En Español | Plain Text Version | Skip Menus

You are here:  Home > Training > Summer Program in the Neurological Sciences and Other Neuroscience Research

The Correlation between T2/T1 Lesion Load on 3T MRI and Cognitive Deficits in Multiple Sclerosis

Skip secondary menu

Project summary

Project purpose and hypothesis
Multiple Sclerosis (MS) is a demyelinating, chronic autoimmune disease that affects the central nervous system. It is characterized by the destruction and repair of myelin and often associated with axonal injury. The pathogenic mechanisms of MS are still partially unknown, but genetic and environmental factors both contribute to the development of the disease. MS symptoms include optic neuritis, motor and sensory deficits, diplopia, cognitive impairment, and depression. Based upon the disease course and evolution over time, three main types of MS are known, namely relapsing remitting (RR), secondary progressive (SP), and primary progressive (PP).

Magnetic Resonance Imaging (MRI) plays an important role in diagnosing and monitoring MS patients over time because of its high sensitivity to white matter (WM) lesions. In fact MS lesions can be easily identified both in the brain and the spinal cord. Furthermore, by using contrast agent, MRI has the ability to detect the appearance of new lesions even during periods of apparent clinical stability.

Ever since its introduction in clinical ground in the late 80s MRI techniques have improved over time. The strength of the magnetic filed has increased from 0.5 to 7 Tesla (T), although the vast majority of the clinical centers still cope with 1.5T MRI. Compared with lower field strengths, 3T MRI offers higher signal-to-noise ratios while maintaining a relatively high spatial resolution. The latter was proven to enable disclosing as much as 21% more inflammatory MS lesions of the WM.

As previously addressed, an important manifestation of MS with a potential high social impact is given by the occurrence of the cognitive impairment. Cognitive disability can be measured by many different tests, one of them being the Paced Auditory Serial Addition Task (PASAT). In this specific test attention and concentration, two of the most affected abilities in MS patients, are measured.

The aim of present study was to measure the T2 and T1 lesion volume (LV) on images obtained on 3T-MRI in a group of MS patients followed at our clinic. Subsequently, correlation between T2/T1 LV and cognitive deficits as measured by the PASAT score were sought. The working hypothesis was that a stronger correlation between macroscopic MS lesions and the PASAT score may exist as sustained by the more accurate (greater) T2/T1 LV estimation on 3T-MRI.

Materials and Methods:

Patient Population and Study Design
The study was performed at NIB-NINDS upon IRB approval and included 24 MS patients (20 RR and 4 SP). Each patient underwent a 3T -MRI which was followed or preceded by one week by the PASAT. The PASAT consists of listening numbers on a tape, remembering the last number heard, and summing it with the previous number heard. The score that is obtained is between 0 and 60, with 60 being the better score.

Image Acquisition
Each patient underwent a MS protocol brain MRI scan at the NIB clinic, including the following sequences: Fast Spin-Echo (FSE) T2-weighted (TR/TE 5300ms/122ms, matrix 512 x 512, FOV 220 x 220 mm2, ETL 22), FLAIR (TR/TE/TI 8800ms/142ms/2200ms, matrix 512 x 512, FOV 220 x 220 mm2), and T1-weighted (TR/TE 700ms/10ms, matrix 240 x 240, FOV 220 x 220 mm2). For all the modalities 46 slices of 3 mm thickness with the same positioning were acquired.

Image Analysis
For each patient, lesions were first manually highlighted on T2 and T1 image hard copies, using also FLAIR as reference, and then checked by a trained neurologist. These lesions were then transferred to a workstation. At this point a volume of interest (VOI) for each lesion was outlined using a semi-automated technique included in MIPAV, a medical imaging software. The total LV (i.e. the sum of all VOIs for each patient) was then computed using the statistic generator provided in the software.

Statistical Analyses
First, separate univariate regression analyses were calculated for both the T2 and T1 LV comparing them to the PASAT score. As these preliminary analysis showed significant results, as second model of multivariate regression analyses taking into account possible confounding factors such as age, disease duration, and level of education were included in the correlation analyses model. The Statview software was used to generate statistics.

Results:

The univariate regression model showed that T2 (p=0.0136) and T1 (p=0.0037) LV were both significantly correlated with the PASAT score, explaining 24.6% and 36.5% of the PASAT variance, respectively. Those results are depicted also in Figure 1 and Figure 2 here below.

                        Figure 1                                                            Figure 2

When multivariate models were obtained, both correlations were weakened to an insignificant level. The models included T2/T1 LV, age, level of education, and years of MS as explanatory variables. The statistics here showed that 50.1% and 63.7% of the variance is explained by all the demographic/clinical covariates and the T2/T1 LV, respectively. However, in both instances, the only significant variable was disease duration. Those results are described in details in the Table 1 and Table 2 below.

Table 1. PASAT vs. Age, Education, Onset Period, and T2 Lesion Volume

  Coefficient Std. Error Std. Coeff. t-Value P-Value
Intercept 65.080 16.083 65.080 4.047 0.0008
Age -0.107 0.217 -0.099 -0.494 0.6273
Onset Period -0.701 0.296 -0.564 -2.367 0.0293
Education -0.241 0.784 -0.053 -0.307 0.7621
T2 LV -1.438E-4 1.916E-4 -0.153 -0.751 0.4625

Table 2. PASAT vs. Age, Education, Onset Period, and T1 Lesion Volume

  Coefficient Std. Error Std. Coeff. t-Value P-Value
Intercept 65.217 14.516 65.217 4.493 0.0004
Age -0.205 0.209 -0.179 -0.980 0.3425
Onset Period -0.608 0.255 -0.504 -2.378 0.0311
Education 0.031 0.742 0.007 0.042 0.9672
T1 LV -0.001 3.323E-4 0.281 -1.517 0.1500

Conclusions:

The results from the study show that the higher amount of WM lesion load disclosed by using a 3T-MRI allows the identification of a positive correlation between lesion extent and the cognitive deficits in MS patients.

However, such a correlation is weakened to an insignificant level when the length of disease duration is considered as a possible determinant factor.

The latter indicates that the disease duration likely carries potential biological factors that are representative of a higher correlation. These factors are intrinsic to the disease and may include brain atrophy and axonal injury not otherwise disclosed by the LV measurements.

Last updated December 27, 2007