,NTF.RNATIOIALJO~RRAI.

OF

Antimicrobial Agents
ELSEVIER
International Journal of Antimicrobial Agents 6 (1996) 161-168

Aspergillosis: diagnosis and treatment

David W. Denning*
Department of Infecdons Diseases and Tropical Medicine. North Manchester General Hospital, Manchester, M8 6RB, UK, and University of Manchester, Department of Medicine, Hope Hospital, Salford, Manchester, M6 8HD, UK

Accepted 13 November 1995

Abstract The incidence of invasive aspergillosis is increasing rapidly in the developed world with two Aspergillus spp., A. jiumigatus and A. Jlavus, causing the majority of infections (85-90% and 5SlO%, respectively). The major risk factors are profound neutropenia (5 1000 x lo6 cells/L), prolonged neutropenia, neutrophil function deficits, and corticosteroid therapy. Useful diagnostic techniques include sputum culture, CT scan, bronchoscopy with microscopy and culture, percutaneous lung biopsy, open lung biopsy and serology. Invasive aspergillosis has an almost 100% mortality rate if untreated. Amphotericin B is the usual first-line therapy although it is associated with a high failure rate. Itraconazole (2400 mg daily) is a useful alternative and surgical resection may be life saving in some cases. The efficacy of the initial therapy is critical for improving mortality rates.
Keywords:

Aspergillosis;

Diagnosis;

Treatment

1. Introduction

entation is not discussed, primarily because it is often nonspecific.

Aspergillus produces a number of different diseases, in the normal and the immunocompromised host, including allergic disease, saprophytic disease, superficial infections and invasive infections such as pulmonary aspergillosis. Invasive aspergillosis, as an opportunistic infection, was first described in 1953. It has become an increasingly common cause of opportunistic fungal infections. The incidence of invasive aspergillosis is increasing rapidly in the western world while that of aspergilloma is decreasing. Aspergillus is generally regarded as an opportunistic pathogen although there are many instances of invasive aspergillosis in non-immunocompromised patients. In this article, diagnostic and therapeutic approaches are reviewed. :[n this article the clinical presboth

2. Incidence In a cancer patient autopsy study conducted in 12 centres in North America and Japan, Bodey et al. [l] showed that 30% of all fungal infections in cancer patients were due to Aspergihs. In a study in Frankfurt by Groll et al. [2] over 10 000 autopsies (a 75% autopsy rate) performed between 1978 and 1992 were analysed. During this period systemic mycoses detected at autopsy rose from 1.5% to 6% (percentage of all autopsies) and aspergillosis as the cause of death rose from 17% to 60%. There has been a significant shift in the epidemiology of fungal infections and aspergillosis is now a very difficult therapeutic problem for many hospital practitioners.

3. Risk factors
*Correspondence to: Department of Infections Diseases and Tropical Medicine, North Manchester General Hospital, Manchester MS 6RB, UK. Tel.: (+44-161) 720 2734; Fax: (+44-161) 720 2732.
0924-8579/96/$32.00 0 1996 Elsevier Science B.V. All rights reserved 0924-8579(95)00042-9

The most important losis are, in decreasing

risk factors for invasive aspergilorder of importance, profound

SSDI

162

D. K Denningllnternational Journal of Antimicrobial Agents 6 (1996) 161-168

neutropenia (C 1000 x lo6 cells/L), prolonged neutropenia [3], neutrophil function deficits [4] (usually combined with macrophage or other cellular immune deficits) as in chronic granulomatous disease and AIDS [5], corticosteroid therapy, graft-versus-host disease [6] and/ or rejection in transplantation [7] (which may or may not be an independent variable as it reflects additional immunosuppression). Less important risk factors appear to be diabetes mellitus, alcohol excess, influenza, prematurity and exposure to Aspergillus in large quantities, for example from living in a rural environment. Patients with chronic respiratory disease requiring treatment with corticosteroids are also at higher than average risk of invasive aspergillosis. Certain groups of patients colonized with Aspergillus are at increased risk, particularly patients with nasal colonization prior to chemotherapy for acute leukaemia [8] and patients with colonization of the tracheobronchial tree prior to lung transplantation. Corticosteroid therapy reduces pulmonary macrophage killing of Aspergillus conidia and neutrophil killing of Aspergillus hyphae [7,9]. The detrimental effect of corticosteroids on neutrophil function can be reversed in vitro by granulocyte colony stimulating factor and gamma interferon [IO]. In addition, corticosteroids directly increase the growth rate of Aspergillus fumigatus and A. Jlavus. Under the influence of hydrocortisone Aspergillus hyphae extend at the astonishing rate of 1-2 cm/h [l 11.

4.1. Culture

4. Diagnosis

In highly immunocompromised patients, particularly neutropenic patients, a CT scan of the chest should be the first definitive investigation when the possibility of invasive pulmonary aspergillosis is suspected. A CT scan is more sensitive than a chest X-ray but not as easy to perform. Some radiographic lesions are virtually diagnostic in this context. Typical radiological signs of invasive aspergillosis include the halo sign (low attenuation area around an infiltrate) or air crescent lesion, [12,13] a pleural-based, sharply angular or triangular lesion or a pleural-based lesion with a pneumothorax. Pneumothorax is not common but usually indicates Aspergilhs if in association with a consolidation or a cavity, although occasionally it may be due to Mucorales. In heart transplantation particularly, nodules in the lung may occur months after transplantation. About half of these are due to Aspergillus and the other half are due to Nocardia [14]. Cavitation may not occur and they are often asymptomatic and found on routine chest X-rays or a CT scan. A percutaneous lung biopsy can be performed to make a specific diagnosis.

The two major Aspergillus species which cause disease are Aspergillus fumigatus (about 85-90% of all Aspergillus infections) and Aspergillusj?avus (5-10%). Two other species, A. terreus and A. niger, together account for approximately 5% of infections whilst many other rare species have been reported as causing disease. The significance of Aspergillus cultured from sputum can be difficult to assess [ 151 since positive sputum cultures of Aspergillus occur in l-16% of healthy people. Chronic lung disease and living in a rural environment increase the frequency of colonization. In most patients with Aspergillus in the sputum no disease is present but in an immunocompromised patient a positive sputum culture is highly significant. Thus the positive predictive value for sputum culture is high but the negative predictive value is low in most immunocompromised patients with features suggestive of invasive aspergillosis. In solid organ transplant patients, Aspergillus tracheobronchitis has been described with no symptoms and a normal chest X-ray, thus a positive Aspergillus culture should prompt immediate bronchoscopy. Invasive aspergillosis occurs in about 4% of patients with AIDS but recovery of Aspergillus from respiratory sites in such patients often reflects colonization [16]. Studies examining the sensitivity and specificity of combinations of positive findings in immunocompromised patients, such as a positive respiratory culture and an abnormal chest radiograph, are urgently required. Positive cultures of nasal swabs for Aspergillus are uncommon but useful in leukaemia. Isolation of Aspergillus from the nose has a 90% correlation with invasive pulmonary aspergillosis during neutropenia in leukaemic patients. A positive nasal culture prior to chemotherapy for leukaemia carries a high risk of invasive pulmonary aspergillosis [8]. However, surveillance swabs have a low yield. A positive nasal swab for Aspergillus in a context other than leukaemia may or may not be significant unless there is overt sinus or nasal disease. Thirty cases of Aspergillus fungaemia have been reported in the literature and we have seen a further two cases [17]. Approximately a third of cases were related to cardiac surgery where patients receiving valve replacements acquired Aspergillus endocarditis post-operatively as a result of contaminated operating room air. About another third occurred during neutropenia, this is a particularly high risk group and the overall mortality was 56%. As expected all the patients who did not receive therapy died. In a recent study, Martin0 et al. [18] showed that Aspergillus fungaemia comprised 8% of all fungaemia cases in leukaemia. A potential problem is pseudofungaemia. The patient is not infected but false positive blood cultures are obtained because of contamination in the laboratory. Aspergillus pseudofungaemia is

D. W Denningllnternational

Journal of Antimicrobial Agents 6 (1996)

161-168

163

more likely if the patient is not in a typical host group and/or the clinical course is not compatible. However, it should be assumed that the patient is infected and treated as such until confirmed otherwise. 4.2. Bronchoscopy Fibreoptic bronchoscopy is now an essential procedure for the diagnosis of pulmonary aspergillosis including Aspergillus airways disease and diffuse (e.g. nonfocal) invasive pulmonary aspergillosis. In leukaemia and bone marrow transplantation patients the most useful procedure is bronchoalveolar lavage (BAL) or a bronchial wash [6,19-221. Samples should be processed for both culture and microscopy (or cytology). Transbronchial biopsy is a simple and extremely productive procedure if airways disease is present, but it has limited value for invasive pulmonary aspergillosis. Brushings contribute little to the diagnosis. The positive yield for A:;pergiEIusvaries depending on the pattern of disease and the host group. In leukaemia and bone marrow transplantation, cytology and culture of BAL fluid yield the diagnosis in approximately O-67% of patients (6,19-22). In solid organ transplant patients, the yield is probably lower, except in lung transplant patients in whom airways disease is probably more frequent [23]. Radiologically, diffuse disease is more likely to yield a positive result than focal disease (100% versus 0% in one study of bone marrow transplant patients) [6]. The utility of BAL for diagnosing invasive aspergillosis in AIDS varies from 0% to 100% [24,25] but in the study in which it was lOO%, a positive BAL was the criterion for inclusion in the study. Transbronchial biopsy often gives false negative results in cases of invasive pulmonary aspergillosis. This may be due to sampling error and in particular the depth of the biopsy. Bronchoscopy with bronchial biopsy and culture is the only means. of making the diagnosis of Aspergillus airways disease diagnosis antemortem. Biopsy of loose material in the tracheal lumen, plaques or ulcers will often reveal necrotic cartilage invaded by hyphae. 4.3. Percutaneous lung biopsy Percutaneous needle biopsy is indicated in patients with focal disease, particularly in the periphery of the lung, as these lesions are not accessible to the bronchoscope and bronchoscopy is often negative. The platelet count should be above 30 :< 109/L or platelet transfusions should be given. Significant bullous or emphysematous lung disease in the area of interest is a contraindication. Specimens should be processed for culture and cytology or histology. The yield with respect to Aspergihs has not been accurately assessed in prospective studies but it

yielded the diagnosis in 5 AIDS patients with invasive aspergillosis in whom BAL was negative [24]. A pneumothorax requiring percutaneous drainage occurred after about 2% of procedures. 4.4. Open lung biopsy Open lung biopsy is occasionally used to establish the diagnosis of invasive pulmonary aspergillosis when the diagnosis is thought to be a malignancy or other infection such as Pneumocystis. No studies of the utility of open lung biopsy for the diagnosis of invasive aspergillosis have so far been performed and its role as a diagnostic technique is presently unclear. However, when positive it yields unequivocal evidence of invasive aspergillosis. 4.5. Serology The antibody response to Aspergillus infection has been well characterised in allergic bronchopulmonary aspergillosis and aspergilloma. Over 95% of patients with an aspergilloma have detectable precipitating IgG antibodies and some have IgM antibodies as well. Successful surgical removal of an aspergilloma results in a subsequent decrease of Aspergillus antibody to low or undetectable levels. Aspergillus antibody is occasionally detectable at the time of diagnosis of invasive aspergillosis in neutropenic patients but may be positive in patients following solid organ transplantation depending on the test used. The frequency of antibody positivity in other host groups, such as chronic granulomatous disease or AIDS is unknown. Aspergillus antigen tests have only been carefully examined in leukaemia and bone marrow transplant patients. Reproducibility of the galactomannan latex test in serum is less than ideal with a sensitivity of approximately 30%. Several samples may need to be tested before positive results are obtained since antigenaemia is often shortlived. Antigen positivity increases closer to death with progressive invasive aspergillosis. Unfortunately, many studies on antigen detection for invasive aspergillosis have not accurately described the temporal relationship of a positive antigen test and the early symptoms of invasive aspergillosis and initiation of treatment. 4.6. Sinus aspergillosis Sinus aspergillosis occurs in leukemic, bone marrow transplant and relatively non-immunocompromised patients. To diagnose invasive Aspergihs sinusitis a biopsy is required. Usually part of the nose, which is accessible externally, is affected in addition to the sinuses. Any nasal discharge or any crusted lesions should be cultured.

164

D. W Denningllnternational

Journal of Antimicrobial Agents 6 (1996) 161-168

4.7. Cerebral aspergillosis

Cerebral aspergillosis is usually confirmed at death. Aspiration of a suspicious lesion in the brain will usually yield Aspergillus on culture or the presence of hyphae on histological examination. Unfortunately, the histological appearance alone may not be distinctive enough to diagnose Aspergillus as the hyphae are similar to those of many other pathogens including Pseudallescheria boydii.

5. Treatment

The rate of progression of invasive aspergillosis varies widely. In liver and bone marrow transplant patients and patients with profound neutropenia the course of disease from first clinical or radiological abnormality to death is usually lo-14 days. The diagnosis is often difficult to establish and several days can elapse between consideration of the diagnosis and partial or complete confirmation. A critical window of opportunity may be missed if treatment is not started early, with fatal consequences for the patient. The exact timing of when treatment should be initiated for invasive aspergillosis is debatable.

Table 1 Criteria for initiation of therapy for invasive aspergillosis or increasing the dose of amphotericin B to at least 1 mg/kg/day in neutropenia (< 1000 x 106/L) including aplastic anemia [23]
l

l l

Isolation of Aspergillus from any site in&ding nose swab, blood, BAL etc. New pulmonary infiltrates on chest X-ray Infiltrates on CT scan of chest showing characteristic features, e.g. halo/crescent sign (i) (ii) pleural based sharply angulated lesion (iii) pleural based lesion with pneumothorax Persistent fevers (> 7 days) with any localising clinical features and not responding to antibiotics, e.g. chest pain (i) (ii) dry cough (iii) facial/sinus pain (iv) epistaxis hoarseness I:!, new skin lesions consistent with aspergillosis Any sudden intracranial event including stroke or fit with or without fever

Antifungal treatment should be started in neutropenic patients when Aspergillus is isolated from any site. The appropriate criteria for the initiation of therapy in neutropenic patients are shown in Table 1 [26]. Hoarseness, which is well described in children and may also be due to Candida, is a feature of Aspergillus epiglottitis. Skin lesions, which are relatively common in candidaemia, are also found in invasive aspergillosis. They are red initially and not painful but then turn black rapidly and are like ecthyma gangrenosum of Pseudomonas. They may also be confused with Fusarium skin lesions. Any sudden intracranial event, such as a stroke or a fit, with or without fever, is also an indication to treat. In most patients, empiric amphotericin B therapy will have been given because of persistent fever after about 5 days of broad-spectrum antibiotics. If a dose of less than 1 mg/kg/day of conventional amphotericin B has been used the dose should be increased to l-l .25 mg/kg/ day. If, however, empiric amphotericin B therapy is started at 1 mg/kg/day, then a decision to switch from this treatment to alternative therapies should be made using the criteria in Table 1. The criteria for initiation of therapy are slightly different in solid organ or allogenic bone marrow transplantation (Table 2) because there is a higher frequency of Aspergillus tracheobronchitis in this population and a much greater differential diagnosis for cases of pneumonitis [23]. If pulmonary disease occurs months after transplantation it is more likely to present as a nodular disease which may indicate Aspergillus, Nocardia or the development of lymphoma [ 141. Immediately following liver transplantation, interpretation of the chest X-ray is difficult because of changes following surgery. In addition, solid organ transplantation patients are more likely to get Aspergillus in operative sites (e.g. mediastinitis, wound drainage, wound infection) which does not occur in leukaemia patients. As with neutropenic patients, any sudden intracranial event, such as a stroke or a fit with or without fever, in transplantation patients is an indication to start treatment. In transplant patients, however, there is a larger range of potential diagnoses such as Listeria, meningitis, cryptococcal meningitis, progressive multifocal leukoencephalopathy (PML) or cerebral infarction (especially liver transplant recipients). Invasive aspergillosis occurs during or immediately after the short period of aplasia following allogenic bone marrow transplantation, or weeks or months after transplantation in
Table 3 Criteria for initiation
l

Table 2 Criteria for initiation of therapy for invasive aspergillosis in solid organ transplantation or allogeneic bone marrow transplantation [23]
l

l l

Radiological infiltrate with isolation of Aspergihs from respiratory secretions, BAL or blood Evidence of ulcers/bronchitis on bronchoscopy with hyphae visualised in BAL Histological or cytological evidence of hyphae from any tissue Any sudden intracranial event including stroke or fit with or without fever

of therapy

for invasive aspergillosis

in AIDS [23]

Respiratory symptoms, abnormal chest X-ray or bronchoscopy and isolation of Aspergihs or visualisation of hyphae in BAL Cerebral lesions unresponsive to therapy for toxoplasmosis or a single enhancing lesion consistent with aspergillosis Histological or cytological evidence of hyphae from any site (culture not obtained or pending)

D. W! Denningllnternational

Journal of Antimicrobial Agents 6 (1996)

161-168

165

graft-versus-host disease. Disease progression in these patients can be very rapid. In AIDS patients invasive aspergillosis tends to be a relatively subacute disease and it is usually possible to make a positive diagnosis without having to start empiric therapy (Table 3) [26] The major exception, however, is intracranial disease. Many patients and their physicians are unwilling to undertake a brain biopsy or aspiration and therefore empirical therapy may be appropriate if there is a single lesion or small number of lesions, without the characteristic features of lymphoma (ventricular extension) or progressive multifocal encephalopathy, not responding to anti-Toxoplirsma therapy. 5.1. Antifungal therapy The standard therapeutic agent for the treatment of invasive aspergillosis is conventional amphotericin B. The success rate of amphotericin B therapy varies substantially between different patient populations but is generally poor. There are two major considerations in the initial choice of therapy: (1) Can the patient take oral medication and is he/she likely to absorb it?
Table 4 First and second-line Agent Amphotericin Itraconazole B

(2) Is the patient taking cyclosporin or is renal dysfunction already present? If the patient can take oral therapy and is not taking drugs that induce the metabolism of P450 enzymes (e.g. rifampin, phenytoin, carbamazepine, phenobarbitone) then itraconazole is a reasonable first choice (Table 4). There is extensive experience with this agent in the treatment of invasive aspergillosis [27-321. Other factors that might reduce absorption and therefore efficacy include HZ-antagonists, omeprazole, patients with intestinal problems (such as graft-versus-host disease) and patients with AIDS. The dose of itraconazole should be 200 mg tds for 4 days followed by 200 mg twice daily, all doses taken with food. Itraconazole serum concentrations should be checked after 5-10 days. Limited data suggests that higher doses, e.g. 800 mg/day, of itraconazole are more efficacious for cerebral aspergillosis (unpublished data). The new itraconazole oral solution will improve bioavailability. The relative toxicity of itraconazole compared with amphotericin B is mild and different (Table 5). If the patient is not suitable for oral therapy then intravenous therapy is required. Conventional amphotericin B should be used at a minimum dose of 0.8-1.0

therapy

for invasive

aspergillosis Initial dose 0.8-l .25 mg/kg/d (IV) 200 mg tds for 4 d then 200 mg bd Comments Considered first-line therapy but high failure rate. Significant interaction with cyclosporin Useful if patient eating and not on P450 inducers. Significant interaction with cyclosporin. Levels should be measured to ensure adequate absorption Less toxic than amphotericin B Less toxic than amphotericin Less toxic than amphotericin B B

Amphotericin Liposomal Amphotericin

B colloidal amphotericin

dispersion B

(PO) 4-6 mg/kg/d (IV) 45 mg/kg/d (IV) 5 mg/kg/d (IV)

B lipid complex

None

of these agents

or regimens

have been compared

in controlled

trials, but all have been shown

to be partially

effective

Table 5 Major side effects of itraconazole

and amphotericin Itraconazole

B Amphotericin B Lipid-associated amphotericin B + + or + +a ? -

Nausea Rash Infusion related fever, chills Fluid retention Abnormal liver function tests Hypokalaemia Hypomagnesaemia Nephrotoxicity Anaemia

+ + + + + +b -

+ + +a + +c +

a Can be ameliorated by concurrent amiloride use. b Only in cyclosporin-treated patients, but can be avoided by halving cyclosporin Worsened by salt depletion, cyclosporin and other nephrotoxic agents.

dose when itraconazole

started.

166

D. CI? DenninglInternationaI

Journal of Antimicrobiul

Agents 6 (1996)

161-168

mg/kg/day, regardless of renal dysfunction, except in neutropenic patients in whom l-l.25 mg/kg/day is appropriate (Table 4). If renal dysfunction is or is likely to be, as in cyclosporin-treated patients, a major problem then one of the new preparations of amphotericin B in doses of 46 mg/kg/day may be appropriate (Table 4). These dosage levels should be continued for at least two weeks or until a therapeutic response has been obtained clinically and radiologically. The apparent severity of the illness should not be used as a guide for choosing oral itraconazole or intravenous amphotericin B. Therapeutic responses have been obtained in profoundly neutropenic patients with extensive disease using oral itraconazole alone when amphotericin B therapy is often ineffective [27,32,33]. There is debate about whether the combination of amphotericin B and itraconazole is appropriate. There are theoretical reasons for antagonism since itraconazole inhibits the production of ergosterol to which amphotericin B binds as part of its mode of action. There is some limited animal mode1 work which suggests that antagonism may be a real phenomenon although the magnitude of this effect is difficult to ascertain [34]. The present recommendation is that treatment should be with one or other drug. If itraconazole is the drug of choice but there is concern about its absorption, the initial use of both agents until adequate itraconazole levels have been achieved and absorption confirmed is a reasonable option, 5.2. Switching
therapy

munocompromised and until there has been either complete or near complete resolution of disease. In neutropenic patients this means continuing treatment until the neutrophil count is above 1000 x 106/L and until there is resolution of disease radiographically. This may take weeks or months. Eradication of disease may require surgery. In patients whose immune status improves gradually over a long period of time (e.g. a solid organ transplant patient treated for many months) therapy should be continued until there is a complete radiological and clinical response. AIDS patients and patients who have received an allogenic bone marrow transplant should probably receive treatment for years if they can tolerate and absorb itraconazole. If unable to take itraconazole, long-term intermittent amphotericin B is the only alternative, apart from surgical resection. 5.4. Surgery In invasive pulmonary aspergillosis surgery may be life saving [35]. Most clinicians reserve lung resection for patients with persistent lung shadows who have to undergo subsequent bone marrow transplantation or more aggressive chemotherapy, those with significant haemoptysis [36] or those with lesions impinging on the great vessels or major airways in whom fatal haemoptysis is frequent. Erosion of invasive aspergillosis into the pulmonary vessels is not uncommon and is a rapidly fatal complication that can be averted by surgery. In patients with invasive sinus aspergillosis, a biopsy of the affected area is necessary for diagnosis. Surgical debridement is useful for treatment but should not be performed during neutropenia as haemorrhage and other operative complications are frequent [35]. In cerebral aspergillosis, because most of the lesions are deep in the brain, surgery only has a role for diagnostic aspiration or biopsy. Surgery is an essential part of the management of patients with Aspergillus endophthalmitis, endocarditis and possibly osteomyelitis.

Unfortunately, a large percentage of patients fail to respond to the first therapy which is administered. In patients with rapidly progressive disease, failure of the first regimen means a fatal outcome, therefore a good initial choice is critical. However, for patients with more slowly progressive disease the evaluation of the response after lo-20 days of therapy can be made using clinical, radiological and other criteria to decide whether therapy should be changed. The less immunocompromised the patient, the more time there is to evaluate the response. Many patients will respond to an alternative therapy having failed the primary therapy, thus all patients should be given more than one form of therapy if a suboptimal response is obtained. Many doctors and patients prefer to switch from initial intravenous amphotericin B as primary therapy to oral itraconazole as follow-on and maintenance therapy. This is a useful strategy provided that the patient has no problems with itraconazole absorption and they respond to amphotericin B initially. 5.3. Duration of therapy> Patients should be treated for as long as they are im-

6. Conclusions Invasive aspergillosis carries a nearly 100% mortality if untreated. The only exceptions are patients whose immunocompromising factors are removed (e.g. permanent cessation of corticosteroids). The attributable mortality of invasive aspergillosis is from 50-100% in collected series of treated patients [35]. Survivors have all received at least 2 weeks of therapy. There are substantial variations between host groups and within host groups. The response rate to amphotericin B deoxycholate varies according to the type of patient and the length of time that it is possible to administer treatment. In bone marrow transplant patients the response rate is only around 5-

D. W Denningllnternational

Journal of Antimicrobial Agents 6 (1996) 161-168

167

10%. In leukemic neutropenic patients it is around 60%, but many of these patients die before that, so the realistic response rate is 2530%. In aplastic anaemia, only about 1 in 10 patients respond. Heart transplant and renal transplant patients have better response rates (e.g. ~50%) but liver transplant patients have poor response rates of 5510%. Response rates with itraconazole are similar. In leukaemia, achieving complete remission is critical to patient survival [37]. Cerebral aspergillosis still has a mortality exceeding 99% in the immunocompromised patient. Many factors influence the outcome of invasive aspergillosis. Those which lead to a better outcome include resolution of neutropenia, leukaemia in remission, heart or kidney transplantation (as opposed to liver or bone marrow transplantation), focal pulmonary disease, early initiation of therapy or change of therapy if the initial choice is unsuccessful and appropriate surgery. Clearly, new agents for the treatment of invasive aspergillosis are urgently required. The role of G-CSF and other cytokines as adjunctive therapy is unclear at present. The efficacy of the primary therapy chosen is a critical factor in improving mortality rates.

References
G, Bueltmann B, Duguid W et al. Fungal infections in cancer patients: an internaticnal autopsy survey. Eur J Clin Microbiol Infect Dis 1992; 11:99-l 09. PI Groll A, Shah PM, Mentzel CH, Schneider M. Changing pattern of invasive mycoses at autopsy. Abstract J109, Interscience Conference on Antimicrobial Agents and Chemotherapy, Orlando, FL. American Society for Microbiology, October 1994. [31Gerson SL, Talbot GH, Hurwitz S, Strom BL, Lusk EJ, Cassileth PA. Prolonged granulocytopenia: the major risk factor for invasive pulmonary aspergillosis in patients with acute leukemia. Ann Intern Med 1984; 100:345-35 1. [41Mouy R, Fischer A, Vilmer E, Seger R, Griscelli C. Incidence, severity, and prevention of infections in chronic granulomatous disease. J Pediatr 1989;114:555-560. [51Khoo S, Denning DW. Aspergillus infection in the acquired immune deficiency syndrome. (Clin Infect Dis 1994; 19(suppl 1):541548. PI McWhinney PHM, Kibbler CC, Hamon MD et al. Progress in the diagnosis and management of aspergillosis in bone marrow transplantation: thirteen years experience. Clin Infect Dis 1993;17:397404. [7] Gustafson TL, Schaffner W, Lavely GB, Stratton CW, Johnson HK. Hutcheson RH. Invasive aspergillosis in renal transplant recipients. Correlation with corticosteroid therapy. J Infect Dis 1983;148:230-234. [8] Martin0 P, Raccah R, Gentile G et al. Aspergillus colonization of the nose and pulmonary aspergillosis in neutropenic patients: a retrospective study. Haematologica 1989;74:263-265. [9] Schaffner A, Douglas H, Braude A. Selective protection against conidia by mononuclear and against mycelia by polymorphonuclear phagocytes in resistance to Aspergillus. J Clin Invest 1982;69:617631. [lo] Roilides E, Unlig K, Venzon D, Pizza PA, Walsh TJ. Prevention Bodey 111

of corticosteroid-induced suppression of human polymorphonuclear leukocyte-induced damage of Aspergih fumigatus hyphae by granulocyte colony-stimulating factor and gamma interferon. Infect Immun 1993;61:48704877. [l l] Ng TTC. Robson GD, Denning DW. Hydrocortisone-enhanced growth of Aspergillus spp.: implications for pathogenesis. Microbiology 1994;140:2475-2480. [12] Midiri M, Lo Casto A, Cascio A, Scarlata F, De Maria M. Invasive pulmonary aspergillosis: CT findings. Med J Infect Parasit Dis 1994$X:59-62. [13] Taccone A, Occhi M, Garaventa A, Manfredini L. Viscoli C. CT of invasive pulmonary aspergillosis in children with cancer. Pediatr Radio1 1993;23: 177-l 80. LB, Schulman LL. Austin JHM. Lung nodules and v41 Haramati masses after cardiac transplantation. Radiology 1993;188;491l 497. 1151Barnes AJ, Denning DW. Aspergillus - significance as a pathogen. Rev Med Microbial 1993;4:176-180. I161Purse11 KJ, Telzak EE, Armstrong D. Aspergillus species colonization and invasive disease in patients with AIDS. Clin Infect Dis 1992;14:141-148. 1171Duthie R, Denning DW. Aspergillus fungemia. Two cases and review. Clin Infect Dis. 1995;20:5988605. [I81Martin0 P, Girmenia C, Micozzi A et al. Fungemia in patients with leukemia. Am J Med Sci 1993;306:2255232. [I91Albelda SM, Talbot GH, Gerson SL, Miller WT, Cassileth PA. Role of fibreoptic bronchoscopy in the diagnosis of invasive pulmonary aspergillosis in patients with acute leukemia. Am J Med 1984;76:1027-1034. PO1 Kahn FW, Jones JM. England DM et al. The role of bronchoalveolar lavage in the diagnosis of invasive pulmonary aspergillosis. Am J Clin Path01 1986;86:518-523. Pll Levy H, Horak DA, Tegtmeier BR, Yokota SB, Forman SJ. The value of bronchoalveolar lavage and bronchial washings in the diagnosis of invasive pulmonary aspergillosis. Resp Med 1992;86:2433248. WI Saito H, Anaissie EJ, Morice RC, Dekmezian R. Bodey GP. Bronchoalveolar lavage in the diagnosis of pulmonary infiltrates in patients with acute leukemia. Chest 1988;94:7455749. tra1231Kramer MR. Denning DW, Marshall SE et al. Ulcerative cheobronchitis following lung transplantation: a new form of invasive aspergillosis. Am Rev Resp Dis 1991;144:552-556. S. Scolaro M, Norris S. Edelstein D, v41 Denning DW. Follansbee Stevens DA. Pulmonary aspergillosis in AIDS. N Engl J Med 1991;324:654662. v51 Lortholary 0, Meyohas MC, DuPont B et al. Invasive aspergillosis in patients with acquired immunodeficiency syndrome: report of 33 cases. Am J Med 1993;95:1777187. of invasive aspergillosis. J Infect WI Denning DW. The treatment 1994;28 (suppl 1):25-33. I, Verhagen CAH, Fibbe WE. De 1271van t Wout JW. Novakova Pauw BE, Van der Meer JWM. The efficacy of itraconazole against systemic fungal infection in neutropenic patients: a randomised comparative study with amphotericin B. J Infect 1991; 22145552. G et al. The treatment P81 De Beule K, De Doncker P, Cauwenbergh of aspergillosis and aspergilloma with itraconazole. clinical results of an open international study (1982-1987). Mycoses 1988;31:476 485. 1291Kreisel W, Kochling G, Schilling CV et al. Therapy of invasive aspergillosis with itraconazole: improvement of therapeutic efficacy by early diagnosis. Mycoses 1991;34:3855394. [301Jennings TS, Hardin TC. Treatment of aspergillosis with itraconazole. Ann Pharmacother 1993;27:12061211. therapy in aspergillosis: study in 49 pa1311DuPont B. Itraconazole tients. J Am Acad Dermatol 1990;23:607-614.

168

D. W DenninglInternationai Journal of Antimicrobial Agents 6 (1996) 161-168 [35] Denning DW, Stevens DA. Antifungal and surgical treatment of invasive aspergillosis: review of 2121 published cases. Rev Infect Dis 1990;12:1147-1201. [36] Kibbler CC, Milkins SR, Bhamra A, Spiteri MA, Noone P, Prentice HG. Apparent pulmonary mycetoma following intensive aspergillosis in neutropenic patients. Thorax 1988;43:108-112. [37] Ribrag V, Dreyfus F, Venot A, Leblong V, Lanore JJ, Varet B. Prognostic factors of invasive pulmonary aspergillosis in leukemic patients. Leuk Lymph 1993;10:317-321.

[32] Denning DW, Lee JY, Hostetler JS et al. NIAID mycoses study group multicenter trial of oral itrazonazole therapy for invasive aspergillosis. Am J Med 1994;97:135-144. [33] Caillot D, Casasnovas 0, Couaillier JF et al. Improvement of invasive pulmonary aspergillosis prognosis in neutropenic patients by using early CT-scan and itraconazole. Abstract 528, 35th ICAAC, San Francisco, 1995. [34] Sugar AM. Use of amphotericin B with azole antifungal drugs: What are we doing? Antimicrob Agents Chemother 1995;39:19071912.