The detection of galactomannan GM
The detection of galactomannan (GM), a polysaccharide that is released during Aspergillus growth, is a useful and reliable non-invasive diagnostic test for screening and management of aspergillosis and is often more sensitive than culture (Schelenz et al., 2015, Fisher et al., 2013). The level of galactomannan in body fluids such as serum and BAL is positively associated with the fungal burden of Aspergillus (Fisher et al., 2013). Two studies have shown evaluated GM detection in sputum, one in patients with invasive aspergillosis (Kimura et al., 2009), one in cystic fibrosis (Baxter et al., n.d.). The optimal cut-off for the galactomannan index (GMI), as determined by constructing receiver operating characteristic (ROC) curve, and for the diagnosis of invasive pulmonary aspergillosis had been set at equal or >0.5 for serum and 1.0 for BAL respectively (Schelenz et al., 2015, Kimura et al., 2009, Khorvash et al., 2014, Morrissey et al., 2013, He et al., 2011, Richardson and Warnock, 2012). There have been considerable variations in the sensitivity and specificity of GMI detections in body fluids (Fisher et al., 2013). In BAL fluid, the sensitivity ranges between 29 and 100% (Hope et al., 2005), whilst it ranges between 17 and 100% in serum (Schelenz et al., 2015). There is considerable controversy surrounding the optimal cut-off for respiratory secretion GMI values. A cut-off value of 1.2 in sputum was proposed for Aspergillus GM in IPA among haematological patients with sensitivity and specificity set at 100 and 62.2% respectively (Kimura et al., 2008, Kimura et al., 2009).
Real-Time PCR (RT-PCR) has now become an important diagnostic procedure for the diagnosis of aspergillosis (Fraczek et al., 2014). Detection of Aspergillus DNA has been applied to most body fluids including blood, BAL and sputum specimens using commercially available kits, for example MycAssay®Aspergillus (Baxter et al., 2013) (Lab21, Cambridge, UK). Both GM and Aspergillus PCR have shown higher sensitivity than culture for the detection of Aspergillus species in sputum in CF patients (Baxter et al., 2013a, Baxter et al., 2011) and PCR in non-CF patients (Denning et al., 2011, Langridge et al., 2016). A Chlorhexidine digluconate australia threshold (Ct) value<36.0 is regarded as positive for Aspergillus DNA in RT-PCR in respiratory fluids (Tuon, 2007). However, there have been many inconsistencies in the reported RT-PCR Ct and GM index values for positive detection of Aspergillus (Kimura et al., 2009).
The survival of patients with aspergillosis is highly dependent on the ability to make an early and accurate diagnosis. Since the isolation of organisms is rarely successful, this study was designed to determine the GM index and PCR Ct values to evaluate the cut-off values for positive sputum detection of Aspergillus in patients with ABPA and CPA.
Materials and methods
Discussion In sputum, the gold standard to confirm the presence of Aspergillus spp. remains microscopy and culture, but these methods are not highly sensitive, however positivity in microscopy and culture of sputum is not always indicative of disease (Schelenz et al., 2015). In our study, there was no perfect relationship between the GMI and PCR for the detection of Aspergillus in the sputum of either patients with ABPA or CPA. This is reflected in the mean GMI level of 3.81, and the mean transformed Ct of 1.49. This result does not concur with previous studies on the diagnosis of pulmonary aspergillosis that validated a GM index of 1.0 for BAL (Khorvash et al., 2014, Morrissey et al., 2013, Wichmann and Kluge, 2013) and 1.2 for sputum (16). In the present study, forty-seven (29.6%) patients had GMI<1.0 whilst about 60.1% of them were negative for Aspergillus DNA. Even though PCR is highly sensitive and specific, most of the patients in whom Aspergillus DNA was not detected in their sputum had detectable GM, with a GMI which ranged from 2.0 to 14.0. A similar situation is seen with our ‘control’ samples in which 33% had high GM indices. In fact the term control is potentially misleading as GPs in the UK tend to submit samples for problematic patients and the two patients with high GM indices could have some form of pulmonary or airways aspergillosis, such as Aspergillus bronchitis. The prior study evaluating GM in sputum from patients suspected of having invasive aspergillosis indicated that GMI sensitivity may be improved by testing multiple samples and two consecutive positive GMI readings should be considered diagnostic (He et al., 2011). It was also discovered that PCR was more sensitive than GMI in our study as confirmed by others (Fraczek et al., 2014, Buchheidt et al., 2004) though both biomarkers were significantly correlated.