Neonatal brachial plexus palsy

INTRODUCTION — The brachial plexus (figure 1) is a network of nerve fusions and divisions that originate from cervical and upper thoracic nerve roots and terminate as named nerves that innervate muscles and skin of the shoulder and arm. The first clinical description of neonatal brachial plexus palsy (NBPP) was reported in the 1760s [1]. In the late 1800s, the different types of NBPP were defined; Duchenne and Erb in separate reports described upper trunk nerve injury to the C5 and C6 nerve roots, now called Erb palsy or Duchenne-Erb palsy [2,3], and Klumpke described lower trunk injury involving the C8 and T1 nerve roots [4]. Subsequently, palsy of all the nerve roots from C5 to T1 was reported (picture 1).

The major controversies about NBPP revolve around the etiologies and the management. Once the diagnosis is established, what studies should be performed? How often does shoulder dystocia or obstetrical intervention contribute to the etiology? What is the role of rehabilitation therapies? When should nerve surgery be considered? When is orthopedic intervention needed? What is the prognosis for complete or partial recovery? Unfortunately, the literature provides a variety of different answers to the questions posed and controversy remains. In addition, the quality of the published evidence about NBPP is suboptimal; high-quality population-based studies with prospective analysis, sufficient follow-up, and clear assessment are scarce [5].

This topic will review the etiology, epidemiology, clinical features, diagnosis, management, and prognosis of NBPP, which is also known as obstetric brachial plexus palsy and birth-related brachial plexus palsy. Other brachial plexopathies are discussed separately. (See "Brachial plexus syndromes".)

ETIOLOGY — Potential mechanisms of neonatal brachial plexus palsy (NBPP) include stretching/traction, compression, infiltration, and oxygen deprivation [6]. Of these, stretching is considered the most common mechanism. In addition, stretching may contribute to brachial plexus injury in cases where the principal cause is one of the other mechanisms. The etiology of NBPP traditionally has been attributed to iatrogenic lateral traction on the fetal head, typically when shoulder dystocia impedes delivery [7-9]. Downward lateral traction (ie, bending the neck away from the anterior shoulder and toward the posterior shoulder) causes increased stretching of the brachial plexus compared with downward axial traction (ie, applying force parallel to the fetal spine) [10].

However, NBPP can occur even when axial traction is properly applied; the occurrence of NBPP following birth does not automatically indicate that the practitioner applied forces or maneuvers that caused the nerve injury [10]. The forces of uterine contraction and maternal pushing alone are probably sufficient to cause excessive traction on the brachial plexus [10,11]. In addition, antepartum factors may predispose to NBPP, including uterine abnormalities such as Müllerian anomalies and fibroids that can result in fetal malpositioning and compression [12].

Furthermore, many cases of brachial plexus injury occur independently of shoulder dystocia or excessive force by the provider:

●A substantial proportion of NBPP cases are not associated with antecedent shoulder dystocia [13]. (See 'Epidemiology' below.)

●In one report, a transient NBPP occurred in the posterior arm after a spontaneous vaginal delivery without shoulder dystocia or traction to the fetal head [14].

●NBPP has been observed after delivery by cesarean section [11,15,16].

●A prenatal insult has been documented in a number of affected infants [17,18].

Injuries to peripheral nerves can be defined as neurapraxic, axonotmetic, and neurotmetic, based upon the severity and extent of injury to the structural components of the peripheral nerve, including Schwann cells, axons, and surrounding connective tissue. (See "Traumatic peripheral neuropathies", section on 'Classification and pathophysiology'.)

●Neurapraxia is usually caused by a mild injury (eg, ischemia, mechanical compression, metabolic or toxic factors) that results in focal demyelination, but no loss of axonal integrity in the region of injury; this type of injury typically has a favorable prognosis for recovery.

●Axonotmesis typically occurs as a result of crush injuries, nerve stretch injuries or percussion injuries; the axon and myelin sheath are locally but irreversibly damaged. However, the surrounding stroma, including the endoneurium and perineurium, remains intact; the recovery potential is intermediate.

●Neurotmesis most often occurs in association with severe lesions. The axon, myelin sheath, and surrounding stroma are all irreversibly damaged. The external continuity of the injured nerve is usually disrupted (eg, as in nerve root avulsion). No significant regeneration occurs with such a lesion unless surgical repair is performed.

While potentially useful for prognosis, there is no gold standard for differentiating these types of lesions in NBPP [19]. The most reliable method is probably surgical exploration, which is generally undertaken only for select patients who have poor spontaneous recovery by three to nine months of age and therefore may benefit from nerve repair (see 'Surgery for severe cases' below). Electrodiagnostic studies (see 'Electrodiagnostic studies' below) and neuroimaging with magnetic resonance imaging or computed tomographic myelography (see 'Neuroimaging' below) can identify some cases of nerve root avulsion (neurotmesis) but have not been considered reliable enough for prognostic purposes in the first months of life (see 'Prognosis' below). Thus, prognosis in most cases of NBPP is estimated, based upon clinical parameters.

EPIDEMIOLOGY — Neonatal brachial plexus palsy (NBPP) is uncommon, with an incidence that ranges from 0.04 to 0.3 percent of live births [9,20-26]. In both a 2014 systematic review of 40 retrospective studies from the American College of Obstetricians and Gynecologists (ACOG) [27] and a study that analyzed a database of pediatric discharges from the United States [9], the cumulative incidence of NBPP was 0.15 percent.

The 2014 ACOG systematic review made the following additional observations:

●Among deliveries complicated by shoulder dystocia, the rate of transient NBPP ranged from 1 to 17 percent, while the rate of persistent NBPP at one year or more after birth ranged from 0.5 to 1.6 percent [28]

●Among deliveries without documented shoulder dystocia, the incidence of NBPP was approximately 0.9 percent [27]

●With cesarean delivery, the incidence of NBPP ranged from 0.03 to 0.15 percent [27]

Risk factors — The only established risk factor for NBPP is shoulder dystocia [10]. (See "Shoulder dystocia: Intrapartum diagnosis, management, and outcome".)

Other possible risk factors include substantial maternal weight gain, maternal diabetes, multiparity, fetal macrosomia/high birth weight, fetal malposition, labor induction, labor abnormalities, operative vaginal delivery, and previous pregnancy complicated by shoulder dystocia or NBPP [9,10,21,29-34]. However, none of these factors have demonstrated a consistent, statistically significant predictive value for the occurrence of NBPP [10,12,35-37].

Of note, probably the majority of NBPP cases occur in vaginal deliveries without shoulder dystocia or other putative risk factors [10,12]. Cesarean delivery and twin or multiple birth mates have been associated with a decreased risk of NBPP [9].

Prevention — There are no proven measures to predict or prevent NBPP [10,12]. The occurrence of shoulder dystocia itself cannot be accurately predicted based upon antenatal risk factors or labor abnormalities. Therefore, clinicians should be prepared for possible shoulder dystocia in all vaginal deliveries and be cognizant of the various procedures that have been shown to be effective for delivering the impacted shoulders. (See "Shoulder dystocia: Intrapartum diagnosis, management, and outcome", section on 'Management'.)

Despite the lack of proven interventions for preventing NBPP, there are several clinical situations where the American College of Obstetricians and Gynecologists suggests that practitioners consider an alteration of usual obstetric management [10,36]:

●Suspected fetal macrosomia with an estimated fetal weight >5000 g in pregnant patients without diabetes or >4500 g in others with diabetes

●Prior recognized shoulder dystocia, especially when associated with a severe neonatal injury

●Midpelvic operative vaginal delivery with a fetal birth weight >4000 g

CLINICAL FEATURES — Most cases of neonatal brachial plexus palsy (NBPP) are unilateral; bilateral involvement occurs in approximately 5 percent of cases [38]. Five different patterns of nerve involvement have been described [21]:

●C5 and C6 injury (Erb palsy) accounts for approximately 50 percent of cases. Weakness involves the deltoid and infraspinatus muscles (mainly C5) and biceps (mainly C6). As a result, the upper arm is adducted and internally rotated, and the forearm is extended, while hand and wrist movement are preserved.

●C5, C6, and C7 injury (Erb palsy plus) accounts for approximately 35 percent of cases and manifests as adduction and internal rotation of the arm, extension and pronation of the forearm, and flexion of the wrists and fingers, sometimes referred to as the "waiter's tip" posture (picture 1).

●C5 to T1 injury usually presents with arm paralysis and some sparing of finger flexion.

●Severe damage to all C5 to T1 roots is characterized by a flail arm and Horner syndrome.

●C8 and T1 injury (true Klumpke palsy) is the most infrequent pattern and manifests as isolated hand paralysis and Horner syndrome.

In practice, NBPP can be divided into the common upper brachial plexus palsy involving injury to C5, C6, and occasionally C7 (Erb palsy) and the less common total brachial plexus palsy involving all the roots from C5 to T1, which is often erroneously called Klumpke palsy. Some have questioned whether true Klumpke palsy (ie, C8 and T1 root injury with weakness of the distal arm only) occurs at all in the newborn period [38].

Lesions associated with NBPP include clavicle and humerus fractures, shoulder subluxation, cervical spine subluxation, cervical spinal cord injury, and facial palsy [38]. These occur in ≤10 percent of cases. Involvement of the phrenic nerve (arising from C3, C4, and C5) with diaphragmatic paralysis accompanies a minority of cases of upper plexus (Erb) palsy [39].

There is a higher incidence of torticollis [40] and early speech delay [41] in infants with NBPP and so multidisciplinary care is important. (See 'Management' below.)

EVALUATION AND DIAGNOSIS — The diagnosis of neonatal brachial plexus palsy (NBPP) is made by the finding of arm weakness at birth with a distribution consistent with a brachial plexus injury [6]. In many cases, the diagnosis is straightforward. However, determining the pattern of weakness in neonates may be difficult, in part because normal neonates tend to have relatively limited arm movements right after delivery. In addition, the distribution of weakness with some types of NBPP is similar to nonbrachial plexus lesions.

Adjuncts that may be useful for therapeutic decisions about the need for nerve repair or reconstruction include electrodiagnostic studies and neuroimaging. Because of the potential impact on prognosis and medicolegal issues, practitioners should try to determine the time when the brachial plexus injury occurred. (See 'Timing of the palsy' below.)

Clinical assessment — The initial diagnostic work-up of NBPP starts with a complete family, maternal, and perinatal history [42]. The infant should be evaluated by clinical examination and radiographic studies for fractures or any other injury [10,43]. The neurologic examination (see "Neurologic examination of the newborn") should include observation of spontaneous movements, passive and active range of motion, stimulated motor and sensory responses, and assessment of reflexes to look for signs of focal or global neurologic deficits.

●Certain postures (eg, "waiter's tip") are associated with particular NBPP injuries (see 'Clinical features' above)

●An asymmetric Moro reflex should arouse suspicion for NBPP

●Ptosis and miosis (Horner syndrome) point to lower trunk involvement

●Asymmetric chest cavity expansion and impaired oxygenation or feeding suggest NBPP associated with diaphragmatic paralysis secondary to phrenic nerve injury [39] (see "Diaphragmatic paralysis in the newborn")

●The early presence of restricted passive range of motion suggests that the peripheral nerve injury occurred in utero or that another musculoskeletal condition is responsible, since contractures and joint subluxations will develop several months after nerve injury [44,45]

●Hemiparesis or global neurologic deficits raise suspicion for a central nervous system etiology [46]

Timing of the palsy — In order to estimate the time of onset of brachial plexus injury, practitioners should record the events leading up to delivery, fetal positioning, site of the brachial plexus injury, anatomic relationship of the affected brachial plexus to the anterior shoulder at delivery, neonatal bruising, associated injuries (eg, bone fractures), presence and location of the caput succedaneum, Apgar scores, and cord blood gas results [42]. (See "Neonatal birth injuries".)

Brachial plexus injuries that occurred well before delivery may have specific associated findings involving the affected arm, such as atrophy of the hand and arm muscles and the pectoralis major muscle, joint contractures, and bone demineralization [42].

Electrodiagnostic studies — Electrodiagnostic studies (electromyography and nerve conduction studies) can help to determine the localization and severity of nerve injury associated with NBPP [47]. Nevertheless, there is no clearly established clinical role for electrodiagnostic studies in determining the etiology, timing of injury, or prognosis of NBPP [42], although controversy persists about these issues [21,48-50]. As examples, some experts argue that evidence of chronic denervation by electromyography within one week of birth would be most consistent with a brachial plexus lesion that occurred in utero, whereas serial electromyography studies showing no fibrillation potentials within 24 hours after birth followed by their appearance at 48 hours after birth would suggest that the brachial plexus lesion occurred during labor and delivery [49]. However, there are few data to support or refute these conclusions. The potential utility of electromyography for estimating the prognosis of NBPP is discussed below. (See 'Prognosis' below.)

Neuroimaging — Imaging of the cervical spinal cord can detect evidence of proximal nerve root avulsion, ie, a neurotmetic lesion with poor prognosis if untreated (see 'Etiology' above), as suggested by the finding of a pseudomeningocele (traumatic meningocele) or a post-traumatic neuroma (a disorganized proliferation of regenerating axons, Schwann cells, and perineural cells).

Although there is some variation, most studies suggest that both magnetic resonance imaging (MRI) and computed tomographic (CT) myelography have a low to moderate sensitivity and a high specificity for the detection of pseudomeningocele [51-53]. In addition, one report found that MRI had a high sensitivity and specificity for the detection of post-traumatic neuroma but was not helpful for precise anatomic localization to the level of the brachial plexus trunk or division level [51]. As CT requires a lumbar puncture to instill contrast and exposes the infant to ionizing radiation, MRI imaging is increasingly preferred over CT for the evaluation of NBPP [53].

Ultrasound can visualize the C5 to C8 nerve roots and brachial plexus in neonates [54], and has been used to evaluate NBPP in small numbers of infants [54-58]. As an example, a case report observed that ultrasound of the brachial plexus in the supraclavicular fossa performed at five months of age detected a small neuroma involving the upper trunk as well as atrophy of the supraspinatus and infraspinatus muscles and laxity of the posterior glenohumeral joint [56]. Another report that employed ultrasound to evaluate infants (average age 14 months) with NBPP suggested that muscle backscatter (a measure of echo intensity) could identify muscle injury and that muscle thickness could differentiate moderate from severe impairment [55]. However, more study is needed to determine whether ultrasound is useful for the evaluation of NBPP.

MANAGEMENT — Infants and children with NBPP should be referred for physical and occupational therapy as soon as the diagnosis is suspected. Surgical intervention is advocated in select cases if functional recovery does not ensue, but there is no consensus regarding the utility or timing of surgery.

Physical and occupational therapy — The initial management of NBPP involves measures designed to prevent contractures, including passive range-of-motion exercises at all relevant joints beginning in the latter half of the first week after birth, supportive splints as needed to prevent finger flexion or elbow contractures, and promotion of muscle strengthening and normal function [38,47].

Serial casting for an elbow flexion contracture has been proposed as a treatment of NBPP [59]. In a prospective study, 41 patients with NBPP and elbow flexion contractures of 30 degrees or more received serial casting until the contracture was 10 degrees or less, for a maximum of eight weeks [60]. Among 20 patients, there were 37 recurrences, and 4 patients showed a loss of flexion mobility. Patient satisfaction with the procedure was only moderate. More research is needed into this form of therapy before it can be recommended.

Surgery for severe cases — Surgical intervention for NBPP is advocated in select cases with severe nerve injury or if functional recovery does not ensue in three to nine months, but there is no consensus regarding the utility or timing of surgery [61-69]. Secondary surgeries such as tendon transfer or contracture release may be evaluated beginning at 18 to 24 months. Early referral to a center with expertise in the management of NBPP may improve outcomes [10].

The absence of functional recovery, based upon serial clinical examination over several months, is the mainstay for selecting patients who might benefit from surgical intervention, but the time required to make this assessment eliminates the possibility of better outcomes from very early surgery. Magnetic resonance imaging (MRI) can aid in determining the severity of nerve injury earlier than clinical examination. In a prospective study, nine infants with NBPP underwent three-dimensional volumetric proton density MRI without sedation to evaluate the cervical spine and brachial plexus at three months of age [70]. Four of the nine children had surgery based solely on serial clinical examination. A radiologic score based upon MRI findings for the number and severity of injured spinal nerve roots at three months of age distinguished patients who had surgery from those who did not. The utility of this approach requires confirmation in larger studies.

Of note, the treatment programs for NBPP are largely institution specific and are based upon mainly small studies or retrospective or anecdotal data. The problem is exacerbated by a lack of multicenter prospective studies and randomized trials. In various reviews and reports, indications for surgical nerve repair or reconstruction include the following:

●Panplexopathy and preganglionic nerve root lesions [10,66]

●Neurotmetic lesions or nerve root avulsions [47]

●Incomplete functional recovery:

•Lack of antigravity strength of elbow flexion by three months of age [71-73]

•Absent or severely impaired hand function at three months in an infant with flail arm at birth [74]

•Inability of infant to bring a cookie to the mouth with the affected limb at nine months (the cookie test) [75,76]

•Inability of the infant to use the affected limb to remove a towel covering the face at six months (the towel test) [77]

•Reasonably good hand function but a persistent deficit of active wrist extension, weak shoulder elevation, and absent shoulder external rotation at six months [66]

The outcomes of studies reporting surgical repair for NBPP are not directly comparable due to variability of the anatomic lesions and differences in surgical techniques and outcome assessments [68]. With these caveats in mind, reports of "good" outcomes after surgery have ranged from approximately 25 to 80 percent [78-81]. As a general rule, C8-T1 lesions reduce the likelihood of surgery to restore useful hand function [47]. Nerve transfers from donor intercostal nerves or ulnar motor fascicles to the musculocutaneous nerve can restore elbow flexion [68,82-85]. Secondary soft-tissue shoulder reconstructive surgery to correct internal rotation contracture may further improve function [65].

Additional treatment options

Botulinum toxin injections — Botulinum toxin injections have been used to treat contractures and muscle imbalance associated with NBPP, but data are limited and retrospective [86-88]. The clinical utility of this intervention needs further study before it can be recommended.

Counseling and mental health support — Children with NBPP should be assessed for behavioral and mental health issues. Some children with NBPP have associated psychosocial problems in addition to functional limitations, both of which can interfere with quality of life [89-91]. In a study that used focus-group interviews of 48 children (ages 8 to 18 years) with NBPP and their families, limitations in the areas of activity and participation (eg, problems with sports) were most common [89]. Older children reported fewer problems related to their disability, perhaps because of improved coping strategies. In another report that compared 42 children and families with NBPP with 43 healthy controls and families, children with NBPP had higher problem scores on multiple parameters including withdrawal, anxiety/depression, social problems, thought problems, delinquent behavior, and aggressive behavior; mothers of children with NBPP had increased maternal distress [90].

Awareness of these potential issues can facilitate screening and referral to support groups and mental health specialists when indicated. (See "Children and youth with special health care needs".)

PROGNOSIS — The natural history of neonatal brachial plexus palsy (NBPP) is not precisely defined. Spontaneous recovery occurs over one to three months in many if not most cases of NBPP, but persistent functional impairment is seen in 18 to 50 percent of patients [24,68,92]. Some children with persistent symptoms at three months may recover by one year [25]. This wide variation among reports is likely due to a number of issues confounding the literature, which include differences in the lesions that occur with brachial plexus injury, the interval and length of follow-up, the definition of recovery, and potential bias caused by referral patterns (eg, many neonates with upper brachial plexus palsy recover spontaneously and are not referred to centers that publish most studies) [38,68].

The more extensive initial injury associated with total brachial plexus palsy (panplexopathy) points to a less favorable outcome compared with upper (Erb) brachial plexus palsy. As an example, a prospective population-based study from Sweden followed 93 children with NBPP from birth to 18 months of age [24]. Overall, full functional recovery by age 18 months was regained in 82 percent. Children with C5 and C6 palsy had the highest functional recovery rate (61 of 64 [95 percent]), children with C5 to C7 palsy had an intermediate rate (7 of 11 [64 percent]), and children with C5 to T1 palsy had the lowest rate (3 of 14 [21 percent]).

In various studies, favorable prognostic indicators for NBPP included the following:

●Early clinical improvement (ie, onset of recovery within two to four weeks) [34,38]

●Elbow flexion at three months of age [24,71,93]

●Normal or near-normal strength in elbow flexion, shoulder external rotation, and forearm supination at three months of age [24]

●Recovery of antigravity strength in biceps, triceps and deltoid muscles by 4.5 months [38]

In contrast, the absence of spontaneous improvement over time is associated with a diminishing potential for recovery [68].

Electrodiagnostic studies, usually performed at three months of age, have not been considered reliable for predicting prognosis because motor unit action potentials (MUAPs) are frequently observed at that age in clinically paralyzed muscles, which ordinarily are associated with an absence of MUAPs [21]. However, a prospective study of 48 newborns with paralysis of elbow flexion found that absence of MUAPs in the deltoid and biceps muscles at one month of age on needle electromyography had high sensitivity for predicting absent elbow flexion at three months [48]. Additionally, a 2020 review of 16 observational studies including 747 children concluded electrodiagnostic studies may provide additive prognostic value to clinical and radiographic assessments and advocated for their performance by an experienced neurophysiologist with expertise in pediatric studies [94].

LITIGATION — Malpractice litigation related to NBPP is an important issue [95]. In a retrospective study of 51 consecutive children with NBPP, litigation was pursued in approximately half of the patients [96]. Surgery was the only obstetric variable associated with litigation; the proportion of children who had brachial nerve surgery was higher in the litigated compared with the non-litigated group (46 versus 8 percent). There were no differences in demographics, peripartum characteristics, operative versus spontaneous vaginal birth, shoulder dystocia, or high birth weight. In an earlier analysis of the same cohort of 51 children, psychosocial factors associated with malpractice litigation included the perception that the sustained birth injury was unnecessary, that the information received in the perinatal period was inadequate, and that family or caregiver concerns were ignored in the perinatal period or were not adequately addressed [97]. The extent of NBPP was not associated with litigation.

SUMMARY AND RECOMMENDATIONS

●Etiologies – Potential mechanisms of neonatal brachial plexus palsy (NBPP) include stretching/traction, compression, infiltration, and oxygen deprivation. Of these, stretching is considered the most common mechanism. However, NBPP can occur even when axial traction is properly applied; the occurrence of NBPP following birth does not automatically indicate that the practitioner applied forces or maneuvers that caused the nerve injury. Many cases of brachial plexus injury are not due to shoulder dystocia or excessive force by the provider. (See 'Etiology' above.)

●Risk factors – Neonatal brachial plexus palsy is uncommon. The only established risk factor for NBPP is shoulder dystocia. There are no proven measures that can predict or prevent NBPP. (See 'Epidemiology' above.)

●Clinical features – NBPP can be divided into the common upper brachial plexus palsy involving injury to C5, C6, and occasionally C7 (Erb palsy) and the less common total brachial plexus palsy involving all the roots from C5 to T1.

Upper palsy due to C5 and C6 injury manifests as adduction and internal rotation of the arm and forearm extension, while hand and wrist movement are preserved. When C7 is additionally involved in upper palsy, there is also flexion of the wrists and fingers (picture 1). Total brachial plexus palsy presents with arm paralysis, sometimes accompanied by a Horner syndrome. (See 'Clinical features' above.)

●Diagnosis – The diagnosis of NBPP is made clinically when arm weakness at birth fits a distribution consistent with a brachial plexus injury. In many cases, the diagnosis is straightforward. Adjuncts that may be useful for therapeutic decisions about the need for surgical nerve repair or reconstruction include electrodiagnostic studies and imaging. Because of the potential impact on prognosis and medicolegal issues, practitioners should try to determine the time when the brachial plexus injury occurred. (See 'Evaluation and diagnosis' above.)

●Management – A period of physical and occupational therapy and observation for evidence of recovery is often employed. Surgical intervention is advocated in select cases if functional recovery does not ensue in three to nine months, but there is no consensus regarding the utility or timing of surgery. Early referral to a center with expertise in the management of NBPP may improve outcomes. (See 'Management' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Barry Russman, MD, who contributed to earlier versions of this topic review.