Document k87wrMvne6kVZqMkrxLxvKz0

'0022'-202X/85/8506-0522$02.00/0 * THEJOURNAL OF INVESTIGATIVE DERMATOLO85G:5Y22, -526, 1985 Copyright 0 1985by The Williams & Wilkins Co. Penetration of Benzene Through Human Skimi IRVIN H.BLANK, PH.D. AND DANIEL J. MCAULIFFE,B.S. Department of Dermutobgy, Harvard Medical School, Massachusetts General Hospital, Boston, Massachusetts, U.S.A. Although it is known that benzene may be absorbed from inhaled air, the amount that may enter the system by percutaneous absorption is less well established. We have measured the penetration of benzene through human abdominal skin in vitro from solutions in water, gasoline, hexadecane, and isooctane and found permeability constants which averaged 111.0,1.4,0.9,and 3.7 x lo-' cm-h-', respectively. The stratum corneum/ water partition coefficient for benzene has been measured and averages 30.0. The partition coefficients for the other vehicles are very low and cannot be measured by the method used for water. A new method ispresented for calculating these coefficients, however, and they are 0.11,0.14,0.17,and 0.19 for gasoline, hexane, isooctane, and hexadecane. The flus of benzene through epidermis in vitro from air saturated with benzene at 31OC averages 1.0p1 ern-'.h-'. Solvents may alter the barrier characteristics of the stratum corneum. Polar and nonpolar molecules probably traverse the stratum corneum via different pathways. By measuring the change in the permeability constants for tritiated water (a polar molecule) a n d for benzene (a relatively nonpolar molecule) before and after exposure to different solvents, alterations of the polar and nonpolar pathways have been shown to differ. Since benzene penetrates normal intact human skin more rapidly than many small organic molecules, and is potentially toxic, the skin should be considered a portal of entry for benzene. Good hygiene should be maintained and care taken to avoid lengthy exposure to solvents containing benzene. If benzene enters the system it is potentially hematotoxic and carcinogenic. For some years, industries such as the rubber, printing, and plastics industries have recognized its possible toxicity and have taken steps to reduce exposure to benzene. Snyder [l] has written a historical perspective of risks from industrial exposure to benzene. While it has been recognized that liquid benzene, as such or in various solvents, may contact the skin and be absorbed percutaneously,the major absorption has been thought to result from inhalation. Rusch et a1 [2],citing earlierwork by Teisinger et al, state that 46% of inhaled benzene is absorbed. If a respiratory rate of 16 per min and a tidal volume of 0.5 liters are assumed, 7.5 pl of benzene would be absorbed each hour through the lungs of a person working in an environment of 10 PPm. Limited quantitative data are availableon the penetration of benzene through the skin. Hanke et a1 [3] found that benzene penetrated the skin of the human forearm at the rate of 0.4 mg.cm-2.h-'. They also found that 10 mg of benzene were Manuscript received January 29,1985; accepted for publication July 17,1985. This work was supported by grants from the National Institutes of Health (5-ROl-AM15461),the Arthur 0.and Gullan M. Wellman Founriation, and the American Petroleum Institute. Reprint requests to. Irvin H. Blank, Ph.D., Wellman Research Laboratory of Photomedicine, Massachusetts General Hospital, Boston, Massachusetts 02114. absorbed by whole-body (nude) expos re to benzene vapor br 7 h. Cesaro [4] found little evidence of p e r c u b m u by the trunk and limb of 8 subjects following e- Q h l 4benzene. Conca and Maltagliati [5]also found little of absorption following 25- to 35-min exposure oft h hpa *and forearms. Maibach and Anjo [6] found only ttbout 0.172 0.1- the applied dose of liquid benzene penetrated the monkeys. Lazarew et al [7] measured 138mg of exhaled air of rabbits held with their feet i m m e a tb, for 2 h. Most of these studies of the percutaneous a b benzene have been conducted in vivo in humans. pranr lN however, studied penetration of liquid 'xnzene in 3 in vivo and in vitro. When he applied 5 pl of benzew I#tcr' of skin, 0.2% or legs of the applied dose penetrated. & ecla* cluded that "percutaneous absorption of benzene is man than in either the mini-pig or mc nkey" and that good overall agreement with respect to the magnitude 0th- zene absorption between the in vitro tind in vivo aeb of Franz recognized that a very high pcrcentage of the dose that he used evaporated. When he placed larger of benzene on monkey skin in vitro, he observed a pe- rate of 0.15 pl-cm-2.h-'. For human skin in vitro, 0.9 d d benzene per cm2penetrated in about 3 h when 520pl.m* mr applied. In most of the in vivo experiments the amount of penetrating has been determined by observing the d benzene or its metabolites in the blood and/or the amoprrt the urine or exhaled air. Occasionally the removal of from the blood by fatty tissues has been observed. The d benzene once it penetrates the epidermis and reach & bloodstream is complex; it may be metabolized to vuiop, products, it may partition out of the blood and into &g tissues, or it may be excreted. Therefore it is difficultto quantitatively the amount of benzene found in the bo$ a excreted to the amount penetrating the skin. For studying the penetration of benzene through tb we have chosen an in vitro technique We recognize P ir not firmly established that this in vitro technique the in vivo situation. In the in vivo situation, the benzenewbkb penetrates the stratum corneum diffuses quickly to the PrpiE lary dermis where it may be taken up by the bloodsasam rad systemically distributed. Our working hypothesis is that di& sion through the stratum corneum is the rate-limiting step, rad that, since this is a passive process thiough nonliving tisrUq it is rate-limiting both in vivo and in titre. Benzene hat3 water solubility (0.2 ml per dl); were it less water-soluble, itb possible that diffusion through the stniturn corneum a t M be the rate-limiting step. With in vitro techniques it is possible to investigate rndtiph *parameters that are more difficult to investigate with h techniques. The parameters that have been studied am:(1) effect of the vehicle on the rate of penetration of bentenr;Q) the effect of the vehicle on the barrier characterhtic6 @ penetration of benzene from the vapor phase; (4) the P d # P coefficient, and (5) the diffusion constant. We have a d to quantify the rate of penetration of benzene acroBLl aepidermis from various vehicles and from environmentd order to be able to estimate the total amount that the body through the skin. MATERIALS AND METHODS PENETRATION OF BENZENE THROUGH HUMAN SKIN 523 4at once, it was wrapped in aluminum foil and held at refrigerator b p e r a t u r e (4C). When only the stratum corneum was needed, the Clpided cells could be rubbed from the stratum corneum with a moist &tan-tipped applicator if the epidermis was first floated on a trypsin mlution [lo]. The stratum corneum was dried over Drierite when it was used for determining partition coefficients. Purr Vapor Materials i We studied benzene penetration from 4 nonpolar vehicles (hexadecane, isooctane, hexane, and gasoline), from water, from air, and from liquid benzene. Gasoline, which commonly contains some ben- I m e , is a nonpolar vehicle of considerable interest because skin expolpre is not unusual in certain occupations. The benzene used had been dyle-distilled and was obtained from Burdick and Jackson Laborato&q certified isooctane was obtained from Fisher Scientific Co., hex- decane, 99%, from Aldrich Chemical Co.; and tritiated water from I New England Nuclear Co. The gasoline was a standardizedreference m p l e (PS-6) obtained from the American Petroleum Institute which contained 2% benzene. i bipment Glass diffusion chambers and the techniques used for measuring ~ t a n e o u sabsorption were described by Scheuplein [Ill. We used mralsizes of diffusion chambers, some of which had a well. Most -only the volumes of the donor and receptor compartments ranged I from 2.0-3.0 ml. A temperature of 31C was used, which is a reasonable m e for the temperature of the cutaneous surface in vivo. The 1 -owing receptor (0.1% NaCl solution) was stirred continuously ritha Teflon-coated magnet. The weak sodium chloride solution was the receptor because we often checked for "holes"in our stratum i OOmeum specimens by measuring its electrical conductivity. since benzene has a low boiling point and a high vapor pressure, it Wily lost to the atmosphere from aqueous solutions. Therefore, i chrmbers were tightly capped at all times. Glass caps were generally '4but if a septum was needed for sampling, the rubber septum had 'Teflon shield (Pierce Chemical Co., Rockford, Illinois), since benzene -9 soluble in rubber. t For measuring the penetration of benzene from saturated aqueous &ion or from the vapor phase and occasionally from pure benzene, &a &%on chambers used are shown diagrammatically in Fig 1.For rOudying the penetration of benzene from an aqueous solution, the b r chamber was filled with saturated aqueous benzene to a level * the side arm and a thin layer of pure benzene placed on top of m h t e d aqueous solution. As benzene was lost from the solution bywmtration through the skin and into the receptor, the stirred donor r'k t continuously saturated from the benzene layer on its surface. seldom, if ever, exposed to air saturated with benzene, but *.r to study vapor transport by observing penetration from a U t of benzene placed in the well of the donor chamber and to evaporate. 1 A packad Scintillation Spectrometer, Model 3330, was used for g- radioactivity of tritiated water. &uoPenetration i at amount of benzene in the aqueous receptor was quantified by w b g 2-r1 samples of the receptor directly onto a 1.5% OVlOl ' b d methyl silicone-Varian) on Chromasorb column of a Varian I b9.. chromatograph. From a knowledge of the volume of the -m and the concentration of benzene in the receptor at intervals the beginning of the experiment, the flux (J.) of benzene %d! be Walu_a-te-d--. h i v e diffusion across the stratum corneum may be e x F F d to Fick's law, which states that flux of a substance is proportional difference in concentration on the two sides of the membrane: J. = hAC. (1) eJa Iflux, AC, = difference in concentration, and k, = permeaconstant. I aPemeability constant may be thought of as flux normalized for i 2 34 Time (hours) FIG 1. Flux of benzene from pure benzene 0,from benzene vapor A,and from water saturated with benzene through human epidermis into 0.1% saline solution. concentration. In vivo, any benzene that penetrates the stratum corneum may be expected to diffuse through the viable epidermis and enter the capillaries in the papillary dermis. Benzene and other molecules diffuse much more rapidly through the epidermis than through the compact stratum corneum and are subsequently removed from the skin by the blood flow. Thus, the concentration of benzene in the epidermis will be very low. In in vitro experiments the concentration of the benzene in the fluid on the receptor side of the chamber is kept very low compared to its concentration in the solution on the donor side. In our calculations, AC, has been equated to the donor concentration. In this paper, we treat the concentration gradient of benzene across the stratum corneum as the driving force, even though the flux of benzene is driven by the gradient in chemical activity across this tissue. We are assuming that the internal and external surface layers of the stratum corneum have the same activity coefficient for benzene, and that, since the system is closed and the receptor is always water, both layers are equally hydrated at equilibrium. The permeability constant may be meaningfully expanded as follows: where K, = partition coefficient, which is solubility in stratum corneum ' solubility in vehicle (donor) D, = diffusion constant, and 6 = membrane (stratum corneum) thick- ness. Partition coefficients were determined by a method described by Scheuplein I l l ] in which a known weight of dry stratum corneum is allowed to come to equilibrium with a measured volume of weak solution of benzene. The concentrations of benzene in the solution before the stratum corneum is added and after equilibrium has been reached are determined. From the difference in concentrations, the amount taken up by the stratum corneum is calculated. It is important that the difference in concentration represents only benzene entering the stratum corneum; loss by evaporation or solution into a rubber septum must be avoided. Instead of the container used by Scheuplein, we used 0.2-ml microvials, obtained from Pierce Chemical Co., with Teflon-shielded rubber septums. The vials were completely filled so that no benzene evaporated into any head space. The Teflon shield was punctured for sampling only once at the end of the experiment. This technique may be used for aqueous solutions of benzene for which the partition coefficient is high but not for solutions of benzene in E ' 524 BLANK AND MCAULIFFE hy&marbons for which partition coefficients are Very low. We believe, however, that such coefficientscan be calculated by a method described in Results. The exact thickness of the stratum corneum for each piece of skin is difficult to determine experimentally but the overall range of thicknew is known and small errors in this parameter will not significantly influence overall conclusions. Fully hydrated stratum corneum is considerably thicker than dry stratum corneum. These equations assume that neither the penetrant nor the vehicle alters the barrier capacity of the stratum corneum. Unfortunately, this is rarely true: water, for instance, changes its thickness; many vehicles may delipidizethe tissue. These changes may affect not only the barrier capacity but also the partition coefficient. Possible alterations in the barrier capacity of the stratum corneum were determined by observing the permeability constanta for benzene (nonpolar) and tritiated water (polar), before and after contact with various vehicles. Penetration rates of these two substances were measured simultaneously from a saturated aqueous solution of benzene to which a trace of tritiated water was added. Permeability constants could be obtained in 3 h, then the donor and receptor fluids removed, a vehicle kept in contact with the stratum corneum for the desired period, removed, and permeability constants for benzene and tritiated water again determined. This method measures barrier characteristics of strongly hydrated stratum corneum since penetration is measured with water in both donor and receptor. RESULTS Flux Fig 1shows the flux of benzene from pure benzene, from air saturated with benzene vapor, and from a saturated aqueous solution through human epidermis into 0.1% saIine solution. Note that steady-state flux is quickly established after short lag periods. There is no indication of any membrane alteration which causes any significantchange in the slopes of the curves. From multiple experiments of this type, the fluxes of benzene from these 3 systems are: pure benzene 2.11 f 1.08, benzene vapor 1.04 & 0.37,and aqueous solution 0.22 f 0.05 p1acm-2. h-'. Vehicles From flux measurements, the permeability constants (cm. h-') for benzene, penetrating from hexadecane, isooctane, and gasoline,each 5%(v/v) solutions,and from a saturatedaqueous solution, through epidermal samples into 0.1% saline, can be calculated and are shown in Table I. For the hydrocarbon vehicles, penetration is more rapid from isooctane than from hexadecane. When used as a vehicle, gasoline seemsto function more similarly to hexadecane than to isooctane. Standard deviations are somewhat large. More experiments might change the means and decreasethe SD somewhat,but we do not believe the general conclusions would be different. b- B 0c c0 0 (3 -.0c-) 0 u) 0 (3 VEHICLE mFIG 2. Change in permeability constants (k,) for benzene rad tritiated water (0a)fter 3-h contact with various vehicles. Error are SD. It is seen that the permeability constant for benzene from various vehicles differs. Permeability is strongly dependent oo partition coefficient, which varies significantly for &&mnt vehicles. Also,the vehicle may alter the bamer characteof the stratum corneum. Alterations of the Barrier Characteristicsof the Skin Fig 2 shows the ratios of the permeability constants benzene and tritiated water taken before and after 3-h conm with various vehicles. A ratio of 1 signifies no change in capacity following vehicle contact. A very weak salt s o b caused little or no change in the permeability constants d either molecule; butanol tkd hexane make the skin some& more permeable to both molecules; gasoline and benzene MLI the skin more permeable to water but not to benzene. H a m , it is seen that different vehicles may alter the polar and 11011polar pathways differently. t TABLE I. Permeability constants' (kp X IO3 cm. h-') for benzene penetratingfrom four vehicles Heradecane Isooctane Gasoline Water Mean fSD 4.4,4.2, 4.0b 152 4.9.4.9 141 5.3, 4.2 115 1.7 83 1.5 133 0.5, 1.4 3.1, 2.2 141 1.4, 1.6 2.4, 1.4 94 1.2, 2.8 111 0.8, 1.1 1.0, 1.3 82 0.6 1.4 84 0.5, 0.6 131 0.8 1.0 93 0.8 84 0.7,o.a 0.9, 1.1 0.94 f 0.38 3.73 f 1.26 1.40 f 0.58 111.1 & 25.9 Flux normalized for concentration. For the hydrocarbon vehicles only, all samples on a single horizontal line were from a single specimen of skin. Partition Coefficientsfor Benzene in Polar and NonpOtar Vehicles As we measured the partition coefficient for benzene betwesn water and hydrated stratum corneum, we found variation from 20 to 35 with a mean of 30. The partition coefficient between hexadecane and stratum corneum is low and cannot be sat& factorily determined by the method described. However, witb ,the stratum corneum/water partition coefficient known, tb stratum corneum/hexadecane partition coefficient can be al- culated in the following manner. In measuring the penetration of benzene from a 5% solutiao in hexadecane through epidermis into 0.1% NaCl mluthlb penetration occurs rapidly; steady state flux is usually r e d d during the fmt hour. However, as the concentration of I " f E iin the receptor continues to increase, the rate of pen decreases until equilibrium is reached, after which timb iconcentration of benzene in the receptor remains co1185Qc This equilibrium is reached at a receptor concentration than saturation for an aqueous solution (2 pl-ml-'). The tor concentration at equilibrium is identical to the co tion of benzene in water which has been allowed to ., . I. Dec. 1985 Donor concentration, Cd pl.ml-' Flux, Js pl cm-'. h-' Permeability constant x loJ cm.h-' Partitioncoefficient, K, Thickness 6 rm Diffuaon coefficient D.. x 10scm'.s-' PENETRATION OF BENZENE THROUGH HUMAN SKIN 525 TABLE 11. Penetration of benzene through epidermis from various vehicles Vehicle Water Hexadecane Isooctane Hexane Camline 2 50 50 50 50 Experimental 0.22 0.05 0.19 0.12 0.07 Experimental 111 0.94 3.73 2.4 1.4 kp= JJCd 30 0.19 0.17 0.14 0.11 Experimental and calculated 40 40 40 40 40 Estimated 4.1 5.5 24.4 19.0 14.0 D, = kp6/K, equilibrium with the donor without the presence of the skin. For 5% benzene in hexadecane, the aqueous concentration at equilibrium is 0.31 pl. ml-'. At equilibrium,the concentrations of benzene in the stratum corneum at the donor boundary and at the receptor boundary areidentical, since if they were different a net flux would result and at equilibrium there is no net flux. The concentration in the stratum corneum at the receptor boundary is the concentration in the receptor multiplied by the stratum corneum/ water partition coefficient. For this system, it is 0.31 pl*ml-' x 30 = 9.3 pl.mI-l, and we have said that this must be the concentration in the tissue at the donor boundary also. The concentration of benzene in the hexadecane donor is 50 pl. d-'and thus the stratum corneum/hexadecane partition coefficient is 9.3150= 0.19. Using the same method, it is possible to determine the "ttum corneum/solvent partition coefficient for any solvent ?miscible with water. Partition coefficients for benzene in y e , hexane, and gasoline are 0.17,0.14,and 0.11,respechvely. The stratum corneum is a relatively good "solvent" for benzene; it is much better than water but not as good as many nonpolar, organic liquids. DiffusionConstant Diffusion constants of molecuIes moving across the stratum Corneum are not measured directly. If lag time and thickness are known, diffusion constants can be calculated. We have found, however, that lag times are difficult to determine accu- -9ratelyo.fwpeercmhoeasebitloitcyaclcounlsattaendtsif,fpuasirotinticoonnsctoaenftfsicfireonmtsa, aknndowesl-- -tad thickness (see Eq. 2). We believe that the aqueous 'baptor strongly hydrates the stratum corneum in the diffusion *bere and that 40 pm is a reasonable estimate of its thick- Table I1is a summary table showingthese various param- *m forbenzene penetrating human epidermis from 5different and entering a 0.1% saline solution on the receptor Diffusion constants for benzene, penetrating from hydro-n solvents, vary from 4.1 to 24.4 x lo-' cm2.s-'. These all high compared to the diffusion constant for water in corneum, which is 5 x IO-'' cm2.s-' Ill].The vehicles *illenter the stratum corneum as well as the benzene, and this '4ay account for some of the differences in diffusion constants 'Or benzene when presented to the skin in different vehicles. DISCUSSION ?e Solubility characteristics of benzene are such that it is %taken up by the stratum corneum. Once in the stratum it does not meet many restraining forces to impede le mfJvement and diffuses easily. The permeability constant br bnzene, as determined in vitro, is higher than that of many olecules, particularly those having one or more quantitative permeability data are difficult to is tempting to apply in vitro data to in vivo situations. We recognize that differences exist in the two systems, particularly with respect to hydration of the stratum corneum. Our in vitro closed system is more closely matched by an in vivo system in which there is occlusion. There are occupations in which some areas of the skin, such as the hands, may be covered for some time with an occlusive oily film. More quantitative data are required in order to know the effect of hydration of the stratum corneum and the effect of alteration of the in vivo water concentration profile on the penetration of benzene. Even though these uncertainties exist, and more data are needed to support the Franz conclusion that there is good overall agreement between in vitro and in vivo data (81,we have chosen to calculate from our in vitro data and from inhalation data of others the amount of benzene entering the body under specific environmental conditions. An adult working in ambient air containing 10ppm of benzene, with 100 cm2 of glabrous skin in contact with gasoline containing 5% benzene, and his entire skin (2 mZ)in contact with ambient air, will absorb in an hour, 7.5 p1 of benzene from inhalation, 7.0pl from contact with gasoline, and 1.5 pl from body exposure to ambient air. Since our in vitro techniques measure the penetration of benzene through strongly hydrated stratum corneum, the calculated flux may be higher than under some in vivo conditions. Nevertheless, it seems that unless good hygiene is maintained and care is taken to prevent lengthy exposure to solvents containing benzene, significant amounts of benzene may enter the body through the skin. We have not addressed the fate of the benzene once it penetrates the stratum corneum and thus, from this study alone, the potential toxicity of benzene that has entered through the skin cannot be determined. It seems apparent, however, that if one is concerned about the systemiceffects of benzene, contact of the skin with benzene and solvents containing benzene should be avoided as much as possible. The authors wish to thank Paul J. Gamin, Amoco Corporation, Peter H. Craig, Mobil Oil Corporation, and Steven Jacques, Massachusetts General Hospital, for their reviews of this manuscript and their very helpful suggestions. REFERENCES 1. Snyder R:The benzene problem in historical perspective. Fundam Ap 1 Toxicol4692-699.1984 2. Ru&$ GM, k n g BK, Laskin S: Benzene metabolism. J Toxicol Environ Health 2 (sup~l):23-36,1977 93. Hanke J, Dubiewicz T, lotrowski J: The absorption of benzene through the skin in man. Med Pr 12413-426,1961 94. CesaroAN: ISpercutaneousbenzene absorption possible? Medicina del Lavora 37:151-156,1946 *5. Conca GL,Maltagliati A: Study of the transcutaneous absorption Howard Maibach, M.D., prepared a literature review of the percutaneous absorptionof petroleum hydywbons for the American Petroleum Institute. References marked are from this review with the permission of Dr. Maibach and the American Petroleum Institute. .. 526 BLANK AND MCAULIFFE of benzene. Medicina del Lavora 46:194-198,1955 6. Maibach HI, Anjo DM:Percutaneous penetration of benzene and benzene contained in solvents used in the rubber industry. Arch Environ Health 36256-260,1981 '7. Laserew NV, Brussilowskaja AJ, Lavroff IN, Lifschitz FB: On the rmeability of the skin to benzin and benzene. Archiv fur Eygiene und Bakteriologie 106:112-122,1931 38. Franz TJ: Percutaneous absorption of benzene, The Toxicolo Of Petroleum Hydrocarbons. Edited by HN MacFarland, CE old sworth, J A MacGregor, RW Call, ML Cave. Washington, DC, American Petroleum Institute, 1982 9. Blank IH, Griesemer RD, Gould E: The penetration of a ,,,ti. cholinesterase agent (sarin) into skin. I. Rate of penetration excised human skin. J Invest Dermatol 29:299-309,1957 10. Kligman AM, Christophers E: Preparation of isolated ehect, al human stratum corneum. Arch Dermatol&702-705,1~ 11. Scheuplein RJ: Mechanism of percutaneous absorption. 1. of penetration and influence of solubility. J Invest hrmrto 45~334-345,1965