plexus injuries (BPIs) can be inconsistent. A dilemma exists when the use of critical donor nerves for nerve transfers may lead to unreliable outcomes that would waste the donor nerve. The purpose of this study was to evaluate the long-term outcomes of elbow and shoulder function in patients with BPIs receiving nerve reconstruction in the delayed setting (i.e., 6-12 months after injury). METHODS Data from patients with delayed BPIs who received a nerve transfer (including proximal and distal nerve transfer/grafting) at a tertiary medical center were retrospectively collected from January 1999 to March 2020. Demographics, extent of injury, mechanism of injury, and reconstructive methods were collected. Patients were categorized into two groups: non-pan-plexus BPI (C5-6, C5-7, and C5-8) and pan-plexus BPI (C5-T1). Acceptable outcome was defined as elbow flexion ≥ M3 status or shoulder abduction ≥ 60°. RESULTS Sixty-four patients were included in the study. The average time from injury to nerve reconstruction was 236 (range 180-441) days, and the average follow-up time was 66 months. In the non-pan-plexus BPI group (n = 43 patients), 74.4% of patients demonstrated M3 elbow flexion, and 48.8% of patients demonstrated M4 elbow flexion. Double fascicular transfer yielded better results and faster recovery than a single fascicular transfer. In the pan-plexus BPI group (n = 21 patients), 38.1% of patients reached M3 elbow flexion and 23.8% attained M4 elbow flexion. In the non-pan-plexus BPI group, the recovery rate of acceptable shoulder abduction was 53.5%, but only 23.5% of pan-plexus patients with BPI achieved acceptable shoulder abduction. CONCLUSIONS Nerve reconstruction can effectively restore functional elbow flexion and acceptable shoulder abduction in non-pan-plexus patients with BPI in the delayed setting. However, neither acceptable elbow flexion nor shoulder abduction could be consistently achieved in pan-plexus BPI. Judicious use of the donor nerves in pan-plexus injuries is required, in addition to preserving a donor nerve for a backup plan such as free-functioning muscle transplantation or tendon transfers.

中文翻译：

---

臂丛神经损伤延迟神经重建：风险值得吗？

目的 臂丛神经损伤 (BPI) 去神经时间 6 个月后的神经重建可能不一致。当使用关键的供体神经进行神经移植时，可能会导致不可靠的结果，从而浪费供体神经，这就存在一个困境。本研究的目的是评估 BPI 患者在延迟环境（即受伤后 6-12 个月）接受神经重建的肘部和肩部功能的长期结果。方法 回顾性收集1999年1月至2020年3月在三级医疗中心接受神经移植（包括近端和远端神经移植/移植）的延迟性BPI患者的资料。人口统计学、损伤程度、损伤机制和重建方法被收集。患者被分为两组：非全丛 BPI（C5-6、C5-7 和 C5-8）和全丛 BPI (C5-T1)。可接受的结果定义为肘部屈曲≥M3状态或肩部外展≥60°。结果 64 名患者被纳入该研究。从受伤到神经重建的平均时间为236天（范围180-441）天，平均随访时间为66个月。在非全丛 BPI 组（n = 43 例患者）中，74.4% 的患者表现出 M3 肘部屈曲，48.8% 的患者表现出 M4 肘部屈曲。双束转移比单束转移产生更好的结果和更快的恢复。在全丛 BPI 组（n = 21 名患者）中，38.1% 的患者达到 M3 肘部屈曲，23.8% 的患者达到 M4 肘部屈曲。在非全丛BPI组中，可接受的肩外展恢复率为53.5%，但全丛BPI患者中只有23.5%达到了可接受的肩外展。结论 神经重建可以有效恢复延迟治疗的非全丛神经 BPI 患者的肘部屈曲功能和可接受的肩部外展。然而，在全丛 BPI 中，无法始终如一地实现可接受的肘部屈曲和肩部外展。除了保留供体神经用于备用计划（例如自由功能的肌肉移植或肌腱移植）之外，还需要在全神经丛损伤中明智地使用供体神经。