Over the past few decades, the incidence of invasive mold disease (IMD) has increased significantly and the fungal spectrum of IMD has broadened. According to literature, over 100,000 IMD cases occur each year, and these are associated with high morbidity and mortality in immunocompromised patients who have hematological malignancy and transplantation (Ruiz-Camps and Jarque, 2014; Ruhnke et al., 2015; Pegorie et al., 2017; Springer et al., 2018).

The members of the order Mucorales and genus Aspergillus are the most common opportunistic pathogens of IMD (Oren and Paul, 2014). Because of the significantly different antifungal susceptibilities and the complexity of the population of patients at risk, management of patients with IMD which lacks typical clinical manifestations has become increasingly complex. Therefore, early and reliable diagnostic methods are necessary for effective treatment.

Currently, the gold standard for the diagnosis of invasive fungal infections depends on histopathology and culture. However, culture is time-consuming and may fail if the potential microbial causes are not considered during sample collection, so formalin-fixed, paraffin-embedded (FFPE) bioptic material is collected for subsequent histological diagnosis. FFPE tissues obtained from patients with proven IMDs are frequently used to detect the etiology of invasive mycoses (Tarrand et al., 2003; Munoz-Cadavid et al., 2010; Babouee Flury et al., 2014). While histopathology can prove invasive fungal infections, the analytical correctness of histological findings is no more than 79% (Sangoi et al., 2009). Therefore, preliminary histological results should be interpreted cautiously (Guarner and Brandt, 2011) and supported by the culture whenever possible. In addition, histopathological observations of fungal shape and arrangement may not be sufficient for accurate identification of the Mucorales and Aspergillus if only a limited quantity of fungal hyphae is present.

In recent years, considerable efforts have been made to develop more sensitive and specific tools and protocols for IMD diagnosis. It is reported that PCR-based techniques, including conventional, semi-nested and real-time PCR, can be used to identify fungal agents in FFPE tissue (Bialek et al., 2005; Rickerts et al., 2006; Walsh et al., 2011; Springer et al., 2016a). RQ-PCR is very suitable for detecting the DNA of FFPE samples which are easily degraded. There are reports using the 18SrRNA gene and the 28SrRNA gene regions to detect and distinguish mucormycosis and invasive aspergillosis (Bialek et al., 2005; Walsh et al., 2011; Springer et al., 2016a; Gade et al., 2017).

The objective of this study was to evaluate RQ-PCR protocols by the use of TaqMan technology for detection and identification of Mucorales and Aspergillus in FFPE samples, targeting the 18SrRNA gene and the 28SrRNA gene. Identification of Mucorales and Aspergillus was analyzed by the combination of Mucorales RQ-PCR (Mucorales18SrRNA and Mucorales28SrRNA) and Aspergillus RQ-PCR (Aspergillus18SrRNA and Aspergillus28SrRNA).

The study involved 76 patients (F/M, 42/34; age, 50.95 ± 17.31 years) with following comorbidities: 23 (30.26%) had diabetes, 16 (21.05%) had hypertension, 10 (13.16%) had bronchiectasis, 8 (10.53%) had solid tumor, 7 (9.21%) had Chronic obstructive pulmonary disease (COPD), 5 (6.58%) had hematological malignancy, 5 (6.58%) had tuberculosis, and 16 (21.05%) had others. Slides stained with PAS or GMS were considered, respectively, positive if magenta or brown-black fungal hyphae with morphological features were observed ( Figure 2 ). Positive special staining with GMS and/or PAS were 24/29 (82.76%). Positive culture cases were 12/52 (23.08%). Positive 1, 3-beta-D-glucan assay cases were 9/40 (22.50%). Positive Galactomannan assay cases were 12/42 (28.57%). The principal site of infections was in lungs (61 cases), next were in other sites (including one ileum, three nasal cavity, four maxillary sinus, two trachea, one external auditory canal, one toe and three main bronchus). CT of the chest was obtained in 61 patients with pulmonary infection. There was a wide spectrum of radiological findings, with the most common being 26 nodules, followed by 20 mass, 16 cavitation, 7 consolidation, 5 pleural effusion, and 3 air crescent sign. Result of bronchofibroscopy were obtained in 56/61 patients with pulmonary infection: 18 patients were normal, and other patients were mainly necrosis, luminal stenosis, and purulent secretion.

The detection and identification of Mucorales and Aspergillus from FFPE samples played an important role in the diagnosis and management of aspergillosis and mucormycosis, whereas microscopy, serology, and culture were restricted by several disadvantages (Jensen et al., 1997; Frater et al., 2001; Rickerts et al., 2007; Hofman et al., 2010; Hamilos et al., 2011). Numerous RQ-PCR assays have been described for detection of Mucorales (Bernal-Martinez et al., 2013; Millon et al., 2013; Lengerova et al., 2014) and Aspergillus (Lass-Florl et al., 2011; Fricke et al., 2012; Paholcsek et al., 2014; Pini et al., 2015). In this study, we evaluated two Mucorales RQ-PCR assays, two Aspergillus RQ-PCR assays, and four combined tests to rapidly detect and identify Mucorales and Aspergillus.

More recently, the RQ-PCR of Mucorales from 268 serum samples and 12 FFPE samples was a promising test method with sensitivity of 91% targeting 18SrRNA gene (Springer et al., 2016b) and 86% targeting 28SrRNA gene (Springer et al., 2016a). In this study, two Mucorales RQ-PCR only had a sensitivity of 75%. The lower sensitivity may be relevant to small sample size and different sample types. The RQ-PCR detecting Mucorales also had a specificity of 100% in FFPE samples but a lower sensitivity of 56% by Hata et al. (Hata et al., 2008). The specificity of our two Mucorales RQ-PCR assays was 97.1% and 94.20%, respectively. As reported in the literature, the RQ-PCR specificity was 87.5% (Springer et al., 2016a). In Mucorales RQ-PCR assays, the diagnostic parameters and LoD values for the different assays indicated that Mucorales18SrRNA would provide the best diagnostic accuracy. These results support the findings of Springer et al. (Springer et al., 2016a; Springer et al., 2016b).

In our study, the specificity and sensitivity of Aspergillus18S rRNA were 87.88% and 73.33%, respectively, which is consistent with the results of previously published studies (Hadrich et al., 2011; Luong et al., 2011). The MycAssay™ Aspergillus real-time PCR kit was tested on tissues by the manufacturer with 15 different Aspergillus spp., including multiple strains of Aspergillus fumigatus, Aspergillus niger, Aspergillus terreus, and Aspergillus nidulans, having a sensitivity of 82% and a specificity of 79% (Lass-Florl et al., 2011). In our study, the specificity of two Aspergillus RQ-PCR assays was elevated, whereas sensitivity was reduced.

However, some limitations and several considerations indicate that it has some drawbacks in the Mucorales RQ-PCR or Aspergillus RQ-PCR alone. Our studies have evaluated the utility of detection of Mucorales RQ-PCR and Aspergillus RQ-PCR. Our finding, combining the results of these two tests gave optimal specificity and PPV, could be used to detect and identify Mucorales and Aspergillus and may provide a solution when gold standard tests were conflicting. Given the ubiquity of Aspergillus and Mucorales in the environment, combining Mucorales RQ-PCR with Aspergillus RQ-PCR would give clinicians greater confidence in detecting and identifying them at the same time and reduce the false positive and/or negative rate. The best combination was the M28A18, with a sensitivity of 67.90%, a specificity of 100%, which had the highest diagnostic potential with FFPE samples. This is inconsistent with the conclusion that Mucorales18SrRNA is better than Mucorales28SrRNA in diagnostic significance, which may be related to the small sample size. In all samples, the number of true positives of M18A18 was very similar to M28A18 (54 VS. 55). Their diagnostic significance needs to be further evaluated in future studies with larger sample sizes.

This study has several limitations, which may be the cause of some “abnormal results”. First, the sample size is too small, especially that the mucormycosis sample is only 21 cases from 19 patients. Due to the limited number of FFPE samples and strains, the very low number of Aspergillus and Mucorales species was tested, the number of Aspergillus and Mucorales species that can be detected by these primers needs to be further evaluated. Second, RQ-PCR assays require strictly positive and negative controls. The limitations of false positive and false negative errors due to amplification and contamination for assessing the value of a molecular diagnostic test have been eloquently highlighted (Mies, 1994; Cataloluk et al., 2003). In some samples, fluorescence signals higher than the positive Cq cutoff was detected, which may be caused by false negatives due to the low number of fungal hyphae. Third, the use of mold-active drugs may affect detection result. There are conflicting reports about the effect of antifungal therapy on the performance of tests. Antifungal therapy has been reported to both decrease (Reinwald et al., 2012) and increase (Musher et al., 2004) the diagnostic performance of RQ-PCR. Furthermore, “abnormal results” may be caused by a condition other than Mucormycosis and Aspergillosis or by drug treatment. Fourthly, Formalin fixed and paraffin wax embedded tissues can cause DNA degradation, only short sequences can be amplified from this type of tissue (Bonin et al., 2003). Although the amplified sequences of the four RQ-PCRs in this study were less than 200bp, which weakened the influence of DNA fragmentation, it may still reduce the sensitivity. Fifth, due to funding reasons, we did not use commercialized kits for the detection of Mucorales and Aspergillus DNA in clinical samples. Finally, Misclassification using these inconsistent criteria of RQ-PCR can occur for many reasons, e.g., detection of pathogenic fungi may be missed due to the diversity of fungal species and test samples, and there are multiple laboratorial protocols. Each suffers from different disadvantages such as vulnerability to contamination or limited detection of selected species or genera.

In conclusion, this preliminary study showed that the two Aspergillus RQ-PCR assays and two Mucorales RQ-PCR assays had high potential for the diagnosis of Mucorales and Aspergillus in FFPE samples. We envisage Aspergillus RQ-PCR and Mucorales RQ-PCR combination approach as a nearpatient test, allowing an immediate detection and identification of Mucorales and Aspergillus, with RQ-PCR results being available within a short time for samples of mucormycosis with aspergillosis. This combination approach can provide useful information when a small number of fungi are present, or the histological diagnosis is difficult in mucormycosis with aspergillosis samples. In the future, these assays may be used as a screening tool to detect other types of samples suspected of having mucormycosis and/or aspergillosis, such as serum, bronchoalveolar lavage fluid (BALF), and cytological samples. The results of our study should be validated in multicenter studies to develop tests for this clinical application.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by West China Hospital Ethics Committee of Sichuan University. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

YJ and FY and conceived and designed the experiments. XJ analyzed the data and wrote the manuscript. All authors contributed to the article and approved the submitted version.