British Journal of Dermatology 2001; 145: 597±601. Photoallergic contact dermatitis is uncommon A.DARVAY, I.R.WHITE, R.J.G.RYCROFT, A.B.JONES, J.L.M.HAWK AND J.P.McFADDEN Department of Environmental Dermatology, St John's Institute of Dermatology, St Thomas' Hospital, Lambeth Palace Road, London SE1 7EH, U.K. Accepted for publication 31 January 2001 Summary Background Despite the enormous increase in sunscreen use, allergic contact (AC) and photoallergic (PA) reactions to ultraviolet (UV) filters are considered rare. Objectives To analyse the data from 2715 patients who underwent photopatch testing at St John's Institute of Dermatology during the period 1983±98. Methods A retrospective analysis of all positive photopatch test episodes was undertaken with the results retrieved from the environmental dermatology database and further verified with the original archived patch test documentation for each individual patient. Results In 111 patients with positive reactions (4 1%), there were 155 AC or PA reactions to allergens in the photopatch test series. Eighty PA reactions were observed in 62 (2 3%) patients (32 men and 30 women, age range 28±75 years), with UV filters accounting for 52 positive reactions (65%), drugs 16 (20%), musk ambrette 11 (14%) and the antiseptic trichlorocarbanilide one (1%). The most common UV filter photoallergen was benzophenone-3 with 14 positive results, followed by benzophenone-10 (n ˆ 9), isopropyl dibenzoylmethane (n ˆ 6), p-aminobenzoic acid (PABA) (n ˆ 5), octyl dimethyl PABA (n ˆ 5), butyl methoxydibenzoylmethane (n ˆ 4), isoamyl methoxycinnamate (n ˆ 2), ethyl methoxycinnamate (n ˆ 2), octyl methoxycinnamate (n ˆ 2), amyl dimethyl PABA (n ˆ 2) and phenylbenzimidazole sulphonic acid (n ˆ 1). A similar number of AC reactions to UV filters was detected in this study. Thus 49 patients (1 8%) had a total of 75 reactions: 51 due to UV filters and 24 as a result of exposure to fragrances and therapeutic agents. Benzophenone-10 accounted for 13 AC reactions and benzophenone-3 for eight reactions. Twentytwo patients had a PA reaction alone, whereas 19 patients had chronic actinic dermatitis and 15 patients polymorphic light eruption (PLE) in addition. Thus, 34 of the 62 patients (55%) had a preceding underlying photodermatosis. Conclusions These results show a low yield of positive photopatch tests. Thus, despite the large increase in the use of UV filters over the last decade, the development of PA reactions remains rare. Furthermore, most of the common UV filter photoallergens identified in this study, including PABA, amyl dimethyl PABA and benzophenone-10, are now rarely used in sunscreen manufacture, while isopropyl dibenzoylmethane was voluntarily removed from the market in 1993. Currently, benzophenone-3 is the commonest contact photoallergen still in widespread use. In contrast, the UVB filter octyl methoxycinnamate, used in a number of sunscreens, produced only two positive PA reactions in 12 years of testing. Nevertheless, although these reactions are extremely rare, patients with photodermatoses such as PLE and chronic actinic dermatitis do represent a group of patients at increased risk of developing photoallergy. Further photopatch test series should be regularly reviewed and updated, as the relevance of individual photoallergens changes over time. Currently, there is no evidence that PA reactions represent a common clinical problem. Key words: benzophenone-3, photoallergy, photodermatosis, photopatch test, sunscreens, ultraviolet filters Correspondence: Dr A.Darvay. E-mail: amritdarvay@doctors.org.uk q 2001 British Association of Dermatologists 597
598 A.DARVAY et al. In the last 10 years there has been a great increase in the use of ultraviolet (UV) filters, with the main aim of alleviating the carcinogenic potential of solar radiation. These agents are also added to cosmetic products such as moisturizers and lipsticks to try and reduce premature ageing of the skin and to prolong their shelf life by protecting against photodegradation. As a result of this increase in consumer exposure to UV filters, several reports have appeared in the literature regarding allergy and photoallergy to sunscreens. 1±5 UV filters can be classified into two major types: chemical and physical. 6 The physical UV filters, zinc oxide and titanium dioxide, reflect and scatter UV radiation and are not sensitizers, whereas chemical UV filters reduce the amount of UV radiation reaching the stratum corneum by absorption, and can cause sensitization. The chemical UVB filters include p-aminobenzoic acid (PABA) and its ester derivatives octyl dimethyl PABA and amyl dimethyl PABA, and cinnamate derivatives. Chemicals absorbing UVB and some UVA radiation include the benzophenones. Butyl methoxydibenzoylmethane predominantly absorbs UVA radiation. Microfine titanium dioxide can, however, reflect and absorb UV radiation. As the chemical UV filters are used in relatively high concentrations, it has been postulated that sensitization may occur fairly readily, although several studies have not, in fact, recorded any significant allergy or photoallergy to these UV filters. 7,8 Other photoallergens are of predominantly historical interest. Antibacterial agents used in soaps, e.g. Table 1. Ultraviolet filters, allergens and photoallergens with International Nomenclature of Cosmetic Ingredients (INCI) and trade names contained in the light series at St John's Institute of Dermatology INCI name Trade name February 1985±May 1986 PABA Benzophenone-10 Mexenone Amyl dimethyl PABA Escalol 506 June 1986±March 1987 PABA Benzophenone-10 Isopropyl dibenzoylmethane a Eusolex 8020 Butyl methoxydibenzoylmethane Parsol 1789 Octyl dimethyl PABA Escalol 507 Octyl methoxycinnamate Parsol MCX Phenyl methyl benzoxazol Witisol Patch test concentration 2% in petrolatum. a Isopropyl dibenzoylmethane was usually used in conjunction with 3-(4-methylbenzylidene) camphor (Eusolex 6300) and the trade name of this combination was Eusolex 8021. PABA, p-aminobenzoic acid. halogenated salicylanilides such as trichlorocarbanilide and chlorinated phenols such as fentichlor, were found to be potent photoallergens in the 1960s and 1970s and are now seldom used. Probably the best-known photoallergen, the fragrance musk ambrette, 9 was long ago removed from the market-place, although it remains in the photopatch test series, as it can still be found in products from Asia and may cause prolonged photosensitivity reactions. This retrospective study was undertaken to assess the results of photopatch testing in 2715 patients over a 15-year period (1983±98). Our aims were to assess the percentage of patients with photoallergic (PA) reactions, their underlying diagnoses and the number of photoallergens, paying particular attention to UV filters, which constitute the majority of the photopatch test series. Subjects and methods Patients referred for patch and photopatch testing were under investigation for a presumed photosensitivity disorder, gave a history of a worsening itchy rash on exposed sites, exhibited signs of a chronic eczematous eruption on exposed sites or were suspected of having chronic actinic dermatitis (CAD). Patients who gave a history suggestive of allergic contact (AC) or PA reactions were included and tested with relevant contact allergens and photoallergens as shown in Tables 1±3. On day 0 the standard photoallergens were applied to the patient's back in duplicate. On day 2 the patches were removed and one series irradiated with 5 J cm 22 of broadband UVA from wall-mounted Philips TL 44D 25\09N fluorescent tubes, although 2 5 J cm 22 was used if, from the history, there were Table 2. St John's Institute Contact Dermatitis Clinic scheme of allergens for photopatch testing, 1991 (`light series') Concentration Allergen in petrolatum (%) Benzocaine 5 Chlorpromazine 0 1 Musk ambrette 5 Promethazine 1 Benzophenone-3 2 Benzophenone-10 2 Butyl methoxydibenzoylmethane 2 Ethyl methoxycinnamate 2 Isopropyl dibenzoylmethane 2 3-(4-methylbenzylidene) camphor 2 Octyl dimethyl PABA 2 Octyl methoxycinnamate 2 PABA 2 PABA, p-aminobenzoic acid.
PHOTOALLERGIC CONTACT DERMATITIS IS UNCOMMON 599 Table 3. St John's Institute Contact Dermatitis Clinic scheme of allergens for photopatch testing, 1998 (`light series') Concentration Allergen in petrolatum (%) Butyl methoxydibenzoylmethane 10 Isoamyl methoxycinnamate 10 Ethyl dimethyl aminobenzoate 10 Musk ambrette 5 Octyl methoxycinnamate 10 Phenyl benzimidazole sulphonic acid 10 p-aminobenzoic acid 10 Benzophenone-3 10 Benzophenone-4 10 3-(4-Methybenzylidene) camphor 10 clear episodes of severe photosensitivity, or the patient was thought to have CAD, as very severe sunburn-type reactions could result if higher doses of UVA were used. In addition, between 1985 and 1987, 280 patients were tested with 1 J cm 22 as early studies had shown that to elicit musk ambrette PA reactions only a small dose was needed. 9 The control series was covered and UV exposure avoided. On day 4 the reactions were assessed using a standard visual scoring system based on the International Contact Dermatitis Research Group criteria. 10 A PA reaction was diagnosed when the irradiated sites were positive, with negative non-irradiated sites. An AC reaction was diagnosed when the irradiated sites and the non-irradiated sites were both positive. If a more strongly positive irradiated site compared with control was found, indicating a possible combined AC and PA reaction, the result was classified as AC if both sites had palpable erythema and PA if the control site showed erythema only and the irradiated site was palpable. Of the 2715 patients analysed, those with a positive AC or PA reaction had their details retrieved from the patch-test database. Of these, 111 patients had their original case notes reviewed to check the validity of the database results, to ascertain the underlying diagnosis and to determine how it was reached. Results Table 4 summarizes the positive results for UV filters, fragrances, therapeutic agents and antiseptics, for both PA and AC reactions. Of a total of 111 patients, 62 (2 3%; 32 men and 30 women, age range 28±75 years) had 80 positive PA reactions, of which 37% were to current relevant photoallergens. A further 49 (1 8%) patients had 75 positive AC reactions. Thus, there was a total of 155 positive reactions in 111 patients. Further analysis of the positive PA reactions Table 4. Photoallergic (PA) and allergic contact (AC) reactions to ultraviolet (UV) filters, fragrances, therapeutic agents and antiseptics PA AC UV filters Benzophenone-3 14 8 Benzophenone-10 9 13 Isopropyl dibenzoylmethane 6 4 PABA 5 7 Octyl dimethyl PABA 5 3 Butyl methoxydibenzoylmethane 4 3 Isoamyl methoxycinnamate 2 3 Ethyl methoxycinnamate 2 2 Amyl dimethyl PABA 2 4 Octyl methoxycinnamate 2 4 Phenylbenzimidazole 1 0 sulphonic acid Fragrances Musk ambrette 11 4 Therapeutic agents Chlorpromazine 6 11 Promethazine 10 9 Antiseptics Trichlorocarbanilide 1 0 Total 80 75 PABA, p-aminobenzoic acid. showed 52 (65%) to be due to UV filters, 11 (14%) to musk ambrette, 16 (20%) to chlorpromazine and promethazine, and one (1%) to trichlorocarbanilide. Of the 75 AC reactions, 51 were due to UV filters (68%), four to fragrances (5%) and 20 to drugs (27%). The most common UV filter photoallergen was benzophenone-3 with 14 positive results, followed by benzophenone-10 (n ˆ 9), isopropyl dibenzoylmethane (n ˆ 6), PABA (n ˆ 5), octyl dimethyl PABA (n ˆ 5), butyl methoxydibenzoylmethane (n ˆ 4), isoamyl methoxycinnamate (n ˆ 2), ethyl methoxycinnamate (n ˆ 2), amyl dimethyl PABA (n ˆ 2), octyl methoxycinnamate (n ˆ 2) and phenylbenzimidazole sulphonic acid (n ˆ 1). The benzophenones accounted for 23 of 52 positive PA reactions (44%) and 21 of 51 positive AC reactions (41%). Table 5 shows a breakdown of the final diagnoses after photopatch testing in the 62 patients with PA reactions. Over half of these patients had prior underlying photodermatoses, in particular CAD (19 patients) and polymorphic light eruption (PLE; 15 patients). Table 5. Final diagnoses after photopatch testing in the 62 patients with photoallergic reactions Diagnosis n Photoallergic contact dermatitis alone 22 Plus chronic actinic dermatitis 19 Plus polymorphic light eruption 15 Plus atopic eczema 2 Unknown 4
600 A.DARVAY et al. Discussion This retrospective study assessed our experience of photopatch testing in 2715 patients undergoing investigation for suspected photosensitivity during the period 1983±98. On the rare occasions that chemicals elicited a positive PA response, UV filters, in particular benzophenone-3 and benzophenone-10, were the most common. However, benzophenone-10 and other UV filters represented in the study, including PABA and amyl dimethyl PABA, are now rarely used in sunscreen manufacture, while isopropyl dibenzoylmethane was voluntarily removed from the market in 1993. Octyl methoxycinnamate remains a commonly used UVB filter found in many commercial sunscreens, but only two PA reactions were found despite its inclusion in the photopatch test series since 1986. Overall, 62 (2 3%) patients with suspected photoallergy exhibited a PA reaction, and of these, 65% were due to UV filters, with benzophenone-3 being the commonest photoallergen with 14 positive results. It is the most commonly used UV filter in the U.S.A., with absorption spectra in both the UVA and UVB wavelengths. Forty-nine patients (1 8%) had a total of 75 AC reactions and, of these, 51 were due to UV filters. These results suggest that UV filter photoallergy and contact allergy are rare in clinical practice, despite an enormous increase in their use over the last 15 years. Certain photoallergens are of more historical interest, as they have been removed from the market, and consequently the British Photodermatology Group (BPG) has published a reduced standard series of six photoallergens incorporating important sensitizers. 11 This series includes butyl methoxydibenzoylmethane, benzophenone-3, octyl methoxycinnamate, octyl dimethyl PABA, musk ambrette and PABA. However, over the years our series has remained more detailed in order to monitor the emergence of potential photoallergens (Tables 1±3). A U.K. study by English et al. 1 found that 5% of 280 patients photopatch tested with UV filters had a positive AC reaction, although only three of these patients had a PA reaction. Thirteen years later, our study from the same unit reports the rate to be similar (4 1%), with 2 3% of patients now having a positive PA reaction. Bell and Rhodes also recently published their experience of photopatch testing 167 patients to the BPG standard photopatch test series. 12 They recorded 16 patients (10%) with a total of 20 PA reactions and found eight positive reactions to butyl methoxydibenzoylmethane and five reactions to both benzophenone-3 and octyl methoxycinnamate. In a study from Singapore, 7 only two PA reactions were found in 61 patients over a 5-year period, while in a 7-year study from Sweden, 13 7 9% of patients studied showed positive reactions, 80% of which were PA. Journe et al. 4 reported 39 positive UV filter PA reactions from 370 patients (10 5%) over a period of 5 years. Nater and De Groot 14 reported a 1±2% incidence of side-effects resulting from UV filter use in the general population, with a low-to-medium risk index, and Wennersten et al. 15 found a UV filter reaction in only 0 1% of their patient population. Fischer and BergstroÈm 16 found a 5% UV filter allergy rate after evaluation of customers' complaints regarding sunscreens sold in Sweden. Most of these reactions were AC, contact urticaria and subjective irritation rather than true PA reactions. Some of these studies have also investigated AC reactions to UV filters, and once again there was a low risk of sensitization. This has also been confirmed in our study, where only 1 8% of patients tested showed an AC reaction to UV filters. However, there are reports suggesting that PA reaction rates are higher, ranging from 7% to 20% (mean 12 2%). 17±21 However, these studies have included a wide variety of compounds, including antiseptics such as fenticlor, diphenylphenol and bithionol, and fragrances such as Myroxylon pereirae (balsam of Peru) and a fragrance mix. Many of these compounds may have produced a phototoxic irritant reaction rather than a true PA response. Furthermore, many of them are now obsolete, and therefore these and similar chemicals have not been included in our series for many years. The current `light series' for photopatch testing (Table 3) at St John's Institute of Dermatology reflects the predominance of UV filters. Musk ambrette remains in the series even though it was withdrawn from the European market many years ago; however, sensitization occasionally occurs as it is still used in some fragrance products manufactured in Asia. The relevance of the PA reactions to promethazine and chlorpromazine could not be fully established and may have been phototoxic in origin; 22 these therapeutic agents have now been removed from our photopatch test series. Interestingly, over 50% of patients with photoallergy had an underlying photodermatosis, predominantly CAD and PLE. These patients were diagnosed in the photobiological unit at St John's Institute of Dermatology, using a monochromatic irradiation source (Applied Physics, U.K.) and solar-simulated irradiation (Oriel,
PHOTOALLERGIC CONTACT DERMATITIS IS UNCOMMON 601 U.K.). Other studies have shown a similar association. 10,23,24 The reasons for this are unclear, but the increase in sensitization may be due to repeated application on injured skin, thus increasing the risk of absorption and subsequent sensitization or, alternatively, patients with CAD may be more prone to sensitization, as characterized by their high contact sensitivity rate. 25 In addition, repeated application of high concentrations of UV filters is more likely in patients with photodermatoses, and so perhaps such individuals are at greater risk of sensitization. Therefore, in this group of patients the potential for PA sensitization to UV filters should be borne in mind during evaluation. This study on photopatch testing suggests that photoallergy is rare and, although UV filters are essentially the only currently relevant photoallergens, even these compounds have a very low potential for both sensitization and photosensitization. Furthermore, these results are from a population with suspected photodermatoses rather than the general population, and therefore the true incidence of PA reactions to sunscreen products is likely to be even lower. There are clearly both temporal and geographical variations in photoallergens to which an individual is likely to be exposed, and because of these factors there is no single simple series of allergens to screen for photoallergy in all countries. However, vigilance is required to detect sensitizing chemicals, and thus the photopatch test series will continue to evolve over the years, although the clinical relevance of PA responses is currently waning. Photopatch testing itself will continue to be a useful tool in distinguishing patients with suspected allergic airborne contact dermatitis, CAD and PA contact dermatitis. Thus, the patients at greatest potential risk appear to be those with photodermatoses such as PLE and CAD, and this should be borne in mind when assessing such individuals. With the removal of the common sensitizers, however, PA reactions are unlikely to pose a common clinical problem in the future, although we recommend continued vigilance and periodic updating of the photopatch test series to reflect the introduction of new UV filters. References 1 English JSC, White IR, Cronin E. Sensitivity to UV filters. Contact Dermatitis 1987; 17: 159±62. 2 Knobler E, Almeida L, Ruzkowski AM et al. Photoallergy to benzophenone. Arch Dermatol 1989; 125: 801±4. 3 Marguery MC, Bazex J. Photoallergic contact dermatitis due to 2-hydroxy, 4-methoxybenzophenone (oxybenzone): report of four cases. Br J Dermatol 1989; 121 (Suppl. 34): 59±60. 4 Journe F, Marguery MC, Rakotondrazafy J et al. UV filter sensitisation: a 5 year study. Acta Derm Venereol (Stockh) 1999; 79: 211±13. 5 Schauder S, Ippen H. Contact and photocontact sensitivity to sunscreens. Review of a 15-year experience and of the literature. Contact Dermatitis 1997; 37: 221±32. 6 Food and Drug Administration. UV filter drug products for overthe-counter human drugs: proposed safety, effective and labelling conditions. Fed Reg 1978; 43: 38206±69. 7 Ang P, Ng S, Goh CL. UV filter allergy in Singapore. Am J Contact Dermatitis 1998; 9: 42±4. 8 Dromboole SH, Maibach HI. UV filter agent intolerance: contact and photocontact sensitisation and contact urticaria. J Am Acad Dermatol 1990; 22: 1068±78. 9 Cronin E. Photoallergic reactions to musk ambrette. Contact Dermatitis 1984; 11: 88±92. 10 Fregert S. Manual of Contact Dermatitis, 2nd edn. Copenhagen: Munksgaard, 1981. 11 British Photodermatology Group. Workshop report. Photopatch testingðmethods and indications. Br J Dermatol 1997; 136: 371±6. 12 Bell HK, Rhodes LE. Photopatch testing in photosensitive patients. Br J Dermatol 2000; 142: 589±90. 13 Berne B, Ros AM. 7 years experience of photopatch testing with UV filter allergens in Sweden. Contact Dermatitis 1998; 38: 61±4. 14 Nater JP, De Groot AC. Unwanted Effects of Cosmetics and Drugs Used in Dermatology, 2nd edn. Amsterdam: Elsevier, 1985:360. 15 Wennersten G, Thune P, Jansen CT et al. Photocontact dermatitis: current status with emphasis on allergic contact photosensitivity (CPS) occurrence, allergens and practical phototesting. Semin Dermatol 1986; 5: 277±89. 16 Fischer T, BergstroÈm K. Evaluation of customers' complaints about UV filter cosmetics sold by the Swedish pharmaceutical company. Contact Dermatitis 1991; 25: 319±22. 17 DeLeo VA, Suarez S, Maso MJ. Photoallergic contact dermatitis: results of photopatch testing in New York, 1985 to 1990. Arch Dermatol 1992; 128: 1513±18. 18 Thune P, Jansen C, Wennersten G et al. The Scandinavian multicenter photopatch study 1980±85: final report. Photodermatol 1988; 5: 261±9. 19 Fotiades J, Soter NA, Lim HW. Results of evaluation of 203 patients for photosensivity in a 7.3 year period. J Am Acad Dermatol 1995; 33: 597±602. 20 HoÈlzle E, Neumann N, Hausen B et al. Photopatch testing: the 5- year experience of the German, Austrian and Swiss Photopatch Test Group. J Am Acad Dermatol 1991; 25: 59±68. 21 Menze J, Muller SA, Connolly SM. Photopatch testing: a six-year experience. J Am Acad Dermatol 1998; 18: 1044±7. 22 Wennersten G, Thune P, Brodthagen H et al. The Scandinavian multicenter photopatch study: preliminary results. Contact Dermatitis 1984; 10: 305±9. 23 Thune P, Eeg-Larsen T. Contact and photocontact allergy in persistent light reactivity. Contact Dermatitis 1984; 11: 98±107. 24 Addo HA, Frain-Bell W. Persistence of allergic contact sensitivity in subjects with photosensitivity dermatitis and actinic reticuloid syndrome. Br J Dermatol 1987; 117: 555±9. 25 Menage H, du P, Ross JS, Norris PG et al. Contact and photocontact sensitization in chronic actinic dermatitis: sesquiterpene lactone mix is an important allergen. Br J Dermatol 1995; 132: 543±7.