Chronic rhinosinusitis (CRS) is characterized by inflammation of the nasal mucosa and one or more paranasal sinuses for temporal intervals extending 8–12 weeks [1-3]. In Canada, CRS is an increasing public health burden and accounted for approximately 3 million dispensed prescriptions in 2006 [4,5]. Several Canadian epidemiological surveys have reported the prevalence of CRS to affect approximately 5% of the general population [5,6]. In those studies, the prevalence of CRS was higher in women compared to men and eastern provinces had the highest prevalence of CRS [5,6]. In Canada, like the United States, CRS causes significant morbidity and reduced workplace productivity [2,5,7]. In the National Population Health Survey, Canadian survey participants that were identified with CRS had a self-perceived health status comparable to participants with other chronic conditions such as cancer, arthritis, asthma and inflammatory bowel disease [5]. A reduction in the mental health of survey participants with CRS was also reported and associated with a higher prevalence of depression, anti-depressant use, and number of visits to mental health professionals [5]. These results highlight the emerging societal burden of CRS; however, fewer studies have characterized the allergic sensitizations associated with cases of CRS in Canada.

Studies of CRS etiology have shown that a progressive accumulation of mucus can result in remodeling and obstruction of the sinus ostia, bone erosion, facial pain, swelling, and can co-occur with or without polyp formation [1]. Although CRS may have an infectious or non-infectious basis [1,8], allergic sensitization has been suggested in several studies as an important contributing factor to the underlying pathology [3,6,9-11]. Total IgE levels have been shown to correlate with the extent of CRS disease on sinus CT [12]. Inflammation, swelling and decreased mucociliary clearance associated with allergic rhinitis and eosinophilic inflammation have been postulated to enhance paranasal sinus ostia destruction [6,9,10]. This pathology may increase the risk of CRS development in sensitized patients but to date, the role of allergens and allergic sensitization remains uncharacterized and requires further clinical evaluation.

The coexistence of allergic rhinitis and CRS has been previously reported in the literature [9,10,13,14] including in a Canadian study [6]. The prevalence of allergic sensitization has been reported to be as high as 84% in patients with CRS [9,13] and a major subset of this patient group can be sensitized to more than one allergen [9]. Perennial allergen sensitization is more prevalent in CRS patients [14] with dust mite and mold allergens among the most frequently reported sensitizing sources [9,14]. The role of fungi in CRS has been the focus of many research studies as these microbial contaminants are commonly recovered from eosinophilic/allergic mucin. The association with fungi has been controversial but is sometimes described as mediating host immune reactions leading to the development of CRS [8,15]. Despite these experimental findings, the contribution of fungi and other perennial allergens to CRS in Canada has remained poorly characterized.

In this study, we document the prevalence of allergic sensitization to a selection of regionally prevalent perennial and seasonal allergen sources in a cohort of Canadian CRS patients. In addition, we compare the prevalence of allergic sensitization to a wider panel of fungi implicated in CRS. The identification of allergic sensitization to regionally prevalent allergens may be an additional diagnostic marker of CRS in clinical populations.

Of the 158 study participants recruited into the study, 101 had a confirmed clinical history of CRS and 57 control subjects had a clinical history of CIU but not CRS. In the CRS group, there were 59 females and 42 males. The age range was 19–70 and the mean age was 44.5. In the control group, there were 28 females and 29 males with an age range 33–61 and the mean age was 47. A total of 31 SPT antigen extracts were tested and 72 out of 99 (73%) CRS patients and 14 out of 44 (32%) CIU control patients reacted to at least one SPT extract (data not shown).

The SPT extracts most likely to elicit a positive SPT response in the CRS population were ragweed (40%), A. alternata (36%), cat (29%), dust mite (29%), and June grass (30%; Table  1). In the control population, ragweed (22%), cat (11%), June grass (21%), and dust mite (16%) extracts were the most common (Table  1). The SPT reactivity to the A. alternata extract was 3 fold higher in CRS compared to the control population. In the serological analysis, sIgE for the grass mix (21%) and tree mix (24%) was lower in CRS compared to control patients; however, there was a greater prevalence of fungal positive patients in CRS (21%) compared to the control population (16%) but this was not statistically significant (Table  2).

In the chi-squared analysis, positive SPT responses to A. alternata were more prevalent in patients with CRS compared to controls (odds ratio (OR): 4.34, 95% confidence interval: 1.57, 12.02; Table  1). Sensitization to other perennial and seasonal SPT extracts including cat (OR: 3.23, 95% CI: 1.16, 9.02) and ragweed (OR: 2.31, 95% CI: 1.02, 5.19) was additionally identified to be more prevalent in the CRS group compared to the control group (Table  1). An elevated OR (1.4) was also observed following the analysis of IgE ImmunoCAP mold mix serological data; however, this was not statistically significant (Table  2). Differences between CRS and control patients were observed for other perennial and seasonal allergen sources including dog (OR: 3.17, 95% CI: 0.68, 14.72), dust mite (OR: 2.19, 95% CI: 0.88, 5.48), and timothy grass (OR: 3.16, 95% CI: 0.88, 11.34); however, these differences did not reach statistical significance and that may be due to the small sample size (Table  1). No other statistically significant differences were identified between the other allergen extracts tested in the SPT and serological analysis (Table  2).

In the ANOVA analysis of sIgE, the mean allergen specific units (kUA/L) measured for CRS and CIU control patients did not vary for grass (p = 0.67), mold (p = 0.34) or tree mix (p = 0.46) ImmunoCAP antigens (Table  3). Similarly, there were no statistically significant differences between mean sIgG titers (mg/mL) in CRS and control patients for C. herbarum (p = 0.84), A. alternata (p = 0.92), S. atra (p = 0.13), and A. versicolor (p = 0.22; Table  4).

In this study, 73% of CRS patients had a positive SPT to at least one of the tested allergen extracts. Sensitization to A. alternata, cat and ragweed were the most prevalent sensitizations in CRS patients compared to the control population. Sensitization to dust mite or timothy grass was also noted but these results were not statistically significant. These results highlight the potential importance of allergic sensitization to A. alternata and other perennial or seasonal allergen sources in CRS patients.

The prevalence of CRS patients having at least one positive SPT (73%) was comparable to results reported in a study that was located in close proximity (Cleveland, OH) to Toronto [9]. The identification of increased prevalence of fungal allergy in CRS was also identified in Cleveland as well as in southern regions of the United States [8,14,18] but not in Europe [19]. The serological results reported in the present study varied from previous studies reporting up to a 5 fold increase in IgG titers as well as increases in IgE⁺ cells (mast, plasma cells) in CRS patients [20,21]. The lack of concordance between SPT and serological studies in CRS patients may be due to increased sensitivity of SPT but also the variability of allergen extracts, as well as concurrent exposure of the subjects to antigens derived from a varying spectrum of perennial and seasonal allergen sources (tree, grass, insect, and fungal) [9]. Allergic sensitization to staphylococcal and streptococcal toxins has been found to be as high as 78% in CRS patients [22]. While these factors were not evaluated in the present study, they could also influence the pathology and prevalence of sensitization in CRS patients. Taking these limitations into consideration, the results of the present study suggest that evidence of allergic sensitization, in particular fungal hypersensitivity, could potentially be used as diagnostic criteria for CRS [8,18]. Currently, the Canadian clinical practice guidelines for CRS recognize allergy testing in cases of CRS but the guidelines do not recognize specific allergens [4].

In the present study, 36% of the CRS study participants were sensitized to A. alternata. These SPT results are comparable to other studies that have identified sensitization to Alternaria in CRS patients [20]. A. alternata has been previously implicated in cases of CRS in the United States [20] and it is among the most ubiquitous phyloplane fungal species in Toronto [23,24]. The presence of A. alternata conidia in the upper respiratory tract has been demonstrated in microscopic and immunologic studies of allergic mucin derived from CRS subjects [25-27]. In vitro studies have also shown that A. alternata extracts induce the recruitment of eosinophils, induce interferon-y, IL-5 and IL-13 in peripheral blood mononuclear cells [20,28]; and regulates the expression of surfactant protein D and LL-37 in fungal sensitized CRS patients [29,30]. It has been hypothesized that allergens released from Alternaria conidia and germinating conidia may have a functional role in CRS but this has remained controversial and is the subject of continued research [8,31].

Sensitization to cat and ragweed were additionally identified as prevalent sensitizations in the present study, and sensitization to these allergens has been previously reported in a CRS study in Washington D.C. [32]. Cat allergy is an important source of perennial allergen in inner city environments. Interestingly, Canadian aerobiological surveys have shown that the highest concentration of ragweed pollen occur in southern Ontario [33].

Our results deviate from previous studies that have identified dust mite to be an important perennial allergen associated with CRS [9,13,19,34]. Gutman and colleagues identified 73% of CRS patients in Cleveland to be sensitized to dust mite [9]. A study of 200 patients in San Francisco who had undergone functional endoscopic sinus surgery for chronic or recurrent rhinosinusitis showed that 22% were sensitized to house dust mite [13]. Freudenberger et al. [34] also identified 68% of CRS patients in New York to exhibit a positive immediate skin test reaction to dust mite extract. Although dust mite sensitization can be as high as 60% in Toronto [35], the prevalence of sensitization to dust mite only approached 30% in the present study and was not more prevalent in patients with CRS. To date, the role of seasonal and perennial allergy in CRS requires further evaluation; however, the results of this study suggest that in addition to Alternaria, cat and ragweed allergens may additionally contribute to swelling and blockage of sinus ostia leading to mucous buildup, secondary infection, and the development of CRS [4].

There were several limiting factors associated with the design of this study that should be considered when interpreting these results. The association of other seasonal and perennial allergens to CRS such as dust mite, dog and timothy grass may have been identified if a larger number of study participants and controls were recruited into the study. Selection and recruitment of patients included CRS patients with and without nasal polyps and allergic fungal sinusitis which may have resulted in the over estimation of allergic sensitization in this population of Canadian CRS patients [10,11,36]. However, our objective was to study the overall prevalence of allergy sensitization in CRS. Geographic variability in CRS and allergic sensitization are also variables that may have influenced the results of this study [5,6]. Surveys of CRS in Canada have shown a greater prevalence of CRS in eastern provinces, in particular Nova Scotia [5]. We hypothesize that the allergen burden in Nova Scotia may differ to the Toronto metropolitan environment due to a varying spectrum of tree, grass, weed, insect and mold allergens and as a result, the profile of allergic sensitization may vary substantially. Future studies should include larger patient numbers and utilize a wider panel of perennial and seasonal allergens that reflect the allergen burden of the local study environment.

In conclusion, the results of this study demonstrate a general association between CRS and allergic sensitization. In the Toronto metropolitan area, A. alternata was prominent but cat and ragweed allergens were also associated with CRS. Taken together, these results suggest exposure to regionally specific allergens may drive a process resulting in CRS in some patients, perhaps with a genetic predisposition or epidermal barrier dysfunction. Future studies should be designed in other regions of Canada to identify and characterize the specific sensitization profiles of CRS patients.

The authors declare no conflict of interest or competing interests.

BJG is a co-investigator with input on study design, interpreted statistical results, and prepared the manuscript. DHB is a co-investigator with input on study design and implementation, and assisted in the preparation and review of the manuscript. ZG enrolled patients, conducted regulatory work, and assisted in the collection of data, CSB conducted data analysis, reviewed patient charts, and reviewed the manuscript. RM recruited patients, conducted data analysis, reviewed patient charts, and reviewed the manuscript. TAB conducted the Phadia ImmunCAP serological analysis and reviewed the manuscript. MLK conducted the statistical analysis, interpreted and discussed results, and reviewed the manuscript. GLS is the principle investigator and contributed to the enrolment of patients, data capture and processing and has reviewed and commented on various drafts of the manuscript. All authors have read and approved the final manuscript.