What is the added value of patch testing with 30 fragrance allergens in addition to the European Baseline series?

Abstract Background Patch testing with the fragrance allergy markers in the European baseline series (EBS) does not identify all patients with fragrance allergy. Hydroperoxides of linalool and limonene have been shown to be useful allergens in detecting fragrance sensitization. Objectives To evaluate the added value of testing with 30 fragrance allergens in addition to the EBS. Methods All patients with suspected fragrance allergy who underwent patch testing at the Amsterdam University Medical Centers between November 2019 and January 2021 to the EBS and fragrance series were included. Results Of 323 patients tested, 162 (50.2%) were found to be fragrance sensitized. The most sensitizing single allergens were the hydroperoxides of linalool (1.0 and 0.5% pet.) and limonene (0.3 and 0.2% pet.). Testing with the hydroperoxides of linalool and limonene identified 62 fragrance‐sensitized patients (38.3%) who could not be detected by the common fragrance markers. Of all fragrance‐sensitized patients, 21 (13.0%) would have been missed when not testing with the fragrance series. Conclusions Patch testing with the fragrance series in addition to the EBS is valuable. To reduce the risk of false‐negative reactions, it is advisable to test the hydroperoxides of linalool and limonene.


| INTRODUCTION
Fragranced substances are considered one of the most frequent causes of contact allergy. 1,2 In the general European population, sensitization to fragrances varies between 3.9% and 5.5% and has been increasing over the past decades. 2,3 This rising trend might be a consequence of the increased use of cosmetics and toiletries containing fragrances. Fragrances can also be found in cleansing agents, textiles, tobacco and in industrial settings. 4 Due to these various applications, contact with fragrance-related allergens remains hard to avoid.
Cosmetic products sold on the European market are required to be labeled if they contain a certain concentration of 26 fragrance substances known to be contact allergens in humans. 5 Only 14 of these 26 substances are presented in the European baseline series (EBS) as part of the fragrance mixes I and II (FMI and FMII). 4,6 Previous studies showed that the EBS is not able to identify all patients with fragrance allergy. [7][8][9][10] The fragrance mixes may even fail to identify its own constituents. 7,11 As the exposure to environmental allergens is constantly changing over time, patch-test series need to adapt as well in order to remain relevant. 12 Common terpenes linalool and limonene are considered one of the most frequent fragrance ingredients. 13 Moreover, dllimonene is also used as a solvent and industrial degreasing agent. 14 Although they are uncommon fragrance allergens in their pure forms, oxidation transforms the prehaptens linalool and limonene into potent allergens. 15,16 These oxidized terpenes with stable concentrations of the main allergic hydroperoxides have been shown to be useful tools in detecting fragrance sensitization. 17 Recently, various contact allergy groups have therefore advised the inclusion of oxidized linalool and dl-limonene in the baseline series. 18,19 However, due to their irritant potential, inclusion is still under discussion concerning the EBS. 20 Since 2019, the terpene hydroperoxides are routinely tested at the Amsterdam University Medical Centers (AUMC). In addition, oil of turpentine, a substance that is used as a raw material in the perfume industry, has been routinely tested due to increased sensitization rates. 21 The aim of this study was to assess the added value of performing patch testing with the labeled fragrance substances, the hydroperoxides of linalool and limonene, and oil of turpentine in addition to the EBS. Other objectives were to report the sensitization rates of the tested allergens and to analyze co-reactions between the fragrance allergens.

| Patients
In this retrospective cohort study, all patients with suspected fragrance allergy who had undergone patch testing with the EBS, the labeled fragrance substances, the hydroperoxides of linalool and limonene, and oil of turpentine (see Appendix 1 for all tested allergens) at the AUMC between November 2019 and January 2021 were analyzed. Patients who were <18 years of age and patients with angryback reactions were excluded. Approval was obtained from the Medical Ethics Review Committee of the Academic Medical Center (reference number W20_555 #20.618).  23 The patch test reactions were scored as negative (À), questionable (?), irritant (IR), and allergic (+, ++, +++). All reactions + or higher were regarded as positive. Clinical relevance of positive patch-test reactions was registered as certain, probable, possible, unlikely, or unknown. "Probable" and "certain" scores of relevance were considered as clinically relevant.

| Linalool and limonene
Of the patients sensitized to linalool hydroperoxide 1.0% pet., 60.3% had a co-reaction to its less concentrated form (0.5% pet. had a concomitant reaction to the less concentrated form (0.2% pet.). Concomitant reactions to the fragrance series in patients sensitized to oxidized linalool or oxidized limonene varied between 32.1% and 39.4% ( Table 2).
The role of the oxidized terpenes as fragrance allergy markers was measured in a correlation analysis. No relevant significant associations with other individual allergens or markers were found (Appendix 2).

| Oil of turpentine
Oil of turpentine yielded a sensitization rate of 1.2%, as four patients

| DISCUSSION
In this retrospective cohort study, the added value of testing with the hydroperoxides of linalool and limonene, the EU-labeled fragrance allergens, and oil of turpentine, in addition to the EBS, was evaluated.
If only the EBS was tested, 34.0% of the sensitized patients would not have had their allergies fully defined and 13.0% would have remained undetected as fragrance allergic. Without patch testing with the hydroperoxides of linalool and limonene, 38.3% of all fragrancesensitized patients would be missed. Therefore, it is valuable to perform patch testing with the fragrance series in addition to the EBS.
Patch testing with the hydroperoxides of linalool and limonene in the EBS will reduce the risk of false-negative reactions, vs testing with nonoxidized linalool and limonene.

| Fragrance sensitization
Fragrance sensitization was present in about half of our patch-tested patients (50.2%) and was significantly associated with facial dermatitis.
Lower rates of fragrance allergy (7.6%-26.9%) were reported in previous studies. [8][9][10]25 Referral bias and selection bias might be the cause of this difference. In addition, the individual constituents of FMI (except for cinnamal) in this study were patch tested at double the concentrations used than in two previous studies. 10,25 By testing at higher concentrations, an improved diagnostic ability is expected. 6,11,26,27 In one previous study, the hydroperoxides of linalool (1% pet.) and limonene (0.3% pet.) were also tested. 25 The higher sensitization rate in the current study (50.2% vs 15.7%) can be explained partly by patch testing linalool and limonene hydroperoxide in the two recommended dilutions instead of one. 20 Besides that, Myroxylon pereirae and methyl-2-octynate were not included as fragrance allergens in the Danish study.

| Individual fragrance allergens
Among the EU-labeled allergens, citral (6.2%), Evernia furfuracea (4.6%), and Evernia prunastri (4.0%) were the most sensitizing. This is in line with previous reports, especially for the last two allergens. [8][9][10]25,29,31 The high sensitization rate of Evernia prunastri is attributable mostly to its allergic compounds chloroatranol and atranol. 32 Due to their hyperallergenic potential, cosmetics containing these compounds are no longer allowed to be sold in the EU from August 2021. 33,34 The patients sensitized to Evernia furfuracea can be divided into two subgroups: those with sensitization to common constituents as also present in Evernia prunastri, and those sensitized to oxidized resin acids as present in the contaminating tree bark which is used in the extraction process. This latter subgroup was identified by positive reactions to colophonium. 35 Five of the 15 patients (33.3%) sensitized to Evernia furfuracea had positive co-reactions to colophonium. Two of these double-sensitized patients (40%) reacted negative to Evernia prunastri. The high co-sensitization rate to colophonium among patients sensizited to Evernia furfuracea in combination with the fact that not all patients were sensitized to Evernia prunastri supports the heterogeneity in this sensitized group.
Citral is a mixture of the two aldehydes geranial and neral. These compounds have also been found in the oxidation process of geraniol.
Just like linalool and limonene, geraniol has the potential to form allergenic compounds by autooxidation on air exposure. 36 Because coreactions between citral or its components and oxidized geraniol are recently specified in literature, sensitization to citral can be explained partly by this co-reactivity. 37 In a recent Swedish study, oxidized geraniol yielded a significantly higher sensitization rate than its pure form. 38 Patch testing with oxidized geraniol may therefore be of value. Due to oxidized geraniol currently being unavailable for patch testing, this cannot be confirmed by the current study. Oil of turpentine was tested as well, owing to the interest in this substance as a fragrance allergen, and yielded a sensitization rate of 1.2%. Seventyfive percent of the patients with positive patch-test reactions to oil of turpentine had concomitant positive patch-test reactions to fragrance markers: FMI, Myroxylon pereirae, and a recommended addition to the EBS: linalool hydroperoxide 0.5% pet. Only one of the four sensitized cases in the current study remained undetected by the fragrance markers; therefore, the added value of patch testing this substance consecutively is questionable.

| Limitations
Selection bias might have been introduced by including patients suspected of having fragrance allergies. Referral bias and missing clinical relevance data are limitations of this study. In addition, the sources of exposure were not investigated.

| CONCLUSION
In conclusion, testing with the hydroperoxides of linalool and limonene in each two dilutions has been useful to identify fragrance-