Neurosarcoidosis in a patient treated with tumor necrosis factor alpha inhibitors

Abstract

We describe the case of a 36-year-old Caucasian man with a 9-year history of HLA-B27-positive ankylosing spondylitis, who presented with new seizures. He had a non-focal neurological exam. Blood tests showed mild leukocytosis and elevation in liver transaminases. Initial head CT showed hypodensity in the left frontal subcortical white matter. MRI of the brain showed leptomeningeal enhancement and several hyperintense lesions (Fig. 1a–c). He had been on adalimumab (Humira ) for 3 years, but this was discontinued as a repeat MRI showed progression and increased size of the lesions 1 month later. He was then treated with intravenous SoluMedrol at 1 g/day for 5 days with prednisone taper and maintained on methotrexate. CSF studies showed mild lymphocytic leukocytosis and elevated protein. EEG showed bihemispheric slowing; left worse than right. Brain biopsy showed well-formed non-caseating granulomas consistent with neurosarcoidosis (Fig. 2a–c). Leading up to this presentation, he had also been treated with etanercept (Enbrel ) for 2 years prior to adalimumab. Tumor necrosis factor alpha inhibitors (TNFAIs) include etanercept (Enbrel )—a TNFa p75 soluble receptor fusion protein; and adalimumab (Humira ) and infliximab (Remicade )—both monoclonal antibody to TNFa. Despite the efficacy of TNFAIs in many autoimmune arthropathies and refractory sarcoidosis, treatment with these agents may have a paradoxical effect, as seen in this case. Neurosarcoidosis, in the presence of multisystem organ involvement and neurologic deficits, has an incidence of approximately 1 per 100,000 and is present in 5 % of systemic sarcoidosis cases [1]. It has been found strictly confined to the CNS in approximately 0.2 per 100,000. CNS involvement usually reveals granulomatous infiltration of the meninges and parenchyma, most prominent at the base of the brain [2, 3]. The etiology of sarcoidosis remains unknown; however, it may possibly be linked to genetic susceptibility with increased Th1 immune response to environmental factors [4]. TNFa is one of the main cytokines responsible for initiation and formation of granulomas; therefore, besides cytotoxic drugs such as methotrexate, azathioprine, and cyclosporine, TNFAIs have been used in some refractory cases of sarcoidosis [5, 6]. Combination treatment with mycophenolate mofetil and infliximab was shown to be effective in treating neurosarcoidosis [7]. Considering this known pathological correlation and efficacy of TNFAIs, it is somewhat surprising and paradoxical that TNFAIs would cause granulomatosis. Tong et al. [8] summarized a total of 37 cases of sarcoidlike granuloma development after TNFAIs therapy. This phenomenon was reported in all three TNFAIs, suggesting this is a ‘‘class effect’’ [8–10]. Almost all cases are peripheral sarcoidosis of the lungs and skin [8]. There was only one case of neurosarcoidosis reported in a 41-year-old female being on infliximab and methotrexate for rheumatoid arthritis [11]. In our case, the temporal relationship between neurosarcoidosis development while being on etanercept and adalimumab, and resolution of the disease while being off the agents, is highly suggestive of causality. We did not rechallenge the patient with another TNFAI as there has been previous reports on recurrence of sarcoidlike granulomas with rechallenging [8, 12]. Y. Mao-Draayer (&) Neurology Department, University of Michigan, Ann Arbor, MI 48103, USA e-mail: maodraay@med.umich.edu