David M. Simpson, MD


July 22, 2004


I have a patient who presented with altered mentation, incontinence, and gait difficulty. MRI showed hydrocephalus but no localized gadolinium-enhancing lesions. Cerebrospinal fluid (CSF) glucose was 36 mg/dL with 15 white cells, all of which were lymph. Shunt surgery was planned, but preoperative evaluation revealed positive HIV with a CD4+ cell count of 16 cells/mcL; headache/stiff neck worsened. Should empirical therapy be started? What about highly active antiretroviral therapy (HAART)?

Nobuyoshi Minemura, MD

Response from David M. Simpson, MD

The first critical step in neurologic diagnosis is neuroanatomic localization of the lesion. Of importance in description of the neurologic examination is whether there are signs of elevated intracranial pressure, ie, a depressed level of consciousness and papilledema. The value of CSF opening pressure would add further helpful information.

It is important to determine whether the neuroradiologic studies demonstrate hydrocephalus of a communicating or noncommunicating form. In the latter case, an obstructive lesion, such as a structural block in the ventricular system, would cause dilation of the ventricles above the lesion. The most likely situation in this case is communicating hydrocephalus. Further evidence for the diagnosis of communicating hydrocephalus is the combination of a dilated ventricular system and relatively small cortical sulci. If the cortical sulci are enlarged in proportion to the ventricular enlargement, this is termed hydrocephalus ex vacuo, which is representative of generalized cerebral atrophy, a common finding in HIV-infected patients.

The cause of hydrocephalus in this case is uncertain. The pace of temporal progression of the neurologic deficits would add helpful information concerning potential etiologies, further diagnostic studies, and treatment options. A chronic and insidious course is typical of normal-pressure hydrocephalus, which may be on a primary (ie, idiopathic) or secondary basis. In an HIV-infected patient, a secondary process is far more likely because of an infectious or granulomatous process. Inflammatory cells and their byproducts may obstruct CSF flow and absorption, resulting in hydrocephalus. With rapid clinical progression, one must consider an acute infectious process.

The presence of headache and stiff neck, together with CSF pleocytosis, suggests meningeal irritation. Although the CSF white blood cell count is only modestly elevated, severely immunosuppressed patients with low CD4+ cell counts may be unable to mount a brisk inflammatory response, even in the setting of fulminant infection. Thus, infectious processes, such as cryptococcal, tuberculous, cytomegalovirus or varicella zoster meningitis, must be excluded. Alternatively, neoplastic lesions, such as lymphomatous meningitis, may present in this fashion.

In the absence of a definable specific etiology, HIV-related meningitis is the likely diagnosis. An elevated CSF HIV viral load would add further support to this diagnosis. Management options include HAART, particularly with agents that have maximal CSF penetration. If clinical deterioration due to progressive hydrocephalus occurs even in the setting of effective antiviral therapy, then CSF shunt might be of benefit. Other methods to control elevated intracranial pressure, such as corticosteroids and acetazolamide, may be helpful but are of unproven efficacy in this setting.