Neurosarcoidosis: A Review

Abstract & Commentary

Source: Kidd D, Beynon HL. The neurological complications of systemic sarcoidosis. Sarcoidosis Vasc Diffuse Lung Dis. 2003;20:85-94.

Sarcoidosis is a multisystem granulomatous disease most often affecting the lungs, eyes, skin, and liver, with neurological involvement in 5-10%. Cranial neuropathy is the most common clinical manifestation, ranging from 60-73%, usually affecting the facial nerve (frequency up to 64%), frequently bilaterally. Sarcoid granulomata may affect the facial nerve in the parotid gland, at the skull base, in the internal auditory canal, or at the brainstem. Other cranial nerves may be affected, including the optic nerve presenting as progressive painless loss of vision or as painful optic neuropathy indistinguishable clinically from demyelinating optic neuritis. Lower cranial nerves IX, X, and XII are involved less frequently.

Cerebral hemisphere involvement occurs either as meningoencephalitis, presenting with altered mental status including psychiatric symptoms, fluctuating focal deficits, or impaired levels of consciousness, or rarely as a focal mass with presentation depending on location of the lesion. Most commonly, intracranial involvement affects the pituitary and hypothalamus and the area around the third ventricle. Diabetes insipidus is most common, but obesity, altered sleep and temperature function, and involvement of the optic chiasm with visual field defects are also seen. Erosion of the sella turcica is rare.

Chronic basal granulomatous leptomeningitis is a common autopsy finding and may be asymptomatic. More often, multiple cranial neuropathies or hydrocephalus due to foraminal obstruction result. Acute meningitis, cerebral infarction, or involvement of the cerebellum or brainstem occur rarely.

Spinal cord involvement most commonly presents as cervical myelopathy or conus medullaris syndrome with severity varying from mild to complete transverse myelitis. Cauda equina involvement, predominantly sensory and painless, also occurs and is associated with elevated protein and cells in the spinal fluid.

Chronic large fiber sensorimotor neuropathy, purely sensory neuropathy, a Guillain-Barré-like syndrome, or mononeuritis multiplex all occur in sarcoidosis. Granulomas may be seen on nerve biopsy but perineural vasculitis is also described. Painful slowly progressive myopathy is seen, as is a nodular, frequently asymptomatic form with palpable lesions in muscle that enhance on magnetic resonance imaging.

Absent any evidence from double-blind, placebo-controlled treatment trials, therapy encompasses steroids and immunosuppressants as steroid-sparing agents including azathioprine, methotrexate, cyclosporine, cyclophosphamide, and chlorambucil. Not all patients respond. Radiotherapy may benefit large, treatment-resistant, mass lesions.


Sarcoidosis remains an etiologic enigma. Recently, a thought-provoking article argued cogently that sarcoid might not be the consequence of a specific etiologic agent. Rather, it is the end result of an abnormal immune response to any of a variety of different exposures.1 Sarcoid granulomas would result from abnormal antigen processing due to an impaired T-cell immune response. Absent the immune system’s ability to efficiently clear the invading antigen, granuloma formation would be the "fallback position."

Ample clinical evidence exists in support of this hypothesis. Mortality among steroid-treated sarcoid patients is higher than in nonsteroid treated patients. Perhaps steroids promote such morbidity by impairing an already paretic T-cell system. Conversely, patients with erythema nodosum, implying the presence of a brisk immune response, do better. Those with combined variable immunodeficiency, on the other hand, frequently develop sarcoid, again implicating T lymphocytes. Parenthetically, this hypothesis also explains why no etiologic agent has been identified.

It does not explain all the clinical data, however.2 Sarcoidosis-like reactions occur in HIV patients but only after the administration of highly active antiretroviral therapy (HAART), making this an example of sarcoid occurring in a relatively privileged situation immunologically rather than underprivileged. Why do sarcoid granulomas develop in allografts following transplantation, given that these tissues are presumably free of inciting antigens? Questions remain, but the theory is intriguing and would preclude our ever finding a cause for sarcoid. Which, ironically, would be the ultimate proof that the theory is on the mark. — Michael Rubin, Professor of Clinical Neurology, New York Presbyterian Hospital-Cornell Campus, Assistant Editor, Neurology Alert.


1. Reich JM. Chest. 2003;124:367-371.

2. Judson MA. Chest. 2003;124:6-8.