Sensorimotor Outcomes of Upper Extremity End-to-Side Nerve Transfers: A Meta-analysis

Background: End-to-side nerve transfer (ETSNT) for treatment of peripheral nerve injuries is controversial given the myriad anatomic locations, injury types, and indications. Efficacy of ETSNT remains debated. We hypothesized differences in age, sex, transfer location, and time to surgery influence outcomes.

Methods: We performed a search of the PubMed database for ETSNT in the upper extremity from 1988 to 2018. Age, sex, transfer location, time to surgery, donor and recipient axons, and strength and sensation outcomes as measured by Medical Research Council scale were extracted from articles. Meaningful recovery was classified as Medical Research Council Grade 3 or greater. Association between meaningful recovery and younger (<25) and older (≥25) patients, injury mechanism, sex, transfer location, donor axons, and recipient axons were calculated using a χ 2 or Fisher exact test. A logistic mixed effect model was used with time to surgery, age (categorical), transfer location, and injury type as a fixed effect, and a random paper effect was included to account for correlation among patients from the same paper.

Results: One hundred fifteen patients from 11 studies were included. Neither age (continuous variable, P = 0.68) nor time to surgery ( P = 0.28) affected meaningful recovery. Injury mechanism, sex, and younger age (<25 vs ≥25 years) were not associated with meaningful recovery. Within the brachial plexus ETSNT demonstrated median M4 ± 1 postoperative strength, with trunks/cords as the primary axon donor ( P = 0.03). The musculocutaneous nerve demonstrated promising but variable results in 31 patients with median strength M3 ± 4. Digital nerves consistently demonstrated meaningful sensory recovery as both donor and recipient axons (15 of 15, 100%). Logistic regression analysis demonstrated that odds of meaningful recovery after ETSNT are significantly greater for transfers within the brachial plexus compared with the distal arm (odds ratio, 41.9; 95% CI, 1.1-1586.7, P = 0.04), but location does not significantly affect meaningful recovery ( P = 0.22).

Conclusions: Patients undergoing ETSNT for digital nerve injury demonstrated meaningful recovery. End-to-side nerve transfer seems to be more efficacious when performed within the brachial plexus. This study did not find sex, injury mechanism, or time to surgery to significantly affect meaningful recovery. Additional study is needed to better evaluate the effectiveness of ETSNT in the upper extremity.