Treatment of hypersensitivity pneumonitis : Contact avoidance versus corticosteroid treatment

OBJECTIVE: To compare the early response to oral corticosteroid treatment with contact avoidance in patients with acute hypersensitivity pneumonitis.

C0Nc1.us10, : Les coni costcr,i"icles acl mini stres ,,rulcmcnt a fa ibles doses ct la suppress io n de !'ex positio n sont tou .sdeu x ck:s traitcrncnts initi aux adcq uats de la pnc' utll(>pa1hi c cl 'hypcrscnsibili tc.Les res ultats ,IL• !' ana lyse du lavage brond10-alvco lairc rcstaient annrmau x malgrc u11~ amel iorati on clinique : ecllc technique n' a pas rourni d"informati ons ,•li,riqul's utiks pour• k su11i ,k la pncurnopathie ct"hypnsensibilite.H YPERS E SITIVITY P"llTNIO'.\ITISIS :\ DELA Yl::D -TY PE aliL'rgic reaction of the lungs in response to the inhalation of a variety of foreign particles .usually organiL• in nature (I).1\L'lllely.the disease mani!"csts i1sclf hy a febrile reaction aecompanieJ hy dyspnea .cough and chest t ightncss (2.1 ).These .,ymptomsusually occur .1 to 8 h al"tcr contact with the otlenJing allergen and suhsiJc over the next 24 to -1-8 h if contact is avoiJcd .\Vhcn possible.contact ,1voidance is the ohvious lreatmcnl of choice for hypersensitivit y pneumonitis.However.when the contact cannot be avoided the acute manifestations can be controlled by syslemic aJminisiration of corticosteroids./\!!hough steroids control acute manikstations.their use Joes nol moJi fy the lung term outcome of the di,case (4.5L IL is our L•urrcnl practice to suggest conlacl avoidance as the initial treatment of hypersensitivity pncumonilis and.when this is nol feasible.we administer short course., of oral corticosteroids to conlrol thc Jisease.It is not known which of these two approaches gives Lhc hesl initial response.nor their effects on lung cells as assL•ssed hy bronchoalveo lar la vage (BA L).The present study was done to compare lhL• short term effect of these two forms of treatment on the initial clinical.functiunal.radiulogical and BAL cell recovery in arnk hypersensiti vity pncumunilis.The results show Lh,1t relatively small dail y Jose., uf ural corticosteroids are as effective as contact avoidance in controlling acute disease.However.a marked bronchoalvcular lymphocytosis persi sts after one nHllllh or either rorm or lll'atmenl.
range 24 to 67. or the five females.Lwo were in Lhe prednisone group and three were in the conlacl avoidance group.Mean ages were similar !"or each group: prednisone 40± 11 years anJ nonconlaL"l 47± 1.3 years.Twenty -s ix patients were nonsmokers and lwo were ex-smokers (lunger lhan one year).The diagnosis was hased on clinical , radiological.functional.UAL and serological criteria (6).Since some of these subjects could not avoid contact (L1.<;ually farmers) they were not randomized into either arm of tl1c study.All subjects we re encouraged Lo avoid contact as the only form or initial treatment.Subjects who could m>l avoid Lhe offending environment ( J l) suhjL'.Cl1 in this study) were given oral corlicostcroicls (20 mg of prcJnisone daily ror one nmnlh ): thosL' who L•ould ceast: conlacl rcccivL•d no other rmm or treatment (n=lJl. All suhjects wcrL' studied twin•: at diag nosis and at lhl' end of one nHrnth llf cilhL-r contact avoidance or cmlicu steroid Lrcalmrnl.Al each visit all suhjech had a c!iniGil history and physical examination.lu ng function measurL'• ments.postcnianterior and bter;il chest film, and HAL.Both lung function te.-Ls were mL•asurcJ in Lhc same laboratory on the same pulmonary function equipment.Tests obtained inclulk'd: lung volumes hy body plethysmography.forced expiratory tfows and single breath carbon dio:-.:iJcdiffusion capacity (DLCO) .Results arc rcporlcJ as a l)L'rcentage of predicted values (7_8).;\II chL'sl film., were read blindly by one or the authors who is a certified B reader.hims were scored as previously described (9).Rridly.lung t"iL'IJs were subdivid~,J inlo six rL•gions: uppL'r, middle and lower !"or each lung.Ead1 or these regions w,ts given a score oro 10 4 fur the presence anJ intensity ot"infillratcs.With this sy.,Lem.normal lungs ha\'L' ,1 score or() while a score of 2-1signifies marked diffuse infiltrates thniu ghonl both lung fields .The BAL was done with 300 1111 .or sterile normal s.dinc injected in si:-.: 50 111L aliquots, each instillation followed hy gL'ntlc aspirati on.The lavage tluicl wa,; kept on ice until centrifugation on cytospin.Cells were counted by Di!T Quik (Baxter) and dillerenlials obtained on Ciinnsa and este rase stained preparations.
Statistical analysis: Data were analy1.eJusing the statistical package SAS (SAS Inst itute Inc.No rth Carolina).Resul t, WL're expressed as mean ± SD for pulmonary function anJ a, mean ± SEM for the graphical represe nlalion or BAL cells.Original data from the BAL cell analysis were transformed into ranks.Thu.,, for eac h variable.the ob.,L' rvalions were replaced by their ra nks because no va riance stahili1,ing transformation was encountered.Data from pul11101wry functinm were compared using an analysis of covaria nce 1m transformed values where the concomitant variables arc Lile initial values as they were correlated with their respcL"li ve changes.The square root or each prpportiun was transfunm'd tu its arc sine.Thus_ the resultant data have an underlying distribution that is normal.All reported P v:iluL•s arc two-tail ed and were nrnsidcrcd significant ir they were less than (l.05. Results of pulmonary functions at diagnosis (initial) and at one month follow-up (mean ± SD) for subjects treated with oral corticosteroids (prednisone) or by contact avoidance

RESULTS
inetcen subjects rl'ccived l'Orticostero id treatment and l•,mtinued their wor!... on the rarm whi le the rL•maining ni ne 11.::r.::treated by contact avoidance only.Su bjects in both ~roups were or sim ilar age.one ex-smoker was in L'ach tr,•atmcnt group, and or the fiv e females .three were in the con tac t avo idance group and two were in the predni sone group.All subjects imprmed with ei ther form or treatment and all subjects reported an improve ment in their dys pnea.\ lh.'r one month.10 patien ts repo rted that they we re now had, to normal and no longer dyspneic: rive were in the ,1lntaet avoidance gro up and five were in the cort icosteroid treatment gro up .All l"cve r and rnalai ,e reso lved by the secllll d vis it.At diag nos is.inspi ratory crack les were heard in 24 ,uhjects; after one month these phys irnl signs had di sapp.:.1rcJ in 17 ( 10 in the corticosteroid treated group.scvrn in tile rnntact avoidance group).Res ults of pulmona ry functio ns at diagnosis and at fo llow-u r are presented in Tahlc I.Both ~roups were initially similar.Lung !unction improved in must subj ects wi th either treatme nt.DLCO was the lung function para meter that was in itially most abnormal.It significantly improved in each group .Figure I sho ws individual chan~cs in DLCO for all subj ec ts who were treated with rr~dn isonc and seven or the nine who ceased con tact.Fol-h111-up data for this p:1ra mete r were not ava il ab le in two ,ubjccts.Forced expiratory vulume in I s (FEY 1) and rurccd 1ital capac ity (FYC) also significa ntly improved in the predni,lllh' group (P=fl.004and 0.000X.respec tivel y).The irn-pn11 cment in these paramete rs tor the contac t avoi dance  A similarity was seen in both groups for the BAL cells (total.lymphocytes and alveolar macrophages) (Figure 2).However, neutrophils were higher in the contact avoidance group.Repeat BAL showed a mild decrease in BAL ce ll s after one month, simi lar in those who had received prednisone and in those treated by contact avoidance only (Figure 226 2).Agai n, the only BAL cell that differed was the m•utrophil.with a sign ificant decrease in the contact avoidance group but no changes in the cortico teroid treatment group.Individual values for the percentage of ly mphocytes at diagnosis and at follow-up for each group are shown in igure 3.These ce ll s were not signi fi cantly different one month after treatment from the initial values.The percentage BAL fluid return was similar for both study groups at the first (50± 12% for the predn isone group and 50± 13% for the rnntact avoida nce group) and at the follo w-up tudics (57± I I 'li versus 55± 15~.respectivel y); this difference was not statistically significant (P=0.07).
Chest film at diagnosi s and at one month follo w-up were avai lable for review in 22 subjects.Initial radi ographic •core~ were sim ilar in both groups.One month later scores had decreased significantly and remained sim ilar in both group, (Figure 4).The only correlation between variables was between lung diffu sion capacity and chest rad iographic (CXR ) score at di ag nosis (P=0.0014)(Figure 5 ).There were no correlations between an y of the BA L cell s parameters and lung functi on or CXR scores at either diagnosis or at one month foll ow-up, whether tested for the group as a whol e or for the two groups separately.

DISCUSSION
The present study shows that rel at ively small doses of co11icoste roids and contact avoidance are both effective in the initial treatment of acute hypersensitivity pneumoniti s.Although all .ubject improved by either treatment and some had actually re turned to norma l, BAL, after one month of corticosteroids or contact avoidance, was still markedly abnormal, wi th a persi sting lymphocytic alvcolitis.Although Can Respir J Vol 1 No 4 Winter 1994 ,iJc effects were tninitnal.one would uhviou sly recommen d contact avoidance when possible: howe ver. in certain cases.c,pccially farlllcrs.cpntact avoidance was often soc i ucwnomically impractical.In such circumstances.cortico-,tcroid treat ment seems lo he a valid ,iltcrnative.The often 4 uotcJ recom me nded daily dose of corticosteroids in the trcattrn:nt of acute hypersen sit ivity pncunwnitis is higher than the 20 mg per day given in th is study (5): in one revicv.: the Jose is not specified ( 10).We have not compared the relative effects or two doses or corticosteroids 011 Putcomc.Hl111•cvc r. since 20 mg daily gives a similar improvement in 1110,t of the clinical parameters evaluated as contact avoid-,mcc (all subjects who continued contal'l had no recurrence of arntc febril e sy mptoms and noted improvement in their clinical condition).we bl'iil.'ve it is reasonable to treat th is di sease 11ith lower doses rather than higher.T his may be especi all y rdc1•an1 since farmers o!'tcn require cnrticl1sll.'roidsfor up to ,i\ months every winter that their disease !lairs up.Anothn way to decrease further the potential effects of long 1cr111 n1rticostcroid therapy may be to gi ve the dru g 011 alternate Jays ( 11) or e VL'lltually u.,e targetl'd Jd ivery mctlwds ( 12).
We do not know whether the long term outcome would Jitfrr between the twu grnups.Based on previous studies it is ,~ry unli kely that either form of initial treatment would ,ign ificant ly modify lllllL'll!llc:corticosteroi ds do not improve long term outcome ( 4.5 ).i\ long term outcome study is impo,sible for our group uf subjects.Most of those who d10,e to cease con tad have done so on a permane nt basis and their outcome m,1y he di !Tc rent because of that deci sion.not h(causc of the initial treatment.Farmers who stayed on the IJmt are for the mos\ part still in contact and may have h~1d ,,r may eventually ha ve reL'l1rrcnces, ci thL' r acute or subacutc in nature.
Subjects were 1H1t ramlu111i1.cdinto either treatmen t arm l\lr reasons already disc ussed.The grnups.although unequal in num ber and in till'.proportion lit' farme r's lun g. were ,im ibr in terms of all clinical parameters compared.Therefore.we believe that se lection bias was not responsible for the rdativcly small differe nces in the outcome of each group.
An interesting observation in this study is the pe rsistence 111 a hi gh in te n. ity lymphocytic alvculitis after one month , J.-,pitc the ma rked clinic~1I.functional and radiolog ical imprtll'L'lllClll.Since the 1x:rccntages uf lymphocytes were similar in both studies.we do not belie ve that the higher BAL tl ui<l return was responsible t'ur this findin g.This finding .:omiboratesour prcvilHI S studies showing that BAL lymphocytosis persists in farme rs with a history or far mer' s lung.r,pccially those who stay un the farm and in asymptomatic ,~ropo,itivc Canners ( 13).Perhaps variables other than the numhcr of lymphocytes in the BAL can di ffere ntiate the two 11-1-16).
One sign ificant difference between the tWll grou ps both at Jiagnosis and especially in response to till' treatment arlll was th~ level of neutrophil s in the BAL fluid.The initial higher nu mber of these celh rnuld be explained by the fact that ,ill  especially since all of these subjects were clinically illlprnvcd and since these cel ls returned to normal values in the off contact group.Two hypotheses to c:xplain this find ing arc that continued exposurL' contributed towards maintaini ng the neu trop hi lia or that corticosteroids were responsi ble for this finding.Corticosteroids arc known to increase blood lcukocytus is by demarginating these cells./\ similar mechani sm ma y take place in the lungs.

CONCLUSIONS
Although BAL fi ndings remain ab11urmal.both contact avoidanL•c and low dose corticosteroids arc effective in controlling symptoms and improving lung function in acute hyperse nsiti vity pncumoni tis.Contact avoidance remains the oh vitHtS treatment when feasible.B::i.scd on the marked improve tne nt in our patien ts in either arm or this study.it is unli ke ly that higher doses of corticosteroids or the addition or rnrticostcroids in subjects who cease rnntact would significan tly add lo the initial treatment of the disease.