Correlation Between Vacuum Phenomenon on CT and Fluid on MRI in Degenerative Disks

Transcription

1 Musculoskeletal Imaging Original Research D Anastasi et al. Vacuum Phenomenon on CT and Fluid on MRI of Degenerative Disks Musculoskeletal Imaging Original Research Melvin D Anastasi 1 Christof Birkenmaier 2 Gerwin P. Schmidt 1 Bernd Wegener 2 Maximilian F. Reiser 1 Andrea Baur-Melnyk 1 D Anastasi M, Birkenmaier C, Schmidt GP, Wegener B, Reiser MF, Baur-Melnyk A Keywords: intervertebral disk degeneration, MRI, spinal osteochondrosis, spine DOI: /AJR Received December 11, 2010; accepted after revision April 8, Institute of Clinical Radiology, Klinikum Grosshadern, Ludwig-Maximilians-Universität, Marchioninistrasse 15, Munich 81377, Germany. Address correspondence to M. D Anastasi (melvin.danastasi@med.uni-muenchen.de). 2 Department of Orthopedic Surgery, Klinikum Grosshadern, Ludwig-Maximilians-Universität, Munich, Germany. AJR 2011; 197: X/11/ American Roentgen Ray Society Correlation Between Vacuum Phenomenon on CT and Fluid on MRI in Degenerative Disks OBJECTIVE. The purpose of this study is to correlate the presence of intradiscal vacuum phenomenon on CT to that of intradiscal fluid on MRI. SUBJECTS AND METHODS. In a prospective study, 20 patients with lumbar vacuum phenomenon on CT underwent two MRI examinations. One was performed after mobilization, and the other was performed after 6 hours of bed rest. T2-weighted turbo-spin echo (TSE), STIR, T1-weighted TSE, and four consecutive T2-weighted TSE sequences were performed on a 1.5-T scanner. Ninety-five discal segments were assessed for the presence of intradiscal fluid or hyperintense signal on the T2-weighted MRI examinations and were correlated with the presence of vacuum phenomenon on CT, degenerative endplate abnormalities on CT, and edema on MRI. RESULTS. Sixty-nine of 95 discal segments (72.6%) showed vacuum phenomenon on CT. Sixteen of those 69 discal segments (23.1%) showed intradiscal fluid (n = 12) or hyperintense signal (n = 4) on MRI examinations performed after mobilization. Forty-one of 69 discal segments (59.4%) with vacuum phenomenon showed intradiscal fluid (n = 29) or hyperintense signal (n = 12) on MRI examinations performed after bed rest. Seventeen segments showed only fluid after bed rest. Nine segments showed more fluid after bed rest than after mobilization. Three segments showed an unchanged amount of fluid. There was a significant correlation between the presence of intradiscal fluid and the amount of bone marrow edema on MRI and the presence of degenerative endplate abnormalities on CT, respectively. CONCLUSION. The replacement of intradiscal vacuum phenomenon by intradiscal fluid is a time- and position-dependent dynamic process and is related to Modic type 1 degenerative disk disease and degenerative endplate changes. I ntervertebral vacuum phenomenon is a common finding on CT examinations of degenerative disk disease. The presence of intradiscal fluid on MRI is rare in degenerative disk disease and primarily raises the suspicion of spondylodiscitis. However, in some cases, elevated signal or fluidlike signal can be present because of fissures in the degenerated disks [1]. Various studies have suggested that there is a dynamic physical continuum between vacuum phenomenon, gas, and fluid [2 5]. This continuum depends on the duration and intensity of spontaneous or stress-induced opening of an anatomic cavity, the surrounding structures, and perfusion. We aimed to correlate intradiscal vacuum phenomenon on CT to the signal intensities and changes in signal intensities in dynamic serial MRI measurements of the lumbar spine. In addition, we aimed to examine whether there is a correlation between the presence of intradiscal fluid and the presence of degenerative changes of the adjacent vertebral endplates (i.e., roughening and fissuring) on CT and the amount of adjacent bone marrow edema, respectively. Subjects and Methods Approval from the ethics commission to conduct this study was obtained. The study was performed in accordance with the Declaration of Helsinki. Patients Twenty patients (15 women and five men; mean age, 71 years; age range, years) from the orthopedic department, who had undergone CT of the lumbar spine at our institution during their inpatient stay between November 2008 and November 2010, were enrolled in this study. Patients gave their oral and written informed consent. Inclusion criteria were the presence of an intervertebral vacuum 1182 AJR:197, November 2011

2 Vacuum Phenomenon on CT and Fluid on MRI of Degenerative Disks phenomenon on CT at one or more disk levels. Recruitment of patients for the study was limited to inpatients in the orthopedic wards. Exclusion Criteria Patients with acute vertebral fractures, known or suspected spondylodiscitis, recent spinal surgery (< 4 months), a history of malignant tumors, as well as patients who were clinically unstable or with any known contraindication to MRI were excluded. A total of 40 patients had to be excluded, as follows: five patients had a not previously known contraindication to MRI (e.g., shrapnel and some types of metallic stents), 17 patients did not wish to participate in the study, 10 patients claimed they did not feel able to walk for the required 10 minutes before the MRI examination after mobilization, three patients were unable to complete the MRI examination because of discomfort (e.g., claustrophobia), and five patients who initially agreed to participate in the study did not wish to continue doing so after the first MRI examination. CT CT scans were performed on a 64-MDCT scanner (Somatom Sensation 64, Siemens Healthcare). The following parameters were used: effective tube current, 300 ma; 120 kv; pitch, 0.9; and collimation, mm and mm. Axial and sagittal multiplanar images were reconstructed with a slice thickness of 2 mm. The CT datasets were assessed by two experienced musculoskeletal radiologists in a consensus reading for the presence and location of osteochondrosis and intervertebral vacuum phenomenon. Each vertebral motion segment was also assessed for the presence of degenerative abnormalities of the endplates (i.e., roughening and fissuring) and their location. A vertebral motion segment can be defined as the smallest functional unit of the spine and consists of two vertebrae and the interconnecting soft tissue [6]. MRI The subjects then underwent two separate MRI examinations with identical sequences, prospectively, on a 1.5-T system (Avanto, Siemens Healthcare). The following sagittal sequences were performed: T2-weighted turbo-spin echo (TSE) (TR/TE, 4960/97; slice thickness, 3.5 mm; and FOV, 320 mm), STIR (TR/TE, 3640/67; slice thickness, 3.5 mm; and FOV, 320 mm), T1- weighted TSE (TR/TE, 508/9.6; slice thickness, 3.5 mm; and FOV, 320 mm), and four consecutive T2-weighted TSE sequences (TR/TE, 4960/97; slice thickness, 3.5 mm; and FOV, 320 mm). Total scanning time was 19 minutes 57 seconds. A Fig year-old woman with osteochondrosis and vacuum phenomenon in multiple discal segments. A, Sagittal CT of lumbar spine after mobilization shows intervertebral vacuum phenomena and degenerative abnormalities of endplates at several levels. B, Image shows sagittal T1-weighted MRI of lumbar spine obtained after mobilization. C, Sagittal STIR image shows vertebral bone marrow edema at several levels, particularly T12 L1, L2 3, and L3 4. (Fig. 1 continues on next page) B C AJR:197, November

3 D Anastasi et al. The two MRI examinations were performed within 3 days of the CT examination during the inpatient stay. The first examination was performed in the evening after mobilization. On the day of this examination, the patients were encouraged to undertake their normal daily activities in the ward. Just before the examination, patients walked along the corridor adjacent to the MRI waiting area together with the author performing the examination for 10 minutes. After the examination, the patients returned to the ward where they were asked to lie supine in bed for 6 hours before the second examination in the early morning. The patients were also asked not to go to the bathroom during this time and were provided with urine bottles or bowls. In the early morning (before breakfast), the patients were transported in bed from the ward to our MRI section. Each patient was transferred in turn from the bed to a nonmagnetic D Fig. 1 (continued) 71-year-old woman with osteochondrosis and vacuum phenomenon in multiple discal segments. D, Sagittal T2-weighted image shows discrete hyperintense signal (arrow) at T12 L1. E and F, STIR (E) and T2-weighted (F) images obtained after 6 hours of bed rest now show fluid-isointense signal at T12 L1 and L2 3 disk levels (arrowheads, F). MRI stretcher and from the stretcher to the scanner table without standing up. The second MRI examination was then performed. The MRI scans were assessed by the same readers for the presence of intradiscal fluid or intradiscal hyperintense signal, and their respective locations were recorded. Intradiscal fluid was diagnosed as such if its signal intensity on T2-weighted images was equivalent to that of CSF. An intradiscal hyperintense signal was described where the signal intensity of the disk on T2-weighted images was higher than that of other normal disks, but not equivalent to the signal intensity of CSF. This was correlated to the location of intervertebral vacuum phenomenon on CT. The two MRI examinations were read in the same way and were compared to assess differences in the presence and amount of intradiscal fluid or hyperintense signal after mobilization and after bed rest. Each vertebral E motion segment was assessed for the presence and amount of bone marrow edema along the endplates (Modic type 1 changes). If bone marrow edema was present, the maximum vertical height of bone marrow hyperintense signal on STIR for each of two vertebrae adjacent to a discal segment was measured and calculated as a percentage of the total vertebral height. The percentage values for the two vertebrae were added, the resulting value (of a maximum of 200) being an indicator of the amount of bone marrow edema in the vertebrae adjacent to a discal segment. A total of 95 vertebral motion segments were evaluated. Statistical Analysis All patient and image interpretation data were collected in a database and analyzed with statistical data collection software (Excel, Microsoft). A score representing the amount of bone marrow edema for each vertebral motion F 1184 AJR:197, November 2011

4 Vacuum Phenomenon on CT and Fluid on MRI of Degenerative Disks segment was calculated as described previously in this article. The presence and amount of bone marrow edema and the presence of degenerative abnormalities of the endplates on CT were correlated to the presence of intradiscal fluid or hyperintense signal and were assessed for statistical significance by means of logistic regression models with random intercepts using SAS-Procedure PROC GLIMMIX (SAS version 9.2 for Windows, SAS Institute). A Results Sixty-nine of 95 (72.6%) evaluated discal segments showed an intradiscal vacuum phenomenon. The discal segments with vacuum phenomenon were distributed as follows: T12 L1 (n = 6), L1 2 (n = 8), L2 3 (n = 11), L3 4 (n = 13), L4 5 (n = 16), and L5 S1 (n = 15). Fifty-one segments (53.7%) showed degenerative abnormalities of the adjacent endplates. After mobilization, 16 of 69 segments with vacuum phenomenon showed a fluidisointense signal (n = 12) or hyperintense signal (n = 4). After bed rest, the presence of fluid and hyperintense signal increased significantly (p < ). Forty-one of 69 (59.4%) segments with vacuum phenomenon on CT showed intradiscal fluid (29/41) or hyperintense signal (12/41) (Figs. 1 and 2). There was no fluid signal in a disk level without intervertebral vacuum phenomenon on CT. The remaining discal segments (without fluid signal or hyperintense signal) showed normal disk signal intensity (n = 15) or dehydrated degenerative disk disease (n = 39). Seventeen discal segments showed fluid only after bed rest. Nine segments showed more fluid after bed rest when compared with the examination performed after mobilization. Three segments had an unchanged amount of fluid on both examinations. Of the 12 segments with hyperintense signal after bed rest, nine (75%) showed a hyperintense signal exclusively after bed rest, whereas three segments showed a stronger hyperintense signal after bed rest than after mobilization (Fig. 2). Fig year-old woman with osteochondrosis and vacuum phenomenon of lumbar spine. A, Sagittal CT image of lumbar spine obtained after mobilization shows vacuum phenomena in L2 3, L3 4, mildly in L4 5, and L5 S1 disks as well as degenerative abnormalities of vertebral endplates at L2 3 and L3 4 levels. B D, T1-weighted (B), STIR (C), and T2-weighted (D) images show Modic type 2 changes of L2 3 and L3 4 endplates. T2-weighted image (D) shows no hyperintense signal or fluid in discal segments of L2 3 and L3 4 levels (arrowheads). E and F, MRI was performed after 6 hours of bed rest. Sagittal STIR (E) and T2-weighted (F) images show now marked hyperintense signal in discal spaces L2 3 and L3 4 (arrowheads, E and F). (Fig. 2 continues on next page) B C AJR:197, November

5 D Anastasi et al. During the time course of the MRI examination after mobilization, there were five discal segments in three patients that showed a very discrete amount of fluid in the first T2- weighted sequence but then exhibited a significant increase in the second T2-weighted sequence 9 minutes after commencement of the examination. This was in turn followed by a slow progressive increase in fluid signal until the last T2-weighted sequence. There were two discal segments in two patients that showed no fluid signal on the first T2-weighted sequence but showed a fluid signal on the second T2-weighted sequence 9 minutes after commencement of the examination, followed by a slow progressive increase in discal fluid until the last sequence (Fig. 3). D Fig. 2 (continued) 77-year-old woman with osteochondrosis and vacuum phenomenon of lumbar spine. A, Sagittal CT image of lumbar spine obtained after mobilization shows vacuum phenomena in L2 3, L3 4, mildly in L4 5, and L5 S1 disks as well as degenerative abnormalities of vertebral endplates at L2 3 and L3 4 levels. B D, T1-weighted (B), STIR (C), and T2-weighted (D) images show Modic type 2 changes of L2 3 and L3 4 endplates. T2-weighted image (D) shows no hyperintense signal or fluid in discal segments of L2 3 and L3 4 levels (arrowheads). E and F, MRI was performed after 6 hours of bed rest. Sagittal STIR (E) and T2-weighted (F) images show now marked hyperintense signal in discal spaces L2 3 and L3 4 (arrowheads, E and F). In contrast, when the same patients underwent the MRI examination after bed rest (without prior mobilization), there was no difference in fluid signal or hyperintense signal during the time course of the examination. There was no increased signal in disks that did not have a vacuum phenomenon. The presence of bone marrow edema, the amount of bone marrow edema, and the presence of degenerative abnormalities of the endplates on CT were correlated to the presence of intradiscal fluid or hyperintense signal and were assessed for statistical significance, as described previously in this article. Twenty-six of 95 total evaluated segments showed type 1 Modic changes (edema) adjacent to the endplates. All of the segments with E type 1 Modic changes adjacent to the endplates were associated with a vacuum phenomenon on CT. The mean score of edema along the adjacent endplates was of 200. Statistical Significance The presence of fluid or hyperintense signal was significantly correlated with the presence (p < ) and amount of edema along the endplates (p < ). In 51 segments, degenerative abnormalities of the endplates were detected on CT. Forty-four of 69 segments with intradiscal vacuum showed degenerative abnormalities of the endplates on CT. The presence of degenerative abnormalities of the endplates on CT was significantly correlated with the presence of fluid or hyper- F 1186 AJR:197, November 2011

6 Vacuum Phenomenon on CT and Fluid on MRI of Degenerative Disks intense signal (p < ). The presence of edema on MRI (type 1 Modic changes) was significantly correlated to degenerative abnormalities of the endplates on CT (p < ). Discussion The term intervertebral vacuum phenomenon refers to the radiographic appearance of gaseous collections in intervertebral disk spaces at one or more spinal levels, most commonly in the lumbar region [7]. It was first described by Magnusson in 1937 [8] and is a common finding that can be observed in 1 3% of spinal radiographs but can reach a prevalence of 20% in elderly individuals [7]. The phenomenon is accentuated on radiographs obtained during spinal extension and A Fig year-old woman (same patient as in Fig.1). MRI was performed after mobilization. A, At level of T12 L1 (arrowhead) there is initially no fluid in first T2-weighted image. B, Fluid (arrowhead) first appears in second T2-weighted sequence 9 minutes after commencement of scan and shows minimal progressive increase in remaining T2-weighted sequences (data not shown). is obscured on those obtained during flexion, with the same observation applying for supine versus upright radiographs [1]. Magnusson reported that the creation of intervertebral vacuum, subsequently filled by gas, requires a reduction of barometric pressure within the disk of up to 0.05 atmosphere ( Pa) [8]. CT has a higher sensitivity in the detection of vacuum phenomena of the spine than does radiography [1]. Ford et al. [9] analyzed the gas contained in a degenerated disk by gas chromatography, which was found to contain 90 92% nitrogen. Vacuum phenomena are most commonly seen in patients with intervertebral osteochondrosis, where dehydration of the disk and enlarging clefts within the nucleus pulposus lead to the accumulation of gas B from the surrounding tissues, appearing as a radiolucent area in the discal space in radiographs and CT scans. Intervertebral vacuum phenomena may also be seen in cases of tears of the peripheral fibers of the annulus fibrosus and in Schmorl node formation. A similar phenomenon which is termed the intravertebral vacuum cleft was described in osteoporotic vertebral compression fractures. This is thought to arise as a result of intraosseous cleft formation, which most probably is due to nonunion within the fracture cleft (i.e., subchondral pseudarthrosis). Formerly, this has been attributed to ischemic osteonecrosis of bone, also known as Kümmell disease [10]. Time-dependent accumulation of fluid within the intravertebral clefts has been described elsewhere [3, 11], with the amount of fluid depending on the duration of supine positioning [2, 3]. The presence of the intravertebral fluid sign is a strong predictive factor for the benign cause of a fracture [12]. Clinically, these phenomena are indicative of vertebral motion segment instability. Degenerated disks usually show low signal intensity on T2-weighted spin-echo sequences. Gas collections exhibit a signal loss on MRI due to the lack of water protons. However, MRI is not very sensitive in detecting gas [1] compared with CT. Fluidlike or hyperintense signal within a disk usually is indicative of spondylodiscitis [13]. However, in a retrospective study of 100 MRI examinations of the lumbar spine in patients without evidence of spinal infection, Schweitzer and el-noueam [14] could show hyperintense signal in 12% of cases. In the present study, we found that the accumulation of fluid depends on the postural status of the patient before MRI and is associated with the presence of vacuum phenomenon, which is a sign for a disrupted unstable disk. Fluid accumulation did not occur in any segment without intervertebral gas collections. Accumulation of gas is a very rapid process, whereas the accumulation of fluid is obviously a relatively slower dynamic process. In the present study, 23.2% of the segments with intervertebral vacuum phenomenon on CT showed a fluid-isointense or hyperintense signal on MRI after mobilization. After bed rest, the presence of fluid and hyperintense signal increased significantly to 59.4%. We have found a significant correlation between the presence of intradiscal fluid and the amount of bone marrow edema and degenerative abnormalities of the endplates, respectively. This finding is supported by the fact that histopathologic sections of type 1 Modic AJR:197, November

7 D Anastasi et al. changes have disruption and fissuring of the endplate and vascularized fibrous tissues within the adjacent bone marrow [15]. The pathogenic mechanisms causing Modic changes are not very clear. According to the literature, there may be biomechanical and biochemical causes [16]. The biomechanical mechanism involves fissuring and microfractures of the endplate due to uneven distribution of loads across the disk resulting from disk degeneration. Recent microfractures will show a hypointense signal on T1-weighted images and a hyperintense signal on T2-weighted images. This phenomenon might, therefore, reflect edema and vascularization after cumulative trauma and an inflammatory response after microfracture in the endplates. The correlation between Modic changes and altered mechanical stress is supported by the observation of the conversion of type 1 to type 2 after stabilization [16]. The causes of type 1 Modic changes may also represent various stages of an inflammatory process in the endplate [17]. In 1986, Crock [18] hypothesized that, in certain individuals, particularly after repeated trauma to the intervertebral disks, a syndrome develops resulting from the production of inflammatory substances in the damaged disk tissues. Diffusion of these substances through the vertebral endplates may then result in a local inflammatory reaction, giving rise to back pain. Burke et al. [19] in 2002 found higher levels of inflammatory mediators (e.g., interleukin-6, interleukin-8, and prostaglandin E2) in the disks of patients with low back pain than in those of patients with sciatica. Ohtori et al. [20] reported in 2006 that protein gene product 9.5 immunoreactive nerve fibers and tumor necrosis factor immunoreactive cells in the endplates from patients with Modic changes were present significantly more often than in normal endplates on MRI. Moreover, the number of tumor necrosis factor immunoreactive cells in endplates with type 1 changes was higher than in endplates with type 2 changes [20]. This suggests that Modic type 1 changes may be the result of inflammation mediated by inflammatory cytokines released from degenerative disks, with type 1 changes representing more active inflammation than type 2 and 3 changes. We hypothesize that the inflammatory response present in Modic type 1 changes gives rise to increased permeability of capillaries in the adjacent bone marrow. Vacuum phenomena are brought about by distraction forces applied across tissues, which may lower local pressures enough that certain gases in steady-state solution within tissues, such as nitrogen, reach a pressure point where they leave solution and enter a gaseous phase in the local anatomy. At that point, the gas fills the disintegrated disk space. We postulate that fluid transudate flows slowly into the vacuum cleft from perfused neighboring structures, such as the endplates and the bone marrow. We hypothesize that, on CT, which takes only a few minutes, the time spent in the supine position is too short for fluid to flow into the vacuum cleft. At rest, there is an influx of fluid from the surrounding tissue, inflammation, and microvessels into the discal cleft, with the gas in the vacuum phenomenon reentering into solution. This is seen in our serial T2 measurements and the comparison of MRI performed after mobilization and after bed rest. Type 1 Modic changes with the presence of hyperintense signal or intradiscal fluid on T2-weighted sequences can mimic spondylodiscitis [21]. In advanced cases of spondylodiscitis, the intradiscal fluid may be accompanied by epidural or paravertebral abscess formation, in which case the diagnosis is clear. However, in the early stages of spondylodiscitis, one may only have intradiscal fluid and some mild adjacent bone marrow edema. Clinical signs and symptoms and inflammatory markers may be equivocal, especially in elderly patients, rendering the differentiation from Modic type 1 changes difficult. We think that, in such cases, if there is a vacuum phenomenon on CT at the same level, performing two MRI scans, one after bed rest and another after mobilization, may provide more information. If the fluid disappears in the MRI after mobilization, infection with intradiscal abscess can be excluded. This may be a topic for further studies. Of segments with intervertebral vacuum phenomenon, 40.6% did not show hyperintense signal or fluidlike signal on T2- weighted images after bed rest. We suppose that, in these cases, there was no adequate contact to adjacent fluid-containing structures or to a vascular network that would have allowed an appreciable fluid inflow during the period of our MRI sequences. To our knowledge, only one other study, published by Wang et al. [22], found fluid replacement of intradiscal vacuum phenomenon on MRI after supine positioning. Those authors performed MRI examinations on 10 patients with multiple vacuum phenomena at 0, 1, and 2 hours while remaining continuously supine. CT was not performed in that study. They found a progressive replacement of vacuum phenomenon by hyperintense fluid in 81% of the vertebral segments containing vacuum phenomenon. This fluid was shown to be consistent with a transudate [22]. In contrast, in the present study, 42.0% of all segments with vacuum phenomenon showed replacement by fluid after bed rest, whereas 17.4% of all segments with vacuum phenomenon showed an intradiscal hyperintense signal below fluid signal intensity after bed rest. The reason for this difference is probably that we relied on CT to detect intervertebral vacuum phenomena, which is more sensitive than the MRI used by Wang et al., thereby obtaining a greater percentage of vacuum phenomena that were not replaced by fluid. In addition, in contrast to our study, six patients with osteoporotic compression fractures at one or more levels were included in the study by Wang et al. It has been shown by Linn and coworkers [3] that there is a continuum of vacuum phenomenon and fluid between the intravertebral fracture cleft and the disk space. It is postulated that fracture edema leads to the fluid accumulation. Therefore, in the present study, we excluded patients with acute osteoporotic vertebral compression fractures. Our study has some limitations. The number of patients (20) was relatively small, and several patients had to be excluded for reasons explained previously. Further studies may be necessary to support our conclusions. The intradiscal fluid observed in the study was not sampled, and its composition is therefore unknown, but we think it is likely to be a transudate, which would be consistent with the findings of previous related studies [3, 22]. Patient weight was not taken into consideration. This affects the axial load of the disk and therefore intradiscal pressure during standing, which may result in morerapid extrusion of fluid from the intradiscal space. On the day of the first MRI examination after mobilization, the patients were encouraged to undertake their normal daily activities in the ward. This is a variable that differed from patient to patient so that there may have been differences in the amount of energy expended. However, all patients were mobilized for 10 minutes in the corridor before the first MRI examination. We do not think this affected our conclusions because we were able to show a significant difference between the presence of fluid or hyperintense signal in the disks after bed rest and after mobilization AJR:197, November 2011

8 Vacuum Phenomenon on CT and Fluid on MRI of Degenerative Disks In conclusion, the replacement of intradiscal vacuum phenomenon by intradiscal fluid in degenerative disks is a dynamic process that is time and position dependent. Other factors, such as the individual anatomic and physiologic situation, may also play a role. Fluid or hyperintense signal is seen more commonly in patients after prolonged bed rest. The presence of fluid or hyperintense signal is correlated to the presence and amount of bone marrow edema and degenerative endplate abnormalities. However, an awareness of this phenomenon may, in some cases, be helpful in differentiating intradiscal fluid collections because of the described phenomenon from early spondylodiscitis, especially when bandlike bone marrow edema is present, though this would require validation through further study. Our findings also underline the important role of dynamic imaging in the study of functional phenomena in pathologic spinal conditions. Acknowledgment We thank Alexander Crispin for his invaluable assistance with the statistical part of the study. References 1. Coulier B. The spectrum of vacuum phenomenon and gas in spine. JBR-BTR 2004; 87: Malghem J, Maldague B, Labaisse MA, et al. Intravertebral vacuum cleft: changes in content after supine positioning. Radiology 1993; 187: Linn J, Birkenmaier C, Hoffmann RT, Reiser M, Baur-Melnyk A. The intravertebral cleft in acute osteoporotic fractures: fluid in magnetic resonance imaging vacuum in computed tomography? Spine (Phila Pa 1976) 2009; 34:E88 E93 4. Yoshida H, Shinomiya K, Nakai O, Kurosa Y, Yamaura I. Lumbar nerve root compression caused by lumbar intraspinal gas: report of three cases. Spine (Phila Pa 1976) 1997; 22: Grunshaw ND, Carey BM. Case report: gas within a cervical vertebral body. Clin Radiol 1994; 49: Leone A, Guglielmi G, Cassar-Pullicino VN, Bonomo L. Lumbar intervertebral instability: a review. Radiology 2007; 245: Resnick D, Niwayama G, Guerra J Jr, Vint V, Usselman J. Spinal vacuum phenomena: anatomical study and review. Radiology 1981; 139: Magnusson W. Über Die Bedingungen Des Hervortretens Der Wirklichen Gelenkspalte Auf Dem Röntgenbilde. Acta Radiol 1937; 18: Ford LT, Gilula LA, Murphy WA, Gado M. Analysis of gas in vacuum lumbar disc. AJR 1977; 128: Kümmell H. Die rarefizierende Ostitis der Wirbelkörper. Deutsche Med 1895; 21: Dupuy DE, Palmer WE, Rosenthal DI. Vertebral fluid collection associated with vertebral collapse. AJR 1996; 167: Baur A, Stabler A, Arbogast S, Duerr HR, Bartl R, Reiser M. Acute osteoporotic and neoplastic vertebral compression fractures: fluid sign at MR imaging. Radiology 2002; 225: Baur-Melnyk A, Triantafyllou M, Birkenmaier C, Reiser M. Degenerative diseases of the spine: rare and often unrecognized causes of pain syndromes (in German). Radiologe 2006; 46: Schweitzer ME, el-noueam KI. Vacuum disc: frequency of high signal intensity on T2-weighted MR images. Skeletal Radiol 1998; 27: Modic MT, Masaryk TJ, Ross JS, Carter JR. Imaging of degenerative disk disease. Radiology 1988; 168: Zhang YH, Zhao CQ, Jiang LS, Chen XD, Dai LY. Modic changes: a systematic review of the literature. Eur Spine J 2008; 17: Mitra D, Cassar-Pullicino VN, McCall IW. Longitudinal study of vertebral type-1 end-plate changes on MR of the lumbar spine. Eur Radiol 2004; 14: Crock HV. Internal disc disruption: a challenge to disc prolapse fifty years on. Spine (Phila Pa 1976) 1986; 11: Burke JG, Watson RWG, McCormack D, Dowling FE, Walsh MG, Fitzpatrick JM. Intervertebral discs which cause low back pain secrete high levels of proinflammatory mediators. J Bone Joint Surg Br 2002; 84: Ohtori S, Inoue G, Toshinori I, et al. Tumor necrosis factor-immunoreactive cells and PGP 9.5-immunoreactive nerve fibers in vertebral endplates of patients with discogenic low back pain and Modic type 1 or type 2 changes on MRI. Spine 2006; 31: Stabler A, Baur A, Kruger A, Weiss M, Helmberger T, Reiser M. Differential diagnosis of erosive osteochondrosis and bacterial spondylitis: magnetic resonance tomography (MRT) (in German). Rofo 1998; 168: Wang HJ, Chen BB, Yu CW, Hsu CY, Shih TT. Alteration of disc vacuum contents during prolonged supine positioning: evaluation with MR image. Spine (Phila Pa 1976) 2007; 32: AJR:197, November