Patch testing in patients with psoriasis: results of a 30-year retrospective study

Dear Editor, Psoriasis and allergic contact dermatitis (ACD) are common, chronic T‐lymphocyte‐mediated conditions. Although patients with psoriasis are commonly referred for patch testing, there are conflicting views on the relationship between contact allergy and psoriasis. Previous studies have suggested ACD is less common in psoriasis1 and that people with psoriasis are more difficult to sensitize.2,3,4 Others have suggested ACD is no less common in psoriasis than in other skin diseases (excluding atopic dermatitis).5 Our study was performed to ascertain if the frequency of ACD differed between individuals with psoriasis and those without. We performed a retrospective study of 2349 patients with psoriasis referred to our unit at the St John's Institute of Dermatology over a 30‐year period (1984–2014). Patients with psoriasis were identified through their history and a clinical examination performed by a dermatologist during the initial clerking within the cutaneous allergy clinic (prior to patch testing). Patch testing was performed using Finn Chambers® (SmartPractice, Phoenix, AZ, U.S.A.) on Scanpor® tape (SmartPractice) and readings were made on day 2/3 and 4/5 and labelled as positive if showing at least a ≥ 1+ reaction. Information was collected from our comprehensive local database on reactions to allergens contained within our extended European baseline series (as used during the pertinent periods). Our extended series contains several additional allergens that are routinely included in patch tests performed at our unit. Comparison was made with a group of patients without psoriasis (the nonpsoriasis group, n = 39 910) patch tested during the same time frame. There were 784 men and 1565 women in the psoriasis group and 15 532 men and 24 378 women in the nonpsoriasis group. Of note, 727 (30·9%) in the psoriasis group self‐reported coexistent atopy, compared with 12 177 (30·5%) in the nonpsoriasis group. Figure 1 and Table S1 (see Supporting Information) summarize our main findings. In the nonpsoriasis group, the commonest allergens were nickel (17·1%), fragrance mix I (7·1%), cobalt (6·0%), methylisothiazolinone (6·0%) and colophonium (4·2%). In the psoriasis group, these also represented the commonest allergens but in the following order: nickel (17·6%), fragrance mix I (7·0%), methylisothiazolinone (6·1%), cobalt (5·3%) and colophonium (4·2%). The only difference in the top 10 allergens for both groups was para‐phenylenediamine (PPD), which had a lower frequency of ACD in the psoriasis group (2·7%) than in the nonpsoriasis group (3·6%) and was replaced by fragrance mix II in the psoriasis group (3·4% psoriasis group, 2·3% nonpsoriasis group). There were no statistically significant differences in the frequency of ACD between the psoriasis and nonpsoriasis groups for each of the allergens in the extended baseline series (Fisher's exact test, assuming a 5% significance level). Comparison of contact allergy frequency in the psoriasis group for the 10 most common allergens in the nonpsoriasis group. Our study represents the largest analysis of patch testing in patients with psoriasis to date, conducted over an extended period of testing. We demonstrate that the rates of ACD are comparable between patients with psoriasis and individuals without psoriasis and are largely positive for the same allergens. A large Danish study previously found a significantly reduced likelihood of a positive patch test in patients with psoriasis.6 This was in comparison with a population of patients referred for patch testing [odds ratio (OR) 0·58, 95% confidence interval (CI) 0·49–0·68] and with individuals participating in a population‐based cross‐sectional study at three time points (OR 0·64, 95% CI 0·42–0·98). Individuals with psoriasis, like those with atopic dermatitis, are chronically exposed to topical therapies and may therefore be at higher risk of developing ACD. One study found 6·5% of patients with psoriasis were allergic to dithranol and the rate of ACD to topical medicaments and their additives was far higher in those with psoriasis than in a group of patients referred for patch testing.7 Our study focused on allergens within our extended European baseline series, and an increased rate of ACD in people with psoriasis to allergens outside these cannot be excluded. Identifying whether subgroups of patients with psoriasis are more likely to develop ACD should be considered in future work. Previous studies have highlighted an incidence of ACD in palmoplantar psoriasis that far exceeds that in other forms.8 Further considerations include looking at whether people with psoriasis are more likely to be multisensitizers and whether longevity of psoriasis influences the risk of developing ACD. To dissect the immunological overlap between ACD and psoriasis, one group analysed the effect of patch testing on psoriatic plaques in patients with known nickel sensitivity.9 They found that the ACD reaction was typical with a maximum intensity after 7 days and that allergen challenge did not affect the appearance of the underlying psoriatic plaque (as graded by the local Psoriasis Area and Severity Index score). This further underlines the hypothesis that psoriasis and ACD are driven by distinct and independent immune mechanisms, and would support our findings in which the presence of one disease does not affect the development of the other. Finally, individuals with psoriasis appear to react to a similar profile of allergens to those without psoriasis, further suggesting that people with psoriasis are not distinct in the manner in which they respond to cutaneous allergens.