Cyclosporine for Behçet's uveitis: is it associated with an increased risk of neurological involvement?
Nøgleord
Abstrakt
OBJECTIVE
The immunosuppressant cyclosporine is widely used to treat Behçet's disease (BD). The aim of this study was to determine whether cyclosporine increases the risk of neurological involvement in BD.
METHODS
Patient files from the Ophthalmology Department for the period 2000-2005 were screened retrospectively, and the occurrence of neurological involvement and its relationship to ocular severity were evaluated.
RESULTS
A total of 454 patients with BD were seen at the Ophthalmology Department in this period, including 24 who had been referred from the Neurology Department. Excluded from the study were 47 patients who did not have uveitis and 114 patients with an inadequate follow-up. The remaining 269 patients had been treated with either cyclosporine (Group I, n=92), other immunosuppressants (Group II, n=132), or no treatment other than colchicine (Group III, n=45). Patients with neurological symptoms were sent to the Neurology Department for evaluation: 20 from Group I [10 with primary headache, 1 with depression, 1 with sinus thrombosis, and 8 with parenchymal neurological involvement (pNBD)]; 13 from Group II [10 with primary headache, 1 with pre-morbid epilepsy, 1 with sinus thrombosis, and 1 with pNBD]; and 5 from Group III with primary headache. The frequency of pNBD was significantly higher in Group I, and included atypical features such as seizures and MRI lesions as well as the typical symptoms of brainstem involvement and pleocytosis. Eye involvement tended to be more severe in Group I, and the difference remained significant both when milder cases were excluded from the analysis (6 vs. 0; p=0.03) and when severe cases were excluded (p=0.04). pNBD was significantly more frequent in patients on cyclosporine alone than in those receiving cyclosporine plus another agent.
CONCLUSIONS
In patients with Behçet's uveitis, cyclosporine seems to be associated with an increased risk of developing pNBD, although the reason for this is unknown. A prospective trial is needed to shed light on this problem.
