|Home | About | Journals | Submit | Contact Us | Français|
While the majority of patients with brachial plexus birth palsy (BPBP) demonstrate spontaneous recovery, some will have persistent neurological deficits and functional limitations. Microsurgical repair and/or reconstruction of the brachial plexus are often recommended for these patients; however, the optimal timing of microsurgery is unknown. Furthermore, the long-term outcomes of microsurgery, as compared with natural history and secondary reconstructive surgery, are unknown. The purpose of this article is to present the rationale for the multicenter prospective study of the treatment of BPBP. The current study protocol is provided.
First described by Smellie and further characterized by Erb, Duchenne, and Klumpke, brachial plexus birth palsy (BPBP) may vary in severity and degree of involvement, ranging from transient neuropraxia to complete cervical nerve root avulsion of part or all of the brachial plexus.1,2,3,4,5,6 Risk factors include large gestational size, prolonged labor, difficult and/or instrumented delivery, and breech presentation.7,8 Despite advances in obstetrics, the incidence of BPBP has remained between 0.1 and 0.4% of live births.8,9,10
The natural history of BPBP remains largely unknown because of the paucity of studies of affected patients followed from birth to maturity. It has been estimated that 80 to 90% of patients recover spontaneously within the first 2 months of life, with subsequent normal upper extremity function.8,9,10,11,12,13,14 Patients with persistent neurological deficits after 3 to 6 months of age, however, are at high risk for permanent neurological dysfunction.6,11,12,14,15
Currently, the prognosis for long-term upper extremity function is dependent upon the timing and rate of neurological recovery. Biceps function has been utilized by many, and the return of active elbow flexion by the age of 3 to 6 months has been shown to prognosticate long-term brachial plexus recovery.11,14,16,17 In addition, Michelow et al reported that using recovery of elbow flexion in combination with elbow, wrist, and digital extension increases the accuracy of predicting long-term prognosis.12 Laurent et al have similarly utilized deltoid, triceps, and biceps function to predict neurological recovery.18 Microsurgical reconstruction of the brachial plexus is generally considered for infants with persistent neurological deficits and poor long-term prognoses, as indicated by these parameters.
There is significant controversy regarding the appropriate timing for microsurgery in infants with BPBP and upper extremity weakness. Current recommendations for the timing of microsurgical repair range between the ages of 3 to 9 months,10,11,12,14,15,16,19 and the range of repairs cited within the literature extends from 1 to 24 months of age.11,15,18,19,20,21,22,23,24,25,26,27
A study performed at our institution highlights the current clinical dilemma.14 Thirty-nine of 66 infants with BPBP demonstrated recovery of biceps function between the third and sixth months of life. Although none of these 39 infants ultimately developed “normal” upper extremity function, it is unknown whether these infants would have had better—or worse—outcomes with microsurgical repair at 3 months of age. Furthermore, it is unknown whether patients who undergo microsurgery at 3 months of age fare better than those who do not undergo early microsurgery but do have subsequent or later secondary shoulder procedures in the form of soft tissue releases, tendon transfers, or humeral osteotomies. Indeed, nine patients with biceps recovery during the fourth, fifth, or sixth month of life treated with latissimus dorsi and teres major tendon transfers to the rotator cuff without prior microsurgical intervention had global shoulder function nearly equivalent to that of patients with spontaneous biceps recovery at 2 to 3 months of age.14 Clearly, earlier surgery would have resulted in a major difference in the number of surgical procedures performed, each with its concomitant risks and all with unknown influence on the ultimate functional outcome. Unfortunately, no prospective study has been performed analyzing the outcomes of microsurgery as a function of age.
Compounding this controversy is the use of numerous methods and scoring schemes for predicting and assessing the outcome of BPBP. To date, no standard evaluation system has been established, nor has there been a consensus on what constitutes a good or acceptable functional outcome.
Finally, because affected infants have varying patterns of brachial plexus injury, patients with BPBP have different rates of predicted neurological recovery and ultimate function. For example, infants with isolated upper plexus injuries (e.g., C5-C6, C5-C7) have better prognoses than patients with total plexus involvement and an associated Horner's syndrome.14,22 Therefore, the pattern of injury must be accounted for when considering the timing for microsurgical intervention.
The limited scientific information regarding the best treatment and timing of treatment for infants with BPBP has led to an emotionally charged environment of opinions among treating hand and upper extremity surgeons. This has made it difficult for parents and physicians to make the best choices in the care of these children.
At the core of this problem lies the fact that most of the current understanding of the natural history of BPBP is based upon retrospective case series. These studies have provided an abundance of useful clinical information, particularly given the relatively low incidence of BPBP. However, case series are often anecdotal, inherently susceptible to selection and treatment biases, subject to limited or incomplete follow-up of patients, and often difficult to compare with other series.
Based upon the aforesaid factors, it is apparent that the experience of a single surgeon during a single professional career will never be able to determine the optimal age for microsurgical reconstruction in infants with BPBP and upper extremity weakness. There are simply too few patients with too many different patterns of injury undergoing too many different types of reconstruction to allow a single surgeon to provide a scientific answer to this fundamental question. As a result, a prospective, multicenter study is needed.
Such a study would have to fulfill several requirements, however, to achieve its desired goal. First, there would have to be an adequate number of patients enrolled, determined by a prestudy power analysis using fundamental assumptions about the incidence of BPBP and number of patients ultimately requiring microsurgery. Second, valid and reliable means of evaluating upper extremity function in BPBP must be established and universally implemented. Third, there must be established indications for microsurgical intervention, followed by all participating centers. Fourth, clinical follow-up at standardized intervals using established outcomes instruments is mandatory, ideally to skeletal maturity. Fifth, the study design should be one that is complicit with the principle of clinical equipoise; treatment of all patients enrolled in the study should adhere to current standards of care regarding BPBP. And finally, dedicated and skilled surgeons must agree to be participants as well as critical evaluators of the study results. With the support of the Pediatric Society of North America Clinical Trials Network and the American Society for Surgery of the Hand Outcomes Studies Grant, we have embarked on just such a study.
The current study design is a multicenter prospective comparative cohort study, designed to answer the following clinical question: Is there a significant difference in the long-term outcome of infants with BPBP and upper extremity weakness between those who undergo microsurgical reconstruction at 3 months of age and those who undergo microsurgery at 6 months of age?
Prior to entertaining the details of the study protocol, a few comments regarding the study design must be made. Currently, it is accepted that the randomized clinical trial (RCT) with concurrent controls is the “gold standard” of evidence-based medicine. Randomization of patients with prospective data collection allows the highest degree of internal validity (i.e., provides the most valid conclusions) and minimizes the effects of bias and confounding variables. However, for the purposes of this study, an RCT design was not chosen for several reasons. First, as there is clear evidence that infants with absence of biceps function by the age of 3 to 6 months benefit from microsurgical reconstruction, it was not deemed ethical to randomly assign a group of patients to serve as concurrent controls. Although information on the natural history of BPBP might be obtained, patients in such a control group would not be eligible for microsurgery when deemed otherwise appropriate, thus violating the principle of clinical equipoise. Second, the process of individual patient randomization to microsurgery at either 3 or 6 months of age is problematic. Individual surgeons and centers have historically developed their own notions of the appropriate timing for microsurgery, and these guidelines can be supported by the existing literature.10,11,12,14,15,16,18,19,20,21,24,25,26,27,28 Rather than require these individual surgeons or centers to violate their own notions of standards of care, the decision was made to construct a multicenter prospective study comparing two cohorts of patients—those undergoing surgery at 3 months and 6 months of age. The randomization protocol is now center randomized with each institution following its current standard of care. Practically, this study design allows greater enrollment of eligible surgeons and centers.
All infants and children diagnosed with BPBP at participating centers are eligible for enrollment in this study. Participating physicians and centers are recruited, organized, and coordinated in conjunction with the American Society for Surgery of the Hand and the Pediatric Orthopaedic Society of North America (POSNA) Clinical Trials Network. The diagnosis of BPBP is made on the basis of clinical history, physical examination, and routine radiographic studies by participating surgeons. Patients with coexisting congenital upper extremity abnormalities, neuromuscular pathology, history of previous surgery on either upper extremity, or inability to undergo therapy, surgery, or outpatient follow-up are excluded from participation.
All patients with BPBP presenting prior to the age of 3 months are evaluated by a participating pediatric hand and/or upper extremity surgeon (Fig. 1A) A detailed clinical history and physical examination are performed with each patient's visit. Routine radiographs are performed only for evaluation for possible fracture or dislocation, obviating the need for unnecessary diagnostic testing requiring sedation or radiation exposure. Patients' functional status is classified according to the Toronto Test score12 (Table 1), Hospital for Sick Children Active Movement Scale19 (AMS) (Table 2), modified Mallet classification28,29 (Fig. 2), and the musculoskeletal outcomes data evaluation and management system (MODEMS) Pediatric Outcomes Instrument.30,31 The validity and reliability of these outcomes instruments have been previously determined.30,32 The patients with spontaneous recovery of their BPBP prior to the age of 3 months (group A) will continue to be observed until a minimum of 2 years of age and ideally to 5 years of age. This group will serve as the first natural history subgroup.
At the beginning of the study, participating surgeons and institutions will be self-designated as 3- or 6-month centers, corresponding to the age at which microsurgical repair will be performed. At the age of 3 months, the patients at 3-month centers with BPBP and absent biceps function or Toronto Test score less than 3.5 will undergo microsurgical repair (group B) (Fig. 1A). These patients will serve as the 3-month microsurgery subgroup. The patients with absent biceps function or Toronto Test scores less than 3.5 at 6-month centers will continue to be observed (group C). Microsurgery will be performed using standard surgical techniques.24 Postoperative clinical follow-up will be performed in all patients from the age of 3 months to a minimum of 2 years, and functional status according to the Toronto Test score, AMS, and Mallet classification schemes will be documented during these regular outpatient visits.
At the age of 6 months, patients will be reassessed (Fig. 1B). Patients from group C with persistent absence of biceps function or Toronto Test scores less than 3.5 will undergo microsurgical repair at 6 months of age (group D). These patients will serve as the 6-month microsurgery subgroup. Patients who have regained biceps function or have Toronto Test scores greater than 3.5 will continue to be observed (group E) and will serve as the second natural history subgroup. Patients in all study groups will continue to be evaluated at regular intervals until a minimum age of 2 years and again, ideally, to 5 years of age.
At 1, 1.5, and 2 years of age, all patients from groups A through E will be evaluated for upper extremity function using the Toronto Test score, AMS, Mallet classification, and MODEMS scoring schemes (Fig. 1C). Plain radiographs, computed tomography (CT) scans, or magnetic resonance imaging (MRI) of the glenohumeral joint will be performed in patients with shoulder dysfunction, as evidenced by internal rotation contracture, external rotation weakness, or abduction weakness, in preparation for possible secondary surgical reconstruction. In the absence of underlying glenohumeral joint deformity, patients with functional limitations will undergo tendon transfer of the latissimus dorsi and teres major muscles to the rotator cuff as previously described (subgroup I).6,33,34,35,36 Patients with glenohumeral deformity in the setting of shoulder dysfunction will undergo tendon transfers with potential glenohumeral joint reduction (subgroup 2).6 Patients with an irreducible glenohumeral joint and severe deformity will undergo humeral derotational osteotomy.37,38,39 Patients without shoulder dysfunction will not undergo additional radiographic or surgical procedures and will continue to be observed (subgroup 3). All patients undergoing reconstructive procedures for shoulder dysfunction and/or glenohumeral joint deformity will be observed for a minimum of 2 years postoperatively.
Final upper extremity function will be graded according to the Toronto Test score, AMS, and Mallet classification at greater than 5 years of age. In addition, the overall health and functional status will be assessed using the pediatric MODEMS instrument at the conclusion of the study.
Based on a total of 650 infants diagnosed with BPBP at birth, an expected 20% will have persistent biceps weakness or composite Toronto Test scores of less than 3.5 points at the age of 3 months.14,40 Of these 130 infants, 65 patients will undergo microsurgical repair at the age of 3 months (group B), based upon equal allotment of patients at 3- and 6-month centers. The other 65 infants at 6-month centers (group C) will be observed until the age of 6 months. We expect that about 30 to 35% of these patients will have persistent biceps deficits and Toronto Test scores of less than 3.5 at 6 months of age; these 20 patients will then undergo microsurgical repair (group D).14,40 A power analysis indicated that sample sizes of 65 and 20 patients in groups B and D, respectively, would provide 82% power for detecting a mean difference of 3 points in the composite Mallet score (range 5–25) at 2 years of age between the groups using a two-group Student t-test based on unequal n's and a two-tailed significance level of 0.05.41 Here, we assume a common standard deviation of 4 points and an effect size of 0.75. In addition, these sample sizes will provide 87% power to detect a mean difference of 2 points between the groups in Toronto score, assuming a common standard deviation of 2.5 points (effect size=0.8) based on a two-group Student t-test and a two-tailed α level of 0.05. Power analysis was performed using version 4.0 of the nQuery Advisor software package (Statistical Solutions, Boston, MA).
It should be emphasized that not all newborns with BPBP are seen by hand and upper extremity surgeons for evaluation and treatment. Based on natural history studies, about 20% of all patients with BPBP at birth have persistent deficits at the age of 3 months.8,9,10,11,13,14 As a result, only 130 patients with absent biceps function or Toronto Test scores less than 3.5 at the age of 3 months would need to be enrolled to obtain the appropriate statistical power to achieve the study's primary objective. Assuming a 15% rate of patient dropout or loss of follow-up, a total of 150 patients will be required to achieve the study's primary objective.
Although the primary aim of this prospective multicenter investigation is to determine the optimal age, if any, for microsurgical reconstruction in infants with BPBP and persistent upper extremity weakness, several secondary objectives may also be achieved. By study design, clinical and radiographic information on all BPBP patients from participating centers may be entered into this standardized, prospective database. In addition to the infants younger than 3 to 6 months of age who undergo microsurgical reconstruction of the brachial plexus, this database would include patients who demonstrate spontaneous neurological recovery early in life and patients who present later in life with chronic BPBP. With continued observation and treatment according to the current standards of care, all of these patients may be observed until skeletal maturity. Recent advances in Internet-based information technologies will facilitate the collection and analysis of these data. Further information regarding the natural history of BPBP and the functional outcomes of secondary reconstructive procedures may thus be obtained.
In this fashion, several secondary study questions may be addressed. First, the functional outcomes between patients undergoing early microsurgery versus late secondary reconstruction may be compared. The results of early microsurgery combined with secondary shoulder reconstruction may also be compared against secondary shoulder reconstruction alone. Furthermore, the long-term functional results of patients undergoing surgical treatment may be compared with the natural history of spontaneous recovery.
In addition, more recent published reports have established that progressive glenohumeral joint deformity and posterior shoulder instability develop in the setting of long-standing BPBP with internal rotation contracture and external rotation weakness.34,36,42,43,44 Patients enrolled in this multicenter, prospective study will undergo CT and MRI imaging of the shoulder to define further the natural history of glenohumeral joint development in chronic BPBP and the remodeling effects, if any, of extra-articular and intra-articular reconstructive procedures.
Again, if this information is obtained from multiple centers in a prospective fashion, adequate statistical power may be achieved to enhance greatly our understanding of BPBP and our ability to provide the best functional results from surgical treatment. This would ultimately allow a scientific, evidence-based approach to the evaluation and treatment of patients with BPBP. Given the emotionally charged atmosphere that may often surround families and care providers of these patients, this study would provide critical information not only to surgeons, but also to patients, families, and primary care physicians. The impact will thus be made in all aspects of the care of patients with BPBP.
At the time of this article's writing, a total of 114 patients have been enrolled in this ongoing multicenter prospective study. There are nine centers presently involved, from four countries and from five different states within the United States. An additional 11 centers are currently seeking Institutional Review Board (IRB) approval to join this prospective multicenter effort. Eighty study patients are younger than 6 months of age. Clearly, there are too few data at this stage for definitive conclusions. Of note, data safety monitoring will occur to detect any statistical differences in the groups.
For further information regarding participation in this multicenter prospective study, please contact Dr. Peter M. Waters, Department of Orthopaedic Surgery, Children's Hospital Boston, 300 Longwood Avenue, Hunnewell 2, Boston, MA 02115.
This investigation is supported by the Pediatric Orthopaedic Society of North America Clinical Trials Network and the American Society for Surgery of the Hand Outcomes Studies Grant. The authors wish to acknowledge and thank Ms. Laurie Travers and Ms. Annie Kuo for their administrative assistance with this ongoing investigation.