1. NTNU Det medisinske fakultet White Matter Abnormalities seen on Diffusion Tensor Imaging Relate to Neuroimpairments in VLBW Children at 15 Years of Age Jon Skranes 1, Torgil R Vangberg 2, Arild Kristoffersen 2, Siri Kulseng 1, Marit Indredavik 3, Kari Anne I. Evensen 1, Marit Martinussen 2, David Tuch 5, Anders Dale 5, Olav Haraldseth 2, Torstein Vik 4 and Ann Mari Brubakk 1 1 Dep of Laboratory Medicine, Children’s and Woman’s Health; 2 Dep of Circulation and Medical Imaging; 3 Dep of Neuroscience; 4 Dep of Public Health and General Practice, Faculty of Medicine, Norwegian University of Science and Technology, Trondheim, Norway; 5 Martinos Center, MGH, Boston, USA

2. NTNU Det medisinske fakultet BACKGROUND VLBW (birth weight < 1500 grams) children are at increased risk of neuroimpairments caused by perinatal injury like periventricular leukomalacia (PVL). Brain DTI may detect subtle changes in white matter microstructure that is not seen on conventional MRI. OBJECTIVE To compare cerebral DTI findings with the results from extensive psychiatric, behavioural, cognitive and motor assessments in VLBW adolescents at age 15. We hypothesized that reduced fractional anisotropy in specific areas of white matter was related with abnormal test results.

3. NTNU Det medisinske fakultet MATERIAL 34 VLBW children 49 non-SGA term controls Examined at 15 years of age in a population based study METHODS Movement-ABC Developmental test of Visual-Motor Integration (VMI) with supplementary tests: Visual perception and Motor coordination Psychiatric interview (Kaufman-SADS), ADHD rating scale Subtests from WISC-III: Arithmetic, Vocabulary, Picture arrangement, Block design

4. NTNU Det medisinske fakultet Child characteristics VLBWControl Child characteristics Mean(SD)Mean(SD) n=34n=49 Birth weight (g) 1232*(217)3631(472) Gestational age (weeks) Male Female 29.3* 16 18 (2.7) 39.3 18 31 (1.4) MRI assessment age (years) 15.1(0.6) 15.3(0.5) * p <0.05 versus controls. SD: standard deviation

5. NTNU Det medisinske fakultet Diffusion-Tensor Imaging MRI technique for in-vivo characterization of 3D white matter microstructure. –Measures magnitude and direction of water diffusion in biological tissue in 3D. DTI scans were performed for calculation of fractional anisotropy (FA) maps for each individual and for group comparisons.

6. NTNU Det medisinske fakultet METHODS cont. Fractional Anisotropy (FA): The extent to which diffusion is directionally restricted, i.e. anisotropic. Scalar measure: 0 to 1 For isotropic diffusion ( λ 1 = λ 2 = λ 3), FA is zero, and in the case where there is a strongly preferred direction of diffusion ( λ 1 >> λ 2 = λ 3), FA approaches one. Mean FA values for anatomical areas of significant differences between VLBW and controls were identified and then compared with clinical data for the VLBW adolescents (low performers vs. normal performers).

7. NTNU Det medisinske fakultet (A) Water molecules in the brain are constantly moving (i.e., in Brownian motion). When motion is unconstrained, as in the ventricles, seen in the MR image on the left, diffusion is isotropic, i.e. motion occurs equally and randomly in all directions. (B) When motion is constrained, as in white matter tracts (right), diffusion is anisotropic, i.e. motion is oriented more in one direction than another (e.g., along the y axis rather than along the x axis). Margaret Rosenbloom et al. July 2004

8. NTNU Det medisinske fakultet Image Acquisition Scanner: –1.5 T Siemens Symphony Structural scan: –T1-weighted MPRAGE-sequence –in-plane resolution 1 x 1 mm –Slice thickness 1.33 mm DTI scans: –Twice-refocused spin echo sequence –6 directions with b = 1000 s/mm 2 –in-plane resolution 1.78 x 1.78 mm –20 contiguous slices, 5 mm thick –Repeated 6 times T1-weighted MPRAGE scan, showing approximate location of DTI slice packet (black lines).

9. NTNU Det medisinske fakultet Diffusion Tensor Calculation Motion correction of DTI series using 12-parameter affine transform The six DTI acquisitions are averaged Diffusion tensor and FA maps calculated from averaged DTI volume. DTI scans 1 - 6 Motion correction and averaging Averaged volume Diffusion tensor Calculation of diffusion tensor FA map S 0 volume

10. NTNU Det medisinske fakultet Image Post Processing SPM2 used for post processing and statistical analysis Custom template created from structural images of all subjects FA maps normalized to custom template Prior to statistical analysis FA maps smoothed with 4 mm FWHM Gaussian FA S 0 Anatomical 3.Transformation 1. Coregister Normalized anatomical 2. Normalization Normalized FA 4. Normalization

11. NTNU Det medisinske fakultet RESULTS The VLBW teenagers had significantly lower mean FA values in several anatomical white matter areas compared with controls: – internal capsule (mainly PLIC) – corpus callosum (anterior and posterior) – centrum semiovale (central parietofrontal w.m.) – peripheral parietofrontal white matter – periventricular occipital white matter

12. NTNU Det medisinske fakultet Location, size and mean FA values in clusters where the VLBW group had significantly lower FA values than controls (p<.05). Location of significant clusters cluster size (voxels) Mean FA in cluster VLBWControl Capsula interna left 16590.3880.440 Capsula interna right 15220.3950.452 Corpus callosum post 4290.4390.515 Corpus callosum ant 5050.4300.505 Centrum semiovale left 7100.3960.446 Centrum semiovale right 3950.4060.457 Peripheral wm left 2350.3820.428 Peripheral wm right 1440.3710.415 Occipital wm left 2120.4140.469

13. NTNU Det medisinske fakultet Results VLBW < Control Corpus callosum (green arrows) RL Regions with reduced FA in VLBW group. Underlay is average of anatomical images from VLBW and control group.

14. NTNU Det medisinske fakultet Results VLBW < Control Internal capsule (mostly PLIC) RLRL

15. NTNU Det medisinske fakultet Results VLBW < Control Occipital white matter (green arrows) RLRL

16. NTNU Det medisinske fakultet Results VLBW < Control (right side) Peripheral white matter Centrum semiovale

17. NTNU Det medisinske fakultet Relationship between motor test results and FA values in different anatomical regions in VLBW adolescents Total ABC score Manual dext. score Ball skills score 5th p. 5th p. 5th p. (8) (22) (3) (29) (4) (27) Anatomical region Capsula interna 0.358 0.400 Periph wm right 0.343 0.384 0.335 0.379 Posterior corpus call 0.370 0.454 P <.05

18. NTNU Det medisinske fakultet Relationship between visuo-perceptual and behavioural test results and FA values in different anatomical regions VMI visual perc. ADHD / ADD -1SD yes no (7) (23) (8) (26) Anatomical region Capsula interna 0.355 0.407 Peripheral wm 0.367 0.396 0.346 0.393 Post. corpus callosum 0.406 0.464 0.359 0.464 Centrum semiovale 0.386 0.421 Occipital wm 0.391 0.431 0.359 0.431

19. NTNU Det medisinske fakultet Correlations between WISC III subtest scores and FA values in different anatomical regions in 33 VLBW adolescents WISC III subtest scores Arithmetic Pict. arrangement Block design Anatomical region Capsula interna,416** Corpus callosum post part,381**,344 #,399** Centrum semiovale,411** Peripheral wm left,571*,418**,472* Occipital wm left,388** * p < 0.01, ** p < 0.05, # p = 0.05 Abbreviations: wm: white matter; WISC: Wechsler Intelligence Scale; Pict.: picture Correlations by Spearman’s rho

20. NTNU Det medisinske fakultet Discussion The VLBW group has reduced anisotropy in several wm areas which are known to be affected by premature birth (PVL, asphyxia) periventricular occipital white matter centrum semiovale corpus callosum posterior limb of internal capsule The reduced anisotropy may be due to: reduced myelination fiber disorganization fewer axons

21. NTNU Det medisinske fakultet Conclusions / Speculations Low FA values in specific brain areas seem to relate to motor, cognitive, perceptual and behavioural impairments in VLBW adolescents. The neuroimpairments may be due to disturbed white matter microstructure and connectivity of commissural and association fibres between different cortical areas. The disrupted connectivity may influence neuronal networks and thereby neurodevelopment and higher cognitive functions in VLBW adolescents.

22. NTNU Det medisinske fakultet