from aspirin-sensitive asthmatic subjectsthorax 1996;51:64-70 effect ofendobronchial aspirin...

Thorax 1996;51:64-70

Effect of endobronchial aspirin challenge on

inflammatory cells in bronchial biopsy samplesfrom aspirin-sensitive asthmatic subjects

Shuaib Nasser, Pandora E Christie, Rudolf Pfister, Ana R Sousa, Andrew Walls,Michael Schmitz-Schumann, Tak H Lee

Department of AlIlergyand RespiratoryMedicine, UMDS,Guy's Hospital,London SEI 9RT, UKS NasserA R SousaT H Lee

HochgebirgsklinikDavos-Wolfgang,SwitzerlandP E ChristieR PfisterM Schmitz-Schumann

Departm ent of

Immunopharma-cology, SouthamptonGeneral Hospital,SouthamptonS016 6YD, UKA Walls

Correspondence to:Professor T H Lee.

Received 26 May 1995Returned to authors24 August 1995Revised version received7 September 1995Accepted for publication12 September 1995

AbstractBackground - The aspirin-induced bron-choconstriction in patients with aspirin-

sensitive asthma is caused by cysteinylleukotriene release. The cellular source ofthe leukotrienes is unknown. The inflam-matory cell infiltrate in bronchial biopsysamples from seven aspirin-sensitive asth-matic (ASA) subjects and eight non-ASAsubjects before and after local challengewith lysine aspirin was therefore ex-

amined.Methods - Using flexible bronchoscopy,airway mucosal biopsy samples were takenand lysine aspirin solution was placeddirectly onto a carina of the contralaterallung. Twenty minutes later a second seriesof biopsy samples was taken from the siteof the local endobronchial lysine aspirin

challenge. The biopsy samples were doubleimmunostained with a rabbit polyclonalantibody to the enzyme 5-lipoxygenase andmonoclonal antibodies to mast cells(AA1), neutrophils (NP57), macrophages(EBMII), T lymphocytes (anti-CD3), andtotal (BMK13) and activated eosinophils(EG2).Results -A decrease in both absolute mastcell numbers staining with mast cell tryp-tase (AA1) and the percentage ofmast cellsco-immunostaining with 5-lipoxygenasewas seen in the ASA patients after lysineaspirin challenge compared with the non-

ASA control group. There was also an

increase in the numbers of activatedeosinophils (EG2) in the ASA subjectscompared with the non-ASA group. Nochanges were observed in the totalnumbers of macrophages (EBMIl),neutrophils (NP57), total eosinophils(BAK13), and T lymphocytes (anti-CD3)after challenge with lysine aspirin.Conclusions - The decrease in numbers ofmast cells staining for tryptase and theincrease in activated eosinophils afterendobronchial challenge with lysine as-

pirin may represent degranulation ofthesecell types, and may be an early event as-

sociated with aspirin-sensitive reactionsin ASA subjects.(Thorax 1996;51:64-70)

Keywords: aspirin-sensitive asthma, endobronchiallysine aspirin, mast cells, eosinophils.

Asthmatic patients with sensitivity to aspirinand other non-steroidal anti-inflammatorydrugs are often corticosteroid dependent andhave accompanying symptoms of rhinosinusitisincluding rhinorrhoea, nasal congestion, an-osmia, loss of taste, and recurrent severe nasalpolyposis.' Upon ingestion of aspirin or othercyclo-oxygenase inhibitors, these individualshave increased cysteinyl leukotriene release asdetected in the urine,23 nasal lavage fluid,45and bronchial lavage fluid6 which does notoccur in aspirin-tolerant subjects. Inhibitionof the synthesis or the action of cysteinylleukotrienes by a 5-lipoxygenase inhibitor78 orleukotriene receptor antagonists9"' leads to in-hibition of the fall in forced expiratory volumein one second (FEV,) and amelioration of thenaso-ocular effects induced by aspirin in as-pirin-sensitive patients, strongly supporting theinvolvement of cysteinyl leukotrienes in thisform of asthma. Despite the evidence for theinvolvement of cysteinyl leukotrienes in aspirinsensitivity, their cellular source is unknown andthere are no previous data on the nature of theinflammatory process in the airways of patientswith aspirin-sensitive asthma.We have already examined the basal in-

flammatory cell infiltrate in bronchial biopsysamples from 12 aspirin-sensitive asthmatic(ASA) subjects and eight non-ASA controlindividuals" and found no significant differ-ence in the total number of cells staining for5-lipoxygenase between the two groups. As apercentage of total 5-lipoxygenase immuno-staining cells, there were significantly moremast cells and total eosinophils and significantlyfewer total macrophages in the bronchial biopsysamples from ASA subjects than from non-ASA subjects. The numbers of neutrophils, Tlymphocytes, and activated eosinophils weresimilar for the two groups. The increased num-bers of eosinophils and mast cells identified inthe bronchial tissue from ASA subjects may bethe source(s) of the enhanced cysteinyl leuko-triene production observed in these uniqueindividuals.

In order to assess the nature of the cellularpopulation following lysine aspirin pro-vocation, we have now studied the inflam-matory cell infiltrate and the cellular expressionof 5-lipoxygenase in bronchial biopsy tissuesfrom asthmatic subjects with aspirin sensitivitybefore and after endobronchial lysine aspirinchallenge and have compared the results withthose in control non-ASA individuals.

64

on August 17, 2021 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.51.1.64 on 1 January 1996. D

ownloaded from

http://thorax.bmj.com/

Mast cells and eosinophils in aspirin-induced asthma

Table 1 Clinical characteristics of aspinin-sensitive subjects

FEV, FEy, (%o Threshold dose of lysine Skin EosinophilsSubject no. Sex Age Drugs (7) predicted) aspirin (mg/ml) test (103p/l)1 M 57 nBDP 3-76 104 25 0.352 F 56 BUD, S, T, IB 2-1 91 2-5 0 393 M 39 BDP, S, T, nBUD 4-11 109 2-5 + 1-014 M 56 BDP, S, IB, F 3-52 104 7-5 + 0-345 M 54 PRED, BUD, S, T 3-2 93 2-5 + 0-366 F 26 BDP, S 2-93 87 25 + 0-327 M 24 BUD, nBUD, S 3-08 81 1 25 0-32Mean (SE) 44 6 (5-6) 3-24 (0 25) 95 6 (3-9) 9-5 (4-1) 0-44 (0-1)

PRED = prednisolone; BDP = beclomethasone dipropionate; BUD = budesonide; n = nasal route; S = salbutamol; T= theophylline; IB = ipratropium bromide; F =fenoterol.

Table 2 Clinical characteristics of non-aspirin-sensitive subjects

Subject FEV, FEV, Skin Eosinophilsno. Sex Age Drugs (1) (% predicted) test (1JU/pl)1 F 53 BDP, DSCG, R 3-4 133 0-142 M 35 DSCG, R 4-57 102 + 0-413 M 49 BUD, S 5-3 126 + 0-234 M 34 S, T, DSCG 3-68 82 + 0-295 M 25 IB, F 3.93 95 + 0-926 M 25 BUD, IB, F 3-52 78 + 0-087 M 30 BDP, S 2-28 52 + 0-468 F 37 PRED, BUD, nBUD, S, T 1-59 54 + 0.19Mean (SE) 36 (3 6) 3-81 (0 36) 90-2 (10-2) 0 34 (0-13)

PRED=prednisolone; BDP=beclomethasone dipropionate; BUD=budesonide; n=nasal route; S=salbutamol; T=theophylline; IB=ipratropium bromide;DSCG=disodium cromoglycate; F=fenoterol; R=reproterol.

MethodsSUBJECTSTwo groups ofnon-smoking asthmatic subjectswere studied during their stay at the Hoch-gebirgsklinik Davos-Wolfgang. Seven ASA sub-jects (five men) ofmean age 44-6 (range 24-57)years with a mean (SE) FEV1 of 95-6 (3-9)%predicted values and a control group of eight(six men) non-ASA subjects of mean age 36(range 25-53) years and a mean FEV, of 90-2(10-2)% predicted values underwent endo-bronchial lysine aspirin challenge. Of theseven ASA subjects, one took prednisoloneorally, six took regular inhaled corticosteroids,and three took regular nasal corticosteroids(table 1). Of the eight non-ASA subjects, onetook prednisolone orally, five took regular in-haled corticosteroids, and one subject took reg-ular nasal corticosteroids; three subjects wereon no regular glucocorticoid treatment (table2). All subjects were taking inhaled or oral 12agonists as required.

STUDY DESIGNThe diagnosis of aspirin sensitivity was madeon the basis of history and confirmed by a20% or greater fall in FEV, by lysine aspirininhalation challenge. The non-ASA controlsubjects also underwent lysine aspirin challengeand, if negative, this was confirmed by in-cremental oral aspirin challenge to 500 mgaspirin. All subjects were skin prick tested tocommon aeroallergens and atopy was definedas the presence of at least two positive reactions(a weal 3 mm greater than saline control) toskin prick tests with the following allergens:cat fur, dog hair, grass pollen, and Derma-tophagoides pteronyssinus. Blood was taken foreosinophil count. At least one week after aspirinchallenge the patients underwent bron-choscopy and had bronchial biopsy samplestaken before and after endobronchial lysineaspirin challenge. The biopsy samples were

immediately snap frozen in embeddingmedium and stored at -80°C. Each subjectunderwent only one endobronchial challengeprocedure.

All subjects gave written and informedconsent and the study was approved by theHochgebirgsklinik Davos-Wolfgang ethicscommittee and performed in accordance withthe Helsinki declaration.

LYSINE ASPIRIN INHALATION CHALLENGEMeasurements of FEV, were made using aspirometer (Micromedical Ltd, Rochester,Kent, UK). Three recordings of FEV, weremade at each time point and the highest valuewas used. Lysine aspirin inhalation challengewas performed by the method of Schmitz-Schumann and co-workers.'2 Lysine aspirin(Synthelabopharma, Lausanne, Switzerland),a powder containing 900 mg lysine acetyl-salicylate and 100 mg glycine, was diluted in5 ml water to produce a solution of 180 mg/ml(0-55 mol/1) lysine acetylsalicylate (equivalentto 100 mg/ml acetylsalicylic acid). This solutionwas diluted with sodium chloride to producea range of increasing doubling concentrationsfrom 1-25 mg/ml to 25 mg/ml (3-8 mmol/I to76 1 mmol/l). Nebulised solutions of 1 ml lys-me aspirin were administered at room tem-perature using a Heyer nebuliser (Carl HeyerGmbH, Bad Ems, Germany). Delivery of airto the nebuliser was regulated to an output of8 1/min and under these conditions the ne-buliser delivers droplets with a mass medianaerodynamic diameter of 5,um. After baselinemeasurements of FEVI, the subjects inhaledthe aerosol via a mouthpiece during normaltidal breathing. If the decrease in FEV, fol-lowing nebulised normal saline was <10% thesubjects underwent lysine aspirin challenge.Incremental doubling concentrations of lysineaspirin were used for bronchial challenge at 30minute intervals, with FEV, measurements at

65



j.com/


ownloaded from


Nasser, Christie, Pfister, Sousa, Walls, Schmitz-Schumann, Lee

15 and 30 minutes until a fall in FEV, of at least20% was achieved (threshold concentration oflysine aspirin). A fall in FEV, of between 10%and 20% was allowed to return to within 10%of baseline before further lysine aspirin ad-ministration. Further measurements of FEV,were continued for at least four hours after apositive reaction and salbutamol was ad-ministered if necessary.

BRONCHOSCOPYBronchoscopies were all performed at the sametime of day (08.00 hours) and by the sameindividual in accordance with current workingparty guidelines.13 Thirty minutes before theprocedure all patients received 1 25mg sal-butamol by nebulisation. After premedicationwith 05 mg atropine intramuscularly, 10 mgdiazepam orally, and topical preservative-free2% lignocaine, a fibreoptic bronchoscope(Olympus model BF-IT20D, Tokyo, Japan)was inserted orally and two or three mucosalbiopsy samples were taken using non-toothedforceps from a right middle or lower lobe seg-mental carina. Supplemental oxygen was sup-plied via a nasal cannula at 4 1/min throughoutthe procedure and pulse and arterial oxygensaturation were monitored using a pulse ox-imeter. After endobronchial challenge withlysine aspirin via the bronchoscope the in-strument was removed. It was re-introduced20 minutes later after additional local an-aesthesia, if necessary. Further biopsy sampleswere taken from the sites of the endobronchialchallenges in the left lung. All subjects weremonitored clinically during and for four hoursafter the procedure with measurements ofFEV,and to allow recovery from the sedation.

ENDOBRONCHIAL LYSINE ASPIRIN CHALLENGEOne millilitre of a solution of lysine aspirincorresponding to the threshold concentrationwas drawn into a soft plastic catheter and in-troduced into the bronchoscope. The cathetertip was first positioned on the carina betweenthe lingula and the upper division bronchusand then on the carina of the left upper lobe.At each of the two sites one drop (20 ,ul) oflysine aspirin solution, at the previously de-termined threshold concentration, was placedgently on the carina and dispersed by lateralmovements of the instrument without touchingthe bronchial mucosa. In the non-ASA subjectsthe same protocol was followed but the max-imum challenge concentration of 25 mg/mllysine aspirin was used. No previous reportsof endobronchial lysine aspirin challenge wereavailable and we therefore performed the chal-lenge in two ASA and two non-ASA subjectsin preliminary studies, in the manner describedabove. Immunohistochemical analyses of thebiopsy samples, carried out in a blinded man-ner, revealed that this dose of lysine aspirinelicited an apparent increase in the numberof cells staining with 5-lipoxygenase in ASAsubjects compared with baseline. There wereno changes in the two control non-ASA biopsysamples.

ANTIBODIESThe antibody to the enzyme 5-lipoxygenasewas a rabbit polyclonal (kindly donated byJ Evans, Merck Frosst Laboratories, Dorval,Quebec, Canada) and used at a dilution of 1/1000. All the primary antibodies used for cellidentification were mouse monoclonal IgG1antibodies and comprised the following: AAl(Dako, High Wycombe, UK), a mast cell tryp-tase marker used at a dilution of 1/50; EBM 11(Dako), a pan-macrophage marker recognisingCD68 used at 1/100 dilution; anti-CD3(Dako), a pan-T lymphocyte marker used at1/50 dilution; NP57 (Dako), an anti-humanneutrophil elastase antibody used at 1/600dilution; BMK13 (Cymbus Bioscience Ltd,Chilworth, UK), a pan-eosinophil marker re-cognising major basic protein used at 1/40dilution; EG2 (Kabi Pharmacia, MiltonKeynes, UK) which detects secreted eosinophilcationic protein and therefore is a marker foractivated eosinophils and was used at 1/100dilution. The secondary antibodies were a goatanti-mouse alkaline phosphatase conjugatedantibody (Sera-lab, Crawley Down, UK) andused at a dilution of 1/100, and a swine anti-rabbit biotinylated antibody used at a dilutionof 1/200.

IMMUNOSTAININGThe bronchial biopsy samples were im-mediately mounted in OCT and snap frozenusing isopentane (BDH) cooled down in liquidnitrogen (Bayer, Basingstoke, UK). Thesamples were stored at - 80°C until furtheranalysis. Sections 5,um thick were cut fromfrozen bronchial biopsy samples onto slidescoated with 3-aminopropyltriethoxy-silane(Sigma, Dorset, UK) to improve adhesion.'4The sections were cut from the centre of thebiopsy sample to ensure as large a cross sec-tional area as possible for immunostaining.They were then fixed in acetone and stored at- 70°C until used for immunohistochemistry.One section was examined for each cell marker.The biopsy samples were double immuno-stained using a rabbit polyclonal antibody tothe enzyme 5-lipoxygenase concurrently withone monoclonal antibody to each cell type:neutrophils (NP57); macrophages (EBMl 1);total (BMK13) and degranulated eosinophils(EG2); mast cells (AA1), and T lymphocytes(anti-CD3). Immunohistochemical analysiswas carried out using both indirect avidin-biotin complex and alkaline phosphatase anti-alkaline phosphatase techniques. Endogenousbiotin activity was abolished by incubating thesections with 0-01% avidin for 10 minutesfollowed by 0 01% biotin for 10 minutes. Eachsection was then treated with 1 /10 normal swineserum (Dako) for 10 minutes to block non-specific sites. The section was then incubatedwith a mixture of the anti-5-lipoxygenase andone of each of the cell marker primary anti-bodies for 90 mintues at room temperature.This was followed by incubation with 1/200biotinylated swine anti-rabbit and 1/100 alka-line phosphatase conjugated goat anti-mousesecondary antibody for 45 minutes. Endo-

66



j.com/


ownloaded from



genous peroxidase activity was abolished byincubation with a mixture of lOmM glucosewith 0O065 mg/ml sodium azide and 1 Fig/mlglucose oxidase for 30 minutes at 370C.15Avidin-biotin complex (Dako) was preparedaccording to the manufacturer's instructionsand incubated on the sections for 30 minutes.The immunoperoxidase colour reaction was

developed by incubation for five minutes withdiaminobenzidine (0 75 mg/ml) and hydrogenperoxide (30,u 0O-1% H202/ml). This was fol-lowed by a 15 minute incubation with alkalinephosphatase buffer containing 100mM amino-methylpropanediol, 1 mM magnesium sul-phate, and 1 mM magnesium sulphate, and1 mM levamisole to block the endogenousalkaline phosphatase activity ofbronchial tissue.The alkaline phosphatase reaction was de-veloped by incubation with 5-bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium(Sigma) for 15 minutes in the dark. The sec-

tions were counterstained with methyl green fortwo minutes and then mounted permanently inDPX (BDH) after dehydrating first in in-cremental concentrations of ethanol followedby acetone and toluene. A positive controlusing nasal polyp tissue and a negative controlwithout primary antibody was included in eachstaining run.

QUANTITATION OF CELL NUMBER

The cell counts were performed by a singleindividual who was blinded to the aspirin sensi-tivity status of the patients. The total numberof positive staining cells in the whole of thesubepithelium was counted. The intra-observercoefficient of variation for cell counts was con-

sistently less than 11% as calculated from re-

peated counting of 10 sections. The hue-saturation-intensity (HSI) system of colourimage analysis'6 was used in the measurementsof the subepithelial area of the biopsy sample.This was performed on an image analyser con-

sisting of a PC computer containing colourimage processing boards (Data Translation,Marlboro, MA, USA) with colour Freelancesoftware (Slight Systems, Hove, UK). The col-our image was acquired from the microscopevia a JVC 3 chip colour TC camera. The HSIthresholds for the tissue were programmed intothe computer. The subepithelial area was thenoutlined manually by the use of a mouse facilityin the software. Automatic detection of thetissue was performed within this manually de-fined zone and the computer calculated a nu-

merical value for the area of tissue measured

(in mm2). From a knowledge ofthe cell numberand area, the number of positive staining cellswas calculated in each biopsy sample and foreach cell type. The area of the biopsy samplesranged from 0 3 to 1 mm2.

DATA ANALYSISResults are presented as mean (SE). Statisticalanalyses were performed with SPSS for Win-dows. A figure for the difference in cell numbersper mm2 and the percentage of total 5-lip-oxygenase positive cells co-immunostainingwith each cell type was calculated before andafter endobronchial lysine aspirin challenge andcompared between ASA and non-ASA groups

using the Mann-Whitney U non-parametrictest for independent samples.

ResultsSUBJECTSThe baseline characteristics of the two groups

of subjects are presented in tables 1 and 2.There were no significant differences in FEV,between the ASA and non-ASA subjects, andno statistical differences in peripheral bloodeosinophil count were seen between the two

groups. Four of the seven ASA subjects andsix of the eight non-ASA subjects were skinprick test positive to the aeroallergens tested.Bronchoscopy, endobronchial lysine aspirinchallenge, and bronchial biopsy were well tol-erated by all individuals and there were no

adverse effects.

CELL NUMBERS

After endobronchial challenge there was a de-crease in mast cell numbers staining for mastcell tryptase in the ASA subjects challengedwith lysine aspirin from 30 4 (I 1 1)/mm2 to14 1 (8 6)/mm2, and this change was sig-nificantly different (p<005) from the changein the non-ASA subjects (fig 1, table 3).The number of total eosinophils (BMK13)

in the ASA group was 1 48-6 (28- 1)/mm2 beforeand 171 (40)/mm2 after challenge with lysineaspirin, which did not represent a significantdifference from the change in the non-ASAgroup. There was an increase in EG2 positiveeosinophils in six of the seven ASA subjectsafter lysine aspirin challenge, whereas only one

non-ASA subject demonstrated an increase inthe numbers of EG2 positive cells after chal-lenge. The change in the numbers of EG2

Table 3 Mean (SE) cell numbers per mm2 for each cell marker before and after local endobronchial challenge with lysine aspirin in aspirin-sensitiveasthmatic (ASA) and non-ASA subjects

Mast cells stainingfor Total eosinophils Activated eosinophils Macrophages Neutrophils T lymphocytestryptase (AAI)* (BMK) (EG2)* (EBMII) (NP57) (anti-CD3)

Before After Before After Before After Before After Before After Before After

ASAsubjects 30 4 (11-1) 14-1 (8 6) 149 (28-1) 171 (40) 28-3 (25 3) 31-7 (9 7) 106 (39-1) 85-4 (30 7) 132 (73) 235 (73 3) 76-9 (38-4) 105 (41-4)Non-ASAsubjects 10-6 (8 7) 4-1 (2) 59-4 (20) 115 (45) 26-4 (15) 6-4 (5 4) 169 (44) 52-8 (13-1) 117 (39-5) 185 (57 9) 91-1 (64 8) 37 (17-6)

* p<005 when the changes before and after lysine aspirin challenge in ASA subjects are compared with non-ASA subjects.

67



j.com/


ownloaded from



AAA1 (mast cells)

BBMK13 (total eosinophils)

p=NSr

CEG2 (activated eosinophils)

180

100

80

60

40

20

0

Before After Before After Before After Before Afterlysine lysine lysine lysine lysine lysine lysine lysineaspirin aspirin aspirin aspirin aspirin aspirin aspirin aspirin

ASA Non-ASA ASA Non-ASA

p<0o05

ABefore Afterlysine lysineaspirin aspirin

ASA

Before Afterlysine lysineaspirin aspirin

Non-ASA

DEBM11 (macrophages)

800

600

400

200

0Before After Before Afterlysine lysine lysine lysineaspirin aspirin aspirin aspirin

ASA Non-ASA

ENP57 (neutrophils)

600

500

400

300

200

100

0Before After Before Afterlysine lysine lysine lysineaspirin aspirin aspirin aspirin

ASA Non-ASA

FAnti-CD3 (T lymphocytes)

p=NS


ASA


Non-ASA

Figure 1 Results obtained from seven aspirin-sensitive asthmatic (ASA) and eight non-ASA subjects both before and after endobronchial challengewith lysine aspirin. Total number of cellslmm2 of bronchial biopsy tissue identified immunohistochemically as (A) mast cells (AAI); (B) total eosinophils(BMK13); (C) degranulated eosinophils (EG2); (D) macrophages (EBM11); (E) neutrophils (NP57); (F) T lymphocytes (anti-CD3). Means are

represented by a horizontal bar.

positive cells in the ASA group was significantlydifferent from the non-ASA subjects (p<005).Compared with the non-ASA group, there

was no significant change in numbers ofmacro-phages, neutrophils, or T lymphocytes in theASA group after challenge with lysine aspirin.

1200 p=NS

1000 _

800

E

E 600

400-

200

0

Before After Before Afterlysine lysine lysine lysineaspirin aspirin aspirin aspirin

ASA Non-ASA

Figure 2 Total number of cellslmm2 staining for 5-lipoxygenase in bronchial biopsy tissue obtained from sevenaspirin-sensitive asthmatic (ASA) and eight non-ASAsubjects both before and after endobronchial challenge withlysine aspirin. Means are represented by a horizontal bar.

CELLS STAINING FOR 5-LIPOXYGENASEBefore endobronchial challenge the total num-bers ofcells immunostaining for 5-lipoxygenasein the bronchial bronchial biopsy samples fromthe seven ASA subjects were similar to thosein biopsy samples from the non-ASA controlsubjects (fig 2). After endobronchial challengethere was no significant change in the numbersof 5-lipoxygenase staining cells in the ASAsubjects compared with the change in the non-

ASA subjects.The percentage of 5-lipoxygenase positive

cells co-immunostaining with the mast cellmarker AA1 decreased from 14-7 (5-1)% to 5(2 3)% in the ASA group after endobronchiallysine aspirin challenge which was significantlydifferent (p<005) from the change in the non-ASA group (3A4 (3K1)% to 2-3 (1-0)%).The percentage of 5-lipoxygenase positive

cells co-immunostaining for eosinophils(BMK13) was not significantly different be-tween the ASA and non-ASA groups. However,the percentage of 5-lipoxygenase positive cellsco-immunostaining for activated eosinophils(EG2) increased following challenge with lysineaspirin in five ASA individuals and in only one

non-ASA subject; this difference was sig-nificantly different (p<005).There were no significant changes in the

percentage of 5-lipoxygenase positive cells co-

cN

E

500

400CM

E 300

= 200(D

100

0

-68



j.com/


ownloaded from



immunostaining with other cell types. T lym-phocytes (anti-CD3) did not co-immunostainwith 5-lipoxygenase.

DiscussionThis study shows that 20 minutes after endo-bronchial challenge with lysine aspirin solutionin ASA individuals there was a significant re-duction in the number of mast cells and anincrease in the number of activated eosinophilsthat was not observed in the non-ASA subjects.There is no previous study reporting endo-bronchial lysine aspirin challenge in aspirinsensitive patients. The timing of the biopsy wasdetermined by the maximal bronchoconstrictorreaction after inhalation of lysine aspirin so-lution which normally occurs at the same timefollowing challenge. We used the concentrationof lysine aspirin for endobronchial challenge ineach ASA individual which had already beenfound to cause a 20% reduction in FEVy wheninhaled (the threshold concentration) by thesame subject. For the non-ASA subjects thehighest concentration of 25 mg/ml was used.We reasoned that the total dose in 20 ,ul appliedlocally via the bronchoscope was probably sub-stantially greater than the dose which wouldhave deposited at the same location followingnebulised inhalation of the same concentrationof lysine aspirin solution. The application ofhigher concentrations of lysine aspirin locallymay have induced more profound changes ininflammatory cell populations.We have used well characterised antibodies

for the definition of different inflammatory cellsby immunohistochemistry. The 5-lipoxygenaseantibody used in the study was raised againstpurified human leucocyte 5-lipoxygenase andits specificity is indicated by immunoblot re-cognition of both rat and human leucocyte 5-lipoxygenase at dilutions of 1:200 and by itsability to immunoprecipitate human leucocyte5-lipoxygenase using standard western analysis.A study of the immunolocalisation of 5-lip-oxygenase to the nuclear envelope has recentlybeen published utilising this rabbit polyclonalantibody.'7 Mast cells were identified using anantibody against tryptase, clone AA1, which isa commercially available mouse monoclonalproduct and recognises mast cell tryptase spe-cifically in the concentrations used in this study.

It was surprising to find that there was adecrease in the number ofcells staining with theanti-tryptase antibody following lysine aspirinprovocation, because previous work in allergenchallenge in vitro has indicated that mast celldegranulation is incomplete and that thereshould still be substantial tryptase left withinthe cell. However, no previously publishedstudy has reported on the mast cell appearancesusing the antibody AA1 after either in vivoallergen or aspirin challenges. Furthermore,mast cell tryptase is a water soluble complex,and because cryosections were used in thisstudy, it is possible that during the immuno-staining procedure tryptase from degranulatedmast cells may have washed away. Thus, theapparent decrease in mast cell numbers found

in this study can be explained by the presenceof degranulated "phantom" mast cells.There are several possible explanations for

the apparent decrease in mast cell numbersother than substantial degranulation. Theseinclude the possibility that the ability to detectmast cells in some of the samples failed fortechnical reasons. This seems unlikely becauseit only occurred in ASA individuals after lysineaspirin challenge and there was no significantchange in the non-ASA subjects. The resultsmay be explained by the migration of mastcells out of the subepithelium after challenge.However, the time between biopsies was tooshort for this to have occurred and therefore thispossibility also seems unlikely. Finally, lysineaspirin challenge in ASA patients may haveproduced an element ofmucosal oedema whichwould result in a decrease in the density of allcells counted in the biopsy samples taken afterthe challenge. However, the numbers of mostof the cells did not change and, indeed, thenumbers ofEG2 positive eosinophils increased.Therefore, the most likely explanation for ourresults is that there was considerable localisedmast cell degranulation.There is substantial evidence from the lit-

erature to support the concept for mast celldegranulation in aspirin-induced reactions insensitive individuals. For instance, immuno-histochemical examination of nasal polypsobtained from ASA patients demonstratedabundant eosinophils and degranulated mastcells.'8 Bosso and colleagues have reported in-creased serum histamine and tryptase levelsafter aspirin ingestion in some ASA subjectswho manifest moderate to severe respiratoryreactions associated with aspirin-induced re-actions in other organs.'9 Similarly, Ferrerifound increased levels of LTC4 and histaminein nasal secretions associated with decreases inFEV, in three out of four ASA patients.5 In5-10% of patients with systemic mast cell dis-ease aspirin therapy causes massive mast cellmediator release with an anaphylactoid re-action, suggesting that aspirin can directly re-lease mast cell mediators in certain subjects.20Treatment with cromolyn sodium in 20 sub-jects with aspirin sensitive asthma has beenreported to lead to a 17% improvement inFEVy 50 minutes after inhalation.2' Further-more, pretreatment with cromolyn sodiumeither prevented or delayed asthmatic re-sponses to aspirin ingestion.22 The de-monstration that a putative mast cell mediator,high molecular weight neutrophil chemotacticfactor, was released into the serum of half of 22subjects following aspirin-induced asthmaticreactions lends further support to the conceptof mast cells as the source of mediators re-sponsible for reactions in patients with aspirin-sensitive asthma.2' In a separate study the re-lease of neutrophil chemotactic factor was ab-olished after desensitisation to aspirin in 10subjects with aspirin-induced urticaria.24 Re-cently, Fischer and colleagues have reporteda significant rise in nasal lavage mast cell-derived tryptase levels following aspirin in-gestion in ASA subjects,25 thereby providingdirect evidence for mast cell degranulation.

69



j.com/


ownloaded from



The other important finding from this studywas the increased numbers of activated eos-inophils in six out of seven subjects followingchallenge with lysine aspirin which was notseen in the non-ASA group. Consistent withthis is the absence of an increase in totaleosinophil numbers. This suggests that eitheraspirin itself or some other factor, perhaps mastcell-derived, activates existing eosinophils inthe airways of patients with ASA. There issubstantial support for eosinophil activation inASA reactions. Sladek and colleagues foundincreased eosinophils and release of extra-cellular protein (ECP) in bronchoalveolar lav-age fluid of subjects with ASA compared withcontrol individuals.6 The same authors alsoreported a significant fall in peripheral bloodeosinophil numbers and a rise in serum ECPand tryptase following oral aspirin ingestion inASA subjects, implicating both mast cells andeosinophils in this reaction.26 In a study ofsevensubjects with ASA, Yoshimi and colleaguesreported an increased peripheral blood eos-

inophilia and increased numbers of activatedeosinophils in nasal polyps of these patients.27The total number of cells expressing 5-lip-

oxygenase was the same in ASA subjects andthe controls, both at baseline and after lysineaspirin challenge. Thus, although the 5-lip-oxygenase pathway is upregulated in ASA,this is not due to increased total numbers of5-lipoxygenase protein-containing cells.

In conclusion we suggest that aspirin chal-lenge in ASA subjects degranulates mast cellsand eosinophils, either directly or indirectly viathe release of a secondary mediator. Either oneor both ofthese cell types may be responsible forthe increased synthesis of cysteinyl leukotrienesfound at rest and following aspirin provocationin aspirin intolerance.

This study was supported by the National Asthma Campaign.

1 Samter M, Beers RF. Intolerance to aspirin: clinical studiesand consideration of its pathogenesis. Ann Intern Med1968;68:975-83.

2 Christie PE, Tagari P, Ford-Hutchinson AW, Charlesson S,Chee P, Arm JP, et al. Urinary leukotriene E4 con-centrations increase after aspirin challenge in aspirin-sensitive asthmatic subjects. Am Rev Respir Dis 199 1;143:1025-9.

3 Christie PE, Tagari P, Ford-Hutchinson AW, Black C,Markendorf A, Lee TH. Urinary leukotriene E4 afterlysine-aspirin inhalation in asthmatic subjects. Am RevRespir Dis 1992;146:1531-4.

4 Ortolani C, Mirone C, Fontana A, Folco GC, MiadonnaA, Montalbetti N, et al. Study of mediators of anaphylaxisin nasal fluids after aspirin and sodium metabisulfite nasalprovocation in intolerant rhinitic patients. Ann Allergy1987;59:106-12.

5 Ferreri NR, Howland WC, Stevenson DD, Spiegelberg HL.Release of leukotrienes, prostaglandins, and histamineinto nasal secretions of aspirin-sensitive asthmatics duringreaction to aspirin. Am Rev Respir Dis 1988;137:847-54.

6 Sladek K, Dworski R, Soja J, Sheller JR, Nizankowska E,Oates JA. Eicosanoids in bronchoalveolar lavage fluidof aspirin-intolerant patients with asthma after aspirinchallenge. Am J Respir Crit Care Med 1994;149:940-6.

7 Israel E, Fischer AR, Rosenberg MA, Lilly CM, Callery JC,Shapiro J, et al. The pivotal role of 5-lipoxygenase productsin the reaction of aspirin-sensitive asthmatics to aspirin.Am Rev Respir Dis 1993;148:1447-51.

8 Nasser SM, Bell GS, Foster S, Spruce KE, MacMillan R,Williams AJ, et al. Effect of the 5-lipoxygenase inhibitorZD2138 on aspirin-induced asthma. Thorax 1994;49:749-56.

9 Christie PE, Smith CM, Lee TH. The potent and selectivesulfidopeptide leukotriene antagonist, SK&F 104353, in-hibits aspirin-induced asthma. Am Rev Respir Dis 1991;144:957-8.

10 Dahlen B, Kumlin M, Margolskee DJ, Larsson C, BlomqvistH, Williams VC, et al. The leukotriene-receptor antagonistMK-0679 blocks airway obstruction induced by inhaledlysine-aspirin in aspirin-sensitive asthmatics. Eur Respir J1993;6: 1018-26.

11 Nasser SM, Pfister R, Christie PE, Sousa AR, Barker J,Schmitz-Schumann M., et al. Inflammatory cell popu-lations in bronchial biopies from aspirin-sensitive asth-matic subjects. AmJ Respir Crit Care Med 1995;151:A37.

12 Schmitz-Schumann M, Schaub E, Virchow C. Inhalationprovocation test with lysine-acetylsalicylic acid in patientswith analgetics-induced asthma. Prax Kin Pneumol 1982;36: 17-21.

13 Anonymous. Investigative use of bronchoscopy, lavage andbronchial biopsies in asthma and other airways disease. 7Invest Allergol Clin Immunol 199 1;1:271-7.

14 Maddox PH, Jenkins D. 3-Aminopropyltriethoxysilane(APES): a new advance in section adhesion. J Clin Pathol1987;40: 1256-7.

15 Andrew SM, Jasani B. An improved method for the in-hibition of endogenous peroxidase non-deleterious tolymphocyte surface markers. Application to immuno-peroxidase studies on eosinophil-rich tissue preparations.HistochemJt 1987;19:426-30.

16 Poston RN, Gall NP. Hue saturation (HSI) colour imageanalyses for the quantification of immunoperoxidasestains. In: Hilger A, ed. Micro 90. New York: Elder, HY,1990:525-8.

17 Woods JW, Evans JF, Ethier D, Scott S, Vickers PJ, Heam L,et al. 5-lipoxygenase and 5-lipoxygenase-activating proteinare localized in the nuclear envelope of activated humanleukocytes. J Exp Med 1993;178:1935-46.

18 Yamashita T, Tsuji H, Maeda N, Tomoda K, KumazawaT. Etiology ofnasal polyps associated with aspirin-sensitiveasthma. Rhinology 1989;8(Suppl): 15-24.

19 Bosso JV, Schwartz LB, Stevenson DD. Tryptase and his-tamine release during aspirin-induced respiratory re-actions. JAllergy Clin Immunol 1991;88:830-7.

20 Yocum MW, Butterfield JH, Gharib H. Increased plasmacalcitonin levels in systemic mast cell disease. Mayo ClinProc 1994;69:987-90.

21 Imokawa S, Sato A, Taniguchi M, Toyoshima M, NakazawaK, Hayakawa H. Sodium cromoglycate nebulized solutionhas an acute bronchodilative effect in patients with aspirin-intolerant asthma (AIA). Arerugi (JapaneseJournal ofAller-gology) 1992;41: 1515-20.

22 Martelli NA, Usandivaras G. Inhibition of aspirin-inducedbronchoconstriction by sodium cromoglycate inhalation.Thorax 1977;32:684-90.

23 Szmidt M, Grzelewska-Rzymowska I, Rozniecki J, Grzeg-orczyk J, Neutrophil chemotactic activity after ad-ministration of aspirin during tolerance in patients withasthma and aspirin hypersensitivity. Pneumonologia i Aler-gologia Polska 1991;59:16-21.

24 Grzelewska-Rzymowska I, Szmidt M, Rozniecki J, Grzeg-orczyk J. Aspirin-induced neutrophil chemotactic activity(NCA) in patients with aspirin-sensitive urticaria afterdesensitization to the drug. J Invest Allergol Clin Immunol1994;4:28-3 1.

25 Fischer AR, Rosenberg MA, Lilly CM, Callery JC, RubinP, Cohn J, et al. Direct evidence for a role of the mast cellin the nasal response to aspirin in aspirin-sensitive asthma.J Allergy Clin Immunol 1994;94:1046-56.

26 Sladek K, Szczeklik A. Cysteinyl leukotrienes over-production and mast cell activation in aspirin-provokedbronchospasm in asthma. Eur RespirJ7 1993;6:391-9.

27 Yoshimi R, Takamura H, Takasaki K, Tsurumoto H, Ku-magami H. Immunohistological study of eosinophilic in-filtration ofnasal polyps in aspirin-induced asthma. NipponJ'ibiinkoka Gakkai Kaiho 1993;96: 1922-5.

70



j.com/


ownloaded from


from aspirin-sensitive asthmatic subjectsthorax 1996;51:64-70 effect ofendobronchial aspirin...

Documents