Background/Purpose: Central nervous
system involvement is one of the most serious complications in Behçet’s disease
(BD). This condition is referred to as neuro-Behçet’s disease (NB) and can be classified into
acute NB (ANB) and chronic progressive NB (CPNB) based upon differences in the clinical
course and response to corticosteroid treatment. Brainstem atrophy is significantly
more frequently observed in CPNB than in ANB. It is also noteworthy that
cognitive dysfunction, in addition to truncal ataxia, is frequently observed in
CPNB, and this cannot be accounted for by brainstem
atrophy. In the present study, we examined volumes of the hippocampus in
order to identify the responsible lesions for neurobehavioral changes in CPNB.

Methods: The subjects were 32
patients, including 13
with CPNB (11 males and 2 females, age 51.2 ± 12.1 years old [mean ± SD]), 13 with
Behçet’s
disease
without NB (non-NB) (10 males and 3 females, age 54.4 ± 11.4 years old), and 6
with Alzheimer’s
Disease (AD) (5
males and 1 female, age 78.8 ± 7.5 years old). All patients with BD satisfied
the international classification criteria for Behçet’s disease. CPNB was
defined as intractable, slowly progressive neurobehavioral changes and/or
ataxia accompanied by persistent elevation of interleukin-6 of >20 pg/mL in
cerebrospinal fluid on two different occasions at an interval of at least 2
weeks. All patients with AD
satisfied the Diagnostic and Statistical Manual of Mental
Disorders (DSM)-IV criteria. Sagittal
sections of T1-weighted images on brain magnetic resonance imaging (MRI) were
obtained from each subject. The areas of the midbrain tegmentum and pons were
measured on mid-sagittal sections of T1-weighted images using image analysis
software (Image J ver.1.45: NIH, USA). Severity of gray matter loss in the
hippocampal region and whole brain were investigated using Voxel-Based Specific
Regional Analysis System for Alzheimer’s Disease (VSRAD) software (Eisai Co.,
Ltd) to determine the degrees of hippocampal region atrophy (Z score) and whole-brain
atrophy (WBAI).

Results: The brainstem area was
significantly decreased in CPNB (461.8 ± 87.3 [mean ± SD]) compared with those in AD (661.9 ± 56.1) and
non-NB (666.1 ± 50.6)
(Figure 1, A). VSRAD analysis showed that Z score was significantly increased
in CPNB (1.46
± 0.70) and AD (3.13 ± 1.21) compared with non-NB (0.77 ± 0.40) (Figure 1, B). All
patients with CPNB showed brainstem atrophy, but there was no significant
correlation between the area of brainstem atrophy and Z score. Neither Z score
nor WBAI was correlated with age in CPNB.

Conclusion: These results indicate
that the hippocampus, in addition to the brainstem, is a common site for lesions
in CPNB, accounting for the progressive cognitive dysfunction in this disease.
The lack of correlation between brainstem atrophy and hippocampal atrophy
suggests that predisposing factors might determine the lesion site in CPNB.