Very-High-Dose Caspofungin Combined with Voriconazole To Treat Central Nervous System Aspergillosis: Substantial Penetration of Caspofungin into Cerebrospinal Fluid

LETTER

Central nervous system (CNS) aspergillosis is a difficult-to-treat infection which remains associated with a severe outcome (1, 2). Among antifungals, voriconazole readily penetrates into both the cerebrospinal fluid (CSF) and brain abscesses and has shown encouraging efficacy in human CNS fungal infections (3). Although combination therapy in invasive aspergillosis is controversial, some authors advocate its use in the most severe forms of the disease, including CNS infection (4). It is commonly admitted that penetration of echinocandins into the CNS is poor (5), but human studies are scarce (6). Moreover, combination therapy to treat severe CNS aspergillosis could be an interesting therapeutic option. We report on the case of an immunocompromised patient with CNS aspergillosis successfully treated with a combination of voriconazole and a high dose of caspofungin. Substantial levels of caspofungin were found into the CSF.

Four months after receiving an orthotopic heart transplant, a 42-year-old man weighing 60 kg was admitted to the intensive care unit because of severe sepsis, altered consciousness, and left-sided hemiparesis. Brain computed tomography (CT) and magnetic resonance imaging (MRI) studies revealed a right cerebellar abscess and a left occipital abscess associated with signs of ventriculitis. Thorax CT showed diffuse nodular infiltrates. Renal failure present at admission required renal replacement therapy. The characteristics of CSF at admission (day 1 [D1]) are indicated in Table 1. At D1, galactomannan antigen titers (Platelia Aspergillus enzyme immunoassay; Bio-Rad Laboratory) in blood and CSF were 3 and 6.5, respectively, while PCR (7) was positive for A. fumigatus in both blood and CSF. Samples from bronchoalveolar lavage fluid (BAL fluid) grew A. fumigatus. Bilirubin, liver enzymes, and prothrombin time were all within normal ranges. Two days after admission, the patient was started on a combination of intravenous (i.v.) voriconazole (6 mg/kg body weight twice daily [b.i.d.] as a loading dose followed by 4 mg/kg b.i.d.) and i.v. caspofungin (280 mg as a loading dose followed by 140 mg/day). MICs of voriconazole and caspofungin, determined with Etest strips (AB Biodisk, Solna, Sweden), were 0.05 and 0.09 for the strain isolated from BAL fluid, respectively. Serum and CSF concentrations of antifungals were measured using high-performance liquid chromatography (HPLC).

TABLE 1.

Caspofungin and voriconazole concentrations in cerebrospinal fluid and serum at different time points

Parametera	Day 1	Day 5	Day 11	Day 17b	Day 18b
CSF parameter
WBC/μl	790	680	400	370	70
PMN (%)	80	80	75	60	60
Protein level (g/liter)	2.07	1.74	2.3	2.1	1.75
Glucose level (mmol/liter)	5.2	3.0	3.2	1.6
Serum caspofungin concn (mg/liter)
Trough		10	18.5	16.2	8.2
2 h and 8 h after administration				30.7 and 17.6
CSF caspofungin concn (mg/liter)
Trough		8	3.5	1	1
2 h and 8 h after administration				1.1 and 1
Serum voriconazole concn (mg/liter)
Trough		8.5	5.2	2.3
2 h and 8 h after administration				4 and 2.9
CSF voriconazole concn (mg/liter)
Trough		3.7	3.4	0.9	1
2 h and 8 h after administration				1 and 1.1

^aCSF, cerebrospinal fluid; WBC, white blood cells; PMN, polymorphonuclear cells.

^bOn days 17 and 18, CSF samples were obtained from ventricular drainage.

The patient's neurological status remained poor, and external ventricular drainage was performed to relieve hydrocephalus on day 17. In addition, the patient underwent aspiration of the left occipital abscess. Caspofungin and voriconazole concentrations in cerebral pus were of 1.5 mg/liter and 1.6 mg/liter, respectively. Table 1 shows caspofungin and voriconazole concentrations in CSF and serum at different times. Notably, CSF concentrations of caspofungin were relatively high during the acute stage of the infection, well above the MIC for the pathogen. In contrast, these concentrations were lower in less-inflammatory CSF. These data are in line with data from experimental studies showing that brain caspofungin levels are higher in infected animals (8) and with numerous data showing that the penetration of hydrophilic anti-infective drugs increases with the intensity of blood-brain barrier inflammation. Moreover, in an animal model of CNS aspergillosis, caspofungin significantly reduced the fungal burden (9).

The patient received a 2-month course of voriconazole and caspofungin, followed by oral voriconazole alone for 6 months, without any side effects attributable to antifungals. We conclude that relatively high concentrations of caspofungin may be obtained in the CSF provided that high doses are administered.