The significance of the two biomarkers for the detection of
The significance of the two biomarkers for the detection of Aspergillus antigen and DNA in the sputum of patients cannot be over-emphasised. The mean GMI value was higher among CPA patients than the ABPA. This might not be unconnected with the fact that CPA is associated with lung damage and pre-existing lung diseases that allow the airway or damaged lung tissue to be colonized with Aspergillus spp. (Zmeili and Soubani, 2007, Kimura et al., 2009, Smith and Denning, 2011, Schweer et al., 2014). ABPA, on the other hand, is characterised by intense inflammation of the airway and colonization by conidia, Psalmotoxin 1 and hyphal fragments of Aspergillus (;). This fact is also supported in our study where seventeen CPA patients produced a positive sputum culture of A. fumigatus, while only three ABPA patients had positive sputum cultures. This is also in agreement with previous studies where some of the patients with Aspergillus isolation from respiratory secretions could not fulfil the diagnostic criteria of ABPA (Liu et al., 2013). The characteristics and consistencies of sputum that have been adjudged to be a non-invasive alternative to bronchoscopy and BAL in the diagnosis of respiratory infections were studied in this audit. The sputum classification adopted by the MRCM was easier to assess and evaluate than the previous Geckler classification used in a previous study (Kimura et al., 2009). Most of the sputa were purulent and mucopurulent in nature. Purulent and blood stained sputum samples were significantly associated with positive cultures and strong PCR signals among CPA than ABPA patients, while GMI>6.5 was more often positive in purulent and blood-stained samples as well as muco-purulent sputum samples. Both PCR signals and GM were usually negative in watery-salivary and mucoid samples, raising the prospect for laboratories of rejecting these samples for further analysis or at least reporting the sputum character with the result, indirectly suggesting that a negative result is a potential false negative. Since we have no working gold standard for the diagnostic utility of GMI in our patient groups, we created three possible constructs for the analysis of GM values. Construct A “true positivity “considered TCt of >0.0 with positive microscopy or culture while Construct B considered positivity with TCt of ≥2.0, in addition to positive culture or microscopy. These constructs only captured about 10% of patients and so more than 90% of the patients were negative but might reasonably considered not be “true negatives“. Not all the patients who had either microscopy or culture positive were positive for both GM antigen and detection of Aspergillus DNA. These categories of patients have not been truly captured. We need further studies to be able to define the true positivity criteria for their diagnosis. We also considered “true negativity” with negative PCR, GM<1.0 and TCt of 0.0. In this category, about 45% of patients were truly negative. With all the inconsistencies in the results of the GM and PCR, we might need further studies to validate the usefulness of measurement of GM antigen in the sputum for the screening and diagnosis of CPA or ABPA. To minimize these effects, we constructed the ROC curve for the two constructs and true negativity to find the optimal cut-off for the positive GM. The optimal cut-off value of sputum GMI as estimated from the ROC curve for both ABPA and CPA in the two possible constructs was 6.5 (Fig. 2). Using the cut-off value of 6.5, the sensitivity and specificity of galactomannan antigen detection were 68.8 and 79.7% for construct A and 71.4 and 79.3% for construct B respectively. However, the PPV was below 30% and the NPV were between 96 and 97% (Table 3). This finding is in contrast with the previous study where 1.2 was the optimal cut-off for sputum in the diagnosis of IPA (Kimura et al., 2009). Our small ‘control’ group may not be true controls. Obtaining sputum from normal people, who would not naturally produce sputum, is challenging, and bronchoscopy or induced sputum samples are not a reasonable substitute for this purpose.