Parenchymal infection (abscess/infarction)

These are, broadly, two radiological manifestations of Aspergillus infection of the central nervous system. These patterns are either one or more contrast enhancing abscesses with surrounding oedema or a hypodense 'infarction'. Usually cerebral aspergillosis is haematogenous in origin but it may follow neurosurgical procedures or be a due to direct extension from sinus, ear or ocular disease. Meningitis is rare and spinal cord involvement is barely reported.

Several factors mitigate against the successful treatment of cerebral aspergillosis. These are:


  • non specific clinical presentation
  • non specific radiological abnormalities
  • late diagnosis
  • cerebral aspergillosis is usually a manifestation of disseminated disease
  • relatively poor blood -brain barrier penetration of antifungal drugs
  • vascular occlusion by fungal infection, further reducing antifungal penetration.

Amphotericin B penetrates poorly into brain tissue. High doses of lipid-associated amphotericin B may be slightly better, but low doses do not penetrate brain tissue at all (Clark et al, 1991; Proffitt et al, 1991). Flucytosine penetrates well into the CSF and probably brain (Polak, 1979). Itraconazole is highly lipophilic and although its CSF penetration is poor, its brain tissue penetration is reasonable if serum concentrations are detectable. In animal models of disseminated Aspergillus infection, itraconazole was successful in reducing the tissue burden of Aspergillus in the brain (Patterson et al, 1993).

Immunocompromised patients

Cerebral aspergillosis in the immunocompromised host is often not diagnosed until autopsy, and carries a mortality of about 99% (Denning, 1996; Pagano et al, 1996). A small number of reports have documented improvement and resolution of cerebral aspergillosis in immunocompromised patients in this setting (Burton et al, 1972; Henze et al, 1982; Kwong et al, 1987; Fu et al, 1988; Boriani et al, 1989; Goodman et al, 1989; Epstein et al, 1991; Green et al, 1991; Kloss et al, 1991; Kreisel et al, 1991; Camarata et al, 1992; Avet-Loiseau et al, 1994; Polo et al, 1992; Gelpi et al, 1994; Coleman et al, 1995; Sanchez et al, 1995; Mrowka et al, 1997; Khoury et al, 1997; Mahlknecht et al, 1997; Schwartz et al, 1997). Reduction in iatrogenic immunosuppression or recovery from neutropenia is an important component of success, if possible. This has included sacrificing renal allografts.

A wide variety of systemic antifungal agents were given in the successfully treated patients. Success was most commonly related to a combination of amphotericin B in combination with either or both of 5-flucytosine and rifampicin, high doses of itraconazole, moderate to high doses of liposomal amphotericin B or (latterly) voriconazole (unlicensed). Rifampicin is problematic as it induces the metabolism of cyclosporine, tracolimus, itraconazole and voriconazole and is not recommended.

Aside from early diagnosis, the most successful regimens have been:

  • amphotericin B (lmg/kg/d) with flucytosine (100-150mg/kg/d adjusted for renal function)
  • lipid associated amphotericin B Amphocil/Amphotec, Abelcet or AmBisome (=5mg/kg/d) preferably with flucytosine
  • itraconazole (800mg daily in divided doses)
  • voriconazole (6mg/kg initially then 3-4mg/kg every 12 hrs).

There is no objective way to select between these regimens at present, other than availability. Some clinicians combine itraconazole with an amphotericin B preparation and flucytosine. There is no place for local (e.g. intrathecal) amphotericin B therapy.


Many successfully treated patients have had open or stereotactic drainage of their cerebral lesions and in some cases total removal of a single abscess was achieved. It is not clear whether surgical debridement plays a part in response for several reasons: 1)a positive diagnosis of cerebral aspergillosis tends to promote more intensive and appropriate therapy: 2) patients who undergo biopsy may be fitter and more likely to survive and 3) these patients tend to have abscesses rather than infarction alone which is probably a manifestation of cerebral aspergillosis with a slighly better outcome. For these reasons a clear recommendation for or against radical surgery cannot be made, and in any case the site of disease has a huge impact on surgical feasibility and likely subsequent neurological deficits.

Non- immunocompromised patients

In the non-immunocompromised patient cerebral aspergillosis is rare. It is most commonly related to direct extension of disease from the ear or from the sinuses. It may be a post operative infection. Sometimes there is an extensive meningitic component with only a little parenchymal disease. It has a better prognosis than disease in the immunocompromised patient with a 33% response to amphotericin B therapy (Denning, 1996).

Successful management has almost always involved systemic antifungal therapy. The exception is a tiny number of cases in which a single mass lesion is completely removed surgically with the expectation that it would be a tumour and the diagnosis was subsequently proven. In these cases, the patients apparently did well without antifungal therapy but the follow up period was often short. The antifungal regimen recommended is the same for immunocompromised patients. Therapy should be continued for months or years until there has been no further reduction in the size of lesions for at least 3 months.

Treatment of postneurosurgical cerebral aspergillosis or meningitis entails removal of hardware if present and systemic antifungal therapy (Feely 1977; Morioka 1990; Kim 1993; Darrasjoly 1996). Intrathecal amphotericin has been utilized in few cases (Morioka 1990; Darrasjoly 1996) but is not recommended because the complications of intrathecal amphotericin B are formidable and the likely benefit small.

Prepared by:

T.F. Patterson
Department of Medicine/Infectious Diseases
University of Texas Health Science Center at San Antonio
7703 Floyd Curl Drive
Mail Code 7881
San Antonio
TX 78229-3900 USA.

David W. Denning FRCP FRCPath FIDSA FMedSci
Professor of Medicine and Medical Mycology
Director, National Aspergillosis Centre
Education and Research Centre
University Hospital of South Manchester (Wythenshawe Hospital)
Southmoor Road
Manchester M23 9LT UK