Donald V Belsito
Division of Dermatology
University of Kansas Medical Center
Kansas City. KS
The incidence of iatrogenic allergic contact dermatitis (IACD) induced by topical medications varies considerably from one country to another, primarily because of differences in prescribing habits.(1–4) In addition, the drugs responsible for IACD change with time, as medications exit and enter the marketplace.(2–8) Individuals at greatest risk for developing IACD are dermatitic patients applying topical agents, healthcare workers and employees in the pharmaceutical industry.
While the intrinsic chemical structure of the medicament determines its sensitising power, highly sensitising agents that are little used in clinical practice or that penetrate the stratum corneum poorly can be associated with a low incidence of IACD. In contrast, medicaments of low sensitising capacity that permeate readily through the stratum corneum, or are applied with occlusion or vehicles that enhance penetration, may induce significant amounts of iatrogenic disease, especially if they are widely used.
Topical antibiotic preparations are frequently used for treatment and/or prevention of cutaneous infections. As a result, the development of IACD to the more-sensitising molecules has ensued.(9) Among these, neomycin, an aminoglycoside antibiotic, is particularly notorious as a contact sensitiser, with incidence rates among patients undergoing patch testing for eczematous dermatitis ranging above 10%.(4) Cross-sensitivity to other aminoglycosides is common.(8,10) Neomycin is often combined with bacitracin and polymyxin B and marketed as a “triple antibiotic”. As a result, it commonly coreacts with bacitracin.(8)
Bacitracin is a polypeptide that has been variably assessed as a weak-to-potent sensitiser.(11) After neomycin, it is the most common topical medicament inducing IACD, with an incidence rate among eczematous patients approaching 10%.(4,12) Anaphylactic shock due to IgE-mediated allergic reactions to bacitracin has also been reported.(12)
The cyclic decapeptide polymyxin B is, along with neomycin and bacitracin, the third component in “triple antibiotic” preparations. Despite its common use in combination with the relatively frequent sensitisers neomycin and bacitracin, there are only rare reports of IACD to polymyxin B.(13) Other topical antibiotics in the marketplace that are also rare sensitisers include erythromycin, clindamycin, silver sulfadiazine, metronidazole, povidone-iodine, sodium fusidate and mupirocin.(14)
Because of issues surrounding IACD to neomycin and bacitracin, Smack et al(15) conducted a study on 884 evaluable volunteers who had undergone dermatological surgery. Four-hundred-and-forty individuals treated their surgical wound with white petrolatum, while 444 treated it with bacitracin. There were no differences in the demographic composition of the two groups. The incidence of postoperative infection and the rapidity of healing did not differ between the two groups. While 0.9% of the group treated with bacitracin developed an IACD, none among the white petrolatum group had this complication; however, the difference was not statistically significant (p=0.12). Nonetheless, these authors advocated the use of white petrolatum for uncomplicated postoperative wound care in the ambulatory setting.(15)
Topical corticosteroids have been used since Sulzberger and Witten introduced hydrocortisone in the 1950s.(16) The first case report of allergic contact dermatitis from hydrocortisone was published several years thereafter.(17) In 1972, Alani and Alani(18) tested a series of patients to hydrocortisone and reported allergic reactions in approximately 0.6% of these patients; however, individuals with stasis dermatitis were particularly prone to develop IACD to corticosteroids and had a much higher incidence rate of disease. In the late 1980s, with the introduction of tixocortol pivalate as a marker for topical corticosteroid allergy,(19) various international contact dermatitis groups have been tracking the incidence of IACD to corticosteroids. Over the course of the past decade, it has become apparent that, among European and North American patients undergoing patch testing, the incidence of reactions to corticosteroids approaches 3% in most countries.(4,20)
The more common corticosteroids responsible for IACD are nonhalogenated molecules, such as hydrocortisone and budesonide. Reactions to tixocortol pivalate, which initially was used to screen for all corticosteroid allergies, indicate allergy to the hydrocortisone group,(21) although one-third of these patients will also be found to be allergic to the hydrocortisone-17-butyrate group and other steroid molecules.(20) In addition to tixocortol pivalate, budesonide is commonly used as a screening allergen for corticosteroid reactivity, especially for reactions to triamcinolone acetonide, amcinonide, fluocinonide, etc.(22) Recently, hydrocortisone- 17-butyrate has been introduced as a screening agent for steroid molecules containing an ester substitution at C17 and/or C21 with molecular size greater than an acetate group. Among these steroids are betamethasone dipropionate, betamethasone valerate and hydrocortisone- 17-butyrate.(22)
Coopman et al(22) grouped corticosteroids into four main allergenic classes, based upon cross-reactivity patterns reported in the literature and observed in their clinics. While most practitioners find these groupings predictive, others have questioned their reliability.(23) Nonetheless, within a given allergenic group of steroids, halogenated molecules are less likely to induce IACD than nonhalogenated steroids. Wilkinson and Jones(24) demonstrated that corticosteroids with the highest arginine-binding capacity had the greatest ability to induce sensitisation, and that halogenation of the steroid molecule reduced this ability, perhaps explaining the lower incidence of contact dermatitis to halogenated corticosteroids.
In contrast to reports of IACD to topical corticosteroids, the frequency of allergic reactions to inhaled corticosteroids has been less studied. Nonetheless, case reports of IACD following the use of nasal sprays containing corticosteroids are becoming increasingly more common. The first cases of IACD to inhaled corticosteroids were reported by Boujnah-Khouadja et al in 1984.(25) Subsequently, Isaksson et al reported allergic reactions to tixocortol pivalate in 1.4% of patients with asthma or rhinitis who used inhaled corticosteroids, compared with 0.9% of patients without asthma or rhinitis.(26) Although the difference in this latter study was not significant (p<0.2%), the authors speculated that the slightly higher incidence of allergic reactions to corticosteroids among the asthma/ rhinitis patients might be due to inadvertent topical exposure, as well as to the higher frequency of atopic dermatitis in the asthma/rhinitis patients. In prior studies, it had been demonstrated that atopic eczema was a risk factor for corticosteroid allergy.(27)
Over the past several decades, topical nonsteroidal anti-inflammatory drugs (NSAIDs) have proliferated in the marketplace. While the topical application of these medicaments decreases the incidence of gastrointestinal and other systemic side-effects,(28) it increases the potential for IACD. Although most of these drugs would appear to have a low sensitising potential,(29,30) some, such as phenylbutazone, suprofen and benoxaprofen, have been withdrawn because of their sensitising properties.(31)
Following the widespread use of topical NSAIDs, numerous case reports of allergic and photoallergic contact dermatitis to these medicaments have appeared in the literature.(31) Allergic reactions have been reported from pyrazolone derivatives, propionic acid derivatives, indomethacin, fenamates, bufexamac and piroxicam. Most of these reactions were allergic contact in nature; however, photoallergic contact dermatitis has also developed following exposure to propionic acid derivatives (tiaprofenic acid, ketoprofen, carprofen and benoxaprofen) and piroxicam.(31) A study of ketoprofen-containing gels conducted by the French drug surveillance system reported a frequency of cutaneous adverse events (photosensitisation and contact allergy) of 0.008–0.023 per 1,000 applications.(32)
Among patients allergic to the pyrazolone or propionic acid derivatives, cross-reactivity between these two groups may occur.(31) However, cross-reactions among patients allergic to indomethacin, fenamates and the piroxicam groups of NSAIDs appear to be rare. Piroxicam has been reported to cross-react with thimerosal, since these molecules share the thiosalicylate moiety.(33,34) Patients known to be allergic to topical NSAIDs should avoid systemic administration of the same group. Case reports have documented systemic contact dermatitis to oxyphenbutazone.(35) In addition, some patients topically sensitised to tiaprofenic acid developed photosensitisation after taking ketoprofen and/or tiaprofenic acid orally.(36,37)
Topical anaesthetics can be broadly classified into two groups: the esters and the amides. The esters, especially the aminobenzoate esters benzocaine and tetracaine, are, in most countries, the more common sensitisers. The most recent data from the North American Contact Dermatitis Group(4) revealed that 1.7% of 5,833 patients undergoing patch testing for eczematous dermatitides were found to be allergic to benzocaine. In Germany, where benzocaine is more frequently used, the incidence of allergic reactions to this anaesthetic has been reported to be as high as 10%. (39) Because of their topical effectiveness, a multitude of different products for the treatment of pruritus, burns, haemorrhoids and other conditions contain benzocaine and/or tetracaine. Unsurprisingly, IACD to topical anaesthetics is particularly common among patients with anogenital disorders.
Because of the chemical structure of benzocaine, it can cross-react not only with other aminobenzoate anaesthetics (eg, tetracaine), but also with the hair dye paraphenylenediamine. In some studies, up to 14% of benzocaine-sensitive patients were also allergic to this hair dye.(40) Individuals who react to both benzocaine and paraphenylenediamine are more likely to cross-react to various sulpha-based drugs, the structure of which is an intermediate between the aminobenzoate and the paraphenylenediamine structures.(8) Among these sulpha-based drugs are the sulphonamides, sulphonylureas and thiazide diuretics.
Despite their increasing use in various topical preparations, most of the amide anaesthetics (lidocaine, mepivacaine, bupivacaine and prilocaine) are uncommon causes of IACD. Nonetheless, Sidhu et al(40) performed a retrospective analysis of 5,464 patients undergoing patch testing to a mixture of 5% benzocaine, 2.5% tetracaine and 2.5% dibucaine. Ninety-one (1.7%) of the 5,464 patients tested had allergic reactions to this mixture; 63 of these 91 patients subsequently underwent testing to the specific components of the mixture. Surprisingly, the amide anaesthetic dibucaine was responsible for the greatest number of caine mix allergic reactions (42.2%), followed by tetracaine (35.9%) and benzocaine (34.4%). Of interest, cross-reactivity between benzocaine and tetracaine appeared to be uncommon (13.6%) in this study,(40) which emphasises the need to test with the specific anaesthetic to which the patient is exposed.
The reason for the high incidence of dibucaine allergy in patients from the UK likely results from usage patterns. Sidhu et al(40) quoted data from the International Medical Statistics Group, which revealed that, in 1997, the total unit sales of dibucaine throughout the UK were 1,238,000 vs 467,000 for benzocaine. Thus, as demonstrated by data from Australia indicating that 49 of the 62 reported cases of IACD to the amide anaesthetic lignocaine worldwide were from that country,(41) where the number of over-the- counter lignocaine-containing topical anaesthetics abound, healthcare professionals must be cognisant of the patterns of anaesthetic use in their region. As previously mentioned, even molecules with a low potential for sensitisation can cause significant problems when they are frequently used, especially on damaged skin.
Among topical antihistamines, phenothiazine derivatives are the most notorious for inducing allergic and photoallergic contact dermatitis.(42,43) Individuals allergic to one phenothiazine have a high likelihood of cross-reacting to other members of the group.(42,44) Topically sensitised patients risk a systemic eczematous contact dermatitis following oral exposure to phenothiazines.(42–44)
After the phenothiazines, the ethanolamine-based diphenhydramine appears to induce the greatest number of cases of IACD to topical antihistamines, perhaps because of its widespread use. Although the incidence of reactivity to topically applied diphenhydramine is difficult to assess,(45) a study was carried out in England.(46) Twelve out of the 117 cases of dermatitis medicamentosa that were evaluated during a one-year period resulted from exposure to a lotion containing diphenhydramine hydrochloride. Individuals sensitised to diphenhydramine may react topically or systemically following exposure to other ethanolamine- based antihistamines.
Ethylenediamine-based antihistamines are not widely used in topical preparations. However, ethylenediamine dihydrochloride is used as a stabiliser in various topical preparations, and is also used as a hardener in plastic resin systems.(8) Some patients topically sensitised to ethylenediamine from these exposures may develop systemic contact dermatitis when exposed to ethylenediamine-based antihistamines such as hydroxyzine hydrochloride.(47)
In the past decade, doxepin hydrochloride, a tricyclic antihistamine that is structurally similar to phenothiazine, has been introduced into the marketplace as a 5% cream. Since its introduction, case reports of IACD have appeared. Of 97 patients evaluated at the Cleveland Clinic, 17 were found to have relevant positive reactions to doxepin cream, 13 of whom were allergic to the doxepin itself.(48) In most of these cases, the duration of treatment exceeded the recommended eight days and/or patients were applying the medicament to damaged skin.(49) Following cutaneous sensitisation to doxepin, systemic contact dermatitis after oral ingestion may occur.(50)
Contact dermatitis may occur where transdermal patches have been applied. In most cases, this is an irritant phenomenon due to the adhesives or to the occlusive nature of these delivery systems.
However, IACD has been reported to develop in 15–50% of patients using transdermal clonidine.(51,52) Curiously, most of these patients can subsequently tolerate systemic clonidine,(53) which is in sharp contrast to the systemic contact dermatitis observable following challenge of some patients topically sensitised to antibiotics, NSAIDs, anaesthetics or antihistamines. In the Transdermal Nicotine Study Group, 11 of 664 (1.7%) patients were determined to have developed true nicotine allergy from these patches. (54) Given the dearth of reports in the literature and the widespread use of transdermal nitroglycerin, scopolamine, oestrogen and testosterone patches, these medicaments seem to be rare contact sensitisers.
In addition to the active medicament, inactive ingredients in a topical preparation can account for iatrogenic disease. Sensitisation to components such as lanolin alcohol, formaldehyde-releasing preservatives, parabens, propylene glycol and fragrances are commonly seen, especially among patients applying these materials to dermatitic skin.(55) Thimerosal, a mercuric derivative of thiosalicylic acid, was previously used as a preservative in many consumer products, especially ophthalmic and otolaryngologic medications and vaccines, where it caused a significant incidence of IACD. However, as a result of IACD and of environmental concerns, its use has declined dramatically during the past two decades. Thus, while many patients undergoing patch testing demonstrate allergic reactions to thimerosal,(4) most of these are not relevant to current disease.(56) Thimerosal allergy is mostly found among women and healthcare workers, two populations with increased likelihood of vaccination with thimerosal- containing vaccines.(56) However, despite the apparent sensitisation to mercurials induced by vaccination, vaccination of individuals with thimerosal or other mercury allergies does not seem to be problematic.(57–59)
IACD to topical medicaments and to their inactive ingredients is common, especially among patients applying these substances to eczematous skin. Individuals who fail to respond to appropriate topical therapy, or whose condition worsens with this therapy, should be strongly suspected of having developed IACD to the medicament(s). Only by patch testing can the allergic aetiology be diagnosed. Treatment depends on removing the offending allergen from the patient’s environment. With successful avoidance, the iatrogenic complications should resolve spontaneously.
- Arch Dermatol 1972;106:335-7.
- Angelini G. Topical drugs. In: Rycroft RJG, Menné T, Frosch PJ, et al, editors. Textbook of contact dermatitis. Berlin/New York: Springer-Verlag; 1982:476-502.
- Storrs FJ. Contact dermatitis caused by drugs. In: van Arsdel PP Jr, editor. Immunology and allergy clinics of North America. Philadelphia: WB Saunders; 1991;3:509-23.
- Am J Contact Dermat 2003;14:59-62.
- Contact Dermatitis 1985;12:263-9.
- Contact Dermatitis 1977;3:327-32.
- Semin Dermatol 1982;1:49-57.
- Rietschel RL, Fowler JF Jr. Fisher’s contact dermatitis. Philadelphia: Lippincott, Williams & Wilkins; 2000.
- Semin Dermatol 1990;9:250-4.
- Hsu S, Quan LT. Topical antibacterial agents. In: Wolverton SE, editor. Comprehensive dermatological drug therapy. Philadelphia: WB Saunders; 2001. p. 472-96.
- Contact Dermatitis 1994;31:92-4.
- Dermatol Surg 2004;30:521-4.
- J Am Acad Dermatol 1990;23:646-7.
- J Dermatol 2003;30:644-8.
- JAMA 1996;276:972-7.
- J Invest Dermatol 1952;19:101-2.
- Hautarzt 1959;10:42-3.
- Ann Allergy 1972;30:181-5.
- Contact Dermatitis 1989;20:158.
- Contact Dermatitis 1996;35:40-4.
- Lancet 1991;337:761-2.
- Br J Dermatol 1989;121:27-34.
- Contact Dermatitis 2000;42:59-63.
- Br J Dermatol 1996;135:225-30.
- Contact Dermatitis 1984;11:83-7.
- Contact Dermatitis 1999;40:327-8.
- Contact Dermatitis 1992;26:182-91.
- Yearbook of Dermatology 1992;185-216.
- Br J Clin Pharmacol 1983;16:633-8.
- Contact Dermatitis 1990;23:336-40.
- Contact Dermatitis 1993;29:57-64.
- Therapie 1998;53:137-44.
- Contact Dermatitis 1991;25:30-4.
- Contact Dermatitis 1987;17:73-9.
- Contact Dermatitis 1985;12:236-7.
- Contact Dermatitis 1989;21:345-6.
- Allergologie 1984;7:143-4.
- Dermatol Ther 2004;17:20-7.
- Trans St Johns Hosp Dermatol Soc 1969;55:17.
- Am J Contact Dermat 1999;10:57-61.
- Contact Dermatitis 1998;39:265-6.
- JAMA 1955;157:909-10.
- J Invest Dermatol 1957;28:329-38.
- Contact Dermatitis Newsletter 1972;11:278-9.
- Arch Dermatol 1974;110:249-52.
- BMJ 1961;1:1366-7.
- Arch Dermatol 1976;112:525-6.
- Arch Dermatol 1996;132:515-8.
- J Am Acad Dermatol 2003;48:294-6.
- J Drugs Dermatol 2003;2:409-10.
- NZ Med J 1986;99:711-4.
- Chest 1988;93:941-5.
- J Am Acad Dermatol 1990;22:811-4.
- JAMA 1991;266:3133-8.
- Ostomy Wound Manage 2000;46 Suppl 1A:65s-74s.
- J Am Acad Dermatol 2001;45:23-7.
- Contact Dermatitis 1991;24:6-10.
- Contact Dermatitis 1988;18:229-33.
- Am J Contact Dermat 2002;13:3-9.