In this study a GMI of was associated
In this study, a GMI of 1.0 was associated with high sensitivity above 90%, specificity below 30% and higher negative predictive value. However, this cut-off couldn\'t have been optimal because of the lower level of Youden J Index which was <0.2 in both constructs (Ruopp et al., 2008). This inconsistency was also supported by varied sensitivities that ranged from 27 to 100% in serum and 60–100% in sputum (Fisher et al., 2013, Kimura et al., 2009, He et al., 2011). The reasons for these observations might be linked to sample variability, the load of organism, other medications, antifungal resistance in the context of antifungal therapy and patient host status (Kimura et al., 2009, He et al., 2011). In order to validate the optimal cut-off for Aspergillus GM in sputum, it is suggested that long term prospective studies involving larger populations of ABPA or CPA patients with diverse predisposing factors should be carried out. In addition to the non-invasive nature of sputum collection, a previous study had confirmed that GM could be more easily detected in sputum than in other respiratory fluids (Kimura et al., 2009). Since there was no adequate control group in the present study, it was technically difficult to assign the optimal cut–off values for the GMI.
Aspergillus spp. especially A. fumigatus and Penicillium spp. were isolated in about 13.2% of the sputa cultured from CPA and ABPA patients in our study. We had only three culture positive sputa in ABPA patients. This rate was low compared to that found in previous studies where the culture rate from sputum was as high as 56–80% (Denning et al., 2011, Denning et al., 2003b). High volume culture that was adopted by the MRCM could improve significantly the yield of Aspergilli from sputum samples (Fraczek et al., 2014). There is a positive correlation between the median GM of 6.11 and positive culture of sputum. This result supported the optimal cut-off value of 6.5 of our GM analysis. >50% of patients with culture positive sputum had a GMI >6.0.
A strongly positive Azithromycin with TCt of ≥2.0 was significantly associated with positive sputum culture. A GM value of >6.5 was strongly associated with about 50% of strongly positive PCR results. As noted previously, a few of the positive sputum cultures did not concur with the corresponding positive detection of Aspergillus DNA by PCR. Due to variability in the sensitivity of PCR and increased significance of adopting TCt of >0.0 as weak signal of Aspergillus detection, there is a need for urgent standardization and validation of PCR tests across all laboratories as advocated by other studies (Denning et al., 2011, Fisher et al., 2013).
Sputum consistencies could be a good predictor of positivity for microscopy, culture, GMI values and Aspergillus detection by PCR. Mucopurulent, purulent and blood-stained sputum had higher GM and PCR values. However, our study has shown that higher mean GMI values and TCt values for PCR were significantly associated with purulent sputum samples.
In conclusion, mucopurulent, purulent and blood stained sputum had higher GM and PCR values. A GMI value of >6.5 was associated with ~50% of strongly positive PCR values. We have no control group in our study which is required for setting the optimal cut-off. There are inconsistencies for both GM and PCR, so within patient reproducibility may be problematic (data not shown). Based on the results presented in this study, it is not clear whether GM is useful in sputum for diagnosis or for following therapy given the wide range of values, which did not correspond with PCR or culture.
Future direction To be able to have the definitive optimal cut-off value for the detection of Aspergillus DNA by PCR and galactomannan antigen, more samples are needed for both index cases and the control groups. It is not clear what the source of the GM found in such high quantities in sputum is. In the future studies, possible confounding variables such as antibiotics (piperacillin-tazobactam, ampicillin), many opportunistic fungi, some solutions containing gluconate, food containing cereals products and soya beans should be evaluated (Girmenia et al., 2007, Lescher-Bru et al., 1998).