Direct evidence for a role of the mast cell in the nasal response to aspirin in aspirin- sensitive asthma
Andrew R. Fischer, MD, a, b Mitchell A. Rosenberg, MD, a, b, c
Craig M. Lilly, MD, a, b Joan C. Callery, RN, a Paul Rubin, MD, d
Judith Cohn, MD, PhD, a Martha V. White, MD, e Yasushi Igarashi, MD, e
Michael A. Kaliner, MD, e Jeffrey M. Drazen, MD, a, b, c and Elliot Israel, MD ~, b. c
Boston, Mass., Abbot t Park, Ill., and Bethesda, Md.
Background: A subset of patients with asthma experience adverse nasoocular reactions after ingestion of aspirin or agents that inhibit cyclooxygenase. Recent evidence has implicated the leukotrienes in the nasoocular reaction, but the cellular sources and mechanism of activation are unknown. We used nasal lavage with and without a 5-lipoxygenase inhibitor, zileuton, to define the role of leukotrienes and to profile nasal cellular activation during this reaction. Methods: A group of eight patients with asthma shown to have adverse reactions to aspirin documented by a 15% or greater decrease in forced expiratory volume in 1 second, accompanied by an elevation in urinary leukotriene E 4 after ingestion of aspirin, received aspirin or placebo in a study with a crossover design. Nasal symptoms and nasal tryptase, histamine, leukotriene, and eosinophil cationic protein levels were evaluated. Serum tryptase and urinary histamine levels were also assessed. Subjects were then randomized to receive a week of treatment with zileuton or placebo, according to a double-blind, crossover design followed by aspirin challenge and measurement of the same mediators. Results: Aspirin ingestion produced a marked increase in nasal symptoms from a baseline symptom score of 2.1 +_ 0.7 to a maximum of 8.4 +_ 1.2 (p < 0.0007). Aspirin ingestion produced a mean maximal increase in nasal tryptase of 3.5 +- 2.6 ng/ml, whereas placebo ingestion produced a mean maximal increase of 0.1 + 0.2 ng/ml (p < 0.05, aspirin vs placebo). Mean maximal nasal histamine increased 1.73 +- 1.16 ng/ml versus 0.08 +_ 0.08 ng/ml from baseline (p < 0.05, aspirin vs placebo). Aspirin produced a mean maximal increase in nasal leukotriene value of 152 pg/ml versus a 16 pg/ml decrease after placebo ingestion (p < O. 05). Zileuton treatment blocked the increase in nasal symptoms after aspirin ingestion (maximum nasal symptom score of 1.6 ++_ 0.6 with zileuton vs 5.5 +- 0.9 with placebo [p < 0.0053]). It also blocked the rise in nasal tryptase (p = 0.011) and nasal leukotriene (p < 0.05) levels after aspirin ingestion. Zileuton treatment had no significant effect on the recovery of nasal histamine. Conclusion: The increase in nasal symptoms in aspirin-sensitive patients with asthma after aspirin ingestion is associated with increases in nasal tryptase, histamine, and cysteinyl leukotriene levels. This mediator profile is consistent with mast cell activation during the nasal response to aspMn and suggests that 5-lipoxygenase products are essential for the nasal response to aspMn. (J ALLERGY CLIN IMMUNOL 1994;94:1046-56.)
From the Combined Program in Pulmonary and Critical Care Medicine: aBeth Israel and bBrigham and Women's Hospi- tals, Department of Medicine, Harvard Medical School, Boston; °Longwood Area Adult Asthma Center, Boston; aAbbott Laboratories, Abbott Park; and ~the Allergic Dis- eases Section, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda.
Supported in part by a grant from Abbott Laboratories and by National Institutes of Health grants AI-U01-31599, HL-T32- 07633, and M01-RR01032-17 to the Harvard-Thorndike Gen- eral Clinical Research Center of the Beth Israel Hospital.
1046
Ingest ion of aspirin or other agents that inhibit the cyclooxygenase enzyme system (i.e., nonsteroi-
dal an t i inf lammatory drugs), elicits respiratory,
Received for publication Oct. 21, 1993; revised June 8, 1994; accepted for publication June 10, 1994.
Reprint requests: Elliot Israel, MD, Respiratory Division, Brigham and Women's Hospital, 75 Francis St., Boston, MA 02115.
J ALLERGY CLIN {MMUNOL Fischer et al. 1047 VOLUME 94, NUMBER 6
Abbreviations used ANOVA: Analysis of variance
ECP: Eosinophil cationic protein FEV~: Forced expiratory volume in 1 sec-
ond LTE4: Leukotriene E 4
nasal, and gastrointestinal symptoms, as well as dermal changes in a subset of patients with asth- ma? The sensitivity to cyclooxygenase inhibitors has led to the hypothesis that nonsteroidal antiin- flammatory drugs may be causing upregulation of the 5-1ipoxygenase pathway and its attendant prod- ucts, the leukotrienes, in these patients. In fact, several groups have shown an increase in urinary leukotriene E 4 (LTE4) after aspirin ingestion or inhalation of lysine-aspirin in aspirin-sensitive pa- tients with asthma. 2-5 It has also been demon- strated that pharmacologic blockade at the level of the cysteinyl leukotriene receptor(s) can blunt the bronchospastic response to aspirin. 6, 7 Importantly, we have recently shown that inhibition of 5-1ipoxy- genase blocks not only the respiratory but also the gastrointestinal and dermal reactions to aspirin in aspirin-sensitive patients with asthma. 8 Although these results establish the importance of 5-1ipoxy- genase products in mediating reactions to aspirin, the cellular source and mechanism of release of these mediators remain unclear.
Both mast cells and eosinophils are present in the nasal mucosa of patients with aspirin sensitiv- ity.9, 10 Both cell types are capable of producing leukotrienes, n Tryptase is an enzyme specific to mast cells, ~2 and eosinophil cationic protein (ECP) is specific to eosinophils. 13 To assess the possible involvement of these two cell lines in aspirin- mediated reactions, we assayed the nasal lavage of the subjects undergoing aspirin challenges s for tryptase, ECP, and other inflammatory mediators. Additionally, we report the effect of zileuton, a potent and selective inhibitor of 5-1ipoxygen- ase,14.15 on the nasal response and mediator pro- duction after oral aspirin administration in these aspirin-sensitive patients with asthma.
METHODS Patients
Eight nonsmoking subjects with aspirin-induced asthma, documented by a 15% or greater decrease in forced expiratory volume in i second (FEV1) and an increase in urinary LTE 4 excretion after aspirin inges-
tion, but not after placebo ingestion, were included in this study; data on pulmonary function changes and ur ina ry L Y E 4 excretion have been previously reported. 8 Each subject gave written informed consent to partici- pate in the protocol, which was approved by the Com- mittee on Clinical Investigations of the Beth Israel Hospital.
Aspirin challenge and drug administration protocols
The protocols used for aspirin challenge, lung func- tion monitoring and urine collection, have been previ- ously outlined in detail. 8 Briefly, subjects (Table I) were initially screened with a single-blind oral aspirin chal- lenge, which was followed by a placebo-c0ntrolled, cross- over aspirin challenge to confirm aspirin sensitivity; the latter was termed the aspirin characterization phase. Once aspirin sensitivity was established, two more aspirin challenges took place during treatment with zileuton or placebo in a randomized, double-blind, crossover fash- ion (i.e., the "double-blind zileuton" phase).
The night before each challenge, subjects were admit- ted to the General Clinical Research Center of the Beth Israel Hospital for urine collection and observation. Subjects receiving nasal steroids continued to take them up to and including the time of the evening dose on the day before each challenge. Each subject's dose of in- haled or nasal steroids was unchanged for at least 4 weeks before each challenge. Individuals who were taking theophylline withheld it for at least 4 days before each challenge. Subjects received neither systemic ste- roids nor any form of cromolyn sodium for at least 30 days before each challenge.
Acetaminophen and antacids were permitted for re- lief of symptoms during the study. Intranasal oxymeta- zoline hydrochloride (Afrin; Schering-Plough, Kenil- worth, N.J.) was permitted during the double-blind zileuton phase; this medication has been shown not to affect mediator release in nasal secretions, 16 although it may reduce nasal symptoms. When used, it was admin- istered after lavage collections were completed or at equivalent times on both double-blind days.
Single-blind challenge Subjects were initially screened with a single-blind
oral aspirin challenge with previously described meth- ods. 2 Subjects ingested increasing doses of aspirin every 2 hours, starting at a dose of 10 mg or 30 rag. A dose of aspirin that would cause at least a 15% decrease in FEV1 was determined from this screening challenge.
Aspirin characterization phase The aspirin characterization phase consisted of chal-
lenges with either aspirin or placebo (lactose) with a crossover design separated by at least 5 days (Fig. 1). A two-dose, subthreshold and threshold, challenge design was used to ensure the safety of the subjects. The mean threshold dose of aspirin was 90 mg (range, 20 to 300 rag).
1048 Fischer et al. J ALLERGY CLIN IMMUNOL DECEMBER 1994
Serum tryptase (t) : " tl I Nasal lavage (~) : ~,I I
r \ \ I \ \ f , , r , I , I , I , I , I , I 20:00 08:00 0 +2 +4 +6 +8 +10 +12 +14 +16
Clock time Relative time (hrs)
t Subthreshold Threshold ASA/placebo ASA/placebo
t t t
FIG. 1. Schematic representation of the aspirin characterization phase. Subjects received either aspirin or placebo in a crossover manner, each administered on a single day as outlined above. At t ime "zero" subthreshold aspirin or placebo dosing took place, fol lowed by threshold dosing 2 hours later. All urine was collected throughout the study day at the intervals indicated. Nasal symptom assessment, nasal lavage, and serum for tryptase measurement occurred at the t ime points noted. Adapted from Israel et al. Am Rev Respir Dis 1993;148:1447-51. By permission.
T A B L E I. Character ist ics of each subject (mod i f i ed f r om reference 8)
Subject Age History of Prestudy FEV 1 Prestudy FEV 1 No. (yr) Sex A t o p y * nasal polyps Medicat ions (L) (% predicted)
101 52 M - + B/IS/NS/T 3.28 89 102 32 M + + B/IS/NS/T 3.04 72 103 42 M + + B/IS/NS/T 3.58 100 104 23 M + + B/IS/NS/T 3.62 83 105 46 F + + B/IS/T 3.04 121 106 40 M + + B/IS 3.81 101 107 42 M + + B/IS/NS/T 3.29 88 108 41 F + - B 3.07 116
B, [32-agonists; IS, inhaled steroids; NS, nasal steroids; 7;, theophylline. *Atopy is defined by a positive prick test response (>3 mm wheal) to any one of a battery of 18 common aeroallergens or a positive
intradermal test result. Modified from Israel et al. Am Rev Respir Dis 1993;148:1447-51. By permission.
D o u b l e - b l i n d z i l e u t o n p h a s e
Subjects were randomized to receive treatment with zileuton (600 mg four times daily) or placebo for 6 to 8 days, followed by aspirin challenge with a randomized, double-blind, crossover design (Fig. 2). Each arm of this phase was separated by at least 7 days, and each aspirin challenge always began at the same time of day. The 7-day minimum interval was adequate for complete washout of zileuton. 14 The aspirin doses were identical for each subject during each arm of this phase. The mean threshold dose of aspirin was 91.25 mg (range, 30 to 300 mg).
N a s a l a n d g a s t r o i n t e s t i n a l s y m p t o m
a s s e s s m e n t
Subjects scored their nasal symptoms at the time points indicated in Figs. 1 and 2. Four symptoms (itchy nose, congestion, sneezing, and runny nose) were scored. Symptoms were scored on a scale of 0 (no symptoms) to 4 (very severe symptoms) by each subject's completion of a separate questionnaire at each time point. Investigators did
not ask or answer questions or interpret the questionnaire in any manner. The reported symptom score was obtained by summing the individual scores.
Gastrointestinal symptoms included nausea, vomiting, diarrhea, and abdominal pain. Presence or absence of any of these four symptoms was assessed at the time of nasal symptom assessment, and the presence of any one of the four was considered to be a positive response as previously reported. 8
N a s a l l a v a g e p r o t o c o l
Nasal lavage was done with the patient seated in the "sniff" position. The same nostril was used for all collections on any given study day. The initial lavage (30 minutes before ingestion of subthreshold aspirin or placebo) was discarded on every study day and is not included in Figs. 1 and 2. Thirty minutes later, the nose was lavaged a second time, and this fluid was recorded as the baseline lavage.
A 10 ml syringe connected to extension tubing con- taining a total of 8 ml of 0.9% saline solution at ambient
J ALLERGY CLIN IMMUNOL F i s c h e r e t a l . 1 0 4 9 VOLUME 94, NUMBER 6
Serum tryptase (1) : ti ,1- t 1 1
Nasallavage({): ~'i { ! + ~ { + {
Nasal symptom score(e) : e', * 6 • • • •
URINE COLLECTION :\~rl D- I-q~--~\ #2 : •
20:00 08:00 0 +2 +4 Clock time
Subthreshold Threshold ASA ASA
Zileuton/ Zileuton/ placebo placebo
#3 #4 #5
I I I I I I i I i I +6 +8 +10 +12 +14
Relative time (hrs)
-I
i I +16
FIG. 2. Schematic representation of the double-blind zileuton phase. In a crossover design subjects were treated with zileuton or placebo for 1 week and received their last 2 doses on the day of their aspirin challenge as indicated. The procedures performed on the last day of each arm are indicated. At t ime "zero" subthreshold aspirin was administered, fol lowed 2 hours later by threshold aspirin. All urine was collected the night before and during the study day at the intervals indicated. Nasal symptom assessment, nasal lavage, and serum for tryptase measure- ment occurred at the t ime points noted. Adapted from Israel et al. Am Rev Respir Dis 1993;148:1447-51. By permission.
temperature was used for the lavage procedure. The extension tubing was inserted atraumatically along the inferior turbinate to a depth of 5 cm. Subjects held both nares closed as the lavage fluid was inserted. After 30 seconds, the fluid was gently aspirated into a syringe and transferred to polypropylene tubes. The tubes were labeled and frozen at - 8 0 ° C until analysis was per- formed as described below.
Urine and serum collection protocols Urine was collected for histamine determination dur-
ing the intervals indicated in Figs. i and 2. Samples were collected in polypropylene bottles. Volumes were mea- sured and aliquots were stored at - 80 ° C until analysis was performed as described below.
Serum was obtained for tryptase assay at the points noted on Figs. 1 and 2. Blood was drawn, allowed to stand at room temperature for 15 minutes, and subjected to centrifugation. Aliquots were then transferred to polypropylene tubes and frozen at - 80 ° C until analysis was performed as described below.
Assay of mediators Nasal tryptase was measured in the retentate, and
histamine was measured in the filtrate of the lavage fluid after it was processed as described by Castells and Schwartz. 17 Tryptase levels (nasal and serum) were measured with an immunoradiometric sandwich assay with a commercially available assay kit (Pharmacia Di- agnostics, Piscataway, N.J.) with a lower limit of detec- tion of 0.5 ng/ml. Histamine levels (nasal and urinary) were measured with a radioimmunoassay kit (Immuno- tech International, Westbrook, Maine) with a lower limit of detection of 0.1 nmol/L (0.022 ng/ml). ECP levels were measured with a radioimmunoassay kit (Pharmacia Diagnostics) with a lower limit of detection of 2.0 ng/ml.
Cysteinyl leukotrienes were measured with commercially available assay kit materials (Advanced Magnetics, Waltham, Mass.) and techniques as described previ- ously. 18 The lower limit of detection of this assay was 27 pg/ml. All of the above assays were carried out in duplicate. The final concentration was the average of the two concentrations obtained by comparison with a freshly prepared standard curve.
One patient (patient 101) did not have nasal lavage performed during the aspirin characterization phase of the study. All eight patients had lavage performed during the double-blind zileuton phase. During the aspirin characterization phase five of the remaining seven patients had lavage assayed for determination of tryptase and histamine levels. The other two patients (nos. 105 and 106) had such severe nasal congestion that adequate (->1 ml) lavage recovery did not occur.
Nasal lavage was assayed for cysteinyl leukotrienes in seven of the patients during the aspirin characterization phase. One patient (no. 106) had lavage performed only at baseline and one other collection interval. Patients 103, 105, and 108 were each missing lavage for one collection point in one of the two arms of the aspirin characterization phase. The results of the available lavages were used for chi square statistical analysis.
One patient (no. 106) had adequate lavage only during the baseline and one other collection interval during the aspirin characterization phase for assay of ECP. This patient was omitted from the analysis of variance (ANOVA) statistical analysis of the ECP re- sults.
One patient discarded his urine during the final collection interval of the placebo arm of the double- blind zileuton phase. The corresponding value from the zileuton arm of the double-blind zileuton phase was omitted from pairwise analysis. Urine samples from
1050 Fischer et al. J ALLERGY CLIN IMMUNOL DECEMBER 1994
*(p < 0.0002) (paired ttest, for mean maximum symptom score, aspirin vs placebo. See Fig. 1 and text for details.
another subject could not be assayed during part of the placebo arm of the aspirin characterization phase, and this subject was excluded from pairwise analysis as appropriate. Inclusion or exclusion of these data did not affect the statistical significance of the results.
Statistical analysis Results are expressed as means _+ SEM. Mean leu-
kotriene levels are reported with the value of the lower limit of detection of the assay when values were assayed at or below this level. An undetectable level of nasal tryptase was averaged as 0.025 ng/ml (one half of the lower limit of detection of the assay for a 10-fold concentrated specimen). Significance between groups was calculated by ANOVA (2-factor with replication) for nasal tryptase, nasal histamine, and ECP data. Nasal symptoms and urinary histamine levels were analyzed by paired Student's t tests. Detection of serum tryptase levels and nasal leukotrienes was analyzed categorically by chi square analysis with Yates' correction when a value below the limit of detection of the respective assay was considered as absence of the mediator. Ap value of 0.05 or less was considered to indicate a significant difference between groups.
RESULTS Patients studied
Eight subjects entered the double-blind part of the protocol, and all completed. FEV1 and urinary LTE 4 values were reported previouslyS; the patient numbers in this report correspond to those in the prior report.
Aspirin characterization phase Nasal symptoms. Subjects experienced a pro-
found increase in nasal symptoms after aspirin administration (Fig. 3). The patients reached their peak nasal symptom score at an average of 160 _+ 30 minutes after threshold aspirin dosing. At baseline there was no difference in nasal symptoms during either arm of this phase. The mean of the prechallenge symptom scores was 2.1 _+ 0.7 on the aspirin challenge day and 1.4 + 0.4 on the placebo challenge day (p > 0.3). Aspirin caused a significant increase in nasal symptoms from a baseline score of 2.1 + 0.7 to a mean of the maximum symptom scores of 8.4 _+ 1.2 (p < 0.0007). Placebo caused no significant increase in nasal symptoms from a baseline score of t.4 + 0.4 to the mean of the maximal nasal symptom scores of 1.6 _ 0.4.
tase recovery compared with placebo challenge (p < 0.05) (Table II). Subjects reached their maximum nasal tryptase level 140 _+ 10 minutes after threshold aspirin administration. Among the patients with nasal lavage assayed for tryptase (see Methods section), there was no significant differ- ence in the mean baseline level of tryptase during the aspirin or placebo arm, 0.5 + 0.1 ng/ml and 0.4 _+ 0.2 ng/ml, respectively (Table II). The mean of the maximum tryptase levels during placebo administration was 0.5 _+ 0.1 ng/ml compared with 4.0 _ 2.7 ng/ml during aspirin administration. Hence, there was an increase of 3.5 _+ 2.6 ng/ml after aspirin ingestion compared with an increase of 0.1 -+ 0.2 ng/ml after placebo administration (p < 0.05, aspirin vs placebo, ANOVA).
Cysteinyl leukotrienes. Aspirin provocation caused a significant increase in leukotriene recov- ery from the nasal lavage of the subjects after aspirin ingestion compared with placebo (p < 0.05) (Table I1). There was no significant difference in the baseline level of cysteinyl leukotrienes on the day subjects received aspirin or placebo. The max- imum nasal leukotriene level reached 180.6 _+ 101.6 pg/ml after aspirin ingestion compared with 31.0 _+ 4.0 pg/ml after ingestion of placebo. This represented an increase of 152 pg/ml after aspirin administration versus 16 pg/ml after placebo ad- ministration (p < 0.05, aspirin vs placebo, chi square).
Histamine. Aspirin provocation also produced a significant increase in the concentration of hista- mine recovered from the lavagate compared with placebo (p < 0.05) (Table II). This maximum
J ALLERGY CLIN IMMUNOL Fischer et al. 1051 VOLUME 94, NUMBER 6
TABLE II. Nasal lavage mediators: Aspirin characterization phase
*p < 0.05 according to ANOVA after aspirin ingestion versus after placebo ingestion. tP < 0.05 according to chi square analysis after aspirin ingestion versus after placebo ingestion.
TABLE III. Serum tryptase levels and gastrointestinal symptoms during the aspirin characterization phase
Placebo Aspirin
Subject Base Max Base Max No. (ng/ml) (ng/ml) GI* (ng/ml) (ng/ml) GI*
*Presence or absence of gastrointestinal symptoms is denoted by plus sign or minus sign.
tNA: serum tryptase assay was not performed at these points.
increase in histamine occurred at 150 + 10 minutes after threshold aspirin dosing. There was no dif- ference in the baseline histamine levels during the aspirin or placebo arm. After aspirin administra- tion, histamine increased to a mean of the maxi- mum levels of 1.81 + 1.19 ng/ml compared with 0.14 + 0.08 ng/ml after placebo administration. This represented an increase of 1.73 +_ 1.16 ng/ml after aspirin ingestion compared with 0.08 -+ 0.08 ng/ml after placebo ingestion (p < 0.04, aspirin vs placebo, ANOVA).
ECP. There was a statistically significant differ- ence between ECP values after aspirin and placebo administration (p < 0.05) (Table II). However, there was significant variability in the baseline ECP levels between the aspirin and placebo arms such that the mean of the baseline ECP concentrations before placebo administration did not differ from the maximum concentration of ECP after aspirin ingestion.
Serum tryptase. Seven of the subjects had serum assayed for tryptase during the aspirin character- ization phase (Table III). At baseline, serum tryptase was not detectable for any subject except for patient 107 who had a level of 0.7 ng/ml during the placebo arm of the aspirin characterization phase. None of the subjects had a level of tryptase above 0.7 ng/ml after placebo ingestion. In con- trast, three of seven subjects had detectable serum levels of tryptase after aspirin ingestion (range, 3.6 to 9.1 ng/ml). All of these patients had gastroin- testinal symptoms. A fourth subject (patient 107) had a detectable serum tryptase level (3.5 ng/ml) during his initial single-blind challenge (data not shown), which was accompanied by a pronounced clinical reaction, including gastrointestinal symp- toms. This severe reaction necessitated a reduction in his aspirin dosage during subsequent visits, and tryptase was not detected at this lower dose.
Urine histamine
There was no difference in urinary histamine excretion at baseline or with aspirin challenge. At baseline, the mean histamine concentration was 34.8 + 9.9 txg/gm creatinine during the aspirin arm and 43.6 _+ 18.2 Ixg/gm creatinine during the placebo arm (p = 0.41). The maximum histamine concentration did not differ for the aspirin or placebo arm, 69.9 -+ 28.8 Ixg/gm creatinine and 70.4 _+ 35.6 txg/gm creatinine, respectively.
Double-bl ind zi leuton phase
Nasal symptoms. The increase in nasal symp- toms after aspirin ingestion was blocked by zileu- ton (Fig. 4). After 1 week of zileuton treatment, baseline nasal symptom scores, 1.0 + 0.4 versus 1.6 + 0.5, zileuton versus placebo, respectively, were unchanged (p = 0.31). There was a threefold difference in mean maximum nasal symptom scores after aspirin ingestion during the placebo arm versus the zileuton treatment arm (p < 0.0053). Aspirin ingestion caused an increase in nasal symptoms from a baseline score of 1.6 +_ 0.5
11)52 F ischer e t al. J ALLERGY CLIN IMMUNOL DECEMBER 1994
9 ~
oo 8- 69 7 ~
oE6 i 3-
o /I • E i i " P l acebo Baseline 1 2 3 4 5 6 Zlleuton
t t Time (hrs) Max
FIG. 4. Effect of aspirin on nasal symptoms during the double-blind zileuton phase (O = aspirin/placebo, • = aspirin/zileuton). The nasal symptom score, expressed as mean _+ SEM, is plotted against time after aspirin dosing. The mean of the maximum symptom scores (_+ SEM) for all eight subjects during each arm of this phase is shown at the right of the figure and indicated by Max. The threshold dose of aspirin was administered 2 hours after the subthreshold dose, and the points of administration are indicated by arrows, t p < 0.0053 (paired t test, for mean maximum symptom score, aspirin/placebo vs aspi- rin/zileuton). See Fig. 2 and text for details.
to a maximum of 5.5 _ 0.9 during placebo treat- ment (p < 0.0015). In contrast, aspirin ingestion caused no significant increase in nasal symptoms from a baseline score of 1.0 _+ 0.4 to a maximum of 1.6 _+ 0.6 during treatment with zileuton (p = 0.25) .
Nasal lavage med ia tors
The results of the assays for tryptase, cysteinyl leukotrienes, histamine, and ECP are shown in Fig. 5.
Tryptase. There was no difference in the baseline level of tryptase after 1 week of treatment with zileuton or placebo, 0.4 _+ 0.1 and 0.4 + 0.1 ng/ml. After threshold-dose aspirin challenge, the in- crease in nasal tryptase recovery was blocked by zileuton (Fig. 5, A) (p < 0.012 zileuton vs placebo, ANOVA). During the placebo arm, subjects reached their peak nasal tryptase level at 130 + 20 minutes after threshold aspirin ingestion. From a baseline value of 0.4 _+ 0.1 ng/ml during both the zileuton and placebo arms, the mean maximum tryptase value after aspirin administration in- creased to 2.8 + 0.9 ng/ml during placebo treat- ment compared with 0.7 -+ 0.2 ng/ml during zileu- ton treatment.
Cysteinyl leukotrienes. There was an increase in the recovery of cysteinyl leukotrienes from nasal lavage during the placebo treatment arm after aspirin challenge; this increase was significantly inhibited by zileuton treatment (Fig. 5, B) (p < 0.05, zileuton vs placebo, chi square). The baseline
FIG. 5. Nasal lavage mediator assessments dur ing the double-b l ind zi leuton phase for (A) tryptase, (B) cysteinyl leukotr ienes, (C) histamine, and (D) ECP. The concentra- t ion (mean _+ SEM) is shown at each col lect ion t ime point. T ime 0 corresponds to subthreshold aspir in dosing. T ime 2 hours corresponds to threshold aspir in dosing. The value at each interval reflects the mean _+ SEM concen- t rat ion for all e ight subjects except as indicated by "n = " in charts A and C. The mean of the m a x i m u m media tor concentrat ions is shown at the r ight of each f igure and indicated by Max (see text for details).
levels of cysteinyl leukotrienes were nearly identi- cal during the zileuton and placebo arms, 27.0 _+ 0 pg/ml and 27.4 _+ 0.4 pg/ml. The maximum nasal leukotriene levels were 150.1 + 82.6 pg/ml during
J ALLERGY CLIN IMMUNOL F i sche r et al. 1053 VOLUME 94, NUMBER 6
placebo t reatment versus 65.2 _+ 37.2 pg/ml during t reatment with zileuton.
Histamine. Zileuton t reatment did not alter his- tamine recovery after threshold aspirin ingestion (Fig. 5, C) (p = 0.6 zileuton vs placebo, ANOVA). Baseline levels of histamine were !.09 + 1.02 ng/ml and 0.09 _ 0.05 ng/ml during the zileuton and placebo t reatment arms, respectively.* After threshold aspirin dosing, the maximum histamine levels were 0.27 +_ 0.22 ng/ml and 0.33 _+ 0.26 ng/ml for the zileuton and placebo treatment arms, respectively.
EC7 ~. Zileuton treatment did not alter E C P recovery compared with placebo (Fig. 5, D) (p = 0.9 zileuton vs placebo, ANOVA). The baseline levels of ECP were 19.8 _+ 14.0 ng/ml and 21.0 _+ 15.8 ng/ml with zileuton and placebo treatment, respectively. The maximum ECP levels were 45.5 _+ 16.9 ng/ml and 51.7 _+ 26.1 ng/ml with zileuton and placebo treatment, respectively.
Serum tryptase. Serum tryptase levels during the double-blind zileuton phase are shown in Table IV. All subjects had tryptase levels of 0.5 ng/ml or less at baseline during both arms of this phase. Two of eight subjects had detectable serum tryptase levels after receiving aspirin with placebo. This increase was detected at all time points beyond 2 hours after threshold aspirin dosing. Tryptase was not detected in these two subjects while they received aspirin with zileuton except at the final sampling point (i.e., 4 hours after thresh- old ASA dosing) in patient 103. The failure to de tec t tryptase in the serum of patient 108 was likely due to the fact that the subject received a lower dose of aspirin (40 mg vs 60 mg) in this phase compared with the aspirin characterization phase.
Ur ine h is tamine
Pretreatment with zileuton produced a lower, although not statistically significant, mean baseline histamine level. The mean histamine concentra- tion was 25.5 _+ 6.6 Ixg/gm creatinine after treat- ment with zileuton versus 35.0 + 10.3 Ixg/gm crea- tinine after placebo treatment (29 = 0.11). Zileuton treatment produced no significant effect on hista- mine excretion after aspirin ingestion. The maximum histamine levels were 73.9 + 43.5 txg/gm creatinine during zileuton treatment and 44.8 _+ 9.5 Ixg/gm creatinine during placebo treatment (p = 0.48).
*These values reflect the fact that one subject (no. 104) had a baseline nasal histamine concentration of 7.21 ng/ml during the zileuton arm. Excluding this outlier, the mean of the baseline histamine concentrations during the zileuton arm was 0.07 -+ 0.04 ng/ml.
TABLE IV. Serum tryptase and gastrointestinal symptoms during the double-blind zileuton phase
*Presence or absence of gastrointestinal symptoms is denoted by a plus sign or a minus sign.
*Undetectable except at the final time point 4 hours after threshold aspirin ingestion.
D I S C U S S I O N
5-Lipoxygenase products of arachidonic acid metabolism such as the cysteinyl leukotrienes are of importance in the pathologic reaction to aspirin in the subset of aspirin-sensitive patients with asthma who hyperexcrete LTE 4 after aspirin chal- lenge. Cysteinyl leukotrienes are potent broncho- constrictors, induce mucus secretion, and increase vascular permeability. 19-2a They have been found in nasal secretions of patients with aspirin sensitivity after aspirin challenge, patients with chronic rhi- nosinusitis, and atopic subjects after nasal allergen challenge. 16,23-a6 Overproduction of these com- pounds could account for both the upper and lower airway abnormalities in the clinical syn- drome of aspirin sensitivity. Our data indicate that leukotrienes are critical to the upper airway re- sponse to aspirin and that during this response mast cells are activated and thus may serve as one source of these leukotrienes.
The role of mast cells and the mediators they release in the response to aspirin has been unclear. Histamine, a mast cell product, has been detected in the plasma after aspirin challenge. 27, 28 Ferreri et al. 23 have shown increases in nasal histamine after aspirin challenge in three of four aspirin-sensitive patients with asthma, whereas Kowalski et al. 2s showed a greater than twofold increase in nasal histamine in four of nine aspirin-sensitive patients with asthma. However, despite the reported recov- ery of histamine in these biologic fluids, t reatment with antihistamines has produced equivocal re-
1054 Fischer et al. J ALLERGY CLIN IMMUNOL DECEMBER 1994
sponses in aspirin challenges. 29-31 In addition, be- cause nasal histamine may arise from mast cells or basophils and because both cell types have been described in the nasal mucosa, 32 the aforemen- tioned studies cannot be used to differentiate the cellular source of histamine production. Others have suggested that mast cells are not involved in the aspirin reaction in the nose. Picado et al. z4 showed no alteration in prostaglandin D2 in the nasal lavage of aspirin-sensitive patients with asthma after aspirin ingestion. Prostaglandin D2 has been consid- ered indicative of mast cell activation, 33 and Picardo et al.24 interpreted this finding as evidence against heightened mast cell sensitivity.
In our study we used tryptase, an enzyme that is specific to mast cells, 12 as an indicator of mast cell activation. We found a significant increase in the recovery of tryptase in the nasal lavage of aspirin- sensitive patients with asthma after aspirin inges- tion. This demonstrates that mast cells are indeed activated during the aspirin-induced nasal reac- tion. The failure of Picado et al. 24 to detect a mast cell product might be due to several factors. Aspi- rin itself inhibits the cyclooxygenase enzyme re- sponsible for the production of prostaglandin D 2. It is also possible that the stimulus to mast cell activation may result in selective release or pro- duction of mast cell mediators? 4
In addition to elevations of nasal tryptase, we were able to detect elevations in serum tryptase in four of eight patients, suggesting more widespread mast cell activation. Elevations occurred in the aspirin characterization phase and in one (patient 107) during the single-blind phase. In the latter case the severity of the clinical reaction was such that an aspirin dose alteration was necessary and no subsequent elevation in serum tryptase was detected. In concert with the findings of others, these subjects had severe clinical responses includ- ing gastrointestinal symptoms 35, 36 consonant with the suggestions of Sladek and Szczeklik 36 that the more severe the reaction, that is, the higher the dose of aspirin administered, the more widespread is the mast cell activation.
Eosinophils may be involved in the aspirin re- sponse because they are capable of leukotriene production, I1 and the nasal polyps of aspirin-sen- sitive patients with asthma contain increased num- bers of eosinophils. 9 However, our measurement of ECP did not clearly demonstrate eosinophil activation during the aspirin reaction. Although we detected an increase in nasal ECP after aspirin ingestion compared with placebo, this must be considered of questionable clinical relevance be-
cause the mean maximal concentration after aspi- rin ingestion was nearly equal to the mean concen- tration before placebo administration (Table II). In addition, such an increase was not observed during the placebo arm of the double-blind zileu- ton phase. Others have also had difficulty demon- strating eosinophil activation in aspirin-induced responses. Although no prior investigation has measured an eosinophil granule protein in the nasal lavage of aspirin-sensitive patients with asthma after aspirin challenge, Sladek and Szczek- lik 36 noted a trend toward elevated ECP levels in the serum after aspirin ingestion, which did not achieve statistical significance. However, in a re- cent study with bronchoalveolar lavage fluid Sladek et al. 37 found a decrease in ECP levels after lysine-aspirin inhalation. The precise role of eosi- nophils in the aspirin response in the nose requires further investigation with less variable markers of eosinophil activation.
Our results with zileuton, a potent and selective 5-1ipoxygenase inhibitor, 14,15 demonstrate that 5-1i- poxygenase products are critical to the nasoocular response evoked by aspirin in aspirin-sensitive patients. Zileuton treatment blocked the appear- ance of nasal symptoms to the level achieved with placebo alone (Figs. 3 and 4). The results with zileuton also provide further insight into the cellu- lar origin and role of inflammatory mediators in aspirin-induced reactions. As expected, zileuton treatment blunted the increase in cysteinyl leuko- trienes recovered in the nasal lavage. However, we also found that zileuton blocked the recovery of tryptase from nasal lavage, suggesting that mast cell activation in the nasal mucosa was suppressed by treatment with the 5-1ipoxygenase inhibitor as well. Trends in the data suggest that this effect carried over to the systemic release of tryptase, where zileuton treatment blunted the increase in serum tryptase in those subjects who had prior serum tryptase elevations. The notable exception was patient 103. However, tryptase was only de- tected at the final sampling point, 4 hours after the threshold dose of aspirin and possibly at a time when zileuton levels had fallen.
Zileuton's effect on nasal and serum tryptase after aspirin ingestion suggests that 5-1ipoxygenase products, such as leukotrienes, act in autocrine or paracine fashion to cause the release of preformed mast cell mediators such as tryptase. However, other explanations need to be considered. It is possible that the observed effects relate to the potency of leukotrienes in inducing alterations in vascular permeability51 In the absence of leuko-
J ALLERGY CLIN IMMUNOL F i sche r et al. 1055 VOLUME 94, NUMBER 6
t r ienes, t ryp tase may not diffuse into a subjec t ' s
nasal secre t ions because of a lack of a d e q u a t e
permeabi l i ty . Al te rna t ive ly , it is poss ib le that zi leu- ton acts d i rec t ly to s tabi l ize mast cell m e m b r a n e s in add i t ion to its ac t ion as a 5-1ipoxygenase inhib- i tor. A l t h o u g h we do not have any da t a to address this ques t ion directly, a s t ruc tura l ly s imilar com- p o u n d has been shown to p reven t Toxoplasrna s p e c i e s - i n d u c e d mas t cell deg ranu la t ion in vitro. 3a
C o n s o n a n t with the act ivat ion of mas t cells, as ind ica ted by our findings with t ryptase , we were also able to de tec t a small but significant increase in nasa l h i s tamine af ter aspir in chal lenge. Such an increase was not no t ed in the urine, poss ibly because of i n a d e q u a t e re lease or r ap id deg rada - t ion. Dur ing the z i leu ton t r e a t m e n t phase of the study, i n a d e q u a t e amount s of nasal h i s tamine were de t ec t ed to assess the effect of z i leu ton on the re lease of this med ia to r . The r eason for the lower a m o u n t of h i s tamine de t ec t ed is unclear . His ta- mine is d e g r a d e d rap id ly and diffuses quickly into the b l o o d s t r e a m and on to t issue surfaces. The inc reased n u m b e r of lavages dur ing the double - b l ind z i l eu ton phase of the p ro toco l may have had a d i s p r o p o r t i o n a t e effect on h i s tamine c o m p a r e d with t ryptase , which diffuses m o r e slowly. 39
In summary , a l though we were only able to s tudy eight subjects, our da t a have shown tha t mas t cells, which are a known source of leukot r ienes , a re ac t iva ted in the nasal r e sponse to aspir in as dem- ons t r a t ed by the de tec t ion of nasal t ryptase af ter aspir in chal lenge. Cysteinyl l euko t r i enes and hista- mine, which can be p r o d u c e d by mast cells, were d e t e c t e d as well. The occur rence of nasal symp- toms, as well as act ivat ion of mas t cells, in r e sponse to aspir in was b locked by z i leuton, an inhib i tor of 5-1ipoxygenase. This conf i rms that 5-1ipoxygenase p roduc t s a re cri t ical to the d e v e l o p m e n t of A S A - induced reac t ions in the nose. I t also suggests that 5-1ipoxygenase p roduc t s may have a role in the act ivat ion of mas t cells dur ing this react ion .
We thank Dr. Marianne Castells for her invaluable advice concerning the analysis of tryptase. We also thank Janet Hurwitz, BA, MT(ASCP), and Amy Hall for expert technical assistance; Bernard Ransil, MD, PhD, for his statistical advice; Lori Scala, RPH, for prepara- tion of the aspirin and placebo doses; the staff of the General Clinical Research Center; and Carmen M. Hall for assistance in preparation of the manuscript.
REFERENCES
1. Szczeklik A, Gryglewski RJ, Czerniawska-Mysik G. Rela- tionship of inhibition of prostaglandin biosynthesis by an-
algesics to asthma attacks in aspirin-sensitive patients. Br Med J 1975;1:67-9.
2. Christie PE, Tagari P, Fordhutchinson AW, et al. Urinary leukotriene-E4 concentrations increase after aspirin chal- lenge in aspirin-sensitive asthmatic subjects. Am Rev Respir Dis 1991;143:1025-9.
3. Knapp HR, Sladek K, FitzGerald GA. Increased excretion of leukotriene E 4 during aspirin induced asthma. J Lab Clin Invest 1992;119:48-51.
4. Kumlin M, Dahlen B, Bjorck T, Zetterstrom O, Granstrom E, Dahlen SE. Urinary excretion of leukotriene E 4 and ll-dehydro-thromboxane B2 in response to bronchial prov- ocations with allergen, aspirin, leukotriene D 4 and hista- mine in asthmatics. Am Rev Respir Dis 1992;146:96-103.
5. Christie PE, Tagari P, Ford-Hutchinson AW, et al. Urinary leukotriene g 4 after lysine-aspirin inhalation in asthmatic subjects. Am Rev Respir Dis 1992;146:1531-4.
6. Christie PE, Smith CM, Lee TH. The potent and selective sulfidopeptide leukotriene antagonist, SK&F 104353, inhib- its aspirin-induced asthma. ~m Rev Respir Dis 1991;144: 957-8.
7. Dahlen B, Kumlin M, Margolskee D J, et al. The leuko- triene-receptor antagonist MK-0679 blocks airway ob- struction induced by bronchial provocation with lysine- aspirin in aspirin-sensitive asthmatics. Eur Respir J 1993;6: 1018-26.
8. Israel E, Fischer A, Rosenberg M, et al. The pivotal role of 5-1ipoxygenase products in the reaction of aspirin-sensitive asthmatics to aspirin. Am Rev Respir Dis 1993;148:1447- 51.
9. Yamashita T, Tsuji H, Maeda N, Tomoda K, Kumazawa T. Etiology of nasal polyps associated with aspirin-sensitive asthma. Rhinology 1989;8:15-24.
10. Takasaka T, Kaku Y, Hozawa K. Mast cell degranulation in nasal polyps. Acta Otolaryngol 1986;430:39-48.
11. Shore S, Austen KF, Drazen JM. Eicosanoids and the lung. In: L'Enfant C, Massaro D, eds. Lung biology in health and disease: lung cell biology. New York: Marcel Dekker, 1989:1011-89.
12. Schwartz LB. Monoclonal antibodies against human mast cell tryptase demonstrate shared antigenic sites on subunits of tryptase and selective localization of the enzyme to mast cells. J Immunol 1985;134:526-31.
13. Olsson I, Venge P, Spitznagel NK, Lehrer RI. Arginine-rich cationic proteins of human eosinophil granules. Lab Invest 1977;36:493-500.
14. Rubin P, Dube L, Braeckman R, et al. Pharmacokinetics, safety, and ability to diminish leukotriene synthesis by zileuton, an inhibitor of 5-1ipoxygenase. Agents Actions Suppl 1991;35:103-16.
15. Carter GW, Young PR, Albert DH, et al. 5-Lipoxygenase inhibitory activity of zileuton. J Pharmacol Exp Ther 1991; 256:929-37.
16. Creticos PS, Peters SP, Adkinson NF, et al. Peptide leu- kotriene release after antigen challenge in patients sensitive to ragweed. N Engl J Med 1984;310:1626-30.
17. Castells M, Schwartz LB. Tryptase levels in nasal-lavage fluid as an indicator of the immediate allergic response. J ALLERGY CLIN IMMUNOL 1988;82:348-55.
18. Igarashi Y, Lundgren JD, Doerfler ME, Shelhamer JH, Kaliner MA, White MV. Human neutrophil-derived hista- mine-releasing activity (HRA-N) causes the release of serotonin but not arachidonic acid metabolites from rat basophilic leukemia cells. J ALLERGY CLIN IMMUNOL 1992; 89:1085-97.
1056 Fischer et al. J ALLERGY CLIN IMMUNOL DECEMBER 1994
19. Drazen JM, Austen KF, Lewis RA, et al. Comparative airway and vascular activities of leukotrienes C-1 and D in vivo and in vitro. Proc Natl Acad Sci U S A 1980;77:4354-8.
20. Marom Z, Shelhamer JH, Bach MK, Morton DR, Kaliner M. Slow-reacting substances, leukotrienes C4 and D4, in- crease the release of mucus from human airways in vitro. Am Rev Respir Dis 1982;126:449-51.
21. Dahlen SE, Bjork J, Hedqvist P. Leukotrienes promote plasma leakage and leukocyte adhesion in postcapillary venules: in vivo effects with relevance to the acute inflam- matory response. Proc Natl Acad Sci U S A 1981;78:3887- 91.
22. Hoover RL, Karnovsky MJ, Austen KF, Corey EJ, Lewis RA. Leukotriene B 4 action on endothelium mediates aug- mented neutrophil/endothelial adhesion. Proc Natl Acad Sci U S A 1984;81:2191-3.
23. Ferreri NR, Howland WC, Stevenson DD, Spiegelberg HL. Release of leukotrienes, prostaglandins, and histamine into nasal secretions of aspirin-sensitive asthmatics during reac- tion to aspirin. Am Rev Respir Dis 1988;137:847-54.
24. Picado C, Ramis I, Rosello J, et al. Release of peptide leukotriene into nasal secretions after local instillation of aspirin in aspirin-sensitive asthmatic patients. Am Rev Respir Dis 1992;145:65-9.
25. Kowalski ML, Sliwinska-Kowalska M, Igarashi Y, et al. Respiratory pathophysiologic responses. Nasal secretions in response to acetylsalicylic acid. J ALLERGY CLIN IMMU- NOL 1993;91:580-98.
26. Knapp HR. Reduced allergen-induced nasal congestion and leukotriene synthesis with an orally active 5-1ipoxyge- nase inhibitor. N Engl J Med 1990;323:1745-8.
27. Stevenson DD, Arroyave CM, Bhat KN, Tan EM. Oral aspirin challenges in asthmatic patients: a study of plasma histamine. Clin Allergy 1976;6:493-505.
28. Szmidt M, Grzelewska-Rzymowska I, Rozniecki J, Kowal- ski ML, Rychlicka I. Histaminemia after aspirin challenge in aspirin-sensitive asthmatics. Agents Actions 1981;11: 105-7.
29. Delaney JC. The effect of ketotifen on aspirin-induced asthmatic reactions. Clin Allergy 1983;13:247-51.
30. Szczeklik A, Serwonska M. Inhibition of idiosyncratic reac- tions to aspirin in asthmatic patients by clemastine. Thorax 1979;34:654-7.
31. Szczeklik A, Czerniawska-Mysik G, Scrwonska M, Kuklin- ski P. Inhibition by ketotifen of idiosyncratic reactions to aspirin. Allergy 1980;35:421-4.
32. Kawabori S, Denburg JA, Schwartz LB, et al. Histochemi- cal and immunohistochemical characteristics of mast cells in nasal polyps. Am J Respir Cell Mol Biol 1992;6:37-43.
33. Naclerio RM, Proud D, Togias AG, et al. Inflammatory mediators in late antigen-induced rhinitis. N Engl J Med 1985;313:65-70.
34. Eggleston PA, Kagey-Sobotka A, Proud D, et al. Disasso- ciation of the release of histamine and arachidonic acid metabolites from osmotically activated basophils and hu- man lung mast cells. Am Rev Respir Dis 1990;141:960-4.
36. Sladek K, Szczeklik A. Cysteinyl leukotrienes overproduc- tion and mast cell activation in aspirin-provoked broncho- spasm in asthma. Eur Respir J 1993;6:391-9.
37. Sladek K, Dworski R, Soja J, et al. Eicosanoids in bron- choalveolar lavage fluid of aspirin-intolerant patients with asthma after aspirin challenge. Am J Respir Crit Care Med 1994;149:940-6,
38. Henderson WR, Chi Emil Y. Mast cell-mediated toxo- plasma gondii cytotoxicity: role of 5-1ipoxygenase products [Abstract]. The 8th International Conference on Prostag- landins and Related Compounds; Montreal, Canfida 1992: 372.
39. Schwartz LB. Cellular inflammation in asthma: neutral proteases of mast cells. Am Rev Respir Dis 1992;145:S18- $21.
