Inclusion criteria
We retrospectively reviewed infants with NBPP and adults with NNBPI for which a nerve transfer was performed. All patients were operated upon consecutively in the Peripheral Nerve Unit within the Department of Neurosurgery, University of Buenos Aires School of Medicine between January 1st, 2002 and December 31st, 2018. To be included in the study, both NBPP and NNBPP had to have undergone an extra or intra-plexal nerve transfer to regain a lost neurological function, but not nerve grafting for the same function. Therefore, in all included cases the specific nerve transfer was the only nerve surgical technique used. The patients who underwent grafting were excluded because if a nerve root was reconstructed with grafts and subsequently a nerve transfer was performed to recover the same or a similar function, it is not possible to determine if the recovery was either due to the root repair or to the nerve transfer.
The primary outcome was the PGS score, ranging from 1 to 4 (Table 1)[3]. Briefly, PGS score 1 represented the lowest independent volitional control, with British Medical research Council (MRC) grade 4 obtained in response to the donor command and MRC grade 0 in response to the acceptor command (minimum brain plasticity), whereas PGS score 4 was no noticeable contraction in response to the donor command and MRC score 4 in response to the acceptor command (maximum brain plasticity). Each muscle reinnervated following the transfer was tested three times in.
We also assessed patient compliance with the rehabilitation therapy using the 4-point Rehabilitation Quality Scale (RQS) [3, 14-16]. An RQS score of 1 represents patients who failed to attend any rehabilitation therapy or attended less than once weekly. Patients who had rehabilitation therapy at a regular center more than once weekly were classified as RQS of 2, and a score of 3 was attained when patients exhibited good adherence to a rehabilitation program at a nonspecialized neurorehabilitation center with periodic assessments at a specialized neuro-rehabilitation center. Patients who exhibited good adherence to a rehabilitation program at our specialized institution were assigned an RSQ score of 4 (Table 2).
Exclusion criteria were: 1/ patients who were lost to follow up; 2/ a clinical final result of the nerve transfer of MRC grade < 4; 3/ surgical repair more than 12 months after the trauma; 4/ NNBPP patients who had a concurrent brain contusion which might have affected their potential for brain plasticity, or 5/ follow-up of less than 12 months. Finally, we studied the influence of the compliance of each patient to the rehabilitation plan (table 2), the time from trauma to surgery and the time of follow-up, in order to determine whether these independent factors predict a response in the PGS score.
The study was performed in full accordance with the Declaration of Helsinki II and our institution´s ethics committee. All eligible patients -or responsible parents in children or NNBPP cases) were asked to participate in our study protocol, which included a throughout clinical examination. Written informed consent was obtained from each patient prior to study participation. Patient demographic characteristics — like gender, age, time from trauma to surgery, and the duration of follow-up — were recorded at the time of assessment.
Surgical strategies and techniques
General descriptions of the brachial plexus surgery and also of our rehabilitation program have recently been published [16-18]. All nerve transfers included a complete or a partial nerve section of the donor nerve and a direct coaptation with the recipient nerve. Nerve transfers only were used when proximal roots for grafting were unavailable as assessed by preoperative MRI and intraoperative inspection. Post-operative evaluations were performed every six months by at least two of the authors.
Post-operative clinical evaluation
Post-operative evaluations were performed every six months on a regular basis by at least two of the authors to reduce ascertainment bias. After a general clinical evaluation, we determined the PGS score. Postoperative evaluations were performed every 6 months
by at least two of the authors.
Statistical analysis
Continuous variables (age at surgery in months, time of follow-up in months, compliance to rehabilitation scale and plasticity scale) were summarized as means with standard deviations (SD) and minimum to maximum ranges, then tested for normality of distribution using the Shapiro-Wilk’s test. Since all four continuous variables were non-normally distributed and were being compared between subject groups (NBPP versus NNBPP), Kruskal-Wallis one-way analysis of variance (KW-ANOVA) was used to compare Medians and distributions. Distributions for both PGS scores and compliance were further compared using Pearson χ2 analysis or Fisher’s Exact test, as indicated. Within each subject group, the degree and significance of correlations between the four continuous variables were calculated using Pearson correlation coefficients, with r values < 0.30 considered weak, from 0.30-0.69 moderate, and ≥ 0.70 strong correlations. To identify predictors of the final neuroplasticity score, simple linear regression analysis was performed with the four independent variables — subject group, time to surgery, length of follow-up, and compliance score entered by forward entry. All tests were two-tailed, with p ≤ 0.050 set as the a-priori criterion for statistical significance.