The Effects of Vehicles on the Human Dermal Irritation Potentials of Allyl Esters

Abstract

Allyl esters, frequently used in the fragrance industry, often contain a certain percentage of free allyl alcohol. Allyl alcohol is known to have a potential for delayed skin irritation. Also present in the finished product are different solvent systems, or vehicles, which are used to deliver the fragrances based upon their intended application. This study was conducted to determine whether different vehicles affect the skin irritation potential of five different allyl esters. The allyl esters tested were allyl amyl glycolate, allyl caproate, allyl (cyclohexyloxy)acetate, allyl cyclohexylpropionate, and allyl phenoxyacetate in the vehicles diethyl phthalate, 3:1 diethyl phthalate:ethanol, and 1:3 diethyl phthalate:ethanol at concentrations of 0.1%, 0.5%, 1.0%, and 2.0% (w/w). A modified cumulative irritation test was conducted in 129 human subjects. Test materials (0.3 ml) were applied under occlusion to skin sites on the back for 1 day (24 h) using Hill Top chambers. Irritation was assessed at 1, 2, 4, and 5 days following application of test materials. Cumulative irritation scores varied considerably among test materials. There were no delayed irritation observations. The highest irritation scores were observed at the 2.0% concentration for all test materials. The irritation scores for allyl amyl glycolate, allyl (cyclohexyloxy)acetate, and allyl phenoxyacetate were highest in 1:3 diethyl phthalate:ethanol, thus the resulting calculated no-observed-effect levels, 0.12%, 0.03%, and 0%, respectively, were much lower for this vehicle compared to the diethyl phthalate vehicle, 0.33%, 0.26%, 0.25%, respectively. These data showed a trend for lower concentration thresholds to induce irritation when higher levels of ethanol were used in the vehicle.

Irritant contact dermatitis has been described as a nonimmunologic inflammatory response characterized by erythema, edema, or corrosion, following single or repeated dermal applications of a substance to an identical site (Mathias 1987). Clinically, irritation may appear in a variety of forms, highlighting the heteromorphous processes by which it progresses (Weltfriend et al. 1996). In the delayed form of acute irritant contact dermatitis, inflammation may not be seen until 12 to 24 h after exposure. This delayed onset often leads to confusion with allergic contact dermatitis, which usually has the same clinical characteristics (Weltfriend et al. 1996). In the case of allergic contact dermatitis, only a minute quantity of the material is necessary to elicit overt reactions, which is distinct from irritant contact dermatitis reactions, which are proportional to the dose applied. It was reported that some allyl esters used in perfumery exhibited a delayed type of irritant reaction (Opdyke 1977a, 1977b, 1977c, 1977d, 1977e).

Allyl esters are materials frequently used in the fragrance industry in a variety of fragrance compounds (Calkin and Jellinek 1994). These materials are produced by direct esterification of allyl alcohol with the corresponding acid (Arctander 1969). As a result of their manufacture, these esters often contain a certain percentage of free allyl alcohol. Allyl alcohol is known to have a delayed skin irritation potential and the presence of allyl alcohol in the esters may affect their irritation potential (Opdyke 1977f). Concern over the delayed irritant potential of allyl alcohol prompted the International Fragrance Association (IFRA) to recommend the use of allyl esters only when the level of free allyl alcohol is less than 0.1% (IFRA 1977). In the study reported here, five esters were selected to investigate their irritant potential—allyl amyl glycolate (AAG), allyl cyclohexylpropionate (ACP), allyl caproate (AC), allyl (cyclohexyloxy)acetate (ACA), and allyl phenoxyacetate (APA).

Dermal irritation and sensitization studies previously conducted on humans with these five allyl esters are summarized in Table 1. Of those studies, only a limited number have been conducted with ACP, AC, and APA, all using petrolatum or a base cream as the vehicles. A larger number of studies have been conducted with AAG and ACA. The irritation results for these compounds varied, due to differences both in the concentrations used and the vehicle chosen. In the repeated-insult patch test conducted with AAG (RIFM 1982), there were many irritant reactions observed in the beginning of induction phase, thus the concentration of AAG was reduced for later patch applications. This was a nonstandard study and due to the high number of irritant reactions observed, no conclusions can be made from this study regarding the sensitization potential of AAG. ACA and APA both produced reactions in sensitization and maximization studies, though these were considered to be irritant in nature (RIFM 1974a, 1998). Basketter et al. (1998) investigated the skin irritation of allyl esters observed in patch tests and use tests, though the specific allyl esters present in the formulation were not identified.

The vehicle in which an irritant is administered to the skin has been shown to have an effect on the material’s irritant potential, though this effect has not been well studied (Fleming and Bergfeld 1990). In a repeated-insult patch test, irritation was found to increase with increasing levels of ethanol (EtOH) in distilled water (dH₂O). Rarely, EtOH alone induced sensitization reactions in humans (Stotts and Ely 1977). In addition, the severity of the irritant reaction has been observed to be dependent on the concentration of the test material applied (Patrick and Maibach 1989). The Research Institute for Fragrance Materials (RIFM) has conducted a series of studies to evaluate the diethyl phthalate (DEP):EtOH vehicle system for its utility in the murine local lymph node assay and investigated the potential differences in skin sensitization resulting from its use. Four fragrance materials, p-t-butyl-alpha-methylhydrocinnamic aldehyde, geraniol, eugenol, and hydroxycitronellal were tested in the vehicles DEP, 3:1 DEP:EtOH, 1:3 DEP:EtOH, and EtOH. The strength of the observed sensitization responses were observed to vary with the vehicle; however, the results did not show any clear pattern of one vehicle over another regarding skin sensitization (Lalko, Isola, and Api 2004). To examine the effect of different vehicle formulations on the irritation potential of the selected allyl esters for this study, RIFM employed a human predictive irritation test utilizing the vehicles DEP, 3:1 DEP:EtOH, and 1:3 DEP:EtOH. The vehicles were selected because they are representative of the matrix in which human skin exposure to commercial fragrance materials is likely to occur and are employed at the RIFM in both hazard and safety assessments (Api 2002). The concentrations chosen, 0.1%, 0.5%, 1.0%, and 2.0%, were selected based upon historical data (Table 1).

MATERIALS AND METHODS

Chemicals

The test materials allyl cyclohexylpropionate (ACP), allyl caproate (AC), and allyl phenoxyacetate (APA) were supplied by Haarmann & Reimer (H&R), Teterboro, NJ. International Flavors and Fragrances (IFF), Union Beach, NJ, supplied allyl amyl glycolate (AAG), allyl (cyclohexyloxy)acetate (ACA), and the vehicles DEP, 3:1 DEP:EtOH, and 1:3 DEP:EtOH. All five test materials were 98% pure and structures are shown in Figure 1. The positive control, sodium lauryl sulfate (SLS; purity 98%) was supplied by Aldrich, St. Louis, MO.

Method

Prior to enrollment, each potential subject completed a medical history questionnaire and informed consent was obtained. Individuals were excluded from the study if they exhibited any dermatological or other medical or physical condition that would preclude topical application of the test material. In addition, pregnant and/or nursing women were excluded from testing. One hundred thirty subjects meeting criteria for inclusion were enrolled and were randomly divided into five test groups containing 26 subjects.

Test materials were applied using 25 mm Hill Top Chamber patches (Hill Top Research, Miamiville, OH). Hill Top Chamber patches consist of a flexible molded plastic chamber with a double rim that fits close to the skin. The chamber is lined with a nonwoven Webril pad, and the entire patch is held in place by semiocclusive tape. An aliquot of 0.3 ml of the test material was dispensed onto the patch and the patch was affixed to the back of each subject. Each fragrance material was tested in each vehicle at 0.1%, 0.5%, 1.0%, and 2.0% (w/w) (Table 2). The positive control, sodium lauryl sulfate, was tested at 1% in distilled water (dH₂O) and at 1% in each of the three vehicles. Each vehicle was tested as its own control and the negative control was physiological saline (Table 2). Each subject received a single application of 20 patches, which included the test material at four concentrations in each vehicle, vehicle controls, and positive and negative controls. The subjects were instructed not to expose their backs to sunlight while on the study. They could bathe or shower in the normal manner, but were asked to avoid excessive wetting while the patches where in place. After 1 day (24 h), the subjects returned to the test facility and the patches were removed. Each test area was rinsed with dH₂O and scored for irritation using the Draize scoring system (Draize, Woodard, and Calvery 1944) modified by Phillips et al. (1972; Table 3). Subjects were again evaluated 2, 4, and 5 days (48, 96, and 120 h) following patch application. After the sites were scored on the day 5 laboratory visit, the study was concluded.

The irritation scores for each observation period were added to the scores from the previous observation periods to obtain a cumulative irritation score (CIS). In this format, the irritation potential of each test material could be better quantified. The raw data were subjected to an analysis of variance (ANOVA) (Nie et al. 1975). The calculated no-observed-effect levels (NOELs) were determined by linear regression.

RESULTS

One hundred thirty subjects were enrolled and 129 completed the study. One subject did not complete the study because the patches did not remain on the back for the required 24-h period. The CISs varied greatly among the test materials, with ACP causing less irritation than any other allyl ester (Figure 2 and Table 4). The mean values for the first observed irritation scores (Tables 5 to 9) were much narrower in comparison to the CISs and were 2 days or less. Therefore, none of the test materials appeared to produce a delayed irritation response in any of the subjects. A maximum irritation score of 4, based upon the modified Draize scoring system (Phillips et al. 1972; Table 3), was not observed on any individual subject for any test material in any group on any given day. The only irritation scores of level 3 that were observed occurred from the positive control, 1% SLS in dH₂O in 28/129 subjects and 1% SLS in DEP in 4/129 subjects. SLS at a concentration of 1% in dH₂O was considerably more irritating than the SLS in any of the other vehicles tested (Table 10). The highest irritation score observed for any of the allyl esters tested was 2. Generally, the irritation scores for the allyl esters waned from day 1 to day 5. In the case of 0.1% AC in 1:3 DEP:EtOH, the increased score on day 4 was due to a level 1 reaction observed in a second subject, not an increased reaction. No irritation was observed with saline.

The dose-response data, showing the concentration of test material plotted against the concentration of EtOH in the vehicle, are presented in Figure 2. These contour plots graphically illustrate the trends of the cumulative irritation reactions for each test material. It should be noted that the contour curves were based upon the limited individual data sets available for each test material. APA, ACA, and to a lesser extent AAG showed a trend towards greater CISs with increasing concentrations of EtOH in the vehicle, which appeared most pronounced for 1.0% and 2.0% of test material. ACP induced little to no irritation with any of the vehicles. AC had the highest CIS when DEP was the vehicle.

The highest CISs were observed at the highest concentration (2%) of test material in the 1:3 DEP:EtOH vehicle for all test materials with the exception of AC (Table 4, Figure 2). APA in 1:3 DEP:EtOH was the most irritating test material, with a CIS of 67. The 1% SLS positive control had the highest CIS, 122, which was observed when the vehicle was dH₂O (Table 10). Results of the ANOVA showed no statistical differences in the CISs based solely upon vehicle selection for all test materials across any given concentration level (Table 4). There were, however, significant differences in CIS’s based upon the concentration levels within test material groups. A concentration of 2% AAG in 1:3 DEP:EtOH was significantly more irritating than 0.1% and 0.5% AAG in all three vehicles. Additionally, 2% AAG in DEP was significantly more irritating than 0.1% and 0.5% AAG in DEP and 3:1 DEP:EtOH and 0.1% AAG in 1:3 DEP:EtOH (Table 4).

Test materials ACP, AC, and ACA, regardless of the vehicle employed or concentration level, had no significant differences in the CIS (Table 4). In the APA test group, the cumulative irritation score for 2% APA in 1:3 DEP:EtOH was significantly greater than the score for 0.1%, 0.5%, or 1.0% APA in either DEP or 3:1 DEP:EtOH, or 0.1% APA in 1:3 DEP:EtOH.

When the CISs for each fragrance material were used to rank the vehicles according to their ability to alter skin irritation potential, the data indicate a trend for higher levels of irritation when the vehicle was 1:3 DEP:EtOH.

AAG: 1:3 DEP:EtOH > DEP > 3:1 DEP:EtOH

ACP: 1:3 DEP:EtOH > 3:1 DEP:EtOH = DEP

AC: DEP > 3:1 DEP:EtOH > 1:3 DEP:EtOH

ACA: 1:3 DEP:EtOH > DEP > 3:1 DEP:EtOH

APA: 1:3 DEP:EtOH > DEP > 3:1 DEP:EtOH

The NOELs for irritation, both the observed and calculated values, were much lower when the vehicle was 1:3 DEP:EtOH (Table 11). The observed and calculated NOELs with the vehicle 1:3 DEP:EtOH did not exceed 0.12% for any test material. For DEP and 3:1 DEP:EtOH, the observed and calculated NOELs ranged between 0.24–1.00% and 0.07–1.00%, respectively.

DISCUSSION

This study examined the effect of different vehicle formulations on the irritation potential of the selected allyl esters. RIFM employed a human predictive irritation test utilizing the vehicles DEP, 3:1 DEP:EtOH, and 1:3 DEP:EtOH. The greatest exposure to fragrance materials occurs through dermal application of alcohol-based products (Cadby, Troy, and Vey 2002). Therefore, the EtOH-containing vehicles used in this study represent realistic vehicles for the study of the dermal irritation potential of fragrance materials.

The dermal irritation and sensitization studies that have been previously conducted on humans with these five allyl esters are summarized in Table 1. There are clear data depicting the irritation potential of AAG, AC, ACA, and APA at certain concentrations. AAG demonstrated no irritation in the two studies listed. In the sensitization or maximization studies listed in Table 1, AAG and ACP induced no reactions. The reactivity observed for AAC, ACA, and APA were determined to be irritant in nature (RIFM 1974a, 1975, 1998).

In the studies described in Table 1, a variety of vehicles were utilized. In only a few of these studies was it possible to directly compare the effects of different vehicles. AAG tested at 0.5% and ACA tested at 0.2% induced no irritation or sensitization in repeated insult patch tests, whether the vehicle was DEP or 1:3 DEP:EtOH (RIFM 1989a, 1989b). When AAG was tested in a repeated-insult patch test at 0.8%, no irritation or sensitization was observed when the vehicle was DEP (RIFM 1990a), whereas 1:3 DEP:EtOH induced irritation in the majority of the subjects (RIFM 1990b). In a single patch irritation test, ACA was tested at 1%, 5%, and 10% in EtOH or olive oil. At all three concentrations there was a greater number of reactions induced by EtOH than by olive oil (RIFM 1989c).

Based on the outcome of the cumulative irritation study reported here, a trend towards a lower CIS when higher levels of EtOH were used in the solvent system was observed. Similarly, it has been shown that EtOH alone is capable of inducing sensitization on rare occasions (Stotts and Ely 1977). This suggests that, in some cases, the vehicle may alter a substance’s irritation potential.

Allyl esters have shown delayed-type irritant reactions in earlier reports (Opdyke 1977a, 1977b, 1977c, 1977d, 1977e). This study was designed to look at five related allyl esters to see if this effect could be reproduced. Another objective was to determine if the current IFRA Standard, which limits the amount of free allyl alcohol in allyl esters to less than 0.1% (IFRA 1977), needs to be revised or a new standard needs to be issued. There were no delayed irritation observations as evidenced by irritation induction of 2 days or less for the mean day of first observed irritation. The primary variable affecting the outcome of the study was the concentration of the test material rather than vehicle, with the highest degree of irritation observed at the highest concentration levels. Because delayed irritation was not observed for any of these allyl esters, the current IFRA Standard does not need to be changed.