Fluticasone Propionate Nasal Spray Is Superior to Montelukast for Allergic Rhinitis While Neither Affects Overall Asthma Control^* FREE TO VIEW

The prevalence of asthma and allergic rhinitis in the United States is estimated to be 10% and 10 to 20%, respectively.¹² In addition, > 75% of asthma patients have symptoms of allergic rhinitis.³⁴ Annually, asthma results in approximately 14.5 million days missed from work and a similar number of days missed from school, totaling 28.5 million days of absenteeism.² The total direct and indirect costs for asthma and rhinitis have been estimated to exceed $23 billion dollars annually.⁵⁶ The combined burden of asthma and rhinitis for patients, providers, and payors is significant and is a target for treatment simplification and cost reduction.

The traditional management of asthma and allergic rhinitis has been to individualize the treatment of the upper and lower respiratory tract with specific pharmacotherapies. However, epidemiologic, anatomic, pathologic, and physiologic findings common to both conditions exist, leading to the hypothesis that the upper and lower airways are linked (“one-airway” hypothesis).⁷ It follows from this hypothesis that a therapeutic approach targeting one site may significantly improve the other site. However, this premise has not been widely evaluated.

The purpose of this study was to investigate prospectively the effect of rhinitis therapy on asthma outcome measures in patients with seasonal allergic rhinitis (SAR) and persistent asthma. If the one-airway hypothesis is true, there may be additional benefit in overall asthma control when symptoms of allergic rhinitis are adequately controlled.

Eligible subjects were at least 15 years of age, had a history of SAR for at least two allergy seasons, and a positive skin test response during screening to the relevant seasonal allergen. Patients also had a diagnosis of persistent asthma (as defined by the American Thoracic Society⁸) and were receiving daily asthma treatment for at least 3 months preceding the study. Patients screened for the study may have been using either fluticasone propionate/salmeterol (FSC), an inhaled corticosteroid (ICS) only, or nonsteroidal asthma medications (ie, short-acting or long-acting β-agonists, anticholinergics, or cromolyn) alone or concurrently with ICS without interruption of the dose and regimen for at least 30 days prior to screening. If patients were receiving an ICS, the total daily dose was limited to one of the following: fluticasone propionate metered-dose inhaler ≤ 220 μg, fluticasone propionate via Diskus inhaler (GlaxoSmithKline; Research Triangle Park, NC) ≤ 250 μg, budesonide ≤ 400 μg, beclomethasone dipropionate ≤ 420 μg, flunisolide ≤ 1,000 μg, or triamcinolone acetonide ≤ 1,000 μg. At study entry, patients currently receiving FCS 100/50 μg bid (Advair Diskus 100/50; GlaxoSmithKline) as asthma therapy were required to have an FEV₁ ≥ 80% of predicted, and those receiving an allowed asthma therapy other than FSC 100/50 μg bid were required to have an FEV₁ between 65% and 95% of the predicted value at visit 1 based on race-adjusted predicted normal values for age ≥ 18 years of Crapo et al,⁹ or predicted normal values for ages 15 to 17 years of Polgar and Promadhat.¹⁰ Patients could not have been using any anti-inflammatory medications to control nasal symptoms for 4 weeks prior to or at any time during the study.

Exclusion criteria included the following: pregnancy and/or lactation; history of life-threatening asthma; asthma hospitalization within 6 months of screening; significant concurrent diseases, including a recent respiratory tract infection; and recent nasal surgery or anatomic defects of the nose, such as a deviated septum or nasal septal perforation. Oral, intranasal, ocular, or parenteral corticosteroids and leukotriene modifiers were prohibited for 4 weeks prior to the screening visit. Patients also were excluded if they had received more than two courses of oral or parenteral corticosteroids within 6 months of screening. Additional medications excluded prior to screening and throughout the study included intranasal or ocular cromolyn, short and long-acting antihistamines, nasal decongestants, and intranasal anticholinergics.

The 4-week, randomized, double-blind, parallel-group study (protocol SAM40066) was approved by the institutional review boards for each of the 92 investigative sites in the United States. Each patient signed a written informed consent document before enrollment and before any study procedures were performed.

Eligible patients with a history of both persistent asthma and SAR entered a 7- to 14-day screening period to document the coexistence of asthma and rhinitis. During the screening period, patients continued their prestudy asthma medications and replaced their current short-acting β₂-agonist with albuterol hydrofluoroalkane to be used as needed for the relief of asthma symptoms throughout the study. Patients used daily diary cards to record daytime and nighttime asthma and rhinitis symptoms. Asthma symptoms were self-evaluated using a 0- to 5-point Likert scale, with 0 representing no symptoms and 5 representing severe symptoms. In addition, daily morning and evening peak expiratory flow (PEF), albuterol use, and number of nighttime awakenings due to asthma were recorded. Patients were instructed in the use of the peak flowmeter (Mini-Wright; Clement Clark; London, UK). Morning and evening PEF (best effort of three attempts) were measured before taking any medications and after recording diary symptoms.

Daytime rhinitis symptoms were evaluated through self-assessment of four different nasal symptoms using a visual analog scale. A score of 0 to 100 (0 represented no symptoms and 100 indicated severe symptoms) was possible for each daytime individual nasal symptom score (D-INSS) of nasal congestion, itching, runny nose, and sneezing. The sum of the individual scores produced a daytime total nasal symptom score (D-TNSS) ranging from 0 to 400. On awakening, the patient also assessed overnight nasal symptoms related to stuffy nose, sleep difficulty due to nasal symptoms, and frequency of nighttime awakenings due to nasal symptoms using a 0 to 3 integer scale (summation of individual scores produced the nighttime total nasal symptom score [N-TNSS], ranging from 0 to 9).

During screening, each patient currently receiving FSC as asthma therapy was required to demonstrate asthma symptom scores ≥ 1 on ≤ 4 of the 7 days immediately prior to randomization (visit 2). Each patient currently receiving an allowed asthma therapy other than FSC 100/50 μg bid was required to demonstrate albuterol hydrofluoroalkane use or have an asthma symptom scores ≥ 2 on ≥ 3 of the 7 days immediately prior to randomization. In addition, each patient was required to demonstrate a D-TNSS of ≥ 200 on ≥ 4 days during this same 7-day period. Patients who did not meet both the asthma and rhinitis symptom criteria during screening were discontinued from the study.

Patients who met both the asthma and rhinitis criteria were randomly assigned to one of the following three treatments for rhinitis for 4 weeks: (1) fluticasone propionate aqueous nasal spray (FPANS), 200 μg qd, plus a placebo capsule; (2) overencapsulated montelukast tablets 10 mg qd plus vehicle placebo aqueous nasal spray; or (3) placebos for both active treatments. Patients self-administered two sprays per nostril and one capsule in the evening during the study period. All prestudy asthma medications (excluding albuterol hydrofluoroalkane) were discontinued following randomization, and each patient was provided open-label FSC 100/50 μg bid for 4 weeks.

At 27 sites, 24-h urine samples were collected prior to randomization and at the end of the 4-week treatment period. Subjects were instructed to start the collection 24 h prior to the randomization visit, and again 24 h prior to the final study visit. All evaluable samples (defined as those collected within the appropriate time interval with a documented start and stop time) had the total urine volume recorded, and a 2-mL sample was frozen and shipped to a central laboratory for analysis.

Because this study employed two primary measures to demonstrate the dual objectives of the study, the selection of sample size took both measures into account. Using a two-sided t test and a significance level of α = 0.05, it was estimated that 244 subjects per treatment group would provide 90% power to detect a difference between treatments in change from baseline D-TNSS of 25 given an SD of 85, and that 133 subjects per treatment group would provide 90% power to detect a difference between treatments in change from baseline morning PEF of 20 L/min given an SD of 45 L/min. Accordingly, the sample size for this study, in order to demonstrate superiority in terms of change from baseline in D-TNSS and equivalence in terms of change from baseline in morning PEF, was estimated to be 244 subjects per group.

Each rhinitis measure was assessed in terms of the difference between treatments in the mean change from baseline D-TNSS, averaged over weeks 1 and 2, and over weeks 1 to 4. Baseline was defined as the arithmetic average of the data recorded on the 4 days immediately preceding randomization. Data were summarized in terms of least-squares estimates.

Each asthma measure was assessed in terms of the difference between treatments in the mean change from baseline to end point. Baseline was defined as the arithmetic average of the data recorded on the day of randomization and the 6 days immediately preceding it. End point was defined as the average of the last 7 days of available on-treatment data. Data were summarized in terms of least-squares estimates. Analysis of asthma measures included data only from those subjects whose pre-enrollment asthma therapy did not include FSC 100/50 μg or an ICS plus an inhaled long-acting β₂-agonist.

All hypothesis tests were two tailed. Treatment differences were assessed in terms of an analysis of covariance model that included baseline as the covariate and terms for treatment, investigator, and prior asthma therapy. Rhinitis results were evaluated for statistical superiority in terms of a p value, assessed at the α =0.05 level; asthma results were evaluated for statistical equivalence in terms of a two-sided 95% confidence interval around the difference between treatments in change from baseline. Confidence intervals were determined through use of analysis of covariance models that included baseline as the covariate and terms for treatment, investigator, and prior asthma therapy. Subject-rated overall satisfaction with treatment was assessed for treatment differences using a van Elteren test, stratified by investigator, and previous asthma therapy.

Of 1,551 patients screened, a total of 863 were randomly assigned to treatment and 805 patients completed the study. Discontinuations were similar in each group with 7%, 6%, and 8% withdrawing in the FPANS plus FSC group, montelukast plus FSC group, and placebo plus FSC group, respectively. The most common reasons for study discontinuation were protocol violations and adverse events.

Baseline demographic and pulmonary function characteristics were similar across groups (Table 1 ). Mean FEV₁ at baseline was approximately 81% of predicted normal for all three treatment groups, and the majority of patients had been treated for their asthma with short-acting β₂-agonists only (59%) [Table 1]. In each treatment group, at least 95% of subjects were ≥ 80% compliant with study medications, and no patient was discontinued from the study because of study medication noncompliance.

A significant improvement (p ≤ 0.001) from baseline at end point in morning PEF, ranging from 31.6 to 34.0 L/min, was seen for all treatment groups when baseline asthma treatments were replaced with open-label FSC (Fig 1 ). However, there were no differences in mean change from baseline for the primary asthma outcome, and change from baseline in morning PEF at end point, when FPANS, montelukast, or placebo were used concurrently with FSC (Fig 1, Table 2 ). Furthermore, there were no significant differences among the groups in morning PEF or other measures of asthma control including evening PEF, asthma symptom-free days, percentage of albuterol-free days, or the number of nighttime awakenings due to asthma (Table 2). At end point, as shown in Table 2, subjects in each treatment group still experienced asthma symptoms severe enough to require use of their rescue medication on at least 63.6% of the days. This indicates that any supplemental improvement in asthma symptoms achieved with the addition of upper airway medications could have been detected, if present.

To evaluate the effect of asthma severity on response to treatment, asthma outcomes were analyzed ad hoc using lung function at baseline to classify asthma severity (ie, FEV₁ < 80% vs FEV₁ ≥ 80% of predicted at baseline). Fifty-nine percent (59%) of patients enrolled in the study had an FEV₁ > 80% of predicted at the time of randomization. Treatment effects were similar for both severity cohorts for all asthma outcomes including morning PEF, evening PEF, percentage of symptom-free days, and percentage of rescue-free days (Table 3 ). As was seen in the total population, there were no statistically significant differences among treatments for any of the end points

The primary end point for control of the upper airways in patients with coexistent asthma and SAR was mean change from baseline over weeks 1 and 2 (days 2 to 15) in subject-rated D-TNSS. The D-TNSS and D-INSS results are presented in Figure 2 , Figure 3 , and Table 4 . Baseline values for D-TNSS and D-INSS were similar across the groups. For both the D-TNSS and the four individual daytime nasal specific outcomes (congestion, rhinorrhea, sneezing, and itching), FPANS plus FSC was superior to both montelukast plus FSC and placebo plus FSC (p ≤ 0.001). Montelukast plus FSC was superior to placebo plus FSC in terms of D-TNSS and itching and sneezing symptom scores, but was comparable to placebo plus FSC with respect to congestion and rhinorrhea symptom scores.

A secondary analysis of mean change in D-TNSS and individual nasal symptom scores over weeks 1 through 4 was also performed. The results of these analyses were similar to those of the primary analyses that were conducted over weeks 1 to 2 only (Table 4).

The N-TNSS is a composite of three individual nighttime symptoms (N-INSS) assessed each morning: nasal congestion on awakening, difficulty in going to sleep because of nasal symptoms, and nighttime awakenings because of nasal symptoms. There were no statistical differences among treatments for N-TNSS or N-INSS at baseline. Statistically significant differences were seen in N-TNSS and N-INSS favoring FPANS plus FSC over montelukast plus FSC and placebo plus FSC (p ≤ 0.002) [Table 5 , Fig 4, 5 ] . Montelukast plus FSC was not significantly different from placebo plus FSC for either N-TNSS or any N-INSS (Table 5). The analysis of N-TNSS and N-INSS over weeks 1 through 4 was similar to weeks 1 and 2 for all treatment comparisons (Table 5).

After 4 weeks of treatment with blinded FPANS, montelukast, or placebo and open-label FSC 100/50 μg, patients rated their overall satisfaction with study medications (Table 6 ). There was a significant difference in patient satisfaction among patients who had received FPANS plus FSC compared with montelukast plus FSC and placebo plus FSC (p < 0.001). Sixty-nine percent (69%) of patients receiving FPANS plus FSC were satisfied or very satisfied with their treatment, compared with 55% and 51% of patients receiving montelukast plus FSC or placebo plus FSC, respectively.

In general, all treatments were well tolerated, and the incidence of adverse events was similar across the groups (36%, 40%, and 42% for FPANS plus FSC, montelukast plus FSC, and placebo plus FSC, respectively). The most common occurring adverse events included headache (9%, 14%, 13%), sore throat (3%, 4%, 3%), epistaxis (3%, 2%, 4%), dyspepsia (2%, 4%, 2%), and back pain (2%, 1%, 3%) for FPANS plus FSC, montelukast plus FSC, and placebo plus FSC, respectively. All other adverse events occurred at an incidence ≤ 2%. Drug-related adverse events as assessed by the investigator occurred infrequently. The most common drug-related adverse events (incidence > 1% in any treatment arm) included epistaxis (2%, 1%, 3%), sore throat (≤ 1%, 1%, 2%), and headache (2%, 4%, 4%) for FPANS plus FSC, montelukast plus FSC, and placebo plus FSC, respectively.

A total of 164 patients (58 patients in the FPANS plus FSC group, 51 patients in the montelukast plus FSC group, and 55 patients in the placebo plus FSC group) were included in the cortisol population. This population included patients with urine volumes > 600 mL for female or > 800 mL for male subjects in addition to creatinine greater than the lower threshold limit (mean − 2.5 × SD), collection time periods of 24 ± 4 h, and no use of additional ICS during treatment or use of oral, injectable, ophthalmic, intranasal, or topical (> 1%) corticosteroid medications within 30 days of screening or during treatment. The geometric mean urinary cortisol excretion at treatment week 4 was comparable across the treatment groups (FPANS plus FSC, 15.67 μg/24 h; montelukast plus FSC, 11.99 μg/24 h; and placebo plus FSC, 13.99 μg/24 h).

Asthma exacerbations, defined by any asthma-related event that required treatment with asthma medications beyond study medications, occurred infrequently in all groups: one patient (< 1%), three patients (1%), and four patients (1%) receiving FPANS plus FSC, montelukast plus FSC, and placebo plus FSC, respectively. Respiratory infections were the most common suspected cause of exacerbations. One patient each in the FPANS plus FSC and montelukast plus FSC groups and two patients in the placebo group discontinued the study because of worsening asthma.

The purpose of this study was to evaluate if there would be additional benefit in overall asthma control when coexistent symptoms of allergic rhinitis were optimally treated. Although asthma and allergic rhinitis share common epidemiologic, anatomic, pathologic, and physiologic features, there was no evidence in this study that treatment of the upper airway conferred significant benefit to the lower airway. Specifically, in the present study, control of the upper airway was significantly greater with FPANS compared with montelukast and placebo for both daytime and nighttime TNSS and INSS. However, overall asthma control, as measured by lung function, symptoms, and rescue albuterol use, were indistinguishable when either FPANS, montelukast, or placebo was added to FSC for the treatment of allergic rhinitis (data not shown).

The superior clinical efficacy of FPANS compared with montelukast and placebo in relieving SAR symptoms demonstrated in the present study has also been documented previously in other studies.^11121314 In addition, intranasal corticosteroids have been shown to be more effective in controlling the symptoms of allergic rhinitis when compared with antihistamines.¹⁵¹⁶ Taken together, these observations are consistent with evidence-based guidelines that identify and recommend intranasal corticosteroids as the most effective class of medications for controlling the multifaceted symptomatology associated with allergic rhinitis.¹⁷

The concept that asthma and allergic rhinitis are manifestations of one disease entity has led to work examining whether concurrent symptoms of the upper and lower airways can be controlled by targeting treatment to only one compartment. For example, treatment of the upper airway with antihistamines has proven to have marginal benefit in controlling lower airway symptoms.¹⁸ Likewise, although leukotriene modifiers such as montelukast are indicated for the treatment of both allergic rhinitis and asthma, their benefit to the upper airway appears to be marginal compared with placebo when they are used to control the lower airways.¹⁹²⁰ Corticosteroids, the broadest-spectrum anti-inflammatory treatment available for the treatment of the upper or lower airways, effectively control upper and lower airway symptoms when administered topically. In this regard, there is some evidence to suggest that treating the upper airways with intranasal corticosteroids confers benefit to the lower airways as manifested by improvements in bronchial hyperresponsiveness,²¹²² asthma symptoms,²² and reduced asthma-related hospitalizations and emergency department visits.²³ In contrast to these studies, a benefit to the lower airways was not observed with either intranasal corticosteroids or montelukast in the present study despite marked differences in the effects of intranasal corticosteroids vs montelukast on upper airway symptoms.

It is likely that the differences in design between the current study and prior studies explains these somewhat disparate results. Specifically, the design of the present study included specific treatments for the upper and lower airway diseases, while the former studies²¹²²²³ targeted therapy for treatment of the upper airway. This may, in part, explain why no additional benefit of treating the upper airways was observed in the lower airways in the current study. Although some previous studies²¹²²²³ indicate that effective treatment of the upper airway symptoms can, to varying extents, positively impact the lower airways, our study suggests that treatment of the upper airways is not adequate for control of the lower airways but rather effective treatment of the individual diseases offers the best prospect for optimal control.

The present study establishes that when asthma is effectively treated with FSC, the addition of montelukast 10 mg/d or FPANS 200 μg/d has no additive benefit on overall asthma control. The lack of added benefit to the lower airway when adding montelukast to single and dual controller therapy has been previously reported by Robinson, et al.²⁴ One potential criticism of the study by Robinson et al²⁴ and our study is that baseline asthma characteristics were too mild to allow differences in asthma outcomes to be distinguished among the three treatment groups. However, in our study patients at baseline were experiencing symptoms on > 90% of days and using rescue albuterol on > 80% of days. In addition, an ad hoc analysis showed asthma outcomes were comparable among the three groups, even for patients with more severe lung dysfunction at baseline (ie, baseline FEV₁ < 80% predicted). This, coupled with the fact that normalization of lung function or ablation of asthma symptoms did not completely occur in any of the groups, suggests that the failure to demonstrate a difference in asthma outcomes among the three treatment groups cannot be explained by asthma being so mild that it would not be possible to detect additional improvements in asthma outcomes regardless of the treatment.

Adverse events in this short-term study were minimal and comparable among groups. Perhaps more importantly, 24-h urinary cortisol excretion concentrations at baseline (ie, at a time when > 68% of patients across the groups were not receiving an ICS) and after 4 weeks of treatment with either FPANS plus FSC, montelukast plus FSC, or placebo plus FSC were comparable, and there was no evidence of a differential effect among the three treatment groups. The lack of effect of fluticasone propionate on hypothalamic-pituitary-adrenal axis function in this study, as assessed by 24-h urinary cortisol excretion, is consistent with previous studies.^2526272829

Despite the strong evidence for an immunologically integrated “one airway,” this study suggests that the applicability of this hypothesis to clinical treatment of rhinitis and asthma may be limited. In this study, in which asthma was treated with FSC, it was not possible to demonstrate further improvement in asthma symptoms or pulmonary function by the addition of either a topical nasal corticosteroid (FPANS) or a systemic leukotriene receptor antagonist (montelukast). However, the topical nasal corticosteroid was significantly more effective in reducing nasal symptoms than the leukotriene receptor antagonist. This indicates that in patients with allergic rhinitis and asthma, the disease in each organ should be treated with the most effective medication for that condition to obtain optimal control of both.

The following investigators participated in this study: N. Amar, Waco, TX; C. Andrews, San Antonio, TX; J. Anon, Erie, PA; R. Arastu, San Antonio, TX; D. Atkinson, Oklahoma City, OK; A. Aven, Arlington Heights, IL; D. Bernstein, Cincinnati, OH; W. Busse, Madison, WI; S. Christensen, Salt Lake City, UT; J. Condemi, Rochester, NY; J. Cook, Wenatchee, WA; J. Craig, Cincinnati, OH; P. Creticos, Baltimore, MD; C. Dual, Metairie, LA; R DeGarmo, Greer, SC; A. Driver, Sellersville, PA; D. Dvorin, Forked River, NJ; M. Ellis, Orange, CA; G. Erdy, Newburgh, IN; A. Finn, Jr., Charleston, SC; G. Fino, Pittsburgh, PA; S. Galant, Orange, CA; S. Gawchick, Upland, PA; R. Gilman, East Providence, RI; P. Goldberg, Indianapolis, IN; D. Gossage, Knoxville, TN; G. Gottschlich, Cincinnati, OH; L. Greos, Englewood, CO; M. Gutierrez, Austin, TX; R. Harker, LeMoyne, PA; A. Heller, San Jose, CA; D. Henry, Salt Lake City, UT; M. Hollie, Chattanooga, TN; M Hudelson, Flower Mound, TX; T. Hunt, Austin, TX; J. Jacobs, Walnut Creek, CA; R. Jacobs, San Antonio, TX; H. Kaiser, Plymouth, MN; S. Kelsen, Philadelphia, PA; E. Kent, Jr., South Burlington, VT; E. Kerwin, Medford, OR; C. LaForce, Raleigh, NC; A. Lapey, West Quincy, MA; T. Lee, Atlanta, GA, J. Leflein, Ypsilanti, MI; M. Livezey, Lilburn, GA; R. Lockey, Tampa, FL; D. Lorch, Brandon, FL; M. Manning, Scottsdale, AZ; B. Martin, San Antonio, TX; J. McFreely, Berkeley, CA; B. Miller, Killeen, TX; S. Miller, North Darmouth, MA; W. Moore, Winston-Salem, NC; H. Nelson, Denver, CO; M. Noonan, Portland, OR; A. Ober, North Andover, MA; N. Ostrom, San Diego, CA; A. Patel, Pueblo, CO; A. Patel, Colorado Springs, CO; J. Pearle, Fullerton, CA; J. Pinnas, Tuscon, AZ; W. Pleskow, Encinitas, CA; S. Pollard, Louisville KY; K. Pudi, Simpsonville, SC; M. Radbill, Bensalen, PA; N. Rai, Tacoma, WA; G. Raphael, Bethesda, MD; P. Ratner, San Antonio, TX; M. Reid, Napa, CA; R. Rhoades, Martinez, GA; J. Rumbyrt, Arvada, CO; N. Rupp, Charleston, SC; R. Saff, Tallahassee, FL; G. Salazar, San Antonio, TX; L. Scholar, Walla Walla, WA; R. Settipane, Providence, RI; S. Shah, Collegeville, PA; R. Shusman, Springfield, PA; J. Sibille, Sunset, LA; B. Sigal, Winston-Salem, NC; R. Sterling; Orangeburg, SC; R. Stoloff, Plattsburgh, NY; R. Sussman, Springfield, NJ; M. Tarpay, Oklahoma City, OK; I. Tripathy, Rolla, MO; M. Vandewalker, Columbia, MO; J. Westerman, Jasper, AL; K. Wingert, Fresno, CA; J. Wolfe, San Jose, CA; and R. ZuWallack, Hartford, CT.

The authors thank Laura Sutton, PharmD, for assistance in preparing the manuscript.