Intercostal Nerve Transfer for Biceps Reinnervation in Obstetrical Brachial Plexus Palsy: A Preferred Reporting Items for Systematic Reviews and Meta-Analysis for Individual Patient Data Systematic Review using Individualized Fusion and Comparison to Supraclavicular Exploration and Nerve Grafting

Introduction: The objective of this study was to search existing literature on nerve reconstruction surgery in patients with obstetric brachial plexus palsy to determine whether treatment with supraclavicular exploration and nerve grafting produced better elbow flexion outcomes compared to intercostal nerve transfer. Methods: This study was a systematic review following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis for Individual Patient Data guidelines. A systematic search was conducted using multiple databases. An ordinal regression model was used to analyze the effect of using supraclavicular exploration and nerve grafting or intercostal nerve on elbow flexion with the two scores measured: elbow flexion Medical Research Council scores and Toronto active movements scale scores for elbow flexion. Results: A final patient database from 6 published articles consisted of 83 supraclavicular exploration and nerve grafting patients (73 patients with Medical Research Council and 10 patients with Toronto score) and 7 published articles which consisted of 131 intercostal nerve patients (84 patients with Medical Research Council and 47 patients with Toronto scores). Patients who underwent supraclavicular exploration and nerve grafting presented with an average Medical Research Council score of 3.9 ± 0.72 and an average Toronto score of 6.2 ± 2.2. Patients who underwent intercostal nerve transfer presented with an average Medical Research Council score of 3.9 ± 0.71 and an average Toronto score of 6.4 ± 1.2. There was no statistical difference between supraclavicular exploration and nerve grafting and intercostal nerve transfer when utilizing Medical Research Council elbow flexion scores (ordinal regression: 0.3821, standard error: 0.4590, p = 0.2551) or Toronto Active Movement Scale score for elbow flexion (ordinal regression: 0.7154, standard error: 0.8487, p = 0.2188). Conclusion: Regardless of surgical intervention utilized (supraclavicular exploration and nerve grafting or intercostal nerve transfers), patients had excellent outcomes for elbow flexion following obstetric brachial plexus palsy when utilizing Medical Research Council or Toronto scores for elbow flexion. The difference between these scores was not statistically significant. Type of study/Level of evidence: Therapeutic Study: Investigating the Result of Treatment/level III.


Introduction
With an estimated incidence between 0.4 and 4 per 1000 live births obstetric brachial plexus palsy (OBPP) is rare but can be a debilitating injury at birth.The extent of nerve involvement is different among patients and can be cate gorized into the following syndromes: upper trunk (C5 C6 ± C7) and complete (C5T1).While complete palsy presents with effects on shoulder, elbow, wrist, and hand movement, patients with upper trunk palsy most promi nently have a lack of active shoulder abduction and elbow flexion. 1,2While most OBPP cases may result in spontane ous recovery, there are a variety of primary and secondary surgical interventions that can be used to treat OBPP to improve function of the affected limb. 3Primary surgical interventions include neurolysis, nerve grafting and nerve transfers.
Traditionally, primary surgery to improve elbow flex ion in OBPP involves a supraclavicular incision, subse quent exploration of the brachial plexus to identify viable proximal roots and then attachment of these roots to the trunk, cord, or peripheral nerve through a nerve graft. 4,5hile proximal nerve transfers are performed near the site of injury in the supra or infraclavicular fossa, distal nerve transfers are those performed beyond the brachial plexus zone and near the neuromuscular junction.
Proximal nerve transfers usually include diagnosing the lesion via brachial plexus exploration and dissection, fol lowed by surgical intervention.These procedures have been well studied and produce excellent outcome.This operation presents with some disadvantages including lon ger recovery times and greater technical demands as com pared to nerve transfers. 6nlike proximal nerve transfers, distal nerve transfers are performed away from the site of injury and closer to the target muscle, do not require nerve grafts, result in shorter surgery times and, skills wise, are within grasp for a surgeon without high volume exposure to OBPP cases due to decreased technical demand and subjectivity of the procedure.Although there are clear benefits of distal nerve transfers, these are at the expense of full donor nerve func tion and postoperative complications such as respiratory failure. 6The distal procedure of intercostal nerve (ICN) transfer for reconstruction of the musculocutaneous nerve (MCN) has been used for adult brachial plexus injury.
More recent studies have shown it to be effective in obstet ric brachial palsy patients, with 70%-90% of patients achieving greater than or equal than Medical Research Council (MRC) M3 strength of biceps. 7,8lbow flexion is critical to a child's development and to many activities of daily living, from eating to buttoning a shirt.OBPP involving the C5C6 nerve roots and global palsy without recovery of C5C6 frequently results in reduced to absent elbow flexion. 9Thus, the reconstructive strategies currently employed in improving elbow flexion in OBPP patients are of considerable interest.Given that there has been no conclusive determination as to whether one procedure may be superior to the other, there is a sig nificant need to analyze the current data.
The aim of this study was to analyze all available lit erature and assess whether supraclavicular exploration with nerve grafting or ICN transfer is more effective in improving elbow flexion utilizing a variety of outcome measures.

Literature search
This study was conducted under Preferred Reporting Items for Systematic Reviews and MetaAnalysis for Individual Patient Data (PRISMAIPD) guidelines.First, a system atic search of the literature was conducted using PubMed, Cochrane, Web of Science, and the Cumulative Index to Nursing and Allied Health Literature (CINAHL) data bases.Specific search terms including "brachial plexus," "injury," "palsy," "nerve plexus," "upper plexus," "pediat ric," and "surgery" were used.The complete collection of Boolean searches is provided in Supplemental Appendix A. From the initial set of articles, duplicates were removed, followed by an abstract and fulltext screening.In these screenings to build the preliminary database, English text studies on brachial plexus surgery in pediatric patients were identified.The exclusion criteria for these studies were as follows: (1) studies that were not full text; (2) stud ies classified as commentaries, review papers, or editori als; (3) studies that were nonhuman or had less than three participants; (4) studies which had full texts that were inaccessible.The fulllength texts were accessed online.For completion, the references of all selected articles were crosschecked.If these articles were not previously included and fulfilled the inclusion/exclusion criteria, they were included in the database.
From this preliminary database, articles were then screened for relevance to this study's specific objective.Only studies which investigated patients who received ICN transfers or treatment with supraclavicular explora tion and nerve grafting (SENG) were included.Studies with ICN nerve transfers and SENG were then subdivided into groups based upon the type of outcome measures used to evaluate elbow flexion.These included the Medical Research Council (MRC) Scale for Muscle Strength (elbow flexors) and elbow flexion Toronto Active Move ment Scale (Table 1).

Data extraction
Data extracted from articles in the preliminary database included number of patients, patient characteristics, fol lowup, outcomes following surgery, and information about any secondary procedures.Data extraction was con ducted by all the authors.The authors of articles that did not have comprehensive data were contacted for additional information.If these authors did not respond or could not provide additional data, these studies were excluded from the present analysis.

Statistical methods
We employed an ordinal regression model to analyze the effect of using SENG or ICN on elbow flexion with the two scores measured: elbow flexion MRC and elbow flex ion Toronto Active Movement Scale.We choose ordinal regression because it is designed for order response cate gorical variable and allows for the independent variable (SENG/ICN method in our case) to be either continuous or discrete.We used a logit link function which is the most widely used in statistical application literature, and the model is formulated as: log(Pr(Score≤j)Pr(Score>j)) = β0 j;+β1Method+ɛ.

Results
The PRISMAIPD workflow used to identify studies is shown in Figure 1.Initially, 2936 texts were identified using the search terms.After a title screening, texts that were irrelevant to pediatric brachial plexus palsies, case reports, traumatic injuries, metaanalyses, editorials, sec ondary surgeries, reviews, nonmicrosurgical operations, and unavailable full texts were excluded, resulting in an initial index of 93 full texts.An additional 18 texts were identified by crossreferences.This produced a prelimi nary database of 111 texts.Afterwards, texts were screened for relevancy to SENG and ICN transfers.This produced 49 texts.Following this, the authors screened the published texts for results reported (elbow flexion MRC and elbow flexion Toronto scores) (Figure 2).In addition, the authors filtered out articles that reported grouped results or nonstandardized scores.For example, some authors chose to report scores of as MRC scores 4 and 5 as excellent, 2 and 3 as adequate, and 1 as unsuccessful.Attempts were made to reach out to these authors for a more detailed report of patient outcomes; however, this was unsuccessful in most instances.A final patient data base from six published articles consisted of 83 SENG patients (73 patients with MRC and 10 patients with Toronto score) and seven published articles which con sisted of 131 ICN patients (84 patients with MRC and 47 patients with Toronto scores).
][18] Data gathered from one text of patients undergoing SENG present an average Toronto score of 6.2 ± 2.2. 19Data gath ered from two texts of patients undergoing ICN transfers present an average Toronto score of 6.4 ± 1.2. 20,21The regression value between SENG and ICN transfers when comparing MRC scores is MRC scores is 0.3821 (standard error 0.4590, p = 0.2551).The regression value between SENG and ICN transfers when comparing Toronto scores is 0.7154 (standard error 0.8487, p = 0.2188).There is no statistical significance between SENG and ICN transfers when utilizing MRC or Toronto scores.

Discussion
Currently, there is little guidance for the use of different surgical procedures for patients undergoing repair for OBPP.Treatments for patients with upper nerve palsies emphasize the restoration of active shoulder abduction and elbow flexion.Traditionally, the use of SENG has pro duced favorable outcomes for patients.However, newer procedures that include the use of ICN transfers have not been extensively compared to proximal surgeries for elbow flexion in the pediatric population.We have found that both procedures result in excellent patient outcomes as a function of elbow flexion measured using the MRC score.
The use of SENG provides surgeons and patients with advantages for repair.This surgical procedure has been the standard for brachial plexus lesions in adults and children.Multiple studies have shown that this procedure has pro vided patients with a safe and effective treatment option for upper palsies.Furthermore, the surgical window allows for not only repair but also diagnosis and visualization of the lesion.This allows repair of multiple deformities and inju ries that may be present in patients during one operation.While these procedures have been well studied and pro duce excellent outcomes, this operation presents with some   disadvantages including the morbidity of longer dissections in an already traumatized area, longer operative times, and the necessity of subjective factors such as visual assess ment of lesions in continuity and relying on intraoperative nerve action potentials. 6,22In addition, SENG procedure is not appropriate in all patients, especially in more extensive injuries with poor quality of proximal donors accompanied by root avulsion.Distal nerve surgeries can provide patients with an improved postsurgical recovery experience as patients tend to have a faster recovery period.Furthermore, there is strong evidence that in the adult population, intra plexus dual nerve transfers provided better outcomes for shoulder and elbow functions in patients with traumatic upper plexus palsy compared to traditional nerve graft ing.However, this procedure does present with some technical and clinical difficulty.While distal transfers spare some operative morbidity due to their distance from the site of injury, this same distance can be a draw back, preventing diagnostic visualization of the affecting lesion. 4This coupled with potentially decreased donor nerve function and lower Mallet scores can call into question the value of ICN procedures compared to SENG, 6 taking into account that pros and cons of both procedures can only be debated for a group of patients where the both options exist, as opposed to patients with poor proximal donors where SENG cannot be performed.Furthermore, complications of ICN transfers have been explored and include pleural tears, wound infections, pleural effusions, acute respiratory distress syndrome, and wound seromas. 23Because of the lack of long term studies, it is unclear the total effect of ICN transfers in the long term such as chest deformity or influence on ventila tion at an older age.With the literature showing cases of deformity at 20 years post-operatively, there should be hesitancy to make ICN transfer a firstline treatment for pediatric populations. 17thers have expressed the concern of impaired respira tory function following ICN transfer with concomitant phrenic nerve transection.This concern may be unfounded as the literature has shown comparable pulmonary function in adults after ICN transfer.In a study of 42 adult patients undergoing phrenic nerve transfers and ICN trans fers, it was found that there was no significant difference in phrenic nerve transfers and multiple ICN transfers com pared to phrenic nerve transfers alone. 24In addition, other studies concluded that 10 adult patients undergoing simul taneous phrenic and ICN transfers did not produce clini cally evident respiratory dysfunction postoperatively. 25o our knowledge, there are no studies describing these complications in pediatric patients.
While conducting this study, many of the shortcomings regarding literature data presentation were highlighted.Some of these shortcomings can be attributed to a lack of studies containing children undergoing ICN transfers.In addition, a lack of standardization when compiling results contributed to the limited number of studies that can be compared.Data are often grouped differently between authors as some choose to report results as a function of time, procedure, location, or final outcome.This lack of standardization makes it difficult for surgeons to compare treatment options when choosing the operations that are best for their patients.Furthermore, very few authors pro vide individual data on patient outcomes.This can make it difficult to compare results and identify differences between selection criteria and functional results between different sets of literature.These issues are especially prevalent in the literature surrounding the treatment and assessment of brachial plexus injuries.Previous attempts have been made to find a consensus on how to report data using the international Plexus Outcome Study Group (iPLUTO) project. 26Some of the recommendations include the use of a data set and timing protocol when providing data in all scientific papers.There are several limitations to this study.Limited sample sizes and inability to personally and empirically assess each patient hindered our ability to provide concrete insights and recommendations.Furthermore, nerve graft ing to the anterior division of the upper trunk and ICN transfer to the MCN are not always utilized for different types and severities of injury.This prevented direct com parison between the two surgical options and can intro duce bias.The use of ICN transfers for obstetric brachial plexus palsies continues to be an area of unknown for cli nicians.Comprehensive and comparative studies for this treatment option are sparse.For physicians and surgeons to be fully equipped to help their patients, there needs to be a standardized, systemic approach for result presentation.

Conclusion
Both SENG and ICN transfers produce favorable out comes for patients suffering from OBPP.There are a pleth ora of functional outcomes used to evaluate shoulder, elbow, and wrist function following surgical intervention.Elbow flexion is of particular interest as it is crucial in motor skill development of an infant.In this study, SENG and ICN transfers produced equivalent outcomes when utilizing MRC elbow flexion scores or Toronto active movement scale scores for elbow flexion.
late presenting patients, median nerve utilization for biceps reinnervation is an excellent option with nearly all patients obtaining full motion at the final post-operative period El-Gammal et al. 20 Age at surgery = 41 months mean (range = 18 months-4 years) for all 19 cases 5 ICN transfers versus 6 Oberlin transfers 11 total elbow flexion cases Assessment of motor recovery using Toronto Active Movement Scale: poor < 4, fair = 5, good = 6, and full = 7 Poor and fair = unsatisfactory Good and full = satisfactory Active elbow flexion-increased from 2.7 degrees preop → 91.8 degrees post-op = average Average grade of 0.

Table 1 .
ICN and SENG studies.thedifference between ICN versus MPN to MCN in order to reinnervate bicep muscle.Although the results in the MPN-MCN group were slightly better than in the ICN-MCN group, possibly due to more axons, larger size, and exclusive motor function, there was no statistically significant difference between the two groups.Suggestion is that MPN-MCN is the first choice for treatment