TwitterRSSFacebookDirectors Blog
  Disorders A - Z:   A    B   C    D    E    F    G    H    I    J    K    L    M    N    O    P    Q    R    S    T    U    V    W    X    Y    Z

You Are Here: Home  »  Training  »  Summer Program in the Neurological Sciences and Other Neuroscience Research  »  2013 Award Winners  » 

Skip secondary menu

Quantifying spinal cord cross-sectional area in HTLV-1 associated myelopathy/tropical spastic paraparesis and multiple sclerosis

Winston Liu Photo

University of Maryland (Maryland)

Winston Liu, Raya Massoud, Giovanna Brunetto, Daniel S. Reich, Govind Nair, and Steve Jacobson

Introduction:  Human  T-cell  lymphotropic  virus-1  associated  myelopathy/tropical  spastic  paraparesis (HAM/TSP)  and  multiple  sclerosis  (MS)  can  both  lead  to  a  progressive  inflammatory  myelopathy.  Current advances in MRI have enhanced its use as a tool to monitor such disease processes. The typical radiologic finding in HAM/TSP is thoracic cord atrophy without abnormal cord signal. Although spinal cord atrophy can sometimes be qualitatively detected on routine clinical MRI, a robust and sensitive method to quantify changes in spinal cord size might better characterize  the disease severity and progression.  Such a method could be used to not only develop an imaging marker for disease, but also to serve as a surrogate end point in clinical trials.

Materials  and  Methods:  MRI  was  performed  on  18  HAM/TSP,  9  MS  (8  relapsing  remitting,  1  primary progressive) and 10 healthy volunteers on a Siemens 3T Skyra system equipped with head-neck and spine array coils. Cross-sectional area of the spinal cord was measured using a novel and fast algorithm developed in-house that traces axial contours of the spinal cord, perpendicular  to the cord-edge along each point from C1 to T10 vertebral body segments in mid-sagittal  T1-weighted  images. Total post processing time was <5 minutes with minimal user input. HAM/TSP and MS patient disability was measured through the Scripps Neurologic Rating Scale (SNRS), Extended Disability  Status Scale (EDSS), Hauser Ambulation  Index (AI) and peripheral  blood pro-viral  loads  (PVL)  of  patients  were  measured  using  digital  droplet  PCR.  A  pairwise  Tukey  Honestly Significant Difference test was used to compare cross-sectional areas across disease types.  A Pearson correlation test  adjusted  for  multiple  comparisons  was  used  to  assess  relationships  between  disability  scores  and  cross- sectional area and a p<0.05 was considered significant.

Results and Discussion: Average cross-sectional area in both the T-spine and C-spine were significantly lower in HAM/TSP patients (c-spine: 50.75 ± 10.02 mm2; T-spine: 24.76 ± 5.01 mm2) as compared to healthy controls (C- spine: 72.02 ± 5.792 mm2, p<0.0001; T-spine: 38.8 ± 6.048 mm2, p<0.0001). However, cross-sectional area was only significantly different in the T-spine when comparing HAM/TSP with MS patients (C-spine: 61.84 ± 9.54

mm2; T-spine: 34.84 ± 5.72 mm2, p<0.001). In HAM/TSP, the cross-sectional areas from multiple cord segments

correlated  with  disease  duration  (C1-C3,  C6,  T1-T2),  EDSS  (T2),  and  AI  (T2,  T6).  Trends  were  observed between HAM/TSP average cross-sectional area and SNRS, but no correlation was observed between HAM/TSP cross-sectional area and PVL. MS cross-sectional area correlated with SNRS (C1-C5, T1-T7) and EDSS (C1-C3, T1-T6).  Imaging  artifacts  due  to  swallowing  (typically  affecting  C6-T1),  and  abdominal  motion  (typically affecting T9-T12) increased the unreliability of the area calculation.

Conclusions: These results suggest that the pattern of spinal cord tissue damage is specific to the underlying inflammatory disease, a finding that has direct implications for the use of average cross-sectional spinal cord area as a surrogate end point for clinical trials. This comparative  study shows the utility of investigating  the entire cord,  which  will  improve  our  understanding   of  the  patterns  of  cord  atrophy  in  inflammatory   diseases. Furthermore, longitudinal studies are underway to help improve our understanding of the disease process.

Acknowledgements: This study was supported by the Intramural Research Program of NINDS. Special thanks to the NINDS Summer Internship Program, and Kory Johnson for his advice relating to the statistics.

Last Modified November 27, 2013