CRSsNP indicates chronic rhinosinusitis without nasal polyps; CRSwNP, chronic rhinosinusitis with nasal polyps; and TDI, threshold discrimination identification.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Schlosser RJ, Mulligan JK, Hyer JM, Karnezis TT, Gudis DA, Soler ZM. Mucous Cytokine Levels in Chronic Rhinosinusitis–Associated Olfactory Loss. JAMA Otolaryngol Head Neck Surg. 2016;142(8):731–737. doi:10.1001/jamaoto.2016.0927
Olfactory loss is a frequent symptom of patients with chronic rhinosinusitis (CRS), but our understanding of how inflammatory cytokines affect olfaction is limited.
To examine whether inflammatory cytokines are present in the olfactory cleft and whether they correlate with objective olfaction.
Design, Setting, and Participants
In this cross-sectional study, patients with CRS underwent quantitative olfactory testing using the Sniffin Sticks test to calculate a composite threshold discrimination identification (TDI) score from October 21, 2013, to November 12, 2015. Nasal mucus was collected using a sponge placed in the olfactory cleft for 5 minutes, and Cytometric Bead Array was used to measure secreted immunomodulatory products. Correlations between TDI score and secreted mediators were then calculated. Data analysis was performed from October 15, 2015, to December 17, 2015.
Main Outcomes and Measures
Composite TDI scores and mean secreted mediator levels in mucus from the olfactory cleft.
Thirty-four patients were enrolled (mean [SD] age, 57.3 [15.7] years; female, 21 [61.8%]; white, 26 [76.5]). The TDI scores were lower in patients with CRS with nasal polyps (CRSwNP) than in patients with CRS without nasal polyps (CRSsNP) (difference, 8.7; 95% CI, 2.5-15.0; P = .007). Interleukin (IL) 5 levels were inversely correlated with TDI scores in patients with CRSwNP and those with CRSsNP (mean [SE] β estimate, −46.56 [15.11]; P = .005), whereas IL-6, IL-7, and vascular endothelial growth factor A were positively correlated with TDI scores only in the CRSwNP cohort. Subscale olfactory TDI scores followed similar correlations to composite TDI scores.
Conclusions and Relevance
In this study, inflammatory cytokines were found in mucus collected from the olfactory cleft. Levels of IL-5, in addition to other cytokines, were associated with objective olfactory function. Further inquiry is needed to establish the source of mucous cytokines and establish whether they play a causal role in olfactory loss.
Olfactory dysfunction is a cardinal symptom of chronic rhinosinusitis (CRS), a disease that affects 12.5% of the adult population across all racial and ethnic groups.1,2 The prevalence of olfactory dysfunction in CRS ranges from 30% to 80%; however, the mechanisms by which olfactory loss occurs are not always clear.3,4 Some patients have significant nasal obstruction and impaired airflow into the olfactory cleft due to polyps or edematous mucosa. Other patients have olfactory loss but without obvious airflow restriction, suggesting additional mechanisms may also be contributory.
Several studies5,6 exploring olfactory loss in CRS have examined specialized olfactory epithelium biopsy specimens, revealing histologic evidence of localized inflammation with lymphocytes, macrophages, and eosinophils. These studies5,6 suggest that the mucosal inflammation typically found in sinus mucosa proper may also be present in the olfactory cleft of some patients with CRS and that this inflammation could contribute to olfactory loss. One of the chief difficulties in studying the human olfactory cleft is the inability to easily perform mucosal biopsies. Although biopsy techniques have been previously described,7,8 the narrow location, associated discomfort, and invasiveness of these procedures limit their use, particularly on a large scale or in the nonoperative setting.
Collection of sinonasal mucus is a technique that has been used to help characterize inflammation in the sinus mucosa proper. Although not a perfect replication of mucosal inflammation, mucus collected from the middle meatus has been found to contain secreted cytokines, which significantly correlate with ethmoid tissue cytokine levels, with differences between those with and without polyps as expected.9 Sinonasal mucus can thus give insights into mucosal inflammation and at a minimum serve as a potential biomarker for discriminatory or predictive studies.
With regard to olfactory dysfunction, elegant animal models have revealed that upregulation of local tumor necrosis factor α (TNF-α) leads to olfactory loss in a reversible fashion.10,11 In humans, an additional study12 reported higher local interleukin (IL) 6 in the nasal mucus of patients with hyposmia from a variety of non-CRS origins when compared with normosmic controls. However, no studies have specifically evaluated cytokine levels in mucus collected directly from the olfactory cleft in patients with CRS. Analysis of olfactory cleft mucus in CRS offers several potential advantages. First, it can be performed in the clinical setting, allowing insights into patients who might not otherwise require surgery. Second, because it does not violate the epithelium, repeated assessments can also be performed over time with less fear of inducing biopsy-related changes or complications, such as scarring, loss of olfaction, or cerebrospinal fluid leak.
Similar to histopathologic findings, we hypothesized that inflammatory mediators would be found in the olfactory cleft in patients with CRS and that specific cytokine levels might correlate with olfactory function. The goals of the current study were thus to (1) determine whether inflammatory mediators could be detected in mucus collected from the olfactory cleft, (2) evaluate whether mediator profile differs between patients with and without polyps, and (3) analyze whether specific mediators correlate with objective olfaction after controlling for polyp status.
Question Can inflammatory mediators be detected in mucus collected from the olfactory cleft in patients with chronic rhinosinusitis?
Findings In a cross-sectional study of 34 patients, 13 of 19 inflammatory mediators were detected in 100% of samples across all patients.
Meaning Inflammatory mediators are found in mucus collected from the olfactory cleft of patients with chronic rhinosinusitis and could contribute to olfactory dysfunction.
Adult patients (≥18 years old) with CRS were prospectively enrolled into a cross-sectional study approved by the Medical University of South Carolina Institutional Review Board from October 21, 2013, to November 12, 2015. Data analysis was performed from October 15, 2015, to December 17, 2015. Written informed consent was obtained from all patients, and data were deidentified. All patients satisfied diagnostic criteria for CRS according to the current Clinical Practice Guideline of the American Academy of Otolaryngology–Head and Neck Surgery, including the presence of at least 2 cardinal symptoms for at least 12 weeks and confirmatory evidence of sinus mucosal inflammation on sinonasal endoscopy or computed tomography.2 Patients receiving systemic corticosteroids in the previous 30 days and those who underwent surgery in the preceding 6 months were excluded. Demographic and exposure information was collected from each patient, including age, sex, race/ethnicity, and active smoking status. Comorbidities of interest were recorded, including the presence of asthma, allergic rhinitis, diabetes mellitus, and prior sinus surgery. Asthma status was determined based on prior physician diagnosis. Polyp status was determined by sinonasal endoscopy and recorded in a binary fashion into those with CRS with nasal polyps (CRSwNP) and CRS without nasal polyps (CRSsNP). Allergic rhinitis was defined as those with a physician diagnosis confirmed with prior positive objective testing (skin prick test or allergen specific IgE antibody test).
Objective olfactory testing was performed using Sniffin Sticks (Burghart Messtechnik GmbH).13 Testing was performed by a trained clinical research coordinator masked to other clinical data. The testing battery performed included odor threshold (OT), odor discrimination (OD), and odor identification (OI). The OT test was performed using dilutions of n-butanol in a single-staircase, triple-forced choice procedure. The OD test used triplets of pens presented in random order, with 2 containing the same odorant and 1 a different odorant. The OI test involved 16 odorants presented at suprathreshold intensity using multiple choice procedures. All participants were blindfolded to avoid visual identification of odorant-containing pens. Each of the 3 individual tests was scored from 0 to 16. The overall results combined and reported as a composite threshold discrimination identification (TDI) score, with higher scores representing better olfaction (range, 0-48).
At enrollment, a polyurethane sponge (Greer Laboratories) was placed into each olfactory cleft under endoscopic guidance and allowed to dwell for 5 minutes. Specifically, sponges were placed into the space between the septum and the middle and superior turbinates. Sponges extended from the anterior plane of the head of the middle turbinate to just in front of the sphenoid sinus. Sponges were then removed and immediately centrifuged at 4°C for 30 minutes. The mucus was then combined from each side, transferred by pipette to a cryotube, and stored at −80°C until use.
The presence of inflammatory cytokines in olfactory mucus was assessed using commercially available Cytometric Bead Array systems with enhanced sensitivity (BD Biosciences). Kits and reagents were purchased for an array of cytokines, chemokines, and immune mediators. Each of these factors has been reported to be altered in patients with CRS vs controls or hypothesized to affect olfactory function in prior studies.14-18 Assays were performed according to manufacturers’ instructions and read on a Guava easyCyte 8HT flow cytometer (EMD Millipore), with analysis performed with FCAP Array Software, version 1.0.1 (BD Biosciences).
Before formal hypothesis tests were conducted, descriptive statistics were generated for all patients, and demographic and comorbidity differences at baseline were tested between the CRSsNP and CRSwNP subgroups. For each mediator, the frequency with which it was detected in mucous samples (ie, above the threshold for sensitivity of the assay) was recorded. Mean values for each mediator were then compared between the CRSsNP and CRSwNP subgroups. To assess differences at baseline for categorical and continuous data, Pearson χ2 tests and t tests were performed, respectively. Where appropriate, Fisher exact tests were performed instead.
Linear models were used to assess the association between olfactory cleft mucous mediator level and olfaction (TDI score) (Figure). An interaction effect was included in the model to determine whether polyp status affected the association between each mediator and total TDI score. A tolerant significance level of P ≤ .15 was used for the interaction effect to determine whether additional tests of association would be performed by polyp status. For those cytokines significantly associated with TDI score on the pooled sample or significant for at least 1 of the polyp status groups, the same associations were tested on the OT, OD, and OI outcomes. This analytic strategy was used to reduce the total number of tests of association and minimize overall type I error. All analyses were performed using SAS statistical software, version 9.4 (SAS Institute Inc). Statistical significance was set at α = .05.
A total of 34 patients with CRS were enrolled, divided nearly evenly between the CRSsNP and CRSwNP subgroups. No significant differences were found between the CRSsNP and CRSwNP groups for demographic factors or medical comorbidities (Table 1). Patients with CRSwNP had significantly lower TDI scores (difference, 8.7; 95% CI, 2.5-15.0; P = .007) compared with patients with CRSsNP, whereas patients with CRSsNP had worse total Sinonsal Outcomes Test 22 scores (difference, 12.5; 95% CI, −0.4 to 25.3; P = .06).
Most olfactory cleft mucous samples contained the cytokines, chemokines, and other inflammatory mediators of interest (Table 2) within the detectable range for the individual assays. The exceptions were IL-13 and CXCL10, which were found in only 58.1% and 41.9% of samples, respectively. Mean values for each mediator are given in Table 3. In general, the SDs are relatively large, indicating variation across patients. Levels of IL-2 were found to be higher in the CRSsNP group (difference, 4.0; 95% CI, 0.4-7.6), whereas total IgE levels were higher in the CRSwNP group (difference, −60 470.3; 95% CI, −116 764.0 to −4177.1). Although not statistically significant, no differences were found between the CRSsNP and CRSwNP groups for IL-5 (difference, −33.0; 95% CI, −71.2 to 5.2), IL-6 (difference, 248.3; 95% CI, −52.4 to 548.9), and TNF-α (difference, 10.6; 95% CI, −0.6 to 21.8).
When all patients with CRS were examined, IL-5 had a highly significant inverse correlation with TDI scores (P < .001) (Table 4) independent of polyp status. When patients with CRSwNP were examined, IL-6, IL-7, and vascular endothelial growth factor A (VEGF-A) had significant positive correlations with TDI scores (P = .02 for IL-6, P = .01 for IL-7, and P = .02 for VEGF-A). With the exception of IL-5, no other cytokines correlated with TDI in the CRSsNP subgroup. Further analysis of these 4 cytokines on OT, OD, and OI revealed findings similar to overall TDI score correlations (Table 5).
This study was a relatively small pilot investigation to determine feasibility and utility of measuring inflammatory mediators in olfactory cleft mucous samples. We were able to find that an array of cytokines (both TH1 and TH2), chemokines, and other inflammatory factors are present within olfactory cleft mucus. Furthermore, there is significant variability in these secreted inflammatory measures across individuals with CRS, as well as differences across subgroups defined by polyp status. Most important, we were able to find a significant association between several cytokines and objective olfactory function, even after controlling for polyp status.
The cross-sectional nature of this study and the lack of control groups do not allow for definitive conclusions to be drawn regarding causality. However, it is worth exploring whether there is biologic plausibility that cytokines in the olfactory cleft could contribute to olfactory loss. We found that a number of olfactory cleft cytokines correlated with objective olfactory TDI scores, the most prominent of which was IL-5. Interleukin 5 was inversely correlated with TDI scores in both the CRSwNP and CRSsNP groups. Although IL-5 has traditionally been reported to be upregulated in TH2-skewed CRSwNP and its association to olfaction in this group is not surprising, it may have a similar role in CRSsNP. One report19 found that even among patients with CRSwNP, this cytokine is differentially expressed, with patients with asthma and CRSwNP having upregulation and others having relatively low levels. It has been predictive of medical and surgical response in CRSwNP, with IL-5 levels greater than 40 pg/mL predicting response to anti–IL-5 therapy.20 In patients with CRSwNP who had recurrence, the median nasal IL-5 level was 482 pg/mL vs 144 pg/mL in patients without recurrence.21
Although it was not surprising to find that IL-5 may play a role in olfaction, we were unable to find any correlation between objective olfaction and a number of other TH2 cytokines, including IL-4 and IL-13. A previous report20 that divided patients with CRSwNP based on nasal mucous IL-5 levels failed to find a difference in general eosinophil markers, including eosinophil cationic protein and eotaxin. It is possible that eosinophils and other TH2 cytokines, such as IL-4 and IL-13, are broadly upregulated in nearly all patients with CRSwNP, rather than having a dichotomous distribution reported for IL-5. This finding would imply that markers such as IL-4 and IL-13 may not be specific for those patients with CRS-related olfactory dysfunction but rather a nonspecific marker of CRSwNP in general.
Several other cytokines, including IL-6, IL-7, and VEGF-A, correlated positively with TDI scores in patients with CRSwNP. Interleukin 6 was previously reported to be inversely correlated with olfactory function12; however, this study was in patients without CRS, so its role may differ in CRS-associated olfactory dysfunction. Interleukin 7 stimulates proliferation of lymphoid cells, including B cells, T cells, and natural killer cells. Its role in CRS has not been previously reported but appears to be worth investigating. VEGF-A induces growth of blood vessels, but it also stimulates monocyte and macrophage migration. It also plays a critical role in neural regeneration because neurons require vascular supply. Vascular endothelial growth factor is up-regulated in CRSwNP and drives epithelial cell proliferation.22 In particular, VEGF-A has neurotrophic effects and has been proposed as a potential therapy for neurodegenerative disorders.23 It is possible that the positive correlation we found is due to increased vascularization and improved olfactory neuron function, but this is purely speculative.
In patients with CRSsNP, IL-5 was the only cytokine found to correlate with TDI scores. Higher levels of IL-5 resulted in worse TDI scores, implying that it plays a similar key role in olfactory function for patients with CRSsNP and CRSwNP. The lack of other olfactory-related cytokines in CRSsNP may be due to its more heterogeneous phenotype and immune profile, typically being reported as a mixed TH1 and TH2 picture. It is possible that this heterogeneity contributed to difficulty finding key cytokines that may play a role in olfaction. Certainly, the relatively small size of this study opens up the possibility of type II error as well. We have previously reported on sinus-specific cytokines and found that many inflammatory cytokines, such as IL-6, IL-4, and IL-10, were upregulated in CRSsNP but not to the same degree as in CRSwNP.9 Thus, larger sample sizes and more precise classifications of CRSsNP may reveal key cytokines in certain subsets of patients. Others have reported upregulation of interferon γ and transforming growth factor β in CRSsNP.14 Although we did not assay for these cytokines in this study, they remain an area of further investigation. Previously, TNF-α was linked to olfactory dysfunction in a murine model10; however, we were unable to find any correlation between TNF-α and objective olfaction in this human study. It is unclear whether this is due to differences in species, use of a genetic model, or need for more precise clinical classification of the various types of CRS.
One important question left unanswered by this study is the location of inflammatory mediator production. One possibility is that mucus collected in the olfactory cleft is simply spillover from that produced in the sinus mucosa proper. In this scenario, the olfactory mucus would resemble mucus collected from the sinuses proper, and its presence or absence in the olfactory cleft would be related to patient-specific differences in anatomy and mucous trafficking. Another possibility is that these mediators are secreted from the olfactory epithelium itself, reflecting direct ongoing inflammation of this specialized epithelium. A better understanding of these questions will require future studies that simultaneously collect mucus from the olfactory cleft and various locations in the sinuses proper. Although we have previously found a significant correlation between cytokine values in middle meatal mucus and formal ethmoid mucosal biopsy specimens,9 repeating this investigation in the olfactory cleft would allow insight into the degree to which findings in olfactory mucus correlate with formal olfactory biopsy specimens.
Although these findings should be considered preliminary, the quantification of inflammatory measures in olfactory mucus appears to be a promising technique. Currently, there are limited means of classifying patients with olfactory loss in the setting of CRS beyond simple polyp status. Furthermore, prognostication from an olfactory standpoint remains limited in CRS, such that physicians are rarely able to offer patients any degree of certainty with regard to olfactory outcomes after treatments, such as sinus surgery, or targeted therapy, such as anti–IL-5 monoclonal antibodies. Certainly, it is worth exploring whether measurement of olfactory cleft inflammatory measures can help classify olfactory dysfunction in the setting of CRS into clinically meaningful groups, informing treatment selection and overall olfactory outcomes.
An array of cytokines, chemokines, and other inflammatory mediators can be detected in mucus collected from the olfactory cleft in patients with CRS, with variability across individual patients and groups with and without polyps. A significant inverse correlation was found between IL-5 levels and objective olfactory function in all patients with CRS, whereas IL-6, IL-7, and VEGF-A were positively correlated with olfaction only in those with CRSwNP. Further study is necessary to confirm these findings, investigate causality, and determine whether these measures are of discriminatory or prognostic value.
Accepted for Publication: March 28, 2016.
Corresponding Author: Zachary M. Soler, MD, MSc, Department of Otolaryngology–Head and Neck Surgery, Medical University of South Carolina, 135 Rutledge Ave, Mail Stop Code 550, Charleston, SC 29425 (email@example.com).
Published Online: May 26, 2016. doi:10.1001/jamaoto.2016.0927.
Author Contributions: Dr Soler had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition, analysis, or interpretation of data: Schlosser, Mulligan, Hyer, Soler.
Drafting of the manuscript: Schlosser, Mulligan, Hyer, Soler.
Critical revision of the manuscript for important intellectual content: Schlosser, Hyer, Karnezis, Gudis, Soler.
Statistical analysis: Hyer, Soler.
Obtained funding: Mulligan, Soler.
Administrative, technical, or material support: Schlosser, Mulligan, Karnezis, Gudis.
Study supervision: Schlosser, Mulligan, Karnezis.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Schlosser reported receiving grant support from OptiNose and IntersectENT and working as a consultant for Olympus and Meda. Dr Soler reported receiving support from grant R03 DC013651-01 from the National Institute on Deafness and Other Communication Disorders and working as a consultant for Olympus. No other disclosures were reported.