0755
Macroscopic Hyperdynamic CSF Flow and CSF Volume Analysis in Children and Adolescents with Congenital Heart Disease1University of Pittsburgh, Pittsburgh, PA, United States
Synopsis
In this study we investigate whether there are differences in CSF volumes between CHDs and healthy controls, as well as examining any correlation between CSF volume and CSF flow, and whether this relationship is altered in CHDs. We showed that certain CSF volumes are elevated in CHD. We also observed that increased flow metrics are correlated with both increased volume in some CSF compartments and decreased volume in other CSF compartments. Lastly, there maybe interplay between CSF flow and volume within CHDs.
Introduction
Our previous investigation uncovered cortical spinal fluid (CSF) flow dynamics abnormalities in children and adolescents born with congenital heart disease (CHD). We previously showed CHD had significantly higher average flow over the lumen of the aqueduct than controls.[1] In this study we investigate whether there are differences in CSF volumes between CHDs and healthy controls. Additionally, we examine if there are any correlation between CSF volume and CSF flow, and if this correlation relationship is altered in CHDs.Methods
A total of 57 children/adolescent participants (CHD=23, 14.4±5.95 y.o.; Healthy Controls=34, 14.4±4.03 y.o.) from part of a larger prospective pediatric connectome study with both T1 volumetric and CSF Flow imaging were included in this study. The imaging acquired on a Siemens 3T Skyra system with 32-channel head coil. The T1 was acquired with the following parameters: TE/TR=3.2/2400 ms; matrix=256x256; Resolution 1.0x1.0x1.0 mm3. The volumetric T1 weighted images were segmented using combination of FreeSurfer[2] and FSL FAST segmentation[3], examining intraventricular CSF (left and right lateral ventricles, Third Ventricle, and Fourth Ventricle) and extra-axial CSF (divided between supratentorial and infratentorial) – and normalized to whole brain volume. The CSF flow was acquired using phase contrast gradient echo at the level of the cerebral aqueduct with the following parameters: velocity encoding=12 cm/s; TR/TE=9.66/30.40 ms; flip angle=15°; matrix=256x256; in plane resolution=0.6mmx0.6mm; slice thickness=5mm;). A circular ROI encompassing the lumen and wall of the aqueduct was used for each study with the surrounding brain parenchyma used as reference. The following CSF flow metrics were calculated using ARGUS software: Average Velocity (AV), Average Flow over Range of Scan (TAF), Average Flow per Minute (AFpM), Total Forward Volume (TFV), Total Reverse Volume (TRV), Net ForwardVolume (NFV).Results
As previously reported, the children/adolescent CHD group compared to healthy controls demonstrated no difference in age, but had higher prevalence of males (p=0.0293).[1] There were no global CSF and whole brain volume differences between CHD and control groups. Examination of normalized CSF volume differences showed CHDs had significantly larger fourth ventricles than controls (p=0.0108; CHD=0.0010±0.0001, controls=0.0013±0.0001). Initial simple regression analysis correlating normalized csf volume to flow measurements revealed significant associations as follows: increased TFV and decreased supratentorial extraaxial CSF (p=0.0467), increased left lateral ventricle (p=0.0396), increased third ventricle (p=0.0098); and increased NFV and decreased supratentorial extraaxial CSF (p=0.0468), increased NFV and decreased fourth ventricle (p=0.0406). Following these observations, ANCOVA analysis was conducted to further examine these correlation relationships in the context of CHD and healthy control groups. While the correlation relationships between increased TFV and decreased supratentorial extraaxial CSF (p=0.0367), increased left lateral ventricle (p=0.04593), increased third ventricle (p=0.0148) are preserved, there were no differences in terms of CHD and healthy controls. However, as shown in Figure 1, this further analysis revealed that as fourth ventricle volume increases the NFV decreases in both CHDs and Controls (p=0.0482), and that CHDs showed higher NFV than controls (p=0.0414).Discussion
While overall global CSF volume was found to be not significantly different between CHD and controls, further examination of CSF compartments showed that the CHD group had larger fourth ventricle than control group. We also observed flow dynamics, specifically TFV and NFV, relate to certain CSF structural compartments. The ANCOVA showed the association between increased fourth ventricle volume and decreased NFV and the significant differences between CHD and controls. This is interesting as our previous study showed macroscopic CSF flow is increased in children and adolescents with CHD compared to controls[1], and we see the same differential dynamic in the CSF flow to volume relationship with CHD greater than controls.Conclusion
These findings show that there is interplay between CSF flow and CSF volume. This relationship between flow dynamics and structural volume is retained and maybe considerably altered in CHD, which bears further investigation.Acknowledgements
No acknowledgement found.References
[1] Lee, V.K., W. T. Reynolds, J. Wallace, R. Hartog, N. Beluk, O. Khalifa, D. Badaly, M. Zahid, R. Ceschin, C. W. Lo, A. Panigrahy. “Macroscopic Hyperdynamic CSF and Ciliary Motion Dysfunction Predict ExecutiveDysfunction in Children and Adolescents with Congenital Heart Disease.” Proceedings of the Twenty-Sevent Annual Joint Meeting of International Society for Magnetic Resonance in Medicine – European Society for Magnetic Resonance in Medicine & Biology Conference, June 16-21, 2018. International Society for Magnetic Resonance in Medicine.
[2] Desikan, R.S., Segonne, F., Fischl, B., Quinn, B.T., Dickerson, B.C., Blacker, D., Buckner, R.L., Dale, A.M., Maguire, R.P., Hyman, B.T., Albert, M.S., Killiany, R.J., 2006. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage 31, 968-980.
[3] Zhang, Y. and Brady, M. and Smith, S. Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm. IEEE Trans Med Imag, 20(1):45-57, 2001.
Figures
