Media Centre

Randomised study of dexamethasone in cryptococcal meningitis shows worse outcomes

March 07 2016

In 451 patients from Vietnam, Thailand, Laos, Indonesia, Uganda and Malawi with cryptococcal meningitis complicating AIDS, adjunctive dexamethasome lead to more disability, more adverse events and slower clearance of C. neoformans from the CSF.

Justin Beardsley from the Wellcome Trust Major Overseas Program in Ho Chi Minh City, Vietnam writing in the New England Journal of Medicine had “hypothesized that dexamethasone would improve outcomes by reducing intracranial pressure and the incidence of inflammatory complications and by decreasing the incidence of IRIS.” Six weeks of dexamethasone at 0.3 mg/Kg intravenously initially or placebo was given in addition to amphotericin B and fluconazole. The dexamethasone regimen was the same as that used for tuberculous meningitis which results in less disability.

At 14 days, CSF opening pressure was lower in the dexamethasone group, but C. neoformans colony counts were higher. At 10 weeks, 42% had died in the placebo arm and 48% in the dexamethasone arm (p=0.45), but a good outcome was only seen in 13% of dexamethasone patients compared to 25% of placebo recipients (p <0.001). These differences were mirrored at 6 months. The study was stopped early because of worse outcomes and more adverse events in the dexamethasone arm. Significantly different adverse events between the two arms are shown in the table:

Adverse event

Dexamethasone (%)

Placebo (%)

P value

Infection or infestation

48 (21)

25 (11)


Gastrointestinal disorder

29 (13)

16 (7)


Renal or urinary disorder

22 (10)

7 (3)


Cardiac disorder

8 (4)




32 (14)

6 (3)



52 (23)

19 (8)



108 (48)

132 (58)



114 (51)

75 (33)


Remarkably, 42 patients were excluded from enrolment because they had already received glucocorticoid therapy. Prior study of raised intracranial pressure has shown no benefit from glucocorticoids, mannitol or acetazolamide, and only repeated lumbar puncture or insertion of a lumbar drain or ventricular drain are clinically useful. Glucocorticoids are recommended for cryptococcal IRIS, but data are scant and also for the very few patients who have a cryptococcoma and substantial cerebral oedema or mass effect. This study did not have the power to evaluate these entities.

The authors conclude: “With no effective adjunctive therapy yet identified, improving access to the most effective antifungal treatments, including flucytosine, must remain a global priority.”

Beardsley J et al, NEJM 2016;374:542-54.