The increase in p-phenylenediamine allergy in Denmark is not explained by an increase in contact allergy to para group chemicals

p-Phenylenediamine (PPD) allergy is common in European dermatitis patients, as suggested by recent publications (1, 2). It is well established that patients with PPD allergy may display cross-reactivity to other chemicals, including benzocaine, procaine, p-aminobenzoic acid sunscreens, sulfonamides, N-isopropyl-N′-phenyl-p-phenylenediamine (IPPD), parabens, and azo and aniline dyes. We previously showed that the prevalence of PPD allergy increased steadily in Denmark from 1.4% in 1989 to 2.4% in 2007 (p = 0.052) (3). Other centres have reported similar increasing trends (4–6). The increase observed in Denmark can probably be explained by increased exposure to hair dyes. However, one may suggest other explanations, including: (i) Increased exposure to PPD-containing temporary henna tattoos; and (ii) Increased prevalence of contact allergy to other ‘para group’ chemicals from the baseline series (benzocaine, parabens, and IPPD), resulting in a ‘false’ increase in PPD allergy not related to hair dye exposure. A recent European multicentre patch test study showed that dermatologists infrequently judged temporary henna tattooing to be causative of PPD allergy (range 0–15.2%, median prevalence 3.9%, and weighted average 4.8%), suggesting that the putative increasing use of temporary henna tattoos may only partly explain the increasing prevalence of PPD allergy (1). This study aimed to test whether the increase in PPD allergy observed in Denmark between 1989 and 2007 was caused by a relative increase in positive patch test reactions to cross-reactants of PPD included in the European baseline series. Dermatitis patients who underwent routine patch testing with the European baseline series between 1 January 1985 and 31 December 2007 were included. The MOAHLFA (male, occupation, atopic dermatitis, hand eczema, leg dermatitis, facial dermatitis, age above 40 years) index was not routinely registered throughout the study period. Thus, MOAHLFA registrations date back to 1994 (but only to 2001 for ‘Facial dermatitis'). Testing was performed throughout the study period with benzocaine 5% in petrolatum, 1% PPD free base in pet., paraben mix 16% in pet., and IPPD 0.1% in pet., using Finn Chambers® (8 mm, Epitest Ltd, Oy, Finland) on Scanpor tape® (Norgesplaster A/S, Alpharma, Vennesla, Norway). Patch test ingredients were delivered from Hermal (Reinbek, Germany). Patch tests were applied to the upper back and were occluded for 2 days. Readings were performed on day 2, on day 3 or day 4, and on day 7, according to the recommendation from the International Contact Dermatitis Research Group (7). Thus, homogeneous redness and infiltration in the entire test area was scored as a 1+ reaction. Homogeneous redness, infiltration and vesicles in the test area was scored as a 2+ reaction, and homogeneous redness, infiltration and coalescing vesicles in the test area was scored as a 3+ reaction. A 1+, 2+ or 3+ reading was interpreted as a positive response. An irritant response, a doubtful (+?) reading or a negative reading was interpreted as a negative response. Data analyses were performed with spss (SPSS, Chicago, IL, USA) for Windows (release 15.0). We investigated whether the increase in PPD allergy could be explained by a relative increase in contact allergy to other para group chemicals. Thus, the number of positive patch test reactions to benzocaine, paraben mix and IPPD in patients with PPD allergy was divided by the number of positive patch test reactions to PPD, and the outcome was then stratified by test year. Furthermore, a logistic regression model with ‘PPD allergy’ as the dependent variable and ‘test-year’ and ‘para group contact allergy’ as the explanatory variables was performed. In this model, an interaction term between study year and para group contact allergy was inserted to test whether the association of PPD and para group contact allergy had changed over the study period. A total of 15 095 dermatitis patients (64.3% women and 35.7% men) aged 4–99 years were patch tested between 1985 and 2007. Their main characteristics have been described previously (8). The MOAHLFA index showed that 12.5% had occupational dermatitis, 12.9% had atopic dermatitis, 31.7% had hand eczema, 9.3% had leg involvement, 19.3% had facial involvement, and 66.5% were above 40 years of age. The overall prevalences of PPD, benzocaine, paraben mix and IPPD allergy were 2.4%, 0.5%, 0.4%, and 0.5%, respectively. The number of positive patch test reactions to benzocaine, paraben mix or IPPD in PPD-allergic patients was divided by the number of positive patch test reactions to PPD and stratified by test year. No overall trends were observed (Fig. 1). Thus, random fluctuations were observed, with peak values in 1995 and 2005 and nadir values in 1991 and 1999. Temporal trends of relative association between para group chemicals and p-phenylenediamine (PPD) in PPD-allergic patients (i.e. the number of positive patch test reactions to benzocaine, paraben mix or N-isopropyl-N′-phenyl-p-phenylenediamine in PPD-allergic patients divided by the number of positive patch test reactions to PPD). A logistic regression analysis with ‘PPD allergy’ as the dependent variable and ‘test-year’, ‘para group allergy’ and an interaction term between test year and para group contact allergy as the explanatory variables did not reveal any significant changes in the association of PPD and para group contact allergy over the study period (p = 0.37). This study suggests that the increase in PPD allergy observed in Danish dermatitis patients between 1989 and 2007 was not explained by a relative increase in contact allergy to para group chemicals. Despite the relatively low number of positive patch test reactions to PPD and para group chemicals, this study is the first to investigate the relationship between positive patch test reactions to PPD and para group chemicals over time. Only a selection of para group chemicals was included in the data analysis. Thus, it is possible that the outcome could have been affected if more para group chemicals had been studied. Although Schnuch et al. recently found that about 12% of PPD-allergic patients had exposure related to shoes/textiles (defined as allergic contact dermatitis following contact with leather, gloves, shoes, or textiles) (9), the number of patients with contact allergy to Disperse Orange 3, Disperse Red 1 and Disperse Red 17 (azo dyes known to cross-react with PPD (10)) was low in our database. Nevertheless, data analysis revealed that the number of positive patch test reactions to Disperse Orange 3, Disperse Red 1 and Disperse Red 17 in PPD-allergic patients divided by the number of positive patch test reactions to PPD did not change over the study period (data not shown). Toluene-2,5-diamine contact allergy was excluded from data analysis, because it was only routinely tested from 2002. Another way to address the study aim would be to identify, in each patient with contact allergy to PPD and para group chemicals, causative exposures and their relative onset and clinical relevance. This would probably be better, but obtaining consistent quality data of this sort is difficult and time-consuming. Further studies are warranted, but we postulate that the increase in PPD allergy in Denmark is not explained by an increase in contact allergy to IPPD, benzocaine, and paraben mix.