Neurology India Medknow Publications on behalf of the Neurological Society of India ISSN: 0028-3886 EISSN: 1998-4022 Vol. 59, Num. 2, 2011, pp. 168-173

Neurology India, Vol. 59, No. 2, March-April, 2011, pp. 168-173

Original Article

Comparison of diffusion tensor image study in association fiber tracts among normal, amnestic mild cognitive impairment, and Alzheimer's patients

Yu-Zhen Zhang¹, Cheng Chang¹, Xiao-er Wei¹, Jian-Liang Fu², Wen-Bin Li¹

¹ Institute of Diagnostic and Interventional Radiology, Shanghai Jiao Tong University, Shanghai, China
² Department of Neurology, The Sixth Affiliated People's Hospital, Shanghai Jiao Tong University, Shanghai, China

Correspondence Address: Wen-Bin Li, No. 600 Yi Shan Road, Shanghai 200233, China, liwenbin@sh163.net

Date of Submission: 02-Oct-2010
Date of Decision: 07-Dec-2010
Date of Acceptance: 27-Dec-2010

Code Number: ni11052

PMID: 21483111

DOI: 10.4103/0028-3886.79129

Abstract

Keywords: Alzheimer's disease, amnestic mild cognitive impairment, dementia, diffusion tensor imaging

Introduction

The condition of amnestic mild cognitive impairment (aMCI) includes defective memory or cognition without the presence of dementia. An estimated 12-15% of patients with aMCI may progress to have Alzheimer's disease (AD) each year, ^[1] and one of the most important issues in clinical management of patients with aMCI is to predict progression to AD. Currently, the diagnosis of aMCI is still mostly based on cognitive and behavioral tests administered in a neurology clinic setting. The ability to quantify cortical changes could be particularly useful in monitoring and even predicting the progression of AD in aMCI subjects. ^[2]

Neuroimaging is valuable in predicting future development of AD in patients with aMCI. Many magnetic resonance (MR) techniques are used in the study of aMCI, such as routine MR imaging (MRI), diffusion weighted imaging (DWI), proton magnetic resonance spectroscopy (1H-MRS), diffusion tensor imaging (DTI), perfusion weighted imaging (PWI), magnetization transfer imaging (MTI), and functional MRI (fMRI). 1H-MRS and DTI are the most widely used of all the methods. ^[3],[4],[5] DTI offers increased sensitivity to alterations in white matter in vivo, which were not apparent using other more conventional MRI protocols. DTI analysis could provide quantitative measurements that could be used to determine the prognosis of aMCI and AD patients. Although mild decreases in anisotropy are a normal result of aging, DTI has shown additional abnormalities in patients with several types of dementia and neurodegenerative disease. ^[6],[7] DTI is the only method that can non-invasively track white matter (WM) fibers and reflect their integrity. ^[4] Diffusivity changes have been observed in the frontal, temporal and parietal WM regions in MCI patients. ^[8],[9],[10] A number of post-mortem investigations also have confirmed WM pathology in AD. ^[11],[12] Brain regions are not isolated entities but instead form highly interconnected neural circuits encompassing remote brain regions. The integrity of network level is believed to be important in higher cognitive function. ^[13],[14] The aim of this study was to investigate whether association fiber tracts between remote cortices were altered in aMCI patients compared with NC and AD.

Material and Methods

Subjects

Sixty subjects (20 NC, 20 aMCI, 20 AD) were enrolled in this study, and all gave written informed consent. Exclusion criteria included structural abnormalities that could produce dementia, such as cortical infarction, tumors, subdural hematoma, brain trauma, epilepsy, alcoholism, psychiatric illness and other systemic diseases that affect brain function. Inclusion criteria for aMCI were: (a) subjective symptoms of memory loss, (b) normal activities of daily living, (c) normal general cognitive function, (d) absence of dementia, and (e) Mini-Mental Status Examination (MMSE) score of at least 24-30. ^[1],[15] Inclusion criteria for AD were: (a) Clinical Dementia Rating Scale (CDR≥1) ^[16] and (b) diagnosis of AD made on the basis of nationally accepted criteria. ^[17] MMSE scores were measured in AD subjects but were not considered as a factor for exclusion. ^[15] NC subjects underwent a cognitive examination; none reported subjective symptoms of cognitive impairment, and all had MMSE scores of at least 27-30.

MRI methods

Routine MRI and DTI were performed on a 3.0 T Intera Achieva MRI (Philips Medical Systems) with an eight-channel phased array head coil. Parameters of routine MRI were: Flip angle 90°, thickness 6-7 mm, gap 0.6 mm, NEX 1. Other parameters were: Axial T1W TSE (TR/TE 2000/20 ms, FOV 250Χ221Χ136 mm, matrix 400Χ250). Axial T2W TSE (TR/TE 3000/80 ms, FOV 250Χ212Χ136 mm, matrix 436Χ295). Axial FLAIR IR (TR/TE 11000/120 ms, FOV 230Χ215Χ/136 mm, matrix 240Χ160). Sagittal T1W SE(TR/TE 2000/20 ms, FOV 230Χ230Χ/136 mm, matrix 296Χ220). Coronal T2W TSE (TR/TE 3000/80 ms, FOV 250Χ/212Χ138 mm, matrix 436Χ300). DWI SE (TR/TE 2634/58 ms, FOV 230Χ230Χ118 mm, matrix: 140Χ136). Single-shot, spin-echo planar imaging (SE EPI) was used in DTI. Scan layers were aligned parallel to the anterior/posterior line with the following settings: TR/TE 6518/60 ms, FOV 256Χ256Χ120 mm, thickness 2 mm, matrix 112Χ112, SENSE factor 2, 15 non-linear directions, b value of 0 and 800 s/mm ² , respectively. Scan time was 4 min and 13 sec.

The various association tracts were identified based on standard atlases, as shown in [Figure - 1]. The post-processing Fiber Trak software was used to measure fractional anisotropy (FA) and apparent diffusion coefficient (ADC) values in different layers of B0 map and color-coded map on the axial images in three groups. In color-coded maps, red represents left and right, green represents before and after, and blue represents top and bottom. An experienced radiologist, blinded to diagnosis, performed the placements of the regions of interest (ROI) for FA and ADC. Each ROI was controlled under two voxels in all tracts.

Three ROIs of IFOF ([Figure - 1]a, along the frontal, fronto-temporal and temporal side) on the right side were measured, each for three times, averaging value was got, the same measurement to the left side. The ROIs of genu of corpus callosum (Gcc) and splenium of corpus callosum (Scc) ([Figure - 1]b, in the middle of Gcc and Scc) were measured three times and averaging value was got. The ROI of SLF II [Figure - 1]c was measured three times on the right side and averaging value was got, the same measurement to the left side. The ROIs of the cingulum tract ([Figure - 1]d and [Figure - 1]e, posterior and middle regions of the cingulum tract) were measured three times on the right side, averaging value was got, the same measurement to the left side. Value of Gcc was set right, Scc was set left. Statistical analysis (paired t test) showed no significant differences between the two sides of IFOF, SLF and cingulum tract, while significant difference was observed between Gcc and Scc. Values of FA and ADC of Gcc and Scc are shown in [Table - 1] and [Table - 2], while values of FA and ADC of IFOF, SLF and the cingulum tract as shown in [Table - 1] and [Table - 2] are the averaging value of the left and right side. Care was taken to minimize inclusion of the cerebrospinal fluid (CSF) and gray matter in ROIs to eliminate partial volume effects.

Statistical analysis

Tests of homogeneity of variance and analysis of variance in clinical data of each group were done. Paired t test was used in left and right FA and ADC values. Analysis of variance and Student-Newman-Keuls (SNK) statistics were done in intergroup comparisons of FA and ADC values.

Results

Clinical data

The clinical data of the three groups are shown in [Table - 3]. Homogeneity of variance tests were done between three groups and the results were as follow: P=0.457 (P>0.05) in age; P=0.496 (P>0.05) in educational years. The results showed homoscedasticity in comparison of age and education in three groups. Then analysis of variance was done, P=0.152 (P>0.05) in age of three groups P=0.116 (P>0.05) in educational years. The results showed no significant differences in the age and education of the three groups.

FA and ADC values

FA values across the three groups are shown in [Table - 1]. As shown in [Table - 4], there were significant group differences in FA values. Relative to NC, aMCI showed decreased FA values in the IFOF and cingulum tract, while AD showed decreased FA values in the IFOF, Gcc, and cingulum tract. AD subjects showed FA values in the cingulum tract that were also significantly lower than aMCI subjects [Figure - 2].

ADC values across the three groups are shown in [Table - 2]. Unlike FA, no significant differences in ADC values were observed across groups, as shown in [Table - 5].

Discussion

The major neuropathological findings of AD include neurofibrillary tangles (NFT) and senile plaques (SP). There are a few neuropathological studies of aMCI from longitudinally studied cohorts. ^{[18],[19],[20],[21],[22],[23]} These studies suggest that aMCI is related to AD pathology and cerebral infarction. It appears that neurofibrillary pathology in medial temporal lobe structures is the major substrate for aMCI and for memory decline and not amyloid plaques. ^[24]

The ROIs selected in the current study were WM tracts that support functions known to be compromised in aMCI and/or AD, including IFOF, corpus callosum (including Gcc and Scc), SLF II and cingulum tract. ^{[5],[14],[25],[26]} The IFOF connects the cortices of the inferior dorsal frontal lobe, temporal lobe and occipital lobe and is involved in spatial information processing, object identification and memory. IFOF also comprises the direct fiber connection between the occipital and frontal lobes. ^[25],[26] The corpus callosum, which contains multiple sub-regions with various functions, serves to connect the cerebral hemispheres. SLF II includes the WM outside the cingulum tract of the superior parietal and superior frontal lobes, which extends to the dorsal premotor and dorsolateral prefrontal regions. ^[14] Finally, the cingulum tract, which is associated with memory function, connects the anterior and posterior cingulate gyrus and descends into the medial temporal lobe. ^[5]

The process of AD is in accordance with sequences as follows: from the entorhinal cortex/entorhinal cortex transition zone to the hippocampus, and then cingulate gyrus to the lateral temporal lobe, frontal and parietal lobe, and finally to the occipital lobe. Recent studies dealing with this topic mention that grey and WM atrophy involved prefrontal, parietal, and temporal lobe areas in NC and prefrontal, cingulate, and parietal lobe areas in aMCI subjects and agreed with the pattern of fiber tract changes. ^{[27],[28],[29]} The current study was conducted to determine whether DTI could detect changes in cerebral association fiber tracts in aMCI and AD subjects and to assess its efficacy for differential diagnosis. Relative to NC, aMCI subjects showed significant decreases in FA values of IFOF and cingulum tract, suggesting that DTI of IFOF and cingulum tract could discriminate aMCI from NC and has the potential to be used as a biomarker of aMCI. AD subjects showed decreases in FA in the IFOF, Gcc, and cingulum tract. In fact, cingulum tract changes in AD subjects were more severe than in aMCI subjects and with significant differences (P<0.05) in intergroup comparison of aMCI and AD subjects [Table - 4]; combined with Gcc, they have the potential to be useful for differential diagnosis. The results may reflect the disease process that aMCI is early-stage Alzheimer's disease.

Lower FA values in the Gcc than in the Scc in all groups suggest involvement of the Gcc in the late process of AD, supporting the "reverse origin" theory, which states that Gcc develops later than Scc, while Gcc is involved earlier than Scc in the pathologic process of AD. ^[30],[31] We found no significant intergroup differences of ADC values, which is in agreement with previous results. ^[32] More research is needed to determine whether ADC values may be useful in the diagnosis of aMCI or AD.

Some investigations of region of interest (ROI) in aMCI and AD patients using the same method, have showed different results in the association fiber tracts concerning to corpus callosum, cingulum tract, SLF and IFOF. ^[33],[34] The deficits of the anterior aspects of Gcc were reported in aMCI patients. ^[35],[36] Bai et al., found significant abnormalities between aMCI and NC on both sides of the IFOF, cingulum tract, and Gcc, which the age, gender, educational years were well matched in aMCI and NC subjects. ^[37] The different results may relate to the heterogeneity of the examined subjects and different MR parameters. The choice of a very small ROI of the IFOF and Gcc may also play a role for incompletely representing the integrity of whole fiber tracts. Intergroup comparison indicated that patients with AD had significantly lower FA values in the IFOF, Gcc and cingulum tract than did patients with NC, suggested neural degeneration of limbic system in AD may cause WM structure destruction. Connective destruction between the frontal, temporal, occipital lobe and limbic system existed at the same time, which supports the "disconnection" theory. ^[35] AD can be regarded as a cortical disconnection syndrome that affects not only cortical neuronal soma but also the axons and dendrites in the cerebral WM. In DTI, elevated ADC values are related to neurodegeneration, whereas decreased FA values indicate disturbed WM homogeneity.

It should be noted that there were several limitations to the present study. Data were analyzed with the ROI method, which is subject to inherent subjectivity. However, the radiologist was blinded to the clinical findings during analysis. Reproducibility of manual drawings, placements and selections of ROIs for DTI data analysis is may also vary across radiologists. Moreover, the ROIs were selected in close proximity but not in overlapping areas. In addition, the DTI applied 15 gradient directions for diffusion encoding, and more accurate FA measurement can be achieved with more advanced MRI systems. The usefulness of the imaging markers found in the current study should be confirmed and further improved for sensitivity in individual patients. Finally, in the current study, MCI subjects were not followed longitudinally to observe whether they progressed to AD, and AD-related biologic signs were not collected. MCI should be best studied in longitudinally followed patients in centers that are experienced in dementia disorders.

In summary, this study demonstrated decreased FA values in the IFOF and cingulum tract of aMCI patients and in the IFOF, Gcc, and cingulum tract of AD patients, suggesting pathologic changes in these regions compared to healthy control subjects. Furthermore, AD patients showed more severe decreases in FA in the cingulum tract than aMCI patients. These findings suggest that differential diagnosis of aMCI and AD could be made using FA values of the Gcc and cingulum tract. Overall, this study suggests that DTI measurements of association fiber tract changes may serve as functional markers for diagnosis and monitoring of aMCI.

Acknowledgment

This research was partly supported by Science and Technology Commission of Shanghai Municipality (Grant No. 08411951200).

References

Copyright 2011 - Neurology India

The following images related to this document are available:

Photo images