The Role of Lumbar Puncture in the Management of Elevated Intracranial Pressure in Patients with AIDS-Associated Cryptococcal Meningitis

Sir—We read with great interest the study by Graybill et al. [1] regarding the diagnosis and management of elevated intracranial pressure in patients with AIDS and cryptococcal meningitis, and we agree that high intracranial pressure without any evidence of obstructive hydrocephalus is a frequent complication of cryptococcal meningitis and is associated with a poor prognosis and more-severe symptoms. However, it seems to us that the authors' recommendation that an “opening pressure >250 mm H2O be treated with large-volume CSF drainage” needs to be considered with caution, despite their observation that this practice was not associated with any major adverse reactions. From January 1986 through December 1999, 470 consecutive HIV-infected patients were evaluated for neurological disorders at the Infectious Diseases Institute of Milan, in Milan, Italy, and underwent a lumbar puncture (LP) as part of the workup. Table 1 shows the final diagnoses and the number of LPs performed. Fatal brain herniation was observed in 2 patients affected by cryptococcal meningitis, both of whom underwent multiple LPs that were aimed at reducing intracranial pressure by means of CSF drainage. HIV-infected patients evaluated for CNS disorders by means of lumbar puncture (LP). Patient 1 was a 26-year-old male injection drug user who had cryptococcal meningitis diagnosed in March 1989 and who experienced a clinical and mycological relapse in November of the same year. The results of a brain CT scan were normal. From 6 November through 11 November, the patient underwent 3 LPs, the last 2 of which were done to treat elevated intracranial pressure. After the third LP was performed (opening pressure, 600 mm H2O), he went into a coma, in association with mydriasis and an absence of response to pain stimuli, and he died 8 h later. Autopsy revealed meningoencephalitis with marked cerebral edema and bulbar-cerebellar herniation. Patient 2 was a 33-year-old homosexual man with AIDS (cutaneous Kaposi's sarcoma) who presented in March 1990 with fever, headache, rigor nucalis, and an altered mental status. The results of a brain CT scan were normal, and an LP revealed the presence of Cryptococcus neoformans. After a second LP was performed 3 days later (opening pressure, 450 mm H2O), the patient became unconscious and had bilaterally dilated and unreactive pupils and a vasomotor reaction throughout the trunk. Although iv mannitol was immediately administered and led to regression of the coma, it resulted in residual amaurosis. The patient was admitted to the intensive care unit, where he experienced several episodes of waxing and waning consciousness and finally died 3 days later. Autopsy revealed meningoencephalitis with marked cerebral edema and cerebellar herniation. Since then, we have stopped using CSF drainage to treat elevated intracranial pressure in patients with cryptococcal meningitis. We believe that cryptococcal meningitis is associated with a higher risk of brain herniation than are any of the other CNS complications observed in patients with AIDS and that great caution is required when frequent LPs are performed to achieve mechanical decompression. Furthermore, the LP should be considered investigational until an “ad hoc” study is performed.