Selective anterior basal enhancement: a distinctive MRI phenotype of neurosarcoidosis mimicking tuberculous meningitis
Abstract
Case Presentation: A 45-year-old Black woman from a tuberculosis-endemic region presented with four days of acute bilateral visual loss and a two-month history of distal limb paresthesias. Best-corrected visual acuity was 20/300 in the right eye and 20/200 in the left, with bilateral relative afferent pupillary defects and normal fundoscopic examination. Brain MRI demonstrated intense, homogeneous gadolinium enhancement selectively involving the optic nerves, optic chiasm, infundibular stalk, and bilateral trigeminal nerves at Meckel's cave, with conspicuous sparing of the intervening basal leptomeninges. CSF analysis revealed 59 leukocytes/mm³ (98% lymphomononuclear), protein 249 mg/dL, and glucose 37 mg/dL. Serum ACE, PPD, CSF GeneXpert, and serum AQP4-IgG were negative. Electrodiagnostic studies disclosed a symmetric distal pure motor polyneuropathy. Mediastinal lymph node biopsy demonstrated non-caseating granulomas, confirming neurosarcoidosis. Initial high-dose corticosteroid therapy produced partial improvement, followed by steroid-refractory disease. Discussion: Neurosarcoidosis most frequently manifests with leptomeningeal disease or multifocal cranial neuropathies. In contrast, selective enhancement confined to the anterior basal neural axis—including the optic pathways, infundibulum, and Meckel's cave while sparing adjacent leptomeninges—is distinctly uncommon. Isolated hypothalamic-pituitary involvement is reported in approximately 12.6% of patients, trigeminal nerve involvement in 5.7%, and asymptomatic trigeminal enhancement in only 3.7%, virtually always occurring alongside additional neuroinflammatory lesions rather than as part of a selective anterior basal pattern. This imaging phenotype is also unusual in neurotuberculosis, where basal cisternal leptomeningitis typically predominates. Therefore, the absence of diffuse leptomeningeal enhancement, despite marked involvement of anterior basal neural structures, represents an important radiological clue against tuberculous meningitis. Although CSF GeneXpert shows high specificity (~98%), its limited sensitivity (50–60%) does not exclude tuberculous meningitis. Likewise, serum ACE has poor diagnostic performance for neurosarcoidosis (overall sensitivity ~51%, falling to 14–35% in isolated disease). In this biomarker-negative setting, MRI pattern recognition narrowed the differential diagnosis and directed tissue biopsy, ultimately establishing the diagnosis. Final Comments: Selective anterior basal enhancement with clinico-radiological dissociation represents an uncommon but highly informative MRI phenotype of neurosarcoidosis. Recognition of this pattern may facilitate earlier diagnostic suspicion and appropriate biopsy selection, particularly in tuberculosis-endemic regions where neurotuberculosis is the principal radiological mimic.