414 Canadian Family Physician | Le Médecin de famille canadien } Vol 67: JUNE | JUIN 2021

Clinical Review

Approach to allergic contact dermatitis caused by topical medicamentsCharles Choi BScPharm Saba Vafaei-Nodeh BSc Jamie Phillips MD Gillian de Gannes MD FRCPC

Editor’s key points} Allergic contact dermatitis (ACD) is a commonly overlooked diagnosis that should be part of the differential diagnosis for any pruritic eczematous eruption. Among topical medicament ingredients, antibiotics are the most common culprit, often in the context of postoperative wound care where ACD might be mistaken for infection. Antibiotics are also the most frequent cause of cosensitization.

} Topical anesthetics are found in an increasingly wide variety of prescription and over-the-counter products and in combination with topical antibiotics and corticosteroids. A comprehensive drug history including all topical products is therefore critical.

} Patch testing is indicated for anyone with suspected ACD to medicaments and can help guide patient counseling on appropriate allergen avoidance.

AbstractObjective To provide an approach to identifying topical medicament ingredients that cause allergic contact dermatitis (ACD) and to recognizing common clinical scenarios in which these ingredients might present.

Sources of information A retrospective chart review was conducted of patients patch tested at the Contact Dermatitis Clinic at St Paul’s Hospital in Vancouver, BC, between November 2016 and June 2019. Data from the North American Contact Dermatitis Group from 2015 to 2016 and The Ottawa Hospital patch test clinic from 2000 to 2010 were also reviewed.

Main message Topical antibiotics are the most common cause of ACD to medicaments and frequently cause cosensitization to multiple allergens. This hypersensitivity reaction is often seen following surgical procedures and should be distinguished from postoperative infection. Corticosteroid allergy is easy to miss and should be suspected in cases of corticosteroid-sensitive dermatoses that worsen despite appropriate treatment. Topical anesthetics and propylene glycol are other causes of ACD found in many prescription and over-the-counter products.

Conclusion Allergic contact dermatitis is easy to miss and should always be considered in cases of eczematous eruptions. A thorough drug history including all topical products—both prescription and over-the-counter—is critical. Patch testing can help identify specific allergens for the patient to avoid.

Vol 67: JUNE | JUIN 2021 | Canadian Family Physician | Le Médecin de famille canadien 415

Clinical Review

A llergic contact dermatitis (ACD) caused by topi-cal medicaments is a common and underrecog-nized entity in the family practice setting. The

differential diagnosis of ACD might be overlooked for a number of reasons, including a lack of awareness of the characteristic features of ACD, anchoring bias toward the original concern for which a medicament was applied, confusion with other causes such as infec-tion, or an incomplete drug history overlooking prescrip-tion or over-the-counter topical products. This article aims to overcome this problem by identifying the types of topical medicament ingredients that cause ACD and by illustrating common scenarios in which each of these might present.

Case descriptionA 37-year-old woman requested elective excision of a benign mole on the right preauricular skin. She was referred to a surgeon who excised the lesion and instructed her to apply ointment daily for moist wound healing. She developed increased erythema, edema, vesiculation, and crusting at the surgical site 1 day after applying an over-the-counter polymyxin B and bacitracin ointment (Figure 1). She described itch but no pain. What would be your next steps in managing this patient?

Sources of informationAfter obtaining approval through the University of British Columbia Research Ethics Board, we performed a

retrospective chart review examining patients who had positive patch test reactions to selected topical medica-ment ingredients at the Contact Dermatitis Clinic at St Paul’s Hospital in Vancouver, BC, from November 2016 to June 2019. More than 800 patients (N = 821) were patch tested to an 80-allergen screening series, plus supplemental allergens when indicated, with readings at 48 hours and a delayed reading at day 8.

We also reviewed patch test data collected by the North American Contact Dermatitis Group (NACDG) between 2015 and 20161 and a retrospective chart review from The Ottawa Hospital patch test clinic in Ontario focusing on ACD to topical medicament ingre-dients from patients patch tested between January 2000 and September 2010.2

Main messageAllergic contact dermatitis is a type IV delayed hyper-sensitivity reaction that occurs when an allergen comes in direct contact with the skin and results in a pruritic eczematous eruption.3 A prior episode of sensitization is required, although this can occur with as little as a single application and is not always evident on history. Sensitization typically takes at least 2 weeks to develop. In acute cases, ACD might present with vesicles, crust-ing, and weeping, whereas chronic cases lead to scaly or lichenified papules and plaques. These changes often appear in a well-circumscribed area where contact with the allergen occurred. However, they can become more generalized and affect nonexposed sites (ie, autoec-zematization). Allergic contact dermatitis to topi-cal medicaments is a frequently overlooked entity that is increasingly relevant to family physicians, as rates appear to be on the rise.4-6

Topical medicaments encompass an array of prod-ucts available both by prescription and over the coun-ter. While they frequently come in the form of creams or ointments, topical medicament ingredients might also be constituents of other products such as eye and ear drops, dressings, or powders. All of these have the potential to cause adverse effects, which include not only ACD but also irritant contact dermatitis or contact urticaria (hives). A comprehensive drug his-tory is necessary whenever there is suspicion that a patient’s skin condition might have been caused by topical medicament ingredients.

We divided the ingredients of topical medicaments into 4 categories: antibiotics, anesthetics, cortico-steroids, and propylene glycol—a component of the vehicle of many topical medicaments. Of 821 patients patch tested, 113 (13.8%) had positive reactions to at least 1 medicament ingredient in these 4 categories. Table 1 summarizes the number of positive patch test results for the 12 topical medicament ingredients included in our study. Table 2 compares the num-ber of positive patch test results for selected topical

Figure 1. Increased erythema, edema, vesiculation, and crusting at a preauricular surgical site in a 37-year-old woman 1 day after applying an over-the-counter polymyxin B and bacitracin ointment

416 Canadian Family Physician | Le Médecin de famille canadien } Vol 67: JUNE | JUIN 2021

Clinical Review

medicament ingredients in our clinic with results from The Ottawa Hospital patch test clinic and the NACDG.

Antibiotics. As illustrated in our case, topical antibi-otics are a common cause of ACD. In our cohort, the 2 common contact allergens were bacitracin and neo-mycin. Combined with the third antibiotic, polymyxin B, these make up 50.7% of the positive patch test reac-tions in our study. This finding is consistent with those of similar studies, which also showed antibiotics to be the most frequent cause of medicament ACD.1,2 Among patch tested patients in North America, neomy-cin and bacitracin were the seventh and eighth most common allergens, respectively.1 Of interest, over-the-counter antibiotics have been shown to be more sen-sitizing compared with their prescribed counterparts.7

Geographic variability has been seen in the sensitization rates to various topical antibiotics, such as higher rates of neomycin allergy in the United States compared with Europe and Canada,8,9 which is also supported by our data. This variability is likely owing to differing prescrib-ing patterns and the commercial availability of various formulations of over-the-counter antimicrobial products. Products with neomycin are available in Canada by pre-scription but not over the counter.

The scenario in our clinical case is frequently encountered owing to the use of topical antibiotics following surgical procedures. We recommend pet-rolatum or other bland emollients for postopera-tive wound care, which have been shown to have an equally low infection rate to bacitracin and other topical antibiotics, without the risk of allergic induc-tion.10-14 Nonetheless, family physicians should be aware that it is still common practice for many physi-cians to recommend topical antibiotics.

Topical antibiotics frequently cause cosensitization, a phenomenon in which a patient becomes sensitized to 2 chemically distinct allergens found in the same source. One common example of this is polymyxin B and baci-tracin, which are often found together in over-the-counter antibiotic ointments. In our cohort, of 13 patients allergic to polymyxin B, 5 (38.5%) were also allergic to bacitracin, and of 33 patients allergic to bacitracin, 5 (15.2%) were also allergic to polymyxin B. Overall, the 3 medicament ingredients with the high-est rates of cosensitization were the 3 antibiotics neo-mycin (22.9%), bacitracin (20.0%), and polymyxin B (18.6%). Other studies also demonstrate high levels of antibiotic cosensitization.2

Topical medicament ingredients can also induce cross-reactions. These occur when patients become sen-sitized to 2 or more different allergens through multiple sources owing to the allergens sharing a similar chemi-cal structure. Rarely, the consequences of these cross-reactions can be severe. As noted in one case report, a patient with a known contact allergy to neomycin devel-oped erythroderma following intravenous gentamicin.15 Both medicament ingredients in this case belonged to the deoxystreptamine class of aminoglycoside antibiot-ics, which are known to have an approximately 50% rate of cross-reactivity.16 Other severe reactions to be aware of include anaphylaxis to bacitracin, which has been described in several case reports.17-19 When prescribing medications, it is important to take a thorough history of the patient’s previous reactions to allergens and to con-sider the possibility of cross-reaction.

Patch testing to a topical antibiotic product and its individual ingredients can help identify the spe-cific allergens to which patients are sensitized. As patch testing is ideally performed on unaffected skin, patients might not generate a reaction to the commer-cial product itself but should react to any individual

Table 1. No. of positive patch test reactions (n = 142 total) for the 12 topical medicament ingredients included in our chart review: Twenty-five of the 113 patients had positive patch test results to > 1 medicament ingredients (n = 23 had 2; n = 2 had 4).

ALLERGEN N % TESTS % PATIENTS

Bacitracin 33 23.2 29.2

Neomycin 26 18.3 23.0

Lidocaine 19 13.4 16.8

Propylene glycol 15 10.6 13.3

Polymyxin B 13 9.2 11.5

Tixocortol-21-pivalate* 12 8.5 10.6

Benzocaine 10 7.0 8.8

Budesonide* 8 5.6 7.1

Desoximetasone* 2 1.4 1.8

Hydrocortisone-17-butyrate* 2 1.4 1.8

Triamcinolone acetonide* 1 0.7 0.9

Clobetasol-17-propionate* 1 0.7 0.9

*Topical corticosteroid.

Table 2. Percentages of patients with positive patch test reactions to selected topical medicament ingredients

ALLERGENVANCOUVER, % (N = 821)

OTTAWA,* % (N = 100)

NACDG 2015-2016, % (N = 5597)

Bacitracin 4.0 44 6.9

Neomycin 3.2 29 7.0

Lidocaine 2.3 12 0.8

Propylene glycol 1.8 Not analyzed 2.8

Tixocortol-21-pivalate

1.5 19 2.1

NACDG—North American Contact Dermatitis Group.*The Ottawa cohort included only patients with positive reactions to ≥ 1 topical medicament ingredients.

Vol 67: JUNE | JUIN 2021 | Canadian Family Physician | Le Médecin de famille canadien 417

Clinical Review

ingredient to which they are sensitized, as these are prepared at a sufficiently high concentration to pro-voke a reaction even on intact skin.

Corticosteroids. Allergic contact dermatitis to topical corticosteroids is an especially difficult diagnostic chal-lenge. Patients might be allergic to the corticosteroid itself or an excipient such as propylene glycol (dis-cussed below). Patients often have a history of atopic dermatitis that might coexist alongside ACD and be difficult to distinguish clinically. In our cohort, 54.2% of patients with corticosteroid allergy had a history of atopy and 61.5% had previous steroid exposure. Corticosteroid allergy is often not considered, as these medications are typically the treatment of choice for eczematous dermatitis. This might lead to increased use and further worsening of the patient’s dermatitis.

Allergic contact dermatitis should be suspected in cases of corticosteroid-sensitive skin disease that does not respond—or that worsens—with appropriate corti- costeroid treatment. Even in an allergic patient, corticosteroids will exert their anti-inflammatory effect, which can lead to an “edge effect” where the periph-ery of treated lesions is more active due to lower con-centrations of the corticosteroid. Corticosteroid-allergic patients might also have severe flaring of their condition when discontinuing their topical corticosteroids, as the benefit of the anti-inflammatory effect is immediately removed but the delayed hypersensitivity reaction per-sists for some time.20

Owing to corticosteroid allergy’s nebulous clini-cal presentation, patch testing is critical to confirm the diagnosis. Six corticosteroid allergens (Table 1) are currently on our 80-allergen screening series. Corticosteroids can be classified into 3 groups based on their chemical structure and reaction patterns.20 If an allergy is identified within one group, corticosteroids from another group can usually be used safely. Cross-reactivity between groups is very rare and most aller-gies are to corticosteroids in group 1 (hydrocortisone type). In our cohort, corticosteroids made up 18.3% of the positive reactions to all topical medicament ingredients, with the most common reaction being to tixocortol-21-pivalate (group 1) (8.5%). There were 2 cases of cross-reactivity between corticosteroids that were all in group 1. There were several cases of cosen-sitization with topical anesthestics and propylene glycol.

Anesthetics. Topical anesthetics are found in a variety of prescription and over-the-counter products. Many family physicians are aware of their presence in anti-itch or numbing preparations but might not realize that they are also found in products for sunburn, aphtha, ulcers, hemorrhoids, earaches, and toothaches. Other sources that might be surprising include denture adhe-sive, personal lubricants, lozenges, and hair removal

products (eg, wax and shaving creams). Anesthetics might also be found in combination with topical antibi-otics and corticosteroids, which can lead to cosensitiza-tion as described above.

Lidocaine was the third most common allergen in our cohort (2.3%). This is increased compared with the incidence of 0.8% seen for patch tested patients in North America between 2015 and 20161 and 0.77% for patients at the University of British Columbia Contact Dermatitis Clinic between 2009 and 2013.21 The change could reflect increased availability and use of lidocaine-containing products. Benzocaine allergy was also found in 1.2% of our cohort. Of 29 patients with topical anes-thetic allergy, 6 showed cosensitization with a topical antibiotic (20.7%) and 2 showed cosensitization with a topical steroid (6.9%).

While patch testing can confirm a contact allergy to topical anesthetics applied directly to the skin, these patients will not necessarily develop an allergic reac-tion to the same anesthetic injected into the dermis or subcutis, as is frequently done for dermatologic and dental procedures. In a study of patients with contact allergy to lidocaine, 2 of 11 tested had reactions to sub-cutaneous challenge and none had reactions to intra-dermal testing.22

Propylene glycol. Propylene glycol is an excipient found in topical medicaments and other personal prod-ucts such as cosmetics, soaps, sunscreens, etc. It is often used as a softening agent, solvent, moisturizer, and preservative. Propylene glycol allergy was found in 1.8% of our cohort and 2.8% of patients patch tested by the NACDG between 2015 and 2016, an increase from 2.2% in 2013 to 2014.1 Eighty percent of our patients with propylene glycol allergy had a history of atopic der-matitis, which might be explained by propylene glycol’s presence in topical eczema treatments including mois-turizers, topical corticosteroids, and topical nonsteroidal anti-inflammatory products, pimecrolimus cream, and crisaborole ointment.

As with corticosteroid allergy, propylene glycol–induced ACD can be difficult to diagnose, and should be suspected in patients with eczematous eruptions that do not respond appropriately to treatment with topical corticosteroids or anti-inflammatory drugs that con-tain propylene glycol. Allergic contact dermatitis due to propylene glycol in nonmedicament products is beyond the scope of this article, but it requires a thorough investigation of the patient’s use of personal products. Patch testing should be undertaken if the diagnosis is suspected. In patients with confirmed allergy who require topical anti-inflammatory drugs, a limited num-ber of propylene glycol–free topical corticosteroids can be used, in addition to calcineurin inhibitor (tacroli-mus) ointment. An article by Al Jasser et al in Skin Therapy Letter6 has a quick reference guide to propylene

418 Canadian Family Physician | Le Médecin de famille canadien } Vol 67: JUNE | JUIN 2021

Clinical Review

glycol–free topical corticosteroids available in Canada, classified by manufacturer, structural class, potency, and vehicle.

Case resolutionBased on the timing and clinical presentation of pruritic erythematous plaques with vesicles and crusting, a diagnosis of ACD to 1 or more of the components of the patient’s ointment was suspected. Infection was less likely, owing to the lack of puru-lent discharge and pain or tenderness. Her ointment was discontinued and she was prescribed 0.05% clo-betasol propionate ointment. After a few days of use, her condition improved dramatically and she was no longer symptomatic (Figure 2).

She was patch tested to her ointment and 5 of its ingredients. The patches were left on for 48 hours and readings were performed at 48 and 72 hours, showing positive reactions (papular erythema) to the ointment and bacitracin (Figures 3 to 6). The patient was counseled to avoid bacitracin-containing products in the future.

ConclusionAllergic contact dermatitis is a commonly overlooked diagnosis that should be part of the differential diag-nosis for any pruritic eczematous eruption. Among topical medicament ingredients, antibiotics are the most common culprit, often in the context of postoperative

wound care, where ACD might be mistaken for infection. Antibiotics are also the most frequent cause of cosensi-tization. Corticosteroid and propylene glycol allergy are

Figure 2. The patient’s preauricular skin after her combination polymyxin B and bacitracin ointment was discontinued and she was prescribed 0.05% clobetasol propionate ointment

Figure 3. The patient’s arm before patch testing

Figure 4. The patient was patch tested to her over-the-counter combination polymyxin B and bacitracin ointment and 5 of its ingredients

Vol 67: JUNE | JUIN 2021 | Canadian Family Physician | Le Médecin de famille canadien 419

Clinical Review

both difficult to diagnose and should be suspected in cases with clinical worsening despite appropriate treat-ment or when substantial flaring occurs with discontinu-ation of treatment. Topical anesthetics are found in an increasingly wide variety of prescription and over-the-counter products and in combination with topical antibi-otics and corticosteroids. A comprehensive drug history

including all topical products is therefore critical. Patch testing is indicated for anyone with suspected ACD to medicaments and can help guide patient counseling on appropriate allergen avoidance.

Mr Choi is a fourth-year medical student and Ms Vafaei-Nodeh is a third-year medi-cal student, both at the University of British Columbia in Vancouver. Dr Phillips is a fourth-year dermatology resident at the University of British Columbia. Dr de Gannes is Medical Director of the Contact Dermatitis Clinic at St Paul’s Hospital in Vancouver.

ContributorsDr de Gannes conceived the idea for the manuscript, provided the data from the Contact Dermatitis Clinic, analyzed and interpreted the data, and oversaw the project. Dr Phillips made substantial editorial revisions to the manuscript and analyzed and interpreted the data. Mr Choi performed the literature review, wrote and submitted the ethics application, extracted data, co-wrote the initial draft of the manuscript, analyzed and interpreted the data, and revised the manuscript. Ms Vafaei-Nodeh extracted data, co-wrote the initial draft of the manuscript, analyzed and interpreted the data, and revised the manuscript.

Competing interestsNone declared

References1. DeKoven JG, Warshaw EM, Zug KA, Maibach HI, Belsito DV, Sasseville D, et al. North

American Contact Dermatitis Group patch test results: 2015-2016. Dermatitis 2018;29(6):297-309.

2. Spring S, Pratt M, Chaplin A. Contact dermatitis to topical medicaments: a retrospective chart review from The Ottawa Hospital patch test clinic. Dermatitis 2012;23(5):210-3.

3. Nguyen HL, Yiannias JA. Contact dermatitis to medications and skin products. Clin Rev Allergy Immunol 2019;56(1):41-59.

4. Warshaw EM, Belsito DV, Taylor JS, Sasseville D, DeKoven JG, Zirwas MJ, et al. North American Contact Dermatitis Group patch test results: 2009 to 2010. Dermatitis 2013;24(2):50-9.

5. Goossens A, Medeiros S. Allergic contact dermatitis from topical medicaments. Expert Rev Dermatol 2008;3(1):37-42.

6. Al Jasser M, Mebuke N, de Gannes GC. Propylene glycol: an often unrecognized cause of allergic contact dermatitis in patients using topical corticosteroids. Skin Therapy Lett 2011;16(5):5-7.

7. Goh CL. Contact sensitivity to topical antimicrobials. (II). Sensitizing potentials of some topical antimicrobials. Contact Dermatitis 1989;21(3):166-71.

8. Elliott JF, Abbas M, Hull P, de Gannes G, Toussi R, Milani A. Decreasing rates of neomycin sensitization in western Canada. J Cutan Med Surg 2016;20(5):446-52. Epub 2015 Dec 23.

9. De Groot AC, Maibach HI. Frequency of sensitization to common allergens: compari-son between Europe and the USA. Contact Dermatitis 2010;62(6):325-9.

10. Smack DP, Harrington AC, Dunn C, Howard RS, Szkutnik AJ, Krivda SJ, et al. Infection and allergy incidence in ambulatory surgery patients using white petrolatum vs bacitracin ointment. A randomized controlled trial. JAMA 1996;276(12):972-7.

11. Draelos ZD, Rizer RL, Trookman NS. A comparison of postprocedural wound care treatments: do antibiotic-based ointments improve outcomes? J Am Acad Dermatol 2011;64(3 Suppl):S23-9. Epub 2011 Jan 17.

12. Campbell RM, Perlis CS, Fisher E, Gloster HM Jr. Gentamicin ointment versus petrola-tum for management of auricular wounds. Dermatol Surg 2006;31(6):664-9.

13. Stout EI, McKessor A. Glycerin-based hydrogel for infection control. Adv Wound Care (New Rochelle) 2012;1(1):48-51.

14. Neuman MG, Nanau RM, Oruña-Sanchez L, Coto G. Hyaluronic acid and wound heal-ing. J Pharm Pharm Sci 2015;18(1):53-60.

15. Guin JD, Phillips D. Erythroderma from systemic contact dermatitis: a complication of systemic gentamicin in a patient with contact allergy to neomycin. Cutis 1989;43(6):564-7.

16. Childs-Kean LM, Shaeer KM, Varghese Gupta S, Cho JC. Aminoglycoside allergic reac-tions. Pharmacy 2019;7(3):124.

17. Lin FL, Woodmansee D, Patterson R. Near-fatal anaphylaxis to topical bacitracin ointment. J Allergy Clin Immunol 1998;101(1 Pt 1):136-7.

18. Vale MA, Connolly A, Epstein AM, Vale MR. Bacitracin-induced anaphylaxis. Arch Dermatol 1978;114(5):800.

19. Blas M, Briesacher KS, Lobato EB. Bacitracin irrigation: a cause of anaphylaxis in the operating room. Anesth Analg 2000;91(4):1027-8.

20. Baeck M, Goossens A. Immediate and delayed allergic hypersensitivity to corticosteroids: practical guidelines. Contact Dermatitis 2012;66(1):38-45. Epub 2011 Oct 5.

21. To D, Kossintseva I, de Gannes G. Lidocaine contact allergy is becoming more preva-lent. Dermatol Surg 2014;40(12):1367-72.

22. Corbo MD, Weber E, DeKoven J. Lidocaine allergy: do positive patch results restrict future use? Dermatitis 2016;27(2):68-71.

This article is eligible for Mainpro+ certified Self-Learning credits. To earn credits, go to www.cfp.ca and click on the Mainpro+ link. This article has been peer reviewed. Can Fam Physician 2021;67:414-9. DOI: 10.46747/cfp.6706414 Cet article se trouve aussi en français à la page 420.

Figure 5. The patient’s arm after the patches were left on for 48 h

Figure 6. The patient’s arm 72 h after patch testing, showing positive reactions (papular erythema) to the over-the-counter polymyxin B and bacitracin ointment and bacitracin