Pediatric Radiology (2022) 52:1648–1657

https://doi.org/10.1007/s00247-021-05180-y

MINISYMPOSIUM: SMALL PARTS AND MUSCULOSKELETAL ULTRASOUND

Ultrasound assessment of glenohumeral dysplasia in infants

Sarah J. Menashe 1 & Anh-Vu Ngo 1 & Marisa B. Osorio 2 & Ramesh S. Iyer 1

Received: 18 May 2021 / Revised: 10 July 2021 / Accepted: 2 August 2021 / Published online: 21 September 2021
# The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2021

Abstract
Brachial plexus birth injury can lead to irreversible neuromuscular dysfunction and skeletal deformity of the upper extremity and
shoulder girdle, ultimately resulting in glenohumeral dysplasia. Diagnosis and treatment of affected infants requires a multi-
disciplinary approach in which imaging plays a vital role. While MRI is excellent for assessing both the shoulder and spine of
these children, it is costly and requires sedation and is thus typically reserved for preoperative planning. US, however, is
inexpensive, dynamic and readily available and provides excellent visualization of the largely cartilaginous glenohumeral joint.
As such, it has become a highly useful modality during early diagnosis and follow-up of children with brachial plexus birth
injuries. In this review, we describe the relevant anatomy of the glenohumeral joint, outlining the normal sonographic appearance
as well as providing tips and tricks for identifying and characterizing pathology.

Keywords Brachial plexus . Brachial plexus birth injury . Glenohumeral dysplasia . Infants . Magnetic resonance imaging .
Musculoskeletal . Ultrasound

Introduction motor innervation to the shoulder, upper chest, arm and hand.
Brachial plexus birth injury most commonly involves the up-
Brachial plexus birth injury has been reported to occur in 1.5– per trunks of the C5–T1 nerve roots, specifically C5–C6,
4 per 1,000 live births, though some report the incidence has though isolated lower trunk injuries rarely occur [5, 6]. The
been decreasing, possibly a result of an increasing rate of most common form of nerve injury — occurring in 70–90%
cesarean sections [1–3]. Typically seen during vaginal deliv- of children with brachial plexus injury — is neurapraxia, a
ery, downward traction on the infant’s head and neck com- focal traumatic “stretching” injury without axonal disruption;
bined with partial obstruction of the shoulders by the maternal this typically resolves spontaneously, usually by 2–3 months
pubis leads to predictable patterns of nerve injury [4]. Many of age. However, the remaining 10–30% of these children
risk factors for brachial plexus birth injury have been de- have residual impairment [7]. In 1987, the Narakas classifica-
scribed and include shoulder dystocia, macrosomia, instru- tion divided obstetric birth palsies into four groups based on
mentation during delivery, and gestational diabetes; some the palsies observed clinically. Widely adopted in clinical
children, however, present in the absence of known predispos- practice, each Narakas group corresponds with the level of
ing conditions [2, 5]. nerve root injury, whereby the higher the classification, the
The brachial plexus is primarily composed of anterior rami more nerve roots are injured [8].
originating from C5–T1 nerve roots, and provides sensory and Unfortunately, brachial plexus birth injury can lead to irre-
versible neuromuscular dysfunction and skeletal deformity of
the shoulder girdle, which can include joint contractures, pos-
terior humeral head subluxation and glenohumeral dysplasia
* Sarah J. Menashe
[email protected] [9–11]. The true incidence of glenohumeral dysplasia, a well-
recognized complication of brachial plexus birth injury in
1
children with persistent neurologic defect at 3 months, varies
Department of Radiology, Seattle Children’s Hospital
and University of Washington School of Medicine,
widely in the literature [6, 12]. However, though a result of the
4800 Sandpoint Way NE, MA.7.220, Seattle, WA 98105, USA initial insult to the brachial plexus, glenohumeral dysplasia is
2
Seattle Children’s Hospital,
considered a separate and evolving pathology with respect to
Medicine and University of Washington School of Medicine, the developing shoulder and can lead to significant shoulder
Seattle, WA, USA dysfunction and morbidity [13–15]. The pathophysiology of
Pediatr Radiol (2022) 52:1648–1657 1649

glenohumeral dysplasia is not entirely understood but is (Sup) the forearm and externally rotating (ER) the shoulder.
thought to be the result of an imbalance between the relatively Our institution adopted the use of the Sup-ER splint, and we
unaffected internal rotators and the paralytic external rotators, employ US monthly to evaluate the glenohumeral relationship
such that when left untreated, posterior subluxation and even- both in and out of the splint to ensure the splint is maintaining
tual dislocation of the humeral head ensues [16]. As with the proper reduction. However, if the Sup-ER splint is unable to
pediatric hip joint, evolving joint incongruency hinders nor- maintain proper glenohumeral alignment and subluxation oc-
mal development of the glenoid, eventually leading to curs, the infant is placed in a spica cast (Fig. 1).
glenohumeral dysplasia [17, 18]. Interestingly, more profound
nerve injury, and correspondingly higher Narakas classifica-
tion in the neonatal period, has been associated with less Ultrasound imaging
glenohumeral dysplasia in older children; this is thought to
be a function of earlier surgical intervention in more severely Imaging of infants with brachial plexus birth injury is done
affected children [19]. primarily with gray-scale static and dynamic US and non-
contrast-enhanced MRI. CT has been used historically,
though it is no longer a preferred modality, particularly in
Clinical presentation, treatment and imaging younger children and infants who are more susceptible to
algorithm the risks of ionizing radiation and in whom unossified carti-
lage is better visualized with other cross-sectional modalities
Our multidisciplinary brachial plexus injury clinic evaluates [5, 25]. MRI has the advantage of being able to evaluate mor-
infants starting at 1 month of age because the majority of phological changes of the glenohumeral joint and shoulder
neurapraxic injuries have resolved by this time. Assessments girdle, in addition to allowing for accurate measurement of
at each clinic visit include passive range of motion, active range many other established parameters of the bony and cartilagi-
of motion and Active Movement Scale scoring, which grades nous glenohumeral joint. As such, MRI is considered the gold
the infant’s ability to move the limb both with gravity eliminat- standard in the preoperative assessment of glenohumeral dys-
ed and against gravity [20]. Infants are followed up at least plasia, and it is often performed in older infants or in more
every 3 months to determine their recovery trajectory, and in- advanced cases of dysplasia to delineate complex anatomy
terventions are based on the algorithm established by the [25]. Nevertheless, MRI has the distinct disadvantage of re-
Hospital for Sick Children in Toronto, Canada [21]. quiring sedation outside of the neonatal period, the long-term
Glenohumeral US screening typically begins at 3 months of effects of which are poorly understood, in addition to being
age because glenohumeral subluxation has been shown to be more costly and, often, less available [26]. Thus, in neonates
present by this age. Infants are screened at 3 months and and infants without severe palsy requiring immediate surgical
6 months of age, regardless of clinical exam, and intervention intervention, US may be used in the screening evaluation of
is based on humeral head alpha angle (as described later) and glenohumeral dysplasia. A rapid, safe, cost-effective, dynamic
passive range of motion. Unfortunately, the clinical exam is not and readily available modality, US is valuable in the early
always predictive of shoulder subluxation, with many infants assessment and follow-up of these children [4, 25, 27]. As
exhibiting abnormal alpha angles despite having full passive noted, in our multidisciplinary brachial plexus injury clinic,
range of motion [19, 22]. If the alpha angle is greater than 40° US comprises the vast majority of imaging in infants with
and there is restricted passive external rotation, infants undergo suspected brachial plexus birth injury, and for routine
manual shoulder reduction, injection of botulinum toxin A into glenohumeral dysplasia monitoring to guide treatment
the pectoralis major, latissimus and subscapularis muscles, and efficacy.
application of a shoulder spica cast for 6 weeks to maintain joint
reduction [23]. Of note, casting for glenohumeral dysplasia General approach
involves abducting and externally rotating the shoulder, similar
to hip spica casting for developmental hip dysplasia. US is Ultrasound of the glenohumeral joint is performed from a
initially repeated when the cast is removed to ensure appropri- posterior approach with a high-frequency linear transducer
ate glenohumeral reduction, then again at 1–2 months post cast (e.g., 10–15 mHz). Infants are typically seated on the exami-
removal to ensure the reduction is maintained. nation table facing away from the sonographer, head and trunk
Prevention of shoulder subluxation can be a challenging supported by the caregiver, though the child might also be
endeavor for children with brachial plexus birth palsy. One placed in the lateral decubitus position. The arm is then
reasonable option to address this issue is a splinting protocol adducted against the body (i.e. elbow against the infant’s side
developed at British Columbia Children’s Hospital called the and wrist against the lower torso) with the shoulder in internal
Sup-ER splint [24]. The splint is designed to keep the shoulder rotation, to accentuate or unmask posterior glenohumeral sub-
passively positioned in glenohumeral reduction by supinating luxation, if it is present. The probe is placed in a transverse
1650 Pediatr Radiol (2022) 52:1648–1657

Fig. 1 Casting. a Clinical image

of 5-month-old boy with upper
trunk brachial plexus palsy who
underwent glenohumeral
reduction, botulinum toxin
injections and placement of
shoulder spica cast to maintain
reduction. b Clinical image of a 3-
month-old girl demonstrates a
Sup-ER splint

oblique plane along the axis of the scapular spine (Fig. 2). positioning and imaging approach to these children.
Both affected and unaffected sides are imaged to provide in- Furtherm ore, the radiologist’s understanding of
ternal controls for comparison. If the child has undergone glenohumeral anatomy and the expected evolution of
splinting or casting, this is usually removed prior to imaging, glenohumeral dysplasia is paramount for accurate
though both in- and out-of-splint imaging can be performed. assessment.
Two to three static images of the affected shoulder can be
helpful in ascertaining the humeral head alpha angle, and cine Normal anatomy
sweeps through the joint can be performed either routinely or
only for troubleshooting purposes. With the probe in place posteriorly, the posterior scapular
Sonographic evaluation of the humeral head with re- spine is readily identified as a linear, highly echogenic struc-
spect to the scapula has been likened to that of the hip, ture, deep to which resultant acoustic shadowing precludes
where evaluation of dysplasia is both reliable and repro- visualization of deeper intrathoracic structures. At the lateral
ducible in experienced hands [28]. Conversely, a lack of margin of the scapula lies the bony glenoid; it is typically
operator expertise can lead to false-negative or -positive echogenic with a sharp posterior margin or corner, similar
exams. It is therefore important that sonographers and to that seen along the superolateral acetabular roof in the
interpreting radiologists alike are trained in the optimal pediatric hip. The normal glenoid fossa angles steeply away
from the transducer such that its anterior margin is difficult to
visualize from this posterior approach. The overlying
glenoid cartilage is a thin, homogeneously hypoechoic,
cap-like structure, which, along with the subjacent bony
glenoid, can be likened to a “steep cliff” on sonographic
images. Immediately superficial to the posterior corner of
the bony glenoid and the overlying cartilaginous cap, one
can identify the posterior component of the glenoid labrum,
a circumscribed, triangular, fibrocartilaginous structure that
is typically more echogenic than the surrounding subcuta-
neous tissues. Finally, the humeral head is recognized
more laterally as a round, largely hypoechoic structure
with a speckled echotexture that is similar to other epiph-
yseal cartilaginous structures. A central rounded
echogenicity with posterior acoustic shadowing might be
present soon after birth and reflects the normal ossifica-
tion center, a structure that increases in size during infan-
cy. The humeral head should be spherical in configuration
with smooth circular margins throughout. The soft tissues
Fig. 2 Clinical image shows glenohumeral US acquisition from the superficial to the echogenic scapula at this level consist
posterior approach with the infant in the seated position, the arm
adducted, flexed at the elbow, and internally rotated with the elbow at
primarily of the deltoid and infraspinatus muscles, though
the infant’s side and the wrist along the lower torso. The US probe is held portions of the supraspinatus might also be evident
in a transverse oblique plane in the axis of the scapular spine (Fig. 3) [27].
Pediatr Radiol (2022) 52:1648–1657 1651

morphology to the posterior head, which can be mistaken for

dysplasia. Care should be taken to ensure the infant’s arm
positioning is optimal (i.e. shoulder in internal rotation) to
avoid this error (Fig. 5).

Glenoid and labrum

With progressive glenohumeral dysplasia, the

osteocartilaginous glenoid can become rounded or irregular
along its posterior margin and angle away from the transducer
less steeply than normal, turning the “cliff” into a more grad-
ual slope (Fig. 6). This appearance is analogous to the dys-
plastic hip, where the superolateral acetabulum has a blunted
appearance and the tectum becomes “shallow.” Similarly, the
Fig. 3 Normal glenohumeral joint. Transverse oblique US image in a 3- posterior glenoid labrum might become irregular or
month-old boy depicts the normal structures of the glenohumeral joint. hypertrophied in appearance (Fig. 6). These findings should
The large circle represents the margins of the unossified humeral head be described in the imaging report; however, it is important to
keep in mind that only the posterior aspect of these structures
Qualitative findings in glenohumeral dysplasia is well seen. If more comprehensive assessment is required,
particularly in older infants who are operative candidates,
The normal anatomy of the glenohumeral joint might be al- glenohumeral morphology remains best assessed by MRI,
tered in glenohumeral dysplasia and should be carefully scru- where glenoid shape can be categorized in the context of the
tinized during the sonographic evaluation. Birch classification system as concave-flat, convex or bicon-
cave [4, 29]. Such classification is important for surgical plan-
Humeral head ning and can be difficult to assess by US.

Humeral head morphology can be affected by glenohumeral Shoulder muscles

dysplasia, whereby progressive dysplasia can lead to humeral
head flattening or even growth arrest [10]. Comparison with When there has been significant brachial plexus birth injury,
the contralateral side is useful in determining whether there is loss of bulk and fatty atrophy of the surrounding musculature
true deficiency in humeral head size because the affected side ensues, typically involving many of the muscles of the shoul-
might be noticeably flatter or smaller than the unaffected side. der and chest, including the deltoid, supraspinatus,
In most cases of mild to moderate dysplasia, the humeral head infraspinatus, subscapularis, teres major and pectoralis mus-
retains its normal spherical configuration, though the affected cles [30, 31]. While readily depicted by MRI, portions of the
epiphysis might have less ossification than its normal contra- infraspinatus and deltoid muscles can also be visualized dur-
lateral counterpart (Fig. 4). Importantly, if the humeral head is ing US examination of the glenohumeral joint; additionally,
imaged in external rotation, the greater tuberosity can come portions of the supraspinatus can also be seen. As such, mus-
into view, giving the spurious appearance of oblong or ovoid cular atrophy can be suggested on US, facilitated by

Fig. 4 Glenohumeral joint in left arm palsy. a, b Transverse oblique US images of the glenohumeral joint in a 3-month-old boy with left arm palsy. Left-
side image (a) demonstrates a smaller humeral head epiphyseal ossification center when compared to the right (b)
1652 Pediatr Radiol (2022) 52:1648–1657

Fig. 5 Suboptimal positioning during glenohumeral head imaging. a, b humeral head, simulating posterior subluxation and apparent dysplasia.
Transverse oblique US images in a 5-month-old girl’s left glenohumeral In intermediate external rotation (b), the humeral head has a less
joint with the shoulder in maximal (a) and intermediate (b) external accentuated oblong configuration rather than the normal spherical
rotation. In maximal external rotation (a), the greater tuberosity shape, which could still be falsely interpreted as dysplasia
(arrows) comes into view and creates an oblong appearance to the

comparisons to the unaffected shoulder or even more remote head displacement. Some authors have used US to measure
images of the affected side. This might be evident as overall glenoid version, or the degree of glenoid sloping as noted on
decreased muscle bulk or increased muscle echogenicity, sug- an axial images, as well [25].
gestive of fatty infiltration (Fig. 7) [32]. Because muscle atro-
phy in this setting reflects sequela of nerve injury, this can take Alpha angle
time to develop and might not be identified in very young
infants or in children with milder dysplasia. Additionally, The humeral head alpha angle is obtained by drawing a line
asymmetrical arm positioning — whether iatrogenic or as a along the posterior margin of the scapula and extending it
result of dysplasia — can lead to perceived differences in through the humeral head, then drawing a second line origi-
muscle bulk that might be inaccurate. For example, if the nating from the posterior glenoid margin and extending it
unaffected side is less internally rotated, the deltoid might be tangential to the posterior margin of the humeral head. The
less stretched over the scapula and appear symmetrical or even angle that is formed along the posterolateral intersection of
smaller than the affected side (Fig. 7). these lines is the humeral head alpha angle. A normal angle
is considered approximately 30° or less, with angles greater
Quantitative measurements in glenohumeral than 30° indicating increasing posterior subluxation of the
dysplasia humeral head [12]. One must take care not to obtain this mea-
surement on an image captured in external rotation because
Measurements of the glenohumeral joint that can be made on this can underestimate the alpha angle and result in a normal
US include the alpha angle and, in neonates, posterior humeral angle measurement in children with mild dysplasia. Likewise,

Fig. 6 Glenoid and labrum. a Transverse oblique US image in a 1-year- (open arrow). b Transverse oblique US image of the normal contralateral
old girl with right-side glenohumeral dysplasia. Image of the right glenohumeral joint in the same girl is provided for comparison. Note both
shoulder shows a rounded configuration of the posterior glenoid margin the sharp margin of the posterior glenoid and the steep slope of the normal
(solid arrow), a shallow visualized glenoid fossa (arrowhead) and an glenoid fossa
irregularly contoured posterior labrum with heterogeneous echogenicity
Pediatr Radiol (2022) 52:1648–1657 1653

Fig. 7 Shoulder muscles. Serial transverse oblique US images in an right side at 7 months of age demonstrates an interval increase in
infant girl with right-side glenohumeral dysplasia. a, b US images at echogenicity and decrease in size of the right infraspinatus muscle,
2 months of age show that although there is right brachial plexus palsy, consistent with fatty atrophy. d Axial dual-echo steady-state (DESS)
the right infraspinatus muscle (a) appears fairly symmetrical or even MRI sequence of bilateral shoulders at 10 months of age confirms the
larger than the normal left side (b). c An additional US image of the atrophy of the right infraspinatus muscle

in external rotation, the greater tuberosity, which is humeral head by the glenoid, but posterior humeral head dis-
posterolaterally located along the proximal humerus, comes placement is recorded as a percentage rather than an angle. To
into view. This can mimic the posterior margin of the humeral make this measurement, the posterior scapular line is ex-
head and lead to overestimating the alpha angle (Fig. 8). tended laterally through the entirety of the humeral head.
A perpendicular line is then drawn from the scapular line
Posterior humeral head displacement to the posterior margin of the humeral head. That value is
divided by the greatest diameter of the humeral head, then
Posterior humeral head displacement can also be assessed by multiplied by 100, resulting in the percentage of the hu-
sonography [28]. This calculation is similar to humeral head meral head that lies posterior to the scapular line. A nor-
alpha angle in that it is a metric for posterior uncovering of the mal value is considered less than 50% [28]. Posterior

Fig. 8 Alpha angle. a, b Transverse oblique US images of bilateral posterior margin of the humeral head is appropriately rounded and
glenohumeral joints in a 3-month-old boy with left glenohumeral smooth, suggesting that the infant’s shoulder is internally rotated rather
dysplasia. The normal alpha angle of 25° in the right shoulder (a) is than externally rotated, which might bring the irregular greater tuberosity
obtained from the intersection of the posterior scapular line and the line into view. The alpha angle on the affected left side (b) is 50°, indicating
tangential to the posterior margin of the humeral head. Note that the posterior subluxation
1654 Pediatr Radiol (2022) 52:1648–1657

humeral head displacement is analogous to femoral head Magnetic resonance imaging

coverage in the hip; however, this percentage refers to the
percentage of the uncovered humeral head rather than the When presenting early with a significant brachial plexus
covered percentage that is typically reported for hips in palsy, or when conservative treatment has failed and the
the setting of suspected dysplasia. Importantly, measure- older infant presents with persistent shoulder dysfunction,
ment of posterior humeral head displacement outside the surgical treatment might be required to optimize upper
neonatal period can be inaccurate because the humeral extremity function and minimize further deformity. MRI
ossification center grows rapidly in early life, resulting should be performed in such cases of advanced dysplasia,
in acoustic shadowing that limits visualization of the an- either in conjunction with or in lieu of US examination.
terior margin of the humeral head. Thus, the humeral head Both shoulders are imaged concurrently, allowing the un-
diameter might not be accurately measured, resulting in a affected side to act as an internal normal control, with an
falsely exaggerated or underestimated posterior humeral emphasis on a three-dimensional (3-D) gradient-recalled
head displacement measurement (Fig. 9). echo (GRE) cartilage-sensitive sequence (i.e. dual-echo
steady-state [DESS]; Siemens, Erlangen, Germany)
(Fig. 10). This sequence offers qualitative assessment of
Glenoid version (glenoscapular angle) glenoid, labrum, humeral head and muscle morphology.
Quantitative measurement including alpha angle, glenoid
While glenoid version measurements have been performed on version and humeral head coverage can be readily per-
US, they have been shown to have relatively poor interrater formed on this sequence, as well.
reliability when compared to MRI and are not routinely per- T1-weighted imaging is useful for marrow signal as well
formed at our institution [25]. Glenoid version is measured by as asymmetry of the shoulder soft tissues, as seen with
extending the posterior scapular line laterally through the muscle atrophy. These findings, in aggregate, allow for
glenoid, then drawing a line tangential to the surface of the imaging grading of the glenohumeral deformity according
glenoid. The angle created by these two lines as measured in to the Waters classification, a system that is useful for
the posteromedial quadrant is then subtracted from 90; the clinical and surgical planning [10]. Finally, the cervical
final number reflects the glenoid version. A negative value spine and brachial plexus nerve roots can be imaged simul-
indicates glenoid retroversion, a normal finding in children taneously. This is facilitated with 3-D steady-state gradient
younger than 8 years. Normal glenoid version angles range echo sequences (e.g., constructive interference in steady
from −23° to 8° (mean −4.2° ± 6.2°), though generally most state, or CISS), which allow for high contrast resolution
glenoid fossae are retroverted before the age of 2 years and between cerebrospinal fluid and soft-tissue structures, with
typically feature less than 10° of retroversion [33]. In more submillimeter isovolumetric voxels to resolve nerve roots.
advanced or complex cases of dysplasia where there is sub- Such high-resolution cervical spine imaging provides ad-
stantial glenoid retroversion or a biconcave glenoid and the ditional information about the levels of nerve root injury
likelihood for surgery is high, MRI is more frequently per- and the presence or absence of pseudomeningoceles [4]
formed for glenohumeral assessment. (Fig. 11).

Fig. 9 Posterior humeral head displacement. a, b Transverse oblique US humeral head diameter (dotted line) and multiplied by 100 to obtain a
images in a 4-month-old girl with shoulder contractures. The posterior percentage. The normal right side (a) has a posterior humeral head
humeral head displacement is obtained by drawing a line parallel to the displacement percentage less than 50%, a normal value, while the
scapular spine and measuring perpendicularly to the posterior margin of affected left-side (b) posterior humeral head displacement percentage is
the humeral head (solid line). This value is divided by the maximal greater than 50%, considered abnormal
Pediatr Radiol (2022) 52:1648–1657 1655

Fig. 10 MRI in advanced dysplasia. Axial dual-echo steady-state (DESS) posterior humeral head subluxation. The right glenoid (solid arrow) is
image in a 3-year-old boy with left glenohumeral dysplasia demonstrates normal; note that the normal glenoid fossa is perpendicular to the long
left glenoid retroversion (posteromedial sloping; dotted arrow) with axis of the scapula

Surgical treatment injured, considerations are made for nerve transfers rather than
reconstruction. With this technique, the site of the nerve injury
When surgery is required, the persistent weakness pattern, is bypassed, the existing nerve connection to the muscle is not
Narakas classification and MRI findings are considered for disrupted, and the nerve supply to the muscle is augmented by
treatment planning. Infants with more than upper trunk in- coaptation of a healthy adjacent nerve to the target peripheral
volvement, or in whom there is suspicion of nerve root avul- nerve [35]. This technique has been shown to have similar
sions or ruptures, should undergo nerve reconstruction. Often results to nerve reconstruction with the benefit of less opera-
with the surgical dissection, a neuroma involving one or more tive time and no lower extremity scar from sural nerve har-
nerve roots at the trunk level of the brachial plexus is found. vesting [35]. Outcomes literature for rebalancing the shoulder
This is removed and replaced by sural nerve grafts that allow muscles after nerve surgery reported that about 60% of pa-
for axon regeneration and reorganization to achieve sufficient tients have functional shoulder motion and do not require
nerve supply to the affected muscles [34]. If C5–6 are solely secondary tendon transfer surgery [36].

Fig. 11 MRI for

pseudomeningoceles. a, b
Coronal (a) and sagittal (b)
constructive interference steady
state (CISS) MR images in a 4-
month-old boy with left brachial
plexus palsy demonstrate left-side
pseudomeningoceles at the C6–7
and C7–T1 levels (arrows)
1656 Pediatr Radiol (2022) 52:1648–1657

Conclusion 15. Moukoko D, Ezaki M, Wilkes D, Carter P (2004) Posterior shoul-

der dislocation in infants with neonatal brachial plexus palsy. J
Bone Joint Surg Am 86:787–793
Brachial plexus birth injury can result in permanent nerve 16. Pearl ML (2009) Shoulder problems in children with brachial plex-
damage and resultant upper extremity dysfunction, and there- us birth palsy: evaluation and management. J Am Acad Orthop
fore these injuries require early and regular clinical and imag- Surg 17:242–254
ing assessment during infancy. US is an excellent tool for 17. Bigliani LU, Kelkar R, Flatow EL et al (1996) Glenohumeral sta-
bility. Biomechanical properties of passive and active stabilizers.
screening and subsequent follow-up of children who suffer Clin Orthop Relat Res Sep 1996:13–30
clinically evident nerve dysfunction, and it provides critical 18. Kotlarsky P, Haber R, Bialik V, Eidelman M (2015) Developmental
qualitative and quantitative data to guide surgical and non- dysplasia of the hip: what has changed in the last 20 years? World J
surgical management. Orthop 6:886–901
19. Iorio ML, Menashe SJ, Iyer RS et al (2015) Glenohumeral
dysplasia following neonatal brachial plexus palsy: presenta-
tion and predictive features during infancy. J Hand Surg 40:
Declarations 2345–2351.e1
20. Curtis C, Stephens D, Clarke HM, Andrews D (2002) The active
Conflicts of interest None movement scale: an evaluative tool for infants with obstetrical bra-
chial plexus palsy. J Hand Surg 27:470–478
21. Borschel GH, Clarke HM (2009) Obstetrical brachial plexus palsy.
Plast Reconstr Surg 124:144e–155e
References 22. Justice D, Rasmussen L, Di Pietro M et al (2015) Prevalence of
posterior shoulder subluxation in children with neonatal brachial
1. Lalka A, Gralla J, Sibbel SE (2020) Brachial plexus birth injury: plexus palsy after early full passive range of motion exercises.
epidemiology and birth weight impact on risk factors. J Pediatr PM R 7:1235–1242
Orthop 40:e460–e465 23. Ezaki M, Malungpaishrope K, Harrison RJ et al (2010)
2. Foad SL, Mehlman CT, Ying J (2008) The epidemiology of neo- Onabotulinum toxin a injection as an adjunct in the treatment of
natal brachial plexus palsy in the United States. J Bone Joint Surg posterior shoulder subluxation in neonatal brachial plexus palsy. J
Am 90:1258–1264 Bone Joint Surg Am 92:2171–2177
3. Van der Looven R, Le Roy L, Tanghe E et al (2020) Risk factors for 24. Verchere C, Durlacher K, Bellows D et al (2014) An early shoulder
neonatal brachial plexus palsy: a systematic review and meta-anal- repositioning program in birth-related brachial plexus injury: a pilot
ysis. Dev Med Child Neurol 62:673–683 study of the Sup-ER protocol. Hand 9:187–195
4. Menashe SJ, Tse R, Nixon JN et al (2015) Brachial plexus birth 25. Donohue KW, Little KJ, Gaughan JP et al (2017)
palsy: multimodality imaging of spine and shoulder abnormalities Comparison of ultrasound and MRI for the diagnosis of
in children. AJR Am J Roentgenol 204:W199–W206 glenohumeral dysplasia in brachial plexus birth palsy. J
5. Waters PM (1997) Obstetric brachial plexus injuries: evaluation Bone Joint Surg Am 99:123–132
and management. J Am Acad Orthop Surg 5:205–214 26. Yu CK, Yuen VMY, Wong GT, Irwin MG (2013) The effects of
6. van der Sluijs JA, van Ouwerkerk WJR, Manoliu RA, Wuisman anaesthesia on the developing brain: a summary of the clinical
PIJM (2004) Secondary deformities of the shoulder in infants with evidence. F1000Res 2:166
an obstetrical brachial plexus lesions considered for neurosurgical 27. Hunter JD, Franklin K, Hughes PM (1998) The ultrasound diagno-
treatment. Neurosurg Focus 16:E9 sis of posterior shoulder dislocation associated with Erb’s palsy.
7. Malessy MJA, Pondaag W (2009) Obstetric brachial plexus inju- Pediatr Radiol 28:510–511
ries. Neurosurg Clin N Am 20:1–14 28. Vathana T, Rust S, Mills J et al (2007) Intraobserver and interob-
8. Al-Qattan MM, El-Sayed AAF, Al-Zahrani AY et al (2009) server reliability of two ultrasound measures of humeral head posi-
Narakas classification of obstetric brachial plexus palsy revisited. tion in infants with neonatal brachial plexus palsy. J Bone Joint
J Hand Surg Eur Vol 34:788–791 Surg Am 89:1710–1715
9. Nikolaou S, Peterson E, Kim A et al (2011) Impaired growth of 29. Birch R (1998) Birth lesions of the brachial plexus. In: Birch R,
denervated muscle contributes to contracture formation following Bonney G, Wynn Parry C (eds) Surgical disorders of the peripheral
neonatal brachial plexus injury. J Bone Joint Surg Am 93:461–470 nerves. Churchill Livingstone, London, pp 209–234
10. Waters PM, Smith GR, Jaramillo D (1998) Glenohumeral deformi- 30. Pons C, Sheehan FT, Im HS et al (2017) Shoulder muscle atrophy
ty secondary to brachial plexus birth palsy. J Bone Joint Surg Am and its relation to strength loss in obstetrical brachial plexus palsy.
80:668–677 Clin Biomech 48:80–87
11. Pöyhiä TH, Nietosvaara YA, Remes VM et al (2005) MRI of rota- 31. van Gelein Vitringa VM, van Kooten EO, Mullender MG et al
tor cuff muscle atrophy in relation to glenohumeral joint incongru- (2009) An MRI study on the relations between muscle atrophy,
ence in brachial plexus birth injury. Pediatr Radiol 35:402–409 shoulder function and glenohumeral deformity in shoulders of chil-
12. Pöyhiä TH, Lamminen AE, Peltonen JI et al (2010) Brachial plexus dren with obstetric brachial plexus injury. J Brachial Plex Peripher
birth injury: US screening for glenohumeral joint instability. Nerve Inj 4:5
Radiology 254:253–260 32. Sofka CM, Haddad ZK, Adler RS (2004) Detection of muscle at-
13. Ruchelsman DE, Grossman JAI, Price AE (2011) Glenohumeral rophy on routine sonography of the shoulder. J Ultrasound Med 23:
deformity in children with brachial plexus birth injuries. Bull 1031–1034
NYU Hosp Jt Dis 69:36–43 33. Mintzer CM, Waters PM, Brown DJ (1996) Glenoid version in
14. Olofsson PN, Chu A, McGrath AM (2019) The pathogenesis of children. J Pediatr Orthop 16:563–566
glenohumeral deformity and contracture formation in obstetric bra- 34. Lin JC, Schwentker-Colizza A, Curtis CG, Clarke HM (2009) Final
chial plexus palsy — a review. J Brachial Plex Peripher Nerve Inj results of grafting versus neurolysis in obstetrical brachial plexus
14:e24–e34 palsy. Plast Reconstr Surg 123:939–948
Pediatr Radiol (2022) 52:1648–1657 1657

35. O’Grady KM, Power HA, Olson JL et al (2017) Comparing the Publisher’s note Springer Nature remains neutral with regard to jurisdic-
efficacy of triple nerve transfers with nerve graft reconstruction in tional claims in published maps and institutional affiliations.
upper trunk obstetric brachial plexus injury. Plast Reconstr Surg
140:747–756
36. Segal D, Cornwall R, Little KJ (2019) Outcomes of spinal
accessory-to-suprascapular nerve transfers for brachial plexus birth
injury. J Hand Surg 44:578–587