European Journal of Radiology 82 (2013) 127–138

Contents lists available at SciVerse ScienceDirect

European Journal of Radiology

journal homepage: www.elsevier.com/locate/ejrad

Review

Biomechanics of the spine. Part II: Spinal instability

Roberto Izzo a,∗ , Gianluigi Guarnieri a , Giuseppe Guglielmi b , Mario Muto a
a
Neuroradiology Department, “A.Cardarelli” Hospital, Napoli, Italy
b
Department of Radiology, University of Foggia, Foggia, Italy

a r t i c l e i n f o a b s t r a c t

Article history: Spine stability is the basic requirement to protect nervous structures and prevent the early deterioration
Received 3 July 2012 of spinal components. All bony and soft spinal components contribute to stability, so any degenerative,
Received in revised form 20 July 2012 traumatic or destructive lesion to any spinal structure gives rise to some degree of instability.
Accepted 21 July 2012
Degenerative instability is considered a major cause of axial and radicular pain and is a frequent indica-
tion for surgery. Nevertheless the assessment of instability remains difficult in both clinical and imaging
Keywords:
settings.
Spine
All static imaging modalities, even conventional MR, the most accurate technique, are unreliable in
Biomechanics
Spinal instability
assessing instability and chronic pain due to degenerative spine.
Spinal degeneration Dynamic-positional MR is considered the most sophisticated imaging modality to evaluate abnormal
Spinal trauma spinal motion and instability.
Spinal tumours In spinal traumas, as multi-detector CT yields high-resolution reconstructions in every spatial plane, it
CT, MR will detect even the tiniest fractures revealing potentially unstable lesions, often avoid the routine use
of MR. Nevertheless, MR remains the only modality that will directly and routinely assess soft tissue
changes. Unfortunately the objectivity of MR in assessing the integrity of ligaments is not rigorously
defined and its use in routine protocols to clear blunt spinal injuries remains controversial.
There are no evidence-based guidelines currently available to assess the risk of spinal instability in the
setting of neoplastic spinal disease, so predicting the risk of a pathological fracture or the timing of a
collapse remains challenging even when the lesions are well-characterized by neuroimaging.
Diagnostic difficulties lead to controversy in the choice of the best treatment in all forms of spinal
instability.
© 2012 Elsevier Ireland Ltd. All rights reserved.

1. Introduction
Traumatic, neoplastic and degenerative instability are impor-
tant cause of spinal pain and disability.
Degenerative instability, in particular, is considered a major
cause of axial and radicular acute and chronic pain and is a fre-
quent indication for surgery. Nevertheless, an accurate definition
of instability and the best diagnostic approach remain matters of
debate and have given rise to controversy in the choice of the best
treatment.
Even though MR is the most accurate imaging modality to
study the degenerative spine, conventional recumbent imaging
is often not reliable in assessing instability and the source of
acute and chronic pain [1]. Positional-upright MR is the most
∗ Corresponding author at: Neuroradiology Department, “A.Cardarelli” Hospital,
Viale Cardarelli 9, 80131 Napoli, Italy. Tel.: +39 0817473116.
E-mail addresses: roberto1766@interfree.it (R. Izzo),
gianluigiguarnieri@hotmail.it (G. Guarnieri), g.gugliemi@unifg.it (G. Guglielmi),
mutomar@tiscali.it (M. Muto).
sophisticated imaging tool to assess instability but the high cost
involved hampers its widespread availability so that conventional
dynamic radiographs remain the simplest and most widely used
diagnostic reference [2].
Traumatic spinal injuries affect a complex structure consist-
ing of soft and bony components having different susceptibility
and healing potential: this complexity contributes to difficulties
in classifying traumas and in assessing instability and the efficacy
of various treatments.
Conventional radiology is inadequate to assess the stabil-
ity of fractures [3]. As multi-detector CT yields high-resolution
reconstructions in every spatial plane, it will detect even the
tiniest fractures revealing potentially unstable lesions. While
conventional radiology and CT can only indirectly evalu-
ate lesions affecting discs and ligaments and are of limited
value for prognosis and therapy, MR is the only imaging
modality that directly and routinely assesses changes in liga-
ments.
Nevertheless, the objectivity of MR in assessing the integrity of
ligaments and predicting mechanical instability remains contro-
versial [4,5].

0720-048X/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ejrad.2012.07.023
128 R. Izzo et al. / European Journal of Radiology 82 (2013) 127–138

Even more in the setting of spinal tumours the definition and

management of instability are still debated as the restoration or
maintenance of spinal stability is an important objective in the
surgical or conservative management of patients with neoplastic
spinal disease [6].
The biomechanical effect of spinal tumours and the mecha-
nisms of vertebral neoplastic collapse are not fully understood
[7]. A comprehensive classification system to define neoplastic
spinal instability has recently been proposed to guide the decision-
making process [8].
This paper is an overview of the current concepts on the spinal
stability, focusing on the basis of degenerative, traumatic and onco-
logic instability and the how the different imaging modalities can
contribute to their assessment.

2. Degenerative instability

Spinal degeneration represents a major public health problem

and an enormous economic burden on society. Degenerative insta-
bility is also a common cause of axial and radicular pain and a
frequent indication for surgery. Degenerative instability has been
defined as a change in vector forces in the relations between the
functional units of the spine (FSU), generating abnormal, imbal-
anced, paradoxical movements.
A degenerative primum movens, generally a disc derangement,
triggers disorders of movement which, in turn, increase the original
bony, articular and biomechanical abnormalities and extend them Fig. 1. Parasagittal FSE T2-weighted (TR/TE = 2200/120 ms) MR image showing
first centrifugally to other joints of the same level (three-articular extended Modic-type I changes involving the endplates and subchondral areas of
L2 and L3 vertebral bodies. When extensive, Modic I changes are often associated
complex) and then to those of adjacent segments transforming a
with segmental instability and painful disc.
segmental into a regional pathology.
The degenerating disc undergoes progressive biochemical and
structural changes that impair the biomechanical properties. While
the disc becomes increasingly fibrous, the interspace progressively
and its site, as well as aggravating activities and movements in
collapses [9]. Owing to disc collapse the annulus and ligaments
specific directions and space quadrants.
become lax and redundant favouring canal stenosis, anterior, pos-
Several microinstability patterns have been described: in flex-
terior and vertical subluxation of the vertebral bodies and raising
ion, extension, lateral bending and multidirectional types [13].
of the superior subjacent facets towards the superior foramina up
Unfortunately, the directional nature of instability, well-defined
to eventual neoarthrosis with the peduncles [10]. Neoarthrosis of
for other joints, remains poorly defined in the spine because the
the raised facets with the peduncles promotes facet remodelling
motion is three-dimensional and includes coupled movements. The
and osteophytosis eventually leading to further foraminal stenosis
site and type of injury within the motion segment again influences
[10]. This creates a vicious circle by which the disease is cyclically
the pattern of instability, but movement dysfunction tends to occur
self-sustaining (self-induced degenerative disease).
in multiple directions [13].
Instability is not an all-or-nothing phenomenon, but is always
Spondylolisthesis is considered a special type of instability
present in different degrees and forms in degenerative disease, con-
because the degree of forward displacement can progress over
ditioning its symptoms and evolution. Kirkaldy-Willis and Farfan
time [14]. While forward subluxation is primarily a pathology of
[11] distinguish three main biomechanical and clinical phases in the
the posterior joints and is most common at L4–L5, retrolisthesis is
evolution of degenerative instability forming a so-called “degener-
primarily a disorder of the disc space and prevails at L3–L4.
ative cascade”, without a clear-cut division:
Disc degeneration can also lead to segmental instability and ver-
tebral shift in the coronal plane. Lateral listhesis and angulation can
2.1. Dysfunction, instability, and restabilization be associated with lateral wedging of the vertebral body and asym-
metric degeneration of the facet joints resulting in degenerative
The first phase of dysfunction is marked by the appearance of scoliosis.
intermittent nonspecific low back pain coinciding with the ini- Acquired isthmic spondylolisthesis can also occur as a com-
tial changes in the discs and facet joints. During the instability plication of degenerating processes, brought on by the opposite
phase, disc degeneration is of intermediate degree and disc space colliding action of the vertically slipped facets upon an interposed
narrowing leads to laxity of the annulus, capsules and ligaments isthmus abutting a stress fracture [10]. Isthmic defects compromise
resulting in abnormal movement of the motion segment up to the ability of the posterior elements to stabilize the FSU, generating
anterolisthesis or retrolisthesis. At this stage pain becomes more instability.
persistent. At lumbar level segmental instability may be purely Several imaging findings can be found in a unstable spine includ-
a movement syndrome, with no apparent bony lesions (microin- ing endplate oedema (Modic-type 1 changes) (Fig. 1), peduncle and
stability) in which directional patterns of motion develop with isthmus oedema (Fig. 2) [10], traction spurs (Fig. 3) [15], extended
impaired muscle control generating symptoms, or may be asso- discal vacuum associated with mild disc space narrowing, joint
ciated with antero-retrolisthesis [12]. effusion or vacuum along with facet joint gapping over 1 mm [16],
The syndromes of microinstability have been defined on the synovial cysts (Fig. 4), annular tears, anterolisthesis, and retrolis-
basis of clinical observations, the supposed mechanism of injury thesis (Fig. 5) [11].
 R. Izzo et al. / European Journal of Radiology 82 (2013) 127–138
Fig. 2. Parasagittal FSE STIR (TR/TE/TI = 2730/120/150 ms) MR image showing type
I hyperintense oedematous vertebral marrow changes involving the pedicles, pars
interarticularis and articular facet processes at L4 and L5 levels secondary to degen-
erative unstable lishthesis.
Fig. 3. Sagittal midline FSE T2-weighted MR image (TR/TE = 2450/120 ms) show-
ing retrolisthesis of L3 and L4 associated with traction spurs. Traction spurs start
2–3 mm away from the vertebral edges and have a more horizontal orientation,
unlike claw osteophytes which develop closer to the endplate and have a pro-
nounced curved shape, directed toward the opposite endplate.
Traction spurs are considered to be due to increased tensile
stresses exerted by the anterior longitudinal ligament (ALL) and
Sharpey’s fibers upon bone insertions in case of increased abnor-
mal motion [15]. Traction spurs typically develop 2–3 mm apart
from vertebral edges with a purely horizontal orientation, unlike
claw osteophytes which develop closer to the endplate and have
a curved shape, directed towards the opposite endplate (Fig. 3).
Claw osteophytes occur in more advanced stages of degenerative
processes where they serve to self-stabilize the motion segment.
129
Fig. 4. Axial FSE T2-weighted (TR/TE = 2340/130 ms) MR image showing a synovial
cyst complicating a mild spondylolisthesis of L5 secondary to hypertrophic remod-
elling and subluxation of the facets with joint effusion on the left. The cyst worsens
the spinal canal stenosis.
Fig. 5. Sagittal lateral FSE T2-weighted (TR/TE = 2400/110 ms) image. Spondyloly-
sis and spondylilysthesis of L5 with oedema of the endplates. Lytic anterolisthesis
does not modify central canal size but provokes deformation and stenosis of the
intervertebral foramina with eventual compression of the nerve ganglion between
the posterolateral disc annulus and the advanced pars interarticularis. The degree of
foraminal stenosis can differ side to side, namely in case of rotational, asymmetrical
listhesis.
Modic type-I changes are more common in case of painful insta-
bility, tend to convert into more stable types II–III after fusion, and
may persist or reappear in case of pseudarthrosis (Fig. 1). The asso-
ciation with instability partly explains why the most severe type-I-
Modic changes (involving from 25% up to 50% and more of vertebral
body height) have been considered valuable indicators of symp-
tomatic disc pathology with a positive predictive value of 100%
[17].
Rarely, endplate osteochondrosis can have an aggressive evolu-
tion simulating infectious or tumoral processes (Fig. 6).
Patients with degenerative spondylolisthesis have significantly
larger facet joint effusions with the probability of anterolisthe-
sis being proportional to effusion size [16]. The most significant
association can be found in subjects with a mobile, intermit-
tent, low-grade antelisthesis in comparison to manifest and more
advanced cases [16].
While intermittent slippages cause increased movement of the
articular processes between the standing posture and the supine
position, the oldest forms undergo a stabilization because of the
130 R. Izzo et al. / European Journal of Radiology 82 (2013) 127–138
Fig. 6. (a and b) A 40-year-old man complaining of worsening low back pain. Sagittal
midline reformatted CT (a) and FSE axial T1-weighted fat-sat (TR/TE = 400/20 ms) (b).
MR image after gadolinium administration (b) show large erosion and geographic
lytic changes of the L4 inferior endplate and vertebral body along with oedema,
vascular congestion responsible for intense contrast enhancement, and sclerosis.
There are no disc signal modifications or contrast enhancement. A needle biopsy
demonstrated the benign inflammatory nature of the lesion.
development of osteophytes, capsule fibrosis and disc collapse with
reduced joint effusions.
In the absence of measurable translation on supine MRI, the
largest facet effusions (>1.5 mm) are highly predictive of degen-
erative spondylolisthesis at L4–L5. A facet effusion >1 mm would
still be an indication for dynamic radiographs or MR to diagnose
an occult degenerative spondylolisthesis that can be missed with
supine positioning [16].
Nevertheless, the specificity and the clinical relevance of all
findings discussed for diagnosing instability are not consistent in
the different reports and need to be established definitively [18].
Conventional imaging findings are in effect only indirect signs of
instability and even though MR is the most accurate modality for
the study of degenerative spine, conventional recumbent imaging
cannot be reliable in assessing instability and chronic pain.
Open MR systems allow positional-dynamic studies in either
standing or seated positions to detect increased and abnormal
intersegmental movements [2]. A comparison of the relative rota-
tion and translation of adjacent motion segments in neutral,
extended and flexed position may disclose dysfunctional move-
ments which can worsen or uncover a central canal, lateral recesses
and/or foraminal stenosis, disc protrusion or extrusion correlating
with increased levels of pain [2,19].
Weishaupt et al., in a cohort of patients studied by functional
MRI, reported a change in diagnosis in the form of disc hernia-
tion, nerve root impingement or foraminal stenosis in 26.3% of
Fig. 7. Axial CT scan. ‘Wrap around bumper’ osteophytes develop along the capsu-
lar attachments of the facet joints in the advanced stages of degenerative lumbar
spine and in long-standing degenerative spondylolisthesis as an attempt of self-
restabilization. Notice also the hypertrophy and sclerosis of the facet processes.
cases comparing supine and flexion position, in 22.4% passing from
supine to extension position, and in 28.2% between flexion and
extension [19].
An unstable slip will also lead to nerve root compression in
case of unstable degenerative or isthmic spondylolisthesis since the
foramen is increasingly narrowed with worsening listhesis giving
rise to stretching of the already constricted nerve root [20].
By using positional MR in patients with spondylolysis and
spondylolisthesis Niggemann et al. described a third form of insta-
bility (out of anterior and angular instabilities) defined as posterior,
not detectable by flexion–extension radiographs, due to move-
ments in the spondylolytic cleft [20]. Posterior instability may lead
to protrusion in the neuroforamen of a bony spur, flava ligaments
or a cyst arising from the cleft itself with nerve root compression
[20].
As spinal degeneration further progresses, in the final phase
of “restabilization”, the fibrosis of the joint capsules, the forma-
tion of osteophytes, the marked discal collapse and the radial
expansive remodelling of vertebral bodies lead to an overall
reduction of mobility and increased stiffness [11]. During the
restabilization, the final evolution of the degenerative process,
functional limitation, stiffness, immobility, spinal pain can eventu-
ally fade or again the irritation or impairment of nervous structures
lead to pain and deficits of varying severity [11]. The resta-
bilization phase is generally associated with imaging findings
different from those of instability, such as significant disc col-
lapse, radial remodelling of vertebral bodies [10], claw osteophytes,
‘wrap around bumper’ osteophytes (Fig. 7) [21] Modic type-3
changes, facet sclerosis, and neoarthrosis between the spinous
processes [22].
Claw and ‘wrap around bumper’ osteophytes develop along the
insertions of the disc and the facet capsule respectively to enlarge
the joint surfaces. The radial enlargement in the horizontal plane
of the vertebral body between two adjacent collapsed interverte-
bral disks also contribute to self-stabilize the motion segment but
may result in sagittal stenosis of the central spinal canal and lateral
recesses [10].
Pseudoarthrosis between spinous processes of adjacent verte-
bral levels (typically L4–L5) develops with advanced collapse of the
intervertebral disk and spondylolisthesis. At least in the early stages
the spinous processes could have a dynamic supportive function for
the articular facets and disc, forming a fourth functional column
(the extreme column) for acceptance and transmission of loads,
which tends to block further vertical movement of the facets and
horizontal slippage of the vertebral body [22].
 R. Izzo et al. / European Journal of Radiology 82 (2013) 127–138
In agreement with Kirkaldy-Willis and Farfan, Murata et al.
found that patients with back pain who underwent dynamic
radiographs and MRI presented linear or angular hypermobility
associated with normal or only mildly degenerative discs and hypo-
mobility in the case of advanced disc disease [23].
A spontaneous tendency to restabilization may occur even in
case of spondylolisthesis whose presence does not always imply
actual instability [11,24]. Disc collapse and osteophytes may block
the progression of slippage, often with secondary improvement
of pain. In effect, despite widespread opinion, spondylolisthesis is
not always associated with increased motion. A 10-year follow-up
study found progression of slippage only in 30% of cases and 65% of
patients who were initially neurologically normal did not worsen
and could be treated conservatively [25]. Since disc collapse makes
the progression of listhesis less likely, an instrumented fusion is
recommended to prevent progression of listhesis only when the
preoperative disc height is greater than 2 mm [26].
By using an open MR system, McGregor et al. investigated the
kinematics of the lumbar spine in 14 subjects with degenerative
and 15 with isthmic spondylolisthesis without detecting signifi-
cant angular or translational hypermobility compared to subjects
with no history of low back pain [24]. The lack of hypermobility
was partly attributed to a potentially increased activity in the erec-
tor spinae and multifidus muscles that may stabilize and brace the
spine around the level of the defect [24].
Even though positional MRI combines the benefits of conven-
tional MRI and functional radiography, dynamic imaging with an
open MR device presents several drawbacks including the reduced
signal/noise ratio due to the lower field strength, the limited pos-
tures that the patient may hold without moving or experiencing
pain, and possible underestimation of the extent of disease in the
seated position compared with the standing position [2]. Nev-
ertheless, when spinal canal stenosis is strongly suspected and
conventional MR–CT findings and clinical symptoms suggesting
surgery are discordant, reimaging with dynamic MR is recom-
mended to search for major canal stenosis [2].
Comparing the results of traditional and functional MR in a small
cohort of patients, Smith reported abnormal findings detected only
by dynamic studies in 52% of patients with appropriate successful
treatment in all cases [27].
Axial-loaded CT (AL-CT) and MR (AL-MR) simulate the weight-
bearing upright position and depict several findings referred to as
elementary modifications such as the appearance of or increase
in disc bulging, vacuum sign disappearance or appearance, facet
underslipping, longitudinal hypermobility associated with thick-
ening of the flava ligaments, appearance of or increase in listhesis,
vertebral rotational movements around the vertical (Z) axis, and
the appearance or increase of canal and/or foraminal stenosis
[22]. These findings can be observed alone or coexist in vari-
ous patterns referred to as complex dynamic modifications (CDM)
[22].
Normally the loaded lumbar spine exhibits a slight increase in
lordosis, minimal bulging of discs and posterior discal wedging,
without significant changes in interspinous spaces or facet gaps.
Out of this normal findings abnormal motion patterns tend to
evolve in a quite stereotyped way up to degenerative listhesis
[22]. Unfortunately axial-loading imaging cannot reflect postural
changes related to muscle tone and physiological loads that are
not uniform at different levels but increase in the caudal direction
along the lumbar spine, so upright/positional MR is considered to
outweigh axial-loaded MR and CT [28].
AL-CT is considered superior to AL-MR in showing complex
tridimensional movements of the posterior arches thanks to 3D
reconstructions. It is preferred in postoperative imaging and in the
late stages of instability where abnormal movements are expected
to be globally reduced. In case of suspected spinal canal stenosis
131
secondary to instability AL-imaging may guide the treatment
[27,28].
The assumption of a close relationship between movement
abnormalities and pain is the basis of fusion and stabilization
surgery. Surgery currently offers a variety of techniques and devices
that aim to block the abnormal movements of the FSU considered
symptomatic or to restore FSU stability by means of flexible fixers
or discal replacement prostheses.
Despite the advantages of axial-loading and positional MRI,
dynamic radiographs remain the most commonly used reference
in view of surgery because of the ease of execution, the wide avail-
ability and lower cost [29].
A variety of methods have been proposed to assess instability
on dynamic radiographs. On flexion–extension lateral radiographs
of the lumbar spine a global anterior translation over 4.5 mm in the
sagittal plane and a global sagittal plane rotation >15◦ from L1–L2
to L3–L4, >20◦ at L4–L5 and >25◦ at L5–S1 from extension to flexion
seem to provide a good degree of accuracy [30].
In order to avoid errors due to image magnification or distor-
tion, sagittal translation can be better expressed as a percentage of
vertebral width and the area of an angular movement must always
be carefully located in the central radiographic beam [29].
More indicative of instability would be the progression over
time of deformation and misalignments. Despite the common
acceptance and clinical impact of these radiologic signs, the role of
plain functional radiographs remains debated due to several impor-
tant limitations including poor accuracy and sensitivity, lack of any
standardization in technical execution and measurement methods
as well as the wide individual variability in motion characteristics
and the wide overlap of patterns among symptomatic and nor-
mal subjects [31]. The optimal patient position to disclose maximal
motion of the lumbar spine has also been a topic of debate: while
many authors prefer to take measurements in the sitting position,
others use lateral decubitus or traction-compression in lieu of the
classical standing position to maximize the abnormal motion [14].
Whatever the patient’s decubitus, functional plain radiographs
show just the static relations of FSUs in the positions of maximal
flexion and extension with no insights into the quality of ongoing
movement. For this reason Brunton et al. claimed cineradiogra-
phy to be the most accurate technique for recognition of instability
[32]. Kanayama et al. investigated cervical and lumbar motion pat-
terns in normal and pathologic spines using the cineradiographic
technique and observed that normal cervical and lumbosacral seg-
mental motions occur not simultaneously but stepwise, starting
from the upper levels and transmitted in a well-regulated way to
the lower segments [33]. In pathological spines, however, motion
initiates at the unstable segments.
Unfortunately the cineradiography has no routine clinical appli-
cation.
Compared to cineradiography, greater levels of precision in
movement measurements in lateral flexion–extension radiographs
are afforded by distortion-compensated roentgen analysis (DCRA),
an advanced noninvasive imaging method [34]. To date, however,
none of the routine clinical radiographic techniques and measure-
ments proposed has a statistically proved specific association with
clinical symptoms.
Despite all the efforts made in recent years, there is no gold
standard for diagnosis or a clear correlation between symptoms
and imaging: abnormal movements and degenerated discs are
very often present in asymptomatic individuals. Therefore, the
biomechanical pattern and radiographic findings must be always
correlated with the patient’s clinical history and physical examina-
tion to establish a diagnosis of spinal instability.
Patients with lumbar spine instability often complain of chronic
or recurrent pain localized in the low back or radiating to the lower
extremities, with a mechanical–postural basis, being worsened by
132 R. Izzo et al. / European Journal of Radiology 82 (2013) 127–138
lifting, standing, sitting, bending or twisting and relieved by lying
supine. Typical signs of altered movement control within the neu-
tral zone are a good range of spinal mobility but with a “painful arc”
and the inability to return to erect posture from bending without
assistance [11]. Patients tend to develop compensatory movements
to stabilize the dysfunctional motion segment through the acti-
vation of torque-producing muscles which provide global trunk
stabilization. High levels of intra-abdominal pressure are generated
even during low-load tasks.
Through different mechanisms degenerative spondylolisthesis
can generate three pain patterns including mechanical pain, radic-
ular pain or neurogenic claudication. Neurogenic claudication is
provoked by severe canal stenosis with increased epidural pres-
sure, compression of the vessels for the cauda equina and oxygen
deprivation of the nerve roots. Pain along the buttocks and both
legs brought on by walking is often associated with tingling, numb-
ness, and leg weakness and is relieved by resting. Unlike vascular
claudication, pain in neurogenic claudication is also provoked by
standing. Pain presents along with segmental sensory loss after
walking a variable distance and is relieved by flexing the spine.
Abnormal movements detectable in a degenerative spine are not
necessarily the cause of pain.
According to Mulholland, instability is often a myth and conse-
quently fusion surgery is not always the correct choice [35]. In many
cases spinal pain is not linked to dysfunctional movement, but also
and more to an abnormal irregular distribution of loads between
joint surfaces. In fact there exists a poor correlation between clinics
and imaging, and pain may persist even after technically successful
fixations or unexpectedly resolves in cases of pseudarthrosis. While
the most rigid fixation has not improved the pain control the flexi-
ble stabilization techniques are successful [35]. In fact pain seems to
be often elicited primarily from load stresses caused by the posture
or powerful muscle contraction such as that of the erector muscles
during lifting tasks rather than by movement itself [35].
As in other load-bearing joints, an altered distribution of loads
can generate pain in nociceptive endings widely present in degen-
erated discs, vertebral bodies and facet joints. Dynamic stabilizers
provide an intermediate solution between conservative treatment
and traditional fusion surgery and often consent a minimally inva-
sive approach. Their general action consists in redistributing loads
within motion segments, eliminating or reducing any noxious con-
centration and restoring the normal range of motion.
3. Traumatic instability
Spinal traumas affect a complex structure composed of soft
and bony components having different traumatic susceptibility and
healing potential. This complexity hampers the classification of
traumas and the assessment of instability and treatment outcomes.
However, unlike degenerative instability, the relationship between
imaging findings and clinical symptoms tends to be more direct.
The definition of instability in the spinal fractures is the topic of
ongoing research. Although fractures are traditionally divided into
stable and unstable, all spinal components contribute to stability.
Damage to any spinal structure gives rise to some degree of insta-
bility which is not an all or nothing phenomenon, while complete
instability is rare.
A number of biomechanical studies have analysed the contribu-
tions of bony and soft structures to spinal stability and the effects
of trauma. Holdsworth proposed a two-vertical-columns model
consisting of a pressure-resistant ventral column of vertebral bod-
ies and discs and a posterior column of neural arches subject to
tensile stresses [36]. Holdsworth’s model is based on the con-
cept that traumatic failure of the posterior arc ligaments would be
sufficient to create instability in flexion [36]. Denis subsequently
Fig. 8. Three-column concept of spinal stability by Denis. This model emphasizes the
pivotal role of the middle column and the PLL. Abbreviations: SSL = supraspinous lig-
ament; PLL = posterior longitudinal ligament; ALL = anterior longitudinal ligament.
proposed a model formed by three vertical columns which divides
Holdsworth’s anterior column into an anterior column including
the anterior halves of the bodies and disks with the adjacent ALL and
a middle column including the posterior half of the bodies and discs
with posterior longitudinal ligament (PLL) (Fig. 8). The posterior
column, like that of Holdsworth, consists of neural arches and the
posterior ligamentous complex (PLC) including the supraspinous
(SSL), interspinous (ISL), flava ligaments (FL) and facet joint cap-
sules [37].
From his clinical observations Denis concluded that only the
simultaneous failure of at least two columns created situations of
instability [37]. The middle column became pivotal in the deter-
mination of spinal stability [37]. Denis redefined burst fractures as
unstable because of the concomitant damage to both anterior and
middle columns [37]. Denis’s model is useful in evaluating instabil-
ity in all major traumatic mechanisms on both bone and soft tissue,
and remains among the most accepted references today.
Kifune et al. assessed the response of fresh human spine
specimens to incremental impacts produced by axial and compres-
sion/flexion loads in a high-speed trauma apparatus [38]. Flexibility
and load–displacement curves were measured before and after
each impact and range of motion (ROM), neutral zone (NZ), elastic
zone (EZ) instabilities were defined as motion before/motion after
impact ratios [38]. Endplate, wedge and burst occurred after an
average impact energy of 57, 84 and 104 N m, respectively. While
endplate fractures produced no significant changes in motion pat-
terns, the first changes occurred in wedge fractures. Burst fractures
showed the largest changes in instabilities, dramatic in comparison
with those of wedge fractures [38].
While a considerable amount of energy is required to pro-
duce the first injury in a vertebra, a relatively smaller additional
amount is sufficient to transform an initially stable injury into an
unstable lesion [38]. From a clinical point of view, only a small incre-
mental trauma is necessary to convert a stable fracture requiring
conservative treatment to an unstable fracture requiring surgical
stabilization [38]. Among all motion parameters the NZ was the
most sensitive in defining instability [38]. The types of injuries
occurring in the cervical, thoracic and lumbar spine differ because
of segmental structural, biomechanical and failure mechanisms
peculiarities.
Spinal traumas with spinal cord injuries have a severe impact
on society as well as the individual. The cervical spine is the most
frequently injured part of the vertebral column and most cervical
spinal cord injuries occur from unstable cervical spine fractures [39]
(Fig. 9). Only a third of spinal trauma victims initially present with
neurological deficits [40]. Patients with clinically occult unstable
cervical fractures may subsequently develop neurological deficits
in up to 10% of cases [41].
 R. Izzo et al. / European Journal of Radiology 82 (2013) 127–138
Fig. 9. (a and b) Midline sagittal reformatted CT scan (a) and axial CT (b) images of
a car accident victim. Typical tear-drop fracture of C4 with posterior displacement
and rotation of the posterior vertebral body fragment, spinal cord compression and
double fracture of laminae. The extension of bony damage is far better depicted by
CT while MR (not shown) assessed the state of the ligaments.
The traditional initial approach for evaluating the acutely
injured cervical spine has been with conventional radiographs.
However, conventional radiology is inadequate for the assessment
of fracture stability [3]. Cross-table lateral, AP and open-mouth
odontoid views in a large cohort of patients with cervical spine
injuries missed 61% of all fractures, 36% of subluxations and dis-
locations and gave false-negative results in 23% of the patients,
half of whom with unstable injuries [3]. Flexion–extension lat-
eral radiographs have been advocated in the acute setting after
blunt trauma, but to date no large clinical studies have identi-
fied and definitively validated criteria to assess instability due to
wide motion variations between different levels of the same sub-
ject, between normal subjects and with respect to gender and age
[42]. However, since pure ligamentous cervical spine injuries are
rare, and in the acute phase can be masked by neck pain and mus-
cle spasm limiting motion, flexion–extension radiographies are
considered to add little or nothing in the assessment of unstable
lesions.
Even though a relative sagittal translation of more than 2 mm is
generally considered significant, a clinical validation for this crite-
rion is also lacking [42]. In rare cases, dynamic plain radiographs
133
can be useful to assess delayed instability two to four weeks after
trauma.
Multi-detector CT (MDCT) can yield high-resolution reformat-
ting in every spatial plane starting from isotropic image voxels
to detect even the tiniest fractures in the middle and posterior
columns revealing potentially unstable lesions [43]. MDCT also
allows an excellent evaluation of vertebral alignment and the spa-
tial position of dislocated bone fragments, and overcomes MR in
the detection of fractures involving the neural arc and C0–C2 joints
[43].
However, the effective clinical value of subtle incomplete frac-
tures of the vertebral body not detected by radiography and of
isolated lesions involving the transverse or spinous process and
laminae awaits to be defined [39].
In obtunded patients the concern for occult cervical lesions
undetected on initial imaging is particularly critical and the risks
of passive physician-assisted dynamic radiographs considered too
high. Hogan et al. reported on the capability of MDCT to predict soft
tissue injuries and the eventual added value of MR in 366 obtunded
[44]. They found a negative predictive value for CT of 98.9% for
ligamentous lesions and 100% predictive value for unstable cervi-
cal spine injuries with only four cases of undetected ligamentous
lesions, none of which was judged unstable. Their conclusion was
that a normal MDCT study alone may exclude unstable cervical
injuries [44].
Wadhwa et al. assessed the results of incorporating
flexion–extension CT into the trauma protocol of a cohort of
77 patients with blunt cervical trauma, obtaining a sensitivity of
100% for clinically unstable injuries and a far superior visualization
of the cervico-thoracic junction compared with conventional
dynamic radiographs [45].
Due to the greater accuracy, higher speed and reduced patient
manipulation, critical during the golden hour, MDCT is the pre-
ferred imaging modality in acute multi-trauma patients. Despite
the intrinsic higher cost, the appropriate use of CT can even be cost-
effective, related in part to the high percentage of inadequate plain
radiographs leading to additional imaging, but also to the greater
sensitivity to fractures which avoids the catastrophic consequences
of a missed injury in either the cervical and thoracolumbar spine
[46].
Thoracic and lumbar burst fractures often result in an abrupt
change in the patient’s quality of life, and, with persistent pain,
can result in functional loss leading to chronic complications over
time [47]. Burst fractures are characterized by the failure of both
the anterior and middle columns [37]. Daffner et al. indicated the
disruption of the posterior vertebral body cortical line as a reliable
conventional sign in predicting a lesion in the middle column and
a burst fracture with a reported sensitivity in case of pure burst
fractures of 100% [48]. According to Ballock et al., however, about
20% of thoracolumbar burst fractures are under-classified as simple
wedge compression fractures, so CT is indicated even in the case of
mild vertebral body compression and deformation [49].
In fact, in one comparative study, while up to two thirds of
all unstable thoracolumbar fractures were missed by conventional
radiographs, the sensitivity of MDCT reached 97.2% [50].
A yet unsolved question is which types of burst fracture can
be treated conservatively and which must be treated by surgery to
avoid post-traumatic kyphotic deformity. The different parameters
used to evaluate the final result (radiologic, clinical and functional)
contribute to the debate on the ideal treatment of these frac-
tures. Moreover, most studies have failed to establish a correlation
between the degree of spinal canal narrowing and the severity of
the neurological deficit, the stenosis being rather only correlated
with the likelihood of neurologic dysfunction [51]. Since patients
with normal neurological status and those with complete neuro-
logical deficit often present a similar degree of narrowing, at the
134 R. Izzo et al. / European Journal of Radiology 82 (2013) 127–138
time of the trauma, a greater degree of displacement of bone frag-
ments probably occur and causes more severe lesions than those
would be caused by the same bone fragment in its definitive resting
position after the accident.
By using the classic three-column concept of instability the
division of burst fractures into stable and unstable lesions can be
difficult [52].
One study proposed an increased interspinous gap, kyphosis
over 20◦ , sagittal displacement, a body height decrease over 50%,
and facet fractures as radiological hallmarks of mechanical instabil-
ity in burst fractures. However, fractures may be unstable despite
the absence of these signs because of injury to the posterior liga-
ments [52].
The surgical approaches to burst fractures can be anterior, pos-
terior or combined. They do not always prevent the worsening
of thoracolumbar kyphosis neither ensure a better outcome than
conservative treatment. To date a gold standard therapy to pre-
vent progressive kyphosis after a burst fracture is lacking. Kyphosis
occurs gradually under physiological stresses under the influence of
persistent occult instability with symptoms related to loss of spinal
alignment and compensatory adjacent level adjustments.
Relapsing kyphosis or instrumentation failure occur most fre-
quently after short posterior fixations which in recent years have
been widely performed owing to the reduced morbidity and better
residual flexibility of the spine, most important in young patients
[53].
McCormack et al. claimed that the extension of lesions to the
vertebral body and adjacent disk, reflecting the residual load-
bearing capacity of the anterior spinal column, can be predictive
of the outcome of the conservative and surgical approaches and
the risk of instrumentation failure [54]. According to McCormack
et al., the principle that osteosynthesis load-sharing between host
bone and implant in long bone fractures ensures the correct union
avoiding pseudarthrosis and implant failure and is also valuable in
the management of acute spinal fractures [54].
McCormack et al.’s load-sharing classification (LSC) [54] is a
three-point score ranging from a minimum of three up to nine
points based on:
- the amount of damaged vertebral body (on CT sagittal reconstruc-
tions);
- the horizontal scattering of fracture fragments (on CT axial scans);
- the degree of kyphosis correction and secondary gap in the ante-
rior column after surgery.
Either an excessive radial displacement of fragments, prevent-
ing fracture healing, or a residual gap in the anterior column after
instrumentation, being inevitable in the most severe kyphoses,
tend to eliminate the load-sharing of the anterior column under-
going instrumentation to maximal mechanical stress and failure
risk [54]. High-scoring patients having undergone short posterior
fixation or conservative management tend to develop relapsing
kyphosis and eventual device mobilization and failure, whereas
low-scoring patients can avoid the risks of the greater morbidity
of long-segment instrumentation.
Jeffrey et al. reported that the more unstable LSC high-scoring
cases (seven or more points) only evolve satisfactorily after a
combined anterior and posterior surgical approach, while the low-
scoring fractures could be appropriately treated by posterior short
instrumentation [55]. Another factor favouring relapsing kypho-
sis, possibly even more important than BONE damage, seems to
be disc collapse adjacent to the fractured endplate [56]. Serial MR
controls after thoracolumbar fractures suggest that disc collapse is
mainly due to settling and herniation through the damaged end-
plates rather than degeneration since most disks retain a normal
signal [56].
Abnormalities such as anterior or diffuse disc collapse and deep
central disc herniation, namely when associated with high LSC
scores, require an anterior or combined surgical approach and
exclude any conservative choice or short posterior fixation given
the high risk of implant failure and progressive kyphosis [56].
The McCormack classification is effective for detecting fractures
that may lead to correction loss but does not account for disrup-
tion of ligaments whose integrity is also of paramount importance
in the maintenance of spinal stability. Missed ligament and disc
lesions can provoke instability and eventual neurologic damage
as well as relapsing kyphosis and chronic pain in patients treated
conservatively, or cause surgical failure.
Magerl’s classification of thoracolumbar fractures includes three
main group of fractures referred to as A, B and C types, corre-
sponding to increasing degrees of instability. The crucial finding
distinguishing more stable Magerl A (compression) from more
unstable B (distraction) type fractures is the state of the posterior
ligaments [57]. The status of posterior ligaments after an injury is of
great importance for the stability of the injured spine [58]. In burst
fractures the condition of the posterior column, rather than the
middle column, a pivotal component of Denis’s concept, is thought
to be a better indicator of burst fracture instability which increases
remarkably in the case of lesions to the posterior ligaments [58].
Posterior ligament failure is far more common than was esti-
mated in the past and in case of missed diagnosis of a lesion
to these ligaments unsuitable conservative treatment or anterior
surgery can have drawbacks such as instability, progressive defor-
mity, chronic pain, and disability. Leferink et al.’s retrospective
study comparing the surgical and imaging records of 160 patients
with type A and type B spinal fractures reported that about 30% of
type B fractures are under-diagnosed as type A since ligamentary
lesions are often not recognized in plain X-rays and CT scans [59].
Determination of the distance between the spinous processes
can be misleading if a reduction of distraction has occurred.
Knowledge of the state of the posterior ligaments is crucial for an
appropriate treatment strategy as the destabilizing effects of injury
to the posterior ligaments can be annulled with a posterior rather
than anterior fixation.
MR is the only imaging modality that directly and routinely
assesses changes in ligaments. Lee et al. reported the accuracy of MR
in detecting lesions to the ISL, SSL and FL as 97%, 90.9% and 87.9%,
respectively, and recommend the routine use of fat-suppressed T2-
weighted sagittal sequences [4] (Fig. 10).
Similarly, using both discontinuity of the black stripe and
hyperintensity in T2-weighted images as morphological criteria in
diagnosing ISL and SSL injury, Haba et al. obtained intra-operatory
confirmation in 94.6% of cases [60]. However, in a prospective study
on 47 patients with acute cervical spine injury analysed by MR, Jef-
frey et al. reported a low specificity, ranging from 56% for the facet
capsules to 67% for the interspinous ligament, as well as a tendency
to over-read injuries to the PLC with consequent risks of unneces-
sary surgery because of too many false-positives. Among posterior
ligaments the facet capsules were the most difficult to assess [61].
Furthermore, the ALL and PLL, critical components of Denis’s
anterior and middle columns respectively, are often impossible to
resolve as separate structures and visualize in a complete form,
namely in the case of spondylosis, so any apparent discontinuity of
the corresponding black line does not necessarily indicate a loss of
integrity.
Although MRI has improved our understanding of spinal injuries
and is guiding the decisions for treatment, the literature reports
shortcomings that require a critical reassessment of its role in
the prediction of spinal instability. While an evident discontinu-
ity or avulsion of a ligament may be considered a reliable sign of
a lesion, the real significance of signal modifications occurring in
and around apparently continuous ligaments remains to be defined.
 R. Izzo et al. / European Journal of Radiology 82 (2013) 127–138
Fig. 10. (a and b) A 29-year-old male victim of a high speed trauma. Midline refor-
matted CT scan (a) and sagittal FSE STIR-weighted (TR/TE/TI = 2740/130/150 ms) MR
scan (b). Burst fracture of L2 vertebral body with migration of a rotated bony frag-
ment in the spinal canal. Horizontal fracture splitting the neural arch traversing
the spinous process, laminae and peduncles. MR with signal fat suppression bet-
ter demonstrates associated injuries to the supraspinous, interspinous and flava
ligaments. Notice the extended hematoma along paraspinal muscles.
Furthermore, there is a problem of grading the signal changes and
interobserver reliability [5]. On the contrary, a ligament must not
necessarily be torn to become biomechanically ineffective, and the
absence of signal modifications cannot always exclude instability
[5].
135
Several trauma centres which have adopted protocols that use
MRI in association with radiography and high-resolution CT, did not
obtain results that justify its routine inclusion in standard protocols
[62].
To date, the objectivity of MR in assessing the integrity of lig-
aments is not yet rigorously defined and its role in predicting
mechanical instability and its routine use in protocols to assess
blunt spinal injuries remains debated [62]. All these arguments
emphasize not only the need for controlled prospective trials to
prove the real value of MRI in the assessment of spinal soft tis-
sue injuries through further comparisons with surgical findings,
not done in most studies, but also the need to develop new MRI
techniques to improve current imaging capabilities.
4. Neoplastic instability
The restoration or maintenance of spinal stability is an impor-
tant objective in the surgical or conservative management of
patients with spinal tumours. However, the definition and manage-
ment of instability are controversial in this setting too [6]. The Spine
Oncology Study Group (SOSG) defines neoplastic spinal instability
as the loss of spinal integrity associated with movement-related
pain, symptomatic or progressive deformity, and/or neural com-
promise under physiologic loads [8].
The biomechanical effect of spinal tumours and the mechanisms
of vertebral neoplastic collapse remain unsettled [7]. There are no
evidence-based guidelines currently available to assess the risk of
spinal instability in the setting of neoplastic spinal disease, so pre-
dicting the risk of a pathological fracture or the timing of a collapse
remains challenging even when the lesions are well-characterized
by neuroimaging [7].
While the role for surgery in the setting of neurologic deficit
and spinal cord compression with metastatic disease is well-
established, it is important in patients with no deficit to recognize
which situations are unstable or may lead to spinal instability and
neurologic injury to prevent an impending collapse.
Conservative treatments can reduce the tumour mass and pro-
mote the reparative process with calcification or ossification in the
metastatic lesion, but they cannot provide immediate spinal sta-
bility in cases of impending or progressive collapse for which only
surgical treatment guarantees immediate stability. The classical
clinical indications for the stabilization of vertebral lesions include
vertebral body collapse greater than 50%, metastatic involvement
of the pedicle, involvement of one-half of the body (Fig. 11) or of
the anterior and posterior elements [63]. However, direct evidence
in support of these criteria is lacking because there are no stud-
ies specifically designed to assess the natural history of metastatic
spinal diseases.
In the absence of obvious vertebral body collapse or defor-
mity, segmental instability is suspected when axial or mechanical
pain is present [7]. Activity-related or postural axial spinal pain
is an important feature of neoplastic instability so that clinical
and symptom-related criteria have been used as bases of the
“NOMS” (neurologic, oncologic, mechanical instability, systemic
disease) classification system [64]. However pain cannot be the
only feature in defining oncologic instability and this system
which does not account for radiological findings failed to receive
validation.
Mechanical pain differ from pain due to periosteal stretching
or compression of neural elements regarding severity, worsening
during recumbency and response to steroids.
The three-column concept proposed by Denis for spinal frac-
tures is not always applicable to tumours where the patterns
of bone destruction, ligamentous involvement, and neurologic
manifestations are completely different and the quality of the
136 R. Izzo et al. / European Journal of Radiology 82 (2013) 127–138
Fig. 11. (a and b) Axial (a) and reformatted sagittal (b) CT scans. Metastatic lytic
destruction of more than 50% of the L5 vertebral body is considered an indication for
surgery. In the lumbar spine the size of the lytic defect, expressed as a percentage of
tumour occupancy, is considered the most important factor predisposing to collapse.
Pre-existing spondylolysis with light listhesis.
surrounding bone and the ability to heal are generally poor
requiring a specific and different set of criteria for stability
assessment [7].
Nevertheless, just referring to Denis’s concept, Krishnaney et al.
suggested that the effects of vertebral body destruction by a
tumour could best be assessed by dividing the vertebral body
involved into 27 identically shaped adjacent cubes [65]. Destruc-
tion of all bone cubes contained in the middle axial third of the
vertebral body results in gross instability and impending col-
lapse by complete failure of both anterior and middle columns,
whereas bone loss in the middle sagittal vertical one third does
not provoke destabilization because of limited disruption of two
columns [65].
Using a multivariate logistic regression model in 53 patients
with metastatic spines, Taneichi et al. attempted to define the prob-
ability of vertebral collapse and the critical point of impending
collapse detecting significant differences in the timing and inci-
dence of vertebral body collapse between thoracic and lumbar
spine [66]. While the most important predisposing factor to col-
lapse in the thoracolumbar and lumbar vertebrae was the size of
the lytic defect expressed as a percentage of tumour occupancy,
destruction of the costovertebral joint proved to be the over-
coming risk factor in the thoracic spine [66]. In case of lesions
confined within the vertebral body, impending collapse occurred
when the vertebral body involvement was 50–60% in the tho-
racic spine and 35–40% in the thoracolumbar and lumbar spine
[66]. The greater resistance to collapse of the thoracic spine is
due to the rib cage integrity which enlarges the transverse dimen-
sion of the thoracic spine, increasing its moment of inertia and
stiffening the spine versus any kind of rotatory force. Destruction
of the costovertebral joint would favour vertebral body collapse
more than the extension of metastatic tumour within the ver-
tebral body [66]. In fact, neurologic signs often long precede
pain in the thoracic spine whose stability can be preserved until
tumour involvement progresses sufficiently to cause neurologic
deficits, whereas cervical and lumbar metastases lose stability and
become painful even several months before the onset of neurologic
deterioration.
McGowan et al. attempted to find a correlation between the
extent of a pure cancellous defect and the strength of the vertebral
body, and a critical fracture threshold using the cross-section area
of provoked defects as a measure of vertebral compromise [67].
While a linear correlation between vertebral body strength and
the cross-sectional area of cancellous defects was demonstrated
through a wide range of defect sizes, no parameter was found to
determine a distinct fracture threshold [67].
Dimar et al. investigated the combined effects of bone min-
eral density and lytic defects in an experimental study on 54
cadaveric spine specimens subjected to compressive loading after
central drilling of both cortex and spongy bone to better simulate
the indiscriminate destructive tumour action [68]. While defect
size alone did not reliably predict the critical fracture threshold,
its combination with bone mineral density gave a more accu-
rate fracture-related factor and serves as a vertebral stability
index [68].
For the purposes of creating guidelines for clinical decision-
making the SOSG developed a spine instability neoplastic score
(SINS) based on factors including spine tumour location (rigid,
semi-rigid, mobile, junctional), spinal alignment between the
motion segment affected by the tumour (normal versus sublux-
ation or de novo deformity assessed with serial radiographs or
by comparing supine and upright radiographs), bone lesion qual-
ity (lytic, blastic versus mixed) degree of vertebral body collapse,
involvement of posterior elements (none, unilateral versus bilat-
eral), and mechanical and/or postural pain [8]. In this score system,
ranging from a minimum of 2 to a maximum of 18, the radi-
ological findings making the highest contribution to scoring are
subluxation–translation, vertebral collapse over 50%, and bilateral
neural arc involvement [8]. Scores of 0–6 denote “stability,” scores
of 7–12 indicate “indeterminate (possibly impending) instability”,
while scores of 13–18 define “instability.” Patients with SINS scores
of 7–18 need surgical consultation [8]. After further validation tests,
the SINS could become the first reference to guide spine surgeons
and other clinicians in patient care and to facilitate communica-
tion among specialists dealing with primary and secondary spinal
tumours.
5. Conclusions
The assessment of stability in degenerative and neoplastic spine
and in trauma patients is a challenging problem in everyday medi-
cal practice. An insufficient correlation between clinical symptoms
and diagnostic imaging in the degenerative spine hampers not only
preoperative planning but also surgery itself as it is difficult to
compare pathologic findings with radiological reports.
The poor accuracy of all static imaging modalities including MR
in patients complaining of mechanical spinal pain is in part due
to the different positions in which patients experience pain dur-
ing loading and muscular activity and when undergo recumbent
imaging studies.
Even though it remains unsettled, dynamic changes may deter-
mine or significantly affect symptoms in both degenerative and
 R. Izzo et al. / European Journal of Radiology 82 (2013) 127–138
post-traumatic spine and, when correctly detected, improve the
success of surgery [2,19,27,28]. Unfortunately there is no functional
modality in which pathological findings are often incidentally
discovered in asymptomatic patients, whereas normal radiolog-
ical patterns are observed in patients complaining of symptoms
strongly suggestive of instability. Even when present, abnormal
movements are not necessarily the cause of pain: the abnormal and
irregular transfer of loads and their concentration through spinal
joints would in many cases be the real cause of mechanical pain,
just as in other joints in the body [35].
This concept helps to explain the lack of correlation between
degrees of disc degeneration and back pain, since individual
anatomic and secondary load transmission changes vary radically
from one patient to another, and why many patients with low back
pain fail to improve following successful most rigid lumbar fusions.
Whatever the static or functional study, it is always mandatory to
compare the radiologic findings to the clinical symptoms.
The search for instability also remains challenging in spinal
traumas. Although fractures are traditionally divided into stable
and unstable, instability is not an all-or-nothing phenomenon and
complete instability is fortunately rare [57]. Unstable radiologically
occult spinal injuries need to be assessed as accurately as possi-
ble. MDCT can detect even the tiniest fractures in the middle and
posterior columns revealing potentially unstable lesions with high
sensitivity [43]. MR is the only imaging modality that directly and
routinely assesses injuries to discs and ligaments and the effect of
these lesions on stability. MR has a high sensitivity for the assess-
ment of most components of the posterior ligamentous complex
in the setting of acute trauma, but its specificity varies [4,60,61].
The low specificity has to be accounted for by the clinician when
using MR to guide treatment choices. Because of the tendency to
over-read posterior ligament injuries MR should not be used alone
to determine a treatment plan [61]. To date the objectivity of MR
in assessing ligaments has not been rigorously defined and its role
in predicting mechanical instability and routine use in protocols to
assess blunt spinal injuries remain a matter of debate [5].
The definition and management of instability are also contro-
versial in the setting of spinal tumours. The three-column concept
proposed by Denis for spinal fractures is not always applicable
to tumours where the patterns of bone destruction, ligamentous
involvement, and neurologic manifestations are completely differ-
ent and the quality of the surrounding bone and the ability to heal
are generally poor requiring a specific and different set of criteria
for stability assessment [7]. The crucial problem remains that of
recognizing the critical point of impending vertebral body collapse
which only surgery can prevent [67].
Biomechanical data and imaging features are shedding more
light on the concept of spinal instability and the impact of different
bony and soft spinal lesions. Unfortunately our current knowledge
on this complex topic is still far from satisfactory and key pieces
are still missing in our mosaic of knowledge.
Conflict of interest
All authors have read the manuscript, they approved this sub-
mission and declare that they do not have any conflicts of interest
related to this work.
References
[1] Bram J, Zanetti M, Min K, et al. MR abnormalities of the intervertebral disks
and adjacent bone marrow as predictors of segmental instability of the lumbar
spine. Acta Radiologica 1998;39:18–23.
[2] Alyas F, Connell D, Saifuddin A. Upright positional MRI of the lumbar spine.
Clinical Radiology 2008;63:1035–48.
[3] Poonnoose PM, Ravichandran G, McClelland MR. Missed and mismanaged
injuries of the spinal cord. Journal of Trauma 2002;53:314–20.
137
[4] Lee HM, Kim HS, Kim DJ, et al. Reliability of magnetic resonance imaging in
detecting posterior ligament complex injury in thoracolumbar spinal fractures.
Spine 2000;15(2516):2079–84.
[5] Flanders AE. Magnetic resonance imaging of acute spinal trauma. In: Schwartz
ED, Flanders AE, editors. Spinal trauma. Philadelphia: Lippincott Williams &
Wilkins; 2007.
[6] York JE, Gokaslan ZL. Instrumentation of the spine in metastatic disease. In:
Errico TJ, editor. Spine: state of the art reviews, vol. 13. Philadelphia: Hanley &
Belfus; 1999. p. 335–50.
[7] Fourney DR, Gokaslan ZL. Spinal instability and deformity due to neoplastic
conditions. Neurosurgical Focus 2003;14(1):1–7.
[8] Fisher CG, Di Paola CP, Ryken TC, et al. A novel classification system for
spinal instability in neoplastic disease. An evidence-based approach and
expert consensus from the Spine Oncology Study Group. Spine 2010;35(22):
1221–9.
[9] Modic MT, Masaryk TJ, Ross JS, et al. Imaging of degenerative disk disease.
Radiology 1988;168:177–86.
[10] Jinkins JR. Acquired degenerative changes of the intervertebral segments at
and suprajacent to the lombosacral junction. A radioanatomic analysis of the
non discal structures of the spinal column and perispinal soft tissues. European
Journal of Radiology 2004;50(2):134–58.
[11] Kirkaldy-Willis WH, Farfan HF. Instability of the lumbar spine. Clinical
Orthopaedics and Related Research 1982;165:110–23.
[12] Herkowitz HN. Spine update: degenerative lumbar spondylolisthesis. Spine
1995;20:1084–90.
[13] Park AL. Instability: clinical manifestations and assessment. In: Slipman CW,
Derby R, Simeone FA, Mayer TG, editors. Interventional spine an algorithmic
approach. Philadelphia: Saunders; 2008.
[14] Penning L, Blickman JR. Instability in lumbar spondylolisthesis: a radio-
graphic study of several concepts. AJR American Journal of Roentgenology
1980;134:293–301.
[15] Macnab I. The traction spur: an indicator of segmental instability. The Journal
of Bone and Joint Surgery American Volume 1971;53:663–70.
[16] Chaput C, Padon D, Rush J, et al. The significance of increased fluid signal on
magnetic resonance imaging in lumbar facets in relationship to degenerative
spondylolisthesis. Spine 2007;32(17):1883–7.
[17] Weishaupt D, Zanetti M, Hodler J, et al. Painful lumbar disk derangement:
relevance of endplate abnormalities at MR imaging. Radiology 2001;218:
420–7.
[18] Weishaupt D, Zanetti M, Hodler J, et al. MR imaging of the lumbar spine:
prevalence of intervertebral disk extrusion and sequestration, nerve root com-
pression, end plate abnormalities, and osteoarthritis of the facet joints in
asymptomatic volunteers. Radiology 1998;209:661–6.
[19] Weishaupt D, Schmid MR, Zanetti M, Positional MR. imaging of the lumbar
spine: does it demonstrated nerve root compromise not visible a conventional
MR imaging? Radiology 2000;215:247–53.
[20] Niggemann P, Kuchta J, Beyer HK, et al. Spondylolysis and spondylolisthe-
sis prevalence of different forms of instability and clinical implications. Spine
2010;20:1–6.
[21] Fujiwara A, Lim TH, An HS, et al. The effect of disc degeneration and facet
joint osteoarthritis on the segmental flexibility of the lumbar spine. Spine
2000;25(23):3036–44.
[22] Cartolari R. Functional evaluation of the lumbar spine with axial-loaded com-
puter tomography and cine ALCT. Rivista di Neuroradiologia 1997;10:569–84.
[23] Murata M, Morio Y, Kuranobu K. Lumbar disc de generation and segmental
instability: a comparison of magnetic resonance images and plain radiographs
of patients with low back pain. Archives of Orthopaedic and Trauma Surgery
1994;113:297–301.
[24] McGregor AH, Anderton L, Gedroyc WM, et al. The use of interventional open
MRI to assess the kinematics of the lumbar spine in patients with spondylolis-
thesis. Spine 2002;27(14):1582–6.
[25] Matsunaga S, Ijiri K, Hayashi K. Nonsurgically managed patients with degen-
erative spondylolisthesis: a 10- to 18-year follow-up study. Journal of
Neurosurgery 2000;93:194–8.
[26] Matsunaga S, Sakou T, Morizono Y, et al. Natural history of degenerative
spondylolisthesis: pathogenesis and natural course of the slippage. Spine
1990;15:1204–10.
[27] Smith FW. Positional upright imaging of the lumbar spine modifies the man-
agement of low back pain and sciatica. Oxford: European Society of Skeletal
Radiology; 2005.
[28] Hioki A, Miyamoto K, Sakai H, et al. Lumbar axial loading device alters lum-
bar sagittal alignment differently from upright standing position: a computed
tomography study. Spine 2010;35(9):995–1001.
[29] Dupuis PR, Yong-Hing K, Cassidy JD, Kirkaldy-Willis WH. Radiological diagnosis
of degenerative lumbar spinal instability. Spine 1985;10:262–76.
[30] Bernhardt M, White AA, Panjabi MM. Biomechanical considerations of spinal
stability. In: Rothman RH, Simeone FA, editors. The spine. Philadelphia: WB
Saunders; 1992. p. 1167–96.
[31] Leone A, Guglielmi G, Cassar-Pullicino VN, et al. Lumbar intervertebral insta-
bility: a review. Radiology 2007;245:62–77.
[32] Brunton FJ, Wilkinson JA, Wise KSH, et al. Cineradiography in cervical spondy-
losis a a means of determining the level of anterior fusion. The Journal of Bone
and Joint Surgery British Volume 1982;64:399–404.
[33] Kanayama M, Abumi K, Kaneda K, et al. Phase lag of the intersegmental motion
in flexion–extension of the lumbar and lumbosacral spine. An in vivo study.
Spine 1996;21:1416–22.
138 R. Izzo et al. / European Journal of Radiology 82 (2013) 127–138
[34] Leivseth G, Brinckman N, Frobin W, et al. Assessment of sagittal seg-
mental motion in the lumbar spine: a comparison between distortion-
compensated and stereophotogrammetric roentgen analysis. Spine 1998;23:
2648–55.
[35] Mulholland RC. The myth of lumbar instability: the importance of abnor-
mal loading as a cause of low back pain. European Spine Journal 2008;17:
619–25.
[36] Holdsworth FW. Fracture, dislocation and fracture-dislocation of the spine. The
Journal of Bone and Joint Surgery British Volume 1963;45B:6–20.
[37] Denis F. The three columns spine and its significance in the classification of
acute thoracolumbar spinal injuries. Spine 1983;8:817–31.
[38] Kifune M, Panjabi MM, Arand M, et al. Fracture pattern and instability of tho-
racolumbar injuries. European Spine Journal 1995;4:98–103.
[39] Blakmore C, Linnau KF. Controversies in clearing the spine. In: Schwartz ED,
Flandres AE, editors. Spinal trauma imaging, diagnosis and treatment. Philadel-
phia: Lippincott Williams & Wilkins; 2007. p. 63–70.
[40] Fine PR, Kuhlemeier KV, De Vivo MJ, et al. Spinal cord injury: an epidemiologic
perspective. Paraplegia 1979;17:237–50.
[41] Reid DC, Henderson R, Saboe L, et al. Etiology and clinical course of missed spine
fractures. Journal of Trauma 1987;27:980–6.
[42] Knopp R, Parker J, Tashjian J, et al. Defining radiographic criteria for
flexion–extension studies of the cervical spine. Annals of Emergency Medicine
2001;38(1):31–5.
[43] Campbell SE, Phillips CD, Dubovsky E, et al. The value of CT in determin-
ing potential instability of simple wedge-compression fractures of the lumbar
spine. AJNR American Journal of Neuroradiology 1995;16:1385–92.
[44] Hogan GJ, Mirvis SE, Shanmuganathan K, et al. Exclusion of unstable cer-
vical spine injury in obtunded patients with blunt trauma: is MR imaging
needed when multi-detector row CT findings are normal? Radiology 2005;237:
106–13.
[45] Wadhwa R, Shamieh S, Haydel J, et al. The role of flexion and extension com-
puted tomography with reconstruction in clearing the cervical spine in trauma
patients: a pilot study. Journal of Neurosurgery: Spine 2011;14:341–7.
[46] Blackmore CC, Ramsay SD, Mann FA, et al. Cervical spine screening with
CT in trauma patients: a cost-effectiveness analysis. Radiology 1999;212:
117–25.
[47] Altay M, Ozkurt B, Nuri Aktekin C, et al. Treatment of unstable thoracolum-
bar junction burst fractures with short- or long-segment posterior fixation in
Magerl type A fractures. European Spine Journal 2007;16:1145–55.
[48] Daffner RH, Deeb ZL, Rothfus WE. The posterior vertebral body line: importance
in the prediction of burst fractures. AJR American Journal of Roentgenology
1987;148:93–6.
[49] Ballock RT, Mackersie R, Abitbol JJ, et al. Can burst fractures be predicted
from plain radiographs? The Journal of Bone and Joint Surgery British Volume
1992;74(B):147–50.
[50] Wintermark M, Mouhsine E, Theurmann N, et al. Thoracolumbar spine fractures
in patients who have sustained severe trauma: depiction with multi-detector
row CT. Radiology 2003;227:681–9.
[51] Meves R, Avanzi O. Correlation between neurological deficit and spinal canal
compromise in 198 patients with thoracolumbar and lumbar fractures. Spine
2005;30(7):787–91.
[52] Petersilge CA, Emery SE. Thoracolumbar burst fracture: evaluating stability.
Seminars in Ultrasound, CT and MR 1996;17:105–14.
[53] Knop C, Bastian L, Lange U, et al. Complications in surgical treatment of thora-
columbar injuries. European Spine Journal 2002;11(3):214–26.
[54] McCormack T, Karaikovic E, Gaines RW. The load sharing classification of spine
fractures. Spine 1994;19(15):1741–4.
[55] Jeffrey WP, Lane JR, Karaikovic EE, et al. Successful short-segment instru-
mentation and fusion for thoracolumbar spine fractures. Spine 2000;25(9):
1157–69.
[56] Furderer S, Wenda K, Thiem N, et al. Traumatic intervertebral disc lesion-
magnetic resonance imaging as a criterion for or against intervertebral fusion.
European Spine Journal 2001;10:154–63.
[57] Magerl F, Aebi M, Gertzbein SD, et al. A comprehensive classification of thoracic
and lumbar injuries. European Spine Journal 1994;3:184–201.
[58] James KS, Wenger KH, Schlegel JD, et al. Biomechanical evaluation of the sta-
bility of thoracolumbar burst fractures. Spine 1994;19:1731–40.
[59] Leferink VJM, Veldhuis EFM, Zimmerman KW, et al. Classificational problems
in ligamentary distraction type vertebral fractures: 30% of all B-type fractures
are initially unrecognised. European Spine Journal 2002;11:246–50.
[60] Haba H, Taqneiki H, Kotani I, et al. Diagnostic accuracy of magnetic reso-
nance imaging for detecting posterior ligamentous complex injury associated
with thoracic and lumbar fractures. Journal of Neurosurgery: Spine 2003;
99:20–6.
[61] Jeffrey AR, Fisher C, Harrop J, et al. Assessment of the posterior ligamentous
complex following acute cervical spine trauma. The Journal of Bone and Joint
Surgery American Volume 2010;92:583–9.
[62] Sliker CW, Mirvis S, Shanmuganathan K, et al. Assessing cervical spine stability
in obtunded blunt trauma patients: review of medical literature. Radiology
2005;234(3):733–9.
[63] Hammerberg KW. Surgical treatment of metastatic spine disease. Spine
1992;17:1148–55.
[64] Placantonakis DG, Laufer I, Wang JC, et al. Posterior stabilization strategies fol-
lowing resection of cervicothoracic junction tumors: review of 90 consecutive
cases. Journal of Neurosurgery: Spine 2008;9:111–9.
[65] Krishnaney AA, Steinmetz MP, Benzel EC. Biomechanics of metastatic spine
cancer. Neurosurgery Clinics of North America 2004;15:375–80.
[66] Taneichi H, Kaneda K, Takeda N, et al. Risk factors and probability of ver-
tebral body collapse in metastases of the thoracic and lumbar spine. Spine
1997;22:239–45.
[67] McGowan DP, Hipp JA, Takeuchi T, et al. Strength reductions from tra-
becular destruction within thoracic vertebrae. Journal of Spinal Disorders
1993;6:130–6.
[68] Dimar Jr JR, Voor MJ, Zhang YM, et al. A human cadaver model for determination
of pathologic fracture threshold resulting from tumorous destruction of the
vertebral body. Spine 1998;23:1209–14.