Diagnosing lanolin contact allergy with lanolin alcohol and Amerchol L101

Background The prevalence of lanolin contact allergy in dermatitis patients varies from 1.2% to 6.9%. Different lanolin derivatives are used in patch testing. Objectives To determine which combination of lanolin derivatives is most effective in patch testing for the diagnosis of lanolin contact allergy. Methods A retrospective analysis of patients patch tested between 2016 and 2017 was performed. Patients were eligible if they had been tested with lanolin alcohol 30% pet., Amerchol L101 50% pet., and a supplementary series containing other lanolin derivatives. Lanolin alcohol and Amerchol L101 were tested in duplicate. Results Of 594 patients, 28.6% (95% confidence interval [CI]: 25.1%‐32.3%) had a positive patch test reaction to at least one lanolin derivative. Reactions to lanolin alcohol (14.7%, 95%CI: 11.3%‐18.2%) and Amerchol L101 (15.0%, 95%CI: 11.5%‐18.5%) were common in the routinely tested series. Reactions to other test preparations were significantly less frequent (P < 0.05). The addition of Amerchol L101 to lanolin alcohol significantly increased the number of positive cases (odds ratio 1.79, P < 0.001). Conclusions The combination of lanolin alcohol and Amerchol L101 is effective in patch testing for the diagnosis of lanolin contact allergy. Routinely testing with other lanolin derivatives may not be worthwhile, as it detects only a few additional patients.


| INTRODUCTION
Lanolin is a complex mixture of sterols (wool wax alcohols), fatty alcohols and fatty acids with a varying composition. 1 It is derived from a secretion of the sebaceous glands of sheep. [1][2][3] Because of its emollient properties, lanolin is used in cosmetic products and topical medicaments. 2,4 Allergic contact dermatitis caused by lanolin typically develops after repeated or prolonged topical exposure, especially on damaged skin. 5 Atopic dermatitis, leg ulcers and lower-extremity venous stasis dermatitis have been identified as risk factors for the development of lanolin contact allergy. 1,2,4,[6][7][8] The reported prevalence of lanolin sensitization in referred dermatitis patients has varied from 1.2% to 6.9% between several studies. 4,6,7,9,10 Lanolin alcohol 30% pet. is the standard patch test agent for diagnosing lanolin contact allergy, and has been included in the European baseline series (EBS) since 1969. 1,2,6,11 However, supplementary patch testing with other lanolin derivatives seems to improve the identification of lanolin-sensitive patients. 2,10,12 Particular attention is being payed to Amerchol L101 50% pet., which is a mixture of 10% lanolin alcohols and mineral oil. 7,11 Several studies reported more reactions to Amerchol L101 than to lanolin alcohol. 2,10,11,13 Moreover, in a recent multicentre study, 13 a group of 79 969 patients were simultaneously tested with both lanolin alcohol and Amerchol L101: more patients reacted only to Amerchol L101 (2.05%) than only to lanolin alcohol (1.19%) (P < 0.001). This raises the question of which lanolin products (eg, acetylated lanolin, hydrogenated lanolin, or ointments such as Eucerin) may also be used as patch test preparations. The frequencies of reactions to different lanolin derivatives have been studied, but there is limited information on the quality of these test preparations.
The clinical relevance-the responsibility of the putative allergen for the (current or past) dermatitis-and reliability are useful in assessing this quality. The reliability can be assessed by means of the reaction index (RI), introduced by Brasch in 1992, ranging from −1 to 1. 14 He proposed that an ideal patch test should have an optimal discriminatory power. A patch test is acceptable when the RI is >0, with a higher number of positive reactions than irritant and doubtful reactions. 14,15 The primary aim of this study was to determine which combination of lanolin derivatives is most effective in patch testing for the diagnosis of lanolin contact allergy. Therefore, we investigated the reaction prevalences, reliability and clinical relevance of individual lanolin derivatives. Moreover, we analysed the additional value of supplementary testing with lanolin alcohol.

| Patients
A retrospective analysis was performed on data of 594 patients who were patch tested in the VU University Medical Centre (VUmc) between January 1, 2016 and December 31, 2017. Patients were selected if they were routinely tested with the EBS, containing lanolin alcohol, and the routine supplementary series, containing Amerchol L101 (n = 594), and if they were additionally tested with our wool alcohol series, containing seven lanolin preparations, and/or with our topical medicament series, containing six lanolin preparations (Tables 1 and 2 17 Irritant reactions and doubtful (?+) reactions were combined and reported as "questionable" reactions, as it can be difficult to distinguish between these reactions. 6,17 The RI was calculated as (pq)/(p + q), with p being the number of positive reactions to a patch test, and q being the number of questionable reactions to a patch test.
The additional value of supplementary testing was calculated as the difference between reaction frequencies. The decision on relevance was based on criteria as proposed by Johansen et al 16 and Lachapelle et al. 17 Positive reactions with known clinical relevance were assessed as: "certain" when the clinician was convinced that the allergen was causative for the dermatitis; "probable" when there was a strong relationship between the allergen and dermatitis; "possible" when the relationship between the allergen and dermatitis was less clear, but the allergen was nevertheless suspected to have caused the dermatitis; and "unlikely/not" when the allergen was not suspected. 8 If the relevance could not be established, it was recorded as "unknown". 16,17 In this study, "certain" and "probable" scores were combined to reflect the highest relevance scores.

| Statistical analysis
All patient characteristics along with the tested series were included as variables in a multivariate logistic regression analysis. Backward elimination with a significance level of P < 0.05 was used to select the best set of risk factors. The interaction between predictors of lanolin contact allergy was evaluated with Spearman's rho correlation test.
Differences in reaction frequencies were evaluated with the one-sized     for 27.1%, the relevance was unknown. However, Amerchol L101 in the routinely tested series had a high proportion of "unknown" scores (67.2%). On the basis of the reactions with known relevance (excluding "unknown" scores), the proportions of the combined "certain" and "probable" relevance scores ranged from 70.0% for lanolin alcohol and Amerchol L101 in the routinely tested series to 85.0% for cremor lanette, only including the preparations with ≥20 positive reactions. in only a few cases did a positive reaction in one series coincide with a questionable reaction to the analogous agent in the other series. We have to keep in mind that low reproducibility is not uncommon in patch testing. It has been described with several allergens other than lanolin, varying from 36.0% to 53.9%. 2,21 The irritancy of the test preparations and the individual susceptibility may play a role in the low reproducibility. Theoretically, this could imply that only patients with positive reactions to both preparations have to be regarded as allergic to lanolin. However, this does not necessarily mean that a patient with only one positive reaction is not allergic to lanolin. Therefore, the diagnosis must be based on the history of the patient in combination with the test results. Importantly, low reproducibility implies that a patch test can give a negative result in a lanolinsensitive patient. This indicates that repeated patch testing may be needed when the suspicion of lanolin contact allergy is high.

| Prevalence and risk factors
Our study population consisted of patients for whom there was a high suspicion of lanolin contact allergy, as they were selected to be additionally tested with lanolin derivatives. Moreover, the patients were referred for tertiary care. As a result, the prevalence of patients with lanolin contact allergy (28.6%, 95%CI: 25.1%-32.3%) is higher than is normally reported. 4,6,7,9,10 The positive association between atopic dermatitis and lanolin contact allergy has previously been reported. 4,7,8 An impaired skin barrier in atopic patients, together with the prolonged use of lanolin-containing topical medicaments, possibly leads to an increased risk of sensitization. On the other hand, the association may be explained by false-positive reactions attributable to irritation, which is more often seen in atopic skin. 5,8 The association between an age of <40 years and lanolin contact allergy was not in line with the literature, as previous studies reported a positive association between higher age and lanolin contact allergy. 6,7 The location of dermatitis may play a part in the association between age and lanolin contact allergy, for example, leg ulcers in the elderly. We did not include the primary site of dermatitis in our study, as other studies have, which might be the reason for diverging results. However, the patient population with leg ulcers referred to our clinic has, in practice, not been treated with lanolin-containing ointments for many years, owing to the use of modern wound dressings. As a consequence, the number of patients with leg ulcers is low in our clinic.

| Strengths and limitations
A strength of our study is that we investigated which lanolin derivatives contributed to the additional diagnostic value and that we assessed the quality of the test preparations in this regard. 10,12 Moreover, all patients were patch tested in the same clinic and under the same circumstances, resulting in an equal set of outcomes. Our study also has limitations. First, the preparations used for patch testing were obtained from different producers, and might vary in their composition. Second, a bias could have been introduced by the fact that relatively few children and leg ulcer patients were included in our study population. This makes age as a potential risk factor difficult to interpret. Finally, we did not include the strength of the positive reaction (+, ++, or +++) in the analysis, because almost all positive reactions were + reactions.

| CONCLUSION
Lanolin contact allergy is frequently seen in referred dermatitis patients. It is a clinical problem, especially for those using lanolin-