July I953 PEPYS: Basic Mechanisms of Allergic Reactions 35I

ConclusionsIn many normal persons changes in the pulse,

from diminution to complete obliteration, can bebrought about by changes in the position of theshoulder girdle produced by movements of thearm, and in some, throwing back the shoulder onone side will obliterate the pulse on the oppositeside. Probably no one mechanism is responsiblein all these cases. Man's upright position, theshape of the upper thoracic opening as determinedby the costal arrangement and the form anddevelopment of the shoulder girdle and itsassociated muscles are the chief factors involved.The clinical features of the syndrome of the

upper thoracic opening are either neural or vascularin type. The clavicle only rarely plays a part intheir production and costo-clavicular compressionmay be associated with local damage to thesubclavian artery.

Vascular changes with pain in cases of the upperthoracic syndrome call for an exploratory operationto free the artery and the lowest trunk of thebrachial plexus from compression. Anteriorscalenotomy alone will not suffice in most of thesecases; the deep relationships of the artery and thebrachial plexus must be examined and the otherlimb of the vice, if present in the form of a rib,a band or tendinous fibres of the scalenus mediusmust be divided as well.

SummaryAn explanation is given of why the syndrome of

the upper. thoracic opening has no counterpart inthe lower limb.The effect of arm and shoulder girdle move-

ments on the pulse is discussed. Costo-clavicularcompression is shown to be comparatively rare andwhen it occurs mnay lead to vascular changes conse-quent upon damage to the subclavian artery. Theprinciples of operation for the relief of these syn-dromes are stated.

BIBLIOGRAPHYEDEN, K. C. (I939), Brit. J. Surg., 27, III.FALCONER, M. A., and WEDDELL, G. (I943), Lancet, ii, 539.LE VAY, A. D. (X945), Ibid., ii, I64.LEWIS, T., and PI-CKERING, G. (I934), Clinical Sci., 1, 354.MURPHY, J. B. (Itio6), Surg. Gyn. Obst., 3, 514.ROGERS, LAMBERT (I937), Med. Annual., SS. 94.ROGERS, LAMBERT (1938) Id. 56, 92.ROGERS, LAME:kRT (1940) d. 58, 93.ROGERS, LAMBERT (I), Ibid., 59, 67.ROGERS, LAMBERT (194I), Rev. Cirurg. B. Aires, 20, 541.ROGERS, LAMBERT (I944), R.N. Medcal Bull., 8, i8.ROGERS, LAMBERT (I94S), Ibid., 63, 283.ROGERS, LAMBERT, and ALDIS, ARNOLD S. (I947), Brit.

Med. .7., i, 82I.ROGERS, LAMBERT (I948), Med. Annual, 6s, 66.ROGERS, LAMBERT (I949), Brit. Med. _7., ii, 956.ROGERS, LAMBERT (I950) Med. Annual, 68, 6o.ROGERS, LAMBERT (i9Si), Brit. Surg. Practice, 3, Butterworth.STAMMERS F. A. R. (i95o), Lancet, i, 603.TELFORD, E. D., and STOPFORD, J. S. B. (1930), Brit. .7.

Surg., is, 560.TELFORD, E. D., and MOTTERSHEAD, S. (i947), Brit. med.

7., i, 325.TODD, W. (9gi), journ. Anat. and Physiol., 45. 293.TODD, W. (1913), .Lancet, i, I1371.TODD, W. (19I3), .ourn. Anat. and Physiol., 47, 250.

VBASIC MECHANISMS OFALLERGIC REACTIONS

By J. PEPYS, M.R.C.P.(Lond.), M.R.C.P.Ed.From the Dept. of Pathology, Guy's Hospital Medical School and Royal National Throat, Nose and Ear Hospital

IntroductionThe general integration into everyday medical

thought of advances in fundamental knowlege inmedical science is often delayed until it has be-come imperative in order to enhance precision ofdiagnosis and treatment. The role of allergic re-actions has clearly reached this stage and theirbasic mechanisms will be discussed within, thelimitations of our present knowledge in which gapsbridged by new developments remain to be filledin later. The term ' allergy' has lost some of itsprecision and a survey of its origin will help tomake its meaning' clearer. Arising out of thefascinating and important study of immunity to-wards the end of the last'century came v. Behring's

production of diphtheria anti-toxic serum derivedfrom.horses. Its extensive use in humans andguinea pigs soon resulted in numerous, unexpectedand severe reactions which were not related to thedisease and which were termed serum sickness(v. Pirquet and Schick, 1905). At this timePortier and Richet (1902), who were studying theproduction of immunity to actinia toxin in dogs,found that if they were suitably spaced the in-jection of doses of toxin far below the toxic levelproduced severe and often fatal reactions. Thisreaction seemed to be opposed to ' prophylaxis'or 'immunity,' and they termed it ' anaphylaxis'or' removal of protection.' This dual capacity ofthe same agent required explanation and v. Pirquet

copyright. on A

pril 13, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.29.333.351 on 1 July 1953. D

ownloaded from

POSTGRADUATE MEDICAL JOURNAL

(I906) stated ' We are in need of a new, general,non-prejudicial word to designate the altered statewhich an organism achieves after acquaintancewith any organic, living or inanimate poison.' Hesuggested the term ' allergy ' to designate this con-cept of altered capacity to react, and he groupedtogether both decreased reactivity or immunityand increased reactivity or hypersensitivity. Withusage allergy has come to mean hypersensitivity,though v. Pirquet's unifying concept is still main-tained by some workers (Sulzberger, 1940).v. Pirquet recognized the fundamental similarity,in both animals and man, of anaphylaxis andhypersensitivity produced by infection in spite oftheir different clinical manifestations. From thattime on experimental studies in animals havecontributed greatly to our understanding of allergicprocesses in man.

Definition of Allergyv. Pirquet's definition outlines the basic aspects

of allergic processes, and its essential truth after50 years of intensive study is a tribute to hisinterpretation. He defined allergy as the ' ac-quired specific altered capacity to react.' Thisdefinition will be amplified by dissection. 'Ac-quired' implies previous adequate exposure to thesensitizing agent, the antigen or allergen. This isfollowed by an ' incubation period' averaging 8to 14 days before the hypersensitive state isestablished. ' Specific' refers to the chemicalcomposition of the antigen used to elicit reactionsin sensitized subjects. ' Altered capacity to re-act' describes the reactions elicited, which maydiffer entirely from those produced by the initialexposure to the antigen, or may be out of allproportion to the dose required to elicit them,being enhanced in speed and, severity. Allergicreactions fall into two main groups:

i. Immediate type hypersensitivity reactions,which include anaphylaxis, the Arthus reactionand allergic disorders in man such as hay fever,perennial rhinitis, asthma, urticaria and gastro-intestinal allergy. The reactions elicited by theantigen come on within a few minutes and can re-solve completely in one to two hours. The Arthusreaction differs in various respects and will bedescribed later.

2. Delayed type hypersensitivity-reactions, whichinclude the bacterial allergic, fungal, chemical andcontact type reactions. These come on after aperiod of hours, are maximal at 24 to 48 hours andresolve themselves more slowly, taking perhapsseveral weeks before complete return to normal.The participation of antibody in these reactions

was conclusively shown by the passive transfer ofanaphylactic hypersensitivity to normal animals byinjection of the serum of sensitized animals (Otto,

1907; Richet, 1907). Similar proof was providedin human allergic disorders by Prausnitz andKuestner (1921) who found that the injection ofthe serum of the latter, who was sensitive to fish,into the skin of his normal colleague transferredthe hypersensitivity locally so that whealing re-actions were obtained on testing the site somehours later with fish extract. More recently thetransfer of delayed type hypersensitivity passivelyby means of lymphoid cells from sensitized sub-jects has been reported in both animals and man(Chase, I945; Lawrence, I949, I952). The com-bination of antigen and antibody initiates the allergicreaction and these terms will be discussed in moredetail.

Definition of Antigen ('gens' or former ofantibody)Antigens are substances which have two actions:i. They stimulate the production of antibodies

after introduction into the body.2. They react specifically with these antibodies

either in vivo or in vitro or both.Antigens are also called allergens in relation to

clinical allergy, and allergens in turn have beencalled atopens and the corresponding antibodyatopic or reaginic antibody, but these terms are nolonger significant since the distinctions on whichthey were based have disappeared. The termsantigen and allergen will be used here inter-changeably. The importance of protein in anti-gens was soon established and the early workersfound that most proteins could act as antigens andthat many antigens are protein or combine withprotein to exert an antigenic action. Antigens canbe divided into the following types:

i. Complete antigens or full antigens are sub-stances which per se can produce hypersensitivestates.

2. Incomplete or partial antigens combine withprotein to become fully antigenic. Our knowledgehas been greatly extended in recent years due tothe work of Landsteiner (I945), who madechemical substances antigenic by simple combina-tion with protein. He found that many of thesechemical substances, or ' haptens' as he termedthem, possess labile nitro or chloro groups whichcombine readily with protein. By using chemicalpartial antigens of known composition, Land-steiner added precision to the immunologicalmechanisms concerned and contributed a conceptof considerable importance for the understandingof chemical and drug allergy and of the patho-genesis of certain diseases. Polysaccharides arevery effective partial antigens, whilst lipids are farless so and their specificity is not high. The re-port of Obermayer and Pick (I906) that homo-logous protein can be made antigenic by chemical

YilY I195 3352

copyright. on A

pril 13, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.29.333.351 on 1 July 1953. D

ownloaded from

Jl15PEPYS: Basic Mechanisms of Allergic Reactions

alteration before introduction into the body, hasbeen extended by reports that partial antigens canbecome fully antigenic by combining with thebody's own protein after introduction into thebody, and these have been termed ' pro-antigens'(Gell, Harington and Rivers, 1946), or ' deriva-tive antigens' because they are derived from bothchemical and the body's own protein (Chase,1952). This is the method by which hyper-sensitivity to chemicals and drugs arises in theordinary way, and the significance of this conceptwill be discussed further in relation to partialantigens derived from infective agents and fromthe body's own tissue constituents. Once thehypersensitive state has been produced the partialantigens can elicit reactions though there is somedoubt whether they need to combine with proteinto do so. There is evidence that combination withprotein is necessary for the elicitation of delayedskin test reactions, and Eisen, Orris and Belman(1952) found that out of eight 2-4-dinitrophenolcompounds tested only those four which combinedwith protein elicited reactions.

Specificity of AntigensThe specificity of the antigens required to elicit

reactions depends on their chemical structure andantigens from different sources possessing similarsuperficial reacting groups react alike. Specificitymay vary as follows:

I. Species or group specificity. Species specificityis of a high order and reactions may be elicited byonly one compound or by extracts from oneparticular source. Group specificity is presentwhen reactions are elicited by related compoundsor extracts from related sources. Hypersensitivityto the sulphonamides provides examples wherethere is hypersensitivity to only one sulphonamide,or to the group of sulphonamides perhaps evenincluding procaine derivatives which are chemic-ally related.

2. Organ specificity has two aspects: (a) Insubjects hypersensitive for example to injection ofliver extract or insulin the reactions may beelicited by extracts from only a single species,such as the pig (species specificity) or by extractsof the particular organ from many animal species(organ specificity). (b) There is evidence tosuggest that autogenous organ products may pro-duce a hypersensitive state in which reactionsoccur in the specific organ. Hypersensitivity tolens protein is the classical example of this form.The above discussion of derivative antigens hasalready indicated that chemical alteration of thebody's own proteins can render them antigenic.

Definition of AntibodyAntibodies are globulin substances produced in

living tissue in response to antigenic stimulationand reacting specifically with the antigen in vivoor in vitro or both. It is the antibodies concernedwith allergic reactions which will be discussed here,and these can be divided as follows:

i. Complete antibodies. These are also termedmultivalent antibodies since the production ofprecipitates on combination with antigen is thoughtto be due to the formation of large aggregates forwhich multiple reacting groups are needed(Marrack, I951).

2. Incomplete antibodies (Marrack, i95i). Afeature of these is the absence of precipitation onaddition of the antigen, which has led to the term'non-precipitating antibody' being applied tothem. Amongst the explanations for this failureto produce precipitates is the absence of adequatereacting groups for the production of the pre-cipitate aggregates, and for this reason they havealso been termed ' univalent antibodies.' If anantigen is added in one stage to antibody containingserum, larger amounts of antibody are precipitatedthan are produced by the addition of the antigenin fractions and in stages, and the presence ofresidual antibody in the latter case is shown by theproduction of passive anaphylactic sensitization oninjection of the supernatant serum, after removalof the precipitate, into a normal subject (Kabat,1952). It is suggested that the addition of theantigen in one stage leads to the formation of largeaggregates which carry down the ' non-precipita-ting' or incomplete antibody. Thus the sametype of antibody, namely incomplete antibody, canfunction as both passive transfer skin-sensitizingantibody in man and as passive transfer ana-phylactic antibody in animals. In the past thepresence of precipitins in hypersensitive animalsand their absence in allergic man was thought torepresent a difference between hypersensitivity inman and animals, but this recent work emphasizestheir basic similarity.

3. Blocking antibody. One of the problemswhich has not been explained as yet is the pro-duction of skin sensitizing antibody in response tocontact with an allergen, such as pollen, by theordinary route, i.e. by inhalation or contact, andthe production of a ' blocking antibody' whenpollen extracts are given by injection (Cooke,Barnard, Hebald and Stull, I935; Loveless, 1940,1943; Harley, 1937). The blocking antibody (tobe distinguished from the term ' blocking anti-body ' as used in haematological work) preventsthe antigen from eliciting reactions in sensitivesubjects, but it is only temporarily present in theserum of treated subjects or in passive transfertests, in contrast to skin sensitizing antibody whichmay persist for life in actively sensitized subjects.The complexity of the antigens used may be. im-

YUIY I 95 3 353-copyright.

on April 13, 2020 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.29.333.351 on 1 July 1953. Dow

nloaded from

POSTGRADUATE MEDICAL JOURNAL

portant because the blocking antibody may be aresponse to antigenic fractions which are not con-cerned with hypersensitive reactions. The routeof administration may also be responsible for thedifferent types of antibody.

Recent evidence indicates that antibodies areformed in the reticulo-endothelial system, par-ticularly the lymphoid-plasma cell series, and acomprehensive survey is given by Barber andDelaunay (1952). The role of circulating lympho--cytes is not yet clear, though it has been shownthat they produce passive transfer of delayed typehypersensitivity (Chase, 1945), and there is evi-dence that they can also produce passive transferof immediate type hypersensitivity (Chase, 1951).Immediate Type Hypersensitivity

i. Production of the hypersensitive state.2. Antigen-antibody reactions.3. Clinical manifestations: (a) Anaphylactic

shock; (b) local allergic inflammation and (c)general allergic inflammation.

i. Production of the Hypersensitive StateAnaphylactic hypersensitivity can be produced

actively and passively. Active sensitization is pro-duced by administration of the antigen by in-gestion, inhalation or injection. The initiallyharmless antigen can be effective in minuteamounts and, as a rule, provokes no local reaction.Larger protein molecules are more effective thansmaller, and both can be enhanced by adsorptionon to inert substances such as charcoal and kaolinor by injection together with adjuvants such asliquid paraffin or after alum precipitation. Coul-son and Stevens (i949) found that the sub-cutaneous injection of 0.04 tug. of alum precipitatedalbumin into guinea pigs was as effective as twicethat dose given intravenously and 6o times thedose given intraperitoneally in producing hyper-sensitivity. The incubation period is 8 to I4 dayson the average, and once present hypersensitivitycan persist for life. The length of the incubationperiod depends on the nature of the antigen, andDoerr and Berger (1922) report that horse serumalbumin requires a longer period than the euglo-bulin, and that very small doses also lengthen theincubation period. Repeated subcutaneous in-jections of protein antigens at intervals of severaldays into rabbits and guinea pigs leads, after threeto five injections, to a severe, necrotic and haemor-rhagic reaction named after Arthus, who describedit in I903. The reaction in these animals comeson after three to four hours and is not readilyreversible.

Passive transfer of hypersensitivity locally orgenerally is produced by injection of the serum ofsensitized animals into normal recipients and re-

actions are elicited by subsequent administrationof the antigen. This process can be reversed, the-antigen being injected first followed by the in-jection of the antibody containing serum (Wrightand Hopkins, 194I). The sensitivity can betransferred to other species though the type ofreaction elicited may vary, thus, horse anti-pneumococcus serum transfers immediate whealor Arthus type reactions to the guinea pig, theskin reactions being elicited by the specific pneu-mococcal carbohydrate (Mehlman and Seegal,I934; Kabat, 1952). It also transfers the Arthustype sensitivity to the rabbit and the anaphy-lactic type to the dog (Francis and Tillett, 193I).The anaphylactic antibody has been studied byKabat and his colleagues, and they report thatnon-precipitating antibody is as efficient as pre-cipitating antibody in transferring anaphylacticsensitivity to the guinea pig, though non-pre-cipitating antibody does not transfer Arthus typehypersensitivity (Kabat, I952). There is a cor-relation between the precipitin titre and theArthus reaction in both actively and passivelysensitized animals (Opie, 1924), and it is thoughtthat the precipitate damages the vessels. Theamount of antibody N required to transfer localreactions is as little as 0.025 mg. in the rabbit ando.og mg. in the guinea pig (Fischel and Kabat,1947; Benacerraf and Kabat, 1949). The in-jection of 0.03 mg. of antibody N into a 250 g.guinea pig produces anaphylactic sensitivity whichcan be shown on testing 48 hours later whenneither this dose nor five times the dose of anti-body can be detected in the serum by in vitrotests, indicating the sensitivity of the biologicalreaction (Kabat, Coffin and Smith, 1947). Kabathas calculated that in passively sensitized guineapigs the uterus may contain Io-5 mg. of antibodyN and will react by contraction on contact withthe antigen in vitro.

2. Antigen-Antibody ReactionsIn vivo antigen-antibody reactions initiate com-

plex biological changes which determine theclinical picture, and they can be a very sensitiveindicator of the presence of antibody and of thechemical specificity of the antigen. The earlycontroversy regarding cellular or humoral mechan-isms in allergic reactions was settled for a time infavour of the cellular mechanism by Dale'sdemonstration that isolated washed smooth muscletissue from sensitized animals reacted in vitro onaddition of the antigen, and that the tissue maythen be desensitized and not react to furtheraddition of the same antigen. Sollmann andGilbert (I937) and Gilbert (1938) report thatvisible contractions of the pulmonary vessels andthe bronchi of sensitized rabbits can be seen in

,YulY I195 3354

copyright. on A

pril 13, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.29.333.351 on 1 July 1953. D

ownloaded from

PEPYS: Basic Mechanisms of Allergic Reactions

thin sections of lung after the addition of theantigen. Similar reactions of the bronchialmuscle of human asthmatics have been reportedby Rosa and McDowell (I95i) and Schild,Hawkins, Mongar and Herxheimer (1954).The possibility of a humoral mechanism in

anaphylaxis was suggested by the report of Deanand Webb (I924) and Webb (1924) that there is amarked leucopenia in anaphylactic dogs and thatthe capillaries of the lungs are crowded withenormous numbers of leucocytes which are firmlyadherent to the vessels. This aspect has recentlybeen re-investigated by e Silva (1950) who hasshown that the injection of antigen into sensitizedanimals is followed by the rapid formation anddisintegration of microthrombi of platelets andleucocytes in the capillaries. This is followed bythe liberation of histamine and plasma proteaseactivators which, he suggests, pass into the tissuesthere leading to the further liberation of histamineheparin, adenosine and other metabolites. Ex-clusion of the liver from the circulation preventsthe liberation of heparin. The heparin is probablyresponsible for the decreased coagulability of theblood in anphylactic shock. The localization ofthe thrombi may be aided by the contraction of thevascular smooth muscle which is a shock tissue inthis type of reaction. Similar microthrombi havebeen reported in Arthus reactions (Stetson, 195I).In the perfusion of sensitized organs e Silva foundthat the use of whole blood as a vehicle for theantigen led to much greater histamine liberation.In addition to the enzyme release of histaminesuggested by e Silva, the action of other histamineliberators, such as simple diamines described byMcIntosh and Paton (949), must also be con-sidered. The relationship between these reactionsand the thrombocytopenia and leucopenia of acuteallergic reactions remains to be clarified.The most important metabolite so far identified

is histamine, which can reproduce many but not allthe anaphylactic and allergic manifestations. Dale(I948) has suggested that histamine may be ex-trinsic or intrinsic, the extrinsic histamine actingon tissues away from the point of liberation and theintrinsic being liberated in close proximity to thereacting cells, and he suggests that this may explainthe failure of anti-histaminic drugs to prevent theallergic reactions in conditions such as asthma.The presence of histamine in the circulating bloodcells has been discussed by Code (1952), who re-ports that in the rabbit the platelets are rich inhistamine whereas in humans it is contained in thegranulocytes, and he points out that the eosino-phile leucocyte which has been suspected ofcarrying histamine does not always contain it inspite of raised histamine content of theblood.

3. Clinical ManifestationsThe tissue reactions are based on smooth

muscle contraction and increased capillary per-meability both of which are produced by his-tamine. Abell -and Schenck (I938) observeddirectly the vascular reactions produced byaddition of antigen to a transparent chamber in therabbit's ear, and they found that there wasarteriolar spasm, circulatory slowing and formationof white cell thrombi and later diapedesis of whitecells into the tissues. Similar vasospasm has beenreported in various species including the mouse,in which anaphylactic shock produces vaso-constriction of the ear vessels (McMaster andKruse, 195I). The dependence of the Arthus re-action on blood vessels is reported by Rich andFollis (I940) who found that it could not be pro-duced in the avascular cornea, but appeared in thecornea after it had become vascularized' as a resultof damage.

(a) Anaphylactic shock. With time the termanaphylactic shock has been used chiefly to des-cribe the reactions in experimental animals, butsimilar shock occurs in humans' as well. Thereare also increasing reports of allergic conditions inanimals like those occurring in humans, such ashay fever, asthma and eczema (Redding, I949;Wittich, 1949), which have the same features andindicate the common fundamental biologicalnature of the allergic reaction. The clinicalpicture varies in different species and is often de-pendent on the distribution of smooth muscle inthe body and on the vulnerability of the shockorgan and the sensitivity to histamine. Ana-phylactic shock in man presents many of the re-actions found in the commoner experimentalanimals and consists of local and generalizedurticaria and angioneurotic oedema, bronchialspasm and mucosal oedema, gastro-intestinalsymptoms, shock, collapse and perhaps death.Anaphylactic shock has been reported followingthe ingestion of very small amounts of milk inmilk sensitive infants (Campbell, I945). The in-jection of too strong doses of pollen into hay feversubjects can readily be followed by severe shock,and anaphylactic death has been recorded afterskin test doses of various allergens (tTrbach andGottlieb, 1946). Another common example is beeor wasp sting sensitivity in which not only is therea high degree of hypersensitivity to the insect pro-tein but the sting contains hyaluronidase whichmust certainly enhance the rate of absorption ofthe antigen and the severity of the shock.The vulnerability of the shock organ may over-

shadow the clinical picture and in the guinea pigdeath rapidly ensues, due to respiratory failureproduced by bronchospasm and mucosal oedemaleading to acute emphysema and great distension

YUyIY 1953 355

copyright. on A

pril 13, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.29.333.351 on 1 July 1953. D

ownloaded from

POSTGRADUATE MEDICAL JOURNAL

of the lungs. It was the resemblance of this syn-drome to asthma in humans which suggested toearly observers that this disease might have anallergic basis, though very pertinent observationsregarding an' allergic' mechanism in asthma hadbeen made on clinical grounds 50 years previously(Hyde-Salter, i868). In the rabbit death resultsfrom spasm of the pulmonary arterioles leading toacute right heart failure, and in the dog gastro-intestinal symptoms predominate and death fol-lows after some hours. The hepatic veins of thedog contain a high proportion of smooth muscle,and their constriction makes the liver a shockorgan. The exclusion of the liver from the circula-tion prevents the anaphylactic shock.

(b) Local allergic inflammation. Local allergicreactions, like the wheal or oedema of the nasalmucosa, resolve rapidly except for the eosinophileinfiltration which can persist for several days.More severe irreversible reactions are found in theArthus reaction, in which there are necrosis andrupture of the minute blood vessels, thrombi ofleucocytes plugging the capillaries and cellular in-filtration into the tissues. This reaction has beencorrelated with the precipitin titre (Opie, 1924).The blood vessels and circulating white cells par-ticipate in the reaction which can be elicited in allparts of the body except for the avascular cornea(Rich and Follis, 1940). Beatrice Seegal has usedthe reverse Arthus reaction to demonstrate howshock organs can be determined. In this reactiona second intravenous injection of antigen producesa local reaction at the site of the first injection ofantigen. Seegal found that if the antigen was in-jected into the anterior chamber of the eye thesubsequent intravenous injection elicited a re-action in that eye. In addition, the production ofa non-specific inflammation in one eye by theinjection of glycerine leads to localization of theantigen in that eye if the first injection of antigenis given intravenously. Subsequent intravenousinjection of the antigen leads to a reaction in thateye, thus indicating the importance of inflam-matory processes in the production of local sen-sitization and the production of a potential shockorgan should the subject be exposed to the specificantigen again (Seegal and Seegal, 193I, 1933).The injection of antigen into the pericardium ofthe rabbit also produced local sensitization morefrequently than injection elsewhere in the body,and perfusion of the coronary vessels with theantigen decreased the flow by I5 to 45 per cent.(Wilcox and Andrus, I938; Seegal and Wilcox,1940).

(c) General allergic inflammation. Serum sick-ness provides the classical example of generalallergic inflammation and was described with theatmost clarity by v. Pirquet and Schick (1905).

Serum sickness presents the typical incubationperiod averaging 8 to 14 days and produces thesame clinical manifestations found in other types ofhypersensitivity such as drug allergy. Pyrexia,urticaria and anioneurotic oedema, splenic andlymph gland enlargement, arthralgia and arthritis,gastro-intestinal and pulmonary disturbances ap-pear as well as lesions in any part of the body. Theinjected serum proteins can be found in the circula-tion for weeks and the flare-up of the conditionwhich occurs at intervals has been correlated withthe persistence of different serum proteins whichvary in the time they take to elicit reactions. Theforeign serum proteins disappear rapidly at thetime of appearance of antibodies, at which time theserum sickness develops (MacKenzie and Leake,I92I; Seegal, I952). Where the foreign serumproteins persist without antibody production theserum sickness does not appear. The re-injectionof foreign protein into sensitized subjects is fol-lowed by its rapid breakdown and disappearanceas compared with its injection into normal animals(Wright and Laws, 1952), suggesting that thisrapid breakdown and disappearance may be afunction of the hypersensitive state.The production of vascular reactions and of

diffuse lesions resembling periarteritis nodosa fol-lowing the injection of large doses of foreign seruminto rabbits was described by Klinge (1929), andthe allergic concept in relation to these vascularlesions has recently been re-emphasized by Richand Gregory (I943). The collagen lesions foundin serum sickness in association with the vasculardamage has led to suggestions that there may beallergic mechanisms operative in disorders, such asacute lupus erythematosis, scleroderma, malignantnephrosclerosis, rheumatic fever as well as peri-arteritis nodosa. There are, however, reasons forbeing cautious in making a diagnosis of an allergicetiology on the basis of histological evidence alone(Klemperer, I947), though allergy provides atempting explanation for these obscure diseases.The nature of the antigen influences the type of

vascular lesion, and More and McLean (I949) re-port that horse serum results in a diffuse arteritis,whilst bovine gamma globulin produces glomerulo-nephritis and granulomatous endocarditis in ad-dition. Hawn and Janeway (1947) report thatcrystalline bovine albumin appeared to be a poorantigen in rabbits, disappearing slowly from thecirculation and producing disseminated lesionsresembling periarteritis nodossa after two to threeweeks, whereas bovine gamma globulin dis-appeared more rapidly and the lesions found werenot disseminated and were predominantly in themyocardium and glomeruli. The production ofvascular lesions seems to be favoured by the use oflarge doses of antigen which persist for long

356 3fulY I195 3copyright.

on April 13, 2020 by guest. P

rotected byhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.29.333.351 on 1 July 1953. Dow

nloaded from

PEPYS: Basic Mechanisms of Allergic Reactions

periods, perhaps thereby producing an enhanceddegree of hypersensitivity. This resembles in somerespects the forced antigenic stimulation of theArthus reaction with its characteristic localvascular damage.

Delayed Type Hypersensitivity(a) Production of delayed type hypersensitivity.(b) Antigen-antibody reaction.v. Pirquet recognized the ' allergic' nature of

the enhanced speed of reaction of vaccination ofpreviously vaccinated subjects described by Jennerand Bryce (Hektoen, 1932), and of the enhancedspeed and severity of re-infection of tuberculousanimals. Fungi, many other infective agents andchemical sensitizing agents elicit reactions inhypersensitive subjects which come on afterseveral hours and are maximal after 24 to 48 hours.In those diseases in which hypersensitivity plays apart these reactions are first elicited at about thesame time that the first manifestations of thedisease appear (Sulzberger, I940). The problemof the relationship between hypersensitivity andimmunity arises at this point, and it would appearthat since the hypersensitivity is a sine qua non ofthe disease it is harmful. But the altered enhancedreaction which hypersensitivity brings about maybe locally harmful and generally beneficial. Infungal infections, for example, the more severe theinflammatory response the higher the degree ofhypersensitivity, the more rapid the cure and themore difficult the demonstration of the pathogen,and it is claimed that the hypersensitive inflam-matory reaction leads to more rapid destructionand throwing off of the fungi. More indolentfungal infections are more chronic and moredifficult to cure (Sulzberger, I940). The degreeof hypersensitivity and the site of reaction are alsoimportant, and a severe tuberculin skin reactioncannot be compared in its end result with a similarreaction in a vulnerable bronchus. The termsimmunity and hypersensitivity have been usedchiefly to describe the end results of reactions inwhich they participate, but this should not obscurethe fact that they have a common basic mechanism,namely the antigen-antibody combination. Thecomplicating factors such as those briefly des-cribed may result in widely differing end picturesunder varying circumstances.The antigenic complexity of bacteria, which con-

tain protein carbohydrate and lipid, influences thetype of hypersensitivity produced and the re-actions elicited. Bacterial proteins like other pro-teins produce and elicit immediate type hyper-sensitivity and reactions when injected by themselves. They are, however, partially antigenic inrelation to delayed type hypersensitivity since theycannot produce it by themselves, though they

elicit delayed reactions on testing subjects in whomdelayed type hypersensitivity has been producedby infection. This partial antigenic of tuberculin,for example, is important since Rich claims thatdesensitization with tuberculin can remove hyper-sensitivity and leave the immunity. But tuberculinhas been found in avirulent as well as virulentstrains (Seibert and Morley, 1933), and it wouldnot appear to be the factor responsible for in-vasiveness. This, together with its inability toproduce delayed type hypersensitivity, suggeststhat another antigenic fraction may be responsiblefor the immune state and that desensitization withtuberculin is ineffective against it. The findings ofBailly (1950) support this possibility, for she foundthat in animals immunized against haemolyticstreptococci, the intravenous injection of the carbo-hydrate fraction produced desensitization againstit but left the immune state intact, whereas de-sensitization with the protein antigenic fractiondeprived the animals of their immunity. If onlythe carbohydrate antigenic fraction had beenstudied separation of immunity from hyper-sensitivity would appear to have been produced.The polysaccharides of organisms such as

streptococci, pneumococci and meningococci arealso partial antigens in relation to delayed typehypersensitivity since they elicit immediate whealtype reactions in sensitized subjects as a rule, butthey cannot produce the delayed type hyper-sensitivity. Lipids are weakly antigenic, but thereport by Raffel (1946, 1948) that purified tuberclewax injected together with tuberculoprotein pro-duces the delayed type hypersehsitivity may givesome indication of their immunological role.

(a) Production of delayed type hypersensitivity.This can be produced both actively and passively,but unlike immediate type hypersensitivity inwhich the injection of the protein antigen need notprovoke a local reaction, for the production of de-layed type hypersensitivity, either an active in-fection even with attenuated strains, or the use ofadjuvants with bacterial antigens is required toproduce hypersensitivity consistently. Adjuvantssuch as liquid paraffin injected together with killedtubercle bacilli will produce a high degree of de-layed type hypersensitivity, and Raffel (1946, 1948)has recently reported that tubercle wax plustuberculoprotein will also do so. The importanceof the local reaction excited by the adjuvant issuggested by the report of Dienes and Schoenheit(I929) that injection of protein antigen such as eggwhite into tuberculous foci leads to the productionof delayed type hypersensitivity with delayed typereactions on testing subsequently with egg white.Raffel, Arnatid, Dukes and Huang (i949) reportthat the injection of egg white with tubercle waxalso produces delayed type hypersensitivity to-

EYUIY 195 3 357

copyright. on A

pril 13, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.29.333.351 on 1 July 1953. D

ownloaded from

POSTGRADUATE MEDICAL JOURNAL

wards the egg white. It seems that the activeinfection or the adjuvant may assist production ofthe hypersensitive state by prolonging the anti-genic stimulation, or the mononuclear cellular in-filtration produced by the adjuvants may be par-ticipating as suggested by Burnet by taking up theantigen, altering it and setting in motion the anti-body forming mechanism. The significance oflocal retention of the bacterial antigen is indicatedby the report of Derick and Swift (I929), whofound that the intravenous injection of strepto-cocci failed to produce hypersensitivity, whereasinjection into the knee, muscle or peritoneal orpleural cavities did produce hypersensitivity. De-layed type reactions are also obtained during theearly stages of sensitization to foreign sera to besucceeded later by immediate type reactions(Dienes and Simon, 1935; Mote and Jones,1936). These findings have led to the view thatdelayed type hypersensitivity is a stage in thedevelopment of immediate type sensitivity andthis fundamental problem needs further elucida-tion.

Contact type hypersensitivity lends itself tocontrolled study of delayed type hypersensitivity.The sensitizing capacity of the chemical dependson the concentration, and dilute solutions ofprimulin, for example, will sensitize 42 per cent.of subjects whereas ioo per cent. are sensitized byconcentrated solutions (Bloch, 1926). The routeof administration is also important and Chase(1946) found that prior feeding of the chemicalinhibited the later production of cutaneous sen-sitization with it. This refractory state persistsand is specific since it does not prevent sensitiza-tion to other chemicals. The refractory animalscould be passively sensitized by injection of cellsfrom sensitized animals so that they reacted for atime to the chemical under consideration. Chase(I949) interprets this as an indication that the re-fractory state is not due to a blocking antibodywhich would perhaps have prevented the reactions.The mechanism of this inhibition of sensitizationis not known, and it is noteworthy that feeding ofthe chemical after sensitization was found to haveno effect on the hypersensitivity. Landsteiner andChase (1940) and Chase (I94I) have also reportedthat the production of contact sensitization is in-fluenced by heredity since they bred poorly andeasily sensitized strains of guinea pigs.The association of immediate and delayed type

hypersensitivity both of which are produced inbacterial allergy and in hypersensitivity to chemicalpartial antigens, is important in relation to hyper-sensitivity to chemical and antibiotic therapeuticagents. Where these are acting as contact sen-sitizers diagnostic patch tests are effective in pro-ducing delayed type reactions, and with those

agents producing immediate type hypersensitivitysuch as the injection of liver extracts and insulinimmediate type reactions are obtained on testing.These are only a small part of the drug hyper-sensitive group, and improved methods of testingfor the remainder are urgently needed. In hyper-sensitivity to chemical partial antigens the chemicalby itself can elicit delayed type contact reactions onapplication to the skin and may also produce ana-phylactic reactions on injection. Where thechemical by itself produces anaphylactic reactionsit is thought that it combines with the body's ownprotein, but Chase (1952) points out that he ob-tained anaphylactic reactions in certain cases onlywhen the chemical had been conjugated with pro-tein before injection into the body, and he suggeststhat the use of proper conjugates may be of value.Rajka and Hegyi (I950) report that they obtainedimmediate type whealing reactions on injecting theserum of subjects sensitive to sulphonamides,arsenicals and barbiturate into the skin of normalsubjects to whom the drug under investigationwas being given. Both these types of testing aredirected towards the immediate type hyper-sensitivity. It seems that investigations along thelines indicated by the experimental study ofchemical partial antigens will take into account thedifferent-types of hypersensitivity and the natureof the antigens used to produce them, togetherwith the appropriate method of testing.

(b) Antigen-antibody reaction. The location ofthe reacting antibody in immediate type hyper-sensitivity can be determined by the reactionswhich occur in vessels and smooth muscle both invivo and in vitro. With delayed type reactions thesite of the antigen-antibody reaction has onlybecome clearer in recent years. Because of theability of tuberculin to elicit reactions in all partsof the body including the avascular cornea Rich(1940) suggested that all the cells of the body !iresensitized and he supported this with his observa-tions of the toxicity of tuberculin for tissue culturesfrom hypersensitive animals. This cytotoxicaction of tuberculin has since been reported by anumber of workers (Rich and Lewis, 1932; Moenand Swift, I936; Aronson, 193I; Heilman,Feldman and Mann, I944; Fabrizio, 1952),though some of the reports are not entirely con-vincing and Baldridge and Kligman (i951) failedto repeat Rich's findings and cited the technicaldifficulties in assessment of this method. Moen(I936) reported that streptococcal protein istoxic for tissue cultures from streptococcal sen-sitive animals. Raffel et al. (I949) have reportedthat the delayed type sensitivity to egg whiteproduced by injection of egg white with tuberclewax, is associated with a similar cytotoxic actionof egg white for bone marrow tissue cultures

358 Yul-V I 95 3

copyright. on A

pril 13, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.29.333.351 on 1 July 1953. D

ownloaded from

PEPYS: Basic Mechanisms of Allergic Reactions

from sensitized animals. In immediate type hyper-sensitivity the antigens are not toxic for tissuecultures from hypersensitive animals and Raffel'sfindings support the claim that delayed type hyper-sensitivity is produced by protein antigens underthe conditions described (Meyer and Loewenthal,1927).The cytotoxic action of tuberculin has so far

been demonstrated with tissue cultures of reticulo-endothelial tissues, such as lymphoid, splenic,bone marrow, circulating white blood cells andcells obtained from mononuclear cellular exudatesinto the peritoneal and pleural cavities. Cruick-shank (I95i) and Everett, Livingood, Pomerat andFunan (1952) have reported that tuberculin is nottoxic for tissue cultures of skin from hypersensitiveanimals. The emphasis on cells of reticulo-endothelial origin was heightened by the reportsof Landsteiner and Chase (1942) and Chase (I945),that contact and tuberculin hypersensitivity arepassively transferred by injection of mononuclearcellular exudate or lymphoid or splenic cells, de-layed type reactions being obtained on testing afteran appropriate interval. Lawrence (1949, 1952)reported successful passive transfer of tuberculinand streptococcal sensitivity in humans by in-jection of circulating white cells, and he found thatthe intensity of reactions elicited depended on thedosage of cells and the concentration of the antigeninjected.The antibody carrying cell has been identified

as the lymphocyte and Wesslen (I952) has usedpure lymphocyte suspensions from thoracic ductlymph for successful passive transfer whilstKirchheimer et al. (I95i) reported that neutro-phile cellular exudates failed to transfer the hyper-sensitivity. The relationship between the numberof lymphocytes and the dose of antigen has alsobeen reported by Wesslen, who found that the in-tensity of the reaction depended on these quantita-tive factors. The report that passive transfer oftuberculin hypersensitivity was not inhibited bycortisone administration to donor or recipient(Cummings and Hudgins, 1952) though it does in-hibit the tuberculin reaction in actively sensitizedsubjects, indicates once more the importance of thecells in determining the reaction. The belief thatall types of cell are sensitized in delayed typehypersensitivity is not borne out by these findings,which suggest that the delay in the onset of re-action is, in fact, a reflection of the time requiredfor the arrival of enough antibody carrying lym-phocytes and that the blood vessels must par-ticipate functionally by conveying these cells to theneighbourhood of the reaction. Histological evi-dence shows that tuberculin reactions start withinthe first one to two hours (Laporte, I934) and itseems likely that the macroscopic reaction may

only appear when enough sensitized cells have re-acted with the antigen to excite the inflammatoryprocess.

Favour (1947) and his colleagues have reportedthat tuberculin lyses lymphocytes from sensitizedsubjects in vitro within 20 minutes, and this, too,indicates that the antigen antibody reaction is farmore rapid than the macroscopic reaction suggests.Miller and Favour (195i) also reported that thesensitized lymphocytes liberate a plasma factorwhich can sensitize normal cells and make themsusceptible to lysis by tuberculin, though Wesslen(1952) was unable to confirm this. The functionalparticipation of the underlying blood vessels inepidermal reactions in humans is suggested by thefindings of the author in investigations to be re-ported elsewhere. It was found that the intra-dermal injection of adrenaline with tuberculininhibits the reaction, and histological study ofthree and 24 hour skin biopsy specimens from theseand control areas shows a striking decrease in theinfiltrating cells in the adrenalin tested areas. Itseems likely that the vasoconstriction produced bythe adrenaline temporarily lessens the local circula-tion and slows the access of circulating antibodycarrying cells. By the time the adrenalin effect hasworn off after i0 to I2 hours, the local concentra-tion of tuberculin has probably decreased below.thereacting level. Whilst the resting reticulo-endothelial cells must certainly take part in thelater stages of the reaction, it would seem that it isthe immigration of circulating lymphocytes thatbrings about the main reaction.The influence of agents which modify tuberculin

reactions may be exerted at various points. Cor-tisone, for example, is a lymphopenio agent thusreducing the number of available antibody carry-ing cells, and in histological study of cortisone in-hibited tuberulin reactions Gell and Hinde (O95 )report that the decrease in mononuclear cells is themost prominent feature. Cortisone also enhancesthe vasoconstrictor tone of blood vessels and re-duces the endothelial damage in reactions totubercle bacilli (Ebert and Wissler, I951; Ebert,1952; Humphrey, 195I). The secondary anti-body response is also reduced by cortisone (Hal-pern et al., I952) though this aspect is not yet clearowing to the variety of experimental conditionsused by different workers. The non-specific anti-inflammatory action of cortisone would act afterthe combination of antigen and antibody and de-crease the reaction in that way (Michael andWhorton, i951; Dougherty and Schneebeli, 1950).Irradiation also inhibits tuberculin reactions andLennox, Dempster and Boag (I952) suggests thatthis is due to an alarm stimulus with adreno-cortical response, i.e. a cortisone response, butirradiation also decreases the number of lympho-

iYUIY 195 3 359

copyright. on A

pril 13, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.29.333.351 on 1 July 1953. D

ownloaded from

POSTGRADUATE MEDICAL JOURNAL

cytes directly and the quantitative reduction ofantibody carrying cells may be a cause of the in-hibition as well. A possible similar mechanism forthe ' desensitizing ' action of the agents, describedby Long, Miles and Perry (195I) D'Arcy Hart,Long and Rees (1952) and Cornforth and Long(1952) which inhibit tuberculin reactions does notappear to have been investigated. Among thesethe sphingomyelin group have been reported toproduce a lymphopenia (Fisher, Harington andLong, 195i), but there is so far no indication of themode of action of dehydroascorbic acid, of theorganic phosphates such as glucose i phosphateand glucose 6 phosphate and of the ergot deriva-tive, lysergic acid, which acts on the hexose mono-phosphate of the blood, or of the surface actingagents such as the polyoxyethylene ethers. Thesemay act at various and differing points along thesequence of antigen and antibody combinationalready described, or on the inflammatory re-sponse of the tissues to the combination of antigenand antibody.

Antihistaminic drugs do not produce a lympho-penia and their inhibitory action on tuberculin re-action in guinea pigs reported by Sarber (I948) andamply confirmed by the author may be due to theactions of these drugs on blood vessels at thetuberculin test site. This inhibition is produced ifthe drug is given in adequate dosage and wellbefore the test is carried out. It seems possiblethat by their inhibition of vascular reactions theantihistaminic drugs may impair the functionalparticipation of the vessels in bringing about thereaction thus reducing the available number ofantibody containing cells, or they may reduce thevascular inflammatory response which follows thecombination of antigen and antibody. Therationale of these inhibitory factors will only becomprehensible when it is understood at- whichpoint the drugs used will exert their action, andthe actions of cortisone for example are directednot only against the condition being treated butagainst the defence mechanisms as a whole, leadingto enhancement of infection, bacterial and virus,which is being reported with increasing frequencyin experimental animals.

Hypersensitivity to the Body's own TissueProductsThe concept of ' horror autotoxicus' of the

early immunologists implied that the body couldnot produce antibodies against its own tissue con-stituents, but this has had to be modified in re-cent years by the demonstration that homologousprotein and the body's own protein can act aspartial antigens. The development of organspecific sensitization to the body's own protein wassuggested by Verhoeff and Lemoine (1922),

Courtney (1929) and Woods (I940), who claimedthat lens protein and uveal tissue protein liberatedby trauma could produce a hypersensitive statein which the antibodies would combine with andproduce reactions in the particular organ. Theysuggested that this is the explanation of the endo-phthalmitis phaco-anaphylactica and sympatheticophthalmia which after an interval of days followtrauma to the lens or uveal tissue. Positive skintests to lens protein derived from various species,i.e. organ specificity, have been reported in thesecases and ' desensitization' with lens protein be-fore operation has been stated to decrease the in-cidence of reactions (Goodman, 1935). The ex-perimental production of organ specific hyper-sensitivity with homologous tissue was reported byBurky (I934), who injected bacterial productssuch as staphylococcus toxin as adjuvants withlens protein and muscle. In sensitized animalstrauma to the lens or muscle produced reactionsout of proportion to the injury, and it is suggestedthat the liberation of the organ protein by traumaprovided the antigen for combination with theantibody produced by sensitization. Similar re-sults have been reported by Lucic (I939) using amixture of uveal tissue* and toxin as the antigen.

Whitfield (1922) suggested that auto-sensitiza-tion to skin breakdown products occurs in skindiseases and. leads to extension and persistence ofthe condition, and Hecht, Sulzberger and Weil(1943) have produced anti-skin antibodies byinjecting skin extracts plus staphylococcal toxininto experimental animals. Auto-antibodies tohomologous skin have been demonstrated byZoutendyk and Gear (1950) in patients with lupuserythematosis.The production of disseminated encephalo-

myelitis by injection of homologous brain plusadjuvants such as liquid paraffin and tuberclebacilli has been reported by Morgan (1946, 1947)and Kabat, Wolf and Bezer (1946, I947). Anti-bodies to normal lung tissue have been reportedin atypical primary pneumonia-by Thomas et al.(I943), and to human heart in rheumatic fever byCavelti (1945)-With increasing refinement of bacterial ant

tissue antigens these problems will probably beclarified, since some findings such as the pro-duction of experimental nephritis by these methodsdescribed by Cavelti and Cavelti (1945) has notbeen confirmed by Humphrey (1948). This pos-sibility is suggested by the report of Glynn andHolborrow (I952) that the injection of a mixtureof chondroitin sulphate, which is a normal con-stituent of the collagen of the body, together withhaemolytic streptococci produced a hypersensitivestate in which mesenchymal lesions such asarthritis were produced and immediate whealing

360 YUyIY 1953

copyright. on A

pril 13, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.29.333.351 on 1 July 1953. D

ownloaded from

&Sl 1953 PEPYS: Basic Mechanisms of Allergic Reactions 36I

reactions were obtained with the carbohydratepartial antigen. The amount of carbohydrate re-quired was minute and the findings are highlysignificant since the ease of formation of the antigenfrom the chondroitin seems within the range ofnormal infection. Where such a hypersensitivestate has been established it seems that both libera-tion of the tissue partial antigen or infection withthe specific organism could elicit reactions in theparticular tissue from which the partial antigen isobtained.

Clinical Allergic Disorders in ManThere is no explanation for the characteristic

ease of sensitization of so-called allergic subjects,and no difference in the basic antigen-antibodymechanisms has been demonstrated between' allergic' and sensitized non-allergic human sub-jects, or between them and experimental hyper-sensitivity in animals. The antigen-antibodyreaction initiates all the phenomena of allergicconditions in man under controlled experimentalconditions, but in clinical practice these reactionsare complicated by many other factors which aresecondary. The concept of the ' allergic equilib-rium' makes it possible to assess these factorsobjectively. The ' allergic state ' and the ' allergiccondition' should be distinguished, and sincewe have no means of altering the constitutionalallergic background the management of thecondition must be directed towards maintaininga ' sub-clinical' allergic state. Increased allergicinsult or the supervention of factors such asinfection, endocrine, psychosomatic and othercauses may overcome the clinical thresholdand precipitate the condition. At this stagethe antigen-antibody reaction may remain theinitiating factor and a direct allergic approach ispossible, but where the clinical condition has ac-quired its own momentum, precipitating factors ofmany sorts can maintain the disorder. The tippingof the balance in favour of the subject by a de-crease of the allergic insult frequently leads totolerance of the secondary factors which then nolonger produce the clinical disorder. The largenumber of possible combinations of factors whichcan operate to the patient's disadvantage also con-tain instances where control of secondary factorsreduces the condition to a sub-clinical level, andthis accounts for the many and varied methodsof treatment of allergic disorders for which goodresults are claimed.The controlled experimental studies which

have been- described, however, indicate thatdecreased exposure or avoidance of the antigenwill prevent the development of sensitization,a point of considerable importance in childrenin allergic families in whom the hereditary

factors enhance the likelihood and early appearanceof clinical allergic disorders. To take only twoexamples of the application of our understandingof this aspect, protection from the recognizedcommoner allergens is therefore of value in in-fancy and is of importance in regard to domesticand occupational exposure in later life. Where sen-sitization has already been established avoidanceor decreased exposure to the antigen is once againthe most fundamental method of control, for thefull exercise ofwhich the environmental contacts ofthe patient should be thoroughly familiar to thephysician. ' Desensitization ' or, to use a pre-ferable term, ' hyposensitization ' leading to de-crease in sensitivity or increased tolerance isanother method of decreasing the effects of theantigens where continued exposure cannot beavoided. To attempt to approach clinical allergicdisorders only by non-specific methods and bypalliative treatments which do not influence theantigen-antibody mechanism on which the con-dition is based, is to deprive the subject of thebenefit of a large body of refined experimental andclinical study. In the same way allergic studycannot afford to overlook the clinical importance ofthe precipitating factors which in themselves mayhave no bearing on antigen-antibody relationships.

AcknowledgmentsMy thanks are due to Professor G. Payling

Wright for his generous advice and criticism, andto Dr. A. Mohun for reading this paper. Theauthor is in receipt of a grant from the MedicalResearch Council.

BIBLIOGRAPHYABELL, R. G., and SCHENCK, H. P. (I938), J. Immunol., 34, I9g.ARONSON, J. D. (r931), Y. exp. Med., 54, 387.ARTHUS, M. (1903), C.R. Soc. Biol., 55, 8I7.BAILLY, M. D. (I950), Y. Immunol., 64, 245.BALDRIDGE, C. D., and KLIGMAN A. M. ('95I), Amer. Rev.

Tuberc., 63, 674.BARBER, M., and DELAUNAY, A. (I952), Postgrad. med. 3'., 28,

351.BENACERRAF, B., and KABAT, E. A. (I949), Y. Immunol., 62,

97, 517.BLOCH, B., and STEINER-WOURLISCH, A. (I926), Arch. f.

Dermat., 152, 283.BURKY, E. L. (1934) 7. Allergy, 5, 466.CAMPBELL, C. A. (I945), Canad. Med. Assocn. J7., 52, 280.CAVELTI, P. A. (I945), Proc. Soc. exp. Biol., N.Y., 6o, 379.CAVELTI, P. A., and CAVELTI, E. S. (I945), Arch. Path., 40,

I58, I63.CHASE, M. W. (I94I),7. exp. Med. 73, 7I.CHASE, M. W. (I942), Proc. Soc. exp. Biol., N.Y., 49, 688.CHASE, M. W. (I945), Ibd., 59, 134.CHASE, M. W. (1946), Ibid., 6I, 257.CHASE, M. W. (I949), Soc. Amer. Bacteriologists, I949 MeetingCHASE, M. W. (i9si), Fed. Proc., I0, 404.CHASE, M. W. (I952), Amer. 3. Med., 13, 357.CODE, C. F. (I952), Physiol. Rev., 32, 47.COOKE, R. A., BARNARD, J. H., HEBALD, S., and STULL, A.

(I935),J. exp. Med., 62, 733.CORNFORTH, J. W., and LONG, D. A. (1952), Lancet, i, 950.COULSON, E. J., and STEVENS, H. ('949), 7. Immunol., 6i,

}, I It 119.

copyright. on A

pril 13, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.29.333.351 on 1 July 1953. D

ownloaded from

362 POSTGRADUATE MEDICAL JOURNAL July I953

COURTNEY, R. H. (I929), Amer. J. Ophth., 12, 20.CRUICKSHANK, C. N. D. (i95I), Nature, i68, 2o6.CUMMINGS, M. M., and HUDGINS, P. C. (I952), J. Immunol.,

69, 331.DALE, H. (I948), Brit. med. J., ii, 28I.D'ARCY HART, P., LONG, D. A., REES, R. J. W. (I952), Ibid.,

i, 68o.DEAN, H. R., and WEBB, R. A. (I924), J. Path. and Bact., 27, 65.DERICK, C. L., and SWIFT, H. F. (I929), J. exp. Med., 49, 615.DIENES, L., and SCHOENHEIT, E. W. (1929), Amner. Retv.

Tuberc., 20, 92.DIENES, L., and SIMON, F, A. (I935), 7. Immunol., 28, 321.DOERR, R., and BERGER, M. (I922), Zeit. f. Hyg., 96, I9I.DOUGHERTY, T. F., and SCHNEEBELI, C. F. (I950), Proc.

Soc. exp. Biol., N.Y., 75, 854.EBERT, R. H., and WISSLER, R. W. (i95i), 7. Lab. and Clin.

Med., 38, 497.EBERT, R. H. (1952), Amer. Rev. Tuberc., 65, 64.EISEN, H. N., ORRIS, L., and BELMAN, S. (I952), . exp. Med.,

95, 473.e SILVA, R. (I950), Ann. N.Y. Acad. Sciences, 50, 145.EVERETT, E. T., LIVINGOOD, C. S., POMERAT, C. M., and

FUNAN, H. V. (I952), J. Invest. Dermatol., x8, 193.FABRIZIO, A. M. (1952), Amer. Rev. Tuberc., 65, 250.FAVOUR, C. B. (I947), Proc. Soc. exp. Biol., N.Y., 65, 269.FISCHEL, E. E., and KABAT, E. A. (1947), J. Immunol., 55, 337.FISHER, N., HARINGTON, C., and LONG, D. A. (iosi),

Lancet, 1i, 522.FRANCIS, T., JUN., and TILLETT, W. S. (1931), J. exp. Med.,

54, 587.GELL, P. G. H., HARINGTON, C. R., and RIVERS, R. P.

(I946), Brit. J. exp. Path., 27, 267.GELL, P. G. H., and HINDE, I. T. (I9s5), Ibid., 32, 5S6.GILBERT, A. J. (1938), J. Pharm. and exp. Therap., 62, 228.GLYNN, L. E., and HOLBORROW, E. J. (1952), Lancet, 1i, 449.GOODMAN, E. L. (I935), Arch. Ophth., 14, 90.HALPERN, B. N., MAURIC, G., HOLTZER, A., and BRIOT,

M. (i9si), Acta. allerg. Kbh., 4, 207.HARLEY, D. (I937), Y. Path. and Bact., 44, 589.HAWN, C. VAN Z., and JANEWAY, C. H. (I947), J. exp. Med.,

85, 571.HECHT, R., SUZLBERGER, M. B., and WEIL, H. (I943),

Ibid., 78, 59.HEILMAN, D. H., FELDMAN, W. H., and MANN P. C. (1944),

Amer. Rev. Tuberc., 50, 334.HEKTOEN, L. (1932), Arch. Path., 14, 837.HUMPHREY, J. H. (I948), J. Path. and Bact., 6o, 21 1.HUMPHREY, J. H. (I95I), Brit. Y. exp. Path., 32, 274.HYDE-SALTER, H. H. (I868), 'Asthma,' Churchill, London.KABAT, E. A., WOLF, A., and BEZER, A. E. (1946), Science,

104, 362.KABAT, E. A., WOLF, A., and BEZER, A. E. (I947), J. exp. Med.,

85, 117.KABAT, E. A., COFFIN, G. S., and SMITH, D. J. (I947), 7.

Immunol., 62, 97, 517.KABAT, E. A. (1952), Amer. Y. Med., 13, 352.KIRCHHEIMER, W. F., HESS, A. R., and SPEARS, R. G.

(I95I), Amer. Rev. Tuberc., 64, Si6.KLEMPERER, P. (I947),' Allergy,' R. A. Cooke, Saunders, Phil.,

P. 76.KLINGE, F. (I929), ' Allergie,' W. Berger and R. Hansen, Thieme,

Leipzig.LANDSTEINER, K., and CHASE, M. W. (I940), Proc. 3rd Inter-

nat. Congress Microbiol., 4, 772.LANDSTEINER, K., and CHASE, M. W. (1942), Proc. Soc. exp.

Biol., N.Y., 49, 688.LANDSTEINER, K. (1945), 'The Specificity of Serological Re-

actions,' Harvard Univ. Press, Camb., Mass.LAPORTE, R. (1934), Ann. Inst. Pasteur, 53, 598.LAWRENCE, H. S. (I949), Proc. Soc. exp. Biol., N.Y., 71, 5i6.LAWRENCE, H. S. (1952), Y. Immunol., 68, 159.LENNOX, B., DEMPSTER, W. J., and BOAG J. W. (1952),

Brit. Y. exp. Path., 33, 380.LONG, D. A., MILES, A. A., and PERRY, W. L. M. (195s),

Lancet, i, io85.LONG, D. A., MILES, A. A., and PERRY, W. L. M. (1951),

Ibid., ii, 902.LOVELESS, M. H. (I940),.7. Immunol., 38, 25.LOVELESS, M. H. (I94I), Ibid., 41, 15.LOVELESS, M. H. (1942), Ibid., 44, I.LOVELESS, M. H. (1943), Ibid., 47, i6s.LUCIC, H. (1939), Proc. Soc. exp. Biol., 40, 273.

MAcKENZIE, G. M., and LEAKE, W. H. (1921), J. exp. Med.,33, 6oi.

MARRACK, J. R. (i95i), Internat. Arch. Allergy., 2, 264.McINTOSH F. C., and PATON, W. D. M. (1949), J. Physiol.,

109, 190.McMASTER, P. D., and KRUSE, H. (i9si),J. exp. Mcd., 94, 323.MAHLMAN, J., and SEEGAL, B. C. (I934), J. Immunol., 26, I.MEYER, K., and LOEWENTHAL, H. (1927), Z. f. Immunitats-

forsch. u. exp. Therap., 54, 420.MICHAEL, M., and WHORTON, C. M. (i95i), Proc. Soc. exp.

Biol., N.Y., 76, 754.MILLER, J. M., and FAVOUR, C. B. (I95I), J. exp. Med., 93, 1.MOEN, J. K., and SWIFT, H. F. (I936), Ibid., 64, 339, 943.MOEN, J. K. (I936), 7. exp. Med. 64, 355.MORE, R. H., and MACLEAN, C. R. (I949), Amer. 7. Path., 25,

413.MORGAN, I. M. (1946), 7. Bact., 5S, 614.MORGAN I. M. (I947), Y. exp. Med., 85, 13I.MOTE, J. R., and JONES, T. D. (I936), Y. Immunol., 30, 149.OBERMAYER, F., and PICK, E. P. (I906), Wien. Klin. Wchnschr.,

19, 327.OPIE, E. L. (1924),.7. Immunol., 9, 231, 247, 255, 259.OTTO, R. (I907), Muenchen med. Wchnschr., 54, I665.PORTIER, P., and RICHET, C. (1902), Compt. Rend. Soc. de

Biol., 54, 170.PRAUSNITZ, C., and KUESTNER, H. (I92I), Zentralbl. f.

Bakt. (Orig.), 86, x6o.RAFFEL, S. (1946), Amer. Rev. Tuberc., 54, 564.RAFFEL, S. (1948),.7. Infect. Dis., 32, 267.RAFFEL, S., and FORNEY, J. E. (I948), Y. exp. Med., 88, 485.RAFFEL, S., ARNAUD, L. E., DUKES, C. D., and HUANG

J. S. (I949), Ibid., 90, 53.RAJKA, E., and HEGYI, E. (I950), Internat. Arch. Allergy, 1, 243.REDDIN, L. (I949), Annals N.Y. Acad. Sciences, 50, 692.RICH, A. R., and LEWIS, M. R. (1932), Bull. Johns Hopkins Hosp.,

50, 1 5.RICH A. R., and FOLLIS, R. H., JUN. (1940), Ibid., 66, io6.RICH A. R., and GREGORY, J. E. (I943), Ibid., 72, 65.RICH, A. R., and GREGORY, J. E. (1943), Ibid., 73, 239.RICHET, C. (1907), Ann. Inst. Pasteur, 21, 497.ROSA, L. M., and MAcDOWALL, R. J. S. (i95i), Acta. allerg.

Kbh., 4, 293.SARBER, R. W. (1948), Amer. Rev. Tuberc., 57, 504.SCHILD, H. O., HAWKINS, D. F., MONGAR, J. L., and

HERXHEIMER, H. (I95i), Lancet, Ri, 376.SEEGAL, D., and SEEGAL, B. C. (1931), Y. exp. Med., 54, 249,

265.SEEGAL, B. C., and SEEGAL, D. (I933), 7. Immunol., 25, 221.SEEGAL, B. C., and WILCOX, H. B., JUN. (1940), Arch. Path.,

30, 416.SEEGAL, B. C. (1952), Amer. Y. Med., 13, 354.SEIBERT, F. B., and MORLEY, N. (1933), Y. Immunol., 24, 149.SOLLMAN, T., and GILBERT, A. J. (I937), 7. Pharm. and exp.

Therap., 6I, 272.STETSON, C. A. (T9SI), 7. exp. Med., 94, 347.SULZBERGER, M. B. (1940), 'Dermatologic Allergy,' Thomas,

Springfield.THOMAS, L., CURNEN, E. C., MIRICK, G. S., ZEIGLER,

J. E UN., and HORSFALL, F. L., JUN. (I943), Proc. Soc.expA.1t:l., N. Y., 52, I 2 1.

URBACH, E., and GOTTLIEB, P. M. (1946), 'Allergy,' Gruneand Stratton, New York.

VERHOEFF, F. H., and LEMOINE, A. N. (1922), Amer. Y.Ophth., S, 700.

v. PIRQUET, C., and SCHICK, B. (1905), 'Serum Sickness,'Williams & Williams, Baltimore,*95i.

v. PIRQUET C. (I906), Muenchen med. Wchnschr., 30, 1457.WEBB, R. A. (1924), Y. Path. and Bact., 27, 79.WESSLMN, T. (1952), Acta. Tuberc. Scand., 26, 38, 175.WESSLtN, T. (1952), Acta. Dermato-Venereol., 32, 195.WHITFIELD, A. (I922), Brit. J. Dermat., 34, 331.WILCOX, H. B., JUN., and ANDRUS, E. C. (1938), . exp. Med.,

67, I 69.WITTICH, F. W. (I949), 'Progreas in Allergy,' P. Kallos, Karger,

Basel.WOODS (1940), 'Modern Trends in Ophthalnology,' Butter-

worth & Co., London.WRIGHT, G. P., and HOPKINS, S. J. (I94I), . Path. and Bact.,

53, 243.WRIGHT, G. P., and LAWS, J. 0. (I952), Brit.J7. exp. Path.,

33, 343.ZOUTENDYK, A., and GEAR, J. H, S. (I950), Brit. med.J7., ii,

"175.

copyright. on A

pril 13, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.29.333.351 on 1 July 1953. D

ownloaded from