Clinical utility of Aspergillus galactomannan and PCR in bronchoalveolar lavage fluid for the diagnosis of invasive pulmonary aspergillosis in patients with haematological malignancies

Abstract

Interpretation of Aspergillus galactomannan (GM) and PCR results in bronchoalveolar lavage (BAL) fluid for the diagnosis of invasive pulmonary aspergillosis (IPA) in patients with haematological malignancies requires clarification. A total of 116 patients underwent BAL for investigation of new lung infiltrates: 40% were neutropenic, 68% and 36% were receiving mould-active antifungal agents and β-lactam antibiotics. The diagnosis of proven IPA (n = 3), probable IPA (n = 15), and possible invasive fungal disease (IFD, n = 50) was made without inclusion of GM results. BAL GM (at cut-off of 0.8) had lower diagnostic sensitivity for IPA than PCR (61% versus 78%) but higher specificity (93% versus 79%). Both tests had excellent negative predictive values (85–90%), supporting their utility in excluding IPA. The use of BAL GM and PCR results increased the certainty of Aspergillus aetiology in 7 probable IPA cases where fungal hyphae were detected in respiratory samples by microscopy, and upgraded 24 patients from possible IFD to probable IPA. Use of BAL GM and PCR improves the diagnosis of IPA.

Introduction

Invasive pulmonary aspergillosis (IPA) is a life-threatening infection in patients undergoing stem cell transplantation or chemotherapy for haematological malignancies with mortality rates of 36–75% (Kontoyiannis et al., 2010, Steinbach et al., 2012). Timely antifungal therapy may improve clinical outcomes, but establishing an early diagnosis of IPA is difficult (Rano et al., 2001, von Eiff et al., 1995). The current ‘gold standard’ for the diagnosis of proven IPA in these patients requires histological evidence of tissue infiltration or culture of Aspergillus from biopsy tissue taken from a sterile site (De Pauw et al., 2008). Yet many patients cannot undergo such invasive procedures, and culture is insensitive (~50%) (Cuenca Estrella et al., 2011). Non–culture-based serological and molecular diagnostic platforms have improved sensitivity for the earlier diagnosis of IPA (Mennink-Kersten et al., 2004, White et al., 2010) allowing for more targeted antifungal therapy.

In haematology patients, Aspergillus galactomannan (GM) enzyme-linked immunosorbent assay (ELISA) and nucleic acid detection by PCR are rapid, sensitive diagnostic tests for IPA (Khot et al., 2008, Maertens et al., 2009). The clinical utility of GM and Aspergillus PCR testing of serum/blood samples has been confirmed in a randomised controlled trial (Morrissey et al., 2013). Detection of Aspergillus GM and/or DNA in bronchoalveolar lavage (BAL) fluid may have superior sensitivity over serum/whole blood for IPA as samples are taken from the immediate vicinity of the infection and consequently may have a higher fungal burden. Furthermore, given the time taken for Aspergillus to invade across the alveolar capillary barrier [within the initial 24 h after lung infection as suggested in an in vitro model (Hope et al., 2007)], testing of GM and/or DNA in BAL fluid may provide earlier positive results than testing on serum. These tests provide a more rapid result than culture and additionally have better potential to confirm the aetiology of a suspected invasive fungal disease (IFD).

However, considerable variations in the reported test performance, including sensitivity rates of BAL GM and PCR, have been noted (Luong et al., 2010, Maertens et al., 2009, Musher et al., 2004, Racil et al., 2011, Reinwald et al., 2012a, Reinwald et al., 2012b), These variations are likely due to differences in patient characteristics and antimicrobial prescribing practices between studies. Additionally, the optimal interpretive optical density (OD) index cut-off for positivity of BAL GM has not been fully determined, and different values have been used between studies (Heng et al., 2013). The lack of standardisation of PCR methods has made it difficult to compare its clinical utility in BAL fluid across studies. Many centres include BAL GM and PCR testing in their clinical management algorithms, but few have reported their results, and those that have are single-centre experiences, which may lack generalizability. Therefore, we undertook a study to assess the utility of GM and PCR testing in BAL fluid, for the diagnosis of IPA in high-risk haematology patients undergoing diagnostic bronchoscopy in 3 Australian hospitals.