Tests To Confirm Drug Allergy

The classic allergic response occurs in patients who have been previously exposed to the drug. It is mediated by drugspecific IgE antibodies on the surfaces of mast cells or basophils, interacting with the drug to induce an array of chemical mediators. The response appears typically 30 minutes after drug administration – peaks in a few more minutes and subsides over a few hours.

Anaphylactoid’ reactions have similar clinical features to anaphylaxis, but they are not IgE-mediated. The time relationships may differ and prior drug exposure may be absent.

Other adverse immune responses include

  • cytotoxic reactions mediated by complement activation e.g. haemolytic anaemia
  • reactions mediated by immune complexes and complement activation e.g. penicillin-induced serum sickness
  • delayed hypersensitivity reactions triggered not by antibody but by T-lymphocytes, specifically sensitised to recognise a particular antigen.

Generalised reactions

1. Anaphylaxis

This is the most feared response with its potential for sudden death. Its features include urticaria, angioedema, asthma, laryngeal oedema, hypotension, cardiac arrhythmias and gastrointestinal upset. The entire reaction is abrupt, explosive, but brief.

2. Serum sickness

This reaction comprises fever, urticaria, joint involvement and lymphadenopathy. It typically develops 1-3 weeks after drug therapy is commenced and may continue for weeks after the drug is stopped.

3. Drug induced autoimmunity

Some drugs are capable of inducing autoantibodies which may not result in disease but can cause lupus erythematosus, hepatitis, etc.

Single organ involvement

Numerous drugs induce immunological reactions in particular organs e.g. penicillin (acute urticaria), neomycin (contact dermatitis), methyldopa (haemolytic anaemia) and penicillamine (nephrotic syndrome).

Evaluation of patients with suspected drug allergy


This is the cardinal diagnostic tool. Key points include

  • suspicion that an unexplained clinical event may be caused by a drug
  • complete and accurate documentation of all drugs, including nonprescription ones, taken over the previous month.
  • temporal relationships between drug administration and the onset of symptoms or signs.
  • correlation of the clinical manifestations with known reactions induced by a particular drug.
  • previous tolerance of the drug (sensitisation).
  • prior history of a similar reaction to the same or crossreacting drug.

A few points should be stressed regarding the temporal relationships. It is virtually impossible to react allergically on the first exposure to a drug never previously taken unless the patient has had a crossreacting drug or unknown subclinical exposure to the drug or related compounds such as meat tenderisers (chymopapain) or quaternary ammonium compounds in cosmetics (muscle relaxants). Reactions rarely occur within the first 7 days of treatment; rather, within 2-4 weeks. Further, if a drug has been taken continuously for a year or more, it is unlikely to cause a reaction. However, the classic IgE-mediated drug allergy typically appears after the first dose of a new course.

Patients who report an allergic response may have had an adverse reaction to the drug. Adverse reactions to drugs are common and drug allergy represents only about 10% of them. Certain clinical features help to distinguish allergy from other reactions:

  • only a small number of patients react in this way
  • minute doses may trigger a response
  • the signs and symptoms are similar to those of allergic responses triggered by other substances (foods, animals, insect stings)
  • they develop soon after the drug is started and resolve soon after it is withdrawn
  • they are quite different from the usual pharmacological actions of the drug
  • specific tests may define the allergic response

Principles of testing

Testing is primarily restricted to the detection of drugspecific IgE antibodies.

1. Skin testing

Finding a suitable test reagent is a problem as most drugs have a low molecular weight and are not immunogenic. Therefore, they must form a macromolecular complex, presumably in combination with a host protein. Furthermore, drug metabolites may cause the reaction. In short, for most drugs, the antigenic determinants are unknown.

General approach

Immediate type skin tests are the most rapid, convenient and reproducible method of detecting drug-specific IgE antibodies. They are most accurate in evaluating reactions induced by drugs which are proteins (insulin, chymopapain, foreign antisera). Although skin tests for lower molecular weight compounds such as sulphonamides are of questionable value, exceptions include the major antigenic determinant of penicillin (benzylpenicilloyl (BPO) compound) and some drugs used in general anaesthesia.


Prick testing should be done initially. The drug is appropriately diluted (e.g. 1:1000) with normal saline, a drop placed on the ventral surface of the forearm and the skin pricked through the drop. A positive response, read at 20 minutes, is a wheal 2 mm or greater than that of a normal saline control, and a surrounding flare. If negative, intradermal tests are done with serial dilutions. The forearm is injected intradermally with 0.05 mL of the lowest concentration (1:10 000 to 1:100 depending on the drug and the patient’s history). A positive response at 20 minutes is a wheal and flare, with the wheal being larger than that raised by the initial injection and the normal saline control. If negative, the next (higher, usually 10-fold) concentration is injected.


  • The tests are usually without risk, but anaphylaxis has occurred.
  • They should be done by personnel familiar with the theory and practice of the procedure.
  • The tester must be aware of appropriate starting dilutions as well as the concentrations likely to induce irritant (nonallergic) responses.
  • False negative results may occur from
    • inappropriate test reagent
    • too short a time interval between the adverse reaction and test — at least two weeks is needed because a severe anaphylactic event temporarily depletes the circulating IgE antibodies and also the chemical mediators from the mast cells.
  • The patient may exhibit an anxiety or vasovagal response.
  • Appropriate means of reversing a severe reaction must be readily available.
  • Informed consent and accurate documentation are essential.

In vivo testing

Test doses can be used when there is an unconvincing history of drug allergy, but an IgE reaction is difficult to exclude. The initial dose is much lower, e.g. 1:100, than the usual therapeutic dose. Subsequent challenges are of higher concentrations than those used in the therapeutic technique of desensitisation. Frequency of dosing is determined by the type of the alleged drug reaction, at 30 minute intervals for IgE-mediated reactions and 24-48 hours for a delayed response such as dermatitis.

In vitro testing

This has the advantage that adverse reactions to testing can be avoided. The most widely used is the radioallergosorbent test (RAST) which measures circulating drugspecific IgE antibodies. It is generally less specific and less sensitive than skin testing, thus limiting its clinical usefulness.

Special considerations


Allergy to penicillin is the best studied drug reaction. Anaphylaxis most commonly occurs between the ages of 20 and 49 years, but children and the elderly are not exempt. It is more likely to occur when the drug is given parenterally rather than orally. With the passage of time, 85% of patients `lose’ their hypersensitivity.1,2 Thus, allergy to penicillin (and presumably to other drugs) is not necessarily lifelong.

Penicillinallergic patients have 10 times the risk of other people of reacting adversely to other antibiotics.

Testing: BPO accounts for 95% of the metabolites of penicillin. Other metabolites, `minor determinants’, are also capable of inducing anaphylactic responses. In Australia, the skin test reagent BPO-polylysine is only available on the Special Access Scheme and commercial preparations of minor antigenic determinants are not available at all. A useful additional test reagent is freshly prepared benzylpenicillin (BP) (6000 U/mL) which may also detect patients allergic to minor determinants. A RAST for BPO is available, but a result is usually required quickly and a RAST result takes a few days. Testing with BPO alone fails to detect a significant number of patients at risk, but the addition of BP will improve the detection rate to 80%.3 This rate can be increased by additional testing with semisynthetic penicillins and particularly with preparations of their minor determinants.2,4 There is still a significantly high false negative group, so subsequent treatment with penicillin in testnegative patients must be carried out with careful monitoring. To date, no skin testnegative patient is known to have developed a severe reaction. That testing is useful has been indicated by studies of patients undergoing testing and subsequent treatment. In patients with positive tests, there is a high incidence of subsequent reactivity; those with negative tests have a very low incidence of reactions.


Maculopapular rashes induced by ampicillin and amoxycillin are seen commonly in patients with infectious mononucleosis, cytomegalovirus and chronic lymphatic leukaemia. They typically develop 48 hours or more after drug administration and persist for a week or so. Unless the rash is clearly urticarial, it is unlikely to be IgE-mediated.


Cephalosporins and penicillin share the beta-lactam ring in their basic structure. However, clinical crossreactivity occurs in approximately 10% of patients. For patients allergic to penicillin, their chances of reacting to the cephalosporins appear no greater than those of reacting to other unrelated antibiotics.5

Sulphonamides and other antibiotics

Despite the severity and frequency of allergic reactions, especially the trimethoprim/sulfamethoxazole combinations, there are no specific diagnostic reagents which have been developed for commercial use to identify patients at risk of anaphylaxis. If desired, testing could be done with individual antibiotics as outlined above. There is a small potential for the Stevens-Johnson syndrome when testing for sulphonamides.

Multiple drug allergy

Many patients are being recognised as having allergic reactions to a number of antigenicallyunrelated drugs, antibiotics in particular. Indeed, one study of 312 penicillinallergic patients found that 21% of them experienced reactions to non-beta-lactam antibiotics.6 Clinical management is difficult, but should follow these basic principles:

  • preference should be given to antibiotics known to be tolerated
  • patients who have reacted to one type of penicillin will not necessarily react to other types
  • of the new classes of beta-lactams, the carbapenems (imipenem) crossreact with the minor determinants of penicillin, whereas the monobactams (astreonam) do not
  • third generation’ cephalosporins have a lower incidence of allergic reactions than the `first and second generation’ drugs

Drugs used in general anaesthesia

Intradermal testing is highly accurate in identifying patients with a high risk of anaphylaxis to induction drugs (e.g. thiopentone) and to muscle relaxants (e.g. alcuronium and suxamethonium). Crossreactivity between muscle relaxants is often present.

Local anaesthetics

These drugs are unlikely to cause IgE-mediated responses. Skin testing is probably warranted, with concentrations increasing to a typical therapeutic dose.

Radiographic contrast media (RCM)

Anaphylactoid reactions develop in 2-3% of patients. They may be more common with the use of ionic media. Most are mild (e.g. urticaria), but fatalities have been recorded. The aetiology is obscure, but IgE antibodies have never been demonstrated. Some features are

  • they often occur on the first exposure
  • specific testing is useless
  • patients with seafood allergy are at no greater risk
  • pretreatment with antihistamines and steroids will reduce the risk of serious reactions.


Bovine and pork insulins differ from human insulin in three and one amino acid sequences respectively. Adverse reactions to human (recombinant DNA) insulin are less likely to develop, but there have been isolated reports of patients allergic to animal insulins who have had cutaneous reactivity to the human insulin preparation.

There are two adverse immune responses: allergy and resistance. Large local swellings due to allergy are common and usually ease with continued treatment aided by antihistamines. Systemic reactions are rare and may require desensitisation. RAST and skin tests may be done. Insulin resistance is also rare and results from antiinsulin IgG antibodies which neutralise exogenous insulin.


Anaphylaxis has occurred in 1% of patients treated by chemonucleolysis with chymopapain. As it often develops on first exposure, sensitisation may have resulted from meat tenderisers. Skin testing immediately before treatment is essential. The testing is done in sequence: prick tests of 1 mg/mL, 10 mg/mL and an intradermal test of 0.2 mL (100 micrograms/mL). If each is negative, the risk of anaphylaxis is minimal.


Allergic reactions have been reported in up to 17% of patients treated with streptokinase. An intradermal test with 0.1 mL of 1000 IU/mL, if positive at 15 minutes, should detect those at risk of anaphylaxis.

Vaccines in eggsensitive patients

Small amounts of egg protein may be found in vaccines for influenza, measlesmumpsrubella, and yellow fever. Although allergic reactions in eggsensitive patients given these vaccines are rare, when in doubt, preliminary testing, firstly by prick and then intradermally, can be done.

Aspirin and other nonsteroidal antiinflammatory drugs (NSAIDs)

Anaphylactoid reactions to these drugs are not mediated through an immune mechanism. No objective testing is feasible and test dosing carries a high risk of severe reactions.

General management principles

General management principles for patients with a history of allergy to a specific drug are

  • obtain full clinical details of the reaction
  • if allergy is suspected, use an alternative, non-cross-reacting drug (available in most situations)
  • if none appropriate, refer to someone competent in testing for drug allergy
  • testing for suspected penicillin allergy should ideally employ different kinds of penicillins and cephalosporins as well
  • if tests are negative, give a test dose first where full resuscitation facilities are available
  • treat minor allergic responses (especially rashes) symptomatically if continuation of the drug is considered essential; however, be vigilant for Stevens-Johnson syndrome which is potentially fatal
  • provide adequate follow-up of patients
  • if the drug is required again, objective testing may be repeated
  • know how to treat anaphylaxis
  • if tests are positive, avoid the drug; however, desensitisation according to accepted protocols may be used as a last resort.