ARLINGTON HEIGHTS, ILL--Adverse drug reactions are becoming increasingly common as more and more medications are approved for use each year. In response to what it terms "a major health problem in the United States," a consortium of three national allergy organizations has issued a practice parameter for the management of both drug hypersensitivity and similar nonimmunologically mediated drug reactions.

"This is a very timely issue because of the large number of medical errors. It is estimated that there are between 44,000 and 98,000 deaths among hospitalized patients in the United States [each year] due to medical errors," said I. Leonard Bernstein, MD, a clinical professor of medicine and environmental health at the University of Cincinnati College of Medicine in Ohio. "To a large extent, many of the medical errors that we see, unfortunately, are due to problems from adverse reactions to drugs." A recent meta-analysis concluded that at least 76,000 deaths each year are due to adverse drug reactions.

The practice parameter defines an adverse drug reaction as any nontherapeutic consequence of drug use--other than treatment failure, intentional or accidental poisoning, and drug abuse. Dr. Bernstein, one of the chief editors of the practice parameter, noted that about 25% of all adverse drug reactions can be attributed to drug hypersensitivity, pseudoallergy, or idiosyncrasy/intolerance.

Because some of these reactions can be life-threatening, early recognition is crucial. However, both accurate diagnosis and appropriate management may be delayed because confusion exists over definitions related to adverse drug reactions. "Doctors and patients alike sometimes cannot tell the difference between true allergic drug reactions and other adverse reactions," Dr. Bernstein said. For instance, "drug hypersensitivity is an immunologically mediated response to pharmacologic agents and/or excipients within a formulation. This needs to be distinguished from drug idiosyncrasy and drug intolerance, which have nonimmunologically mediated effects."

Likewise, pseudoallergic or anaphylactoid reactions are immediate systemic reactions that mimic anaphylaxis "but are nonimmune. These are caused by the effect of the drug itself upon the release of harmful mediators from human cells, such as mast cells and peripheral blood basophils," Dr. Bernstein explained.

THE PRACTICE PARAMETER

The new practice parameter was sponsored jointly by the American Academy of Allergy, Asthma, and Immunology; the American College of Allergy, Asthma, and Immunology; and the Joint Council of Allergy, Asthma, and Immunology. It contains two parts:

• An executive summary and management algorithm designed to serve as a concise summary for everyday clinical decision-making (see Figure 1).
• An evidence-based commentary that details the diagnosis, treatment, and prevention of drug hypersensitivity, pseudoallergic, and idiosyncratic/intolerance reactions.

Particular attention is given to immunologic reactions to ß-lactam antibiotics, because penicillin and its analogs are the most frequent cause of allergic drug reactions in this country. The practice parameter also outlines strategies that can be used for desensitization, and it reviews adverse drug reactions associated with human immunodeficiency virus (HIV) infection.

ANAPHYLACTIC REACTIONS

"Anaphylactic reactions are of great concern because they are potentially life-threatening," said Richard Nicklas, MD, another of the practice parameter's chief editors and a clinical professor of medicine at George Washington Medical Center in Washington, DC. "There are many different medications that have been documented as producing anaphylaxis, but most of them do so only rarely." Unfortunately, at least one very common drug can produce anaphylaxis, he added. "The classic example of a medication that can produce anaphylaxis is penicillin."

However, a patient who takes penicillin and develops a rash will not necessarily have an anaphylactic reaction when treated again with penicillin. If there is any question about penicillin allergy and the patient needs to be treated with penicillin, the best course of action is to perform skin testing, explained Dr. Nicklas. "The immediate hypersensitivity skin test for penicillin is very reliable," he noted. "The negative predictive value is very high. Only 1% to 3% of patients [with a negative test] will have a reaction if given penicillin, and reported cases in which this has occurred have been very mild."

Dr. Nicklas added, however, that "a fairly significant percentage of patients" who have a positive skin test will not react to penicillin. "Nevertheless, if you have a positive reaction, you should assume that the patient is allergic to penicillin."

In most cases, a patient who is allergic to penicillin can be treated with another antibiotic. However, cross-reactivity may complicate drug selection. For example, carbapenems are cross-reactive to penicillin; cross-reactivity may also occur with cephalosporins, although the risk is lower with third-generation agents than with first-generation drugs. Still, Dr. Nicklas warned that even though the "cross-reactivity between penicillin and cephalosporins is low, severe and life-threatening reactions can occur in penicillin-allergic patients who receive a cephalosporin. Thus, one has to be cautious about giving a cephalosporin to patients who are documented as being allergic to penicillin." When there is no acceptable substitute antibiotic for a patient who is allergic to penicillin, desensitization is indicated.

THINKING OUTSIDE THE BOX

Unlike penicillin allergy, which is mediated by immunoglobulin E antibodies, many immunologically mediated drug reactions cannot be easily classified. "In the past, we've relied on the Gell-Coombs classification system, which outlines four categories of hypersensitivity," noted Rebecca Gruchalla, MD, PhD, who is another of the chief editors of the practice parameter as well as an associate professor of internal medicine and director of the Division of Allergy and Immunology at the University of Texas Southwestern Medical Center in Dallas. "However, the practicing physician will find that most drug reactions do not fit nicely into one of these four categories," she explained.

The reason, in part, is "that there is probably more than one immune mechanism occurring," said Dr. Gruchalla. "We have to start thinking out of the box. That is, hopefully, what new research will help us do--sort out these reactions and classify them in better ways."

One approach that has been used recently is classification by the predominant tissue or organ system affected. For example, hypersensitivity reactions in the lungs may cause alveolar or interstitial pneumonitis, edema, granulomatous reactions, and fibrosis.

NONIMMUNOLOGIC REACTIONS

Although the practice parameter acknowledges that the mechanisms underlying many hypersensitivity reactions are still unclear, it continues to classify nonimmunologically mediated reactions, such as pseudoallergy, separately. Causes of pseudoallergic reactions commonly include opiates and radiocontrast media. The practice parameter reports both good news and bad news about pseudoallergic reactions.

"Unfortunately, there's not a whole lot we can do about them" once they occur, said Dr. Gruchalla. "Luckily, though, in this day and age we have better types of radiocontrast media. Because we have better agents, the frequency [of pseudoallergic reactions] is actually decreased." In addition, an effective premedication protocol has been developed that can prevent most reactions.

Physicians also need to be on the alert for idiosyncratic drug reactions, all three of the experts interviewed for this article suggested. The practice parameter defines idiosyncrasy as "an unexpected and unpredictable effect that is unrelated to the intended pharmacologic action of a drug. It is nonimmune but reproducible if [the drug is] readministered."

Common causes of such idiosyncratic drug reactions include aspirin and other nonsteroidal anti-inflammatory drugs. Dr. Nicklas noted that these drugs "have the potential to produce life-threatening reactions in some patients, particularly patients with asthma. It does not appear to be an allergic reaction." He added, "The fact that a medication is not necessarily associated with a true allergic reaction does not mean that it can be given safely without concern about serious types of reactions."

Patients with severe reactions to aspirin and other nonsteroidal anti-inflammatory drugs "can be desensitized, just [as they] can be for penicillin, even though the reactions are different," Dr. Nicklas said.

AIDS PATIENTS PARTICULARLY SUSCEPTIBLE

Drug hypersensitivity reactions are especially common in HIV-infected patients. "We're unsure as to why this group tends to be more reactive to drugs, but there could be a variety of factors," Dr. Gruchalla said. "These drugs may be more likely to be metabolized to reactive forms in these individuals than in other individuals." She added that HIV-infected patients may also "have a slow acetylator phenotype, as well as a glutathione deficiency. Both of these may play an important role in the liver's ability to rid the body of toxic drug metabolites."

Furthermore, "we used to think it was only the sulfonamides that were causing problems in patients with AIDS. But there are many others," Dr. Gruchalla said. Antimycobacterial agents, pentamidine, phenytoin, and zidovudine are known to cause adverse reactions in this group of patients. AIDS patients are also known to have coexisting infections that "may predispose these patients to drug reactions," she said. Still, "depending upon the kind of reaction, some of these drugs can actually be reintroduced using a desensitization procedure."

DRUG ALLERGY TESTING: HELP OR HARM?

Given the scope of the problem presented by adverse drug reactions, is it worth sending patients for drug allergy testing? Dr. Gruchalla is leery about the claims made by many laboratories that perform such testing. "Allergic drug reactions are so diverse," she explained. "In many instances, we don't know the particular immunogenic agent. What's sitting in the bottle on the shelf is not what is important in the body. In many cases, you don't know what antigen to use in the testing process, and you don't even know what to test for, because the mechanism is not known." As an alternative, Dr. Gruchalla recommends that physicians consult with an allergist who has expertise in the area of drug allergy before ordering a battery of tests.

CONSUMER AWARENESS

It is not only physicians who must learn more about adverse drug reactions. Patients need to become more aware of drug hypersensitivity, too. "The next goal of the task force should be to write similar guidelines for consumers," Dr. Bernstein said. "There is so much direct advertising by the drug manufacturers on television and other media, and possible side effects are minimized. Most of it is in small print. Patients with allergies who push their doctors to prescribe these medications may be putting themselves at risk. Consumers need to identify their own specific risks in terms of history and genetic susceptibility for any new drug."

Moreover, "no one can predict the number of adverse drug reactions or the extent of drug hypersensitivity from the database that companies use for the approval of a new drug," Dr. Bernstein continued. "Typically, only about 2,000 to 3,000 patients are involved in [trials for] any new drug application. In order to really have a true gauge of the incidence of adverse drug reactions, you have to study large populations. And that can't be done until the drug is marketed," he pointed out. Over half of FDA-approved drugs will produce serious side effects after they are marketed. Therefore, in order to reduce the incidence of adverse drug reactions, Dr. Bernstein suggests a national postmarketing drug surveillance program.

--Bob Kronemyer

Reference
1. Joint Task Force on Practice Parameters. Disease management of drug hypersensitivity: a practice parameter. Ann Allergy Asthma Immunol. 1999;83:665-700.

Return to table of contents