Sei sulla pagina 1di 7

Emerg Radiol DOI 10.1007/s10140-016-1460-8

Emerg Radiol DOI 10.1007/s10140-016-1460-8 ORIGINAL ARTICLE Acute traumatic intraosseous fluid sign predisposes to dynamic

ORIGINAL ARTICLE

ORIGINAL ARTICLE

Acute traumatic intraosseous fluid sign predisposes to dynamic fracture mobility

Troy A. Hutchins 1 & Richard H. Wiggins 1 & Jill M. Stein 1 & Lubdha M. Shah 1

Received: 5 September 2016 /Accepted: 24 October 2016 # American Society of Emergency Radiology 2016

Abstract The intraosseous fluid sign (IFS) in chronic os- teoporotic vertebral fractures is attributed to fluid accumu- lation within non-healing intervertebral clefts. IFS can also be seen in acute traumatic fractures, not previously de- scribed. We hypothesize a pathophysiological mechanism for the acute traumatic intraosseous fluid sign (ATIFS) and its predisposition to dynamic fracture mobility with axial loading on upright radiographs. Retrospective analysis was performed of 41 acute thoracic and lumbar compression or stable burst fractures with both supine CT and upright plain films completed within 1 week of each other. The presence of an intravertebral cleft wit h fluid attenuation and verte- bral body height loss was assessed on CT scans. Changes in the fractured vertebral body height and angulation were measured on upright radiographs. The ATIFS was identi- fied in 18 (44%) of the 41 acute fractures. Mean kyphotic angle increase was significantly greater ( p = 0.000) for ATIFS fractures (8.2°, SD ±4 .2) than fractures without ATIFS (1.6°, SD ±3.4). There was significantly greater mean anterior ( p = 0.0009) and central ( p = 0.026) height loss in ATIFS fractures (4.3 mm, SD ±3.76 and 1.89 mm, SD ±4.44, respectively) compared to fractures without ATIFS (0.59 mm, SD ±2.24 and 0.52 mm, SD ±2.01, respectively). The IFS can be seen in acute traumatic ver- tebral fractures and show dynamic mobility. These ATIFS fractures show statistically significant greater mean height loss ratio differences and have significantly greater

* Troy A. Hutchins troy.hutchins@hsc.utah.edu

1 Department of Radiology, University of Utah Health Sciences Center, 30 North, 1900 East, #1A071, Salt Lake City, UT 84132-2140, USA

changes in kyphotic angulation on upright imaging when compared to fractures without ATIFS.

Keywords Compression fracture . Burst fracture . Vertebral body . Spine

Introduction

Horizontal linear fluid signal intensity on magnetic resonance (MRI) within an osteoporotic vertebral fracture is known as the intraosseous fluid sign (IFS), which represents fluid accu- mulation within an intravertebral cleft [1]. Such intravertebral clefts within fractured osteoporotic vertebral bodies have been well-described [27] and are attributed to avascular necrosis with incomplete fracture healing [3, 8]. The etiology of fluid accumulation within an intravertebral cleft has been suggested to be chronic immobility [8], with movement of fluid into the cleft through negative pressure [4, 6, 8, 9]. The clinical implications of fluid within an intravertebral cleft remain under debate [7, 8, 10], with theories attributing the content of the cleft to the stage or severity of fracture [7, 10]. In patients with unhealed osteoporotic fractures and per- sistent pain, a fluid sign has been associated with increased dynamic mobility, higher patient pain scores, and worse per- formance status [8]. Prior studies have evaluated the fluid sign in the context of patients with osteoporotic fractures present- ing for vertebroplasty treatment [5, 11]. These studies have concluded that those patients with unhealed, painful, mobile osteoporotic compression fractures preferentially benefit from vertebroplasty [8]. In acute traumatic fractures, an intervertebral cleft can also accumulate fluid with the imaging appearance of the IFS, which has not previously been characterized to our knowl- edge. The purpose of this study was to determine whether

IFS, which has not previously been characterized to our knowl- edge. The purpose of this study

Emerg Radiol

such acute traumatic intraosseous fluid sign (ATIFS) fractures are also predisposed to dynamic fracture mobility.

Materials and methods

Subjects

An IRB-approved retrospective query of the radiology infor- mation system was performed at a level 1 trauma center for the term B fracture^ during a 3-year period from March 2009 to March 2012 for all CT studies of the thoracic or lumbar spine. Those patients with compression or burst fractures of the tho- racic or lumbar spine less than 1 week in age based on imaging and clinical history and those who also had comparison up- right radiographs performed within 1 week of the CT study were included. The electronic medical record was reviewed for each patients history, definitive treatment, and clinical follow-up. Chronic or indeterminate-age fractures by imaging and history and severe scoliosis limiting accurate measure- ments and imaging evaluation were excluded. Vertebral frac- tures requiring operative treatment for neurologic deficit or severe osseous deformity [12] were also excluded from our subject group.

Imagingdefinitions and criteria

All CT imaging was performed on a 64-slice scanner (Siemens AG, Erlangen, Germany) with helical acquisition and 0.6 mm collimation. All CT scans were reviewed in the axial, sagittal, and coronal planes in bone algorithm for the presence or absence of acute (less than 1 week) fracture, using features such as cortical disruption, irregular linear lucency without corticated margins, and height loss. The fractures were radiographically characterized accord- ing to the Denis classification of spinal fractures [13], where compression fractures involve only the anterior column and burst fractures involve the anterior and middle columns. Fractures involving the posterior column, including flexion- distraction and fracture dislocation injuries, were excluded based on the inherent instability predisposing such injuries for operative management due to mechanical instability and associated neurologic compromise. Also, excluded were pa- tients with imaging features suggestive of disruption of the posterior ligamentous complex with consequent vertebral col- umn instability, such as axial compression greater than 50% and kyphotic angle greater than 25° [12, 14, 15], which may indicate the need for operative stabilization. Fractures were then evaluated for the presence or absence of ATIFS on CT, defined as a linear area of fluid attenuation and disruption/ absence of trabeculation within a vertebral body fracture cleft [16]. The average Hounsfield unit of the IFS was measured using a region of interest (ROI) of 0.5 cm 2 on the sagittal and

of interest (ROI) of ≤ 0.5 cm 2 on the sagittal and coronal reconstructions, with care

coronal reconstructions, with care to exclude the adjacent bone. Average Hounsfield units of the L1 vertebral body (or a non-fractured adjacent vertebral body) was also measured in all patients with a standard ROI size of 0.5 cm 2 on sagittal reconstructions, in order to qu antify bone mineralization [1719]. All radiographic imaging was performed according to published standardized protocols, with upright anterior- posterior and lateral views obtained to include the affected vertebral body fracture.

Measurement of dynamic mobility

Vertebral body (VB) height and kyphotic angulation were calculated on supine sagittal CT and upright lateral radio- graphic images. The radiographic image was calibrated to a reference normal VB height on the corresponding supine CT. The anterior, central, and posterior VB heights were measured to the nearest millimeter for each fracture on both upright lateral radiograph and supine midline sagittal CT [ 8 , 11 ]. The mean dynamic height loss difference at each anterior, central, and posterior location was compared between cases with or without IFS. Posterior VB height of the adjacent nor- mal VB was also measured to the nearest millimeter. A height loss ratio of the minimum fractured vertebral body height (min) relative to the posterior height of the adjacent normal (NL) VB was calculated (1[min/NL posterior]) for both up- right radiograph and supine CT images. A dynamic height loss ratio difference was then calculated as the supine CT height loss ratio minus the upright radiograph height loss ratio [20]. Kyphotic angulation was measured from the superior and in- ferior endplates of the fractured VB on upright and supine images [21]. The change in kyphotic angle was calculated as the supine CT angle minus the upright radiograph angle.

Statistical analysis

For group comparisons, categorical variables were compared using a chi-square test, or Fisher s exact test, as appropriate. For continuous variables, the Students t test was used. All data was analyzed using two-tailed test and a P value of <0.05 was considered significant. All statistical analysis was performed using Stata 14 for Mac (Version 14.0, StataCorp, College Station, TX).

Results

The RIS query for B fracture^ of the thoracic or lumbar spine yielded 193 CT exams. Twenty-eight trauma patients (15 males, 13 females) (age range 1595 years, mean 55.07 (SD ±23.02)) meeting the inclusion criteria were found to have 41 acute thoracic or lumbar vertebral fractures. Seventeen (41%) thoracic and 24 (59%) lumbar spine fractures were identified.

Emerg Radiol

Table 1 Demographics of trauma patients with spine fractures. There were 41 fractures in 28 patients

 

IFS present

IFS absent

Total fractures

18

23

Age (mean, range) (year)

61.2 (2795)

49.7 (1578)

Gender

12 female, 6 male

8 female, 15 male

Osteoporosis present

202.4 SD±130.3

179.8 SD+60.1

(HU 135) Fracture level (number)

L1(6) L2,L3,T11(2)

L1(6) L2,L4(3) T6,T10,T12(2)

T7,T8,T10,T12,L4,L5(1) T1,T4,T7,T8,T11(1)

Fracture type

Compression (7), burst (11)

Compression (18), burst (5)

Treatment

Conservative only (7), +brace (9), +brace (1), kyphoplasty(1)

Conservative only (5), +brace (16), +brace (1)

Of the 41 fractures, 25 (61%) were classified as compression- type and 16 (39%) as burst-type. Underlying marrow pathol- ogy was evident in five patients with 12 (29%) vertebral frac- tures, of which five had ATIFS (Table 1). The ATIFS was identified on CT in 18 (44%) fractures and was more common in burst-type fractures (69%) compared to compression-type fractures (28%) (p = 0.010) (Fig. 1). There were 23 (56%) vertebral fractures without ATIFS (Fig. 2). The Hounsfield unit measured in the L1 VB (or adjacent non-fractured VB) ranged 51 to 416 with mean 189.7 HU, SD ±96.6 [18]. Although this includes the five patients with the 12 pathologic fractures, there was also no significant difference in the bone densities of those patients with ATIFS fractures and those without the imaging finding (p = 0.465). There was no significant difference between the CT and radiographic measurements of the posterior height of a normal adjacent VB (mean difference 0.0317 mm, SD 0.378). Those vertebral fractures with ATIFS had significantly greater dynamic anterior ( p = 0.0009) and central (p = 0.026) height loss on upright imaging relative to supine imaging. Mean anterior height loss was 4.3 mm (SD ±3.76) for ATIFS fractures, compared to 0.59 mm (SD ±2.24) for fractures without ATIFS. Similarly, mean central height loss was 1.89 mm (SD ±4.44) for ATIFS fractures, and mean 0.52 mm (SD ±2.01) for fractures without ATIFS. The pos- terior VB dynamic height loss was not significantly different in fractures with or without ATIFS (p = 0.8514); mean height loss was 0.18 mm (SD ±2.11) for ATIFS fractures and 0.30 mm (SD ±2.02) without ATIFS (Table 2). There was a statistically significant greater dynamic mean height loss ratio difference in ATIFS fractures (p = 0.021), which was best demonstrated on upright radiographs as compared to supine imaging (p = 0.0095) (Table 3). Baseline mean kyphotic angle for all fractures was 6.2 (SD ±7.6) degrees on supine CT and 10.7 (SD ±7.2) degrees on upright radiographs. The dynamic change in kyphotic angle was greater in ATIFS fractures (p = 0.000). On upright relative to supine imaging, the mean increase in kyphotic angle for

ATIFS fractures was 8.2° (SD ±4.2) and 1.6° (SD ±3.4) in fractures without ATIFS.

Imaging and clinical follow-up

Twenty-six fractures had imaging follow-up for up to18 months, 11 of which were unchanged. Fifteen frac- tures (58%) displayed mild interim in height loss. Twenty-seven of the fractures had clinical follow-up, 18 (67%) of which were correlated with decreased pain. Of these 18 fractures, only 4 fractures (22%) displayed ATIFS. There was significant correlation between pain and ATIFS in our trauma population ( p = 0.04) at up to 12 months clinical follow-u p. All of these fractures were treated non-operatively, with the exception of two fractures in one elderly patient who underwent kyphoplasty for per- sistent pain. Notably, this patient had low bone mineral

sistent pain. Notably, this patient had low bone mineral Fig. 1 a Sagittal CT reconstruction of

Fig. 1 a Sagittal CT reconstruction of the lumbar spine, obtained with the patient in supine position, demonstrates a two-column fracture of the L3 vertebral body with an intraosseous fluid sign (arrow). The height of the vertebral body is minimally decreased in comparison to the adjacent normal vertebral levels on this image. b Sagittal STIR image with the same patient in supine position shows hyperintense signal throughout the L3 vertebral body marrow in keeping with edema. Focal fluid signal is noted in the intravertebral cleft (arrow),corresponding the low attenuating collection on the CT. c Upright lateral radiograph reveals significant height loss of the L3 vertebral body as compared to the supine imaging modalities ( arrow ). The upright axial loading has compressed the serosanguineous fluid in the posttr aumatic intravertebral cleft with consequent dynamic fracture mobility

compressed the serosanguineous fluid in the posttr aumatic intravertebral cleft with consequent dynamic fracture mobility

Emerg Radiol

Emerg Radiol Fig. 2 a Sagittal CT reconstruction of the lumbar spine, obtained with the patient

Fig. 2 a Sagittal CT reconstruction of the lumbar spine, obtained with the patient in supine position, demonstrates an anterior column compression fracture of the L1 vertebral body without intravertebral fluid attenuation to suggest an intraosseous fluid sign (arrow). b Lateral upright radiograph shows preservation of the L1 vertebral body height ( arrow ) and no dynamic fracture mobility with axial loading related to upright positioning

density and two ATIFS fractures; one of which also displayed gas and therefore was compatible with an acute on chronic fracture. In this patient, the dynamic vertebral height loss ratio measured 20.5% with 10° angulation. Seven of the 27 fractures (26%) had confounding sources of back pain, which were not directly attributable to the fracture by clinical assessment.

Discussion

The trauma population with acute vertebral fractures in our study is distinct from the population of unhealed osteoporotic vertebral fractures with intraosseous fluid sign reported previ- ously [5, 8, 11]. In our patient cohort with acute compression or burst fractures, 44% had acute traumatic intraosseous fluid sign (ATIFS). Patients with ATIFS fractures showed statisti- cally significant greater mean height loss ratio difference and significantly greater increased kyphotic angulation on upright imaging compared to fractures without ATIFS. There was no significant difference in bone mineral density between ATIFS and non-ATIFS fractures. The ATIFS in our patients was more common in burst fractures, in keeping with the findings of

Baur et al., who demonstrated that the fluid sign is correlated with an increased severity of vertebral body fracture [1]. As in our cases, it was also observed that a fluid sign occurs where there is the most severe compression of the spongiosa at the fractured endplate [1]. Previous studies of the IFS are in the context of subacute to chronic osteoporotic compression fractures with fluid accu- mulating within a non-healing cleft. In our patient population in the setting of acute trauma, however, there is no preexisting cleft within the vertebral body to account for the fluid accu- mulation. Therefore, the fluid-filled cleft is thought to occur at the time of the injury. We hypothesize a traumatic mechanism of axial compression and flexion followed by equal or greater hyperextension. During flexion/compression, the vertebral body fractures and loses height. The fracture results in weak- ened medullary bone near the vertebral endplate with a hori- zontal linear band of impacted trabeculae; this is in keeping with the proposed mechanism by Baur et al [1]. The anterior longitudinal ligament adjacent to this zone of impacted bone buckles anteriorly. In our proposed mechanism, there is sub- sequent hyperextension with a slight distraction component, during which the weakened linear band of fractured bone sep- arates as the vertebral body regains some of its original height and the anterior longitudinal ligament pulls taut but does not rupture. This creates a horizontal linear cavity that then fills with fluid and/or hemorrhage, seen on imaging as the ATIFS (Fig. 3). The distraction component allows the serosanguinous fluid to fill the cleft between the disrupted trabeculae. A pre- dominant compression mechanism only would allow edema to disperse into the marrow [22]. Studies have shown that there is some degree of B settling^ and increasing kyphotic angulation over time with simple compression fractures or stable burst fractures and that this does not correlate with pain [23]. Similarly, greater than half of the fractures with imaging follow-up in our series showed mild progression of height loss without associated pain. An increase in fracture kyphotic angle greater than 10° on upright relative to supine positioning has been suggested as an indi- cation for operative intervention [24]. While the ATIFS was associated with greater vertebral height loss and angulation compared to those fractures without the ATIFS in our patients, the mean increased angulation in ATIFS fractures was less than this aforementioned 10-degree threshold. The dynamic

Table 2 Mean, standard deviation (SD), and range for the anterior, central, and posterior vertebral body (VB) heights (all fractures grouped), and the posterior height of a normal adjacent VB, on upright and supine images

Supine

(mm mean, SD, and range)

Upright

(mm mean, SD, and range)

Anterior

21.1±5.7 (5.530.9)

18.9±5.7 (7.733.2)

Center

17.9±4.4 (8.325.3)

17.4±5.2 (5.626.7)

Posterior

25.3±4.3 (14.833.3)

25.0±4.8 (14.535.8)

Posterior normal VB

26.3±4.4 (18.733.9)

26.3±4.2 (18.433.5)

– 33.3) 25.0±4.8 (14.5 – 35.8) Posterior normal VB 26.3±4.4 (18.7 – 33.9) 26.3±4.2 (18.4 –

Emerg Radiol

Table 3 Height loss on supine and upright imaging and dynamic height loss

 

Supine

Upright

mean, SD, (95% CI)

mean, SD, (95% CI)

Ratio min/normal

0.33±0.15 (0.290.38)

0.37±0.16 (0.320.42)

Ratio min/normal with IFS

0.35±0.17 (0.270.43)

0.44±0.17 (0.350.53)

Ratio min/normal without IFS

0.32±0.13 (0.260.38)

0.31±0.13 (0.250.37)

Dynamic mean height loss ratio difference with IFS

0.91±0.18

Dynamic mean height loss ratio difference without IFS

0.10±0.90

mobility of the ATIFS fracture with upright imaging can be predicted, and our results show that it not a sign of progression or instability. The observation that the IFS correlates with increased dy- namic mobility in the setting of traumatic fractures is consis- tent with prior observations that osteoporotic fractures with intravertebral clefts display abnormal mobility in patients pre- senting for vertebroplasty [5, 8, 14]. Dynamic mobility has been attributed to increased pain and decreased performance status in prior investigations [8]. We found a significant asso- ciation between the ATIFS and persistent pain on follow-up. Studies have concluded that those patients with unhealed, painful, mobile osteoporotic fractures preferentially benefit from vertebroplasty [8]. The patient with two ATIFS fractures and persistent pain shared similar demographics with those patients shown to benefit from percutaneous cement augmen- tation [14, 15]. Whether acute or chronic, the IFS represents a fluid- filled horizontal linear intrao sseous cavity. Therefore, ATIFS and osteoporotic IFS will behave similarly and

ATIFS and osteoporotic IFS will behave similarly and Fig. 3 Graphic images above and representative CT

Fig. 3 Graphic images above and representative CT images below for hypothesized ATIFS mechanism. Normal vertebral body ( a, e). With flexion-compression, there is vertebral body fracture and loss of height with linear band of impacted trabeculae (b, f) and buckling of the anterior longitudinal ligament (yellow arrow). With subsequent hyperextension, the weakened fractured bone separates while the anterior longitudinal ligament pulls taut, creating a horizontal linear cavity that then fills with fluid and/or hemorrhage (c, g)

show a loss of height when imaging is performed with weight-bearing. However, when an acute fracture occurs, the bone surrounding the intraosseous cavity is more likely to be viable relative to the bone surrounding a chronic frac- ture with a non-healing IFS cleft. It is therefore possible that despite the similar appearances, ATIFS fractures may behave differently than IFS fractures with respect to subse- quent healing and pain on follow-up. The majority of our patients with clinical follow- up (67%) had decreased or resolved pain attributable to the fracture. This may be due to differences in the pathophysiology of the IFS in acute traumatic fractures as oppos ed to unhealed osteoporotic fractures. The clinical medical record was the reference for determining the improvement or persistence of pain, particularly whether it was directly attributable to the trau- matic fracture(s). Given the multifactorial causes of back pain, this can be challenging to filter. Although a few of our patients had underlying marrow pathology, which may introduce heterogeneity in the imaging appearance of the ATIFS, there was no significant difference in bone densities of those patients with ATIFS and those with- out the imaging finding (p = 0.465). Tumor cell-filled marrow reportedly makes IFS rare in malignant fractures (6% of neo- plastic fractures in the study by Baur et al. [1]). However, it is important to be aware that IFS is not a definitive imaging sign of benignity. In our two patients (five ATIFS fractures) with underlying bone marrow pathology, the traumatic etiology of the fractures may have contributed to the greater percentage of ATIFS (42%). Marrow abnormality, whether due to osteopo- rosis or tumor, can weaken the vertebral column with vulner- ability to fracture, which can demonstrate ATIFS. An important point to consider in our methodology is the use of the dynamic height loss ratio as it allowed for an internal comparison. A direct comparison of values by different modalities (i.e., supine CT and upright ra- diographs) is limited given issues with magnification, parallax, etc. This is reflected by the absolute numbers that were measured for anterior, central, and posterior height loss on supine CT versus upright radiograph im- ages, some of which were negative values. For example, the mean B increased ^ height in the center of the vertebral body with upright positioning as compared to the supine positioning for fractures without ATIFS ( 0.52 mm, SD

body with upright positioning as compared to the supine positioning for fractures without ATIFS ( −

Emerg Radiol

±2.01) is thought to be due to parallax. A more accurate delineation of the cortical margins is possible on CT. Although there was no significant difference between the CT and radiographic measurements of the posterior height of a normal adjacent vertebral body, the center of the fractured vertebral b ody may have been positioned caudal to the x-ray beam center, with relative beam di- vergence distorting radiographic landmarks. This paral- lax effect may contribute to the difficulty determining the cortical margin of the center of the fractured verte- bral body. This was less problematic for the anterior and posterior margins. Of note, the parallax can be exagger- ated in large patients [ 25 ]. Little is known about the evolution of ATIFS on im- aging, but as stated earlier, the IFS is evident in subacute and chronic osteoporotic fractures. Although the major- ity of upright imaging was performed within 1 2 days of the trauma (four patients with eight fractures had upright radiographics greater than 1 day but within 1 week), the clinical scenario (e.g., pain , co-morbidities, transfer from outside institution, etc.) determined the timing of the upright radiographs. Bone marrow edema in vertebral compression fractures on MRI, which is more sensitive than CT, resolves after 1 3 months [ 26 ] but can persist for longer than 3 months [ 27 ]. Therefore, it may be hy- pothesized that the marrow edema and fluid within the intravertebral cleft would be present at least 1 week. MRI can be helpful in detecting the IFS, and although MRI examinations were available in 14 fractures, we did not choose to include that imaging modality in this study. At our institution as at many ot hers, MRI is reserved for those patients with concern for neurological injury or those that have additional imaging findings suggestive of mechanical instability. MRI has a role in the detection of ligamentous injury, modifying the fracture classifica- tion and therefore treatment approach [ 28 ]. However, recent literature is controversial regarding the utility of MRI in the acute trauma setting with less sensitivity and specificity for the integrity of the posterior ligamentous complex than previously reported [ 23 ] and marrow ede- ma [ 22 , 29 ]. As CT is the standard for evaluation of acute spine trauma, the focus of this study is on the more com- monly encountered manifestation of ATIFS on CT and radiographs. Further investigations will be helpful to evaluate the sensitivity of MRI for detection of ATIFS in comparison to CT. A recognized limitation of this study is the relatively small sample size of fractures from a single institution with limited clinical follow-up in some patients, as is common in the trauma population. Future studies are necessary to determine the clinical significance of the ATIFS with respect to fracture healing, pain on follow- up, and optimal treatment.

fracture healing, pain on follow- up, and optimal treatment. Conclusion The ATIFS was observed in just

Conclusion

The ATIFS was observed in just under half of the acute traumatic vertebral fractures i n our study. It was associated with both statistically significant dynamic vertebral body height loss and increased angulation, paralleling findings previously described in IFS osteoporotic fractures. It is important to recognize that the dynamic changes of ATIFS fractures on upright imaging are expected and should not be misinterpreted as indicating fracture progres- sion or instability. The mecha nism of these ATIFS fractures may be distinct from the osteoporotic fractures with IFS and dynamic mobility. Future studies are necessary to de- termine the clinical significance of the ATIFS with respect to fracture healing, pain on follow-up, and optimal treatment.

Compliance with ethical standards

Conflict of interest The authors declare that they have no conflict of interest.

References

1. Baur A, Stabler A, Arbogast S, Duerr HR, Bartl R, Reiser M (2002) Acute osteoporotic and neoplastic vertebral com- pression fractures: fluid sign at MR imaging. Radiology 225(3):730 735. doi: 10.1148/radiol.2253011413

2. Bhalla S, Reinus WR (1998) The linear intravertebral vacuum: a sign of benign vertebral collapse. AJR Am J Roentgenol 170(6):

15631569. doi:10.2214/ajr.170.6.9609175

3. Dupuy DE, Palmer WE, Rosenthal DI (1996) Vertebral fluid col- lection associated with vertebral collapse. AJR Am J Roentgenol 167(6):15351538. doi:10.2214/ajr.167.6.8956592

4. Malghem J, Maldague B, Labaisse MA, Dooms G, Duprez T, Devogelaer JP, Vande Berg B (1993) Intravertebral vacuum cleft:

changes in content after supine positioning. Radiology 187(2):483487. doi:10.1148/radiology.187.2.8475295

5. Mirovsky Y, Anekstein Y, Shalmon E, Peer A (2005) Vacuum clefts of the vertebral bodies. AJNR Am J Neuroradiol 26(7):16341640

6. Sarli M, Perez Manghi FC, Gallo R, Zanchetta JR (2005) The vacuum cleft sign: an uncommon radiological sign. Osteoporosis international: a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 16(10):1210 1214.

7. Yu CW, Hsu CY, Shih TT, Chen BB, Fu CJ (2007) Vertebral osteonecrosis: MR imaging findings and related changes on adja- cent levels. AJNR Am J Neuroradiol 28(1):4247

8. Kawaguchi S, Horigome K, Yajima H, Oda T, Kii Y, Ida K, Yoshimoto M, Iba K, Takebayashi T, Yamashita T (2010) Symptomatic relevance of intravertebral cleft in patients with oste- oporotic vertebral fracture. J Neurosurg Spine 13(2):267 275.

9. Linn J, Birkenmaier C, Hoffmann RT, Reiser M, Baur-Melnyk A (2009) The intravertebral cleft in acute osteoporotic fractures: fluid in magnetic resonance imaging-vacuum in computed tomography? Spine 34(2):E88E93. doi:10.1097/BRS.0b013e318193ca06

Emerg Radiol

10. Jang JS, Kim DY, Lee SH (2003) Efficacy of percutaneous vertebroplasty in the treatment of intravertebral pseudarthrosis as- sociated with noninfected avascular necrosis of the vertebral body. Spine 28(14):15881592

11. McKiernan F, Faciszewski T (2003) Intravertebral clefts in osteo- porotic vertebral compression fractures. Arthritis Rheum 48(5):

14141419. doi:10.1002/art.10984

12. Alexandru D, So W (2012) Evaluation and management of verte- bral compression fractures. Permanente J 16(4):4651

13. Denis F (1983) The three column spine and its significance in the classification of acute thoracolumbar spinal injuries. Spine 8(8):

817831

14. Wood KB, Li W, Lebl DS, Ploumis A (2014) Management of thoracolumbar spine fractures. Spine J: Off J North Am Spine Soc 14(1):145164. doi:10.1016/j.spinee.2012.10.041

15. Ghobrial GM, Jallo J (2013) Thoracolumbar spine trauma: review of the evidence. J Neurosurg Sci 57(2):115122

16. Wiggins MC, Sehizadeh M, Pilgram TK, Gilula LA (2007) Importance of intravertebral fracture clefts in vertebroplasty outcome. AJR Am J Roentgenol 188(3):634 640. doi: 10.2214/AJR.06.0542

17. Pickhardt PJ, Pooler BD, Lauder T, del Rio AM, Bruce RJ, Binkley N (2013) Opportunistic screening for osteoporosis using abdominal computed tomography scans obtained for other indications. Ann Intern Med 158(8):588 595. doi: 10.7326/0003-4819-158-8-201304160-00003

18. Schreiber JJ, Anderson PA, Rosas HG, Buchholz AL, Au AG (2011) Hounsfield units for assessing bone mineral density and strength: a tool for osteoporosis management. J Bone Joint Surg Am 93(11):10571063. doi:10.2106/JBJS.J.00160

19. Schreiber JJ, Anderson PA, Hsu WK (2014) Use of computed to- mography for assessing bone mineral density. Neurosurg Focus 37(1):E4. doi:10.3171/2014.5.FOCUS1483

20. Teng MM, Wei CJ, Wei LC, Luo CB, Lirng JF, Chang FC, Liu CL, Chang CY (2003) Kyphosis correction and height restoration ef- fects of percutaneous vertebroplasty. AJNR Am J Neuroradiol

24(9):18931900

21. Alanay A, Pekmezci M, Kara eminogullari O, Acaroglu E, Yazici M, Cil A, Pijnenburg B, Genc Y, Oner FC (2007)

Radiographic measurement of t he sagittal plane deformity

in patients with osteoporotic spinal fractures evaluation of

intrinsic error. Eur Spine J: Off Publ Eur Spine Soc Eur

Spinal Deformity Soc Eur Section Cervical Spine Res Soc 16(12):2126 2132. doi: 10.1007/s00586-007-0474-z

22. Brinckman MA, Chau C, Ross JS (2015) Marrow edema variability

in acute spine fractures. Spine J: Off J North Am Spine Soc 15(3):

23. Wood K, Buttermann G, Mehbod A, Garvey T, Jhanjee R, Sechriest

V (2003) Operative compared with nonoperative treatment of a

thoracolumbar burst fracture without neurological deficit. A pro- spective, randomized study. J Bone Joint Surg Am 85-A(5):773

781

24. Mehta JS, Reed MR, McVie JL, Sanderson PL (2004) Weight- bearing radiographs in thoracolumbar fractures: do they influence management? Spine 29(5):564567

25. Auerbach JD, Namdari S, Milby AH, White AP, Reddy SC, Lonner BS, Balderston RA (2008) The parallax effect in the evaluation of range of motion in lumbar total disc replacement. SAS J 2(4):184188. doi:10.1016/SASJ-2008-0020-RR

26. Baker LL, Goodman SB, Perkash I, Lane B, Enzmann DR (1990) Benign versus pathologic compression fractures of vertebral bodies: assessment with conventional spin-echo, chemical-shift, and STIR MR imaging. Radiology 174(2):

27. Voormolen MH, van Rooij WJ, van der Graaf Y, Lohle PN, Lampmann LE, Juttmann JR, Sluzewski M (2006) Bone marrow edema in osteoporotic vertebral compression fractures after percu- taneous vertebroplasty and relation with clinical outcome. AJNR Am J Neuroradiol 27(5):983988

28. Winklhofer S, Thekkumthala-Sommer M, Schmidt D, Rufibach K, Werner CM, Wanner GA, Alkadhi H, Hodler J, Andreisek G (2013) Magnetic resonance imaging frequently changes classification of acute traumatic thoracolumbar spine injuries. Skelet Radiol 42(6):

779786. doi:10.1007/s00256-012-1551-x

29. Lensing FD, Bisson EF, Wiggins RH 3rd, Shah LM (2014) Reliability of the STIR sequence for acute type II odontoid frac- tures. AJNR Am J Neuroradiol 35(8):16421646. doi:10.3174/ajnr.

STIR sequence for acute type II odontoid frac- tures. AJNR Am J Neuroradiol 35(8):1642 – 1646.