Bacterial findings in optimised sampling and characterisation of S. aureus in chronic rhinosinusitis

Chronic rhinosinusitis (CRS) affects more than 10 % of the European population [1]. The condition is characterised by nasal congestion, nasal discharge, headache, facial fullness, and changes in smell and taste lasting longer than 3 months, and is verified by nasal endoscopy and/or computerised tomography of the sinuses [2, 3]. The disease is probably multifactorial, and different factors have been suggested to affect its development, including environmental factors and host factors [4‐7]. Reduced ventilation of the sinuses due to blockage of the ostiomeatal complex in the middle nasal meatus is thought to be one factor [6, 8]. Reduced oxygen pressure in the sinus and absorption of oxygen might promote bacterial growth [9]. Allergy and asthma are suggested to enhance mucosal swelling and cause obstruction of the ostium, thereby predisposing for CRS. Immune deficiencies, cell membrane sodium and chloride channel malfunction, and ciliary dysfunction may also be contributing factors [10]. Important environmental factors in development of CRS are probably microbial, especially fungi and bacteria contributing to chronic mucosal inflammation [7]. Bacterial biofilms are often found in sinuses in patients with CRS undergoing sinus surgery and are associated with more severe disease [11, 12]. Biofilms are complex structures composed of communities of microbs embedded within an extracellular matrix, predominantly polysaccharides. Bacterial involvement is well accepted in the pathogenesis of acute sinusitis, where Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis are the most common bacterial findings [5, 13]. The role of bacteria in CRS is less clear, and findings of Staphylococcus aureus, coagulase-negative staphylococci (CoNS), and anaerobes seem to dominate to various extents in different studies [13‐19]. However, a Brazilian study of 62 samples from maxillary sinuses of CRS patients found no anaerobes, and Pseudomonas aeruginosa was the most commonly found bacterium [20]. The variability in microbial presence in different studies might be a result of differences in culturing techniques, contamination of samples, patient selection, ethnic origin of the patients, and pre-treatment regimens. In addition, handling of samples can affect growth due to the high sensitivity of the anaerobes. S. aureus in maxillary sinus cultures has been reported in about 25 % of patients with CRS [21]. A meta-analysis supports the role of S. aureus in asthma and allergic rhinitis [22], and S. aureus has also been shown to have an association with inflammatory diseases, such as atopic dermatitis [23, 24]. S. aureus displays a wide range of virulence factors; among these, staphylococcal enterotoxins and toxic shock syndrome toxin-1 have been demonstrated to activate the immune system and affect proinflammatory cells by acting as superantigens. Some studies have suggested that chronic rhinosinusitis with nasal polyposis (CRSwNP) has a relationship with S. aureus infection and especially with staphylococcal enterotoxins as a modulator of the disease [6, 25‐27]. The aim of this study was to investigate the bacterial spectrum patients with CRS, and especially the presence of S. aureus, and to characterise S. aureus isolated from the nares and maxillary sinus of CRS patients in comparison with samples from the nares in healthy controls. Another aim was to evaluate an optimised culturing technique for the maxillary sinus.

Initial enrolment comprised 43 patients with CRS and 58 healthy controls. One patient was excluded due to having titanic dental implants reaching the maxillary sinus, and one of the controls was excluded due to an age below 18 years. The study group therefore consisted of 42 patients and 57 controls. Mean age was almost equal between the groups (51.5 and 50.0 years). The patient group was 46.0 % male, and the control group was 40.5 % male. Polyposis was present in 24 of the 42 (57 %) patients and none of the controls. All samples collected from the nares showed bacterial growth, though one sample in the patient group was missing. Fourteen different bacteria were identified. Table 1 shows the bacterial findings in patients and controls. The most common aerobic bacteria were CoNS, isolated from the maxillary sinus in 18/42 (43 %) patients. CoNS was found in nares in 17/42 (40 %) patients, and in 43/57 (75 %) controls (p = 0.0008). S. aureus was found in both nares and maxillary sinus in 15/42 (36 %) patients using the protected brush technique and in 18/42 (43 %) patients using the cotton-tipped aluminium swab technique (p = 0.66). S. aureus was isolated in the nares in 24/42 (57 %) patients and 16/57 (28 %) controls (p = 0.004). S. aureus findings in the maxillary sinus did not differ significantly between CRSwNP patients and CRSsNP patients (p = 0.35). Facultative anaerobic bacteria were found in five samples from nares of CRS patients and in two samples from controls. There was no significant difference between the patients and controls regarding the presence of facultative anaerobic bacteria in nares (p = 0.13). Maxillary sinus cultures using the protected brush technique showed two species present in 6/42 (14.3 %) patients, but none with three or more, while cultures using the conventional technique showed mixed flora in 14/42 (33 %) patients. At least one species of aerobic bacteria was cultured from 41/42 (98 %) patients, and anaerobic/facultative anaerobic bacteria was cultured from the maxillary sinus in 5/42 (12 %) patients. The nares showed mixed flora in 7/57 (12 %) controls and 14/42 (33 %) patients. Three samples (7 %) showed no growth, all collected with the brush technique from the maxillary sinus in CRS patients. In a comparison between the two sampling techniques, there was consistency in growth in 23/42 (56 %) patients. In 22/42 (52 %) of the brush samples, there was growth of only one species (Table 2). In one case, S. pneumoniae was found with the cotton-tipped aluminium swab and H. influenzae with the protected brush, and in another case, S. aureus was found with the cotton-tipped aluminium swab and H. influenzae with the protected brush. Furthermore, in two cases, α-streptococci and anaerobic Gram-positive cocci, respectively, were found with the protected brush but not with the cotton-tipped aluminium swab.

There was a wide variation in the distribution of spa types (Table 3; Fig. 3). We found identical spa types of S. aureus from the nares and from the maxillary sinus in 17/18 patients (94 %). The remaining patient with S. aureus isolates showed two unrelated spa types in the two areas: t084 and t189.

The isolation rate of S. aureus in maxillary sinus cultures from patients with CRS has been reported to be about 25 % [21]. In this study, S. aureus was found in 15/42 (36 %) of the maxillary sinus cultures from CRS patients with both the optimised sampling technique (protected brush) and standard cotton-tipped aluminium swab. In another three patients, S. aureus were found using standard cotton-tipped aluminium swab but not with the brush. This indicates a contamination rate of 3/18 (17 %) for S. aureus. In 17/42 (40.5 %) samples, bacterial findings with standard cotton-tipped aluminium swab differ from findings using protected brush. However, CoNS in combination with other bacterial findings using cotton-tipped aluminium swab compromised most of these differences. The protected brush technique was superior to the standard cotton-tipped aluminium swab in reducing the contamination of CoNS and S. aureus when culturing maxillary sinus. However, the clinical relevance of using this optimised technique as the protected brush is questionable. In almost all patients cultured with the cotton-tipped aluminium swab, this culturing technique seemed to be adequate for collecting clinically relevant samples, and thus have a low impact on antibiotic treatment (Table 3). S. aureus is thought to persistently colonise the nares of about 20 % of the population (range 12–30 %), and approximately 30 % of the remainder is intermittent carriers (range 16–70 %) [31, 32]. Wertheim et al. regarded the nose as the major site of S. aureus carriage, and from here, the organism is thought to spread to other parts of the body [32]. S. aureus and CoNS were the predominant bacterial findings in the maxillary sinus of our patients with CRS. In addition, the number of patients with CRS with S. aureus in the nares was statistically significantly higher than the number of controls with S. aureus in the nares, while the number of patients with CoNS was statistically significantly lower compared to controls. The high S. aureus colonisation rate of the nares in patients with CRS may reflect its importance in the CRS disease. A hypothesis based on our findings might be that S. aureus counteracts the CoNS in the nares and promotes colonisation of S. aureus in the maxillary sinus through transport of bacteria across the mucus conjoining the nares with the sinus.

Furthermore, identical spa types in both the nares and maxillary sinus were found in all patients with findings of S. aureus in both locations (with one exception), which supports the theory that the nares can be the primary site from where the bacteria can spread and colonise the maxillary sinus. S. aureus has in certain circumstances the ability to produce enterotoxins acting as superantigens that bind to T cells and exaggerate disease severity and expression [25], and cause serious invasive diseases such as sepsis with or without infective endocarditis and necrotising pneumonia as well as mild diseases, such as superficial skin and soft tissue infections. S. aureus enterotoxins may play a role in the severity of CRS, especially CRSwNP [27, 33, 34]. The ability of superantigens to enhance inflammatory reactions [25] might have a connection to the predominance of S. aureus in CRS, again especially CRSwNP [35]. An increased immune response to S. aureus enterotoxins has been demonstrated in nasal polyposis tissue, resulting in more pronounced eosinophilic inflammation and higher local immunoglobulin E production against staphylococcal enterotoxins in patients affected by CRSwNP [26, 36]. Our study included 23 patients with CRSwNP and 19 with CRSsNP. We could not find any differences in bacterial findings in nares or maxillary sinus when comparing these groups, but the two groups were small and comparison is difficult. Brook et al., who studied the microbiology of the maxillary sinus in 48 patients with CRS, also found no difference between CRSwNP and CRSsNP [37] which recently also was shown by Brook et al. [38]. Another study showed no significant difference in the prevalence of genes encoding virulence factors between isolates from patients with CRS and healthy controls [39]. However, a large meta-analysis by Ou et al. based on 12 case–control studies with a total of 340 cases and 178 controls demonstrated a relationship between the presence of S. aureus superantigens and the persistence and severity of CRSwNP [27].

The high rate of S. aureus in the maxillary sinus in patients with CRS in our study strengthens the importance of S. aureus for the pathogenesis of the disease, but its role needs to be further investigated. Larger studies could hopefully lead to valuable knowledge about the role of S. aureus in CRS. Sweden is a low endemic area for methicillin-resistant S. aureus (MRSA), and there was only one MRSA isolate in our study; this is probably also due to a generally low prescription of antibiotics in Sweden, which leads to low incidence of resistant isolates. All participants were living in Sweden, but we have no information on ethnic origin, which may have been of interest. Some patients were included in conjunction with sinus surgery, and some had a previous history of sinus surgery and were included as outpatients at the otolaryngology clinic. However, a previous study indicates that endoscopic sinus surgery does not change the bacterial flora, though it does change the presence of bacterial biofilms in the sinus [40].

Contamination with CoNS was common using cotton-tipped aluminium swab technique. However, the protected brush technique did reduce the contamination rate but does not significantly improve the reliability of bacteriological diagnostics for practical clinical use. The risk of unnecessary antibiotic treatment is, therefore, thought to be low when using the easy-to-handle cotton-tipped aluminium swab. Furthermore, S. aureus was found in 36 % of the maxillary sinus samples from the CRS patients, and the number of patients with S. aureus in the nares was statistically significantly higher than the number of controls with S. aureus in the nares, while the number of patients with CoNS in the nares was statistically significantly lower compared to controls. These findings, together with a very high rate of identical spa types of S. aureus from patients with S. aureus in both nares and maxillary sinus, might indicate colonisation of the maxillary sinus from the nares and presumably a relationship between S. aureus and CRS.