Postgraduate Medical Journal (1985) 61, 925-933

Drug eruptionsS.K. Goolamali

Central Middlesex Hospital, London and Northwick Park Hospital and Clinical Research Centre, Harrow,Middlesex, UK.

Introduction

The last four decades have seen an enormous increasein the use of potent and toxic drugs as therapy. Theconsiderable advances that these drugs have providedin the control and prevention of disease are matchedby their potential to produce unwanted side effectsmany of which are reflected in the skin.These side effects cannot in the main be predicted.

Tests in animals may not reliably be extrapolated tothe human and only occasionally can the occurrence ofside effects be gauged from the structural formula of anew drug. With some fifty new drugs developed eachyear the practitioner is challenged simply to keep pacewith their therapeutic indications. The results of onesurvey suggested that 30% of prescriptions fromgeneral practitioners was based on data supplied bydrug companies. In 1985 it is also increasingly com-mon to find that patients, when faced with a chronicdisease have sought several medical opinions or havetaken refuge in remedies which rely more on advertis-ing than their therapeutic effectiveness. The availableinformation suggests then that drug reactions arelikely to be common.

In a recent study (Black & Somers, 1984), drug-related illness accounted for some 6% of hospitaladmissions in a year. It has been estimated that theaverage inpatient in a British hospital receives fivedrugs whilst one in an American hospital receives nine.Since a large majority of allergic reactions affect theskin the incidence of drug-induced cutaneous erup-tions probably reflects the risk of allergic drug reac-tions in general. Reactions to drugs in over twentythousand medical inpatients were studied in a largecollaborative study (Arndt & Jick, 1976). Skin reac-tions occurred in a little over 2%. Despite the lack ofdata on non-hospitalized patients it is clear that even ifonly 2% of patients receiving drugs experience a 'rash'it is a number that we can all ill afford.

Mechanisms of adverse drug reactions

An adverse drug reaction is commonly defined as anyresponse to a drug that is unintended and occurs atdoses used in the prophylaxis or therapy of disease.Adverse reactions may occur as a result of inap-propriate metabolism, idiosyncrasy or hypersen-sitivity to a drug. Impaired metabolism may resultfrom altered oxidation or acetylation ofa drug as maybe seen in slow-acetylators in isoniazid-induced acne.An idiosyncratic reaction is an inbuilt abnormalresponse to a drug such as may occur when bar-biturates or sulphonamides are given to a patient withporphyria. Hypersensitivity or allergy to a drug ismediated by an antigen-antibody reaction. Previousexposure and sensitization to the drug is aprerequisite though it is possible for a drug to induceantibody formation without clinical symptoms ofhypersensitivity.

Hypersensitivity reactions are normally classifiedinto four types (Gell & Coombs, 1963) - Type 1(immediate or anaphylactic), Type 2 (cytotoxic orautoallergic), Type 3 (immune complex disease) andType 4 (cell-mediated or delayed). The vast majority ofdrug reactions are due to hypersensitivity and thisreview will be restricted to a brief discussion of theseallergic responses and a description of some of thecommon skin reactions provoked by drugs.

Type 1 reaction

This is mediated by IgE or reagin. IgE antibodies areproduced in response to a drug or drug-haptencomplex and attach themselves preferentially to themast cell. On exposure to antigen there is mediatorrelease from the mast cell which then results in thewell-described reactions of urticaria, angio-oedema,asthma and, in severe cases, anaphylaxis.

Type 2 reaction

This may present with thrombocytopenia, leucopeniaor haemolytic anaemia. In this reaction the antigen(drug) combines with a platelet or red cell, forexample, and stimulates the formation ofantibodies tothe drug-cell combination. Sedormid purpura, quinine

© The Fellowship of Postgraduate Medicine, 1985

Correspondence: S.K. Goolamali, M.D., F.R.C.P., North-wick Park Hospital, Watford Road, Harrow, MiddlesexHA1 3UJ, UK.

copyright. on A

pril 1, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.61.720.925 on 1 O

ctober 1985. Dow

nloaded from

926 S.K. GOOLAMALI

purpura, many of the haemolytic anaemias, includingthat due to high dose penicillin therapy (Petz &Fudenberg, 1966) and methyldopa, are examples ofthis type of reaction. Some 10 to 20% of patientsreceiving methyldopa develop a positive Coombs' testbut only 0.5 to 1.0% develop a haemolytic anaemia.Approximately 70% of reported drug-inducedimmune haemolytic anaemias are a result of methyl-dopa. Leucopenia of immunological origin often withthe presence of anti-leucocyte agglutinins is not in-frequently a reaction of drugs.

Type 3 reaction

The initial exposure to the drug sensitizes and resultsin antibody, usually IgM or IgG, production. After aninterval antigen-antibody complexes form and aredeposited in the peripheral circulation. Complement isactivated and causes accumulation of neutrophilicleucocytes which release lysosomal enzyme that des-troys tissues. The commonly recognized systemicreaction in this category is the serum sickness syn-drome. Fever, arthralgia and lymphadenopathy aretypical features and classically arise from hypersen-sitivity to foreign serum. A similar clinical picture mayresult 1 to 3 weeks after treatment with penicillin,para-aminosalicylic acid or sulphonamides. Vas-culitis, when a feature of the Type 3 reaction, is oftenattributed to sulphonamides but may also be causedby phenytoin, thiouracils, chlorpromazine, aspirin,gold and penicillin.

Type 4 reaction

This T-cell mediated or delayed hypersensitivity typeof reaction is the basis of an allergic dermatitis oftendue to a topical medication. In a large European study(Bandmann et al., 1972), a third ofthe patients with anallergic contact dermatitis showed a hypersensitivityto a topical medicament. Sensitization is more likely tooccur if the medication is applied to damaged skin.The importance of this 'contact' allergy is that itprohibits the systemic use of the same or related drug.If the offending drug is mistakenly taken it produces awidespread dermatitis. Amongst the contact sen-sitizers ethylenediamine hydrochloride used as astabilizer in some creams is of particular importance.It cross-reacts with drugs used systemically such asaminophylline which consists of two-thirds theo-phylline and one-third ethylenediamine. Some antihis-tamines which are derived from ethylenediamine mayalso cause a generalized eruption if given to a sen-sitized patient (Fisher, 1976). Another recognizedsensitizer is benzocaine found in anti-pruritic prepara-tions, in local applications for the treatment ofhaemorrhoids and in some sunburn remedies. Ben-zocaine is a derivative of para-aminobenzoic acid

(PABA) and therefore may cross-react with otherPABA compounds such as procaine and dubucaine.Topical antihistamine, freely available as an over-the-counter preparation, is a commmon sensitizer as isneomycin especially when applied to chronic stasisdermatitis or otitis externa (Leyden& Kligman, 1979).

Types of cutaneous reactions

Skin reactions to drugs are diverse. In a series of 464cases during 1966-1970, Kuokkanen (1972) foundthat exanthematous eruptions were the most commonand accounted for 46% of reactions. Urticariaoccurred in 23% of cases, fixed drug eruption in 10%,erythema multiforme in some 5% and other reactionsin less than 5% each. In another more recent study(Kauppinen & Stubb, 1984) exanthemas again formedthe largest group, 42% of 446 inpatients, fixed erup-tions occurred in 21%, urticaria-angioedema in 12.5%and erythema multiforme with Stevens-Johnson syn-drome accounted for approximately 6%. The remain-der comprised eczema, toxic epidermal necrolysis,photosensitivity reactions, purpura and the systemiclupus erythematosus (SLE) syndrome.Exanthematous eruption

This type of reaction can be caused by almost anydrug. Itch may or may not be associated and differen-tiation from a viral exanthem may at times be difficult.The greatest incidence occurs with the penicillins andthe eruption commonly consists of maculo-papularerythematous lesions. The risk ofdeveloping penicillinhypersensitivity is increased in patients with a previoushistory of other drug allergy and in those ofwhom thedrug is given parenterally. In one study (Shapiro et al.,1969) a rash was reported in 4.5% of 622 patientstreated with penicillins other than ampicillin, in 9.5%of 422 patients treated with ampicillin and in 1.8% of2941 patients not receiving either of these drugs.Although ampicillin may provoke urticaria, it com-monly causes a morbilliform eruption (Figure 1)which lasts a few days. An increased incidence ofampicillin rash occurs in infectious mononucleosis,cytomegalovirus mononucleosis, lymphatic leukaemiaand in viral respiratory infections (Almeyda & Levan-tine, 1972). In these conditions the rash may also besomewhat prolonged.Other drugs commonly associated with an exanth-

ematous eruption include barbiturates, sulphon-amides and allied anti-diabetic and diuretic drugs,phenytoin, erythromycin, allopurinol and gold salts.Gold salts were first used in France in the 1920s and1930s though their efficacy as an anti-rheumatic drugwas confirmed very much later (Fraser, 1945). Twoforms of these drugs are in common use, the aqueous

copyright. on A

pril 1, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.61.720.925 on 1 O

ctober 1985. Dow

nloaded from

DRUG ERUPTIONS 927

Figure 1 Ampicillin reaction.

based sodium aurothiomalate and the oil basedaurothioglucose. Both drugs are administered in-tramuscularly. Overall, approximately 30% ofpatients started on chrysotherapy may, at least tem-porarily, have to discontinue the drug because of sideeffects, the most common being a rash or mucousmembrane lesions. The most common eruption iserythematous, maculo-papular and pruritic, but a

lichen-planus like eruption or a pityriasis rosea-likerash also occurs (Penneys et al., 1974). After thereaction has settled gold therapy can generally be

Figure 2 Penicillin induced urticaria.

restarted without incident (Klinefelter, 1975).Allopurinol, an analogue of hypoxanthine, is an

inhibitor of xanthine oxidase and is used widely in theprophylaxis of gout. Its major side effect is an

erythematous maculopapular eruption thoughoccasionally urticaria and toxic epidermal necrolysishave been seen. A drug rash has been noted in some2% of patients taking allopurinal alone (BostonCollaborative Drug Surveillance Program, 1972) butthe incidence rose to 22.4% in patients treated simul-taneously with both allopurinol and ampicillin.

Table I Hypersensitivity drug reactions

Type Reaction Drug commonly responsible1 Urticaria: Penicillin

Angio-oedema: AspirinAsthma: Foreign seraAnaphylaxis: Penicillin

2 Thrombocytopenicpurpura: Quinine

Granulocytopenia: GoldHaemolytic anaemia: Methyldopa

3 Vasculitis: SulphonamidesSerum sickness: PenicillinL.E. syndrome: Hydralazine, Procainamide

4 Contact dermatitis: EthylenediamineNeomycinTopical antihistamine

copyright. on A

pril 1, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.61.720.925 on 1 O

ctober 1985. Dow

nloaded from

928 S.K. GOOLAMALI

Urticaria and angioedema

Urticaria or 'hives' presents as well-defined, eryth-ematous, raised often oedematous patches in the skin.The lesions are usually multiple, of variable size andinvariably pruritic. Individual wheals seldom lastmore than 24 to 48 hours which is a useful diagnosticsign. Episodes lasting less than 6 weeks are arbitrarilyclassified as acute and those longer than 6 weeks aschronic. Angioedema is characterized by plaques ofsubcutaneous oedema which may be painful andtender. Urticaria results from degranulation of mastcells in the dermis, releasing histamine and othermediators. Penicillin and its derivatives are the mostcommon causes ofa Type 1 reaction in which urticariamay be a prominent feature (Figure 2, Table I).Cephalosporins, sulphonamides, blood products andvaccines prepared in eggs may also induce an urticariaas part of the Type 1 hypersensitivity reaction. Avariety of drugs may induce histamine release by adirect action on mast cells. Morphine sulphate, curareand polymyxin antibiotics belong to this group.Urticaria is often a feature of the serum sicknesssyndrome, and, finally, aspirin and non-steroidalanti-inflammatory agents may induce urticaria (Ros etal., 1976). It is postulated that these drugs inhibitprostaglandin production and the lack of the in-hibitory action of prostaglandins allows the develop-ment of urticaria.

In a 17 year prospective study of the clinicalcharacteristics of patients with a history of allergy toaspirin (Speer et al., 1981), the most common manifes-tation was urticaria-angioedema. In these patients90% were also sensitive to inhalants (76%), foods,(74%) and other drugs (43%). The ten most commondrugs which may induce urticaria as reported to theCommittee on Safety of Medicines in the UnitedKingdom between the years 1964 and 1983 (Griffin,1983) are shown in Table II.

Fixed drug eruption

The term fixed drug eruption (l'eruption fixe) wasintroduced by Brocq (1894) when he described threepatients in whom an erythematous pigmented erup-tion developed after ingesting antipyrine. It usuallypresents as a solitary pruritic erythematous maculewhich forms into an oedematous plaque. Vesicles andbullae may develop later. When the acute phasesubsides there is residual pigmentation which becomesmore prominent with each exposure to the causativedrug. With repeated attacks existing lesions increase insize and new lesions appear. The lesions may appearanywhere on the body though the lips, hands and glanspenis (Figure 3) are commonly affected. The length oftime from re-exposure to a drug and onset of symp-toms varies between half an hour to 8 hours (Stubb,

Table II The top .ten causes of drug-induced urticariareported to the Committee on Safety of Medicines

1964-1983 (Griffin, 1983)

CotrimoxazoleFenbufenBenoxaprofen*AmpicillinFenclofenacFeprazoneCimetidineAlclofenac*Nalidixic acidAmoxycillin* Drug now discontinued.

1976) Antihistamines do not prevent a fixed drugeruption. In general a single drug is responsible butsome patients may exhibit a fixed drug eruption toseveral drugs. In a recent review of the literature(Korkij & Soltani, 1984) 68 drugs had been reported asresponsible for a fixed drug eruption. Common offen-ders include quinine, sulphonamides, phenolphtha-lein, barbiturates, oxyphenbutazone and chlor-diazepoxide. Phenolphthalein is considered a safe andreliable drug and hence found in proprietary pur-gatives, and because it is tasteless it is included in thosesold in the form of chewing gum or chocolate.The mechanism which determines the localization

!tIFigure 3 Balano-posthitis caused by fixed drug eruptionshowing lesions.

copyright. on A

pril 1, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.61.720.925 on 1 O

ctober 1985. Dow

nloaded from

DRUG ERUPTIONS 929

of the skin lesions remains unexplained. Patch testingon unaffected skin with the drug responsible for thefixed eruption has invariably produced negative re-sults whilst occasionally a positive response has beenobtained when the patch test has been applied topreviously affected but 'healed' skin (Welsh, 1961). Ablood-borne mediator has been identified as a possiblecause in a phenolphthalein-induced fixed eruption(Wyatt et al., 1972) but this fails to explain why thelesions are so localized.

Erythema multiformeThe diagnostic lesion is the iris or target lesion (Figure4). As the name multiforme implies it may present withseveral different types of lesions. These may consist ofurticated papules or, as in the severe form of thedisease, bullae, which become widespread and areassociated with ulceration of the oral and genitalmucosa-the Stevens-Johnson syndrome.Erythema multiforme is considered to be a hyper-

sensitivity reaction commonly to a drug or to anunderlying virus or bacterial infection. The common-est associations are with preceding herpes simplex ormycoplasma infection. Streptococcal infection, infec-tious mononucleosis, tuberculosis, sarcoidosis, X-raytherapy of malignant disease and sometimes malig-nancy itself are other recognized causes. The mostcommon drug causes of erythema multiforme aresulphonamides, phenytoin, barbiturates, phenyl-butazone, sulphonylureas, penicillin and salicylates.

Figure 4 Erythema multiforme.

In a recent review of the literature 40 drugs had beenimplicated as the cause of erythema multiforme (Huffet al., 1983). Sulphonamides were the most frequentoffenders. Sulphonamide-associated erythema mul-tiforme usually appears a week or two after therapybut may occur within hours in patients who havepreviously been sensitized to the drug. Stevens-John-son syndrome may also be caused by the long actingsulphonamides (Carroll et al., 1966), penicillins,diphenylhydantoin and chlorpropamide (Bianchine etal., 1968).

Photosensitivity

Photosensitivity is an all-encompassing term to des-cribe untoward reactions to non-ionizing radiation.There are two main types - phototoxicity andphotoallergy. A phototoxic reaction relies on a highconcentration of a photosensitizing agent in the skinand occurs when a sufficient quantity of light ofappropriate wavelength reaches the skin. Everyone iscapable of exhibiting a phototoxic reaction whichpresents as erythema, urticaria and oedema within 24hours of light exposure. The eruption is confined tolight exposed areas commonly the face, V area ofneck,dorsum of the hands and the feet. Acute erythema isusually followed by hyperpigmentation. A phototoxicreaction may occur on first administration and sub-sides quickly on withdrawal of the drug. Drugsrecognized as causes include the tetracycline groupespecially 'demethylchlortetracycline, nalidixic acid,chlorothiazide, sulphonamides and griseofulvin.A photoallergic reaction is less common and may

occur with small quantities of a drug. Photoallergymay be of the 'immediate' type and then commonlypresents as solar urticaria or produces an eczema as aType 4 reaction (Horio, 1984). The latter variety is themore frequently encountered. The eruption, unlike thephototoxic reaction, may spread to areas which havenot been light exposed. Normally there is a delay of48hours before the onset ofthe reaction. Although drugsadministered systemically will induce photoallergy it isimportant to recognize that a photoallergic dermatitismay be caused by a topically applied photosensitizer.In practice it may be difficult to differentiate between aphototoxic and a photoallergic reaction in an in-dividual patient especially as similar drugs produceboth types of reaction. Those drugs which may causephotoallergy include chlorpromazine, chlorothiazide,sulphonamides and griseofulvin.LE syndromeA syndrome clinically indistinguishable from systemiclupus erythematosus may be provoked by a varietyof drugs. Those commonly identified includehydralazine, procainamide, isoniazid and phenytoin.

copyright. on A

pril 1, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.61.720.925 on 1 O

ctober 1985. Dow

nloaded from

930 S.K. GOOLAMALI

The development of SLE with these drugs appears tobe dose related and with hydralazine a dose greaterthan 200 mg/d (300mg in rapid acetylators) is morelikely to induce the syndrome.The syndrome may present with a variety of skin

signs which include erythema with the familiar'butterfly' rash, urticaria, purpura, photodermatitis,alopecia, livedo reticularis, cutaneous vasculitis, hy-perpigmentation and nail fold telangiectasia. Arthral-gia and pleurisy also occur but renal and centralnervous system involvement is uncommon. Tests forantinuclear factor are invariably positive and LE cellsare commonly found. Antibodies to single stranded-DNA may be found but those against double strandedDNA are rare. A genetic predisposition to the syn-drome has been claimed because the HLA antigenDRW-4 is commonly identified in patients. Indeed thecombination of HLA-DRW-4, female sex, slowacetylator status and a minimum dose of 200 mg/d ofhydralazine almost inevitably led to the syndrome inone study (Batchelor et al., 1980). Other drugs knownto provoke the syndrome are griseofulvin, methyl-dopa, chlorpromazine, para-aminosalicylic acid andthiouracil.

Lichen planus-like drug eruptionSeveral drugs produce an eruption identical with orvirtually indistinguishable from lichen planus (Figure5). Lesions in the mouth are more frequent in 'true'lichen planus but the distribution of the eruption overthe body is similar in the two varieties. In addition thelesions in both types heal with pigmentation, andhistology, in most cases, cannot separate a drug-induced lichen planus from that which arises de novo.Gold and organic arsenicals were amongst the first tobe recognized as drugs which produced a lichenoidrash. Arsenic is now considered outdated as atherapeutic measure though gold has gained inpopularity in recent years for the treatment ofrheumatoid arthritis. Mepacrine was found to inducea lichenoid eruption as long ago as the Second WorldWar when troops who took the drug as an anti-malarial agent developed the typical rash.Chloroquine will also cause a lichenoid eruption andthiazide diuretics, chlorothiazide and hydrochloroth-iazide (Harber et al., 1959) are known to cause aphotosensitive lichenoid eruption. Amiphenazole, arespiratory stimulant, may cause bone marrow depres-sion with long term use but lichenoid eruptions mayalso be induced. Methyldopa includes in its potentialside effects a lichenoid rash.

Drug-induced alopeciaHair growth is normally divided into three stages:telogen when the hair is resting, anagen when it is

Figure 5 Drug-induced lichen planus.

Figure 6 Fixed drug eruption.

copyright. on A

pril 1, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.61.720.925 on 1 O

ctober 1985. Dow

nloaded from

DRUG ERUPTIONS 931

actively growing and catagen when it is involuting. Inthe adult, approximately 85% of the scalp hair is inanagen, 14% in telogen and 1% in catagen. With hairloss associated with the contraceptive pill, the telogenhairs are shed - a phenomenon known as telogeneffluvium. Acute disease or emotional stress mayprovoke a similar hair loss. With cyclophosphamidetherapy some anagen follicles enter catagenprematurely and inhibition of mitosis results in aconstriction of the hair shaft. These hairs may be shed4-6 days after the first effective dose. Anti-cancerdrugs commonly cause alopecia of the scalp thoughwith long term use of chemotherapeutic agents theremay also be loss of axillary and pubic hair. Severeloss is common with cyclophosphamide, thenitrosoureas (e.g. lomustine) and doxorubicin. Thehair returns on stopping the drug though occasionallyit is of a different colour and texture than previously(Falkson& Schulz, 1981). Heparin, the coumarins andindandiones all cause a telogen effluvium 2 to 3months after starting anticoagulant therapy. Anti-thyroid drugs, thiouraoils and carbimazole, may alsoinduce alopecia. Loss of scalp hair occurred in fivewomen who were given carbimazole in doses varyingbetween 15 and 60mg/d for 4 to 40 weeks. Allimproved when carbimazole was discontinued or thedose reduced (Papadopoulos & Harden, 1966).

Drug-induced hyperpigmentationThe antimalarials cause pigmentary changes in some25% of patients receiving these drugs for more than 3to 4 months (Levantine & Almeyda 1973). Mepacrineproduces a diffuse yellowish discolouration of theskin. It occurs in most patients who receive the drugand fades a few weeks or several months after therapyis discontinued. Chloroquine may whiten hair and atthe same time stain the skin a bluish-grey colour.Prolonged high doses of chlorpromazine and relatedphenothiazines produce hyperpigmentation, par-ticularly in sun-exposed areas of skin (Satanove,1965). The pigmentation is cumulative and fades onlya little in winter. The skin colour changes range fromtan in the early stages to slate-grey later and finally topurple. Minocycline, a tetracycline derivative, used inthe treatment of acne vulgaris, may cause patchy ordiffuse pigmentation normally described as blue-blackin colour (McGrae & Zelickson, 1980; Simons &Morales, 1980).Chloasma is a well recognized side effect of oral

contraceptives. In one study some 3 of 10 patientsreceiving the drug developed the pigmentation (Res-nik, 1967). Discontinuation of the oral contraceptivemay cause partial remission in pigment though inmany the chloasma remains unchanged.

Hyperpigmentation is also a well recognized com-

plication ofseveral anti-cancer drugs. Those common-ly reported include bleomycin, busulphan, cyclophos-phamide, systemic fluorouracil, hydroxyurea andmithramycin. The hyperpigmentation associated withthese drugs appears to be unrelated to increasedACTH or melanocyte stimulating hormone activity.Bleomycin induced pigmentation occurs in approx-imately 30% ofpatients receiving the drug and may bediffuse, patchy or linear. The patchy pigmentation isoften prominent on pressure areas-the elbows, kneesand buttocks (Blum et al., 1973). Long term therapywith busulphan can produce a diffuse pigmentationakin to Addison's disease (Harrold, 1966). Cyclophos-phamide hyperpigmentation may be widespread(Harrison & Wood, 1972) or localized to the palmssoles or nails (Shah et al., 1978). Mithramycinproduces hyperpigmentation in 35.% ofpatients on thedrug whilst long term treatment with hydroxyureainduces both alopecia and pigmentation (Kennedy etal., 1975).Where are we after 60 years?Some sixty years ago Henry Semon in his article onDrug Eruptions in this Journal (Semon, 1926) des-cribed the problem of halogen acne, the lesions whicharsenic may induce, the use of mercury ointment astreatment ofpediculosis pubis and the development ofidiosyncrasy to aspirin and quinine. In 1985 only thelatter two drugs survive as therapeutic agents. Thediagnosis of a drug eruption is no easier today than itwas six decades ago. In many ways it is more difficultwith the profusion ofdrugs and plethora ofrecognizedside effects. Penicillin allergy can now be reliablyconfirmed by 'scratch' and 'prick' skin tests andintradermal techniques using benzylpenicilloylpolylysine conjugate and a mixture ofminor determin-ants of penicillin (Chandra et al., 1980). Patch testingis valuable in the diagnosis of a contact dermatitis butcan give misleading results when used to diagnosereactions not of the cell mediated-delayed hypersen-sitivity type. Laboratory tests otherwise have notprovided a fail-safe method of isolating a single drugfrom many as the cause of a specific drug reaction.

In the main the best the moder clinician can do is tosuspect a drug and reproduce the rash by reintroduc-ing the drug, the so called provocation test. In practicethis is not always possible nor desirable. The physicianis often forced to react to the eruption as an amateursleuth who has been invited to decide the odds on aparticular suspect drug as the cause of the cutaneousoffence. Clearly there is urgent need for research intothe vexed problem of drug eruptions which occasion-ally cause death, undoubtedly increase the length ofinpatient stay for a considerable number of patientsand are a source ofphysical and psychological discom-fort for many.

copyright. on A

pril 1, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.61.720.925 on 1 O

ctober 1985. Dow

nloaded from

932 S.K. GOOLAMALI

References

ALMEYDA, J. & LEVANTINE, A. (1972). Drug reactions XIX.Adverse cutaneous reactions to the penicillins-ampicillinrashes. British Journal of Dermatology, 87, 293.

ARNDT, K.A. & JICK, H. (1976). Rates ofcutaneous reactionsto drugs. A report from the Boston Collaborative DrugSurveillance Program. Journal of the American MedicalAssociation, 235, 913.

BANDMANN, H.J., CALNAN, C.J., CRONIN, E., FREGERT, S.,HJORTH, N., MAGNUSSON, B., MAIBACH, H., MALTEN,K., MENEGHINI, C.L., PIRILA, V. & WILKINSON, D.S.(1972). Dermatitis from applied medicaments. Archives ofDermatology, 106, 335.

BATCHELOR, J.R., WELSH, K.I., MANSILLA TINOCO, R.,DOLLERY, C.T., HUGHES, G.R., BERNSTEIN, R., RYAN,P., NAISH, P.F., ABER, G.M., BING, R.F. & RUSSELL, G.I.(1980). Hydralazine-induced systemic lupus eryth-ematosus: influence ofHLA-DB and sex on susceptibility.Lancet, i, 1107.

BIANCHINE, J.R., MACARAEG, P.V.J., LASAGNA, L., AZAR-NOFF, D.L., BRUNK, F., HVIDBERG, E.F. & OWEN, J.A.(1968). Drugs as etiologic factors in the Stevens-Johnsonsyndrome. American Journal of Medicine, 44, 390.

BLACK, A.J. & SOMERS, K. (1984). Drug related illnessresulting in hospital admission. Journal of the RoyalCollege of Physicians of London, 18, 40.

BLUM, R.H., CARTER, S.K. & AGRE, K. (1973). A clinicalreview of bleomycin- a new antineoplastic agent. Cancer,31, 903.

BOSTON COLLABORATIVE DRUG SURVEILLANCEPROGRAM. (1972). Excess of ampicillin rashes associatedwith allopurinol or hyperuricaemia. New England Journalof Medicine, 286, 505.

BROCQ, L. (1894). Eruption erythemato-pigmentee fixe due al'antipyrine. Annales de Dermatologie et de Syphiligraphie(Paris) 5, (Series 3), 308.

CARROLL, O.M., BRYAN, P.A. & ROBINSON, R.J. (1966).Stevens-Johnson syndrome associated with long-actingsulphonamides. Journal of the American Medical Associa-tion, 195, 691.

CHANDRA, R.K., JOGLEKAR, S.A & TOMAS, E. (1980).Pencillin allergy:antipenicillin IgE antibodies and im-mediate hypersensitivity skin reactions employing majorand minor determinants of penicillin. Archives ofDiseasein Childhood, 55, 857.

FALKSON, G. & SCHULZ, E.J. (1981). Changes in hairpigmentation associated with cancer chemotherapy. Can-cer Treatment Reports, 65, 529.

FISHER, A.A. (1976). Allergic dermatitis medicamentosa. Thesystemic contact-type variety. Cutis, 18, 637.

FRASER, T.N. (1945). Gold treatment in rheumatoid arth-ritis. Annals of the Rheumatic Diseases, 4, 71.

GELL, P.G.H. &COOMBS, R.R.A. (1963). In ClinicalAspects ofImmunology, 1st Edition. Blackwell Scientific Publica-tions: Oxford.

GRIFFIN, J.P. (1983). Drug induced allergic and hypersen-sitivity reactions. Practitioner, 227, 1283.

HARBER, L.C., LASHINSKY, A.M. & BAER, R.L. (1959). Skinmanifestations of photosensitivity due to chlorothiazideand hydrochlorothiazide Journal of Investigative Der-matology, 33, 83.

HARRISON, B.M. &WOOD, C.B.S. (1972). Cyclophosphamideand pigmentation. British Medical Journal, 2, 352.

HARROLD, B.P. (1966). Syndrome resembling Addison'sdisease following prolonged treatment with busulfantherapy, British Medical Journal, 1, 463.

HORIO, T. (1984). Photoallergic reaction: classification andpathogenesis. International Journal of Dermatology, 23,376.

HUFF, J.C., WESTON, W.L. & TONNESEN, M.G. (1983).Erythema multiforme: a critical review of characteristics,diagnostic criteria and causes. Journal of the AmericanAcademy of Dermatology, 8, 763.

KAUPPINEN, K. & STUBB, S. (1984). Drug eruptions:causative agents and clinical types. Acta DermatologicaVenereologica (Stockholm), 64, 320.

KENNEDY, B.J., SMITH, L.R. & GOLTZ, R.W. (1975). Skinchanges secondary to hydroxyurea therapy. Archives ofDermatology, 111, 183.

KLINEFELTER, H.F. (1975). Reinstitution of gold therapy inrheumatoid arthritis after mucocutaneous reactions. Jour-nal of Rheumatology, 2, 21.

KORKIJ, W. & SOLTANI, K. (1984). Fixed drug eruption: abrief review. Archives of Dermatology, 120, 520.

KUOKKANEN, L. (1972). Drug eruptions: A series of 464cases in the Department of Dermatology, University ofTurku, Finland, during 1966-1970. Acta Allergologica,27, 407.

LEVANTINE, A. & ALMEYDA, J. (1973). Drug inducedchanges in pigmentation. British Journal of Dermatology,89, 105.

LEYDEN, J.J. & KLIGMAN, A.M. (1979). Contact dermatitisto neomycin sulphate. Journal of the American MedicalAssociation, 242, 1276.

McGRAE, J.D. & ZELICKSON, A.S. (1980). Skin pigmentationsecondary to minocycline therapy, Archives of Der-matology, 116, 1262.

PAPADOUPOULOS, S. & HARDEN, McG. R. (1966). Hair lossin patients treated with carbimazole. British MedicalJournal, 2, 1502.

PENNEYS, N.S., ACKERMAN, A.B. & GOTTLIEB, N.L. (1974).Gold dermatitis. A clinical and histopathological study.Archives of Dermatology, 109, 372.

PETZ, L.D. & FUDENBERG, G.H.H. (1966). Coombs positivehemolytic anaemia caused by penicillin administration.New England Journal of Medicine, 274, 171.

RESNIK, S. (1967). Melasma induced by oral contraceptivedrugs Journal of the American Medical Association, 199,601.

ROS, A.M., JUHLIN, L. & MICHAELSSON, G. (1976). Afollow-up study of patients with recurrent urticaria andhypersensitivity to aspirin, benzoates and azo dyes. BritishJournal of Dermatology, 95, 19.

SATANOVE, A. (1965). Pigmentation due to phenothiazinesin high and prolonged dosage. Journal of the AmericanMedical Association, 191, 263.

SEMON, H.C. (1926). Drug eruptions. Postgraduate MedicalJournal, 1, 144.

SHAH, P.C., RAO, K.R.P. & PATEL, A.R. (1978). Cyclophos-phamide induced nail pigmentation. British Journal ofDermatology, 98, 675.

copyright. on A

pril 1, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.61.720.925 on 1 O

ctober 1985. Dow

nloaded from

DRUG ERUPTIONS 933

SHAPIRO, S., SISKIND, D.V., SLONE, D., LEWIS, G.P. & JICK,H. (1969). Drug rash with ampicillin and other penicillins.Lancet, ii, 969.

SIMONS, J.C. & MORALES, A. (1980). Minocycline andgeneralized cutaneous pigmentation. Journal of theAmerican Academy of Dermatology, 3, 244.

SPEER, F., DENISON, T.R. & BAPTIST, J.E. (1981). Aspirinallergy. Annals of Allergy, 46, 123.

STUBB, S. (1976). Blood leucocytes with special reference to

basophils and eosinophils during provocation tests in fixederuption and drug exanthema. Acta Dermatologica Ven-ereologica, 56, 1.

WELSH, A.L. (1961). The fixed eruption. Charles Thomas:Springfield, Illinois.

WYATT, E., GREAVES, M.W. & SONDERGAARD, J. (1972).Fixed drug eruption (phenolphthalein): evidence for ablood borne mediator. Archives of Dermatology, 8, 763.

copyright. on A

pril 1, 2020 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.61.720.925 on 1 O

ctober 1985. Dow

nloaded from