Table of Contents   This Chapter   All Chapters

INTRODUCTION TO DRUG ERUPTIONS

Adverse drug reactions are common, and a skin eruption is a feature in about 30% of cases. About 2% of hospital inpatients may develop a drug rash (also termed toxicoderma in some countries). Unfortunately, most cutaneous drug eruptions are of a non-specific morphology (Fig. 18.1); additionally, most of those that occur in a hospital setting are in individuals who are ill and taking many drugs. Diagnosis can therefore be difficult, especially as few drug eruptions are amenable to any form of testing other than the response to drug withdrawal (and sometimes to rechallenge). However, the diagnosis and identification of the likely cause can often be made with reasonable confidence based on a few basic principles.

Figure 18.1 A typical non-specific acute maculopapular drug eruption. In this case, the cause was cimetidine, but numerous drugs cause the same reaction pattern. (Courtesy of Dr. G. Dawn.)

Mechanisms of drug eruption

There are several mechanisms of drug eruption, some of which correlate with particular patterns of reaction. Some of these are listed with examples in Table 18.1. Recent research has identified the fact that some clinical patterns may relate to specific patterns of T-cell or cytokine activation;for example, a high proportion of CD8+ cells in the epidermis tend to correlate with development of blistering. This has allowed the type IV immunologic reactions to be divided into those that tend to activate monocytes (type IVa), eosinophils (IVb), or neutrophils (IVd); type IVc, involving cytotoxic functions, takes part in all type IV reactions.

    Host factors may also influence the risk or the severity of drug eruptions. These include the following.

Figure 18.2  Semiconfluent rash on the arms due to phenytoin, associated with fever and malaise. Anticonvulsant reactions are commonly relatively severe.

Figure 18.3  Vasculitic-looking rash on the leg of a patient prescribed amoxicillin for a viral infection. Many such patients do not have rash related to other penicillins, and they may subsequently tolerate ampicillin derivatives without any adverse effect.

Figure 18.4  A similar eruption to that in Figure 18.3, demonstrating accentuation at a pressure area under the elastic of a sock.

Infectious mononucleosis—ampicillin and derivatives are a common cause of drug rash (Figs 18.3 and 18.4). However, if given to patients with infectious mononucleosis, the likelihood of rash increases to over 90%. A similar but less dramatic increase applies if ampicillin is used in the context of other viral infections (especially cytomegalovirus), compared with use of the same drug for bacterial infection.

HIV infection—there is an increased risk of severe adverse reactions to sulfonamides and some other antibiotics in patients with HIV infection. A deficiency of glutathione allows increased levels of a toxic hydroxylamine derivative of sulfamethoxazole, a drug that patients with AIDS may take on a prolonged basis for prophylaxis against pneumocystis pneumonia. Drug-induced lichenoid photoreactions in patients with HIV infection may be more common in black patients.

Hepatitis B infection may predispose to the dapsone hypersensitivity syndrome.

Mastocytosis—several drugs non-specifically lead to mast cell degranulation and may trigger systemic symptoms in patients with significant mastocytosis (see list in Ch.11).

Collagen vascular disease—a papular eruption closely linked to use of methotrexate has been reported. Griseofulvin-triggered lupus erythematosus has been suggested to be more frequent in patients who have anti-SSA or anti-SSB antibodies.

Atopy—reactions to gold may possibly be more common.

Approach to making the diagnosis

Most drug eruptions have a maculopapular or ‘urticated' morphology, and start within a few days up to about 2 weeks after taking the relevant drug. Thus most drug eruptions are relatively simple to diagnose (Figs 18.5 and 18.6). The more acute-onset urticarial eruptions are even simpler to suspect. However, there are several confounding problems.

Figure 18.5  Some adverse drug reactions are predictable and dose-related, such as cheilitis due to systemic retinoids for acne.

Figure 18.6  Mild eczematous changes may also occur due to retinoids, usually on the dorsum of the hands and forearms, but are less well recognized than the cheilitis shown in Figure 18.5.

Figure 18.7  Inverse pattern of drug eruption. Rarely does a drug eruption affect the body folds preferentially. Here the bathing trunk area and axilla are affected.

     The basic questions listed in the introduction should therefore be applied. In patients who have started several drugs in close temporal proximity, and who have a non-specific morphology of eruption, it is helpful to have an idea of the relative risk of rash due to the various implicated drugs. It is essential to document timing of the eruption in relation to any suspect drugs, and to include non-prescribed medications in the inquiry. For example, an antibiotic or a non-steroidal antiinflammatory drug (NSAID) started a week before onset of a maculopapular eruption is a more likely cause than a beta-blocker started 2 months previously, partly because the former drugs are a more common cause of rash, and partly because the timing fits better with the pattern of eruption.

    The diagnosis is often supported by the response to withdrawal of the suspect drug (Fig.18.10), most eruptions improving over a few days or weeks (although some may initially worsen, usually for no more than a few days). Some patterns, such as quinine-induced photosensitivity, may take months to settle. For most drug reactions, there is no simple diagnostic test. Patch testing can be useful for identifying some drug reactions (such as to carbamazepine) but is considered unreliable for others (such as to local anesthetics) and is simply unhelpful in most instances. Prick tests and intradermal tests are not routinely helpful, and have a potential risk, but can be performed for penicilloyl residues if there are important reasons requiring use of a penicillin rather than of an alternative antibiotic. Lymphocyte transformation tests, basophil activation tests, and other laboratory tests are not routinely available. Rechallenge is generally not performed unless the drug is important to the patient (e.g. to treat a condition where there are few alternatives), and the expected reaction is not likely to be severe. Challenge testing can be a very time-consuming process but may be required occasionally (e.g. in some cases of local anesthetic reaction).

Figure 18.8  A classic pattern but often not recognized. This eruption is due to co-danthramer, a laxative that is metabolized to dithranol (anthralin) in the bowel. It therefore causes irritation (red) and staining (orange-brown) of the skin of the buttocks and posterior thighs, typically in immobile patients.

Figure 18.9  Occasionally, replacing a drug that has caused a skin reaction may be followed by a reaction to the second drug. This example shows a fixed drug eruption due to carbamazepine, with a background exanthem pattern due to phenytoin.

Figure 18.10  Withdrawal of a suspect drug is usually followed by resolution of the eruption within a few days or weeks. A sunburn-like desquamation is common during this phase, and may be generalized (a), although it is most obvious on palms and soles, where the skin is thicker (b). (Panel a from Lawrence CM, Cox NH . Physical Signs in Dermatology, 2nd edn. London : Mosby, 2002.)

Assessment and treatment overview

The most common patterns of drug eruption will respond to drug withdrawal and administration of a topical corticosteroid and oral antihistamines as required on a symptomatic basis. Antihistamines are most useful for urticarial reactions, whereas short-term potent topical corticosteroids are generally the most effective way to treat maculopapular eruptions.

    Some eruptions require other specific measures, for example anaphylaxis or severe urticaria (see Ch. 9), toxic epidermal necrolysis (treated along the lines used for thermal burns, with particular attention to fluid balance, secondary infection, and skin fragility), or erythroderma (which may be complicated by impaired temperature control, high-output cardiac failure, and malabsorption).

    Desensitization may be used for some drug eruptions, for example urticarial reactions to sulfasalazine or penicillins. Desensitization to penicillin is not routine, and is generally performed only in situations where there is no easy alternative to treat a specific infection; the process requires careful intensive care unit monitoring for safety.

    In many situations, it is important to consider whether an alternative drug can be substituted. For example, in an infection, an antibiotic can usually be chosen for which cross-reactions would be unlikely. In the case of NSAIDs, most selective cyclooxygenase (COX)-2 inhibitors have been shown to be safe in the vast majority of patients with urticarial or with pseudoallergic reactions to traditional NSAIDs.

    Generally, when considering therapy, it is also important to consider severity issues. These include the following.

White/Cox: Diseases of the Skin, 2ed.(c) 2006, Elsevier Inc. All rights reserved.