Media Centre

Diagnosing coccidioidal meningitis using CSF (1,3)-beta-D-glucan testing

September 14 2016

Diagnosing coccidioidal meningitis is challenging, due to infrequent positive test results and often a delayed positivity. If left untreated, the mortality is 90% in 1 year and 100% in 2 years. Currently, diagnosis depends on isolating coccidioides from a cerebral spinal fluid (CSF) sample, but this test is only positive in around 1/3 of patients. A positive result for anti-coccidioidal antibodies in  CSF provides a presumptive diagnosis, but the result may often be negative in the earlier stages of  disease. β-glucan levels in  CSF may be another avenue to explore, as there is no reason for β-glucan to be present in the CSF in the absence of invasive fungal diseases.
 In a recent study therefore, the diagnostic performance of β-glucan levels in CSF samples was assessed in the context of suspected coccidioides meningitis (Stevens et al 2016)

Twenty-six  CSF specimens were collected over 15 years from 21 patients considered to have coccidioidal meningitis. 11 specimens were taken from patients with no clinical evidence for a CNS or systemic fungal diagnosis. β-1,3 glucan assays were performed with the Fungitell kit (Associates of Cape Cod, Falmouth, MA). In addition, the coccidioidal antibody assay and the coccidioidal polymerase chain reaction assay were performed on all samples.

β-glucan concentrations in CSF samples ranged from 18-3300pg/ml in coccidioidal meningitis patients, while the range was <3.9-103pg/ml in the control group. A cut-off of 31pg/ml was used, according to the bottom of the range directed by the commercial kit. Just one coccidioidal meningitis patient had a CSF level below this cut-off, while two control patients had a level above. This gives a sensitivity of 96%, a specificity of 82%, a positive predictive value of 93% and a negative predictive value of 90%. CSF samples with B-1,3 glucan levels around103pg/ml were shown to be extremely likely to have coccidioidal meningitis (when other mycoses can be excluded).

To summarise, a high level of discrimination and accuracy was found for this test. Previous studies have reported sensitivities of 53% and 50% in acute and past coccidioidomycosis. Sensitivity in disseminated and meningeal coccidioidomycosis has been reported at 83%. These results in combination with the current data suggest that sensitivity is increased in a compartmentalised infection. Testing CSF for β-glucan in meningitis cases of unknown aetiology could assist the clinician to a diagnosis of fungal meningitis and a  positive result could lead to a better travel history, or to  testing CSF for coccidioidal antibodies. 

Electron microscopy of Coccidioides immitis

Article  J. Clin. Microbiol. doi:10.1128/JCM.01224-16   Stevens . D  et al 2016