Document 93vE9QgnJyO4M34dqNoY0Q6D3

I L Occup Environ M e d 1998;55:795-804 REVIEWS 795 Proposal for the assessment of quantitative dermal exposure limits in occupational environments: part 1 . Development of a concept to derive a quantitative dermal occupational exposure limit P M J Bos, D H Brouwer, H Stevenson, P J Boogaard, W L A M de Kort, J J van Hemmen Abstract occupational situation where dermal ex- Dermal uptake of chemicals at the work- pos& contributes notably to the systemic place may contribute considerably to the exposure. Further research should show total internal exposure and so needs to be whether this concept is more generally regulated. At present only qualitative applicable. warning signs-the "skin notations"-are (Occup Environ Med 1998;55:795-804) available as instruments. An attempt was made to develop a quantitative dermal Keywords: dermal exposure; occupational exposure limits; industrialhygiene occupational exposure limit (DOEL) .complementary to respiratory occupa- BACKGROUND FOR DERMAL OCCUPATIONAL tional exposure limits (OELs) The DOEL EXPOSURE LIMITS refers to the total dose deposited on the At the workplace toxic substances may enter 1` skin during a working shift. Based on the body through the respiratory tract, through available data and experience a theoreti- the gastrointestinaltract, and through the skin. cal procedure for the assessment of a It is assumed that the major routes of entry are DOEL was developed. A DOEL was the airways and the skin, although the intestinal derived for cyclophosphamide and 4,4- route may not be negligible due to the so called methylene &aniline (MDA) according to hand-mouth shunt and, secondary ingestion t h i s procedure. The DOEL for M D A was after respiratory exposure. tested for applicability in an actual occu- To protect people form detrimental effects pational exposure scenario. An integrated due to exposure to chemicals, several limit approach is recommended for situations values have been developed. For oral exposure in which both dermal and respiratory acceptable daily intake values (ADIs) have exposures contribute considerably to the been adopted for the general population, but internal exposure of the worker. The these are of limited value to workers. Several starting point should be an internal health types of quantitative occupational exposure TNO Nutrition and Food Research Institute, Division of Toxicology,Zeist, T h e Netherlands based occupational exposure limit-that is, the maximum dose to be absorbed without leading to adverse systemic effects. The proposed assessment of an limits (OELs) have been derived to protect workers from adverse health effects of toxic substances at work. Basically, if exposure is kept below the limit, no adverse health events PMJBos external DOEL is then either based on are expected in the workers (table 1). For D H Brouwer absorption rate or absorption percentage. respiratory exposure, the threshold limit value H Stevenson WLAMdeKort J J van Hemmen The estimation of skin penetration seems (TLV, set by the American Conference of Govto be of crucial importance in this con- ernmental Industrial Hygienists (ACGIH)) is cept. If for a specific substance a maximal most often adopted worldwide. In the Nether- Shell Research and Tecnhnology Centre, Shell International Chemicals BV, Department of Molecular Toxicology, Amsterdam, T h e Netherlands absorption rate can be estimated a maxi- mal skin surface area to be exposed can be assessed which may then serve the purpose of a DOEL. As long as the actual skin surface exposed is smaller than t h i s maximal skin surface area the internal OEL will not be exceeded, and therefore, no lands, maximum accepted concentrations (MAC values) are in use. As well as external exposure limit values, internal values exist for several xenobiotics for systemic exposure, known as biological limit values (BLVs). Examples are the biological exposure index (set by the ACGIH)' and the P J Boogaard systemic health problems would be ex- German "BAT-Werte" (set by the "Deutsche pected, independent of the dermal dose/ Forschungsgemeinschaft").2 Recently, the Correspondence to: Dr J J van Hemmen, TNO unit area. If not, the DOEL may be Health and Safety Executive in the United Numtion and Food Research interpreted as the product of dermal dosel Kingdom has introduced biological monitoring h t i N t e , Division of Toxicology,Zeist, The Netherlands unit area (mglcm') and exposed skin guidance values for six substances.' The main surface area (cm'). The proposed concept advantages of a BLV relate to (a) its independfor a DOEL is relevant and can be made ence of the route of entry, and (b) its use in Accepted 16 July 1998 applicable for health surveillance in the assessing an overall health risk, as monitoring of Tobk 1 Some characteritrics of available exposure limits .Rmrte of enm ~~~ Respiratory rracr Garminresnnd macr Skin Miscellaneous OT combined Name Qualitative or quantitative Target population Dimensions Monitoring methods Maximum accepted Acceptable daily intake (ADD Skin denotation concentration (MAC) Threshold limit value (TLW Quantitative Quantitative Qualitative Working population mg/m' parts per million @pm) fibres d m ' Environmental monitoring (EM) Personal air sampling (PAS) General population meng food mgkg body weight Food residues or contaminaotsin Combination with food intake data No specificworker monitoring method Working population Not applicable; however likely to be assessed as mg (mdan3 For example, environmenul surface wipe-off; patches, gloves, coveralls;tracer methods; skin washings; or skin snipping Biological limit value; (BEI, BAT-Wene, biological monitoring guidance value) Quantitative Working population or general population (a) mg/l blood, mg/l urine, mdm' exhaled air (b) cholinesteraseinhibition, zinc protoporphyrin,DNA adducts, mutations, e o Biologicalmedia:blood, urine, exhaled air, faeces, hair internal (systemic) exposure provides infor- mation on the result of exposure through all possible routes of entry. Disadvantages are that local effects are not covered and only a few adequate monitoring methods based on human toxicokinetic data are available. Specific strategies for the sampling of biological fluids may be necessary. Furthermore, the position of blood samplingrelative to the skin exposuremay be critical for the outcome,' and possible skin metabolism is not generally taken into consideration.' No quantitative limit values exist that may protect against adverse effects from uptake through dermal exposure. A qualitative "skin notation" is generally the only available warning against absorption of chemical substances through the skin. This, however, only points to a potential risk but gives no indication of the actual contribution of skin exposure to the body burden. For example, 2-methoxyethanol has a skin notation, but it was estimated by Pausten- bach that dermal uptake by exposure of one hand for 30 minutes was 4.5 times the amount taken up by inhalation for eight hours to the TLV of 5 ppm.6 Furthermore, clear discrepan- cies in the denotation practice exist between countries. Besides, the absence of a skinnotation does not imply that risk due to skin exposurecan be ignored.' The scientific basis underlying the skin notation in the TLV list of the ACGIH is not always well In The Nether- lands, a skin notation is applied by the Dutch Expert Committee on Occupational Standards (DECOS) predominantly based on a strategy proposed by European Centre for Ecotoxicology and Toxicology of Chemicals (ECETOC).9It is well known that uptake through the skin of sev- eral chemicals can contribute considerably to the internal exposure. For many chemicals-for example, pesticides-the skin is the predomi- nant route of entry. For adequate protection of workers against adverse health effects of these compounds, it is important to regulate exposure through dermal uptake rather than, or as well as, pulmonary uptake, with preference for a quanti- tative variable." The importance of regulating dermal exposure has also recently been stressed at a European meeting funded by the European Community," which led to&e institution of the Dermal Exposure Network. At the request of the Dutch Ministry of Social Affairs and Employment a concept was developed for a more quantitative health based dermal occupational exposure limit (DOEL). A practically applicable DOEL may serve as a useful tool in the policy of protecting workers' health and provide the government with a valuable and applicable instrument for the regulation of skin exposure at the workplace as well as other OELs. The process was divided into three phases, (a) the development of such a concept, (b) assessment of a DOEL for two substances according to this concept, and (c) testing the applicability of one of these DOELs in an actual occupational environment. The present report describes the first two phases. The third phase is presented in an accompanying paper." The present reports are considered to serve as a starting point for the development of DOELs, and are not meant to be the final guidance for the derivation for a DOEL. General aspects of setting priorities of substances and the different levels at which a DOEL may be set (internal or external exposure) are considered, and the toxicological basis for this is discussed. As a DOEL is meant to be an instrument to keep the internal dose below a certain level the estimation of the actual dose taken up will be crucial in thisconcept. This can be approached in two ways, absorption rate and absorption percentage. Both possibilities are presented followed by a discussion about the difficulties associated with the estimation of dermal uptake in the occupational situation. The uptake is dependent on many factors that will vary in different occupational settings. As illustrative examples, DOELs are calculated according to the proposed concept for cyclophosphamideand 4,4'methylene dianiline (MDA). PRIOIUTY SETIWG FOR THE ASSESSMENT OF DERMAL OCCUPATIONALEXPOSURE LIMITS A DOEL, in combination with other types of standards, may provide a more complete set of exposure limits, which together allow for a risk assessment of a workers' health relative to chemical exposure at work (fig 1). It will be neither possible nor necessaryto assess a DOEL for every possible substance. Assessment of a DOEL will be useful for all chemicals for which adequate protection of the worker demands 797 Systemic effect .. ..7,.. Biologicali limit i value j I patterns indicative of skin uptakecyclophosphamide2'-may clearly point to the necessity of controlling dermal exposure rather than or as well as respiratory exposure. internal dose Basic elements for a dermal occupational exposure limit OUTLINE AND DEFINITIONS .-. The 0EL.s are assessed to control the internal exposure of the worker exposed to chemicals, Intestines Skin to serve to keep the internal exposure below a concentration or dose at which no adverse ;Maximum I Acceptablei Skin i ;j Docecrumpaaltional ;accepted j denotation: ;exposure health effects are expected, or below an accepted level. Similarly, a DOEL should rep- Concentration; intake i i ; l j y ~ j ~ ~res~en!t the maximum amount of substance (mg) deposited on the skin surface within a I 1External dose given time (usually a workshift), without giving rise to adverse systemic health effects. Figure 1 Exposure mutes and limit values. For the dermal mute the limit (skin denotation) i s of a qualitative nature. The dermal occupariaal exposure limit (DOEL) couldjiU Fenske defined dermal exposure as the product of skin loading rate (mass per skin sur- this gap. face area per unit time) and area exposed (cm2)."Dermal exposure is expressed in units monitoring skin exposure on a regular basis. of mass per unit time Ogm).Cheme and Rob- Therefore, criteria should be developed to set a ertson proposed an alternative definition which priority list of chemicals.Priority settingmay be recognises the biological process involved in based on data obtained from Merent view- skin absorption-that is, the concentration of points and sources. Chemicals for which a skin the substance at the skin ~urface.'~ notation could be assessed as well as an OEL for Dermal exposure at the workplace is gener- respiratory exposure may be selected. For ally intermittent. Due to the reservoir function example, Fiserova-Bergerova et ul proposed the of the skinJz4internal exposure may continue comparison of rates of skin penetration and pul- even after the external exposure has ended. monary uptake at the level of an OEL as a crite- Occupational exposure may be due to contact rion for relevant dermal uptake.I3l4 However, with contaminated surfaces, gases, aerosols, these chemicals are generally selected according liquids, and dusts. Two expressions of dermal to criteria for priority setting for respiratory exposure have been defined.25Firstly, potential exposure-for example, based on the prevention dermal exposure, which is the total amount of of respiratory tract irritation-which are not chemical (mg) deposited on the worker, either necessarily applicable to dermal exposure. on the (protective) clothing or the bare The ECETOC and the US Environmental (uncovered) skin. Secondly, actual dermal Protection Agency (US EPA) published docu- exposure, which is defined as the amount of a ments that may provide some useful information chemical actually coming into contact with the on this topic.9l5 Further, specific groups of sub- bare (uncovered) skin, including the h c t i o n stances related to either structure or effect may transferring through (protective) clothing to be identified for which a DOEL is either neces- the underlying skin, and which is, therefore, sary or unnecessary. Examples of substances available for percutaneous absorption. Prefer- related to structure are the glycol ethers,I6 ably, insight on actual exposure conditions whereas examples of those related to effect are (frequency and duration of exposure, skin sur- defatting agents which may either be taken up face area, and location) should be present for quickly themselves, or increase skin uptake of the derivation of an accurate DOEL. Other- other compounds present. wise, a DOEL may turn out to be too For physicochemical variables, a molecular conservative because of several conservative weight 4 0 0 or a log P, between -1 and 4 may assumptions that have to be made. indicate potentially considerabledermal uptake; vapour pressure and boiling point are also POSSIBLE LEVELS FOR SETTING A DERMAL important ~ariab1es.I~ For toxicodynamic OCCUPATIONAL EXPOSURE LlMlT variables, dermal toxicity data compared with It is obvious that a DOEL has to be assessed so those obtained by other routes may provide that testing for compliance can be performed suitable information-for example, comparison with (relative) ease. Theoretically, DOELs may of a dermal LD, and intravenous LD50may be be set at the following levels: indicative, except in those cases where the rate of The internal level (setting a BLV) appearance in the systemic circulation is impor- The level on the skin surface (mg tant. Also, actual or estimated occupational deposited on the skin) exposure conditions will also provide important The level in the occupational environment information for priority setting. Models for the (the amount of chemical present on sur- estimation of dermal exposure for specific faces of working equipment, or pesticide activities have been developed; description of residues). these models are beyond the scope of this report. As already mentioned, the first level, assess- Occupational case reports of systemic toxic ment of a BLV, has clear advantages and is the effectsin the absence of respiratory exposure- level to be considered first. However, it is at the for example, MDA'*-r urinary excretion moment, limited in its use. The third level has Bos, Brouwet, -S ,e1 a1 been used to investigate crop -g a- ties where skin contact may o c m *&pesticide residues on the crop.2bIn other occupational environments, it Was thought that it would be more complex to set a DOEL at this level and to test it for compliance. Furthermore, regdating skin exposure at the third level is considered to be a derivation of the second level. Therefore, this report focuses on the second level, whether an applicable and useful external DOEL can be assessed for the amount of chemical deposited on the skin. It is hoped that the DOEL will be a useful tool in controlling the internal exposure as a result of skin uptake and should thereforebe related to either the maximal internal dose that is expected not to affectthe health of the worker, or to a generally accepted risk level. BASIS FOR A DERMAL OCCUPATIONAL EXPOSURE LIMIT The maximal internal dose (based on the no adverse observed effect level (NOAEL)) can either be derived from human data (preferably) or from animal data. Such data should relate the external dermal dose directly to health effects, leading automatically to an acceptable value of a DOEL. Suitable human data are usually lacking,therefore, data for other exposure routes or animal data may be used to derive a DOEL. Acceptance of a pivotal role for the internal exposure (concentration of the toxicant in the central compartment) is then essential in the process of the extrapolations from animal to human and route to route. If based on animal data, data from dermal toxicity experimentsare preferred, provided that experimental exposure conditions resemble those at the workplace (exposed surface area, dose or concentration per surface area, exposure time, climatological conditions, etc)." The assessment of a DOEL fiom these experiments through direct extrapolation will generally include high to low dose extrapolation. It should then be taken into account that the absorption percentage may increase with a decreasing dermal area dose (0,mgkm'). In the case of a high D, (infinite dose1*)a considerable amount of the substancemay not be absorbed during the exposure period or workshift and may be finally wiped off;the absorption percentage will then be much less than 100% of the applied dose. How- ever, for a low D, (finite dose") the exposure period may be long enough for the deposited amount of substance to be (almost) completely (or maximally) absorbed. Thus an absorption percentage experimentally derived for a specific dermal dose/unit area cannot be generally applied to other doses." '*Linear extrapolation of an estimated absorption percentage to lower dermal doses may, therefore, result in an underestimation of the amount absorbed. It was con- cluded by ECETOC that absorption data expressed as a percentage of applied dose absorbed per unit of time are relevant only to a particular dose and a particular time." In the practice of setting standards, OELs for systemiceffects have often to be based on oral or respiratory animal experiments, as suitable dermal toxicity studies are usually absent for most chemicals.In either case, route to route extrapo- derivation is given in a recent report:") lne HBR-OEL, is defined as the maximal interna1 dose (mglday) not leading to adverse health effectsfor the worker. A N o ~ Lis translated into an HBR-OEL, with toxicokinetic data (mainly bioavailability data) for the exposure route the NOAEL is based on. An external DOEL is derived fiom an HBR-OEL,, with toxicokinetic data for dermal exposure-for example, dermal permeability constant, penetration flux, or absorption percentage. Because route to route extrapolation is the method extended farthest and most often applied in risk assessment for dermal exposure in the occupational situation, the proposed concept is based on this method. The final procedure proposed will also be applicable for direct extrapolation. Proposal for a standard:the dermal occupationalexposure limit Crucial for the translation of an HBR-OEL, into an external DOEL as defined in the previous section, is the estimation of the dermal uptake which may be based on estimated or measured absorption rate or absorption percentage. ASSESSMENT OF A DOEL BASED ON ABSORPTION RATE (FLUX) If adequate data on absorption rate are available, a refined derivation of the DOEL is possible. For the undamaged skin, the variable for the steady state absorption rate is the flux3 (mg/(cmzxhour)),defined as D A C pick's fmt law of diffusion), where K is the permeability constant, and AC is the concentration gradient across the stratum c o m e ~ m . ' ~ The absorption rate is, among others, dependent on DA (actually on the concentration in the vehiculum),15and will increase with increasing D, until a steady state flux is reached. A further increase of DAwill then not result in a higher rate of uptake. For the purpose of the assessment of a DOEL, the maximal flux derived under exposure conditions relevant for the occupational situation U-,+J should be the basis. The internal dose (mg) is then determined by the product J,,,,,,x'lkA, where T is duration of exposure (hours/day)-that is, the time from the onset of dermal exposure until the exposure is ended and the skin is cleaned, and A is the exposed skin surface area (cm'). The maximal internal dose is then equal to J m m ~ , x l k AS. tarting from a maximal accepted internal dose (HBR0 E L r(mglday)), it follows: J,,,,x~~HBR-OEL,, or ASHBR-OEL/U,,,,XT). So, under the assumption of a specified exposure time (default: T=8 hoursiday) the internal dose depends only on the exposed skin surface area A. This means that a maximal allowable exposed skin surface area (Ammc)an 799 experimental conditions, including the expo- A Da =. b (ID):A a ~ sure period and the concentration.`' '" Without eQ Da < b (FD): Da A c DOEL ~ further knowledge, the default value for 0n maximal dermal absorption is set at loo%, 0 Ub unless experimental data or physicochemical 9e parameters may point to a lower maximal -9 absorption percentage (see next section). The m HBR-OEL,,, (mg'day) may then be translated n into an external DOEL by dividing the HBR-OEL,,, through the absorbed fraction (F) a Exposed skin surface (A) of the substance: HBR-OEL,,,/F. This external DOEL can then be interpreted as the product of two variables, the dermal d o s e h i t area and Figure 2 Graphical presentatian of the dermal occupational exposure limit (DOEL) rekuive w the dermal doselunit area (DJ and the exposed skin surface area (A). riteAUC (shaded area) representc the "safe"valuesfor D,,xA. For D,, 2 b (when a maximalflux rekvantfor the occupational situation (3-d b reached), the DOEL can be set asA , (=a), and is indepedmt ofthe dermal doselunit area. As long as A<a, the absorbed abse is not expecred w give rire w &e sysrmtic health e&n. IfA'a, either the exposure time or D, should be reduced k,the DOEL can be expressed as the multiplicationof the h a l doselunit area and the e x p o d skin surface area D,,xA. the exposed surface area (A; cm') for a given work shift-that is, D,,xA. For testing whether exposure conditions comply with the external DOEL two ap- proaches are now possible. Firstly, a single default value for A can be set for every occupa- tional situation and 4,can then be determined at the workplace, or secondly, both parameters may be determined at the actual workplace and the product can be calculated. In the first situ- be defmed (fig 2 in which the theoretical ation, 0,is the single variable to be monitored DOEL expressed as DAxA is graphically for compliance. This is a rather conservative presented), so that ifASA,,, no adverse health and rigid approach. In the second situation effects are to be expected because the HBR- both variables are to be monitored and actual OEL,, will then not be exceeded. If D A 3b (the exposure conditions can be taken into account. dermal area dose ar which 3ma;ias creached) a A default value for A may be based on maximal penetration rate is reached, a further knowledge of the actual occupational exposure increase of -0, will not lead to a higher situation. Estimates for the surface area for di- absorbed dose during a specified T. The ferent parts of the body have been pre~ented.'~ amount taken up is then independent of DAand For a given value of the DOEL (defined as depends only on the exposed surface area A. DAxA(mglday)) DAis allowed to be higher for Thus as long as A<u (fig 2), the absorbed dose smaller values of A. As already noted, if D, will not exceed the HBR-OEL,; the DOEL increases, the absorption percentage will stay can then simply be set as A,, (=u,calculated as equal or will decrease when J,,,, is reached. HBR-OE&jUma;sD). In case & A m , the So, if for a given DOEL the actual A is much internal exposure can be reduced by either smaller than the default value on which the diminishing the exposure time T or reducing DOEL is based the amount of substance the flux3. Then, as stated above,Jis dependent absorbed will be overestimated,the overestima- on DA and the fluxcan be reduced by decreas- tion will increase as the actual A deviates more ing 0,. This situation is similar to that from the default value. described later with an absorption percentage as the starting point, the external DOEL can SUMMARY then be interpreted as the product D,xA (see The above considerations have been summa- later). As long as the value for DAxAassessed rised in table 2. If 3,=, can be estimated, a for an occupational situation lies in the shaded maximal skin surface area to be exposed can be area of the curve (the AUC in fig 2), the HBR- calculated and the DOEL can be expressed as O E L , will not be exceeded. A, (cm'). As long as ASA,,, no adverse health effects are to be expected. If A>A,,, D., ASSESSMENT OF A DOEL BASED ON ABSORPTION (mg/cm2)has to be estimated and the DOEL PERCENTAGE can be expressed as D,xA (mg). If no adequate data are available to estimate an IfJme;mits unknown, the skin uptake has to be appropriate absorption rate a DOEL has to be estimated with a relevaxit absorption percent- based on absorption percentage. As already age. The DOEL can then be interpreted as mentioned, an absorption percentage esti- D,xA. It is recommended that a DOEL should mated for a specific D,4is not commonly appli- be derived relative to a standard exposed cable for other values of D,+;the absorption surface area. If the actual exposed skin surface percentage may increase with decreasing 0,. area deviates from this standard area, the Dq The percentage absorbed depends on the can be adjusted accordingly. For instance, for a Table 2 Summary of the possibilitiesfor the setting of a dermal occupational exposure limit (DOEL) given maximal value for D,xA (the DOEL) D,, is allowed to be higher if A decreases. But then the absorption percentage may decrease, which Starting puim DOEL expressedas means that if the appropriate data are available, Biological monitoring Biological limit value Maximal penetration rate Am,(ifJmy., is known and applicable) D.,xA (ifJ*-,-. is unknown) Absorption percentage D,,XA a relatively high DOEL may be set for occupational settings where the actual exposed skin surface area is expected to be small, and a rela- A-xposed surface area;A,,=rnaximal surface area to be exposed;Jm7,,,,=maximalflux derived tively low DOEL for a worst case default value under exposure conditions relevant for the occupationalsituation;DA=dermal area dose. for A. The estimation of dermal up& absorption percentage will be more appliaMu1e with a finite dermal dosdunit area, w h e a s approach with an absorption rate (flw)can k applied in situations where an infinite dermal dosehnit area is present." To illustrate this D O E L are calculated for MDA and cyclo- phosphamide in a later section. Estimation ofdermal penetration OCCCTATIONALFACTORS AFFECTING DERMAL PENETRATION In the procedure for the assessment of a DOEL as proposed in the previous section, the estimation of dermal uptake, either expressed as a flux or as an absorption percentage, is of crucial importance. Comprehensive reviews on dermal penetration of chemical substances already '*e x i ~ t . " ' ~24 29-31 These reviews teach that factors affecting dermal penetration can be divided into three distinct categories: (a) substance related factors-for example, physi- cochemical properties such as molecular weight, octanol-water partition coefficient, volatility, and polarity (ionisability); (b) situation related factor-for example, environmental temperature, humidity, the presence or absence of occlusive material (clothing, gloves), the time frame of the exposure (duration and frequency)); (c) skin related factors-for example, anatomical site, physically damaged skin. The compound will in practice be irregularly divided over the exposed skin area. However, the DOEL will usually relate to a continuous 8 hour shift exposure with a constant exposure level. Generally, it can be assumed that when the amount of substance is irregularly divided over the exposed skin surface area, the amount taken up will be equal or less than when the division is regular (under the assumption that the flux is roughly equal over the entire exposed surface area). As a worst case assumption, it is assumed that the total amount is present on the skin during the entire workshift. In assessing a DOEL it is assumed that dermal penetration is measured for the pure substance. If a substance is part of a mixture, dermal penetration may vary greatly with the actual composition. Also, skin contact may affect the solubility of the compound in the vehicle, and, therefore, the absorption." It is important to note that not only the dermal dosehnit area, but the concentration in the vehicle at the skin surface is of importance.15 For equal absolute amounts of a substance applied per cm2,different concentrationsin any vehicle at the skin surface may lead to different absorption rates or percentages. This may at least partly explain the difference in absorption percentages for MDA found by two groups of investigators." 34 In the present report, it is assumed that the concentration in the vehicle (solution) is more or less constant. The starting point is a pure substance or a constant concentration of the substance at the skin surface. Furthermore, it should be realised that occupational exposure may be due to contact with contaminated surfaces, gases, aerosols, liquids, ESTIMATION OF DERMAL P E m n O N The estimation of dermal penetration is difficult due to the fact that absorption is influenced by several substance, situation, and skin dependent variables. Recently, US EPA and ECETOC have discussed several in vitro and in vivo techniques for the estimation of skin penetration in detail.I5 The reliability of the estimated DOEL will increase if skin uptake can be estimated more precisely for the occupational exposure conditions. The US EPA proposed several equations for the estimation of the dermally absorbed dosefday for aqueous solutions and v a p o ~ r s . ~ ~ The US EPA based the estimation of dermal penetration predominantly on the estimation of the permeability qonstant; equations for the estimation of this constant were presented for aqueous solutions and for vapours for steady state and non-steady state situations. The theoretical and experimental considerationsfor the estimation of an appropriate permeability constant in different exposure situations were reviewed in detail. The US EPA considered non-steady state conditions to characterise actual exposure more closely than do steady state conditions, especially for exposure peri- ods which are relatively short compared with the lag time necessary to reach a steady state flux.'5It is recommended that the usefulness and applicability of these approaches for the occupational situation be studied. L u n g and Paustenbach reviewed some important principles involved in the assessment of percutaneous absorption and discussed some possibilities for a quantitative determination of chemical uptake through the skin." Although a mathematical model may provide a proper tool for estimating percutaneous absorption, the available models still need further validation (see previous section). The same authors proposed three indirect methods for the estimation of the dermal bioavailability of a substance: by comparison of the area under the plasma concentration-time curve after cutaneous and intravenous administration (preferably performed with labelled material), by estimating the total amount excreted relative to the administered dose, and by measurement of the amount of substance remaining on the skin at the end of e~posure.~' The first method may be the most accurate; SO 1 however, these data are seldom available in risk DERIVATION OF A DOEL FOR CYCLOPHOSPH.L\UDE assessment. Risk evaluation Recently, a tiered approach for the estima- Key study for the risk evaluation of c y d o p h ~ - tion of dermal absorption has been descrild." phamide was the oral experiment with rats by Briefly, this approach starts with a default d u e Schmiihl and Habs." Under the assumpnon of of 100% for dermal absorption when no data 100% oral absorption and based on the pn34- are available. In the next tier molecular uwght ously mentioned method for derivanon d or the log Pm were proposed to discrimmate HBC-OCRV for genotoxic carcinogens' an between poorly absorbed substances, and sub- incidence of 0.97/mg cyclophosphamide stances for which 100% absorption may be a absorbedkglday for malignant tumouw w a s reasonable estimate. It was stated that although calculated for male rats. The incidence per mg no clear relation was presented, absorption absorbed cyclophosphamide per day for the would be considerably <100%if the molecular worker (70 kg; 40 year exposure for five weight is greater than 500. For log Pm, daydweek) is then 5.3~10-'.Reference d u e s maximum absorption was associated with for a working life (40 years) additional values between 1 and 2, whereas for log P, mortality incidences of 4/1000 and 4,100 values <-1 or >4 the dermal absorption was are requested by the Dutch Ministry of Social considered to be <10%." l9 Affairs and Employment. The daily internal Formulas for the calculation of the flu.or cyclophosphamide doses associated with these the permeation coefficient based on the log Pm excess cancer levels are 0.75 mg and 7.5 pg, and the molecular weight have been respectively. proposed.13However, the US EPA warns that the log P, will not be a valid parameter for the Dermal absorption estimation of lipophilicity for certain classes of Key studies for the estimation of dermal chemicals-for example, nitro phenol^.'^ absorption were the volunteer studies of Hirst Recently, the validity of five of these models et &,'I the studies with cancer parienw by has been evaluated by comparison with wri- Mouridsen et al: " and the study with rats by mental in vitro permeation coefficients by Sessink et ~ 1Th.e a~nim~al study showed a uri- Wilschut et ~Tlh.ey'co~nsidered a revised ver- nary cyclophosphamide excretion of 5%-7% sion ofan unpublished model by Robinson to after dermal and intravenous administration, be the most appropriate for the estimation of indicating 100% dermal absorption. The skin penetration from aqueous solutions. How- human studies indicated that after intravenous ever, the usefulness of this model in actual and dermal administration, about 10%and 1% occupational situations remains to be investi- of the administered dose, respectively, was gated. excreted in the urine in 24 hours as cyclophos- The usefulness of in vitro and in vivo test phamide. The human and rat data are difficult systems has been reviewed recently." I' The to compare, due to significant differences in results of in vitro studies for the determination of skin penetration are difficult to compare with those obtained in vivo. Standardised experimental conditions and the use of reference compounds for calibration should improve the comparability of the test systems. In general, animal skin seems to be more perme- able than human skin." A single default correction factor cannot be derived, because the extent of overestimation seems to be specific to the agent and animal. We recommended that animal absorption data is considered as an overestimation of absorption in humans." l9 exposure conditions. In the rat study cyclophosphamide was applied in a glycerol suspension which penetrates easily into the skin, and therefore may have enhanced the absorption. In the human volunteer study, cyclophosphamide was applied in methanol which evaporated within 30 seconds, the skin was covered and washed with water and soap after 6 hours. As the exposure conditions in the human volunteer study were considered to be more comparable with the occupational conditions (watery solutions; exposure for 6 hours before cleaning) this study was used as a starting point for the estimation of the absorption. Testing of the applicability of the concept Based on these human studies the total The applicability of the procedure described urinary excretion of cyclophosphamide within for setting a DOEL was tested by assessing a 4 days after intravenous administration (0.02 DOEL for two genotoxic carcinogens cyclo- mgkg) was about 13%. A total of 2%-3% phosphamide and MDA. The principle is simi- cyclophosphamide was estimated to be ex- lar for compounds for which a threshold value creted in urine after dermal application of 1mg I for the expression of toxic effects can be (100 pgkm'; occlusion).So, urinary cyclophos- assessed. The assessments based on literature phamide excretion after dermal application i i searches are briefly described, more details are was maximally 25%-30% of that after a r reported separately (in Dutch)." For both comparable intravenous dose. I compounds, relevant dermal toxicity data were absent; an HBR-OEL, was derived as a start- Estimation of a DOEL for actual exposure 1 ing point. These values were calculated accord- For preparation and application of cyclophos- J ing to the Dutch method for the calculation of phamide as a drug it is expected that dermal I health based calculated occupational cancer exposure is limited to the hands and lower I risk values (HBC-OCRV) for genotoxic arms, an area of about 2000 cm'. Starting from I carcinogen^,'^ as adopted by DECOS, a an absorption percentage of 30%, a daily intercommission of the Health Council of the nal dose of 0.75 mg equals a 0,of (750x100/ J Netherlands. 30)/2000=1 pgkm'. However, the absorption percentage of 30% was estimated based on a ence value equals a dermal dose/unit area of 80 D,,of 100 pg/cm'. Considering the facr that ngtcm'. absorption percentage may increase with de- creasing D,,an absorption percentage of 100% Discussion and conclusions is assumed at a dermal doseiunit area of about 1 pgkm2.Therefore, the DOEL interpreted as D,xA, was set at 0.75 mgiday. For an estimated maximum value for A of 2000 cm' the D.,will The present report presents a first attempt to develop a procedure for the assessment of a relevant and useful quantitative parameter for controlling dermal exposure. It is meant as a be 750/2000=0.4 pgkm?. Similarly, the cyclo- starting point for further discussion to develop phosphamide dose of 7.5 pg associatedwith the a practically suitable and applicable DOEL for lower reference value equals a D,, of 4 ndcm'. occupational situations. The importance of the conmbution of skin uptake to the total human DERIVATION OF A DOEL FOR 4,4`-METHYLEhT exposure for many chemicals is widely Dm recognised." Is Is Initiatives in the United Risk evaluation States'@" and the establishment of a European The oral NTP study was the main source of network on dermal exposure supported by the data for the evaluation of the carcinogenic risk European Commission should provide impor- of MDA.45Under the assumption of 100% oral tant information in the near future for absorption and based on the method for refinement of the proposed procedure. For derivation of a HBC-OCRV for genotoxic car- some chemicals in the workplace and for cinogens already mentioned" an incidence was certain workplaces skin uptake will be the pre- found of 4.67x10-' per mg MDA absorbedlkgi dominant or only exposure route. day for neoplastic noduli for male rats. The Dermal exposure is merely controlled by a incidence per mg absorbed MDA per day for a qualitative parameter. One of the reasons for worker (70 kg; 40 year exposure for five this is that the problems associated with quan- daydweek) is then 0 . 2 5 ~ 1 0 -T~h. e daily inter- titative standard setting for dermal exposure nal MDA doses associated with the reference are thought to be complex. It is acknowledged values of 4/1000 and 4/100 000 are 16 mg and that for the proposed procedure several as- 0.16 mg, respectively. sumptions and simplificationshad to be made. However, many of the drawbacks and assump- Dermal absorption tions mentioned are not unique to dermal The penetration of MDA in the skin is fast but exposure and are basically not different from not complete, even after application of low those associated with oral or respiratory expo- doses about 50% could be washed off. Gener- sure. For instance, in setting standards for ally, the absorption percentage decreased with occupational respiratory exposure, combined increasing doses although the absolute amount exposure is seldom dealt with, ventilation rate taken up remained more or less the same. In (and, therefore, uptake) may be concentration vitro studies showed that about twice as much dependent," work load and working conditions MDA penetrates human skin as rat skin. may influence breathing rate and depth, and Occlusion increased the amounts taken up by the ratio of mouth versus nose breathing may twofold to 2.5-f0ld.~~ also be of importance. Furthermore, pharma- Absorption studies with volunteers have cokinetic data for respiratory exposure are been carried out. Exposure conditions in most often lacking and default values for absorption studies deviate from occupational conditions. have to be used. All these factors, which are In one experiment with exposure conditions seldom accounted for, will determine the resembling those at the workplace MDA (10% pulmonary uptake of chemical substances, and solution (w/v) in ethanol; DA= 0.6 mg/cmz)was therefore the internal dose. applied to the forearm of two volunteers, with- Also, in respiratory exposure the amount out occlusion.After three hours 41% and 47%, actually absorbed is for some substances respectively, could be washed Based on related to the proportions of differently sized these results an absorption percentage of 55% particles, defining the inhalable and respirable was derived for MDA. fraction. Certain groups of substances-for example, fibres, dusts, and aerosols-require a Estimation of a DOELfor actual exposure specific approach. It may, therefore, be relevant In general, occupationalskin exposure is limited for health risk assessment to consider whether to hands and lower arms, an area of 2000 cm'. a substance is taken up by the nasal mucosa or Startingfrom an absorption percentage of 55%, the alveoli. a daily internal dose of 16 mg equals a D, of Thus, also for other routes of exposure, (16 000~100/55)/2,000=15 pgkm'. However, absorption will be dependent on characteristics the absorptionpercentage of 55% was estimated related to substances and situations and to based on a D, of 0.6 mgkm2.Considering the physiological variables. Nevertheless, many fact that the absorption percentage for MDA OELs for respiratory exposure serve as useful will increase with decreasing dermal doselunit tools for regulating exposure to prevent adverse area, an absorption percentage of 100% is health effects in workers. assumed at a dermal d o s e h i t area of about 15 In table 2, the possible assessments of a p&mz. Therefore, the DOEL interpreted as quantitative DOEL are summarised. The total D,xA was set at 16 mg a day. For an estimated internal exposure (including dermal) may be maximum value forA of 2000 cm' the D,will be controlled by a BLV. The suirability ofa BLV 16 000/2000=8 pgkm'. Similarly, the MDA for this purpose will be dependent on the use- dose of 0.16 mg associated with the lower refer- fulness of specificvariables in biological tissue 803 and the availability of an appropriate m%hod for analysis. Often, a specific stratem for sampling biological fluids and analysis will be demanded, especially when both respiratory and dermal exposure are of importance. Both routes will probably result in different internal exposure patterns and, therefore, different urinary excretion patterns. If an external DOEL is considered to be appropriate we recommend assessment of t h i s DOEL relative to the surface area of exposed skin. Generally,the DOEL will be derived from an HBR-OEL, derived from animal toxicity data. Translation to an external DOEL is pref- erably performed with an appropriate estimation of the penetration rate of the substance under occupational conditions. If a J,,,, can be estimated it may be possible to define a maximum skin surface area (A,, (an'))to be exposed for a given T. If the actual exposed surface area is SA,,,=no health risk is indicated for dermal exposure, independent of the dermal dose/unit area. If a J , , , cannot be derived or ifA>A,, the approach will be the same as when an absorption percentage serves as a basis, the DOEL can then be interpreted as DAxA. Compliance can then be tested by monitoring either both D, and A or only DA starting from a default value forA. Iffor a given DOEL A decreases, D, is allowed to increase. Hence, if the actual exposed surface area-for example, hand only exposure-is much smaller and the D,is allowed to be, and the DAis much higher than the initial values on which the DOEL is based, lower absorption percentages can be used in assessment of health risk. This means that for a relatively small area of exposed skin the DOEL expressed as DAxA,may be set at a higher level. So, for a specific substance the DOEL may be set at different levels depending on the actual surface area of exposed skin. A quantitative DOEL can be used to control dermal exposure, but also, in combination with dermal exposure data to rank substances for possible risks at the workplace. Also, it might be helpful to manufacturers of personal protective equipment for the development of appropriate products. The present concept is mainly developed for substances that act systemically.In general, the occurrence of local effects will be predomi- nantly dependent on D,,the actual concentra- tion at the skin surface. Therefore, if a dose-effect relation for local effects is available, local effects may be regulated by assessing a maximum value for 0,as well as a maximal value for the DOEL expressed as D,xA. A quantitative DOEL will be a useful type of exposure limit complementary to respiratory OELs at the workplace. Together they may provide a more complete set of exposure limits which allow assessment of a worker's health risk relative to chemical exposure at work. This is at least the case orthose occupational situa- tions where the uptake through respiratory exposure is far less than through dermal exposure. An integrated approach, summation of the internal exposures through both routes, is recommended for situations where both exposure routes contribute considerably to the internal exposure of the worker. In that case, a DOEL as well as a respiratory OEL may be set, taking into account that combined exposure through the dermal and respiratory route at the same time should not result in a health risk. Exposure to both routes should be controlled together. A concept for a procedure for this should be developed taking the route specific kinetics into account. It will, however, be difficult to assess whether the critical effect is related to dose or concentration. If related to dose, the absorbed doses may be summed, whereas, if related to concentration it should be realised that the internal exposure pattern (height, onset, and duration of measured concentrations of blood and tissue) may be completely different for both routes. In our institute this is currently under study. A DOEL derived according to the proposed concept may already be useful as is illustrated by the evaluation of an external DOEL for MDA in the workplace investigated." In this study a significant correlation was found between cumulative MDA excretion in urine and the results of the dermal exposure measurements (assessed by hand washing). Testing for compliance was performed both by comparison between the MDA excreted in the urine and a BLV, and by comparison between the estimated external dermal exposure (hand wash method) and the external DOEL as derived in this report. Both comparisons were in close agreement, the exposure was estimated to be 20%-25% of the BLV and DOEL, respectively.I* As stated before, one of the first steps now is to develop criteria to set a priority list of chemicals for which a DOEL is a useful and necessary tool in occupational health risk management. For t h i s purpose the publications of ECETOC and US EPA and others may provide some basic elements, but these need further development and evaluation.' 'I Of crucial importance for a proper assessment of a DOEL for systemic effects is the estimation of dermal uptake. Ifno relevant data on dermal absorption are available a default value of 100% may be used. If relevant physicochemical data are available a maximal expected absorption of ~ 1 0 0 %may be calculated. Some proposals have been made for the estimation of dermal penetrati~n.'I*~I9 'I 'b Further, recent comparison of a few models based on molecular weight and log P,. led to a proposal for an additional model for the estimation of skin penetration for aqueous solutions." A combined evaluation of the applicability of the proposed models and procedures for general and specific occupational environments may provide some useful tools for a protocol for the estimation of dermal penetration at the workplace and, therefore, of a more refined DOEL. In summary, although several assumptions and simplifications have been made, the proposed concept is considered workable in an occupational environment. The gxtent and complexity of the problems associated with the assessment of a DOEL do not mean that applicable and relevant standards cannot be derived t c i t for skin exposure. Also during standard setting for respiratory and oral exposure, interpreta- tional and applicability problems often have to be dealt with. Tentative DOELs were derived for two carcinogens, the procedure will be similar for non-neoplastic agents. Further- more, the DOEL seemed to be relevant and applicable for MDA in an actual occupational setting." We recommend the testing of this approach for more substances in practice. This study was financially supported by the Dutch Ministry of Social Affairs and Employment.The important contributionsof NJ van Sittert (Shell International Chemicals BV, Department of Molecular Toxicology, Amsterdam) and H van der Waal (Shell Nederland Chemie BV, Biomedical Laboratory, Pernis) in parts of this work are gratefully acknowledged, as well as the valuable contributionsof JJG Opdam,J Kriise, and W K de Raat (TNO Nutrition and Food ResearchInstitute,Zeist). One of the referees of the present manuscript is gratefully acknowledged for his valuable comments. 1 American Conference of Governmental Industrial Hygienists (ACGIH). Documencarion of rhe threshold limit values and biological exposure indices, 6th ed. Cincinnati, USA ACGIH, 1991. 2 Greim H, Lehnen G, eds. Biologkche Arbeiuswff-TohanzWru (BAT-RZme) und Exposinbnr~uivalenufirkrebsaeu- gende Ahkuswfle (EM)W.einheim, Germany: VCH Ver- lagsgesellschaft mbH, 1996. 3 Wilson HK. Recent policy and technical developments in biological monitoring in the United Kingdom. Sci Total E n v i m 1997;199:1916. 4 Aitio A, Pekari K, JiirvsaloJ. Skin absorption as a source of error in biological monitoring. Scand3 U M Envimn Health 1984i10317-20. 5 Kao J, Carver MP. Cutaneous metabolism of xenobiotics. Drug Merab Rev 1990;22:363-410. 6 Paustenbach DJ. Assessment of the developmental risks resulting from occupational exposureto selectglycol ethers within the semiconductorindustry.3 T d Environ Health 1988;23:29-75. 7 Grandjean P, Berlin A, Gilbert M, et d.Preventing percuta- neous absorption of indusnial chemicals: the skin denota- tion. A m 3Ind Med 1988;1497-107. 8 Scansetti G, Piolatto G, Rubino G. Skin notation in the context of workplace exposure standards. A m J Ind Mul 1988;14:725-32. 9 European Centre for Ecotoxicology and Toxicology of Chemicals. Snmgy for assigning a skin noradon Brussels: ECETOC, 1993 (Revised ECETOC Document No 31.) 10 Fenske R, Van Hemmen, JJ. Occupational skin exposure to chemical substances: setting limits [editorial]. Ann &cup Hyg 1994;38:333-6. 11 Dost AA. Meeting report. A European meeting held to discuss dermal exposure monitoring and related issues. Brussels: 21-23 June 1994.Ann O c q Hyg 1995;39:24155. 12 Brouwer D, Hoogendoorn L,Bos PMJ, et d.Proposal for the assessmentof quantitative dermal exposure limits in the occupational environment. Part II. Feasibility study for application in an exposure scenariofor M D A by two different dermal exposure sampling methods. &cup E m ' m Med 1997;55:000400. 13 Fiserova-Bergerova V, Pierce JT,Droz PO. Dermal absorp tion potential of industrial chemicals: criteria for skin notation.A m 3 l n d M c d 1990;17:617-35. 14 Fiserova-Bergerova V. Relevance of occupational skin exposure.Ann Occup Hyg 1993;37:673-85. 15 Environmental Protection Agency. Dermal exposure assessment:Rinciples and applicanbnr finten'm npo+]. Washington, DC: U S Environmental Protection Agency, 1992. 16 Kezic S,Mahieu K, Monster AC, n al. Dermal absorption of vaporous and liquid 2-methoxyethanol and 2-ethoxyethanol in volunteers.OccupE m ' m Med 199754: 3843. 17 Johanson G, Boman A. Percutaneous absorption of 2-butoxyethanol vapour in human subjects. Br 3Ind Med 1991;48:788-92. 18 European Centre for Ecotoxicology and Toxicology of Chemicals. Percuranoous absorption. Brussels: ECETOC, 1993:l-80. (Monograph No 20.) 19 Stevenson H, Opdam JJG, Van Ommen B, et d . Romolfor the esrimarion of dermal absorption according a tiered approach on behalf of the risk assessment of pcrricider. Zeist, The Netherlands:TNO, 1994. (TNO-repon V94.129.) 20 Bronaugh RL, Barton CN. Prediction of human percutaneous absorption with physicochemical data. In:Wang RGM, Knaak JB, Maibach HI, eds. Health risk assessment. Dermal and tnhalamn exposure and absorpaon of wxicanu. Boca Raton, USA CRC Press, 1993. 21 Bos PMJ. Testing the possibdiw for the assessmenfof a dermal occupational exposure limit: a case sn+. Cydophorphamide. Zeist, The Netherlands: TNO Food and Nutrition Research, 1994. (In Dutch.) 22 Fenske R. Dermal exposure assessment techniques. Ann OmP Hyg 1993;37:687-706. 23 Cherrie, JWR,obertson A. Biologically relevant assessment of dermal exposure. Ann Occup Hyg 1995;39:387-92. 24 Hot~hkissS A M .Skin absorpdon of occupationalchemicals. In: Handbook of occupanbnal hygiene. Kingston upon Thames, UK:Groner, 1993. 25 Chester G. Revised guidance document for the conduct of field studies of exposure to pesticides in use. In: Curry PB, Iyengar S,Maloney P, Maroni M, eds. M e W of pesticide exposure assessment. London: Plenum Ress, 1995. 26 Van Hemmen JJ, Van Golstein Brouwers YGC,Brouwer DH. Pesticide exposure and re-entry in agridture. In: Curry PB, Iyengar S,M o n e y PA, Maroni M,eds. Methods of pesricide exposure assusmmz. London: Plenum Press, 1995. 27 De Raat WK,Kriise J, Stevenson H, et al. Ourline of a p w w lfor dermal-ro-dmnd arrapolation. TNO-regon V97.972, Zeist, The Netherlands: 1997. (TNO-report V97.972.) 28 Hakkert BC, Stevenson H, Bos PMJ, et al. Methods for the establishment of hedth-based recommended occupariaal expo- sure limits for existing subsrances. Zeist, The Netherlands: 1996. (TNO-report V96.463.) 29 Wester RC, Maibach HI, Cutaneous pharmacokinetics: 10 steps to percutaneous absorption. Drug Metab Rev 1983;14:169-205. 30 Grandjean P. Skin pene&: Haeardour chemicals at 4. London: Taylor and Francis, 1990. 3I Paustenbach DJ, hung H-W. Techniques for assessing the health risks of dermal contact with chemicals in the environment.In: Wang RGM, Knaak JB,Maibach HI, eds. Health risk assessment Dermal and inhalation exposure and absorprionofrmicanrs. B o a Raton, USA: CRC Press, 1993. 32 De Carvalho M, Falson-Reg F, Eynard I, et al. Changes in vehicle composition during skin permeation studies. In: Brain KR, James VJ,Walters KA, eds. Rediction ofpercum- mu(penenanOn. Vol 3b. Cardirr, UK: STS Publishing, 1993:251-54. 33 Boogaard PJ, Van der Waal H. BWb@ monitor& ofdermal exposure w 4,4'-diaminodiphenyl nuthar (MMJby dnmni- nation of M M in hydrdyrd urine- a human wlunuer study. Shell Biomedical -, preliminory i n d rrpon. Rot- terdam, The Netherlands: Shell Biomedical Laboratory, 1994. 34 Brunmark P. M e W fm assessment of urpmvn w ammodc amines and isocyanates by air moniroring andbionrclrkc~[doc- wral dissemuion]. Lund, Sweden:University of Lund, 1994. 35 Kompaore F,Tsuruta H.I n vivo Merences b e e n hi, black, and white in the s m m corneum barrier function.Int Arch Occup Envinm Hurlrh 1993;65:S223-5. 36 h u n g H-W, Paustenbach DJ. Techaiquu for estimatingthe percutaneous absorption of chemicals due to occupational and environmental exposure. Appl 0ccup E m i m Hy8 1994;9 187-97. 37 Wilschut A, Ten Berge WF,Robinson PJ, et al. Estimating skin permeation. The validation of five mpthematical skin permeation models. C h s p h e r e 1995;W1271-96. 38 Boogaard PJ. Testing rhe possibility f m the assessment of a dermal occupatiad exposure limic a case study. 4,4'-mrhylone dianilinc. Den Haag, The Netherlands: SheU Internationale Petroleum Maatschappij B, 1994. (Unpublished report, in Dutch.) 39 Health Council of the Netherlands: Dutch Expert Com- mittee on Occupational Standards (DECOS): CcJculaEing cancer risk. The Hague: Health Council of the Netherlands, 1995. (Publication No 1995/06WGD.) 40 Schmiihl D, Habs M. Carcinogenic acdon of low-dosc cyclophosphamidegiven orally to SpragucDawley rats in a Metime experiment. InrJCanc~r1979;23:706-12. 41 Hirst M, Tse S,Mills DG, er al. Occupational exposure to cyclophosphamide.b n c n 1984;i: 186-8. 42 Mouridsen HT, Faber 0, Skovsted L The biotransforma- tion of cyclophosphamide in man: analysis of the variation in normal subjects.A m Pharmawl 1974;35:98-106. 43 Mouridsen HT, Faber 0, Skovsted L The metabolism of cyclophosphamide. Dose dependency and the effect of long-term treatment with cyclophosphamide. Cancer 1976; 37~665-70. 44 Sessink PJM, Van den Brock PHH, Bos Rp.Urinary cyclo- phosphamide excretion in rats after intra tracheal, dermal, oral, and intravenous administration of cyclophosphamide. 3 A H l Tmicl1991;11:125-8. 45 NTP. Technical Repon on the carcinogenee studies of 4,4'-methylene diandine dihydrochlorido in F344lN mu and B6C3F, mice (drinking water srdies). NTP,1983. (Cas No 135524k8.) 46 Hotchkiss S A M , Houitt P, Caldwell I. Percutaneous absorption of 4,4'-merhylene-bis-(2-chloroaniline) and 4,C-mcthylcne dianilinc through rat and human skin in virro. T&&xdoogv in vitro 1993;2:141-8. 47 Boogaard PJ, Van der Waal, dgm~alexposure w 4 , 4 ' + m M H. B d ~ @ ~ maolniwring of di@-Mahme (MM) b detormtnanon of M a 4 m bd*Sed wine- a human duntee study. Rotterdam, ?'be NctherlPadr: Shell Biomedical ~ p b o ~ ~ t opr yr, c l i i n t e d rrporq 1994. 48 Narie Y.Sensory initmiom by lirbomc chrmicpla, crit ~ n , 3973;2:299-363. i1;