Cross-Reactivity and Sulfonamide Antibiotics
http://www.100md.com
《新英格兰医药杂志》
To the Editor: The article by Strom et al. (Oct. 23 issue)1 suffers from the classic shortcomings of retrospective studies of adverse drug reactions and particularly from the use of the term "allergic" to describe a myriad of outcomes grouped together as allergic drug reactions. The authors' broad definition of hypersensitivity or allergic reactions (in Supplementary Appendix 1) includes an "unspecified adverse effect of a drug, medicinal agent, or biologic substance" and introduces so much noise that the results are essentially useless clinically. Even their "narrow definition" includes highly diverse immunologic and nonimmunologic outcomes. True immunologic reactions will be swamped by reported adverse outcomes that are nonimmunologic. The authors suggest that the high reported rate of reactivity to penicillin in subjects identified as having sulfonamide "allergy" may identify a group of subjects at higher than average risk for antibiotic "allergic" reactions. It is far more likely that they identified patients who were predisposed to report adverse reactions to antibiotics — a common phenomenon in practice that is related to patients' vigilance. Without objective data to distinguish among true allergic reactions, different types of immunologic reactions, and nonimmunologic adverse events, the kind of information reported by Strom et al. does little to clarify the issues involved.
Andrew Saxon, M.D.
David Geffen School of Medicine at UCLA
Los Angeles, CA 90095-1680
asaxon@mednet.ucla.edu
Eric Macy, M.D.
Kaiser Permanente
San Diego, CA 92111
References
Strom BL, Schinnar R, Apter AJ, et al. Absence of cross-reactivity between sulfonamide antibiotics and sulfonamide nonantibiotics. N Engl J Med 2003;349:1628-1635.
To the Editor: Strom and colleagues discuss possible allergic cross-reactions between sulfonamide antibiotics and sulfonamide "nonantibiotics," irrespective of their chemical class. The authors consider as sulfonamide nonantibiotics compounds that are not sulfonamides in a narrower sense. For example, the antileprosy drug dapsone is a diphenylsulfone with para-amino groups at C4 and C4'. The antihypertensive agent diazoxide should be considered a benzothiadiazine-1,1-dioxide because sulfonamides with that ring structure (e.g., chlorothiazide) also have a free sulfonamide group at carbon 7. Sumatriptan, however, which is not mentioned by the authors, is a sulfonamide-containing compound (N-methyl-indol-methansulfonamide).
It is important to remember, in addition to the chemical structure of compounds, their pharmacodynamic and pharmacokinetic properties. Glibenclamide, for example, is a long-chain sulfonylurea derivative that is completely different from hydrochlorothiazide. A common sulfa allergy to different chemical classes of drugs cannot reasonably be expected.
Heinz G. Endres, M.D.
Ruhr University Bochum
D-44780 Bochum, Germany
heinz.endres@ruhr-uni-bochum.de
To the Editor: Strom and colleagues correctly warn of the increased possibility of allergic reactions to drugs in patients who have a history of such reactions to another drug. Table 1 of their article lists sulfonamide nonantibiotic drugs as compounds that should prompt concern. Three commonly used analgesics also belong on this list: celecoxib (Celebrex), rofecoxib (Vioxx), and valdecoxib (Bextra) — all of which contain an SO2NH2 moiety.
Howard J. Wetstone, M.D.
77 Kenwood Cir.
Bloomfield, CT 06002
h.wetstone@comcast.net
The authors reply: Drs. Saxon and Macy correctly point out that we were not studying immunologically proven drug reactions. To the contrary, the study was designed to examine conditions that clinicians see in practice and observations on which they base their medical decisions. It is for that reason that we defined our outcome as "one or more codes for a hypersensitivity or allergic reaction." Although we probably included some nonallergic events in our broad definition, repeating the analyses with the narrow definition yielded similar results. Although the broad definition included the code "unspecified adverse effect of a drug, medicinal agent, or biologic substance" — a point that is of concern to Drs. Saxon and Macy — not a single patient had this particular diagnosis. Furthermore, among 411 patients with "allergic reactions," only 47 (11.4 percent) had codes for an "adverse reaction to sulphonamides," an "adverse reaction to bendrofluazide," an "abnormal drug reaction," an "adverse drug reaction," a "drug adverse effect," a "drug abnormal adverse reaction," or a "drug reaction."
Finally, Drs. Saxon and Macy are concerned that our results may reflect data from patients who were predisposed to reporting adverse antibiotic reactions. The General Practice Research Database is a data base of medical records that includes information on all medical care received,1 in contrast to a spontaneous-reporting data base, which is subject to selective underreporting.2,3 With regard to the possibility that patients selectively reported adverse reactions to their physician, it should be noted that our results did not differ when the outcomes were restricted to hospitalization, which should have been recorded reliably.
Dr. Endres correctly notes the heterogeneity of the group of drugs we considered as sulfonamide nonantibiotics. However, it is important to note that our results persisted in each of four separate subanalyses: thiazides only, loop diuretics only, sulfonylureas only, and other sulfonamide nonantibiotics. Furthermore, in clinical practice, patients are considered sulfa-allergic in a generic sense, without regard to the distinctions made by Dr. Endres.
Dr. Wetstone points out that some of the cyclooxygenase-2 inhibitors are sulfonamide nonantibiotics. Indeed, this is true of celecoxib and valdecoxib, though not of rofecoxib.4 We were unable to include these drugs in our study, however, because neither celecoxib nor valdecoxib was available in the United Kingdom at the time of the study and because use of rofecoxib in the United Kingdom was still limited.
Brian L. Strom, M.D.
Rita Schinnar, M.P.A.
Andrea J. Apter, M.D.
University of Pennsylvania School of Medicine
Philadelphia, PA 19104
bstrom@cceb.med.upenn.edu
References
Garcia Rodriguez LA, Perez-Gutthann S, Jick S. The UK General Practice Research Database. In: Strom BL, ed. Pharmacoepidemiology. 3rd ed. Chichester, England: John Wiley, 2000:375-85.
Kennedy DL, Goldman SA, Lillie RB. Spontaneous reporting in the United States. In: Strom BL, ed. Pharmacoepidemiology. 3rd ed. Chichester, England: John Wiley, 2000:151-74.
Wiholm BE, Olsson S, Moore N, Waller P. Spontaneous reporting systems outside the US. In: Strom BL, ed. Pharmacoepidemiology. 3rd ed. Chichester, England: John Wiley, 2000:175-92.
Bingham CO III. Development and clinical application of COX-2 selective inhibitors for the treatment of osteoarthritis and rheumatoid arthritis. Cleve Clin J Med 2002;69:Suppl 1:SI-5.
Andrew Saxon, M.D.
David Geffen School of Medicine at UCLA
Los Angeles, CA 90095-1680
asaxon@mednet.ucla.edu
Eric Macy, M.D.
Kaiser Permanente
San Diego, CA 92111
References
Strom BL, Schinnar R, Apter AJ, et al. Absence of cross-reactivity between sulfonamide antibiotics and sulfonamide nonantibiotics. N Engl J Med 2003;349:1628-1635.
To the Editor: Strom and colleagues discuss possible allergic cross-reactions between sulfonamide antibiotics and sulfonamide "nonantibiotics," irrespective of their chemical class. The authors consider as sulfonamide nonantibiotics compounds that are not sulfonamides in a narrower sense. For example, the antileprosy drug dapsone is a diphenylsulfone with para-amino groups at C4 and C4'. The antihypertensive agent diazoxide should be considered a benzothiadiazine-1,1-dioxide because sulfonamides with that ring structure (e.g., chlorothiazide) also have a free sulfonamide group at carbon 7. Sumatriptan, however, which is not mentioned by the authors, is a sulfonamide-containing compound (N-methyl-indol-methansulfonamide).
It is important to remember, in addition to the chemical structure of compounds, their pharmacodynamic and pharmacokinetic properties. Glibenclamide, for example, is a long-chain sulfonylurea derivative that is completely different from hydrochlorothiazide. A common sulfa allergy to different chemical classes of drugs cannot reasonably be expected.
Heinz G. Endres, M.D.
Ruhr University Bochum
D-44780 Bochum, Germany
heinz.endres@ruhr-uni-bochum.de
To the Editor: Strom and colleagues correctly warn of the increased possibility of allergic reactions to drugs in patients who have a history of such reactions to another drug. Table 1 of their article lists sulfonamide nonantibiotic drugs as compounds that should prompt concern. Three commonly used analgesics also belong on this list: celecoxib (Celebrex), rofecoxib (Vioxx), and valdecoxib (Bextra) — all of which contain an SO2NH2 moiety.
Howard J. Wetstone, M.D.
77 Kenwood Cir.
Bloomfield, CT 06002
h.wetstone@comcast.net
The authors reply: Drs. Saxon and Macy correctly point out that we were not studying immunologically proven drug reactions. To the contrary, the study was designed to examine conditions that clinicians see in practice and observations on which they base their medical decisions. It is for that reason that we defined our outcome as "one or more codes for a hypersensitivity or allergic reaction." Although we probably included some nonallergic events in our broad definition, repeating the analyses with the narrow definition yielded similar results. Although the broad definition included the code "unspecified adverse effect of a drug, medicinal agent, or biologic substance" — a point that is of concern to Drs. Saxon and Macy — not a single patient had this particular diagnosis. Furthermore, among 411 patients with "allergic reactions," only 47 (11.4 percent) had codes for an "adverse reaction to sulphonamides," an "adverse reaction to bendrofluazide," an "abnormal drug reaction," an "adverse drug reaction," a "drug adverse effect," a "drug abnormal adverse reaction," or a "drug reaction."
Finally, Drs. Saxon and Macy are concerned that our results may reflect data from patients who were predisposed to reporting adverse antibiotic reactions. The General Practice Research Database is a data base of medical records that includes information on all medical care received,1 in contrast to a spontaneous-reporting data base, which is subject to selective underreporting.2,3 With regard to the possibility that patients selectively reported adverse reactions to their physician, it should be noted that our results did not differ when the outcomes were restricted to hospitalization, which should have been recorded reliably.
Dr. Endres correctly notes the heterogeneity of the group of drugs we considered as sulfonamide nonantibiotics. However, it is important to note that our results persisted in each of four separate subanalyses: thiazides only, loop diuretics only, sulfonylureas only, and other sulfonamide nonantibiotics. Furthermore, in clinical practice, patients are considered sulfa-allergic in a generic sense, without regard to the distinctions made by Dr. Endres.
Dr. Wetstone points out that some of the cyclooxygenase-2 inhibitors are sulfonamide nonantibiotics. Indeed, this is true of celecoxib and valdecoxib, though not of rofecoxib.4 We were unable to include these drugs in our study, however, because neither celecoxib nor valdecoxib was available in the United Kingdom at the time of the study and because use of rofecoxib in the United Kingdom was still limited.
Brian L. Strom, M.D.
Rita Schinnar, M.P.A.
Andrea J. Apter, M.D.
University of Pennsylvania School of Medicine
Philadelphia, PA 19104
bstrom@cceb.med.upenn.edu
References
Garcia Rodriguez LA, Perez-Gutthann S, Jick S. The UK General Practice Research Database. In: Strom BL, ed. Pharmacoepidemiology. 3rd ed. Chichester, England: John Wiley, 2000:375-85.
Kennedy DL, Goldman SA, Lillie RB. Spontaneous reporting in the United States. In: Strom BL, ed. Pharmacoepidemiology. 3rd ed. Chichester, England: John Wiley, 2000:151-74.
Wiholm BE, Olsson S, Moore N, Waller P. Spontaneous reporting systems outside the US. In: Strom BL, ed. Pharmacoepidemiology. 3rd ed. Chichester, England: John Wiley, 2000:175-92.
Bingham CO III. Development and clinical application of COX-2 selective inhibitors for the treatment of osteoarthritis and rheumatoid arthritis. Cleve Clin J Med 2002;69:Suppl 1:SI-5.