- Email: [email protected]
Diffusion Tensor Magnetic Resonance Imaging and Fiber Tractography of the Sacral Plexus in Children with Spina Bifida
Author's Accepted Manuscript Diffusion tensor MRI and fiber tractography of the sacral plexus in children with spina bifida Wieke Haakma, Pieter Dik, Bennie ten Haken, Martijn Froeling, Rutger A.J. Nievelstein, Inge Cuppen, Tom P.V.M. de Jong, Alexander Leemans
PII: DOI: Reference:
S0022-5347(14)03417-X 10.1016/j.juro.2014.02.2581 JURO 11432
To appear in: The Journal of Urology Accepted Date: 10 February 2014 Please cite this article as: Haakma W, Dik P, ten Haken B, Froeling M, Nievelstein RAJ, Cuppen I, de Jong TPVM, Leemans A, Diffusion tensor MRI and fiber tractography of the sacral plexus in children with spina bifida, The Journal of Urology® (2014), doi: 10.1016/j.juro.2014.02.2581. DISCLAIMER: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our subscribers we are providing this early version of the article. The paper will be copy edited and typeset, and proof will be reviewed before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to The Journal pertain. All press releases and the articles they feature are under strict embargo until uncorrected proof of the article becomes available online. We will provide journalists and editors with full-text copies of the articles in question prior to the embargo date so that stories can be adequately researched and written. The standard embargo time is 12:01 AM ET on that date.
ACCEPTED MANUSCRIPT Title page Diffusion tensor MRI and fiber tractography of the sacral plexus in children with spina bifida
RI PT
Running title: DTI of the sacral plexus in spina bifida patients Authors
Wieke Haakma, MSc1,2, Pieter Dik, PhD3, Bennie ten Haken, PhD4, Martijn Froeling, PhD1, Rutger A.J. Nievelstein, PhD1, Inge Cuppen, PhD5, Tom P.V.M. de Jong, PhD3, Alexander
SC
Leemans, PhD6
M AN U
Affiliations 1
Department of Radiology, University Medical Center Utrecht, Utrecht, the Netherlands
2
Department of Department of Forensic Medicine and Comparative Medicine Lab, Aarhus
University, Aarhus, Denmark
Department of Pediatric Urology, University Children's Hospitals UMC Utrecht, Utrecht, the
TE D
3
Netherlands 4
University of Twente, MIRA Institute for Biomedical Technology and Technical Medicine,
5
Department of Pediatric Neurology, University Children's Hospitals UMC Utrecht, Utrecht,
AC C
the Netherlands 6
EP
Neuro Imaging, Enschede, the Netherlands
Image Sciences Institute, University Medical Center Utrecht, Utrecht, the Netherlands
Address for correspondence: Pieter Dik University Medical Center Utrecht Department of Pediatric Urology
ACCEPTED MANUSCRIPT Lundlaan 6 3508 AB, Utrecht the Netherlands Telephone: +31 887554004
RI PT
Fax number: +31 887555348 [email protected]
Word count abstract: 250
AC C
EP
TE D
M AN U
Word count manuscript (excluding abstract): 2410
SC
Key terms: Diffusion tensor imaging; Microstructural properties; Sacral plexus; Spina bifida; Fiber tractography
1
ACCEPTED MANUSCRIPT Abstract Purpose: It is still largely unknown how neural tube defects in spina bifida (SB) affect the nerves at the level of the sacral plexus. Visualizing the sacral plexus in 3D could improve our
RI PT
anatomical understanding regarding neurological problems in SB patients. We investigate anatomical and microstructural properties of the sacral plexus of SB patients with diffusion tensor imaging (DTI) and fiber tractography (FT).
SC
Materials and methods: 10 SB patients (8-16 years) underwent DTI on a 3T magnetic resonance imaging system (MRI). Anatomical 3D reconstructions of the sacral plexus of 10
M AN U
SB patients were obtained. FT was performed with the diffusion MRI-toolbox ExploreDTI to determine the fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD) in the sacral plexus of SB patients. Results were compared to 10 healthy controls.
TE D
Results: Nerves of SB patients showed to a large extent asymmetry and disorganisation compared to healthy controls. Especially at the level of the myelomeningocele, it was difficult to find a connection with the cauda equina. Furthermore, the MD, AD, and RD values in S1-S3
EP
were significantly lower in the SB patients.
AC C
Conclusions: This 3T MRI study shows for the first time asymmetry and disorganization of the sacral plexus in 10 SB patients with DTI and FT. The observed difference in diffusion values shows that these methods can be used to identify nerve abnormalities. We expect that this technique could provide a valuable contribution to a better analysis and understanding of the problems of SB patients in the future.
2
ACCEPTED MANUSCRIPT Introduction The incidence of spina bifida (SB) worldwide ranges from 0.3-4.5 per 1000 births [1, 2]. Patients with SB generally suffer from neurogenic bladder dysfunction and affected sensory and motor innervation of the lower limbs [3]. The contraction and relaxation of the bladder
RI PT
sphincter is coordinated by the central and peripheral nervous system. In SB patients, this sphincter function can be affected [4]. Without treatment, the level to which this function is compromised determines the prognosis of these patients. Early diagnosis is critical to prevent
SC
further neurologic impairment, but can be quite arduous due to the complex anatomical configuration and high inter-subject variability of the peripheral sacral branches [5, 6]. At
M AN U
present, there are no reliable in vivo and noninvasive routine clinical techniques available to determine how the sacral plexus is organized on an anatomical or structural level in SB patients.
A technique that allows for 3D visualization and structural characterization of nerve tissue is
TE D
diffusion tensor imaging (DTI) [7-9]. DTI is an MRI technique that is sensitive to the random movement of diffusing water molecules, the so-called Brownian motion. The diffusion is
EP
more pronounced along the nerves than across their main orientation, causing diffusion to show a high degree of anisotropy in such fibrous tissue [10]. This anisotropy can be
AC C
quantified by applying a diffusion-weighted MRI acquisition protocol with multiple diffusion gradient orientations and, subsequently, estimating the diffusion tensor [7]. This enables the possibility to reconstruct the 3D architecture of peripheral nerves noninvasively, and is referred to as fiber tractography (FT) [11]. Although DTI has been used in several studies on peripheral nerves [12, 13], it has been rarely used in the lumbosacral region [5]. The potential value of DTI to quantify peripheral nerve injury or dysfunction looks promising, but it is still questionable to what extent this can be translated to a clinical setting given the numerous fiber constituents that may modulate the
3
ACCEPTED MANUSCRIPT observed DTI results [10]. The four diffusion parameters that are commonly used to investigate the tissue microstructural properties are: 1) the fractional anisotropy (FA), which is high when water molecules are moving predominantly along one direction, 2) the mean diffusivity (MD), which is the average of all eigenvalues (the overall amount of diffusion),
RI PT
3) the axial diffusivity (AD), which is equal to the largest eigenvalue, and 4) the radial diffusivity, which is defined as the average of the second and third eigenvalues (see Fig. 1).
SC
[FIG 1]
While DTI [7, 8] has been widely used to investigate white matter tracts in the brain (e.g., [11,
M AN U
14-16]), its application to peripheral nerves is still limited, mainly due to the bigger challenges related to data acquisition [5, 12, 13]. Takagi et al. examined the nerve regeneration of the sciatic nerve in rats following contusive injury. The correlation of fractional anisotropy (FA) values with both histological and functional changes observed in
TE D
that work demonstrates the potential clinical value of DTI in peripheral nerve damage and repair, with lower FA values indicating damage to peripheral nerves [17]. Van der Jagt et al. have recently shown promising progress in reconstructing and analyzing the peripheral sacral
EP
nerves using DTI and FT [5].
AC C
In this work, we investigated the sacral plexus in 10 children with SB with neurogenic bladder dysfunction using DTI and FT and compared the results with healthy controls. Our hypothesis is that for these SB patients, the microstructural properties of the sacral plexus would show abnormal values of DTI parameters in nerve regions where the tissue structure has been affected. Our results demonstrate that with DTI and FT (i) peripheral sacral nerves in SB patients with neurogenic bladder dysfunction can be reconstructed and visualized with great detail, (ii) the microstructural tissue organization of the sacral plexus can be characterized in vivo and noninvasively, and (iii) peripheral nerve tissue abnormalities can be identified at
4
ACCEPTED MANUSCRIPT specific locations in the sacral plexus. We believe that investigating nerve tissue properties with DTI and FT in addition to conventional MRI can be helpful to better understand the mechanism of disturbed innervations of the bladder and lower limb muscles in children with
RI PT
SB. Materials and methods Data acquisition
SC
Local institutional review board approval was obtained for this study and written informed consent was given prior to the MRI examination. Ten patients with SB and neurogenic
M AN U
bladder dysfunction, 6 girls, 4 boys, mean age 11.4 years (range 8-16 years) were included. Neural tube defects (myelomeningocele) were mostly located in the lumbar sacral region (L5S2). To reduce variability within the patient group and keep the acquisition time to a minimum, MRI was performed from the level of L4 to the pelvic floor region. The
TE D
healthy controls were the participants from the study by van der Jagt et al [5]. All subjects were examined with a 3 Tesla MR system (Achieva; Philips Healthcare, Best, The Netherlands) using a 16-channel phased-array surface coil. Diffusion weighted images and
EP
anatomical 3D Turbo Spin Echo (3D-TSE) T2-weighted images were obtained using the
AC C
acquisition protocols as described previously (van der Jagt et al [5]).
Data processing and analysis The DTI data sets were processed using the ExploreDTI diffusion MRI toolbox (www.ExploreDTI.com) [18] as follows: 1) The data were corrected for subject motion and for eddy current induced geometrical distortions [19], 2) diffusion tensors and subsequently diffusion parameters (MD, FA, AD, and RD) were calculated using the iteratively weighted linear regression procedure [20], and 3) a deterministic streamline tractography approach was
5
ACCEPTED MANUSCRIPT used [21] to reconstruct the fiber pathways. In each nerve, a ‘SEED’ region-of-interest (ROI) was placed in the middle of the nerve. Then, a second ROI (‘AND’ region) was selected in the nerve root near the cauda equina and another ‘AND’ ROI more distally along the nerve where it is still traceable. By placing these ‘AND’ regions according to the 3D TSE T2
RI PT
dataset, fiber trajectories can be obtained along the entire nerve. Combining a low FA threshold (i.e., 0.001) in combination with these ‘AND’ ROIs provides a feasible way to reconstruct tracts with a high reproducibility [5]. Finally, the accuracy of the anatomical
SC
location of the fiber tracts was evaluated with the anatomical (3D TSE T2-weigthed) images
M AN U
by a pediatric neuroradiologist with more than 15 years of experience in MRI.
Statistical analyses
The FA, MD, AD, and RD values of the nerves L4 to S3 of SB patients were investigated and compared with the values in healthy controls as obtained in the study by van der Jagt et al [5].
TE D
The variability between the nerves of SB patients was investigated per nerve using the Kruskall-Wallis test. Furthermore, the nerves of the left and right side were compared using a non-parametric Mann-Whitney U test. The same test was also used to investigate the
EP
differences between nerves of SB patients and healthy controls. The analysis was performed
Results
AC C
using SPSS version 21.0 (SPSS Inc. Chicago, IL, USA).
With tractography it was possible to obtain 3D anatomical reconstructions of all 10 SB patients (see supplemental material). Fig. 2 shows the FT results of lumbosacral plexus of a typical SB patient with pseudo-color encoding representing the magnitude of the diffusion parameters. Table 1 summarizes the diffusion measures for the SB patients and the healthy controls. As
6
ACCEPTED MANUSCRIPT the MD values appear to be most indicative to show abnormalities in the nerves (see Table 1), the fiber tracts are color coded for MD in the remaining figures. [FIG 2]
RI PT
Fiber tract evaluation The sacral plexus of SB patients shows asymmetry and disorganization compared to healthy controls. In two SB patients, nerves at the level of L5 could not be reconstructed with FT despite
SC
the fact that they were visible on the anatomical T2-weighted images (see Fig. 3B). In none of the SB patients, S4 and S5 could be traced, which was also the case for the healthy control group.
M AN U
At this level, they were in most cases also not visible on the anatomical T2-weighted images. [FIG 3]
In the SB patients, it was difficult to locate a connection to the cauda equina, especially at the
healthy controls (Fig. 4B).
EP
[FIG 4]
TE D
level of the myelomeningocele (Fig. 4A). By contrast, this connection was clearly visible in the
Diffusion parameters
AC C
Statistical testing revealed no significant difference in FA, MD, AD, and RD between the left and right side at the levels L4-S3 (p>0.05). Therefore, both sides were pooled for each nerve level. Furthermore, no significant differences were found between individual nerves per level (p>0.05). The MD at the level of L4 of the SB patients was comparable to healthy controls as shown in Table 1. However, from the level of L5 downward, a significantly lower MD was found in SB patients compared to healthy controls at the levels S1-S3. Similar differences were found for the AD and RD parameters.
7
ACCEPTED MANUSCRIPT [TABLE 1] Innervation of the bladder A trajectory originating from S2-S4 and continuing to the bladder was found in three patients
RI PT
(see Fig 5A for an example in one patient). Although this could not be confirmed by the anatomical T2-weighted images, it is likely that this is the pudendal nerve, as it shows a similar course as presented on a schematic anatomy configuration of the sacral plexus (Fig. 5B).
SC
[FIG 5]
M AN U
Discussion
To the best of our knowledge, this is the first 3-Tesla DTI study that is able to reconstruct the 3D architecture of the sacral plexus in 10 SB patients, detailing the individual pathway
TE D
trajectories and the microstructural properties of L4-S3. Differences in both anatomy (in particular, asymmetry and disorganization) and mean diffusivity values in the sacral nerves were found between SB patients and healthy controls.
EP
Interpretation
AC C
Fiber tractography provides striking images of the sacral plexus non-invasively [5]. However, this technology has not been used extensively for investigating peripheral nerves as DTI and FT are not straightforward and trivial to apply. In addition, one of the major difficulties of DTI is interpretation [22]. Abnormalities in fiber reconstructions and diffusion parameters can also be caused due to deformations and artifacts in the data or limitations of the imaging technique itself [23, 24]. Clinical relevance
8
ACCEPTED MANUSCRIPT In this study, MD values from the level of L5 to caudal are lower in SB patients than controls, whereas for L4, which is still intact, the MD values were similar between both subject groups. The reduction in MD could involve reduced intrinsic diffusion in the intraaxonal space due to cytoskeletal breakdown resulting in an increased viscosity [10]. It can
RI PT
also be related to the myelomeningocele which in these patients is mostly located at the level of L5-S1. At that level, it was also difficult to find a connection to the cauda equina. With some abuse of terminology, if nerves at that level are not working properly, this might be
SC
compared to an electric cable which is not adequately plugged into the socket. To investigate whether the lower MD values of these nerves are caused by the myelomeningocele, these
higher level (thoracic or high lumbar).
M AN U
results could be compared with SB patients where the myelomeningocele is located at a
DTI was able to show a trajectory going to the bladder, which is likely to be the pudendal nerve. The anatomical image was not able to visualize this trajectory. As the pudendal nerve
TE D
originates from S2-S4 it is expected that the pudendal nerve will also have lower MD values. Future studies investigating the correlation between urodynamic findings and DTI metrics in
EP
the sacral plexus of SB patients may support this hypothesis. The patient group can be extended by investigating other sacral malformations such as anorectal malformation [25].
AC C
Lower lesions are often associated with detrusor sphincter dyssynergy compared to higher lesions. DTI could play an important role in diagnosis and therapy in future neuroanastomotic procedures in these patients. Limitations Although the data have been corrected for subject motion and geometrical distortions, the data could still be misaligned due to nonlinear behavior of these artifacts [5, 9]. In addition, due to the relatively large voxel size (3x3x3 mm) partial volume effects will have an effect on
9
ACCEPTED MANUSCRIPT the diffusion parameters and smaller nerve bundles, especially at S3 as they are smaller in diameter than the other nerves [23, 26]. The control group consists of healthy adults. Although the diffusion properties may change with age, such as in brain white matter fiber bundles [27], we expect that the differences in AD and RD will be relatively low in
RI PT
comparison to the differences in SB patients compared to healthy adults. Future work
SC
To determine to what extent diffusion parameters represent functionality of nerves in SB patients, other patient groups with neurological disorders can be investigated. In newborn SB
M AN U
patients, DTI can be performed before and after closing the spinal cord as a comparison tool for localizing potential nerve damage. By improving the resolution of DTI we believe that in the future, it can be possible to visualize small nerves in these infants. Other pathologies in which DTI can be used to visualize the peripheral nerves include multiple
TE D
sclerosis [28], and paraplegia. Finally, DTI could potentially be applied to locate the nerves for sacral nerve stimulation, neurostimulator implants, and neuroanastomis procedures [29] in the sacral plexus. Combining information of an electromyogram with DTI tractography and
AC C
of peripheral nerves.
EP
diffusion parameters may improve the understanding of how to interpret diffusion parameters
At this moment, DTI cannot replace conventional anatomical imaging modalities, i.e. 2D T1and T2-weighted sequences or high-resolution 3D TSE protocols, but it can be used as a complementary tool to better detect and understand the neurological problems in SB patients. Further adjustments of the registration process by fusing the anatomical and DTI information may improve the analysis of the nerves at each level. Large scale studies in patients with different types of SB are necessary to further optimize the technique and better determine the
10
ACCEPTED MANUSCRIPT exact role of DTI in the diagnosis and follow-up of patients with SB. We expect that abnormal diffusion parameters can be indicative for affected or dysfunctional nerves.
Conclusion
RI PT
This 3T MRI study shows for the first time asymmetry and disorganization of the sacral plexus in 10 SB patients using DTI and FT. These abnormalities indicate that the sacral plexus of SB patients is different from the healthy controls. The observed difference in diffusion
SC
values shows that these methods can be used to identify nerve abnormalities. Combining 3D TSE, DTI, and FT and correlating diffusion parameters with neurological problems of SB
M AN U
patients, is expected to provide a valuable contribution to a better analysis and diagnosis of these patients in the future.
Acknowledgment
TE D
We would like to thank Niels Blanken (MRI radiographer, department of radiology) for his
AC C
EP
help in acquiring the MRI data.
11
ACCEPTED MANUSCRIPT References
8. 9. 10. 11. 12. 13.
14. 15. 16.
17. 18. 19. 20. 21. 22.
23.
RI PT
7.
SC
6.
M AN U
4. 5.
TE D
3.
EP
2.
de Jong TPVM, Chrzan R, Klijn AJ, et al. Treatment of the neurogenic bladder in spina bifida. Pediatr Nephrol. 2008; 23: 889-896. Dik P, Klijn AJ, van Gool JD, et al. Early start to therapy preserves kidney function in spina bifida patients. European urology 2006; 49: 908-913. Özek MM, Cinalli G, Maixner WJ. Spina bifida - management and outcome. New York: Springer 2008. de Groat WC. A neurologic basis for the overactive bladder. Urology 1997; 50: 36-52. van der Jagt PKN, Dik P, Froeling M, et al. Architectural configuration and microstructural properties of the sacral plexus: A diffusion tensor MRI and fiber tractography study. NeuroImage 2012; 62: 1792-1799. Matejcik V. Anatomical variations of lumbosacral plexus. Surgical and Radiologic Anatomy 2010; 32: 409-414. Basser PJ, Mattiello J, LeBihan D. MR diffusion tensor spectroscopy and imaging. Biophysical journal 1994; 66: 259-267. Jones DK, Leemans A. Diffusion Tensor Imaging. In: Modo M, Bulte JWM, editors. Magnetic Resonance Neuroimaging: Humana Press 2011; 127-144. Tournier JD, Mori S, Leemans A. Diffusion tensor imaging and beyond. Magnetic resonance in medicine 2011; 65: 1532-1556. Beaulieu C. The basis of anisotropic water diffusion in the nervous system - a technical review. NMR in biomedicine 2002; 15: 435-455. Mori S, van Zijl PCM. Fiber tracking: principles and strategies - a technical review. NMR in biomedicine 2002; 15: 468-80. Hiltunen J, Suortti T, Arvela S, et al. Diffusion tensor imaging and tractography of distal peripheral nerves at 3 T. Clinical Neurophysiology 2005; 116: 2315-2323. Vargas MI, Viallon M, Nguyen D, et al. Diffusion tensor imaging (DTI) and tractography of the brachial plexus: feasibility and initial experience in neoplastic conditions. Neuroradiology 2010; 52: 237-245. Wakana S, Jiang H, Nagae-Poetscher LM, et al. Fiber tract-based atlas of human white matter anatomy. Radiology 2004; 230: 77-87. Reijmer YD, Leemans A, Caeyenberghs K, et al. Disruption of cerebral networks and cognitive impairment in Alzheimer disease. Neurology 2013; 80: 1370-1377. Reijmer YD, Brundel M, de Bresser J, et al. Microstructural white matter abnormalities and cognitive functioning in type 2 diabetes: a diffusion tensor imaging study. Diabetes care 2013; 36: 137-144. Takagi T, Nakamura M, Yamada M, et al. Visualization of peripheral nerve degeneration and regeneration: monitoring with diffusion tensor tractography. NeuroImage 2009; 44: 884-892. Leemans A, Jeurissen B, Sijbers J, et al. ExploreDTI: a graphical toolbox for processing, analyzing, and visualizing diffusion MR data. Proc Intl Soc Mag Reson Med 2009; 17: 3536. Leemans A, Jones DK. The B-matrix must be rotated when correcting for subject motion in DTI data. Magnetic resonance in medicine 2009; 61: 1336-1349. Veraart J, Sijbers J, Sunaert S, et al. Weighted linear least squares estimation of diffusion MRI parameters: Strengths, limitations, and pitfalls. NeuroImage 2013; 81: 335-346. Basser PJ, Pajevic S, Pierpaoli C, et al. In vivo fiber tractography using DT-MRI data. Magnetic Resonance in Medicine 2000; 44: 625-632. Deprez S, Billiet T, Sunaert S, et al. Diffusion tensor MRI of chemotherapy-induced cognitive impairment in non-CNS cancer patients: a review. Brain Imaging and Behavior 2013; 7: 409435. Vos SB, Jones DK, Viergever MA, et al. Partial volume effect as a hidden covariate in DTI analyses. NeuroImage 2011; 55: 1566-1576.
AC C
1.
12
ACCEPTED MANUSCRIPT
RI PT
29.
SC
28.
M AN U
27.
TE D
26.
EP
25.
Tournier JD, Calamante F, King MD, et al. Limitations and requirements of diffusion tensor fiber tracking: An assessment using simulations. Magnetic Resonance in Medicine. 2002; 47: 701-708. Boemers TM, Van Gool JD, De Jong TPVM, et al. Urodynamic evaluation of children with the caudal regression syndrome (caudal dysplasia sequence). Journal of Urology 1994; 151: 10381040. Hogan Q. Size of human lower thoracic and lumbosacral nerve roots. Anesthesiology. 1996; 85: 37-42. Bennett IJ, Madden DJ, Vaidya CJ, et al. Age-related differences in multiple measures of white matter integrity: A diffusion tensor imaging study of healthy aging. Human brain mapping 2010; 31: 378-390. Roosendaal SD, Geurts JJG, Vrenken H, et al. Regional DTI differences in multiple sclerosis patients. NeuroImage 2009; 44: 1397-1403. Rasmussen MM, Clemmensen D, Rawashdeh YF, et al. Surgical reinnervation with nerve anastomosis technique for neurogenic bladder and bowel dysfunction. Ugeskrift for laeger 2011; 173: 2412-2415.
AC C
24.
13
ACCEPTED MANUSCRIPT Legend figures Figure 1: In fluid the random movement of water molecules is isotropic. Diffusion along the nerves (represented as axial diffusivity or λ1) is higher than perpendicular to the nerves (represented as radial diffusivity or mean of λ2 and λ3). λ1, λ2, and λ3 are the eigenvalues (the length of the diffusion) in a particular orientation. This diffusion orientation preference is called anisotropy.
RI PT
Figure 2: 3D anatomy of the lumbar and sacral nerves at the level of L3-S3 of a typical spina bifida patient. Color maps of mean diffusivity, fractional anisotropy, axial diffusivity, and radial diffusivity are shown along the fiber pathways. A red color represents a high diffusivity and blue color a low diffusivity.
SC
Figure 3: Lower lumbar and sacral nerves, A) healthy control, B) spina bifida patient with myelomeningocele from the level of L5 to caudal. L5 on the left side could not be reconstructed (indicated with “1”).
M AN U
Figure 4: A) Posterior view of a spina bifida patients with myelomeningocele at the lumbosacral level of L5-S1, the sacral nerves do not connect with the cauda equina (indicated with the arrows), B) healthy control in which the nerves show a connection to the cauda equina.
TE D
Figure 5: A) Anterior view of a spina bifida patient with myelomeningocele at the lumbar level of L1-L4, the pudendal nerve is indicated with “PN”, and the sciatic nerve with “SN”, B) schematic overview of the lumbosacral nerves. The sacral branches and pudendal nerve in Fig. 5A correspond with the anatomy shown in Fig. 5B.
AC C
EP
Table 1: Mean and standard deviation of the diffusion parameters (fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD)) of the lumbar and sacral nerves of spina bifida (SB) patients and healthy controls at the level of L4-S3.
ACCEPTED MANUSCRIPT Table 1 Nerve
-3
FA
Diffusivity (mm²/s) x 10 MD
AD
RD
SB
Healthy
SB
Healthy
SB
Healthy
SB
L4
0.28±0.05
0.28±0.04
1.34±0.22
1.32±0.16
1.72±0.22
1.70±0.18
1.15±0.21
1.14±0.15
L5
0.31±0.03
0.29±0.04
1.42±0.21
1.31±0.24
1.86±0.24
1.72±0.31
1.19±0.20
1.11±0.21
S1
0.26±0.03
0.27±0.05
1.83±0.24**
1.40±0.22**
2.31±0.27**
1.78±0.23**
1.59±0.22**
1.22±0.23**
S2
0.23±0.03
0.25±0.05
1.73±0.29**
1.36±0.21**
2.13±0.36*
1.70±0.21*
1.530.26*
1.19±0.22*
S3
0.22±0.03
0.26±0.05
1.62±0.32*
1.33±0.23*
1.98±0.39*
1.66±0.35*
1.43±0.29*
1.17±0.22*
SC
RI PT
Healthy
AC C
EP
TE D
M AN U
*: p<0.01, **: p< 0.001
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
axial diffusivity
DTI
diffusion tensor imaging
FA
fractional anisotropy
FT
fiber tractography
MD
mean diffusivity
MRI
magnetic resonance imaging
RD
radial diffusivity
SB
spina bifida
TSE
turbo spin echo
AC C
EP
TE D
M AN U
SC
AD
RI PT
Key of definitions for abbreviations
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
PII: DOI: Reference:
S0022-5347(14)03417-X 10.1016/j.juro.2014.02.2581 JURO 11432
To appear in: The Journal of Urology Accepted Date: 10 February 2014 Please cite this article as: Haakma W, Dik P, ten Haken B, Froeling M, Nievelstein RAJ, Cuppen I, de Jong TPVM, Leemans A, Diffusion tensor MRI and fiber tractography of the sacral plexus in children with spina bifida, The Journal of Urology® (2014), doi: 10.1016/j.juro.2014.02.2581. DISCLAIMER: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our subscribers we are providing this early version of the article. The paper will be copy edited and typeset, and proof will be reviewed before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to The Journal pertain. All press releases and the articles they feature are under strict embargo until uncorrected proof of the article becomes available online. We will provide journalists and editors with full-text copies of the articles in question prior to the embargo date so that stories can be adequately researched and written. The standard embargo time is 12:01 AM ET on that date.
ACCEPTED MANUSCRIPT Title page Diffusion tensor MRI and fiber tractography of the sacral plexus in children with spina bifida
RI PT
Running title: DTI of the sacral plexus in spina bifida patients Authors
Wieke Haakma, MSc1,2, Pieter Dik, PhD3, Bennie ten Haken, PhD4, Martijn Froeling, PhD1, Rutger A.J. Nievelstein, PhD1, Inge Cuppen, PhD5, Tom P.V.M. de Jong, PhD3, Alexander
SC
Leemans, PhD6
M AN U
Affiliations 1
Department of Radiology, University Medical Center Utrecht, Utrecht, the Netherlands
2
Department of Department of Forensic Medicine and Comparative Medicine Lab, Aarhus
University, Aarhus, Denmark
Department of Pediatric Urology, University Children's Hospitals UMC Utrecht, Utrecht, the
TE D
3
Netherlands 4
University of Twente, MIRA Institute for Biomedical Technology and Technical Medicine,
5
Department of Pediatric Neurology, University Children's Hospitals UMC Utrecht, Utrecht,
AC C
the Netherlands 6
EP
Neuro Imaging, Enschede, the Netherlands
Image Sciences Institute, University Medical Center Utrecht, Utrecht, the Netherlands
Address for correspondence: Pieter Dik University Medical Center Utrecht Department of Pediatric Urology
ACCEPTED MANUSCRIPT Lundlaan 6 3508 AB, Utrecht the Netherlands Telephone: +31 887554004
RI PT
Fax number: +31 887555348 [email protected]
Word count abstract: 250
AC C
EP
TE D
M AN U
Word count manuscript (excluding abstract): 2410
SC
Key terms: Diffusion tensor imaging; Microstructural properties; Sacral plexus; Spina bifida; Fiber tractography
1
ACCEPTED MANUSCRIPT Abstract Purpose: It is still largely unknown how neural tube defects in spina bifida (SB) affect the nerves at the level of the sacral plexus. Visualizing the sacral plexus in 3D could improve our
RI PT
anatomical understanding regarding neurological problems in SB patients. We investigate anatomical and microstructural properties of the sacral plexus of SB patients with diffusion tensor imaging (DTI) and fiber tractography (FT).
SC
Materials and methods: 10 SB patients (8-16 years) underwent DTI on a 3T magnetic resonance imaging system (MRI). Anatomical 3D reconstructions of the sacral plexus of 10
M AN U
SB patients were obtained. FT was performed with the diffusion MRI-toolbox ExploreDTI to determine the fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD) in the sacral plexus of SB patients. Results were compared to 10 healthy controls.
TE D
Results: Nerves of SB patients showed to a large extent asymmetry and disorganisation compared to healthy controls. Especially at the level of the myelomeningocele, it was difficult to find a connection with the cauda equina. Furthermore, the MD, AD, and RD values in S1-S3
EP
were significantly lower in the SB patients.
AC C
Conclusions: This 3T MRI study shows for the first time asymmetry and disorganization of the sacral plexus in 10 SB patients with DTI and FT. The observed difference in diffusion values shows that these methods can be used to identify nerve abnormalities. We expect that this technique could provide a valuable contribution to a better analysis and understanding of the problems of SB patients in the future.
2
ACCEPTED MANUSCRIPT Introduction The incidence of spina bifida (SB) worldwide ranges from 0.3-4.5 per 1000 births [1, 2]. Patients with SB generally suffer from neurogenic bladder dysfunction and affected sensory and motor innervation of the lower limbs [3]. The contraction and relaxation of the bladder
RI PT
sphincter is coordinated by the central and peripheral nervous system. In SB patients, this sphincter function can be affected [4]. Without treatment, the level to which this function is compromised determines the prognosis of these patients. Early diagnosis is critical to prevent
SC
further neurologic impairment, but can be quite arduous due to the complex anatomical configuration and high inter-subject variability of the peripheral sacral branches [5, 6]. At
M AN U
present, there are no reliable in vivo and noninvasive routine clinical techniques available to determine how the sacral plexus is organized on an anatomical or structural level in SB patients.
A technique that allows for 3D visualization and structural characterization of nerve tissue is
TE D
diffusion tensor imaging (DTI) [7-9]. DTI is an MRI technique that is sensitive to the random movement of diffusing water molecules, the so-called Brownian motion. The diffusion is
EP
more pronounced along the nerves than across their main orientation, causing diffusion to show a high degree of anisotropy in such fibrous tissue [10]. This anisotropy can be
AC C
quantified by applying a diffusion-weighted MRI acquisition protocol with multiple diffusion gradient orientations and, subsequently, estimating the diffusion tensor [7]. This enables the possibility to reconstruct the 3D architecture of peripheral nerves noninvasively, and is referred to as fiber tractography (FT) [11]. Although DTI has been used in several studies on peripheral nerves [12, 13], it has been rarely used in the lumbosacral region [5]. The potential value of DTI to quantify peripheral nerve injury or dysfunction looks promising, but it is still questionable to what extent this can be translated to a clinical setting given the numerous fiber constituents that may modulate the
3
ACCEPTED MANUSCRIPT observed DTI results [10]. The four diffusion parameters that are commonly used to investigate the tissue microstructural properties are: 1) the fractional anisotropy (FA), which is high when water molecules are moving predominantly along one direction, 2) the mean diffusivity (MD), which is the average of all eigenvalues (the overall amount of diffusion),
RI PT
3) the axial diffusivity (AD), which is equal to the largest eigenvalue, and 4) the radial diffusivity, which is defined as the average of the second and third eigenvalues (see Fig. 1).
SC
[FIG 1]
While DTI [7, 8] has been widely used to investigate white matter tracts in the brain (e.g., [11,
M AN U
14-16]), its application to peripheral nerves is still limited, mainly due to the bigger challenges related to data acquisition [5, 12, 13]. Takagi et al. examined the nerve regeneration of the sciatic nerve in rats following contusive injury. The correlation of fractional anisotropy (FA) values with both histological and functional changes observed in
TE D
that work demonstrates the potential clinical value of DTI in peripheral nerve damage and repair, with lower FA values indicating damage to peripheral nerves [17]. Van der Jagt et al. have recently shown promising progress in reconstructing and analyzing the peripheral sacral
EP
nerves using DTI and FT [5].
AC C
In this work, we investigated the sacral plexus in 10 children with SB with neurogenic bladder dysfunction using DTI and FT and compared the results with healthy controls. Our hypothesis is that for these SB patients, the microstructural properties of the sacral plexus would show abnormal values of DTI parameters in nerve regions where the tissue structure has been affected. Our results demonstrate that with DTI and FT (i) peripheral sacral nerves in SB patients with neurogenic bladder dysfunction can be reconstructed and visualized with great detail, (ii) the microstructural tissue organization of the sacral plexus can be characterized in vivo and noninvasively, and (iii) peripheral nerve tissue abnormalities can be identified at
4
ACCEPTED MANUSCRIPT specific locations in the sacral plexus. We believe that investigating nerve tissue properties with DTI and FT in addition to conventional MRI can be helpful to better understand the mechanism of disturbed innervations of the bladder and lower limb muscles in children with
RI PT
SB. Materials and methods Data acquisition
SC
Local institutional review board approval was obtained for this study and written informed consent was given prior to the MRI examination. Ten patients with SB and neurogenic
M AN U
bladder dysfunction, 6 girls, 4 boys, mean age 11.4 years (range 8-16 years) were included. Neural tube defects (myelomeningocele) were mostly located in the lumbar sacral region (L5S2). To reduce variability within the patient group and keep the acquisition time to a minimum, MRI was performed from the level of L4 to the pelvic floor region. The
TE D
healthy controls were the participants from the study by van der Jagt et al [5]. All subjects were examined with a 3 Tesla MR system (Achieva; Philips Healthcare, Best, The Netherlands) using a 16-channel phased-array surface coil. Diffusion weighted images and
EP
anatomical 3D Turbo Spin Echo (3D-TSE) T2-weighted images were obtained using the
AC C
acquisition protocols as described previously (van der Jagt et al [5]).
Data processing and analysis The DTI data sets were processed using the ExploreDTI diffusion MRI toolbox (www.ExploreDTI.com) [18] as follows: 1) The data were corrected for subject motion and for eddy current induced geometrical distortions [19], 2) diffusion tensors and subsequently diffusion parameters (MD, FA, AD, and RD) were calculated using the iteratively weighted linear regression procedure [20], and 3) a deterministic streamline tractography approach was
5
ACCEPTED MANUSCRIPT used [21] to reconstruct the fiber pathways. In each nerve, a ‘SEED’ region-of-interest (ROI) was placed in the middle of the nerve. Then, a second ROI (‘AND’ region) was selected in the nerve root near the cauda equina and another ‘AND’ ROI more distally along the nerve where it is still traceable. By placing these ‘AND’ regions according to the 3D TSE T2
RI PT
dataset, fiber trajectories can be obtained along the entire nerve. Combining a low FA threshold (i.e., 0.001) in combination with these ‘AND’ ROIs provides a feasible way to reconstruct tracts with a high reproducibility [5]. Finally, the accuracy of the anatomical
SC
location of the fiber tracts was evaluated with the anatomical (3D TSE T2-weigthed) images
M AN U
by a pediatric neuroradiologist with more than 15 years of experience in MRI.
Statistical analyses
The FA, MD, AD, and RD values of the nerves L4 to S3 of SB patients were investigated and compared with the values in healthy controls as obtained in the study by van der Jagt et al [5].
TE D
The variability between the nerves of SB patients was investigated per nerve using the Kruskall-Wallis test. Furthermore, the nerves of the left and right side were compared using a non-parametric Mann-Whitney U test. The same test was also used to investigate the
EP
differences between nerves of SB patients and healthy controls. The analysis was performed
Results
AC C
using SPSS version 21.0 (SPSS Inc. Chicago, IL, USA).
With tractography it was possible to obtain 3D anatomical reconstructions of all 10 SB patients (see supplemental material). Fig. 2 shows the FT results of lumbosacral plexus of a typical SB patient with pseudo-color encoding representing the magnitude of the diffusion parameters. Table 1 summarizes the diffusion measures for the SB patients and the healthy controls. As
6
ACCEPTED MANUSCRIPT the MD values appear to be most indicative to show abnormalities in the nerves (see Table 1), the fiber tracts are color coded for MD in the remaining figures. [FIG 2]
RI PT
Fiber tract evaluation The sacral plexus of SB patients shows asymmetry and disorganization compared to healthy controls. In two SB patients, nerves at the level of L5 could not be reconstructed with FT despite
SC
the fact that they were visible on the anatomical T2-weighted images (see Fig. 3B). In none of the SB patients, S4 and S5 could be traced, which was also the case for the healthy control group.
M AN U
At this level, they were in most cases also not visible on the anatomical T2-weighted images. [FIG 3]
In the SB patients, it was difficult to locate a connection to the cauda equina, especially at the
healthy controls (Fig. 4B).
EP
[FIG 4]
TE D
level of the myelomeningocele (Fig. 4A). By contrast, this connection was clearly visible in the
Diffusion parameters
AC C
Statistical testing revealed no significant difference in FA, MD, AD, and RD between the left and right side at the levels L4-S3 (p>0.05). Therefore, both sides were pooled for each nerve level. Furthermore, no significant differences were found between individual nerves per level (p>0.05). The MD at the level of L4 of the SB patients was comparable to healthy controls as shown in Table 1. However, from the level of L5 downward, a significantly lower MD was found in SB patients compared to healthy controls at the levels S1-S3. Similar differences were found for the AD and RD parameters.
7
ACCEPTED MANUSCRIPT [TABLE 1] Innervation of the bladder A trajectory originating from S2-S4 and continuing to the bladder was found in three patients
RI PT
(see Fig 5A for an example in one patient). Although this could not be confirmed by the anatomical T2-weighted images, it is likely that this is the pudendal nerve, as it shows a similar course as presented on a schematic anatomy configuration of the sacral plexus (Fig. 5B).
SC
[FIG 5]
M AN U
Discussion
To the best of our knowledge, this is the first 3-Tesla DTI study that is able to reconstruct the 3D architecture of the sacral plexus in 10 SB patients, detailing the individual pathway
TE D
trajectories and the microstructural properties of L4-S3. Differences in both anatomy (in particular, asymmetry and disorganization) and mean diffusivity values in the sacral nerves were found between SB patients and healthy controls.
EP
Interpretation
AC C
Fiber tractography provides striking images of the sacral plexus non-invasively [5]. However, this technology has not been used extensively for investigating peripheral nerves as DTI and FT are not straightforward and trivial to apply. In addition, one of the major difficulties of DTI is interpretation [22]. Abnormalities in fiber reconstructions and diffusion parameters can also be caused due to deformations and artifacts in the data or limitations of the imaging technique itself [23, 24]. Clinical relevance
8
ACCEPTED MANUSCRIPT In this study, MD values from the level of L5 to caudal are lower in SB patients than controls, whereas for L4, which is still intact, the MD values were similar between both subject groups. The reduction in MD could involve reduced intrinsic diffusion in the intraaxonal space due to cytoskeletal breakdown resulting in an increased viscosity [10]. It can
RI PT
also be related to the myelomeningocele which in these patients is mostly located at the level of L5-S1. At that level, it was also difficult to find a connection to the cauda equina. With some abuse of terminology, if nerves at that level are not working properly, this might be
SC
compared to an electric cable which is not adequately plugged into the socket. To investigate whether the lower MD values of these nerves are caused by the myelomeningocele, these
higher level (thoracic or high lumbar).
M AN U
results could be compared with SB patients where the myelomeningocele is located at a
DTI was able to show a trajectory going to the bladder, which is likely to be the pudendal nerve. The anatomical image was not able to visualize this trajectory. As the pudendal nerve
TE D
originates from S2-S4 it is expected that the pudendal nerve will also have lower MD values. Future studies investigating the correlation between urodynamic findings and DTI metrics in
EP
the sacral plexus of SB patients may support this hypothesis. The patient group can be extended by investigating other sacral malformations such as anorectal malformation [25].
AC C
Lower lesions are often associated with detrusor sphincter dyssynergy compared to higher lesions. DTI could play an important role in diagnosis and therapy in future neuroanastomotic procedures in these patients. Limitations Although the data have been corrected for subject motion and geometrical distortions, the data could still be misaligned due to nonlinear behavior of these artifacts [5, 9]. In addition, due to the relatively large voxel size (3x3x3 mm) partial volume effects will have an effect on
9
ACCEPTED MANUSCRIPT the diffusion parameters and smaller nerve bundles, especially at S3 as they are smaller in diameter than the other nerves [23, 26]. The control group consists of healthy adults. Although the diffusion properties may change with age, such as in brain white matter fiber bundles [27], we expect that the differences in AD and RD will be relatively low in
RI PT
comparison to the differences in SB patients compared to healthy adults. Future work
SC
To determine to what extent diffusion parameters represent functionality of nerves in SB patients, other patient groups with neurological disorders can be investigated. In newborn SB
M AN U
patients, DTI can be performed before and after closing the spinal cord as a comparison tool for localizing potential nerve damage. By improving the resolution of DTI we believe that in the future, it can be possible to visualize small nerves in these infants. Other pathologies in which DTI can be used to visualize the peripheral nerves include multiple
TE D
sclerosis [28], and paraplegia. Finally, DTI could potentially be applied to locate the nerves for sacral nerve stimulation, neurostimulator implants, and neuroanastomis procedures [29] in the sacral plexus. Combining information of an electromyogram with DTI tractography and
AC C
of peripheral nerves.
EP
diffusion parameters may improve the understanding of how to interpret diffusion parameters
At this moment, DTI cannot replace conventional anatomical imaging modalities, i.e. 2D T1and T2-weighted sequences or high-resolution 3D TSE protocols, but it can be used as a complementary tool to better detect and understand the neurological problems in SB patients. Further adjustments of the registration process by fusing the anatomical and DTI information may improve the analysis of the nerves at each level. Large scale studies in patients with different types of SB are necessary to further optimize the technique and better determine the
10
ACCEPTED MANUSCRIPT exact role of DTI in the diagnosis and follow-up of patients with SB. We expect that abnormal diffusion parameters can be indicative for affected or dysfunctional nerves.
Conclusion
RI PT
This 3T MRI study shows for the first time asymmetry and disorganization of the sacral plexus in 10 SB patients using DTI and FT. These abnormalities indicate that the sacral plexus of SB patients is different from the healthy controls. The observed difference in diffusion
SC
values shows that these methods can be used to identify nerve abnormalities. Combining 3D TSE, DTI, and FT and correlating diffusion parameters with neurological problems of SB
M AN U
patients, is expected to provide a valuable contribution to a better analysis and diagnosis of these patients in the future.
Acknowledgment
TE D
We would like to thank Niels Blanken (MRI radiographer, department of radiology) for his
AC C
EP
help in acquiring the MRI data.
11
ACCEPTED MANUSCRIPT References
8. 9. 10. 11. 12. 13.
14. 15. 16.
17. 18. 19. 20. 21. 22.
23.
RI PT
7.
SC
6.
M AN U
4. 5.
TE D
3.
EP
2.
de Jong TPVM, Chrzan R, Klijn AJ, et al. Treatment of the neurogenic bladder in spina bifida. Pediatr Nephrol. 2008; 23: 889-896. Dik P, Klijn AJ, van Gool JD, et al. Early start to therapy preserves kidney function in spina bifida patients. European urology 2006; 49: 908-913. Özek MM, Cinalli G, Maixner WJ. Spina bifida - management and outcome. New York: Springer 2008. de Groat WC. A neurologic basis for the overactive bladder. Urology 1997; 50: 36-52. van der Jagt PKN, Dik P, Froeling M, et al. Architectural configuration and microstructural properties of the sacral plexus: A diffusion tensor MRI and fiber tractography study. NeuroImage 2012; 62: 1792-1799. Matejcik V. Anatomical variations of lumbosacral plexus. Surgical and Radiologic Anatomy 2010; 32: 409-414. Basser PJ, Mattiello J, LeBihan D. MR diffusion tensor spectroscopy and imaging. Biophysical journal 1994; 66: 259-267. Jones DK, Leemans A. Diffusion Tensor Imaging. In: Modo M, Bulte JWM, editors. Magnetic Resonance Neuroimaging: Humana Press 2011; 127-144. Tournier JD, Mori S, Leemans A. Diffusion tensor imaging and beyond. Magnetic resonance in medicine 2011; 65: 1532-1556. Beaulieu C. The basis of anisotropic water diffusion in the nervous system - a technical review. NMR in biomedicine 2002; 15: 435-455. Mori S, van Zijl PCM. Fiber tracking: principles and strategies - a technical review. NMR in biomedicine 2002; 15: 468-80. Hiltunen J, Suortti T, Arvela S, et al. Diffusion tensor imaging and tractography of distal peripheral nerves at 3 T. Clinical Neurophysiology 2005; 116: 2315-2323. Vargas MI, Viallon M, Nguyen D, et al. Diffusion tensor imaging (DTI) and tractography of the brachial plexus: feasibility and initial experience in neoplastic conditions. Neuroradiology 2010; 52: 237-245. Wakana S, Jiang H, Nagae-Poetscher LM, et al. Fiber tract-based atlas of human white matter anatomy. Radiology 2004; 230: 77-87. Reijmer YD, Leemans A, Caeyenberghs K, et al. Disruption of cerebral networks and cognitive impairment in Alzheimer disease. Neurology 2013; 80: 1370-1377. Reijmer YD, Brundel M, de Bresser J, et al. Microstructural white matter abnormalities and cognitive functioning in type 2 diabetes: a diffusion tensor imaging study. Diabetes care 2013; 36: 137-144. Takagi T, Nakamura M, Yamada M, et al. Visualization of peripheral nerve degeneration and regeneration: monitoring with diffusion tensor tractography. NeuroImage 2009; 44: 884-892. Leemans A, Jeurissen B, Sijbers J, et al. ExploreDTI: a graphical toolbox for processing, analyzing, and visualizing diffusion MR data. Proc Intl Soc Mag Reson Med 2009; 17: 3536. Leemans A, Jones DK. The B-matrix must be rotated when correcting for subject motion in DTI data. Magnetic resonance in medicine 2009; 61: 1336-1349. Veraart J, Sijbers J, Sunaert S, et al. Weighted linear least squares estimation of diffusion MRI parameters: Strengths, limitations, and pitfalls. NeuroImage 2013; 81: 335-346. Basser PJ, Pajevic S, Pierpaoli C, et al. In vivo fiber tractography using DT-MRI data. Magnetic Resonance in Medicine 2000; 44: 625-632. Deprez S, Billiet T, Sunaert S, et al. Diffusion tensor MRI of chemotherapy-induced cognitive impairment in non-CNS cancer patients: a review. Brain Imaging and Behavior 2013; 7: 409435. Vos SB, Jones DK, Viergever MA, et al. Partial volume effect as a hidden covariate in DTI analyses. NeuroImage 2011; 55: 1566-1576.
AC C
1.
12
ACCEPTED MANUSCRIPT
RI PT
29.
SC
28.
M AN U
27.
TE D
26.
EP
25.
Tournier JD, Calamante F, King MD, et al. Limitations and requirements of diffusion tensor fiber tracking: An assessment using simulations. Magnetic Resonance in Medicine. 2002; 47: 701-708. Boemers TM, Van Gool JD, De Jong TPVM, et al. Urodynamic evaluation of children with the caudal regression syndrome (caudal dysplasia sequence). Journal of Urology 1994; 151: 10381040. Hogan Q. Size of human lower thoracic and lumbosacral nerve roots. Anesthesiology. 1996; 85: 37-42. Bennett IJ, Madden DJ, Vaidya CJ, et al. Age-related differences in multiple measures of white matter integrity: A diffusion tensor imaging study of healthy aging. Human brain mapping 2010; 31: 378-390. Roosendaal SD, Geurts JJG, Vrenken H, et al. Regional DTI differences in multiple sclerosis patients. NeuroImage 2009; 44: 1397-1403. Rasmussen MM, Clemmensen D, Rawashdeh YF, et al. Surgical reinnervation with nerve anastomosis technique for neurogenic bladder and bowel dysfunction. Ugeskrift for laeger 2011; 173: 2412-2415.
AC C
24.
13
ACCEPTED MANUSCRIPT Legend figures Figure 1: In fluid the random movement of water molecules is isotropic. Diffusion along the nerves (represented as axial diffusivity or λ1) is higher than perpendicular to the nerves (represented as radial diffusivity or mean of λ2 and λ3). λ1, λ2, and λ3 are the eigenvalues (the length of the diffusion) in a particular orientation. This diffusion orientation preference is called anisotropy.
RI PT
Figure 2: 3D anatomy of the lumbar and sacral nerves at the level of L3-S3 of a typical spina bifida patient. Color maps of mean diffusivity, fractional anisotropy, axial diffusivity, and radial diffusivity are shown along the fiber pathways. A red color represents a high diffusivity and blue color a low diffusivity.
SC
Figure 3: Lower lumbar and sacral nerves, A) healthy control, B) spina bifida patient with myelomeningocele from the level of L5 to caudal. L5 on the left side could not be reconstructed (indicated with “1”).
M AN U
Figure 4: A) Posterior view of a spina bifida patients with myelomeningocele at the lumbosacral level of L5-S1, the sacral nerves do not connect with the cauda equina (indicated with the arrows), B) healthy control in which the nerves show a connection to the cauda equina.
TE D
Figure 5: A) Anterior view of a spina bifida patient with myelomeningocele at the lumbar level of L1-L4, the pudendal nerve is indicated with “PN”, and the sciatic nerve with “SN”, B) schematic overview of the lumbosacral nerves. The sacral branches and pudendal nerve in Fig. 5A correspond with the anatomy shown in Fig. 5B.
AC C
EP
Table 1: Mean and standard deviation of the diffusion parameters (fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD)) of the lumbar and sacral nerves of spina bifida (SB) patients and healthy controls at the level of L4-S3.
ACCEPTED MANUSCRIPT Table 1 Nerve
-3
FA
Diffusivity (mm²/s) x 10 MD
AD
RD
SB
Healthy
SB
Healthy
SB
Healthy
SB
L4
0.28±0.05
0.28±0.04
1.34±0.22
1.32±0.16
1.72±0.22
1.70±0.18
1.15±0.21
1.14±0.15
L5
0.31±0.03
0.29±0.04
1.42±0.21
1.31±0.24
1.86±0.24
1.72±0.31
1.19±0.20
1.11±0.21
S1
0.26±0.03
0.27±0.05
1.83±0.24**
1.40±0.22**
2.31±0.27**
1.78±0.23**
1.59±0.22**
1.22±0.23**
S2
0.23±0.03
0.25±0.05
1.73±0.29**
1.36±0.21**
2.13±0.36*
1.70±0.21*
1.530.26*
1.19±0.22*
S3
0.22±0.03
0.26±0.05
1.62±0.32*
1.33±0.23*
1.98±0.39*
1.66±0.35*
1.43±0.29*
1.17±0.22*
SC
RI PT
Healthy
AC C
EP
TE D
M AN U
*: p<0.01, **: p< 0.001
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
axial diffusivity
DTI
diffusion tensor imaging
FA
fractional anisotropy
FT
fiber tractography
MD
mean diffusivity
MRI
magnetic resonance imaging
RD
radial diffusivity
SB
spina bifida
TSE
turbo spin echo
AC C
EP
TE D
M AN U
SC
AD
RI PT
Key of definitions for abbreviations
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT