Diagnostic pitfalls: a case of neurosarcoidosis mimicking tuberculous meningitis

Abstract

We report on a 44-year-old Caucasian male patient who developed recurrent unilateral vitreous hemorrhages in February 2008. Ophthalmologists assumed a diagnosis of Eales disease (Fig. 1a). In July 2008, he complained of exhaustion, general weakness and gait unsteadiness. In November 2008, he suffered from acute vestibular dysfunction, which responded to intravenous corticosteroid treatment. On admission to our hospital in December 2008, he showed a spinal syndrome with paraspasticity and gait ataxia. His past medical history was unremarkable, but his mother had been treated for tuberculosis in the 1950. Cranial and spinal MRI showed no intraparenchymal abnormalities in the CNS, but widespread nodular, contrast-enhancing meningeal lesions around the cervicothoracic spinal cord (Fig. 1b) and brain stem (Fig. 1c) extending to the cerebellar surface, cervical spinal roots and right trigeminal and vestibular nerve. External MRI 1 week before found only a minor contrast enhancement on the brain stem surface. Chest computer tomography revealed bihilar lymphadenopathy. Blood tests (CRP, erythrocyte sedimentation rate, calcium, ACE, sIL2-R, auto-antibody screening) were normal, except for lymphopenia (0.44/nl, normal CD4/ CD8 ratio). Cerebrospinal fluid (CSF) showed lymphocytic pleocytosis (99 cells/ll) (immune phenotyping in Fig. 1d), increased protein (464 mg/dl) and lactate levels (54.1 mg/dl), decreased glucose (30.8 mg/dl), intrathecal IgA synthesis (16%), oligoclonal bands and increased sIL2-R (1,303 IU/ml). Bronchoalveolar lavage demonstrated inflammation with lymphocytosis (normal CD4/CD8 ratio). Non-caseating epithelioid cell granulomata were discovered in biopsies of an infracarinal lymph node and lung tissue (Fig. 1e). Microbiological tests ruled out a broad spectrum of bacterial, viral and fungal infections. Cytopathological examination of CSF and BAL found no malignant cells. Tuberculin skin test (PPD5), analysis of blood and CSF by T-SPOT.TB and QuantiFERON-TB Gold In-Tube test (QFT-G-IT), microscopy (Ziehl-Neelson staining for acidfast bacteria) and PCR for tuberculous mycobacteria showed no evidence of latent or active mycobacterial infection (5 9 CSF, 3 9 sputum, 3 9 urine, 2 9 BAL, 2 9 lymph node biopsy). After exclusion of malignancies and other infectious or systemic autoimmune diseases, the imperfect sensitivity of mycobacterial diagnostic tests prevented definitive distinction of the remaining differential diagnoses: CNS tuberculosis and neurosarcoidosis. The diagnostic dilemma led to a pragmatic treatment decision including F. Scheibe (&) O. Wengert L. Harms K. Ruprecht Department of Neurology, Charite–Universitatsmedizin Berlin, Chariteplatz 1, 10117 Berlin, Germany e-mail: franziska.scheibe@charite.de