DOI:10.2214/AJR.05.2159
AJR 2007; 188:733-738
© American Roentgen Ray Society
Sonography of the Neonatal Spine: Part 1, Normal Anatomy, Imaging Pitfalls, and Variations That May Simulate Disorders
Lisa H. Lowe1,2,
Andrew J. Johanek1,3 and
Charlotte W. Moore1,2
1 Department of Radiology, The University of Missouri-Kansas City, Kansas City,
MO.
2 Department of Radiology, Children's Mercy Hospital and Clinics, 2401 Gillham
Rd., Kansas City, MO 64108.
3 Department of Radiology, St. Luke's Hospital, Kansas City, MO.
Received December 16, 2005;
accepted after revision February 28, 2006.
Address correspondence to L. H. Howe
(lhlowe{at}cmh.edu).
Awarded a Bronze Medal poster exhibit at the 2005 annual meeting of the
American Roentgen Ray Society, New Orleans, LA.
CME
This article is available for CME credit. See
www.arrs.org
for more information.
FOR YOUR INFORMATION
The reader's attention is directed to part 2 accompanying this article,
titled "Sonography of the Neonatal Spine: Part 2, Spinal
Disorders," which begins on
page 739.
FOR YOUR INFORMATION
This article is available for CME credit. See
www.arrs.org
for more information.
Abstract
OBJECTIVE. Our objective is to discuss neonatal spine sonography
with emphasis on imaging pitfalls and normal variants that may simulate
disease and to distinguish them from true spinal disorders.
CONCLUSION. Sonography of the neonatal spine is now accepted as a
highly sensitive, readily available screening study that can be used to
evaluate various anomalies of the lumbar spine in most infants younger than 4
months.
Keywords: neonatal imaging • neuroradiology • pediatric radiology • sonography • spine
Introduction
Although MRI has been considered the imaging gold standard, recent
advances in sonography have allowed its image quality to improve significantly
enough that its diagnostic value is equal to that of MRI
[1]. Sonography can now
characterize nearly all spinal anomalies sufficiently in the first days of
life. This allows the clinical determination of whether the lesion requires
urgent intervention or whether further radiologic evaluation with studies such
as MRI can be delayed until therapeutic intervention is more imminent. In part
1 of this pictorial essay, we discuss lumbar spine embryology, sonography
techniques and indications, normal anatomy, and developmental variations and
pitfalls that may simulate disease. Part 2 covers abnormal entities.
Embryology
To understand spine anomalies, knowledge of embryonic development is
necessary. The CNS starts to form during the third gestational week, beginning
with the process known as neurulation (Fig.
1A). Next, canalization occurs at the distal end of the neural
tube in the caudal cell mass, resulting in an ependyma-lined neural tube that
unites with the rest of the spinal cord to form the conus medullaris and
ventriculus terminalis (Fig.
1B). Finally, at 38 days of gestation, retrogressive
differentiation occurs (Fig.
1C), forming the filum terminale
[2-5].
View larger version (21K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A —Schematics illustrate three stages of spinal cord development.
Neurulation (closure of neural tube) is process of progression from neural
plate to neural groove to neural tube. (Reprinted with permission from Sadler
T. Langman's medical embryology, 5th ed. Baltimore, MD: Lippincott
Williams & Wilkins, 1985:335
[5])
|
|
View larger version (15K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B —Schematics illustrate three stages of spinal cord development.
Canalization occurs when multiple microcysts form and coalesce in caudal cell
mass (arrows), which fuses to distal neural tube
(arrowheads), forming primitive spinal cord. (Reprinted with
permission from Barkovich AJ. Normal development of the neonatal and infant
brain, skull and spine. In: Barkovich AJ. Pediatric neuroimaging, 3rd
ed. Philadelphia, PA: Lippincott Williams & Wilkins, 2000:624
[2])
|
|
View larger version (16K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1C —Schematics illustrate three stages of spinal cord development.
Retrogressive differentiation (programmed cell death) is process whereby
caudal cell mass and neural tube regress in size to form fetal conus
medullaris, ventriculus terminalis, and filum terminale. Note labeled
structures. (Reprinted with permission from Barkovich AJ. Normal development
of the neonatal and infant brain, skull and spine. In: Barkovich AJ.
Pediatric neuroimaging, 3rd ed. Philadelphia, PA: Lippincott Williams
& Wilkins, 2000:624
[2])
|
|
Technique and Indications
Images are obtained in the longitudinal and transverse planes using a
linear 5-12-MHz transducer (Fig.
2A,
2B). The vertebral level is
determined by counting down from the 12th rib and confirmed by counting up
from the L5-S1 junction or the tip of coccyx. If the vertebral level is
unclear, correlation with radiographs (possibly with a marker) may help. Color
or power Doppler sonography may be used as an adjunct to better characterize
soft-tissue masses found on the skin or in the spinal canal.
View larger version (111K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A —1-week-old boy with normal lumbar spine sonogram and history of
unilateral renal agenesis. Transverse lumbar sonogram shows normal anatomy as
labeled. V = vertebra, transverse process (arrowhead).
|
|
View larger version (47K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B —1-week-old boy with normal lumbar spine sonogram and history of
unilateral renal agenesis. Longitudinal lumbar sonogram shows normal anatomy
as labeled. Note central echoic complex (arrowheads), a normal
finding that results from interface of central end of anterior median fissure
and not central spinal canal.
|
|
Indications for lumbar spine sonography include multiple congenital
anomalies placing an infant at increased risk, complicated sacral dimple
(location above the gluteal crease, bottom of pit not seen, possible drainage
from dimple, and presence of skin stigmata), softtissue mass suspected of
being spina bifida occulta, determination of reason for failed lumbar
puncture, and location of CSF that may be tapped
[1,
6]. Low-risk lesions include
simple midline dimples (< 5 mm in diameter, within 2.5 cm of the anus, no
other cutaneous stigmata). High-risk lesions include atypical dimples (> 5
mm in diameter, > 2.5 cm above the anus), hemangiomas, cutis aplasia, hairy
patches, and skin tags [7]. MRI
is the study of choice when surgical therapy is required, such as with open
spinal dysraphism or obvious CSF drainage from a skin dimple or sinus
tract.
Normal Variants That May Simulate Disorders
Several common normal variants that may be confused with disorders on
lumbar spine sonography will be discussed, including ventriculus terminalis,
filar cyst, prominent filum terminale, cauda equina pseudomass, pseudosinus
tract, and dysmorphic coccyx. Familiarity with these variations can prevent
misinterpretation and referrals for unneeded additional clinical or imaging
evaluation.
Ventriculus Terminalis
The ventriculus terminalis, often seen on sonography and MRI in children
younger than 5 years, is due to incomplete fetal regression of the embryonic
terminal ventricle in the conus medullaris
[6,
8] (Fig.
3A,
3B).
View larger version (122K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A —1-month-old boy with ventriculus terminalis who was referred for
deep sacral dimple and who is developmentally normal at 18 months.
Longitudinal sonogram of spine reveals distention of distal lumbar spinal
canal just above conus medullaris (arrowhead). Size smaller than 5 mm
and stability over time distinguish this normal variant from small syrinx.
|
|
View larger version (132K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B —1-month-old boy with ventriculus terminalis who was referred for
deep sacral dimple and who is developmentally normal at 18 months. Sagittal
T2-weighted MR image at age 7 months shows stable distention of distal spinal
canal (arrowhead), excluding syrinx.
|
|
Filar Cyst
The so-called filar cyst is an interesting incidental finding that has only
recently been studied, perhaps at least partly because of being detected more
often with improved sonographic equipment. Although this lesion has been
scantly described in the radiology literature, it is often visible on
sonography. There is no autopsy description of a filar cyst, which begs the
questions of its origin and its validity as an entity
[3,
4]. In addition, the
nomenclature for this lesion is confusing in that it has been termed both
"ventriculus terminalis" and "filar cyst" by various
authors
[6-8].
We prefer the latter term, filar cyst, to specify the filar location from the
conus medullaris location of the ventriculus terminalis.
Possible explanations for the origin of the filar cyst include that perhaps
the normal arachnoid reflections form a pseudocystlike structure or that it is
a true ependyma-lined cystic embryonic remnant (possibly indistinguishable
from the ventriculus terminalis) that is disrupted by the act of opening the
dura during autopsy. Regardless of its origin, it is a normal variant that
alone has no known clinical significance and that does not require additional
imaging [4]. If MRI is
performed, in our experience, the filar cyst is less reliably visible than on
sonography. Strict imaging criteria for filar cysts should be applied
(location midline, within filum, just below conus; fusiform shape,
well-defined, hypoechoic appearance of a simple cyst) to avoid the potential
for underdiagnosing a true disorder (Figs.
4A,
4B and
5A,
5B).
View larger version (112K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4A —Filar cyst in 14-day-old girl with deep sacral dimple and normal
motor development. Transverse sonogram of proximal cauda equina shows
well-defined, midline, cystic collection (arrow). Note normal ventral
and dorsal nerve root bundles (arrowheads).
|
|
View larger version (99K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4B —Filar cyst in 14-day-old girl with deep sacral dimple and normal
motor development. Longitudinal sonogram reveals well-defined fusiform
"cyst" in midline (arrow) just below conus medullaris.
Also note prominent echogenic central spinal canal (arrowhead), a
normal variant seen in some children.
|
|
View larger version (89K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5A —Filar cyst in 5-week-old boy with multiple anomalies who had been
followed up with MRI at age 2 months. Longitudinal sonogram of filum and cauda
equina (arrowhead) shows unusually long filar cyst
(calipers). Despite its length, it meets criteria for filar cyst:
location just below conus medullaris, fusiform shape, well defined, thin
walled, and hypoechoic.
|
|
View larger version (86K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 5B —Filar cyst in 5-week-old boy with multiple anomalies who had been
followed up with MRI at age 2 months. Longitudinal T2-weighted MR image shows
ill-defined filar cyst (arrows) that is better seen on
sonography.
|
|
Prominent Filum Terminale
A prominent filum terminale may cause concern when it stands out as
particularly echogenic in comparison with other nerve roots. It is
distinguished as normal by its thickness and typical midline course
[1]
(Fig. 6).
"Pseudomass" due to Positional Nerve Root Clumping
Positional clumping of the nerve roots occurs when an infant is scanned in
the decubitus position. Rescanning the child prone will cause the
"mass" to disappear as the nerve roots return to their normal
position (Fig. 7A,
7B).
View larger version (113K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7A —Positional pseudomass in 2-week-old boy with left renal agenesis who
was scanned in left decubitus position. Transverse sonogram shows clumping of
nerve roots (arrows) on left due to left decubitus position.
|
|
View larger version (94K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 7B —Positional pseudomass in 2-week-old boy with left renal agenesis who
was scanned in left decubitus position. Longitudinal sonogram also reveals
masslike appearance of nerve roots (arrows). Prone images (not shown)
were normal.
|
|
Pseudosinus Tract
Another common normal variant is a pseudosinus tract, which is seen on
sonography as a residual cordlike region composed of fibrous tissue extending
from a skin dimple to the coccyx (Fig.
8). True dermal sinus tracts rarely occur at the tip of the coccyx
and are typically found in a more cranial location. However, a careful search
should be made for any mass or fluid along the course of the fibrous tract. If
CSF is draining via a dimple, then a true sinus tract is likely, and MRI is
the imaging technique of choice.
View larger version (128K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 8 —Pseudosinus tract in 12-day-old infant with dimple in gluteal
crease. Longitudinal sonogram shows cartilaginous, hypoechoic, dorsally
curving tip of coccyx (arrowhead), from which hypoechoic cordlike
structure (curved arrow) extends caudally and terminates at base of
skin dimple (straight arrow).
|
|
Dysmorphic Coccyx
The tip of the coccyx can vary widely in shape, and in some cases may mimic
a mass when palpated on physical examination (Fig.
9A,
9B).
View larger version (132K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 9A —Misshapen coccyx in two neonatal girls, each with palpable
"lump" beneath sacral dimple in gluteal crease. Longitudinal
sonogram of coccyx in 2-week-old girl shows hypoechoic cartilaginous tip
(arrowheads), which is acutely angulated dorsally as it extends
toward skin surface. Palpated "lump" was tip of coccyx.
|
|
View larger version (141K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 9B —Misshapen coccyx in two neonatal girls, each with palpable
"lump" beneath sacral dimple in gluteal crease. Longitudinal
sonogram of coccyx in 2-week-old girl reveals it is straightened, with loss of
its normal ventral curve. Hypoechoic cartilaginous tip (arrowhead)
extends dorsally toward skin surface, causing clinically palpable
"lump."
|
|
Conclusion
Neonatal spinal sonography is a useful screening technique for occult
spinal anomalies; it can characterize normal anatomy and normal variants that
may simulate disorders. Familiarity with these findings will prevent
misinterpretation and inappropriate referrals.
References
- Dick EA, Patel K, Owens CM, De Bruyn R. Spinal ultrasound in
infants. Br J Radiol 2002;75
: 384-392[Abstract/Free Full Text]
- Barkovich AJ. Normal development of the neonatal and infant brain,
skull, and spine. In: Barkovich AJ. Pediatric
neuroimaging, 3rd ed. Philadelphia, PA: Lippincott Williams &
Wilkins, 2000: 50-54
- Gamble HJ. Electron microscope observations upon the conus
medullaris and filum terminale of human fetuses. J
Anat 1971; 110:173
-179[Medline]
- Malas MA, Salbacak A, Buyukmumcu M, Seker M, Koyluoglu B, Karabulut
AK. An investigation of the conus medullaris termination level during the
period of fetal development to adulthood. Kaibogaku
Zasshi 2001; 76:453
-459[Medline]
- Sadler T. Langman's medical embryology, 5th
ed. Baltimore, MD: Lippincott Williams & Wilkins, 1985:334
-337, 343-345
- Unsinn KM, Geley T, Freund MC, Gassner I. US of the spinal cord in
newborns: spectrum of normal findings, variants, congenital anomalies, and
acquired diseases. RadioGraphics 2000;20
: 923-938[Abstract/Free Full Text]
- Kriss VM, Desai NS. Occult spinal dysraphism in neonates:
assessment of high-risk cutaneous stigmata on sonography.
AJR 1998; 171:1687
-1692[Abstract/Free Full Text]
- Coleman LT, Zimmerman RA, Rorke LB. Ventriculus terminalis of the
conus medullaris: MR findings in children. Am J
Neuroradiol 1995; 16:1421
-1426[Abstract]
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
Related articles in AJR:
- Sonography of the Neonatal Spine: Part 2, Spinal Disorders
- Lisa H. Lowe, Andrew J. Johanek, and Charlotte W. Moore
AJR 2007 188: 739-744.
[Abstract]
[Full Text]
Top
Abstract
Introduction
