SEARCH

SEARCH BY CITATION

Keywords:

  • brachial plexus;
  • cervical nerve root;
  • neonates;
  • obstetric injury;
  • ultrasound

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Introduction: In this exploratory study we investigated whether ultrasound can visualize the neonatal cervical roots and brachial plexus. Methods: In 12 healthy neonates <2 days old, the neck region was studied unilaterally with ultrasound using a small-footprint 15–7-MHz transducer. Results: The C5–C8 nerve roots and brachial plexus could be imaged with sufficient delineation of the root exits to assess their integrity. The brachial plexus was more difficult to discern from the surrounding area in neonates compared with adults, especially in the interscalene region because of the smaller amount of connective tissue in and surrounding muscles and nerves. In addition, the large deposits of brown fat make for a different ultrasound appearance of the neonatal neck compared with adults. Conclusions: Ultrasound of the neonatal cervical nerve roots is feasible and may be used as a non-invasive screening technique to assess nerve root integrity in obstetric brachial plexus injury. Muscle Nerve 51: 35–41, 2015

Abbreviations
CSA

cross-sectional area

CT

computed tomography

OBPI

obstetric brachial plexus injury

Obstetric brachial plexus injury (OBPI), also called neonatal brachial plexus injury, is the most common peripheral nerve injury in infants,[1] affecting an estimated 0.5–3 per 1000 live births. About 70% of affected children have complete recovery. In cases of severe nerve damage without signs of improvement, surgical intervention is indicated. During the first weeks of life it is difficult to determine which children will need future surgical treatment, as reversible nerve damage (neurapraxia and axonotmesis) and irreversible nerve damage (neurotmesis, including root avulsion) have similar clinical presentations. Also, although after 1 month severe obstetric brachial plexus lesions can be identified with high sensitivity, the specificity of the proposed algorithm is only approximately 70%.[2, 3] As the effectiveness of reconstructive surgery for neurotmesis and root avulsions has decreased over time, new, preferably non-invasive diagnostic instruments that allow prediction of OPBI outcome in an early stage would be helpful.[4]

Previous studies have used imaging techniques to depict the cervical nerve roots and brachial plexus in OBPI, although validated diagnostic imaging studies are sparse. Root avulsions can be detected with computed tomography (CT) myelography,[5] but the diagnostic accuracy of this technique is unknown, and its application is limited in newborns because of its invasive nature. Assessment of injury at the level of the brachial plexus proper with magnetic resonance imaging (MRI) can be challenging, because the plexus cannot be visualized in the standard scan planes (i.e., coronal, sagittal, and transversal).[6, 7] MRI can show associated abnormalities such as myelomeningocele,[8] but it poses specific difficulties in infants due to relatively long scanning times and the need for immobilization and sedation.

Recent advances in ultrasound technology allow visualization of peripheral nerves with high resolution.[9] Because ultrasound is a bedside tool that can be used without sedation, even in the very young, it is a potentially attractive diagnostic technique for infants with OBPI. Brachial plexus ultrasound imaging has been shown to be feasible in adults.[10-12] So far, it is unclear whether the small nerves of the brachial plexus and cervical nerve roots of neonates can also be visualized. Their appearance and size are currently unknown. In this study we aimed to investigate the feasibility of neonatal cervical root and brachial plexus ultrasound using a high-resolution, small-footprint ultrasound probe. When feasible, this technique could be used next in a diagnostic study to determine its accuracy for detection of neurotmesis or root avulsions in OBPI in newborn infants.

METHODS

  1. Top of page
  2. ABSTRACT
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Participants

Twelve healthy Caucasian neonates, <2 days old, were enrolled in the study after written informed consent was obtained from their parents. The study was approved by the regional ethics committee. Subjects characteristics are presented in Table 1. No infant had signs of OPBI, clavicle fracture, or other peripartum problems.

Table 1. Demographic features of participants (N = 12)
Gender8 boys / 4 girls
Gestational age, median (range)40 (36−41) weeks
Weight, median (range)3720 (2680−4640) g
Height, median (range)50 (48−53) cm
Head circumference, median (range)35 (33−37.8) cm
Side of measurement4 left / 8 right
Reason for admission to the clinicCaesarean section, n = 8; vacuum extraction, n = 2; possible neonatal infection, n = 2

Ultrasound Measurements

All measurements were made using a broadband linear 15–7-MHz small-footprint probe (Philips IU22; Philips, Best, The Netherlands). System settings were adjusted to create the optimum image for differentiating nerve from surrounding tissue using the available imaging enhancement techniques. The infant was placed supine, with the head slightly extended and rotated to the contralateral side (Fig. 1). Because of the exploratory nature of the study and time constraints, each infant was measured unilaterally (Table 1). The side of measurement depended on the location of the bed with respect to the ultrasound device. The integrity of the cervical roots and brachial plexus was determined by following all nerve roots from the foramina to the clavicle in both a transverse and longitudinal plane. Images were captured at 3 locations (nerve root at the foraminal exit, interscalene region, and supraclavicular region), and measurements were made online. The supraclavicular brachial plexus with its typical “cluster-of-grapes” shape was easily recognizable as a starting point for the measurements for identification of other brachial plexus and root levels.

image

Figure 1. With a small-footprint transducer it is possible to evaluate the neonatal brachial plexus in the longitudinal plane. The child is placed supine with the head slightly extended, which, for reference, is shown in the photograph of this 2-month-old boy.

Download figure to PowerPoint

Supraclavicular Region

The location for measuring the brachial plexus in the supraclavicular region was based on the protocol described by van Geffen and co-workers.[10] The center of the probe was positioned at the midclavicular level parallel to the clavicle and moved gradually more laterally to identify the typical clustered nerve structures (Fig. 2). In this way the brachial plexus could be identified close (Fig. 2, left) or just cranial (Fig. 2, right) to the first rib. The cross-sectional area (CSA) of the nerves at this level was determined by delineating the entire cluster using the caliper function available on the ultrasound device. This method includes all peri- and epineural fibrous tissue within this region of the brachial plexus.

image

Figure 2. The brachial plexus in the supraclavicular area appears as a “cluster-of-grapes” formation (circle) at a depth of approximately 1 cm between the anterior (AS) and middle scalene (MS) muscles (right image) or, slightly more laterodistally, above the first rib. Both the anterior and middle scalene muscles and the plexus are surrounded by brown fat (arrows), appearing as fine granular hyperechoic tissue.

Download figure to PowerPoint

Interscalene Region

The superior, middle, and inferior trunks of the brachial plexus were identified between the anterior and middle scalene muscles by tracing the nerve segment in the “cluster-of-grapes” structure back in the proximal, that is, craniomedial, direction. The CSA was determined for each separate nerve segment by tracing within the hyperechoic rim when possible.

Cervical Nerve Roots

The C5–C8 nerve roots were measured in the transversal and longitudinal planes. The shape of the vertebra was used to identify root level.[13] Also, the entrance of the vertebral artery was determined, because in the majority (94%) of patients it enters at the level of the C6 nerve root.[14] In each measurement, an attempt was made to capture all roots in the same plane in a single image (Fig. 3). If not possible, individual nerve roots were imaged (Fig. 4). The nerve diameter was measured with electronic calipers at the root exit from the foramen.

image

Figure 3. In a longitudinal plane the nerve roots are visible as black lines, without any fascicular structure. In this image the vertebral artery is visible outside the foramen at the level of the C7 vertebra. SCM, sternocleidomastoid muscle; BF, brown fat; AV, vertebral artery.

Download figure to PowerPoint

image

Figure 4. Longitudinal images of the C5–C8 nerve roots (arrow). After exiting the neuroforamen, the roots take a sharp bend caudally, especially at the C5 and C6 levels. This makes it difficult to visualize all the nerve roots in a single plane.

Download figure to PowerPoint

Analysis of Findings

Statistical analysis was performed using Excel 2003 (Microsoft Corp., Redmond, Washington). Medians and ranges of diameters and CSAs were calculated.

RESULTS

  1. Top of page
  2. ABSTRACT
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

In all but 2 subjects the measurements could be performed without any problems, as these 10 children slept throughout the whole 45 minutes it took, on average, to complete the examination. In 2 subjects the examination could not be completed, and only the supraclavicular and interscalene levels were measured, because the infants became hungry and needed to be breast-fed.

Supraclavicular Region

The “cluster-of-grapes” structure in the supraclavicular area was recognizable in all subjects (Fig. 2). It was located at a depth of between 0.5 and 1 cm, depending on the infant's position. Quite different from what is seen in adults, the neonatal brachial plexus and other structures in the neck are surrounded by brown fat,[15] which shows up as a region with a fine granular hyperechoic aspect (Fig. 2). Due to the small size of all structures and low reflectivity of the nerves, muscles, and surrounding fibrous tissue in most subjects, it was difficult to sharply delineate these structures from each other. In these difficult cases we chose to estimate the CSA of the brachial plexus elements at this level using the ellipse caliper function, as free tracing would probably not provide extra accuracy. The median CSA of the clustered structure was 7.2 (range 4.1–18) mm2 (Table 2), which, for reference, is about the size of an adult median nerve in the forearm.

Table 2. Cross-sectional area (transversal plane) and diameter (longitudinal plane) of cervical roots at foraminal exit and supraclavicular brachial plexus
 Diameter (mm)Cross-sectional area (mm2)
 NaMedian (range)NaRange
  1. NA, not assessed.

  2. a

    The number of measurements is lower than the total number of subjects, because it was not possible to reliably measure each structure in all participants.

C591.5 (0.7–2.3)22–3.9
C6111.5 (0.9–2.2)22–2.4
C7101.6 (1.0–2.7)31.4–4.2
C841.5 (1.3–1.8)22.1–4.1
Brachial plexusNANA74.1–18

Interscalene Region

The nerves at this location also appeared to have little perineural fibrous tissue, which made it difficult to differentiate them from surrounding muscle and fibrous tissue. Following a learning curve, they could be visualized from the third subject onward (n = 10). In 4 subjects it was possible to capture the C5–C8 roots and brachial plexus nerve trunks in 1 image, where they appeared as black dots between the anterior and middle scalene muscles (Fig. 5). In the other subjects the individual trunk elements were only visible as round dots or single black lines that ran in line with the direction of the ultrasound beam toward the root foraminal exit (Fig. 6). The CSAs of the nerve roots in the interscalene area were not measured.

image

Figure 5. Transverse image of the C5–C7 nerve roots in the interscalene area (left: infant; right: adult). They appear as black dots with a “traffic light” appearance. In infants, this is only seen in a minority of subjects, as often the roots cannot be visualized in a single plane; in adults, this is an easily recognized location and a good starting point for examination of the brachial plexus. SCM, sternocleidomastoid muscle; JV, jugular vein; CA, carotid artery; AS, anterior scalene muscle; MS, middle scalene muscle.

Download figure to PowerPoint

image

Figure 6. Transverse image of the C5 nerve root in the interscalene area. Due to the size and direction of the nerve it was not possible to obtain a true transverse image in the interscalene region. In this image, the nerve is visible, exiting the C5 foramen and running into the interscalene region. AS, anterior scalene muscle; MS, middle scalene muscle.

Download figure to PowerPoint

Nerve Roots

The C5–C7 vertebrae were identified easily based on their shape or, in a longitudinal plane, based on the entrance level of the vertebral artery (Fig. 7). In a longitudinal plane, the extraforaminal position and exit of the C5, C6, and C7 roots could be visualized easily in all subjects (Figs. 3 and 4). The C8 root required additional experience, but it could be visualized from the third subject onward. The T1 root level could not be identified positively in any of the subjects.

image

Figure 7. The C6 root and vertebra were good landmarks for determining the level of measurement. Both the shape of the C6 vertebra (left picture) and the entrance of the vertebral artery (AV) were used. SCM, sternocleidomastoid muscle; BF, brown fat.

Download figure to PowerPoint

No major size differences were found between the different nerve root levels, whereas, in adults, C5 usually has the smallest diameter and C7 the largest. The longitudinal images also show that the C5–C8 nerve roots appeared as a sharp delineated black line that tended to run a more or less horizontal course for the first 3–4 mm and then make a sharp bend, running at a 125–140° angle toward the supraclavicular region. The median diameter of the roots was 1.5 mm (range 0.7– 2.7 mm; Table 2). For comparison, an adult nerve root has a diameter of approximately 3.5 mm.[16] After a learning curve, the nerve roots could also be identified in a transverse plane leaving the foramen from the 4th subject onward (Fig. 8). The median nerve root CSA was 2.1 mm2 (range 1.4–4.2 mm2; Table 2). For comparison, an adult nerve root has a CSA of approximately 10 mm2.16

image

Figure 8. Transverse images of the C5–C8 nerve roots at the level of the neuroforamen (arrow).

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. ABSTRACT
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

In this exploratory study we showed that ultrasound visualization of the C5–C8 neonatal cervical nerve roots and brachial plexus is feasible, with sufficient delineation of the root exits to assess their integrity. In our experience more operator training is required for neonates than adults before all neural elements can be identified reliably. Using a small-footprint transducer (e.g., a probe surface of ≈0.5 × 2 cm) it was possible to capture the cervical roots in a single longitudinal image in most neonates. Nerve roots appeared as thick black lines entering the foramina and, as seen in most adults, no individual nerve fascicles could be discerned within the roots in any subject. The smaller amount of connective tissue in and surrounding muscles and nerves in neonates, which provide ultrasound reflections and image detail in adults, made it more difficult to discern the brachial plexus trunks from the surrounding tissue, especially in the interscalene area. In addition, the presence of large amounts of brown fat produces an ultrasound appearance of the neonatal neck unlike that seen with adults.

The extraforaminal position of the C5–C8 nerve roots and the continuity of the first 1.5 cm of peripheral nerve could be imaged well in most subjects. C4 and T1 could not be detected reliably in any infant. More scanning experience and the use of a higher frequency transducer (up to 20 MHz) may improve recognition of peripheral nervous system elements in this population. In contrast to the use of the interscalene region in adults, with the “traffic light” appearance of the trunks as a landmark for scanning the brachial plexus (Fig. 7), our results do not support this approach in neonates, as the lack of reflective (connective) tissue provided little contrast between the scalene muscles and the neural elements. Further studies in OPBI patients must determine whether it is possible to visualize the interscalene roots with sufficient resolution to detect traumatic transection or neuroma. Fortunately, the brachial plexus in the supraclavicular region was easily visualized in every infant, showing the same variation in shape of the “grape cluster” as seen in adults.[17]

In adults, ultrasound of the brachial plexus has been shown to be a useful imaging technique for preoperative diagnosis of brachial plexus lesions. It can differentiate between pre- and postganglionic lesions, showing high predictive values for detection of mass lesions or entrapment,[18] as well as root avulsions after traumatic brachial plexus injury.[19, 20] Because the ultrasound aspect of the nerve roots in neonates was very similar to that in adults, it seems plausible that nerve avulsions could also be visualized in neonates using high-resolution ultrasound. Because the majority of patients with severe OBPI show root avulsions on CT myelography,[5] ultrasound could potentially replace this invasive imaging technique and the exposure to ionizing radiation in this vulnerable population. Furthermore, detection of root avulsion not only affects the decision to perform surgery but also affects the specific surgical approach.[5] In adults with traumatic brachial plexopathy, the diagnostic accuracy of CT myelography and MRI is 85% and 52%, respectively, for determining root avulsion.[21] These values are lower than the predictive values for nerve ultrasound in adult brachial plexus injury.[19] In neonates with suspected OBPI, CT myelography was found to have a positive predictive value of 50% and a negative predictive value of 93% for detecting root avulsion.[22] The sensitivity for detecting root avulsion by finding a pseudomeningocele was low for MRI (50%), but with high specificity (almost 100%).[6] As it is uncertain whether the resolution of ultrasound is sufficient for visualization of intraspinal pathology in neonates, we cannot predict its use for detection of intraforaminal pseudomeningocele. Ultrasound will probably continue to focus on extraforaminal nerve integrity and pathology. Moreover, prospective assessments will need to determine the clinical significance of these findings, as MRI studies have shown that root avulsions in neonates have an uncertain relationship with functional outcome.[8]

In conclusion, in this pilot study we have shown that it is possible to visualize the neonatal cervical nerve roots and brachial plexus with ultrasound, making it an attractive, non-invasive bedside tool to potentially replace the more cumbersome imaging techniques used for infants with OBPI. Further study is warranted to determine the possibilities of ultrasound for detection and definition of nerve trauma in OPBI patients as compared with other imaging modalities.

REFERENCES

  1. Top of page
  2. ABSTRACT
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES