Contents lists available at ScienceDirect

Ann Allergy Asthma Immunol 111 (2013) 337e341

Characterization and treatment of patients with chronic rhinosinusitis andnasal polypsCynthia R. Fountain, MA *; Pamela A. Mudd, MD y; Vijay R. Ramakrishnan, MD y; Stefan H. Sillau, MS z;Todd T. Kingdom, MD y; and Rohit K. Katial, MD x

*University of Colorado School of Medicine, Aurora, ColoradoyDepartment of Otolaryngology, University of Colorado, Aurora, ColoradozUniversity of Colorado, Aurora, ColoradoxDivision of Allergy and Immunology, National Jewish Health, Denver, Colorado

A R T I C L E I N F O

Article history:Received for publication April 30, 2013.Received in revised form July 10, 2013.Accepted for publication July 13, 2013.

A

BaOti

Reprints: Rohit K. Katial, MD, Professor of MedicCO 80206; E-mail: KatialR@njhealth.org.Disclosures: Authors have nothing to disclose.Funding Sources: None.

1081-1206/13/$36.00 - see front matter � 2013 Ahttp://dx.doi.org/10.1016/j.anai.2013.07.017

B S T R A C T

ackground: Patients with chronic rhinosinusitis (CRS) and nasal polyps (NPs) may be subdivided intospirin-sensitive (AS) and aspirin-tolerant (AT) populations. These cohorts are not well characterized.bjective: To examine phenotypic characteristics and determine the extent of medical/surgical interven-ons in patients with CRSþNP and to compare the AS with the AT subset in the CRSþNP sample.

Methods: Retrospective chart review was performed at a tertiary academic respiratory hospital. Dataincluded patient demographics, asthma severity, peripheral eosinophilia, Lund-Mackay computed tomo-graphic score, symptomatic dysosmia, and therapeutic interventions.Results: Of the 182 patients included, 81 had aspirin sensitivity (45%) and 101 had aspirin tolerance (55%).Asthma was present in 94% of patients with CRSþNP (100% in AS subgroup vs 89% in AT subgroup, P ¼ .001).Eighty-eight percent of the CRSþNP sample had moderate to severe persistent asthma. In the AS and ATsubgroups, asthma severity was similar (P > .6). The CRSþNP sample showed a mean computed tomographicscore of 14.0 (44% with eosinophilia and 46% with dysosmia). More severe sinus disease was noted in the ASgroup (Lund-Mackay computed tomographic scores, P ¼ .002; olfactory symptoms, P ¼ .001). Serumeosinophil levels were not statistically different between groups (51% in AS group, 39% in AT group, P > .1).Conclusion: This study is one of the broadest reviews of patients with CRSþNP, with unique findings in thehigh prevalence of asthma in AS and AT patients, greater olfactory dysfunction in AS patients, and a minorityof patients with CRSþNP and circulating eosinophils. Most AS patients do not have increased circulatingeosinophils, as is often believed. These results shed further light on the association between asthma andupper respiratory tract disease in those with nasal polyposis.� 2013 American College of Allergy, Asthma & Immunology. Published by Elsevier Inc. All rights reserved.

Introduction

Nasal polyps (NPs) are found in approximately 20% of all patientswith chronic rhinosinusitis (CRS), and CRS plus NPs (CRSþNP)affects 0.5% to 4% of the population worldwide.1 Although thepathophysiology underlying NP formation is not fully understood,systemic, local, and environmental factors appear to contribute. Themost abundant inflammatory cell population within polyp tissue isconstituted by eosinophils, and most are activated. As such,systemic eosinophilia may be associated with CRSþNP.2 CRSþNP isoften associated with asthma2e4 and frequently manifests a severedisease phenotype requiring multiple hospitalizations, medical

ine, 1400 Jackson Street, Denver,

merican College of Allergy, Asthma &

interventions, and surgeries.5e8 A subgroup of patients withCRSþNP is intolerant of nonsteroidal anti-inflammatory drugs,developing a profound respiratory reaction after the ingestion ofcyclooxygenase-1einhibiting nonsteroidal anti-inflammatorydrugs. This aspirin-sensitive (AS) disorder is more refractory totreatment compared with aspirin-tolerant (AT) control patientswith CRSþNP.8,9

Although several studies have investigated the clinical charac-teristics of patients with CRSþNP, most have included smallsamples and/or limited characterizations of those samples(Table 1). The primary objectives of the present investigation wereto examine phenotypic disease characteristics and severity and todetermine the extent of medical and surgical interventions inpatients with CRSþNP, all within a single large sample, who wereselected from the patient population at National Jewish Health(Denver, Colorado). A secondary objective was to compare the ASand AT subsets of the CRSþNP sample.

Immunology. Published by Elsevier Inc. All rights reserved.

Table 1Studies describing patients with CRSþNP

Study Patients with CRSþNP (AS/AT) Clinical characteristics evaluated

Alobid et al13 (2011) 245 (59/186) asthma, CT scores, olfactory dysfunctionAwad et al5 (2008) 91 (41/50) asthma, CT scores, medical therapies, ER visitsBatra et al9 (2003) 17 (9/8) asthma, CT scores, medical therapies, tissue inflammation/edemaCeylan et al6 (2007) 19 asthma, medical therapiesKim and Kountakis8 (2007) 208 (10/198) asthma, surgeries, NP recurrenceKonstantinidis et al18 (2010) 27 olfactory dysfunctionLoehrl et al21 (2006) 34 (34/d) asthma, medical therapies, ER visitsMcMains and Kountakis16 (2006) 125 (15/110) CT scores, medical therapies, surgeries, NP recurrenceMendelsohn et al7 (2011) 549 (63/486) asthma, NP recurrence, surgeriesPoznanovic et al22 (2007) 303 CT scores, peripheral eosinophiliaSeybt et al4 (2007) 40 asthma, surgeries, olfactory dysfunctionSzczeklik et al20 (2000) 500 (500/d) asthma, medical therapies

Abbreviations: AS, aspirin sensitivity; AT, aspirin tolerance; CRS, chronic rhinosinusitis; CT, computed tomographic; ER, emergency room; NP, nasal polyps.

Table 2Presence and severity of asthma in patients with chronic rhinosinusitis and nasalpolyps

Asthma severity n (%) AS (%) AT (%) P value

Intermittent 9 (5) 4 (5) 5 (6)Mild persistent 12 (7) 4 (5) 8 (9)Moderate persistent 41 (25) 18 (22.5) 23 (26)Severe persistent 105 (63) 54 (67.5) 51 (59) .62a

Total 171 (94)b 81 (100) 90 (89) .001c

Abbreviations: AS, aspirin sensitivity; AT, aspirin tolerance.aFisher exact test.bAsthma severity could not be determined from the charts of 4 patients.cc2 test.

C.R. Fountain et al. / Ann Allergy Asthma Immunol 111 (2013) 337e341338

Methods

This study was approved by the National Jewish Health insti-tutional review board. A retrospective chart review was performedof patients treated at a tertiary care respiratory hospital from 2008to 2010. The medical records database for International Classifica-tion of Diseases, Ninth Revision codes corresponding to CRSþNP(471.9 and 473.9) was queried, and patients were included in thestudy if the medical chart review confirmed the diagnoses of CRSand NPs. Patients were excluded if they had a diagnosis of cysticfibrosis or immunodeficiency. One hundred eighty-two patientswere included in the final analysis.

Data were collected on patient demographics, presence andseverity of asthma as defined by the National Heart, Lung, andBlood Institute,10 absolute peripheral eosinophil counts, Lund-Mackay computed tomographic (CT) staging score,11 and presenceof dysosmia/anosmia. Eosinophilia was defined as a serum eosin-ophil level higher than 0.45 � 109 cells/L, the upper limit of normalin the authors’ laboratory. If multiple laboratory values wereavailable for a given patient, the largest value was used for statis-tical comparison, as a representative marker for maximal diseaseseverity. CT scans of the paranasal sinuses were scored by a singleresearcher who was blinded to the remaining data using thescoring system (0, no opacity; 1, partial opacity; 2, total opacity foreach sinus and ostiomeatal unit) of Lund and Mackay,11 resulting ina bilateral total score of 0 to 24. If a patient had multiple CT imagesavailable, the highest score was used for statistical analysis. Ifanosmia/dysosmia was noted in the medical chart, then “yes,olfactory symptoms were present” was recorded in data collection.If the chart noted a patient denying decreased sense of smell, or ifthe chart neglected to mention olfactory symptoms, then thepatient was recorded as not having olfactory symptoms. Thisdecision minimizes potential differences, making apparent signif-icant differences truly significant, as discussed further below.

To determine the extent of medical and surgical interventions,steroid-saline nasal rinse use (budesonide inhalation suspension0.5 mg/2 mL), leukotriene modifier use, and long-term oral corti-costeroid (OCS) use were recorded as “yes” or “no” for each chart. Inaddition, the number of oral steroid bursts per 12-month period(only for those not using long-term OCSs) and the number of sinussurgeries were recorded. Leukotriene modifier use was affirmed ifmontelukast sodium or zileuton use was noted in the chart. Steroidbursts were defined as courses lasting shorter than 4 weeks. Long-term corticosteroid use was defined as any oral course beyond thescope of a burst.

During analysis, percentages for the presence of aspirin sensi-tivity, asthma, eosinophilia, olfactory symptoms, steroid-salinerinse use, leukotriene modifier use, long-term OCS use, and sinussurgery use were determined for the CRSþNP cohort. For thosepatients with CRSþNP and asthma, percentages of patients with

intermittent, mild persistent, moderate persistent, and severepersistent asthma were determined. For parametric variables, suchas Lund-Mackay CT scores, absolute serum eosinophils, number ofannual steroid bursts, and number of surgeries (in those patientshaving undergone surgery; those without surgery were excluded),mean values, standard deviations (SDs), standard errors of themean, and 95% confidence intervals were calculated.

Statistical analysis of the AS vs AT subgroups was performed bycomparing continuous variables using t tests. Serum eosinophils,steroid bursts, and number of surgeries exhibited non-gaussiandistributions; so nonparametric Wilcoxon tests were performed.To correct the skew, square root transformations were applied tothese variables and t tests were performed on the transformedvariables. Associations between AS and AT subgroup categoricalvariables were tested using c2 tests. Associations between AS andAT asthma severity were tested using Fisher exact tests owing tosmaller samples within each category.

Results

Asthma Presence and Severity

The overall CRSþNP cohort included 182 patients from NationalJewish Health (mean age 52.4 years, range 11e88 years). Fifty-twopercent (n ¼ 94) of patients were men. Ninety-four percent(n¼ 171) of the CRSþNP cohort (n¼ 182) were asthmatic, of whom167 had asthma severity documented (Table 2). Sixty-three percent(n ¼ 105) of these patients had severe persistent, 25% (n ¼ 41) hadmoderate persistent, 7% (n¼ 12) hadmild persistent, and 5% (n¼ 9)had intermittent asthma.

Sinus Disease

Sinus disease burden showed a mean Lund-Mackay CT score of14.0 (SD 6.2; Table 3). Eosinophilia was found in 44% (n ¼ 58 of 131charts with eosinophil data), with a mean serum eosinophil level of0.51 � 109 cells/L (SD 0.59, normal range 0.0e0.45 � 109 cells/L;Tables 3 and 4).

Table 5Interventions

Total (%) AS (%) AT (%) P valuea

Steroidesaline rinses 102/182 (56) 48 (59) 54 (53) .43Leukotriene modifier use 115/182 (63) 61 (75) 54 (53) .002Long-term oral corticosteroid use 42/182 (23) 18 (22) 24 (24) .81Sinus surgery 138/182 (76) 66 (81) 72 (71) .11

Abbreviations: AS, aspirin sensitivity; AT, aspirin tolerance.ac2 test used for statistical comparison.

Table 6Interventions measured using continuous data

n Mean (difference) CI P valuea

Annual steroid burstsCRS+NP 140 1.85 1.49e2.21AS 63 2.59 1.96e3.21AT 77 1.25 0.88e1.61

Table 3Sinus CT scores and eosinophil counts

Lund-Mackay CT scoreSerum eosinophillevel ( � 109 cells/L)

n Mean CI n Mean CI

CRSþNP 149a 14.0 13.0e15.0 131a 0.51 0.41e0.61AS 69 15.7 14.3e17.0 61 0.60 0.41e0.79AT 80 12.6 11.2e13.9 70 0.44 0.34e0.53Percentage ofdifference (P)

22 (.002)b 1.14e5.04 31 (.14)c �0.05 to 0.37

Abbreviations: AS, aspirin sensitivity; AT, aspirin tolerance; CI, confidence interval;CRS, chronic rhinosinusitis; CT, computed tomographic; NP, nasal polyps.aThe number for Lund-Mackay scores is smaller than the overall sample size (182)because only 149 charts had CT imaging available to be scored. Similarly, thenumber for serum eosinophils is 131 because only 131 charts had absolute serumeosinophils recorded.bStatistical analysis performed with t test.cStatistical analysis performed with t test and confirmed with Wilcoxon test andsquare root transformations applied to the t test.

C.R. Fountain et al. / Ann Allergy Asthma Immunol 111 (2013) 337e341 339

Evidence of olfactory dysfunction was found in 46% (n ¼ 84;Table 4). In the remaining 54%, dysosmia was denied at inquiry orwas not queried of the patient.

Steroid-saline nasal rinses were used in 56% (n ¼ 102), leuko-triene modifiers in 63% (n ¼ 115), long-term OCSs in 23% (n ¼ 42),and surgery in 76% (138) of the 182 patients with CRSþNP (Table 5).All 182 charts were queried for nasal corticosteroid use, and 116commented on their use. Of these 116, 105 were reported to betaking commercial nasal corticosteroids. One hundred thirty-eightof 182 patients underwent surgery, averaging 2.48 previoussurgeries in those who underwent surgery (those who did notundergo surgery were excluded; Table 6). In the 77% withouta history of long-term OCS use (n ¼ 140), the average number ofsteroid bursts per year was 1.85.

Aspirin Tolerant vs Aspirin Sensitive

Subgroup analysis was performed in 81 AS patients (45%) and101 AT patients (55%). AS patients showed a higher prevalence ofasthma (100% vs 89% in AT group, P ¼ .001; Table 2). Similarpercentages of patients with intermittent, mild persistent,moderate persistent, and severe persistent asthma were found inthe AS and AT subcohorts (P > .6). The vast majority of AS and ATasthmatics had severe persistent asthma (67.5% and 59%).

Markers of disease severity, such as Lund-Mackay CT scores andolfactory symptoms, suggested more extensive disease in the ASgroup (Tables 3 and 4). Lund-Mackay CT scores were 22% higher inthe AS group (mean score 15.7 vs 12.6, P ¼ .002). Olfactorysymptom prevalence was 59% vs 36% in the AS vs AT group(P ¼ .001). Eosinophilia was found in 51% of AS patients (mean0.60� 109 cells/L, SD 0.75) and 39% of AT patients (mean 0.44� 109

cells/L, SD 0.39), showing a 31% difference in absolute eosinophils.Neither differences in the percentage of patients with eosinophilia(P ¼ .16) nor differences in mean absolute eosinophils between the2 groups (P¼ .14) proved to be statistically significant, although thetrend was toward higher eosinophil levels in the AS group.

In general, medical and surgical interventions were performedmore frequently in AS patients, including frequency of steroid

Table 4Presence of eosinophilia and olfactory dysfunction

n (%) AS (%) AT (%) P valuea

Eosinophilia 58/131 (44) 31 (51) 27 (39) .16Olfactory dysfunction 84/182 (46) 48 (59) 36 (36) .001

Abbreviations: AS, aspirin sensitivity; AT, aspirin tolerance.ac2 test used for statistical comparison.

bursts (2.6 vs 1.3 annually, P < .001), prior surgeries (3.0 vs 2.0,P ¼ .007), and leukotriene modifier use (75% vs 53%, P ¼ .002;Tables 5 and 6). In contrast, steroid-saline nasal rinse use (P ¼ .43),long-term OCS use (P ¼ .81), and a history of sinus surgery (P ¼ .11)did not occur at a statistically different frequency between the ASand AT cohorts. Despite similar percentages of patients witha history of at least 1 prior surgery (AS 81%, AT 71%), of those whounderwent surgery, AS patients underwent significantly moresurgeries.

Discussion

This study adds to the literature the largest chart review ofpatients with CRSþNP that looks at phenotypic disease character-istics and severity, in addition to the extent of medical and surgicalinterventions in patients with CRSþNP, all within in a singlesample. No other study has reported a high prevalence of asthma inpatients with CRSþNP and in the AT subgroup. It is important torealize that patients with CRP plus nasal polyposis were investi-gated specifically as opposed to all patients with CRS, as most otherstudies have done. Furthermore, this broad analysis reports thepercentages (39%e51%) of patients with CRSþNP and increasedcirculating eosinophils. Such data have not previously beenreported and shed further light on the association between asthmaand upper respiratory tract disease in those with nasal polyposis.However, there is the perception that most AS patients may haveincreased circulating eosinophils,12 but the present data suggestedotherwise (51%).

In the CRSþNP sample, 94% of patients had asthma, 88% ofwhom had moderate to severe persistent asthma. In comparison,Alobid et al13 described 53% asthmatics among 245 patients withCRSþNP, of whom 55% had moderate to severe persistent asthma.Seybt et al4 found only 40% asthmatics in their sample of40 patients with CRSþNP.

Perhaps the substantially larger percentage of CRSþNP asth-matics and of moderate to severe asthmatics in the present studycan be attributed to several factors working in concert. First,a skewed population at the authors’ tertiary academic respiratory

AS � AT 1.34 (70%) 0.63e2.06 <.001Number of surgeriesCRS+NP 138 2.48 2.14e2.82AS 66 2.97 2.34e3.60AT 72 2.03 1.75e2.30AS � AT 0.94 (38%) 0.26e1.62 .007

Abbreviations: AS, aspirin sensitivity; AT, aspirin tolerance; CI, confidence interval;CRS, chronic rhinosinusitis; NP, nasal polyps.aStatistical analysis performed with t test and confirmed with Wilcoxon test andsquare root transformations applied to the t test.

C.R. Fountain et al. / Ann Allergy Asthma Immunol 111 (2013) 337e341340

hospital is likely. Nonetheless, patients with CRSþNP in this samplewere referred for upper respiratory tract disease, not asthma.Second, the authors selected a subgroup with CRS, namely thosewith NPs, rather than all comers. Third, the large percentage ofCRSþNP asthmatics may be attributed in part to a large percentage(45%) of AS patients, in contrast to the study of Alobid et al,13 whichhad only 24% AS patients. Moreover, the present results showed90% moderate to severe persistent asthma within the AS subgroup.Similarly, Ceylan et al6 reported 100%moderate to severe persistentasthma in 12 AS patients. As more AS patients are included in theCRSþNP sample, their highly prevalent moderate to severepersistent asthma will weigh more heavily on the CRSþNP group’soverall asthma severity. Nevertheless, it is apparent that the ATsubgroup also yielded a larger percentage (85%) with moderate tosevere asthma. The AT and AS groups would be expected to showa high prevalence of asthma because CRSþNP has been described asa systemic disease rather than a local process. Ten Brinke et al14

evaluated chronic sinus disease based on CT scores and reporteda correlation with increasing sinus scores and markers of asthma,such as functional residual capacity, exhaled nitric oxide, andcirculation eosinophils, all suggesting a systemic connection.

Eosinophil data, for absolute serum counts and the presence ofan elevated count, were collected. For 182 patients with CRSþNP,131 charts with eosinophil data yielded a mean of 0.51 �109 cells/L(<0.45�109 cells/L considered normal). Forty-four percent of thesepatients had peripheral eosinophilia (serum absolute count abovenormal range). Peripheral eosinophilia was found in 51% of ASpatients and 39% of AT patients. Thework of Zeitz12 has argued thatperipheral eosinophilia would be expected in most AS patients.However, unexpectedly, the authors found that only 44% of patientswith CRSþNP had increased eosinophils. In addition, they found nodifference in the AS and AT groups. This is an important finding for2 reasons. First, it suggests that not all cases of CRSþNP arenecessarily detectable by increased circulating eosinophils, assometimes believed. Second, it suggests that the disease basis inthe 2 groups (AS and AT) has a similar pathophysiology.

The present Lund-Mackay CT scores were compared with thosein the literature. McMains and Kountakis15 found that the meanpreoperative CT score in 125 patients undergoing revision sinussurgeries for CRS (with and without NPs) was 13.4 � 0.7. They alsoreported that patients with NPs and asthma had higher CT scoresthan those without.15 In accord with the finding of McMains andKountakis, the mean CT score in the present study was 14.0 � 1.0.There was a significant difference in AS vs AT CT scores (15.7 � 1.4vs 12.6 � 1.4, P ¼ .002), which is consistent with the literature.Awad et al5 and McMains and Kountakis16 found significantdifferences in their respective AS and AT groups (preoperatively, 19,SD 4.82 vs 14, SD 6.8, P ¼ .006; postoperatively, 14, SD 7.4 vs 5, SD2.08, P < .001 in the former and 20.2 � 1.9 vs 12.6 � 0.7, P < .001 inthe latter). Awad et al5 hypothesized that less improvement wasseen in the AS group compared with the AT group because thesinonasal disease in this subset is marked by more inflammatorycells and mediators, contributing to a more extensive and resistantdisease process, indicating the systemic nature of this disease.

In assessing disease impact, the authors looked at olfactorydeficits, finding that 46% of patients with CRSþNP developedsubjective dysosmia. In AS vs AT patients, 59% vs 36% (P¼ .001) hadolfactory dysfunction. In comparison, Kim and Kountakis8 found40% vs 23% with anosmia in AS vs AT patients and all patients inthat study had asthma. In contrast, Alobid et al13 found no signifi-cant differences between AS and AT olfactory dysfunction(as measured by Barcelona Smell Test 24 and a loss-of-smellquestionnaire). Differences between the present data and those ofAlobid et al may be attributed the use of a subjective measurement(patient query) in the present study and the use of an objectivemeasurement (Barcelona Smell Test 24) in the study by Alobid et al.

Nevertheless, Downey et al17 found that subjective measurementsof anosmia correlated with objective University of PennsylvaniaSmell Identification test results, thereby reinforcing the authors’argument that subjective measurements are valid and that dysos-mia is more prevalent in AS patients.

Several other studies have similarly argued that greater olfac-tory dysfunction is indicative of greater sinonasal disease. Kon-stantinidis et al18 showed that olfactory dysfunction correlatedwith inflammation severity in patients with polyps. Bhattachar-yya19 reported that dysosmia is predictive of radiographic sinusdisease. Downey et al17 found loss of smell to be predictive ofpersistent mucosal inflammation after sinus surgery. Therefore, it isreasonable to conclude that subjective reporting of olfactorydysfunction is another measurement of sinus disease severity, inaddition to CT scoring, and that AS patients experience more severesinus disease than AT patients.

Looking at interventions in the CRSþNP sample, the authorsfound that 56% used steroid-saline rinses, 63% used leukotrienemodifiers, and 23% used long-term OCSs. In comparison, 59% of theCRSþNP sample in the study by Batra et al9 used long-term OCSspreoperatively and 12% postoperatively. The present results forlong-term OCS use are within the pre- and postoperative ranges ofthose reported by Batra et al, indicating agreement because thepresent chart review did not take into account the timing ofsurgeries in relation to OCS treatment.9 Szczeklik et al20 and Loehrlet al21 looked at AS samples, noting 41% to 74% long-term OCS usevs 22% in the present sample. Szczeklik et al20 clearly defined long-term OCS use, but Loehrl et al21 did not and may have over-estimated long-term OCS use based on their study design. Thesedata clearly show high usage of adjunctive therapies in an attemptto improve disease control. However, these patients still undergoa significant number of surgeries, highlighting an unmet need.

Endoscopic sinus surgery was the final intervention investi-gated, recording a positive history of surgery and the number oflifetime surgeries per patient. The present results agree with 1study4 in the prevalence of any surgical history but suggest fewerprocedures per patient than in other studies.4,15 Seybt et al4

reported that 76% of asthmatics with CRS and NPs and thosewithout NPs underwent surgical treatment.4 The present study alsofound 76%, but in 94% of asthmatic patients with CRSþNP. Seybtet al4 reported a mean of 2.9 surgeries per asthmatic patient withCRS vs 2.5 per patient with CRSþNP in the present study. McMainsand Kountakis15 reported 1.9 � 0.1 surgeries per endoscopic sinussurgical patient. Given that the patients in the study by McMainsand Kountakis were about to undergo an additional surgery andthat their group included patients without NPs, whereas thepresent study did not, the authors deduce that the study byMcMains and Kountakis predicts larger numbers of surgeries forthe CRSþNP population compared with the present study.

Subgroup analysis of surgery data showed that 81% of AS and71% of AT patients had at least 1 surgery (P¼ .11). Similarly, Kim andKountakis8 reported that 80% of their AS sample had undergone atleast 1 endoscopic sinus surgery. Of the present 66 AS patients witha history of surgery, themean number of surgeries was 3.0; of 72 ATpatients, it was 2.0 (P ¼ .007). Supporting these results, Mendel-sohn et al7 found that AS patients were more likely to have recur-rence and undergo additional surgeries (odds ratio 2.7, confidenceinterval 1.5e3.2, P < .01). The present study and the study byMendelsohn et al clearly show that surgery is not curative in thiscohort and multiple procedures are undertaken owing to theunderlying systemic nature of the disease.

Limitations should be acknowledged. Given the retrospectivestudy design, it was often difficult to contextualize measurementssuch as Lund-Mackay scores, serum eosinophils, and olfactorysymptoms in relation to medical and surgical interventions.Nevertheless, coordinating oral steroids with CT imaging, serum

C.R. Fountain et al. / Ann Allergy Asthma Immunol 111 (2013) 337e341 341

eosinophils, or olfactory symptoms, for example, would haveamplified differences between the AS and AT subgroups. However,even without such coordination, significant differences were notedin CT scores and olfactory symptoms. In addition, although it isunlikely that symptoms of dysosmiawere queried and documentedconsistently by medical staff, counting patients with potentialdysosmia who were supposedly not queried about olfactorysymptoms as not having symptoms minimizes the prevalence ofdysfunction in the present study. The fact that the authors stillfound 46% with dysosmia in the overall CRSþNP sample anda significant difference in AS vs AT dysosmia, while minimizing anypotential difference through the study design, is noteworthy.Recording the total number of surgeries in a patient’s lifetime (vssurgeries per unit of time) may underestimate the data of youngerpatients with severe disease who eventually will undergo moresurgical procedures as their disease progresses. However, incom-plete surgical histories made efforts to determine surgeries per unittime a fruitless undertaking.

Given that all patients in the present sample were identifiedthrough their treatment at a tertiary care respiratory hospital,a referral bias likely exists, because many may have been specifi-cally referred for their underlying sinus disease. In addition, inclassifying patients with CRSþNP as having aspirin sensitivity, itshould be noted that diagnosis by aspirin challenge was verified in57% of AS patients. The remaining 43% of AS charts noted aspirinsensitivity by history of worsening asthma in response to aspirin orother nonsteroidal anti-inflammatory drugs on multiple occasionsbut were not formally challenged. This may be viewed as a short-coming in this study.

Despite inherent limitations, the authors add to the literaturethe largest multiparameter review of patients with CRSþNP and itssubset analysis of large AS and AT samples. This study is unique inthat it is a simultaneous analysis of multiple disease parametersand their interactions to include eosinophilia, surgery, olfactorydysfunction, medication use, and aspirin intolerance. The authorshave shown unique findings in the high prevalence of asthma in ASand AT patients and a greater degree of olfactory dysfunction in ASpatients and reported a small percentage of patients with circu-lating eosinophils in the CRSþNP group. Therefore, physicianstreating patients with NPs are recommended to screen for asthma,dysosmia, and aspirin sensitivity but should not rely on circulatingeosinophils to discriminate between these 2 sinus disease pheno-types. AS patients seem more likely to undergo more extensiveinterventions than AT controls, although the 2 groups have highuse. Given the high prevalence of revision surgeries and unclearbest practices for medical management, prospective randomizedcontrolled trials looking at medical management for postoperativepatients with CRSþNP and include AS and AT subgroups are

recommended because disease burden is significant and thereremains a clear unmet need. The present data characterize suchpatients and the state of current interventions.

References

[1] Rudmik L, Schlosser RJ, Smith TL, Soler ZM. Impact of topical nasal steroidtherapy on symptoms of nasal polyposis. Laryngoscope. 2012;122:1431e1437.

[2] Picado C. Aspirin intolerance and nasal polyposis. Curr Allergy Asthma Rep.2002;2:488e493.

[3] Fokkens W, Lund V, Bachert C, et al. EAACI Position Paper on rhinosinusitisand nasal polyps: executive summary. Allergy. 2005;60:583e601.

[4] Seybt MW, McMains KC, Kountakis SE. The prevalence and effect of asthma onadults with chronic rhinosinusitis. Ear Nose Throat J. 2007;86:409e411.

[5] Awad OG, Lee JH, Fasano MB, Graham SM. Sinonasal outcomes after endo-scopic sinus surgery in asthmatic patients with nasal polyps: a differencebetween aspirin-tolerant and aspirin-induced asthma? Laryngoscope. 2008;118:1282e1286.

[6] Ceylan E, Gencer M, San I. Nasal polyps and the severity of asthma. Respir-ology. 2007;12:272e276.

[7] Mendelsohn D, Jeremic G, Wright ED, Rotenberg BW. Revision rates afterendoscopic sinus surgery: a recurrence analysis. Ann Otol Rhinol Laryngol.2011;120:162.

[8] Kim J, Kountakis SE. The prevalence of Samter’s triad in patients undergoingfunctional endoscopic sinus surgery. Ear Nose Throat J. 2007;86:396.

[9] Batra PS, Kern RC, Tripathi A, et al. Outcome analysis of endoscopic sinussurgery in patients with nasal polyps and asthma. Laryngoscope. 2003;113:1703e1706.

[10] Guidelines for the Diagnosis and Management of Asthma. Expert Panel Report3. NIH Publication 07-4051. Bethesda, MD: National Institutes of Health;2007.

[11] Lund VJ, Mackay IS. Staging in rhinosinusitis. Rhinology. 1993;31:183e184.[12] Zeitz H. Bronchial asthma, nasal polyps, and aspirin sensitivity: Samter’s

syndrome. Clin Chest Med. 1988;9:567e576.[13] Alobid I, Cardelus S, Benitez P, et al. Persistent asthma has an accumulative

impact on the loss of smell in patients with nasal polyposis. Rhinology. 2011;49:519e524.

[14] Ten Brinke A, Grootendorst DC, Schmidt JT, et al. Chronic sinusitis in severeasthma is related to sputum eosinophilia. J Allergy Clin Immunol. 2002;109:621e626.

[15] McMains KC, Kountakis SE. Revision FESS: objective and subjective surgicaloutcomes. Am J Rhinol. 2005;19:344e347.

[16] McMains KC, Kountakis SE. Medical and surgical considerations in patientswith Samter’s triad. Am J Rhinol. 2006;20:573e576.

[17] Downey LL, Jacobs JB, Lebowitz RA. Anosmia and chronic sinus disease.Otolaryngol Head Neck Surg. 1996;115:24e28.

[18] Konstantinidis I, Witt M, Kaidoglou K, Constantinidis J, Gudziol V. Olfactorymucosa in nasal polyposis: implications for FESS outcome. Rhinology. 2010;48:47e53.

[19] Bhattacharyya N. Clinical and symptom criteria for the accurate diagnosis ofchronic rhinosinusitis. Laryngoscope. 2006;116(suppl 110):1e22.

[20] Szczeklik A, Nizankowska E, Duplaga M. Natural history of aspirin-inducedasthma. AIANE Investigators. European Network on Aspirin-InducedAsthma. Eur Respir J. 2000;16:432e436.

[21] Loehrl TA, Ferre RM, Toohill RJ, Smith TL. Long-term asthma outcomes afterendoscopic sinus surgery in aspirin triad patients. Am J Otolaryngol. 2006;27:154e160.

[22] Poznanovic SA, Kingdom TT. Total IgE levels and peripheral eosinophilia:correlation with mucosal disease based on computed tomographic imagingof the paranasal sinus. Arch Otolaryngol Head Neck Surg. 2007;133:701e704.