Abstract
Objective
To compare systemic allergen sensitivity and local allergen sensitivity in the sinonasal tissue of patients with a recently identified subtype of chronic rhinosinusitis strongly associated with allergy: central compartment atopic disease (CCAD).
Study Design
Prospective cohort study.
Setting
Academic tertiary care rhinology clinic.
Methods
Fifteen participants with endoscopic and radiographic evidence of CCAD underwent systemic allergy testing with skin testing and measurement of serum specific immunoglobulin E (sIgE) to 15 regionally common aeroallergens. Local allergen sensitivity was determined by measuring sIgE to these same 15 allergens in their sinonasal tissue. sIgE testing was performed by ImmunoCAP assay.
Results
Of the 15 participants, 14 were sensitive to at least 1 allergen locally in the central compartment and systemically on skin or serum testing. Among all participants, 4 were sensitive to allergens on central compartment sIgE testing that they were not sensitive to on skin and serum sIgE testing (range, 1-8 discordant allergens). Comparisons between local and systemic aeroallergen sensitivity results showed statistically significant correlations (P < .05) ranging from weak to strong.
Conclusion
Systemic allergy testing is recommended in the initial workup for CCAD. Local allergen sensitivities may be present in a subset of patients with CCAD. Further study of the clinical significance of these sensitivities should be undertaken in CCAD, with evaluation of the role of medical therapies and allergen immunotherapy in the treatment of CCAD.
Keywords
Chronic rhinosinusitis (CRS) has traditionally been characterized phenotypically by the presence or absence of nasal polyps.1,2 The association of CRS with another common sinonasal inflammatory disorder, allergic rhinitis (AR), has long been questioned. The inflammatory signatures of certain endotypes of CRS and AR seem to provide a plausible biologic basis for an association between these conditions.3,4 Numerous clinical studies of this possible association have been undertaken. A comprehensive review in 2014, however, revealed contradictory findings. 5
The role of allergy is even debated in allergic fungal rhinosinusitis (AFRS), which was classically defined by Bent and Kuhn as occurring only in patients with systemic type I hypersensitivity to inhaled fungal elements. 6 One study compared patients with AFRS with patients who had similar phenotypes (nasal polyps and eosinophilic mucus) but no fungal colonization/allergy, and it found no significant clinical or immunologic difference between the groups. 7 This same research group and others, however, identified a subset of these patients without immunoglobulin E (IgE) to fungi in their serum but with fungal IgE in their eosinophilic mucus and sinonasal tissue.8,9 When patients have symptoms of AR but negative results on traditional systemic allergy testing, they are diagnosed with non-AR (NAR). 10 However, in some patients with NAR, allergen specific IgE (sIgE) is present locally. Powe et al described this as local AR (LAR). 11 A recent meta-analysis found that 7.4% to 13.4% of patients with NAR had sIgE locally in the nasal cavity. 12 Allergen immunotherapy (AIT) was shown in a small series to be effective in the treatment of LAR. 13
A recently described CRS subtype, central compartment atopic disease (CCAD), is characterized by polypoid changes of the central sinonasal compartment (the middle and superior turbinates and/or the posterior-superior nasal septum) with unique radiographic findings: central opacification in the nasal cavities with peripheral clearing of the paranasal sinuses.14-17 The hallmark of CCAD is its strong association with allergy. Studies have shown allergy sensitization in 74% to 100% of patients with CCAD.14-16,18 The presence of multifocal middle turbinate edema has a specificity of 94.7% for inhalant allergy. 19 It is thought that chronic exposure of the central compartment to inhalant allergens via nasal airflow leads to polypoid inflammatory changes in exposed mucosa.20,21
The studies of allergy in CCAD used systemic measures of allergy status—specifically, skin testing and/or in vitro serologic testing of allergen sIgE.14-16,18,19 Local allergen sensitivity in the sinonasal tissue of patients with CCAD has not yet been studied, although correlation between local and systemic allergen sensitivity is frequently assumed. We therefore hypothesize that (1) local and systemic allergen sensitivity is correlated in CCAD and (2) local allergen sensitivity is present without evidence of systemic sensitivity in a subset of patients.
Methods
Approval was obtained from the Emory University Institutional Review Board prior to initiation of the study (No. 00114540). Consecutive patients of the Emory Sinus, Nasal and Allergy Center (October 2019–September 2020) who were ≥18 years of age and had radiographic and endoscopic evidence of CCAD were identified for recruitment.15,17 Patients previously diagnosed with CCAD undergoing routine follow-up during this time were also recruited. Typical CRS and allergy symptomatology was assessed in the patient’s history. Patients currently undergoing treatment with biologic therapy or those who were pregnant were excluded from the study. Patients currently undergoing treatment with AIT or those who previously completed a course of AIT were not excluded from the study, as they could still show sensitivity on allergy testing and were presenting with subjective and objective manifestations of CCAD. Patients who had undergone nasal or sinus surgery as well as those receiving medical therapy with topical intranasal corticosteroids or antihistamines were included. Patients were counseled about the research protocol, and written informed consent was obtained. No participants were harmed, and there were no adverse events.
Local aeroallergen sensitivity in the central compartment was assessed by measuring sIgE levels to 15 common regional aeroallergens with the Local Antibody Mucosal Brush Diagnostic (LAMB-Dx) test (Immunovent, LLC). Under endoscopic visualization with topical anesthesia, the bilateral areas of the central compartment with the most prominent mucosal edema or polyposis were sampled with the LAMB-Dx mucosal brush. The contents of the LAMB-Dx brush were suspended in phosphate buffered saline in an Eppendorf tube, which was then sealed and stored at −80 °C. Samples were processed and tested by Immunovent, LLC, which measures sIgE with the ImmunoCAP system for the selected aeroallergens with testing performed by Phadia Immunology Reference Laboratory (Thermo Fisher Scientific Inc).
Systemic allergen sensitivity was assessed with 2 methods: in vitro serologic testing and skin testing. Participants with medical contraindications to skin testing (current beta-blocker use, severe or uncontrolled asthma, dermatographism, prior anaphylaxis to skin testing) underwent serologic testing only. Serologic testing was performed by measuring sIgE levels in participants’ serum to the same 15 regionally significant aeroallergens measured with the LAMB-Dx mucosal brush test. Serum was obtained through standard venipuncture and laboratory processing practices and was performed by personnel from the Emory Medical Laboratory and study staff. Samples were stored at −80 °C. Serum samples were also tested by Phadia Immunology Reference Laboratory via the ImmunoCAP system. In addition, systemic inflammatory markers were obtained: serum total IgE and serum eosinophil count.
Allergy skin testing was performed in the Emory Sinus, Nasal and Allergy Center per the modified quantitative testing (MQT) method.22,23 The standard clinical testing panel with 28 antigens was used, as well as positive (histamine) and negative (glycerin) controls (Stallergenes Greer Inc). The 15 antigens included in serologic and local sIgE testing were part of the 28-antigen skin test panel. Medications that interfere with allergy skin testing, such as oral and intranasal antihistamines, were stopped prior to testing. 24
Allergy testing by the MQT method is a semiquantitative skin-testing technique that uses a combination of skin prick testing (SPT) and selected intradermal tests to arrive at an endpoint.22,23 The endpoint is a numeric value (0-6) that signifies the degree of reactivity to a specific allergen on skin testing, with 6 being the most reactive. The endpoint is classically delineated when an intradermal injection grows 2 mm as compared with a negative wheal or 2 mm greater than the negative control. MQT employs a combination of SPT wheal measurements and the results from specific dilutions of intradermal tests to arrive at an endpoint for each allergen tested. This decreases the number of intradermal injections as compared with what would be required in classic intradermal dilutional testing or skin endpoint titration.
The skin-testing endpoint serves as a marker for immunotherapy vial mixing, indicating a safe starting dose for subcutaneous immunotherapy. As sublingual immunotherapy (SLIT) does not require an extended escalation phase, quantitative allergy skin testing is not required, and many patients opt for nonquantitative SPT alone without added intradermal testing. Participants recruited for this study who had previously undergone allergy testing in preparation for SLIT had only SPT wheal measurements available. To reduce risk to these participants, repeat skin testing was not performed. Positive results were recorded as an endpoint of 6 or 3 to 5 and negative results as an endpoint of 0. Results with an endpoint between 3 and 5 were extrapolated to an endpoint of 3, the least reactive endpoint still indicating a positive result.
For purposes of data analysis, allergy skin testing by MQT or extrapolation of SPT was considered positive if an endpoint ≥3 was obtained. Serum testing was considered positive if a result was class II or greater (≥0.7 kUA/L). Local sIgE testing has an established cutoff in LAR (≥0.145 kUA/L), associated with a positive predictive value of 100% and a negative predictive value of 66.67%. 25
Statistical analyses were performed with G*Power version 3.1 (Heinrich-Heine-Universität) and SPSS Statistics version 26.0 (IBM Inc). Sample size analysis revealed that achieving a moderate effect size with an alpha level of 0.05 and a power of 0.8 based on a repeated measures analysis of variance would require 10 patients to achieve statistical significance. Comparisons between local and systemic aeroallergen sensitivity were performed with Spearman correlation; results were not normally distributed as assessed by visual inspection of normal Q-Q plots. Comparisons between median sIgE levels in the central compartment and serum were performed with a Wilcoxon signed rank test. Sensitivity and specificity of serum sIgE and central compartment sIgE were calculated by utilizing allergy skin testing as the gold standard.
Results
Fifteen participants were recruited for this study. Demographic and clinical characteristics are shown in Table 1 . Seven participants had previously undergone AIT with treatment lengths from 3 months to 4 years. Of these 7 participants, 6 had completed these treatments at least 8 years prior to the start of this study (range, 8 to >30 years).
Demographics and Clinical Characteristics (15 Participants).
Median (range).
Of the 15 participants, 9 underwent complete evaluation with all 3 forms of allergy testing described here: skin testing by MQT, serum sIgE by ImmunoCAP, and local sinonasal sIgE by ImmunoCAP. Two participants had undergone allergy SPT designed for SLIT and did not undergo repeat skin testing with the MQT method. They still had adequate data for skin, serum, and local sinonasal allergy test analysis. This testing was completed 10 months prior to recruitment for 1 participant, who did not proceed with initiation of SLIT. The other participant underwent testing 2.5 years prior, initiated SLIT at that time, and completed 2 years of therapy, at which time it was discontinued. He restarted SLIT 4 months prior to recruitment and was receiving treatment at the time of the study. Four participants had medical comorbidities that precluded them from allergy skin testing. Fourteen participants were using topical intranasal corticosteroids and/or antihistamines prior to study initiation.
Of 15 participants, 14 were sensitive to at least 1 allergen locally in the central sinonasal compartment and systemically on skin or serum testing. One participant was not sensitive to any of the tested allergens via local or systemic sIgE. Skin testing was unable to be completed in this participant due to medical comorbidities. Binary results for all participants are shown in Figure 1 . Of 14 participants who demonstrated systemic allergen sensitivity, 4 (28.6%) were sensitive to allergens on central compartment sIgE testing that they were not sensitive to on skin and serum sIgE testing (range, 1-8 discordant allergens). The median number of positive antigens on skin testing was 10, as compared with 6 on serum sIgE testing and 14 with central compartment sIgE testing ( Table 2 ). Of the 11 participants who underwent all 3 forms of allergy testing, 4 (36.4%) were sensitive to more antigens on local sIgE testing than on either form of systemic testing. Systemic inflammatory markers and quantitative testing results are presented in Table 3 . There was wide variability in these markers and results. Several participants had serum total IgE levels multiple times the upper limit of normal, while 1 had a level higher than the limit of the assay (5000 kU/L). The median sIgE level for the central compartment was 0.24 kUA/L, as compared with 0.14 kUA/L for serum, a statistically significant difference (z = 5.06, P < .001).
Aggregate Number of Positive Antigens by Testing Method for Each Participant.
Abbreviation: sIgE, specific immunoglobulin E.
Skin testing performed in preparation for sublingual immunotherapy and full quantitative results not available.
Participant unable to undergo skin testing due to contraindication.
Systemic Inflammatory Markers and Systemic/Local Allergy Testing Results.
Abbreviations: IQR, interquartile range; sIgE, specific immunoglobulin E.

Local and systemic allergy testing results (a/b/c): skin (a), serum sIgE (b), and central compartment sIgE (c). Shading indicates concordance: green, total concordance; red, discordance; yellow, negative skin and serum sIgE, positive central compartment sIgE. sIgE, specific immunoglobulin E.
Comparisons between local and systemic aeroallergen sensitivity showed correlations ranging from weak to strong; all were statistically significant ( Table 4 ). The strongest correlation was between central compartment and serum sIgE test results (ρ = 0.698, P < .0005). The correlation between skin testing and local central compartment sIgE testing results was weakest yet still statistically significant (ρ = 0.353, P < .0005). Also shown in Table 4 is a subgroup analysis to evaluate if prior AIT or surgery could influence these results. Correlations were similar between participants who had undergone AIT and those who were untreated. Correlations were somewhat stronger in patients who had not undergone surgery as compared with those who had. With skin testing as the current gold standard, the sensitivity and specificity of serum and central compartment sIgE testing are shown in Table 5 .
Spearman Correlation Between Local Markers of Allergy With Subgroup Analyses Based on Immunotherapy and Surgical Status. a
Abbreviations: AIT, allergen immunotherapy; sIgE, specific immunoglobulin E.
Each correlation: P < .01 (unless noted otherwise).
P < .05.
Sensitivity and Specificity: Serum and Central Compartment sIgE vs Skin Testing as Gold Standard.
Abbreviation: sIgE, specific immunoglobulin E.
Discussion
The mainstay of treatment for CRS has been endoscopic sinus surgery and topical corticosteroid therapy. While highly effective in many patients, this was likely initially driven by a limited understanding of the pathophysiology of the disease and its subtypes/endotypes. The most widely used phenotypic classification of CRS, with and without nasal polyps, illustrates this. This classification system masks a far more complex picture, as CRS encompasses many disease states and inflammatory profiles. The question of whether allergy is implicated in the pathogenesis of CRS has likely been obscured by this classification system.
More accurately, allergy is likely an element in the pathogenesis of some subtypes of CRS. CCAD is one of these subtypes. The question of whether local allergen sensitivity may play a role in the pathogenesis of CCAD has not been studied. This study aimed to assess the local allergen sensitivity profile of the central sinonasal compartment and compare this with systemic allergen sensitivity profiles. Understanding this association or lack thereof will provide the foundation for study of the role of AIT in the treatment of patients with CCAD.
In the current study, participants with radiographic and endoscopic evidence of CCAD underwent systemic (skin and/or serum sIgE) and local sinonasal allergy testing to 15 commonly encountered aeroallergens. Participant demographics and comorbidities (asthma) were similar to prior studies of patients with CCAD.15,26 The same assay and reference laboratory were used for both forms of sIgE testing. Of 15 participants, 14 (93.3%) had evidence of allergen sensitivity on at least 1 type of testing, in line with previous reports of CCAD.14-16 In the 11 participants who completed all 3 types of testing, 8 (72.7%) were sensitive to allergens on all 3 forms. Quantitative results demonstrated statistically significant correlations among the tests, the strongest between serum and local sIgE testing. Correlations remained statistically significant and similar to overall correlations when subgroup analysis was performed according to prior immunotherapy treatment and surgical status. The median sIgE level for the central compartment (0.24 kUA/L) was higher than that for serum (0.14 kUA/L), demonstrating a statistically significant difference (z = 5.06, P < .001). Increased levels of sIgE in the central compartment vs serum is further evidence that CCAD is an exuberant local manifestation of allergy.
Of 14 participants, 4 (28.6%) who demonstrated systemic allergen sensitivity were positive to allergens on local sIgE testing but negative on skin and serum sIgE testing, evidence of local allergen sensitivity. This may be another piece of evidence indicating that sinonasal tissue is producing sIgE locally in CCAD. The clinical significance of these results deserves investigation in future studies. Interestingly, 6 participants in this cohort had undergone AIT for AR for at least a year, including 2 of the 4 participants with evidence of possible local allergy. Local allergy, inadequate length of AIT, development of new allergen sensitivities, and/or a lack of topical corticosteroid treatment may contribute to this progression. All may be important considerations in the treatment of patients with CCAD. Systemic allergy testing has been recommended in the workup for CCAD. 16 Based on these findings, local allergy testing may be an important adjunct in the workup.
The sensitivity of serum sIgE testing and the specificity of central compartment sIgE testing are lower than desired. However, these calculations rely on the assumption that the gold standard test—allergy skin testing in this case—is truly a gold standard. While skin testing is the most widely accepted allergy test, the presence of LAR indicates that this may not be the case. Further study of the clinical significance of local allergen sensitivities should be undertaken with study of the role of medical therapies in the treatment of CCAD (AIT, topical antihistamines and corticosteroids, and oral antihistamines). Additionally, while the role of AIT based on possible local allergen sensitivity has been investigated, this should be studied in patients with CCAD and evidence of LAR. 13
There are limitations to this study. The cut point for a positive vs negative local sIgE test result is not well established. Some studies have utilized the assay’s limit of detectability (0.1 kUA/L) as the cut point, while others have determined it to be 0.12 kUA/L.27,28 This study utilized a more conservative value from the literature on LAR, ≥0.145 kUA/L. 25 The testing methodology from the LAR literature was not the LAMB-Dx test, as used in this study, but rather the ImmunoCAP assay. False-positive results are also possible. It could be that the positive antigens on local sIgE testing of the central compartment but negative on both systemic tests could be false positive results. Additional studies of patients with LAR and CCAD with larger sample sizes would help establish if these findings are true results and not simply false positives.
Other limitations in this study were that certain patients had undergone nonquantitative SPT designed for SLIT and were not retested; furthermore, those with medical comorbidities contraindicating allergy skin testing underwent only serum and local sIgE testing. Nonetheless, the results showed significant correlations between systemic and local allergy testing, increased levels of sIgE in local sinonasal tissue as compared with serum, and allergen sensitivities at the local level that were not detected systemically in patients with CCAD.
Conclusion
Systemic allergy testing is recommended in the initial workup for CCAD. Local allergy may be present in a subset of patients with CCAD. Further study of the clinical significance of local allergen sensitivities should be undertaken in CCAD, with evaluation of the role of medical therapies and AIT in the treatment of CCAD.
Footnotes
Acknowledgements
We recognize William Reisacher, MD, and Immunovent, LLC, for technical assistance with local allergen sensitivity testing.
This article was presented at the 2020 Annual Meeting of the American Academy of Otolaryngic Allergy; October 23, 2020 (virtual).
