ALLERGY
`CLINICAL
`IMMUNOLOGY
`
`AND
`
`~
`
`SYMPOSI
`PROCEEDINGS ON
`DRUG ALLERGY: PREVENTION,
`DIAGNOSIS, TREATMENT
`
`December 1, 1983
`W ashington, D.C.
`
`1-'diton
`Robert A. Goldstein, M.D., Ph.D., and Roy Patterson, M .D.
`
`Editonal A11i.1w111
`Janet Ayres
`
`!,po111or"d /11
`The National Institute of Allergy and Infectious Diseases
`National Center for Drugs and Biologics (FDA)
`in coopcrution 11"11/,
`American Academy of Allergy and Immunology
`Asthma and Allergy Foundation of America
`S11pponed /J)
`Boehringer lngelheim
`Schering Corporation
`Pu blished monthly by The C. V . M osby Company
`
`ISSN 0091-6749
`
`

`

`THE JOURNAL OF
`
`AND
`
`ALLERGY
`CLINICAL IMMUNOLOGY
`October, 1984
`
`VOLUME 74
`
`NUMBER 4
`PART2
`
`SYMPOSIUM
`PROCEEDINGS ON
`DRUG ALLERGY: PREVENTION,
`DIAGNOSIS, TREATMENT
`
`December 1, 1983
`Washington, D.C.
`
`-IEAL TH SCIENCES UBRi-,i=:Y
`University of Wisconsin
`1305 Linden Dr .. Madison, Wis. 53706
`
`OCT 2 2 1984
`
`Editors
`Robert A. Goldstein, M.D., Ph.D., and Roy Patterson, M.D.
`
`Editorial Assistalll
`Janet Ayres
`
`Sponsored by
`The National Institute of Allergy and Infectious Diseases
`National Center for Drugs and Biologics (FDA)
`in cooperation with
`American Academy of Allergy and Immunology
`Asthma and Allergy Foundation of America
`
`

`

`Faculty
`
`N. Franklin Adkinson, M.D.
`The Johns Hopki11s University School of Medicine
`Baltimore, Md.
`
`Robert A. Goldstein, M.D., Ph.D.
`National Institute of Allergy and Infectious Diseases
`National lnstitures of Health
`Bethesda, Md.
`
`Paul A. Greenberger, M.D.
`Northwestern University
`Chicago, Ill.
`
`Hershel Jick, M.D.
`Boston Medical School
`Boston University
`Boston. Mass.
`
`Allen P. Kaplan, M.D.
`State University of New York at Stony Brook
`Stony Brook, N.Y.
`
`Richard M. Krause, M.D.
`National Institute of Allergy and Infectious Diseases
`National Institutes of Health
`Bethesda. Md.
`
`Kenneth P. Mathews, M.D.
`University of Michigan
`A1111 Arbor, Mich.
`
`J. Michael McGinnis, M.D.
`Public Health Service
`U.S. Department of Health and Human Services
`Washington, D.C.
`
`Paul D. Parkman, M.D.
`National Center for Drugs and Biologics, FDA, PHS
`U.S. Department of Health and Human Services
`Bethesda, Md.
`
`Roy Patterson, M.D.
`Northwestern University
`Medical School
`Chicago, Ill.
`
`Michael Schatz, M.D.
`Kaiser-Permanente Medical Center and
`University of San Diego S/hool of Medicine
`San Diego, Calif.
`
`Albert L. Sheffer, M.D.
`Harvard Medical School
`Boston, Mass.
`
`Ronald A. Simon, M.D.
`Scripps Clinic and Research Foundation and
`University of California San Diego
`School of Medicine
`San Diego, Calif.
`
`Dorothy D. Sogn, M.D.
`National lnstirure of Allergy and Infectious Diseases
`National Institutes of Health
`Bethesda, Md.
`
`Donald D. Stevenson, M.D.
`Scripps Clinic and Research Foundation
`La Jolla, Calif.
`
`Timothy J. Sullivan, M.D.
`University of Texas
`Health Sciences Center
`Dallas, Texas
`
`Eng M. Tan, M.D.
`W. M. Keck Foundation and
`Scripps Clinic and Research Foundation
`La Jolla, Calif.
`
`

`

`THE JOURNAL OF
`
`ALLERGY
`CLINICAL IMMUNOLOGY
`
`AND
`
`VOLUME 74
`
`MOSBY -
`
`Copyright © 1984 by The C. V. Mosby Company
`
`NUMBER 4
`PART2
`
`CONTENTS
`October, 1984
`
`Foreword
`
`Robert A. Goldstein. M.D., Ph.D .. Berhesdt1, Md.
`
`Introductory remarks
`
`J. Micht1el McGinnis. M.D .. Washi1tgron, D.C.
`Paa/ D. Parkman, M.D .. Bethesda, Md.
`Richard M . Krause. M.D .. At/all/a, Ga.
`
`549
`
`551
`552
`553
`
`Contents continued on page 4A
`
`Vol. 74. No. 4 , Part 2, October, 1984. THE JOURNAL OF ALLERGY AND CLINICAL it,!MUNOLOGY is published monthly by The C . V. Mosby
`Company, 11830 Westline Industrial Drive, St. Louis, Mo. 63146. POSTMASTER: Send address changes to The C. V. Mosby Co., 11830
`Westline Industrial Drive, St. Louis, Mo. 63146.
`
`1985 Annual subscription rates
`
`U.S .A.
`
`lnternatioMI
`(surface mail)
`All regions
`
`l111ernariona/
`(airmail)•
`
`Region I
`
`Region 2
`
`Region 3
`
`$75.00
`$40.00
`$32.00
`
`lnstitutionalt
`Individual+
`Student, resident+
`
`$93.50
`$1 12.50
`$13-3.65
`$173.25
`$58.50
`$ 77.50
`$ 98.65
`$119.75
`$50.50
`$ 69.50
`$ 90.65
`$1 11.75
`Single copies are $5.00. Remittances should be made by check, draft, or post office or express money order, in U.S. funds, drawn through a
`U.S. bank, payable to this 1oURNAL.
`•Airmail breakdown - Domestic: First-class and Priority rates for the U.S. and possessions are available upon request. Region /: Central
`America, islands, and mainland colonies of European countries in The ~ricas. Region 2: South America, Europe, Egyp(, Africa (bordering
`the Mediterranean). Region 3: Asia. Australasia, Africa (other than Mediterranean), Middle East, Far East, The Pacific. U.S.S.R . (and
`constituent Republics).
`tlnstitutional (multiple-reader) subscriptions are available to public and private libraries, schools, hospitals, and clinics; city, county, state,
`provincial, and national government bureaus and departments; and all commercial and private institutions and organizations.
`+Individual subscriptions and all student-rate subscriptions must be in the names of, billed to, and paid by individuals. All student-rate requests
`must indicate training status and name of institution.
`Subscriptions may begin at any time.
`Second-class postage paid at St. Louis, Mo. and additional mailing offices. Printed in the U.S.A. Copyright © 1984 by The C . V. Mosby
`Company.
`
`October, 1984
`
`Page 3A
`
`

`

`CONTENTS
`CONTINUED
`
`Adverse drug reactions: The.magnitude of the problem
`
`555
`
`Hershel l ick, M .D ., Waltham, Mass.
`
`Clinical spectrum of allergic and pseudoallergic drug reactions
`
`558
`
`Kenneth P. Mathews, M.D., Ann Arbor, Mich.
`
`Risk factors for drug allergy
`
`N. Franklin Adkinson, Jr .. M .D ., Baltimore, Md.
`
`Drug-induced skin disease
`
`Allen P. Kaplan , M .D., Stony Brook, N .Y.
`
`Management of adverse drug reactions
`
`Alben L . Sheffer, M .D., and Doris S. Pennoyer, M .D ., Boston, Mass.
`
`Penicillin allergy
`
`Dorothy D. Sogn, M .D., Bethesda , Md.
`
`Allergic reactions to antimicrobial agents: A review of reactions to
`drugs not in the beta lactam antibiotic class
`
`Timothy J. Sullivan, M .D., Dallas, Texas
`
`Contrast media reactions
`
`Paul A. Greenberger , M.D., Chicago, Ill.
`
`Skin testing and incremental challenge in the evaluation of adverse
`reactions to local anesthetics
`
`Michael Schatz, M .D., San Diego, Calif.
`
`Diagnosis, prevention, and treatment of adverse reactions to
`aspirin and nonsteroidal anti-inflammatory drugs
`
`Donald D. Ste venson , M .D., La Jolla, Calif.
`
`Adverse reactions to drug additives
`
`Ronald A . Simon. M .D ., La Jolla , Calif.
`
`Autoallergic reactions induced by procainamide
`
`Eng M . Tan, M.D., and Robert L. Rubin, Ph .D., La Jolla, Calif.
`
`567
`
`573
`
`580
`
`589
`
`594
`
`600
`
`606
`
`617
`
`623
`
`631
`
`Page 4A
`
`October, 1984
`
`

`

`CONTENTS
`
`CONTINUED
`
`Proteins: Chymopapain and insulin
`
`Leslie C. Grammer, M.D., and Roy Patterson, M.D., Chicago, Ill.
`
`Early recognition of allergic reactions to new drugs
`
`Roy Patterson, M.D. , Chicago, Ill.
`
`Summary
`
`635
`
`641
`
`643
`
`Robert A. Goldstein, M.D., Ph .D., and Roy Patterson, M.D., Bethesda. Md .. and Chicago, 1/1.
`
`Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by The C. V.
`Mosby Company for libraries and other users ~gistercd with the Copyright Clearance Center (CCC) Transactional Reporting Service,
`provided that 104' per page is paid directly to CCC, 21 Congress St., Salem, MA 01970; phone (617) 744-3350. The U.S. Government
`~tains a nonexclusive, royalty-f= license in and to any copyright on articles authored by government employees. 0091-6749/ 84
`$00.00 + 10.
`
`October, 1984
`
`Page 5A
`
`

`

`This material may be protected by Copyright law (Title 17 U.S. Code)
`
`Risk factors for drug allergy
`
`N. Franklin Adkinson, Jr., M.D. Baltimore, Md.
`
`Judging from available data on penicillin allergy there appear to be multiple definable risk
`factors for drugcinduced immunopathology. The induction of a drug-specific immune response
`can be influenced by the age of the patient; underlying genetic or metabolic factors, which may
`restrict ability to initiate a drug-specific immune re.1ponse; the chemical properties of the drug,
`largely its protein reactivity; and the dose and duration of treatment as well as the route of drug
`administration. The elicitation of drug-induced immunopathology is a function of the persistence
`of a drug-.1pecific immune response: the frequency of drug treatmelll and its dose and duration ;
`and probable constitutional factors that may determine in an importalll way the efficiency with
`which a drug-specific immune response can be translated illlo a systemic allergic reaction .
`(J ALLERGY CUN IMMUNOL 74 :567, /984 .)
`
`Even with highly sensitizing drugs, only a small
`percentage of patients with multiple drug exposures
`develop adverse reactions that can be attributed to a
`drug-specific immune response. In a large population
`of clinic patients with sexually transmitted disease in
`Baltimore, we have prospectively observed more than
`8000 patients for evidence of allergic reactions after
`penicillin treatment. In that setting we found an in(cid:173)
`cidence of posttreatment "allergic" reactions of only
`1.45% among unselected patients without histories of
`prior penicillin allergy. What factors contribute to the
`risk of these unusual allergic reactions to drugs? This
`article attempts to address this clinically important
`question. Most of what we know regarding risk factors
`for drug allergy has been derived from studies of al(cid:173)
`lergy to the penicillin group of drugs-far and away
`the most prevalent cause of allergic drug reactions.
`The principles involved in penicillin allergy will hope(cid:173)
`fully be applicable to other allergenic drugs as well.
`Some influences worthy of consideration are listed
`in Table I. It is useful to separate, insofar as possible,
`effects on the induction of a drug-specific immune
`response from those contributing to the elicitation of
`a clinically overt reaction in a previously sensitized
`subject. The table indicates my rather crude assess(cid:173)
`ment of the likelihood that each of seven factors will
`influence the induction or elicitation p~ases of drug
`allergy. The terms "likely" and "unlikely" have been
`
`From the Johns Hopkins University School of Medicine at the Good
`Samaritan Hospital , Baltimore, Md.
`Reprint requests: N. Franklin Adkinson, Jr. , M.D., Johns Hopkins
`University School of Medicine at The Good Samaritan Hospital,
`5601 Lock Raven Blvd., Baltimore, MD 21239.
`
`chosen rather than the more concrete "yes" or "no"
`to emphasize the fact that in almost all cases the con(cid:173)
`clusions to be drawn are soft and are drawn from
`limited data and often indirect observations. Unfor(cid:173)
`tunately few attempts have been made to definitively
`assess the quantitative effects of the patient and treat(cid:173)
`ment variables about to be discussed .
`A good illustration of this point is the first relevant
`patient factor-age . Age is judged "likely" to exert
`a significant influence on the induction of drug allergy
`largely because of two clinical observations. First, it
`is generally agreed (without much firm supportive
`data) that pediatric populations suffer fewer allergic
`reactions to penicillin than adults who receive com(cid:173)
`parable doses . 1 Younger age is likely to be associated
`with less cumulative drug exposure, and the contri(cid:173)
`bution of this factor has not been sorted out. Second,
`there is evidence from studies of Hymenoptera allergy
`that immunoglobulin (lg) E antibody responses in chil(cid:173)
`dren, as assessed by skin or radioallergosorbent tests,
`dissipate more rapidly than those of adults. 2 This may
`be true for drug-specific lgE as well, although one
`cross-sectional analysis did not support this notion .3
`There are unfortunately no prospective studies in chil(cid:173)
`dren of the prevalence of penicilloyl lgE antibody
`responses or their rate of disappearance.
`Similarly, age is judged "unlikely" to have a major
`effect on the elicitation of drug allergy solely on the
`grounds that severe drug-induced immunopathology
`can and does occur in small children. After the neo(cid:173)
`natal period, there is little evidence to suggest that
`children lack any of the mechanisms of immunopa(cid:173)
`thology observed in adults . The possibility of a quan(cid:173)
`titative distinction remains . This is clearly an area
`
`567
`
`

`

`588 Adkinson
`
`J. ALLERGY CUN. IMMUNOL
`OCTOBER 1984
`
`TABLE I. Personal and treatment factors contributing to drug-induced immunopathology
`
`Personal factors
`I. Age
`2 . Genetic/constitutional factors
`a. Specific immune responsiveness
`b. Atopy (re: IgE mediation)
`3. Persistence of immune response
`Treatment factors
`I. Chemical properties of drug
`2. Dose, duration of treatment
`3. Route of drug administration
`4 . Frequency of drug treatment
`
`Influence on
`
`Induction of
`drug-specific
`immune response
`
`Ellcltatlon of
`dlnlcally overt
`reaction
`
`Likely
`
`Likely
`Unlikely
`
`Likely
`Likely
`Likely
`Unlikely
`
`Unlikely
`
`Likely
`Likely
`
`Likely
`Likely
`Not clear
`Likely
`
`where additional age-related research is needed, not
`only with regard to the genesis of drug allergy, but
`concerning the expression of immunopathologic re(cid:173)
`actions in general.
`Several constitutional or genetic factors are likely
`to be important detenninants of drug allergy. Exper(cid:173)
`imental studies in animals have indicated a genetic
`basis for the ability to respond immunologically to a
`variety of haptens and low-molecular-weight poly(cid:173)
`mers. •-6 To pursue this hypothesis in man, we have
`studied the IgG and lgE antibody responses after ad(cid:173)
`ministration of penicillin to 60 hospitalized patients
`who received at least 2 gm for 10 or more days. The
`antibody response specific for the major penicilloyl
`9 Eighteen
`detenninant was determined serologically. 7
`-
`percent of the patients made IgG alone, 20% made
`both IgG and IgE, but a majority mounted no de(cid:173)
`tectable serologic response. These data suggest that
`only about half of the adult population can respond
`to penicillin administration with a detectable lgG or
`IgE antibody response. There are of course numerous
`possible explanations for apparent nonresponsiveness
`to high-dose penicillin administration. 10 Whatever the
`ultimate explanation, however, it is reasonable to ex(cid:173)
`pect that nonresponders will have substantially re(cid:173)
`duced risk of not only lgE-mediated allergy but other
`immunologic reactions to penicillin as well.
`Another genetic constellation worthy of consider(cid:173)
`ation with regard to IgE-dependent drug reactions is
`the constitutional diathesis that has traditionally been
`called "atopy." The atopic disposition carries with it
`a higher than expected incidence of allergic rhinitis,
`allergic asthma (usually with onset in childhood), and/
`
`or atopic dermatitis. A useful marker of the atopic
`state-is an elevated total serum IgE. 11
`There are two lines of evidence to support the as(cid:173)
`sertion that atopy has little effect on the induction of
`a drug-specific IgE response. First, we did not observe
`any difference in the total serum lgE levels of the IgE
`responders and nonresponders in the study of peni(cid:173)
`cilloyl immune responsiveness mentioned above.*
`The second observation pertinent to this point con(cid:173)
`cerns the rates of clinically assessed atopy as indicated
`by responses to questions regarding prior history of
`seasonal rhinitis, allergic asthma, or atopic dermatitis
`among 4069 patients who were skin tested for peni(cid:173)
`cillin allergy. All patients had negative histories of
`penicillin allergy and were skin tested with penicilloyl
`polylysine and a minor detenninant penicillin mixture.
`Use of these two reagents is thought to be sufficient
`to detect most of the clinically important IgE antibody
`16 The results in(cid:173)
`12
`responses to penicillin antigens. 3·
`-
`dicate that there is no difference in the frequency of
`atopy by history in skin test- positive as opposed to
`skin test- negative subjects (Fig. I). These findings
`are compatible with the suggestion that an atopic back(cid:173)
`ground does not predispose to a higher rate of drug(cid:173)
`specific lgE antibody responses.
`On the other hand, an atopic background is found
`with a high frequency among patients suffering severe
`18 It even appears
`and fatal anaphylactic reactions. 17
`•
`that atopic individuals are at greater risk for serious
`pseudoallergic reactions such as those to contrast me(cid:173)
`dia where lgE-dependent mechanisms are not in-
`
`* Adkinson NF, Wheeler B: Unpubli~hed data.
`
`

`

`VOI.UME74
`NUMBER 4, PART 2
`
`Risk factors for drug allergy 589
`
`TABLE II. lgE-dependent reactions among
`patients with negative penicillin skin tests
`
`History of prior
`reaction
`
`Anaphylaxis
`Urticaria
`Exanthem
`Other/uncertain
`None
`
`Rate of lgE-dependent
`reactions after
`treatment• 1%)
`
`15.4t (2/ 13)
`1.si (15/192)
`2.4§ (3/ 126)
`0.5 (1/202)
`0.45 ( 18/3996)
`
`*During first 72 hours of treatment. Of the 21 reactions among
`history-positive patients, 16 involved transient urticaria! reac(cid:173)
`tions, three cases included angioedema, and three cases mani(cid:173)
`fested mild dyspnea and/or dysphagia. There were no cardio(cid:173)
`vascular alterations in any case.
`tp = 0.002.
`tp < 0.0001.
`§p = 0 .049.
`
`volved. 19 Whatever its basis, the atopic constitution
`appears to carry with it a substantially increased risk
`of serious allergic reactions once an lgE antibody re(cid:173)
`sponse has been mounted.
`Genetic or . constitutional factors other than atopy
`may also play a role in the elicitation of IgE-dependent
`reactions. We have recently obtained data that support
`the concept that a conglomeration of patient factors
`relating to an individual 's ability to translate an lgE
`antibody response into a clinically manifested allergic
`reaction may constitute an important risk factor. As
`part of a prospective, randomized study of penicillin
`skin testing, we treated and observed for reaction 3996
`patients with negative histories of prior penicillin reac(cid:173)
`tivity and negative penicillin skin tests. Among these
`patients, 18 had reactions within the first 72 hours of
`therapy that were judged on clinical grounds to be
`probably IgE mediated. This translates to an IgE-de(cid:173)
`pendent reaction rate of 0.45% among history-nega(cid:173)
`tive, skin test- negative subjects. In the same clinical
`population, we also skin tested and treated 533 in(cid:173)
`dividuals with histories of prior penicillin reactions,
`but whose penicillin skin tests were currently shown
`to be unambiguously negative. The observed 72-hour
`reactions rates for probable IgE-mediated reactions arc
`as shown in Table II. The important observation here
`is that the rates for probable IgE-mediated reactions
`were significantly higher among patients with histories
`of urticaria and anaphylaxis than among patients with
`histories of other penicillin allergy or with negative
`histories.
`It is important to recall that all of the patients in
`the first two categories had previously sustained prob-
`
`35
`
`30
`
`25
`
`>, ! 20
`I us
`
`l.
`
`10
`
`5
`
`N•Ttl
`
`N•
`3994
`
`Polttlve
`PENICILLIN SKIN TESTS
`
`FIG. 1. Frequency of atopy among 4069 subjects skin test(cid:173)
`ed for penicillin allergy. Subjects were randomly selected
`from clinic patients with sexually transmitted disease who
`required penicillin therapy for an active infection. All sub(cid:173)
`jects denied prior penicillin reaction. Atopy was deter(cid:173)
`mined by any positive response to questions regarding
`personal history of allergic asthma, seasonal rhinitis, and/
`or atopic dermatitis. The rate of atopy did not differ be(cid:173)
`tween skin test-positive and skin test-negative subjects
`(p = 0.78; Fisher's exact test).
`
`able lgE-mediated reactions and had unequivocally
`negative penicillin skin tests prior to their treatment.
`Hence, their reactions must be presumed due to very
`small amounts of lgE antibodies that are undetectable
`by standard skin testing techniques. While such a
`small amount of drug-specific antibody may be of no
`clinical consequence for most patients, it appears to
`be capable of mediating urticaria) and anaphylactic
`reactions more readily in patients who have sustained
`such reactions in the past. It may be that some or all
`of these reactions in skin test- negative patients are
`not lgE dependent, but are pseudoallergic or idiosyn(cid:173)
`cratic in origin. Nevertheless, the fact remains that
`patients who previously sustained anaphylactic or ur(cid:173)
`ticaria) reactions to penicillin are at greater risk for
`recurrence of such reactions if they arc retreated(cid:173)
`even when IgE antibodies are undetectable by skin
`testing. The relationship of this constitutional risk fac(cid:173)
`tor to atopy, as classically defined, remains to be de(cid:173)
`termined.
`A third factor that clearly influences the cumulative
`risk of penicillin allergy is the persistence of drug(cid:173)
`specific antibodies once they are formed . We are con(cid:173)
`ducting an ongoin~ study to assess prospectively the
`
`

`

`(:3 Procaine P9nicillin (N•2988)
`(3.00 IM)
`0 Amplclllln (N•820)
`(3.50 P.OJ
`E1'Jllenzathine l'llniclllin (N■ l79)
`(1.50 IM)
`
`570 Adkinson
`
`100
`
`80
`
`20
`
`loE Mediated
`
`Probably
`lmmunolo0ic
`
`Al I
`
`FIG. 2. Comparison of reaction rates for three penicillin
`treatment regimens. All subjects gave negative histories
`of prior penicillin reactions; none were skin tested prior
`to treatment. Treatment regimens were chosen on clinical
`grounds, independently of study. Reactions within 72
`hours of initiation of treatment were designated on clinical
`criteria to be lgE mediated (urticaria, anaphylaxis). The
`" probably immunologic" category included lgE-mediated
`reactions, plus pruritus and/or nonurticarial skin erup(cid:173)
`tions. The " all" category included reactions of any type
`attributed by the patient to penicillin treatment. In all three
`categories, the reaction rates for oral ampicillin and pro(cid:173)
`caine penicillin did not differ by chi-square analysis
`(p = 0.23 to 0.89), while that for bicillin was significantly
`higher (p = 0.003 to 0.01 ).
`
`natural history of penicillin allergy in man. This in(cid:173)
`volves the periodic longitudinal follow-up of patients
`who have positive penicillin skin tests. More than 50
`such patients are currently being evaluated with rou(cid:173)
`tine penicillin skin tests and serologic determinations
`of penicilloyl IgG and lgE antibodies every 6 to 8
`weeks. Although the study is incomplete, available
`data indicate that the penicilloyl IgE antibody half(cid:173)
`life in serum ranges from as short as 10 days to an
`indeterminantly long interval (> 1000 days). The rate
`of decline of penicillin antibodies cannot as yet be
`related to any clinical or immunologic parameters,
`including the recentness of penicillin therapy, anti(cid:173)
`body titer at entry, or the patient's history of prior
`penicillin reaction. Furthermore, the half-life of pen(cid:173)
`icilloyl IgE antibody is not significantly correlated
`with the half-life of IgE tetanus antibody in the same
`patients. This finding suggests that the persistence of
`penicilloyl IgE antibodies depends on some factors
`
`J. ALLERGY CUN. IMMUNOL.
`OCTOBER 1984
`
`that are independent of those controlling the metab(cid:173)
`olism of IgE antibodies in general. Further study will
`be needed to determine what these host factors are
`and the role, if any, of occult exposure to penicillin
`or cross-reacting antigens in the environment. It is
`obvious, however, that whatever the reason, the per(cid:173)
`sistence of IgE antibody over long periods of time
`carries with it a cumulative risk of penicillin allergy
`much greater than that experienced by patients whose
`drug-specific antibody levels decline rapidly.
`There are at least four treatment-related factors that
`may influence the rate of induction and/or elicitation
`of allergic drug reactions. The first of these is the
`chemical properties of the drug. We now know quite
`conclusively from multiple lines of evidence that this
`is a pivotal determinant of drug allergies, with a strong
`influence on both the induction of a drug-specific im(cid:173)
`mune response and the subsequent elicitation of im(cid:173)
`munopathology. In work by Herman Eisen20 and his
`colleagues in the early 1950s , the rate of contact sen(cid:173)
`sitization by simple chemicals was found to be directly
`related to the rate of conjugation of those chemicals
`to large-molecular-weight carriers, usually proteins.
`This concept has now achieved the status of immu(cid:173)
`nologic dogma. In general this principle has been
`borne out clinically. Highly allergenic drugs invari(cid:173)
`ably have significant protein reactivity, and those
`drugs that are relatively inert with regard to protein
`interactions enjoy a very low incidence of allergic side
`effects. 2 1. 22
`The numerous ways in which penicillin can interact
`covalently with proteins has been well worked out
`thanks to a decade of pioneering work by Drs. Charles
`20 It
`Parker and Bernard Levine and their associates. 21
`·
`should be pointed out that the covalent attachment of
`penicillin to proteins occurs at a measurable rate under
`physiologic conditions without the need for enzyme,
`28 Hence, it is possible to detect pen(cid:173)
`or cofactors. 27
`·
`icilloated proteins in the serum of nearly every patient
`receiving moderate doses of penicillin drugs.* Thus
`the failure to observe a penicilloyl immune response
`in a patient on long-term penicillin therapy is not likely
`to be explained by the patient's inability to produci.:
`drug protein conjugates. This situation with penicillin
`stands in contrast to that of other drugs such as iso(cid:173)
`niazid and hydralazine where a genetically determined
`capacity to acetylate the drug may contribute to its
`immunogenicity in some patients. 22
`29
`·
`The dose and duration of drug treatment are likely,
`within limits, to affect the induction of a drug-specifiL(cid:173)
`immune response. Evidence of this comes largely
`
`• Adkinson NF. Wheeler B: Unpublished data.
`
`

`

`YOLUME 1,
`NUMBER 4, PART 2
`
`Risk factors for drug allergy 571
`
`12
`
`•
`
`1
`
`•
`
`35
`
`from studies in experimental animals. The dose and
`duration of drug treatment may also influence the elic(cid:173)
`itation of immunopathology. For example, drug-in(cid:173)
`duced cytopenias, interstitial nephritis, and serum
`sickness reactions occur more frequently with higher(cid:173)
`dose, longer-term therapy. J0.
`32 Even lgE-dependent re(cid:173)
`actions are dose related; anecdotal cases are reported
`in which patients tolerate a given level of treatment,
`but manifest acute allergic reactions when the dose is
`increased. 33 Furthermore, the duration of treatment
`defines the temporal period of risk and thereby influ(cid:173)
`ences immunopathology independent of any effect it
`may have on immune responsiveness.
`Another risk factor is the route of drug administra(cid:173)
`tion. Regarding the induction phase, the use of topical
`penicillin has been largely abandoned due in large
`part to the high rate of contact sensitization observed
`during its early use in toothpastes and mouthwashes
`in Europe. 17 Otherwise there is little evidence that the
`risk of sensitization is greater for one route of treat(cid:173)
`ment than for another.
`Two clinical surveys have found that penicillin re(cid:173)
`actions were elicited almost twice as often with par(cid:173)
`enteral administration as with oral administration. 34
`In reviewing the world literature, Sullivan et al. 33 not(cid:173)
`ed that each year only six anaphylactic deaths occurred
`after oral administration of 13-lactam drugs compared
`ID hundreds of penicillin-induced fatalities after par(cid:173)
`enteral administration. However, our recent experi(cid:173)
`ence suggests that with careful follow-up, reaction
`rates for similar doses of oral and parenteral penicillin
`may be comparable.
`Fig. 2 shows results from a prospective treatment
`protocol in which three treatment regimens were com(cid:173)
`pared: procaine penicillin, 3 gm intramuscularly; am(cid:173)
`picillin, 3.5 gm orally as a single dose; and benzathine
`penicillin, 1.5 gm intramuscularly. One gram of pro(cid:173)
`benecid was given orally along with the parenteral
`procaine penicillin and the oral ampicillin. Reactions
`occurring within 72 hours of penicillin treatment were
`documented. Results are similar for all reported re(cid:173)
`actions, those judged of probable immunologic origin
`and those judged clinically to be likely lgE mediated.
`In each case there is no difference in the reaction rate
`observed with comparable doses of procaine penicillin
`Ind orally administered ampicillin. The reaction rates
`for benzathine penicillin were significantly higher, de(cid:173)
`spite its lower dose, suggesting that the benzathine
`additive is responsible for the excess reaction rate.
`This finding of a higher reaction rate for benzathine
`penicillin and other depo-type penicillins is in agree(cid:173)
`men with other available studies. 36
`This failure to demonstrate an increased reaction
`
`·
`
`rate for intramuscular procaine penicillin could be due
`to the excess rate of reactions associated with ampi(cid:173)
`cillin as opposed to benzylpenicillin, or to the use of
`probenecid to boost serum levels. Nevertheless, these
`results raise the question of whether the reported dif(cid:173)
`ferences in reaction rates for oral versus parenteral
`penicillin may not be due more likely to a difference
`in usual dose than the route of administration.
`Finally, there is little evidence that the frequency
`of drug treatment affects the likelihood of sensitiza(cid:173)
`tion. On the other hand, the frequency of drug usage
`has a marked influence on the likelihood of eliciting
`an allergic reaction. In the case of IgE-dependent pen(cid:173)
`icillin allergy, most patients lose their lgE antibody
`and hence their sensitivity over a finite period of
`time.' · 10
`• Thus, the more frequently the drug is
`administered, the more likely that an IgE-dependent
`reaction will be provoked. By analogy, this may also
`apply to other forms of drug-induced immunopath(cid:173)
`ology.
`It should be clear from the tentative nature of all
`of the conclusions drawn that much important work
`has yet to be undertaken. In considering future re(cid:173)
`search directions, an emphasis on patient-related vari(cid:173)
`ables rather than the treatment-related factors is to be
`desired. This is because drug allergy is rare enough
`that treatment-related risk factors are unlikely to have
`a significant impact on the way we use drugs clini(cid:173)
`cally-that is, on the dose, duration, route, and fre(cid:173)
`quency of prescribed drug therapy. An exception to
`this generalization would be knowledge of the chem(cid:173)
`ical properties of a drug, particularly its protein reac(cid:173)
`tivity, since such information may impact on drug
`design and premarket estimates of allergenic potential.
`On the other hand, patient-related factors are of im(cid:173)
`mense practical importance since they may identify
`individuals who have generally increased risk for mul(cid:173)
`tiple allergic drug reactions or for multiple types of
`immunopathology. In the case of penicillin allergy,
`for example, if high-risk patient subsets can be iden(cid:173)
`tified, then routine penicillin skin testing before treat(cid:173)
`ment can be made vastly more cost-effective.
`Since drug-specific immune responses occur much
`more frequently than drug-induced immunopathology,
`attention should be focused on the patient factors that
`are responsible for translating a drug-specific immune
`response into an allergic reaction. These factors need
`to be delineated and quantified. As concerns drugs
`whose prevalence of use or reaction rates are less than
`for 13-lactam antibiotics, collaborative research efforts
`involving laboratory, clinical, and epidemiologic stud(cid:173)
`ies will be necessary for findings to be conclusive.
`Such studies are likely to be successful only if un-
`
`

`

`572 Adkinson
`
`J. ALLERGY CUN. IMMUNOL.
`OCTOBER 1984
`
`dertaken as a joint venture involving interested in(cid:173)
`vestigators, federal agencies, and the pharmaceutical
`industry.
`
`REFERENCES
`
`I. Berkowitz M, Glaser J, Johnstone DE: The incidence of allergy
`to drugs in pediatric practice. Ann Allergy 11:561, 1953
`2. Sch