Results
The majority of CRS patients had relatively severe disease on radiologic scoring, 66% had polyposis, and the majority were undergoing revision endoscopic sinus surgery, reflecting the tertiary nature of the surgical practice (see Table 1).
Number of Species Detected
Molecular organism detection using the Ibis T-5000 was positive in 100% of CRS patients and controls. A total of 33 different bacterial species were identified in CRS patients by the Ibis system, with a mean of 3.0 (2.0–4.0) species detected per patient (see Additional file 1: Table S1). In control patients, 5 different organisms were detected with a mean of 2.0 (1.0–3.0) per patient. There was a trend of increasing mean isolates per patient from controls 2.0 (1.0–3.0), CRS without nasal polyposis (CRSsNP) 2.5 (1.0–3.0), to CRS with nasal polyposis (CRSwNP) 3.2 (2.0–4.0) but this was not statistically significant (p > 0.05, Kruskal-Wallis test, Dunn's post-hoc comparison). Ibis analysis detected fungi in only 4 CRSwNP patients, and no fungi were detected in CRSsNP patients or controls.
Conventional culture was positive in 73% of CRS patients, with an average of 1.3 organisms per patient detected. 12 different organisms were identified. Cultures were positive from 33% of control patients. Only one organism (Staphylococcus epidermidis) was cultured from this subject group.
Diversity
30/38 (79%) CRS patients, and 3/6 (50%) of control patients, had more than one bacterial species detected on the mucosa using the Ibis molecular diagnostic. Staphylococcus aureus was the most commonly detected organism in CRS patients (23/38, 61%), followed by Staphylococcus epidermidis (21/38, 55%), and Propionibacterium acnes (14/38, 37%, see Table 2). Nocardia asteroides (9/38, 24%), Haemophilus influenzae (5/38, 13%) and Pseudomonas aeruginosa (3/38, 8%) were detected less commonly. In control patients, Propionibacterium acnes was the most commonly detected organism (5/6, 85%), followed by Staphylococcus epidermidis, present in 4/6 (67%) of patients. Less commonly detected were Staphylococcus aureus 2/6 (33%), Nocardia asteroides 1/6 (17%) and Streptococcus agalactiae 1/6 (17%). Anaerobic species were detected in 18/38 (47%) of CRS patients, and 5/6 (83%) of control patients.
Conventional cultures produced polymicrobial results in 21/38 (55%) CRS patients, whereas no control patients had polymicrobial flora using this method. CRS patients yielded predominantly S. aureus (14/38) and Staphylococcus epidermidis/coagulase-negative Staphylococcus (CNS) (11/38). Only Staphylococcus epidermidis/CNS was cultivated from 2 control patients. No growth was reported for 9/38 CRS patients, and 4/6 control patients.
Fungi were rarely detected in CRSwNP patients regardless of technique used, and were not detected in CRSsNP patients or controls. Analysis with the Ibis biosensor showed only two species - Aspergillus fumigatus (3), and Bipolaris papendorfii (1). Culture detected only two of the Ibis fungal positive patients (both Aspergillus fumigatus) plus one Penicillium chrysogenum, and one Trichosporon in other patients, which were not detected by the biosensor. Fungal-specific FISH analysis was positive in the three A. fumigatus patients detected by the Ibis biosensor, the P. chrysogenum culture-positive patient, and two additional patients, which were negative by the other techniques. Regardless of detection technique, all patients in whom fungi was detected had nasal polyposis.
Detection of S. Aureus Using Ibis Biosensor, FISH, and Culture
We used S. aureus as a model organism for studying the characteristics of the three different detection methods. The biosensor detected S. aureus in 61% of CRS patients, biofilm was positive in 50% (see Figure 1), and conventional culture was positive in 37%. FISH analysis had a sensitivity of 78% (±18) for the detection of S. aureus compared to Ibis, with a specificity of 93% (± 6). In contrast, conventional culture detected S. aureus with a sensitivity of 61% (±17), and a specificity of 93% (+16). S. aureus was detected in control patients only by molecular detection (2/6, 33%). Neither conventional culture nor FISH analysis was positive for S. aureus in these patients.
(Enlarge Image)
Figure 1.
Staphylococcus aureus FISH probe tagged with Alexa 488, analyzed with confocal scanning laser microscopy at 80×magnification.
Relative Microbial Density
We determined the relative numbers of the bacterial species using the Ibis biosensor, which provides a genomes/well measure. Results for the top ten organisms are shown in Table 2. The most commonly detected organisms in CRS patients, S. aureus and S. epidermidis, had markedly different microbial densities, with S. aureus organisms present in much higher numbers than S. epidermidis based on bacterial genomes detected. Some of the less commonly detected organisms such as Corynebacterium pseudodiphtheriticum, Moraxella catarrhalis, Streptococcus pneumoniae, and Pseudomonas aeruginosa, were found in relatively high numbers when detected on the mucosa. The abundance of S. aureus was much greater in CRS patients compared to controls, however other organisms such as S. epidermidis and P. acnes were detected in similar quantities in both patient groups. When all organisms were considered, CRS patients had significantly greater bacterial genomes per sample than control patients (p < 0.05, Mann–Whitney U test - see Table 2).
Are FISH and Culture Detection Related to Microbial Abundance
We again used the organism S. aureus to compare detection techniques based on microbial numbers. The quantitative genomic analyses of the 25 (CRS and control) samples in which S. aureus was detected by the Ibis biosensor, were compared to the detection of the organism by FISH and conventional culture to determine if these latter detection methods are dependent on the abundance of organisms to provide a positive finding (see Figure 2). The number of S. aureus genomes per sample was significantly higher in those specimens that tested positive for S. aureus by FISH and conventional culture (p < 0.05, Kruskal-Wallis test, Dunn's post-hoc comparison). However there was no statistically significant difference between FISH and culture detection sensitivity based on S. aureus genomes per sample (p > 0.05, see Figure 1).
(Enlarge Image)
Figure 2.
The relationship betweenS. aureusabundance, and detection by FISH and culture. The detection of S. aureus by FISH and culture is dependant on the abundance of the bacteria as measured by the number of bacterial genomes present in the sample.