Introduction

Although preservative chemicals are an indispen-sable component of a wide range of personal care products, they nevertheless frequently possess the properties necessary for them to behave as a skin sensitizer [1,2]. Preservative chemicals are usually identified as positive in predictive tests for skin sensitization [3,4]. Furthermore, over many decades, human skin exposure to preservatives has often been associated with the causation of a degree of allergic contact dermatitis (ACD) [5–7]. Since the skin-sensitizing hazard can be identi-fied, it is reasonable to ask why preservatives cause ACD. Two obvious possibilities arise, that the risk they present to human health has not been properly assessed and that risk management is insufficient. This paper discusses both of these aspects, but with particular emphasis on the former.

A recent publication reported the outcome of a retrospective evaluation of a quantitative risk assessment (QRA) process for skin sensitization for a range of preservatives in several types of products [8]. QRA for skin sensitization uses information on the relative potency of a sensitizing chemical to determine a level for the inclusion of a sensitizing substance in a formulation that should be below the threshold for the induction of skin sensitiza-tion in humans. The details and application of this QRA approach have been extensively published elsewhere [9–14]. They have been adopted by the fragrance industry in the last few years as a more effective way to determine the maximum accept-able level of sensitizing fragrance chemicals in a wide range of product categories [15].

Quantitative risk assessment is attractive and appears to offer a degree of precision previously lacking in skin-sensitization risk assessment.

Cutaneous and Ocular Toxicology, 2010; 29(1): 4–9

C R I T I C A L R E V I E W

Methyldibromoglutaronitrile: skin sensitization and quantitative risk assessment

David A. Basketter

DABMEB Consultancy Ltd., Sharnbrook, Bedfordshire, UK

AbstractPreservatives can be a frequent cause of allergic contact dermatitis (ACD). A quantitative risk assess-ment (QRA) method for identifying safe exposure levels has been suggested as a more effective tool for this purpose. This work assesses the validity of QRA by its retrospective application to the sensitizing preservative methyldibromoglutaronitrile (MDGN), which has recently been associated with unaccept-able exposure levels in consumer products. Using a recently published QRA analysis of 4 preservatives in 5 consumer product types, the accuracy of the predictions for MDGN was assessed in light of what is known clinically about the nature and incidence of ACD to this material. Based on a local lymph node assay (LLNA) EC3 value (concentration of test chemical required to provoke a 3-fold increase in lymph node cell proliferation) of 0.9% in a weight-of-evidence approach to the identification of thresholds for the induction of skin sensitization, it can be determined that the acceptable levels of exposure to MDGN in a range of products range from as little as 25 ppm to in excess of 10,000 ppm. Thus, proactive use of QRA, used conservatively and in combination with expert judgment, would have limited the problem of ACD to this new preservative that is known to have caused problems on the consumer market.

Keywords: Preservatives; methyldibromoglutaronitrile; skin sensitization; induction thresholds; local lymph node assay; EC3 value; quantitative risk assessment

Address for Correspondence: Dr. David A. Basketter, DABMEB Consultancy Ltd., Sharnbrook, Bedfordshire MK44 1LQ, UK. Tel./Fax: +44-1234782944. E-mail: david.basketter@ukonline.co.uk

(Received 17 July 2009; revised 17 September 2009; accepted 20 September 2009)

ISSN 1556-9527 print/ISSN 1556-9535 online © 2010 Informa UK LtdDOI: 10.3109/15569520903351151 http://www.informahealthcare.com/cot

Cutaneous and Ocular Toxicology

1

2010

4

29

9

1556-95271556-9535© 2010 Informa UK Ltd10.3109/15569520903351151

17 July 200920 September 200917 September 2009

COT

435293

Cut

aneo

us a

nd O

cula

r T

oxic

olog

y D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Uls

ter

at J

orda

nsto

wn

on 1

1/13

/14

For

pers

onal

use

onl

y.

MDGN allergy and quantitative risk assessment 5

Nevertheless, it has to be recognized that underlying the QRA is actually a combination of science and expert judgment. Uncertainties exist in the estimation of both allergen potency and skin exposure. The extent to which QRA is a valid approach can be judged, at least in part, by applying it to situations in which sensitizing materials have already been used in defined markets and then associated with the pres-ence or absence of effects. In the present work, the preservative methyldibromoglutaronitrile (MDGN) was selected as a case study. MDGN has been associ-ated with use in product types and at levels that have led to an unacceptable frequency of ACD [16–19]. Would QRA have predicted this?

Materials and methods

Identification of no-expected-sensitization-induction levels

The no-expected-sensitization-induction level (NESIL, expressed in µg/cm2) was derived from the local lymph node assay (LLNA) EC3 value (concen-tration of test chemical required to provoke a 3-fold increase in lymph node cell proliferation) of 0.9% published in 2005 [4] using the approach described by Safford [20]. In brief, the human NESIL was estimated by applying the mouse-to-human factor derived from the previously published correlation [12] as defined by the following linear regression equation:

log NESIL g/cm 1.16 log EC3 g/cm 0.64102

102µ = µ −

where the conversion from percentage values to the area dose (based on application of 25 µL of test solution to a mouse ear of 1 cm2) is achieved via the following equation:

EC3 g/cm EC3% 0.025 1,000,000 (g to g)100 1

2µ = × × µ×

Other data may be available for MDGN, but, in the spirit of using this as a real example, only the LLNA data were used, since new European Union (EU) regulations normally will mean that this is the only predictive test data available for new preservative chemicals [21,22].

Exposure data

Six product types representing a variety of expo-sure types/situations were selected: shampoo, face cream, nonaerosol deodorant, a lipstick, a body

lotion, and a liquid soap. These are the products assessed previously with the addition of liquid soap, still using the underlying considerations in that publication, which makes use of industry meas-urements of consumer exposure, including the 90th percentile of the dose (Table 1) [8]. For liquid soap, use may be both occupational and domestic. Frequency of use occupationally is hard to ascertain, but certainly 50 uses in an 8-hour work shift seem to be quite possible [23]. If one adds to this domestic use of 10 times per day, then at 1 g per use, the total daily exposure can be, at 60 g, surprisingly high. This may appear excessive, but it is apparently within the bounds of possibility and so has been used in the calculations in this paper.

Safety assessment factors

Safety assessment factors (SAFs) (for interindividual variability, matrix effects, and use considerations) are used to extrapolate from the experimental exposures from which the benchmark data have been generated and relate them to real-life con-sumer product–exposure scenarios [10,11,14] In the present QRA, for consistency the SAFs used (Table 2) were exactly as described previously in order to test their predictivity [8]. The SAFs for liquid soap were taken from those for the similar rinse-off product, shampoo.

Quantitative risk assessment

The NESIL was divided by the calculated overall SAF to determine the acceptable exposure level (AEL) (Table 3). Two of the elements of the overall SAF (interindividual variability and use considera-tions) for each product type were the same as those determined from the QRA fragrance materials task force document [14]. For the matrix SAF, a factor of 3 was taken (actually √10 = approximately 3.16). The SAFs were multiplied together for each prod-uct type to determine the overall SAF. The AEL was translated into a maximum consumer exposure level (CEL) to MDGN by assuming the maximum safe level would be where the AEL to CEL ratio was 1. In the QRA calculations, numbers have normally been expressed to no more than 3 significant figures. In all calculations, dose is expressed as µg/cm2.

Results

The LLNA EC3 value of 0.9% translates to a NESIL of 120 µg/cm using the conservative approach

Cut

aneo

us a

nd O

cula

r T

oxic

olog

y D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Uls

ter

at J

orda

nsto

wn

on 1

1/13

/14

For

pers

onal

use

onl

y.

6 David A. Basketter

described by Safford [20]. Division of this fig-ure by the SAFs (from Table 2) produces the AEL. Normally, this estimate is then compared with the proposed CEL. The hypothesis is that if the CEL is no greater than the AEL, then the product use should not be associated with the induction of skin sensitization, although, of course, this allows no margin of error. However, in this work, the logic has been manipulated such that when AEL = CEL, this predicts the absolute maximum safe use level of a sensitizing ingredient, in this case MDGN. These results are shown in Table 3. As would be expected, the theoretical maximum safe concen-tration of MDGN varies widely, from 35 ppm in a lipstick to > 10,000 ppm in a daily-use shampoo. The final column of Table 3 displays the values a toxicologist might actually choose using the out-come of the QRA as a starting point, but also based on a combination of additional factors, includ-ing a proposed upper limit of 1,000 ppm based on microbiological considerations and raw material costs, as well as on a level of concern about the

particular product type, the use of the preservative in other products types and/or by other compa-nies, and the recognition that the QRA is not (yet) a precision tool. The fact that the body lotion had a lower maximum limit than the face cream results in a much lower limit for the face cream than the QRA-predicted maximum, the underlying logic being that these products might be used interchangeably by the consumer and thus the lower value must be applied to both in practice.

Discussion

In any textbook or review on the causes of ACD, preservatives are high on the list [7,24–26]. Although this is not altogether surprising, since effective preservatives are almost always low-molecular-weight reactive organic chemicals, which makes them ideal haptens, it does in reality reflect a use of preservative chemicals that exceeds safe levels [27]. A recent publication examining the levels of a range of preservatives in cosmetics and using a QRA tool showed that, at least for some product types,

Table 1. Exposure data used to calculate dose per unit area.Product type

Amount useda (g/day)

Retentionb factor

Surface areab cm2

Dose/unit areab (mg/cm2)/day

Shampoo 10.5 0.01 1,430 0.07Face cream 1.54 1 565 2.73Nonaerosol deodorant

1.51 1 200 7.55

Body lotion 7.82 1 15,670 0.5Lipstick 0.057 1 4.8 11.8Liquid soap 60.0 0.1 840 7.1aThe 90th percentile, taken from reference 8 (Basketter et al. 2008), except for the liquid soap (see the Materials and methods section of text).bTaken from reference 8 (Basketter et al. 2008).

Table 2. Default safety assessment factors used for different product types.Product type Human SAF Matrix SAF Use SAF Overall SAFa

Shampoo 10 3 3 100Face cream 10 3 3 100Nonaerosol deodorant

10 3 10 300

Body lotion 10 3 10 300Lipstick 10 3 10 300Liquid soap 10 3 3 100aSafety assessment factor (SAF) values taken from reference 8 (Basketter et al. 2008) and rounded to the nearest hundred.

Table 3. Quantitative risk assessment for methyldibromoglutaronitrile in different product types.

Product type NESILa (µg/cm2) SAFb AELc (µg/cm2)Daily product used

(µg/cm2)Theoretical maximum use

level in product (ppm)e

Probable maximum use level in product (ppm)f

Shampoo 120 100 1.2 70 >10,000 1,000Face cream 120 100 1.2 2,730 450 50Nonaerosol deodorant

120 300 0.4 7,550 50 25

Body lotion 120 300 0.4 5,000 80 50Lipstick 120 300 0.4 11,800 35 25Liquid soap 120 100 1.2 7,100 175 100aNo-expected-sensitization-induction level (NESIL) derived from local lymph node assay (LLNA)-predicted human repeat insult patch test (HRIPT) no-observed-effect level (NOEL).bSensitization assessment factor (SAF) from Table 2.cAcceptable exposure level (AEL) (NESIL divided by SAF).dDosimetry on skin arising from daily use of product expressed in µg of product per cm2 (from reference 8 [Basketter et al. 2008] and Table 1).eDerived by division of the AEL by the daily product use level and rounded to the nearest 5.fEstimated value that a practicing toxicologist would actually set, recognizing that quantitative risk assessment (QRA) is not a precision tool.

Cut

aneo

us a

nd O

cula

r T

oxic

olog

y D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Uls

ter

at J

orda

nsto

wn

on 1

1/13

/14

For

pers

onal

use

onl

y.

MDGN allergy and quantitative risk assessment 7

typical levels of preservatives would be close to, or even somewhat above, the theoretical maximum safe level [8]. This is consistent with the clinical pic-ture, which shows that preservative chemicals are often common causes of ACD, including in asso-ciation with cosmetics [23–25]. However, the paper on preservatives and QRA excluded a currently important cause of ACD to a preservative, MDGN. This chemical, originally permitted at up to 1,000 ppm in cosmetics by the EU, after an initial period where few problems were reported [28] has been associated with an epidemic of allergy in the last 15 years [16–19,29,30]. Causative products have included leave-on and rinse-off products [19]. The question addressed in this present publication is whether, had it been available at the time (i.e., the end of the last century), the careful application of QRA would have limited the use levels of MDGN and helped to avoid this epidemic.

A review in 2005 identified the causative products in Denmark associated with ACD to MDGN [19]. This showed that skin cream, liquid soaps, and shampoo accounted for 68%; the discussion below focuses on these. It is recognized that MDGN allergy also arises from a range of other products not considered here, but the review here has been limited to cosmetics [31,32]. Taking the most important product category from the Danish review, creams, which accounted for almost one third of cases, 2 products in this present analysis cover this category, face cream and body lotion (data are not shown, but a hand cream would not differ). While the maximum use level pre-dicted by the QRA for a face cream is approximately 450 ppm, the actual use level would be limited to 50 ppm for the reasons already explained. Nevertheless, this raises 2 important points: First, determining a safe use level of a preservative for a single product in isolation is unwise. For a raw material that may be used in multiple product types, the experienced toxicologist must make a judgment about safe use. A similar consideration would apply, for example, to fragrance and any other commonly used ingredient. Secondly, at 50 ppm, the CEL would be almost an order of magnitude lower than the QRA-predicted AEL for a face cream—many toxicologists would describe this as a “safety margin of 10,” but to do so would be to fundamentally misunderstand skin-sensitization QRA, where those safety margins have already been introduced via the SAFs. As a reminder, the purpose of this paper is to do a practical test to assess whether in the case of MDGN those factors are adequate, always bearing in mind that there is also a dependence on the correct measurement of skin-sensitization potency in the LLNA. Having said that, then, it is of interest that in the previous

publication, particularly for products with pro-longed skin contact, such as creams, AEL to CEL ratios were often less than 1, indicating a real level of risk [8]. In the present case considering the use of MDGN and creams, the suggested maximum use level of 50 ppm would actually have an AEL to CEL ratio greater than 1, indicating that it should be safe in the marketplace. Concentrations of MDGN asso-ciated with ACD in this product category have been reported to be in the range of approximately 150–400 ppm [18,19]. This indicates that either there is only a little room for maneuver in the QRA or that the use of mouse data somewhat underestimates the true sensitizing potency of MDGN. Nevertheless, proper application of QRA appears to have identified that the use level should have been distinctly lower than the product concentrations that were actually deployed in practice.

The second most important product category in the Danish review was liquid soaps, said to account for almost a fourth of cases of ACD to MDGN. Here product use concentrations have been reported to be similar to those used in the skin creams [18,19]. QRA predicts a maximum safe level of a little over 175 ppm, although in practice a maximum working level of 100 ppm would likely be selected, for the reasons already alluded to in the Results section. This level is made low by the particularly high use frequency, but the clinical evidence is that this would not be wholly protective, and hindsight suggests, there-fore, that the SAF for use should actually be 10, to take account of the damaging effect of such frequent exposure to surfactant. This would effectively reduce the maximum allowed level to 60 ppm, which the precautionary toxicologist might well reduce to 50 ppm for product use in practice. However, this is a product use area where exposure data are limited, particularly occupationally, and it must be noted that the figures reported in this paper do effectively suggest that 6 g of soap remains on the hands, which may be excessive.

The third category of importance was shampoo, attributed to 1 in 7 of ACD cases [19]. The less frequent use and more thorough rinsing of this product category (compared with liquid soap) mean that the maximum use level in theory could be greater than 1%! This would not occur in practice as preservatives are typically an expensive raw material, and such levels would not be required from a microbiological perspective. A more realistic level of 1,000 ppm is suggested in practice. Even this level, however, must be too high, since it is equiva-lent to the maximum level originally allowed by the EU, and yet there were clinical cases of ACD to MDGN associated with shampoo use. Of course, it

Cut

aneo

us a

nd O

cula

r T

oxic

olog

y D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Uls

ter

at J

orda

nsto

wn

on 1

1/13

/14

For

pers

onal

use

onl

y.

8 David A. Basketter

must be accepted that some of these might in reality be elicitation of ACD in an individual sensitized by another product. Nevertheless, the judgments made in a number of cases seem to be that the shampoo may have been a primary cause of the ACD [18,19]. Accordingly, it is clear that the QRA for MDGN in shampoo gives a misleading degree of confidence, whether through the application of insufficient safety factors or through the underestimation of the potency of MDGN.

For the remaining products considered in this paper, the retrospective QRA ultimately suggests that a working upper concentration limit of 25 ppm would be appropriate for lipstick and nonaerosol deodorant. There is little clinical evidence that there have been problems with MDGN in these product categories, but this might well have arisen simply via limited use of this preservative in such product types. However, it is interesting to contrast this value with the original EU limit of 1,000 ppm. If nothing else, it helps to reinforce the message that even where there is a regulatory limit for use of preservatives in cosmetics, this does not imply that it is always a safe limit. The responsibility still lies with the company safety assessor to judge what is an acceptable use level. Taking a specific example, then, the frequent use of MDGN in the presence of detergents has been demonstrated to result in a > 6-fold higher risk, clearly offering a learning opportunity in relation to the assignment of SAFs [33]. In this case, it indicates that the matrix and/or use SAF would need to be higher, 10 rather than 3.

One issue highlighted by this analysis is the possibility that the LLNA may be underestimating the relative potency of MDGN. The EC3 value used, 0.9%, is the lowest published figure (i.e., most potent); in alternate experiments, values up to twice this have been reported, effectively halving the potency, which would double the ppm limits calculated in this paper [34]. Nevertheless, these data do suggest a potency similar to that of isoeugenol, a well-recognized and important skin sensitizer in humans. It is worth noting that MDGN was reported to be negative in the guinea pig maximization test, the regulatory standard method used prior to the advent of the LLNA. Furthermore, MDGN was only weakly sensitizing in other non-standard methods [35–37]. Thus, the guinea pig, in this case, was even less accurate a predictor for humans, and no doubt these older data, in part, account for the relatively high levels of this preservative originally allowed within the EU. This serves as a reminder that both mouse and guinea pig assays are not perfect predictors and indicates the need for a degree of cau-tion to be applied to data interpretation, particularly when data are used to derive what are considered

to be safe use levels for consumers. Ultimately, con-sumer and dermatologist feedback (including proac-tive surveillance, where appropriate) will provide an indication of the accuracy of a risk assessment, but of course the aim should always be to avoid the genera-tion of ACD such that there are no clinical data [38].

Conclusion

A retrospective analysis of skin-sensitization QRA applied to MDGN indicates that it would have served to limit use levels to a degree, although for many products associated with MDGN allergy, not by more than a modest factor. This indicates a need to be conservative in the assignment of safety assessment factors, but also to recognize that relative potency predictions from animal models, while a very valu-able tool, are imperfect. Then, for ingredients such as preservatives, which may have widespread use in multiple products, additional allowance may be necessary, with the effect of reducing further the maximum use concentration of a preservative. QRA provides a tool for this since the daily skin dose from different products is calculated and can be combined to examine what a more realistic over-all exposure might be. Both for individual product use and products in combination, new approaches to probabilistic assessment seem likely to offer the prospect of a more meaningful evaluation of real-life exposure [20, 39]. Ultimately, though, it must be borne in mind that relative potency prediction, QRA, probabilistic exposure assessment, and so forth, are just tools to guide professional risk safety evaluation, not a replacement for it.

Acknowledgements

Declaration of interest

The author has no conflicts of interest and received no funding for the preparation of this manuscript.

References

1. Rodford RA. Toxicology of preservatives. In: Morpeth FF, ed. Preservation of Surfactant Formulations. Springer, London 1995:311–36.

2. Kabara JJ, ed. Cosmetic and Drug Preservation. Principles and Practice. Cosmetic Science and Technology Series. Volume 1. New York, NY/Basel, Switzerland: Marcel Decker Inc. 1984.

3. Basketter DA, Gilmour NJ, Wright ZM, Walters T, Boman A, Lidén C. Biocides: characterisation of the allergenic hazard of methylisothiazolinone. Cutan Ocul Toxicol 2003;22:187–99.

Cut

aneo

us a

nd O

cula

r T

oxic

olog

y D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Uls

ter

at J

orda

nsto

wn

on 1

1/13

/14

For

pers

onal

use

onl

y.

MDGN allergy and quantitative risk assessment 9

4. Gerberick GF, Ryan CA, Kern PS, Schlatter H, Dearman RJ, Kimber I, et al. Compilation of historical local lymph node assay data for the evaluation of skin sensitization alterna-tives. Dermatitis 2005;16:157–202.

5. Cronin E. Preservatives and antibacterials. In: Contact Dermatitis. Edinburgh, Scotland: Churchill Livingstone. 1980:664–713.

6. Sasseville D. Hypersensitivity to preservatives. Dermatol Ther 2004;17:251–63.

7. Rietschel RL, Fowler JF. Preservatives and vehicles in cosmetics and toiletries. In: Fisher’s Textbook of Contact Dermatitis. 6th edition; Hamilton, ON, Canada: BC Decker Inc. 2008:266–318.

8. Basketter DA, Clapp CJ, Safford BJ, Jowsey IR, McNamee PM, Ryan CA, et al. Preservatives and skin sensitisation quantita-tive risk assessment: risk benefit considerations. Dermatitis 2008;19:20–7.

9. Gerberick GF, Robinson MK, Felter SP, White I, Basketter DA. Understanding fragrance allergy using an exposure-based risk assessment approach. Contact Dermatitis 2001; 45:333–40.

10. Felter SP, Robinson MK, Basketter DA, Gerberick GF. A review of the scientific basis for default uncertainty factors for use in quantitative risk assessment of the induction of aller-gic contact dermatitis. Contact Dermatitis 2002;47:257–66.

11. Felter SP, Ryan CA, Basketter DA, Gerberick GF. Application of the risk assessment paradigm to the induc-tion of allergic contact dermatitis. Regul Toxicol Pharmacol 2003;37:1–10.

12. Basketter DA, Clapp C, Jefferies D, Safford RJ, Ryan CA, Gerberick GF. Predictive identification of human skin sensi-tisation thresholds. Contact Dermatitis 2005;53:260–7.

13. Basketter DA, Angelini G, Ingber A, Kern P, Menné T. Nickel, chromium and cobalt in consumer products: revis-iting safe levels in the new millennium. Contact Dermatitis 2003;49:1–7.

14. Api AM, Basketter DA, Cadby PA, Cano M-F, Ellis G, Gerberick GF. Dermal sensitization quantitative risk assessment (QRA) for fragrance ingredients. Regul Toxicol Pharmacol 2008;54:3–23.

15. International Fragrance Association, 2007: http://www.ifraorg.org/GuideLines.asp. Accessed May 01, 2009

16. de Groot AC, de Cock PA, Coenraads PJ, van Ginkel CJ, Jagtman BA, van Joost T, et al. Methyldibromoglutaronitrile is an important contact allergen in The Netherlands. Contact Dermatitis 1996;34:118–20.

17. Wilkinson JD, Shaw S, Andersen KE, Brandao FM, Bruynzeel DP, Bruze M, et al. Monitoring levels of preserva-tive sensitivity in Europe. A 10 year overview (1991–2000). Contact Dermatitis 2002;46:207–10.

18. Zachariae C, Rastogi S, Devantier C, Menné T, Johansen JD. Methyldibromo glutaronitrile: clinical experience and exposure-based risk assessment. Contact Dermatitis 2003; 48:150–4.

19. Johansen JD, Veien NK, Laurberg G, Kaaber K, Thormann J, Lauritzen M, Avnstorp C. Contact allergy methyldibromo glutaronitrile—data from a “front line” network. Contact Dermatitis 2005;52:138–41.

20. Safford RJ. The Dermal Sensitisation Threshold—a TTC approach for allergic contact dermatitis. Regul Toxicol Pharmacol 2008;51:195–200.

21. REACH Commission of the European Communities. Regulation (EC) No. 1907/2006 of the European Parliament and of the Council of 18 December 2006 concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH), establishing a European Chemicals

Agency, amending Directive 1999/45/EC and repealing Council Regulation (EEC) No. 793/93 and Commission Regulation (EC) No. 1488/94 as well as Council Directive 76/769/EEC and Commission Directives 91/155/EEC, 93/67/EEC, 93/105/EC and 2000/21/EC. Off J Eur Union, L 396/1 of 30.12.2006. http://europa.eu.int/comm/enterprise/reach/index_en.htm. Accessed May 01, 2009

22. European Chemicals Agency (ECHA). Endpoint specific guidance. In: Guidance on Information Requirements and Chemical Safety Assessment. Helsinki, Finland: European Chemicals Agency. 2008:256–88.

23. Miljoministeriet/Miljostyrelsen. Survey of liquid hand soaps, including health and environmental assessments. Survey of chemical substances in consumer products, no. 69, 2006. http://www2.mst.dk/udgiv/publications/2006/87-7052-062-3/html/default_eng.htm. Accessed May 01, 2009

24. Cronin E. Contact Dermatitis. Edinburgh, Scotland: Churchill Livingstone. 1980.

25. Frosch PJ, Menné T, Lepoittevin J-P. Contact Dermatitis. 4th edition; Heidelberg, Germany: Springer-Verlag. 2006.

26. Kanerva L, Elsner P, Wahlberg JE, MAibach HI. Condensed Handbook of Occupational Dermatology. Heidelberg, Germany: Springer-Verlag. 2004.

27. Divkovic M, Pease CM, Gerberick GF, Basketter DA. Hapten-protein binding: from theory to practical application in the in vitro prediction of skin sensitisation. Contact Dermatitis 2005;53:189–200.

28. Jacobs M-C, White IR, Rycroft RJG, Taub N. Patch testing with preservatives at St John’s from 1982 to 1993. Contact Dermatitis 1995;33:247–54.

29. Zachiariae C, Johansen JD, Rastogi SC, Menné T. Allergic contact dermatitis from methyldibromo glutaronitrile—clinical cases from 2003. Contact Dermatitis 2005;52:6–8.

30. Thyssen JP, Johansen JD, Menné T. Contact allergy epidem-ics and their controls. Contact Dermatitis 2007;56:185–9.

31. Wong CS, Beck M. Occupational contact allergy to methyldibromo glutaronitrile in abrasive cleansers and work creams. Contact Dermatitis 2001;44:311–2.

32. Marcano ME, Heras F, Conde-Salazar L. Occupational allergic contact dermatitis to methyldibromoglutaronitrile in hand degreasing toilet paper. Contact Dermatitis 2007; 57:126–7.

33. Pedersen LK, Haslund P, Johansen JD, Held E, Volund A, Agner T. Influence of a detergent on skin response to meth-yldibromo glutaronitrile in sensitized individuals. Contact Dermatitis 2004: 50:1–5.

34. Basketter DA, Gerberick GF, Kimber I. The local lymph node assay EC3 value: status of validation. Contact Dermatitis 2007;57:70–5.

35. Bruze M, Gruvberger B, Agrup G. Sensitization studies in the guinea pig with the active ingredients of Euxyl® K400. Contact Dermatitis 1988;18:37–9.

36. Wahlkvist H, Boman A, Montelius J, Wahlberg JE. Sensitizing potential in mice, guinea pig and man of the preservative Euxyl® K400 and its ingredient methyldibromo glutaroni-trile. Contact Dermatitis 1999;41:330–8.

37. Hausen BM. The sensitizing potency of Euxyl® K400 and its components 1,2-dibromo-2,4-dicyanobutane and 2- phenoxyethanol. Contact Dermatitis 1993;28:149–53.

38. Jowsey I. Proactive surveillance of contact allergies: an important component of the risk management strategy for skin sensitizers. Contact Dermatitis 2007;56:305–10.

39. European Cosmetics Association (COLIPA) Exposure Task Force. European Consumer Exposure to Cosmetic Products, 2005. http://www.cremesoftware.com/information/cases-tudies/colipa.php. Accessed May 01, 2009

Cut

aneo

us a

nd O

cula

r T

oxic

olog

y D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Uls

ter

at J

orda

nsto

wn

on 1

1/13

/14

For

pers

onal

use

onl

y.