With the steadily increasing production of potent drugs which affect all aspects of bodily function and metabolism the side effects of medical treatment have become a serious problem. Up to 20 per cent of patients admitted to medical wards have been reported to suffer a reaction to drugs used in treatment and 5 per cent of medical patients are actually admitted because of an adverse drug reaction. At least in such cases the drugs taken are usually known, but when one takes into account the cross reactions which can occur between dyes and preservatives in food and drugs which have either been ingested or applied to the skin and one also considers the ability of such chemicals to sensitise the skin to light, the possible sources of harm to the patient seem almost limitless.
The most important step in diagnosis is an accurate history of any medicament whether being taken orally, by injection, inunction or pessary. It is useless merely asking the patient if they have been taking any drugs as this conjures up visions of opium and hashish. Leading questions must be asked about the more common drugs of addiction such as as aspirin, aperients and even saccharine. The dates on which therapy was started and stopped should always be recorded, but the fact that administration stopped before the appearance of a reaction does not rule the drug out, as some drugs such as gold salts are very slowly excreted and others such as the penicillins give rise to delayed hypersensitivity reactions.
No medicament which has a pharmacological action can be dismissed as being free from side-effects and the fact that the patient has taken a drug continuously foryears, such as phenobarbitone for epilepsy, does not make development of sensitisation impossible.
Allergic reactions are the commonest cause of drug eruptions. Only those drugs which incorporate proteins or polysaccharides act as complete antigens; most drugs are haptens and they, or their metabolites or contaminants become bound to protein to form an antigen leading to antibody formation. When the drug has been administered parenterally the resultant eruption is an allergic reaction in the dermis giving rise to urticaria or exanthema. If the drug or a chemically related substance is, or has in the past, been applied to the skin the hapten protein complex takes place in the dermis and the reaction is eczematous.
Hypersensitivity reactions have been classified into four types but many such reactions do not clearly fit into any one type.
Type I is an anaphylactic reaction in which the antibody involved is IgE. The IgE antibodies have an affinity for cell membranes and can set in motion an enzyme reaction which leads to release of histamine and kinins from mast cells and basophils. Massive release of histamine may cause an anaphylactic reaction with bronchiolar constriction, oedema of the larynx and glottis, vasodilation and shock. The accompanying skin eruption is urticarial. Such a reaction may follow injection of penicillin or serum in a sensitised subject. For this reason even intradermal skin testing is hazardous as a test of penicillin sensitivity and if the patient is said to be sensitive it is best to use other antibiotics unless penicillin is the only possible life-saving drug.
Type II reactions. Here the antigen attaches itself to cell surfaces, the antibodies involved are either IgG or IgM, the antigen—antibody complex takes up complement and the cell is destroyed. The mechanism is best illustrated by experiments performed by Ackroyd with a hypnotic Sedormid which is no longer in use. Sedormid acts as a hapten combining with platelets; in the presence of complement the Sedormid- platelet combination reacts with antibody causing agglutination and lysis of the platelets. The resultant thrombocytopenia causes a purpuric eruption. However Ackroyd also showed that the application of Sedormid to the skin as a patch test gave rise to local purpura in the absence of thrombocytopenia suggesting that the drug attached itself to capillary endothelial cells and that these were damaged in the resultant allergic reaction. Purpura as a drug eruption may therefore be due to thrombocytopenia or capillary damage or both.
Type III reactions. Here the drug in the form of an antigen reacts with antibodies of the IgG or IgM classes to form antigen-antibody complexes in tissue spaces. These ‘immune complexes’ take up complement and are deposited in blood vessel walls or basement membranes causing inflammation. The most clear out example of this reaction is serum-sickness in which antibodies build up during a period of about 10 days and the symptoms and signs of pyrexia, urticaria, lymphadenopathy and arthralgia then appear. This is also the classical allergic reaction to penicillin but related drugs may give rise to exanthema rather than urticaria.
Type IV reactions (delayed type). In this type humoral antibodies are not found and the mechanism is clearest in epidermal reactions. Here the drug has been applied to the skin, combining with proteins to form an antibody. After a delay of about 48 hours sensitised T. lymphocytes and macrophages invade the affected area of skin. Activation of the lymphocytes transforms them into large blast cells undergoing mitosis.
At the same time a number of soluble factors are released which act as mediators of the hypersensitivity reaction which in this case is eczematous. Once this pattern of sensitivity is established it remains true to type whatever the route of the antigen, thus if neomycin has in the past produced an area of contact dermatitis the subsequent injection of streptomycin, which is closely related, will again cause the skin to break out in an eczematous eruption, initially at the original site of the neomycin dermatitis.
To what extent the exanthematous eruptions are due to Type III or Type IV reactions is not yet worked out, it seems likely that more than one type of reaction at the same time could explain the bizarre reactions which drugs often cause.
It is impossible to list all the described reactions to commonly used drugs but it is useful to consider the main clinical manifestations and to have some idea of the drugs most likely to be incriminated in each group.
Exanthematous eruptions. These are the commonest skin reactions and produce an erythema or widespread morbilliform eruption. In severe cases, especially if the drug is slowly excreted because of impaired liver or kidney function, the rash may be accompanied by fever, lymphadenopathy and a sore mouth. Barbiturates, especially the slowly excreted phenobarbitone produce this rash and also tend to produce a confluent erythema on the palms, soles and flexures. Sulphonamides also produce a morbilliform eruption and although less used by themselves are used in combination in drugs such as Cotrimoxazole. Chlorpropamide, phenylbutazone, glutethimide (Doriden) and amitriptyline are other commonly used drugs likely to produce such eruptions.
Ampicillin produces a maculo-papular eruption, often bluish in colour from an element of purpura, which appears up to 2 to 4 weeks after the drug was first administered. Nearly all patients with glandular fever develop such a reaction if given ampicillin as do many with lymphatic leukaemia.
Urticaria. Penicillin has already been mentioned as the cause of a serum sickness type of reaction and it is important to appreciate that it follows at least 10 days after administration of the drug except in the anaphylactic reactions when it is immediate. Sometimes in the delayed reactions the interval can be up to 4 weeks. Contamination of milk with penicillin can cause further outbreaks as may rarely the moulds of cheese. Aspirin is a cause of allergic urticaria and may also act as a histamine release agent so that it provokes urticarial outbreaks in those whose urticaria is due to other causes. Sufferers from this disease must therefore avoid all aspirin containing drugs. Those sensitive to aspirin may also react to azo dyes and benzoic acid preservatives found in food.
Vesico-bullous reactions. Bulla formation is largely a matter of the degree of reaction and most of the exanthema can progress to blister formation but sulphonamides and barbiturates are the common offenders, the blistering often starting on the hands and feet. Bullous lesions may appear at the sides of pressure or trauma in drug-induced coma, particularly barbiturate coma but also with other drugs.
Purpura. Some of the commonly used drugs producing purpura are aspirin, phenylbutazone, thiazide diuretics, quinine and quinidine. However in many of the exanthematous drug eruptions there is some degree of purpura which often gives them a startlingly bluish colour which to the tutored eye suggests a drug rash.
Carbromal produces a characteristic purpura accompanied by a slightly scaly reticulate erythema. Meprobamate may also produce a similar reaction.
Fixed drug eruptions. These are so-called because on each exposure to the drug the lesions recur on the same site. They arise within an hour or two of taking the drug as discs of erythema, either single or multiple and in severe reactions may produce a solitary bulla at the centre of the disc. The lesion subsides quickly when the drug is stopped leaving a disc of pigmentation which may persist until the next attack. Lesions can occur on the buccal mucosa producing raw red patches or in the urethra causing urethritis. Phenolphthalein is the commonest cause, being present in many purgatives, especially ’the ‘chocolate laxatives’ and agar compounds. More obscurely it may be present as colouring matter in toothpaste and icing sugar. Phenacetin, phenazone, barbiturates, sulphonamides, tetracyclines and chlordiazepoxide are other commonly used drugs causing fixed eruptions.
Acne. Papules and pustules on the face may be produced by the halogens; iodides and bromides are not often used alone nowadays but may be part of the structure of a more complex chemical such as radio-opaque materials and still cause acne. Isoniazid and ethionamide also cause acne.
Lichenoid and psoriasiform eruptions. Eruptions morphologically indistinguishable from lichen planus are caused by a number of chemically unrelated drugs. Often the rash is florid and rather scaly and the mouth lesions less frequently present. With slowly excreted drugs such as gold salts, organic arsenicals and mepacrine resolution of the rash may take months. Chloroquine, amiphenazole (Daptazole) and para-aminosalicyclic acid also cause such an eruption, as less commonly do chlorothiazide and phenothiazides.
Practolol may cause eczematous eruptions or exfoliative dermatitis but a more characteristic eruption is psoriasiform and affecting especially the palms, soles and bony promontaries. The drug also aggravates existing psoriasis and similar aggravation of psoriasis is a common effect of chloroquine therapy.
Light sensitivity. Some drugs in sufficiently high dose and with sufficient exposure to ultra-violet light cause tissue damage in the light exposed areas. The resultant phototoxic reaction causes what looks like acute sunburn. There is a high incidence of this reaction during administration of demethylchlortetracycline but it can occur with other tetracyclines, chlorpromazines, chlorothiazide and sulphonamides.
Drug induced allergic reactions to ultra-violet light are delayed in onset and like other allergies are not dose dependent. Sulphonamide drugs, the sulphonylureas, chlorpromazine and all the antibiotics can produce such sensitivity to light which may persist for months or years after the drug has been stopped.
Exfoliative dermatitis. If the early signs of a drug eruption are ignored and administration continued or if the drug is slowly excreted the patient may develop generalised exfoliative dermatitis; the whole skin becoming red, oedematous and shedding large amounts of scale. There is usually associated generalised lymphadenopathy. Gold salts, streptomycin and phenylbutazone are among the common causes.
Toxic epidermal necrolysis. It has already been mentioned that in children nearly all cases are due to staphylococcal infection of the skin. In adults the usual cause is drugs and the prognosis is grave. The patient becomes acutely ill and pyrexial, there is sudden onset of widespread erythema which rapidly develops profuse vesicles and bullae. The blisters coalesce in some areas to lift off the epidermis, in others the slightest pressure slides off sheets of skin like that of an over-ripe peach leving the appearance of severe extensive burns. One third of cases die in the first 3 to 4 days. The drugs most commonly incriminated are phenylbutazone and its relatives, sulphonamides, barbiturates and cytostatic drugs.
Eczematous reactions. Most eczematous reactions are due to previous sensitisation by external applications of a drug followed by systemic administration of the same drug or an analogue. However methyldopa can provoke a picture of seborrhoeic dermatitis or aggravate existing eczema, the eruption subsiding only when the drug is stopped.
Other reactions which may be drug induced can also be toxic reactions to infective and other factors. They include erythema multiforme, erythema nodosum, polyarteritis nodosa and lupus erythematous-like phenomena.
Diagnosis. When a single drug is being administered, stopping treatment is usually followed by rapid improvement though in the case of slowly excreted drugs such as phenobarbitone and mepacrine the eruption may continue to get worse. In these circumstances any febrile reaction usually subsides before the rash, indicating improvement. When multiple drugs are being given it is desirable to stop them all but in some patients this may be hazardous and the most likely offender has to be judged by the pattern of the reaction and stopped. Here again the subsidence of a drug fever may indicate success before the rash improves.
Test doses are usually safe in fixed drug eruptions provided that a fraction of the usual dose is given and sometimes the known causes of fixed drug eruptions have to be administered in series when the likely culprit is not detectable in the history. Test dosing in other types of drug reaction is hazardous and although penicillin analogues have been used to test for penicillin sensitivity such testing should only be used when that drug is vital and should be done under close observation with adrenaline and intravenous hydrocortisone ready for treating an anaphylactic reaction.
Patch tests are only useful when the reaction is epidermal, producing an eczematous rash and due to previous sensitisation by application of the drug to the skin. Most reactions are dermal and only in fixed drug eruptions when a solution is applied to one of the lesions for 48 hours is a positive reaction likely.
Scratch and intracutaneous tests not only expose the patient to some risk of anaphylaxis but are also unreliable as most drugs are haptens and not complete antigens, thus a negative reaction does not rule out sensitivity.
In vitro tests include the lymphocyte transformation test, the macrophage inhibition test and the basophil degranulation test but are too unreliable to be used in diagnosis.
Treatment. Urticarial eruptions provoked by such drugs as penicillin can usually be controlled by an adequate dose of antihistamines. The patient should rest in bed and if the reaction is severe an adult can be given triprolidine hydrochloride (Actidil) 5 mg three times daily and promethazine hydrochloride (Phenergan) mg 50 at night.
The dose can be reduced slowly once the urticaria is controlled but in some cases—particularly in atopic subjects—a small dose may have to be given for weeks or months. Only if antihistamines fail should prednisolone be given as an addition to and not a substitute for the antihistamines. Prednisolone 30 mg daily is usually adequate in an adult and the dose can be slowly reduced once the eruption is controlled, leaving the patient on antihistamines.
Other widespread eruptions should be treated with bed rest, a blood count performed to exclude bone marrow suppression and if itching is troublesome antihistamines may relieve this though they will have little effect on the course of the eruption. Locally shake lotions with 1 per cent phenol may give symptomatic relief; local corticosteroids are useless.
If the drug is slowly excreted or its breakdown and excretion impaired by poor liver or kidney function, especially in the elderly, or if administration of the drug has not been stopped promptly after the onset of a reaction, severe generalised erythema, ulceration of the mucosa and vaso-motor collapse may occur. The blood pressure, fluid balance and serum electrolytes must be measured regularly as the toxic effects of the drug combined With low blood pressure may produce severe kidney damage. When the subject or the drug make deterioration possible it is best to commence treatment with corticosteroids as soon as the diagnosis is made, giving prednisolone 30 or 40 mg daily or if there is vomiting or diarrhoea giving hydrocortisone intravenously.
Desensitisation is a dangerous and often unsuccessful process where dermal reactions are concerned and fortunately it is nearly always unnecessary as alternative drugs are available. If the original reaction was mild, desensitisation can be successful in the tetracyclines giving 10 mg orally as the initial and test dose and doubling this daily until therapeutic levels are reached.
Epidermal sensitivity to drugs as penicillin can mean that doctors or nurses in contact with patients taking the drug develop dermatitis and have difficulty in continuing their work. In such cases an attempt at desensitisation is justifiable starting with an infinitesimally small dose such as 10 units of penicillin.
Sensitivity to the offending drug is lifelong and it is important to impress upon the patient the need to warn any doctor prescribing for him of the allergy. Failure to do so may result not only in recurrence of the skin eruption but in anaphylactic shock and possibly sudden death.