Creutzfeldt–Jakob disease (CJD) is a rapidly progressive, fatal and transmissible neurodegenerative disorder. The clinical features of typical CJD are progressive dementia, generalised myoclonus and, in the later stages, akinetic mutism. Characteristic findings in laboratory examinations include periodic sharp wave complexes (PSWCs) on electroencephalogram and elevated 14-3-3 protein in CSF. However, these tests cannot be applied to assessment of the progression of the lesion. Brain biopsy confirms the diagnosis of CJD but it does not reveal how the lesion develops in the brain.

MR examinations are routinely performed in Japan for patients with suspected CJD. Diffusion weighted imaging (DWI) is now accepted as the most useful imaging modality for diagnosing patients with CJD, even in the early stages or without PSWCs.1 DWI demonstrates hyperintensity for lesions in the striatum and the cerebral cortex, with high sensitivity.2 However, the underlying pathomechanism of the abnormal signals remains unclear.

Diffusion tensor imaging (DTI) is a new MR technique that can indirectly assess the integrity of tissue. The diffusivity and anisotropy of water molecular displacement are quantified by apparent diffusion coefficient (ADC) and fractional anisotropy (FA), respectively. To our knowledge, however, the utility of DTI in CJD has not been reported. Here we demonstrate the potential of DTI for the diagnosis and analysis of the underlying pathology of CJD, by estimating values of ADC and FA.

We obtained approval from the Clinical Trial Centre of Developmental Therapeutics at our institution. MR images were obtained in three sporadic CJD (sCJD) patients who fulfilled the WHO criteria for probable sCJD, and in three clinically normal control subjects seen from 2004 to 2007 at our department. Age range of patients with sCJD (71.7 (±1.5) years) and control subjects (67.7 (±2.3) years) were similar. Patients showed the phenotype of classical sCJD. Although pathology was not obtained, two of the three patients who underwent the PRNP gene study were homozygous for methionine at codon 129.

All cases underwent DTI with a 3 T clinical MR unit (Signa 3TVhi; GE Medial Systems, Milwaukee, Wisconsin, USA). DTI was obtained by single shot, spin echo, echo planar imaging with sampling of the six axis diffusion tensor. Thirty axial slices covering the whole brain were imaged for a total acquisition time of 124 s using the following acquisition parameters: 10000/87.7 (TR/TE), field of view 28×28 cm, image matrix 128×128 pixels, section thickness 4 mm gapless and b value of 1000 s/mm².

Isotropic diffusion weighted images, ADC maps and FA maps were computed on a network workstation (AW4.3; GE Medical Systems), and distorted correction was conducted by the FUNCTOOL 2.6.0 (GE Medical Systems) package. Values of ADC and FA were calculated from the three orthogonal directions. Regions of interest (ROI) were manually selected to encircle the specific brain region while excluding the adjacent CSF. ROI were traced on EPIT2 weighted images acquired as part of the epiDWI sequence. ROI were selected in the head of the caudate nucleus, putamen, globus pallidus, thalamus, posterior limb of the internal capsule (PLIC) and deep white matter in each patient (fig 1A). ROI measurements of the corresponding anatomical structures from both the left and right hemispheres were averaged, and representative ADC and FA values were obtained for each anatomical region. FA values were scored on a scale from 0 to 1. The overall difference in each ADC and FA value between the sCJD group and control group was analysed individually by the use of the unpaired two tailed Student’s t test. A p value of 0.05 or smaller was regarded as statistically significant.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 (A) Apparent diffusion coefficient (ADC, left) and fractional anisotropy (FA, right) maps. Circles indicate the examined regions of interest (ROI). ROI were selected in the head of the caudate nucleus, putamen, globus pallidus, thalamus, posterior limb of the internal capsule and deep white matter. (B) Comparison of ADC values in patients and controls. Mean ADC values of the ROI in each group are indicated. ADC values in patients with sporadic Creutzfeldt–Jakob disease (sCJD) were decreased in the striatum. (C) Comparison of FA values in patients and controls. Mean FA values of the ROI in each group are indicated. FA values in patients with sCJD were lower than those in controls in the internal capsule, but were still within the normal range. *Significant difference between patients and controls (p<0.05).

The ADC values measured from different ROI of both patients with sCJD and control subjects are shown in fig 1B. Patients with sCJD had statistically lower ADC values than control patients in the striatum (p = 0.01). The trend towards decreased ADC values in the caudate nucleus, putamen, globus pallidus and thalamus correlated well with previously published reports. We found a statistically significant difference in the PLIC for comparison of FA values between patients with sCJD and control subjects (p = 0.019). However, both of these FA values in the PLIC were within the normal range. Therefore, we conclude there was no evidence of any abnormality of anisotropy in either group.

Discussion

In this study, we tested the feasibility of DTI, a new technique that provides information on tissue damage in patients with CJD.

We showed here that mean ADC values of the striatum were significantly lower in patients with CJD than in controls. ADC values usually decrease in lesions where DWI shows hyperintensity in patients with CJD, as has been reported previously.3 The underlying mechanisms for ADC changes in patients with CJD remain to be elucidated. It has been hypothesised that patients with CJD show abnormalities on DWI or ADC because of activated microglia, spongiform changes or accumulation of hydrophobic and abnormal prion protein.

FA is believed to reflect the integrity of the fibre tracts in tissue. FA is usually decreased in lesions of many neurodegenerative diseases, including Alzheimer’s disease4 or Parkinson’s disease,5 probably because of the damage to fibre tracts. However, in our study, mean FA values of patients with CJD were not significantly different from those of controls. These results of ADC and FA values indicate that damage to the fibre tracts is not as intense as those of neurons in patients with CJD in its early stages.

In conclusion, we demonstrated that ADC and FA values on DTI could be used to assess the pathophysiology of patients with CJD. It is clear that ADC values can be used to diagnose CJD while FA values, in combination with ADC values, can be useful for assessment of the pattern of disease progression, at least in the early stages.