Neuroradiology (2014) 56:1069–1078 DOI 10.1007/s00234-014-1433-0
DIAGNOSTIC NEURORADIOLOGY
Functional and quantitative magnetic resonance myelography of symptomatic stenoses of the lumbar spine Knut Eberhardt & Oliver Ganslandt & Andreas Stadlbauer
Received: 25 June 2014 / Accepted: 11 September 2014 / Published online: 23 September 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Introduction The objective of this study was to demonstrate that functional, quantitative magnetic resonance myelography (MRM) allows standardized diagnosis of symptomatic lumbar spinal stenoses which show at least equal detectability compared to functional myelography and postmyelographic CT (pmCT) based on intra- and postoperative findings. Methods We investigated 43 volunteers and 47 patients with symptomatic lumbar spinal stenoses using MRM in normal position as well as in flexion and extension in a standard whole-body MR scanner. Twenty volunteers were additionally examined under axial loading. All patients were investigated by functional myelography and pmCT and 10 patients had a functional lumbar MRM postoperatively. Range of motion and cerebrospinal fluid (CSF) volumes in normal position, flexion, extension, and under axial loading (volunteers) were assessed for each segment. Detectability was determined by using intraoperative findings, and postoperative freedom of symptoms was correlated with CSF volume changes in MRM. Results The ranges of motion in a standard whole-body MR scanner provide adequate scope for investigations into function (flexion and extension) in both volunteers and patients. Axial loading was associated with a mechanism of extension, albeit to a far smaller extent. Detectability of lumbar stenoses was 100 % for MRM, 58 % for conventional myelography, K. Eberhardt (*) MRI Center of Excellence, District Hospital Castle of Werneck, Balthasar-Neumann-Platz 1, D-97440 Werneck, Germany e-mail: [email protected] O. Ganslandt : A. Stadlbauer Department of Neurosurgery, University of Erlangen-Nuremberg, Erlangen, Germany A. Stadlbauer Department of Radiology and Nuclear Medicine, Medical University Vienna, Vienna, Austria
and 68 % for pmCT. Postoperative changes in CSF volume of levels with stenoses in MRM strongly correlated with freedom of symptoms (R=0.772). Conclusion This MRM method allows for exact diagnosis and reproducible quantification of stenoses, motion-related changes, and spondylolistheses of the lumbar spine. It may be useful for early detection of alterations in order to avoid neuronal compression. Keywords Functional MR myelography . Motion function . Flexion . Extension . Lumbar spine
Introduction For successful treatment, meaningful representation of the function-dependent and often spatially very complex changes of lumbar spinal stenosis is needed by the surgeon. A 3D visualization of the cerebrospinal fluid (CSF) volume of the nerve roots sheath provides additional information which may be helpful to prevent from overestimation of lumbar spine stenoses [1, 2]. Standardized assessment of the functional stability of the facet joints is important to determine both the surgical technique and the postoperative follow-up. Conventional functional myelography has long been the method of choice for diagnosing lumbar spinal stenosis and is still an important method for investigating the influence of hyperextension and hyperflexion on the extent of the stenosis [3]. It is still the only routine method for detecting the morphological correlates of a functional, symptomatic lumbar spinal stenosis so far [4] and it is the only accurate imaging technique for patients with spinal metallic implants, which can cause artifacts on magnetic resonance imaging (MRI) and computed tomography (CT). Most cases of spinal stenosis are unambiguously clearly diagnosed by imaging techniques (MRI or CT) performed in
1070
supine position because the measured values of the dural sac are clearly pathological and may only worsen in motion function in charge or extension. However, in ambiguous cases or patients, in whom discrepancies between clinical symptoms and supine imaging occurred [5–10], further exploration is required through imaging in motion function. One of the few reliable prognostic signs is the block of contrast flow, which is a good predictor of the successful outcome of decompression surgery [11]. Conventional myelography is limited by its inability to determine the cause of block or compression and to visualize extrathecal nerve root compression. The combination with a CT scan performed after myelography (postmyelographic CT (pmCT)) compensates for these limitations [3]. The pmCT is a complement to conventional myelography and helps to assess the dural sac (diameters or cross-sectional areas) and the bony status (which is especially important in older patients) of the surgical area. MRI provides a very good soft tissue contrast, allows any slice orientation, and does not produce ionizing radiation [12, 13]. Therefore, MRI is the preferred imaging modality for the radiological assessment of lumbar spinal stenosis [14]. Open MRI systems enable a functional MRI investigation of spinal flexion and extension during the application of axial loading or even in the supine position [15]. However, the excess to and the number of these MR systems is limited, and their magnetic field strength is in general low (less than 1 T). This study introduces a concept of standardized highresolution MRI using a conventional whole-body MR scanner in combination with adequate post-processing to investigate the lumbar spine in motion function (flexion, extension, and compression) and compares it with conventional myelography, pmCT, and intra- and postoperative findings. The questions to be answered here were the following: (i) Does the range of motion in the MR scanner provide adequate scope for investigations into function? (ii) Does the MR myelography (MRM) offer adequate preoperative diagnosis compared with conventional myelography and pmCT? (iii) How good is the consensus of the results of preoperative imaging methods with the intraoperative findings? (iv) How good is the consistency of the findings of postoperative MRM with the postoperative outcome?
Materials and methods Phantom, volunteers, and patients Measurements using a self-developed, CuSO4 solution-filled Plexiglas phantom (volume of the liquid, 31.75 mm3) were performed in order to assess the detectability of volumetric
Neuroradiology (2014) 56:1069–1078
methods (Fig. 1). We examined 43 healthy volunteers (average age of 38.0±9.0 years), 24 volunteers were females (37.6 ±8.1 years), and 19 volunteers were males (38.7±10.1 years). We included 47 patients (average age of 67.7±13.1 years) with symptomatic lumbar spinal stenosis; 25 patients were females (69.8±9.2 years), and 22 patients were males (65.4± 13.1 years). We included patients with clinical symptoms after a routine MRI examination, which was performed previously and demonstrated a stenosis. The populations of volunteers and patients did significantly differ in terms of age (p100 %) with intermittent claudication; 2=low residual symptoms (especially pain or sensory disturbances); and 3=no improvement.
Fig. 4 Functional lumbar MR myelography of a volunteer in flexion, normal position, and extension without (four columns left) and with superposition of anatomical background information (three columns right)
Statistics The statistical analyses were performed with statistical software (SPSS, Chicago, IL, USA). To check intraindividual differences and the range of motion between the imaging method’s liquid volume, a Wilcoxon signed rank test was used and a Mann-Whitney U test was used to compare the significance of the tests between patients and controls. Pearson correlation coefficient was calculated to determine the degree
Neuroradiology (2014) 56:1069–1078
1073
of linear relationship between two parameters. Comparison of plain films with 2D pmCT and 3D MRM data may cause inaccuracies, which was avoided by the formation of quotients. The statistical analyses were based on a significance level of p=0.05.
Results Functional range of motion The results for the functional range of motion, flexion, and extension in 43 volunteers are listed in Table 1. The average range of motion ranged from 13.0° (L1-2) to 16.7° (L5-S1) compared with 11.9° (L1-2) to 17.0° (L5-S1) at Dvorak et al. [18]. Figure 5 demonstrates the procedure for determining the functional range of motion for a 78-year-old patient with multi-segmental central lumbar spinal stenosis and degenerative spondylolisthesis. The segmental functional range of motion, flexion, and extension of the lumbar spine in 47 patients with lumbar spinal stenosis is also shown in Table 1. The functional ranges of motion were for MRM between 7.6° (L1-2) and 11.41° (L5-S1) and for conventional myelography, between 5.9° (L1-2) and 9.2° (L5-S1). There was no significant difference in the range of motion of patients between conventional myelography and MRM.
31.75 mm3. These findings were compared with the CT data for adjustments of the MRI examinations. Table 2 shows the changes of CSF volume due to functional motion as differences in CSF volumes in flexion, extension, and compression compared to the normal position in the 43 volunteers and 47 patients with lumbar spinal stenosis. For volunteers, CSF volumes increased in flexion relative to the normal position between 5.3 % (L2-3) and 9.6 % (L4-5) and decreased in extension between 9.2 % (L1-2) and 21.5 % (L5-S1) and under compression between 6.7 % (L1-2) and 12.0 % (L5-S1) when compared to normal position. The results for compression and extension correlated strongly with coefficients between R=0.983 (L1-2) and R=0.996 (L3-4). From this, it is noted that compression is effective mainly through an extension mechanism; however, it is less extensive when compared to direct extension. We therefore decided to abandon applying compression to patients. The results for the patients show that the CSF volume increased in flexion between 9.1 % (L2-3) and 18.1 % (L3-4) and decreased in extension between 10.6 % (L1-2) and 16.7 % (L3-4) with respect to normal position. Figure 6 shows the results for CSF volumetry in movement function for the same patient as in Fig. 5, another example of a 75-year-old patient with central stenosis at L3-4, and leftsided lateral stenosis in segment L4-5 is shown in Fig. 7. Detectability
CSF volumetry in motion function The validation of automated volumetric measurements with the phantom showed deviations of the volume values of CuSO4 solution from 0.09 to 2.39 % with a total volume of
Table 3 summarizes the findings and sensitivities of the imaging methods compared with intraoperative findings in patients with degenerative spondylolisthesis and lumbar spinal stenosis. With MRM, more stenotic segments (n=41) were
Table 1 Functional range of motion (ROM), flexion, and extension of the lumbar spine for 43 volunteers, 47 patients, and values from the literature Lumbar levels L1-2
L2-3
L3-4
L4-5
L5-S1
Volunteers flexion
88.6°±2.5°
92.9°±3.8°
96.0°±2.6°
95.7°±3.0°
102.9°±3.5°
Volunteers extension Volunteers ROM Pat. MRM flexion Pat. MRM extension Pat. MRM ROMa Pat. cMyelo flexion Pat. cMyelo extension Pat. cMyelo ROMb Literaturec
101.6°±2.8° 13.0°±2.0° 90.7°±3.3° 98.3°±4.0° 7.6°±3.3° 91.7°±2.9° 97.6°±4.6° 5.9°±2.5° 11.9°
106.6°±3.3° 13.7°±2.8° 94.9°±3.4° 101.6°±4.1° 6.7°±3.3° 94.7°±4.7° 100.6°±7.2° 5.9°±2.7° 14.5°
109.9°±3.7° 13.9°±2.7° 98.1°±5.5° 105.0°±5.6° 6.9°±3.7° 97.7°±2.7° 103.6°±6.8° 5.9°±3.4° 15.3°
111.1°±3.6° 15.4°±3.1° 97.6°±5.5° 107.6°±5.2° 10.0°±4.1° 98.7°±6.8° 107.3°±6.9° 8.6°±5.3° 18.2°
119.6°±3.9° 16.7°±3.0° 104.7°±8.4° 116.1°±7.5° 11.4°±5.7° 106.1°±7.3° 115.3°±6.8° 9.2°±4.4° 17.0°
Functional range of motion is the difference of angles between flexion and extension in degrees a
Pat. MRM ROM=range of motion of patients determined on magnetic resonance myelography
b
Pat. cMyelo ROM=range of motion of patients determined on conventional myelography
c
Dvorak et al. [18]
1074
Neuroradiology (2014) 56:1069–1078
Fig. 5 Determination of the functional range of motion of a 78-year-old female patient with multi-segmental central lumbar spinal stenosis and degenerative spondylolisthesis. Measurement of listhesis as described previously by Dupuis13 on conventional radiograms (top and bottom left) and functional lumbar MR examination (SINOP sequence, in-phase) in flexion (top middle and right) and extension (bottom middle and right) shows the same functional range of motion. An increased anterolisthesis at the level L4-5 from 3 to 10 mm is also recognizable. Note: angle la=lb= 3°, angle lc=1d=6°. Increase of anterolisthesis: distance 2c=2d= 0.7 cm, distance 2a=2b=1.0 cm
detected as with conventional myelography (n=23) including pmCT (n=28). Here, both axillary and peripheral nerve root sheaths were better assessed in MRM compared with conventional myelography and pmCT. The detectability of MRM for lumbar spinal stenosis was 100 %, which corresponded to the intraoperative gold standard. The detectability of conventional myelography was 58 %, and that of pmCT was 68 %. While spondylolisthesis was reliably detected with MRM, conventional myelography, and pmCT (detectability=100 % for all methods), examinations in motion function were only possible with MRM and conventional myelography. The
axillary region and the peripheral nerve root sheaths were depicted and evaluated in MRM only. The detectability of MRM (100 %) corresponded to the gold standard, i.e., the intraoperative findings. Conventional myelography and pmCT had lower detectability due to false-positive findings for lumbar spinal stenosis caused by overemphasis of liquid compression and due to methodological and technical reasons for spondylolistheses, i.e., assessment of the nerve roots in the lateral projection is not possible and functional pmCT examinations are not feasible (increased radiation exposure to the patient).
Table 2 Functional changes of CSF volumes in MRM in flexion, extension, and compression compared to normal position at 43 volunteers and 47 patients with lumbar spinal stenosis Level
Volunteers F–N [mm3]
Volunteers N–E [mm3]
Volunteersa N–C [mm3]
Patients F–N [mm3]
Patients N–E [mm3]
L1-2 L2-3 L3-4 L4-5 L5-S1
149 (6.3 %) 116 (5.3 %) 168 (7.5 %) 227 (9.6 %) 190 (7.9 %)
204 (9.2 %) 293 (14.0 %) 266 (12.8 %) 309 (14.5 %) 475 (21.5 %)
148 (6.7 %) 172 (8.2 %) 180 (8.6 %) 238 (11.9 %) 264 (12.0 %)
230 (12.2 %) 172 (9.1 %) 202 (18.1 %) 165 (15.4 %) 193 (11.7 %)
200 (10.6 %) 204 (10.9 %) 186 (16.7 %) 121 (11.2 %) 226 (13.8 %)
Functional changes of CSF volumes as difference of volume between function (flexion, extension, and compression) and normal position. Changes of CSF volumes relative to normal position in percentage terms are shown in parentheses F flexion, N normal position, E extension, C compression a
Twenty volunteers were investigated under compression
Neuroradiology (2014) 56:1069–1078
1075
Fig. 6 Functional volumetry of a 78-year-old female patient with multi-segmental central lumbar spinal stenosis and degenerative spondylolistheses (same patient as in Fig. 5) in extension (left) and flexion (right). The CSF volume nearly doubles in flexion at the stenosed levels (L2-3, 86.8 % and L3-4, 78.9 %) and increased only by 24.8 % at level L4-5 (red arrow), however
Postoperative outcome A total of 10 patients were pre- and postoperatively examined by MRM and one patient additionally by conventional myelography. These 10 patients underwent laminectomy due to lumbar spinal stenosis. The postoperative examinations were performed 3–6 months after surgery. Figure 8 shows pre- and postoperative MRM examinations of a 75-year-old patient with bilateral stenosis of the lumbar spine. A correlation analysis between the score of degree of postoperative freedom from symptoms and the postoperative change in CSF volume in the stenotic levels on MRM for the 10 patients gave a correlation coefficient of R=0.772. This indicates that the postoperative improvement on the extent of postoperative decompression showed a strong correlation.
Table 3 Detectability of the methods compared with intraoperative findings
Fig. 7 Functional volumetry of a 75-year-old female patient with central stenosis at level L3-4 and left-sided lateral stenosis at level L4-5. Functional lumbar MR myelography in flexion (top row), extension (middle row), and conventional myelography in flexion (bottom row) shows consistently a strong defect at level L3-4 on right-handed side (yellow arrows) and an axillary defect on left-handed side at level L4-5 (red arrows)
Criteria for assessment Detectability Detectability MRM Conv. myelo./ pmCT
Detectability Surgery
Stenoses recognized Axilla definable Nerves depicted Listhesis recognized Increase in function Nerve root compression
41 (100 41 (100 41 (100 15 (100 n.a. 15 (100
41 (100 %) 38 (93 %) 38 (93 %) 15 (100 %) 15 (100 %) 15 (100 %)
23/28 (56/68 %) 5/15 (12/36 %) 5/15 (12/36 %) 15 (100 %) 15/0 (100/0 %) n.a.
%)a %)a %)a %) %)
The figures are the number of findings; the percentage related to the gold standard of intraoperative findings are in parentheses MRM MR myelography, conv. myelo./pmCT conventional myelography/ postmyelographic CT, n.a. not applicable a
Two patients were operated a second time due to worsening of symptoms after first surgery. Whereby, three stenoses, which were not found during first surgery but in preoperative MRM, were confirmed during second surgery
1076
Neuroradiology (2014) 56:1069–1078
Fig. 8 Pre- and postoperative results of lumbar MR myelography of a 75-year-old female patient with lateral stenosis at level L4-5. Preoperative MR myelography showed a signal cancelation at level L4-5 (left dorsolateraly); the CSF volume in this segment was 1413 mm3. This value increased after 7 days postoperatively to 2096 mm3 and normalized at 1945 mm3 after 4 months. The slightly lower value for the CSF volume at the second postoperative measurement was explained due to scar tissue which has been formed in the meantime
Discussion In this study, we demonstrated the possibility to study the functional range of motion of the lumbar spine in a conventional whole-body MR scanner. Our MR methods achieved sufficient data quality even without the magnetic field’s isocenter. This approach guarantees detectability in preoperative diagnosis lumbar motion function, which was even superior to the results from conventional myelography and pmCT. Very good correlations between pre- and postoperative functional MRM with both intraoperative findings and postoperative outcome were found. We were able to quantitatively define CSF sub-volumes in functional positions by using direct volume rendering. Function-dependent measurements were described in numerous publications using open MRI systems with mostly low field strength [19] or only under axial load [20]. However, in all in vivo studies published so far, only 2D analyses (spinal distances and areas) were carried out [21, 22], but this strategy is associated with large scattering and poor reproducibility of
values. 3D analyses were described only in in vitro measurements on cadavers [23, 24] and in animal models [25] and are therefore not directly comparable with our results. In volumetry studies, shifts in functional positions or rotational deformities (scoliosis) play not an essential role since variations between imaging plane and spinal axis to be compensated have considerable influence in 2D parameters [19]. The methodical error of or volumetric approach was found with 0.9 to 2.7 % using data from a phantom. A further important advantage of our method is that the peripheral CSF sheaths can be assessed quantitatively too. The values given in the literature [22, 24] for the range of motion show a relatively large range but are in good accordance with our findings. Studies were carried out with open MR scanners using vertically aligned magnetic fields of low field strength which allowed for some degree of movement [19]. However, the spatial resolution of the MR data, which can be achieved with these MR scanners, is far too insufficient to perform detailed analysis or 3D examinations [1]. Furthermore, studies in sitting or in the lateral position without
Neuroradiology (2014) 56:1069–1078
position fixing devices allow only poorly controlled and not standardized examinations. Limited mobility in older patients with lumbar spinal stenosis found in our study population is confirmed by other studies [26, 27]. Our results also demonstrated that compression of the lumbar spine is associated and comparable with lordosis in extension. The resulting reduction in volume of the spinal canal under compression relative to the normal position is low compared with that in extension. We therefore decided to completely dispense with applying functional studies under compression, i.e., axial load. The differences in total range of motion between patients and volunteers can be explained by relieving posture due to pain and limited mobility. However, one has to keep in mind that the differences in range of motion and CSF volumes can be partly influenced by age differences between volunteers and patients too. Therefore, we decided to present no p values of significance for the differences between volunteers and patients. Our results are confirmed by Madsen et al. [28] but are also in contrast to other groups that see relevance for axial load MRI by correlation with clinical parameters [29, 30]. However, it must be expressly pointed out that these measurement studies have only been conducted under motion function under axial load, and neither under flexion or extension. Coulier et al. [31] demonstrated that the upright myelography and especially during upright extension was the most efficient technique to detect pathological narrowing of the dural sac. They found only minor discrepancies between the neutral position and the extension showing that already when the patient was upright, there was spontaneous extension. It also appeared from their results that flexion was of limited value and therefore of no great interest in all diagnostic techniques. It was also empirically demonstrated that 60 to 80 mm2 was also the correct landmark of lumbar stenosis and that this measure of surface was more specific and more reproducible than the conventional measure of the diameter only. Our visualization results of MRM in patients with lumbar spinal stenosis were in stronger accordance with the intraoperative findings. Thus, MRM showed a higher detectability than conventional myelography and pmCT. Stenoses were overrated in pmCT and therefore lead to increased falsepositive findings as demonstrated previously [32]. The advantages of our MRM approach for the assessment of stenoses in the lumbar spine compared to conventional myelography and pmCT have been shown in previous studies [16, 33]. In all cases of high grade stenosis, height and length of the main stenosis could be sufficiently localized and determined using our MRM approach. Spondylolisthesis could be as easily detected and evaluated. In contrast to conventional myelography, we were able to assess axilla region and peripheral root sheaths with functional motion MRM, which is also sometimes challenging in conventional MRI [34]. Furthermore, our study showed that the change in CSF volume on MRM as a result of lumbar decompressive surgery strongly
1077
correlates with postoperative clinical findings. These results, however, are contradictory to other postoperative MRI studies [35–38], which described that early postoperative MR controls lead to false-positive results and late MR controls to false-negative results. This may be explained by circumscribed local changes (adhesions between the dura and muscle in area of the laminectomy or circumscribed arachnoiditis) which could lead to incorrect results in imaging. To sum up, one could say that for preoperative evaluation, the volume change between flexion and extension is the most significant parameter. Our results for volunteers under compression demonstrated that compression is effective mainly through an extension mechanism. However, it is less extensive when compared to direct extension. Furthermore, extension is easier to implement and is better tolerated by the subjects compared to compression. Flexion is preferable to normal position due to the larger absolute CSF volumes. Changes in CSF volume between flexion and extension allow for estimation of functional variability and remaining CSF volume to relieve the nerve roots. For postoperative evaluation, the change between pre- and postoperative CSF volume in flexion is most meaningful because it reveals the maximum increase in CSF volume caused by surgery. The patients included in our study represented a population of patients with severe stenoses but not with a mild stenosis. In this regard, the results of the conventional myelography were not disastrous but may reflect the disadvantage of conventional myelography in patients with severe spinal stenosis. In conclusion, clinically reliable assignment of functionally related symptoms (e.g., of instabilities) to main stenosis height and length is often challenging because of the diversity of symptoms. Since decompressive surgery is usually the first choice for treatment of neural compression, medical imaging has to significantly and reliably contribute to preoperative treatment planning. Necessary diagnostic criteria can be fulfilled by this advanced MRM approach, which allows for sufficient visualization of neural structures under motion function. Ethical standards and patient consent We declare that all human and animal studies have been approved by the Ethics Committee of the University Erlangen and have therefore been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. We declare that all patients gave informed consent prior to inclusion in this study. Conflict of interest We declare that we have no conflict of interest.
References 1. Eberhardt K, Ganslandt O, Stadlbauer A (2013) Improved magnetic resonance myelography using image fusion. Fortschr Röntgenstr 185:333–339
1078 2. Hans FJ, Reinges MH, Krings T (2004) Lumbar nerve root avulsion following trauma: balanced fast field-echo MRI. Neuroradiol 46: 144–147 3. Siebert E, Prüss H, Klingebiel R et al (2009) Lumbar spinal stenosis: syndrome, diagnostics and treatment. Nat Rev Neurol 5:392–403 4. Saifuddin A (2000) The imaging of lumbar spinal stenosis. Clin Radiol 55:581–594 5. Adamova B, Vohanka S, Dusek L (2005) Dynamic electrophysiological examination in patients with lumbar spinal stenosis: is it useful in clinical practice? Eur Spine J 14:269–276 6. Amundsen T, Weber H, Lilleas F et al (1995) Lumbar spinal stenosis. Clinical and radiologic features. Spine (Phila Pa 1976) 20:1178–1186 7. Bal S, Celiker R, Palaoglu S, Cila A (2006) F wave studies of neurogenic intermittent claudication in lumbar spinal stenosis. Am J Phys Med Rehabil 85:135–140 8. Boden SD, Davis DO, Dina TS et al (1990) Abnormal magneticresonance scans of the lumbar spine in asymptomatic subjects. A prospective investigation abnormal lumbar magnetic-resonance spine scans of the in asymptomatic. J Bone Jt Surg 72:403–408 9. Moon E-S, Kim H-S, Park J-O et al (2005) Comparison of the predictive value of myelography, computed tomography and MRI on the treadmill test in lumbar spinal stenosis. Yonsei Med J 46: 806–811 10. Weinstein JN, Tosteson TD, Lurie JD et al (2008) Surgical versus nonsurgical therapy for lumbar spinal stenosis. N Engl J Med 358: 794–810 11. Herno A, Airaksinen O, Saari T, Miettinen H (1994) The predictive value of preoperative myelography in lumbar spinal stenosis. Spine (Phila Pa 1976) 19:1335–1338 12. Klasen J, Antoch G, Blondin D (2011) MR imaging of the abdomen in pregnancy. Röfo 183:514–522 13. Komlosi P, Zhang Y, Leiva-Salinas C et al (2014) Adaptive statistical iterative reconstruction reduces patient radiation dose in neuroradiology CT studies. Neuroradiol 56:187–193 14. Issack PS, Cunningham ME, Pumberger M et al (2012) Degenerative lumbar spinal stenosis: evaluation and management. J Am Acad Orthop Surg 20:527–535 15. Alyas F, Connell D, Saifuddin A (2008) Upright positional MRI of the lumbar spine. Clin Radiol 63:1035–1048 16. Eberhardt K, Ganslandt O, Stadlbauer A (2014) Improved magnetic resonance myelography of the lumbar spine using image fusion and volumetry. J Neurosurg Spine 20:220–226 17. Dupuis PR, Yong-Hing K, Cassidy JD, Kirkaldy-Willis WH (1985) Radiologic diagnosis of degenerative lumbar spinal instability. Spine (Phila Pa 1976) 10:262–276 18. Dvorak J, Panjabi MM, Chang DG et al (1991) Functional radiographic diagnosis of the lumbar spine. Flexion-extension and lateral bending. Spine (Phila Pa 1976) 16:562–571 19. Vitzthum HE, Konig A, Seifert V (2000) Dynamic examination of the lumbar spine by using vertical, open magnetic resonance imaging. J Neurosurg 93:58–64 20. Meakin JR, Smith FW, Gilbert FJ, Aspden RM (2008) The effect of axial load on the sagittal plane curvature of the upright human spine in vivo. J Biomech 41:2850–2854 21. Hansson T, Suzuki N, Hebelka H, Gaulitz A (2009) The narrowing of the lumbar spinal canal during loaded MRI: the effects of the disc and ligamentum flavum. Eur Spine J 18:679–686
Neuroradiology (2014) 56:1069–1078 22. Huang KY, Lin RM, Lee YL, Li JD (2009) Factors affecting disability and physical function in degenerative lumbar spondylolisthesis of L4-5: evaluation with axially loaded MRI. Eur Spine J 18:1851–1857 23. Bisschop A, van Royen BJ, Mullender MG et al (2012) Which factors prognosticate spinal instability following lumbar laminectomy? Eur Spine J 21:2640–2648 24. Fujii R, Sakaura H, Mukai Y et al (2007) Kinematics of the lumbar spine in trunk rotation: in vivo three-dimensional analysis using magnetic resonance imaging. Eur Spine J 16:1867–1874 25. Chan DD, Khan SN, Ye X et al (2011) Mechanical deformation and glycosaminoglycan content changes in a rabbit annular puncture disc degeneration model. Spine (Phila Pa 1976) 36:1438–1445 26. Harvey SB, Smith FW, Hukins DW (1998) Measurement of lumbar spine flexion-extension using a low-field open-magnet magnetic resonance scanner. Invest Radiol 33:439–443 27. McGregor AH, McCarthy ID, Dore CJ, Hughes SP (1997) Quantitative assessment of the motion of the lumbar spine in the low back pain population and the effect of different spinal pathologies of this motion. Eur Spine J 6:308–315 28. Madsen R, Jensen TS, Pope M et al (2008) The effect of body position and axial load on spinal canal morphology: an MRI study of central spinal stenosis. Spine (Phila Pa 1976) 33:61–67 29. Kanno H, Endo T, Ozawa H et al (2012) Axial loading during magnetic resonance imaging in patients with lumbar spinal canal stenosis: does it reproduce the positional change of the dural sac detected by upright myelography? Spine (Phila Pa 1976) 37:E985–E992 30. Kanno H, Ozawa H, Koizumi Yet al (2012) Dynamic change of dural sac cross-sectional area in axial loaded magnetic resonance imaging correlates with the severity of clinical symptoms in patients with lumbar spinal canal stenosis. Spine (Phila Pa 1976) 37:207–213 31. Coulier B (2000) Evaluation of lumbar canal stenosis: decubitus imaging methods versus flexion-extension myelography and surface measurements versus the diameter of the dural sac. J Belge Radiol Belgisch Tijdschr voor Radiol 83:61–67 32. Zander DR, Lander PH (1998) Positionally dependent spinal stenosis: correlation of upright flexion-extension myelography and computed tomographic myelography. Can Assoc Radiol J 49:256–261 33. Eberhardt KE, Hollenbach HP, Tomandl B, Huk WJ (1997) Threedimensional MR myelography of the lumbar spine: comparative case study to X-ray myelography. Eur Radiol 7:737–742 34. Weishaupt D, Schmid MR, Zanetti M et al (2000) Positional MR imaging of the lumbar spine: does it demonstrate nerve root compromise not visible at conventional MR imaging? Radiol 215:247–253 35. Awwad EE, Smith KR Jr (1999) MRI of marked dural sac compression by surgicel in the immediately postoperative period after uncomplicated lumbar laminectomy. J Comput Assist Tomogr 23:969–975 36. Herno A, Partanen K, Talaslahti T et al (1999) Long-term clinical and magnetic resonance imaging follow-up assessment of patients with lumbar spinal stenosis after laminectomy. Spine (Phila Pa 1976) 24: 1533–1537 37. Leonardi MA, Zanetti M, Saupe N, Min K (2010) Early postoperative MRI in detecting hematoma and dural compression after lumbar spinal decompression: prospective study of asymptomatic patients in comparison to patients requiring surgical revision. Eur Spine J 19: 2216–2222 38. Ross JS (1999) MR imaging of the postoperative lumbar spine. Magn Reson Imaging Clin N Am 7:513–524, viii
DIAGNOSTIC NEURORADIOLOGY
Functional and quantitative magnetic resonance myelography of symptomatic stenoses of the lumbar spine Knut Eberhardt & Oliver Ganslandt & Andreas Stadlbauer
Received: 25 June 2014 / Accepted: 11 September 2014 / Published online: 23 September 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Introduction The objective of this study was to demonstrate that functional, quantitative magnetic resonance myelography (MRM) allows standardized diagnosis of symptomatic lumbar spinal stenoses which show at least equal detectability compared to functional myelography and postmyelographic CT (pmCT) based on intra- and postoperative findings. Methods We investigated 43 volunteers and 47 patients with symptomatic lumbar spinal stenoses using MRM in normal position as well as in flexion and extension in a standard whole-body MR scanner. Twenty volunteers were additionally examined under axial loading. All patients were investigated by functional myelography and pmCT and 10 patients had a functional lumbar MRM postoperatively. Range of motion and cerebrospinal fluid (CSF) volumes in normal position, flexion, extension, and under axial loading (volunteers) were assessed for each segment. Detectability was determined by using intraoperative findings, and postoperative freedom of symptoms was correlated with CSF volume changes in MRM. Results The ranges of motion in a standard whole-body MR scanner provide adequate scope for investigations into function (flexion and extension) in both volunteers and patients. Axial loading was associated with a mechanism of extension, albeit to a far smaller extent. Detectability of lumbar stenoses was 100 % for MRM, 58 % for conventional myelography, K. Eberhardt (*) MRI Center of Excellence, District Hospital Castle of Werneck, Balthasar-Neumann-Platz 1, D-97440 Werneck, Germany e-mail: [email protected] O. Ganslandt : A. Stadlbauer Department of Neurosurgery, University of Erlangen-Nuremberg, Erlangen, Germany A. Stadlbauer Department of Radiology and Nuclear Medicine, Medical University Vienna, Vienna, Austria
and 68 % for pmCT. Postoperative changes in CSF volume of levels with stenoses in MRM strongly correlated with freedom of symptoms (R=0.772). Conclusion This MRM method allows for exact diagnosis and reproducible quantification of stenoses, motion-related changes, and spondylolistheses of the lumbar spine. It may be useful for early detection of alterations in order to avoid neuronal compression. Keywords Functional MR myelography . Motion function . Flexion . Extension . Lumbar spine
Introduction For successful treatment, meaningful representation of the function-dependent and often spatially very complex changes of lumbar spinal stenosis is needed by the surgeon. A 3D visualization of the cerebrospinal fluid (CSF) volume of the nerve roots sheath provides additional information which may be helpful to prevent from overestimation of lumbar spine stenoses [1, 2]. Standardized assessment of the functional stability of the facet joints is important to determine both the surgical technique and the postoperative follow-up. Conventional functional myelography has long been the method of choice for diagnosing lumbar spinal stenosis and is still an important method for investigating the influence of hyperextension and hyperflexion on the extent of the stenosis [3]. It is still the only routine method for detecting the morphological correlates of a functional, symptomatic lumbar spinal stenosis so far [4] and it is the only accurate imaging technique for patients with spinal metallic implants, which can cause artifacts on magnetic resonance imaging (MRI) and computed tomography (CT). Most cases of spinal stenosis are unambiguously clearly diagnosed by imaging techniques (MRI or CT) performed in
1070
supine position because the measured values of the dural sac are clearly pathological and may only worsen in motion function in charge or extension. However, in ambiguous cases or patients, in whom discrepancies between clinical symptoms and supine imaging occurred [5–10], further exploration is required through imaging in motion function. One of the few reliable prognostic signs is the block of contrast flow, which is a good predictor of the successful outcome of decompression surgery [11]. Conventional myelography is limited by its inability to determine the cause of block or compression and to visualize extrathecal nerve root compression. The combination with a CT scan performed after myelography (postmyelographic CT (pmCT)) compensates for these limitations [3]. The pmCT is a complement to conventional myelography and helps to assess the dural sac (diameters or cross-sectional areas) and the bony status (which is especially important in older patients) of the surgical area. MRI provides a very good soft tissue contrast, allows any slice orientation, and does not produce ionizing radiation [12, 13]. Therefore, MRI is the preferred imaging modality for the radiological assessment of lumbar spinal stenosis [14]. Open MRI systems enable a functional MRI investigation of spinal flexion and extension during the application of axial loading or even in the supine position [15]. However, the excess to and the number of these MR systems is limited, and their magnetic field strength is in general low (less than 1 T). This study introduces a concept of standardized highresolution MRI using a conventional whole-body MR scanner in combination with adequate post-processing to investigate the lumbar spine in motion function (flexion, extension, and compression) and compares it with conventional myelography, pmCT, and intra- and postoperative findings. The questions to be answered here were the following: (i) Does the range of motion in the MR scanner provide adequate scope for investigations into function? (ii) Does the MR myelography (MRM) offer adequate preoperative diagnosis compared with conventional myelography and pmCT? (iii) How good is the consensus of the results of preoperative imaging methods with the intraoperative findings? (iv) How good is the consistency of the findings of postoperative MRM with the postoperative outcome?
Materials and methods Phantom, volunteers, and patients Measurements using a self-developed, CuSO4 solution-filled Plexiglas phantom (volume of the liquid, 31.75 mm3) were performed in order to assess the detectability of volumetric
Neuroradiology (2014) 56:1069–1078
methods (Fig. 1). We examined 43 healthy volunteers (average age of 38.0±9.0 years), 24 volunteers were females (37.6 ±8.1 years), and 19 volunteers were males (38.7±10.1 years). We included 47 patients (average age of 67.7±13.1 years) with symptomatic lumbar spinal stenosis; 25 patients were females (69.8±9.2 years), and 22 patients were males (65.4± 13.1 years). We included patients with clinical symptoms after a routine MRI examination, which was performed previously and demonstrated a stenosis. The populations of volunteers and patients did significantly differ in terms of age (p100 %) with intermittent claudication; 2=low residual symptoms (especially pain or sensory disturbances); and 3=no improvement.
Fig. 4 Functional lumbar MR myelography of a volunteer in flexion, normal position, and extension without (four columns left) and with superposition of anatomical background information (three columns right)
Statistics The statistical analyses were performed with statistical software (SPSS, Chicago, IL, USA). To check intraindividual differences and the range of motion between the imaging method’s liquid volume, a Wilcoxon signed rank test was used and a Mann-Whitney U test was used to compare the significance of the tests between patients and controls. Pearson correlation coefficient was calculated to determine the degree
Neuroradiology (2014) 56:1069–1078
1073
of linear relationship between two parameters. Comparison of plain films with 2D pmCT and 3D MRM data may cause inaccuracies, which was avoided by the formation of quotients. The statistical analyses were based on a significance level of p=0.05.
Results Functional range of motion The results for the functional range of motion, flexion, and extension in 43 volunteers are listed in Table 1. The average range of motion ranged from 13.0° (L1-2) to 16.7° (L5-S1) compared with 11.9° (L1-2) to 17.0° (L5-S1) at Dvorak et al. [18]. Figure 5 demonstrates the procedure for determining the functional range of motion for a 78-year-old patient with multi-segmental central lumbar spinal stenosis and degenerative spondylolisthesis. The segmental functional range of motion, flexion, and extension of the lumbar spine in 47 patients with lumbar spinal stenosis is also shown in Table 1. The functional ranges of motion were for MRM between 7.6° (L1-2) and 11.41° (L5-S1) and for conventional myelography, between 5.9° (L1-2) and 9.2° (L5-S1). There was no significant difference in the range of motion of patients between conventional myelography and MRM.
31.75 mm3. These findings were compared with the CT data for adjustments of the MRI examinations. Table 2 shows the changes of CSF volume due to functional motion as differences in CSF volumes in flexion, extension, and compression compared to the normal position in the 43 volunteers and 47 patients with lumbar spinal stenosis. For volunteers, CSF volumes increased in flexion relative to the normal position between 5.3 % (L2-3) and 9.6 % (L4-5) and decreased in extension between 9.2 % (L1-2) and 21.5 % (L5-S1) and under compression between 6.7 % (L1-2) and 12.0 % (L5-S1) when compared to normal position. The results for compression and extension correlated strongly with coefficients between R=0.983 (L1-2) and R=0.996 (L3-4). From this, it is noted that compression is effective mainly through an extension mechanism; however, it is less extensive when compared to direct extension. We therefore decided to abandon applying compression to patients. The results for the patients show that the CSF volume increased in flexion between 9.1 % (L2-3) and 18.1 % (L3-4) and decreased in extension between 10.6 % (L1-2) and 16.7 % (L3-4) with respect to normal position. Figure 6 shows the results for CSF volumetry in movement function for the same patient as in Fig. 5, another example of a 75-year-old patient with central stenosis at L3-4, and leftsided lateral stenosis in segment L4-5 is shown in Fig. 7. Detectability
CSF volumetry in motion function The validation of automated volumetric measurements with the phantom showed deviations of the volume values of CuSO4 solution from 0.09 to 2.39 % with a total volume of
Table 3 summarizes the findings and sensitivities of the imaging methods compared with intraoperative findings in patients with degenerative spondylolisthesis and lumbar spinal stenosis. With MRM, more stenotic segments (n=41) were
Table 1 Functional range of motion (ROM), flexion, and extension of the lumbar spine for 43 volunteers, 47 patients, and values from the literature Lumbar levels L1-2
L2-3
L3-4
L4-5
L5-S1
Volunteers flexion
88.6°±2.5°
92.9°±3.8°
96.0°±2.6°
95.7°±3.0°
102.9°±3.5°
Volunteers extension Volunteers ROM Pat. MRM flexion Pat. MRM extension Pat. MRM ROMa Pat. cMyelo flexion Pat. cMyelo extension Pat. cMyelo ROMb Literaturec
101.6°±2.8° 13.0°±2.0° 90.7°±3.3° 98.3°±4.0° 7.6°±3.3° 91.7°±2.9° 97.6°±4.6° 5.9°±2.5° 11.9°
106.6°±3.3° 13.7°±2.8° 94.9°±3.4° 101.6°±4.1° 6.7°±3.3° 94.7°±4.7° 100.6°±7.2° 5.9°±2.7° 14.5°
109.9°±3.7° 13.9°±2.7° 98.1°±5.5° 105.0°±5.6° 6.9°±3.7° 97.7°±2.7° 103.6°±6.8° 5.9°±3.4° 15.3°
111.1°±3.6° 15.4°±3.1° 97.6°±5.5° 107.6°±5.2° 10.0°±4.1° 98.7°±6.8° 107.3°±6.9° 8.6°±5.3° 18.2°
119.6°±3.9° 16.7°±3.0° 104.7°±8.4° 116.1°±7.5° 11.4°±5.7° 106.1°±7.3° 115.3°±6.8° 9.2°±4.4° 17.0°
Functional range of motion is the difference of angles between flexion and extension in degrees a
Pat. MRM ROM=range of motion of patients determined on magnetic resonance myelography
b
Pat. cMyelo ROM=range of motion of patients determined on conventional myelography
c
Dvorak et al. [18]
1074
Neuroradiology (2014) 56:1069–1078
Fig. 5 Determination of the functional range of motion of a 78-year-old female patient with multi-segmental central lumbar spinal stenosis and degenerative spondylolisthesis. Measurement of listhesis as described previously by Dupuis13 on conventional radiograms (top and bottom left) and functional lumbar MR examination (SINOP sequence, in-phase) in flexion (top middle and right) and extension (bottom middle and right) shows the same functional range of motion. An increased anterolisthesis at the level L4-5 from 3 to 10 mm is also recognizable. Note: angle la=lb= 3°, angle lc=1d=6°. Increase of anterolisthesis: distance 2c=2d= 0.7 cm, distance 2a=2b=1.0 cm
detected as with conventional myelography (n=23) including pmCT (n=28). Here, both axillary and peripheral nerve root sheaths were better assessed in MRM compared with conventional myelography and pmCT. The detectability of MRM for lumbar spinal stenosis was 100 %, which corresponded to the intraoperative gold standard. The detectability of conventional myelography was 58 %, and that of pmCT was 68 %. While spondylolisthesis was reliably detected with MRM, conventional myelography, and pmCT (detectability=100 % for all methods), examinations in motion function were only possible with MRM and conventional myelography. The
axillary region and the peripheral nerve root sheaths were depicted and evaluated in MRM only. The detectability of MRM (100 %) corresponded to the gold standard, i.e., the intraoperative findings. Conventional myelography and pmCT had lower detectability due to false-positive findings for lumbar spinal stenosis caused by overemphasis of liquid compression and due to methodological and technical reasons for spondylolistheses, i.e., assessment of the nerve roots in the lateral projection is not possible and functional pmCT examinations are not feasible (increased radiation exposure to the patient).
Table 2 Functional changes of CSF volumes in MRM in flexion, extension, and compression compared to normal position at 43 volunteers and 47 patients with lumbar spinal stenosis Level
Volunteers F–N [mm3]
Volunteers N–E [mm3]
Volunteersa N–C [mm3]
Patients F–N [mm3]
Patients N–E [mm3]
L1-2 L2-3 L3-4 L4-5 L5-S1
149 (6.3 %) 116 (5.3 %) 168 (7.5 %) 227 (9.6 %) 190 (7.9 %)
204 (9.2 %) 293 (14.0 %) 266 (12.8 %) 309 (14.5 %) 475 (21.5 %)
148 (6.7 %) 172 (8.2 %) 180 (8.6 %) 238 (11.9 %) 264 (12.0 %)
230 (12.2 %) 172 (9.1 %) 202 (18.1 %) 165 (15.4 %) 193 (11.7 %)
200 (10.6 %) 204 (10.9 %) 186 (16.7 %) 121 (11.2 %) 226 (13.8 %)
Functional changes of CSF volumes as difference of volume between function (flexion, extension, and compression) and normal position. Changes of CSF volumes relative to normal position in percentage terms are shown in parentheses F flexion, N normal position, E extension, C compression a
Twenty volunteers were investigated under compression
Neuroradiology (2014) 56:1069–1078
1075
Fig. 6 Functional volumetry of a 78-year-old female patient with multi-segmental central lumbar spinal stenosis and degenerative spondylolistheses (same patient as in Fig. 5) in extension (left) and flexion (right). The CSF volume nearly doubles in flexion at the stenosed levels (L2-3, 86.8 % and L3-4, 78.9 %) and increased only by 24.8 % at level L4-5 (red arrow), however
Postoperative outcome A total of 10 patients were pre- and postoperatively examined by MRM and one patient additionally by conventional myelography. These 10 patients underwent laminectomy due to lumbar spinal stenosis. The postoperative examinations were performed 3–6 months after surgery. Figure 8 shows pre- and postoperative MRM examinations of a 75-year-old patient with bilateral stenosis of the lumbar spine. A correlation analysis between the score of degree of postoperative freedom from symptoms and the postoperative change in CSF volume in the stenotic levels on MRM for the 10 patients gave a correlation coefficient of R=0.772. This indicates that the postoperative improvement on the extent of postoperative decompression showed a strong correlation.
Table 3 Detectability of the methods compared with intraoperative findings
Fig. 7 Functional volumetry of a 75-year-old female patient with central stenosis at level L3-4 and left-sided lateral stenosis at level L4-5. Functional lumbar MR myelography in flexion (top row), extension (middle row), and conventional myelography in flexion (bottom row) shows consistently a strong defect at level L3-4 on right-handed side (yellow arrows) and an axillary defect on left-handed side at level L4-5 (red arrows)
Criteria for assessment Detectability Detectability MRM Conv. myelo./ pmCT
Detectability Surgery
Stenoses recognized Axilla definable Nerves depicted Listhesis recognized Increase in function Nerve root compression
41 (100 41 (100 41 (100 15 (100 n.a. 15 (100
41 (100 %) 38 (93 %) 38 (93 %) 15 (100 %) 15 (100 %) 15 (100 %)
23/28 (56/68 %) 5/15 (12/36 %) 5/15 (12/36 %) 15 (100 %) 15/0 (100/0 %) n.a.
%)a %)a %)a %) %)
The figures are the number of findings; the percentage related to the gold standard of intraoperative findings are in parentheses MRM MR myelography, conv. myelo./pmCT conventional myelography/ postmyelographic CT, n.a. not applicable a
Two patients were operated a second time due to worsening of symptoms after first surgery. Whereby, three stenoses, which were not found during first surgery but in preoperative MRM, were confirmed during second surgery
1076
Neuroradiology (2014) 56:1069–1078
Fig. 8 Pre- and postoperative results of lumbar MR myelography of a 75-year-old female patient with lateral stenosis at level L4-5. Preoperative MR myelography showed a signal cancelation at level L4-5 (left dorsolateraly); the CSF volume in this segment was 1413 mm3. This value increased after 7 days postoperatively to 2096 mm3 and normalized at 1945 mm3 after 4 months. The slightly lower value for the CSF volume at the second postoperative measurement was explained due to scar tissue which has been formed in the meantime
Discussion In this study, we demonstrated the possibility to study the functional range of motion of the lumbar spine in a conventional whole-body MR scanner. Our MR methods achieved sufficient data quality even without the magnetic field’s isocenter. This approach guarantees detectability in preoperative diagnosis lumbar motion function, which was even superior to the results from conventional myelography and pmCT. Very good correlations between pre- and postoperative functional MRM with both intraoperative findings and postoperative outcome were found. We were able to quantitatively define CSF sub-volumes in functional positions by using direct volume rendering. Function-dependent measurements were described in numerous publications using open MRI systems with mostly low field strength [19] or only under axial load [20]. However, in all in vivo studies published so far, only 2D analyses (spinal distances and areas) were carried out [21, 22], but this strategy is associated with large scattering and poor reproducibility of
values. 3D analyses were described only in in vitro measurements on cadavers [23, 24] and in animal models [25] and are therefore not directly comparable with our results. In volumetry studies, shifts in functional positions or rotational deformities (scoliosis) play not an essential role since variations between imaging plane and spinal axis to be compensated have considerable influence in 2D parameters [19]. The methodical error of or volumetric approach was found with 0.9 to 2.7 % using data from a phantom. A further important advantage of our method is that the peripheral CSF sheaths can be assessed quantitatively too. The values given in the literature [22, 24] for the range of motion show a relatively large range but are in good accordance with our findings. Studies were carried out with open MR scanners using vertically aligned magnetic fields of low field strength which allowed for some degree of movement [19]. However, the spatial resolution of the MR data, which can be achieved with these MR scanners, is far too insufficient to perform detailed analysis or 3D examinations [1]. Furthermore, studies in sitting or in the lateral position without
Neuroradiology (2014) 56:1069–1078
position fixing devices allow only poorly controlled and not standardized examinations. Limited mobility in older patients with lumbar spinal stenosis found in our study population is confirmed by other studies [26, 27]. Our results also demonstrated that compression of the lumbar spine is associated and comparable with lordosis in extension. The resulting reduction in volume of the spinal canal under compression relative to the normal position is low compared with that in extension. We therefore decided to completely dispense with applying functional studies under compression, i.e., axial load. The differences in total range of motion between patients and volunteers can be explained by relieving posture due to pain and limited mobility. However, one has to keep in mind that the differences in range of motion and CSF volumes can be partly influenced by age differences between volunteers and patients too. Therefore, we decided to present no p values of significance for the differences between volunteers and patients. Our results are confirmed by Madsen et al. [28] but are also in contrast to other groups that see relevance for axial load MRI by correlation with clinical parameters [29, 30]. However, it must be expressly pointed out that these measurement studies have only been conducted under motion function under axial load, and neither under flexion or extension. Coulier et al. [31] demonstrated that the upright myelography and especially during upright extension was the most efficient technique to detect pathological narrowing of the dural sac. They found only minor discrepancies between the neutral position and the extension showing that already when the patient was upright, there was spontaneous extension. It also appeared from their results that flexion was of limited value and therefore of no great interest in all diagnostic techniques. It was also empirically demonstrated that 60 to 80 mm2 was also the correct landmark of lumbar stenosis and that this measure of surface was more specific and more reproducible than the conventional measure of the diameter only. Our visualization results of MRM in patients with lumbar spinal stenosis were in stronger accordance with the intraoperative findings. Thus, MRM showed a higher detectability than conventional myelography and pmCT. Stenoses were overrated in pmCT and therefore lead to increased falsepositive findings as demonstrated previously [32]. The advantages of our MRM approach for the assessment of stenoses in the lumbar spine compared to conventional myelography and pmCT have been shown in previous studies [16, 33]. In all cases of high grade stenosis, height and length of the main stenosis could be sufficiently localized and determined using our MRM approach. Spondylolisthesis could be as easily detected and evaluated. In contrast to conventional myelography, we were able to assess axilla region and peripheral root sheaths with functional motion MRM, which is also sometimes challenging in conventional MRI [34]. Furthermore, our study showed that the change in CSF volume on MRM as a result of lumbar decompressive surgery strongly
1077
correlates with postoperative clinical findings. These results, however, are contradictory to other postoperative MRI studies [35–38], which described that early postoperative MR controls lead to false-positive results and late MR controls to false-negative results. This may be explained by circumscribed local changes (adhesions between the dura and muscle in area of the laminectomy or circumscribed arachnoiditis) which could lead to incorrect results in imaging. To sum up, one could say that for preoperative evaluation, the volume change between flexion and extension is the most significant parameter. Our results for volunteers under compression demonstrated that compression is effective mainly through an extension mechanism. However, it is less extensive when compared to direct extension. Furthermore, extension is easier to implement and is better tolerated by the subjects compared to compression. Flexion is preferable to normal position due to the larger absolute CSF volumes. Changes in CSF volume between flexion and extension allow for estimation of functional variability and remaining CSF volume to relieve the nerve roots. For postoperative evaluation, the change between pre- and postoperative CSF volume in flexion is most meaningful because it reveals the maximum increase in CSF volume caused by surgery. The patients included in our study represented a population of patients with severe stenoses but not with a mild stenosis. In this regard, the results of the conventional myelography were not disastrous but may reflect the disadvantage of conventional myelography in patients with severe spinal stenosis. In conclusion, clinically reliable assignment of functionally related symptoms (e.g., of instabilities) to main stenosis height and length is often challenging because of the diversity of symptoms. Since decompressive surgery is usually the first choice for treatment of neural compression, medical imaging has to significantly and reliably contribute to preoperative treatment planning. Necessary diagnostic criteria can be fulfilled by this advanced MRM approach, which allows for sufficient visualization of neural structures under motion function. Ethical standards and patient consent We declare that all human and animal studies have been approved by the Ethics Committee of the University Erlangen and have therefore been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. We declare that all patients gave informed consent prior to inclusion in this study. Conflict of interest We declare that we have no conflict of interest.
References 1. Eberhardt K, Ganslandt O, Stadlbauer A (2013) Improved magnetic resonance myelography using image fusion. Fortschr Röntgenstr 185:333–339
1078 2. Hans FJ, Reinges MH, Krings T (2004) Lumbar nerve root avulsion following trauma: balanced fast field-echo MRI. Neuroradiol 46: 144–147 3. Siebert E, Prüss H, Klingebiel R et al (2009) Lumbar spinal stenosis: syndrome, diagnostics and treatment. Nat Rev Neurol 5:392–403 4. Saifuddin A (2000) The imaging of lumbar spinal stenosis. Clin Radiol 55:581–594 5. Adamova B, Vohanka S, Dusek L (2005) Dynamic electrophysiological examination in patients with lumbar spinal stenosis: is it useful in clinical practice? Eur Spine J 14:269–276 6. Amundsen T, Weber H, Lilleas F et al (1995) Lumbar spinal stenosis. Clinical and radiologic features. Spine (Phila Pa 1976) 20:1178–1186 7. Bal S, Celiker R, Palaoglu S, Cila A (2006) F wave studies of neurogenic intermittent claudication in lumbar spinal stenosis. Am J Phys Med Rehabil 85:135–140 8. Boden SD, Davis DO, Dina TS et al (1990) Abnormal magneticresonance scans of the lumbar spine in asymptomatic subjects. A prospective investigation abnormal lumbar magnetic-resonance spine scans of the in asymptomatic. J Bone Jt Surg 72:403–408 9. Moon E-S, Kim H-S, Park J-O et al (2005) Comparison of the predictive value of myelography, computed tomography and MRI on the treadmill test in lumbar spinal stenosis. Yonsei Med J 46: 806–811 10. Weinstein JN, Tosteson TD, Lurie JD et al (2008) Surgical versus nonsurgical therapy for lumbar spinal stenosis. N Engl J Med 358: 794–810 11. Herno A, Airaksinen O, Saari T, Miettinen H (1994) The predictive value of preoperative myelography in lumbar spinal stenosis. Spine (Phila Pa 1976) 19:1335–1338 12. Klasen J, Antoch G, Blondin D (2011) MR imaging of the abdomen in pregnancy. Röfo 183:514–522 13. Komlosi P, Zhang Y, Leiva-Salinas C et al (2014) Adaptive statistical iterative reconstruction reduces patient radiation dose in neuroradiology CT studies. Neuroradiol 56:187–193 14. Issack PS, Cunningham ME, Pumberger M et al (2012) Degenerative lumbar spinal stenosis: evaluation and management. J Am Acad Orthop Surg 20:527–535 15. Alyas F, Connell D, Saifuddin A (2008) Upright positional MRI of the lumbar spine. Clin Radiol 63:1035–1048 16. Eberhardt K, Ganslandt O, Stadlbauer A (2014) Improved magnetic resonance myelography of the lumbar spine using image fusion and volumetry. J Neurosurg Spine 20:220–226 17. Dupuis PR, Yong-Hing K, Cassidy JD, Kirkaldy-Willis WH (1985) Radiologic diagnosis of degenerative lumbar spinal instability. Spine (Phila Pa 1976) 10:262–276 18. Dvorak J, Panjabi MM, Chang DG et al (1991) Functional radiographic diagnosis of the lumbar spine. Flexion-extension and lateral bending. Spine (Phila Pa 1976) 16:562–571 19. Vitzthum HE, Konig A, Seifert V (2000) Dynamic examination of the lumbar spine by using vertical, open magnetic resonance imaging. J Neurosurg 93:58–64 20. Meakin JR, Smith FW, Gilbert FJ, Aspden RM (2008) The effect of axial load on the sagittal plane curvature of the upright human spine in vivo. J Biomech 41:2850–2854 21. Hansson T, Suzuki N, Hebelka H, Gaulitz A (2009) The narrowing of the lumbar spinal canal during loaded MRI: the effects of the disc and ligamentum flavum. Eur Spine J 18:679–686
Neuroradiology (2014) 56:1069–1078 22. Huang KY, Lin RM, Lee YL, Li JD (2009) Factors affecting disability and physical function in degenerative lumbar spondylolisthesis of L4-5: evaluation with axially loaded MRI. Eur Spine J 18:1851–1857 23. Bisschop A, van Royen BJ, Mullender MG et al (2012) Which factors prognosticate spinal instability following lumbar laminectomy? Eur Spine J 21:2640–2648 24. Fujii R, Sakaura H, Mukai Y et al (2007) Kinematics of the lumbar spine in trunk rotation: in vivo three-dimensional analysis using magnetic resonance imaging. Eur Spine J 16:1867–1874 25. Chan DD, Khan SN, Ye X et al (2011) Mechanical deformation and glycosaminoglycan content changes in a rabbit annular puncture disc degeneration model. Spine (Phila Pa 1976) 36:1438–1445 26. Harvey SB, Smith FW, Hukins DW (1998) Measurement of lumbar spine flexion-extension using a low-field open-magnet magnetic resonance scanner. Invest Radiol 33:439–443 27. McGregor AH, McCarthy ID, Dore CJ, Hughes SP (1997) Quantitative assessment of the motion of the lumbar spine in the low back pain population and the effect of different spinal pathologies of this motion. Eur Spine J 6:308–315 28. Madsen R, Jensen TS, Pope M et al (2008) The effect of body position and axial load on spinal canal morphology: an MRI study of central spinal stenosis. Spine (Phila Pa 1976) 33:61–67 29. Kanno H, Endo T, Ozawa H et al (2012) Axial loading during magnetic resonance imaging in patients with lumbar spinal canal stenosis: does it reproduce the positional change of the dural sac detected by upright myelography? Spine (Phila Pa 1976) 37:E985–E992 30. Kanno H, Ozawa H, Koizumi Yet al (2012) Dynamic change of dural sac cross-sectional area in axial loaded magnetic resonance imaging correlates with the severity of clinical symptoms in patients with lumbar spinal canal stenosis. Spine (Phila Pa 1976) 37:207–213 31. Coulier B (2000) Evaluation of lumbar canal stenosis: decubitus imaging methods versus flexion-extension myelography and surface measurements versus the diameter of the dural sac. J Belge Radiol Belgisch Tijdschr voor Radiol 83:61–67 32. Zander DR, Lander PH (1998) Positionally dependent spinal stenosis: correlation of upright flexion-extension myelography and computed tomographic myelography. Can Assoc Radiol J 49:256–261 33. Eberhardt KE, Hollenbach HP, Tomandl B, Huk WJ (1997) Threedimensional MR myelography of the lumbar spine: comparative case study to X-ray myelography. Eur Radiol 7:737–742 34. Weishaupt D, Schmid MR, Zanetti M et al (2000) Positional MR imaging of the lumbar spine: does it demonstrate nerve root compromise not visible at conventional MR imaging? Radiol 215:247–253 35. Awwad EE, Smith KR Jr (1999) MRI of marked dural sac compression by surgicel in the immediately postoperative period after uncomplicated lumbar laminectomy. J Comput Assist Tomogr 23:969–975 36. Herno A, Partanen K, Talaslahti T et al (1999) Long-term clinical and magnetic resonance imaging follow-up assessment of patients with lumbar spinal stenosis after laminectomy. Spine (Phila Pa 1976) 24: 1533–1537 37. Leonardi MA, Zanetti M, Saupe N, Min K (2010) Early postoperative MRI in detecting hematoma and dural compression after lumbar spinal decompression: prospective study of asymptomatic patients in comparison to patients requiring surgical revision. Eur Spine J 19: 2216–2222 38. Ross JS (1999) MR imaging of the postoperative lumbar spine. Magn Reson Imaging Clin N Am 7:513–524, viii
