Article Text

Download PDFPDF

The breathing hand: obstetric brachial plexopathy reinnervation from thoracic roots?
  1. S M Friedenberg1,
  2. R C Hermann2
  1. 1Department of Neurology, Geisinger Medical Center, Danville, Pennsylvania
  2. 2Department of Neurology, Mayo Clinic, Rochester, Minnesota
  1. Correspondence to:
 Dr. Scott Friedenberg
 Geisinger Medical Center, Department of Neurology (MC 14-05), 100 North Academy Avenue, Danville, PA 17822, USA


It has been found that in cases of obstetric brachial plexopathy, injured phrenic nerve or C3/4/5 roots may sprout into the adjacent injured upper and middle trunks of the brachial plexus. This aberrant regeneration produces co-contraction of the diaphragm and proximal upper limb muscles. This phenomenon, referred to as respiratory synkinesis or “the breathing arm”, may not be limited to the upper cervical roots. We present two cases, identified through electromyographic investigations, of respiratory synkinesis selectively affecting intrinsic hand muscles, and propose that upper thoracic roots and their intercostal nerves may also produce respiratory synkinesis, resulting in a “breathing hand.” This novel brand of synkinesis indicates that obstetric brachial plexus neuropathies can have quite proximal nerve injury in all trunks. The findings in our patients may not be entirely unique. The time required to develop distal muscle synkinesis and the subtle nature of our findings may suggest that with time and the assistance of EMG the breathing hand may be more common. When considering brachial plexus surgery, the significance of respiratory synkinesis should not be overlooked as its presence indicates injury at a root or proximal trunk level and may come from either nerves destined for the diaphragm or for the intercostal muscles.

  • synkinesis
  • nerve regeneration
  • brachial plexus
  • EMG
  • electromyograph
  • obstetric brachial plexopathy
  • MRC, Medical Research Council

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Case 1

A 1 year old baby was evaluated for her obstetric brachial plexus neuropathy and subsequent surgical repair of the upper portion of her brachial plexus. Her birth at 37 weeks (4.3 kg) was complicated by prolonged labour, left shoulder dystocia, and respiratory distress necessitating cardiopulmonary resuscitation. Neurological examination at 6 months revealed complete paralysis of the left upper extremity. CT myelography demonstrated a C7 pseudomeningocele. Surgery was pursued to try to identify sites for potential neurolysis or nerve grafting.

At surgery the brachial plexus appeared grossly stretched, with the upper trunk being pulled down such that the takeoff of the suprascapular nerve was at the level of the clavicle. The upper trunk also contained a large proximal neuroma. Intraoperative somatosensory evoked potentials were performed, showing cortical responses as well as limb muscle contraction with C5 and C6 root stimulation, but not with C7 root stimulation. In an effort to provide clinically useful proximal limb function, sural nerve fascicular grafts were made from the C5 root to the axillary nerve and from the C6 root to the anterior portion of the upper trunk. The spinal accessory nerve was directly anastomosed to the suprascapular nerve.

At 1 year of age, the child’s left arm was areflexic with Medical Research Council (MRC) grade 1-2/5 strength for the entire extremity. Nerve conduction studies at that time showed an absent median sensory response and low amplitude ulnar motor (0.7 mV), ulnar sensory (3 μV), and musculocutaneous motor (0.1 mV) responses. Needle examination demonstrated severe denervation (fibrillation potentials) and reinnervation (volitional activation of large, complex motor unit potentials) in her biceps brachii, deltoid, extensor digitorum communis, infraspinatus, and first dorsal interosseous muscles, and thus supported the presence of injury in all portions of the brachial plexus. Needle examination of the first dorsal interosseous muscle at rest showed no spontaneous motor unit potential activation, although there were large motor unit potentials with deep inspiration. At 33 months of age, the child could elevate the shoulder to 70°, had MRC grade 3/5 finger flexion and extension, and could appreciate sensation in the left upper extremity. On deep inspiration, intrinsic hand muscle contraction was visible to the naked eye. The patient’s family remarked that they had in fact noticed that the more excitable their daughter was, the more noticeable were her hand contractions.

Case 2

We evaluated a girl for her obstetric brachial plexus neuropathy. Born at 39 weeks (weighing 11 pounds), her 24-hour vaginal labour was complicated by left shoulder dystocia, meconium aspiration and respiratory failure. Throughout her infancy, her left upper extremity was paretic. She also had a mild left Horner’s syndrome (miosis and ptosis). During the course of her early childhood, her left arm function slowly improved.

Examination at the age of 8 years, 11 months demonstrated an atrophic, areflexic left upper limb with MRC grade 4/5 strength in the left triceps and upper trunk innervated muscles. The remaining muscle groups had MRC grade 2-3/5 strength, with the intrinsic hand muscles having the weakest contraction. Deep respiration did not evoke visible limb contraction. Mild left ptosis and miosis were present.

Left upper extremity nerve conduction studies showed low amplitude median (3 μV) and ulnar (5 μV) sensory responses, normal amplitude median (9.4 mV) and musculocutaneous (9.4 mV) motor responses, and a borderline ulnar motor response (4.5 mV, normal for age >4.1 mV). Needle examinations supported a diffuse, chronic brachial plexopathy with large, complex motor unit potentials under volitional control in the biceps brachii, deltoid, triceps, and first dorsal interosseus muscles. Cervical paraspinal needle examination was normal. Needle examination showed volitional activation of large motor unit potentials in the first dorsal interosseous and deltoid muscles. Deep inspiration evoked large motor unit potentials (respiratory synkinesis) in the first dorsal interosseous and deltoid muscles.


Respiratory synkinesis in the upper extremity has been attributed to aberrant mid-cervical roots and the phrenic nerve (due to their close proximity to the upper and middle trunks).1–3 It seems unlikely that phrenic nerve axons are responsible for our two patients’ synkinesis given the long and circuitous course that would be required to selectively link high cervical roots or the phrenic nerve to their inferior trunks or medial cords. We propose that the first and possibly second thoracic roots and their intercostal nerves (which contribute to the inferior trunk),4 including their thoracic intercostal branches, were damaged during the children’s deliveries. Thus, injured nerves destined for respiratory muscles would lie in very close proximity to the inferior trunks’ empty nerve sheaths. This anatomy (figs 1A and B) could produce preferential lower trunk respiratory synkinesis. The deltoid synkinesis of Case 2 could reflect thoracic root rather than phrenic nerve axons extension (fig 1B).

Figure 1

The proposed course of thoracic roots and intercostal nerves mediating respiratory synkinesis. (A) Case 1: locations of sural nerve grafts (solid line) and the proposed course of aberrant thoracic roots and intercostal nerves (broken line). (B) Case 2: the proposed aberrant course of the thoracic roots and intercostal nerves (broken line). Note two possible sources of synkinesis in the deltoid muscle, thoracic intercostal nerves or phrenic nerve (solid line).

Thoracic root mediated synkinesis may not be an entirely novel phenomenon. Intercostal nerve associated synkinesis (to latissimus dorsi) has been reported following thoracotomy, where the proximity of the lesion was also felt to contribute to the evolution of synkinesis.5 In the case of obstetric brachial plexus neuropathy, it must be kept in mind firstly that neurophysiological testing typically occurs early after injury (for surgical planning). At this early stage, proximal sprouts may not have had sufficient time to establish connections with distal inferior trunk innervated muscles. Secondly, the synkinesis in one of our patients was not a robust phenomenon and would have gone undetected without the sensitivity of electromyography.

Our patient’s neuropathies would be best categorised as Erb-klumpke palsies, involving roots from CS to TI. The presence of an Erb-klumpke palsy generally indicates that there has been a significant injury to the proximal nerve or nerve root.6 The presence of thoracic root mediated synkinesis would further imply, as it does in phrenic nerve mediated synkinesis,3 that nerve disruption is quite proximal. Therefore, the severity and proximal location of the injury could limit the potential for nerve regeneration and may further explain why this phenomenon has not previously been reported. Cases of diffuse limb synkinesis, including lower trunk innervated muscles have been reported,12 and it has been proposed that nerves to accessory muscles may also contribute to synkinesis. It is reasonable then to consider that respiratory synkinesis following a severe brachial plexopathies may reflect aberrant regeneration from nerves originating in the cervical or thoracic cord, or both.

Our electromyographic studies were not extensive because of the children’s young age, and this limitation must be recognised. Extensive needle examination may have revealed more diffuse synkinesis, although it would not necessarily discount the possibility of thoracic root contribution to their syndromes. Phrenic nerve or thoracic root stimulation studies may have been helpful in separating these possibilities. In addition, simultaneous recordings from respiratory and intrinsic hand muscles (as was performed in a case of arm-diaphragm synkinesis2) could have been helpful in documenting the synkinesis.

Our cases reflect the unfortunate 10–20% of patients with obstetric brachial plexopathy who have a poor outcome and often require surgical consideration.7 Proper surgical planning includes obtaining a firm understanding of the patient’s anatomy.6 Respiratory synkinesis should be sought in all portions of the brachial plexus, as its presence indicates that the plexus lesion is likely quite proximal3 and involves high cervical roots, phrenic nerves or thoracic roots, and/or intercostal nerves. Clinically, the existence of synkinesis indicates an injury that was severe enough to disrupt multiple proximal nerve elements, and that while the distal receiving trunk of the brachial plexus was injured, the continuity of the nerve sheath was sufficient to provide the milieu for nerve regeneration. Finally, identifying synkinesis indicates that the aberrant nerve regeneration had been progressing for quite some time. It is uncertain, however, whether the presence of synkinesis portends a particular prognosis, either with respect to conservative or surgical management. This issue may merit further consideration, especially in cases where phrenic and intercostal nerve/thoracic root grafting9 are being proposed.



  • Competing interest: none declared