Medical Mycology Month 2011, Early Online, 14

Impact of galactomannan testing on the prevalence of invasive aspergillosis in patients with hematological malignancies
MARTIN HOENIGL*, HELMUT J. F. SALZER*, REINHARD B. RAGGAM, THOMAS VALENTIN*, ANGELIKA ROHN, ALBERT WLFLER, KATHARINA SEEBER, WERNER LINKESCH & ROBERT KRAUSE* *Section of Infectious Diseases , Division of Pulmonology, Medical University of Graz, Graz, Austria, Clinical Institute of Medical and Chemical Laboratory Diagnostics , Medical University Hospital of Graz, Graz, Austria, Division of Hematology, Medical University of Graz, Graz, Austria, and AGES PharmMed and Federal Office for Safety in Health Care , Austrian Agency for Health and Food Safety,Vienna, Austria

Galactomannan (GM) is a polysaccharide component of the cell wall of Aspergillus spp. and is released into the hosts circulation by growing hyphae. GM testing of patients with hematological malignancies has been rarely considered in recent epidemiologic studies of invasive mould infections (IMIs). The aim of the investigation was to analyze the impact of GM testing on the reported prevalence of IMI by comparing detection rates of IMI before and after the introduction of this diagnostic procedure. Prevalence of IMI was assessed by conducting a prospective single-centre study over seven months in 2010. Results obtained were then compared to those obtained with a representative collection of patients assessed by the same investigators at the same institution over seven months in 2007, i.e., prior to the introduction of GM testing. We found that, in general, detection rates of invasive aspergillosis (IA) and invasive mould infections increased significantly after the introduction of GM analysis. This study may therefore indicate that GM testing has a significant impact on the reported prevalence of IMI. Broad usage of such testing in patients with hematological malignancies may be able to produce a realistic picture of IMI rates when current diagnostic criteria are applied. Keywords invasive mould infection (IMI), aspergillosis, hematological malignancy, galactomannan testing, voriconazole

Introduction
The inability to consistently achieve early and reliable diagnoses of invasive mould infections (IMIs) remains one of the central problems in the management of patients with hematological malignancies. The European Organization for Research and Treatment of Cancer Invasive Fungal Infections Cooperative Group (EORTC) and the Mycoses Study Group of the National Institute of Allergy and Infectious Disease (MSG) have had some success in tackling
Received 21 March 2011; Received in final revised form 2 May 2011; Accepted 2 July 2011 Correspondence: Martin Hoenigl, Section of Infectious Diseases, Division of Pulmonology, Medical University of Graz, Auenbruggerplatz 20, 8036 Graz, Austria. Tel.: 43 316 385 3150; fax: 43 316 385 4622; E-mail: martin.hoenigl@medunigraz.at

this problem by establishing consensus definitions for defining opportunistic IMI [1]. However, when using these criteria the reported prevalence of IMI significantly depends on both the use of galactomannan (GM) antigen testing and the rate at which invasive diagnostic procedures are performed. GM is a polysaccharide component of the cell wall of Aspergillus spp. that is released into the circulation of the host by growing Aspergillus hyphae. Invasive aspergillosis (IA) accounts for the majority of IMIs in patients with hematological malignancies [2]. A test for IA should not be too invasive nor a burden for the already weakened immunocompromised patient. Serum GM detection through the use of the Platelia Aspergillus enzyme immunoassay (EIA) seems to have potential of meeting both of these requirements and has proven to be a promising tool in the diagnosis of IA [3,4].
DOI: 10.3109/13693786.2011.603102

2011 ISHAM

Hoenigl et al.

EORTC/MSG criteria define possible, probable and proven invasive mould infections. The latter requires histopathologic, cytopathologic or microscopic evidence of the fungus in sterile samples primarily obtained at autopsy or through needle biopsies. However, these procedures are often hindered by the severe medical condition of patients at risk for IMI [1]. Therefore differentiation between possible and probable IMI is of particular clinical relevance for early initiation of adequate therapeutic regimens. Possible IMI is defined as fulfilment of clinical and host criteria, while probable IMI requires additional mycological evidence. While recovery of the etiologic agent in cultures is often falsely negative in cases of IA, GM testing offers an uncomplicated and rapid way of getting this evidence and, therefore, accounts as the mycological evidence in the vast majority of probable IA in recent studies [5,6]. Differentiation between possible and probable IMI is also crucial for estimating the real burden of IMI as rates are consistently based on only probable and proven cases, whereas the number of possible cases is rarely described [5]. However, frequencies of GM testing, as well as modalities of invasive diagnostic procedures do not only vary significantly but are also usually not reported, even in recent groundbreaking multicenter studies [5]. Therefore, many probable cases may not have been included in previous epidemiological studies and the actual burden of IMI in patients with hematological malignancies may still be underestimated [7]. The aim of this study was to compare the reported prevalence of IMI before and after the introduction of GM testing in our clinical setting.

an index 0.5 was considered positive. In a Cochrane database systemic review, overall sensitivity of 78% (6189%) and overall specificity of 81% was reported for samples with a 0.5 cut-off values, with most investigations reporting a marked increase of specificity when the cut-off value was set at 0.7 [3,9,10]. The study was approved by the local ethics committee. Statistical analysis was performed using SPSS, version 17 (SPSS Inc., Chicago, IL, USA). A P-value of less than 0.05 was considered statistically significant.

Results
In 2010, the patients in 18% (129/729) of cases with hematological malignancies received systemic antifungal therapy, with those in 3.29% (24/729) of cases developing IMI. Specifically, 38% (9/24) of these cases were classified as possible, 54% (13/24) as probable and 8% (2/24) as proven IMI according to EORTC/MSG criteria [1]. IA was the most common IMI accounting for 73% (11/15) of probable and proven cases. Fusarium solani, Absidia spp., Geosmithia argillacea and Penicillium spp. were each involved in one probable or proven IMI. In 73% (94/129) of cases in which systemic antifungal therapy was employed, serum GM testing was conducted which resulted in positive results in ten cases. Probable IA was diagnosed in 9/10 of these cases and autopsy proven IMI due to Geosmithia argillacea was identified in one patient with GM index test results 21.9. GM index was 0.7 in 7/10 cases and between 0.5 and 0.7 in the other 3 cases. In 89% (84/94) of cases in which the samples were tested for GM levels were negative. Possible false positive GM test results due to treatment with piperacillin/tazobactam or echinocandin were not observed in our patient population. In our study collective positive predictive value for GM testing was 100%, negative predictive value 97.6%. The number of patients receiving systemic antifungal therapy was comparable to those treated in 2007. Interestingly, rates of probable or proven IMI (P 0.01), as well as rates of probable and proven IA (P 0.02) were significantly higher in 2010 when compared to 2007. Chest computer tomography was performed in 62% (80/129) and bronchoscopy in 19% (24/129) of cases. Fourteen percent of cases underwent allogeneic hematopoietic stem cell transplantation (HSCT), 5% autologous HSCT and 17% had undergone allogeneic HSCT prior to current admission. Six of 24 patients with IMI had a fatal outcome of which five died within four weeks of diagnosis and another after 10 weeks. Two of the six patients died of other causes not related to IMI. Results are listed in Table 1.
2011 ISHAM, Medical Mycology, Early Online, 14

Materials and methods

A prospective single-centre study was conducted at the Division of Hematology, Medical University of Graz, Austria, to assess the prevalence of IMIs over seven months in 2010. The objective was to analyze the impact of GM testing on the reported prevalence of such infections. Results obtained were then compared to a representative patient population assessed by the same investigators at the same institution over seven months in 2007 prior to the introduction of GM testing [8]. GM testing was routinely performed twice a week, the study population included only patients 18 years of age and patients were identified and screened by clinical rounds, chart reviews and surveys of electronic documents twice a week. Serum GM concentrations were determined by the Platelia Aspergillus EIA (Bio-Rad, Vienna, Austria) test kits in accord with the manufacturers instructions. Test variable was the titre of the Aspergillus GM antigen. Results were recorded as an index relative to the optical density of the control sample obtained in the same run with

Galactomannan testing and prevalence of invasive aspergillosis

Table 1 Demographic data, underlying diseases, invasive mould infection (IMI) and outcome. Antifungal therapy 2007e Number Sex male female Age median age Underlying diseases AML NHL MDS ALL CLL MM others HSCT autologous allogeneic Bronchoscopy Prolonged Neutropenia 117/690 (17%) 70/117 (60%) 47/117 (40%) 51.2 52/117 24/117 12/117 12/117 11/117 7/117 6/117 (44%) (21%) (10%) (10%) (9%) (6%) (5%) Antifungal therapy 2010 129/729 (18%) 85/129 (66%) 44/129 (34%) 53.5 63/129 24/129 19/129 16/129 10/129 4/129 9/129 (49%) (19%) (15%) (12%) (8%) (3%) (7%) Pa IMI 2007e 27/690 (4.9%) 14/27 (52%) 13/27 (48%) 56.7 15/27 (56%) 6/27 (22%) 3/27 (11%) 0 5/27 (19%) 0 0 0 6/27 (22%) 13/27(48%) 9/27 (33%) Pc

IMI 2010 24/729 (3.8%) 19/24 (79%) 5/24 (21%) 53.6 11/24 1/24 4/24 7/24 2/24 1/24 1/24 (46%) (4%) (17%) (29%) (8%) (4%) (4%)

0.002

11/117 (9%) 12/117 (10%) 36/117 (31%) 42/117 (36%)

6/129 (5%) 18/129 (14%) 24/129 (19%) 57/129 (44%)

0.03

1/24 (3%) 5/24 (17%) 13/24 (54%) 16/24 (67%)b

0.02

Antifungal agents considered gold standard in case of IMI lip amph B voriconazole Status of IMI proven probable possible Causative pathogen Aspergillus spp. Other moulds Unknownd Breakthrough Infection Outcome Improvement Survival 12 weeks 7/117 (6%) 14/117 (12%) 9/129 (7%) 18/129 (14%) 4/27 (15%) 4/27 (15%) 2/27 (7%) 1/27 (4%) 24/27 (89%) 2/27 (7%) 1/27 (4%) 24/27 (89%) 1/27 (4%) 18/27 (67%) 18/27 (67%) 7/24 (29%) 11/24 (46%) 2/24 (8%) 13/24 (54%) 9/24 (38%) 11/24 (46%) 4/24 (17%) 9/24 (38%) 4/24 (17%) 20/24 (83%) 18/24 (75%) 0.02

0.001 0.001 0.002 0.001

ALL, acute lymphoblastic leukaemia; AML, acute myeloid leukaemia; CLL, chronic lymphocytic leukaemia; HSCT, haematopoietic stem cell transplantation; IMI, invasive mould infection; lip amph B, liposomal amphotericin B; MDS, myelodysplastic syndrome; MM, multiple myeloma; NHL, non-Hodgkin lymphoma. aP value displayed in case of significant difference between 2007 and 2010 in patient collective with antifungal therapy P 0.05. bSignificantly associated with IMI in 2010 P 0.05. cP value displayed in case of significant difference between 2007 and 2010 in patient collective with IMI P 0.05. dCausative pathogen is unknown in case of possible IMI. eData from 2007 previously published [8].

Discussion
We found a significant increase in IA diagnosed after the introduction of GM testing when compared to a representative collection of patients evaluated prior to the use of this method. The increase was caused by a shift in the classification of cases from possible, meaning causative pathogen remains unknown, to probable IMI [8]. Possible explanations for the increase include excessive use of antifungal agents, but we found no evidence of this situation as breakthrough infections did not significantly increase. Furthermore, there was not a change in the route of exposure as no construction work was
2011 ISHAM, Medical Mycology, Early Online, 14

performed at the institution during the study period. Therefore the increase in the reported diagnosis of IA and IMI must be directly associated with the introduction of GM testing, as positive GM test results fulfil the mycological criteria for the diagnosis of probable IA. Without GM testing the majority of these cases may have been classified as possible IMI. However, it could be suggested that in absence of GM-testing the rate of invasive diagnostic procedures, like bronchoscopies, might increase so as to obtain the mycological evidence to support a diagnosis of IMI. Consequently, the detection of moulds in culture might increase and as a result, these

Hoenigl et al.

cases might still have been classified as probable IMI. Indeed, we observed a significantly higher rate of bronchoscopies before the introduction of GM-testing. The fact that the number of probable and proven IMI increased 5-fold after introduction of GM testing may argue for its use as the primary diagnostic method in case of suspected IMI. Broad GM testing might demonstrate that the present reported rates of 3.4% of probable and proven invasive fungal infections among patients undergoing HSCT is a vast under estimation of IMI. Furthermore, the frequency of the use of GM testing is unknown but is probably currently quite low [5]. We found a rate of 2.61% of probable and proven IMI among patients with hematological malignancies and considering that only 26/729 patients had HSCT, and another 51/729 had HSCT prior to current admission, a considerably higher rate of IMI among HSCT patients may be suggested. Our study further indicates that GM testing may also have an impact on early initiation of adequate antifungal therapy and may therefore have an impact on patients survival. We found a significantly higher rate of voriconazole therapy in patients with IMI after introduction of GM testing when compared to the similar patient population of 2007. As voriconazole is considered the gold standard in treatment of IA, early initiation of therapy with this antifungal might well affect patient outcome when compared to those cases lacking mycologic evidence which are generally treated with echinocandins. Although rates of both, improvement on antifungal therapy and survival among patients with IMI were higher in 2010 when compared to 2007, the difference was not significant in our study patient population. Considering that Aspergillus resistance to various antifungal agents poses an increasing risk, however, the impact of mycological detection on outcome and survival may definitely be on the rise [11]. There is, however, one consideration which has to be taken into account when interpreting the results. Although the two patient populations were comparable and possible confounding factors were ruled out to the best of our knowledge, this is a non-randomized observational study and results should therefore be interpreted carefully. In conclusion, our study may indicate that GM testing has a significant impact on the prevalence of IMI in patients with hematological malignancies. Broad usage of GM testing seems to be essential to produce a realistic estimation of IMI prevalence when current EORTC/MSG criteria applied.
This paper was first published online on Early Online on 7 September 2011.

Acknowledgement
This study was funded in part by Merck Investigator Studies Program (MISP: 39543) research grant. The original data of this manuscript (data obtained from 2010 and comparison with previously published data from 2007) has not been presented previously. Declaration of interest: M. Hoenigl received a research grant from Merck Investigator Studies Program (MISP). The other authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

References
1 De Pauw B, Walsh TJ, Donnelly JP, et al. Revised definitions of invasive fungal disease from the 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) Consensus Group. Clin Infect Dis 2008; 46: 18131821. 2 Warnock DW, Hajjeh RA, Lasker BA. Epidemiology and prevention of invasive aspergillosis. Curr Infect Dis Rep 2001; 3: 507516. 3 Leeflang MM, Debets-Ossenkopp YJ, Visser CE, et al. Galactomannan detection for invasive aspergillosis in immunocompromized patients. Cochrane Database Syst Rev 2008; 4: CD007394. 4 Maertens J, Van Eldere J, Verhaegen J, et al. Use of circulating galactomannan screening for early diagnosis of invasive aspergillosis in allogeneic stem cell transplant recipients. J Infect Dis 2002; 186: 12971306. 5 Kontoyiannis DP, Marr KA, Park BJ, et al. Prospective surveillance for invasive fungal infections in hematopoietic stem cell transplant recipients, 20012006: Overview of the Transplant-Associated Infection Surveillance Network (TRANSNET) Database. Clin Infect Dis 2010; 50: 10911100. 6 Perkhofer S, Lass-Florl C, Hell M, et al. The Nationwide Austrian Aspergillus Registry: a prospective data collection on epidemiology, therapy and outcome of invasive mould infections in immunocompromised and/or immunosuppressed patients. Int J Antimicrob Agents 2010; 36: 531536. 7 Hoenigl M, Valentin T, Salzer HJ, Zollner-Schwetz I, Krause R. Underestimating the real burden of invasive fungal infections in hematopoietic stem cell transplant recipients? Clin Infect Dis 2010; 51: 253254; author reply 254255. 8 Hoenigl M, Zollner-Schwetz I, Sill H, et al. Epidemiology of invasive fungal infections and rational for antifungal therapy in patients with haematological malignancies. Myc 2010 (in press). 9 Hadrich I, Mary C, Makni F, et al. Comparison of PCR-ELISA and Real-Time PCR for invasive aspergillosis diagnosis in patients with hematological malignancies. Med Mycol 2011; 49: 489494. 10 Orlopp K, von Lilienfeld-Toal M, Marklein G, et al. False positivity of the Aspergillus galactomannan Platelia ELISA because of piperacillin/tazobactam treatment: does it represent a clinical problem? J Antimicrob Chemother 2008; 62: 11091112. 11 Arendrup MC, Mavridou E, Mortensen KL, et al. Development of azole resistance in Aspergillus fumigatus during azole therapy associated with change in virulence. PLoS One 2010; 5: e10080.

2011 ISHAM, Medical Mycology, Early Online, 14