Cross-Reactivity of Fusarium spp. in the Aspergillus Galactomannan Enzyme-Linked Immunosorbent Assay

Abstract

Nine of 11 hematological patients with disseminated/deep-seated Fusarium infection tested at least twice for Aspergillus galactomannan (GM) had repeated positive results in the absence of Aspergillus isolation in culture. The centrifuged supernatants of 12 Fusarium isolates were tested by a GM enzyme-linked immunosorbent assay (EIA). All the isolates produced positive reactions when tested undiluted. These results show cross-reactivity of Fusarium spp. with Aspergillus GM that may constitute a drawback with respect to the specificity of the Platelia EIA.

TEXT

Fusarium is a large genus of hyaline filamentous molds that have emerged as pathogen of immunocompetent and immunocompromised hosts. While Fusarium infections, such as onychomycosis, peritonitis in association with peritoneal dialysis, and keratitis as a consequence of corneal trauma or contact lens wear, in immunocompetent patients tend to be superficial or locally invasive, disseminated Fusarium infection remains an uncommon and yet severe opportunistic infection among highly immunocompromised patients (9). The European Confederation of Medical Mycology (ECMM) recently launched a survey on this infection in Europe.

The treatment of Fusarium infection in immunocompromised hosts is a frustrating task due to the limited susceptibility of Fusarium species to antifungal agents. Most of the species have shown in vitro resistance to azoles, polyenes, and echinocandins; numerous experimental studies in animals carried out to evaluate different antifungal treatments have had disappointing results, and a favorable outcome was found in less than 50% of the cases (5, 7, 9, 14, 15, 19).

The detection of Aspergillus galactomannan (GM) using a sandwich enzyme-linked immunosorbent assay (EIA) is now included among the microbiological criteria of the revised European Organization for Research and Treatment of Cancer/Invasive Fungal Infections Cooperative Group and the National Institute of Allergy and Infectious Diseases Mycoses Study Group (EORTC/MSG) classification of invasive aspergillosis (2). The clinical sensitivity of GM EIA is variable, ranging from 29% to 100% according to the host group and the underlying pathological processes (13). A major problem is the occurrence of false-positive results related in some cases to cross-reactivity with other fungi. The epitope detected by the EB-A2 monoclonal antibody employed in the EIA is not exclusively present in Aspergillus species. This epitope is contained also in antigens of other molds, including Penicillium, Paecilomyces, Trichothecium, Geotrichum, and Myceliophthora, of dimorphic fungi, such as Blastomyces dermatitidis and Histoplasma capsulatum, and also of yeasts (1, 3, 4, 8, 20, 21). Cross-reactivity of Fusarium spp. in the GM assay is controversial and seems to be related to the method used and the tested species (1, 6, 17). While cross-reactivity of F. oxysporum was found using Pastorex Aspergillus antigen latex agglutination but not using the Platelia test (1, 6, 16), F. solani was reported to lack cross-reactivity in both the assays (1, 6, 16).

The aims of the present study were to check GM index results in patients with disseminated Fusarium infection included in the ECMM survey conducted in Italy by the Federazione Italiana di Micopatologia Umana e Animale (FIMUA) and to test the cross-reactivity of different Fusarium spp. in the GM EIA.

Aspergillus GM index results of patients with Fusarium disseminated or deep-seated infection notified during the ECMM-FIMUA survey were checked in the microbiological database.

A total of 12 Fusarium isolates—2 F. oxysporum, 3 F. proliferatum, 3 F. verticillioides, and 4 F. solani species complex (FSSC), namely, 1 F. falciforme and 3 F. sp. cucurbitae MPV—identified by sequencing the elongation factor 1α gene (10, 11, 12) were studied. Seven isolates were from the patients included in the study. The other five, namely, four isolates from blood and one from a skin biopsy specimen, were from patients not included in this study, and the GM test was not performed for these patients. One Aspergillus fumigatus isolate was used as a control.

The isolates were grown on Potato dextrose agar for 72 h at room temperature. A loop of biomass was transferred in 50 ml of Sabouraud broth and incubated in the dark at room temperature on a rotary shaker (20 rpm). After 72 h at room temperature, 15 ml of the supernatant was centrifuged at 2,000 rpm for 5 min. The centrifuged supernatants (undiluted and diluted 1:100) were tested twice by Platelia Aspergillus EIA (Bio-Rad, Marne La Coquette, France).

Eleven patients were tested at least twice for GM (Table 1). Eight with hematological malignancy had proven disseminated Fusarium infection, documented by isolation of the fungus from blood or by the presence of hyphae at histology of a skin biopsy specimen combined with growth of Fusarium in culture. Another three neutropenic patients had probable infections according to the EORTC/MSG criteria, as hyphae were seen by direct microscopy of respiratory tract samples or pleural fluid and Fusarium spp. were isolated in culture. Nine of these 11 patients had repeated positive GM test results, with an index ranging from 0.50 to 7.7 in the absence of isolation of Aspergillus spp. in culture of bronchial secretions or of other respiratory tract specimens (Table 1).

Table 1.

Serum GM index results for patients with disseminated/deep-seated Fusarium infection^a

Patient	Sex	Age (yr)	Predisposing factor(s)	Fusarium infecting isolate	Positive biological sample(s)	Serum GM index range
1	M	29	Allo-HSCT	F. oxysporum	Pleural fluid	0.69–0.90
2	M	63	Leukemia	F. oxysporum	Bronchial secretions	1.40–2.30
3	M	55	AML	F. proliferatum	Blood	0.89–0.86
4	F	61	Allo-HSCT	F. proliferatum	Blood	1.37–2.33
5	F	56	AML, allo-HSCT	F. proliferatum	Blood	0.7–2.15
6	M	8	Non-Hodgkin's lymphoma	F. proliferatum	Blood	0.53–7.7
7	M	19	ALL, auto-HSCT	F. proliferatum	Blood	0.54–1.45
8	F	57	Postchemotherapy aplasia, RAEB	FSSC	Purulent nasal discharge	0.50–0.60
9	M	41	Non-Hodgkin's lymphoma, allo-HSCT	F. verticillioides	Skin biopsy	Negative
10	M	5	ALL	F. verticillioides	Blood	Negative
11	F	9	ALL, auto-HSCT	F. verticillioides	Blood	0.70–4.16

^aM, male; F, female; allo-HSCT, allogeneic hematopoietic stem cell transplantation; AML, acute myeloid leukemia; ALL, acute lymphoblastic leukemia; RAEB, refractory anemia with excess blasts; auto-HSCT, autologous hematopoietic stem cell transplantation; FSSC, Fusarium solani species complex.

The means of the indices of reactivity in the Platelia assay of fungal cultures are reported in Table 2. All the Fusarium isolates produced positive reactions (index ≥ 0.5) when tested undiluted. The highest values (approximately 4) were detected in two FSSC isolates, one of which (F. falciforme) tested positive (0.55) also at the 1:100 dilution. The A. fumigatus isolate, tested as a control, showed a positive GM index when tested undiluted and diluted 1:100.

Table 2.

Reactivities of exoantigens from 12 clinical isolates of Fusarium species and from one A. fumigatus isolate in the Platelia Aspergillus assay

IUM isolatea	Patient (source)	Fusarium sp.	Index of reactivity in Platelia assay
			Undiluted reaction mixture	Reaction mixture diluted 10−2
07-0123	1 (pleural fluid)	F. oxysporum	2.35	0.25
09-1022	2 (bronchial secretions)	F. oxysporum	1.68	0.30
10-0049	4 (blood)	F. proliferatum	2.48	0.24
09-0505	3 (blood)	F. proliferatum	1.49	0.35
09-0782	7 (blood)	F. proliferatum	2.04	0.37
10-0035	(blood)	F. sp. cucurbitae MPV	2.05	0.18
10-0036	(blood)	F. sp. cucurbitae MPV	2.50	0.42
07-0259	(blood)	F. sp. cucurbitae MPV	3.97	0.36
08-0221	(skin biopsy)	F. falciforme	4.14	0.55
09-0620	(blood)	F. verticillioides	1.84	0.38
09-0780	10 (blood)	F. verticillioides	1.15	0.22
09-0781	11 (blood)	F. verticillioides	1.82	0.22
11-0088		A. fumigatus	6.05	1.72

^aIUM, Igiene Università Milano culture collection.

Monitoring of GM serum levels is currently used for early diagnosis of invasive aspergillosis due to its standardization and simple application in routine practice. The reported values of sensitivity and specificity of the Platelia Aspergillus assay differ widely in the literature. A 93% specificity for proven and probable cases was reported in a meta-analysis (13), whereas a low level of specificity is reported in some experiences (18).

Among the different causes of false-positive results, including the use of antibiotics such as piperacillin and amoxicillin, infections caused by mold containing cross-reactive GM have been reported. Cross-reactivity with antigens released by a large number of fungi has been reported in the literature (1, 3, 4, 8, 20, 21). Test specificity with exoantigens of Fusarium cultures is controversial (1, 6, 16).

Positive GM test results for patients with Fusarium infection have been anecdotally reported in the literature (18). In our survey, 9 out of 11 patients with disseminated or deep-seated Fusarium infection had repeated positive GM test results in the absence of isolation of Aspergillus spp.

To verify the hypothesis that a cross-reaction with exoantigens released by Fusarium might be the cause of Aspergillus GM positivity, we tested 12 Fusarium isolates, representing the species that most frequently cause infection, for their reactivity in the Platelia Aspergillus assay. All the Fusarium tested isolates showed positive indices of reactivity, even if lower than that shown by the A. fumigatus isolate used as control. The reactivity indices of Fusarium cultures ranged from 1.15 (F. verticillioides) to 4.14 (F. falciforme).

Our cases as well as in vitro studies show that Fusarium exoantigens may be cross-reactive in the Platelia Aspergillus assay. Therefore, a positive GM test in an immunocompromised host may represent invasive aspergillosis or another fungal infection, including Fusarium etiology. On the other hand, this unexpected reactivity of the test could be considered a useful diagnostic and prognostic tool in the management of Fusarium infections, as no antigen system for detection of this fungal infection exits.

In conclusion, these results show cross-reactivity of Fusarium spp. with Aspergillus GM that may constitute a drawback with respect to the specificity of the EIA. Therefore, a positive GM result, as with the presence of septate acute-angle branching hyphae in tissue and a positive β-d-glucan test, cannot help medical practitioners reach a specific diagnosis. This may represent a crucial point for the choice of antifungal therapy and for the evaluation of treatment outcome of the use of antifungals, such as echinocandins, that have been shown to be inactive in vitro against all Fusarium spp.